{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.00085, "eval_steps": 500, "global_step": 8500, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 1e-07, "grad_norm": 13.864434242248535, "learning_rate": 0.0, "loss": 73.5625, "step": 1 }, { "epoch": 2e-07, "grad_norm": 13.48910903930664, "learning_rate": 1e-08, "loss": 73.625, "step": 2 }, { "epoch": 3e-07, "grad_norm": 14.2167387008667, "learning_rate": 2e-08, "loss": 73.625, "step": 3 }, { "epoch": 4e-07, "grad_norm": 13.33409595489502, "learning_rate": 3.0000000000000004e-08, "loss": 73.625, "step": 4 }, { "epoch": 5e-07, "grad_norm": 13.449651718139648, "learning_rate": 4e-08, "loss": 73.625, "step": 5 }, { "epoch": 6e-07, "grad_norm": 13.80512809753418, "learning_rate": 5.0000000000000004e-08, "loss": 73.6875, "step": 6 }, { "epoch": 7e-07, "grad_norm": 13.694576263427734, "learning_rate": 6.000000000000001e-08, "loss": 73.5, "step": 7 }, { "epoch": 8e-07, "grad_norm": 13.790714263916016, "learning_rate": 7e-08, "loss": 73.625, "step": 8 }, { "epoch": 9e-07, "grad_norm": 13.956605911254883, "learning_rate": 8e-08, "loss": 73.6875, "step": 9 }, { "epoch": 1e-06, "grad_norm": 13.796408653259277, "learning_rate": 9e-08, "loss": 73.6875, "step": 10 }, { "epoch": 1.1e-06, "grad_norm": 13.963505744934082, "learning_rate": 1.0000000000000001e-07, "loss": 73.5625, "step": 11 }, { "epoch": 1.2e-06, "grad_norm": 13.888561248779297, "learning_rate": 1.1e-07, "loss": 73.5625, "step": 12 }, { "epoch": 1.3e-06, "grad_norm": 13.5466890335083, "learning_rate": 1.2000000000000002e-07, "loss": 73.625, "step": 13 }, { "epoch": 1.4e-06, "grad_norm": 13.878573417663574, "learning_rate": 1.3e-07, "loss": 73.5, "step": 14 }, { "epoch": 1.5e-06, "grad_norm": 13.623698234558105, "learning_rate": 1.4e-07, "loss": 73.5625, "step": 15 }, { "epoch": 1.6e-06, "grad_norm": 13.813441276550293, "learning_rate": 1.5000000000000002e-07, "loss": 73.5625, "step": 16 }, { "epoch": 1.7e-06, "grad_norm": 13.076991081237793, "learning_rate": 1.6e-07, "loss": 73.5625, "step": 17 }, { "epoch": 1.8e-06, "grad_norm": 13.786111831665039, "learning_rate": 1.7e-07, "loss": 73.625, "step": 18 }, { "epoch": 1.9e-06, "grad_norm": 13.549301147460938, "learning_rate": 1.8e-07, "loss": 73.6875, "step": 19 }, { "epoch": 2e-06, "grad_norm": 13.712065696716309, "learning_rate": 1.9e-07, "loss": 73.5, "step": 20 }, { "epoch": 2.1e-06, "grad_norm": 13.899330139160156, "learning_rate": 2.0000000000000002e-07, "loss": 73.625, "step": 21 }, { "epoch": 2.2e-06, "grad_norm": 13.786901473999023, "learning_rate": 2.0999999999999997e-07, "loss": 73.5625, "step": 22 }, { "epoch": 2.3e-06, "grad_norm": 13.647011756896973, "learning_rate": 2.2e-07, "loss": 73.5625, "step": 23 }, { "epoch": 2.4e-06, "grad_norm": 13.713528633117676, "learning_rate": 2.3000000000000002e-07, "loss": 73.5625, "step": 24 }, { "epoch": 2.5e-06, "grad_norm": 13.645662307739258, "learning_rate": 2.4000000000000003e-07, "loss": 73.625, "step": 25 }, { "epoch": 2.6e-06, "grad_norm": 14.283074378967285, "learning_rate": 2.5e-07, "loss": 73.625, "step": 26 }, { "epoch": 2.7e-06, "grad_norm": 13.786810874938965, "learning_rate": 2.6e-07, "loss": 73.6875, "step": 27 }, { "epoch": 2.8e-06, "grad_norm": 13.583864212036133, "learning_rate": 2.7e-07, "loss": 73.625, "step": 28 }, { "epoch": 2.9e-06, "grad_norm": 13.53901481628418, "learning_rate": 2.8e-07, "loss": 73.625, "step": 29 }, { "epoch": 3e-06, "grad_norm": 13.56877613067627, "learning_rate": 2.9000000000000003e-07, "loss": 73.5625, "step": 30 }, { "epoch": 3.1e-06, "grad_norm": 13.632527351379395, "learning_rate": 3.0000000000000004e-07, "loss": 73.6875, "step": 31 }, { "epoch": 3.2e-06, "grad_norm": 13.880523681640625, "learning_rate": 3.1e-07, "loss": 73.5625, "step": 32 }, { "epoch": 3.3e-06, "grad_norm": 13.455181121826172, "learning_rate": 3.2e-07, "loss": 73.6875, "step": 33 }, { "epoch": 3.4e-06, "grad_norm": 13.400262832641602, "learning_rate": 3.3e-07, "loss": 73.5625, "step": 34 }, { "epoch": 3.5e-06, "grad_norm": 13.204479217529297, "learning_rate": 3.4e-07, "loss": 73.5625, "step": 35 }, { "epoch": 3.6e-06, "grad_norm": 13.526925086975098, "learning_rate": 3.5e-07, "loss": 73.5625, "step": 36 }, { "epoch": 3.7e-06, "grad_norm": 13.56838321685791, "learning_rate": 3.6e-07, "loss": 73.5625, "step": 37 }, { "epoch": 3.8e-06, "grad_norm": 13.788740158081055, "learning_rate": 3.7e-07, "loss": 73.625, "step": 38 }, { "epoch": 3.9e-06, "grad_norm": 13.793792724609375, "learning_rate": 3.8e-07, "loss": 73.5625, "step": 39 }, { "epoch": 4e-06, "grad_norm": 14.137533187866211, "learning_rate": 3.8999999999999997e-07, "loss": 73.6875, "step": 40 }, { "epoch": 4.1e-06, "grad_norm": 13.904379844665527, "learning_rate": 4.0000000000000003e-07, "loss": 73.625, "step": 41 }, { "epoch": 4.2e-06, "grad_norm": 13.895687103271484, "learning_rate": 4.1e-07, "loss": 73.625, "step": 42 }, { "epoch": 4.3e-06, "grad_norm": 14.146393775939941, "learning_rate": 4.1999999999999995e-07, "loss": 73.5625, "step": 43 }, { "epoch": 4.4e-06, "grad_norm": 13.730606079101562, "learning_rate": 4.3e-07, "loss": 73.625, "step": 44 }, { "epoch": 4.5e-06, "grad_norm": 13.854805946350098, "learning_rate": 4.4e-07, "loss": 73.5625, "step": 45 }, { "epoch": 4.6e-06, "grad_norm": 13.894522666931152, "learning_rate": 4.5e-07, "loss": 73.5625, "step": 46 }, { "epoch": 4.7e-06, "grad_norm": 13.2392578125, "learning_rate": 4.6000000000000004e-07, "loss": 73.6875, "step": 47 }, { "epoch": 4.8e-06, "grad_norm": 13.556278228759766, "learning_rate": 4.7e-07, "loss": 73.75, "step": 48 }, { "epoch": 4.9e-06, "grad_norm": 13.716812133789062, "learning_rate": 4.800000000000001e-07, "loss": 73.625, "step": 49 }, { "epoch": 5e-06, "grad_norm": 13.964197158813477, "learning_rate": 4.900000000000001e-07, "loss": 73.5625, "step": 50 }, { "epoch": 5.1e-06, "grad_norm": 13.632363319396973, "learning_rate": 5e-07, "loss": 73.6875, "step": 51 }, { "epoch": 5.2e-06, "grad_norm": 13.38712215423584, "learning_rate": 5.100000000000001e-07, "loss": 73.5625, "step": 52 }, { "epoch": 5.3e-06, "grad_norm": 13.229913711547852, "learning_rate": 5.2e-07, "loss": 73.5, "step": 53 }, { "epoch": 5.4e-06, "grad_norm": 13.719653129577637, "learning_rate": 5.3e-07, "loss": 73.625, "step": 54 }, { "epoch": 5.5e-06, "grad_norm": 12.99260425567627, "learning_rate": 5.4e-07, "loss": 73.5625, "step": 55 }, { "epoch": 5.6e-06, "grad_norm": 13.961441040039062, "learning_rate": 5.5e-07, "loss": 73.5625, "step": 56 }, { "epoch": 5.7e-06, "grad_norm": 13.631799697875977, "learning_rate": 5.6e-07, "loss": 73.5625, "step": 57 }, { "epoch": 5.8e-06, "grad_norm": 13.392004013061523, "learning_rate": 5.7e-07, "loss": 73.75, "step": 58 }, { "epoch": 5.9e-06, "grad_norm": 13.643326759338379, "learning_rate": 5.800000000000001e-07, "loss": 73.625, "step": 59 }, { "epoch": 6e-06, "grad_norm": 13.313652992248535, "learning_rate": 5.9e-07, "loss": 73.5, "step": 60 }, { "epoch": 6.1e-06, "grad_norm": 13.296246528625488, "learning_rate": 6.000000000000001e-07, "loss": 73.5625, "step": 61 }, { "epoch": 6.2e-06, "grad_norm": 12.835416793823242, "learning_rate": 6.1e-07, "loss": 73.5625, "step": 62 }, { "epoch": 6.3e-06, "grad_norm": 13.500370025634766, "learning_rate": 6.2e-07, "loss": 73.5, "step": 63 }, { "epoch": 6.4e-06, "grad_norm": 14.06345272064209, "learning_rate": 6.3e-07, "loss": 73.5625, "step": 64 }, { "epoch": 6.5e-06, "grad_norm": 13.467910766601562, "learning_rate": 6.4e-07, "loss": 73.5625, "step": 65 }, { "epoch": 6.6e-06, "grad_norm": 13.793481826782227, "learning_rate": 6.5e-07, "loss": 73.8125, "step": 66 }, { "epoch": 6.7e-06, "grad_norm": 13.709341049194336, "learning_rate": 6.6e-07, "loss": 73.75, "step": 67 }, { "epoch": 6.8e-06, "grad_norm": 13.779867172241211, "learning_rate": 6.7e-07, "loss": 73.5625, "step": 68 }, { "epoch": 6.9e-06, "grad_norm": 13.970793724060059, "learning_rate": 6.8e-07, "loss": 73.625, "step": 69 }, { "epoch": 7e-06, "grad_norm": 13.877311706542969, "learning_rate": 6.900000000000001e-07, "loss": 73.625, "step": 70 }, { "epoch": 7.1e-06, "grad_norm": 13.815545082092285, "learning_rate": 7e-07, "loss": 73.6875, "step": 71 }, { "epoch": 7.2e-06, "grad_norm": 13.491872787475586, "learning_rate": 7.1e-07, "loss": 73.625, "step": 72 }, { "epoch": 7.3e-06, "grad_norm": 13.690380096435547, "learning_rate": 7.2e-07, "loss": 73.5625, "step": 73 }, { "epoch": 7.4e-06, "grad_norm": 13.530372619628906, "learning_rate": 7.3e-07, "loss": 73.5625, "step": 74 }, { "epoch": 7.5e-06, "grad_norm": 13.24216079711914, "learning_rate": 7.4e-07, "loss": 73.5625, "step": 75 }, { "epoch": 7.6e-06, "grad_norm": 13.833526611328125, "learning_rate": 7.5e-07, "loss": 73.5625, "step": 76 }, { "epoch": 7.7e-06, "grad_norm": 13.154213905334473, "learning_rate": 7.6e-07, "loss": 73.5625, "step": 77 }, { "epoch": 7.8e-06, "grad_norm": 13.55800724029541, "learning_rate": 7.699999999999999e-07, "loss": 73.5625, "step": 78 }, { "epoch": 7.9e-06, "grad_norm": 13.533388137817383, "learning_rate": 7.799999999999999e-07, "loss": 73.5625, "step": 79 }, { "epoch": 8e-06, "grad_norm": 12.90389633178711, "learning_rate": 7.900000000000001e-07, "loss": 73.4375, "step": 80 }, { "epoch": 8.1e-06, "grad_norm": 13.699131965637207, "learning_rate": 8.000000000000001e-07, "loss": 73.625, "step": 81 }, { "epoch": 8.2e-06, "grad_norm": 14.151235580444336, "learning_rate": 8.100000000000001e-07, "loss": 73.5625, "step": 82 }, { "epoch": 8.3e-06, "grad_norm": 13.401898384094238, "learning_rate": 8.2e-07, "loss": 73.625, "step": 83 }, { "epoch": 8.4e-06, "grad_norm": 12.971834182739258, "learning_rate": 8.3e-07, "loss": 73.5625, "step": 84 }, { "epoch": 8.5e-06, "grad_norm": 13.448446273803711, "learning_rate": 8.399999999999999e-07, "loss": 73.5, "step": 85 }, { "epoch": 8.6e-06, "grad_norm": 13.639290809631348, "learning_rate": 8.500000000000001e-07, "loss": 73.6875, "step": 86 }, { "epoch": 8.7e-06, "grad_norm": 13.647253036499023, "learning_rate": 8.6e-07, "loss": 73.5, "step": 87 }, { "epoch": 8.8e-06, "grad_norm": 13.423108100891113, "learning_rate": 8.7e-07, "loss": 73.625, "step": 88 }, { "epoch": 8.9e-06, "grad_norm": 13.234737396240234, "learning_rate": 8.8e-07, "loss": 73.5625, "step": 89 }, { "epoch": 9e-06, "grad_norm": 13.644418716430664, "learning_rate": 8.9e-07, "loss": 73.6875, "step": 90 }, { "epoch": 9.1e-06, "grad_norm": 13.497346878051758, "learning_rate": 9e-07, "loss": 73.6875, "step": 91 }, { "epoch": 9.2e-06, "grad_norm": 13.384448051452637, "learning_rate": 9.100000000000001e-07, "loss": 73.625, "step": 92 }, { "epoch": 9.3e-06, "grad_norm": 13.391413688659668, "learning_rate": 9.200000000000001e-07, "loss": 73.625, "step": 93 }, { "epoch": 9.4e-06, "grad_norm": 13.153121948242188, "learning_rate": 9.3e-07, "loss": 73.625, "step": 94 }, { "epoch": 9.5e-06, "grad_norm": 13.874184608459473, "learning_rate": 9.4e-07, "loss": 73.5, "step": 95 }, { "epoch": 9.6e-06, "grad_norm": 13.332117080688477, "learning_rate": 9.5e-07, "loss": 73.625, "step": 96 }, { "epoch": 9.7e-06, "grad_norm": 13.225434303283691, "learning_rate": 9.600000000000001e-07, "loss": 73.5625, "step": 97 }, { "epoch": 9.8e-06, "grad_norm": 13.648529052734375, "learning_rate": 9.7e-07, "loss": 73.5, "step": 98 }, { "epoch": 9.9e-06, "grad_norm": 14.294005393981934, "learning_rate": 9.800000000000001e-07, "loss": 73.625, "step": 99 }, { "epoch": 1e-05, "grad_norm": 13.486104965209961, "learning_rate": 9.9e-07, "loss": 73.5625, "step": 100 }, { "epoch": 1.01e-05, "grad_norm": 13.966443061828613, "learning_rate": 1e-06, "loss": 73.6875, "step": 101 }, { "epoch": 1.02e-05, "grad_norm": 13.408844947814941, "learning_rate": 1.0099999999999999e-06, "loss": 73.625, "step": 102 }, { "epoch": 1.03e-05, "grad_norm": 13.945192337036133, "learning_rate": 1.0200000000000002e-06, "loss": 73.5625, "step": 103 }, { "epoch": 1.04e-05, "grad_norm": 13.554696083068848, "learning_rate": 1.03e-06, "loss": 73.5, "step": 104 }, { "epoch": 1.05e-05, "grad_norm": 13.722262382507324, "learning_rate": 1.04e-06, "loss": 73.5, "step": 105 }, { "epoch": 1.06e-05, "grad_norm": 14.362868309020996, "learning_rate": 1.0500000000000001e-06, "loss": 73.8125, "step": 106 }, { "epoch": 1.07e-05, "grad_norm": 12.963215827941895, "learning_rate": 1.06e-06, "loss": 73.5625, "step": 107 }, { "epoch": 1.08e-05, "grad_norm": 13.040894508361816, "learning_rate": 1.07e-06, "loss": 73.5625, "step": 108 }, { "epoch": 1.09e-05, "grad_norm": 13.496689796447754, "learning_rate": 1.08e-06, "loss": 73.625, "step": 109 }, { "epoch": 1.1e-05, "grad_norm": 12.973114013671875, "learning_rate": 1.0900000000000002e-06, "loss": 73.625, "step": 110 }, { "epoch": 1.11e-05, "grad_norm": 13.376940727233887, "learning_rate": 1.1e-06, "loss": 73.625, "step": 111 }, { "epoch": 1.12e-05, "grad_norm": 13.423301696777344, "learning_rate": 1.11e-06, "loss": 73.5, "step": 112 }, { "epoch": 1.13e-05, "grad_norm": 14.047898292541504, "learning_rate": 1.12e-06, "loss": 73.6875, "step": 113 }, { "epoch": 1.14e-05, "grad_norm": 13.22636890411377, "learning_rate": 1.13e-06, "loss": 73.5625, "step": 114 }, { "epoch": 1.15e-05, "grad_norm": 13.150336265563965, "learning_rate": 1.14e-06, "loss": 73.5625, "step": 115 }, { "epoch": 1.16e-05, "grad_norm": 13.312559127807617, "learning_rate": 1.15e-06, "loss": 73.5, "step": 116 }, { "epoch": 1.17e-05, "grad_norm": 13.54313850402832, "learning_rate": 1.1600000000000001e-06, "loss": 73.5, "step": 117 }, { "epoch": 1.18e-05, "grad_norm": 13.342724800109863, "learning_rate": 1.17e-06, "loss": 73.6875, "step": 118 }, { "epoch": 1.19e-05, "grad_norm": 13.630057334899902, "learning_rate": 1.18e-06, "loss": 73.5625, "step": 119 }, { "epoch": 1.2e-05, "grad_norm": 13.717558860778809, "learning_rate": 1.19e-06, "loss": 73.5, "step": 120 }, { "epoch": 1.21e-05, "grad_norm": 14.034920692443848, "learning_rate": 1.2000000000000002e-06, "loss": 73.5625, "step": 121 }, { "epoch": 1.22e-05, "grad_norm": 12.964286804199219, "learning_rate": 1.21e-06, "loss": 73.5, "step": 122 }, { "epoch": 1.23e-05, "grad_norm": 13.391942024230957, "learning_rate": 1.22e-06, "loss": 73.625, "step": 123 }, { "epoch": 1.24e-05, "grad_norm": 13.302276611328125, "learning_rate": 1.23e-06, "loss": 73.5, "step": 124 }, { "epoch": 1.25e-05, "grad_norm": 13.081567764282227, "learning_rate": 1.24e-06, "loss": 73.625, "step": 125 }, { "epoch": 1.26e-05, "grad_norm": 13.869545936584473, "learning_rate": 1.25e-06, "loss": 73.5625, "step": 126 }, { "epoch": 1.27e-05, "grad_norm": 13.455565452575684, "learning_rate": 1.26e-06, "loss": 73.5, "step": 127 }, { "epoch": 1.28e-05, "grad_norm": 13.791558265686035, "learning_rate": 1.2700000000000001e-06, "loss": 73.625, "step": 128 }, { "epoch": 1.29e-05, "grad_norm": 13.645962715148926, "learning_rate": 1.28e-06, "loss": 73.625, "step": 129 }, { "epoch": 1.3e-05, "grad_norm": 13.296918869018555, "learning_rate": 1.29e-06, "loss": 73.5625, "step": 130 }, { "epoch": 1.31e-05, "grad_norm": 13.747169494628906, "learning_rate": 1.3e-06, "loss": 73.5625, "step": 131 }, { "epoch": 1.32e-05, "grad_norm": 13.228446006774902, "learning_rate": 1.3100000000000002e-06, "loss": 73.5, "step": 132 }, { "epoch": 1.33e-05, "grad_norm": 13.962552070617676, "learning_rate": 1.32e-06, "loss": 73.5, "step": 133 }, { "epoch": 1.34e-05, "grad_norm": 13.579384803771973, "learning_rate": 1.33e-06, "loss": 73.5625, "step": 134 }, { "epoch": 1.35e-05, "grad_norm": 13.230710983276367, "learning_rate": 1.34e-06, "loss": 73.5, "step": 135 }, { "epoch": 1.36e-05, "grad_norm": 13.567021369934082, "learning_rate": 1.35e-06, "loss": 73.5, "step": 136 }, { "epoch": 1.37e-05, "grad_norm": 12.811543464660645, "learning_rate": 1.36e-06, "loss": 73.5, "step": 137 }, { "epoch": 1.38e-05, "grad_norm": 13.733097076416016, "learning_rate": 1.37e-06, "loss": 73.625, "step": 138 }, { "epoch": 1.39e-05, "grad_norm": 13.566929817199707, "learning_rate": 1.3800000000000001e-06, "loss": 73.5625, "step": 139 }, { "epoch": 1.4e-05, "grad_norm": 13.628262519836426, "learning_rate": 1.39e-06, "loss": 73.6875, "step": 140 }, { "epoch": 1.41e-05, "grad_norm": 13.57292652130127, "learning_rate": 1.4e-06, "loss": 73.5625, "step": 141 }, { "epoch": 1.42e-05, "grad_norm": 13.723799705505371, "learning_rate": 1.41e-06, "loss": 73.625, "step": 142 }, { "epoch": 1.43e-05, "grad_norm": 14.213798522949219, "learning_rate": 1.42e-06, "loss": 73.625, "step": 143 }, { "epoch": 1.44e-05, "grad_norm": 13.248021125793457, "learning_rate": 1.43e-06, "loss": 73.5625, "step": 144 }, { "epoch": 1.45e-05, "grad_norm": 13.645977020263672, "learning_rate": 1.44e-06, "loss": 73.5625, "step": 145 }, { "epoch": 1.46e-05, "grad_norm": 13.61572551727295, "learning_rate": 1.45e-06, "loss": 73.5, "step": 146 }, { "epoch": 1.47e-05, "grad_norm": 13.234169960021973, "learning_rate": 1.46e-06, "loss": 73.5625, "step": 147 }, { "epoch": 1.48e-05, "grad_norm": 13.495967864990234, "learning_rate": 1.47e-06, "loss": 73.6875, "step": 148 }, { "epoch": 1.49e-05, "grad_norm": 13.479613304138184, "learning_rate": 1.48e-06, "loss": 73.5, "step": 149 }, { "epoch": 1.5e-05, "grad_norm": 12.886178016662598, "learning_rate": 1.49e-06, "loss": 73.5, "step": 150 }, { "epoch": 1.51e-05, "grad_norm": 13.639323234558105, "learning_rate": 1.5e-06, "loss": 73.5625, "step": 151 }, { "epoch": 1.52e-05, "grad_norm": 13.812483787536621, "learning_rate": 1.51e-06, "loss": 73.5, "step": 152 }, { "epoch": 1.53e-05, "grad_norm": 13.071154594421387, "learning_rate": 1.52e-06, "loss": 73.5, "step": 153 }, { "epoch": 1.54e-05, "grad_norm": 13.37407112121582, "learning_rate": 1.5300000000000002e-06, "loss": 73.5625, "step": 154 }, { "epoch": 1.55e-05, "grad_norm": 13.398472785949707, "learning_rate": 1.5399999999999999e-06, "loss": 73.5625, "step": 155 }, { "epoch": 1.56e-05, "grad_norm": 13.664745330810547, "learning_rate": 1.55e-06, "loss": 73.625, "step": 156 }, { "epoch": 1.57e-05, "grad_norm": 13.450276374816895, "learning_rate": 1.5599999999999999e-06, "loss": 73.5, "step": 157 }, { "epoch": 1.58e-05, "grad_norm": 13.383707046508789, "learning_rate": 1.57e-06, "loss": 73.5, "step": 158 }, { "epoch": 1.59e-05, "grad_norm": 13.524046897888184, "learning_rate": 1.5800000000000001e-06, "loss": 73.4375, "step": 159 }, { "epoch": 1.6e-05, "grad_norm": 13.396870613098145, "learning_rate": 1.59e-06, "loss": 73.5625, "step": 160 }, { "epoch": 1.61e-05, "grad_norm": 13.882829666137695, "learning_rate": 1.6000000000000001e-06, "loss": 73.5, "step": 161 }, { "epoch": 1.62e-05, "grad_norm": 13.542072296142578, "learning_rate": 1.6099999999999998e-06, "loss": 73.5625, "step": 162 }, { "epoch": 1.63e-05, "grad_norm": 12.320090293884277, "learning_rate": 1.6200000000000002e-06, "loss": 73.5, "step": 163 }, { "epoch": 1.64e-05, "grad_norm": 13.312244415283203, "learning_rate": 1.6300000000000003e-06, "loss": 73.5625, "step": 164 }, { "epoch": 1.65e-05, "grad_norm": 12.99962043762207, "learning_rate": 1.64e-06, "loss": 73.5, "step": 165 }, { "epoch": 1.66e-05, "grad_norm": 12.987727165222168, "learning_rate": 1.65e-06, "loss": 73.4375, "step": 166 }, { "epoch": 1.67e-05, "grad_norm": 13.491228103637695, "learning_rate": 1.66e-06, "loss": 73.5, "step": 167 }, { "epoch": 1.68e-05, "grad_norm": 13.407848358154297, "learning_rate": 1.67e-06, "loss": 73.5625, "step": 168 }, { "epoch": 1.69e-05, "grad_norm": 13.65124797821045, "learning_rate": 1.6799999999999998e-06, "loss": 73.5, "step": 169 }, { "epoch": 1.7e-05, "grad_norm": 13.30240249633789, "learning_rate": 1.6900000000000001e-06, "loss": 73.5625, "step": 170 }, { "epoch": 1.71e-05, "grad_norm": 13.703505516052246, "learning_rate": 1.7000000000000002e-06, "loss": 73.625, "step": 171 }, { "epoch": 1.72e-05, "grad_norm": 13.372552871704102, "learning_rate": 1.71e-06, "loss": 73.5625, "step": 172 }, { "epoch": 1.73e-05, "grad_norm": 13.747980117797852, "learning_rate": 1.72e-06, "loss": 73.5, "step": 173 }, { "epoch": 1.74e-05, "grad_norm": 13.47564697265625, "learning_rate": 1.73e-06, "loss": 73.5625, "step": 174 }, { "epoch": 1.75e-05, "grad_norm": 13.97725772857666, "learning_rate": 1.74e-06, "loss": 73.5, "step": 175 }, { "epoch": 1.76e-05, "grad_norm": 13.73171615600586, "learning_rate": 1.7500000000000002e-06, "loss": 73.5, "step": 176 }, { "epoch": 1.77e-05, "grad_norm": 13.646438598632812, "learning_rate": 1.76e-06, "loss": 73.5625, "step": 177 }, { "epoch": 1.78e-05, "grad_norm": 12.986227989196777, "learning_rate": 1.7700000000000002e-06, "loss": 73.625, "step": 178 }, { "epoch": 1.79e-05, "grad_norm": 13.07104778289795, "learning_rate": 1.78e-06, "loss": 73.5, "step": 179 }, { "epoch": 1.8e-05, "grad_norm": 13.796416282653809, "learning_rate": 1.79e-06, "loss": 73.5, "step": 180 }, { "epoch": 1.81e-05, "grad_norm": 13.549040794372559, "learning_rate": 1.8e-06, "loss": 73.5625, "step": 181 }, { "epoch": 1.82e-05, "grad_norm": 13.482290267944336, "learning_rate": 1.81e-06, "loss": 73.5, "step": 182 }, { "epoch": 1.83e-05, "grad_norm": 13.399819374084473, "learning_rate": 1.8200000000000002e-06, "loss": 73.5, "step": 183 }, { "epoch": 1.84e-05, "grad_norm": 13.660837173461914, "learning_rate": 1.83e-06, "loss": 73.5625, "step": 184 }, { "epoch": 1.85e-05, "grad_norm": 13.732992172241211, "learning_rate": 1.8400000000000002e-06, "loss": 73.5625, "step": 185 }, { "epoch": 1.86e-05, "grad_norm": 13.626218795776367, "learning_rate": 1.8499999999999999e-06, "loss": 73.5625, "step": 186 }, { "epoch": 1.87e-05, "grad_norm": 13.294013977050781, "learning_rate": 1.86e-06, "loss": 73.5, "step": 187 }, { "epoch": 1.88e-05, "grad_norm": 12.994363784790039, "learning_rate": 1.87e-06, "loss": 73.5625, "step": 188 }, { "epoch": 1.89e-05, "grad_norm": 13.95550537109375, "learning_rate": 1.88e-06, "loss": 73.5625, "step": 189 }, { "epoch": 1.9e-05, "grad_norm": 13.144396781921387, "learning_rate": 1.8900000000000001e-06, "loss": 73.5, "step": 190 }, { "epoch": 1.91e-05, "grad_norm": 12.912415504455566, "learning_rate": 1.9e-06, "loss": 73.5, "step": 191 }, { "epoch": 1.92e-05, "grad_norm": 13.303735733032227, "learning_rate": 1.91e-06, "loss": 73.5625, "step": 192 }, { "epoch": 1.93e-05, "grad_norm": 13.176830291748047, "learning_rate": 1.9200000000000003e-06, "loss": 73.5, "step": 193 }, { "epoch": 1.94e-05, "grad_norm": 13.223044395446777, "learning_rate": 1.93e-06, "loss": 73.4375, "step": 194 }, { "epoch": 1.95e-05, "grad_norm": 13.411727905273438, "learning_rate": 1.94e-06, "loss": 73.5, "step": 195 }, { "epoch": 1.96e-05, "grad_norm": 13.083850860595703, "learning_rate": 1.95e-06, "loss": 73.5625, "step": 196 }, { "epoch": 1.97e-05, "grad_norm": 13.732102394104004, "learning_rate": 1.9600000000000003e-06, "loss": 73.5625, "step": 197 }, { "epoch": 1.98e-05, "grad_norm": 13.143880844116211, "learning_rate": 1.9699999999999998e-06, "loss": 73.4375, "step": 198 }, { "epoch": 1.99e-05, "grad_norm": 13.069990158081055, "learning_rate": 1.98e-06, "loss": 73.4375, "step": 199 }, { "epoch": 2e-05, "grad_norm": 13.723267555236816, "learning_rate": 1.99e-06, "loss": 73.5, "step": 200 }, { "epoch": 2.01e-05, "grad_norm": 13.735518455505371, "learning_rate": 2e-06, "loss": 73.6875, "step": 201 }, { "epoch": 2.02e-05, "grad_norm": 13.416829109191895, "learning_rate": 2.0100000000000002e-06, "loss": 73.5, "step": 202 }, { "epoch": 2.03e-05, "grad_norm": 13.453069686889648, "learning_rate": 2.0199999999999997e-06, "loss": 73.5625, "step": 203 }, { "epoch": 2.04e-05, "grad_norm": 13.582813262939453, "learning_rate": 2.03e-06, "loss": 73.5625, "step": 204 }, { "epoch": 2.05e-05, "grad_norm": 13.080293655395508, "learning_rate": 2.0400000000000004e-06, "loss": 73.5, "step": 205 }, { "epoch": 2.06e-05, "grad_norm": 13.556493759155273, "learning_rate": 2.05e-06, "loss": 73.5, "step": 206 }, { "epoch": 2.07e-05, "grad_norm": 12.98852825164795, "learning_rate": 2.06e-06, "loss": 73.5625, "step": 207 }, { "epoch": 2.08e-05, "grad_norm": 13.484770774841309, "learning_rate": 2.07e-06, "loss": 73.5, "step": 208 }, { "epoch": 2.09e-05, "grad_norm": 13.057182312011719, "learning_rate": 2.08e-06, "loss": 73.5625, "step": 209 }, { "epoch": 2.1e-05, "grad_norm": 13.500996589660645, "learning_rate": 2.09e-06, "loss": 73.5, "step": 210 }, { "epoch": 2.11e-05, "grad_norm": 13.239738464355469, "learning_rate": 2.1000000000000002e-06, "loss": 73.5625, "step": 211 }, { "epoch": 2.12e-05, "grad_norm": 13.07319450378418, "learning_rate": 2.11e-06, "loss": 73.5625, "step": 212 }, { "epoch": 2.13e-05, "grad_norm": 13.956418991088867, "learning_rate": 2.12e-06, "loss": 73.5625, "step": 213 }, { "epoch": 2.14e-05, "grad_norm": 13.399481773376465, "learning_rate": 2.13e-06, "loss": 73.4375, "step": 214 }, { "epoch": 2.15e-05, "grad_norm": 13.294981002807617, "learning_rate": 2.14e-06, "loss": 73.5, "step": 215 }, { "epoch": 2.16e-05, "grad_norm": 13.160533905029297, "learning_rate": 2.15e-06, "loss": 73.5, "step": 216 }, { "epoch": 2.17e-05, "grad_norm": 13.458478927612305, "learning_rate": 2.16e-06, "loss": 73.4375, "step": 217 }, { "epoch": 2.18e-05, "grad_norm": 13.246068000793457, "learning_rate": 2.17e-06, "loss": 73.5, "step": 218 }, { "epoch": 2.19e-05, "grad_norm": 13.33403491973877, "learning_rate": 2.1800000000000003e-06, "loss": 73.5, "step": 219 }, { "epoch": 2.2e-05, "grad_norm": 13.391315460205078, "learning_rate": 2.1899999999999998e-06, "loss": 73.5, "step": 220 }, { "epoch": 2.21e-05, "grad_norm": 13.164493560791016, "learning_rate": 2.2e-06, "loss": 73.5, "step": 221 }, { "epoch": 2.22e-05, "grad_norm": 13.39545726776123, "learning_rate": 2.21e-06, "loss": 73.5, "step": 222 }, { "epoch": 2.23e-05, "grad_norm": 13.700584411621094, "learning_rate": 2.22e-06, "loss": 73.5, "step": 223 }, { "epoch": 2.24e-05, "grad_norm": 12.7564115524292, "learning_rate": 2.2300000000000002e-06, "loss": 73.5, "step": 224 }, { "epoch": 2.25e-05, "grad_norm": 13.187485694885254, "learning_rate": 2.24e-06, "loss": 73.5, "step": 225 }, { "epoch": 2.26e-05, "grad_norm": 12.73024845123291, "learning_rate": 2.25e-06, "loss": 73.5, "step": 226 }, { "epoch": 2.27e-05, "grad_norm": 13.395713806152344, "learning_rate": 2.26e-06, "loss": 73.4375, "step": 227 }, { "epoch": 2.28e-05, "grad_norm": 13.149252891540527, "learning_rate": 2.27e-06, "loss": 73.5, "step": 228 }, { "epoch": 2.29e-05, "grad_norm": 12.85241413116455, "learning_rate": 2.28e-06, "loss": 73.4375, "step": 229 }, { "epoch": 2.3e-05, "grad_norm": 13.827969551086426, "learning_rate": 2.29e-06, "loss": 73.5, "step": 230 }, { "epoch": 2.31e-05, "grad_norm": 13.54524040222168, "learning_rate": 2.3e-06, "loss": 73.5625, "step": 231 }, { "epoch": 2.32e-05, "grad_norm": 12.822538375854492, "learning_rate": 2.31e-06, "loss": 73.375, "step": 232 }, { "epoch": 2.33e-05, "grad_norm": 13.30932903289795, "learning_rate": 2.3200000000000002e-06, "loss": 73.5, "step": 233 }, { "epoch": 2.34e-05, "grad_norm": 13.388375282287598, "learning_rate": 2.33e-06, "loss": 73.4375, "step": 234 }, { "epoch": 2.35e-05, "grad_norm": 12.996811866760254, "learning_rate": 2.34e-06, "loss": 73.5, "step": 235 }, { "epoch": 2.36e-05, "grad_norm": 12.992647171020508, "learning_rate": 2.35e-06, "loss": 73.5, "step": 236 }, { "epoch": 2.37e-05, "grad_norm": 13.552807807922363, "learning_rate": 2.36e-06, "loss": 73.4375, "step": 237 }, { "epoch": 2.38e-05, "grad_norm": 13.842035293579102, "learning_rate": 2.37e-06, "loss": 73.4375, "step": 238 }, { "epoch": 2.39e-05, "grad_norm": 13.160901069641113, "learning_rate": 2.38e-06, "loss": 73.375, "step": 239 }, { "epoch": 2.4e-05, "grad_norm": 13.619793891906738, "learning_rate": 2.39e-06, "loss": 73.4375, "step": 240 }, { "epoch": 2.41e-05, "grad_norm": 13.451335906982422, "learning_rate": 2.4000000000000003e-06, "loss": 73.5, "step": 241 }, { "epoch": 2.42e-05, "grad_norm": 13.234610557556152, "learning_rate": 2.41e-06, "loss": 73.5, "step": 242 }, { "epoch": 2.43e-05, "grad_norm": 13.145569801330566, "learning_rate": 2.42e-06, "loss": 73.375, "step": 243 }, { "epoch": 2.44e-05, "grad_norm": 13.658345222473145, "learning_rate": 2.43e-06, "loss": 73.4375, "step": 244 }, { "epoch": 2.45e-05, "grad_norm": 13.420488357543945, "learning_rate": 2.44e-06, "loss": 73.5, "step": 245 }, { "epoch": 2.46e-05, "grad_norm": 13.162190437316895, "learning_rate": 2.4500000000000003e-06, "loss": 73.4375, "step": 246 }, { "epoch": 2.47e-05, "grad_norm": 13.59197998046875, "learning_rate": 2.46e-06, "loss": 73.5, "step": 247 }, { "epoch": 2.48e-05, "grad_norm": 13.829290390014648, "learning_rate": 2.47e-06, "loss": 73.5625, "step": 248 }, { "epoch": 2.49e-05, "grad_norm": 13.297616004943848, "learning_rate": 2.48e-06, "loss": 73.5, "step": 249 }, { "epoch": 2.5e-05, "grad_norm": 13.626100540161133, "learning_rate": 2.49e-06, "loss": 73.4375, "step": 250 }, { "epoch": 2.51e-05, "grad_norm": 13.326560020446777, "learning_rate": 2.5e-06, "loss": 73.5, "step": 251 }, { "epoch": 2.52e-05, "grad_norm": 13.659133911132812, "learning_rate": 2.51e-06, "loss": 73.4375, "step": 252 }, { "epoch": 2.53e-05, "grad_norm": 13.558892250061035, "learning_rate": 2.52e-06, "loss": 73.4375, "step": 253 }, { "epoch": 2.54e-05, "grad_norm": 12.769811630249023, "learning_rate": 2.53e-06, "loss": 73.5, "step": 254 }, { "epoch": 2.55e-05, "grad_norm": 13.625076293945312, "learning_rate": 2.5400000000000002e-06, "loss": 73.5, "step": 255 }, { "epoch": 2.56e-05, "grad_norm": 13.309321403503418, "learning_rate": 2.5499999999999997e-06, "loss": 73.4375, "step": 256 }, { "epoch": 2.57e-05, "grad_norm": 13.672002792358398, "learning_rate": 2.56e-06, "loss": 73.4375, "step": 257 }, { "epoch": 2.58e-05, "grad_norm": 13.397356033325195, "learning_rate": 2.5700000000000004e-06, "loss": 73.5, "step": 258 }, { "epoch": 2.59e-05, "grad_norm": 13.15006160736084, "learning_rate": 2.58e-06, "loss": 73.375, "step": 259 }, { "epoch": 2.6e-05, "grad_norm": 12.989713668823242, "learning_rate": 2.59e-06, "loss": 73.5, "step": 260 }, { "epoch": 2.61e-05, "grad_norm": 13.889228820800781, "learning_rate": 2.6e-06, "loss": 73.5, "step": 261 }, { "epoch": 2.62e-05, "grad_norm": 13.123197555541992, "learning_rate": 2.61e-06, "loss": 73.375, "step": 262 }, { "epoch": 2.63e-05, "grad_norm": 13.125618934631348, "learning_rate": 2.6200000000000003e-06, "loss": 73.4375, "step": 263 }, { "epoch": 2.64e-05, "grad_norm": 13.075492858886719, "learning_rate": 2.63e-06, "loss": 73.4375, "step": 264 }, { "epoch": 2.65e-05, "grad_norm": 13.195477485656738, "learning_rate": 2.64e-06, "loss": 73.25, "step": 265 }, { "epoch": 2.66e-05, "grad_norm": 13.082707405090332, "learning_rate": 2.65e-06, "loss": 73.25, "step": 266 }, { "epoch": 2.67e-05, "grad_norm": 12.753524780273438, "learning_rate": 2.66e-06, "loss": 73.4375, "step": 267 }, { "epoch": 2.68e-05, "grad_norm": 13.304618835449219, "learning_rate": 2.67e-06, "loss": 73.4375, "step": 268 }, { "epoch": 2.69e-05, "grad_norm": 13.050411224365234, "learning_rate": 2.68e-06, "loss": 73.4375, "step": 269 }, { "epoch": 2.7e-05, "grad_norm": 13.13429069519043, "learning_rate": 2.69e-06, "loss": 73.375, "step": 270 }, { "epoch": 2.71e-05, "grad_norm": 13.40479850769043, "learning_rate": 2.7e-06, "loss": 73.375, "step": 271 }, { "epoch": 2.72e-05, "grad_norm": 13.249833106994629, "learning_rate": 2.7100000000000003e-06, "loss": 73.4375, "step": 272 }, { "epoch": 2.73e-05, "grad_norm": 13.233908653259277, "learning_rate": 2.72e-06, "loss": 73.4375, "step": 273 }, { "epoch": 2.74e-05, "grad_norm": 13.229077339172363, "learning_rate": 2.73e-06, "loss": 73.125, "step": 274 }, { "epoch": 2.75e-05, "grad_norm": 13.314671516418457, "learning_rate": 2.74e-06, "loss": 73.3125, "step": 275 }, { "epoch": 2.76e-05, "grad_norm": 13.741567611694336, "learning_rate": 2.75e-06, "loss": 73.5, "step": 276 }, { "epoch": 2.77e-05, "grad_norm": 12.917357444763184, "learning_rate": 2.7600000000000003e-06, "loss": 73.4375, "step": 277 }, { "epoch": 2.78e-05, "grad_norm": 13.566869735717773, "learning_rate": 2.77e-06, "loss": 73.5, "step": 278 }, { "epoch": 2.79e-05, "grad_norm": 13.380074501037598, "learning_rate": 2.78e-06, "loss": 73.4375, "step": 279 }, { "epoch": 2.8e-05, "grad_norm": 13.790977478027344, "learning_rate": 2.7900000000000004e-06, "loss": 73.3125, "step": 280 }, { "epoch": 2.81e-05, "grad_norm": 13.409905433654785, "learning_rate": 2.8e-06, "loss": 73.375, "step": 281 }, { "epoch": 2.82e-05, "grad_norm": 13.063041687011719, "learning_rate": 2.81e-06, "loss": 73.375, "step": 282 }, { "epoch": 2.83e-05, "grad_norm": 13.631767272949219, "learning_rate": 2.82e-06, "loss": 73.375, "step": 283 }, { "epoch": 2.84e-05, "grad_norm": 13.662729263305664, "learning_rate": 2.83e-06, "loss": 73.375, "step": 284 }, { "epoch": 2.85e-05, "grad_norm": 13.42518424987793, "learning_rate": 2.84e-06, "loss": 73.4375, "step": 285 }, { "epoch": 2.86e-05, "grad_norm": 13.8087739944458, "learning_rate": 2.8500000000000002e-06, "loss": 73.5, "step": 286 }, { "epoch": 2.87e-05, "grad_norm": 13.71200180053711, "learning_rate": 2.86e-06, "loss": 73.3125, "step": 287 }, { "epoch": 2.88e-05, "grad_norm": 13.234236717224121, "learning_rate": 2.87e-06, "loss": 73.4375, "step": 288 }, { "epoch": 2.89e-05, "grad_norm": 13.711640357971191, "learning_rate": 2.88e-06, "loss": 73.375, "step": 289 }, { "epoch": 2.9e-05, "grad_norm": 13.486573219299316, "learning_rate": 2.89e-06, "loss": 73.375, "step": 290 }, { "epoch": 2.91e-05, "grad_norm": 13.30084228515625, "learning_rate": 2.9e-06, "loss": 73.4375, "step": 291 }, { "epoch": 2.92e-05, "grad_norm": 13.337886810302734, "learning_rate": 2.91e-06, "loss": 73.375, "step": 292 }, { "epoch": 2.93e-05, "grad_norm": 13.75003719329834, "learning_rate": 2.92e-06, "loss": 73.375, "step": 293 }, { "epoch": 2.94e-05, "grad_norm": 14.156579971313477, "learning_rate": 2.9300000000000003e-06, "loss": 73.5, "step": 294 }, { "epoch": 2.95e-05, "grad_norm": 13.412623405456543, "learning_rate": 2.94e-06, "loss": 73.125, "step": 295 }, { "epoch": 2.96e-05, "grad_norm": 12.932133674621582, "learning_rate": 2.9499999999999997e-06, "loss": 73.25, "step": 296 }, { "epoch": 2.97e-05, "grad_norm": 13.164972305297852, "learning_rate": 2.96e-06, "loss": 73.1875, "step": 297 }, { "epoch": 2.98e-05, "grad_norm": 13.32223129272461, "learning_rate": 2.97e-06, "loss": 73.3125, "step": 298 }, { "epoch": 2.99e-05, "grad_norm": 13.821455001831055, "learning_rate": 2.98e-06, "loss": 73.4375, "step": 299 }, { "epoch": 3e-05, "grad_norm": 13.473308563232422, "learning_rate": 2.99e-06, "loss": 73.5, "step": 300 }, { "epoch": 3.01e-05, "grad_norm": 13.666665077209473, "learning_rate": 3e-06, "loss": 73.3125, "step": 301 }, { "epoch": 3.02e-05, "grad_norm": 13.532960891723633, "learning_rate": 3.01e-06, "loss": 73.1875, "step": 302 }, { "epoch": 3.03e-05, "grad_norm": 13.009042739868164, "learning_rate": 3.02e-06, "loss": 73.375, "step": 303 }, { "epoch": 3.04e-05, "grad_norm": 13.482487678527832, "learning_rate": 3.0300000000000002e-06, "loss": 73.3125, "step": 304 }, { "epoch": 3.05e-05, "grad_norm": 13.215324401855469, "learning_rate": 3.04e-06, "loss": 73.1875, "step": 305 }, { "epoch": 3.06e-05, "grad_norm": 13.712039947509766, "learning_rate": 3.05e-06, "loss": 73.3125, "step": 306 }, { "epoch": 3.07e-05, "grad_norm": 13.075980186462402, "learning_rate": 3.0600000000000003e-06, "loss": 73.25, "step": 307 }, { "epoch": 3.08e-05, "grad_norm": 13.958653450012207, "learning_rate": 3.0700000000000003e-06, "loss": 73.3125, "step": 308 }, { "epoch": 3.09e-05, "grad_norm": 12.661245346069336, "learning_rate": 3.0799999999999997e-06, "loss": 73.25, "step": 309 }, { "epoch": 3.1e-05, "grad_norm": 13.312789916992188, "learning_rate": 3.0900000000000005e-06, "loss": 73.125, "step": 310 }, { "epoch": 3.11e-05, "grad_norm": 13.056986808776855, "learning_rate": 3.1e-06, "loss": 73.375, "step": 311 }, { "epoch": 3.12e-05, "grad_norm": 13.239110946655273, "learning_rate": 3.11e-06, "loss": 73.25, "step": 312 }, { "epoch": 3.13e-05, "grad_norm": 13.4395170211792, "learning_rate": 3.1199999999999998e-06, "loss": 73.375, "step": 313 }, { "epoch": 3.14e-05, "grad_norm": 13.566024780273438, "learning_rate": 3.13e-06, "loss": 73.3125, "step": 314 }, { "epoch": 3.15e-05, "grad_norm": 13.323553085327148, "learning_rate": 3.14e-06, "loss": 73.1875, "step": 315 }, { "epoch": 3.16e-05, "grad_norm": 13.510903358459473, "learning_rate": 3.15e-06, "loss": 73.375, "step": 316 }, { "epoch": 3.17e-05, "grad_norm": 13.496997833251953, "learning_rate": 3.1600000000000002e-06, "loss": 73.375, "step": 317 }, { "epoch": 3.18e-05, "grad_norm": 12.994120597839355, "learning_rate": 3.17e-06, "loss": 73.1875, "step": 318 }, { "epoch": 3.19e-05, "grad_norm": 13.158904075622559, "learning_rate": 3.18e-06, "loss": 73.1875, "step": 319 }, { "epoch": 3.2e-05, "grad_norm": 12.74785327911377, "learning_rate": 3.19e-06, "loss": 73.25, "step": 320 }, { "epoch": 3.21e-05, "grad_norm": 13.40980339050293, "learning_rate": 3.2000000000000003e-06, "loss": 73.25, "step": 321 }, { "epoch": 3.22e-05, "grad_norm": 13.221853256225586, "learning_rate": 3.21e-06, "loss": 73.0625, "step": 322 }, { "epoch": 3.23e-05, "grad_norm": 13.401768684387207, "learning_rate": 3.2199999999999997e-06, "loss": 73.1875, "step": 323 }, { "epoch": 3.24e-05, "grad_norm": 12.98742961883545, "learning_rate": 3.2300000000000004e-06, "loss": 73.25, "step": 324 }, { "epoch": 3.25e-05, "grad_norm": 13.389715194702148, "learning_rate": 3.2400000000000003e-06, "loss": 73.25, "step": 325 }, { "epoch": 3.26e-05, "grad_norm": 13.654870986938477, "learning_rate": 3.25e-06, "loss": 73.125, "step": 326 }, { "epoch": 3.27e-05, "grad_norm": 12.968194961547852, "learning_rate": 3.2600000000000006e-06, "loss": 73.25, "step": 327 }, { "epoch": 3.28e-05, "grad_norm": 13.770217895507812, "learning_rate": 3.27e-06, "loss": 73.125, "step": 328 }, { "epoch": 3.29e-05, "grad_norm": 13.335156440734863, "learning_rate": 3.28e-06, "loss": 73.25, "step": 329 }, { "epoch": 3.3e-05, "grad_norm": 12.833738327026367, "learning_rate": 3.29e-06, "loss": 73.1875, "step": 330 }, { "epoch": 3.31e-05, "grad_norm": 13.088351249694824, "learning_rate": 3.3e-06, "loss": 73.1875, "step": 331 }, { "epoch": 3.32e-05, "grad_norm": 13.253767967224121, "learning_rate": 3.31e-06, "loss": 73.125, "step": 332 }, { "epoch": 3.33e-05, "grad_norm": 13.729750633239746, "learning_rate": 3.32e-06, "loss": 73.0625, "step": 333 }, { "epoch": 3.34e-05, "grad_norm": 13.646214485168457, "learning_rate": 3.3300000000000003e-06, "loss": 73.1875, "step": 334 }, { "epoch": 3.35e-05, "grad_norm": 13.651313781738281, "learning_rate": 3.34e-06, "loss": 73.3125, "step": 335 }, { "epoch": 3.36e-05, "grad_norm": 13.795608520507812, "learning_rate": 3.35e-06, "loss": 73.375, "step": 336 }, { "epoch": 3.37e-05, "grad_norm": 13.318341255187988, "learning_rate": 3.3599999999999996e-06, "loss": 73.125, "step": 337 }, { "epoch": 3.38e-05, "grad_norm": 13.053764343261719, "learning_rate": 3.3700000000000003e-06, "loss": 73.1875, "step": 338 }, { "epoch": 3.39e-05, "grad_norm": 13.565455436706543, "learning_rate": 3.3800000000000002e-06, "loss": 73.1875, "step": 339 }, { "epoch": 3.4e-05, "grad_norm": 13.372631072998047, "learning_rate": 3.3899999999999997e-06, "loss": 73.125, "step": 340 }, { "epoch": 3.41e-05, "grad_norm": 13.144970893859863, "learning_rate": 3.4000000000000005e-06, "loss": 73.125, "step": 341 }, { "epoch": 3.42e-05, "grad_norm": 13.336371421813965, "learning_rate": 3.41e-06, "loss": 73.1875, "step": 342 }, { "epoch": 3.43e-05, "grad_norm": 13.389857292175293, "learning_rate": 3.42e-06, "loss": 73.0625, "step": 343 }, { "epoch": 3.44e-05, "grad_norm": 13.658300399780273, "learning_rate": 3.43e-06, "loss": 73.1875, "step": 344 }, { "epoch": 3.45e-05, "grad_norm": 13.079704284667969, "learning_rate": 3.44e-06, "loss": 73.125, "step": 345 }, { "epoch": 3.46e-05, "grad_norm": 12.891420364379883, "learning_rate": 3.45e-06, "loss": 73.0625, "step": 346 }, { "epoch": 3.47e-05, "grad_norm": 13.24589729309082, "learning_rate": 3.46e-06, "loss": 73.3125, "step": 347 }, { "epoch": 3.48e-05, "grad_norm": 12.827505111694336, "learning_rate": 3.4700000000000002e-06, "loss": 73.1875, "step": 348 }, { "epoch": 3.49e-05, "grad_norm": 13.241964340209961, "learning_rate": 3.48e-06, "loss": 73.125, "step": 349 }, { "epoch": 3.5e-05, "grad_norm": 12.945211410522461, "learning_rate": 3.49e-06, "loss": 73.25, "step": 350 }, { "epoch": 3.51e-05, "grad_norm": 13.305326461791992, "learning_rate": 3.5000000000000004e-06, "loss": 73.125, "step": 351 }, { "epoch": 3.52e-05, "grad_norm": 13.148869514465332, "learning_rate": 3.5100000000000003e-06, "loss": 73.1875, "step": 352 }, { "epoch": 3.53e-05, "grad_norm": 13.825087547302246, "learning_rate": 3.52e-06, "loss": 73.0625, "step": 353 }, { "epoch": 3.54e-05, "grad_norm": 13.412226676940918, "learning_rate": 3.5299999999999997e-06, "loss": 73.1875, "step": 354 }, { "epoch": 3.55e-05, "grad_norm": 12.815329551696777, "learning_rate": 3.5400000000000004e-06, "loss": 73.125, "step": 355 }, { "epoch": 3.56e-05, "grad_norm": 13.162130355834961, "learning_rate": 3.55e-06, "loss": 73.125, "step": 356 }, { "epoch": 3.57e-05, "grad_norm": 13.487090110778809, "learning_rate": 3.56e-06, "loss": 73.0625, "step": 357 }, { "epoch": 3.58e-05, "grad_norm": 13.002684593200684, "learning_rate": 3.57e-06, "loss": 73.125, "step": 358 }, { "epoch": 3.59e-05, "grad_norm": 13.55034065246582, "learning_rate": 3.58e-06, "loss": 73.125, "step": 359 }, { "epoch": 3.6e-05, "grad_norm": 13.577582359313965, "learning_rate": 3.59e-06, "loss": 73.0625, "step": 360 }, { "epoch": 3.61e-05, "grad_norm": 13.844265937805176, "learning_rate": 3.6e-06, "loss": 73.3125, "step": 361 }, { "epoch": 3.62e-05, "grad_norm": 13.737598419189453, "learning_rate": 3.61e-06, "loss": 73.0, "step": 362 }, { "epoch": 3.63e-05, "grad_norm": 12.993484497070312, "learning_rate": 3.62e-06, "loss": 73.0625, "step": 363 }, { "epoch": 3.64e-05, "grad_norm": 13.497552871704102, "learning_rate": 3.63e-06, "loss": 73.1875, "step": 364 }, { "epoch": 3.65e-05, "grad_norm": 13.50912857055664, "learning_rate": 3.6400000000000003e-06, "loss": 73.25, "step": 365 }, { "epoch": 3.66e-05, "grad_norm": 13.057828903198242, "learning_rate": 3.65e-06, "loss": 73.0625, "step": 366 }, { "epoch": 3.67e-05, "grad_norm": 13.317353248596191, "learning_rate": 3.66e-06, "loss": 73.125, "step": 367 }, { "epoch": 3.68e-05, "grad_norm": 13.579947471618652, "learning_rate": 3.6700000000000004e-06, "loss": 73.125, "step": 368 }, { "epoch": 3.69e-05, "grad_norm": 13.080233573913574, "learning_rate": 3.6800000000000003e-06, "loss": 73.125, "step": 369 }, { "epoch": 3.7e-05, "grad_norm": 13.080486297607422, "learning_rate": 3.69e-06, "loss": 73.125, "step": 370 }, { "epoch": 3.71e-05, "grad_norm": 13.311331748962402, "learning_rate": 3.6999999999999997e-06, "loss": 73.125, "step": 371 }, { "epoch": 3.72e-05, "grad_norm": 13.139957427978516, "learning_rate": 3.7100000000000005e-06, "loss": 73.125, "step": 372 }, { "epoch": 3.73e-05, "grad_norm": 13.586795806884766, "learning_rate": 3.72e-06, "loss": 73.25, "step": 373 }, { "epoch": 3.74e-05, "grad_norm": 13.488346099853516, "learning_rate": 3.73e-06, "loss": 73.125, "step": 374 }, { "epoch": 3.75e-05, "grad_norm": 12.536556243896484, "learning_rate": 3.74e-06, "loss": 73.0, "step": 375 }, { "epoch": 3.76e-05, "grad_norm": 13.239697456359863, "learning_rate": 3.75e-06, "loss": 73.125, "step": 376 }, { "epoch": 3.77e-05, "grad_norm": 13.392301559448242, "learning_rate": 3.76e-06, "loss": 73.0625, "step": 377 }, { "epoch": 3.78e-05, "grad_norm": 13.62486743927002, "learning_rate": 3.77e-06, "loss": 73.125, "step": 378 }, { "epoch": 3.79e-05, "grad_norm": 13.354044914245605, "learning_rate": 3.7800000000000002e-06, "loss": 73.125, "step": 379 }, { "epoch": 3.8e-05, "grad_norm": 13.179764747619629, "learning_rate": 3.79e-06, "loss": 73.0625, "step": 380 }, { "epoch": 3.81e-05, "grad_norm": 12.997714042663574, "learning_rate": 3.8e-06, "loss": 73.0, "step": 381 }, { "epoch": 3.82e-05, "grad_norm": 13.98970890045166, "learning_rate": 3.8100000000000004e-06, "loss": 73.0, "step": 382 }, { "epoch": 3.83e-05, "grad_norm": 12.927799224853516, "learning_rate": 3.82e-06, "loss": 73.125, "step": 383 }, { "epoch": 3.84e-05, "grad_norm": 13.921652793884277, "learning_rate": 3.83e-06, "loss": 73.0, "step": 384 }, { "epoch": 3.85e-05, "grad_norm": 13.006847381591797, "learning_rate": 3.8400000000000005e-06, "loss": 73.0, "step": 385 }, { "epoch": 3.86e-05, "grad_norm": 13.084640502929688, "learning_rate": 3.85e-06, "loss": 73.0625, "step": 386 }, { "epoch": 3.87e-05, "grad_norm": 12.917254447937012, "learning_rate": 3.86e-06, "loss": 73.0625, "step": 387 }, { "epoch": 3.88e-05, "grad_norm": 13.406126022338867, "learning_rate": 3.87e-06, "loss": 73.0625, "step": 388 }, { "epoch": 3.89e-05, "grad_norm": 13.393180847167969, "learning_rate": 3.88e-06, "loss": 73.0, "step": 389 }, { "epoch": 3.9e-05, "grad_norm": 13.257850646972656, "learning_rate": 3.89e-06, "loss": 73.0, "step": 390 }, { "epoch": 3.91e-05, "grad_norm": 13.004161834716797, "learning_rate": 3.9e-06, "loss": 73.0, "step": 391 }, { "epoch": 3.92e-05, "grad_norm": 13.656088829040527, "learning_rate": 3.910000000000001e-06, "loss": 73.0625, "step": 392 }, { "epoch": 3.93e-05, "grad_norm": 13.50597095489502, "learning_rate": 3.920000000000001e-06, "loss": 73.0625, "step": 393 }, { "epoch": 3.94e-05, "grad_norm": 13.430299758911133, "learning_rate": 3.93e-06, "loss": 73.0, "step": 394 }, { "epoch": 3.95e-05, "grad_norm": 13.662679672241211, "learning_rate": 3.9399999999999995e-06, "loss": 73.125, "step": 395 }, { "epoch": 3.96e-05, "grad_norm": 13.804597854614258, "learning_rate": 3.95e-06, "loss": 73.0625, "step": 396 }, { "epoch": 3.97e-05, "grad_norm": 13.883934020996094, "learning_rate": 3.96e-06, "loss": 73.0625, "step": 397 }, { "epoch": 3.98e-05, "grad_norm": 14.236750602722168, "learning_rate": 3.97e-06, "loss": 73.0, "step": 398 }, { "epoch": 3.99e-05, "grad_norm": 13.254412651062012, "learning_rate": 3.98e-06, "loss": 73.0, "step": 399 }, { "epoch": 4e-05, "grad_norm": 12.98829460144043, "learning_rate": 3.99e-06, "loss": 72.9375, "step": 400 }, { "epoch": 4.01e-05, "grad_norm": 12.85839557647705, "learning_rate": 4e-06, "loss": 73.0625, "step": 401 }, { "epoch": 4.02e-05, "grad_norm": 13.245451927185059, "learning_rate": 4.01e-06, "loss": 73.1875, "step": 402 }, { "epoch": 4.03e-05, "grad_norm": 13.241745948791504, "learning_rate": 4.0200000000000005e-06, "loss": 73.0, "step": 403 }, { "epoch": 4.04e-05, "grad_norm": 13.75320053100586, "learning_rate": 4.03e-06, "loss": 73.0, "step": 404 }, { "epoch": 4.05e-05, "grad_norm": 13.509017944335938, "learning_rate": 4.0399999999999994e-06, "loss": 73.0, "step": 405 }, { "epoch": 4.06e-05, "grad_norm": 13.018754005432129, "learning_rate": 4.05e-06, "loss": 73.0, "step": 406 }, { "epoch": 4.07e-05, "grad_norm": 13.013233184814453, "learning_rate": 4.06e-06, "loss": 73.0, "step": 407 }, { "epoch": 4.08e-05, "grad_norm": 13.415353775024414, "learning_rate": 4.07e-06, "loss": 73.0, "step": 408 }, { "epoch": 4.09e-05, "grad_norm": 13.008545875549316, "learning_rate": 4.080000000000001e-06, "loss": 73.0, "step": 409 }, { "epoch": 4.1e-05, "grad_norm": 13.839768409729004, "learning_rate": 4.09e-06, "loss": 73.0, "step": 410 }, { "epoch": 4.11e-05, "grad_norm": 12.9841947555542, "learning_rate": 4.1e-06, "loss": 73.0, "step": 411 }, { "epoch": 4.12e-05, "grad_norm": 13.155670166015625, "learning_rate": 4.11e-06, "loss": 73.0, "step": 412 }, { "epoch": 4.13e-05, "grad_norm": 13.276348114013672, "learning_rate": 4.12e-06, "loss": 72.9375, "step": 413 }, { "epoch": 4.14e-05, "grad_norm": 13.74177360534668, "learning_rate": 4.13e-06, "loss": 73.0, "step": 414 }, { "epoch": 4.15e-05, "grad_norm": 13.342119216918945, "learning_rate": 4.14e-06, "loss": 73.0, "step": 415 }, { "epoch": 4.16e-05, "grad_norm": 13.325584411621094, "learning_rate": 4.15e-06, "loss": 73.0, "step": 416 }, { "epoch": 4.17e-05, "grad_norm": 13.430407524108887, "learning_rate": 4.16e-06, "loss": 73.0, "step": 417 }, { "epoch": 4.18e-05, "grad_norm": 12.990144729614258, "learning_rate": 4.17e-06, "loss": 73.0, "step": 418 }, { "epoch": 4.19e-05, "grad_norm": 13.573437690734863, "learning_rate": 4.18e-06, "loss": 73.0, "step": 419 }, { "epoch": 4.2e-05, "grad_norm": 13.521942138671875, "learning_rate": 4.1900000000000005e-06, "loss": 73.0625, "step": 420 }, { "epoch": 4.21e-05, "grad_norm": 13.434293746948242, "learning_rate": 4.2000000000000004e-06, "loss": 73.0, "step": 421 }, { "epoch": 4.22e-05, "grad_norm": 13.26038932800293, "learning_rate": 4.2099999999999995e-06, "loss": 73.0, "step": 422 }, { "epoch": 4.23e-05, "grad_norm": 13.418600082397461, "learning_rate": 4.22e-06, "loss": 73.0, "step": 423 }, { "epoch": 4.24e-05, "grad_norm": 13.086088180541992, "learning_rate": 4.23e-06, "loss": 72.9375, "step": 424 }, { "epoch": 4.25e-05, "grad_norm": 13.842316627502441, "learning_rate": 4.24e-06, "loss": 73.0, "step": 425 }, { "epoch": 4.26e-05, "grad_norm": 13.62370777130127, "learning_rate": 4.250000000000001e-06, "loss": 72.9375, "step": 426 }, { "epoch": 4.27e-05, "grad_norm": 13.237730026245117, "learning_rate": 4.26e-06, "loss": 72.875, "step": 427 }, { "epoch": 4.28e-05, "grad_norm": 13.425525665283203, "learning_rate": 4.27e-06, "loss": 72.9375, "step": 428 }, { "epoch": 4.29e-05, "grad_norm": 13.190109252929688, "learning_rate": 4.28e-06, "loss": 73.0, "step": 429 }, { "epoch": 4.3e-05, "grad_norm": 12.757128715515137, "learning_rate": 4.2900000000000004e-06, "loss": 73.0, "step": 430 }, { "epoch": 4.31e-05, "grad_norm": 13.181131362915039, "learning_rate": 4.3e-06, "loss": 72.9375, "step": 431 }, { "epoch": 4.32e-05, "grad_norm": 13.78049087524414, "learning_rate": 4.31e-06, "loss": 72.8125, "step": 432 }, { "epoch": 4.33e-05, "grad_norm": 13.295443534851074, "learning_rate": 4.32e-06, "loss": 72.875, "step": 433 }, { "epoch": 4.34e-05, "grad_norm": 13.687934875488281, "learning_rate": 4.33e-06, "loss": 72.875, "step": 434 }, { "epoch": 4.35e-05, "grad_norm": 13.669114112854004, "learning_rate": 4.34e-06, "loss": 72.75, "step": 435 }, { "epoch": 4.36e-05, "grad_norm": 13.830716133117676, "learning_rate": 4.35e-06, "loss": 72.9375, "step": 436 }, { "epoch": 4.37e-05, "grad_norm": 13.447793960571289, "learning_rate": 4.360000000000001e-06, "loss": 72.6875, "step": 437 }, { "epoch": 4.38e-05, "grad_norm": 13.497492790222168, "learning_rate": 4.37e-06, "loss": 72.8125, "step": 438 }, { "epoch": 4.39e-05, "grad_norm": 13.606331825256348, "learning_rate": 4.3799999999999996e-06, "loss": 73.0, "step": 439 }, { "epoch": 4.4e-05, "grad_norm": 13.782898902893066, "learning_rate": 4.39e-06, "loss": 72.875, "step": 440 }, { "epoch": 4.41e-05, "grad_norm": 13.709412574768066, "learning_rate": 4.4e-06, "loss": 72.875, "step": 441 }, { "epoch": 4.42e-05, "grad_norm": 13.186978340148926, "learning_rate": 4.41e-06, "loss": 72.6875, "step": 442 }, { "epoch": 4.43e-05, "grad_norm": 13.370944023132324, "learning_rate": 4.42e-06, "loss": 72.875, "step": 443 }, { "epoch": 4.44e-05, "grad_norm": 13.442479133605957, "learning_rate": 4.43e-06, "loss": 72.9375, "step": 444 }, { "epoch": 4.45e-05, "grad_norm": 13.615199089050293, "learning_rate": 4.44e-06, "loss": 72.75, "step": 445 }, { "epoch": 4.46e-05, "grad_norm": 13.638628005981445, "learning_rate": 4.45e-06, "loss": 72.6875, "step": 446 }, { "epoch": 4.47e-05, "grad_norm": 13.779143333435059, "learning_rate": 4.4600000000000005e-06, "loss": 72.625, "step": 447 }, { "epoch": 4.48e-05, "grad_norm": 13.44798469543457, "learning_rate": 4.47e-06, "loss": 72.625, "step": 448 }, { "epoch": 4.49e-05, "grad_norm": 13.578337669372559, "learning_rate": 4.48e-06, "loss": 72.8125, "step": 449 }, { "epoch": 4.5e-05, "grad_norm": 14.065568923950195, "learning_rate": 4.49e-06, "loss": 72.75, "step": 450 }, { "epoch": 4.51e-05, "grad_norm": 13.366068840026855, "learning_rate": 4.5e-06, "loss": 72.625, "step": 451 }, { "epoch": 4.52e-05, "grad_norm": 13.885690689086914, "learning_rate": 4.51e-06, "loss": 72.5625, "step": 452 }, { "epoch": 4.53e-05, "grad_norm": 13.551187515258789, "learning_rate": 4.52e-06, "loss": 72.6875, "step": 453 }, { "epoch": 4.54e-05, "grad_norm": 13.381160736083984, "learning_rate": 4.530000000000001e-06, "loss": 72.625, "step": 454 }, { "epoch": 4.55e-05, "grad_norm": 13.608617782592773, "learning_rate": 4.54e-06, "loss": 72.6875, "step": 455 }, { "epoch": 4.56e-05, "grad_norm": 13.639422416687012, "learning_rate": 4.55e-06, "loss": 72.625, "step": 456 }, { "epoch": 4.57e-05, "grad_norm": 13.869024276733398, "learning_rate": 4.56e-06, "loss": 72.5, "step": 457 }, { "epoch": 4.58e-05, "grad_norm": 13.874001502990723, "learning_rate": 4.57e-06, "loss": 72.625, "step": 458 }, { "epoch": 4.59e-05, "grad_norm": 13.535733222961426, "learning_rate": 4.58e-06, "loss": 72.5, "step": 459 }, { "epoch": 4.6e-05, "grad_norm": 13.238990783691406, "learning_rate": 4.59e-06, "loss": 72.5, "step": 460 }, { "epoch": 4.61e-05, "grad_norm": 13.706893920898438, "learning_rate": 4.6e-06, "loss": 72.5, "step": 461 }, { "epoch": 4.62e-05, "grad_norm": 13.469103813171387, "learning_rate": 4.61e-06, "loss": 72.5625, "step": 462 }, { "epoch": 4.63e-05, "grad_norm": 13.195649147033691, "learning_rate": 4.62e-06, "loss": 72.5, "step": 463 }, { "epoch": 4.64e-05, "grad_norm": 13.530961990356445, "learning_rate": 4.6300000000000006e-06, "loss": 72.5625, "step": 464 }, { "epoch": 4.65e-05, "grad_norm": 13.53665542602539, "learning_rate": 4.6400000000000005e-06, "loss": 72.5, "step": 465 }, { "epoch": 4.66e-05, "grad_norm": 12.728484153747559, "learning_rate": 4.6499999999999995e-06, "loss": 72.5625, "step": 466 }, { "epoch": 4.67e-05, "grad_norm": 13.709749221801758, "learning_rate": 4.66e-06, "loss": 72.5, "step": 467 }, { "epoch": 4.68e-05, "grad_norm": 13.372309684753418, "learning_rate": 4.67e-06, "loss": 72.5625, "step": 468 }, { "epoch": 4.69e-05, "grad_norm": 13.057953834533691, "learning_rate": 4.68e-06, "loss": 72.5, "step": 469 }, { "epoch": 4.7e-05, "grad_norm": 13.628046989440918, "learning_rate": 4.69e-06, "loss": 72.4375, "step": 470 }, { "epoch": 4.71e-05, "grad_norm": 13.151557922363281, "learning_rate": 4.7e-06, "loss": 72.5, "step": 471 }, { "epoch": 4.72e-05, "grad_norm": 13.463614463806152, "learning_rate": 4.71e-06, "loss": 72.5625, "step": 472 }, { "epoch": 4.73e-05, "grad_norm": 13.199589729309082, "learning_rate": 4.72e-06, "loss": 72.5, "step": 473 }, { "epoch": 4.74e-05, "grad_norm": 13.053608894348145, "learning_rate": 4.7300000000000005e-06, "loss": 72.5, "step": 474 }, { "epoch": 4.75e-05, "grad_norm": 13.356788635253906, "learning_rate": 4.74e-06, "loss": 72.5, "step": 475 }, { "epoch": 4.76e-05, "grad_norm": 13.800188064575195, "learning_rate": 4.75e-06, "loss": 72.5, "step": 476 }, { "epoch": 4.77e-05, "grad_norm": 13.535609245300293, "learning_rate": 4.76e-06, "loss": 72.4375, "step": 477 }, { "epoch": 4.78e-05, "grad_norm": 12.961105346679688, "learning_rate": 4.77e-06, "loss": 72.5, "step": 478 }, { "epoch": 4.79e-05, "grad_norm": 13.4580659866333, "learning_rate": 4.78e-06, "loss": 72.5, "step": 479 }, { "epoch": 4.8e-05, "grad_norm": 13.220132827758789, "learning_rate": 4.79e-06, "loss": 72.5, "step": 480 }, { "epoch": 4.81e-05, "grad_norm": 13.80113697052002, "learning_rate": 4.800000000000001e-06, "loss": 72.375, "step": 481 }, { "epoch": 4.82e-05, "grad_norm": 13.622671127319336, "learning_rate": 4.8100000000000005e-06, "loss": 72.5, "step": 482 }, { "epoch": 4.83e-05, "grad_norm": 13.397513389587402, "learning_rate": 4.82e-06, "loss": 72.5, "step": 483 }, { "epoch": 4.84e-05, "grad_norm": 14.030313491821289, "learning_rate": 4.83e-06, "loss": 72.4375, "step": 484 }, { "epoch": 4.85e-05, "grad_norm": 13.639060020446777, "learning_rate": 4.84e-06, "loss": 72.4375, "step": 485 }, { "epoch": 4.86e-05, "grad_norm": 13.650350570678711, "learning_rate": 4.85e-06, "loss": 72.3125, "step": 486 }, { "epoch": 4.87e-05, "grad_norm": 13.899084091186523, "learning_rate": 4.86e-06, "loss": 72.3125, "step": 487 }, { "epoch": 4.88e-05, "grad_norm": 13.809762001037598, "learning_rate": 4.87e-06, "loss": 72.375, "step": 488 }, { "epoch": 4.89e-05, "grad_norm": 13.373944282531738, "learning_rate": 4.88e-06, "loss": 72.25, "step": 489 }, { "epoch": 4.9e-05, "grad_norm": 13.971879005432129, "learning_rate": 4.89e-06, "loss": 72.4375, "step": 490 }, { "epoch": 4.91e-05, "grad_norm": 13.291943550109863, "learning_rate": 4.9000000000000005e-06, "loss": 72.4375, "step": 491 }, { "epoch": 4.92e-05, "grad_norm": 13.483219146728516, "learning_rate": 4.9100000000000004e-06, "loss": 72.375, "step": 492 }, { "epoch": 4.93e-05, "grad_norm": 13.614572525024414, "learning_rate": 4.92e-06, "loss": 72.375, "step": 493 }, { "epoch": 4.94e-05, "grad_norm": 13.22291374206543, "learning_rate": 4.929999999999999e-06, "loss": 72.25, "step": 494 }, { "epoch": 4.95e-05, "grad_norm": 13.640573501586914, "learning_rate": 4.94e-06, "loss": 72.375, "step": 495 }, { "epoch": 4.96e-05, "grad_norm": 12.912464141845703, "learning_rate": 4.95e-06, "loss": 72.1875, "step": 496 }, { "epoch": 4.97e-05, "grad_norm": 13.510415077209473, "learning_rate": 4.96e-06, "loss": 72.25, "step": 497 }, { "epoch": 4.98e-05, "grad_norm": 13.389180183410645, "learning_rate": 4.970000000000001e-06, "loss": 72.125, "step": 498 }, { "epoch": 4.99e-05, "grad_norm": 13.56809139251709, "learning_rate": 4.98e-06, "loss": 72.25, "step": 499 }, { "epoch": 5e-05, "grad_norm": 13.981008529663086, "learning_rate": 4.99e-06, "loss": 72.0625, "step": 500 }, { "epoch": 5e-05, "eval_loss": 9.004996299743652, "eval_runtime": 364.2712, "eval_samples_per_second": 27.452, "eval_steps_per_second": 1.716, "step": 500 }, { "epoch": 5.01e-05, "grad_norm": 12.755992889404297, "learning_rate": 5e-06, "loss": 72.125, "step": 501 }, { "epoch": 5.02e-05, "grad_norm": 13.567408561706543, "learning_rate": 5.01e-06, "loss": 72.1875, "step": 502 }, { "epoch": 5.03e-05, "grad_norm": 13.13305377960205, "learning_rate": 5.02e-06, "loss": 72.125, "step": 503 }, { "epoch": 5.04e-05, "grad_norm": 13.306023597717285, "learning_rate": 5.03e-06, "loss": 72.0625, "step": 504 }, { "epoch": 5.05e-05, "grad_norm": 13.41368579864502, "learning_rate": 5.04e-06, "loss": 72.1875, "step": 505 }, { "epoch": 5.06e-05, "grad_norm": 13.642829895019531, "learning_rate": 5.05e-06, "loss": 72.0625, "step": 506 }, { "epoch": 5.07e-05, "grad_norm": 13.747808456420898, "learning_rate": 5.06e-06, "loss": 72.0625, "step": 507 }, { "epoch": 5.08e-05, "grad_norm": 13.421524047851562, "learning_rate": 5.070000000000001e-06, "loss": 72.125, "step": 508 }, { "epoch": 5.09e-05, "grad_norm": 13.79678726196289, "learning_rate": 5.0800000000000005e-06, "loss": 72.0, "step": 509 }, { "epoch": 5.1e-05, "grad_norm": 13.362953186035156, "learning_rate": 5.09e-06, "loss": 72.0, "step": 510 }, { "epoch": 5.11e-05, "grad_norm": 13.886173248291016, "learning_rate": 5.0999999999999995e-06, "loss": 72.0, "step": 511 }, { "epoch": 5.12e-05, "grad_norm": 13.49032211303711, "learning_rate": 5.11e-06, "loss": 72.0625, "step": 512 }, { "epoch": 5.13e-05, "grad_norm": 13.661921501159668, "learning_rate": 5.12e-06, "loss": 72.0625, "step": 513 }, { "epoch": 5.14e-05, "grad_norm": 13.80624771118164, "learning_rate": 5.13e-06, "loss": 72.0625, "step": 514 }, { "epoch": 5.15e-05, "grad_norm": 13.84620475769043, "learning_rate": 5.140000000000001e-06, "loss": 72.0, "step": 515 }, { "epoch": 5.16e-05, "grad_norm": 13.160120010375977, "learning_rate": 5.15e-06, "loss": 72.0625, "step": 516 }, { "epoch": 5.17e-05, "grad_norm": 13.24973201751709, "learning_rate": 5.16e-06, "loss": 72.125, "step": 517 }, { "epoch": 5.18e-05, "grad_norm": 14.063905715942383, "learning_rate": 5.17e-06, "loss": 72.0, "step": 518 }, { "epoch": 5.19e-05, "grad_norm": 13.477499008178711, "learning_rate": 5.18e-06, "loss": 72.0, "step": 519 }, { "epoch": 5.2e-05, "grad_norm": 13.442255020141602, "learning_rate": 5.19e-06, "loss": 72.0, "step": 520 }, { "epoch": 5.21e-05, "grad_norm": 13.900424003601074, "learning_rate": 5.2e-06, "loss": 72.0, "step": 521 }, { "epoch": 5.22e-05, "grad_norm": 14.005378723144531, "learning_rate": 5.21e-06, "loss": 72.0, "step": 522 }, { "epoch": 5.23e-05, "grad_norm": 13.34553337097168, "learning_rate": 5.22e-06, "loss": 71.9375, "step": 523 }, { "epoch": 5.24e-05, "grad_norm": 13.30509090423584, "learning_rate": 5.23e-06, "loss": 71.9375, "step": 524 }, { "epoch": 5.25e-05, "grad_norm": 13.428346633911133, "learning_rate": 5.240000000000001e-06, "loss": 72.0, "step": 525 }, { "epoch": 5.26e-05, "grad_norm": 13.481391906738281, "learning_rate": 5.2500000000000006e-06, "loss": 72.0, "step": 526 }, { "epoch": 5.27e-05, "grad_norm": 13.900728225708008, "learning_rate": 5.26e-06, "loss": 72.0, "step": 527 }, { "epoch": 5.28e-05, "grad_norm": 12.912371635437012, "learning_rate": 5.2699999999999995e-06, "loss": 72.0, "step": 528 }, { "epoch": 5.29e-05, "grad_norm": 13.814693450927734, "learning_rate": 5.28e-06, "loss": 72.0, "step": 529 }, { "epoch": 5.3e-05, "grad_norm": 13.151686668395996, "learning_rate": 5.29e-06, "loss": 72.0, "step": 530 }, { "epoch": 5.31e-05, "grad_norm": 13.657590866088867, "learning_rate": 5.3e-06, "loss": 71.9375, "step": 531 }, { "epoch": 5.32e-05, "grad_norm": 13.253593444824219, "learning_rate": 5.31e-06, "loss": 71.9375, "step": 532 }, { "epoch": 5.33e-05, "grad_norm": 13.740532875061035, "learning_rate": 5.32e-06, "loss": 71.9375, "step": 533 }, { "epoch": 5.34e-05, "grad_norm": 14.07413387298584, "learning_rate": 5.33e-06, "loss": 71.875, "step": 534 }, { "epoch": 5.35e-05, "grad_norm": 13.681927680969238, "learning_rate": 5.34e-06, "loss": 72.0, "step": 535 }, { "epoch": 5.36e-05, "grad_norm": 13.324553489685059, "learning_rate": 5.3500000000000004e-06, "loss": 71.9375, "step": 536 }, { "epoch": 5.37e-05, "grad_norm": 13.417908668518066, "learning_rate": 5.36e-06, "loss": 71.9375, "step": 537 }, { "epoch": 5.38e-05, "grad_norm": 12.827329635620117, "learning_rate": 5.37e-06, "loss": 71.9375, "step": 538 }, { "epoch": 5.39e-05, "grad_norm": 13.676140785217285, "learning_rate": 5.38e-06, "loss": 71.9375, "step": 539 }, { "epoch": 5.4e-05, "grad_norm": 13.923197746276855, "learning_rate": 5.39e-06, "loss": 71.875, "step": 540 }, { "epoch": 5.41e-05, "grad_norm": 13.1552152633667, "learning_rate": 5.4e-06, "loss": 71.875, "step": 541 }, { "epoch": 5.42e-05, "grad_norm": 13.265161514282227, "learning_rate": 5.410000000000001e-06, "loss": 71.9375, "step": 542 }, { "epoch": 5.43e-05, "grad_norm": 13.648545265197754, "learning_rate": 5.420000000000001e-06, "loss": 71.875, "step": 543 }, { "epoch": 5.44e-05, "grad_norm": 14.054911613464355, "learning_rate": 5.43e-06, "loss": 71.75, "step": 544 }, { "epoch": 5.45e-05, "grad_norm": 13.410968780517578, "learning_rate": 5.44e-06, "loss": 71.8125, "step": 545 }, { "epoch": 5.46e-05, "grad_norm": 14.150003433227539, "learning_rate": 5.45e-06, "loss": 71.8125, "step": 546 }, { "epoch": 5.47e-05, "grad_norm": 13.140257835388184, "learning_rate": 5.46e-06, "loss": 71.75, "step": 547 }, { "epoch": 5.48e-05, "grad_norm": 13.407617568969727, "learning_rate": 5.47e-06, "loss": 71.875, "step": 548 }, { "epoch": 5.49e-05, "grad_norm": 13.361193656921387, "learning_rate": 5.48e-06, "loss": 71.8125, "step": 549 }, { "epoch": 5.5e-05, "grad_norm": 13.742301940917969, "learning_rate": 5.49e-06, "loss": 71.75, "step": 550 }, { "epoch": 5.51e-05, "grad_norm": 13.647250175476074, "learning_rate": 5.5e-06, "loss": 71.8125, "step": 551 }, { "epoch": 5.52e-05, "grad_norm": 13.841861724853516, "learning_rate": 5.51e-06, "loss": 71.625, "step": 552 }, { "epoch": 5.53e-05, "grad_norm": 13.940226554870605, "learning_rate": 5.5200000000000005e-06, "loss": 71.5625, "step": 553 }, { "epoch": 5.54e-05, "grad_norm": 14.098892211914062, "learning_rate": 5.53e-06, "loss": 71.875, "step": 554 }, { "epoch": 5.55e-05, "grad_norm": 13.767340660095215, "learning_rate": 5.54e-06, "loss": 71.8125, "step": 555 }, { "epoch": 5.56e-05, "grad_norm": 13.485721588134766, "learning_rate": 5.55e-06, "loss": 71.5625, "step": 556 }, { "epoch": 5.57e-05, "grad_norm": 13.09888744354248, "learning_rate": 5.56e-06, "loss": 71.75, "step": 557 }, { "epoch": 5.58e-05, "grad_norm": 13.47565746307373, "learning_rate": 5.57e-06, "loss": 71.75, "step": 558 }, { "epoch": 5.59e-05, "grad_norm": 13.305706977844238, "learning_rate": 5.580000000000001e-06, "loss": 71.5625, "step": 559 }, { "epoch": 5.6e-05, "grad_norm": 13.260443687438965, "learning_rate": 5.59e-06, "loss": 71.5625, "step": 560 }, { "epoch": 5.61e-05, "grad_norm": 13.876309394836426, "learning_rate": 5.6e-06, "loss": 71.625, "step": 561 }, { "epoch": 5.62e-05, "grad_norm": 13.163431167602539, "learning_rate": 5.61e-06, "loss": 71.5625, "step": 562 }, { "epoch": 5.63e-05, "grad_norm": 13.940244674682617, "learning_rate": 5.62e-06, "loss": 71.5625, "step": 563 }, { "epoch": 5.64e-05, "grad_norm": 14.002641677856445, "learning_rate": 5.63e-06, "loss": 71.625, "step": 564 }, { "epoch": 5.65e-05, "grad_norm": 13.585519790649414, "learning_rate": 5.64e-06, "loss": 71.5, "step": 565 }, { "epoch": 5.66e-05, "grad_norm": 13.694162368774414, "learning_rate": 5.65e-06, "loss": 71.625, "step": 566 }, { "epoch": 5.67e-05, "grad_norm": 13.502567291259766, "learning_rate": 5.66e-06, "loss": 71.625, "step": 567 }, { "epoch": 5.68e-05, "grad_norm": 13.50383186340332, "learning_rate": 5.67e-06, "loss": 71.625, "step": 568 }, { "epoch": 5.69e-05, "grad_norm": 13.493083953857422, "learning_rate": 5.68e-06, "loss": 71.5625, "step": 569 }, { "epoch": 5.7e-05, "grad_norm": 13.770458221435547, "learning_rate": 5.690000000000001e-06, "loss": 71.4375, "step": 570 }, { "epoch": 5.71e-05, "grad_norm": 13.740925788879395, "learning_rate": 5.7000000000000005e-06, "loss": 71.5625, "step": 571 }, { "epoch": 5.72e-05, "grad_norm": 13.495709419250488, "learning_rate": 5.7099999999999995e-06, "loss": 71.5625, "step": 572 }, { "epoch": 5.73e-05, "grad_norm": 14.037687301635742, "learning_rate": 5.72e-06, "loss": 71.5, "step": 573 }, { "epoch": 5.74e-05, "grad_norm": 13.584688186645508, "learning_rate": 5.73e-06, "loss": 71.5625, "step": 574 }, { "epoch": 5.75e-05, "grad_norm": 13.670408248901367, "learning_rate": 5.74e-06, "loss": 71.5, "step": 575 }, { "epoch": 5.76e-05, "grad_norm": 13.831013679504395, "learning_rate": 5.75e-06, "loss": 71.5, "step": 576 }, { "epoch": 5.77e-05, "grad_norm": 13.588455200195312, "learning_rate": 5.76e-06, "loss": 71.5, "step": 577 }, { "epoch": 5.78e-05, "grad_norm": 14.016668319702148, "learning_rate": 5.77e-06, "loss": 71.5, "step": 578 }, { "epoch": 5.79e-05, "grad_norm": 13.188124656677246, "learning_rate": 5.78e-06, "loss": 71.5625, "step": 579 }, { "epoch": 5.8e-05, "grad_norm": 14.161450386047363, "learning_rate": 5.7900000000000005e-06, "loss": 71.5, "step": 580 }, { "epoch": 5.81e-05, "grad_norm": 13.826251983642578, "learning_rate": 5.8e-06, "loss": 71.4375, "step": 581 }, { "epoch": 5.82e-05, "grad_norm": 13.522049903869629, "learning_rate": 5.81e-06, "loss": 71.5, "step": 582 }, { "epoch": 5.83e-05, "grad_norm": 13.675570487976074, "learning_rate": 5.82e-06, "loss": 71.5625, "step": 583 }, { "epoch": 5.84e-05, "grad_norm": 13.925761222839355, "learning_rate": 5.83e-06, "loss": 71.5, "step": 584 }, { "epoch": 5.85e-05, "grad_norm": 13.596080780029297, "learning_rate": 5.84e-06, "loss": 71.5, "step": 585 }, { "epoch": 5.86e-05, "grad_norm": 13.48724365234375, "learning_rate": 5.85e-06, "loss": 71.5, "step": 586 }, { "epoch": 5.87e-05, "grad_norm": 13.393336296081543, "learning_rate": 5.860000000000001e-06, "loss": 71.3125, "step": 587 }, { "epoch": 5.88e-05, "grad_norm": 13.0198392868042, "learning_rate": 5.87e-06, "loss": 71.5, "step": 588 }, { "epoch": 5.89e-05, "grad_norm": 14.366602897644043, "learning_rate": 5.88e-06, "loss": 71.3125, "step": 589 }, { "epoch": 5.9e-05, "grad_norm": 13.80266284942627, "learning_rate": 5.89e-06, "loss": 71.4375, "step": 590 }, { "epoch": 5.91e-05, "grad_norm": 14.043784141540527, "learning_rate": 5.899999999999999e-06, "loss": 71.4375, "step": 591 }, { "epoch": 5.92e-05, "grad_norm": 14.366748809814453, "learning_rate": 5.91e-06, "loss": 71.5, "step": 592 }, { "epoch": 5.93e-05, "grad_norm": 13.664860725402832, "learning_rate": 5.92e-06, "loss": 71.4375, "step": 593 }, { "epoch": 5.94e-05, "grad_norm": 12.838698387145996, "learning_rate": 5.93e-06, "loss": 71.4375, "step": 594 }, { "epoch": 5.95e-05, "grad_norm": 13.87798023223877, "learning_rate": 5.94e-06, "loss": 71.3125, "step": 595 }, { "epoch": 5.96e-05, "grad_norm": 13.455665588378906, "learning_rate": 5.950000000000001e-06, "loss": 71.3125, "step": 596 }, { "epoch": 5.97e-05, "grad_norm": 13.53802490234375, "learning_rate": 5.96e-06, "loss": 71.4375, "step": 597 }, { "epoch": 5.98e-05, "grad_norm": 14.010590553283691, "learning_rate": 5.9700000000000004e-06, "loss": 71.5, "step": 598 }, { "epoch": 5.99e-05, "grad_norm": 13.441306114196777, "learning_rate": 5.98e-06, "loss": 71.375, "step": 599 }, { "epoch": 6e-05, "grad_norm": 13.627511024475098, "learning_rate": 5.989999999999999e-06, "loss": 71.4375, "step": 600 }, { "epoch": 6.01e-05, "grad_norm": 13.531881332397461, "learning_rate": 6e-06, "loss": 71.4375, "step": 601 }, { "epoch": 6.02e-05, "grad_norm": 13.889076232910156, "learning_rate": 6.010000000000001e-06, "loss": 71.3125, "step": 602 }, { "epoch": 6.03e-05, "grad_norm": 13.621227264404297, "learning_rate": 6.02e-06, "loss": 71.25, "step": 603 }, { "epoch": 6.04e-05, "grad_norm": 13.696484565734863, "learning_rate": 6.030000000000001e-06, "loss": 71.375, "step": 604 }, { "epoch": 6.05e-05, "grad_norm": 13.138842582702637, "learning_rate": 6.04e-06, "loss": 71.3125, "step": 605 }, { "epoch": 6.06e-05, "grad_norm": 14.300238609313965, "learning_rate": 6.05e-06, "loss": 71.25, "step": 606 }, { "epoch": 6.07e-05, "grad_norm": 13.133456230163574, "learning_rate": 6.0600000000000004e-06, "loss": 71.4375, "step": 607 }, { "epoch": 6.08e-05, "grad_norm": 13.548200607299805, "learning_rate": 6.0699999999999995e-06, "loss": 71.375, "step": 608 }, { "epoch": 6.09e-05, "grad_norm": 13.798404693603516, "learning_rate": 6.08e-06, "loss": 71.25, "step": 609 }, { "epoch": 6.1e-05, "grad_norm": 13.428898811340332, "learning_rate": 6.09e-06, "loss": 71.375, "step": 610 }, { "epoch": 6.11e-05, "grad_norm": 13.93244457244873, "learning_rate": 6.1e-06, "loss": 71.25, "step": 611 }, { "epoch": 6.12e-05, "grad_norm": 14.050407409667969, "learning_rate": 6.11e-06, "loss": 71.375, "step": 612 }, { "epoch": 6.13e-05, "grad_norm": 13.533824920654297, "learning_rate": 6.120000000000001e-06, "loss": 71.3125, "step": 613 }, { "epoch": 6.14e-05, "grad_norm": 13.453178405761719, "learning_rate": 6.13e-06, "loss": 71.25, "step": 614 }, { "epoch": 6.15e-05, "grad_norm": 13.044660568237305, "learning_rate": 6.1400000000000005e-06, "loss": 71.4375, "step": 615 }, { "epoch": 6.16e-05, "grad_norm": 13.782415390014648, "learning_rate": 6.15e-06, "loss": 71.25, "step": 616 }, { "epoch": 6.17e-05, "grad_norm": 14.078405380249023, "learning_rate": 6.1599999999999995e-06, "loss": 71.25, "step": 617 }, { "epoch": 6.18e-05, "grad_norm": 13.621548652648926, "learning_rate": 6.17e-06, "loss": 71.3125, "step": 618 }, { "epoch": 6.19e-05, "grad_norm": 13.817481994628906, "learning_rate": 6.180000000000001e-06, "loss": 71.3125, "step": 619 }, { "epoch": 6.2e-05, "grad_norm": 13.902446746826172, "learning_rate": 6.19e-06, "loss": 71.125, "step": 620 }, { "epoch": 6.21e-05, "grad_norm": 14.227055549621582, "learning_rate": 6.2e-06, "loss": 71.0625, "step": 621 }, { "epoch": 6.22e-05, "grad_norm": 14.249224662780762, "learning_rate": 6.21e-06, "loss": 71.0, "step": 622 }, { "epoch": 6.23e-05, "grad_norm": 13.914899826049805, "learning_rate": 6.22e-06, "loss": 71.1875, "step": 623 }, { "epoch": 6.24e-05, "grad_norm": 13.97021770477295, "learning_rate": 6.2300000000000005e-06, "loss": 71.0, "step": 624 }, { "epoch": 6.25e-05, "grad_norm": 13.667683601379395, "learning_rate": 6.2399999999999995e-06, "loss": 71.0625, "step": 625 }, { "epoch": 6.26e-05, "grad_norm": 13.566893577575684, "learning_rate": 6.25e-06, "loss": 71.1875, "step": 626 }, { "epoch": 6.27e-05, "grad_norm": 14.130186080932617, "learning_rate": 6.26e-06, "loss": 71.0625, "step": 627 }, { "epoch": 6.28e-05, "grad_norm": 13.867085456848145, "learning_rate": 6.269999999999999e-06, "loss": 71.1875, "step": 628 }, { "epoch": 6.29e-05, "grad_norm": 14.15766716003418, "learning_rate": 6.28e-06, "loss": 71.0625, "step": 629 }, { "epoch": 6.3e-05, "grad_norm": 13.474671363830566, "learning_rate": 6.290000000000001e-06, "loss": 71.1875, "step": 630 }, { "epoch": 6.31e-05, "grad_norm": 13.981532096862793, "learning_rate": 6.3e-06, "loss": 71.0, "step": 631 }, { "epoch": 6.32e-05, "grad_norm": 13.78842544555664, "learning_rate": 6.3100000000000006e-06, "loss": 71.0625, "step": 632 }, { "epoch": 6.33e-05, "grad_norm": 14.002915382385254, "learning_rate": 6.3200000000000005e-06, "loss": 71.1875, "step": 633 }, { "epoch": 6.34e-05, "grad_norm": 13.505497932434082, "learning_rate": 6.3299999999999995e-06, "loss": 71.125, "step": 634 }, { "epoch": 6.35e-05, "grad_norm": 13.425030708312988, "learning_rate": 6.34e-06, "loss": 71.0625, "step": 635 }, { "epoch": 6.36e-05, "grad_norm": 13.330291748046875, "learning_rate": 6.350000000000001e-06, "loss": 71.1875, "step": 636 }, { "epoch": 6.37e-05, "grad_norm": 14.178326606750488, "learning_rate": 6.36e-06, "loss": 71.0625, "step": 637 }, { "epoch": 6.38e-05, "grad_norm": 13.922074317932129, "learning_rate": 6.37e-06, "loss": 71.0, "step": 638 }, { "epoch": 6.39e-05, "grad_norm": 13.808489799499512, "learning_rate": 6.38e-06, "loss": 71.0625, "step": 639 }, { "epoch": 6.4e-05, "grad_norm": 13.859967231750488, "learning_rate": 6.39e-06, "loss": 71.0625, "step": 640 }, { "epoch": 6.41e-05, "grad_norm": 14.012020111083984, "learning_rate": 6.4000000000000006e-06, "loss": 71.0, "step": 641 }, { "epoch": 6.42e-05, "grad_norm": 13.656329154968262, "learning_rate": 6.41e-06, "loss": 71.0, "step": 642 }, { "epoch": 6.43e-05, "grad_norm": 13.934961318969727, "learning_rate": 6.42e-06, "loss": 71.0625, "step": 643 }, { "epoch": 6.44e-05, "grad_norm": 13.582793235778809, "learning_rate": 6.43e-06, "loss": 71.0, "step": 644 }, { "epoch": 6.45e-05, "grad_norm": 13.44519329071045, "learning_rate": 6.439999999999999e-06, "loss": 71.0625, "step": 645 }, { "epoch": 6.46e-05, "grad_norm": 14.572875022888184, "learning_rate": 6.45e-06, "loss": 70.9375, "step": 646 }, { "epoch": 6.47e-05, "grad_norm": 13.89630126953125, "learning_rate": 6.460000000000001e-06, "loss": 70.9375, "step": 647 }, { "epoch": 6.48e-05, "grad_norm": 13.762505531311035, "learning_rate": 6.47e-06, "loss": 70.875, "step": 648 }, { "epoch": 6.49e-05, "grad_norm": 13.604170799255371, "learning_rate": 6.480000000000001e-06, "loss": 70.9375, "step": 649 }, { "epoch": 6.5e-05, "grad_norm": 13.724708557128906, "learning_rate": 6.4900000000000005e-06, "loss": 70.9375, "step": 650 }, { "epoch": 6.51e-05, "grad_norm": 13.948774337768555, "learning_rate": 6.5e-06, "loss": 70.9375, "step": 651 }, { "epoch": 6.52e-05, "grad_norm": 14.045595169067383, "learning_rate": 6.51e-06, "loss": 70.9375, "step": 652 }, { "epoch": 6.53e-05, "grad_norm": 13.71200180053711, "learning_rate": 6.520000000000001e-06, "loss": 70.9375, "step": 653 }, { "epoch": 6.54e-05, "grad_norm": 14.408919334411621, "learning_rate": 6.53e-06, "loss": 70.8125, "step": 654 }, { "epoch": 6.55e-05, "grad_norm": 13.958176612854004, "learning_rate": 6.54e-06, "loss": 70.9375, "step": 655 }, { "epoch": 6.56e-05, "grad_norm": 13.734806060791016, "learning_rate": 6.549999999999999e-06, "loss": 71.0625, "step": 656 }, { "epoch": 6.57e-05, "grad_norm": 14.043981552124023, "learning_rate": 6.56e-06, "loss": 70.8125, "step": 657 }, { "epoch": 6.58e-05, "grad_norm": 13.461383819580078, "learning_rate": 6.570000000000001e-06, "loss": 71.0, "step": 658 }, { "epoch": 6.59e-05, "grad_norm": 13.893697738647461, "learning_rate": 6.58e-06, "loss": 70.9375, "step": 659 }, { "epoch": 6.6e-05, "grad_norm": 14.152552604675293, "learning_rate": 6.5900000000000004e-06, "loss": 70.8125, "step": 660 }, { "epoch": 6.61e-05, "grad_norm": 14.377528190612793, "learning_rate": 6.6e-06, "loss": 70.75, "step": 661 }, { "epoch": 6.62e-05, "grad_norm": 13.767778396606445, "learning_rate": 6.609999999999999e-06, "loss": 70.8125, "step": 662 }, { "epoch": 6.63e-05, "grad_norm": 13.831917762756348, "learning_rate": 6.62e-06, "loss": 70.8125, "step": 663 }, { "epoch": 6.64e-05, "grad_norm": 14.295366287231445, "learning_rate": 6.630000000000001e-06, "loss": 70.8125, "step": 664 }, { "epoch": 6.65e-05, "grad_norm": 13.494677543640137, "learning_rate": 6.64e-06, "loss": 71.0, "step": 665 }, { "epoch": 6.66e-05, "grad_norm": 14.52921199798584, "learning_rate": 6.65e-06, "loss": 70.6875, "step": 666 }, { "epoch": 6.67e-05, "grad_norm": 14.000243186950684, "learning_rate": 6.660000000000001e-06, "loss": 70.625, "step": 667 }, { "epoch": 6.68e-05, "grad_norm": 13.905571937561035, "learning_rate": 6.67e-06, "loss": 70.8125, "step": 668 }, { "epoch": 6.69e-05, "grad_norm": 13.985931396484375, "learning_rate": 6.68e-06, "loss": 70.8125, "step": 669 }, { "epoch": 6.7e-05, "grad_norm": 13.165407180786133, "learning_rate": 6.690000000000001e-06, "loss": 70.875, "step": 670 }, { "epoch": 6.71e-05, "grad_norm": 14.570108413696289, "learning_rate": 6.7e-06, "loss": 70.5625, "step": 671 }, { "epoch": 6.72e-05, "grad_norm": 14.332723617553711, "learning_rate": 6.71e-06, "loss": 70.625, "step": 672 }, { "epoch": 6.73e-05, "grad_norm": 13.870111465454102, "learning_rate": 6.719999999999999e-06, "loss": 70.5625, "step": 673 }, { "epoch": 6.74e-05, "grad_norm": 14.580185890197754, "learning_rate": 6.73e-06, "loss": 70.75, "step": 674 }, { "epoch": 6.75e-05, "grad_norm": 13.757917404174805, "learning_rate": 6.740000000000001e-06, "loss": 70.5625, "step": 675 }, { "epoch": 6.76e-05, "grad_norm": 14.098215103149414, "learning_rate": 6.75e-06, "loss": 70.625, "step": 676 }, { "epoch": 6.77e-05, "grad_norm": 13.889370918273926, "learning_rate": 6.7600000000000005e-06, "loss": 70.625, "step": 677 }, { "epoch": 6.78e-05, "grad_norm": 13.543923377990723, "learning_rate": 6.77e-06, "loss": 70.75, "step": 678 }, { "epoch": 6.79e-05, "grad_norm": 14.494497299194336, "learning_rate": 6.7799999999999995e-06, "loss": 70.4375, "step": 679 }, { "epoch": 6.8e-05, "grad_norm": 13.892583847045898, "learning_rate": 6.79e-06, "loss": 70.6875, "step": 680 }, { "epoch": 6.81e-05, "grad_norm": 14.069755554199219, "learning_rate": 6.800000000000001e-06, "loss": 70.5, "step": 681 }, { "epoch": 6.82e-05, "grad_norm": 14.10586166381836, "learning_rate": 6.81e-06, "loss": 70.4375, "step": 682 }, { "epoch": 6.83e-05, "grad_norm": 14.291821479797363, "learning_rate": 6.82e-06, "loss": 70.5625, "step": 683 }, { "epoch": 6.84e-05, "grad_norm": 14.351411819458008, "learning_rate": 6.830000000000001e-06, "loss": 70.5625, "step": 684 }, { "epoch": 6.85e-05, "grad_norm": 13.590216636657715, "learning_rate": 6.84e-06, "loss": 70.4375, "step": 685 }, { "epoch": 6.86e-05, "grad_norm": 14.20474910736084, "learning_rate": 6.8500000000000005e-06, "loss": 70.5, "step": 686 }, { "epoch": 6.87e-05, "grad_norm": 14.491683959960938, "learning_rate": 6.86e-06, "loss": 70.25, "step": 687 }, { "epoch": 6.88e-05, "grad_norm": 14.437154769897461, "learning_rate": 6.87e-06, "loss": 70.3125, "step": 688 }, { "epoch": 6.89e-05, "grad_norm": 14.214030265808105, "learning_rate": 6.88e-06, "loss": 70.375, "step": 689 }, { "epoch": 6.9e-05, "grad_norm": 13.905532836914062, "learning_rate": 6.889999999999999e-06, "loss": 70.4375, "step": 690 }, { "epoch": 6.91e-05, "grad_norm": 14.396810531616211, "learning_rate": 6.9e-06, "loss": 70.3125, "step": 691 }, { "epoch": 6.92e-05, "grad_norm": 13.368425369262695, "learning_rate": 6.910000000000001e-06, "loss": 70.625, "step": 692 }, { "epoch": 6.93e-05, "grad_norm": 14.467592239379883, "learning_rate": 6.92e-06, "loss": 70.125, "step": 693 }, { "epoch": 6.94e-05, "grad_norm": 14.07659912109375, "learning_rate": 6.93e-06, "loss": 70.25, "step": 694 }, { "epoch": 6.95e-05, "grad_norm": 13.678894996643066, "learning_rate": 6.9400000000000005e-06, "loss": 70.25, "step": 695 }, { "epoch": 6.96e-05, "grad_norm": 13.846305847167969, "learning_rate": 6.9499999999999995e-06, "loss": 70.4375, "step": 696 }, { "epoch": 6.97e-05, "grad_norm": 14.344511032104492, "learning_rate": 6.96e-06, "loss": 70.125, "step": 697 }, { "epoch": 6.98e-05, "grad_norm": 13.975683212280273, "learning_rate": 6.970000000000001e-06, "loss": 70.3125, "step": 698 }, { "epoch": 6.99e-05, "grad_norm": 13.725882530212402, "learning_rate": 6.98e-06, "loss": 70.25, "step": 699 }, { "epoch": 7e-05, "grad_norm": 14.02601146697998, "learning_rate": 6.99e-06, "loss": 70.1875, "step": 700 }, { "epoch": 7.01e-05, "grad_norm": 14.149601936340332, "learning_rate": 7.000000000000001e-06, "loss": 70.25, "step": 701 }, { "epoch": 7.02e-05, "grad_norm": 14.683643341064453, "learning_rate": 7.01e-06, "loss": 69.9375, "step": 702 }, { "epoch": 7.03e-05, "grad_norm": 14.377092361450195, "learning_rate": 7.0200000000000006e-06, "loss": 70.0625, "step": 703 }, { "epoch": 7.04e-05, "grad_norm": 13.508744239807129, "learning_rate": 7.03e-06, "loss": 70.125, "step": 704 }, { "epoch": 7.05e-05, "grad_norm": 14.27863883972168, "learning_rate": 7.04e-06, "loss": 70.0, "step": 705 }, { "epoch": 7.06e-05, "grad_norm": 14.04149055480957, "learning_rate": 7.05e-06, "loss": 70.0625, "step": 706 }, { "epoch": 7.07e-05, "grad_norm": 13.799717903137207, "learning_rate": 7.059999999999999e-06, "loss": 70.125, "step": 707 }, { "epoch": 7.08e-05, "grad_norm": 13.9593505859375, "learning_rate": 7.07e-06, "loss": 70.0, "step": 708 }, { "epoch": 7.09e-05, "grad_norm": 14.294474601745605, "learning_rate": 7.080000000000001e-06, "loss": 69.875, "step": 709 }, { "epoch": 7.1e-05, "grad_norm": 14.307499885559082, "learning_rate": 7.09e-06, "loss": 70.0, "step": 710 }, { "epoch": 7.11e-05, "grad_norm": 13.888057708740234, "learning_rate": 7.1e-06, "loss": 70.0, "step": 711 }, { "epoch": 7.12e-05, "grad_norm": 13.905348777770996, "learning_rate": 7.1100000000000005e-06, "loss": 70.0, "step": 712 }, { "epoch": 7.13e-05, "grad_norm": 14.086119651794434, "learning_rate": 7.12e-06, "loss": 69.875, "step": 713 }, { "epoch": 7.14e-05, "grad_norm": 14.221412658691406, "learning_rate": 7.13e-06, "loss": 69.9375, "step": 714 }, { "epoch": 7.15e-05, "grad_norm": 14.204227447509766, "learning_rate": 7.14e-06, "loss": 69.875, "step": 715 }, { "epoch": 7.16e-05, "grad_norm": 14.413752555847168, "learning_rate": 7.15e-06, "loss": 69.75, "step": 716 }, { "epoch": 7.17e-05, "grad_norm": 14.447648048400879, "learning_rate": 7.16e-06, "loss": 69.625, "step": 717 }, { "epoch": 7.18e-05, "grad_norm": 13.275869369506836, "learning_rate": 7.170000000000001e-06, "loss": 69.9375, "step": 718 }, { "epoch": 7.19e-05, "grad_norm": 14.1619873046875, "learning_rate": 7.18e-06, "loss": 69.6875, "step": 719 }, { "epoch": 7.2e-05, "grad_norm": 14.165518760681152, "learning_rate": 7.190000000000001e-06, "loss": 69.6875, "step": 720 }, { "epoch": 7.21e-05, "grad_norm": 13.807438850402832, "learning_rate": 7.2e-06, "loss": 69.75, "step": 721 }, { "epoch": 7.22e-05, "grad_norm": 14.538020133972168, "learning_rate": 7.21e-06, "loss": 69.6875, "step": 722 }, { "epoch": 7.23e-05, "grad_norm": 14.57617473602295, "learning_rate": 7.22e-06, "loss": 69.5625, "step": 723 }, { "epoch": 7.24e-05, "grad_norm": 13.881351470947266, "learning_rate": 7.229999999999999e-06, "loss": 69.625, "step": 724 }, { "epoch": 7.25e-05, "grad_norm": 14.827073097229004, "learning_rate": 7.24e-06, "loss": 69.5625, "step": 725 }, { "epoch": 7.26e-05, "grad_norm": 14.633003234863281, "learning_rate": 7.250000000000001e-06, "loss": 69.375, "step": 726 }, { "epoch": 7.27e-05, "grad_norm": 13.844290733337402, "learning_rate": 7.26e-06, "loss": 69.4375, "step": 727 }, { "epoch": 7.28e-05, "grad_norm": 14.124848365783691, "learning_rate": 7.27e-06, "loss": 69.4375, "step": 728 }, { "epoch": 7.29e-05, "grad_norm": 14.088972091674805, "learning_rate": 7.280000000000001e-06, "loss": 69.5, "step": 729 }, { "epoch": 7.3e-05, "grad_norm": 14.448264122009277, "learning_rate": 7.29e-06, "loss": 69.4375, "step": 730 }, { "epoch": 7.31e-05, "grad_norm": 14.05547046661377, "learning_rate": 7.3e-06, "loss": 69.375, "step": 731 }, { "epoch": 7.32e-05, "grad_norm": 14.225979804992676, "learning_rate": 7.31e-06, "loss": 69.375, "step": 732 }, { "epoch": 7.33e-05, "grad_norm": 14.301802635192871, "learning_rate": 7.32e-06, "loss": 69.25, "step": 733 }, { "epoch": 7.34e-05, "grad_norm": 14.079911231994629, "learning_rate": 7.33e-06, "loss": 69.375, "step": 734 }, { "epoch": 7.35e-05, "grad_norm": 13.954413414001465, "learning_rate": 7.340000000000001e-06, "loss": 69.375, "step": 735 }, { "epoch": 7.36e-05, "grad_norm": 14.633395195007324, "learning_rate": 7.35e-06, "loss": 69.1875, "step": 736 }, { "epoch": 7.37e-05, "grad_norm": 13.73357105255127, "learning_rate": 7.360000000000001e-06, "loss": 69.3125, "step": 737 }, { "epoch": 7.38e-05, "grad_norm": 13.533805847167969, "learning_rate": 7.37e-06, "loss": 69.375, "step": 738 }, { "epoch": 7.39e-05, "grad_norm": 13.967336654663086, "learning_rate": 7.38e-06, "loss": 69.125, "step": 739 }, { "epoch": 7.4e-05, "grad_norm": 13.907071113586426, "learning_rate": 7.39e-06, "loss": 69.25, "step": 740 }, { "epoch": 7.41e-05, "grad_norm": 14.33962345123291, "learning_rate": 7.3999999999999995e-06, "loss": 69.0625, "step": 741 }, { "epoch": 7.42e-05, "grad_norm": 14.036276817321777, "learning_rate": 7.41e-06, "loss": 69.0625, "step": 742 }, { "epoch": 7.43e-05, "grad_norm": 14.373995780944824, "learning_rate": 7.420000000000001e-06, "loss": 69.0, "step": 743 }, { "epoch": 7.44e-05, "grad_norm": 13.279586791992188, "learning_rate": 7.43e-06, "loss": 69.25, "step": 744 }, { "epoch": 7.45e-05, "grad_norm": 13.835576057434082, "learning_rate": 7.44e-06, "loss": 69.0625, "step": 745 }, { "epoch": 7.46e-05, "grad_norm": 14.343767166137695, "learning_rate": 7.450000000000001e-06, "loss": 69.0, "step": 746 }, { "epoch": 7.47e-05, "grad_norm": 14.20533275604248, "learning_rate": 7.46e-06, "loss": 68.875, "step": 747 }, { "epoch": 7.48e-05, "grad_norm": 14.370162010192871, "learning_rate": 7.4700000000000005e-06, "loss": 68.75, "step": 748 }, { "epoch": 7.49e-05, "grad_norm": 14.102258682250977, "learning_rate": 7.48e-06, "loss": 68.8125, "step": 749 }, { "epoch": 7.5e-05, "grad_norm": 14.238746643066406, "learning_rate": 7.4899999999999994e-06, "loss": 68.75, "step": 750 }, { "epoch": 7.51e-05, "grad_norm": 14.459688186645508, "learning_rate": 7.5e-06, "loss": 68.75, "step": 751 }, { "epoch": 7.52e-05, "grad_norm": 13.407461166381836, "learning_rate": 7.510000000000001e-06, "loss": 68.9375, "step": 752 }, { "epoch": 7.53e-05, "grad_norm": 13.310074806213379, "learning_rate": 7.52e-06, "loss": 68.9375, "step": 753 }, { "epoch": 7.54e-05, "grad_norm": 14.59634017944336, "learning_rate": 7.530000000000001e-06, "loss": 68.5625, "step": 754 }, { "epoch": 7.55e-05, "grad_norm": 13.75051212310791, "learning_rate": 7.54e-06, "loss": 68.6875, "step": 755 }, { "epoch": 7.56e-05, "grad_norm": 14.305959701538086, "learning_rate": 7.55e-06, "loss": 68.5625, "step": 756 }, { "epoch": 7.57e-05, "grad_norm": 14.588299751281738, "learning_rate": 7.5600000000000005e-06, "loss": 68.375, "step": 757 }, { "epoch": 7.58e-05, "grad_norm": 14.081939697265625, "learning_rate": 7.5699999999999995e-06, "loss": 68.5, "step": 758 }, { "epoch": 7.59e-05, "grad_norm": 13.225693702697754, "learning_rate": 7.58e-06, "loss": 68.9375, "step": 759 }, { "epoch": 7.6e-05, "grad_norm": 13.850139617919922, "learning_rate": 7.59e-06, "loss": 68.625, "step": 760 }, { "epoch": 7.61e-05, "grad_norm": 13.827278137207031, "learning_rate": 7.6e-06, "loss": 68.5, "step": 761 }, { "epoch": 7.62e-05, "grad_norm": 14.146428108215332, "learning_rate": 7.61e-06, "loss": 68.3125, "step": 762 }, { "epoch": 7.63e-05, "grad_norm": 13.658434867858887, "learning_rate": 7.620000000000001e-06, "loss": 68.4375, "step": 763 }, { "epoch": 7.64e-05, "grad_norm": 14.356409072875977, "learning_rate": 7.63e-06, "loss": 68.3125, "step": 764 }, { "epoch": 7.65e-05, "grad_norm": 14.146637916564941, "learning_rate": 7.64e-06, "loss": 68.3125, "step": 765 }, { "epoch": 7.66e-05, "grad_norm": 13.119014739990234, "learning_rate": 7.650000000000001e-06, "loss": 68.5625, "step": 766 }, { "epoch": 7.67e-05, "grad_norm": 14.340291976928711, "learning_rate": 7.66e-06, "loss": 68.1875, "step": 767 }, { "epoch": 7.68e-05, "grad_norm": 13.567488670349121, "learning_rate": 7.670000000000001e-06, "loss": 68.375, "step": 768 }, { "epoch": 7.69e-05, "grad_norm": 14.072280883789062, "learning_rate": 7.680000000000001e-06, "loss": 68.125, "step": 769 }, { "epoch": 7.7e-05, "grad_norm": 13.906455993652344, "learning_rate": 7.69e-06, "loss": 68.0625, "step": 770 }, { "epoch": 7.71e-05, "grad_norm": 14.400522232055664, "learning_rate": 7.7e-06, "loss": 67.9375, "step": 771 }, { "epoch": 7.72e-05, "grad_norm": 14.000092506408691, "learning_rate": 7.709999999999999e-06, "loss": 68.0, "step": 772 }, { "epoch": 7.73e-05, "grad_norm": 13.759057998657227, "learning_rate": 7.72e-06, "loss": 68.125, "step": 773 }, { "epoch": 7.74e-05, "grad_norm": 13.556885719299316, "learning_rate": 7.73e-06, "loss": 68.1875, "step": 774 }, { "epoch": 7.75e-05, "grad_norm": 13.875627517700195, "learning_rate": 7.74e-06, "loss": 68.0, "step": 775 }, { "epoch": 7.76e-05, "grad_norm": 14.190862655639648, "learning_rate": 7.75e-06, "loss": 67.75, "step": 776 }, { "epoch": 7.77e-05, "grad_norm": 14.008695602416992, "learning_rate": 7.76e-06, "loss": 67.8125, "step": 777 }, { "epoch": 7.78e-05, "grad_norm": 13.735383033752441, "learning_rate": 7.77e-06, "loss": 67.6875, "step": 778 }, { "epoch": 7.79e-05, "grad_norm": 13.905145645141602, "learning_rate": 7.78e-06, "loss": 67.875, "step": 779 }, { "epoch": 7.8e-05, "grad_norm": 13.315927505493164, "learning_rate": 7.79e-06, "loss": 68.0, "step": 780 }, { "epoch": 7.81e-05, "grad_norm": 14.015225410461426, "learning_rate": 7.8e-06, "loss": 67.75, "step": 781 }, { "epoch": 7.82e-05, "grad_norm": 13.459209442138672, "learning_rate": 7.81e-06, "loss": 67.875, "step": 782 }, { "epoch": 7.83e-05, "grad_norm": 13.911338806152344, "learning_rate": 7.820000000000001e-06, "loss": 67.75, "step": 783 }, { "epoch": 7.84e-05, "grad_norm": 13.783612251281738, "learning_rate": 7.83e-06, "loss": 67.8125, "step": 784 }, { "epoch": 7.85e-05, "grad_norm": 14.066487312316895, "learning_rate": 7.840000000000001e-06, "loss": 67.375, "step": 785 }, { "epoch": 7.86e-05, "grad_norm": 13.927901268005371, "learning_rate": 7.850000000000001e-06, "loss": 67.5, "step": 786 }, { "epoch": 7.87e-05, "grad_norm": 13.599596977233887, "learning_rate": 7.86e-06, "loss": 67.5, "step": 787 }, { "epoch": 7.88e-05, "grad_norm": 13.849498748779297, "learning_rate": 7.870000000000001e-06, "loss": 67.4375, "step": 788 }, { "epoch": 7.89e-05, "grad_norm": 13.334601402282715, "learning_rate": 7.879999999999999e-06, "loss": 67.5, "step": 789 }, { "epoch": 7.9e-05, "grad_norm": 13.693037986755371, "learning_rate": 7.89e-06, "loss": 67.375, "step": 790 }, { "epoch": 7.91e-05, "grad_norm": 13.442035675048828, "learning_rate": 7.9e-06, "loss": 67.5, "step": 791 }, { "epoch": 7.92e-05, "grad_norm": 13.878938674926758, "learning_rate": 7.91e-06, "loss": 67.0625, "step": 792 }, { "epoch": 7.93e-05, "grad_norm": 13.525701522827148, "learning_rate": 7.92e-06, "loss": 67.5, "step": 793 }, { "epoch": 7.94e-05, "grad_norm": 13.480079650878906, "learning_rate": 7.93e-06, "loss": 67.3125, "step": 794 }, { "epoch": 7.95e-05, "grad_norm": 13.224320411682129, "learning_rate": 7.94e-06, "loss": 67.375, "step": 795 }, { "epoch": 7.96e-05, "grad_norm": 13.158392906188965, "learning_rate": 7.95e-06, "loss": 67.3125, "step": 796 }, { "epoch": 7.97e-05, "grad_norm": 13.183201789855957, "learning_rate": 7.96e-06, "loss": 67.375, "step": 797 }, { "epoch": 7.98e-05, "grad_norm": 13.698829650878906, "learning_rate": 7.97e-06, "loss": 67.0, "step": 798 }, { "epoch": 7.99e-05, "grad_norm": 13.33995532989502, "learning_rate": 7.98e-06, "loss": 67.0, "step": 799 }, { "epoch": 8e-05, "grad_norm": 13.223627090454102, "learning_rate": 7.990000000000001e-06, "loss": 67.1875, "step": 800 }, { "epoch": 8.01e-05, "grad_norm": 13.501330375671387, "learning_rate": 8e-06, "loss": 66.9375, "step": 801 }, { "epoch": 8.02e-05, "grad_norm": 13.263901710510254, "learning_rate": 8.01e-06, "loss": 66.9375, "step": 802 }, { "epoch": 8.03e-05, "grad_norm": 13.32174301147461, "learning_rate": 8.02e-06, "loss": 66.875, "step": 803 }, { "epoch": 8.04e-05, "grad_norm": 13.487302780151367, "learning_rate": 8.03e-06, "loss": 66.75, "step": 804 }, { "epoch": 8.05e-05, "grad_norm": 13.301477432250977, "learning_rate": 8.040000000000001e-06, "loss": 66.875, "step": 805 }, { "epoch": 8.06e-05, "grad_norm": 13.382040023803711, "learning_rate": 8.05e-06, "loss": 66.625, "step": 806 }, { "epoch": 8.07e-05, "grad_norm": 13.217137336730957, "learning_rate": 8.06e-06, "loss": 66.875, "step": 807 }, { "epoch": 8.08e-05, "grad_norm": 13.310396194458008, "learning_rate": 8.07e-06, "loss": 66.6875, "step": 808 }, { "epoch": 8.09e-05, "grad_norm": 12.70387077331543, "learning_rate": 8.079999999999999e-06, "loss": 67.0, "step": 809 }, { "epoch": 8.1e-05, "grad_norm": 13.554513931274414, "learning_rate": 8.09e-06, "loss": 66.4375, "step": 810 }, { "epoch": 8.11e-05, "grad_norm": 13.117542266845703, "learning_rate": 8.1e-06, "loss": 66.5625, "step": 811 }, { "epoch": 8.12e-05, "grad_norm": 12.712786674499512, "learning_rate": 8.11e-06, "loss": 66.6875, "step": 812 }, { "epoch": 8.13e-05, "grad_norm": 13.098764419555664, "learning_rate": 8.12e-06, "loss": 66.5, "step": 813 }, { "epoch": 8.14e-05, "grad_norm": 13.022997856140137, "learning_rate": 8.13e-06, "loss": 66.4375, "step": 814 }, { "epoch": 8.15e-05, "grad_norm": 12.858806610107422, "learning_rate": 8.14e-06, "loss": 66.625, "step": 815 }, { "epoch": 8.16e-05, "grad_norm": 13.087048530578613, "learning_rate": 8.15e-06, "loss": 66.5, "step": 816 }, { "epoch": 8.17e-05, "grad_norm": 12.747725486755371, "learning_rate": 8.160000000000001e-06, "loss": 66.4375, "step": 817 }, { "epoch": 8.18e-05, "grad_norm": 13.237361907958984, "learning_rate": 8.17e-06, "loss": 66.1875, "step": 818 }, { "epoch": 8.19e-05, "grad_norm": 12.671786308288574, "learning_rate": 8.18e-06, "loss": 66.4375, "step": 819 }, { "epoch": 8.2e-05, "grad_norm": 12.85875129699707, "learning_rate": 8.19e-06, "loss": 66.3125, "step": 820 }, { "epoch": 8.21e-05, "grad_norm": 12.405821800231934, "learning_rate": 8.2e-06, "loss": 66.5, "step": 821 }, { "epoch": 8.22e-05, "grad_norm": 13.07856559753418, "learning_rate": 8.210000000000001e-06, "loss": 66.0625, "step": 822 }, { "epoch": 8.23e-05, "grad_norm": 13.018475532531738, "learning_rate": 8.22e-06, "loss": 66.1875, "step": 823 }, { "epoch": 8.24e-05, "grad_norm": 13.105154037475586, "learning_rate": 8.23e-06, "loss": 65.9375, "step": 824 }, { "epoch": 8.25e-05, "grad_norm": 12.811441421508789, "learning_rate": 8.24e-06, "loss": 66.0, "step": 825 }, { "epoch": 8.26e-05, "grad_norm": 12.24394702911377, "learning_rate": 8.249999999999999e-06, "loss": 66.5, "step": 826 }, { "epoch": 8.27e-05, "grad_norm": 12.856313705444336, "learning_rate": 8.26e-06, "loss": 66.1875, "step": 827 }, { "epoch": 8.28e-05, "grad_norm": 12.286417007446289, "learning_rate": 8.27e-06, "loss": 66.0625, "step": 828 }, { "epoch": 8.29e-05, "grad_norm": 12.537483215332031, "learning_rate": 8.28e-06, "loss": 65.9375, "step": 829 }, { "epoch": 8.3e-05, "grad_norm": 12.375907897949219, "learning_rate": 8.29e-06, "loss": 65.875, "step": 830 }, { "epoch": 8.31e-05, "grad_norm": 12.420113563537598, "learning_rate": 8.3e-06, "loss": 66.0, "step": 831 }, { "epoch": 8.32e-05, "grad_norm": 12.373382568359375, "learning_rate": 8.31e-06, "loss": 65.6875, "step": 832 }, { "epoch": 8.33e-05, "grad_norm": 11.9176607131958, "learning_rate": 8.32e-06, "loss": 65.875, "step": 833 }, { "epoch": 8.34e-05, "grad_norm": 12.355897903442383, "learning_rate": 8.330000000000002e-06, "loss": 65.9375, "step": 834 }, { "epoch": 8.35e-05, "grad_norm": 12.412797927856445, "learning_rate": 8.34e-06, "loss": 65.75, "step": 835 }, { "epoch": 8.36e-05, "grad_norm": 12.29367733001709, "learning_rate": 8.35e-06, "loss": 65.8125, "step": 836 }, { "epoch": 8.37e-05, "grad_norm": 12.362329483032227, "learning_rate": 8.36e-06, "loss": 65.5625, "step": 837 }, { "epoch": 8.38e-05, "grad_norm": 12.16931438446045, "learning_rate": 8.37e-06, "loss": 65.875, "step": 838 }, { "epoch": 8.39e-05, "grad_norm": 12.073834419250488, "learning_rate": 8.380000000000001e-06, "loss": 65.6875, "step": 839 }, { "epoch": 8.4e-05, "grad_norm": 12.422995567321777, "learning_rate": 8.39e-06, "loss": 65.5, "step": 840 }, { "epoch": 8.41e-05, "grad_norm": 12.235881805419922, "learning_rate": 8.400000000000001e-06, "loss": 65.5, "step": 841 }, { "epoch": 8.42e-05, "grad_norm": 11.909436225891113, "learning_rate": 8.41e-06, "loss": 65.6875, "step": 842 }, { "epoch": 8.43e-05, "grad_norm": 12.046891212463379, "learning_rate": 8.419999999999999e-06, "loss": 65.6875, "step": 843 }, { "epoch": 8.44e-05, "grad_norm": 12.069416999816895, "learning_rate": 8.43e-06, "loss": 65.25, "step": 844 }, { "epoch": 8.45e-05, "grad_norm": 11.80577278137207, "learning_rate": 8.44e-06, "loss": 65.375, "step": 845 }, { "epoch": 8.46e-05, "grad_norm": 11.65450382232666, "learning_rate": 8.45e-06, "loss": 65.5, "step": 846 }, { "epoch": 8.47e-05, "grad_norm": 11.70303726196289, "learning_rate": 8.46e-06, "loss": 65.4375, "step": 847 }, { "epoch": 8.48e-05, "grad_norm": 12.221776008605957, "learning_rate": 8.47e-06, "loss": 65.25, "step": 848 }, { "epoch": 8.49e-05, "grad_norm": 11.785761833190918, "learning_rate": 8.48e-06, "loss": 65.4375, "step": 849 }, { "epoch": 8.5e-05, "grad_norm": 12.039600372314453, "learning_rate": 8.49e-06, "loss": 65.25, "step": 850 }, { "epoch": 8.51e-05, "grad_norm": 11.632035255432129, "learning_rate": 8.500000000000002e-06, "loss": 65.4375, "step": 851 }, { "epoch": 8.52e-05, "grad_norm": 11.960968017578125, "learning_rate": 8.51e-06, "loss": 65.125, "step": 852 }, { "epoch": 8.53e-05, "grad_norm": 11.791102409362793, "learning_rate": 8.52e-06, "loss": 65.0625, "step": 853 }, { "epoch": 8.54e-05, "grad_norm": 11.28836727142334, "learning_rate": 8.53e-06, "loss": 65.5, "step": 854 }, { "epoch": 8.55e-05, "grad_norm": 11.553174018859863, "learning_rate": 8.54e-06, "loss": 65.125, "step": 855 }, { "epoch": 8.56e-05, "grad_norm": 11.61713981628418, "learning_rate": 8.550000000000001e-06, "loss": 65.0625, "step": 856 }, { "epoch": 8.57e-05, "grad_norm": 11.704262733459473, "learning_rate": 8.56e-06, "loss": 64.8125, "step": 857 }, { "epoch": 8.58e-05, "grad_norm": 11.702911376953125, "learning_rate": 8.57e-06, "loss": 64.875, "step": 858 }, { "epoch": 8.59e-05, "grad_norm": 11.293460845947266, "learning_rate": 8.580000000000001e-06, "loss": 65.0, "step": 859 }, { "epoch": 8.6e-05, "grad_norm": 11.456600189208984, "learning_rate": 8.589999999999999e-06, "loss": 64.7812, "step": 860 }, { "epoch": 8.61e-05, "grad_norm": 11.203218460083008, "learning_rate": 8.6e-06, "loss": 65.125, "step": 861 }, { "epoch": 8.62e-05, "grad_norm": 11.105107307434082, "learning_rate": 8.61e-06, "loss": 65.0, "step": 862 }, { "epoch": 8.63e-05, "grad_norm": 11.6665620803833, "learning_rate": 8.62e-06, "loss": 64.625, "step": 863 }, { "epoch": 8.64e-05, "grad_norm": 11.519858360290527, "learning_rate": 8.63e-06, "loss": 64.8125, "step": 864 }, { "epoch": 8.65e-05, "grad_norm": 11.364030838012695, "learning_rate": 8.64e-06, "loss": 64.8125, "step": 865 }, { "epoch": 8.66e-05, "grad_norm": 11.232035636901855, "learning_rate": 8.65e-06, "loss": 64.8125, "step": 866 }, { "epoch": 8.67e-05, "grad_norm": 11.68748950958252, "learning_rate": 8.66e-06, "loss": 64.5, "step": 867 }, { "epoch": 8.68e-05, "grad_norm": 10.944489479064941, "learning_rate": 8.67e-06, "loss": 64.875, "step": 868 }, { "epoch": 8.69e-05, "grad_norm": 11.102535247802734, "learning_rate": 8.68e-06, "loss": 64.5, "step": 869 }, { "epoch": 8.7e-05, "grad_norm": 10.927079200744629, "learning_rate": 8.69e-06, "loss": 64.9375, "step": 870 }, { "epoch": 8.71e-05, "grad_norm": 10.939054489135742, "learning_rate": 8.7e-06, "loss": 64.4688, "step": 871 }, { "epoch": 8.72e-05, "grad_norm": 11.497695922851562, "learning_rate": 8.71e-06, "loss": 64.625, "step": 872 }, { "epoch": 8.73e-05, "grad_norm": 11.157756805419922, "learning_rate": 8.720000000000001e-06, "loss": 64.8125, "step": 873 }, { "epoch": 8.74e-05, "grad_norm": 10.892260551452637, "learning_rate": 8.73e-06, "loss": 64.5, "step": 874 }, { "epoch": 8.75e-05, "grad_norm": 10.81840991973877, "learning_rate": 8.74e-06, "loss": 64.625, "step": 875 }, { "epoch": 8.76e-05, "grad_norm": 10.906977653503418, "learning_rate": 8.750000000000001e-06, "loss": 64.4062, "step": 876 }, { "epoch": 8.77e-05, "grad_norm": 10.821674346923828, "learning_rate": 8.759999999999999e-06, "loss": 64.5, "step": 877 }, { "epoch": 8.78e-05, "grad_norm": 10.746541023254395, "learning_rate": 8.77e-06, "loss": 64.5625, "step": 878 }, { "epoch": 8.79e-05, "grad_norm": 11.237996101379395, "learning_rate": 8.78e-06, "loss": 64.125, "step": 879 }, { "epoch": 8.8e-05, "grad_norm": 11.085122108459473, "learning_rate": 8.79e-06, "loss": 64.25, "step": 880 }, { "epoch": 8.81e-05, "grad_norm": 10.644852638244629, "learning_rate": 8.8e-06, "loss": 64.4062, "step": 881 }, { "epoch": 8.82e-05, "grad_norm": 10.829193115234375, "learning_rate": 8.81e-06, "loss": 64.0938, "step": 882 }, { "epoch": 8.83e-05, "grad_norm": 10.780670166015625, "learning_rate": 8.82e-06, "loss": 64.4062, "step": 883 }, { "epoch": 8.84e-05, "grad_norm": 10.732144355773926, "learning_rate": 8.83e-06, "loss": 64.4375, "step": 884 }, { "epoch": 8.85e-05, "grad_norm": 10.749310493469238, "learning_rate": 8.84e-06, "loss": 64.0625, "step": 885 }, { "epoch": 8.86e-05, "grad_norm": 10.903584480285645, "learning_rate": 8.85e-06, "loss": 63.8125, "step": 886 }, { "epoch": 8.87e-05, "grad_norm": 10.870782852172852, "learning_rate": 8.86e-06, "loss": 63.5625, "step": 887 }, { "epoch": 8.88e-05, "grad_norm": 10.494131088256836, "learning_rate": 8.87e-06, "loss": 64.2188, "step": 888 }, { "epoch": 8.89e-05, "grad_norm": 10.62818717956543, "learning_rate": 8.88e-06, "loss": 64.3438, "step": 889 }, { "epoch": 8.9e-05, "grad_norm": 10.806772232055664, "learning_rate": 8.890000000000001e-06, "loss": 63.7812, "step": 890 }, { "epoch": 8.91e-05, "grad_norm": 10.955464363098145, "learning_rate": 8.9e-06, "loss": 63.75, "step": 891 }, { "epoch": 8.92e-05, "grad_norm": 10.537474632263184, "learning_rate": 8.91e-06, "loss": 63.6875, "step": 892 }, { "epoch": 8.93e-05, "grad_norm": 10.190065383911133, "learning_rate": 8.920000000000001e-06, "loss": 64.1562, "step": 893 }, { "epoch": 8.94e-05, "grad_norm": 10.386507987976074, "learning_rate": 8.93e-06, "loss": 63.9375, "step": 894 }, { "epoch": 8.95e-05, "grad_norm": 10.566305160522461, "learning_rate": 8.94e-06, "loss": 63.5, "step": 895 }, { "epoch": 8.96e-05, "grad_norm": 10.46068000793457, "learning_rate": 8.95e-06, "loss": 63.6562, "step": 896 }, { "epoch": 8.97e-05, "grad_norm": 10.569717407226562, "learning_rate": 8.96e-06, "loss": 63.5625, "step": 897 }, { "epoch": 8.98e-05, "grad_norm": 10.547991752624512, "learning_rate": 8.97e-06, "loss": 63.625, "step": 898 }, { "epoch": 8.99e-05, "grad_norm": 10.377073287963867, "learning_rate": 8.98e-06, "loss": 63.4688, "step": 899 }, { "epoch": 9e-05, "grad_norm": 10.192441940307617, "learning_rate": 8.99e-06, "loss": 63.75, "step": 900 }, { "epoch": 9.01e-05, "grad_norm": 10.346672058105469, "learning_rate": 9e-06, "loss": 63.5312, "step": 901 }, { "epoch": 9.02e-05, "grad_norm": 10.225927352905273, "learning_rate": 9.01e-06, "loss": 63.5, "step": 902 }, { "epoch": 9.03e-05, "grad_norm": 10.153120994567871, "learning_rate": 9.02e-06, "loss": 63.6562, "step": 903 }, { "epoch": 9.04e-05, "grad_norm": 10.05332088470459, "learning_rate": 9.03e-06, "loss": 63.6875, "step": 904 }, { "epoch": 9.05e-05, "grad_norm": 10.070221900939941, "learning_rate": 9.04e-06, "loss": 63.5938, "step": 905 }, { "epoch": 9.06e-05, "grad_norm": 10.313018798828125, "learning_rate": 9.05e-06, "loss": 63.3438, "step": 906 }, { "epoch": 9.07e-05, "grad_norm": 10.222672462463379, "learning_rate": 9.060000000000001e-06, "loss": 63.3438, "step": 907 }, { "epoch": 9.08e-05, "grad_norm": 10.298966407775879, "learning_rate": 9.07e-06, "loss": 63.3125, "step": 908 }, { "epoch": 9.09e-05, "grad_norm": 9.815750122070312, "learning_rate": 9.08e-06, "loss": 63.625, "step": 909 }, { "epoch": 9.1e-05, "grad_norm": 9.667336463928223, "learning_rate": 9.090000000000001e-06, "loss": 63.625, "step": 910 }, { "epoch": 9.11e-05, "grad_norm": 9.815160751342773, "learning_rate": 9.1e-06, "loss": 63.4062, "step": 911 }, { "epoch": 9.12e-05, "grad_norm": 10.02402114868164, "learning_rate": 9.110000000000001e-06, "loss": 63.1562, "step": 912 }, { "epoch": 9.13e-05, "grad_norm": 10.039457321166992, "learning_rate": 9.12e-06, "loss": 63.0, "step": 913 }, { "epoch": 9.14e-05, "grad_norm": 10.062280654907227, "learning_rate": 9.13e-06, "loss": 63.2812, "step": 914 }, { "epoch": 9.15e-05, "grad_norm": 9.866779327392578, "learning_rate": 9.14e-06, "loss": 63.5, "step": 915 }, { "epoch": 9.16e-05, "grad_norm": 9.92796516418457, "learning_rate": 9.15e-06, "loss": 63.0938, "step": 916 }, { "epoch": 9.17e-05, "grad_norm": 9.809439659118652, "learning_rate": 9.16e-06, "loss": 63.3438, "step": 917 }, { "epoch": 9.18e-05, "grad_norm": 9.726333618164062, "learning_rate": 9.17e-06, "loss": 63.4062, "step": 918 }, { "epoch": 9.19e-05, "grad_norm": 10.27943229675293, "learning_rate": 9.18e-06, "loss": 62.7812, "step": 919 }, { "epoch": 9.2e-05, "grad_norm": 9.923724174499512, "learning_rate": 9.19e-06, "loss": 63.125, "step": 920 }, { "epoch": 9.21e-05, "grad_norm": 9.776262283325195, "learning_rate": 9.2e-06, "loss": 63.1562, "step": 921 }, { "epoch": 9.22e-05, "grad_norm": 10.05753231048584, "learning_rate": 9.21e-06, "loss": 62.5312, "step": 922 }, { "epoch": 9.23e-05, "grad_norm": 10.067216873168945, "learning_rate": 9.22e-06, "loss": 62.6875, "step": 923 }, { "epoch": 9.24e-05, "grad_norm": 9.938775062561035, "learning_rate": 9.23e-06, "loss": 62.5312, "step": 924 }, { "epoch": 9.25e-05, "grad_norm": 9.898990631103516, "learning_rate": 9.24e-06, "loss": 62.5938, "step": 925 }, { "epoch": 9.26e-05, "grad_norm": 9.89505672454834, "learning_rate": 9.25e-06, "loss": 62.75, "step": 926 }, { "epoch": 9.27e-05, "grad_norm": 9.820269584655762, "learning_rate": 9.260000000000001e-06, "loss": 62.8125, "step": 927 }, { "epoch": 9.28e-05, "grad_norm": 9.785930633544922, "learning_rate": 9.27e-06, "loss": 62.7188, "step": 928 }, { "epoch": 9.29e-05, "grad_norm": 9.649805068969727, "learning_rate": 9.280000000000001e-06, "loss": 62.75, "step": 929 }, { "epoch": 9.3e-05, "grad_norm": 9.778607368469238, "learning_rate": 9.29e-06, "loss": 62.5312, "step": 930 }, { "epoch": 9.31e-05, "grad_norm": 9.530630111694336, "learning_rate": 9.299999999999999e-06, "loss": 62.75, "step": 931 }, { "epoch": 9.32e-05, "grad_norm": 9.770853042602539, "learning_rate": 9.31e-06, "loss": 62.375, "step": 932 }, { "epoch": 9.33e-05, "grad_norm": 9.653392791748047, "learning_rate": 9.32e-06, "loss": 62.6875, "step": 933 }, { "epoch": 9.34e-05, "grad_norm": 9.52197265625, "learning_rate": 9.33e-06, "loss": 62.4688, "step": 934 }, { "epoch": 9.35e-05, "grad_norm": 9.621892929077148, "learning_rate": 9.34e-06, "loss": 62.4375, "step": 935 }, { "epoch": 9.36e-05, "grad_norm": 9.49044132232666, "learning_rate": 9.35e-06, "loss": 63.0938, "step": 936 }, { "epoch": 9.37e-05, "grad_norm": 9.625767707824707, "learning_rate": 9.36e-06, "loss": 62.6875, "step": 937 }, { "epoch": 9.38e-05, "grad_norm": 9.668294906616211, "learning_rate": 9.37e-06, "loss": 62.2812, "step": 938 }, { "epoch": 9.39e-05, "grad_norm": 9.578293800354004, "learning_rate": 9.38e-06, "loss": 62.3125, "step": 939 }, { "epoch": 9.4e-05, "grad_norm": 9.879005432128906, "learning_rate": 9.39e-06, "loss": 62.3125, "step": 940 }, { "epoch": 9.41e-05, "grad_norm": 9.26992130279541, "learning_rate": 9.4e-06, "loss": 62.5938, "step": 941 }, { "epoch": 9.42e-05, "grad_norm": 9.741432189941406, "learning_rate": 9.41e-06, "loss": 62.125, "step": 942 }, { "epoch": 9.43e-05, "grad_norm": 9.542763710021973, "learning_rate": 9.42e-06, "loss": 62.4062, "step": 943 }, { "epoch": 9.44e-05, "grad_norm": 9.331867218017578, "learning_rate": 9.430000000000001e-06, "loss": 62.375, "step": 944 }, { "epoch": 9.45e-05, "grad_norm": 9.504682540893555, "learning_rate": 9.44e-06, "loss": 62.0312, "step": 945 }, { "epoch": 9.46e-05, "grad_norm": 9.386909484863281, "learning_rate": 9.450000000000001e-06, "loss": 62.375, "step": 946 }, { "epoch": 9.47e-05, "grad_norm": 9.543030738830566, "learning_rate": 9.460000000000001e-06, "loss": 62.0938, "step": 947 }, { "epoch": 9.48e-05, "grad_norm": 9.453919410705566, "learning_rate": 9.469999999999999e-06, "loss": 62.375, "step": 948 }, { "epoch": 9.49e-05, "grad_norm": 9.36674690246582, "learning_rate": 9.48e-06, "loss": 62.0312, "step": 949 }, { "epoch": 9.5e-05, "grad_norm": 9.571072578430176, "learning_rate": 9.49e-06, "loss": 62.0312, "step": 950 }, { "epoch": 9.51e-05, "grad_norm": 9.615150451660156, "learning_rate": 9.5e-06, "loss": 61.8125, "step": 951 }, { "epoch": 9.52e-05, "grad_norm": 9.266009330749512, "learning_rate": 9.51e-06, "loss": 62.375, "step": 952 }, { "epoch": 9.53e-05, "grad_norm": 9.239862442016602, "learning_rate": 9.52e-06, "loss": 62.3438, "step": 953 }, { "epoch": 9.54e-05, "grad_norm": 9.388384819030762, "learning_rate": 9.53e-06, "loss": 61.9062, "step": 954 }, { "epoch": 9.55e-05, "grad_norm": 9.385138511657715, "learning_rate": 9.54e-06, "loss": 61.9375, "step": 955 }, { "epoch": 9.56e-05, "grad_norm": 9.419622421264648, "learning_rate": 9.55e-06, "loss": 61.7188, "step": 956 }, { "epoch": 9.57e-05, "grad_norm": 9.560093879699707, "learning_rate": 9.56e-06, "loss": 61.6562, "step": 957 }, { "epoch": 9.58e-05, "grad_norm": 9.4889554977417, "learning_rate": 9.57e-06, "loss": 62.0625, "step": 958 }, { "epoch": 9.59e-05, "grad_norm": 8.988265991210938, "learning_rate": 9.58e-06, "loss": 62.3125, "step": 959 }, { "epoch": 9.6e-05, "grad_norm": 9.52734661102295, "learning_rate": 9.59e-06, "loss": 61.5312, "step": 960 }, { "epoch": 9.61e-05, "grad_norm": 9.097179412841797, "learning_rate": 9.600000000000001e-06, "loss": 62.1562, "step": 961 }, { "epoch": 9.62e-05, "grad_norm": 9.454554557800293, "learning_rate": 9.61e-06, "loss": 61.4688, "step": 962 }, { "epoch": 9.63e-05, "grad_norm": 9.153898239135742, "learning_rate": 9.620000000000001e-06, "loss": 62.125, "step": 963 }, { "epoch": 9.64e-05, "grad_norm": 9.330815315246582, "learning_rate": 9.630000000000001e-06, "loss": 61.6562, "step": 964 }, { "epoch": 9.65e-05, "grad_norm": 9.72275161743164, "learning_rate": 9.64e-06, "loss": 61.3125, "step": 965 }, { "epoch": 9.66e-05, "grad_norm": 9.388081550598145, "learning_rate": 9.65e-06, "loss": 61.4688, "step": 966 }, { "epoch": 9.67e-05, "grad_norm": 9.236139297485352, "learning_rate": 9.66e-06, "loss": 61.5938, "step": 967 }, { "epoch": 9.68e-05, "grad_norm": 9.027604103088379, "learning_rate": 9.67e-06, "loss": 62.0312, "step": 968 }, { "epoch": 9.69e-05, "grad_norm": 9.193624496459961, "learning_rate": 9.68e-06, "loss": 61.5, "step": 969 }, { "epoch": 9.7e-05, "grad_norm": 9.14177131652832, "learning_rate": 9.69e-06, "loss": 61.75, "step": 970 }, { "epoch": 9.71e-05, "grad_norm": 9.21058177947998, "learning_rate": 9.7e-06, "loss": 61.75, "step": 971 }, { "epoch": 9.72e-05, "grad_norm": 9.118620872497559, "learning_rate": 9.71e-06, "loss": 61.75, "step": 972 }, { "epoch": 9.73e-05, "grad_norm": 9.088716506958008, "learning_rate": 9.72e-06, "loss": 61.625, "step": 973 }, { "epoch": 9.74e-05, "grad_norm": 9.206241607666016, "learning_rate": 9.73e-06, "loss": 61.25, "step": 974 }, { "epoch": 9.75e-05, "grad_norm": 9.151651382446289, "learning_rate": 9.74e-06, "loss": 61.2188, "step": 975 }, { "epoch": 9.76e-05, "grad_norm": 9.240182876586914, "learning_rate": 9.75e-06, "loss": 61.4688, "step": 976 }, { "epoch": 9.77e-05, "grad_norm": 9.352987289428711, "learning_rate": 9.76e-06, "loss": 61.25, "step": 977 }, { "epoch": 9.78e-05, "grad_norm": 9.508001327514648, "learning_rate": 9.770000000000001e-06, "loss": 61.0, "step": 978 }, { "epoch": 9.79e-05, "grad_norm": 8.883227348327637, "learning_rate": 9.78e-06, "loss": 62.0, "step": 979 }, { "epoch": 9.8e-05, "grad_norm": 9.131720542907715, "learning_rate": 9.790000000000001e-06, "loss": 61.125, "step": 980 }, { "epoch": 9.81e-05, "grad_norm": 9.582850456237793, "learning_rate": 9.800000000000001e-06, "loss": 60.6875, "step": 981 }, { "epoch": 9.82e-05, "grad_norm": 9.125800132751465, "learning_rate": 9.81e-06, "loss": 61.0625, "step": 982 }, { "epoch": 9.83e-05, "grad_norm": 9.01922607421875, "learning_rate": 9.820000000000001e-06, "loss": 61.125, "step": 983 }, { "epoch": 9.84e-05, "grad_norm": 9.03088092803955, "learning_rate": 9.83e-06, "loss": 61.0625, "step": 984 }, { "epoch": 9.85e-05, "grad_norm": 9.083083152770996, "learning_rate": 9.84e-06, "loss": 61.0312, "step": 985 }, { "epoch": 9.86e-05, "grad_norm": 9.077742576599121, "learning_rate": 9.85e-06, "loss": 61.0312, "step": 986 }, { "epoch": 9.87e-05, "grad_norm": 9.057887077331543, "learning_rate": 9.859999999999999e-06, "loss": 60.875, "step": 987 }, { "epoch": 9.88e-05, "grad_norm": 8.55154800415039, "learning_rate": 9.87e-06, "loss": 62.375, "step": 988 }, { "epoch": 9.89e-05, "grad_norm": 8.761177062988281, "learning_rate": 9.88e-06, "loss": 61.125, "step": 989 }, { "epoch": 9.9e-05, "grad_norm": 9.016249656677246, "learning_rate": 9.89e-06, "loss": 61.2188, "step": 990 }, { "epoch": 9.91e-05, "grad_norm": 8.754048347473145, "learning_rate": 9.9e-06, "loss": 61.1875, "step": 991 }, { "epoch": 9.92e-05, "grad_norm": 9.200346946716309, "learning_rate": 9.91e-06, "loss": 60.5, "step": 992 }, { "epoch": 9.93e-05, "grad_norm": 8.876838684082031, "learning_rate": 9.92e-06, "loss": 61.0, "step": 993 }, { "epoch": 9.94e-05, "grad_norm": 8.757495880126953, "learning_rate": 9.93e-06, "loss": 61.4062, "step": 994 }, { "epoch": 9.95e-05, "grad_norm": 8.726040840148926, "learning_rate": 9.940000000000001e-06, "loss": 61.6562, "step": 995 }, { "epoch": 9.96e-05, "grad_norm": 8.834831237792969, "learning_rate": 9.95e-06, "loss": 61.125, "step": 996 }, { "epoch": 9.97e-05, "grad_norm": 8.835776329040527, "learning_rate": 9.96e-06, "loss": 60.6562, "step": 997 }, { "epoch": 9.98e-05, "grad_norm": 8.862153053283691, "learning_rate": 9.970000000000001e-06, "loss": 61.0312, "step": 998 }, { "epoch": 9.99e-05, "grad_norm": 9.10541820526123, "learning_rate": 9.98e-06, "loss": 60.5625, "step": 999 }, { "epoch": 0.0001, "grad_norm": 9.29044246673584, "learning_rate": 9.990000000000001e-06, "loss": 60.375, "step": 1000 }, { "epoch": 0.0001, "eval_loss": 7.538768768310547, "eval_runtime": 364.1443, "eval_samples_per_second": 27.462, "eval_steps_per_second": 1.716, "step": 1000 }, { "epoch": 0.0001001, "grad_norm": 9.124345779418945, "learning_rate": 1e-05, "loss": 60.4375, "step": 1001 }, { "epoch": 0.0001002, "grad_norm": 9.41726303100586, "learning_rate": 1.001e-05, "loss": 60.0938, "step": 1002 }, { "epoch": 0.0001003, "grad_norm": 8.93173599243164, "learning_rate": 1.002e-05, "loss": 60.6875, "step": 1003 }, { "epoch": 0.0001004, "grad_norm": 9.111990928649902, "learning_rate": 1.0029999999999999e-05, "loss": 60.3438, "step": 1004 }, { "epoch": 0.0001005, "grad_norm": 8.782609939575195, "learning_rate": 1.004e-05, "loss": 61.1562, "step": 1005 }, { "epoch": 0.0001006, "grad_norm": 9.252752304077148, "learning_rate": 1.005e-05, "loss": 60.1562, "step": 1006 }, { "epoch": 0.0001007, "grad_norm": 8.887770652770996, "learning_rate": 1.006e-05, "loss": 60.375, "step": 1007 }, { "epoch": 0.0001008, "grad_norm": 8.99946117401123, "learning_rate": 1.007e-05, "loss": 60.375, "step": 1008 }, { "epoch": 0.0001009, "grad_norm": 9.170857429504395, "learning_rate": 1.008e-05, "loss": 60.2812, "step": 1009 }, { "epoch": 0.000101, "grad_norm": 9.04532241821289, "learning_rate": 1.009e-05, "loss": 60.125, "step": 1010 }, { "epoch": 0.0001011, "grad_norm": 8.962970733642578, "learning_rate": 1.01e-05, "loss": 60.25, "step": 1011 }, { "epoch": 0.0001012, "grad_norm": 8.685812950134277, "learning_rate": 1.0110000000000001e-05, "loss": 60.6562, "step": 1012 }, { "epoch": 0.0001013, "grad_norm": 8.879267692565918, "learning_rate": 1.012e-05, "loss": 60.125, "step": 1013 }, { "epoch": 0.0001014, "grad_norm": 8.940817832946777, "learning_rate": 1.013e-05, "loss": 60.25, "step": 1014 }, { "epoch": 0.0001015, "grad_norm": 8.88219165802002, "learning_rate": 1.0140000000000001e-05, "loss": 60.25, "step": 1015 }, { "epoch": 0.0001016, "grad_norm": 9.43167781829834, "learning_rate": 1.015e-05, "loss": 60.0, "step": 1016 }, { "epoch": 0.0001017, "grad_norm": 9.163987159729004, "learning_rate": 1.0160000000000001e-05, "loss": 59.6562, "step": 1017 }, { "epoch": 0.0001018, "grad_norm": 8.8132905960083, "learning_rate": 1.017e-05, "loss": 60.0312, "step": 1018 }, { "epoch": 0.0001019, "grad_norm": 9.031733512878418, "learning_rate": 1.018e-05, "loss": 59.9688, "step": 1019 }, { "epoch": 0.000102, "grad_norm": 9.099145889282227, "learning_rate": 1.019e-05, "loss": 59.6562, "step": 1020 }, { "epoch": 0.0001021, "grad_norm": 9.057586669921875, "learning_rate": 1.0199999999999999e-05, "loss": 59.75, "step": 1021 }, { "epoch": 0.0001022, "grad_norm": 8.728660583496094, "learning_rate": 1.021e-05, "loss": 60.0938, "step": 1022 }, { "epoch": 0.0001023, "grad_norm": 9.054150581359863, "learning_rate": 1.022e-05, "loss": 59.5, "step": 1023 }, { "epoch": 0.0001024, "grad_norm": 8.890925407409668, "learning_rate": 1.023e-05, "loss": 59.875, "step": 1024 }, { "epoch": 0.0001025, "grad_norm": 8.602814674377441, "learning_rate": 1.024e-05, "loss": 60.0938, "step": 1025 }, { "epoch": 0.0001026, "grad_norm": 8.891504287719727, "learning_rate": 1.025e-05, "loss": 59.4375, "step": 1026 }, { "epoch": 0.0001027, "grad_norm": 8.879639625549316, "learning_rate": 1.026e-05, "loss": 59.9062, "step": 1027 }, { "epoch": 0.0001028, "grad_norm": 9.04471206665039, "learning_rate": 1.027e-05, "loss": 59.3125, "step": 1028 }, { "epoch": 0.0001029, "grad_norm": 8.917312622070312, "learning_rate": 1.0280000000000002e-05, "loss": 59.9375, "step": 1029 }, { "epoch": 0.000103, "grad_norm": 8.734441757202148, "learning_rate": 1.029e-05, "loss": 59.9688, "step": 1030 }, { "epoch": 0.0001031, "grad_norm": 9.322687149047852, "learning_rate": 1.03e-05, "loss": 59.0312, "step": 1031 }, { "epoch": 0.0001032, "grad_norm": 8.85023307800293, "learning_rate": 1.0310000000000001e-05, "loss": 59.5312, "step": 1032 }, { "epoch": 0.0001033, "grad_norm": 8.855387687683105, "learning_rate": 1.032e-05, "loss": 59.5312, "step": 1033 }, { "epoch": 0.0001034, "grad_norm": 9.083724021911621, "learning_rate": 1.0330000000000001e-05, "loss": 59.125, "step": 1034 }, { "epoch": 0.0001035, "grad_norm": 8.690860748291016, "learning_rate": 1.034e-05, "loss": 59.5938, "step": 1035 }, { "epoch": 0.0001036, "grad_norm": 8.5335693359375, "learning_rate": 1.0350000000000001e-05, "loss": 59.875, "step": 1036 }, { "epoch": 0.0001037, "grad_norm": 8.61517333984375, "learning_rate": 1.036e-05, "loss": 59.7812, "step": 1037 }, { "epoch": 0.0001038, "grad_norm": 8.838473320007324, "learning_rate": 1.0369999999999999e-05, "loss": 59.3125, "step": 1038 }, { "epoch": 0.0001039, "grad_norm": 8.471953392028809, "learning_rate": 1.038e-05, "loss": 59.9375, "step": 1039 }, { "epoch": 0.000104, "grad_norm": 8.951537132263184, "learning_rate": 1.039e-05, "loss": 59.4375, "step": 1040 }, { "epoch": 0.0001041, "grad_norm": 9.035890579223633, "learning_rate": 1.04e-05, "loss": 59.0938, "step": 1041 }, { "epoch": 0.0001042, "grad_norm": 9.102971076965332, "learning_rate": 1.041e-05, "loss": 58.9062, "step": 1042 }, { "epoch": 0.0001043, "grad_norm": 9.211016654968262, "learning_rate": 1.042e-05, "loss": 58.5625, "step": 1043 }, { "epoch": 0.0001044, "grad_norm": 8.802969932556152, "learning_rate": 1.043e-05, "loss": 59.125, "step": 1044 }, { "epoch": 0.0001045, "grad_norm": 8.844217300415039, "learning_rate": 1.044e-05, "loss": 59.0, "step": 1045 }, { "epoch": 0.0001046, "grad_norm": 8.752573013305664, "learning_rate": 1.045e-05, "loss": 59.3125, "step": 1046 }, { "epoch": 0.0001047, "grad_norm": 8.805431365966797, "learning_rate": 1.046e-05, "loss": 58.9688, "step": 1047 }, { "epoch": 0.0001048, "grad_norm": 8.648407936096191, "learning_rate": 1.047e-05, "loss": 58.9688, "step": 1048 }, { "epoch": 0.0001049, "grad_norm": 8.364664077758789, "learning_rate": 1.0480000000000001e-05, "loss": 59.5938, "step": 1049 }, { "epoch": 0.000105, "grad_norm": 8.843344688415527, "learning_rate": 1.049e-05, "loss": 59.0, "step": 1050 }, { "epoch": 0.0001051, "grad_norm": 8.329179763793945, "learning_rate": 1.0500000000000001e-05, "loss": 59.4062, "step": 1051 }, { "epoch": 0.0001052, "grad_norm": 8.471250534057617, "learning_rate": 1.051e-05, "loss": 59.3438, "step": 1052 }, { "epoch": 0.0001053, "grad_norm": 8.373708724975586, "learning_rate": 1.052e-05, "loss": 59.375, "step": 1053 }, { "epoch": 0.0001054, "grad_norm": 8.504606246948242, "learning_rate": 1.053e-05, "loss": 59.6562, "step": 1054 }, { "epoch": 0.0001055, "grad_norm": 8.886446952819824, "learning_rate": 1.0539999999999999e-05, "loss": 58.7188, "step": 1055 }, { "epoch": 0.0001056, "grad_norm": 8.751822471618652, "learning_rate": 1.055e-05, "loss": 58.8438, "step": 1056 }, { "epoch": 0.0001057, "grad_norm": 8.406770706176758, "learning_rate": 1.056e-05, "loss": 59.2188, "step": 1057 }, { "epoch": 0.0001058, "grad_norm": 8.74715805053711, "learning_rate": 1.057e-05, "loss": 58.6562, "step": 1058 }, { "epoch": 0.0001059, "grad_norm": 8.3496732711792, "learning_rate": 1.058e-05, "loss": 59.0312, "step": 1059 }, { "epoch": 0.000106, "grad_norm": 8.718629837036133, "learning_rate": 1.059e-05, "loss": 58.4688, "step": 1060 }, { "epoch": 0.0001061, "grad_norm": 8.428848266601562, "learning_rate": 1.06e-05, "loss": 58.9688, "step": 1061 }, { "epoch": 0.0001062, "grad_norm": 9.044172286987305, "learning_rate": 1.061e-05, "loss": 58.0625, "step": 1062 }, { "epoch": 0.0001063, "grad_norm": 8.45057487487793, "learning_rate": 1.062e-05, "loss": 58.7188, "step": 1063 }, { "epoch": 0.0001064, "grad_norm": 8.13118839263916, "learning_rate": 1.063e-05, "loss": 58.9062, "step": 1064 }, { "epoch": 0.0001065, "grad_norm": 8.686187744140625, "learning_rate": 1.064e-05, "loss": 58.9688, "step": 1065 }, { "epoch": 0.0001066, "grad_norm": 8.561543464660645, "learning_rate": 1.0650000000000001e-05, "loss": 58.3438, "step": 1066 }, { "epoch": 0.0001067, "grad_norm": 8.650547981262207, "learning_rate": 1.066e-05, "loss": 58.5, "step": 1067 }, { "epoch": 0.0001068, "grad_norm": 8.45773983001709, "learning_rate": 1.0670000000000001e-05, "loss": 58.625, "step": 1068 }, { "epoch": 0.0001069, "grad_norm": 8.8767671585083, "learning_rate": 1.068e-05, "loss": 57.8438, "step": 1069 }, { "epoch": 0.000107, "grad_norm": 8.721537590026855, "learning_rate": 1.069e-05, "loss": 57.7812, "step": 1070 }, { "epoch": 0.0001071, "grad_norm": 8.56977653503418, "learning_rate": 1.0700000000000001e-05, "loss": 58.2812, "step": 1071 }, { "epoch": 0.0001072, "grad_norm": 8.458526611328125, "learning_rate": 1.0709999999999999e-05, "loss": 58.875, "step": 1072 }, { "epoch": 0.0001073, "grad_norm": 8.45875358581543, "learning_rate": 1.072e-05, "loss": 58.7188, "step": 1073 }, { "epoch": 0.0001074, "grad_norm": 8.583601951599121, "learning_rate": 1.073e-05, "loss": 58.2812, "step": 1074 }, { "epoch": 0.0001075, "grad_norm": 8.412371635437012, "learning_rate": 1.074e-05, "loss": 58.4062, "step": 1075 }, { "epoch": 0.0001076, "grad_norm": 8.315399169921875, "learning_rate": 1.075e-05, "loss": 58.3438, "step": 1076 }, { "epoch": 0.0001077, "grad_norm": 8.517126083374023, "learning_rate": 1.076e-05, "loss": 58.3125, "step": 1077 }, { "epoch": 0.0001078, "grad_norm": 8.215516090393066, "learning_rate": 1.077e-05, "loss": 58.5938, "step": 1078 }, { "epoch": 0.0001079, "grad_norm": 8.485785484313965, "learning_rate": 1.078e-05, "loss": 58.0312, "step": 1079 }, { "epoch": 0.000108, "grad_norm": 8.737732887268066, "learning_rate": 1.079e-05, "loss": 57.9375, "step": 1080 }, { "epoch": 0.0001081, "grad_norm": 8.925642967224121, "learning_rate": 1.08e-05, "loss": 57.2812, "step": 1081 }, { "epoch": 0.0001082, "grad_norm": 8.599075317382812, "learning_rate": 1.081e-05, "loss": 57.9062, "step": 1082 }, { "epoch": 0.0001083, "grad_norm": 8.235535621643066, "learning_rate": 1.0820000000000001e-05, "loss": 58.8125, "step": 1083 }, { "epoch": 0.0001084, "grad_norm": 8.46491527557373, "learning_rate": 1.083e-05, "loss": 58.0938, "step": 1084 }, { "epoch": 0.0001085, "grad_norm": 8.536995887756348, "learning_rate": 1.0840000000000001e-05, "loss": 57.75, "step": 1085 }, { "epoch": 0.0001086, "grad_norm": 8.33989143371582, "learning_rate": 1.085e-05, "loss": 58.4375, "step": 1086 }, { "epoch": 0.0001087, "grad_norm": 8.58334732055664, "learning_rate": 1.086e-05, "loss": 58.0312, "step": 1087 }, { "epoch": 0.0001088, "grad_norm": 8.356263160705566, "learning_rate": 1.0870000000000001e-05, "loss": 58.0, "step": 1088 }, { "epoch": 0.0001089, "grad_norm": 8.417243003845215, "learning_rate": 1.088e-05, "loss": 57.9375, "step": 1089 }, { "epoch": 0.000109, "grad_norm": 8.363197326660156, "learning_rate": 1.089e-05, "loss": 57.875, "step": 1090 }, { "epoch": 0.0001091, "grad_norm": 8.385573387145996, "learning_rate": 1.09e-05, "loss": 57.625, "step": 1091 }, { "epoch": 0.0001092, "grad_norm": 8.56344223022461, "learning_rate": 1.091e-05, "loss": 58.0938, "step": 1092 }, { "epoch": 0.0001093, "grad_norm": 8.359785079956055, "learning_rate": 1.092e-05, "loss": 57.7812, "step": 1093 }, { "epoch": 0.0001094, "grad_norm": 8.217353820800781, "learning_rate": 1.093e-05, "loss": 58.125, "step": 1094 }, { "epoch": 0.0001095, "grad_norm": 8.240918159484863, "learning_rate": 1.094e-05, "loss": 58.2188, "step": 1095 }, { "epoch": 0.0001096, "grad_norm": 8.025176048278809, "learning_rate": 1.095e-05, "loss": 57.6562, "step": 1096 }, { "epoch": 0.0001097, "grad_norm": 7.992165565490723, "learning_rate": 1.096e-05, "loss": 57.875, "step": 1097 }, { "epoch": 0.0001098, "grad_norm": 8.216459274291992, "learning_rate": 1.097e-05, "loss": 57.9062, "step": 1098 }, { "epoch": 0.0001099, "grad_norm": 8.325234413146973, "learning_rate": 1.098e-05, "loss": 57.5625, "step": 1099 }, { "epoch": 0.00011, "grad_norm": 8.516204833984375, "learning_rate": 1.0990000000000002e-05, "loss": 57.1875, "step": 1100 }, { "epoch": 0.0001101, "grad_norm": 8.354333877563477, "learning_rate": 1.1e-05, "loss": 57.4062, "step": 1101 }, { "epoch": 0.0001102, "grad_norm": 8.326759338378906, "learning_rate": 1.1010000000000001e-05, "loss": 56.9375, "step": 1102 }, { "epoch": 0.0001103, "grad_norm": 8.392999649047852, "learning_rate": 1.102e-05, "loss": 57.0, "step": 1103 }, { "epoch": 0.0001104, "grad_norm": 8.289926528930664, "learning_rate": 1.103e-05, "loss": 57.5625, "step": 1104 }, { "epoch": 0.0001105, "grad_norm": 8.27065658569336, "learning_rate": 1.1040000000000001e-05, "loss": 56.9375, "step": 1105 }, { "epoch": 0.0001106, "grad_norm": 7.944595813751221, "learning_rate": 1.105e-05, "loss": 58.125, "step": 1106 }, { "epoch": 0.0001107, "grad_norm": 8.095520973205566, "learning_rate": 1.106e-05, "loss": 57.6875, "step": 1107 }, { "epoch": 0.0001108, "grad_norm": 8.457866668701172, "learning_rate": 1.107e-05, "loss": 56.8438, "step": 1108 }, { "epoch": 0.0001109, "grad_norm": 8.350686073303223, "learning_rate": 1.108e-05, "loss": 57.0, "step": 1109 }, { "epoch": 0.000111, "grad_norm": 8.33674430847168, "learning_rate": 1.109e-05, "loss": 56.75, "step": 1110 }, { "epoch": 0.0001111, "grad_norm": 8.197669982910156, "learning_rate": 1.11e-05, "loss": 57.2812, "step": 1111 }, { "epoch": 0.0001112, "grad_norm": 8.191856384277344, "learning_rate": 1.111e-05, "loss": 56.9688, "step": 1112 }, { "epoch": 0.0001113, "grad_norm": 8.121025085449219, "learning_rate": 1.112e-05, "loss": 57.1875, "step": 1113 }, { "epoch": 0.0001114, "grad_norm": 8.034751892089844, "learning_rate": 1.113e-05, "loss": 57.2188, "step": 1114 }, { "epoch": 0.0001115, "grad_norm": 8.12697982788086, "learning_rate": 1.114e-05, "loss": 57.3125, "step": 1115 }, { "epoch": 0.0001116, "grad_norm": 7.979389667510986, "learning_rate": 1.115e-05, "loss": 57.1562, "step": 1116 }, { "epoch": 0.0001117, "grad_norm": 8.33296012878418, "learning_rate": 1.1160000000000002e-05, "loss": 56.9062, "step": 1117 }, { "epoch": 0.0001118, "grad_norm": 8.188544273376465, "learning_rate": 1.117e-05, "loss": 56.9062, "step": 1118 }, { "epoch": 0.0001119, "grad_norm": 7.855072975158691, "learning_rate": 1.118e-05, "loss": 57.1875, "step": 1119 }, { "epoch": 0.000112, "grad_norm": 8.365545272827148, "learning_rate": 1.119e-05, "loss": 56.4688, "step": 1120 }, { "epoch": 0.0001121, "grad_norm": 8.095410346984863, "learning_rate": 1.12e-05, "loss": 56.875, "step": 1121 }, { "epoch": 0.0001122, "grad_norm": 7.993049144744873, "learning_rate": 1.1210000000000001e-05, "loss": 57.2188, "step": 1122 }, { "epoch": 0.0001123, "grad_norm": 8.100505828857422, "learning_rate": 1.122e-05, "loss": 56.2812, "step": 1123 }, { "epoch": 0.0001124, "grad_norm": 8.246110916137695, "learning_rate": 1.1230000000000001e-05, "loss": 56.4375, "step": 1124 }, { "epoch": 0.0001125, "grad_norm": 8.12435531616211, "learning_rate": 1.124e-05, "loss": 56.5, "step": 1125 }, { "epoch": 0.0001126, "grad_norm": 7.739667892456055, "learning_rate": 1.1249999999999999e-05, "loss": 57.3438, "step": 1126 }, { "epoch": 0.0001127, "grad_norm": 8.352540016174316, "learning_rate": 1.126e-05, "loss": 56.1562, "step": 1127 }, { "epoch": 0.0001128, "grad_norm": 7.80290412902832, "learning_rate": 1.127e-05, "loss": 57.125, "step": 1128 }, { "epoch": 0.0001129, "grad_norm": 7.911022663116455, "learning_rate": 1.128e-05, "loss": 57.0625, "step": 1129 }, { "epoch": 0.000113, "grad_norm": 8.22894287109375, "learning_rate": 1.129e-05, "loss": 56.25, "step": 1130 }, { "epoch": 0.0001131, "grad_norm": 8.299407005310059, "learning_rate": 1.13e-05, "loss": 55.9688, "step": 1131 }, { "epoch": 0.0001132, "grad_norm": 8.011561393737793, "learning_rate": 1.131e-05, "loss": 56.3125, "step": 1132 }, { "epoch": 0.0001133, "grad_norm": 8.105046272277832, "learning_rate": 1.132e-05, "loss": 55.875, "step": 1133 }, { "epoch": 0.0001134, "grad_norm": 7.755193710327148, "learning_rate": 1.133e-05, "loss": 57.2188, "step": 1134 }, { "epoch": 0.0001135, "grad_norm": 7.714022636413574, "learning_rate": 1.134e-05, "loss": 56.8125, "step": 1135 }, { "epoch": 0.0001136, "grad_norm": 8.133477210998535, "learning_rate": 1.135e-05, "loss": 55.6875, "step": 1136 }, { "epoch": 0.0001137, "grad_norm": 8.036331176757812, "learning_rate": 1.136e-05, "loss": 56.0625, "step": 1137 }, { "epoch": 0.0001138, "grad_norm": 7.8458356857299805, "learning_rate": 1.137e-05, "loss": 56.375, "step": 1138 }, { "epoch": 0.0001139, "grad_norm": 8.083093643188477, "learning_rate": 1.1380000000000001e-05, "loss": 56.0625, "step": 1139 }, { "epoch": 0.000114, "grad_norm": 8.058805465698242, "learning_rate": 1.139e-05, "loss": 56.0, "step": 1140 }, { "epoch": 0.0001141, "grad_norm": 7.845074653625488, "learning_rate": 1.1400000000000001e-05, "loss": 56.625, "step": 1141 }, { "epoch": 0.0001142, "grad_norm": 7.745107173919678, "learning_rate": 1.1410000000000001e-05, "loss": 57.0, "step": 1142 }, { "epoch": 0.0001143, "grad_norm": 8.069292068481445, "learning_rate": 1.1419999999999999e-05, "loss": 55.7812, "step": 1143 }, { "epoch": 0.0001144, "grad_norm": 7.823101043701172, "learning_rate": 1.143e-05, "loss": 56.25, "step": 1144 }, { "epoch": 0.0001145, "grad_norm": 7.864339828491211, "learning_rate": 1.144e-05, "loss": 55.9375, "step": 1145 }, { "epoch": 0.0001146, "grad_norm": 7.925984859466553, "learning_rate": 1.145e-05, "loss": 56.4688, "step": 1146 }, { "epoch": 0.0001147, "grad_norm": 7.802701473236084, "learning_rate": 1.146e-05, "loss": 56.375, "step": 1147 }, { "epoch": 0.0001148, "grad_norm": 8.001948356628418, "learning_rate": 1.147e-05, "loss": 55.6875, "step": 1148 }, { "epoch": 0.0001149, "grad_norm": 8.064249038696289, "learning_rate": 1.148e-05, "loss": 55.5625, "step": 1149 }, { "epoch": 0.000115, "grad_norm": 8.19449520111084, "learning_rate": 1.149e-05, "loss": 55.4375, "step": 1150 }, { "epoch": 0.0001151, "grad_norm": 8.22554874420166, "learning_rate": 1.15e-05, "loss": 55.0312, "step": 1151 }, { "epoch": 0.0001152, "grad_norm": 7.867717742919922, "learning_rate": 1.151e-05, "loss": 55.8125, "step": 1152 }, { "epoch": 0.0001153, "grad_norm": 7.817559242248535, "learning_rate": 1.152e-05, "loss": 55.7188, "step": 1153 }, { "epoch": 0.0001154, "grad_norm": 7.744104385375977, "learning_rate": 1.153e-05, "loss": 55.6562, "step": 1154 }, { "epoch": 0.0001155, "grad_norm": 7.854978084564209, "learning_rate": 1.154e-05, "loss": 55.875, "step": 1155 }, { "epoch": 0.0001156, "grad_norm": 7.853444576263428, "learning_rate": 1.1550000000000001e-05, "loss": 55.75, "step": 1156 }, { "epoch": 0.0001157, "grad_norm": 7.83394718170166, "learning_rate": 1.156e-05, "loss": 55.8438, "step": 1157 }, { "epoch": 0.0001158, "grad_norm": 7.851531505584717, "learning_rate": 1.1570000000000001e-05, "loss": 55.4375, "step": 1158 }, { "epoch": 0.0001159, "grad_norm": 7.822463035583496, "learning_rate": 1.1580000000000001e-05, "loss": 55.0938, "step": 1159 }, { "epoch": 0.000116, "grad_norm": 8.192028045654297, "learning_rate": 1.1589999999999999e-05, "loss": 54.6875, "step": 1160 }, { "epoch": 0.0001161, "grad_norm": 7.832354545593262, "learning_rate": 1.16e-05, "loss": 55.0938, "step": 1161 }, { "epoch": 0.0001162, "grad_norm": 7.876185894012451, "learning_rate": 1.161e-05, "loss": 54.875, "step": 1162 }, { "epoch": 0.0001163, "grad_norm": 7.61892557144165, "learning_rate": 1.162e-05, "loss": 56.3125, "step": 1163 }, { "epoch": 0.0001164, "grad_norm": 7.596429824829102, "learning_rate": 1.163e-05, "loss": 55.375, "step": 1164 }, { "epoch": 0.0001165, "grad_norm": 7.708346843719482, "learning_rate": 1.164e-05, "loss": 55.5312, "step": 1165 }, { "epoch": 0.0001166, "grad_norm": 7.724869251251221, "learning_rate": 1.165e-05, "loss": 55.1562, "step": 1166 }, { "epoch": 0.0001167, "grad_norm": 7.922326564788818, "learning_rate": 1.166e-05, "loss": 55.125, "step": 1167 }, { "epoch": 0.0001168, "grad_norm": 7.799662113189697, "learning_rate": 1.167e-05, "loss": 55.0625, "step": 1168 }, { "epoch": 0.0001169, "grad_norm": 7.678478717803955, "learning_rate": 1.168e-05, "loss": 55.1562, "step": 1169 }, { "epoch": 0.000117, "grad_norm": 7.4621901512146, "learning_rate": 1.169e-05, "loss": 55.9375, "step": 1170 }, { "epoch": 0.0001171, "grad_norm": 7.68449068069458, "learning_rate": 1.17e-05, "loss": 55.2812, "step": 1171 }, { "epoch": 0.0001172, "grad_norm": 7.555746078491211, "learning_rate": 1.171e-05, "loss": 55.2188, "step": 1172 }, { "epoch": 0.0001173, "grad_norm": 7.522375583648682, "learning_rate": 1.1720000000000001e-05, "loss": 55.6562, "step": 1173 }, { "epoch": 0.0001174, "grad_norm": 7.526357650756836, "learning_rate": 1.1730000000000001e-05, "loss": 54.9688, "step": 1174 }, { "epoch": 0.0001175, "grad_norm": 7.813880443572998, "learning_rate": 1.174e-05, "loss": 54.5, "step": 1175 }, { "epoch": 0.0001176, "grad_norm": 7.758258819580078, "learning_rate": 1.175e-05, "loss": 54.5, "step": 1176 }, { "epoch": 0.0001177, "grad_norm": 7.648372173309326, "learning_rate": 1.176e-05, "loss": 54.9375, "step": 1177 }, { "epoch": 0.0001178, "grad_norm": 7.699697971343994, "learning_rate": 1.177e-05, "loss": 54.3125, "step": 1178 }, { "epoch": 0.0001179, "grad_norm": 7.569566249847412, "learning_rate": 1.178e-05, "loss": 55.1562, "step": 1179 }, { "epoch": 0.000118, "grad_norm": 7.436944484710693, "learning_rate": 1.179e-05, "loss": 55.3438, "step": 1180 }, { "epoch": 0.0001181, "grad_norm": 7.739448070526123, "learning_rate": 1.1799999999999999e-05, "loss": 54.2188, "step": 1181 }, { "epoch": 0.0001182, "grad_norm": 7.439329624176025, "learning_rate": 1.1809999999999999e-05, "loss": 55.8125, "step": 1182 }, { "epoch": 0.0001183, "grad_norm": 8.071717262268066, "learning_rate": 1.182e-05, "loss": 53.7812, "step": 1183 }, { "epoch": 0.0001184, "grad_norm": 7.629164218902588, "learning_rate": 1.183e-05, "loss": 54.25, "step": 1184 }, { "epoch": 0.0001185, "grad_norm": 7.748568534851074, "learning_rate": 1.184e-05, "loss": 54.2812, "step": 1185 }, { "epoch": 0.0001186, "grad_norm": 7.587509632110596, "learning_rate": 1.1850000000000002e-05, "loss": 54.5625, "step": 1186 }, { "epoch": 0.0001187, "grad_norm": 7.9441351890563965, "learning_rate": 1.186e-05, "loss": 53.9375, "step": 1187 }, { "epoch": 0.0001188, "grad_norm": 7.724300861358643, "learning_rate": 1.187e-05, "loss": 54.1562, "step": 1188 }, { "epoch": 0.0001189, "grad_norm": 7.837823867797852, "learning_rate": 1.188e-05, "loss": 53.5, "step": 1189 }, { "epoch": 0.000119, "grad_norm": 7.670633316040039, "learning_rate": 1.1890000000000001e-05, "loss": 53.9062, "step": 1190 }, { "epoch": 0.0001191, "grad_norm": 7.673746585845947, "learning_rate": 1.1900000000000001e-05, "loss": 54.0938, "step": 1191 }, { "epoch": 0.0001192, "grad_norm": 7.649930000305176, "learning_rate": 1.191e-05, "loss": 54.625, "step": 1192 }, { "epoch": 0.0001193, "grad_norm": 7.45327615737915, "learning_rate": 1.192e-05, "loss": 54.4375, "step": 1193 }, { "epoch": 0.0001194, "grad_norm": 7.5371856689453125, "learning_rate": 1.193e-05, "loss": 53.7188, "step": 1194 }, { "epoch": 0.0001195, "grad_norm": 7.465251922607422, "learning_rate": 1.1940000000000001e-05, "loss": 54.5, "step": 1195 }, { "epoch": 0.0001196, "grad_norm": 7.773534774780273, "learning_rate": 1.195e-05, "loss": 53.4375, "step": 1196 }, { "epoch": 0.0001197, "grad_norm": 7.400661945343018, "learning_rate": 1.196e-05, "loss": 54.25, "step": 1197 }, { "epoch": 0.0001198, "grad_norm": 7.420455455780029, "learning_rate": 1.1969999999999999e-05, "loss": 54.5938, "step": 1198 }, { "epoch": 0.0001199, "grad_norm": 7.698644161224365, "learning_rate": 1.1979999999999999e-05, "loss": 53.8125, "step": 1199 }, { "epoch": 0.00012, "grad_norm": 7.344296932220459, "learning_rate": 1.199e-05, "loss": 54.4688, "step": 1200 }, { "epoch": 0.0001201, "grad_norm": 7.562631130218506, "learning_rate": 1.2e-05, "loss": 53.8125, "step": 1201 }, { "epoch": 0.0001202, "grad_norm": 7.408926486968994, "learning_rate": 1.201e-05, "loss": 53.8125, "step": 1202 }, { "epoch": 0.0001203, "grad_norm": 7.729759216308594, "learning_rate": 1.2020000000000002e-05, "loss": 53.375, "step": 1203 }, { "epoch": 0.0001204, "grad_norm": 7.490784168243408, "learning_rate": 1.203e-05, "loss": 53.5938, "step": 1204 }, { "epoch": 0.0001205, "grad_norm": 7.493447780609131, "learning_rate": 1.204e-05, "loss": 53.7188, "step": 1205 }, { "epoch": 0.0001206, "grad_norm": 7.282375812530518, "learning_rate": 1.205e-05, "loss": 54.3438, "step": 1206 }, { "epoch": 0.0001207, "grad_norm": 7.591827392578125, "learning_rate": 1.2060000000000001e-05, "loss": 53.5, "step": 1207 }, { "epoch": 0.0001208, "grad_norm": 7.682456016540527, "learning_rate": 1.2070000000000001e-05, "loss": 53.7812, "step": 1208 }, { "epoch": 0.0001209, "grad_norm": 7.2463202476501465, "learning_rate": 1.208e-05, "loss": 53.8125, "step": 1209 }, { "epoch": 0.000121, "grad_norm": 7.508241653442383, "learning_rate": 1.209e-05, "loss": 53.3125, "step": 1210 }, { "epoch": 0.0001211, "grad_norm": 7.579042911529541, "learning_rate": 1.21e-05, "loss": 52.8125, "step": 1211 }, { "epoch": 0.0001212, "grad_norm": 7.6091084480285645, "learning_rate": 1.2110000000000001e-05, "loss": 53.1562, "step": 1212 }, { "epoch": 0.0001213, "grad_norm": 7.4466657638549805, "learning_rate": 1.2120000000000001e-05, "loss": 53.2812, "step": 1213 }, { "epoch": 0.0001214, "grad_norm": 7.076037406921387, "learning_rate": 1.213e-05, "loss": 54.3438, "step": 1214 }, { "epoch": 0.0001215, "grad_norm": 7.500237941741943, "learning_rate": 1.2139999999999999e-05, "loss": 53.0, "step": 1215 }, { "epoch": 0.0001216, "grad_norm": 7.2276997566223145, "learning_rate": 1.2149999999999999e-05, "loss": 53.8125, "step": 1216 }, { "epoch": 0.0001217, "grad_norm": 7.537385940551758, "learning_rate": 1.216e-05, "loss": 52.7812, "step": 1217 }, { "epoch": 0.0001218, "grad_norm": 7.664484024047852, "learning_rate": 1.217e-05, "loss": 52.7812, "step": 1218 }, { "epoch": 0.0001219, "grad_norm": 7.303295612335205, "learning_rate": 1.218e-05, "loss": 53.7188, "step": 1219 }, { "epoch": 0.000122, "grad_norm": 7.290215492248535, "learning_rate": 1.2190000000000002e-05, "loss": 53.7188, "step": 1220 }, { "epoch": 0.0001221, "grad_norm": 7.22444486618042, "learning_rate": 1.22e-05, "loss": 53.5, "step": 1221 }, { "epoch": 0.0001222, "grad_norm": 7.554375648498535, "learning_rate": 1.221e-05, "loss": 52.5625, "step": 1222 }, { "epoch": 0.0001223, "grad_norm": 7.6661787033081055, "learning_rate": 1.222e-05, "loss": 52.6562, "step": 1223 }, { "epoch": 0.0001224, "grad_norm": 7.28986120223999, "learning_rate": 1.2230000000000001e-05, "loss": 53.375, "step": 1224 }, { "epoch": 0.0001225, "grad_norm": 7.4713215827941895, "learning_rate": 1.2240000000000001e-05, "loss": 53.375, "step": 1225 }, { "epoch": 0.0001226, "grad_norm": 7.248449325561523, "learning_rate": 1.225e-05, "loss": 53.0625, "step": 1226 }, { "epoch": 0.0001227, "grad_norm": 7.284693717956543, "learning_rate": 1.226e-05, "loss": 53.3125, "step": 1227 }, { "epoch": 0.0001228, "grad_norm": 7.41956901550293, "learning_rate": 1.227e-05, "loss": 52.5625, "step": 1228 }, { "epoch": 0.0001229, "grad_norm": 7.39863920211792, "learning_rate": 1.2280000000000001e-05, "loss": 53.0312, "step": 1229 }, { "epoch": 0.000123, "grad_norm": 7.424099922180176, "learning_rate": 1.2290000000000001e-05, "loss": 53.25, "step": 1230 }, { "epoch": 0.0001231, "grad_norm": 7.63134241104126, "learning_rate": 1.23e-05, "loss": 52.3125, "step": 1231 }, { "epoch": 0.0001232, "grad_norm": 7.46926212310791, "learning_rate": 1.2309999999999999e-05, "loss": 52.9688, "step": 1232 }, { "epoch": 0.0001233, "grad_norm": 7.465059280395508, "learning_rate": 1.2319999999999999e-05, "loss": 52.0, "step": 1233 }, { "epoch": 0.0001234, "grad_norm": 7.630496501922607, "learning_rate": 1.233e-05, "loss": 52.125, "step": 1234 }, { "epoch": 0.0001235, "grad_norm": 7.242209434509277, "learning_rate": 1.234e-05, "loss": 52.9062, "step": 1235 }, { "epoch": 0.0001236, "grad_norm": 7.2926740646362305, "learning_rate": 1.235e-05, "loss": 53.6562, "step": 1236 }, { "epoch": 0.0001237, "grad_norm": 7.212370872497559, "learning_rate": 1.2360000000000002e-05, "loss": 52.8125, "step": 1237 }, { "epoch": 0.0001238, "grad_norm": 7.0620598793029785, "learning_rate": 1.237e-05, "loss": 53.2188, "step": 1238 }, { "epoch": 0.0001239, "grad_norm": 7.333443641662598, "learning_rate": 1.238e-05, "loss": 52.8125, "step": 1239 }, { "epoch": 0.000124, "grad_norm": 7.3336873054504395, "learning_rate": 1.239e-05, "loss": 52.2188, "step": 1240 }, { "epoch": 0.0001241, "grad_norm": 7.147743225097656, "learning_rate": 1.24e-05, "loss": 52.5, "step": 1241 }, { "epoch": 0.0001242, "grad_norm": 7.3426618576049805, "learning_rate": 1.2410000000000001e-05, "loss": 52.3438, "step": 1242 }, { "epoch": 0.0001243, "grad_norm": 7.250710487365723, "learning_rate": 1.242e-05, "loss": 52.1562, "step": 1243 }, { "epoch": 0.0001244, "grad_norm": 7.301087379455566, "learning_rate": 1.243e-05, "loss": 52.625, "step": 1244 }, { "epoch": 0.0001245, "grad_norm": 7.3164191246032715, "learning_rate": 1.244e-05, "loss": 52.0938, "step": 1245 }, { "epoch": 0.0001246, "grad_norm": 7.23098087310791, "learning_rate": 1.2450000000000001e-05, "loss": 52.7812, "step": 1246 }, { "epoch": 0.0001247, "grad_norm": 7.254575252532959, "learning_rate": 1.2460000000000001e-05, "loss": 52.25, "step": 1247 }, { "epoch": 0.0001248, "grad_norm": 7.3596625328063965, "learning_rate": 1.2470000000000001e-05, "loss": 51.7812, "step": 1248 }, { "epoch": 0.0001249, "grad_norm": 7.4117536544799805, "learning_rate": 1.2479999999999999e-05, "loss": 51.4688, "step": 1249 }, { "epoch": 0.000125, "grad_norm": 7.203137397766113, "learning_rate": 1.2489999999999999e-05, "loss": 52.1562, "step": 1250 }, { "epoch": 0.0001251, "grad_norm": 7.224808692932129, "learning_rate": 1.25e-05, "loss": 52.1562, "step": 1251 }, { "epoch": 0.0001252, "grad_norm": 7.343545913696289, "learning_rate": 1.251e-05, "loss": 51.75, "step": 1252 }, { "epoch": 0.0001253, "grad_norm": 7.2418365478515625, "learning_rate": 1.252e-05, "loss": 51.6562, "step": 1253 }, { "epoch": 0.0001254, "grad_norm": 7.007531642913818, "learning_rate": 1.2530000000000002e-05, "loss": 52.4375, "step": 1254 }, { "epoch": 0.0001255, "grad_norm": 6.968234539031982, "learning_rate": 1.2539999999999999e-05, "loss": 52.5312, "step": 1255 }, { "epoch": 0.0001256, "grad_norm": 7.113062381744385, "learning_rate": 1.255e-05, "loss": 52.375, "step": 1256 }, { "epoch": 0.0001257, "grad_norm": 7.2013702392578125, "learning_rate": 1.256e-05, "loss": 51.8125, "step": 1257 }, { "epoch": 0.0001258, "grad_norm": 7.480557441711426, "learning_rate": 1.257e-05, "loss": 51.9375, "step": 1258 }, { "epoch": 0.0001259, "grad_norm": 7.298722743988037, "learning_rate": 1.2580000000000002e-05, "loss": 51.3438, "step": 1259 }, { "epoch": 0.000126, "grad_norm": 7.417407035827637, "learning_rate": 1.259e-05, "loss": 50.875, "step": 1260 }, { "epoch": 0.0001261, "grad_norm": 7.077691078186035, "learning_rate": 1.26e-05, "loss": 51.6562, "step": 1261 }, { "epoch": 0.0001262, "grad_norm": 7.0805535316467285, "learning_rate": 1.261e-05, "loss": 51.75, "step": 1262 }, { "epoch": 0.0001263, "grad_norm": 6.998329162597656, "learning_rate": 1.2620000000000001e-05, "loss": 51.875, "step": 1263 }, { "epoch": 0.0001264, "grad_norm": 7.11593770980835, "learning_rate": 1.2630000000000001e-05, "loss": 51.6875, "step": 1264 }, { "epoch": 0.0001265, "grad_norm": 7.197646617889404, "learning_rate": 1.2640000000000001e-05, "loss": 51.3438, "step": 1265 }, { "epoch": 0.0001266, "grad_norm": 7.040606498718262, "learning_rate": 1.265e-05, "loss": 51.5312, "step": 1266 }, { "epoch": 0.0001267, "grad_norm": 7.308192253112793, "learning_rate": 1.2659999999999999e-05, "loss": 50.9375, "step": 1267 }, { "epoch": 0.0001268, "grad_norm": 7.215775012969971, "learning_rate": 1.267e-05, "loss": 51.2188, "step": 1268 }, { "epoch": 0.0001269, "grad_norm": 7.334770679473877, "learning_rate": 1.268e-05, "loss": 50.6875, "step": 1269 }, { "epoch": 0.000127, "grad_norm": 7.020657062530518, "learning_rate": 1.269e-05, "loss": 51.5938, "step": 1270 }, { "epoch": 0.0001271, "grad_norm": 7.122906684875488, "learning_rate": 1.2700000000000002e-05, "loss": 51.6875, "step": 1271 }, { "epoch": 0.0001272, "grad_norm": 6.957503795623779, "learning_rate": 1.2709999999999999e-05, "loss": 51.5312, "step": 1272 }, { "epoch": 0.0001273, "grad_norm": 7.044886589050293, "learning_rate": 1.272e-05, "loss": 51.0938, "step": 1273 }, { "epoch": 0.0001274, "grad_norm": 6.931931972503662, "learning_rate": 1.273e-05, "loss": 51.8125, "step": 1274 }, { "epoch": 0.0001275, "grad_norm": 7.079439640045166, "learning_rate": 1.274e-05, "loss": 51.375, "step": 1275 }, { "epoch": 0.0001276, "grad_norm": 7.171270847320557, "learning_rate": 1.2750000000000002e-05, "loss": 50.5625, "step": 1276 }, { "epoch": 0.0001277, "grad_norm": 7.095261573791504, "learning_rate": 1.276e-05, "loss": 51.5312, "step": 1277 }, { "epoch": 0.0001278, "grad_norm": 6.948635578155518, "learning_rate": 1.277e-05, "loss": 51.4062, "step": 1278 }, { "epoch": 0.0001279, "grad_norm": 7.283115863800049, "learning_rate": 1.278e-05, "loss": 50.2188, "step": 1279 }, { "epoch": 0.000128, "grad_norm": 7.020801544189453, "learning_rate": 1.2790000000000001e-05, "loss": 51.0938, "step": 1280 }, { "epoch": 0.0001281, "grad_norm": 6.878104209899902, "learning_rate": 1.2800000000000001e-05, "loss": 51.5, "step": 1281 }, { "epoch": 0.0001282, "grad_norm": 7.140067100524902, "learning_rate": 1.2810000000000001e-05, "loss": 50.6562, "step": 1282 }, { "epoch": 0.0001283, "grad_norm": 7.064198017120361, "learning_rate": 1.282e-05, "loss": 50.6562, "step": 1283 }, { "epoch": 0.0001284, "grad_norm": 7.170623302459717, "learning_rate": 1.2829999999999999e-05, "loss": 50.0625, "step": 1284 }, { "epoch": 0.0001285, "grad_norm": 6.972163200378418, "learning_rate": 1.284e-05, "loss": 51.0625, "step": 1285 }, { "epoch": 0.0001286, "grad_norm": 7.093153953552246, "learning_rate": 1.285e-05, "loss": 50.4062, "step": 1286 }, { "epoch": 0.0001287, "grad_norm": 6.831297874450684, "learning_rate": 1.286e-05, "loss": 51.8125, "step": 1287 }, { "epoch": 0.0001288, "grad_norm": 7.019623279571533, "learning_rate": 1.2870000000000002e-05, "loss": 50.8125, "step": 1288 }, { "epoch": 0.0001289, "grad_norm": 6.93576717376709, "learning_rate": 1.2879999999999999e-05, "loss": 50.625, "step": 1289 }, { "epoch": 0.000129, "grad_norm": 7.083858489990234, "learning_rate": 1.289e-05, "loss": 50.1875, "step": 1290 }, { "epoch": 0.0001291, "grad_norm": 7.067101955413818, "learning_rate": 1.29e-05, "loss": 50.0312, "step": 1291 }, { "epoch": 0.0001292, "grad_norm": 7.251895427703857, "learning_rate": 1.291e-05, "loss": 49.7188, "step": 1292 }, { "epoch": 0.0001293, "grad_norm": 6.953596115112305, "learning_rate": 1.2920000000000002e-05, "loss": 50.9062, "step": 1293 }, { "epoch": 0.0001294, "grad_norm": 6.8302788734436035, "learning_rate": 1.293e-05, "loss": 50.7812, "step": 1294 }, { "epoch": 0.0001295, "grad_norm": 6.787026405334473, "learning_rate": 1.294e-05, "loss": 50.8438, "step": 1295 }, { "epoch": 0.0001296, "grad_norm": 6.825897693634033, "learning_rate": 1.295e-05, "loss": 51.25, "step": 1296 }, { "epoch": 0.0001297, "grad_norm": 7.1078362464904785, "learning_rate": 1.2960000000000001e-05, "loss": 50.4062, "step": 1297 }, { "epoch": 0.0001298, "grad_norm": 6.9669599533081055, "learning_rate": 1.2970000000000001e-05, "loss": 50.125, "step": 1298 }, { "epoch": 0.0001299, "grad_norm": 6.832136154174805, "learning_rate": 1.2980000000000001e-05, "loss": 50.4375, "step": 1299 }, { "epoch": 0.00013, "grad_norm": 6.841429710388184, "learning_rate": 1.299e-05, "loss": 50.3125, "step": 1300 }, { "epoch": 0.0001301, "grad_norm": 6.736574649810791, "learning_rate": 1.3e-05, "loss": 51.0625, "step": 1301 }, { "epoch": 0.0001302, "grad_norm": 6.874002456665039, "learning_rate": 1.301e-05, "loss": 50.0938, "step": 1302 }, { "epoch": 0.0001303, "grad_norm": 6.825052738189697, "learning_rate": 1.302e-05, "loss": 50.3125, "step": 1303 }, { "epoch": 0.0001304, "grad_norm": 6.87752103805542, "learning_rate": 1.303e-05, "loss": 50.3125, "step": 1304 }, { "epoch": 0.0001305, "grad_norm": 6.828380107879639, "learning_rate": 1.3040000000000002e-05, "loss": 50.1562, "step": 1305 }, { "epoch": 0.0001306, "grad_norm": 6.835124969482422, "learning_rate": 1.3049999999999999e-05, "loss": 49.9688, "step": 1306 }, { "epoch": 0.0001307, "grad_norm": 7.016767978668213, "learning_rate": 1.306e-05, "loss": 50.0, "step": 1307 }, { "epoch": 0.0001308, "grad_norm": 6.881340503692627, "learning_rate": 1.307e-05, "loss": 49.9062, "step": 1308 }, { "epoch": 0.0001309, "grad_norm": 6.8892316818237305, "learning_rate": 1.308e-05, "loss": 49.8125, "step": 1309 }, { "epoch": 0.000131, "grad_norm": 6.788465976715088, "learning_rate": 1.3090000000000002e-05, "loss": 49.9062, "step": 1310 }, { "epoch": 0.0001311, "grad_norm": 6.9352264404296875, "learning_rate": 1.3099999999999998e-05, "loss": 49.7812, "step": 1311 }, { "epoch": 0.0001312, "grad_norm": 6.924383163452148, "learning_rate": 1.311e-05, "loss": 49.8438, "step": 1312 }, { "epoch": 0.0001313, "grad_norm": 6.814214706420898, "learning_rate": 1.312e-05, "loss": 50.0312, "step": 1313 }, { "epoch": 0.0001314, "grad_norm": 6.863121509552002, "learning_rate": 1.313e-05, "loss": 50.5938, "step": 1314 }, { "epoch": 0.0001315, "grad_norm": 6.9984354972839355, "learning_rate": 1.3140000000000001e-05, "loss": 49.2188, "step": 1315 }, { "epoch": 0.0001316, "grad_norm": 6.792233467102051, "learning_rate": 1.3150000000000001e-05, "loss": 49.6562, "step": 1316 }, { "epoch": 0.0001317, "grad_norm": 6.951003551483154, "learning_rate": 1.316e-05, "loss": 49.3438, "step": 1317 }, { "epoch": 0.0001318, "grad_norm": 6.749107837677002, "learning_rate": 1.317e-05, "loss": 49.5312, "step": 1318 }, { "epoch": 0.0001319, "grad_norm": 6.893924713134766, "learning_rate": 1.3180000000000001e-05, "loss": 49.8438, "step": 1319 }, { "epoch": 0.000132, "grad_norm": 6.869941234588623, "learning_rate": 1.319e-05, "loss": 48.9688, "step": 1320 }, { "epoch": 0.0001321, "grad_norm": 7.000456809997559, "learning_rate": 1.32e-05, "loss": 49.5312, "step": 1321 }, { "epoch": 0.0001322, "grad_norm": 6.8329386711120605, "learning_rate": 1.3210000000000002e-05, "loss": 49.2812, "step": 1322 }, { "epoch": 0.0001323, "grad_norm": 6.7939043045043945, "learning_rate": 1.3219999999999999e-05, "loss": 49.6562, "step": 1323 }, { "epoch": 0.0001324, "grad_norm": 6.9336137771606445, "learning_rate": 1.323e-05, "loss": 49.4688, "step": 1324 }, { "epoch": 0.0001325, "grad_norm": 6.958576202392578, "learning_rate": 1.324e-05, "loss": 49.1875, "step": 1325 }, { "epoch": 0.0001326, "grad_norm": 6.707515716552734, "learning_rate": 1.325e-05, "loss": 50.0312, "step": 1326 }, { "epoch": 0.0001327, "grad_norm": 6.904212474822998, "learning_rate": 1.3260000000000002e-05, "loss": 48.9062, "step": 1327 }, { "epoch": 0.0001328, "grad_norm": 6.822237491607666, "learning_rate": 1.3269999999999998e-05, "loss": 48.9375, "step": 1328 }, { "epoch": 0.0001329, "grad_norm": 6.981447696685791, "learning_rate": 1.328e-05, "loss": 49.0625, "step": 1329 }, { "epoch": 0.000133, "grad_norm": 6.96505880355835, "learning_rate": 1.329e-05, "loss": 48.9062, "step": 1330 }, { "epoch": 0.0001331, "grad_norm": 6.847214221954346, "learning_rate": 1.33e-05, "loss": 49.375, "step": 1331 }, { "epoch": 0.0001332, "grad_norm": 6.874561786651611, "learning_rate": 1.3310000000000001e-05, "loss": 48.9062, "step": 1332 }, { "epoch": 0.0001333, "grad_norm": 6.889916896820068, "learning_rate": 1.3320000000000001e-05, "loss": 48.5625, "step": 1333 }, { "epoch": 0.0001334, "grad_norm": 6.986288070678711, "learning_rate": 1.333e-05, "loss": 48.9688, "step": 1334 }, { "epoch": 0.0001335, "grad_norm": 7.319582462310791, "learning_rate": 1.334e-05, "loss": 49.3125, "step": 1335 }, { "epoch": 0.0001336, "grad_norm": 6.805470943450928, "learning_rate": 1.3350000000000001e-05, "loss": 48.6562, "step": 1336 }, { "epoch": 0.0001337, "grad_norm": 6.76211404800415, "learning_rate": 1.336e-05, "loss": 49.0625, "step": 1337 }, { "epoch": 0.0001338, "grad_norm": 6.805838108062744, "learning_rate": 1.337e-05, "loss": 48.8438, "step": 1338 }, { "epoch": 0.0001339, "grad_norm": 6.896024227142334, "learning_rate": 1.3380000000000002e-05, "loss": 48.6562, "step": 1339 }, { "epoch": 0.000134, "grad_norm": 6.7913737297058105, "learning_rate": 1.3389999999999999e-05, "loss": 48.25, "step": 1340 }, { "epoch": 0.0001341, "grad_norm": 6.825916767120361, "learning_rate": 1.34e-05, "loss": 48.6875, "step": 1341 }, { "epoch": 0.0001342, "grad_norm": 6.727726459503174, "learning_rate": 1.341e-05, "loss": 49.0625, "step": 1342 }, { "epoch": 0.0001343, "grad_norm": 6.678559303283691, "learning_rate": 1.342e-05, "loss": 48.6562, "step": 1343 }, { "epoch": 0.0001344, "grad_norm": 6.7459564208984375, "learning_rate": 1.3430000000000002e-05, "loss": 49.3438, "step": 1344 }, { "epoch": 0.0001345, "grad_norm": 6.67067289352417, "learning_rate": 1.3439999999999998e-05, "loss": 49.3438, "step": 1345 }, { "epoch": 0.0001346, "grad_norm": 6.752988338470459, "learning_rate": 1.345e-05, "loss": 48.5312, "step": 1346 }, { "epoch": 0.0001347, "grad_norm": 6.771851539611816, "learning_rate": 1.346e-05, "loss": 48.9375, "step": 1347 }, { "epoch": 0.0001348, "grad_norm": 6.481204032897949, "learning_rate": 1.347e-05, "loss": 49.5938, "step": 1348 }, { "epoch": 0.0001349, "grad_norm": 6.8165388107299805, "learning_rate": 1.3480000000000001e-05, "loss": 48.875, "step": 1349 }, { "epoch": 0.000135, "grad_norm": 6.703948497772217, "learning_rate": 1.3490000000000001e-05, "loss": 48.1875, "step": 1350 }, { "epoch": 0.0001351, "grad_norm": 6.681395053863525, "learning_rate": 1.35e-05, "loss": 48.125, "step": 1351 }, { "epoch": 0.0001352, "grad_norm": 6.673883438110352, "learning_rate": 1.351e-05, "loss": 48.3438, "step": 1352 }, { "epoch": 0.0001353, "grad_norm": 6.701688766479492, "learning_rate": 1.3520000000000001e-05, "loss": 48.9688, "step": 1353 }, { "epoch": 0.0001354, "grad_norm": 6.5370025634765625, "learning_rate": 1.3530000000000001e-05, "loss": 48.5312, "step": 1354 }, { "epoch": 0.0001355, "grad_norm": 6.6904706954956055, "learning_rate": 1.354e-05, "loss": 47.7188, "step": 1355 }, { "epoch": 0.0001356, "grad_norm": 6.671186923980713, "learning_rate": 1.3550000000000002e-05, "loss": 48.4062, "step": 1356 }, { "epoch": 0.0001357, "grad_norm": 6.853418827056885, "learning_rate": 1.3559999999999999e-05, "loss": 49.3438, "step": 1357 }, { "epoch": 0.0001358, "grad_norm": 6.607544422149658, "learning_rate": 1.357e-05, "loss": 49.0312, "step": 1358 }, { "epoch": 0.0001359, "grad_norm": 6.757382869720459, "learning_rate": 1.358e-05, "loss": 48.125, "step": 1359 }, { "epoch": 0.000136, "grad_norm": 6.590203285217285, "learning_rate": 1.359e-05, "loss": 48.2812, "step": 1360 }, { "epoch": 0.0001361, "grad_norm": 6.539251804351807, "learning_rate": 1.3600000000000002e-05, "loss": 48.0938, "step": 1361 }, { "epoch": 0.0001362, "grad_norm": 6.598182201385498, "learning_rate": 1.3609999999999998e-05, "loss": 48.125, "step": 1362 }, { "epoch": 0.0001363, "grad_norm": 6.8250226974487305, "learning_rate": 1.362e-05, "loss": 47.6562, "step": 1363 }, { "epoch": 0.0001364, "grad_norm": 6.718772888183594, "learning_rate": 1.363e-05, "loss": 48.1875, "step": 1364 }, { "epoch": 0.0001365, "grad_norm": 6.544684886932373, "learning_rate": 1.364e-05, "loss": 48.0312, "step": 1365 }, { "epoch": 0.0001366, "grad_norm": 6.549186706542969, "learning_rate": 1.3650000000000001e-05, "loss": 47.9375, "step": 1366 }, { "epoch": 0.0001367, "grad_norm": 6.598300457000732, "learning_rate": 1.3660000000000001e-05, "loss": 48.25, "step": 1367 }, { "epoch": 0.0001368, "grad_norm": 6.672146797180176, "learning_rate": 1.367e-05, "loss": 47.125, "step": 1368 }, { "epoch": 0.0001369, "grad_norm": 6.587193012237549, "learning_rate": 1.368e-05, "loss": 48.1875, "step": 1369 }, { "epoch": 0.000137, "grad_norm": 6.673519611358643, "learning_rate": 1.369e-05, "loss": 47.5, "step": 1370 }, { "epoch": 0.0001371, "grad_norm": 6.656841278076172, "learning_rate": 1.3700000000000001e-05, "loss": 49.6875, "step": 1371 }, { "epoch": 0.0001372, "grad_norm": 6.472851753234863, "learning_rate": 1.3710000000000001e-05, "loss": 48.3438, "step": 1372 }, { "epoch": 0.0001373, "grad_norm": 6.7288970947265625, "learning_rate": 1.372e-05, "loss": 47.9688, "step": 1373 }, { "epoch": 0.0001374, "grad_norm": 6.470300674438477, "learning_rate": 1.3729999999999999e-05, "loss": 48.9062, "step": 1374 }, { "epoch": 0.0001375, "grad_norm": 6.735988140106201, "learning_rate": 1.374e-05, "loss": 47.75, "step": 1375 }, { "epoch": 0.0001376, "grad_norm": 6.431521892547607, "learning_rate": 1.375e-05, "loss": 48.3438, "step": 1376 }, { "epoch": 0.0001377, "grad_norm": 6.473598957061768, "learning_rate": 1.376e-05, "loss": 48.3438, "step": 1377 }, { "epoch": 0.0001378, "grad_norm": 6.585727214813232, "learning_rate": 1.3770000000000002e-05, "loss": 47.4375, "step": 1378 }, { "epoch": 0.0001379, "grad_norm": 6.473487854003906, "learning_rate": 1.3779999999999999e-05, "loss": 47.6875, "step": 1379 }, { "epoch": 0.000138, "grad_norm": 6.589076519012451, "learning_rate": 1.379e-05, "loss": 47.0, "step": 1380 }, { "epoch": 0.0001381, "grad_norm": 6.590631484985352, "learning_rate": 1.38e-05, "loss": 47.6875, "step": 1381 }, { "epoch": 0.0001382, "grad_norm": 6.402490139007568, "learning_rate": 1.381e-05, "loss": 48.125, "step": 1382 }, { "epoch": 0.0001383, "grad_norm": 6.432614326477051, "learning_rate": 1.3820000000000002e-05, "loss": 47.4688, "step": 1383 }, { "epoch": 0.0001384, "grad_norm": 6.533024311065674, "learning_rate": 1.3830000000000001e-05, "loss": 47.75, "step": 1384 }, { "epoch": 0.0001385, "grad_norm": 6.51165246963501, "learning_rate": 1.384e-05, "loss": 47.625, "step": 1385 }, { "epoch": 0.0001386, "grad_norm": 6.475950717926025, "learning_rate": 1.385e-05, "loss": 47.6562, "step": 1386 }, { "epoch": 0.0001387, "grad_norm": 6.405846118927002, "learning_rate": 1.386e-05, "loss": 48.1562, "step": 1387 }, { "epoch": 0.0001388, "grad_norm": 6.439004898071289, "learning_rate": 1.3870000000000001e-05, "loss": 47.8438, "step": 1388 }, { "epoch": 0.0001389, "grad_norm": 6.551534175872803, "learning_rate": 1.3880000000000001e-05, "loss": 47.125, "step": 1389 }, { "epoch": 0.000139, "grad_norm": 6.39193868637085, "learning_rate": 1.3889999999999999e-05, "loss": 47.4375, "step": 1390 }, { "epoch": 0.0001391, "grad_norm": 6.597825050354004, "learning_rate": 1.3899999999999999e-05, "loss": 46.6562, "step": 1391 }, { "epoch": 0.0001392, "grad_norm": 6.471019268035889, "learning_rate": 1.391e-05, "loss": 47.5625, "step": 1392 }, { "epoch": 0.0001393, "grad_norm": 6.471146106719971, "learning_rate": 1.392e-05, "loss": 47.0, "step": 1393 }, { "epoch": 0.0001394, "grad_norm": 6.441690921783447, "learning_rate": 1.393e-05, "loss": 47.6562, "step": 1394 }, { "epoch": 0.0001395, "grad_norm": 6.480464935302734, "learning_rate": 1.3940000000000002e-05, "loss": 47.625, "step": 1395 }, { "epoch": 0.0001396, "grad_norm": 6.39990758895874, "learning_rate": 1.3949999999999999e-05, "loss": 47.5312, "step": 1396 }, { "epoch": 0.0001397, "grad_norm": 6.367557048797607, "learning_rate": 1.396e-05, "loss": 47.4688, "step": 1397 }, { "epoch": 0.0001398, "grad_norm": 6.547006607055664, "learning_rate": 1.397e-05, "loss": 46.4375, "step": 1398 }, { "epoch": 0.0001399, "grad_norm": 6.42279052734375, "learning_rate": 1.398e-05, "loss": 48.0312, "step": 1399 }, { "epoch": 0.00014, "grad_norm": 6.570935249328613, "learning_rate": 1.3990000000000002e-05, "loss": 47.4375, "step": 1400 }, { "epoch": 0.0001401, "grad_norm": 6.599931240081787, "learning_rate": 1.4000000000000001e-05, "loss": 47.0312, "step": 1401 }, { "epoch": 0.0001402, "grad_norm": 6.265009880065918, "learning_rate": 1.401e-05, "loss": 47.1875, "step": 1402 }, { "epoch": 0.0001403, "grad_norm": 6.438925266265869, "learning_rate": 1.402e-05, "loss": 46.6875, "step": 1403 }, { "epoch": 0.0001404, "grad_norm": 6.475961208343506, "learning_rate": 1.403e-05, "loss": 46.9062, "step": 1404 }, { "epoch": 0.0001405, "grad_norm": 6.573430061340332, "learning_rate": 1.4040000000000001e-05, "loss": 46.3125, "step": 1405 }, { "epoch": 0.0001406, "grad_norm": 6.558287143707275, "learning_rate": 1.4050000000000001e-05, "loss": 46.4062, "step": 1406 }, { "epoch": 0.0001407, "grad_norm": 6.389150619506836, "learning_rate": 1.406e-05, "loss": 47.1562, "step": 1407 }, { "epoch": 0.0001408, "grad_norm": 6.322193145751953, "learning_rate": 1.4069999999999999e-05, "loss": 48.1562, "step": 1408 }, { "epoch": 0.0001409, "grad_norm": 6.224388122558594, "learning_rate": 1.408e-05, "loss": 47.625, "step": 1409 }, { "epoch": 0.000141, "grad_norm": 6.360954761505127, "learning_rate": 1.409e-05, "loss": 47.25, "step": 1410 }, { "epoch": 0.0001411, "grad_norm": 6.193026542663574, "learning_rate": 1.41e-05, "loss": 47.9375, "step": 1411 }, { "epoch": 0.0001412, "grad_norm": 6.543434143066406, "learning_rate": 1.4110000000000002e-05, "loss": 47.7188, "step": 1412 }, { "epoch": 0.0001413, "grad_norm": 6.402957916259766, "learning_rate": 1.4119999999999999e-05, "loss": 47.0625, "step": 1413 }, { "epoch": 0.0001414, "grad_norm": 6.295007705688477, "learning_rate": 1.413e-05, "loss": 47.3125, "step": 1414 }, { "epoch": 0.0001415, "grad_norm": 6.175797462463379, "learning_rate": 1.414e-05, "loss": 48.0625, "step": 1415 }, { "epoch": 0.0001416, "grad_norm": 6.613070011138916, "learning_rate": 1.415e-05, "loss": 45.8125, "step": 1416 }, { "epoch": 0.0001417, "grad_norm": 6.471121311187744, "learning_rate": 1.4160000000000002e-05, "loss": 46.625, "step": 1417 }, { "epoch": 0.0001418, "grad_norm": 6.47176456451416, "learning_rate": 1.4170000000000002e-05, "loss": 46.3438, "step": 1418 }, { "epoch": 0.0001419, "grad_norm": 6.427188396453857, "learning_rate": 1.418e-05, "loss": 46.125, "step": 1419 }, { "epoch": 0.000142, "grad_norm": 6.524816989898682, "learning_rate": 1.419e-05, "loss": 45.9375, "step": 1420 }, { "epoch": 0.0001421, "grad_norm": 6.358154296875, "learning_rate": 1.42e-05, "loss": 47.2812, "step": 1421 }, { "epoch": 0.0001422, "grad_norm": 6.3673624992370605, "learning_rate": 1.4210000000000001e-05, "loss": 46.4375, "step": 1422 }, { "epoch": 0.0001423, "grad_norm": 6.6436357498168945, "learning_rate": 1.4220000000000001e-05, "loss": 46.5938, "step": 1423 }, { "epoch": 0.0001424, "grad_norm": 6.492889404296875, "learning_rate": 1.423e-05, "loss": 46.4062, "step": 1424 }, { "epoch": 0.0001425, "grad_norm": 6.439519882202148, "learning_rate": 1.424e-05, "loss": 46.3125, "step": 1425 }, { "epoch": 0.0001426, "grad_norm": 6.359393119812012, "learning_rate": 1.425e-05, "loss": 46.4688, "step": 1426 }, { "epoch": 0.0001427, "grad_norm": 6.562929153442383, "learning_rate": 1.426e-05, "loss": 46.0312, "step": 1427 }, { "epoch": 0.0001428, "grad_norm": 6.442821025848389, "learning_rate": 1.427e-05, "loss": 45.4375, "step": 1428 }, { "epoch": 0.0001429, "grad_norm": 6.317464351654053, "learning_rate": 1.428e-05, "loss": 46.0625, "step": 1429 }, { "epoch": 0.000143, "grad_norm": 6.314460754394531, "learning_rate": 1.4289999999999999e-05, "loss": 45.9688, "step": 1430 }, { "epoch": 0.0001431, "grad_norm": 6.241281032562256, "learning_rate": 1.43e-05, "loss": 46.4688, "step": 1431 }, { "epoch": 0.0001432, "grad_norm": 6.5193963050842285, "learning_rate": 1.431e-05, "loss": 45.625, "step": 1432 }, { "epoch": 0.0001433, "grad_norm": 6.031685829162598, "learning_rate": 1.432e-05, "loss": 47.7188, "step": 1433 }, { "epoch": 0.0001434, "grad_norm": 6.39438533782959, "learning_rate": 1.4330000000000002e-05, "loss": 45.5312, "step": 1434 }, { "epoch": 0.0001435, "grad_norm": 6.2862324714660645, "learning_rate": 1.4340000000000002e-05, "loss": 46.125, "step": 1435 }, { "epoch": 0.0001436, "grad_norm": 6.472970485687256, "learning_rate": 1.435e-05, "loss": 45.75, "step": 1436 }, { "epoch": 0.0001437, "grad_norm": 6.28604793548584, "learning_rate": 1.436e-05, "loss": 46.0, "step": 1437 }, { "epoch": 0.0001438, "grad_norm": 6.293455600738525, "learning_rate": 1.437e-05, "loss": 47.375, "step": 1438 }, { "epoch": 0.0001439, "grad_norm": 6.2602057456970215, "learning_rate": 1.4380000000000001e-05, "loss": 46.25, "step": 1439 }, { "epoch": 0.000144, "grad_norm": 6.504815101623535, "learning_rate": 1.4390000000000001e-05, "loss": 45.5938, "step": 1440 }, { "epoch": 0.0001441, "grad_norm": 6.3342366218566895, "learning_rate": 1.44e-05, "loss": 45.5, "step": 1441 }, { "epoch": 0.0001442, "grad_norm": 6.301247596740723, "learning_rate": 1.441e-05, "loss": 45.9688, "step": 1442 }, { "epoch": 0.0001443, "grad_norm": 6.273526191711426, "learning_rate": 1.442e-05, "loss": 46.0312, "step": 1443 }, { "epoch": 0.0001444, "grad_norm": 6.2740678787231445, "learning_rate": 1.443e-05, "loss": 47.0, "step": 1444 }, { "epoch": 0.0001445, "grad_norm": 6.325813293457031, "learning_rate": 1.444e-05, "loss": 45.0938, "step": 1445 }, { "epoch": 0.0001446, "grad_norm": 6.318421363830566, "learning_rate": 1.445e-05, "loss": 46.4688, "step": 1446 }, { "epoch": 0.0001447, "grad_norm": 6.3151445388793945, "learning_rate": 1.4459999999999999e-05, "loss": 45.75, "step": 1447 }, { "epoch": 0.0001448, "grad_norm": 6.330325126647949, "learning_rate": 1.447e-05, "loss": 45.7188, "step": 1448 }, { "epoch": 0.0001449, "grad_norm": 6.304328918457031, "learning_rate": 1.448e-05, "loss": 46.5938, "step": 1449 }, { "epoch": 0.000145, "grad_norm": 6.322455883026123, "learning_rate": 1.449e-05, "loss": 45.7812, "step": 1450 }, { "epoch": 0.0001451, "grad_norm": 6.178551197052002, "learning_rate": 1.4500000000000002e-05, "loss": 46.4062, "step": 1451 }, { "epoch": 0.0001452, "grad_norm": 6.315935134887695, "learning_rate": 1.4510000000000002e-05, "loss": 45.2812, "step": 1452 }, { "epoch": 0.0001453, "grad_norm": 6.263783931732178, "learning_rate": 1.452e-05, "loss": 45.0312, "step": 1453 }, { "epoch": 0.0001454, "grad_norm": 6.470102787017822, "learning_rate": 1.453e-05, "loss": 45.1562, "step": 1454 }, { "epoch": 0.0001455, "grad_norm": 6.220240592956543, "learning_rate": 1.454e-05, "loss": 45.8438, "step": 1455 }, { "epoch": 0.0001456, "grad_norm": 6.216402530670166, "learning_rate": 1.4550000000000001e-05, "loss": 45.6875, "step": 1456 }, { "epoch": 0.0001457, "grad_norm": 6.349188327789307, "learning_rate": 1.4560000000000001e-05, "loss": 44.9375, "step": 1457 }, { "epoch": 0.0001458, "grad_norm": 6.409142017364502, "learning_rate": 1.457e-05, "loss": 44.7188, "step": 1458 }, { "epoch": 0.0001459, "grad_norm": 6.316352367401123, "learning_rate": 1.458e-05, "loss": 45.5625, "step": 1459 }, { "epoch": 0.000146, "grad_norm": 6.413899898529053, "learning_rate": 1.459e-05, "loss": 44.75, "step": 1460 }, { "epoch": 0.0001461, "grad_norm": 6.37227201461792, "learning_rate": 1.46e-05, "loss": 44.9062, "step": 1461 }, { "epoch": 0.0001462, "grad_norm": 6.28816032409668, "learning_rate": 1.461e-05, "loss": 45.0, "step": 1462 }, { "epoch": 0.0001463, "grad_norm": 6.313782215118408, "learning_rate": 1.462e-05, "loss": 44.9062, "step": 1463 }, { "epoch": 0.0001464, "grad_norm": 6.226417064666748, "learning_rate": 1.4629999999999999e-05, "loss": 45.5, "step": 1464 }, { "epoch": 0.0001465, "grad_norm": 6.16673469543457, "learning_rate": 1.464e-05, "loss": 46.0938, "step": 1465 }, { "epoch": 0.0001466, "grad_norm": 6.2313361167907715, "learning_rate": 1.465e-05, "loss": 45.375, "step": 1466 }, { "epoch": 0.0001467, "grad_norm": 6.184562683105469, "learning_rate": 1.466e-05, "loss": 45.0938, "step": 1467 }, { "epoch": 0.0001468, "grad_norm": 6.402483940124512, "learning_rate": 1.4670000000000002e-05, "loss": 44.6875, "step": 1468 }, { "epoch": 0.0001469, "grad_norm": 6.2219624519348145, "learning_rate": 1.4680000000000002e-05, "loss": 45.875, "step": 1469 }, { "epoch": 0.000147, "grad_norm": 6.251800537109375, "learning_rate": 1.469e-05, "loss": 44.7188, "step": 1470 }, { "epoch": 0.0001471, "grad_norm": 6.082367897033691, "learning_rate": 1.47e-05, "loss": 45.2812, "step": 1471 }, { "epoch": 0.0001472, "grad_norm": 6.209146499633789, "learning_rate": 1.471e-05, "loss": 45.2812, "step": 1472 }, { "epoch": 0.0001473, "grad_norm": 6.131628036499023, "learning_rate": 1.4720000000000001e-05, "loss": 45.6562, "step": 1473 }, { "epoch": 0.0001474, "grad_norm": 6.09556245803833, "learning_rate": 1.4730000000000001e-05, "loss": 46.0, "step": 1474 }, { "epoch": 0.0001475, "grad_norm": 6.235780239105225, "learning_rate": 1.474e-05, "loss": 44.3125, "step": 1475 }, { "epoch": 0.0001476, "grad_norm": 6.3033037185668945, "learning_rate": 1.475e-05, "loss": 44.9062, "step": 1476 }, { "epoch": 0.0001477, "grad_norm": 6.262908458709717, "learning_rate": 1.476e-05, "loss": 44.7812, "step": 1477 }, { "epoch": 0.0001478, "grad_norm": 6.337582588195801, "learning_rate": 1.4770000000000001e-05, "loss": 44.25, "step": 1478 }, { "epoch": 0.0001479, "grad_norm": 6.241846561431885, "learning_rate": 1.478e-05, "loss": 44.875, "step": 1479 }, { "epoch": 0.000148, "grad_norm": 6.219179153442383, "learning_rate": 1.479e-05, "loss": 44.8125, "step": 1480 }, { "epoch": 0.0001481, "grad_norm": 6.218071937561035, "learning_rate": 1.4799999999999999e-05, "loss": 45.2188, "step": 1481 }, { "epoch": 0.0001482, "grad_norm": 6.311007022857666, "learning_rate": 1.481e-05, "loss": 44.625, "step": 1482 }, { "epoch": 0.0001483, "grad_norm": 6.308942794799805, "learning_rate": 1.482e-05, "loss": 44.6875, "step": 1483 }, { "epoch": 0.0001484, "grad_norm": 6.2515339851379395, "learning_rate": 1.483e-05, "loss": 44.1875, "step": 1484 }, { "epoch": 0.0001485, "grad_norm": 6.152646541595459, "learning_rate": 1.4840000000000002e-05, "loss": 44.9062, "step": 1485 }, { "epoch": 0.0001486, "grad_norm": 6.309966087341309, "learning_rate": 1.4850000000000002e-05, "loss": 44.3125, "step": 1486 }, { "epoch": 0.0001487, "grad_norm": 6.248706340789795, "learning_rate": 1.486e-05, "loss": 44.4062, "step": 1487 }, { "epoch": 0.0001488, "grad_norm": 6.111617565155029, "learning_rate": 1.487e-05, "loss": 44.875, "step": 1488 }, { "epoch": 0.0001489, "grad_norm": 6.138513565063477, "learning_rate": 1.488e-05, "loss": 45.375, "step": 1489 }, { "epoch": 0.000149, "grad_norm": 6.269906520843506, "learning_rate": 1.4890000000000001e-05, "loss": 44.3125, "step": 1490 }, { "epoch": 0.0001491, "grad_norm": 6.300713062286377, "learning_rate": 1.4900000000000001e-05, "loss": 44.3125, "step": 1491 }, { "epoch": 0.0001492, "grad_norm": 6.14302921295166, "learning_rate": 1.491e-05, "loss": 45.0, "step": 1492 }, { "epoch": 0.0001493, "grad_norm": 6.299230575561523, "learning_rate": 1.492e-05, "loss": 44.625, "step": 1493 }, { "epoch": 0.0001494, "grad_norm": 6.294898509979248, "learning_rate": 1.493e-05, "loss": 43.7188, "step": 1494 }, { "epoch": 0.0001495, "grad_norm": 6.1497063636779785, "learning_rate": 1.4940000000000001e-05, "loss": 45.25, "step": 1495 }, { "epoch": 0.0001496, "grad_norm": 6.63415002822876, "learning_rate": 1.4950000000000001e-05, "loss": 44.8438, "step": 1496 }, { "epoch": 0.0001497, "grad_norm": 6.380233287811279, "learning_rate": 1.496e-05, "loss": 43.7812, "step": 1497 }, { "epoch": 0.0001498, "grad_norm": 6.056626319885254, "learning_rate": 1.4969999999999999e-05, "loss": 44.8125, "step": 1498 }, { "epoch": 0.0001499, "grad_norm": 6.122725486755371, "learning_rate": 1.4979999999999999e-05, "loss": 44.9062, "step": 1499 }, { "epoch": 0.00015, "grad_norm": 6.298846244812012, "learning_rate": 1.499e-05, "loss": 43.6875, "step": 1500 }, { "epoch": 0.00015, "eval_loss": 5.494104862213135, "eval_runtime": 364.1692, "eval_samples_per_second": 27.46, "eval_steps_per_second": 1.716, "step": 1500 }, { "epoch": 0.0001501, "grad_norm": 11.855791091918945, "learning_rate": 1.5e-05, "loss": 89.4375, "step": 1501 }, { "epoch": 0.0001502, "grad_norm": 11.970094680786133, "learning_rate": 1.501e-05, "loss": 88.9375, "step": 1502 }, { "epoch": 0.0001503, "grad_norm": 11.825295448303223, "learning_rate": 1.5020000000000002e-05, "loss": 89.9375, "step": 1503 }, { "epoch": 0.0001504, "grad_norm": 12.03028392791748, "learning_rate": 1.503e-05, "loss": 87.875, "step": 1504 }, { "epoch": 0.0001505, "grad_norm": 11.944097518920898, "learning_rate": 1.504e-05, "loss": 89.3125, "step": 1505 }, { "epoch": 0.0001506, "grad_norm": 11.77283763885498, "learning_rate": 1.505e-05, "loss": 89.125, "step": 1506 }, { "epoch": 0.0001507, "grad_norm": 11.934531211853027, "learning_rate": 1.5060000000000001e-05, "loss": 87.5, "step": 1507 }, { "epoch": 0.0001508, "grad_norm": 11.855193138122559, "learning_rate": 1.5070000000000001e-05, "loss": 88.375, "step": 1508 }, { "epoch": 0.0001509, "grad_norm": 11.6827392578125, "learning_rate": 1.508e-05, "loss": 89.3125, "step": 1509 }, { "epoch": 0.000151, "grad_norm": 11.954957008361816, "learning_rate": 1.509e-05, "loss": 87.75, "step": 1510 }, { "epoch": 0.0001511, "grad_norm": 11.718145370483398, "learning_rate": 1.51e-05, "loss": 88.0625, "step": 1511 }, { "epoch": 0.0001512, "grad_norm": 11.530472755432129, "learning_rate": 1.5110000000000001e-05, "loss": 89.25, "step": 1512 }, { "epoch": 0.0001513, "grad_norm": 11.798002243041992, "learning_rate": 1.5120000000000001e-05, "loss": 88.4375, "step": 1513 }, { "epoch": 0.0001514, "grad_norm": 11.953389167785645, "learning_rate": 1.513e-05, "loss": 87.875, "step": 1514 }, { "epoch": 0.0001515, "grad_norm": 11.680356979370117, "learning_rate": 1.5139999999999999e-05, "loss": 88.8125, "step": 1515 }, { "epoch": 0.0001516, "grad_norm": 11.720935821533203, "learning_rate": 1.5149999999999999e-05, "loss": 88.6875, "step": 1516 }, { "epoch": 0.0001517, "grad_norm": 11.798686027526855, "learning_rate": 1.516e-05, "loss": 88.9375, "step": 1517 }, { "epoch": 0.0001518, "grad_norm": 11.694768905639648, "learning_rate": 1.517e-05, "loss": 87.625, "step": 1518 }, { "epoch": 0.0001519, "grad_norm": 11.854416847229004, "learning_rate": 1.518e-05, "loss": 87.5, "step": 1519 }, { "epoch": 0.000152, "grad_norm": 11.804166793823242, "learning_rate": 1.5190000000000002e-05, "loss": 87.5, "step": 1520 }, { "epoch": 0.0001521, "grad_norm": 11.772957801818848, "learning_rate": 1.52e-05, "loss": 88.25, "step": 1521 }, { "epoch": 0.0001522, "grad_norm": 11.960450172424316, "learning_rate": 1.521e-05, "loss": 87.0625, "step": 1522 }, { "epoch": 0.0001523, "grad_norm": 11.786050796508789, "learning_rate": 1.522e-05, "loss": 87.3125, "step": 1523 }, { "epoch": 0.0001524, "grad_norm": 12.02585506439209, "learning_rate": 1.5230000000000002e-05, "loss": 86.0, "step": 1524 }, { "epoch": 0.0001525, "grad_norm": 11.747179985046387, "learning_rate": 1.5240000000000001e-05, "loss": 87.9375, "step": 1525 }, { "epoch": 0.0001526, "grad_norm": 11.714150428771973, "learning_rate": 1.525e-05, "loss": 87.25, "step": 1526 }, { "epoch": 0.0001527, "grad_norm": 11.772976875305176, "learning_rate": 1.526e-05, "loss": 86.5625, "step": 1527 }, { "epoch": 0.0001528, "grad_norm": 11.734100341796875, "learning_rate": 1.527e-05, "loss": 87.125, "step": 1528 }, { "epoch": 0.0001529, "grad_norm": 11.775686264038086, "learning_rate": 1.528e-05, "loss": 87.4375, "step": 1529 }, { "epoch": 0.000153, "grad_norm": 11.796177864074707, "learning_rate": 1.529e-05, "loss": 87.625, "step": 1530 }, { "epoch": 0.0001531, "grad_norm": 11.95287799835205, "learning_rate": 1.5300000000000003e-05, "loss": 86.125, "step": 1531 }, { "epoch": 0.0001532, "grad_norm": 11.887791633605957, "learning_rate": 1.531e-05, "loss": 87.0625, "step": 1532 }, { "epoch": 0.0001533, "grad_norm": 11.548758506774902, "learning_rate": 1.532e-05, "loss": 87.3125, "step": 1533 }, { "epoch": 0.0001534, "grad_norm": 11.616901397705078, "learning_rate": 1.533e-05, "loss": 86.9375, "step": 1534 }, { "epoch": 0.0001535, "grad_norm": 11.714616775512695, "learning_rate": 1.5340000000000002e-05, "loss": 86.625, "step": 1535 }, { "epoch": 0.0001536, "grad_norm": 11.771031379699707, "learning_rate": 1.535e-05, "loss": 85.9375, "step": 1536 }, { "epoch": 0.0001537, "grad_norm": 11.87447738647461, "learning_rate": 1.5360000000000002e-05, "loss": 86.3125, "step": 1537 }, { "epoch": 0.0001538, "grad_norm": 11.68323040008545, "learning_rate": 1.537e-05, "loss": 86.4375, "step": 1538 }, { "epoch": 0.0001539, "grad_norm": 11.763370513916016, "learning_rate": 1.538e-05, "loss": 85.9375, "step": 1539 }, { "epoch": 0.000154, "grad_norm": 11.63352108001709, "learning_rate": 1.539e-05, "loss": 86.375, "step": 1540 }, { "epoch": 0.0001541, "grad_norm": 11.659286499023438, "learning_rate": 1.54e-05, "loss": 86.8125, "step": 1541 }, { "epoch": 0.0001542, "grad_norm": 11.715020179748535, "learning_rate": 1.541e-05, "loss": 87.125, "step": 1542 }, { "epoch": 0.0001543, "grad_norm": 11.713168144226074, "learning_rate": 1.5419999999999998e-05, "loss": 86.0, "step": 1543 }, { "epoch": 0.0001544, "grad_norm": 11.8790283203125, "learning_rate": 1.543e-05, "loss": 87.0, "step": 1544 }, { "epoch": 0.0001545, "grad_norm": 11.551057815551758, "learning_rate": 1.544e-05, "loss": 85.8125, "step": 1545 }, { "epoch": 0.0001546, "grad_norm": 11.429879188537598, "learning_rate": 1.545e-05, "loss": 88.0, "step": 1546 }, { "epoch": 0.0001547, "grad_norm": 11.400771141052246, "learning_rate": 1.546e-05, "loss": 87.3125, "step": 1547 }, { "epoch": 0.0001548, "grad_norm": 11.639250755310059, "learning_rate": 1.5470000000000003e-05, "loss": 86.3125, "step": 1548 }, { "epoch": 0.0001549, "grad_norm": 11.554698944091797, "learning_rate": 1.548e-05, "loss": 86.0625, "step": 1549 }, { "epoch": 0.000155, "grad_norm": 11.33513069152832, "learning_rate": 1.549e-05, "loss": 88.0, "step": 1550 }, { "epoch": 0.0001551, "grad_norm": 11.57883358001709, "learning_rate": 1.55e-05, "loss": 86.4375, "step": 1551 }, { "epoch": 0.0001552, "grad_norm": 11.587578773498535, "learning_rate": 1.5510000000000002e-05, "loss": 87.25, "step": 1552 }, { "epoch": 0.0001553, "grad_norm": 11.631257057189941, "learning_rate": 1.552e-05, "loss": 85.5, "step": 1553 }, { "epoch": 0.0001554, "grad_norm": 11.561725616455078, "learning_rate": 1.5530000000000002e-05, "loss": 85.4375, "step": 1554 }, { "epoch": 0.0001555, "grad_norm": 11.868837356567383, "learning_rate": 1.554e-05, "loss": 85.5, "step": 1555 }, { "epoch": 0.0001556, "grad_norm": 12.030888557434082, "learning_rate": 1.555e-05, "loss": 85.4375, "step": 1556 }, { "epoch": 0.0001557, "grad_norm": 11.721089363098145, "learning_rate": 1.556e-05, "loss": 84.875, "step": 1557 }, { "epoch": 0.0001558, "grad_norm": 11.465258598327637, "learning_rate": 1.5570000000000002e-05, "loss": 85.75, "step": 1558 }, { "epoch": 0.0001559, "grad_norm": 11.625471115112305, "learning_rate": 1.558e-05, "loss": 85.375, "step": 1559 }, { "epoch": 0.000156, "grad_norm": 11.661459922790527, "learning_rate": 1.5589999999999998e-05, "loss": 86.8125, "step": 1560 }, { "epoch": 0.0001561, "grad_norm": 11.460598945617676, "learning_rate": 1.56e-05, "loss": 85.875, "step": 1561 }, { "epoch": 0.0001562, "grad_norm": 11.483807563781738, "learning_rate": 1.561e-05, "loss": 85.6875, "step": 1562 }, { "epoch": 0.0001563, "grad_norm": 11.545580863952637, "learning_rate": 1.562e-05, "loss": 84.9375, "step": 1563 }, { "epoch": 0.0001564, "grad_norm": 11.488518714904785, "learning_rate": 1.563e-05, "loss": 86.5, "step": 1564 }, { "epoch": 0.0001565, "grad_norm": 11.404995918273926, "learning_rate": 1.5640000000000003e-05, "loss": 86.0625, "step": 1565 }, { "epoch": 0.0001566, "grad_norm": 11.470978736877441, "learning_rate": 1.565e-05, "loss": 85.5, "step": 1566 }, { "epoch": 0.0001567, "grad_norm": 11.510252952575684, "learning_rate": 1.566e-05, "loss": 86.5625, "step": 1567 }, { "epoch": 0.0001568, "grad_norm": 11.237356185913086, "learning_rate": 1.567e-05, "loss": 86.5625, "step": 1568 }, { "epoch": 0.0001569, "grad_norm": 11.380836486816406, "learning_rate": 1.5680000000000002e-05, "loss": 86.3125, "step": 1569 }, { "epoch": 0.000157, "grad_norm": 11.34556770324707, "learning_rate": 1.569e-05, "loss": 87.25, "step": 1570 }, { "epoch": 0.0001571, "grad_norm": 11.4042387008667, "learning_rate": 1.5700000000000002e-05, "loss": 85.625, "step": 1571 }, { "epoch": 0.0001572, "grad_norm": 11.649144172668457, "learning_rate": 1.571e-05, "loss": 83.9375, "step": 1572 }, { "epoch": 0.0001573, "grad_norm": 11.32030200958252, "learning_rate": 1.572e-05, "loss": 86.1875, "step": 1573 }, { "epoch": 0.0001574, "grad_norm": 11.432470321655273, "learning_rate": 1.573e-05, "loss": 85.75, "step": 1574 }, { "epoch": 0.0001575, "grad_norm": 11.351807594299316, "learning_rate": 1.5740000000000002e-05, "loss": 85.625, "step": 1575 }, { "epoch": 0.0001576, "grad_norm": 11.332376480102539, "learning_rate": 1.575e-05, "loss": 85.5, "step": 1576 }, { "epoch": 0.0001577, "grad_norm": 11.747312545776367, "learning_rate": 1.5759999999999998e-05, "loss": 83.6875, "step": 1577 }, { "epoch": 0.0001578, "grad_norm": 11.493258476257324, "learning_rate": 1.577e-05, "loss": 85.25, "step": 1578 }, { "epoch": 0.0001579, "grad_norm": 11.303948402404785, "learning_rate": 1.578e-05, "loss": 85.4375, "step": 1579 }, { "epoch": 0.000158, "grad_norm": 11.353314399719238, "learning_rate": 1.579e-05, "loss": 85.5, "step": 1580 }, { "epoch": 0.0001581, "grad_norm": 11.475183486938477, "learning_rate": 1.58e-05, "loss": 85.125, "step": 1581 }, { "epoch": 0.0001582, "grad_norm": 11.416069984436035, "learning_rate": 1.5810000000000003e-05, "loss": 85.3125, "step": 1582 }, { "epoch": 0.0001583, "grad_norm": 11.469791412353516, "learning_rate": 1.582e-05, "loss": 83.625, "step": 1583 }, { "epoch": 0.0001584, "grad_norm": 11.495800018310547, "learning_rate": 1.583e-05, "loss": 84.1875, "step": 1584 }, { "epoch": 0.0001585, "grad_norm": 11.347009658813477, "learning_rate": 1.584e-05, "loss": 84.3125, "step": 1585 }, { "epoch": 0.0001586, "grad_norm": 11.404011726379395, "learning_rate": 1.585e-05, "loss": 84.25, "step": 1586 }, { "epoch": 0.0001587, "grad_norm": 11.331399917602539, "learning_rate": 1.586e-05, "loss": 84.6875, "step": 1587 }, { "epoch": 0.0001588, "grad_norm": 11.486993789672852, "learning_rate": 1.5870000000000002e-05, "loss": 84.8125, "step": 1588 }, { "epoch": 0.0001589, "grad_norm": 11.486628532409668, "learning_rate": 1.588e-05, "loss": 84.125, "step": 1589 }, { "epoch": 0.000159, "grad_norm": 11.41197681427002, "learning_rate": 1.589e-05, "loss": 83.875, "step": 1590 }, { "epoch": 0.0001591, "grad_norm": 11.50651741027832, "learning_rate": 1.59e-05, "loss": 83.375, "step": 1591 }, { "epoch": 0.0001592, "grad_norm": 11.40333080291748, "learning_rate": 1.5910000000000002e-05, "loss": 84.4375, "step": 1592 }, { "epoch": 0.0001593, "grad_norm": 11.41156005859375, "learning_rate": 1.592e-05, "loss": 84.125, "step": 1593 }, { "epoch": 0.0001594, "grad_norm": 11.424681663513184, "learning_rate": 1.5929999999999998e-05, "loss": 83.25, "step": 1594 }, { "epoch": 0.0001595, "grad_norm": 11.179929733276367, "learning_rate": 1.594e-05, "loss": 85.25, "step": 1595 }, { "epoch": 0.0001596, "grad_norm": 11.350379943847656, "learning_rate": 1.595e-05, "loss": 83.625, "step": 1596 }, { "epoch": 0.0001597, "grad_norm": 11.30370044708252, "learning_rate": 1.596e-05, "loss": 83.875, "step": 1597 }, { "epoch": 0.0001598, "grad_norm": 11.342832565307617, "learning_rate": 1.597e-05, "loss": 83.25, "step": 1598 }, { "epoch": 0.0001599, "grad_norm": 11.332225799560547, "learning_rate": 1.5980000000000003e-05, "loss": 83.5625, "step": 1599 }, { "epoch": 0.00016, "grad_norm": 11.266222953796387, "learning_rate": 1.5989999999999998e-05, "loss": 84.1875, "step": 1600 }, { "epoch": 0.0001601, "grad_norm": 11.43062686920166, "learning_rate": 1.6e-05, "loss": 83.5625, "step": 1601 }, { "epoch": 0.0001602, "grad_norm": 11.166046142578125, "learning_rate": 1.601e-05, "loss": 84.5, "step": 1602 }, { "epoch": 0.0001603, "grad_norm": 11.344799041748047, "learning_rate": 1.602e-05, "loss": 83.1875, "step": 1603 }, { "epoch": 0.0001604, "grad_norm": 11.39476203918457, "learning_rate": 1.603e-05, "loss": 83.625, "step": 1604 }, { "epoch": 0.0001605, "grad_norm": 11.016305923461914, "learning_rate": 1.604e-05, "loss": 84.9375, "step": 1605 }, { "epoch": 0.0001606, "grad_norm": 11.331496238708496, "learning_rate": 1.605e-05, "loss": 83.6875, "step": 1606 }, { "epoch": 0.0001607, "grad_norm": 11.173765182495117, "learning_rate": 1.606e-05, "loss": 84.375, "step": 1607 }, { "epoch": 0.0001608, "grad_norm": 11.105717658996582, "learning_rate": 1.607e-05, "loss": 85.1875, "step": 1608 }, { "epoch": 0.0001609, "grad_norm": 11.248224258422852, "learning_rate": 1.6080000000000002e-05, "loss": 83.8125, "step": 1609 }, { "epoch": 0.000161, "grad_norm": 11.4295015335083, "learning_rate": 1.609e-05, "loss": 82.8125, "step": 1610 }, { "epoch": 0.0001611, "grad_norm": 11.350955963134766, "learning_rate": 1.61e-05, "loss": 83.9375, "step": 1611 }, { "epoch": 0.0001612, "grad_norm": 11.247148513793945, "learning_rate": 1.611e-05, "loss": 83.4375, "step": 1612 }, { "epoch": 0.0001613, "grad_norm": 11.086236000061035, "learning_rate": 1.612e-05, "loss": 84.3125, "step": 1613 }, { "epoch": 0.0001614, "grad_norm": 11.098573684692383, "learning_rate": 1.613e-05, "loss": 83.875, "step": 1614 }, { "epoch": 0.0001615, "grad_norm": 11.348047256469727, "learning_rate": 1.614e-05, "loss": 83.1875, "step": 1615 }, { "epoch": 0.0001616, "grad_norm": 11.169997215270996, "learning_rate": 1.6150000000000003e-05, "loss": 83.1875, "step": 1616 }, { "epoch": 0.0001617, "grad_norm": 11.257536888122559, "learning_rate": 1.6159999999999998e-05, "loss": 83.5, "step": 1617 }, { "epoch": 0.0001618, "grad_norm": 11.304291725158691, "learning_rate": 1.617e-05, "loss": 83.9375, "step": 1618 }, { "epoch": 0.0001619, "grad_norm": 11.34911823272705, "learning_rate": 1.618e-05, "loss": 82.9375, "step": 1619 }, { "epoch": 0.000162, "grad_norm": 11.359582901000977, "learning_rate": 1.619e-05, "loss": 82.6875, "step": 1620 }, { "epoch": 0.0001621, "grad_norm": 11.43436336517334, "learning_rate": 1.62e-05, "loss": 81.5, "step": 1621 }, { "epoch": 0.0001622, "grad_norm": 11.373072624206543, "learning_rate": 1.621e-05, "loss": 82.4375, "step": 1622 }, { "epoch": 0.0001623, "grad_norm": 11.195119857788086, "learning_rate": 1.622e-05, "loss": 84.0625, "step": 1623 }, { "epoch": 0.0001624, "grad_norm": 11.049300193786621, "learning_rate": 1.623e-05, "loss": 83.375, "step": 1624 }, { "epoch": 0.0001625, "grad_norm": 11.10819149017334, "learning_rate": 1.624e-05, "loss": 83.4375, "step": 1625 }, { "epoch": 0.0001626, "grad_norm": 11.200508117675781, "learning_rate": 1.6250000000000002e-05, "loss": 82.5625, "step": 1626 }, { "epoch": 0.0001627, "grad_norm": 11.225939750671387, "learning_rate": 1.626e-05, "loss": 82.5625, "step": 1627 }, { "epoch": 0.0001628, "grad_norm": 11.106568336486816, "learning_rate": 1.627e-05, "loss": 83.375, "step": 1628 }, { "epoch": 0.0001629, "grad_norm": 11.150421142578125, "learning_rate": 1.628e-05, "loss": 83.0625, "step": 1629 }, { "epoch": 0.000163, "grad_norm": 11.192241668701172, "learning_rate": 1.629e-05, "loss": 82.5625, "step": 1630 }, { "epoch": 0.0001631, "grad_norm": 11.128030776977539, "learning_rate": 1.63e-05, "loss": 82.6875, "step": 1631 }, { "epoch": 0.0001632, "grad_norm": 11.366748809814453, "learning_rate": 1.631e-05, "loss": 81.0, "step": 1632 }, { "epoch": 0.0001633, "grad_norm": 11.179375648498535, "learning_rate": 1.6320000000000003e-05, "loss": 82.9375, "step": 1633 }, { "epoch": 0.0001634, "grad_norm": 10.959238052368164, "learning_rate": 1.6329999999999998e-05, "loss": 83.75, "step": 1634 }, { "epoch": 0.0001635, "grad_norm": 11.190832138061523, "learning_rate": 1.634e-05, "loss": 82.3125, "step": 1635 }, { "epoch": 0.0001636, "grad_norm": 11.21330738067627, "learning_rate": 1.635e-05, "loss": 81.6875, "step": 1636 }, { "epoch": 0.0001637, "grad_norm": 11.11747932434082, "learning_rate": 1.636e-05, "loss": 81.9375, "step": 1637 }, { "epoch": 0.0001638, "grad_norm": 11.298837661743164, "learning_rate": 1.637e-05, "loss": 82.25, "step": 1638 }, { "epoch": 0.0001639, "grad_norm": 11.277381896972656, "learning_rate": 1.638e-05, "loss": 81.0, "step": 1639 }, { "epoch": 0.000164, "grad_norm": 11.104133605957031, "learning_rate": 1.639e-05, "loss": 82.1875, "step": 1640 }, { "epoch": 0.0001641, "grad_norm": 11.214916229248047, "learning_rate": 1.64e-05, "loss": 82.0625, "step": 1641 }, { "epoch": 0.0001642, "grad_norm": 11.148131370544434, "learning_rate": 1.641e-05, "loss": 82.8125, "step": 1642 }, { "epoch": 0.0001643, "grad_norm": 11.112770080566406, "learning_rate": 1.6420000000000002e-05, "loss": 82.5, "step": 1643 }, { "epoch": 0.0001644, "grad_norm": 11.18635368347168, "learning_rate": 1.643e-05, "loss": 80.75, "step": 1644 }, { "epoch": 0.0001645, "grad_norm": 11.048691749572754, "learning_rate": 1.644e-05, "loss": 83.8125, "step": 1645 }, { "epoch": 0.0001646, "grad_norm": 11.135150909423828, "learning_rate": 1.645e-05, "loss": 82.125, "step": 1646 }, { "epoch": 0.0001647, "grad_norm": 11.023847579956055, "learning_rate": 1.646e-05, "loss": 82.1875, "step": 1647 }, { "epoch": 0.0001648, "grad_norm": 11.032581329345703, "learning_rate": 1.647e-05, "loss": 82.25, "step": 1648 }, { "epoch": 0.0001649, "grad_norm": 10.882695198059082, "learning_rate": 1.648e-05, "loss": 84.9375, "step": 1649 }, { "epoch": 0.000165, "grad_norm": 10.961777687072754, "learning_rate": 1.6490000000000003e-05, "loss": 81.8125, "step": 1650 }, { "epoch": 0.0001651, "grad_norm": 11.20870304107666, "learning_rate": 1.6499999999999998e-05, "loss": 82.1875, "step": 1651 }, { "epoch": 0.0001652, "grad_norm": 11.134272575378418, "learning_rate": 1.651e-05, "loss": 81.5, "step": 1652 }, { "epoch": 0.0001653, "grad_norm": 10.889755249023438, "learning_rate": 1.652e-05, "loss": 83.5625, "step": 1653 }, { "epoch": 0.0001654, "grad_norm": 11.065031051635742, "learning_rate": 1.653e-05, "loss": 82.375, "step": 1654 }, { "epoch": 0.0001655, "grad_norm": 11.27316665649414, "learning_rate": 1.654e-05, "loss": 79.8125, "step": 1655 }, { "epoch": 0.0001656, "grad_norm": 11.198715209960938, "learning_rate": 1.655e-05, "loss": 80.75, "step": 1656 }, { "epoch": 0.0001657, "grad_norm": 11.055648803710938, "learning_rate": 1.656e-05, "loss": 81.9375, "step": 1657 }, { "epoch": 0.0001658, "grad_norm": 11.146472930908203, "learning_rate": 1.657e-05, "loss": 81.0625, "step": 1658 }, { "epoch": 0.0001659, "grad_norm": 11.19427490234375, "learning_rate": 1.658e-05, "loss": 80.5, "step": 1659 }, { "epoch": 0.000166, "grad_norm": 10.967174530029297, "learning_rate": 1.6590000000000002e-05, "loss": 82.875, "step": 1660 }, { "epoch": 0.0001661, "grad_norm": 11.37897777557373, "learning_rate": 1.66e-05, "loss": 79.75, "step": 1661 }, { "epoch": 0.0001662, "grad_norm": 11.097198486328125, "learning_rate": 1.661e-05, "loss": 82.0625, "step": 1662 }, { "epoch": 0.0001663, "grad_norm": 10.96146011352539, "learning_rate": 1.662e-05, "loss": 81.125, "step": 1663 }, { "epoch": 0.0001664, "grad_norm": 11.034396171569824, "learning_rate": 1.6630000000000002e-05, "loss": 80.0625, "step": 1664 }, { "epoch": 0.0001665, "grad_norm": 11.061921119689941, "learning_rate": 1.664e-05, "loss": 81.875, "step": 1665 }, { "epoch": 0.0001666, "grad_norm": 11.300657272338867, "learning_rate": 1.665e-05, "loss": 79.8125, "step": 1666 }, { "epoch": 0.0001667, "grad_norm": 11.197166442871094, "learning_rate": 1.6660000000000003e-05, "loss": 80.375, "step": 1667 }, { "epoch": 0.0001668, "grad_norm": 10.98279094696045, "learning_rate": 1.6669999999999998e-05, "loss": 81.9375, "step": 1668 }, { "epoch": 0.0001669, "grad_norm": 11.043325424194336, "learning_rate": 1.668e-05, "loss": 80.75, "step": 1669 }, { "epoch": 0.000167, "grad_norm": 10.966551780700684, "learning_rate": 1.669e-05, "loss": 80.375, "step": 1670 }, { "epoch": 0.0001671, "grad_norm": 10.982937812805176, "learning_rate": 1.67e-05, "loss": 80.5, "step": 1671 }, { "epoch": 0.0001672, "grad_norm": 11.045061111450195, "learning_rate": 1.671e-05, "loss": 81.125, "step": 1672 }, { "epoch": 0.0001673, "grad_norm": 11.25208854675293, "learning_rate": 1.672e-05, "loss": 80.0625, "step": 1673 }, { "epoch": 0.0001674, "grad_norm": 10.908302307128906, "learning_rate": 1.673e-05, "loss": 80.875, "step": 1674 }, { "epoch": 0.0001675, "grad_norm": 10.964232444763184, "learning_rate": 1.674e-05, "loss": 79.6875, "step": 1675 }, { "epoch": 0.0001676, "grad_norm": 10.88063907623291, "learning_rate": 1.675e-05, "loss": 80.9375, "step": 1676 }, { "epoch": 0.0001677, "grad_norm": 11.139262199401855, "learning_rate": 1.6760000000000002e-05, "loss": 79.4375, "step": 1677 }, { "epoch": 0.0001678, "grad_norm": 11.198766708374023, "learning_rate": 1.677e-05, "loss": 79.125, "step": 1678 }, { "epoch": 0.0001679, "grad_norm": 10.993010520935059, "learning_rate": 1.678e-05, "loss": 81.25, "step": 1679 }, { "epoch": 0.000168, "grad_norm": 11.06545352935791, "learning_rate": 1.679e-05, "loss": 79.4375, "step": 1680 }, { "epoch": 0.0001681, "grad_norm": 10.88327693939209, "learning_rate": 1.6800000000000002e-05, "loss": 81.0625, "step": 1681 }, { "epoch": 0.0001682, "grad_norm": 10.90662956237793, "learning_rate": 1.681e-05, "loss": 81.0625, "step": 1682 }, { "epoch": 0.0001683, "grad_norm": 10.89114761352539, "learning_rate": 1.682e-05, "loss": 81.375, "step": 1683 }, { "epoch": 0.0001684, "grad_norm": 10.992010116577148, "learning_rate": 1.6830000000000003e-05, "loss": 80.6875, "step": 1684 }, { "epoch": 0.0001685, "grad_norm": 11.057424545288086, "learning_rate": 1.6839999999999998e-05, "loss": 79.6875, "step": 1685 }, { "epoch": 0.0001686, "grad_norm": 10.815861701965332, "learning_rate": 1.685e-05, "loss": 80.8125, "step": 1686 }, { "epoch": 0.0001687, "grad_norm": 10.851166725158691, "learning_rate": 1.686e-05, "loss": 81.75, "step": 1687 }, { "epoch": 0.0001688, "grad_norm": 11.063788414001465, "learning_rate": 1.687e-05, "loss": 80.0, "step": 1688 }, { "epoch": 0.0001689, "grad_norm": 10.901667594909668, "learning_rate": 1.688e-05, "loss": 79.8125, "step": 1689 }, { "epoch": 0.000169, "grad_norm": 10.897828102111816, "learning_rate": 1.689e-05, "loss": 79.375, "step": 1690 }, { "epoch": 0.0001691, "grad_norm": 10.9066162109375, "learning_rate": 1.69e-05, "loss": 80.25, "step": 1691 }, { "epoch": 0.0001692, "grad_norm": 11.124065399169922, "learning_rate": 1.691e-05, "loss": 79.4375, "step": 1692 }, { "epoch": 0.0001693, "grad_norm": 10.849020957946777, "learning_rate": 1.692e-05, "loss": 80.25, "step": 1693 }, { "epoch": 0.0001694, "grad_norm": 10.995338439941406, "learning_rate": 1.6930000000000002e-05, "loss": 80.1875, "step": 1694 }, { "epoch": 0.0001695, "grad_norm": 10.745613098144531, "learning_rate": 1.694e-05, "loss": 81.3125, "step": 1695 }, { "epoch": 0.0001696, "grad_norm": 10.822576522827148, "learning_rate": 1.695e-05, "loss": 81.0625, "step": 1696 }, { "epoch": 0.0001697, "grad_norm": 10.834251403808594, "learning_rate": 1.696e-05, "loss": 81.0625, "step": 1697 }, { "epoch": 0.0001698, "grad_norm": 10.99413776397705, "learning_rate": 1.6970000000000002e-05, "loss": 78.5, "step": 1698 }, { "epoch": 0.0001699, "grad_norm": 10.745895385742188, "learning_rate": 1.698e-05, "loss": 79.875, "step": 1699 }, { "epoch": 0.00017, "grad_norm": 10.999711990356445, "learning_rate": 1.699e-05, "loss": 79.3125, "step": 1700 }, { "epoch": 0.0001701, "grad_norm": 10.955674171447754, "learning_rate": 1.7000000000000003e-05, "loss": 80.875, "step": 1701 }, { "epoch": 0.0001702, "grad_norm": 10.888795852661133, "learning_rate": 1.7009999999999998e-05, "loss": 78.9375, "step": 1702 }, { "epoch": 0.0001703, "grad_norm": 10.802494049072266, "learning_rate": 1.702e-05, "loss": 79.625, "step": 1703 }, { "epoch": 0.0001704, "grad_norm": 10.999938011169434, "learning_rate": 1.703e-05, "loss": 79.125, "step": 1704 }, { "epoch": 0.0001705, "grad_norm": 10.716090202331543, "learning_rate": 1.704e-05, "loss": 80.125, "step": 1705 }, { "epoch": 0.0001706, "grad_norm": 10.765728950500488, "learning_rate": 1.705e-05, "loss": 79.9375, "step": 1706 }, { "epoch": 0.0001707, "grad_norm": 10.725311279296875, "learning_rate": 1.706e-05, "loss": 80.1875, "step": 1707 }, { "epoch": 0.0001708, "grad_norm": 10.912590980529785, "learning_rate": 1.707e-05, "loss": 78.875, "step": 1708 }, { "epoch": 0.0001709, "grad_norm": 10.935500144958496, "learning_rate": 1.708e-05, "loss": 79.8125, "step": 1709 }, { "epoch": 0.000171, "grad_norm": 10.742402076721191, "learning_rate": 1.709e-05, "loss": 80.4375, "step": 1710 }, { "epoch": 0.0001711, "grad_norm": 10.83121395111084, "learning_rate": 1.7100000000000002e-05, "loss": 79.1875, "step": 1711 }, { "epoch": 0.0001712, "grad_norm": 11.009101867675781, "learning_rate": 1.711e-05, "loss": 79.5, "step": 1712 }, { "epoch": 0.0001713, "grad_norm": 10.9196138381958, "learning_rate": 1.712e-05, "loss": 79.8125, "step": 1713 }, { "epoch": 0.0001714, "grad_norm": 10.713299751281738, "learning_rate": 1.713e-05, "loss": 79.875, "step": 1714 }, { "epoch": 0.0001715, "grad_norm": 10.738595962524414, "learning_rate": 1.714e-05, "loss": 79.375, "step": 1715 }, { "epoch": 0.0001716, "grad_norm": 10.912589073181152, "learning_rate": 1.715e-05, "loss": 78.625, "step": 1716 }, { "epoch": 0.0001717, "grad_norm": 10.915419578552246, "learning_rate": 1.7160000000000002e-05, "loss": 78.8125, "step": 1717 }, { "epoch": 0.0001718, "grad_norm": 10.834695816040039, "learning_rate": 1.717e-05, "loss": 79.875, "step": 1718 }, { "epoch": 0.0001719, "grad_norm": 10.900665283203125, "learning_rate": 1.7179999999999998e-05, "loss": 78.3125, "step": 1719 }, { "epoch": 0.000172, "grad_norm": 10.814151763916016, "learning_rate": 1.719e-05, "loss": 77.75, "step": 1720 }, { "epoch": 0.0001721, "grad_norm": 10.723401069641113, "learning_rate": 1.72e-05, "loss": 79.625, "step": 1721 }, { "epoch": 0.0001722, "grad_norm": 10.772787094116211, "learning_rate": 1.721e-05, "loss": 78.8125, "step": 1722 }, { "epoch": 0.0001723, "grad_norm": 10.624513626098633, "learning_rate": 1.722e-05, "loss": 79.0, "step": 1723 }, { "epoch": 0.0001724, "grad_norm": 10.845480918884277, "learning_rate": 1.723e-05, "loss": 78.25, "step": 1724 }, { "epoch": 0.0001725, "grad_norm": 10.63790225982666, "learning_rate": 1.724e-05, "loss": 79.3125, "step": 1725 }, { "epoch": 0.0001726, "grad_norm": 10.91887378692627, "learning_rate": 1.725e-05, "loss": 77.8125, "step": 1726 }, { "epoch": 0.0001727, "grad_norm": 11.007935523986816, "learning_rate": 1.726e-05, "loss": 78.3125, "step": 1727 }, { "epoch": 0.0001728, "grad_norm": 10.816855430603027, "learning_rate": 1.7270000000000002e-05, "loss": 77.625, "step": 1728 }, { "epoch": 0.0001729, "grad_norm": 10.639528274536133, "learning_rate": 1.728e-05, "loss": 79.25, "step": 1729 }, { "epoch": 0.000173, "grad_norm": 10.81132698059082, "learning_rate": 1.729e-05, "loss": 77.875, "step": 1730 }, { "epoch": 0.0001731, "grad_norm": 10.696173667907715, "learning_rate": 1.73e-05, "loss": 79.1875, "step": 1731 }, { "epoch": 0.0001732, "grad_norm": 10.654017448425293, "learning_rate": 1.731e-05, "loss": 79.875, "step": 1732 }, { "epoch": 0.0001733, "grad_norm": 10.759590148925781, "learning_rate": 1.732e-05, "loss": 78.875, "step": 1733 }, { "epoch": 0.0001734, "grad_norm": 10.894126892089844, "learning_rate": 1.7330000000000002e-05, "loss": 80.3125, "step": 1734 }, { "epoch": 0.0001735, "grad_norm": 10.793487548828125, "learning_rate": 1.734e-05, "loss": 79.6875, "step": 1735 }, { "epoch": 0.0001736, "grad_norm": 11.036444664001465, "learning_rate": 1.7349999999999998e-05, "loss": 77.9375, "step": 1736 }, { "epoch": 0.0001737, "grad_norm": 10.841407775878906, "learning_rate": 1.736e-05, "loss": 77.75, "step": 1737 }, { "epoch": 0.0001738, "grad_norm": 10.731624603271484, "learning_rate": 1.737e-05, "loss": 79.0, "step": 1738 }, { "epoch": 0.0001739, "grad_norm": 10.66896915435791, "learning_rate": 1.738e-05, "loss": 79.75, "step": 1739 }, { "epoch": 0.000174, "grad_norm": 10.88010025024414, "learning_rate": 1.739e-05, "loss": 79.0625, "step": 1740 }, { "epoch": 0.0001741, "grad_norm": 10.936246871948242, "learning_rate": 1.74e-05, "loss": 79.5625, "step": 1741 }, { "epoch": 0.0001742, "grad_norm": 10.785653114318848, "learning_rate": 1.741e-05, "loss": 77.8125, "step": 1742 }, { "epoch": 0.0001743, "grad_norm": 10.53821086883545, "learning_rate": 1.742e-05, "loss": 79.9375, "step": 1743 }, { "epoch": 0.0001744, "grad_norm": 10.72213363647461, "learning_rate": 1.743e-05, "loss": 80.0, "step": 1744 }, { "epoch": 0.0001745, "grad_norm": 10.682663917541504, "learning_rate": 1.7440000000000002e-05, "loss": 79.0, "step": 1745 }, { "epoch": 0.0001746, "grad_norm": 10.701003074645996, "learning_rate": 1.745e-05, "loss": 79.0, "step": 1746 }, { "epoch": 0.0001747, "grad_norm": 10.708039283752441, "learning_rate": 1.746e-05, "loss": 77.9375, "step": 1747 }, { "epoch": 0.0001748, "grad_norm": 10.948110580444336, "learning_rate": 1.747e-05, "loss": 76.25, "step": 1748 }, { "epoch": 0.0001749, "grad_norm": 10.633480072021484, "learning_rate": 1.748e-05, "loss": 80.0, "step": 1749 }, { "epoch": 0.000175, "grad_norm": 10.546310424804688, "learning_rate": 1.749e-05, "loss": 79.25, "step": 1750 }, { "epoch": 0.0001751, "grad_norm": 10.7278413772583, "learning_rate": 1.7500000000000002e-05, "loss": 77.3125, "step": 1751 }, { "epoch": 0.0001752, "grad_norm": 10.804795265197754, "learning_rate": 1.751e-05, "loss": 77.8125, "step": 1752 }, { "epoch": 0.0001753, "grad_norm": 10.649805068969727, "learning_rate": 1.7519999999999998e-05, "loss": 77.9375, "step": 1753 }, { "epoch": 0.0001754, "grad_norm": 10.849508285522461, "learning_rate": 1.753e-05, "loss": 75.875, "step": 1754 }, { "epoch": 0.0001755, "grad_norm": 10.669038772583008, "learning_rate": 1.754e-05, "loss": 79.1875, "step": 1755 }, { "epoch": 0.0001756, "grad_norm": 10.650232315063477, "learning_rate": 1.755e-05, "loss": 78.125, "step": 1756 }, { "epoch": 0.0001757, "grad_norm": 10.93763542175293, "learning_rate": 1.756e-05, "loss": 75.4375, "step": 1757 }, { "epoch": 0.0001758, "grad_norm": 10.791654586791992, "learning_rate": 1.757e-05, "loss": 77.0, "step": 1758 }, { "epoch": 0.0001759, "grad_norm": 10.705726623535156, "learning_rate": 1.758e-05, "loss": 77.9375, "step": 1759 }, { "epoch": 0.000176, "grad_norm": 10.580390930175781, "learning_rate": 1.759e-05, "loss": 77.625, "step": 1760 }, { "epoch": 0.0001761, "grad_norm": 10.656951904296875, "learning_rate": 1.76e-05, "loss": 77.1875, "step": 1761 }, { "epoch": 0.0001762, "grad_norm": 10.772924423217773, "learning_rate": 1.7610000000000002e-05, "loss": 76.0625, "step": 1762 }, { "epoch": 0.0001763, "grad_norm": 10.663338661193848, "learning_rate": 1.762e-05, "loss": 77.25, "step": 1763 }, { "epoch": 0.0001764, "grad_norm": 10.654496192932129, "learning_rate": 1.763e-05, "loss": 77.0, "step": 1764 }, { "epoch": 0.0001765, "grad_norm": 10.55364990234375, "learning_rate": 1.764e-05, "loss": 79.3125, "step": 1765 }, { "epoch": 0.0001766, "grad_norm": 10.59168529510498, "learning_rate": 1.765e-05, "loss": 77.8125, "step": 1766 }, { "epoch": 0.0001767, "grad_norm": 10.650297164916992, "learning_rate": 1.766e-05, "loss": 77.125, "step": 1767 }, { "epoch": 0.0001768, "grad_norm": 10.719636917114258, "learning_rate": 1.7670000000000002e-05, "loss": 77.0, "step": 1768 }, { "epoch": 0.0001769, "grad_norm": 10.793968200683594, "learning_rate": 1.768e-05, "loss": 75.75, "step": 1769 }, { "epoch": 0.000177, "grad_norm": 10.62430477142334, "learning_rate": 1.769e-05, "loss": 76.3125, "step": 1770 }, { "epoch": 0.0001771, "grad_norm": 10.62021541595459, "learning_rate": 1.77e-05, "loss": 78.4375, "step": 1771 }, { "epoch": 0.0001772, "grad_norm": 10.529915809631348, "learning_rate": 1.771e-05, "loss": 76.5625, "step": 1772 }, { "epoch": 0.0001773, "grad_norm": 10.798215866088867, "learning_rate": 1.772e-05, "loss": 75.4375, "step": 1773 }, { "epoch": 0.0001774, "grad_norm": 10.537800788879395, "learning_rate": 1.773e-05, "loss": 77.125, "step": 1774 }, { "epoch": 0.0001775, "grad_norm": 10.770161628723145, "learning_rate": 1.774e-05, "loss": 76.4375, "step": 1775 }, { "epoch": 0.0001776, "grad_norm": 10.577451705932617, "learning_rate": 1.775e-05, "loss": 76.0, "step": 1776 }, { "epoch": 0.0001777, "grad_norm": 10.794952392578125, "learning_rate": 1.776e-05, "loss": 75.75, "step": 1777 }, { "epoch": 0.0001778, "grad_norm": 10.660951614379883, "learning_rate": 1.777e-05, "loss": 75.5, "step": 1778 }, { "epoch": 0.0001779, "grad_norm": 10.519042015075684, "learning_rate": 1.7780000000000003e-05, "loss": 77.0, "step": 1779 }, { "epoch": 0.000178, "grad_norm": 10.545072555541992, "learning_rate": 1.779e-05, "loss": 76.8125, "step": 1780 }, { "epoch": 0.0001781, "grad_norm": 10.663309097290039, "learning_rate": 1.78e-05, "loss": 76.6875, "step": 1781 }, { "epoch": 0.0001782, "grad_norm": 10.746946334838867, "learning_rate": 1.781e-05, "loss": 77.375, "step": 1782 }, { "epoch": 0.0001783, "grad_norm": 10.625143051147461, "learning_rate": 1.782e-05, "loss": 76.5, "step": 1783 }, { "epoch": 0.0001784, "grad_norm": 10.452560424804688, "learning_rate": 1.783e-05, "loss": 77.5, "step": 1784 }, { "epoch": 0.0001785, "grad_norm": 10.465551376342773, "learning_rate": 1.7840000000000002e-05, "loss": 76.625, "step": 1785 }, { "epoch": 0.0001786, "grad_norm": 10.593033790588379, "learning_rate": 1.785e-05, "loss": 77.5, "step": 1786 }, { "epoch": 0.0001787, "grad_norm": 10.652999877929688, "learning_rate": 1.786e-05, "loss": 76.5, "step": 1787 }, { "epoch": 0.0001788, "grad_norm": 10.6140775680542, "learning_rate": 1.787e-05, "loss": 76.25, "step": 1788 }, { "epoch": 0.0001789, "grad_norm": 10.677515983581543, "learning_rate": 1.788e-05, "loss": 75.875, "step": 1789 }, { "epoch": 0.000179, "grad_norm": 10.523712158203125, "learning_rate": 1.789e-05, "loss": 77.0, "step": 1790 }, { "epoch": 0.0001791, "grad_norm": 10.816093444824219, "learning_rate": 1.79e-05, "loss": 74.8125, "step": 1791 }, { "epoch": 0.0001792, "grad_norm": 10.561775207519531, "learning_rate": 1.791e-05, "loss": 75.5, "step": 1792 }, { "epoch": 0.0001793, "grad_norm": 10.464574813842773, "learning_rate": 1.792e-05, "loss": 76.75, "step": 1793 }, { "epoch": 0.0001794, "grad_norm": 10.453524589538574, "learning_rate": 1.793e-05, "loss": 76.5625, "step": 1794 }, { "epoch": 0.0001795, "grad_norm": 10.519964218139648, "learning_rate": 1.794e-05, "loss": 76.625, "step": 1795 }, { "epoch": 0.0001796, "grad_norm": 10.693997383117676, "learning_rate": 1.7950000000000003e-05, "loss": 74.4375, "step": 1796 }, { "epoch": 0.0001797, "grad_norm": 10.64401626586914, "learning_rate": 1.796e-05, "loss": 75.3125, "step": 1797 }, { "epoch": 0.0001798, "grad_norm": 10.452816009521484, "learning_rate": 1.797e-05, "loss": 78.1875, "step": 1798 }, { "epoch": 0.0001799, "grad_norm": 10.566648483276367, "learning_rate": 1.798e-05, "loss": 75.875, "step": 1799 }, { "epoch": 0.00018, "grad_norm": 10.594141960144043, "learning_rate": 1.799e-05, "loss": 74.875, "step": 1800 }, { "epoch": 0.0001801, "grad_norm": 10.448125839233398, "learning_rate": 1.8e-05, "loss": 76.3125, "step": 1801 }, { "epoch": 0.0001802, "grad_norm": 10.401894569396973, "learning_rate": 1.8010000000000002e-05, "loss": 76.75, "step": 1802 }, { "epoch": 0.0001803, "grad_norm": 10.610258102416992, "learning_rate": 1.802e-05, "loss": 75.1875, "step": 1803 }, { "epoch": 0.0001804, "grad_norm": 10.440401077270508, "learning_rate": 1.803e-05, "loss": 76.25, "step": 1804 }, { "epoch": 0.0001805, "grad_norm": 10.447839736938477, "learning_rate": 1.804e-05, "loss": 77.0, "step": 1805 }, { "epoch": 0.0001806, "grad_norm": 10.623817443847656, "learning_rate": 1.805e-05, "loss": 76.3125, "step": 1806 }, { "epoch": 0.0001807, "grad_norm": 10.760772705078125, "learning_rate": 1.806e-05, "loss": 75.375, "step": 1807 }, { "epoch": 0.0001808, "grad_norm": 10.51712703704834, "learning_rate": 1.807e-05, "loss": 75.9375, "step": 1808 }, { "epoch": 0.0001809, "grad_norm": 10.787428855895996, "learning_rate": 1.808e-05, "loss": 74.125, "step": 1809 }, { "epoch": 0.000181, "grad_norm": 10.413379669189453, "learning_rate": 1.809e-05, "loss": 76.5625, "step": 1810 }, { "epoch": 0.0001811, "grad_norm": 10.707974433898926, "learning_rate": 1.81e-05, "loss": 74.5625, "step": 1811 }, { "epoch": 0.0001812, "grad_norm": 10.540277481079102, "learning_rate": 1.811e-05, "loss": 75.5625, "step": 1812 }, { "epoch": 0.0001813, "grad_norm": 10.474909782409668, "learning_rate": 1.8120000000000003e-05, "loss": 77.1875, "step": 1813 }, { "epoch": 0.0001814, "grad_norm": 10.581918716430664, "learning_rate": 1.813e-05, "loss": 75.625, "step": 1814 }, { "epoch": 0.0001815, "grad_norm": 10.54649543762207, "learning_rate": 1.814e-05, "loss": 75.4375, "step": 1815 }, { "epoch": 0.0001816, "grad_norm": 10.404945373535156, "learning_rate": 1.815e-05, "loss": 75.8125, "step": 1816 }, { "epoch": 0.0001817, "grad_norm": 10.537050247192383, "learning_rate": 1.816e-05, "loss": 75.125, "step": 1817 }, { "epoch": 0.0001818, "grad_norm": 10.64071273803711, "learning_rate": 1.817e-05, "loss": 76.6875, "step": 1818 }, { "epoch": 0.0001819, "grad_norm": 10.442253112792969, "learning_rate": 1.8180000000000002e-05, "loss": 75.375, "step": 1819 }, { "epoch": 0.000182, "grad_norm": 10.428080558776855, "learning_rate": 1.819e-05, "loss": 76.0625, "step": 1820 }, { "epoch": 0.0001821, "grad_norm": 10.296630859375, "learning_rate": 1.82e-05, "loss": 76.4375, "step": 1821 }, { "epoch": 0.0001822, "grad_norm": 10.323344230651855, "learning_rate": 1.821e-05, "loss": 74.875, "step": 1822 }, { "epoch": 0.0001823, "grad_norm": 10.475907325744629, "learning_rate": 1.8220000000000002e-05, "loss": 75.625, "step": 1823 }, { "epoch": 0.0001824, "grad_norm": 10.535412788391113, "learning_rate": 1.823e-05, "loss": 74.75, "step": 1824 }, { "epoch": 0.0001825, "grad_norm": 10.529937744140625, "learning_rate": 1.824e-05, "loss": 74.0625, "step": 1825 }, { "epoch": 0.0001826, "grad_norm": 10.555984497070312, "learning_rate": 1.825e-05, "loss": 74.75, "step": 1826 }, { "epoch": 0.0001827, "grad_norm": 10.664322853088379, "learning_rate": 1.826e-05, "loss": 74.875, "step": 1827 }, { "epoch": 0.0001828, "grad_norm": 10.602211952209473, "learning_rate": 1.827e-05, "loss": 74.0625, "step": 1828 }, { "epoch": 0.0001829, "grad_norm": 10.418312072753906, "learning_rate": 1.828e-05, "loss": 75.0625, "step": 1829 }, { "epoch": 0.000183, "grad_norm": 10.414963722229004, "learning_rate": 1.8290000000000003e-05, "loss": 75.6875, "step": 1830 }, { "epoch": 0.0001831, "grad_norm": 10.508442878723145, "learning_rate": 1.83e-05, "loss": 74.375, "step": 1831 }, { "epoch": 0.0001832, "grad_norm": 10.432555198669434, "learning_rate": 1.831e-05, "loss": 75.4375, "step": 1832 }, { "epoch": 0.0001833, "grad_norm": 10.469473838806152, "learning_rate": 1.832e-05, "loss": 74.5625, "step": 1833 }, { "epoch": 0.0001834, "grad_norm": 10.339816093444824, "learning_rate": 1.833e-05, "loss": 74.875, "step": 1834 }, { "epoch": 0.0001835, "grad_norm": 10.438968658447266, "learning_rate": 1.834e-05, "loss": 74.8125, "step": 1835 }, { "epoch": 0.0001836, "grad_norm": 10.777366638183594, "learning_rate": 1.8350000000000002e-05, "loss": 75.0, "step": 1836 }, { "epoch": 0.0001837, "grad_norm": 10.707802772521973, "learning_rate": 1.836e-05, "loss": 78.0625, "step": 1837 }, { "epoch": 0.0001838, "grad_norm": 10.437213897705078, "learning_rate": 1.837e-05, "loss": 74.3125, "step": 1838 }, { "epoch": 0.0001839, "grad_norm": 10.431440353393555, "learning_rate": 1.838e-05, "loss": 75.875, "step": 1839 }, { "epoch": 0.000184, "grad_norm": 10.38449478149414, "learning_rate": 1.8390000000000002e-05, "loss": 74.125, "step": 1840 }, { "epoch": 0.0001841, "grad_norm": 10.561605453491211, "learning_rate": 1.84e-05, "loss": 77.125, "step": 1841 }, { "epoch": 0.0001842, "grad_norm": 10.490052223205566, "learning_rate": 1.841e-05, "loss": 74.5625, "step": 1842 }, { "epoch": 0.0001843, "grad_norm": 10.40212631225586, "learning_rate": 1.842e-05, "loss": 75.8125, "step": 1843 }, { "epoch": 0.0001844, "grad_norm": 10.394779205322266, "learning_rate": 1.8429999999999998e-05, "loss": 74.1875, "step": 1844 }, { "epoch": 0.0001845, "grad_norm": 10.477581024169922, "learning_rate": 1.844e-05, "loss": 73.5625, "step": 1845 }, { "epoch": 0.0001846, "grad_norm": 10.522961616516113, "learning_rate": 1.845e-05, "loss": 72.875, "step": 1846 }, { "epoch": 0.0001847, "grad_norm": 10.521002769470215, "learning_rate": 1.846e-05, "loss": 73.3125, "step": 1847 }, { "epoch": 0.0001848, "grad_norm": 10.377899169921875, "learning_rate": 1.847e-05, "loss": 74.8125, "step": 1848 }, { "epoch": 0.0001849, "grad_norm": 10.571773529052734, "learning_rate": 1.848e-05, "loss": 74.125, "step": 1849 }, { "epoch": 0.000185, "grad_norm": 10.516894340515137, "learning_rate": 1.849e-05, "loss": 75.125, "step": 1850 }, { "epoch": 0.0001851, "grad_norm": 10.519248008728027, "learning_rate": 1.85e-05, "loss": 73.0, "step": 1851 }, { "epoch": 0.0001852, "grad_norm": 10.265591621398926, "learning_rate": 1.851e-05, "loss": 74.0625, "step": 1852 }, { "epoch": 0.0001853, "grad_norm": 10.392489433288574, "learning_rate": 1.8520000000000002e-05, "loss": 75.8125, "step": 1853 }, { "epoch": 0.0001854, "grad_norm": 10.352034568786621, "learning_rate": 1.853e-05, "loss": 74.5625, "step": 1854 }, { "epoch": 0.0001855, "grad_norm": 10.400627136230469, "learning_rate": 1.854e-05, "loss": 73.9375, "step": 1855 }, { "epoch": 0.0001856, "grad_norm": 10.386198043823242, "learning_rate": 1.855e-05, "loss": 73.5, "step": 1856 }, { "epoch": 0.0001857, "grad_norm": 10.476325035095215, "learning_rate": 1.8560000000000002e-05, "loss": 73.75, "step": 1857 }, { "epoch": 0.0001858, "grad_norm": 10.511929512023926, "learning_rate": 1.857e-05, "loss": 74.25, "step": 1858 }, { "epoch": 0.0001859, "grad_norm": 10.316984176635742, "learning_rate": 1.858e-05, "loss": 73.8125, "step": 1859 }, { "epoch": 0.000186, "grad_norm": 10.564013481140137, "learning_rate": 1.859e-05, "loss": 74.375, "step": 1860 }, { "epoch": 0.0001861, "grad_norm": 10.520541191101074, "learning_rate": 1.8599999999999998e-05, "loss": 72.9375, "step": 1861 }, { "epoch": 0.0001862, "grad_norm": 10.493901252746582, "learning_rate": 1.861e-05, "loss": 73.5, "step": 1862 }, { "epoch": 0.0001863, "grad_norm": 10.289846420288086, "learning_rate": 1.862e-05, "loss": 74.5, "step": 1863 }, { "epoch": 0.0001864, "grad_norm": 10.55647087097168, "learning_rate": 1.863e-05, "loss": 72.5625, "step": 1864 }, { "epoch": 0.0001865, "grad_norm": 10.611162185668945, "learning_rate": 1.864e-05, "loss": 73.1875, "step": 1865 }, { "epoch": 0.0001866, "grad_norm": 10.35433292388916, "learning_rate": 1.865e-05, "loss": 73.3125, "step": 1866 }, { "epoch": 0.0001867, "grad_norm": 10.491514205932617, "learning_rate": 1.866e-05, "loss": 72.1875, "step": 1867 }, { "epoch": 0.0001868, "grad_norm": 10.399983406066895, "learning_rate": 1.867e-05, "loss": 74.4375, "step": 1868 }, { "epoch": 0.0001869, "grad_norm": 10.26793098449707, "learning_rate": 1.868e-05, "loss": 75.6875, "step": 1869 }, { "epoch": 0.000187, "grad_norm": 10.247400283813477, "learning_rate": 1.8690000000000002e-05, "loss": 74.1875, "step": 1870 }, { "epoch": 0.0001871, "grad_norm": 10.544309616088867, "learning_rate": 1.87e-05, "loss": 71.75, "step": 1871 }, { "epoch": 0.0001872, "grad_norm": 10.499302864074707, "learning_rate": 1.871e-05, "loss": 72.0625, "step": 1872 }, { "epoch": 0.0001873, "grad_norm": 10.524739265441895, "learning_rate": 1.872e-05, "loss": 74.0, "step": 1873 }, { "epoch": 0.0001874, "grad_norm": 10.458234786987305, "learning_rate": 1.8730000000000002e-05, "loss": 72.9375, "step": 1874 }, { "epoch": 0.0001875, "grad_norm": 10.530044555664062, "learning_rate": 1.874e-05, "loss": 73.9375, "step": 1875 }, { "epoch": 0.0001876, "grad_norm": 10.297721862792969, "learning_rate": 1.8750000000000002e-05, "loss": 74.1875, "step": 1876 }, { "epoch": 0.0001877, "grad_norm": 10.372282981872559, "learning_rate": 1.876e-05, "loss": 74.25, "step": 1877 }, { "epoch": 0.0001878, "grad_norm": 10.302542686462402, "learning_rate": 1.8769999999999998e-05, "loss": 73.6875, "step": 1878 }, { "epoch": 0.0001879, "grad_norm": 10.455268859863281, "learning_rate": 1.878e-05, "loss": 74.4375, "step": 1879 }, { "epoch": 0.000188, "grad_norm": 10.465432167053223, "learning_rate": 1.879e-05, "loss": 71.75, "step": 1880 }, { "epoch": 0.0001881, "grad_norm": 10.485514640808105, "learning_rate": 1.88e-05, "loss": 73.6875, "step": 1881 }, { "epoch": 0.0001882, "grad_norm": 10.271815299987793, "learning_rate": 1.881e-05, "loss": 74.125, "step": 1882 }, { "epoch": 0.0001883, "grad_norm": 10.41010856628418, "learning_rate": 1.882e-05, "loss": 72.875, "step": 1883 }, { "epoch": 0.0001884, "grad_norm": 10.436948776245117, "learning_rate": 1.883e-05, "loss": 73.5, "step": 1884 }, { "epoch": 0.0001885, "grad_norm": 10.39260482788086, "learning_rate": 1.884e-05, "loss": 72.75, "step": 1885 }, { "epoch": 0.0001886, "grad_norm": 10.365702629089355, "learning_rate": 1.885e-05, "loss": 72.875, "step": 1886 }, { "epoch": 0.0001887, "grad_norm": 10.395852088928223, "learning_rate": 1.8860000000000002e-05, "loss": 71.6875, "step": 1887 }, { "epoch": 0.0001888, "grad_norm": 10.262138366699219, "learning_rate": 1.887e-05, "loss": 72.75, "step": 1888 }, { "epoch": 0.0001889, "grad_norm": 10.43374252319336, "learning_rate": 1.888e-05, "loss": 72.875, "step": 1889 }, { "epoch": 0.000189, "grad_norm": 10.227705001831055, "learning_rate": 1.889e-05, "loss": 73.875, "step": 1890 }, { "epoch": 0.0001891, "grad_norm": 10.353809356689453, "learning_rate": 1.8900000000000002e-05, "loss": 73.125, "step": 1891 }, { "epoch": 0.0001892, "grad_norm": 10.32115364074707, "learning_rate": 1.891e-05, "loss": 73.6875, "step": 1892 }, { "epoch": 0.0001893, "grad_norm": 10.273197174072266, "learning_rate": 1.8920000000000002e-05, "loss": 72.5625, "step": 1893 }, { "epoch": 0.0001894, "grad_norm": 10.563416481018066, "learning_rate": 1.893e-05, "loss": 73.1875, "step": 1894 }, { "epoch": 0.0001895, "grad_norm": 10.399928092956543, "learning_rate": 1.8939999999999998e-05, "loss": 71.4375, "step": 1895 }, { "epoch": 0.0001896, "grad_norm": 10.268082618713379, "learning_rate": 1.895e-05, "loss": 74.1875, "step": 1896 }, { "epoch": 0.0001897, "grad_norm": 10.20687484741211, "learning_rate": 1.896e-05, "loss": 73.0, "step": 1897 }, { "epoch": 0.0001898, "grad_norm": 10.29458999633789, "learning_rate": 1.897e-05, "loss": 72.5, "step": 1898 }, { "epoch": 0.0001899, "grad_norm": 10.40804672241211, "learning_rate": 1.898e-05, "loss": 72.3125, "step": 1899 }, { "epoch": 0.00019, "grad_norm": 10.445226669311523, "learning_rate": 1.899e-05, "loss": 71.25, "step": 1900 }, { "epoch": 0.0001901, "grad_norm": 10.368699073791504, "learning_rate": 1.9e-05, "loss": 71.75, "step": 1901 }, { "epoch": 0.0001902, "grad_norm": 10.655980110168457, "learning_rate": 1.901e-05, "loss": 71.375, "step": 1902 }, { "epoch": 0.0001903, "grad_norm": 10.2596435546875, "learning_rate": 1.902e-05, "loss": 72.5, "step": 1903 }, { "epoch": 0.0001904, "grad_norm": 10.201349258422852, "learning_rate": 1.9030000000000002e-05, "loss": 72.625, "step": 1904 }, { "epoch": 0.0001905, "grad_norm": 10.324286460876465, "learning_rate": 1.904e-05, "loss": 73.375, "step": 1905 }, { "epoch": 0.0001906, "grad_norm": 10.320876121520996, "learning_rate": 1.905e-05, "loss": 71.875, "step": 1906 }, { "epoch": 0.0001907, "grad_norm": 10.326148986816406, "learning_rate": 1.906e-05, "loss": 73.5, "step": 1907 }, { "epoch": 0.0001908, "grad_norm": 10.288737297058105, "learning_rate": 1.9070000000000002e-05, "loss": 72.125, "step": 1908 }, { "epoch": 0.0001909, "grad_norm": 10.349129676818848, "learning_rate": 1.908e-05, "loss": 72.875, "step": 1909 }, { "epoch": 0.000191, "grad_norm": 10.330138206481934, "learning_rate": 1.9090000000000002e-05, "loss": 73.0625, "step": 1910 }, { "epoch": 0.0001911, "grad_norm": 10.37156867980957, "learning_rate": 1.91e-05, "loss": 71.125, "step": 1911 }, { "epoch": 0.0001912, "grad_norm": 10.332508087158203, "learning_rate": 1.9109999999999998e-05, "loss": 71.8125, "step": 1912 }, { "epoch": 0.0001913, "grad_norm": 10.3118257522583, "learning_rate": 1.912e-05, "loss": 73.25, "step": 1913 }, { "epoch": 0.0001914, "grad_norm": 10.078189849853516, "learning_rate": 1.913e-05, "loss": 72.6875, "step": 1914 }, { "epoch": 0.0001915, "grad_norm": 10.437750816345215, "learning_rate": 1.914e-05, "loss": 72.1875, "step": 1915 }, { "epoch": 0.0001916, "grad_norm": 10.279887199401855, "learning_rate": 1.915e-05, "loss": 71.5625, "step": 1916 }, { "epoch": 0.0001917, "grad_norm": 10.300010681152344, "learning_rate": 1.916e-05, "loss": 71.25, "step": 1917 }, { "epoch": 0.0001918, "grad_norm": 10.136702537536621, "learning_rate": 1.917e-05, "loss": 73.875, "step": 1918 }, { "epoch": 0.0001919, "grad_norm": 10.21435260772705, "learning_rate": 1.918e-05, "loss": 71.75, "step": 1919 }, { "epoch": 0.000192, "grad_norm": 10.43912410736084, "learning_rate": 1.919e-05, "loss": 70.25, "step": 1920 }, { "epoch": 0.0001921, "grad_norm": 10.203756332397461, "learning_rate": 1.9200000000000003e-05, "loss": 71.75, "step": 1921 }, { "epoch": 0.0001922, "grad_norm": 10.295417785644531, "learning_rate": 1.921e-05, "loss": 72.1875, "step": 1922 }, { "epoch": 0.0001923, "grad_norm": 10.405128479003906, "learning_rate": 1.922e-05, "loss": 71.5, "step": 1923 }, { "epoch": 0.0001924, "grad_norm": 10.384449005126953, "learning_rate": 1.923e-05, "loss": 72.8125, "step": 1924 }, { "epoch": 0.0001925, "grad_norm": 10.155817031860352, "learning_rate": 1.9240000000000002e-05, "loss": 72.1875, "step": 1925 }, { "epoch": 0.0001926, "grad_norm": 10.310267448425293, "learning_rate": 1.925e-05, "loss": 72.1875, "step": 1926 }, { "epoch": 0.0001927, "grad_norm": 10.375925064086914, "learning_rate": 1.9260000000000002e-05, "loss": 70.75, "step": 1927 }, { "epoch": 0.0001928, "grad_norm": 10.55906867980957, "learning_rate": 1.927e-05, "loss": 73.375, "step": 1928 }, { "epoch": 0.0001929, "grad_norm": 10.122315406799316, "learning_rate": 1.928e-05, "loss": 74.625, "step": 1929 }, { "epoch": 0.000193, "grad_norm": 10.311563491821289, "learning_rate": 1.929e-05, "loss": 72.0625, "step": 1930 }, { "epoch": 0.0001931, "grad_norm": 10.309248924255371, "learning_rate": 1.93e-05, "loss": 72.6875, "step": 1931 }, { "epoch": 0.0001932, "grad_norm": 10.181316375732422, "learning_rate": 1.931e-05, "loss": 70.6875, "step": 1932 }, { "epoch": 0.0001933, "grad_norm": 10.316238403320312, "learning_rate": 1.932e-05, "loss": 71.5625, "step": 1933 }, { "epoch": 0.0001934, "grad_norm": 10.179081916809082, "learning_rate": 1.933e-05, "loss": 72.4375, "step": 1934 }, { "epoch": 0.0001935, "grad_norm": 10.145713806152344, "learning_rate": 1.934e-05, "loss": 71.75, "step": 1935 }, { "epoch": 0.0001936, "grad_norm": 10.175612449645996, "learning_rate": 1.935e-05, "loss": 71.5, "step": 1936 }, { "epoch": 0.0001937, "grad_norm": 10.24749755859375, "learning_rate": 1.936e-05, "loss": 72.5625, "step": 1937 }, { "epoch": 0.0001938, "grad_norm": 10.308296203613281, "learning_rate": 1.9370000000000003e-05, "loss": 70.5, "step": 1938 }, { "epoch": 0.0001939, "grad_norm": 10.151104927062988, "learning_rate": 1.938e-05, "loss": 71.3125, "step": 1939 }, { "epoch": 0.000194, "grad_norm": 10.346338272094727, "learning_rate": 1.939e-05, "loss": 71.625, "step": 1940 }, { "epoch": 0.0001941, "grad_norm": 10.234703063964844, "learning_rate": 1.94e-05, "loss": 70.4375, "step": 1941 }, { "epoch": 0.0001942, "grad_norm": 10.15624713897705, "learning_rate": 1.9410000000000002e-05, "loss": 72.5, "step": 1942 }, { "epoch": 0.0001943, "grad_norm": 10.2576265335083, "learning_rate": 1.942e-05, "loss": 71.8125, "step": 1943 }, { "epoch": 0.0001944, "grad_norm": 10.192061424255371, "learning_rate": 1.9430000000000002e-05, "loss": 71.625, "step": 1944 }, { "epoch": 0.0001945, "grad_norm": 10.456489562988281, "learning_rate": 1.944e-05, "loss": 69.75, "step": 1945 }, { "epoch": 0.0001946, "grad_norm": 10.204558372497559, "learning_rate": 1.945e-05, "loss": 71.3125, "step": 1946 }, { "epoch": 0.0001947, "grad_norm": 10.194465637207031, "learning_rate": 1.946e-05, "loss": 70.25, "step": 1947 }, { "epoch": 0.0001948, "grad_norm": 10.135470390319824, "learning_rate": 1.947e-05, "loss": 71.4375, "step": 1948 }, { "epoch": 0.0001949, "grad_norm": 10.270816802978516, "learning_rate": 1.948e-05, "loss": 72.8125, "step": 1949 }, { "epoch": 0.000195, "grad_norm": 10.145831108093262, "learning_rate": 1.949e-05, "loss": 73.25, "step": 1950 }, { "epoch": 0.0001951, "grad_norm": 10.1314115524292, "learning_rate": 1.95e-05, "loss": 71.4375, "step": 1951 }, { "epoch": 0.0001952, "grad_norm": 10.260855674743652, "learning_rate": 1.951e-05, "loss": 70.0625, "step": 1952 }, { "epoch": 0.0001953, "grad_norm": 10.306357383728027, "learning_rate": 1.952e-05, "loss": 69.9375, "step": 1953 }, { "epoch": 0.0001954, "grad_norm": 10.049440383911133, "learning_rate": 1.953e-05, "loss": 72.5625, "step": 1954 }, { "epoch": 0.0001955, "grad_norm": 10.397737503051758, "learning_rate": 1.9540000000000003e-05, "loss": 70.75, "step": 1955 }, { "epoch": 0.0001956, "grad_norm": 10.084511756896973, "learning_rate": 1.955e-05, "loss": 70.3125, "step": 1956 }, { "epoch": 0.0001957, "grad_norm": 10.107562065124512, "learning_rate": 1.956e-05, "loss": 71.9375, "step": 1957 }, { "epoch": 0.0001958, "grad_norm": 10.259406089782715, "learning_rate": 1.957e-05, "loss": 71.6875, "step": 1958 }, { "epoch": 0.0001959, "grad_norm": 10.093414306640625, "learning_rate": 1.9580000000000002e-05, "loss": 72.5, "step": 1959 }, { "epoch": 0.000196, "grad_norm": 10.273123741149902, "learning_rate": 1.959e-05, "loss": 69.75, "step": 1960 }, { "epoch": 0.0001961, "grad_norm": 10.390420913696289, "learning_rate": 1.9600000000000002e-05, "loss": 70.6875, "step": 1961 }, { "epoch": 0.0001962, "grad_norm": 10.27889633178711, "learning_rate": 1.961e-05, "loss": 70.4375, "step": 1962 }, { "epoch": 0.0001963, "grad_norm": 10.058853149414062, "learning_rate": 1.962e-05, "loss": 71.5, "step": 1963 }, { "epoch": 0.0001964, "grad_norm": 10.128463745117188, "learning_rate": 1.963e-05, "loss": 73.125, "step": 1964 }, { "epoch": 0.0001965, "grad_norm": 10.429844856262207, "learning_rate": 1.9640000000000002e-05, "loss": 69.5625, "step": 1965 }, { "epoch": 0.0001966, "grad_norm": 10.319170951843262, "learning_rate": 1.965e-05, "loss": 69.25, "step": 1966 }, { "epoch": 0.0001967, "grad_norm": 10.247960090637207, "learning_rate": 1.966e-05, "loss": 70.5, "step": 1967 }, { "epoch": 0.0001968, "grad_norm": 10.223762512207031, "learning_rate": 1.967e-05, "loss": 68.75, "step": 1968 }, { "epoch": 0.0001969, "grad_norm": 10.190348625183105, "learning_rate": 1.968e-05, "loss": 69.1875, "step": 1969 }, { "epoch": 0.000197, "grad_norm": 10.337133407592773, "learning_rate": 1.969e-05, "loss": 70.5625, "step": 1970 }, { "epoch": 0.0001971, "grad_norm": 10.41808032989502, "learning_rate": 1.97e-05, "loss": 69.125, "step": 1971 }, { "epoch": 0.0001972, "grad_norm": 9.977437019348145, "learning_rate": 1.9710000000000003e-05, "loss": 71.375, "step": 1972 }, { "epoch": 0.0001973, "grad_norm": 10.139707565307617, "learning_rate": 1.9719999999999998e-05, "loss": 71.4375, "step": 1973 }, { "epoch": 0.0001974, "grad_norm": 10.059530258178711, "learning_rate": 1.973e-05, "loss": 71.875, "step": 1974 }, { "epoch": 0.0001975, "grad_norm": 10.154842376708984, "learning_rate": 1.974e-05, "loss": 69.875, "step": 1975 }, { "epoch": 0.0001976, "grad_norm": 10.15227222442627, "learning_rate": 1.975e-05, "loss": 69.5, "step": 1976 }, { "epoch": 0.0001977, "grad_norm": 10.316965103149414, "learning_rate": 1.976e-05, "loss": 69.5, "step": 1977 }, { "epoch": 0.0001978, "grad_norm": 10.136872291564941, "learning_rate": 1.9770000000000002e-05, "loss": 70.5, "step": 1978 }, { "epoch": 0.0001979, "grad_norm": 10.085726737976074, "learning_rate": 1.978e-05, "loss": 71.0625, "step": 1979 }, { "epoch": 0.000198, "grad_norm": 10.222604751586914, "learning_rate": 1.979e-05, "loss": 69.75, "step": 1980 }, { "epoch": 0.0001981, "grad_norm": 10.206459045410156, "learning_rate": 1.98e-05, "loss": 69.9375, "step": 1981 }, { "epoch": 0.0001982, "grad_norm": 9.955349922180176, "learning_rate": 1.9810000000000002e-05, "loss": 71.5625, "step": 1982 }, { "epoch": 0.0001983, "grad_norm": 10.17930793762207, "learning_rate": 1.982e-05, "loss": 70.75, "step": 1983 }, { "epoch": 0.0001984, "grad_norm": 10.088285446166992, "learning_rate": 1.983e-05, "loss": 71.0625, "step": 1984 }, { "epoch": 0.0001985, "grad_norm": 10.180354118347168, "learning_rate": 1.984e-05, "loss": 70.3125, "step": 1985 }, { "epoch": 0.0001986, "grad_norm": 10.134940147399902, "learning_rate": 1.985e-05, "loss": 69.625, "step": 1986 }, { "epoch": 0.0001987, "grad_norm": 10.140948295593262, "learning_rate": 1.986e-05, "loss": 71.25, "step": 1987 }, { "epoch": 0.0001988, "grad_norm": 10.105887413024902, "learning_rate": 1.987e-05, "loss": 69.125, "step": 1988 }, { "epoch": 0.0001989, "grad_norm": 10.131355285644531, "learning_rate": 1.9880000000000003e-05, "loss": 70.5, "step": 1989 }, { "epoch": 0.000199, "grad_norm": 10.018689155578613, "learning_rate": 1.9889999999999998e-05, "loss": 71.4375, "step": 1990 }, { "epoch": 0.0001991, "grad_norm": 10.111222267150879, "learning_rate": 1.99e-05, "loss": 69.5625, "step": 1991 }, { "epoch": 0.0001992, "grad_norm": 10.133853912353516, "learning_rate": 1.991e-05, "loss": 69.8125, "step": 1992 }, { "epoch": 0.0001993, "grad_norm": 10.056241989135742, "learning_rate": 1.992e-05, "loss": 69.0, "step": 1993 }, { "epoch": 0.0001994, "grad_norm": 10.109450340270996, "learning_rate": 1.993e-05, "loss": 68.75, "step": 1994 }, { "epoch": 0.0001995, "grad_norm": 10.147754669189453, "learning_rate": 1.9940000000000002e-05, "loss": 69.5625, "step": 1995 }, { "epoch": 0.0001996, "grad_norm": 10.053937911987305, "learning_rate": 1.995e-05, "loss": 69.625, "step": 1996 }, { "epoch": 0.0001997, "grad_norm": 10.346978187561035, "learning_rate": 1.996e-05, "loss": 69.125, "step": 1997 }, { "epoch": 0.0001998, "grad_norm": 10.11376667022705, "learning_rate": 1.997e-05, "loss": 68.6875, "step": 1998 }, { "epoch": 0.0001999, "grad_norm": 10.165390968322754, "learning_rate": 1.9980000000000002e-05, "loss": 68.25, "step": 1999 }, { "epoch": 0.0002, "grad_norm": 10.25505256652832, "learning_rate": 1.999e-05, "loss": 70.4375, "step": 2000 }, { "epoch": 0.0002, "eval_loss": 8.524580001831055, "eval_runtime": 403.2967, "eval_samples_per_second": 24.796, "eval_steps_per_second": 1.55, "step": 2000 }, { "epoch": 0.0002001, "grad_norm": 9.967585563659668, "learning_rate": 2e-05, "loss": 71.0, "step": 2001 }, { "epoch": 0.0002002, "grad_norm": 10.217305183410645, "learning_rate": 2.001e-05, "loss": 69.875, "step": 2002 }, { "epoch": 0.0002003, "grad_norm": 10.10339641571045, "learning_rate": 2.002e-05, "loss": 69.5, "step": 2003 }, { "epoch": 0.0002004, "grad_norm": 10.115350723266602, "learning_rate": 2.003e-05, "loss": 71.8125, "step": 2004 }, { "epoch": 0.0002005, "grad_norm": 10.16146183013916, "learning_rate": 2.004e-05, "loss": 68.0, "step": 2005 }, { "epoch": 0.0002006, "grad_norm": 10.037304878234863, "learning_rate": 2.0050000000000003e-05, "loss": 69.625, "step": 2006 }, { "epoch": 0.0002007, "grad_norm": 10.476919174194336, "learning_rate": 2.0059999999999998e-05, "loss": 69.3125, "step": 2007 }, { "epoch": 0.0002008, "grad_norm": 9.933809280395508, "learning_rate": 2.007e-05, "loss": 68.9375, "step": 2008 }, { "epoch": 0.0002009, "grad_norm": 10.033228874206543, "learning_rate": 2.008e-05, "loss": 68.6875, "step": 2009 }, { "epoch": 0.000201, "grad_norm": 9.955198287963867, "learning_rate": 2.009e-05, "loss": 69.8125, "step": 2010 }, { "epoch": 0.0002011, "grad_norm": 10.012125015258789, "learning_rate": 2.01e-05, "loss": 70.125, "step": 2011 }, { "epoch": 0.0002012, "grad_norm": 10.269689559936523, "learning_rate": 2.0110000000000002e-05, "loss": 66.5625, "step": 2012 }, { "epoch": 0.0002013, "grad_norm": 10.17927360534668, "learning_rate": 2.012e-05, "loss": 67.9375, "step": 2013 }, { "epoch": 0.0002014, "grad_norm": 10.166531562805176, "learning_rate": 2.013e-05, "loss": 67.5, "step": 2014 }, { "epoch": 0.0002015, "grad_norm": 9.982065200805664, "learning_rate": 2.014e-05, "loss": 69.0625, "step": 2015 }, { "epoch": 0.0002016, "grad_norm": 10.05953598022461, "learning_rate": 2.0150000000000002e-05, "loss": 67.75, "step": 2016 }, { "epoch": 0.0002017, "grad_norm": 10.072693824768066, "learning_rate": 2.016e-05, "loss": 68.6875, "step": 2017 }, { "epoch": 0.0002018, "grad_norm": 10.192292213439941, "learning_rate": 2.0170000000000002e-05, "loss": 67.9375, "step": 2018 }, { "epoch": 0.0002019, "grad_norm": 10.189484596252441, "learning_rate": 2.018e-05, "loss": 68.1875, "step": 2019 }, { "epoch": 0.000202, "grad_norm": 10.200225830078125, "learning_rate": 2.019e-05, "loss": 67.5, "step": 2020 }, { "epoch": 0.0002021, "grad_norm": 9.984915733337402, "learning_rate": 2.02e-05, "loss": 67.375, "step": 2021 }, { "epoch": 0.0002022, "grad_norm": 10.250639915466309, "learning_rate": 2.021e-05, "loss": 69.5625, "step": 2022 }, { "epoch": 0.0002023, "grad_norm": 10.15843391418457, "learning_rate": 2.0220000000000003e-05, "loss": 68.9375, "step": 2023 }, { "epoch": 0.0002024, "grad_norm": 9.864794731140137, "learning_rate": 2.0229999999999998e-05, "loss": 69.4375, "step": 2024 }, { "epoch": 0.0002025, "grad_norm": 10.130589485168457, "learning_rate": 2.024e-05, "loss": 66.75, "step": 2025 }, { "epoch": 0.0002026, "grad_norm": 9.870321273803711, "learning_rate": 2.025e-05, "loss": 69.5, "step": 2026 }, { "epoch": 0.0002027, "grad_norm": 10.216444969177246, "learning_rate": 2.026e-05, "loss": 68.0625, "step": 2027 }, { "epoch": 0.0002028, "grad_norm": 10.087665557861328, "learning_rate": 2.027e-05, "loss": 67.125, "step": 2028 }, { "epoch": 0.0002029, "grad_norm": 10.021193504333496, "learning_rate": 2.0280000000000002e-05, "loss": 67.4375, "step": 2029 }, { "epoch": 0.000203, "grad_norm": 9.984146118164062, "learning_rate": 2.029e-05, "loss": 67.8125, "step": 2030 }, { "epoch": 0.0002031, "grad_norm": 10.092177391052246, "learning_rate": 2.03e-05, "loss": 67.5, "step": 2031 }, { "epoch": 0.0002032, "grad_norm": 10.105181694030762, "learning_rate": 2.031e-05, "loss": 67.125, "step": 2032 }, { "epoch": 0.0002033, "grad_norm": 10.0947847366333, "learning_rate": 2.0320000000000002e-05, "loss": 68.75, "step": 2033 }, { "epoch": 0.0002034, "grad_norm": 10.02136516571045, "learning_rate": 2.033e-05, "loss": 68.4375, "step": 2034 }, { "epoch": 0.0002035, "grad_norm": 9.922673225402832, "learning_rate": 2.034e-05, "loss": 68.0625, "step": 2035 }, { "epoch": 0.0002036, "grad_norm": 9.925615310668945, "learning_rate": 2.035e-05, "loss": 67.9375, "step": 2036 }, { "epoch": 0.0002037, "grad_norm": 10.053800582885742, "learning_rate": 2.036e-05, "loss": 67.0312, "step": 2037 }, { "epoch": 0.0002038, "grad_norm": 9.972097396850586, "learning_rate": 2.037e-05, "loss": 66.875, "step": 2038 }, { "epoch": 0.0002039, "grad_norm": 10.238794326782227, "learning_rate": 2.038e-05, "loss": 66.1875, "step": 2039 }, { "epoch": 0.000204, "grad_norm": 9.922769546508789, "learning_rate": 2.0390000000000003e-05, "loss": 67.9375, "step": 2040 }, { "epoch": 0.0002041, "grad_norm": 9.971976280212402, "learning_rate": 2.0399999999999998e-05, "loss": 69.1875, "step": 2041 }, { "epoch": 0.0002042, "grad_norm": 10.186835289001465, "learning_rate": 2.041e-05, "loss": 67.4688, "step": 2042 }, { "epoch": 0.0002043, "grad_norm": 9.938898086547852, "learning_rate": 2.042e-05, "loss": 67.75, "step": 2043 }, { "epoch": 0.0002044, "grad_norm": 9.898589134216309, "learning_rate": 2.043e-05, "loss": 68.0625, "step": 2044 }, { "epoch": 0.0002045, "grad_norm": 9.993139266967773, "learning_rate": 2.044e-05, "loss": 66.8438, "step": 2045 }, { "epoch": 0.0002046, "grad_norm": 9.870691299438477, "learning_rate": 2.0450000000000002e-05, "loss": 66.7812, "step": 2046 }, { "epoch": 0.0002047, "grad_norm": 10.106609344482422, "learning_rate": 2.046e-05, "loss": 66.125, "step": 2047 }, { "epoch": 0.0002048, "grad_norm": 9.97212028503418, "learning_rate": 2.047e-05, "loss": 66.6875, "step": 2048 }, { "epoch": 0.0002049, "grad_norm": 10.094063758850098, "learning_rate": 2.048e-05, "loss": 65.375, "step": 2049 }, { "epoch": 0.000205, "grad_norm": 10.00854206085205, "learning_rate": 2.0490000000000002e-05, "loss": 67.1875, "step": 2050 }, { "epoch": 0.0002051, "grad_norm": 9.906662940979004, "learning_rate": 2.05e-05, "loss": 66.8438, "step": 2051 }, { "epoch": 0.0002052, "grad_norm": 9.8468599319458, "learning_rate": 2.051e-05, "loss": 67.8125, "step": 2052 }, { "epoch": 0.0002053, "grad_norm": 9.83792495727539, "learning_rate": 2.052e-05, "loss": 66.75, "step": 2053 }, { "epoch": 0.0002054, "grad_norm": 9.971479415893555, "learning_rate": 2.053e-05, "loss": 67.9375, "step": 2054 }, { "epoch": 0.0002055, "grad_norm": 9.875134468078613, "learning_rate": 2.054e-05, "loss": 66.3125, "step": 2055 }, { "epoch": 0.0002056, "grad_norm": 9.889384269714355, "learning_rate": 2.055e-05, "loss": 65.9062, "step": 2056 }, { "epoch": 0.0002057, "grad_norm": 10.013370513916016, "learning_rate": 2.0560000000000003e-05, "loss": 68.5, "step": 2057 }, { "epoch": 0.0002058, "grad_norm": 9.89936351776123, "learning_rate": 2.0569999999999998e-05, "loss": 66.625, "step": 2058 }, { "epoch": 0.0002059, "grad_norm": 9.941503524780273, "learning_rate": 2.058e-05, "loss": 66.9375, "step": 2059 }, { "epoch": 0.000206, "grad_norm": 9.911052703857422, "learning_rate": 2.059e-05, "loss": 67.75, "step": 2060 }, { "epoch": 0.0002061, "grad_norm": 9.892444610595703, "learning_rate": 2.06e-05, "loss": 66.2188, "step": 2061 }, { "epoch": 0.0002062, "grad_norm": 9.815709114074707, "learning_rate": 2.061e-05, "loss": 66.9688, "step": 2062 }, { "epoch": 0.0002063, "grad_norm": 9.808388710021973, "learning_rate": 2.0620000000000002e-05, "loss": 66.9375, "step": 2063 }, { "epoch": 0.0002064, "grad_norm": 9.86568832397461, "learning_rate": 2.063e-05, "loss": 65.6875, "step": 2064 }, { "epoch": 0.0002065, "grad_norm": 9.770045280456543, "learning_rate": 2.064e-05, "loss": 66.5625, "step": 2065 }, { "epoch": 0.0002066, "grad_norm": 9.898188591003418, "learning_rate": 2.065e-05, "loss": 66.25, "step": 2066 }, { "epoch": 0.0002067, "grad_norm": 9.870433807373047, "learning_rate": 2.0660000000000002e-05, "loss": 65.4688, "step": 2067 }, { "epoch": 0.0002068, "grad_norm": 9.64307975769043, "learning_rate": 2.067e-05, "loss": 67.375, "step": 2068 }, { "epoch": 0.0002069, "grad_norm": 9.807807922363281, "learning_rate": 2.068e-05, "loss": 66.125, "step": 2069 }, { "epoch": 0.000207, "grad_norm": 9.7689208984375, "learning_rate": 2.069e-05, "loss": 65.9688, "step": 2070 }, { "epoch": 0.0002071, "grad_norm": 10.117674827575684, "learning_rate": 2.0700000000000002e-05, "loss": 66.4688, "step": 2071 }, { "epoch": 0.0002072, "grad_norm": 9.678217887878418, "learning_rate": 2.071e-05, "loss": 67.375, "step": 2072 }, { "epoch": 0.0002073, "grad_norm": 9.870391845703125, "learning_rate": 2.072e-05, "loss": 65.8125, "step": 2073 }, { "epoch": 0.0002074, "grad_norm": 9.78575611114502, "learning_rate": 2.0730000000000003e-05, "loss": 65.9062, "step": 2074 }, { "epoch": 0.0002075, "grad_norm": 9.769021034240723, "learning_rate": 2.0739999999999998e-05, "loss": 65.8438, "step": 2075 }, { "epoch": 0.0002076, "grad_norm": 9.902575492858887, "learning_rate": 2.075e-05, "loss": 64.4375, "step": 2076 }, { "epoch": 0.0002077, "grad_norm": 9.719600677490234, "learning_rate": 2.076e-05, "loss": 67.375, "step": 2077 }, { "epoch": 0.0002078, "grad_norm": 9.668708801269531, "learning_rate": 2.077e-05, "loss": 67.5, "step": 2078 }, { "epoch": 0.0002079, "grad_norm": 9.731209754943848, "learning_rate": 2.078e-05, "loss": 64.25, "step": 2079 }, { "epoch": 0.000208, "grad_norm": 9.743050575256348, "learning_rate": 2.0790000000000003e-05, "loss": 65.5312, "step": 2080 }, { "epoch": 0.0002081, "grad_norm": 9.843193054199219, "learning_rate": 2.08e-05, "loss": 65.8125, "step": 2081 }, { "epoch": 0.0002082, "grad_norm": 9.641849517822266, "learning_rate": 2.081e-05, "loss": 65.7188, "step": 2082 }, { "epoch": 0.0002083, "grad_norm": 9.820322036743164, "learning_rate": 2.082e-05, "loss": 64.7812, "step": 2083 }, { "epoch": 0.0002084, "grad_norm": 9.788528442382812, "learning_rate": 2.0830000000000002e-05, "loss": 65.4062, "step": 2084 }, { "epoch": 0.0002085, "grad_norm": 9.630453109741211, "learning_rate": 2.084e-05, "loss": 65.25, "step": 2085 }, { "epoch": 0.0002086, "grad_norm": 9.939208984375, "learning_rate": 2.085e-05, "loss": 66.875, "step": 2086 }, { "epoch": 0.0002087, "grad_norm": 9.670159339904785, "learning_rate": 2.086e-05, "loss": 64.1562, "step": 2087 }, { "epoch": 0.0002088, "grad_norm": 9.668744087219238, "learning_rate": 2.0870000000000002e-05, "loss": 65.3125, "step": 2088 }, { "epoch": 0.0002089, "grad_norm": 9.899395942687988, "learning_rate": 2.088e-05, "loss": 64.4688, "step": 2089 }, { "epoch": 0.000209, "grad_norm": 9.995280265808105, "learning_rate": 2.089e-05, "loss": 64.875, "step": 2090 }, { "epoch": 0.0002091, "grad_norm": 9.69394302368164, "learning_rate": 2.09e-05, "loss": 65.875, "step": 2091 }, { "epoch": 0.0002092, "grad_norm": 9.688862800598145, "learning_rate": 2.0909999999999998e-05, "loss": 65.6562, "step": 2092 }, { "epoch": 0.0002093, "grad_norm": 9.628240585327148, "learning_rate": 2.092e-05, "loss": 65.1875, "step": 2093 }, { "epoch": 0.0002094, "grad_norm": 9.802079200744629, "learning_rate": 2.093e-05, "loss": 63.8438, "step": 2094 }, { "epoch": 0.0002095, "grad_norm": 9.755867004394531, "learning_rate": 2.094e-05, "loss": 63.9062, "step": 2095 }, { "epoch": 0.0002096, "grad_norm": 9.552199363708496, "learning_rate": 2.095e-05, "loss": 63.7812, "step": 2096 }, { "epoch": 0.0002097, "grad_norm": 9.68614387512207, "learning_rate": 2.0960000000000003e-05, "loss": 63.8438, "step": 2097 }, { "epoch": 0.0002098, "grad_norm": 9.63162612915039, "learning_rate": 2.097e-05, "loss": 63.4062, "step": 2098 }, { "epoch": 0.0002099, "grad_norm": 9.658440589904785, "learning_rate": 2.098e-05, "loss": 64.25, "step": 2099 }, { "epoch": 0.00021, "grad_norm": 9.692501068115234, "learning_rate": 2.099e-05, "loss": 63.6875, "step": 2100 }, { "epoch": 0.0002101, "grad_norm": 9.696625709533691, "learning_rate": 2.1000000000000002e-05, "loss": 63.6875, "step": 2101 }, { "epoch": 0.0002102, "grad_norm": 9.643299102783203, "learning_rate": 2.101e-05, "loss": 63.8438, "step": 2102 }, { "epoch": 0.0002103, "grad_norm": 9.580775260925293, "learning_rate": 2.102e-05, "loss": 61.4688, "step": 2103 }, { "epoch": 0.0002104, "grad_norm": 9.692644119262695, "learning_rate": 2.103e-05, "loss": 61.8125, "step": 2104 }, { "epoch": 0.0002105, "grad_norm": 9.629671096801758, "learning_rate": 2.104e-05, "loss": 63.4062, "step": 2105 }, { "epoch": 0.0002106, "grad_norm": 9.542960166931152, "learning_rate": 2.105e-05, "loss": 63.4375, "step": 2106 }, { "epoch": 0.0002107, "grad_norm": 9.571887969970703, "learning_rate": 2.106e-05, "loss": 64.4375, "step": 2107 }, { "epoch": 0.0002108, "grad_norm": 9.673304557800293, "learning_rate": 2.107e-05, "loss": 62.6562, "step": 2108 }, { "epoch": 0.0002109, "grad_norm": 9.493108749389648, "learning_rate": 2.1079999999999998e-05, "loss": 62.75, "step": 2109 }, { "epoch": 0.000211, "grad_norm": 9.350164413452148, "learning_rate": 2.109e-05, "loss": 63.2188, "step": 2110 }, { "epoch": 0.0002111, "grad_norm": 9.659684181213379, "learning_rate": 2.11e-05, "loss": 61.8438, "step": 2111 }, { "epoch": 0.0002112, "grad_norm": 9.644436836242676, "learning_rate": 2.111e-05, "loss": 61.0312, "step": 2112 }, { "epoch": 0.0002113, "grad_norm": 9.57265567779541, "learning_rate": 2.112e-05, "loss": 63.9375, "step": 2113 }, { "epoch": 0.0002114, "grad_norm": 9.555940628051758, "learning_rate": 2.1130000000000003e-05, "loss": 62.1875, "step": 2114 }, { "epoch": 0.0002115, "grad_norm": 9.480903625488281, "learning_rate": 2.114e-05, "loss": 61.5312, "step": 2115 }, { "epoch": 0.0002116, "grad_norm": 9.366453170776367, "learning_rate": 2.115e-05, "loss": 61.7188, "step": 2116 }, { "epoch": 0.0002117, "grad_norm": 9.573132514953613, "learning_rate": 2.116e-05, "loss": 61.5938, "step": 2117 }, { "epoch": 0.0002118, "grad_norm": 9.541674613952637, "learning_rate": 2.1170000000000002e-05, "loss": 61.625, "step": 2118 }, { "epoch": 0.0002119, "grad_norm": 9.315372467041016, "learning_rate": 2.118e-05, "loss": 62.4375, "step": 2119 }, { "epoch": 0.000212, "grad_norm": 9.498926162719727, "learning_rate": 2.119e-05, "loss": 61.6875, "step": 2120 }, { "epoch": 0.0002121, "grad_norm": 9.630905151367188, "learning_rate": 2.12e-05, "loss": 60.3438, "step": 2121 }, { "epoch": 0.0002122, "grad_norm": 9.5501127243042, "learning_rate": 2.121e-05, "loss": 61.8125, "step": 2122 }, { "epoch": 0.0002123, "grad_norm": 9.418230056762695, "learning_rate": 2.122e-05, "loss": 60.9062, "step": 2123 }, { "epoch": 0.0002124, "grad_norm": 9.495058059692383, "learning_rate": 2.1230000000000002e-05, "loss": 60.8438, "step": 2124 }, { "epoch": 0.0002125, "grad_norm": 9.291244506835938, "learning_rate": 2.124e-05, "loss": 62.625, "step": 2125 }, { "epoch": 0.0002126, "grad_norm": 9.416329383850098, "learning_rate": 2.1249999999999998e-05, "loss": 61.2812, "step": 2126 }, { "epoch": 0.0002127, "grad_norm": 9.386695861816406, "learning_rate": 2.126e-05, "loss": 61.2812, "step": 2127 }, { "epoch": 0.0002128, "grad_norm": 9.250471115112305, "learning_rate": 2.127e-05, "loss": 61.4062, "step": 2128 }, { "epoch": 0.0002129, "grad_norm": 9.453954696655273, "learning_rate": 2.128e-05, "loss": 61.7812, "step": 2129 }, { "epoch": 0.000213, "grad_norm": 9.354055404663086, "learning_rate": 2.129e-05, "loss": 60.375, "step": 2130 }, { "epoch": 0.0002131, "grad_norm": 9.544196128845215, "learning_rate": 2.1300000000000003e-05, "loss": 58.6562, "step": 2131 }, { "epoch": 0.0002132, "grad_norm": 9.431273460388184, "learning_rate": 2.131e-05, "loss": 60.0625, "step": 2132 }, { "epoch": 0.0002133, "grad_norm": 9.375019073486328, "learning_rate": 2.132e-05, "loss": 59.375, "step": 2133 }, { "epoch": 0.0002134, "grad_norm": 9.3251371383667, "learning_rate": 2.133e-05, "loss": 63.4062, "step": 2134 }, { "epoch": 0.0002135, "grad_norm": 9.357114791870117, "learning_rate": 2.1340000000000002e-05, "loss": 60.125, "step": 2135 }, { "epoch": 0.0002136, "grad_norm": 9.26671028137207, "learning_rate": 2.135e-05, "loss": 60.9688, "step": 2136 }, { "epoch": 0.0002137, "grad_norm": 9.199756622314453, "learning_rate": 2.136e-05, "loss": 60.125, "step": 2137 }, { "epoch": 0.0002138, "grad_norm": 9.27835464477539, "learning_rate": 2.137e-05, "loss": 59.5312, "step": 2138 }, { "epoch": 0.0002139, "grad_norm": 9.233546257019043, "learning_rate": 2.138e-05, "loss": 59.0312, "step": 2139 }, { "epoch": 0.000214, "grad_norm": 9.225489616394043, "learning_rate": 2.139e-05, "loss": 59.5, "step": 2140 }, { "epoch": 0.0002141, "grad_norm": 9.246992111206055, "learning_rate": 2.1400000000000002e-05, "loss": 59.0, "step": 2141 }, { "epoch": 0.0002142, "grad_norm": 9.223522186279297, "learning_rate": 2.141e-05, "loss": 59.2188, "step": 2142 }, { "epoch": 0.0002143, "grad_norm": 9.294316291809082, "learning_rate": 2.1419999999999998e-05, "loss": 59.2188, "step": 2143 }, { "epoch": 0.0002144, "grad_norm": 9.358565330505371, "learning_rate": 2.143e-05, "loss": 59.3438, "step": 2144 }, { "epoch": 0.0002145, "grad_norm": 9.189288139343262, "learning_rate": 2.144e-05, "loss": 59.5625, "step": 2145 }, { "epoch": 0.0002146, "grad_norm": 9.092613220214844, "learning_rate": 2.145e-05, "loss": 58.9375, "step": 2146 }, { "epoch": 0.0002147, "grad_norm": 9.041653633117676, "learning_rate": 2.146e-05, "loss": 59.4062, "step": 2147 }, { "epoch": 0.0002148, "grad_norm": 9.15941047668457, "learning_rate": 2.1470000000000003e-05, "loss": 58.4375, "step": 2148 }, { "epoch": 0.0002149, "grad_norm": 9.365605354309082, "learning_rate": 2.148e-05, "loss": 56.4375, "step": 2149 }, { "epoch": 0.000215, "grad_norm": 9.246635437011719, "learning_rate": 2.149e-05, "loss": 58.7812, "step": 2150 }, { "epoch": 0.0002151, "grad_norm": 9.01097297668457, "learning_rate": 2.15e-05, "loss": 58.4688, "step": 2151 }, { "epoch": 0.0002152, "grad_norm": 9.123529434204102, "learning_rate": 2.1510000000000002e-05, "loss": 59.2812, "step": 2152 }, { "epoch": 0.0002153, "grad_norm": 9.13985538482666, "learning_rate": 2.152e-05, "loss": 57.4375, "step": 2153 }, { "epoch": 0.0002154, "grad_norm": 9.151568412780762, "learning_rate": 2.153e-05, "loss": 58.7188, "step": 2154 }, { "epoch": 0.0002155, "grad_norm": 9.016154289245605, "learning_rate": 2.154e-05, "loss": 58.6875, "step": 2155 }, { "epoch": 0.0002156, "grad_norm": 9.112479209899902, "learning_rate": 2.155e-05, "loss": 58.0938, "step": 2156 }, { "epoch": 0.0002157, "grad_norm": 8.901529312133789, "learning_rate": 2.156e-05, "loss": 58.3125, "step": 2157 }, { "epoch": 0.0002158, "grad_norm": 9.176918029785156, "learning_rate": 2.1570000000000002e-05, "loss": 56.9688, "step": 2158 }, { "epoch": 0.0002159, "grad_norm": 8.917030334472656, "learning_rate": 2.158e-05, "loss": 59.0, "step": 2159 }, { "epoch": 0.000216, "grad_norm": 8.877838134765625, "learning_rate": 2.1589999999999998e-05, "loss": 57.7812, "step": 2160 }, { "epoch": 0.0002161, "grad_norm": 8.918885231018066, "learning_rate": 2.16e-05, "loss": 58.4375, "step": 2161 }, { "epoch": 0.0002162, "grad_norm": 8.723161697387695, "learning_rate": 2.161e-05, "loss": 57.0938, "step": 2162 }, { "epoch": 0.0002163, "grad_norm": 8.921587944030762, "learning_rate": 2.162e-05, "loss": 56.5938, "step": 2163 }, { "epoch": 0.0002164, "grad_norm": 9.034866333007812, "learning_rate": 2.163e-05, "loss": 56.8438, "step": 2164 }, { "epoch": 0.0002165, "grad_norm": 8.786333084106445, "learning_rate": 2.1640000000000003e-05, "loss": 56.2812, "step": 2165 }, { "epoch": 0.0002166, "grad_norm": 8.873806953430176, "learning_rate": 2.165e-05, "loss": 57.0938, "step": 2166 }, { "epoch": 0.0002167, "grad_norm": 8.80734920501709, "learning_rate": 2.166e-05, "loss": 56.8125, "step": 2167 }, { "epoch": 0.0002168, "grad_norm": 8.858109474182129, "learning_rate": 2.167e-05, "loss": 56.125, "step": 2168 }, { "epoch": 0.0002169, "grad_norm": 8.880186080932617, "learning_rate": 2.1680000000000002e-05, "loss": 56.9062, "step": 2169 }, { "epoch": 0.000217, "grad_norm": 8.803231239318848, "learning_rate": 2.169e-05, "loss": 54.2188, "step": 2170 }, { "epoch": 0.0002171, "grad_norm": 8.84544563293457, "learning_rate": 2.17e-05, "loss": 55.7812, "step": 2171 }, { "epoch": 0.0002172, "grad_norm": 8.792439460754395, "learning_rate": 2.171e-05, "loss": 55.875, "step": 2172 }, { "epoch": 0.0002173, "grad_norm": 8.716792106628418, "learning_rate": 2.172e-05, "loss": 57.2812, "step": 2173 }, { "epoch": 0.0002174, "grad_norm": 8.840030670166016, "learning_rate": 2.173e-05, "loss": 55.6875, "step": 2174 }, { "epoch": 0.0002175, "grad_norm": 8.674586296081543, "learning_rate": 2.1740000000000002e-05, "loss": 55.2812, "step": 2175 }, { "epoch": 0.0002176, "grad_norm": 8.739676475524902, "learning_rate": 2.175e-05, "loss": 54.4375, "step": 2176 }, { "epoch": 0.0002177, "grad_norm": 8.649190902709961, "learning_rate": 2.176e-05, "loss": 55.5312, "step": 2177 }, { "epoch": 0.0002178, "grad_norm": 8.704854011535645, "learning_rate": 2.177e-05, "loss": 55.75, "step": 2178 }, { "epoch": 0.0002179, "grad_norm": 8.929208755493164, "learning_rate": 2.178e-05, "loss": 55.6875, "step": 2179 }, { "epoch": 0.000218, "grad_norm": 8.667614936828613, "learning_rate": 2.179e-05, "loss": 57.1875, "step": 2180 }, { "epoch": 0.0002181, "grad_norm": 8.642916679382324, "learning_rate": 2.18e-05, "loss": 54.5, "step": 2181 }, { "epoch": 0.0002182, "grad_norm": 8.696272850036621, "learning_rate": 2.1810000000000003e-05, "loss": 53.4375, "step": 2182 }, { "epoch": 0.0002183, "grad_norm": 8.590659141540527, "learning_rate": 2.182e-05, "loss": 54.0625, "step": 2183 }, { "epoch": 0.0002184, "grad_norm": 8.629226684570312, "learning_rate": 2.183e-05, "loss": 53.0, "step": 2184 }, { "epoch": 0.0002185, "grad_norm": 8.680209159851074, "learning_rate": 2.184e-05, "loss": 54.5625, "step": 2185 }, { "epoch": 0.0002186, "grad_norm": 8.656569480895996, "learning_rate": 2.1850000000000003e-05, "loss": 53.2812, "step": 2186 }, { "epoch": 0.0002187, "grad_norm": 8.599714279174805, "learning_rate": 2.186e-05, "loss": 54.4688, "step": 2187 }, { "epoch": 0.0002188, "grad_norm": 8.533751487731934, "learning_rate": 2.187e-05, "loss": 54.625, "step": 2188 }, { "epoch": 0.0002189, "grad_norm": 8.478224754333496, "learning_rate": 2.188e-05, "loss": 52.4375, "step": 2189 }, { "epoch": 0.000219, "grad_norm": 8.622870445251465, "learning_rate": 2.189e-05, "loss": 53.8438, "step": 2190 }, { "epoch": 0.0002191, "grad_norm": 8.586084365844727, "learning_rate": 2.19e-05, "loss": 52.6562, "step": 2191 }, { "epoch": 0.0002192, "grad_norm": 8.717585563659668, "learning_rate": 2.1910000000000002e-05, "loss": 54.8125, "step": 2192 }, { "epoch": 0.0002193, "grad_norm": 8.538300514221191, "learning_rate": 2.192e-05, "loss": 53.3125, "step": 2193 }, { "epoch": 0.0002194, "grad_norm": 8.50560188293457, "learning_rate": 2.193e-05, "loss": 53.1562, "step": 2194 }, { "epoch": 0.0002195, "grad_norm": 8.405988693237305, "learning_rate": 2.194e-05, "loss": 52.2188, "step": 2195 }, { "epoch": 0.0002196, "grad_norm": 8.513202667236328, "learning_rate": 2.195e-05, "loss": 52.2188, "step": 2196 }, { "epoch": 0.0002197, "grad_norm": 8.506519317626953, "learning_rate": 2.196e-05, "loss": 50.375, "step": 2197 }, { "epoch": 0.0002198, "grad_norm": 8.570908546447754, "learning_rate": 2.197e-05, "loss": 53.6875, "step": 2198 }, { "epoch": 0.0002199, "grad_norm": 8.383262634277344, "learning_rate": 2.1980000000000003e-05, "loss": 53.25, "step": 2199 }, { "epoch": 0.00022, "grad_norm": 8.407401084899902, "learning_rate": 2.199e-05, "loss": 51.625, "step": 2200 }, { "epoch": 0.0002201, "grad_norm": 8.359652519226074, "learning_rate": 2.2e-05, "loss": 50.9062, "step": 2201 }, { "epoch": 0.0002202, "grad_norm": 8.435853958129883, "learning_rate": 2.201e-05, "loss": 52.4062, "step": 2202 }, { "epoch": 0.0002203, "grad_norm": 8.357525825500488, "learning_rate": 2.2020000000000003e-05, "loss": 52.0938, "step": 2203 }, { "epoch": 0.0002204, "grad_norm": 8.364212989807129, "learning_rate": 2.203e-05, "loss": 50.75, "step": 2204 }, { "epoch": 0.0002205, "grad_norm": 8.356772422790527, "learning_rate": 2.204e-05, "loss": 51.75, "step": 2205 }, { "epoch": 0.0002206, "grad_norm": 8.304450988769531, "learning_rate": 2.205e-05, "loss": 51.9688, "step": 2206 }, { "epoch": 0.0002207, "grad_norm": 8.328248977661133, "learning_rate": 2.206e-05, "loss": 51.7188, "step": 2207 }, { "epoch": 0.0002208, "grad_norm": 8.284815788269043, "learning_rate": 2.207e-05, "loss": 51.375, "step": 2208 }, { "epoch": 0.0002209, "grad_norm": 8.310721397399902, "learning_rate": 2.2080000000000002e-05, "loss": 50.4688, "step": 2209 }, { "epoch": 0.000221, "grad_norm": 8.190423965454102, "learning_rate": 2.209e-05, "loss": 50.5312, "step": 2210 }, { "epoch": 0.0002211, "grad_norm": 8.309791564941406, "learning_rate": 2.21e-05, "loss": 51.9375, "step": 2211 }, { "epoch": 0.0002212, "grad_norm": 8.115564346313477, "learning_rate": 2.211e-05, "loss": 49.5938, "step": 2212 }, { "epoch": 0.0002213, "grad_norm": 8.248490333557129, "learning_rate": 2.212e-05, "loss": 51.2812, "step": 2213 }, { "epoch": 0.0002214, "grad_norm": 8.224411964416504, "learning_rate": 2.213e-05, "loss": 50.875, "step": 2214 }, { "epoch": 0.0002215, "grad_norm": 8.20934009552002, "learning_rate": 2.214e-05, "loss": 48.6562, "step": 2215 }, { "epoch": 0.0002216, "grad_norm": 8.176395416259766, "learning_rate": 2.2150000000000003e-05, "loss": 49.4375, "step": 2216 }, { "epoch": 0.0002217, "grad_norm": 8.101838111877441, "learning_rate": 2.216e-05, "loss": 49.25, "step": 2217 }, { "epoch": 0.0002218, "grad_norm": 8.176382064819336, "learning_rate": 2.217e-05, "loss": 49.7812, "step": 2218 }, { "epoch": 0.0002219, "grad_norm": 8.181842803955078, "learning_rate": 2.218e-05, "loss": 49.75, "step": 2219 }, { "epoch": 0.000222, "grad_norm": 8.1691255569458, "learning_rate": 2.219e-05, "loss": 48.7188, "step": 2220 }, { "epoch": 0.0002221, "grad_norm": 8.243263244628906, "learning_rate": 2.22e-05, "loss": 49.3125, "step": 2221 }, { "epoch": 0.0002222, "grad_norm": 8.107013702392578, "learning_rate": 2.221e-05, "loss": 49.2812, "step": 2222 }, { "epoch": 0.0002223, "grad_norm": 8.07948112487793, "learning_rate": 2.222e-05, "loss": 49.4688, "step": 2223 }, { "epoch": 0.0002224, "grad_norm": 8.074081420898438, "learning_rate": 2.223e-05, "loss": 49.0938, "step": 2224 }, { "epoch": 0.0002225, "grad_norm": 8.558090209960938, "learning_rate": 2.224e-05, "loss": 48.2188, "step": 2225 }, { "epoch": 0.0002226, "grad_norm": 8.176921844482422, "learning_rate": 2.2250000000000002e-05, "loss": 49.0, "step": 2226 }, { "epoch": 0.0002227, "grad_norm": 8.099858283996582, "learning_rate": 2.226e-05, "loss": 47.7812, "step": 2227 }, { "epoch": 0.0002228, "grad_norm": 8.087740898132324, "learning_rate": 2.227e-05, "loss": 49.4375, "step": 2228 }, { "epoch": 0.0002229, "grad_norm": 8.036834716796875, "learning_rate": 2.228e-05, "loss": 47.9062, "step": 2229 }, { "epoch": 0.000223, "grad_norm": 7.898154258728027, "learning_rate": 2.2290000000000002e-05, "loss": 48.4375, "step": 2230 }, { "epoch": 0.0002231, "grad_norm": 8.080243110656738, "learning_rate": 2.23e-05, "loss": 47.8125, "step": 2231 }, { "epoch": 0.0002232, "grad_norm": 7.969535827636719, "learning_rate": 2.231e-05, "loss": 47.4688, "step": 2232 }, { "epoch": 0.0002233, "grad_norm": 7.7585344314575195, "learning_rate": 2.2320000000000003e-05, "loss": 50.0938, "step": 2233 }, { "epoch": 0.0002234, "grad_norm": 7.932236671447754, "learning_rate": 2.2329999999999998e-05, "loss": 48.375, "step": 2234 }, { "epoch": 0.0002235, "grad_norm": 7.899070739746094, "learning_rate": 2.234e-05, "loss": 46.5625, "step": 2235 }, { "epoch": 0.0002236, "grad_norm": 7.868811130523682, "learning_rate": 2.235e-05, "loss": 48.25, "step": 2236 }, { "epoch": 0.0002237, "grad_norm": 7.73252010345459, "learning_rate": 2.236e-05, "loss": 48.0, "step": 2237 }, { "epoch": 0.0002238, "grad_norm": 7.9742937088012695, "learning_rate": 2.237e-05, "loss": 46.3125, "step": 2238 }, { "epoch": 0.0002239, "grad_norm": 7.8397536277771, "learning_rate": 2.238e-05, "loss": 47.1875, "step": 2239 }, { "epoch": 0.000224, "grad_norm": 7.838926315307617, "learning_rate": 2.239e-05, "loss": 47.125, "step": 2240 }, { "epoch": 0.0002241, "grad_norm": 7.7837982177734375, "learning_rate": 2.24e-05, "loss": 47.0, "step": 2241 }, { "epoch": 0.0002242, "grad_norm": 7.938530921936035, "learning_rate": 2.241e-05, "loss": 47.5312, "step": 2242 }, { "epoch": 0.0002243, "grad_norm": 7.856749057769775, "learning_rate": 2.2420000000000002e-05, "loss": 46.4688, "step": 2243 }, { "epoch": 0.0002244, "grad_norm": 8.040616989135742, "learning_rate": 2.243e-05, "loss": 46.6875, "step": 2244 }, { "epoch": 0.0002245, "grad_norm": 7.647137641906738, "learning_rate": 2.244e-05, "loss": 47.5625, "step": 2245 }, { "epoch": 0.0002246, "grad_norm": 8.253006935119629, "learning_rate": 2.245e-05, "loss": 47.1875, "step": 2246 }, { "epoch": 0.0002247, "grad_norm": 7.75338888168335, "learning_rate": 2.2460000000000002e-05, "loss": 48.1562, "step": 2247 }, { "epoch": 0.0002248, "grad_norm": 7.8322954177856445, "learning_rate": 2.247e-05, "loss": 47.9062, "step": 2248 }, { "epoch": 0.0002249, "grad_norm": 7.846690654754639, "learning_rate": 2.248e-05, "loss": 47.4375, "step": 2249 }, { "epoch": 0.000225, "grad_norm": 7.768208980560303, "learning_rate": 2.249e-05, "loss": 45.875, "step": 2250 }, { "epoch": 0.0002251, "grad_norm": 7.7435407638549805, "learning_rate": 2.2499999999999998e-05, "loss": 46.1875, "step": 2251 }, { "epoch": 0.0002252, "grad_norm": 7.56777811050415, "learning_rate": 2.251e-05, "loss": 48.5938, "step": 2252 }, { "epoch": 0.0002253, "grad_norm": 7.642414093017578, "learning_rate": 2.252e-05, "loss": 45.5938, "step": 2253 }, { "epoch": 0.0002254, "grad_norm": 7.837481498718262, "learning_rate": 2.253e-05, "loss": 45.75, "step": 2254 }, { "epoch": 0.0002255, "grad_norm": 7.646183490753174, "learning_rate": 2.254e-05, "loss": 45.4688, "step": 2255 }, { "epoch": 0.0002256, "grad_norm": 7.741044044494629, "learning_rate": 2.255e-05, "loss": 46.375, "step": 2256 }, { "epoch": 0.0002257, "grad_norm": 7.659153461456299, "learning_rate": 2.256e-05, "loss": 47.7812, "step": 2257 }, { "epoch": 0.0002258, "grad_norm": 7.65839147567749, "learning_rate": 2.257e-05, "loss": 45.8438, "step": 2258 }, { "epoch": 0.0002259, "grad_norm": 7.7989091873168945, "learning_rate": 2.258e-05, "loss": 45.6875, "step": 2259 }, { "epoch": 0.000226, "grad_norm": 7.676018714904785, "learning_rate": 2.2590000000000002e-05, "loss": 47.125, "step": 2260 }, { "epoch": 0.0002261, "grad_norm": 7.596715450286865, "learning_rate": 2.26e-05, "loss": 44.5625, "step": 2261 }, { "epoch": 0.0002262, "grad_norm": 7.614014625549316, "learning_rate": 2.261e-05, "loss": 44.7812, "step": 2262 }, { "epoch": 0.0002263, "grad_norm": 7.511618137359619, "learning_rate": 2.262e-05, "loss": 43.5938, "step": 2263 }, { "epoch": 0.0002264, "grad_norm": 7.519399642944336, "learning_rate": 2.2630000000000002e-05, "loss": 44.3125, "step": 2264 }, { "epoch": 0.0002265, "grad_norm": 7.4990458488464355, "learning_rate": 2.264e-05, "loss": 43.8125, "step": 2265 }, { "epoch": 0.0002266, "grad_norm": 7.651999473571777, "learning_rate": 2.265e-05, "loss": 43.1562, "step": 2266 }, { "epoch": 0.0002267, "grad_norm": 7.559002876281738, "learning_rate": 2.266e-05, "loss": 44.0, "step": 2267 }, { "epoch": 0.0002268, "grad_norm": 7.471904277801514, "learning_rate": 2.2669999999999998e-05, "loss": 43.0, "step": 2268 }, { "epoch": 0.0002269, "grad_norm": 7.628164768218994, "learning_rate": 2.268e-05, "loss": 43.5625, "step": 2269 }, { "epoch": 0.000227, "grad_norm": 7.535248756408691, "learning_rate": 2.269e-05, "loss": 44.4688, "step": 2270 }, { "epoch": 0.0002271, "grad_norm": 7.319836139678955, "learning_rate": 2.27e-05, "loss": 43.6875, "step": 2271 }, { "epoch": 0.0002272, "grad_norm": 7.559850215911865, "learning_rate": 2.271e-05, "loss": 45.375, "step": 2272 }, { "epoch": 0.0002273, "grad_norm": 7.592349529266357, "learning_rate": 2.272e-05, "loss": 43.5, "step": 2273 }, { "epoch": 0.0002274, "grad_norm": 7.430238723754883, "learning_rate": 2.273e-05, "loss": 43.5625, "step": 2274 }, { "epoch": 0.0002275, "grad_norm": 7.312240123748779, "learning_rate": 2.274e-05, "loss": 42.2188, "step": 2275 }, { "epoch": 0.0002276, "grad_norm": 7.307692527770996, "learning_rate": 2.275e-05, "loss": 45.0312, "step": 2276 }, { "epoch": 0.0002277, "grad_norm": 7.340839862823486, "learning_rate": 2.2760000000000002e-05, "loss": 41.0, "step": 2277 }, { "epoch": 0.0002278, "grad_norm": 7.33964729309082, "learning_rate": 2.277e-05, "loss": 42.4062, "step": 2278 }, { "epoch": 0.0002279, "grad_norm": 7.304256439208984, "learning_rate": 2.278e-05, "loss": 41.0, "step": 2279 }, { "epoch": 0.000228, "grad_norm": 7.232362747192383, "learning_rate": 2.279e-05, "loss": 41.5625, "step": 2280 }, { "epoch": 0.0002281, "grad_norm": 7.389524459838867, "learning_rate": 2.2800000000000002e-05, "loss": 41.0312, "step": 2281 }, { "epoch": 0.0002282, "grad_norm": 7.263487339019775, "learning_rate": 2.281e-05, "loss": 42.9375, "step": 2282 }, { "epoch": 0.0002283, "grad_norm": 7.427538871765137, "learning_rate": 2.2820000000000002e-05, "loss": 42.3125, "step": 2283 }, { "epoch": 0.0002284, "grad_norm": 7.433857440948486, "learning_rate": 2.283e-05, "loss": 42.8438, "step": 2284 }, { "epoch": 0.0002285, "grad_norm": 7.31108283996582, "learning_rate": 2.2839999999999998e-05, "loss": 43.2188, "step": 2285 }, { "epoch": 0.0002286, "grad_norm": 7.192925930023193, "learning_rate": 2.285e-05, "loss": 41.0938, "step": 2286 }, { "epoch": 0.0002287, "grad_norm": 7.321473121643066, "learning_rate": 2.286e-05, "loss": 42.0938, "step": 2287 }, { "epoch": 0.0002288, "grad_norm": 7.285192012786865, "learning_rate": 2.287e-05, "loss": 43.0, "step": 2288 }, { "epoch": 0.0002289, "grad_norm": 7.195656776428223, "learning_rate": 2.288e-05, "loss": 42.625, "step": 2289 }, { "epoch": 0.000229, "grad_norm": 7.355578899383545, "learning_rate": 2.289e-05, "loss": 44.4062, "step": 2290 }, { "epoch": 0.0002291, "grad_norm": 7.324771881103516, "learning_rate": 2.29e-05, "loss": 42.4688, "step": 2291 }, { "epoch": 0.0002292, "grad_norm": 7.395291328430176, "learning_rate": 2.291e-05, "loss": 42.7188, "step": 2292 }, { "epoch": 0.0002293, "grad_norm": 7.224531650543213, "learning_rate": 2.292e-05, "loss": 41.6562, "step": 2293 }, { "epoch": 0.0002294, "grad_norm": 7.339974880218506, "learning_rate": 2.2930000000000002e-05, "loss": 41.9375, "step": 2294 }, { "epoch": 0.0002295, "grad_norm": 7.286098003387451, "learning_rate": 2.294e-05, "loss": 42.6562, "step": 2295 }, { "epoch": 0.0002296, "grad_norm": 7.221029281616211, "learning_rate": 2.295e-05, "loss": 41.5938, "step": 2296 }, { "epoch": 0.0002297, "grad_norm": 7.31257438659668, "learning_rate": 2.296e-05, "loss": 41.9062, "step": 2297 }, { "epoch": 0.0002298, "grad_norm": 7.069023609161377, "learning_rate": 2.2970000000000002e-05, "loss": 39.7812, "step": 2298 }, { "epoch": 0.0002299, "grad_norm": 7.197727203369141, "learning_rate": 2.298e-05, "loss": 41.125, "step": 2299 }, { "epoch": 0.00023, "grad_norm": 7.125405311584473, "learning_rate": 2.2990000000000002e-05, "loss": 41.5312, "step": 2300 }, { "epoch": 0.0002301, "grad_norm": 7.245908737182617, "learning_rate": 2.3e-05, "loss": 40.875, "step": 2301 }, { "epoch": 0.0002302, "grad_norm": 7.131826877593994, "learning_rate": 2.3009999999999998e-05, "loss": 39.8438, "step": 2302 }, { "epoch": 0.0002303, "grad_norm": 7.024893283843994, "learning_rate": 2.302e-05, "loss": 41.9375, "step": 2303 }, { "epoch": 0.0002304, "grad_norm": 7.236159801483154, "learning_rate": 2.303e-05, "loss": 41.2188, "step": 2304 }, { "epoch": 0.0002305, "grad_norm": 7.09937858581543, "learning_rate": 2.304e-05, "loss": 39.6875, "step": 2305 }, { "epoch": 0.0002306, "grad_norm": 7.0625810623168945, "learning_rate": 2.305e-05, "loss": 39.25, "step": 2306 }, { "epoch": 0.0002307, "grad_norm": 7.200896739959717, "learning_rate": 2.306e-05, "loss": 39.8125, "step": 2307 }, { "epoch": 0.0002308, "grad_norm": 7.114955425262451, "learning_rate": 2.307e-05, "loss": 39.5938, "step": 2308 }, { "epoch": 0.0002309, "grad_norm": 7.020622730255127, "learning_rate": 2.308e-05, "loss": 41.3438, "step": 2309 }, { "epoch": 0.000231, "grad_norm": 7.363964557647705, "learning_rate": 2.309e-05, "loss": 39.0625, "step": 2310 }, { "epoch": 0.0002311, "grad_norm": 7.006573677062988, "learning_rate": 2.3100000000000002e-05, "loss": 40.625, "step": 2311 }, { "epoch": 0.0002312, "grad_norm": 7.15456485748291, "learning_rate": 2.311e-05, "loss": 42.375, "step": 2312 }, { "epoch": 0.0002313, "grad_norm": 6.8626837730407715, "learning_rate": 2.312e-05, "loss": 39.0312, "step": 2313 }, { "epoch": 0.0002314, "grad_norm": 7.112387657165527, "learning_rate": 2.313e-05, "loss": 40.7812, "step": 2314 }, { "epoch": 0.0002315, "grad_norm": 6.934446334838867, "learning_rate": 2.3140000000000002e-05, "loss": 38.75, "step": 2315 }, { "epoch": 0.0002316, "grad_norm": 6.99973726272583, "learning_rate": 2.315e-05, "loss": 39.5625, "step": 2316 }, { "epoch": 0.0002317, "grad_norm": 7.051058292388916, "learning_rate": 2.3160000000000002e-05, "loss": 39.0625, "step": 2317 }, { "epoch": 0.0002318, "grad_norm": 6.987892150878906, "learning_rate": 2.317e-05, "loss": 39.75, "step": 2318 }, { "epoch": 0.0002319, "grad_norm": 7.00262975692749, "learning_rate": 2.3179999999999998e-05, "loss": 40.75, "step": 2319 }, { "epoch": 0.000232, "grad_norm": 6.827770233154297, "learning_rate": 2.319e-05, "loss": 38.4375, "step": 2320 }, { "epoch": 0.0002321, "grad_norm": 7.067201614379883, "learning_rate": 2.32e-05, "loss": 38.25, "step": 2321 }, { "epoch": 0.0002322, "grad_norm": 6.820005416870117, "learning_rate": 2.321e-05, "loss": 38.9688, "step": 2322 }, { "epoch": 0.0002323, "grad_norm": 6.794841766357422, "learning_rate": 2.322e-05, "loss": 39.5625, "step": 2323 }, { "epoch": 0.0002324, "grad_norm": 7.1287431716918945, "learning_rate": 2.323e-05, "loss": 40.1875, "step": 2324 }, { "epoch": 0.0002325, "grad_norm": 7.043614864349365, "learning_rate": 2.324e-05, "loss": 39.625, "step": 2325 }, { "epoch": 0.0002326, "grad_norm": 6.909962177276611, "learning_rate": 2.325e-05, "loss": 38.3125, "step": 2326 }, { "epoch": 0.0002327, "grad_norm": 6.860498428344727, "learning_rate": 2.326e-05, "loss": 38.3438, "step": 2327 }, { "epoch": 0.0002328, "grad_norm": 6.850043296813965, "learning_rate": 2.3270000000000003e-05, "loss": 37.3125, "step": 2328 }, { "epoch": 0.0002329, "grad_norm": 6.880703449249268, "learning_rate": 2.328e-05, "loss": 37.75, "step": 2329 }, { "epoch": 0.000233, "grad_norm": 6.9582014083862305, "learning_rate": 2.329e-05, "loss": 38.8438, "step": 2330 }, { "epoch": 0.0002331, "grad_norm": 7.09195613861084, "learning_rate": 2.33e-05, "loss": 39.4375, "step": 2331 }, { "epoch": 0.0002332, "grad_norm": 6.859071731567383, "learning_rate": 2.3310000000000002e-05, "loss": 37.5312, "step": 2332 }, { "epoch": 0.0002333, "grad_norm": 7.111638069152832, "learning_rate": 2.332e-05, "loss": 40.4375, "step": 2333 }, { "epoch": 0.0002334, "grad_norm": 6.789240837097168, "learning_rate": 2.3330000000000002e-05, "loss": 38.3125, "step": 2334 }, { "epoch": 0.0002335, "grad_norm": 6.8712158203125, "learning_rate": 2.334e-05, "loss": 37.5625, "step": 2335 }, { "epoch": 0.0002336, "grad_norm": 6.944106101989746, "learning_rate": 2.335e-05, "loss": 38.75, "step": 2336 }, { "epoch": 0.0002337, "grad_norm": 6.775470733642578, "learning_rate": 2.336e-05, "loss": 37.0312, "step": 2337 }, { "epoch": 0.0002338, "grad_norm": 6.737112998962402, "learning_rate": 2.337e-05, "loss": 38.5312, "step": 2338 }, { "epoch": 0.0002339, "grad_norm": 6.726865768432617, "learning_rate": 2.338e-05, "loss": 36.0625, "step": 2339 }, { "epoch": 0.000234, "grad_norm": 6.8453145027160645, "learning_rate": 2.339e-05, "loss": 38.0938, "step": 2340 }, { "epoch": 0.0002341, "grad_norm": 6.933478832244873, "learning_rate": 2.34e-05, "loss": 38.6875, "step": 2341 }, { "epoch": 0.0002342, "grad_norm": 6.738262176513672, "learning_rate": 2.341e-05, "loss": 36.75, "step": 2342 }, { "epoch": 0.0002343, "grad_norm": 6.792433738708496, "learning_rate": 2.342e-05, "loss": 37.625, "step": 2343 }, { "epoch": 0.0002344, "grad_norm": 6.631856441497803, "learning_rate": 2.343e-05, "loss": 38.3125, "step": 2344 }, { "epoch": 0.0002345, "grad_norm": 6.73577356338501, "learning_rate": 2.3440000000000003e-05, "loss": 36.9062, "step": 2345 }, { "epoch": 0.0002346, "grad_norm": 6.833475112915039, "learning_rate": 2.345e-05, "loss": 35.0312, "step": 2346 }, { "epoch": 0.0002347, "grad_norm": 7.382632255554199, "learning_rate": 2.3460000000000002e-05, "loss": 37.125, "step": 2347 }, { "epoch": 0.0002348, "grad_norm": 6.943807601928711, "learning_rate": 2.347e-05, "loss": 38.9688, "step": 2348 }, { "epoch": 0.0002349, "grad_norm": 6.659213066101074, "learning_rate": 2.348e-05, "loss": 37.375, "step": 2349 }, { "epoch": 0.000235, "grad_norm": 6.801396369934082, "learning_rate": 2.349e-05, "loss": 37.0625, "step": 2350 }, { "epoch": 0.0002351, "grad_norm": 6.738434314727783, "learning_rate": 2.35e-05, "loss": 36.3438, "step": 2351 }, { "epoch": 0.0002352, "grad_norm": 7.003298282623291, "learning_rate": 2.351e-05, "loss": 37.5625, "step": 2352 }, { "epoch": 0.0002353, "grad_norm": 6.757877349853516, "learning_rate": 2.352e-05, "loss": 35.75, "step": 2353 }, { "epoch": 0.0002354, "grad_norm": 6.6432976722717285, "learning_rate": 2.3530000000000003e-05, "loss": 36.9062, "step": 2354 }, { "epoch": 0.0002355, "grad_norm": 6.571918964385986, "learning_rate": 2.354e-05, "loss": 35.6875, "step": 2355 }, { "epoch": 0.0002356, "grad_norm": 6.710591793060303, "learning_rate": 2.355e-05, "loss": 38.25, "step": 2356 }, { "epoch": 0.0002357, "grad_norm": 6.599053382873535, "learning_rate": 2.356e-05, "loss": 35.9688, "step": 2357 }, { "epoch": 0.0002358, "grad_norm": 6.7090959548950195, "learning_rate": 2.357e-05, "loss": 36.25, "step": 2358 }, { "epoch": 0.0002359, "grad_norm": 6.477239608764648, "learning_rate": 2.358e-05, "loss": 33.7812, "step": 2359 }, { "epoch": 0.000236, "grad_norm": 6.5696635246276855, "learning_rate": 2.359e-05, "loss": 37.0312, "step": 2360 }, { "epoch": 0.0002361, "grad_norm": 6.527729511260986, "learning_rate": 2.3599999999999998e-05, "loss": 34.0938, "step": 2361 }, { "epoch": 0.0002362, "grad_norm": 6.5055460929870605, "learning_rate": 2.3610000000000003e-05, "loss": 36.5625, "step": 2362 }, { "epoch": 0.0002363, "grad_norm": 6.627758502960205, "learning_rate": 2.3619999999999997e-05, "loss": 35.875, "step": 2363 }, { "epoch": 0.0002364, "grad_norm": 6.489551544189453, "learning_rate": 2.3630000000000002e-05, "loss": 35.625, "step": 2364 }, { "epoch": 0.0002365, "grad_norm": 6.604640960693359, "learning_rate": 2.364e-05, "loss": 36.3438, "step": 2365 }, { "epoch": 0.0002366, "grad_norm": 6.45327091217041, "learning_rate": 2.365e-05, "loss": 36.5625, "step": 2366 }, { "epoch": 0.0002367, "grad_norm": 6.5227227210998535, "learning_rate": 2.366e-05, "loss": 35.3438, "step": 2367 }, { "epoch": 0.0002368, "grad_norm": 6.489030838012695, "learning_rate": 2.367e-05, "loss": 34.5312, "step": 2368 }, { "epoch": 0.0002369, "grad_norm": 6.462714672088623, "learning_rate": 2.368e-05, "loss": 34.2812, "step": 2369 }, { "epoch": 0.000237, "grad_norm": 6.642274379730225, "learning_rate": 2.369e-05, "loss": 36.4688, "step": 2370 }, { "epoch": 0.0002371, "grad_norm": 6.523331165313721, "learning_rate": 2.3700000000000003e-05, "loss": 35.9219, "step": 2371 }, { "epoch": 0.0002372, "grad_norm": 6.488804340362549, "learning_rate": 2.371e-05, "loss": 34.9688, "step": 2372 }, { "epoch": 0.0002373, "grad_norm": 6.359453201293945, "learning_rate": 2.372e-05, "loss": 33.7812, "step": 2373 }, { "epoch": 0.0002374, "grad_norm": 6.482349395751953, "learning_rate": 2.373e-05, "loss": 34.6875, "step": 2374 }, { "epoch": 0.0002375, "grad_norm": 6.718332290649414, "learning_rate": 2.374e-05, "loss": 35.3438, "step": 2375 }, { "epoch": 0.0002376, "grad_norm": 6.406708240509033, "learning_rate": 2.375e-05, "loss": 33.5, "step": 2376 }, { "epoch": 0.0002377, "grad_norm": 6.547664642333984, "learning_rate": 2.376e-05, "loss": 34.7188, "step": 2377 }, { "epoch": 0.0002378, "grad_norm": 6.527320384979248, "learning_rate": 2.3769999999999998e-05, "loss": 34.3125, "step": 2378 }, { "epoch": 0.0002379, "grad_norm": 6.474934101104736, "learning_rate": 2.3780000000000003e-05, "loss": 35.0938, "step": 2379 }, { "epoch": 0.000238, "grad_norm": 6.395140171051025, "learning_rate": 2.3789999999999998e-05, "loss": 34.625, "step": 2380 }, { "epoch": 0.0002381, "grad_norm": 6.431519985198975, "learning_rate": 2.3800000000000003e-05, "loss": 34.7656, "step": 2381 }, { "epoch": 0.0002382, "grad_norm": 6.522791385650635, "learning_rate": 2.381e-05, "loss": 35.2969, "step": 2382 }, { "epoch": 0.0002383, "grad_norm": 6.381486892700195, "learning_rate": 2.382e-05, "loss": 33.1562, "step": 2383 }, { "epoch": 0.0002384, "grad_norm": 6.617220878601074, "learning_rate": 2.383e-05, "loss": 35.25, "step": 2384 }, { "epoch": 0.0002385, "grad_norm": 6.242531776428223, "learning_rate": 2.384e-05, "loss": 34.0781, "step": 2385 }, { "epoch": 0.0002386, "grad_norm": 6.455689430236816, "learning_rate": 2.385e-05, "loss": 35.3438, "step": 2386 }, { "epoch": 0.0002387, "grad_norm": 6.385616302490234, "learning_rate": 2.386e-05, "loss": 34.9531, "step": 2387 }, { "epoch": 0.0002388, "grad_norm": 6.4845757484436035, "learning_rate": 2.3870000000000004e-05, "loss": 35.375, "step": 2388 }, { "epoch": 0.0002389, "grad_norm": 6.378547191619873, "learning_rate": 2.3880000000000002e-05, "loss": 34.0625, "step": 2389 }, { "epoch": 0.000239, "grad_norm": 6.211874485015869, "learning_rate": 2.389e-05, "loss": 32.6875, "step": 2390 }, { "epoch": 0.0002391, "grad_norm": 6.296255111694336, "learning_rate": 2.39e-05, "loss": 31.5312, "step": 2391 }, { "epoch": 0.0002392, "grad_norm": 6.200486183166504, "learning_rate": 2.391e-05, "loss": 32.5625, "step": 2392 }, { "epoch": 0.0002393, "grad_norm": 6.326794624328613, "learning_rate": 2.392e-05, "loss": 33.0, "step": 2393 }, { "epoch": 0.0002394, "grad_norm": 6.484475612640381, "learning_rate": 2.393e-05, "loss": 34.5, "step": 2394 }, { "epoch": 0.0002395, "grad_norm": 6.319868087768555, "learning_rate": 2.3939999999999998e-05, "loss": 32.5625, "step": 2395 }, { "epoch": 0.0002396, "grad_norm": 6.2924675941467285, "learning_rate": 2.3950000000000003e-05, "loss": 34.2188, "step": 2396 }, { "epoch": 0.0002397, "grad_norm": 6.409213066101074, "learning_rate": 2.3959999999999998e-05, "loss": 32.7188, "step": 2397 }, { "epoch": 0.0002398, "grad_norm": 6.628047943115234, "learning_rate": 2.3970000000000003e-05, "loss": 35.5625, "step": 2398 }, { "epoch": 0.0002399, "grad_norm": 6.218595027923584, "learning_rate": 2.398e-05, "loss": 32.2656, "step": 2399 }, { "epoch": 0.00024, "grad_norm": 6.413675308227539, "learning_rate": 2.399e-05, "loss": 34.6875, "step": 2400 }, { "epoch": 0.0002401, "grad_norm": 6.191205978393555, "learning_rate": 2.4e-05, "loss": 33.1719, "step": 2401 }, { "epoch": 0.0002402, "grad_norm": 6.357745170593262, "learning_rate": 2.401e-05, "loss": 33.5938, "step": 2402 }, { "epoch": 0.0002403, "grad_norm": 6.139225006103516, "learning_rate": 2.402e-05, "loss": 33.1562, "step": 2403 }, { "epoch": 0.0002404, "grad_norm": 6.132444858551025, "learning_rate": 2.403e-05, "loss": 33.9375, "step": 2404 }, { "epoch": 0.0002405, "grad_norm": 7.256519317626953, "learning_rate": 2.4040000000000004e-05, "loss": 33.7188, "step": 2405 }, { "epoch": 0.0002406, "grad_norm": 6.014881134033203, "learning_rate": 2.4050000000000002e-05, "loss": 33.1406, "step": 2406 }, { "epoch": 0.0002407, "grad_norm": 6.231376647949219, "learning_rate": 2.406e-05, "loss": 32.5469, "step": 2407 }, { "epoch": 0.0002408, "grad_norm": 6.489138126373291, "learning_rate": 2.407e-05, "loss": 34.4375, "step": 2408 }, { "epoch": 0.0002409, "grad_norm": 6.123095989227295, "learning_rate": 2.408e-05, "loss": 31.8125, "step": 2409 }, { "epoch": 0.000241, "grad_norm": 6.073126316070557, "learning_rate": 2.409e-05, "loss": 31.625, "step": 2410 }, { "epoch": 0.0002411, "grad_norm": 6.179654598236084, "learning_rate": 2.41e-05, "loss": 32.7188, "step": 2411 }, { "epoch": 0.0002412, "grad_norm": 6.248153209686279, "learning_rate": 2.4109999999999998e-05, "loss": 33.375, "step": 2412 }, { "epoch": 0.0002413, "grad_norm": 6.112429141998291, "learning_rate": 2.4120000000000003e-05, "loss": 31.0625, "step": 2413 }, { "epoch": 0.0002414, "grad_norm": 6.246798515319824, "learning_rate": 2.4129999999999998e-05, "loss": 32.2344, "step": 2414 }, { "epoch": 0.0002415, "grad_norm": 6.1009111404418945, "learning_rate": 2.4140000000000003e-05, "loss": 32.3906, "step": 2415 }, { "epoch": 0.0002416, "grad_norm": 6.308210849761963, "learning_rate": 2.415e-05, "loss": 32.5938, "step": 2416 }, { "epoch": 0.0002417, "grad_norm": 6.192907333374023, "learning_rate": 2.416e-05, "loss": 32.4375, "step": 2417 }, { "epoch": 0.0002418, "grad_norm": 6.299713611602783, "learning_rate": 2.417e-05, "loss": 31.8906, "step": 2418 }, { "epoch": 0.0002419, "grad_norm": 6.096022129058838, "learning_rate": 2.418e-05, "loss": 31.4219, "step": 2419 }, { "epoch": 0.000242, "grad_norm": 6.1204962730407715, "learning_rate": 2.419e-05, "loss": 33.0625, "step": 2420 }, { "epoch": 0.0002421, "grad_norm": 6.591331958770752, "learning_rate": 2.42e-05, "loss": 32.5781, "step": 2421 }, { "epoch": 0.0002422, "grad_norm": 5.986545562744141, "learning_rate": 2.4210000000000004e-05, "loss": 32.2344, "step": 2422 }, { "epoch": 0.0002423, "grad_norm": 6.159282207489014, "learning_rate": 2.4220000000000002e-05, "loss": 31.9844, "step": 2423 }, { "epoch": 0.0002424, "grad_norm": 6.308074474334717, "learning_rate": 2.423e-05, "loss": 33.1406, "step": 2424 }, { "epoch": 0.0002425, "grad_norm": 6.021403789520264, "learning_rate": 2.4240000000000002e-05, "loss": 30.875, "step": 2425 }, { "epoch": 0.0002426, "grad_norm": 6.032246112823486, "learning_rate": 2.425e-05, "loss": 30.7188, "step": 2426 }, { "epoch": 0.0002427, "grad_norm": 6.226408958435059, "learning_rate": 2.426e-05, "loss": 32.5625, "step": 2427 }, { "epoch": 0.0002428, "grad_norm": 5.9908857345581055, "learning_rate": 2.427e-05, "loss": 30.8906, "step": 2428 }, { "epoch": 0.0002429, "grad_norm": 6.483747959136963, "learning_rate": 2.4279999999999998e-05, "loss": 33.0781, "step": 2429 }, { "epoch": 0.000243, "grad_norm": 6.089707851409912, "learning_rate": 2.4290000000000003e-05, "loss": 30.25, "step": 2430 }, { "epoch": 0.0002431, "grad_norm": 5.99740743637085, "learning_rate": 2.4299999999999998e-05, "loss": 31.4219, "step": 2431 }, { "epoch": 0.0002432, "grad_norm": 6.188370227813721, "learning_rate": 2.4310000000000003e-05, "loss": 31.5, "step": 2432 }, { "epoch": 0.0002433, "grad_norm": 6.126152515411377, "learning_rate": 2.432e-05, "loss": 31.7969, "step": 2433 }, { "epoch": 0.0002434, "grad_norm": 6.105762958526611, "learning_rate": 2.433e-05, "loss": 30.6406, "step": 2434 }, { "epoch": 0.0002435, "grad_norm": 6.155549049377441, "learning_rate": 2.434e-05, "loss": 32.1406, "step": 2435 }, { "epoch": 0.0002436, "grad_norm": 6.254861354827881, "learning_rate": 2.435e-05, "loss": 31.4531, "step": 2436 }, { "epoch": 0.0002437, "grad_norm": 6.401580333709717, "learning_rate": 2.436e-05, "loss": 31.4375, "step": 2437 }, { "epoch": 0.0002438, "grad_norm": 6.648218154907227, "learning_rate": 2.437e-05, "loss": 35.3125, "step": 2438 }, { "epoch": 0.0002439, "grad_norm": 6.059316635131836, "learning_rate": 2.4380000000000004e-05, "loss": 30.5, "step": 2439 }, { "epoch": 0.000244, "grad_norm": 6.266129493713379, "learning_rate": 2.4390000000000002e-05, "loss": 31.3906, "step": 2440 }, { "epoch": 0.0002441, "grad_norm": 5.995891094207764, "learning_rate": 2.44e-05, "loss": 29.6875, "step": 2441 }, { "epoch": 0.0002442, "grad_norm": 5.973202228546143, "learning_rate": 2.4410000000000002e-05, "loss": 29.2344, "step": 2442 }, { "epoch": 0.0002443, "grad_norm": 6.210598468780518, "learning_rate": 2.442e-05, "loss": 30.6875, "step": 2443 }, { "epoch": 0.0002444, "grad_norm": 5.974357604980469, "learning_rate": 2.443e-05, "loss": 30.6719, "step": 2444 }, { "epoch": 0.0002445, "grad_norm": 5.994708061218262, "learning_rate": 2.444e-05, "loss": 31.3594, "step": 2445 }, { "epoch": 0.0002446, "grad_norm": 6.081677436828613, "learning_rate": 2.4449999999999998e-05, "loss": 31.6406, "step": 2446 }, { "epoch": 0.0002447, "grad_norm": 5.915447235107422, "learning_rate": 2.4460000000000003e-05, "loss": 30.3438, "step": 2447 }, { "epoch": 0.0002448, "grad_norm": 6.330596446990967, "learning_rate": 2.4469999999999998e-05, "loss": 30.7656, "step": 2448 }, { "epoch": 0.0002449, "grad_norm": 6.689070224761963, "learning_rate": 2.4480000000000003e-05, "loss": 30.9688, "step": 2449 }, { "epoch": 0.000245, "grad_norm": 6.095787525177002, "learning_rate": 2.449e-05, "loss": 29.5781, "step": 2450 }, { "epoch": 0.0002451, "grad_norm": 5.849099636077881, "learning_rate": 2.45e-05, "loss": 29.8906, "step": 2451 }, { "epoch": 0.0002452, "grad_norm": 5.866293430328369, "learning_rate": 2.451e-05, "loss": 29.7812, "step": 2452 }, { "epoch": 0.0002453, "grad_norm": 5.884515762329102, "learning_rate": 2.452e-05, "loss": 29.5312, "step": 2453 }, { "epoch": 0.0002454, "grad_norm": 5.855321407318115, "learning_rate": 2.453e-05, "loss": 29.4375, "step": 2454 }, { "epoch": 0.0002455, "grad_norm": 5.97646951675415, "learning_rate": 2.454e-05, "loss": 29.7969, "step": 2455 }, { "epoch": 0.0002456, "grad_norm": 5.908134460449219, "learning_rate": 2.4550000000000004e-05, "loss": 30.875, "step": 2456 }, { "epoch": 0.0002457, "grad_norm": 6.08822774887085, "learning_rate": 2.4560000000000002e-05, "loss": 29.7656, "step": 2457 }, { "epoch": 0.0002458, "grad_norm": 5.890263557434082, "learning_rate": 2.457e-05, "loss": 29.6719, "step": 2458 }, { "epoch": 0.0002459, "grad_norm": 5.961470603942871, "learning_rate": 2.4580000000000002e-05, "loss": 30.4219, "step": 2459 }, { "epoch": 0.000246, "grad_norm": 5.92188835144043, "learning_rate": 2.459e-05, "loss": 30.9062, "step": 2460 }, { "epoch": 0.0002461, "grad_norm": 6.031449794769287, "learning_rate": 2.46e-05, "loss": 30.2812, "step": 2461 }, { "epoch": 0.0002462, "grad_norm": 5.944753170013428, "learning_rate": 2.461e-05, "loss": 29.5781, "step": 2462 }, { "epoch": 0.0002463, "grad_norm": 6.007025718688965, "learning_rate": 2.4619999999999998e-05, "loss": 29.3281, "step": 2463 }, { "epoch": 0.0002464, "grad_norm": 6.000064849853516, "learning_rate": 2.4630000000000003e-05, "loss": 30.625, "step": 2464 }, { "epoch": 0.0002465, "grad_norm": 5.913351058959961, "learning_rate": 2.4639999999999998e-05, "loss": 29.4375, "step": 2465 }, { "epoch": 0.0002466, "grad_norm": 6.071257591247559, "learning_rate": 2.4650000000000003e-05, "loss": 29.4688, "step": 2466 }, { "epoch": 0.0002467, "grad_norm": 5.8364577293396, "learning_rate": 2.466e-05, "loss": 29.6406, "step": 2467 }, { "epoch": 0.0002468, "grad_norm": 5.8790812492370605, "learning_rate": 2.467e-05, "loss": 28.7812, "step": 2468 }, { "epoch": 0.0002469, "grad_norm": 6.20643949508667, "learning_rate": 2.468e-05, "loss": 30.9844, "step": 2469 }, { "epoch": 0.000247, "grad_norm": 6.004348278045654, "learning_rate": 2.469e-05, "loss": 29.5781, "step": 2470 }, { "epoch": 0.0002471, "grad_norm": 6.168084621429443, "learning_rate": 2.47e-05, "loss": 30.4531, "step": 2471 }, { "epoch": 0.0002472, "grad_norm": 5.96368408203125, "learning_rate": 2.471e-05, "loss": 28.5312, "step": 2472 }, { "epoch": 0.0002473, "grad_norm": 5.693983554840088, "learning_rate": 2.4720000000000004e-05, "loss": 28.0938, "step": 2473 }, { "epoch": 0.0002474, "grad_norm": 5.846338272094727, "learning_rate": 2.4730000000000002e-05, "loss": 28.9219, "step": 2474 }, { "epoch": 0.0002475, "grad_norm": 5.850586414337158, "learning_rate": 2.474e-05, "loss": 30.0625, "step": 2475 }, { "epoch": 0.0002476, "grad_norm": 6.0289740562438965, "learning_rate": 2.4750000000000002e-05, "loss": 29.2656, "step": 2476 }, { "epoch": 0.0002477, "grad_norm": 5.9219970703125, "learning_rate": 2.476e-05, "loss": 29.5312, "step": 2477 }, { "epoch": 0.0002478, "grad_norm": 5.815984725952148, "learning_rate": 2.4770000000000002e-05, "loss": 28.7031, "step": 2478 }, { "epoch": 0.0002479, "grad_norm": 5.717505931854248, "learning_rate": 2.478e-05, "loss": 27.7344, "step": 2479 }, { "epoch": 0.000248, "grad_norm": 5.685620307922363, "learning_rate": 2.4789999999999998e-05, "loss": 29.9219, "step": 2480 }, { "epoch": 0.0002481, "grad_norm": 6.140514850616455, "learning_rate": 2.48e-05, "loss": 30.0781, "step": 2481 }, { "epoch": 0.0002482, "grad_norm": 5.878580570220947, "learning_rate": 2.4809999999999998e-05, "loss": 30.2969, "step": 2482 }, { "epoch": 0.0002483, "grad_norm": 5.655693531036377, "learning_rate": 2.4820000000000003e-05, "loss": 28.2344, "step": 2483 }, { "epoch": 0.0002484, "grad_norm": 5.6099443435668945, "learning_rate": 2.483e-05, "loss": 28.7031, "step": 2484 }, { "epoch": 0.0002485, "grad_norm": 6.116191864013672, "learning_rate": 2.484e-05, "loss": 28.0, "step": 2485 }, { "epoch": 0.0002486, "grad_norm": 5.786317825317383, "learning_rate": 2.485e-05, "loss": 29.0625, "step": 2486 }, { "epoch": 0.0002487, "grad_norm": 5.7032151222229, "learning_rate": 2.486e-05, "loss": 29.4062, "step": 2487 }, { "epoch": 0.0002488, "grad_norm": 5.732761383056641, "learning_rate": 2.487e-05, "loss": 29.1875, "step": 2488 }, { "epoch": 0.0002489, "grad_norm": 5.618765354156494, "learning_rate": 2.488e-05, "loss": 27.9375, "step": 2489 }, { "epoch": 0.000249, "grad_norm": 5.960012912750244, "learning_rate": 2.4890000000000004e-05, "loss": 29.8438, "step": 2490 }, { "epoch": 0.0002491, "grad_norm": 5.949596881866455, "learning_rate": 2.4900000000000002e-05, "loss": 30.75, "step": 2491 }, { "epoch": 0.0002492, "grad_norm": 5.923252582550049, "learning_rate": 2.4909999999999997e-05, "loss": 29.8281, "step": 2492 }, { "epoch": 0.0002493, "grad_norm": 5.791914939880371, "learning_rate": 2.4920000000000002e-05, "loss": 28.5469, "step": 2493 }, { "epoch": 0.0002494, "grad_norm": 5.746078014373779, "learning_rate": 2.493e-05, "loss": 29.8906, "step": 2494 }, { "epoch": 0.0002495, "grad_norm": 5.993460655212402, "learning_rate": 2.4940000000000002e-05, "loss": 30.0156, "step": 2495 }, { "epoch": 0.0002496, "grad_norm": 6.040581703186035, "learning_rate": 2.495e-05, "loss": 28.1562, "step": 2496 }, { "epoch": 0.0002497, "grad_norm": 5.552365303039551, "learning_rate": 2.4959999999999998e-05, "loss": 27.0781, "step": 2497 }, { "epoch": 0.0002498, "grad_norm": 5.642887115478516, "learning_rate": 2.497e-05, "loss": 27.1719, "step": 2498 }, { "epoch": 0.0002499, "grad_norm": 6.160917282104492, "learning_rate": 2.4979999999999998e-05, "loss": 30.1406, "step": 2499 }, { "epoch": 0.00025, "grad_norm": 6.054183006286621, "learning_rate": 2.4990000000000003e-05, "loss": 29.4688, "step": 2500 }, { "epoch": 0.00025, "eval_loss": 3.3703787326812744, "eval_runtime": 401.8817, "eval_samples_per_second": 24.883, "eval_steps_per_second": 1.555, "step": 2500 }, { "epoch": 0.0002501, "grad_norm": 5.565805435180664, "learning_rate": 2.5e-05, "loss": 28.2031, "step": 2501 }, { "epoch": 0.0002502, "grad_norm": 5.629227638244629, "learning_rate": 2.501e-05, "loss": 26.8125, "step": 2502 }, { "epoch": 0.0002503, "grad_norm": 5.722161769866943, "learning_rate": 2.502e-05, "loss": 27.5625, "step": 2503 }, { "epoch": 0.0002504, "grad_norm": 6.053603649139404, "learning_rate": 2.503e-05, "loss": 30.2344, "step": 2504 }, { "epoch": 0.0002505, "grad_norm": 5.450896263122559, "learning_rate": 2.504e-05, "loss": 26.9062, "step": 2505 }, { "epoch": 0.0002506, "grad_norm": 5.463363170623779, "learning_rate": 2.505e-05, "loss": 26.875, "step": 2506 }, { "epoch": 0.0002507, "grad_norm": 5.679909706115723, "learning_rate": 2.5060000000000004e-05, "loss": 28.3594, "step": 2507 }, { "epoch": 0.0002508, "grad_norm": 5.594815731048584, "learning_rate": 2.5070000000000002e-05, "loss": 28.0625, "step": 2508 }, { "epoch": 0.0002509, "grad_norm": 5.433632850646973, "learning_rate": 2.5079999999999997e-05, "loss": 27.2812, "step": 2509 }, { "epoch": 0.000251, "grad_norm": 5.716043949127197, "learning_rate": 2.5090000000000002e-05, "loss": 27.6562, "step": 2510 }, { "epoch": 0.0002511, "grad_norm": 5.664153575897217, "learning_rate": 2.51e-05, "loss": 26.4688, "step": 2511 }, { "epoch": 0.0002512, "grad_norm": 5.534369945526123, "learning_rate": 2.5110000000000002e-05, "loss": 25.7031, "step": 2512 }, { "epoch": 0.0002513, "grad_norm": 5.515860080718994, "learning_rate": 2.512e-05, "loss": 27.1719, "step": 2513 }, { "epoch": 0.0002514, "grad_norm": 5.656839370727539, "learning_rate": 2.5129999999999998e-05, "loss": 27.2656, "step": 2514 }, { "epoch": 0.0002515, "grad_norm": 5.531663417816162, "learning_rate": 2.514e-05, "loss": 27.375, "step": 2515 }, { "epoch": 0.0002516, "grad_norm": 5.6716461181640625, "learning_rate": 2.5149999999999998e-05, "loss": 28.3594, "step": 2516 }, { "epoch": 0.0002517, "grad_norm": 5.447032451629639, "learning_rate": 2.5160000000000003e-05, "loss": 26.6562, "step": 2517 }, { "epoch": 0.0002518, "grad_norm": 5.604796409606934, "learning_rate": 2.517e-05, "loss": 26.3125, "step": 2518 }, { "epoch": 0.0002519, "grad_norm": 5.371697902679443, "learning_rate": 2.518e-05, "loss": 25.9375, "step": 2519 }, { "epoch": 0.000252, "grad_norm": 5.564846515655518, "learning_rate": 2.519e-05, "loss": 26.1875, "step": 2520 }, { "epoch": 0.0002521, "grad_norm": 6.0198073387146, "learning_rate": 2.52e-05, "loss": 29.3906, "step": 2521 }, { "epoch": 0.0002522, "grad_norm": 5.7497944831848145, "learning_rate": 2.521e-05, "loss": 27.8438, "step": 2522 }, { "epoch": 0.0002523, "grad_norm": 5.617159366607666, "learning_rate": 2.522e-05, "loss": 26.9844, "step": 2523 }, { "epoch": 0.0002524, "grad_norm": 5.888708591461182, "learning_rate": 2.5230000000000004e-05, "loss": 26.5469, "step": 2524 }, { "epoch": 0.0002525, "grad_norm": 5.707100868225098, "learning_rate": 2.5240000000000002e-05, "loss": 26.9062, "step": 2525 }, { "epoch": 0.0002526, "grad_norm": 5.636329650878906, "learning_rate": 2.5249999999999997e-05, "loss": 28.1406, "step": 2526 }, { "epoch": 0.0002527, "grad_norm": 5.696524143218994, "learning_rate": 2.5260000000000002e-05, "loss": 25.8281, "step": 2527 }, { "epoch": 0.0002528, "grad_norm": 5.661008834838867, "learning_rate": 2.527e-05, "loss": 26.9375, "step": 2528 }, { "epoch": 0.0002529, "grad_norm": 5.592834949493408, "learning_rate": 2.5280000000000002e-05, "loss": 25.75, "step": 2529 }, { "epoch": 0.000253, "grad_norm": 5.611721992492676, "learning_rate": 2.529e-05, "loss": 27.1875, "step": 2530 }, { "epoch": 0.0002531, "grad_norm": 5.668675899505615, "learning_rate": 2.53e-05, "loss": 25.8281, "step": 2531 }, { "epoch": 0.0002532, "grad_norm": 5.368985176086426, "learning_rate": 2.531e-05, "loss": 26.125, "step": 2532 }, { "epoch": 0.0002533, "grad_norm": 5.8484110832214355, "learning_rate": 2.5319999999999998e-05, "loss": 27.2188, "step": 2533 }, { "epoch": 0.0002534, "grad_norm": 5.6370415687561035, "learning_rate": 2.5330000000000003e-05, "loss": 26.9375, "step": 2534 }, { "epoch": 0.0002535, "grad_norm": 6.0774827003479, "learning_rate": 2.534e-05, "loss": 28.0469, "step": 2535 }, { "epoch": 0.0002536, "grad_norm": 5.581841945648193, "learning_rate": 2.535e-05, "loss": 25.9375, "step": 2536 }, { "epoch": 0.0002537, "grad_norm": 5.850471019744873, "learning_rate": 2.536e-05, "loss": 26.6094, "step": 2537 }, { "epoch": 0.0002538, "grad_norm": 5.686932563781738, "learning_rate": 2.537e-05, "loss": 25.8125, "step": 2538 }, { "epoch": 0.0002539, "grad_norm": 5.479832649230957, "learning_rate": 2.538e-05, "loss": 25.8906, "step": 2539 }, { "epoch": 0.000254, "grad_norm": 5.452784538269043, "learning_rate": 2.539e-05, "loss": 25.0469, "step": 2540 }, { "epoch": 0.0002541, "grad_norm": 5.936349868774414, "learning_rate": 2.5400000000000004e-05, "loss": 28.0938, "step": 2541 }, { "epoch": 0.0002542, "grad_norm": 5.7082343101501465, "learning_rate": 2.5410000000000002e-05, "loss": 26.1406, "step": 2542 }, { "epoch": 0.0002543, "grad_norm": 5.362971305847168, "learning_rate": 2.5419999999999997e-05, "loss": 26.2188, "step": 2543 }, { "epoch": 0.0002544, "grad_norm": 5.527894496917725, "learning_rate": 2.5430000000000002e-05, "loss": 26.7969, "step": 2544 }, { "epoch": 0.0002545, "grad_norm": 5.354765892028809, "learning_rate": 2.544e-05, "loss": 26.4688, "step": 2545 }, { "epoch": 0.0002546, "grad_norm": 5.410459518432617, "learning_rate": 2.5450000000000002e-05, "loss": 25.8906, "step": 2546 }, { "epoch": 0.0002547, "grad_norm": 5.735154151916504, "learning_rate": 2.546e-05, "loss": 27.8906, "step": 2547 }, { "epoch": 0.0002548, "grad_norm": 5.661908149719238, "learning_rate": 2.547e-05, "loss": 27.6875, "step": 2548 }, { "epoch": 0.0002549, "grad_norm": 5.512752532958984, "learning_rate": 2.548e-05, "loss": 25.75, "step": 2549 }, { "epoch": 0.000255, "grad_norm": 6.202521800994873, "learning_rate": 2.5489999999999998e-05, "loss": 27.4219, "step": 2550 }, { "epoch": 0.0002551, "grad_norm": 5.622015476226807, "learning_rate": 2.5500000000000003e-05, "loss": 25.9219, "step": 2551 }, { "epoch": 0.0002552, "grad_norm": 5.419823169708252, "learning_rate": 2.551e-05, "loss": 25.4844, "step": 2552 }, { "epoch": 0.0002553, "grad_norm": 5.312170028686523, "learning_rate": 2.552e-05, "loss": 24.8438, "step": 2553 }, { "epoch": 0.0002554, "grad_norm": 5.4918951988220215, "learning_rate": 2.553e-05, "loss": 24.9375, "step": 2554 }, { "epoch": 0.0002555, "grad_norm": 5.547058582305908, "learning_rate": 2.554e-05, "loss": 26.5, "step": 2555 }, { "epoch": 0.0002556, "grad_norm": 5.379054069519043, "learning_rate": 2.555e-05, "loss": 25.6562, "step": 2556 }, { "epoch": 0.0002557, "grad_norm": 5.7195868492126465, "learning_rate": 2.556e-05, "loss": 27.6719, "step": 2557 }, { "epoch": 0.0002558, "grad_norm": 5.3323283195495605, "learning_rate": 2.5570000000000004e-05, "loss": 25.2812, "step": 2558 }, { "epoch": 0.0002559, "grad_norm": 5.701965808868408, "learning_rate": 2.5580000000000002e-05, "loss": 26.7344, "step": 2559 }, { "epoch": 0.000256, "grad_norm": 5.689982891082764, "learning_rate": 2.5589999999999997e-05, "loss": 26.9844, "step": 2560 }, { "epoch": 0.0002561, "grad_norm": 5.651844024658203, "learning_rate": 2.5600000000000002e-05, "loss": 26.0781, "step": 2561 }, { "epoch": 0.0002562, "grad_norm": 5.369045257568359, "learning_rate": 2.561e-05, "loss": 25.6406, "step": 2562 }, { "epoch": 0.0002563, "grad_norm": 5.757684707641602, "learning_rate": 2.5620000000000002e-05, "loss": 26.3125, "step": 2563 }, { "epoch": 0.0002564, "grad_norm": 5.466500759124756, "learning_rate": 2.563e-05, "loss": 24.3594, "step": 2564 }, { "epoch": 0.0002565, "grad_norm": 5.516260623931885, "learning_rate": 2.564e-05, "loss": 23.8281, "step": 2565 }, { "epoch": 0.0002566, "grad_norm": 5.385896682739258, "learning_rate": 2.565e-05, "loss": 25.0469, "step": 2566 }, { "epoch": 0.0002567, "grad_norm": 5.386533260345459, "learning_rate": 2.5659999999999998e-05, "loss": 24.125, "step": 2567 }, { "epoch": 0.0002568, "grad_norm": 5.237969398498535, "learning_rate": 2.5670000000000003e-05, "loss": 23.7344, "step": 2568 }, { "epoch": 0.0002569, "grad_norm": 5.933482646942139, "learning_rate": 2.568e-05, "loss": 27.0156, "step": 2569 }, { "epoch": 0.000257, "grad_norm": 5.208066463470459, "learning_rate": 2.569e-05, "loss": 24.875, "step": 2570 }, { "epoch": 0.0002571, "grad_norm": 5.539393901824951, "learning_rate": 2.57e-05, "loss": 25.2969, "step": 2571 }, { "epoch": 0.0002572, "grad_norm": 5.1189284324646, "learning_rate": 2.571e-05, "loss": 24.3281, "step": 2572 }, { "epoch": 0.0002573, "grad_norm": 5.8175249099731445, "learning_rate": 2.572e-05, "loss": 25.2031, "step": 2573 }, { "epoch": 0.0002574, "grad_norm": 5.154368877410889, "learning_rate": 2.573e-05, "loss": 24.4844, "step": 2574 }, { "epoch": 0.0002575, "grad_norm": 5.2388410568237305, "learning_rate": 2.5740000000000004e-05, "loss": 24.5781, "step": 2575 }, { "epoch": 0.0002576, "grad_norm": 5.49668550491333, "learning_rate": 2.5750000000000002e-05, "loss": 24.8906, "step": 2576 }, { "epoch": 0.0002577, "grad_norm": 5.300713062286377, "learning_rate": 2.5759999999999997e-05, "loss": 24.6562, "step": 2577 }, { "epoch": 0.0002578, "grad_norm": 5.376718521118164, "learning_rate": 2.5770000000000002e-05, "loss": 25.2969, "step": 2578 }, { "epoch": 0.0002579, "grad_norm": 5.725271224975586, "learning_rate": 2.578e-05, "loss": 26.4531, "step": 2579 }, { "epoch": 0.000258, "grad_norm": 5.48429012298584, "learning_rate": 2.5790000000000002e-05, "loss": 24.7031, "step": 2580 }, { "epoch": 0.0002581, "grad_norm": 5.384649753570557, "learning_rate": 2.58e-05, "loss": 25.3438, "step": 2581 }, { "epoch": 0.0002582, "grad_norm": 5.443394184112549, "learning_rate": 2.581e-05, "loss": 24.2344, "step": 2582 }, { "epoch": 0.0002583, "grad_norm": 5.318784713745117, "learning_rate": 2.582e-05, "loss": 25.0938, "step": 2583 }, { "epoch": 0.0002584, "grad_norm": 5.316436767578125, "learning_rate": 2.583e-05, "loss": 24.8438, "step": 2584 }, { "epoch": 0.0002585, "grad_norm": 5.334114074707031, "learning_rate": 2.5840000000000003e-05, "loss": 23.9062, "step": 2585 }, { "epoch": 0.0002586, "grad_norm": 5.872819423675537, "learning_rate": 2.585e-05, "loss": 25.7656, "step": 2586 }, { "epoch": 0.0002587, "grad_norm": 5.381070137023926, "learning_rate": 2.586e-05, "loss": 24.4531, "step": 2587 }, { "epoch": 0.0002588, "grad_norm": 5.218656539916992, "learning_rate": 2.587e-05, "loss": 23.9531, "step": 2588 }, { "epoch": 0.0002589, "grad_norm": 5.446895122528076, "learning_rate": 2.588e-05, "loss": 25.7812, "step": 2589 }, { "epoch": 0.000259, "grad_norm": 5.373314380645752, "learning_rate": 2.589e-05, "loss": 24.4688, "step": 2590 }, { "epoch": 0.0002591, "grad_norm": 5.229781150817871, "learning_rate": 2.59e-05, "loss": 24.1562, "step": 2591 }, { "epoch": 0.0002592, "grad_norm": 5.208466529846191, "learning_rate": 2.5910000000000004e-05, "loss": 24.125, "step": 2592 }, { "epoch": 0.0002593, "grad_norm": 6.271269798278809, "learning_rate": 2.5920000000000003e-05, "loss": 25.2344, "step": 2593 }, { "epoch": 0.0002594, "grad_norm": 5.487994194030762, "learning_rate": 2.5929999999999997e-05, "loss": 23.5469, "step": 2594 }, { "epoch": 0.0002595, "grad_norm": 5.253428936004639, "learning_rate": 2.5940000000000002e-05, "loss": 23.6875, "step": 2595 }, { "epoch": 0.0002596, "grad_norm": 5.854215621948242, "learning_rate": 2.595e-05, "loss": 26.0469, "step": 2596 }, { "epoch": 0.0002597, "grad_norm": 5.485320091247559, "learning_rate": 2.5960000000000002e-05, "loss": 24.5, "step": 2597 }, { "epoch": 0.0002598, "grad_norm": 5.7917633056640625, "learning_rate": 2.597e-05, "loss": 25.7031, "step": 2598 }, { "epoch": 0.0002599, "grad_norm": 5.62728214263916, "learning_rate": 2.598e-05, "loss": 25.9375, "step": 2599 }, { "epoch": 0.00026, "grad_norm": 5.214705467224121, "learning_rate": 2.599e-05, "loss": 23.7969, "step": 2600 }, { "epoch": 0.0002601, "grad_norm": 5.2012505531311035, "learning_rate": 2.6e-05, "loss": 24.2656, "step": 2601 }, { "epoch": 0.0002602, "grad_norm": 5.1622772216796875, "learning_rate": 2.6010000000000003e-05, "loss": 23.3594, "step": 2602 }, { "epoch": 0.0002603, "grad_norm": 5.09535026550293, "learning_rate": 2.602e-05, "loss": 22.7969, "step": 2603 }, { "epoch": 0.0002604, "grad_norm": 5.615042209625244, "learning_rate": 2.603e-05, "loss": 25.6875, "step": 2604 }, { "epoch": 0.0002605, "grad_norm": 5.217511177062988, "learning_rate": 2.604e-05, "loss": 24.5312, "step": 2605 }, { "epoch": 0.0002606, "grad_norm": 5.039108753204346, "learning_rate": 2.605e-05, "loss": 22.9531, "step": 2606 }, { "epoch": 0.0002607, "grad_norm": 4.942178726196289, "learning_rate": 2.606e-05, "loss": 22.2344, "step": 2607 }, { "epoch": 0.0002608, "grad_norm": 5.154867649078369, "learning_rate": 2.607e-05, "loss": 23.6094, "step": 2608 }, { "epoch": 0.0002609, "grad_norm": 5.272024154663086, "learning_rate": 2.6080000000000004e-05, "loss": 24.5469, "step": 2609 }, { "epoch": 0.000261, "grad_norm": 5.045335292816162, "learning_rate": 2.609e-05, "loss": 22.7344, "step": 2610 }, { "epoch": 0.0002611, "grad_norm": 5.299790859222412, "learning_rate": 2.6099999999999997e-05, "loss": 23.1875, "step": 2611 }, { "epoch": 0.0002612, "grad_norm": 5.388454437255859, "learning_rate": 2.6110000000000002e-05, "loss": 23.4219, "step": 2612 }, { "epoch": 0.0002613, "grad_norm": 5.153306007385254, "learning_rate": 2.612e-05, "loss": 22.8438, "step": 2613 }, { "epoch": 0.0002614, "grad_norm": 5.241060256958008, "learning_rate": 2.6130000000000002e-05, "loss": 23.3438, "step": 2614 }, { "epoch": 0.0002615, "grad_norm": 5.518542766571045, "learning_rate": 2.614e-05, "loss": 24.3438, "step": 2615 }, { "epoch": 0.0002616, "grad_norm": 5.116453170776367, "learning_rate": 2.615e-05, "loss": 24.3906, "step": 2616 }, { "epoch": 0.0002617, "grad_norm": 5.2315168380737305, "learning_rate": 2.616e-05, "loss": 23.8438, "step": 2617 }, { "epoch": 0.0002618, "grad_norm": 5.10846471786499, "learning_rate": 2.617e-05, "loss": 23.2344, "step": 2618 }, { "epoch": 0.0002619, "grad_norm": 5.099444389343262, "learning_rate": 2.6180000000000003e-05, "loss": 23.5938, "step": 2619 }, { "epoch": 0.000262, "grad_norm": 5.21929407119751, "learning_rate": 2.619e-05, "loss": 24.0469, "step": 2620 }, { "epoch": 0.0002621, "grad_norm": 5.284022808074951, "learning_rate": 2.6199999999999996e-05, "loss": 22.9375, "step": 2621 }, { "epoch": 0.0002622, "grad_norm": 5.495255947113037, "learning_rate": 2.621e-05, "loss": 24.1719, "step": 2622 }, { "epoch": 0.0002623, "grad_norm": 5.054660320281982, "learning_rate": 2.622e-05, "loss": 23.1562, "step": 2623 }, { "epoch": 0.0002624, "grad_norm": 5.133628845214844, "learning_rate": 2.623e-05, "loss": 23.4844, "step": 2624 }, { "epoch": 0.0002625, "grad_norm": 5.0620198249816895, "learning_rate": 2.624e-05, "loss": 22.625, "step": 2625 }, { "epoch": 0.0002626, "grad_norm": 5.08545446395874, "learning_rate": 2.6250000000000004e-05, "loss": 23.4375, "step": 2626 }, { "epoch": 0.0002627, "grad_norm": 5.254687786102295, "learning_rate": 2.626e-05, "loss": 23.0312, "step": 2627 }, { "epoch": 0.0002628, "grad_norm": 5.00596809387207, "learning_rate": 2.6269999999999998e-05, "loss": 22.6875, "step": 2628 }, { "epoch": 0.0002629, "grad_norm": 4.996240615844727, "learning_rate": 2.6280000000000002e-05, "loss": 21.7812, "step": 2629 }, { "epoch": 0.000263, "grad_norm": 5.197492599487305, "learning_rate": 2.629e-05, "loss": 22.1094, "step": 2630 }, { "epoch": 0.0002631, "grad_norm": 5.24721622467041, "learning_rate": 2.6300000000000002e-05, "loss": 23.0938, "step": 2631 }, { "epoch": 0.0002632, "grad_norm": 5.211394786834717, "learning_rate": 2.631e-05, "loss": 22.5625, "step": 2632 }, { "epoch": 0.0002633, "grad_norm": 5.177051067352295, "learning_rate": 2.632e-05, "loss": 23.1562, "step": 2633 }, { "epoch": 0.0002634, "grad_norm": 5.074063301086426, "learning_rate": 2.633e-05, "loss": 22.1719, "step": 2634 }, { "epoch": 0.0002635, "grad_norm": 5.432488918304443, "learning_rate": 2.634e-05, "loss": 24.5938, "step": 2635 }, { "epoch": 0.0002636, "grad_norm": 5.0768256187438965, "learning_rate": 2.6350000000000004e-05, "loss": 21.1719, "step": 2636 }, { "epoch": 0.0002637, "grad_norm": 5.476498126983643, "learning_rate": 2.6360000000000002e-05, "loss": 22.4844, "step": 2637 }, { "epoch": 0.0002638, "grad_norm": 5.315165996551514, "learning_rate": 2.6369999999999997e-05, "loss": 23.0938, "step": 2638 }, { "epoch": 0.0002639, "grad_norm": 5.138946533203125, "learning_rate": 2.638e-05, "loss": 22.9531, "step": 2639 }, { "epoch": 0.000264, "grad_norm": 5.492170333862305, "learning_rate": 2.639e-05, "loss": 23.8281, "step": 2640 }, { "epoch": 0.0002641, "grad_norm": 4.930859088897705, "learning_rate": 2.64e-05, "loss": 20.8906, "step": 2641 }, { "epoch": 0.0002642, "grad_norm": 5.004940032958984, "learning_rate": 2.641e-05, "loss": 22.2656, "step": 2642 }, { "epoch": 0.0002643, "grad_norm": 5.593879699707031, "learning_rate": 2.6420000000000005e-05, "loss": 22.625, "step": 2643 }, { "epoch": 0.0002644, "grad_norm": 5.199212551116943, "learning_rate": 2.643e-05, "loss": 22.1562, "step": 2644 }, { "epoch": 0.0002645, "grad_norm": 5.100207805633545, "learning_rate": 2.6439999999999998e-05, "loss": 22.3594, "step": 2645 }, { "epoch": 0.0002646, "grad_norm": 5.382095813751221, "learning_rate": 2.6450000000000003e-05, "loss": 23.0781, "step": 2646 }, { "epoch": 0.0002647, "grad_norm": 5.44952917098999, "learning_rate": 2.646e-05, "loss": 23.25, "step": 2647 }, { "epoch": 0.0002648, "grad_norm": 4.958463668823242, "learning_rate": 2.6470000000000002e-05, "loss": 21.375, "step": 2648 }, { "epoch": 0.0002649, "grad_norm": 5.032123565673828, "learning_rate": 2.648e-05, "loss": 21.1875, "step": 2649 }, { "epoch": 0.000265, "grad_norm": 5.24052619934082, "learning_rate": 2.649e-05, "loss": 23.0, "step": 2650 }, { "epoch": 0.0002651, "grad_norm": 5.07439661026001, "learning_rate": 2.65e-05, "loss": 21.0, "step": 2651 }, { "epoch": 0.0002652, "grad_norm": 5.466571807861328, "learning_rate": 2.651e-05, "loss": 23.2812, "step": 2652 }, { "epoch": 0.0002653, "grad_norm": 5.178341865539551, "learning_rate": 2.6520000000000004e-05, "loss": 22.4688, "step": 2653 }, { "epoch": 0.0002654, "grad_norm": 5.437129020690918, "learning_rate": 2.6530000000000002e-05, "loss": 21.8438, "step": 2654 }, { "epoch": 0.0002655, "grad_norm": 4.963681697845459, "learning_rate": 2.6539999999999997e-05, "loss": 20.8906, "step": 2655 }, { "epoch": 0.0002656, "grad_norm": 5.6259965896606445, "learning_rate": 2.655e-05, "loss": 22.7031, "step": 2656 }, { "epoch": 0.0002657, "grad_norm": 5.354860782623291, "learning_rate": 2.656e-05, "loss": 22.9062, "step": 2657 }, { "epoch": 0.0002658, "grad_norm": 4.999288082122803, "learning_rate": 2.657e-05, "loss": 22.0, "step": 2658 }, { "epoch": 0.0002659, "grad_norm": 5.121066570281982, "learning_rate": 2.658e-05, "loss": 21.9062, "step": 2659 }, { "epoch": 0.000266, "grad_norm": 5.148739337921143, "learning_rate": 2.6590000000000005e-05, "loss": 22.5156, "step": 2660 }, { "epoch": 0.0002661, "grad_norm": 5.037651538848877, "learning_rate": 2.66e-05, "loss": 21.4844, "step": 2661 }, { "epoch": 0.0002662, "grad_norm": 4.847867965698242, "learning_rate": 2.6609999999999998e-05, "loss": 21.2344, "step": 2662 }, { "epoch": 0.0002663, "grad_norm": 5.017890453338623, "learning_rate": 2.6620000000000003e-05, "loss": 21.4688, "step": 2663 }, { "epoch": 0.0002664, "grad_norm": 5.438411235809326, "learning_rate": 2.663e-05, "loss": 22.8125, "step": 2664 }, { "epoch": 0.0002665, "grad_norm": 5.020574569702148, "learning_rate": 2.6640000000000002e-05, "loss": 21.7031, "step": 2665 }, { "epoch": 0.0002666, "grad_norm": 5.7231268882751465, "learning_rate": 2.665e-05, "loss": 24.125, "step": 2666 }, { "epoch": 0.0002667, "grad_norm": 4.928754806518555, "learning_rate": 2.666e-05, "loss": 20.0156, "step": 2667 }, { "epoch": 0.0002668, "grad_norm": 5.754295825958252, "learning_rate": 2.667e-05, "loss": 22.7812, "step": 2668 }, { "epoch": 0.0002669, "grad_norm": 5.014497756958008, "learning_rate": 2.668e-05, "loss": 19.8281, "step": 2669 }, { "epoch": 0.000267, "grad_norm": 5.082116603851318, "learning_rate": 2.6690000000000004e-05, "loss": 22.4375, "step": 2670 }, { "epoch": 0.0002671, "grad_norm": 5.489953994750977, "learning_rate": 2.6700000000000002e-05, "loss": 22.3438, "step": 2671 }, { "epoch": 0.0002672, "grad_norm": 4.900982856750488, "learning_rate": 2.6709999999999997e-05, "loss": 21.2188, "step": 2672 }, { "epoch": 0.0002673, "grad_norm": 4.959640979766846, "learning_rate": 2.672e-05, "loss": 20.0156, "step": 2673 }, { "epoch": 0.0002674, "grad_norm": 5.18635892868042, "learning_rate": 2.673e-05, "loss": 20.9844, "step": 2674 }, { "epoch": 0.0002675, "grad_norm": 4.935641765594482, "learning_rate": 2.674e-05, "loss": 21.3906, "step": 2675 }, { "epoch": 0.0002676, "grad_norm": 5.382370471954346, "learning_rate": 2.675e-05, "loss": 21.875, "step": 2676 }, { "epoch": 0.0002677, "grad_norm": 4.926283836364746, "learning_rate": 2.6760000000000005e-05, "loss": 21.0469, "step": 2677 }, { "epoch": 0.0002678, "grad_norm": 5.02637243270874, "learning_rate": 2.677e-05, "loss": 19.5938, "step": 2678 }, { "epoch": 0.0002679, "grad_norm": 4.921310901641846, "learning_rate": 2.6779999999999998e-05, "loss": 21.2656, "step": 2679 }, { "epoch": 0.000268, "grad_norm": 5.093906879425049, "learning_rate": 2.6790000000000003e-05, "loss": 21.1406, "step": 2680 }, { "epoch": 0.0002681, "grad_norm": 4.9828782081604, "learning_rate": 2.68e-05, "loss": 21.2969, "step": 2681 }, { "epoch": 0.0002682, "grad_norm": 5.1273908615112305, "learning_rate": 2.6810000000000003e-05, "loss": 21.9688, "step": 2682 }, { "epoch": 0.0002683, "grad_norm": 5.088643550872803, "learning_rate": 2.682e-05, "loss": 20.8594, "step": 2683 }, { "epoch": 0.0002684, "grad_norm": 5.268448829650879, "learning_rate": 2.683e-05, "loss": 21.7656, "step": 2684 }, { "epoch": 0.0002685, "grad_norm": 6.2556071281433105, "learning_rate": 2.684e-05, "loss": 21.6406, "step": 2685 }, { "epoch": 0.0002686, "grad_norm": 5.0257344245910645, "learning_rate": 2.685e-05, "loss": 20.2031, "step": 2686 }, { "epoch": 0.0002687, "grad_norm": 5.192383766174316, "learning_rate": 2.6860000000000004e-05, "loss": 23.3906, "step": 2687 }, { "epoch": 0.0002688, "grad_norm": 5.2807841300964355, "learning_rate": 2.6870000000000002e-05, "loss": 21.75, "step": 2688 }, { "epoch": 0.0002689, "grad_norm": 4.875032901763916, "learning_rate": 2.6879999999999997e-05, "loss": 20.875, "step": 2689 }, { "epoch": 0.000269, "grad_norm": 5.20595645904541, "learning_rate": 2.6890000000000002e-05, "loss": 22.2969, "step": 2690 }, { "epoch": 0.0002691, "grad_norm": 5.164567947387695, "learning_rate": 2.69e-05, "loss": 21.2031, "step": 2691 }, { "epoch": 0.0002692, "grad_norm": 4.97483491897583, "learning_rate": 2.691e-05, "loss": 21.2344, "step": 2692 }, { "epoch": 0.0002693, "grad_norm": 4.959793567657471, "learning_rate": 2.692e-05, "loss": 20.7812, "step": 2693 }, { "epoch": 0.0002694, "grad_norm": 5.119083404541016, "learning_rate": 2.6930000000000005e-05, "loss": 21.9688, "step": 2694 }, { "epoch": 0.0002695, "grad_norm": 5.160106658935547, "learning_rate": 2.694e-05, "loss": 20.7344, "step": 2695 }, { "epoch": 0.0002696, "grad_norm": 5.175765037536621, "learning_rate": 2.6949999999999998e-05, "loss": 21.8594, "step": 2696 }, { "epoch": 0.0002697, "grad_norm": 4.989496231079102, "learning_rate": 2.6960000000000003e-05, "loss": 20.0625, "step": 2697 }, { "epoch": 0.0002698, "grad_norm": 5.033694744110107, "learning_rate": 2.697e-05, "loss": 21.5938, "step": 2698 }, { "epoch": 0.0002699, "grad_norm": 5.112461566925049, "learning_rate": 2.6980000000000003e-05, "loss": 21.2188, "step": 2699 }, { "epoch": 0.00027, "grad_norm": 5.277057647705078, "learning_rate": 2.699e-05, "loss": 22.5781, "step": 2700 }, { "epoch": 0.0002701, "grad_norm": 4.946776866912842, "learning_rate": 2.7e-05, "loss": 19.9531, "step": 2701 }, { "epoch": 0.0002702, "grad_norm": 5.016019344329834, "learning_rate": 2.701e-05, "loss": 20.625, "step": 2702 }, { "epoch": 0.0002703, "grad_norm": 5.005223751068115, "learning_rate": 2.702e-05, "loss": 20.1719, "step": 2703 }, { "epoch": 0.0002704, "grad_norm": 5.02770471572876, "learning_rate": 2.7030000000000004e-05, "loss": 20.7031, "step": 2704 }, { "epoch": 0.0002705, "grad_norm": 5.305837154388428, "learning_rate": 2.7040000000000002e-05, "loss": 21.8906, "step": 2705 }, { "epoch": 0.0002706, "grad_norm": 4.983026504516602, "learning_rate": 2.7049999999999997e-05, "loss": 20.375, "step": 2706 }, { "epoch": 0.0002707, "grad_norm": 4.904435157775879, "learning_rate": 2.7060000000000002e-05, "loss": 20.7969, "step": 2707 }, { "epoch": 0.0002708, "grad_norm": 5.135472774505615, "learning_rate": 2.707e-05, "loss": 21.5625, "step": 2708 }, { "epoch": 0.0002709, "grad_norm": 6.009512901306152, "learning_rate": 2.708e-05, "loss": 21.1562, "step": 2709 }, { "epoch": 0.000271, "grad_norm": 4.726728916168213, "learning_rate": 2.709e-05, "loss": 19.8594, "step": 2710 }, { "epoch": 0.0002711, "grad_norm": 5.085262775421143, "learning_rate": 2.7100000000000005e-05, "loss": 21.5938, "step": 2711 }, { "epoch": 0.0002712, "grad_norm": 4.885828495025635, "learning_rate": 2.711e-05, "loss": 20.6406, "step": 2712 }, { "epoch": 0.0002713, "grad_norm": 4.73411226272583, "learning_rate": 2.7119999999999998e-05, "loss": 20.4531, "step": 2713 }, { "epoch": 0.0002714, "grad_norm": 4.954255104064941, "learning_rate": 2.7130000000000003e-05, "loss": 21.0156, "step": 2714 }, { "epoch": 0.0002715, "grad_norm": 4.919240474700928, "learning_rate": 2.714e-05, "loss": 21.2812, "step": 2715 }, { "epoch": 0.0002716, "grad_norm": 4.868825435638428, "learning_rate": 2.7150000000000003e-05, "loss": 19.2031, "step": 2716 }, { "epoch": 0.0002717, "grad_norm": 4.885971546173096, "learning_rate": 2.716e-05, "loss": 21.4531, "step": 2717 }, { "epoch": 0.0002718, "grad_norm": 4.8488287925720215, "learning_rate": 2.717e-05, "loss": 18.75, "step": 2718 }, { "epoch": 0.0002719, "grad_norm": 5.474164009094238, "learning_rate": 2.718e-05, "loss": 22.2188, "step": 2719 }, { "epoch": 0.000272, "grad_norm": 4.9402265548706055, "learning_rate": 2.719e-05, "loss": 20.375, "step": 2720 }, { "epoch": 0.0002721, "grad_norm": 5.054856777191162, "learning_rate": 2.7200000000000004e-05, "loss": 19.9688, "step": 2721 }, { "epoch": 0.0002722, "grad_norm": 5.139786243438721, "learning_rate": 2.7210000000000002e-05, "loss": 19.9531, "step": 2722 }, { "epoch": 0.0002723, "grad_norm": 4.936676502227783, "learning_rate": 2.7219999999999997e-05, "loss": 20.2344, "step": 2723 }, { "epoch": 0.0002724, "grad_norm": 5.021756172180176, "learning_rate": 2.7230000000000002e-05, "loss": 20.6562, "step": 2724 }, { "epoch": 0.0002725, "grad_norm": 5.148033618927002, "learning_rate": 2.724e-05, "loss": 19.6406, "step": 2725 }, { "epoch": 0.0002726, "grad_norm": 4.9928998947143555, "learning_rate": 2.725e-05, "loss": 20.0625, "step": 2726 }, { "epoch": 0.0002727, "grad_norm": 5.311802864074707, "learning_rate": 2.726e-05, "loss": 21.9688, "step": 2727 }, { "epoch": 0.0002728, "grad_norm": 4.907752990722656, "learning_rate": 2.727e-05, "loss": 20.0469, "step": 2728 }, { "epoch": 0.0002729, "grad_norm": 4.842103958129883, "learning_rate": 2.728e-05, "loss": 20.9219, "step": 2729 }, { "epoch": 0.000273, "grad_norm": 5.0052490234375, "learning_rate": 2.7289999999999998e-05, "loss": 21.4062, "step": 2730 }, { "epoch": 0.0002731, "grad_norm": 5.285139083862305, "learning_rate": 2.7300000000000003e-05, "loss": 20.6406, "step": 2731 }, { "epoch": 0.0002732, "grad_norm": 5.466280460357666, "learning_rate": 2.731e-05, "loss": 20.6719, "step": 2732 }, { "epoch": 0.0002733, "grad_norm": 5.396193504333496, "learning_rate": 2.7320000000000003e-05, "loss": 21.1875, "step": 2733 }, { "epoch": 0.0002734, "grad_norm": 4.7792510986328125, "learning_rate": 2.733e-05, "loss": 18.6094, "step": 2734 }, { "epoch": 0.0002735, "grad_norm": 5.213866233825684, "learning_rate": 2.734e-05, "loss": 20.8594, "step": 2735 }, { "epoch": 0.0002736, "grad_norm": 4.743049144744873, "learning_rate": 2.735e-05, "loss": 18.3438, "step": 2736 }, { "epoch": 0.0002737, "grad_norm": 5.470842361450195, "learning_rate": 2.736e-05, "loss": 21.6875, "step": 2737 }, { "epoch": 0.0002738, "grad_norm": 4.775577545166016, "learning_rate": 2.7370000000000004e-05, "loss": 19.0469, "step": 2738 }, { "epoch": 0.0002739, "grad_norm": 4.959975242614746, "learning_rate": 2.738e-05, "loss": 19.3906, "step": 2739 }, { "epoch": 0.000274, "grad_norm": 4.86496639251709, "learning_rate": 2.7389999999999997e-05, "loss": 20.4062, "step": 2740 }, { "epoch": 0.0002741, "grad_norm": 5.065260410308838, "learning_rate": 2.7400000000000002e-05, "loss": 19.5781, "step": 2741 }, { "epoch": 0.0002742, "grad_norm": 4.815097808837891, "learning_rate": 2.741e-05, "loss": 19.3125, "step": 2742 }, { "epoch": 0.0002743, "grad_norm": 4.8369622230529785, "learning_rate": 2.7420000000000002e-05, "loss": 18.8125, "step": 2743 }, { "epoch": 0.0002744, "grad_norm": 4.735267162322998, "learning_rate": 2.743e-05, "loss": 18.9844, "step": 2744 }, { "epoch": 0.0002745, "grad_norm": 4.8648200035095215, "learning_rate": 2.744e-05, "loss": 19.6875, "step": 2745 }, { "epoch": 0.0002746, "grad_norm": 4.790833950042725, "learning_rate": 2.745e-05, "loss": 19.6406, "step": 2746 }, { "epoch": 0.0002747, "grad_norm": 4.881219863891602, "learning_rate": 2.7459999999999998e-05, "loss": 18.8594, "step": 2747 }, { "epoch": 0.0002748, "grad_norm": 4.830053329467773, "learning_rate": 2.7470000000000003e-05, "loss": 19.0625, "step": 2748 }, { "epoch": 0.0002749, "grad_norm": 4.878846645355225, "learning_rate": 2.748e-05, "loss": 18.7656, "step": 2749 }, { "epoch": 0.000275, "grad_norm": 4.916638374328613, "learning_rate": 2.7490000000000003e-05, "loss": 19.2969, "step": 2750 }, { "epoch": 0.0002751, "grad_norm": 5.475046157836914, "learning_rate": 2.75e-05, "loss": 21.4062, "step": 2751 }, { "epoch": 0.0002752, "grad_norm": 5.810745716094971, "learning_rate": 2.751e-05, "loss": 21.0781, "step": 2752 }, { "epoch": 0.0002753, "grad_norm": 5.02387809753418, "learning_rate": 2.752e-05, "loss": 20.0781, "step": 2753 }, { "epoch": 0.0002754, "grad_norm": 4.774989604949951, "learning_rate": 2.753e-05, "loss": 19.0938, "step": 2754 }, { "epoch": 0.0002755, "grad_norm": 5.187727451324463, "learning_rate": 2.7540000000000004e-05, "loss": 21.2812, "step": 2755 }, { "epoch": 0.0002756, "grad_norm": 4.9502739906311035, "learning_rate": 2.755e-05, "loss": 19.2344, "step": 2756 }, { "epoch": 0.0002757, "grad_norm": 5.580729961395264, "learning_rate": 2.7559999999999997e-05, "loss": 19.6406, "step": 2757 }, { "epoch": 0.0002758, "grad_norm": 4.933372974395752, "learning_rate": 2.7570000000000002e-05, "loss": 18.1406, "step": 2758 }, { "epoch": 0.0002759, "grad_norm": 4.867650985717773, "learning_rate": 2.758e-05, "loss": 18.0469, "step": 2759 }, { "epoch": 0.000276, "grad_norm": 4.838322639465332, "learning_rate": 2.7590000000000002e-05, "loss": 19.2656, "step": 2760 }, { "epoch": 0.0002761, "grad_norm": 5.078263759613037, "learning_rate": 2.76e-05, "loss": 20.875, "step": 2761 }, { "epoch": 0.0002762, "grad_norm": 4.791270732879639, "learning_rate": 2.761e-05, "loss": 18.2969, "step": 2762 }, { "epoch": 0.0002763, "grad_norm": 4.931293487548828, "learning_rate": 2.762e-05, "loss": 19.75, "step": 2763 }, { "epoch": 0.0002764, "grad_norm": 4.926638126373291, "learning_rate": 2.7629999999999998e-05, "loss": 19.4688, "step": 2764 }, { "epoch": 0.0002765, "grad_norm": 4.666944980621338, "learning_rate": 2.7640000000000003e-05, "loss": 19.1875, "step": 2765 }, { "epoch": 0.0002766, "grad_norm": 5.19611930847168, "learning_rate": 2.765e-05, "loss": 20.4531, "step": 2766 }, { "epoch": 0.0002767, "grad_norm": 4.730144023895264, "learning_rate": 2.7660000000000003e-05, "loss": 19.5312, "step": 2767 }, { "epoch": 0.0002768, "grad_norm": 4.687786102294922, "learning_rate": 2.767e-05, "loss": 17.7812, "step": 2768 }, { "epoch": 0.0002769, "grad_norm": 4.741515159606934, "learning_rate": 2.768e-05, "loss": 18.6562, "step": 2769 }, { "epoch": 0.000277, "grad_norm": 4.674302101135254, "learning_rate": 2.769e-05, "loss": 18.75, "step": 2770 }, { "epoch": 0.0002771, "grad_norm": 4.807997226715088, "learning_rate": 2.77e-05, "loss": 19.6719, "step": 2771 }, { "epoch": 0.0002772, "grad_norm": 4.6543169021606445, "learning_rate": 2.7710000000000004e-05, "loss": 17.7656, "step": 2772 }, { "epoch": 0.0002773, "grad_norm": 4.848941326141357, "learning_rate": 2.772e-05, "loss": 18.6094, "step": 2773 }, { "epoch": 0.0002774, "grad_norm": 4.800899028778076, "learning_rate": 2.7729999999999997e-05, "loss": 18.8828, "step": 2774 }, { "epoch": 0.0002775, "grad_norm": 4.714990615844727, "learning_rate": 2.7740000000000002e-05, "loss": 17.7344, "step": 2775 }, { "epoch": 0.0002776, "grad_norm": 4.894636631011963, "learning_rate": 2.775e-05, "loss": 18.5938, "step": 2776 }, { "epoch": 0.0002777, "grad_norm": 5.191262245178223, "learning_rate": 2.7760000000000002e-05, "loss": 21.2969, "step": 2777 }, { "epoch": 0.0002778, "grad_norm": 4.445683002471924, "learning_rate": 2.777e-05, "loss": 16.5938, "step": 2778 }, { "epoch": 0.0002779, "grad_norm": 4.817911624908447, "learning_rate": 2.7779999999999998e-05, "loss": 19.1562, "step": 2779 }, { "epoch": 0.000278, "grad_norm": 4.558908939361572, "learning_rate": 2.779e-05, "loss": 18.0625, "step": 2780 }, { "epoch": 0.0002781, "grad_norm": 4.665267467498779, "learning_rate": 2.7799999999999998e-05, "loss": 17.7109, "step": 2781 }, { "epoch": 0.0002782, "grad_norm": 4.983461856842041, "learning_rate": 2.7810000000000003e-05, "loss": 19.7188, "step": 2782 }, { "epoch": 0.0002783, "grad_norm": 4.549698829650879, "learning_rate": 2.782e-05, "loss": 18.4531, "step": 2783 }, { "epoch": 0.0002784, "grad_norm": 4.595810413360596, "learning_rate": 2.7830000000000003e-05, "loss": 18.5781, "step": 2784 }, { "epoch": 0.0002785, "grad_norm": 4.829601764678955, "learning_rate": 2.784e-05, "loss": 19.1875, "step": 2785 }, { "epoch": 0.0002786, "grad_norm": 4.878640174865723, "learning_rate": 2.785e-05, "loss": 19.5938, "step": 2786 }, { "epoch": 0.0002787, "grad_norm": 5.217353343963623, "learning_rate": 2.786e-05, "loss": 19.2344, "step": 2787 }, { "epoch": 0.0002788, "grad_norm": 4.7430925369262695, "learning_rate": 2.787e-05, "loss": 18.2188, "step": 2788 }, { "epoch": 0.0002789, "grad_norm": 4.997043132781982, "learning_rate": 2.7880000000000004e-05, "loss": 19.3906, "step": 2789 }, { "epoch": 0.000279, "grad_norm": 4.627202033996582, "learning_rate": 2.789e-05, "loss": 17.8281, "step": 2790 }, { "epoch": 0.0002791, "grad_norm": 5.239760875701904, "learning_rate": 2.7899999999999997e-05, "loss": 18.0859, "step": 2791 }, { "epoch": 0.0002792, "grad_norm": 5.030478000640869, "learning_rate": 2.7910000000000002e-05, "loss": 18.6406, "step": 2792 }, { "epoch": 0.0002793, "grad_norm": 4.707918167114258, "learning_rate": 2.792e-05, "loss": 18.7188, "step": 2793 }, { "epoch": 0.0002794, "grad_norm": 4.694299221038818, "learning_rate": 2.7930000000000002e-05, "loss": 18.7188, "step": 2794 }, { "epoch": 0.0002795, "grad_norm": 4.586289405822754, "learning_rate": 2.794e-05, "loss": 17.8438, "step": 2795 }, { "epoch": 0.0002796, "grad_norm": 4.4019246101379395, "learning_rate": 2.795e-05, "loss": 18.3281, "step": 2796 }, { "epoch": 0.0002797, "grad_norm": 4.743744373321533, "learning_rate": 2.796e-05, "loss": 19.375, "step": 2797 }, { "epoch": 0.0002798, "grad_norm": 4.694622039794922, "learning_rate": 2.7969999999999998e-05, "loss": 18.7031, "step": 2798 }, { "epoch": 0.0002799, "grad_norm": 4.7197723388671875, "learning_rate": 2.7980000000000003e-05, "loss": 17.4062, "step": 2799 }, { "epoch": 0.00028, "grad_norm": 4.485874652862549, "learning_rate": 2.799e-05, "loss": 17.3594, "step": 2800 }, { "epoch": 0.0002801, "grad_norm": 4.4137444496154785, "learning_rate": 2.8000000000000003e-05, "loss": 17.0234, "step": 2801 }, { "epoch": 0.0002802, "grad_norm": 4.570250511169434, "learning_rate": 2.801e-05, "loss": 17.7656, "step": 2802 }, { "epoch": 0.0002803, "grad_norm": 4.813685417175293, "learning_rate": 2.802e-05, "loss": 18.4219, "step": 2803 }, { "epoch": 0.0002804, "grad_norm": 4.985946178436279, "learning_rate": 2.803e-05, "loss": 19.5938, "step": 2804 }, { "epoch": 0.0002805, "grad_norm": 4.656802177429199, "learning_rate": 2.804e-05, "loss": 17.7812, "step": 2805 }, { "epoch": 0.0002806, "grad_norm": 4.737152099609375, "learning_rate": 2.8050000000000004e-05, "loss": 18.3594, "step": 2806 }, { "epoch": 0.0002807, "grad_norm": 4.852945804595947, "learning_rate": 2.806e-05, "loss": 17.8281, "step": 2807 }, { "epoch": 0.0002808, "grad_norm": 5.322506427764893, "learning_rate": 2.8069999999999997e-05, "loss": 18.0391, "step": 2808 }, { "epoch": 0.0002809, "grad_norm": 4.702515602111816, "learning_rate": 2.8080000000000002e-05, "loss": 17.8672, "step": 2809 }, { "epoch": 0.000281, "grad_norm": 4.9520745277404785, "learning_rate": 2.809e-05, "loss": 18.8125, "step": 2810 }, { "epoch": 0.0002811, "grad_norm": 4.761645793914795, "learning_rate": 2.8100000000000002e-05, "loss": 18.1719, "step": 2811 }, { "epoch": 0.0002812, "grad_norm": 4.548405647277832, "learning_rate": 2.811e-05, "loss": 16.7422, "step": 2812 }, { "epoch": 0.0002813, "grad_norm": 5.04771614074707, "learning_rate": 2.812e-05, "loss": 18.5781, "step": 2813 }, { "epoch": 0.0002814, "grad_norm": 4.727956295013428, "learning_rate": 2.813e-05, "loss": 17.3281, "step": 2814 }, { "epoch": 0.0002815, "grad_norm": 4.862520217895508, "learning_rate": 2.8139999999999998e-05, "loss": 18.0, "step": 2815 }, { "epoch": 0.0002816, "grad_norm": 4.744198799133301, "learning_rate": 2.8150000000000003e-05, "loss": 16.5469, "step": 2816 }, { "epoch": 0.0002817, "grad_norm": 4.7819342613220215, "learning_rate": 2.816e-05, "loss": 18.1719, "step": 2817 }, { "epoch": 0.0002818, "grad_norm": 4.631181716918945, "learning_rate": 2.8170000000000003e-05, "loss": 16.7109, "step": 2818 }, { "epoch": 0.0002819, "grad_norm": 4.519904613494873, "learning_rate": 2.818e-05, "loss": 16.9453, "step": 2819 }, { "epoch": 0.000282, "grad_norm": 4.762391090393066, "learning_rate": 2.819e-05, "loss": 18.3125, "step": 2820 }, { "epoch": 0.0002821, "grad_norm": 4.468061447143555, "learning_rate": 2.82e-05, "loss": 16.3438, "step": 2821 }, { "epoch": 0.0002822, "grad_norm": 5.211294174194336, "learning_rate": 2.821e-05, "loss": 19.9844, "step": 2822 }, { "epoch": 0.0002823, "grad_norm": 4.713068008422852, "learning_rate": 2.8220000000000004e-05, "loss": 18.8047, "step": 2823 }, { "epoch": 0.0002824, "grad_norm": 4.671262264251709, "learning_rate": 2.823e-05, "loss": 17.125, "step": 2824 }, { "epoch": 0.0002825, "grad_norm": 4.528439998626709, "learning_rate": 2.8239999999999997e-05, "loss": 17.2656, "step": 2825 }, { "epoch": 0.0002826, "grad_norm": 4.864354610443115, "learning_rate": 2.8250000000000002e-05, "loss": 18.6172, "step": 2826 }, { "epoch": 0.0002827, "grad_norm": 5.204596042633057, "learning_rate": 2.826e-05, "loss": 17.8047, "step": 2827 }, { "epoch": 0.0002828, "grad_norm": 4.455136775970459, "learning_rate": 2.8270000000000002e-05, "loss": 16.8438, "step": 2828 }, { "epoch": 0.0002829, "grad_norm": 5.204645156860352, "learning_rate": 2.828e-05, "loss": 19.6719, "step": 2829 }, { "epoch": 0.000283, "grad_norm": 4.897683620452881, "learning_rate": 2.829e-05, "loss": 17.9062, "step": 2830 }, { "epoch": 0.0002831, "grad_norm": 5.170593738555908, "learning_rate": 2.83e-05, "loss": 17.2656, "step": 2831 }, { "epoch": 0.0002832, "grad_norm": 4.3335113525390625, "learning_rate": 2.831e-05, "loss": 16.2266, "step": 2832 }, { "epoch": 0.0002833, "grad_norm": 4.700484275817871, "learning_rate": 2.8320000000000003e-05, "loss": 16.4453, "step": 2833 }, { "epoch": 0.0002834, "grad_norm": 4.625811576843262, "learning_rate": 2.833e-05, "loss": 17.1875, "step": 2834 }, { "epoch": 0.0002835, "grad_norm": 4.619012832641602, "learning_rate": 2.8340000000000003e-05, "loss": 17.4844, "step": 2835 }, { "epoch": 0.0002836, "grad_norm": 4.589936256408691, "learning_rate": 2.835e-05, "loss": 17.5938, "step": 2836 }, { "epoch": 0.0002837, "grad_norm": 4.460606575012207, "learning_rate": 2.836e-05, "loss": 16.5781, "step": 2837 }, { "epoch": 0.0002838, "grad_norm": 4.691569805145264, "learning_rate": 2.837e-05, "loss": 18.1094, "step": 2838 }, { "epoch": 0.0002839, "grad_norm": 4.637815475463867, "learning_rate": 2.838e-05, "loss": 17.2969, "step": 2839 }, { "epoch": 0.000284, "grad_norm": 5.172640323638916, "learning_rate": 2.8390000000000004e-05, "loss": 18.5938, "step": 2840 }, { "epoch": 0.0002841, "grad_norm": 4.628631591796875, "learning_rate": 2.84e-05, "loss": 17.375, "step": 2841 }, { "epoch": 0.0002842, "grad_norm": 4.511446475982666, "learning_rate": 2.8409999999999997e-05, "loss": 16.9766, "step": 2842 }, { "epoch": 0.0002843, "grad_norm": 4.62901496887207, "learning_rate": 2.8420000000000002e-05, "loss": 16.7578, "step": 2843 }, { "epoch": 0.0002844, "grad_norm": 4.6724042892456055, "learning_rate": 2.843e-05, "loss": 17.2656, "step": 2844 }, { "epoch": 0.0002845, "grad_norm": 4.652232646942139, "learning_rate": 2.8440000000000002e-05, "loss": 17.3203, "step": 2845 }, { "epoch": 0.0002846, "grad_norm": 5.208554267883301, "learning_rate": 2.845e-05, "loss": 17.8906, "step": 2846 }, { "epoch": 0.0002847, "grad_norm": 4.68185567855835, "learning_rate": 2.846e-05, "loss": 16.8125, "step": 2847 }, { "epoch": 0.0002848, "grad_norm": 4.696484088897705, "learning_rate": 2.847e-05, "loss": 16.9844, "step": 2848 }, { "epoch": 0.0002849, "grad_norm": 4.645572662353516, "learning_rate": 2.848e-05, "loss": 17.0547, "step": 2849 }, { "epoch": 0.000285, "grad_norm": 4.4747633934021, "learning_rate": 2.8490000000000003e-05, "loss": 16.6953, "step": 2850 }, { "epoch": 0.0002851, "grad_norm": 4.683969020843506, "learning_rate": 2.85e-05, "loss": 17.3516, "step": 2851 }, { "epoch": 0.0002852, "grad_norm": 4.625978946685791, "learning_rate": 2.8510000000000003e-05, "loss": 16.0547, "step": 2852 }, { "epoch": 0.0002853, "grad_norm": 4.4521074295043945, "learning_rate": 2.852e-05, "loss": 16.2656, "step": 2853 }, { "epoch": 0.0002854, "grad_norm": 4.787761211395264, "learning_rate": 2.853e-05, "loss": 17.6016, "step": 2854 }, { "epoch": 0.0002855, "grad_norm": 4.375016689300537, "learning_rate": 2.854e-05, "loss": 16.2188, "step": 2855 }, { "epoch": 0.0002856, "grad_norm": 4.677867889404297, "learning_rate": 2.855e-05, "loss": 18.6406, "step": 2856 }, { "epoch": 0.0002857, "grad_norm": 4.4339680671691895, "learning_rate": 2.856e-05, "loss": 15.9453, "step": 2857 }, { "epoch": 0.0002858, "grad_norm": 4.396962642669678, "learning_rate": 2.857e-05, "loss": 17.0859, "step": 2858 }, { "epoch": 0.0002859, "grad_norm": 5.348623275756836, "learning_rate": 2.8579999999999997e-05, "loss": 16.7266, "step": 2859 }, { "epoch": 0.000286, "grad_norm": 4.652529716491699, "learning_rate": 2.8590000000000002e-05, "loss": 17.1562, "step": 2860 }, { "epoch": 0.0002861, "grad_norm": 4.414595127105713, "learning_rate": 2.86e-05, "loss": 16.1484, "step": 2861 }, { "epoch": 0.0002862, "grad_norm": 4.47542667388916, "learning_rate": 2.8610000000000002e-05, "loss": 17.6094, "step": 2862 }, { "epoch": 0.0002863, "grad_norm": 5.187290191650391, "learning_rate": 2.862e-05, "loss": 17.0391, "step": 2863 }, { "epoch": 0.0002864, "grad_norm": 4.416170120239258, "learning_rate": 2.863e-05, "loss": 16.5938, "step": 2864 }, { "epoch": 0.0002865, "grad_norm": 4.46373176574707, "learning_rate": 2.864e-05, "loss": 16.1719, "step": 2865 }, { "epoch": 0.0002866, "grad_norm": 4.680726051330566, "learning_rate": 2.865e-05, "loss": 16.5234, "step": 2866 }, { "epoch": 0.0002867, "grad_norm": 5.048381805419922, "learning_rate": 2.8660000000000003e-05, "loss": 16.7812, "step": 2867 }, { "epoch": 0.0002868, "grad_norm": 4.405988693237305, "learning_rate": 2.8669999999999998e-05, "loss": 16.5078, "step": 2868 }, { "epoch": 0.0002869, "grad_norm": 4.618431091308594, "learning_rate": 2.8680000000000003e-05, "loss": 16.9062, "step": 2869 }, { "epoch": 0.000287, "grad_norm": 4.291433811187744, "learning_rate": 2.869e-05, "loss": 15.2188, "step": 2870 }, { "epoch": 0.0002871, "grad_norm": 4.670327663421631, "learning_rate": 2.87e-05, "loss": 16.5781, "step": 2871 }, { "epoch": 0.0002872, "grad_norm": 4.583630084991455, "learning_rate": 2.871e-05, "loss": 16.7812, "step": 2872 }, { "epoch": 0.0002873, "grad_norm": 4.470605850219727, "learning_rate": 2.872e-05, "loss": 15.8438, "step": 2873 }, { "epoch": 0.0002874, "grad_norm": 4.351926803588867, "learning_rate": 2.873e-05, "loss": 16.0703, "step": 2874 }, { "epoch": 0.0002875, "grad_norm": 4.7254557609558105, "learning_rate": 2.874e-05, "loss": 16.7109, "step": 2875 }, { "epoch": 0.0002876, "grad_norm": 4.62042236328125, "learning_rate": 2.8749999999999997e-05, "loss": 16.1719, "step": 2876 }, { "epoch": 0.0002877, "grad_norm": 5.063781261444092, "learning_rate": 2.8760000000000002e-05, "loss": 17.4141, "step": 2877 }, { "epoch": 0.0002878, "grad_norm": 4.7523345947265625, "learning_rate": 2.877e-05, "loss": 17.0938, "step": 2878 }, { "epoch": 0.0002879, "grad_norm": 4.958543300628662, "learning_rate": 2.8780000000000002e-05, "loss": 16.6875, "step": 2879 }, { "epoch": 0.000288, "grad_norm": 4.783759117126465, "learning_rate": 2.879e-05, "loss": 17.1562, "step": 2880 }, { "epoch": 0.0002881, "grad_norm": 4.359407424926758, "learning_rate": 2.88e-05, "loss": 16.0625, "step": 2881 }, { "epoch": 0.0002882, "grad_norm": 4.523177623748779, "learning_rate": 2.881e-05, "loss": 16.9141, "step": 2882 }, { "epoch": 0.0002883, "grad_norm": 4.35944128036499, "learning_rate": 2.882e-05, "loss": 16.1953, "step": 2883 }, { "epoch": 0.0002884, "grad_norm": 4.440458297729492, "learning_rate": 2.8830000000000003e-05, "loss": 16.1797, "step": 2884 }, { "epoch": 0.0002885, "grad_norm": 4.846312046051025, "learning_rate": 2.884e-05, "loss": 16.8906, "step": 2885 }, { "epoch": 0.0002886, "grad_norm": 4.49927282333374, "learning_rate": 2.8850000000000003e-05, "loss": 14.9062, "step": 2886 }, { "epoch": 0.0002887, "grad_norm": 4.72593355178833, "learning_rate": 2.886e-05, "loss": 16.0391, "step": 2887 }, { "epoch": 0.0002888, "grad_norm": 4.529187202453613, "learning_rate": 2.887e-05, "loss": 15.8594, "step": 2888 }, { "epoch": 0.0002889, "grad_norm": 4.757946968078613, "learning_rate": 2.888e-05, "loss": 16.5234, "step": 2889 }, { "epoch": 0.000289, "grad_norm": 4.5501556396484375, "learning_rate": 2.889e-05, "loss": 15.9062, "step": 2890 }, { "epoch": 0.0002891, "grad_norm": 4.50446891784668, "learning_rate": 2.89e-05, "loss": 15.5703, "step": 2891 }, { "epoch": 0.0002892, "grad_norm": 5.304258346557617, "learning_rate": 2.891e-05, "loss": 16.5781, "step": 2892 }, { "epoch": 0.0002893, "grad_norm": 4.246376037597656, "learning_rate": 2.8919999999999998e-05, "loss": 15.2344, "step": 2893 }, { "epoch": 0.0002894, "grad_norm": 4.287469387054443, "learning_rate": 2.8930000000000003e-05, "loss": 15.5547, "step": 2894 }, { "epoch": 0.0002895, "grad_norm": 4.571712017059326, "learning_rate": 2.894e-05, "loss": 15.9453, "step": 2895 }, { "epoch": 0.0002896, "grad_norm": 4.485263347625732, "learning_rate": 2.8950000000000002e-05, "loss": 15.7031, "step": 2896 }, { "epoch": 0.0002897, "grad_norm": 4.5893964767456055, "learning_rate": 2.896e-05, "loss": 16.4453, "step": 2897 }, { "epoch": 0.0002898, "grad_norm": 4.300561428070068, "learning_rate": 2.897e-05, "loss": 15.7031, "step": 2898 }, { "epoch": 0.0002899, "grad_norm": 4.651968955993652, "learning_rate": 2.898e-05, "loss": 15.5625, "step": 2899 }, { "epoch": 0.00029, "grad_norm": 4.574719429016113, "learning_rate": 2.899e-05, "loss": 16.3828, "step": 2900 }, { "epoch": 0.0002901, "grad_norm": 4.320968151092529, "learning_rate": 2.9000000000000004e-05, "loss": 16.3906, "step": 2901 }, { "epoch": 0.0002902, "grad_norm": 4.4197998046875, "learning_rate": 2.901e-05, "loss": 15.6328, "step": 2902 }, { "epoch": 0.0002903, "grad_norm": 4.53903341293335, "learning_rate": 2.9020000000000003e-05, "loss": 16.8594, "step": 2903 }, { "epoch": 0.0002904, "grad_norm": 4.367023944854736, "learning_rate": 2.903e-05, "loss": 15.8672, "step": 2904 }, { "epoch": 0.0002905, "grad_norm": 4.487772464752197, "learning_rate": 2.904e-05, "loss": 15.75, "step": 2905 }, { "epoch": 0.0002906, "grad_norm": 4.457683563232422, "learning_rate": 2.905e-05, "loss": 15.5312, "step": 2906 }, { "epoch": 0.0002907, "grad_norm": 4.6332268714904785, "learning_rate": 2.906e-05, "loss": 15.8906, "step": 2907 }, { "epoch": 0.0002908, "grad_norm": 4.323024272918701, "learning_rate": 2.907e-05, "loss": 14.4609, "step": 2908 }, { "epoch": 0.0002909, "grad_norm": 4.742920398712158, "learning_rate": 2.908e-05, "loss": 16.2891, "step": 2909 }, { "epoch": 0.000291, "grad_norm": 4.666335105895996, "learning_rate": 2.9089999999999998e-05, "loss": 16.7188, "step": 2910 }, { "epoch": 0.0002911, "grad_norm": 4.418400764465332, "learning_rate": 2.9100000000000003e-05, "loss": 15.1719, "step": 2911 }, { "epoch": 0.0002912, "grad_norm": 5.078399181365967, "learning_rate": 2.911e-05, "loss": 16.1328, "step": 2912 }, { "epoch": 0.0002913, "grad_norm": 4.4096903800964355, "learning_rate": 2.9120000000000002e-05, "loss": 15.3594, "step": 2913 }, { "epoch": 0.0002914, "grad_norm": 4.572752952575684, "learning_rate": 2.913e-05, "loss": 15.8281, "step": 2914 }, { "epoch": 0.0002915, "grad_norm": 4.665991306304932, "learning_rate": 2.914e-05, "loss": 17.0547, "step": 2915 }, { "epoch": 0.0002916, "grad_norm": 4.440099239349365, "learning_rate": 2.915e-05, "loss": 14.7266, "step": 2916 }, { "epoch": 0.0002917, "grad_norm": 4.6129255294799805, "learning_rate": 2.916e-05, "loss": 16.6406, "step": 2917 }, { "epoch": 0.0002918, "grad_norm": 4.330190181732178, "learning_rate": 2.9170000000000004e-05, "loss": 14.4219, "step": 2918 }, { "epoch": 0.0002919, "grad_norm": 4.353707313537598, "learning_rate": 2.918e-05, "loss": 14.9922, "step": 2919 }, { "epoch": 0.000292, "grad_norm": 4.465911865234375, "learning_rate": 2.9190000000000003e-05, "loss": 15.2891, "step": 2920 }, { "epoch": 0.0002921, "grad_norm": 6.285282611846924, "learning_rate": 2.92e-05, "loss": 18.1562, "step": 2921 }, { "epoch": 0.0002922, "grad_norm": 4.356308937072754, "learning_rate": 2.921e-05, "loss": 16.5234, "step": 2922 }, { "epoch": 0.0002923, "grad_norm": 4.507532596588135, "learning_rate": 2.922e-05, "loss": 15.7656, "step": 2923 }, { "epoch": 0.0002924, "grad_norm": 4.2283854484558105, "learning_rate": 2.923e-05, "loss": 14.875, "step": 2924 }, { "epoch": 0.0002925, "grad_norm": 4.387848377227783, "learning_rate": 2.924e-05, "loss": 15.5469, "step": 2925 }, { "epoch": 0.0002926, "grad_norm": 4.4351701736450195, "learning_rate": 2.925e-05, "loss": 16.0234, "step": 2926 }, { "epoch": 0.0002927, "grad_norm": 4.16475772857666, "learning_rate": 2.9259999999999998e-05, "loss": 15.2734, "step": 2927 }, { "epoch": 0.0002928, "grad_norm": 5.452282905578613, "learning_rate": 2.9270000000000003e-05, "loss": 16.3203, "step": 2928 }, { "epoch": 0.0002929, "grad_norm": 4.437356472015381, "learning_rate": 2.928e-05, "loss": 14.9844, "step": 2929 }, { "epoch": 0.000293, "grad_norm": 4.507708549499512, "learning_rate": 2.9290000000000002e-05, "loss": 15.2578, "step": 2930 }, { "epoch": 0.0002931, "grad_norm": 4.115293502807617, "learning_rate": 2.93e-05, "loss": 14.1875, "step": 2931 }, { "epoch": 0.0002932, "grad_norm": 4.716981887817383, "learning_rate": 2.931e-05, "loss": 16.3359, "step": 2932 }, { "epoch": 0.0002933, "grad_norm": 4.439236640930176, "learning_rate": 2.932e-05, "loss": 15.0234, "step": 2933 }, { "epoch": 0.0002934, "grad_norm": 4.4764323234558105, "learning_rate": 2.933e-05, "loss": 15.2578, "step": 2934 }, { "epoch": 0.0002935, "grad_norm": 4.61359167098999, "learning_rate": 2.9340000000000004e-05, "loss": 15.875, "step": 2935 }, { "epoch": 0.0002936, "grad_norm": 4.462646007537842, "learning_rate": 2.935e-05, "loss": 15.2266, "step": 2936 }, { "epoch": 0.0002937, "grad_norm": 4.612847328186035, "learning_rate": 2.9360000000000003e-05, "loss": 16.4688, "step": 2937 }, { "epoch": 0.0002938, "grad_norm": 4.600231170654297, "learning_rate": 2.9370000000000002e-05, "loss": 16.0078, "step": 2938 }, { "epoch": 0.0002939, "grad_norm": 4.400425910949707, "learning_rate": 2.938e-05, "loss": 14.8672, "step": 2939 }, { "epoch": 0.000294, "grad_norm": 4.245770454406738, "learning_rate": 2.939e-05, "loss": 14.8594, "step": 2940 }, { "epoch": 0.0002941, "grad_norm": 4.454634666442871, "learning_rate": 2.94e-05, "loss": 14.8828, "step": 2941 }, { "epoch": 0.0002942, "grad_norm": 4.388282299041748, "learning_rate": 2.941e-05, "loss": 15.5859, "step": 2942 }, { "epoch": 0.0002943, "grad_norm": 4.288054466247559, "learning_rate": 2.942e-05, "loss": 14.3047, "step": 2943 }, { "epoch": 0.0002944, "grad_norm": 4.181088924407959, "learning_rate": 2.9429999999999998e-05, "loss": 14.1641, "step": 2944 }, { "epoch": 0.0002945, "grad_norm": 4.289495944976807, "learning_rate": 2.9440000000000003e-05, "loss": 13.9844, "step": 2945 }, { "epoch": 0.0002946, "grad_norm": 4.534033298492432, "learning_rate": 2.945e-05, "loss": 16.0938, "step": 2946 }, { "epoch": 0.0002947, "grad_norm": 4.383325576782227, "learning_rate": 2.9460000000000003e-05, "loss": 14.7344, "step": 2947 }, { "epoch": 0.0002948, "grad_norm": 4.408135890960693, "learning_rate": 2.947e-05, "loss": 14.9688, "step": 2948 }, { "epoch": 0.0002949, "grad_norm": 4.906869411468506, "learning_rate": 2.948e-05, "loss": 16.3984, "step": 2949 }, { "epoch": 0.000295, "grad_norm": 4.564929962158203, "learning_rate": 2.949e-05, "loss": 15.4453, "step": 2950 }, { "epoch": 0.0002951, "grad_norm": 4.3303351402282715, "learning_rate": 2.95e-05, "loss": 14.2578, "step": 2951 }, { "epoch": 0.0002952, "grad_norm": 4.338034629821777, "learning_rate": 2.9510000000000004e-05, "loss": 14.9375, "step": 2952 }, { "epoch": 0.0002953, "grad_norm": 4.147040843963623, "learning_rate": 2.952e-05, "loss": 14.5312, "step": 2953 }, { "epoch": 0.0002954, "grad_norm": 4.2163872718811035, "learning_rate": 2.9530000000000004e-05, "loss": 13.7188, "step": 2954 }, { "epoch": 0.0002955, "grad_norm": 4.197270393371582, "learning_rate": 2.9540000000000002e-05, "loss": 14.3906, "step": 2955 }, { "epoch": 0.0002956, "grad_norm": 4.364101886749268, "learning_rate": 2.955e-05, "loss": 14.75, "step": 2956 }, { "epoch": 0.0002957, "grad_norm": 4.737512111663818, "learning_rate": 2.956e-05, "loss": 15.6016, "step": 2957 }, { "epoch": 0.0002958, "grad_norm": 4.405240058898926, "learning_rate": 2.957e-05, "loss": 14.8984, "step": 2958 }, { "epoch": 0.0002959, "grad_norm": 4.504712104797363, "learning_rate": 2.958e-05, "loss": 15.8984, "step": 2959 }, { "epoch": 0.000296, "grad_norm": 4.254082202911377, "learning_rate": 2.959e-05, "loss": 14.3203, "step": 2960 }, { "epoch": 0.0002961, "grad_norm": 4.36760950088501, "learning_rate": 2.9599999999999998e-05, "loss": 14.1719, "step": 2961 }, { "epoch": 0.0002962, "grad_norm": 4.311966896057129, "learning_rate": 2.9610000000000003e-05, "loss": 15.1562, "step": 2962 }, { "epoch": 0.0002963, "grad_norm": 4.4636921882629395, "learning_rate": 2.962e-05, "loss": 15.5078, "step": 2963 }, { "epoch": 0.0002964, "grad_norm": 4.16254997253418, "learning_rate": 2.9630000000000003e-05, "loss": 14.4922, "step": 2964 }, { "epoch": 0.0002965, "grad_norm": 4.497593879699707, "learning_rate": 2.964e-05, "loss": 15.1641, "step": 2965 }, { "epoch": 0.0002966, "grad_norm": 4.532444953918457, "learning_rate": 2.965e-05, "loss": 15.6719, "step": 2966 }, { "epoch": 0.0002967, "grad_norm": 4.701189041137695, "learning_rate": 2.966e-05, "loss": 14.625, "step": 2967 }, { "epoch": 0.0002968, "grad_norm": 4.323169231414795, "learning_rate": 2.967e-05, "loss": 14.1484, "step": 2968 }, { "epoch": 0.0002969, "grad_norm": 4.383123874664307, "learning_rate": 2.9680000000000004e-05, "loss": 14.8359, "step": 2969 }, { "epoch": 0.000297, "grad_norm": 4.308600902557373, "learning_rate": 2.969e-05, "loss": 14.8125, "step": 2970 }, { "epoch": 0.0002971, "grad_norm": 4.598495960235596, "learning_rate": 2.9700000000000004e-05, "loss": 14.0781, "step": 2971 }, { "epoch": 0.0002972, "grad_norm": 4.719667434692383, "learning_rate": 2.9710000000000002e-05, "loss": 15.0, "step": 2972 }, { "epoch": 0.0002973, "grad_norm": 4.504169940948486, "learning_rate": 2.972e-05, "loss": 14.5078, "step": 2973 }, { "epoch": 0.0002974, "grad_norm": 4.3584513664245605, "learning_rate": 2.973e-05, "loss": 14.0703, "step": 2974 }, { "epoch": 0.0002975, "grad_norm": 4.283166885375977, "learning_rate": 2.974e-05, "loss": 13.8203, "step": 2975 }, { "epoch": 0.0002976, "grad_norm": 4.122700214385986, "learning_rate": 2.975e-05, "loss": 14.0938, "step": 2976 }, { "epoch": 0.0002977, "grad_norm": 4.2741169929504395, "learning_rate": 2.976e-05, "loss": 13.8594, "step": 2977 }, { "epoch": 0.0002978, "grad_norm": 4.3709025382995605, "learning_rate": 2.9769999999999998e-05, "loss": 14.625, "step": 2978 }, { "epoch": 0.0002979, "grad_norm": 4.550168037414551, "learning_rate": 2.9780000000000003e-05, "loss": 14.5312, "step": 2979 }, { "epoch": 0.000298, "grad_norm": 4.324486255645752, "learning_rate": 2.979e-05, "loss": 14.7734, "step": 2980 }, { "epoch": 0.0002981, "grad_norm": 4.351390361785889, "learning_rate": 2.9800000000000003e-05, "loss": 14.0312, "step": 2981 }, { "epoch": 0.0002982, "grad_norm": 4.152400016784668, "learning_rate": 2.981e-05, "loss": 14.25, "step": 2982 }, { "epoch": 0.0002983, "grad_norm": 4.2570013999938965, "learning_rate": 2.982e-05, "loss": 14.2734, "step": 2983 }, { "epoch": 0.0002984, "grad_norm": 4.354986190795898, "learning_rate": 2.983e-05, "loss": 14.5547, "step": 2984 }, { "epoch": 0.0002985, "grad_norm": 4.235496520996094, "learning_rate": 2.984e-05, "loss": 13.5781, "step": 2985 }, { "epoch": 0.0002986, "grad_norm": 4.65847110748291, "learning_rate": 2.985e-05, "loss": 15.8984, "step": 2986 }, { "epoch": 0.0002987, "grad_norm": 4.1647186279296875, "learning_rate": 2.986e-05, "loss": 13.7031, "step": 2987 }, { "epoch": 0.0002988, "grad_norm": 4.806039333343506, "learning_rate": 2.9870000000000004e-05, "loss": 15.2188, "step": 2988 }, { "epoch": 0.0002989, "grad_norm": 4.333047389984131, "learning_rate": 2.9880000000000002e-05, "loss": 14.2969, "step": 2989 }, { "epoch": 0.000299, "grad_norm": 5.031891822814941, "learning_rate": 2.989e-05, "loss": 14.6484, "step": 2990 }, { "epoch": 0.0002991, "grad_norm": 4.221193313598633, "learning_rate": 2.9900000000000002e-05, "loss": 14.2812, "step": 2991 }, { "epoch": 0.0002992, "grad_norm": 4.297886371612549, "learning_rate": 2.991e-05, "loss": 14.2656, "step": 2992 }, { "epoch": 0.0002993, "grad_norm": 4.2214274406433105, "learning_rate": 2.992e-05, "loss": 13.9609, "step": 2993 }, { "epoch": 0.0002994, "grad_norm": 4.2055344581604, "learning_rate": 2.993e-05, "loss": 13.9375, "step": 2994 }, { "epoch": 0.0002995, "grad_norm": 4.12393045425415, "learning_rate": 2.9939999999999998e-05, "loss": 14.0234, "step": 2995 }, { "epoch": 0.0002996, "grad_norm": 4.280452251434326, "learning_rate": 2.9950000000000003e-05, "loss": 14.25, "step": 2996 }, { "epoch": 0.0002997, "grad_norm": 4.374136924743652, "learning_rate": 2.9959999999999998e-05, "loss": 15.3047, "step": 2997 }, { "epoch": 0.0002998, "grad_norm": 4.173422336578369, "learning_rate": 2.9970000000000003e-05, "loss": 14.6562, "step": 2998 }, { "epoch": 0.0002999, "grad_norm": 4.103376865386963, "learning_rate": 2.998e-05, "loss": 13.6328, "step": 2999 }, { "epoch": 0.0003, "grad_norm": 4.4414801597595215, "learning_rate": 2.999e-05, "loss": 14.7891, "step": 3000 }, { "epoch": 0.0003, "eval_loss": 1.7186750173568726, "eval_runtime": 407.6178, "eval_samples_per_second": 24.533, "eval_steps_per_second": 1.533, "step": 3000 }, { "epoch": 0.0003001, "grad_norm": 4.194701194763184, "learning_rate": 3e-05, "loss": 13.9688, "step": 3001 }, { "epoch": 0.0003002, "grad_norm": 4.1234259605407715, "learning_rate": 3.001e-05, "loss": 13.2422, "step": 3002 }, { "epoch": 0.0003003, "grad_norm": 4.26988410949707, "learning_rate": 3.002e-05, "loss": 13.5859, "step": 3003 }, { "epoch": 0.0003004, "grad_norm": 4.214878082275391, "learning_rate": 3.003e-05, "loss": 13.6094, "step": 3004 }, { "epoch": 0.0003005, "grad_norm": 4.335902214050293, "learning_rate": 3.0040000000000004e-05, "loss": 13.7266, "step": 3005 }, { "epoch": 0.0003006, "grad_norm": 4.131463050842285, "learning_rate": 3.0050000000000002e-05, "loss": 14.0234, "step": 3006 }, { "epoch": 0.0003007, "grad_norm": 4.131214141845703, "learning_rate": 3.006e-05, "loss": 13.6484, "step": 3007 }, { "epoch": 0.0003008, "grad_norm": 4.261483669281006, "learning_rate": 3.0070000000000002e-05, "loss": 14.4219, "step": 3008 }, { "epoch": 0.0003009, "grad_norm": 4.358345031738281, "learning_rate": 3.008e-05, "loss": 13.8516, "step": 3009 }, { "epoch": 0.000301, "grad_norm": 4.224039077758789, "learning_rate": 3.009e-05, "loss": 14.2969, "step": 3010 }, { "epoch": 0.0003011, "grad_norm": 3.8865697383880615, "learning_rate": 3.01e-05, "loss": 13.7031, "step": 3011 }, { "epoch": 0.0003012, "grad_norm": 4.495300769805908, "learning_rate": 3.0109999999999998e-05, "loss": 14.375, "step": 3012 }, { "epoch": 0.0003013, "grad_norm": 4.00750732421875, "learning_rate": 3.0120000000000003e-05, "loss": 13.0625, "step": 3013 }, { "epoch": 0.0003014, "grad_norm": 4.163105487823486, "learning_rate": 3.0129999999999998e-05, "loss": 14.7656, "step": 3014 }, { "epoch": 0.0003015, "grad_norm": 4.139406204223633, "learning_rate": 3.0140000000000003e-05, "loss": 13.3359, "step": 3015 }, { "epoch": 0.0003016, "grad_norm": 4.277431011199951, "learning_rate": 3.015e-05, "loss": 13.5, "step": 3016 }, { "epoch": 0.0003017, "grad_norm": 4.73074197769165, "learning_rate": 3.016e-05, "loss": 13.8672, "step": 3017 }, { "epoch": 0.0003018, "grad_norm": 3.9893667697906494, "learning_rate": 3.017e-05, "loss": 13.5391, "step": 3018 }, { "epoch": 0.0003019, "grad_norm": 4.163100719451904, "learning_rate": 3.018e-05, "loss": 13.9609, "step": 3019 }, { "epoch": 0.000302, "grad_norm": 4.3695478439331055, "learning_rate": 3.019e-05, "loss": 13.7344, "step": 3020 }, { "epoch": 0.0003021, "grad_norm": 4.551717758178711, "learning_rate": 3.02e-05, "loss": 14.1797, "step": 3021 }, { "epoch": 0.0003022, "grad_norm": 3.9851553440093994, "learning_rate": 3.0210000000000004e-05, "loss": 12.9688, "step": 3022 }, { "epoch": 0.0003023, "grad_norm": 4.359283924102783, "learning_rate": 3.0220000000000002e-05, "loss": 14.0625, "step": 3023 }, { "epoch": 0.0003024, "grad_norm": 5.082442760467529, "learning_rate": 3.023e-05, "loss": 14.6484, "step": 3024 }, { "epoch": 0.0003025, "grad_norm": 4.216992378234863, "learning_rate": 3.0240000000000002e-05, "loss": 13.9453, "step": 3025 }, { "epoch": 0.0003026, "grad_norm": 4.368180751800537, "learning_rate": 3.025e-05, "loss": 14.3828, "step": 3026 }, { "epoch": 0.0003027, "grad_norm": 4.183067798614502, "learning_rate": 3.026e-05, "loss": 12.9141, "step": 3027 }, { "epoch": 0.0003028, "grad_norm": 4.025392532348633, "learning_rate": 3.027e-05, "loss": 14.0156, "step": 3028 }, { "epoch": 0.0003029, "grad_norm": 4.078479290008545, "learning_rate": 3.0279999999999998e-05, "loss": 13.4688, "step": 3029 }, { "epoch": 0.000303, "grad_norm": 4.384459972381592, "learning_rate": 3.0290000000000003e-05, "loss": 14.8984, "step": 3030 }, { "epoch": 0.0003031, "grad_norm": 4.016053199768066, "learning_rate": 3.0299999999999998e-05, "loss": 13.5391, "step": 3031 }, { "epoch": 0.0003032, "grad_norm": 4.216038227081299, "learning_rate": 3.0310000000000003e-05, "loss": 13.7422, "step": 3032 }, { "epoch": 0.0003033, "grad_norm": 4.121150493621826, "learning_rate": 3.032e-05, "loss": 13.5156, "step": 3033 }, { "epoch": 0.0003034, "grad_norm": 4.356866836547852, "learning_rate": 3.033e-05, "loss": 13.9062, "step": 3034 }, { "epoch": 0.0003035, "grad_norm": 4.39754581451416, "learning_rate": 3.034e-05, "loss": 13.7578, "step": 3035 }, { "epoch": 0.0003036, "grad_norm": 4.1412811279296875, "learning_rate": 3.035e-05, "loss": 12.9766, "step": 3036 }, { "epoch": 0.0003037, "grad_norm": 4.127572536468506, "learning_rate": 3.036e-05, "loss": 12.9609, "step": 3037 }, { "epoch": 0.0003038, "grad_norm": 4.17681884765625, "learning_rate": 3.037e-05, "loss": 13.8047, "step": 3038 }, { "epoch": 0.0003039, "grad_norm": 4.279952049255371, "learning_rate": 3.0380000000000004e-05, "loss": 13.7656, "step": 3039 }, { "epoch": 0.000304, "grad_norm": 4.562777996063232, "learning_rate": 3.0390000000000002e-05, "loss": 13.2734, "step": 3040 }, { "epoch": 0.0003041, "grad_norm": 4.5680999755859375, "learning_rate": 3.04e-05, "loss": 13.8828, "step": 3041 }, { "epoch": 0.0003042, "grad_norm": 4.2030439376831055, "learning_rate": 3.0410000000000002e-05, "loss": 13.8203, "step": 3042 }, { "epoch": 0.0003043, "grad_norm": 4.62878942489624, "learning_rate": 3.042e-05, "loss": 15.5625, "step": 3043 }, { "epoch": 0.0003044, "grad_norm": 4.339844226837158, "learning_rate": 3.0430000000000002e-05, "loss": 13.7109, "step": 3044 }, { "epoch": 0.0003045, "grad_norm": 4.207889080047607, "learning_rate": 3.044e-05, "loss": 14.1406, "step": 3045 }, { "epoch": 0.0003046, "grad_norm": 4.4827961921691895, "learning_rate": 3.0449999999999998e-05, "loss": 13.8359, "step": 3046 }, { "epoch": 0.0003047, "grad_norm": 4.142913341522217, "learning_rate": 3.0460000000000003e-05, "loss": 13.0703, "step": 3047 }, { "epoch": 0.0003048, "grad_norm": 3.8258612155914307, "learning_rate": 3.0469999999999998e-05, "loss": 12.3672, "step": 3048 }, { "epoch": 0.0003049, "grad_norm": 4.040428161621094, "learning_rate": 3.0480000000000003e-05, "loss": 12.3984, "step": 3049 }, { "epoch": 0.000305, "grad_norm": 4.279740810394287, "learning_rate": 3.049e-05, "loss": 13.8125, "step": 3050 }, { "epoch": 0.0003051, "grad_norm": 4.1018853187561035, "learning_rate": 3.05e-05, "loss": 12.8125, "step": 3051 }, { "epoch": 0.0003052, "grad_norm": 4.0386834144592285, "learning_rate": 3.051e-05, "loss": 12.5703, "step": 3052 }, { "epoch": 0.0003053, "grad_norm": 4.915343284606934, "learning_rate": 3.052e-05, "loss": 13.9766, "step": 3053 }, { "epoch": 0.0003054, "grad_norm": 4.776984214782715, "learning_rate": 3.053e-05, "loss": 14.0078, "step": 3054 }, { "epoch": 0.0003055, "grad_norm": 4.403343200683594, "learning_rate": 3.054e-05, "loss": 13.8594, "step": 3055 }, { "epoch": 0.0003056, "grad_norm": 4.060154914855957, "learning_rate": 3.0550000000000004e-05, "loss": 13.0781, "step": 3056 }, { "epoch": 0.0003057, "grad_norm": 4.458163261413574, "learning_rate": 3.056e-05, "loss": 13.9844, "step": 3057 }, { "epoch": 0.0003058, "grad_norm": 4.36746072769165, "learning_rate": 3.057e-05, "loss": 14.2422, "step": 3058 }, { "epoch": 0.0003059, "grad_norm": 4.294203758239746, "learning_rate": 3.058e-05, "loss": 13.6406, "step": 3059 }, { "epoch": 0.000306, "grad_norm": 4.018357276916504, "learning_rate": 3.059e-05, "loss": 13.2578, "step": 3060 }, { "epoch": 0.0003061, "grad_norm": 4.065389156341553, "learning_rate": 3.0600000000000005e-05, "loss": 13.3594, "step": 3061 }, { "epoch": 0.0003062, "grad_norm": 4.399044513702393, "learning_rate": 3.061e-05, "loss": 13.2109, "step": 3062 }, { "epoch": 0.0003063, "grad_norm": 4.423494815826416, "learning_rate": 3.062e-05, "loss": 13.0156, "step": 3063 }, { "epoch": 0.0003064, "grad_norm": 4.337902069091797, "learning_rate": 3.063e-05, "loss": 13.8281, "step": 3064 }, { "epoch": 0.0003065, "grad_norm": 4.550607681274414, "learning_rate": 3.064e-05, "loss": 13.8672, "step": 3065 }, { "epoch": 0.0003066, "grad_norm": 4.459248065948486, "learning_rate": 3.065e-05, "loss": 12.2578, "step": 3066 }, { "epoch": 0.0003067, "grad_norm": 4.459362983703613, "learning_rate": 3.066e-05, "loss": 13.7266, "step": 3067 }, { "epoch": 0.0003068, "grad_norm": 4.087619781494141, "learning_rate": 3.0669999999999996e-05, "loss": 13.1875, "step": 3068 }, { "epoch": 0.0003069, "grad_norm": 3.956347942352295, "learning_rate": 3.0680000000000004e-05, "loss": 12.6875, "step": 3069 }, { "epoch": 0.000307, "grad_norm": 3.932720184326172, "learning_rate": 3.069e-05, "loss": 12.0156, "step": 3070 }, { "epoch": 0.0003071, "grad_norm": 5.017730712890625, "learning_rate": 3.07e-05, "loss": 13.1641, "step": 3071 }, { "epoch": 0.0003072, "grad_norm": 4.067151069641113, "learning_rate": 3.071e-05, "loss": 12.5234, "step": 3072 }, { "epoch": 0.0003073, "grad_norm": 4.07366418838501, "learning_rate": 3.0720000000000004e-05, "loss": 13.0547, "step": 3073 }, { "epoch": 0.0003074, "grad_norm": 4.133744239807129, "learning_rate": 3.073e-05, "loss": 12.9453, "step": 3074 }, { "epoch": 0.0003075, "grad_norm": 4.064107894897461, "learning_rate": 3.074e-05, "loss": 12.8281, "step": 3075 }, { "epoch": 0.0003076, "grad_norm": 4.300269603729248, "learning_rate": 3.075e-05, "loss": 13.8203, "step": 3076 }, { "epoch": 0.0003077, "grad_norm": 4.334654331207275, "learning_rate": 3.076e-05, "loss": 13.5625, "step": 3077 }, { "epoch": 0.0003078, "grad_norm": 4.894828796386719, "learning_rate": 3.0770000000000005e-05, "loss": 14.5469, "step": 3078 }, { "epoch": 0.0003079, "grad_norm": 3.7898101806640625, "learning_rate": 3.078e-05, "loss": 12.1797, "step": 3079 }, { "epoch": 0.000308, "grad_norm": 3.9666337966918945, "learning_rate": 3.079e-05, "loss": 12.125, "step": 3080 }, { "epoch": 0.0003081, "grad_norm": 4.044882297515869, "learning_rate": 3.08e-05, "loss": 12.6406, "step": 3081 }, { "epoch": 0.0003082, "grad_norm": 3.975879192352295, "learning_rate": 3.081e-05, "loss": 12.4531, "step": 3082 }, { "epoch": 0.0003083, "grad_norm": 4.260823726654053, "learning_rate": 3.082e-05, "loss": 13.3672, "step": 3083 }, { "epoch": 0.0003084, "grad_norm": 4.446697235107422, "learning_rate": 3.083e-05, "loss": 12.7422, "step": 3084 }, { "epoch": 0.0003085, "grad_norm": 4.369131565093994, "learning_rate": 3.0839999999999996e-05, "loss": 13.3281, "step": 3085 }, { "epoch": 0.0003086, "grad_norm": 4.1950364112854, "learning_rate": 3.0850000000000004e-05, "loss": 13.2266, "step": 3086 }, { "epoch": 0.0003087, "grad_norm": 4.131595611572266, "learning_rate": 3.086e-05, "loss": 13.5, "step": 3087 }, { "epoch": 0.0003088, "grad_norm": 4.693575859069824, "learning_rate": 3.087e-05, "loss": 13.6953, "step": 3088 }, { "epoch": 0.0003089, "grad_norm": 4.081747531890869, "learning_rate": 3.088e-05, "loss": 12.4375, "step": 3089 }, { "epoch": 0.000309, "grad_norm": 4.189774513244629, "learning_rate": 3.0890000000000004e-05, "loss": 12.4453, "step": 3090 }, { "epoch": 0.0003091, "grad_norm": 4.040273666381836, "learning_rate": 3.09e-05, "loss": 11.8047, "step": 3091 }, { "epoch": 0.0003092, "grad_norm": 4.0818047523498535, "learning_rate": 3.091e-05, "loss": 12.4609, "step": 3092 }, { "epoch": 0.0003093, "grad_norm": 4.071326732635498, "learning_rate": 3.092e-05, "loss": 12.0547, "step": 3093 }, { "epoch": 0.0003094, "grad_norm": 3.987217426300049, "learning_rate": 3.093e-05, "loss": 11.9219, "step": 3094 }, { "epoch": 0.0003095, "grad_norm": 4.264516830444336, "learning_rate": 3.0940000000000005e-05, "loss": 14.3672, "step": 3095 }, { "epoch": 0.0003096, "grad_norm": 4.392807960510254, "learning_rate": 3.095e-05, "loss": 13.4609, "step": 3096 }, { "epoch": 0.0003097, "grad_norm": 4.537779808044434, "learning_rate": 3.096e-05, "loss": 13.3359, "step": 3097 }, { "epoch": 0.0003098, "grad_norm": 4.28918981552124, "learning_rate": 3.097e-05, "loss": 13.6328, "step": 3098 }, { "epoch": 0.0003099, "grad_norm": 4.60708475112915, "learning_rate": 3.098e-05, "loss": 13.2578, "step": 3099 }, { "epoch": 0.00031, "grad_norm": 4.293633937835693, "learning_rate": 3.099e-05, "loss": 12.5312, "step": 3100 }, { "epoch": 0.0003101, "grad_norm": 4.165204048156738, "learning_rate": 3.1e-05, "loss": 13.0, "step": 3101 }, { "epoch": 0.0003102, "grad_norm": 4.12003231048584, "learning_rate": 3.1009999999999996e-05, "loss": 12.5469, "step": 3102 }, { "epoch": 0.0003103, "grad_norm": 4.180938243865967, "learning_rate": 3.1020000000000005e-05, "loss": 12.8828, "step": 3103 }, { "epoch": 0.0003104, "grad_norm": 4.337403297424316, "learning_rate": 3.103e-05, "loss": 12.875, "step": 3104 }, { "epoch": 0.0003105, "grad_norm": 4.510208606719971, "learning_rate": 3.104e-05, "loss": 13.625, "step": 3105 }, { "epoch": 0.0003106, "grad_norm": 3.9945015907287598, "learning_rate": 3.105e-05, "loss": 12.2031, "step": 3106 }, { "epoch": 0.0003107, "grad_norm": 4.681159496307373, "learning_rate": 3.1060000000000004e-05, "loss": 13.5078, "step": 3107 }, { "epoch": 0.0003108, "grad_norm": 4.287213325500488, "learning_rate": 3.107e-05, "loss": 12.5391, "step": 3108 }, { "epoch": 0.0003109, "grad_norm": 4.070905685424805, "learning_rate": 3.108e-05, "loss": 12.6094, "step": 3109 }, { "epoch": 0.000311, "grad_norm": 4.0991926193237305, "learning_rate": 3.109e-05, "loss": 11.7734, "step": 3110 }, { "epoch": 0.0003111, "grad_norm": 3.793558120727539, "learning_rate": 3.11e-05, "loss": 11.7422, "step": 3111 }, { "epoch": 0.0003112, "grad_norm": 4.245021820068359, "learning_rate": 3.1110000000000005e-05, "loss": 13.0625, "step": 3112 }, { "epoch": 0.0003113, "grad_norm": 5.156881809234619, "learning_rate": 3.112e-05, "loss": 12.9297, "step": 3113 }, { "epoch": 0.0003114, "grad_norm": 4.007889747619629, "learning_rate": 3.113e-05, "loss": 11.8281, "step": 3114 }, { "epoch": 0.0003115, "grad_norm": 4.115943908691406, "learning_rate": 3.1140000000000003e-05, "loss": 12.2891, "step": 3115 }, { "epoch": 0.0003116, "grad_norm": 4.325364589691162, "learning_rate": 3.115e-05, "loss": 13.0156, "step": 3116 }, { "epoch": 0.0003117, "grad_norm": 4.555175304412842, "learning_rate": 3.116e-05, "loss": 13.9141, "step": 3117 }, { "epoch": 0.0003118, "grad_norm": 4.521339416503906, "learning_rate": 3.117e-05, "loss": 12.9531, "step": 3118 }, { "epoch": 0.0003119, "grad_norm": 4.389863967895508, "learning_rate": 3.1179999999999996e-05, "loss": 13.6094, "step": 3119 }, { "epoch": 0.000312, "grad_norm": 4.105045795440674, "learning_rate": 3.1190000000000005e-05, "loss": 13.3906, "step": 3120 }, { "epoch": 0.0003121, "grad_norm": 4.395541667938232, "learning_rate": 3.12e-05, "loss": 12.8203, "step": 3121 }, { "epoch": 0.0003122, "grad_norm": 4.227506637573242, "learning_rate": 3.121e-05, "loss": 13.3828, "step": 3122 }, { "epoch": 0.0003123, "grad_norm": 4.824154853820801, "learning_rate": 3.122e-05, "loss": 14.5469, "step": 3123 }, { "epoch": 0.0003124, "grad_norm": 4.500116348266602, "learning_rate": 3.1230000000000004e-05, "loss": 13.6484, "step": 3124 }, { "epoch": 0.0003125, "grad_norm": 4.317654609680176, "learning_rate": 3.124e-05, "loss": 13.0781, "step": 3125 }, { "epoch": 0.0003126, "grad_norm": 3.8973302841186523, "learning_rate": 3.125e-05, "loss": 12.1172, "step": 3126 }, { "epoch": 0.0003127, "grad_norm": 4.212470531463623, "learning_rate": 3.126e-05, "loss": 12.4375, "step": 3127 }, { "epoch": 0.0003128, "grad_norm": 4.022023677825928, "learning_rate": 3.127e-05, "loss": 12.2344, "step": 3128 }, { "epoch": 0.0003129, "grad_norm": 4.173778057098389, "learning_rate": 3.1280000000000005e-05, "loss": 12.6328, "step": 3129 }, { "epoch": 0.000313, "grad_norm": 3.9178037643432617, "learning_rate": 3.129e-05, "loss": 11.5859, "step": 3130 }, { "epoch": 0.0003131, "grad_norm": 4.367602825164795, "learning_rate": 3.13e-05, "loss": 13.0469, "step": 3131 }, { "epoch": 0.0003132, "grad_norm": 4.052750587463379, "learning_rate": 3.1310000000000003e-05, "loss": 11.6016, "step": 3132 }, { "epoch": 0.0003133, "grad_norm": 3.8811826705932617, "learning_rate": 3.132e-05, "loss": 11.6562, "step": 3133 }, { "epoch": 0.0003134, "grad_norm": 4.070820331573486, "learning_rate": 3.133e-05, "loss": 11.5, "step": 3134 }, { "epoch": 0.0003135, "grad_norm": 4.166890621185303, "learning_rate": 3.134e-05, "loss": 11.9297, "step": 3135 }, { "epoch": 0.0003136, "grad_norm": 5.06160306930542, "learning_rate": 3.1349999999999996e-05, "loss": 13.4609, "step": 3136 }, { "epoch": 0.0003137, "grad_norm": 4.107486724853516, "learning_rate": 3.1360000000000005e-05, "loss": 12.1719, "step": 3137 }, { "epoch": 0.0003138, "grad_norm": 4.159612655639648, "learning_rate": 3.137e-05, "loss": 12.9219, "step": 3138 }, { "epoch": 0.0003139, "grad_norm": 4.197668552398682, "learning_rate": 3.138e-05, "loss": 12.625, "step": 3139 }, { "epoch": 0.000314, "grad_norm": 4.047123908996582, "learning_rate": 3.139e-05, "loss": 12.1406, "step": 3140 }, { "epoch": 0.0003141, "grad_norm": 4.3158860206604, "learning_rate": 3.1400000000000004e-05, "loss": 12.1016, "step": 3141 }, { "epoch": 0.0003142, "grad_norm": 3.9308483600616455, "learning_rate": 3.141e-05, "loss": 12.2188, "step": 3142 }, { "epoch": 0.0003143, "grad_norm": 4.07874059677124, "learning_rate": 3.142e-05, "loss": 12.0703, "step": 3143 }, { "epoch": 0.0003144, "grad_norm": 4.216789722442627, "learning_rate": 3.143e-05, "loss": 12.9844, "step": 3144 }, { "epoch": 0.0003145, "grad_norm": 4.4545512199401855, "learning_rate": 3.144e-05, "loss": 12.8359, "step": 3145 }, { "epoch": 0.0003146, "grad_norm": 3.9575791358947754, "learning_rate": 3.1450000000000006e-05, "loss": 11.7266, "step": 3146 }, { "epoch": 0.0003147, "grad_norm": 4.6207380294799805, "learning_rate": 3.146e-05, "loss": 13.1094, "step": 3147 }, { "epoch": 0.0003148, "grad_norm": 4.516714096069336, "learning_rate": 3.147e-05, "loss": 14.0156, "step": 3148 }, { "epoch": 0.0003149, "grad_norm": 4.093890190124512, "learning_rate": 3.1480000000000004e-05, "loss": 11.6406, "step": 3149 }, { "epoch": 0.000315, "grad_norm": 4.16150426864624, "learning_rate": 3.149e-05, "loss": 11.8594, "step": 3150 }, { "epoch": 0.0003151, "grad_norm": 4.188538074493408, "learning_rate": 3.15e-05, "loss": 12.3906, "step": 3151 }, { "epoch": 0.0003152, "grad_norm": 4.296939849853516, "learning_rate": 3.151e-05, "loss": 13.1172, "step": 3152 }, { "epoch": 0.0003153, "grad_norm": 4.2411417961120605, "learning_rate": 3.1519999999999996e-05, "loss": 12.0859, "step": 3153 }, { "epoch": 0.0003154, "grad_norm": 4.440285682678223, "learning_rate": 3.1530000000000005e-05, "loss": 13.3047, "step": 3154 }, { "epoch": 0.0003155, "grad_norm": 4.174127578735352, "learning_rate": 3.154e-05, "loss": 12.1875, "step": 3155 }, { "epoch": 0.0003156, "grad_norm": 4.034418106079102, "learning_rate": 3.155e-05, "loss": 12.3594, "step": 3156 }, { "epoch": 0.0003157, "grad_norm": 4.488844871520996, "learning_rate": 3.156e-05, "loss": 12.8984, "step": 3157 }, { "epoch": 0.0003158, "grad_norm": 4.429039478302002, "learning_rate": 3.1570000000000004e-05, "loss": 12.5859, "step": 3158 }, { "epoch": 0.0003159, "grad_norm": 4.253639221191406, "learning_rate": 3.158e-05, "loss": 12.8984, "step": 3159 }, { "epoch": 0.000316, "grad_norm": 4.056889533996582, "learning_rate": 3.159e-05, "loss": 11.7266, "step": 3160 }, { "epoch": 0.0003161, "grad_norm": 4.251730918884277, "learning_rate": 3.16e-05, "loss": 12.3672, "step": 3161 }, { "epoch": 0.0003162, "grad_norm": 4.355313777923584, "learning_rate": 3.161e-05, "loss": 12.4531, "step": 3162 }, { "epoch": 0.0003163, "grad_norm": 4.073619842529297, "learning_rate": 3.1620000000000006e-05, "loss": 12.2344, "step": 3163 }, { "epoch": 0.0003164, "grad_norm": 3.9847664833068848, "learning_rate": 3.163e-05, "loss": 12.0469, "step": 3164 }, { "epoch": 0.0003165, "grad_norm": 4.491983413696289, "learning_rate": 3.164e-05, "loss": 12.6328, "step": 3165 }, { "epoch": 0.0003166, "grad_norm": 4.011117458343506, "learning_rate": 3.1650000000000004e-05, "loss": 11.8047, "step": 3166 }, { "epoch": 0.0003167, "grad_norm": 4.118056297302246, "learning_rate": 3.166e-05, "loss": 12.7812, "step": 3167 }, { "epoch": 0.0003168, "grad_norm": 3.8982393741607666, "learning_rate": 3.167e-05, "loss": 12.0, "step": 3168 }, { "epoch": 0.0003169, "grad_norm": 4.089506149291992, "learning_rate": 3.168e-05, "loss": 11.6484, "step": 3169 }, { "epoch": 0.000317, "grad_norm": 3.823589563369751, "learning_rate": 3.1689999999999996e-05, "loss": 11.6172, "step": 3170 }, { "epoch": 0.0003171, "grad_norm": 4.54476261138916, "learning_rate": 3.17e-05, "loss": 12.0312, "step": 3171 }, { "epoch": 0.0003172, "grad_norm": 3.9650237560272217, "learning_rate": 3.171e-05, "loss": 11.4609, "step": 3172 }, { "epoch": 0.0003173, "grad_norm": 3.9006104469299316, "learning_rate": 3.172e-05, "loss": 11.1328, "step": 3173 }, { "epoch": 0.0003174, "grad_norm": 4.44875431060791, "learning_rate": 3.173e-05, "loss": 12.3359, "step": 3174 }, { "epoch": 0.0003175, "grad_norm": 4.258243083953857, "learning_rate": 3.1740000000000004e-05, "loss": 11.4609, "step": 3175 }, { "epoch": 0.0003176, "grad_norm": 5.055328369140625, "learning_rate": 3.175e-05, "loss": 13.3125, "step": 3176 }, { "epoch": 0.0003177, "grad_norm": 4.481571197509766, "learning_rate": 3.176e-05, "loss": 12.8672, "step": 3177 }, { "epoch": 0.0003178, "grad_norm": 4.156774044036865, "learning_rate": 3.177e-05, "loss": 11.6328, "step": 3178 }, { "epoch": 0.0003179, "grad_norm": 3.8571271896362305, "learning_rate": 3.178e-05, "loss": 11.8516, "step": 3179 }, { "epoch": 0.000318, "grad_norm": 3.8231124877929688, "learning_rate": 3.1790000000000006e-05, "loss": 11.3594, "step": 3180 }, { "epoch": 0.0003181, "grad_norm": 3.8678410053253174, "learning_rate": 3.18e-05, "loss": 11.3672, "step": 3181 }, { "epoch": 0.0003182, "grad_norm": 3.991071939468384, "learning_rate": 3.1809999999999995e-05, "loss": 11.7266, "step": 3182 }, { "epoch": 0.0003183, "grad_norm": 4.208522796630859, "learning_rate": 3.1820000000000004e-05, "loss": 12.8438, "step": 3183 }, { "epoch": 0.0003184, "grad_norm": 4.157339572906494, "learning_rate": 3.183e-05, "loss": 11.5469, "step": 3184 }, { "epoch": 0.0003185, "grad_norm": 4.158442974090576, "learning_rate": 3.184e-05, "loss": 11.9219, "step": 3185 }, { "epoch": 0.0003186, "grad_norm": 3.9469897747039795, "learning_rate": 3.185e-05, "loss": 11.3906, "step": 3186 }, { "epoch": 0.0003187, "grad_norm": 4.372818946838379, "learning_rate": 3.1859999999999997e-05, "loss": 12.2578, "step": 3187 }, { "epoch": 0.0003188, "grad_norm": 4.051171779632568, "learning_rate": 3.187e-05, "loss": 11.3359, "step": 3188 }, { "epoch": 0.0003189, "grad_norm": 3.806670665740967, "learning_rate": 3.188e-05, "loss": 10.9844, "step": 3189 }, { "epoch": 0.000319, "grad_norm": 4.212409019470215, "learning_rate": 3.189e-05, "loss": 12.5156, "step": 3190 }, { "epoch": 0.0003191, "grad_norm": 4.613623142242432, "learning_rate": 3.19e-05, "loss": 12.9609, "step": 3191 }, { "epoch": 0.0003192, "grad_norm": 4.048587322235107, "learning_rate": 3.1910000000000004e-05, "loss": 12.1328, "step": 3192 }, { "epoch": 0.0003193, "grad_norm": 4.698572635650635, "learning_rate": 3.192e-05, "loss": 12.1562, "step": 3193 }, { "epoch": 0.0003194, "grad_norm": 3.892582893371582, "learning_rate": 3.193e-05, "loss": 11.4062, "step": 3194 }, { "epoch": 0.0003195, "grad_norm": 4.513314247131348, "learning_rate": 3.194e-05, "loss": 12.7891, "step": 3195 }, { "epoch": 0.0003196, "grad_norm": 4.335578918457031, "learning_rate": 3.195e-05, "loss": 11.75, "step": 3196 }, { "epoch": 0.0003197, "grad_norm": 4.395205497741699, "learning_rate": 3.1960000000000006e-05, "loss": 12.2109, "step": 3197 }, { "epoch": 0.0003198, "grad_norm": 4.159797668457031, "learning_rate": 3.197e-05, "loss": 12.0, "step": 3198 }, { "epoch": 0.0003199, "grad_norm": 4.623922824859619, "learning_rate": 3.1979999999999995e-05, "loss": 13.2109, "step": 3199 }, { "epoch": 0.00032, "grad_norm": 4.834364414215088, "learning_rate": 3.1990000000000004e-05, "loss": 12.6797, "step": 3200 }, { "epoch": 0.0003201, "grad_norm": 4.239382266998291, "learning_rate": 3.2e-05, "loss": 11.1797, "step": 3201 }, { "epoch": 0.0003202, "grad_norm": 4.145647048950195, "learning_rate": 3.201e-05, "loss": 11.9219, "step": 3202 }, { "epoch": 0.0003203, "grad_norm": 3.926037311553955, "learning_rate": 3.202e-05, "loss": 10.5156, "step": 3203 }, { "epoch": 0.0003204, "grad_norm": 4.056522846221924, "learning_rate": 3.2029999999999997e-05, "loss": 11.8438, "step": 3204 }, { "epoch": 0.0003205, "grad_norm": 4.285017013549805, "learning_rate": 3.204e-05, "loss": 12.3672, "step": 3205 }, { "epoch": 0.0003206, "grad_norm": 4.016689300537109, "learning_rate": 3.205e-05, "loss": 11.2109, "step": 3206 }, { "epoch": 0.0003207, "grad_norm": 3.714569330215454, "learning_rate": 3.206e-05, "loss": 10.5781, "step": 3207 }, { "epoch": 0.0003208, "grad_norm": 4.070149898529053, "learning_rate": 3.207e-05, "loss": 12.9141, "step": 3208 }, { "epoch": 0.0003209, "grad_norm": 4.141865253448486, "learning_rate": 3.208e-05, "loss": 11.5703, "step": 3209 }, { "epoch": 0.000321, "grad_norm": 3.913559913635254, "learning_rate": 3.209e-05, "loss": 11.0781, "step": 3210 }, { "epoch": 0.0003211, "grad_norm": 4.005146503448486, "learning_rate": 3.21e-05, "loss": 11.5703, "step": 3211 }, { "epoch": 0.0003212, "grad_norm": 4.054205417633057, "learning_rate": 3.211e-05, "loss": 12.1016, "step": 3212 }, { "epoch": 0.0003213, "grad_norm": 4.258857250213623, "learning_rate": 3.212e-05, "loss": 11.5312, "step": 3213 }, { "epoch": 0.0003214, "grad_norm": 4.0924835205078125, "learning_rate": 3.2130000000000006e-05, "loss": 11.8984, "step": 3214 }, { "epoch": 0.0003215, "grad_norm": 4.040266990661621, "learning_rate": 3.214e-05, "loss": 11.5391, "step": 3215 }, { "epoch": 0.0003216, "grad_norm": 4.984847068786621, "learning_rate": 3.2149999999999995e-05, "loss": 13.5234, "step": 3216 }, { "epoch": 0.0003217, "grad_norm": 3.912262201309204, "learning_rate": 3.2160000000000004e-05, "loss": 11.625, "step": 3217 }, { "epoch": 0.0003218, "grad_norm": 3.9006030559539795, "learning_rate": 3.217e-05, "loss": 11.1641, "step": 3218 }, { "epoch": 0.0003219, "grad_norm": 4.044284343719482, "learning_rate": 3.218e-05, "loss": 11.5312, "step": 3219 }, { "epoch": 0.000322, "grad_norm": 4.083791732788086, "learning_rate": 3.219e-05, "loss": 12.5625, "step": 3220 }, { "epoch": 0.0003221, "grad_norm": 4.190662384033203, "learning_rate": 3.22e-05, "loss": 11.5469, "step": 3221 }, { "epoch": 0.0003222, "grad_norm": 4.039274215698242, "learning_rate": 3.221e-05, "loss": 11.7031, "step": 3222 }, { "epoch": 0.0003223, "grad_norm": 3.9752583503723145, "learning_rate": 3.222e-05, "loss": 10.8125, "step": 3223 }, { "epoch": 0.0003224, "grad_norm": 4.353899955749512, "learning_rate": 3.223e-05, "loss": 11.1406, "step": 3224 }, { "epoch": 0.0003225, "grad_norm": 4.121684551239014, "learning_rate": 3.224e-05, "loss": 11.9766, "step": 3225 }, { "epoch": 0.0003226, "grad_norm": 4.066223621368408, "learning_rate": 3.225e-05, "loss": 11.3672, "step": 3226 }, { "epoch": 0.0003227, "grad_norm": 4.243929386138916, "learning_rate": 3.226e-05, "loss": 12.2109, "step": 3227 }, { "epoch": 0.0003228, "grad_norm": 3.994262218475342, "learning_rate": 3.227e-05, "loss": 11.4688, "step": 3228 }, { "epoch": 0.0003229, "grad_norm": 3.9189815521240234, "learning_rate": 3.228e-05, "loss": 11.125, "step": 3229 }, { "epoch": 0.000323, "grad_norm": 4.217655181884766, "learning_rate": 3.229e-05, "loss": 11.7969, "step": 3230 }, { "epoch": 0.0003231, "grad_norm": 4.277668476104736, "learning_rate": 3.2300000000000006e-05, "loss": 12.2734, "step": 3231 }, { "epoch": 0.0003232, "grad_norm": 4.184340476989746, "learning_rate": 3.231e-05, "loss": 11.8281, "step": 3232 }, { "epoch": 0.0003233, "grad_norm": 3.999751567840576, "learning_rate": 3.2319999999999995e-05, "loss": 11.4766, "step": 3233 }, { "epoch": 0.0003234, "grad_norm": 4.286815643310547, "learning_rate": 3.2330000000000004e-05, "loss": 11.4453, "step": 3234 }, { "epoch": 0.0003235, "grad_norm": 4.097026824951172, "learning_rate": 3.234e-05, "loss": 11.2734, "step": 3235 }, { "epoch": 0.0003236, "grad_norm": 4.420822620391846, "learning_rate": 3.235e-05, "loss": 11.9766, "step": 3236 }, { "epoch": 0.0003237, "grad_norm": 4.046294689178467, "learning_rate": 3.236e-05, "loss": 10.9688, "step": 3237 }, { "epoch": 0.0003238, "grad_norm": 4.0677809715271, "learning_rate": 3.237e-05, "loss": 11.6641, "step": 3238 }, { "epoch": 0.0003239, "grad_norm": 4.110946178436279, "learning_rate": 3.238e-05, "loss": 11.1562, "step": 3239 }, { "epoch": 0.000324, "grad_norm": 4.498655319213867, "learning_rate": 3.239e-05, "loss": 12.2891, "step": 3240 }, { "epoch": 0.0003241, "grad_norm": 4.171398162841797, "learning_rate": 3.24e-05, "loss": 11.6875, "step": 3241 }, { "epoch": 0.0003242, "grad_norm": 4.259690284729004, "learning_rate": 3.241e-05, "loss": 11.9062, "step": 3242 }, { "epoch": 0.0003243, "grad_norm": 4.196627140045166, "learning_rate": 3.242e-05, "loss": 11.4844, "step": 3243 }, { "epoch": 0.0003244, "grad_norm": 4.299365043640137, "learning_rate": 3.243e-05, "loss": 11.9297, "step": 3244 }, { "epoch": 0.0003245, "grad_norm": 4.672417163848877, "learning_rate": 3.244e-05, "loss": 12.8359, "step": 3245 }, { "epoch": 0.0003246, "grad_norm": 4.864752769470215, "learning_rate": 3.245e-05, "loss": 12.2266, "step": 3246 }, { "epoch": 0.0003247, "grad_norm": 3.9772098064422607, "learning_rate": 3.246e-05, "loss": 11.5391, "step": 3247 }, { "epoch": 0.0003248, "grad_norm": 3.984213352203369, "learning_rate": 3.2470000000000006e-05, "loss": 11.2188, "step": 3248 }, { "epoch": 0.0003249, "grad_norm": 4.254142761230469, "learning_rate": 3.248e-05, "loss": 11.0781, "step": 3249 }, { "epoch": 0.000325, "grad_norm": 4.016782283782959, "learning_rate": 3.2489999999999996e-05, "loss": 10.7266, "step": 3250 }, { "epoch": 0.0003251, "grad_norm": 4.111670970916748, "learning_rate": 3.2500000000000004e-05, "loss": 10.8047, "step": 3251 }, { "epoch": 0.0003252, "grad_norm": 3.729321002960205, "learning_rate": 3.251e-05, "loss": 10.0312, "step": 3252 }, { "epoch": 0.0003253, "grad_norm": 4.472454071044922, "learning_rate": 3.252e-05, "loss": 11.3984, "step": 3253 }, { "epoch": 0.0003254, "grad_norm": 5.1932196617126465, "learning_rate": 3.253e-05, "loss": 12.0156, "step": 3254 }, { "epoch": 0.0003255, "grad_norm": 4.058501243591309, "learning_rate": 3.254e-05, "loss": 10.6094, "step": 3255 }, { "epoch": 0.0003256, "grad_norm": 4.450124263763428, "learning_rate": 3.255e-05, "loss": 11.2656, "step": 3256 }, { "epoch": 0.0003257, "grad_norm": 4.315848350524902, "learning_rate": 3.256e-05, "loss": 11.5625, "step": 3257 }, { "epoch": 0.0003258, "grad_norm": 4.343723773956299, "learning_rate": 3.257e-05, "loss": 12.2266, "step": 3258 }, { "epoch": 0.0003259, "grad_norm": 4.783463478088379, "learning_rate": 3.258e-05, "loss": 12.2188, "step": 3259 }, { "epoch": 0.000326, "grad_norm": 4.320399761199951, "learning_rate": 3.259e-05, "loss": 11.8281, "step": 3260 }, { "epoch": 0.0003261, "grad_norm": 4.1972551345825195, "learning_rate": 3.26e-05, "loss": 11.5156, "step": 3261 }, { "epoch": 0.0003262, "grad_norm": 3.796144485473633, "learning_rate": 3.261e-05, "loss": 10.1094, "step": 3262 }, { "epoch": 0.0003263, "grad_norm": 3.7178866863250732, "learning_rate": 3.262e-05, "loss": 10.7422, "step": 3263 }, { "epoch": 0.0003264, "grad_norm": 4.453627586364746, "learning_rate": 3.263e-05, "loss": 12.5859, "step": 3264 }, { "epoch": 0.0003265, "grad_norm": 4.911516189575195, "learning_rate": 3.2640000000000006e-05, "loss": 12.3359, "step": 3265 }, { "epoch": 0.0003266, "grad_norm": 4.017552375793457, "learning_rate": 3.265e-05, "loss": 10.2344, "step": 3266 }, { "epoch": 0.0003267, "grad_norm": 4.349109649658203, "learning_rate": 3.2659999999999996e-05, "loss": 11.1641, "step": 3267 }, { "epoch": 0.0003268, "grad_norm": 4.3839430809021, "learning_rate": 3.2670000000000004e-05, "loss": 11.7188, "step": 3268 }, { "epoch": 0.0003269, "grad_norm": 4.069872856140137, "learning_rate": 3.268e-05, "loss": 11.5312, "step": 3269 }, { "epoch": 0.000327, "grad_norm": 4.946285724639893, "learning_rate": 3.269e-05, "loss": 12.8594, "step": 3270 }, { "epoch": 0.0003271, "grad_norm": 4.320945739746094, "learning_rate": 3.27e-05, "loss": 11.7188, "step": 3271 }, { "epoch": 0.0003272, "grad_norm": 4.11987829208374, "learning_rate": 3.271e-05, "loss": 11.2891, "step": 3272 }, { "epoch": 0.0003273, "grad_norm": 4.189293384552002, "learning_rate": 3.272e-05, "loss": 11.2812, "step": 3273 }, { "epoch": 0.0003274, "grad_norm": 4.540557384490967, "learning_rate": 3.273e-05, "loss": 11.7266, "step": 3274 }, { "epoch": 0.0003275, "grad_norm": 4.37672233581543, "learning_rate": 3.274e-05, "loss": 11.25, "step": 3275 }, { "epoch": 0.0003276, "grad_norm": 4.313999176025391, "learning_rate": 3.275e-05, "loss": 11.3672, "step": 3276 }, { "epoch": 0.0003277, "grad_norm": 4.196670055389404, "learning_rate": 3.276e-05, "loss": 12.3125, "step": 3277 }, { "epoch": 0.0003278, "grad_norm": 3.7270560264587402, "learning_rate": 3.277e-05, "loss": 10.4453, "step": 3278 }, { "epoch": 0.0003279, "grad_norm": 3.9468445777893066, "learning_rate": 3.278e-05, "loss": 10.5625, "step": 3279 }, { "epoch": 0.000328, "grad_norm": 4.433849334716797, "learning_rate": 3.279e-05, "loss": 12.7969, "step": 3280 }, { "epoch": 0.0003281, "grad_norm": 4.000179290771484, "learning_rate": 3.28e-05, "loss": 10.4688, "step": 3281 }, { "epoch": 0.0003282, "grad_norm": 4.158862113952637, "learning_rate": 3.2810000000000006e-05, "loss": 10.7656, "step": 3282 }, { "epoch": 0.0003283, "grad_norm": 4.051787853240967, "learning_rate": 3.282e-05, "loss": 10.8594, "step": 3283 }, { "epoch": 0.0003284, "grad_norm": 4.009098529815674, "learning_rate": 3.2829999999999996e-05, "loss": 11.3594, "step": 3284 }, { "epoch": 0.0003285, "grad_norm": 4.038074493408203, "learning_rate": 3.2840000000000004e-05, "loss": 11.25, "step": 3285 }, { "epoch": 0.0003286, "grad_norm": 4.692169666290283, "learning_rate": 3.285e-05, "loss": 11.8828, "step": 3286 }, { "epoch": 0.0003287, "grad_norm": 4.001228332519531, "learning_rate": 3.286e-05, "loss": 10.6562, "step": 3287 }, { "epoch": 0.0003288, "grad_norm": 4.636309623718262, "learning_rate": 3.287e-05, "loss": 11.6328, "step": 3288 }, { "epoch": 0.0003289, "grad_norm": 4.129543781280518, "learning_rate": 3.288e-05, "loss": 11.5547, "step": 3289 }, { "epoch": 0.000329, "grad_norm": 4.099421977996826, "learning_rate": 3.289e-05, "loss": 11.0469, "step": 3290 }, { "epoch": 0.0003291, "grad_norm": 4.2602643966674805, "learning_rate": 3.29e-05, "loss": 11.9375, "step": 3291 }, { "epoch": 0.0003292, "grad_norm": 4.596462726593018, "learning_rate": 3.291e-05, "loss": 12.2734, "step": 3292 }, { "epoch": 0.0003293, "grad_norm": 4.166189670562744, "learning_rate": 3.292e-05, "loss": 11.4375, "step": 3293 }, { "epoch": 0.0003294, "grad_norm": 5.362889766693115, "learning_rate": 3.293e-05, "loss": 12.2266, "step": 3294 }, { "epoch": 0.0003295, "grad_norm": 4.621495246887207, "learning_rate": 3.294e-05, "loss": 12.7969, "step": 3295 }, { "epoch": 0.0003296, "grad_norm": 4.488149642944336, "learning_rate": 3.295e-05, "loss": 11.625, "step": 3296 }, { "epoch": 0.0003297, "grad_norm": 3.998244524002075, "learning_rate": 3.296e-05, "loss": 10.5703, "step": 3297 }, { "epoch": 0.0003298, "grad_norm": 4.318847179412842, "learning_rate": 3.297e-05, "loss": 11.5781, "step": 3298 }, { "epoch": 0.0003299, "grad_norm": 4.119663715362549, "learning_rate": 3.2980000000000006e-05, "loss": 10.5, "step": 3299 }, { "epoch": 0.00033, "grad_norm": 3.9446024894714355, "learning_rate": 3.299e-05, "loss": 10.4844, "step": 3300 }, { "epoch": 0.0003301, "grad_norm": 4.198339939117432, "learning_rate": 3.2999999999999996e-05, "loss": 11.1875, "step": 3301 }, { "epoch": 0.0003302, "grad_norm": 4.08008337020874, "learning_rate": 3.3010000000000004e-05, "loss": 10.8828, "step": 3302 }, { "epoch": 0.0003303, "grad_norm": 4.160445213317871, "learning_rate": 3.302e-05, "loss": 10.6875, "step": 3303 }, { "epoch": 0.0003304, "grad_norm": 4.125695705413818, "learning_rate": 3.303e-05, "loss": 11.8438, "step": 3304 }, { "epoch": 0.0003305, "grad_norm": 4.040487766265869, "learning_rate": 3.304e-05, "loss": 11.0938, "step": 3305 }, { "epoch": 0.0003306, "grad_norm": 4.203553676605225, "learning_rate": 3.305e-05, "loss": 11.1328, "step": 3306 }, { "epoch": 0.0003307, "grad_norm": 4.719578742980957, "learning_rate": 3.306e-05, "loss": 11.2891, "step": 3307 }, { "epoch": 0.0003308, "grad_norm": 3.790029525756836, "learning_rate": 3.307e-05, "loss": 9.7578, "step": 3308 }, { "epoch": 0.0003309, "grad_norm": 4.178408622741699, "learning_rate": 3.308e-05, "loss": 11.0547, "step": 3309 }, { "epoch": 0.000331, "grad_norm": 3.88385009765625, "learning_rate": 3.309e-05, "loss": 10.5547, "step": 3310 }, { "epoch": 0.0003311, "grad_norm": 4.04526424407959, "learning_rate": 3.31e-05, "loss": 11.1953, "step": 3311 }, { "epoch": 0.0003312, "grad_norm": 4.0054121017456055, "learning_rate": 3.311e-05, "loss": 10.7812, "step": 3312 }, { "epoch": 0.0003313, "grad_norm": 4.404824733734131, "learning_rate": 3.312e-05, "loss": 11.3281, "step": 3313 }, { "epoch": 0.0003314, "grad_norm": 4.11653470993042, "learning_rate": 3.313e-05, "loss": 10.4531, "step": 3314 }, { "epoch": 0.0003315, "grad_norm": 4.128344535827637, "learning_rate": 3.314e-05, "loss": 10.1562, "step": 3315 }, { "epoch": 0.0003316, "grad_norm": 4.2458600997924805, "learning_rate": 3.3150000000000006e-05, "loss": 11.1094, "step": 3316 }, { "epoch": 0.0003317, "grad_norm": 3.804975986480713, "learning_rate": 3.316e-05, "loss": 10.3438, "step": 3317 }, { "epoch": 0.0003318, "grad_norm": 4.2271809577941895, "learning_rate": 3.3169999999999996e-05, "loss": 11.125, "step": 3318 }, { "epoch": 0.0003319, "grad_norm": 3.787332057952881, "learning_rate": 3.3180000000000004e-05, "loss": 10.2812, "step": 3319 }, { "epoch": 0.000332, "grad_norm": 4.239166259765625, "learning_rate": 3.319e-05, "loss": 10.9141, "step": 3320 }, { "epoch": 0.0003321, "grad_norm": 4.154354572296143, "learning_rate": 3.32e-05, "loss": 10.8984, "step": 3321 }, { "epoch": 0.0003322, "grad_norm": 3.8980112075805664, "learning_rate": 3.321e-05, "loss": 10.875, "step": 3322 }, { "epoch": 0.0003323, "grad_norm": 3.9956798553466797, "learning_rate": 3.322e-05, "loss": 11.0078, "step": 3323 }, { "epoch": 0.0003324, "grad_norm": 4.027379512786865, "learning_rate": 3.323e-05, "loss": 10.4844, "step": 3324 }, { "epoch": 0.0003325, "grad_norm": 4.109697341918945, "learning_rate": 3.324e-05, "loss": 10.7266, "step": 3325 }, { "epoch": 0.0003326, "grad_norm": 4.49432897567749, "learning_rate": 3.325e-05, "loss": 11.5391, "step": 3326 }, { "epoch": 0.0003327, "grad_norm": 4.328161239624023, "learning_rate": 3.3260000000000003e-05, "loss": 11.3203, "step": 3327 }, { "epoch": 0.0003328, "grad_norm": 4.315303325653076, "learning_rate": 3.327e-05, "loss": 10.6797, "step": 3328 }, { "epoch": 0.0003329, "grad_norm": 4.174460411071777, "learning_rate": 3.328e-05, "loss": 11.0234, "step": 3329 }, { "epoch": 0.000333, "grad_norm": 3.93855357170105, "learning_rate": 3.329e-05, "loss": 10.3125, "step": 3330 }, { "epoch": 0.0003331, "grad_norm": 4.022221565246582, "learning_rate": 3.33e-05, "loss": 10.7188, "step": 3331 }, { "epoch": 0.0003332, "grad_norm": 4.30765438079834, "learning_rate": 3.331e-05, "loss": 10.0547, "step": 3332 }, { "epoch": 0.0003333, "grad_norm": 3.8584938049316406, "learning_rate": 3.3320000000000006e-05, "loss": 9.8906, "step": 3333 }, { "epoch": 0.0003334, "grad_norm": 4.225302219390869, "learning_rate": 3.333e-05, "loss": 10.1094, "step": 3334 }, { "epoch": 0.0003335, "grad_norm": 3.9800033569335938, "learning_rate": 3.3339999999999996e-05, "loss": 10.4609, "step": 3335 }, { "epoch": 0.0003336, "grad_norm": 4.5449299812316895, "learning_rate": 3.3350000000000004e-05, "loss": 12.0703, "step": 3336 }, { "epoch": 0.0003337, "grad_norm": 4.289771556854248, "learning_rate": 3.336e-05, "loss": 10.6875, "step": 3337 }, { "epoch": 0.0003338, "grad_norm": 4.595143795013428, "learning_rate": 3.337e-05, "loss": 10.6562, "step": 3338 }, { "epoch": 0.0003339, "grad_norm": 4.658059120178223, "learning_rate": 3.338e-05, "loss": 11.6562, "step": 3339 }, { "epoch": 0.000334, "grad_norm": 4.367034435272217, "learning_rate": 3.339e-05, "loss": 11.3984, "step": 3340 }, { "epoch": 0.0003341, "grad_norm": 3.9704926013946533, "learning_rate": 3.34e-05, "loss": 10.4219, "step": 3341 }, { "epoch": 0.0003342, "grad_norm": 4.128273010253906, "learning_rate": 3.341e-05, "loss": 10.6797, "step": 3342 }, { "epoch": 0.0003343, "grad_norm": 4.291218280792236, "learning_rate": 3.342e-05, "loss": 10.8672, "step": 3343 }, { "epoch": 0.0003344, "grad_norm": 4.879637241363525, "learning_rate": 3.3430000000000003e-05, "loss": 11.1172, "step": 3344 }, { "epoch": 0.0003345, "grad_norm": 4.181098461151123, "learning_rate": 3.344e-05, "loss": 10.0469, "step": 3345 }, { "epoch": 0.0003346, "grad_norm": 3.972691774368286, "learning_rate": 3.345e-05, "loss": 10.3828, "step": 3346 }, { "epoch": 0.0003347, "grad_norm": 4.436202526092529, "learning_rate": 3.346e-05, "loss": 10.8672, "step": 3347 }, { "epoch": 0.0003348, "grad_norm": 4.7815446853637695, "learning_rate": 3.347e-05, "loss": 11.0781, "step": 3348 }, { "epoch": 0.0003349, "grad_norm": 3.906643867492676, "learning_rate": 3.348e-05, "loss": 10.2266, "step": 3349 }, { "epoch": 0.000335, "grad_norm": 3.8373446464538574, "learning_rate": 3.3490000000000006e-05, "loss": 9.6016, "step": 3350 }, { "epoch": 0.0003351, "grad_norm": 3.990563154220581, "learning_rate": 3.35e-05, "loss": 10.5078, "step": 3351 }, { "epoch": 0.0003352, "grad_norm": 3.9109091758728027, "learning_rate": 3.3509999999999996e-05, "loss": 11.3516, "step": 3352 }, { "epoch": 0.0003353, "grad_norm": 4.099769592285156, "learning_rate": 3.3520000000000004e-05, "loss": 10.4609, "step": 3353 }, { "epoch": 0.0003354, "grad_norm": 3.9826912879943848, "learning_rate": 3.353e-05, "loss": 10.0938, "step": 3354 }, { "epoch": 0.0003355, "grad_norm": 4.484297752380371, "learning_rate": 3.354e-05, "loss": 10.7344, "step": 3355 }, { "epoch": 0.0003356, "grad_norm": 4.1387200355529785, "learning_rate": 3.355e-05, "loss": 11.4531, "step": 3356 }, { "epoch": 0.0003357, "grad_norm": 3.7992279529571533, "learning_rate": 3.356e-05, "loss": 9.5625, "step": 3357 }, { "epoch": 0.0003358, "grad_norm": 4.342659950256348, "learning_rate": 3.357e-05, "loss": 11.2422, "step": 3358 }, { "epoch": 0.0003359, "grad_norm": 7.785485744476318, "learning_rate": 3.358e-05, "loss": 11.9844, "step": 3359 }, { "epoch": 0.000336, "grad_norm": 3.9241621494293213, "learning_rate": 3.359e-05, "loss": 10.1875, "step": 3360 }, { "epoch": 0.0003361, "grad_norm": 4.481233596801758, "learning_rate": 3.3600000000000004e-05, "loss": 11.1797, "step": 3361 }, { "epoch": 0.0003362, "grad_norm": 4.896907329559326, "learning_rate": 3.361e-05, "loss": 11.3281, "step": 3362 }, { "epoch": 0.0003363, "grad_norm": 4.73055362701416, "learning_rate": 3.362e-05, "loss": 11.8047, "step": 3363 }, { "epoch": 0.0003364, "grad_norm": 4.517014026641846, "learning_rate": 3.363e-05, "loss": 11.5469, "step": 3364 }, { "epoch": 0.0003365, "grad_norm": 4.136028289794922, "learning_rate": 3.364e-05, "loss": 10.5234, "step": 3365 }, { "epoch": 0.0003366, "grad_norm": 3.85905385017395, "learning_rate": 3.365e-05, "loss": 9.8359, "step": 3366 }, { "epoch": 0.0003367, "grad_norm": 4.074779510498047, "learning_rate": 3.3660000000000006e-05, "loss": 10.7578, "step": 3367 }, { "epoch": 0.0003368, "grad_norm": 4.239984035491943, "learning_rate": 3.367e-05, "loss": 10.25, "step": 3368 }, { "epoch": 0.0003369, "grad_norm": 5.920673847198486, "learning_rate": 3.3679999999999996e-05, "loss": 11.5547, "step": 3369 }, { "epoch": 0.000337, "grad_norm": 3.863361120223999, "learning_rate": 3.3690000000000004e-05, "loss": 9.875, "step": 3370 }, { "epoch": 0.0003371, "grad_norm": 4.209554672241211, "learning_rate": 3.37e-05, "loss": 10.3438, "step": 3371 }, { "epoch": 0.0003372, "grad_norm": 4.092309474945068, "learning_rate": 3.371e-05, "loss": 10.0781, "step": 3372 }, { "epoch": 0.0003373, "grad_norm": 4.802425384521484, "learning_rate": 3.372e-05, "loss": 10.8828, "step": 3373 }, { "epoch": 0.0003374, "grad_norm": 4.862827777862549, "learning_rate": 3.373e-05, "loss": 12.1875, "step": 3374 }, { "epoch": 0.0003375, "grad_norm": 4.167935848236084, "learning_rate": 3.374e-05, "loss": 10.1484, "step": 3375 }, { "epoch": 0.0003376, "grad_norm": 4.243851661682129, "learning_rate": 3.375e-05, "loss": 10.5078, "step": 3376 }, { "epoch": 0.0003377, "grad_norm": 4.010030269622803, "learning_rate": 3.376e-05, "loss": 10.0938, "step": 3377 }, { "epoch": 0.0003378, "grad_norm": 3.8439948558807373, "learning_rate": 3.3770000000000004e-05, "loss": 10.1328, "step": 3378 }, { "epoch": 0.0003379, "grad_norm": 3.887462615966797, "learning_rate": 3.378e-05, "loss": 10.3516, "step": 3379 }, { "epoch": 0.000338, "grad_norm": 3.718141555786133, "learning_rate": 3.379e-05, "loss": 10.2578, "step": 3380 }, { "epoch": 0.0003381, "grad_norm": 4.037032127380371, "learning_rate": 3.38e-05, "loss": 10.2109, "step": 3381 }, { "epoch": 0.0003382, "grad_norm": 4.068556785583496, "learning_rate": 3.381e-05, "loss": 10.3984, "step": 3382 }, { "epoch": 0.0003383, "grad_norm": 4.041705131530762, "learning_rate": 3.382e-05, "loss": 10.1172, "step": 3383 }, { "epoch": 0.0003384, "grad_norm": 4.55108642578125, "learning_rate": 3.3830000000000006e-05, "loss": 11.0312, "step": 3384 }, { "epoch": 0.0003385, "grad_norm": 4.181764602661133, "learning_rate": 3.384e-05, "loss": 10.2734, "step": 3385 }, { "epoch": 0.0003386, "grad_norm": 3.910088539123535, "learning_rate": 3.3849999999999996e-05, "loss": 9.8125, "step": 3386 }, { "epoch": 0.0003387, "grad_norm": 4.057553768157959, "learning_rate": 3.3860000000000004e-05, "loss": 10.2891, "step": 3387 }, { "epoch": 0.0003388, "grad_norm": 4.039722919464111, "learning_rate": 3.387e-05, "loss": 10.2891, "step": 3388 }, { "epoch": 0.0003389, "grad_norm": 3.8159613609313965, "learning_rate": 3.388e-05, "loss": 10.1836, "step": 3389 }, { "epoch": 0.000339, "grad_norm": 3.9674181938171387, "learning_rate": 3.389e-05, "loss": 10.0938, "step": 3390 }, { "epoch": 0.0003391, "grad_norm": 3.9209303855895996, "learning_rate": 3.39e-05, "loss": 10.3672, "step": 3391 }, { "epoch": 0.0003392, "grad_norm": 4.239545822143555, "learning_rate": 3.391e-05, "loss": 9.8203, "step": 3392 }, { "epoch": 0.0003393, "grad_norm": 4.004146575927734, "learning_rate": 3.392e-05, "loss": 9.7422, "step": 3393 }, { "epoch": 0.0003394, "grad_norm": 4.08314847946167, "learning_rate": 3.393e-05, "loss": 10.8047, "step": 3394 }, { "epoch": 0.0003395, "grad_norm": 3.988220453262329, "learning_rate": 3.3940000000000004e-05, "loss": 10.3828, "step": 3395 }, { "epoch": 0.0003396, "grad_norm": 4.089816570281982, "learning_rate": 3.395e-05, "loss": 10.0664, "step": 3396 }, { "epoch": 0.0003397, "grad_norm": 4.212343215942383, "learning_rate": 3.396e-05, "loss": 10.5234, "step": 3397 }, { "epoch": 0.0003398, "grad_norm": 3.9881434440612793, "learning_rate": 3.397e-05, "loss": 10.4922, "step": 3398 }, { "epoch": 0.0003399, "grad_norm": 5.337203502655029, "learning_rate": 3.398e-05, "loss": 11.7422, "step": 3399 }, { "epoch": 0.00034, "grad_norm": 4.544561386108398, "learning_rate": 3.399e-05, "loss": 11.1953, "step": 3400 }, { "epoch": 0.0003401, "grad_norm": 4.235530376434326, "learning_rate": 3.4000000000000007e-05, "loss": 10.4375, "step": 3401 }, { "epoch": 0.0003402, "grad_norm": 3.9115917682647705, "learning_rate": 3.401e-05, "loss": 9.6172, "step": 3402 }, { "epoch": 0.0003403, "grad_norm": 3.863034963607788, "learning_rate": 3.4019999999999996e-05, "loss": 9.2422, "step": 3403 }, { "epoch": 0.0003404, "grad_norm": 3.962681293487549, "learning_rate": 3.4030000000000005e-05, "loss": 10.0391, "step": 3404 }, { "epoch": 0.0003405, "grad_norm": 3.9435994625091553, "learning_rate": 3.404e-05, "loss": 9.75, "step": 3405 }, { "epoch": 0.0003406, "grad_norm": 4.049876689910889, "learning_rate": 3.405e-05, "loss": 9.7734, "step": 3406 }, { "epoch": 0.0003407, "grad_norm": 3.991595983505249, "learning_rate": 3.406e-05, "loss": 9.8672, "step": 3407 }, { "epoch": 0.0003408, "grad_norm": 4.0078325271606445, "learning_rate": 3.407e-05, "loss": 9.6797, "step": 3408 }, { "epoch": 0.0003409, "grad_norm": 3.9059255123138428, "learning_rate": 3.408e-05, "loss": 10.4531, "step": 3409 }, { "epoch": 0.000341, "grad_norm": 3.985442638397217, "learning_rate": 3.409e-05, "loss": 9.875, "step": 3410 }, { "epoch": 0.0003411, "grad_norm": 4.073984622955322, "learning_rate": 3.41e-05, "loss": 9.6016, "step": 3411 }, { "epoch": 0.0003412, "grad_norm": 3.815159797668457, "learning_rate": 3.4110000000000004e-05, "loss": 9.5469, "step": 3412 }, { "epoch": 0.0003413, "grad_norm": 3.9025957584381104, "learning_rate": 3.412e-05, "loss": 9.8906, "step": 3413 }, { "epoch": 0.0003414, "grad_norm": 4.264541149139404, "learning_rate": 3.413e-05, "loss": 9.8906, "step": 3414 }, { "epoch": 0.0003415, "grad_norm": 4.153369903564453, "learning_rate": 3.414e-05, "loss": 9.9922, "step": 3415 }, { "epoch": 0.0003416, "grad_norm": 4.489535331726074, "learning_rate": 3.415e-05, "loss": 11.6719, "step": 3416 }, { "epoch": 0.0003417, "grad_norm": 4.108375549316406, "learning_rate": 3.416e-05, "loss": 10.2109, "step": 3417 }, { "epoch": 0.0003418, "grad_norm": 4.414920806884766, "learning_rate": 3.417e-05, "loss": 10.1953, "step": 3418 }, { "epoch": 0.0003419, "grad_norm": 4.087825775146484, "learning_rate": 3.418e-05, "loss": 9.6406, "step": 3419 }, { "epoch": 0.000342, "grad_norm": 4.137132167816162, "learning_rate": 3.4189999999999996e-05, "loss": 10.1094, "step": 3420 }, { "epoch": 0.0003421, "grad_norm": 4.568292617797852, "learning_rate": 3.4200000000000005e-05, "loss": 11.9609, "step": 3421 }, { "epoch": 0.0003422, "grad_norm": 3.8219153881073, "learning_rate": 3.421e-05, "loss": 8.9688, "step": 3422 }, { "epoch": 0.0003423, "grad_norm": 4.35713529586792, "learning_rate": 3.422e-05, "loss": 10.0078, "step": 3423 }, { "epoch": 0.0003424, "grad_norm": 4.1804423332214355, "learning_rate": 3.423e-05, "loss": 9.9375, "step": 3424 }, { "epoch": 0.0003425, "grad_norm": 4.542559623718262, "learning_rate": 3.424e-05, "loss": 10.3906, "step": 3425 }, { "epoch": 0.0003426, "grad_norm": 4.20182466506958, "learning_rate": 3.425e-05, "loss": 10.0703, "step": 3426 }, { "epoch": 0.0003427, "grad_norm": 3.8954553604125977, "learning_rate": 3.426e-05, "loss": 10.0234, "step": 3427 }, { "epoch": 0.0003428, "grad_norm": 3.8842320442199707, "learning_rate": 3.427e-05, "loss": 9.5703, "step": 3428 }, { "epoch": 0.0003429, "grad_norm": 3.9736902713775635, "learning_rate": 3.428e-05, "loss": 9.7656, "step": 3429 }, { "epoch": 0.000343, "grad_norm": 4.112651824951172, "learning_rate": 3.429e-05, "loss": 10.3047, "step": 3430 }, { "epoch": 0.0003431, "grad_norm": 4.080387115478516, "learning_rate": 3.43e-05, "loss": 10.4062, "step": 3431 }, { "epoch": 0.0003432, "grad_norm": 4.10958194732666, "learning_rate": 3.431e-05, "loss": 9.8359, "step": 3432 }, { "epoch": 0.0003433, "grad_norm": 3.89727783203125, "learning_rate": 3.4320000000000003e-05, "loss": 9.3984, "step": 3433 }, { "epoch": 0.0003434, "grad_norm": 3.99847149848938, "learning_rate": 3.433e-05, "loss": 9.6953, "step": 3434 }, { "epoch": 0.0003435, "grad_norm": 4.055912017822266, "learning_rate": 3.434e-05, "loss": 9.5859, "step": 3435 }, { "epoch": 0.0003436, "grad_norm": 4.387887001037598, "learning_rate": 3.435e-05, "loss": 10.9609, "step": 3436 }, { "epoch": 0.0003437, "grad_norm": 4.093634605407715, "learning_rate": 3.4359999999999996e-05, "loss": 9.4062, "step": 3437 }, { "epoch": 0.0003438, "grad_norm": 4.745052337646484, "learning_rate": 3.4370000000000005e-05, "loss": 10.0938, "step": 3438 }, { "epoch": 0.0003439, "grad_norm": 3.901541233062744, "learning_rate": 3.438e-05, "loss": 9.9844, "step": 3439 }, { "epoch": 0.000344, "grad_norm": 3.945960283279419, "learning_rate": 3.439e-05, "loss": 9.7578, "step": 3440 }, { "epoch": 0.0003441, "grad_norm": 3.955289840698242, "learning_rate": 3.44e-05, "loss": 9.8203, "step": 3441 }, { "epoch": 0.0003442, "grad_norm": 3.791595220565796, "learning_rate": 3.441e-05, "loss": 9.4688, "step": 3442 }, { "epoch": 0.0003443, "grad_norm": 4.786859512329102, "learning_rate": 3.442e-05, "loss": 10.7734, "step": 3443 }, { "epoch": 0.0003444, "grad_norm": 3.852816104888916, "learning_rate": 3.443e-05, "loss": 9.4609, "step": 3444 }, { "epoch": 0.0003445, "grad_norm": 3.9114298820495605, "learning_rate": 3.444e-05, "loss": 10.0781, "step": 3445 }, { "epoch": 0.0003446, "grad_norm": 3.7254295349121094, "learning_rate": 3.445e-05, "loss": 9.8594, "step": 3446 }, { "epoch": 0.0003447, "grad_norm": 4.128079891204834, "learning_rate": 3.446e-05, "loss": 9.7188, "step": 3447 }, { "epoch": 0.0003448, "grad_norm": 3.8521411418914795, "learning_rate": 3.447e-05, "loss": 10.0938, "step": 3448 }, { "epoch": 0.0003449, "grad_norm": 3.7841033935546875, "learning_rate": 3.448e-05, "loss": 9.2031, "step": 3449 }, { "epoch": 0.000345, "grad_norm": 3.6274397373199463, "learning_rate": 3.4490000000000004e-05, "loss": 8.9766, "step": 3450 }, { "epoch": 0.0003451, "grad_norm": 3.8506691455841064, "learning_rate": 3.45e-05, "loss": 9.6016, "step": 3451 }, { "epoch": 0.0003452, "grad_norm": 4.06758975982666, "learning_rate": 3.451e-05, "loss": 9.6484, "step": 3452 }, { "epoch": 0.0003453, "grad_norm": 4.5061798095703125, "learning_rate": 3.452e-05, "loss": 10.5938, "step": 3453 }, { "epoch": 0.0003454, "grad_norm": 4.398696422576904, "learning_rate": 3.4529999999999996e-05, "loss": 10.4062, "step": 3454 }, { "epoch": 0.0003455, "grad_norm": 4.197793483734131, "learning_rate": 3.4540000000000005e-05, "loss": 10.7188, "step": 3455 }, { "epoch": 0.0003456, "grad_norm": 4.172128677368164, "learning_rate": 3.455e-05, "loss": 9.4766, "step": 3456 }, { "epoch": 0.0003457, "grad_norm": 4.2549943923950195, "learning_rate": 3.456e-05, "loss": 9.6719, "step": 3457 }, { "epoch": 0.0003458, "grad_norm": 4.00117301940918, "learning_rate": 3.457e-05, "loss": 9.3828, "step": 3458 }, { "epoch": 0.0003459, "grad_norm": 4.461249828338623, "learning_rate": 3.458e-05, "loss": 9.9141, "step": 3459 }, { "epoch": 0.000346, "grad_norm": 4.20103120803833, "learning_rate": 3.459e-05, "loss": 9.6953, "step": 3460 }, { "epoch": 0.0003461, "grad_norm": 4.055276393890381, "learning_rate": 3.46e-05, "loss": 9.7344, "step": 3461 }, { "epoch": 0.0003462, "grad_norm": 3.7303593158721924, "learning_rate": 3.461e-05, "loss": 9.0312, "step": 3462 }, { "epoch": 0.0003463, "grad_norm": 4.2501420974731445, "learning_rate": 3.462e-05, "loss": 9.4922, "step": 3463 }, { "epoch": 0.0003464, "grad_norm": 4.156157493591309, "learning_rate": 3.463e-05, "loss": 10.5234, "step": 3464 }, { "epoch": 0.0003465, "grad_norm": 3.9639360904693604, "learning_rate": 3.464e-05, "loss": 9.7422, "step": 3465 }, { "epoch": 0.0003466, "grad_norm": 4.065398216247559, "learning_rate": 3.465e-05, "loss": 9.5781, "step": 3466 }, { "epoch": 0.0003467, "grad_norm": 4.593960285186768, "learning_rate": 3.4660000000000004e-05, "loss": 9.9492, "step": 3467 }, { "epoch": 0.0003468, "grad_norm": 3.992659091949463, "learning_rate": 3.467e-05, "loss": 9.2578, "step": 3468 }, { "epoch": 0.0003469, "grad_norm": 4.568031311035156, "learning_rate": 3.468e-05, "loss": 10.0391, "step": 3469 }, { "epoch": 0.000347, "grad_norm": 3.8288345336914062, "learning_rate": 3.469e-05, "loss": 9.3359, "step": 3470 }, { "epoch": 0.0003471, "grad_norm": 4.219034671783447, "learning_rate": 3.4699999999999996e-05, "loss": 9.6016, "step": 3471 }, { "epoch": 0.0003472, "grad_norm": 4.229843616485596, "learning_rate": 3.4710000000000005e-05, "loss": 9.5312, "step": 3472 }, { "epoch": 0.0003473, "grad_norm": 3.826164960861206, "learning_rate": 3.472e-05, "loss": 9.6172, "step": 3473 }, { "epoch": 0.0003474, "grad_norm": 3.8404030799865723, "learning_rate": 3.473e-05, "loss": 8.8984, "step": 3474 }, { "epoch": 0.0003475, "grad_norm": 3.9308347702026367, "learning_rate": 3.474e-05, "loss": 9.4609, "step": 3475 }, { "epoch": 0.0003476, "grad_norm": 3.8785057067871094, "learning_rate": 3.475e-05, "loss": 9.25, "step": 3476 }, { "epoch": 0.0003477, "grad_norm": 4.225211143493652, "learning_rate": 3.476e-05, "loss": 9.957, "step": 3477 }, { "epoch": 0.0003478, "grad_norm": 3.824558734893799, "learning_rate": 3.477e-05, "loss": 8.8125, "step": 3478 }, { "epoch": 0.0003479, "grad_norm": 3.9660282135009766, "learning_rate": 3.478e-05, "loss": 9.2305, "step": 3479 }, { "epoch": 0.000348, "grad_norm": 4.11309289932251, "learning_rate": 3.479e-05, "loss": 9.7188, "step": 3480 }, { "epoch": 0.0003481, "grad_norm": 3.983299732208252, "learning_rate": 3.48e-05, "loss": 9.6719, "step": 3481 }, { "epoch": 0.0003482, "grad_norm": 3.830436944961548, "learning_rate": 3.481e-05, "loss": 9.5273, "step": 3482 }, { "epoch": 0.0003483, "grad_norm": 3.9448773860931396, "learning_rate": 3.482e-05, "loss": 9.5156, "step": 3483 }, { "epoch": 0.0003484, "grad_norm": 4.055325031280518, "learning_rate": 3.4830000000000004e-05, "loss": 8.8555, "step": 3484 }, { "epoch": 0.0003485, "grad_norm": 3.7771012783050537, "learning_rate": 3.484e-05, "loss": 9.1094, "step": 3485 }, { "epoch": 0.0003486, "grad_norm": 4.95332145690918, "learning_rate": 3.485e-05, "loss": 9.7656, "step": 3486 }, { "epoch": 0.0003487, "grad_norm": 4.324108600616455, "learning_rate": 3.486e-05, "loss": 10.3516, "step": 3487 }, { "epoch": 0.0003488, "grad_norm": 4.3664164543151855, "learning_rate": 3.4869999999999996e-05, "loss": 9.8555, "step": 3488 }, { "epoch": 0.0003489, "grad_norm": 4.212305068969727, "learning_rate": 3.4880000000000005e-05, "loss": 9.3047, "step": 3489 }, { "epoch": 0.000349, "grad_norm": 4.726918697357178, "learning_rate": 3.489e-05, "loss": 9.9375, "step": 3490 }, { "epoch": 0.0003491, "grad_norm": 4.540855407714844, "learning_rate": 3.49e-05, "loss": 8.9531, "step": 3491 }, { "epoch": 0.0003492, "grad_norm": 3.9311349391937256, "learning_rate": 3.491e-05, "loss": 8.4844, "step": 3492 }, { "epoch": 0.0003493, "grad_norm": 3.7078661918640137, "learning_rate": 3.492e-05, "loss": 8.4766, "step": 3493 }, { "epoch": 0.0003494, "grad_norm": 4.00162935256958, "learning_rate": 3.493e-05, "loss": 10.0859, "step": 3494 }, { "epoch": 0.0003495, "grad_norm": 4.181024074554443, "learning_rate": 3.494e-05, "loss": 9.793, "step": 3495 }, { "epoch": 0.0003496, "grad_norm": 4.578210830688477, "learning_rate": 3.495e-05, "loss": 10.6016, "step": 3496 }, { "epoch": 0.0003497, "grad_norm": 3.7809059619903564, "learning_rate": 3.496e-05, "loss": 8.1562, "step": 3497 }, { "epoch": 0.0003498, "grad_norm": 6.469610691070557, "learning_rate": 3.497e-05, "loss": 10.7969, "step": 3498 }, { "epoch": 0.0003499, "grad_norm": 3.9823861122131348, "learning_rate": 3.498e-05, "loss": 9.3828, "step": 3499 }, { "epoch": 0.00035, "grad_norm": 3.8631653785705566, "learning_rate": 3.499e-05, "loss": 8.6016, "step": 3500 }, { "epoch": 0.00035, "eval_loss": 1.0911582708358765, "eval_runtime": 406.3204, "eval_samples_per_second": 24.611, "eval_steps_per_second": 1.538, "step": 3500 }, { "epoch": 0.0003501, "grad_norm": 4.085127353668213, "learning_rate": 3.5000000000000004e-05, "loss": 9.5312, "step": 3501 }, { "epoch": 0.0003502, "grad_norm": 3.858337163925171, "learning_rate": 3.501e-05, "loss": 9.0234, "step": 3502 }, { "epoch": 0.0003503, "grad_norm": 4.071375370025635, "learning_rate": 3.502e-05, "loss": 9.2969, "step": 3503 }, { "epoch": 0.0003504, "grad_norm": 4.4632086753845215, "learning_rate": 3.503e-05, "loss": 9.6953, "step": 3504 }, { "epoch": 0.0003505, "grad_norm": 4.041941165924072, "learning_rate": 3.5039999999999997e-05, "loss": 9.3281, "step": 3505 }, { "epoch": 0.0003506, "grad_norm": 5.256340026855469, "learning_rate": 3.5050000000000005e-05, "loss": 10.0703, "step": 3506 }, { "epoch": 0.0003507, "grad_norm": 4.134398460388184, "learning_rate": 3.506e-05, "loss": 8.8789, "step": 3507 }, { "epoch": 0.0003508, "grad_norm": 3.902724504470825, "learning_rate": 3.507e-05, "loss": 9.0977, "step": 3508 }, { "epoch": 0.0003509, "grad_norm": 4.211234092712402, "learning_rate": 3.508e-05, "loss": 10.0, "step": 3509 }, { "epoch": 0.000351, "grad_norm": 4.0008158683776855, "learning_rate": 3.509e-05, "loss": 8.8555, "step": 3510 }, { "epoch": 0.0003511, "grad_norm": 5.827559947967529, "learning_rate": 3.51e-05, "loss": 11.7227, "step": 3511 }, { "epoch": 0.0003512, "grad_norm": 4.666118144989014, "learning_rate": 3.511e-05, "loss": 9.7969, "step": 3512 }, { "epoch": 0.0003513, "grad_norm": 4.6113972663879395, "learning_rate": 3.512e-05, "loss": 10.4609, "step": 3513 }, { "epoch": 0.0003514, "grad_norm": 3.9297800064086914, "learning_rate": 3.513e-05, "loss": 9.6797, "step": 3514 }, { "epoch": 0.0003515, "grad_norm": 4.332104682922363, "learning_rate": 3.514e-05, "loss": 9.3711, "step": 3515 }, { "epoch": 0.0003516, "grad_norm": 4.328232765197754, "learning_rate": 3.515e-05, "loss": 9.9453, "step": 3516 }, { "epoch": 0.0003517, "grad_norm": 4.012649059295654, "learning_rate": 3.516e-05, "loss": 8.9492, "step": 3517 }, { "epoch": 0.0003518, "grad_norm": 4.154623985290527, "learning_rate": 3.5170000000000004e-05, "loss": 10.1797, "step": 3518 }, { "epoch": 0.0003519, "grad_norm": 4.945087432861328, "learning_rate": 3.518e-05, "loss": 10.9375, "step": 3519 }, { "epoch": 0.000352, "grad_norm": 3.8574047088623047, "learning_rate": 3.519e-05, "loss": 8.6875, "step": 3520 }, { "epoch": 0.0003521, "grad_norm": 4.654843807220459, "learning_rate": 3.52e-05, "loss": 9.9141, "step": 3521 }, { "epoch": 0.0003522, "grad_norm": 3.8247666358947754, "learning_rate": 3.521e-05, "loss": 8.5391, "step": 3522 }, { "epoch": 0.0003523, "grad_norm": 4.814696311950684, "learning_rate": 3.5220000000000005e-05, "loss": 10.5859, "step": 3523 }, { "epoch": 0.0003524, "grad_norm": 3.8880372047424316, "learning_rate": 3.523e-05, "loss": 9.0508, "step": 3524 }, { "epoch": 0.0003525, "grad_norm": 4.266276836395264, "learning_rate": 3.524e-05, "loss": 9.4453, "step": 3525 }, { "epoch": 0.0003526, "grad_norm": 3.8841683864593506, "learning_rate": 3.525e-05, "loss": 8.9609, "step": 3526 }, { "epoch": 0.0003527, "grad_norm": 4.504324913024902, "learning_rate": 3.526e-05, "loss": 9.9609, "step": 3527 }, { "epoch": 0.0003528, "grad_norm": 3.907808542251587, "learning_rate": 3.527e-05, "loss": 9.4102, "step": 3528 }, { "epoch": 0.0003529, "grad_norm": 4.008424282073975, "learning_rate": 3.528e-05, "loss": 9.2422, "step": 3529 }, { "epoch": 0.000353, "grad_norm": 4.003440856933594, "learning_rate": 3.529e-05, "loss": 9.6172, "step": 3530 }, { "epoch": 0.0003531, "grad_norm": 3.8157765865325928, "learning_rate": 3.53e-05, "loss": 8.8867, "step": 3531 }, { "epoch": 0.0003532, "grad_norm": 3.9825029373168945, "learning_rate": 3.531e-05, "loss": 9.1055, "step": 3532 }, { "epoch": 0.0003533, "grad_norm": 4.119210720062256, "learning_rate": 3.532e-05, "loss": 9.625, "step": 3533 }, { "epoch": 0.0003534, "grad_norm": 3.9707088470458984, "learning_rate": 3.533e-05, "loss": 8.6953, "step": 3534 }, { "epoch": 0.0003535, "grad_norm": 4.058837413787842, "learning_rate": 3.5340000000000004e-05, "loss": 8.9844, "step": 3535 }, { "epoch": 0.0003536, "grad_norm": 3.7465462684631348, "learning_rate": 3.535e-05, "loss": 9.2578, "step": 3536 }, { "epoch": 0.0003537, "grad_norm": 3.902986526489258, "learning_rate": 3.536e-05, "loss": 8.3711, "step": 3537 }, { "epoch": 0.0003538, "grad_norm": 4.816512107849121, "learning_rate": 3.537e-05, "loss": 9.8281, "step": 3538 }, { "epoch": 0.0003539, "grad_norm": 3.9221103191375732, "learning_rate": 3.538e-05, "loss": 8.418, "step": 3539 }, { "epoch": 0.000354, "grad_norm": 4.291802406311035, "learning_rate": 3.5390000000000005e-05, "loss": 9.7969, "step": 3540 }, { "epoch": 0.0003541, "grad_norm": 3.994638681411743, "learning_rate": 3.54e-05, "loss": 8.918, "step": 3541 }, { "epoch": 0.0003542, "grad_norm": 3.9600296020507812, "learning_rate": 3.541e-05, "loss": 9.5234, "step": 3542 }, { "epoch": 0.0003543, "grad_norm": 4.546225547790527, "learning_rate": 3.542e-05, "loss": 9.7305, "step": 3543 }, { "epoch": 0.0003544, "grad_norm": 4.5136284828186035, "learning_rate": 3.543e-05, "loss": 10.0391, "step": 3544 }, { "epoch": 0.0003545, "grad_norm": 5.081112384796143, "learning_rate": 3.544e-05, "loss": 8.957, "step": 3545 }, { "epoch": 0.0003546, "grad_norm": 4.139482498168945, "learning_rate": 3.545e-05, "loss": 9.1914, "step": 3546 }, { "epoch": 0.0003547, "grad_norm": 3.9847028255462646, "learning_rate": 3.546e-05, "loss": 9.1953, "step": 3547 }, { "epoch": 0.0003548, "grad_norm": 4.230966091156006, "learning_rate": 3.547e-05, "loss": 9.7344, "step": 3548 }, { "epoch": 0.0003549, "grad_norm": 4.030985355377197, "learning_rate": 3.548e-05, "loss": 9.4336, "step": 3549 }, { "epoch": 0.000355, "grad_norm": 3.858213186264038, "learning_rate": 3.549e-05, "loss": 8.6016, "step": 3550 }, { "epoch": 0.0003551, "grad_norm": 4.147913932800293, "learning_rate": 3.55e-05, "loss": 9.5781, "step": 3551 }, { "epoch": 0.0003552, "grad_norm": 3.7448112964630127, "learning_rate": 3.5510000000000004e-05, "loss": 8.3398, "step": 3552 }, { "epoch": 0.0003553, "grad_norm": 4.218708038330078, "learning_rate": 3.552e-05, "loss": 9.3594, "step": 3553 }, { "epoch": 0.0003554, "grad_norm": 4.213053226470947, "learning_rate": 3.553e-05, "loss": 9.0547, "step": 3554 }, { "epoch": 0.0003555, "grad_norm": 4.417523384094238, "learning_rate": 3.554e-05, "loss": 9.7734, "step": 3555 }, { "epoch": 0.0003556, "grad_norm": 4.0552802085876465, "learning_rate": 3.555e-05, "loss": 8.9805, "step": 3556 }, { "epoch": 0.0003557, "grad_norm": 3.6911566257476807, "learning_rate": 3.5560000000000005e-05, "loss": 8.3242, "step": 3557 }, { "epoch": 0.0003558, "grad_norm": 4.035233020782471, "learning_rate": 3.557e-05, "loss": 9.1133, "step": 3558 }, { "epoch": 0.0003559, "grad_norm": 4.085198879241943, "learning_rate": 3.558e-05, "loss": 9.1875, "step": 3559 }, { "epoch": 0.000356, "grad_norm": 4.129579544067383, "learning_rate": 3.559e-05, "loss": 9.4219, "step": 3560 }, { "epoch": 0.0003561, "grad_norm": 3.667135715484619, "learning_rate": 3.56e-05, "loss": 8.2422, "step": 3561 }, { "epoch": 0.0003562, "grad_norm": 4.100320816040039, "learning_rate": 3.561e-05, "loss": 9.1953, "step": 3562 }, { "epoch": 0.0003563, "grad_norm": 3.9646003246307373, "learning_rate": 3.562e-05, "loss": 9.4219, "step": 3563 }, { "epoch": 0.0003564, "grad_norm": 4.220263481140137, "learning_rate": 3.563e-05, "loss": 9.6055, "step": 3564 }, { "epoch": 0.0003565, "grad_norm": 3.8862528800964355, "learning_rate": 3.564e-05, "loss": 8.8086, "step": 3565 }, { "epoch": 0.0003566, "grad_norm": 3.772306442260742, "learning_rate": 3.565e-05, "loss": 8.5469, "step": 3566 }, { "epoch": 0.0003567, "grad_norm": 4.013455867767334, "learning_rate": 3.566e-05, "loss": 9.1953, "step": 3567 }, { "epoch": 0.0003568, "grad_norm": 4.338392734527588, "learning_rate": 3.567e-05, "loss": 10.3477, "step": 3568 }, { "epoch": 0.0003569, "grad_norm": 3.906348943710327, "learning_rate": 3.5680000000000004e-05, "loss": 8.2539, "step": 3569 }, { "epoch": 0.000357, "grad_norm": 4.116657257080078, "learning_rate": 3.569e-05, "loss": 8.9727, "step": 3570 }, { "epoch": 0.0003571, "grad_norm": 4.163711071014404, "learning_rate": 3.57e-05, "loss": 8.5938, "step": 3571 }, { "epoch": 0.0003572, "grad_norm": 4.249398231506348, "learning_rate": 3.571e-05, "loss": 9.9141, "step": 3572 }, { "epoch": 0.0003573, "grad_norm": 4.596261978149414, "learning_rate": 3.572e-05, "loss": 9.3125, "step": 3573 }, { "epoch": 0.0003574, "grad_norm": 3.8098151683807373, "learning_rate": 3.5730000000000005e-05, "loss": 8.5391, "step": 3574 }, { "epoch": 0.0003575, "grad_norm": 3.6537082195281982, "learning_rate": 3.574e-05, "loss": 7.9297, "step": 3575 }, { "epoch": 0.0003576, "grad_norm": 3.924492597579956, "learning_rate": 3.575e-05, "loss": 8.6953, "step": 3576 }, { "epoch": 0.0003577, "grad_norm": 3.9190399646759033, "learning_rate": 3.576e-05, "loss": 8.8711, "step": 3577 }, { "epoch": 0.0003578, "grad_norm": 3.6862311363220215, "learning_rate": 3.577e-05, "loss": 7.75, "step": 3578 }, { "epoch": 0.0003579, "grad_norm": 3.8659183979034424, "learning_rate": 3.578e-05, "loss": 8.3867, "step": 3579 }, { "epoch": 0.000358, "grad_norm": 3.56437611579895, "learning_rate": 3.579e-05, "loss": 8.125, "step": 3580 }, { "epoch": 0.0003581, "grad_norm": 4.0573554039001465, "learning_rate": 3.58e-05, "loss": 8.3633, "step": 3581 }, { "epoch": 0.0003582, "grad_norm": 4.204733371734619, "learning_rate": 3.581e-05, "loss": 8.7305, "step": 3582 }, { "epoch": 0.0003583, "grad_norm": 4.150760650634766, "learning_rate": 3.582e-05, "loss": 8.4297, "step": 3583 }, { "epoch": 0.0003584, "grad_norm": 3.890657424926758, "learning_rate": 3.583e-05, "loss": 8.8594, "step": 3584 }, { "epoch": 0.0003585, "grad_norm": 3.930206060409546, "learning_rate": 3.584e-05, "loss": 8.8047, "step": 3585 }, { "epoch": 0.0003586, "grad_norm": 4.1720709800720215, "learning_rate": 3.5850000000000004e-05, "loss": 9.0977, "step": 3586 }, { "epoch": 0.0003587, "grad_norm": 3.872006416320801, "learning_rate": 3.586e-05, "loss": 8.4922, "step": 3587 }, { "epoch": 0.0003588, "grad_norm": 4.156218528747559, "learning_rate": 3.587e-05, "loss": 9.1328, "step": 3588 }, { "epoch": 0.0003589, "grad_norm": 4.399153232574463, "learning_rate": 3.588e-05, "loss": 8.7031, "step": 3589 }, { "epoch": 0.000359, "grad_norm": 3.917479991912842, "learning_rate": 3.589e-05, "loss": 9.1797, "step": 3590 }, { "epoch": 0.0003591, "grad_norm": 4.007040023803711, "learning_rate": 3.5900000000000005e-05, "loss": 8.9062, "step": 3591 }, { "epoch": 0.0003592, "grad_norm": 4.1533637046813965, "learning_rate": 3.591e-05, "loss": 8.9102, "step": 3592 }, { "epoch": 0.0003593, "grad_norm": 3.6185474395751953, "learning_rate": 3.592e-05, "loss": 8.0586, "step": 3593 }, { "epoch": 0.0003594, "grad_norm": 4.260969638824463, "learning_rate": 3.593e-05, "loss": 9.4062, "step": 3594 }, { "epoch": 0.0003595, "grad_norm": 4.164505958557129, "learning_rate": 3.594e-05, "loss": 8.9219, "step": 3595 }, { "epoch": 0.0003596, "grad_norm": 4.0043416023254395, "learning_rate": 3.595e-05, "loss": 8.3594, "step": 3596 }, { "epoch": 0.0003597, "grad_norm": 4.063927173614502, "learning_rate": 3.596e-05, "loss": 8.3047, "step": 3597 }, { "epoch": 0.0003598, "grad_norm": 3.658076524734497, "learning_rate": 3.597e-05, "loss": 8.2109, "step": 3598 }, { "epoch": 0.0003599, "grad_norm": 3.563760995864868, "learning_rate": 3.598e-05, "loss": 8.1445, "step": 3599 }, { "epoch": 0.00036, "grad_norm": 3.7824599742889404, "learning_rate": 3.599e-05, "loss": 8.1953, "step": 3600 }, { "epoch": 0.0003601, "grad_norm": 4.007077217102051, "learning_rate": 3.6e-05, "loss": 8.5938, "step": 3601 }, { "epoch": 0.0003602, "grad_norm": 4.472359657287598, "learning_rate": 3.601e-05, "loss": 9.9648, "step": 3602 }, { "epoch": 0.0003603, "grad_norm": 3.737208366394043, "learning_rate": 3.6020000000000004e-05, "loss": 7.6875, "step": 3603 }, { "epoch": 0.0003604, "grad_norm": 3.7609057426452637, "learning_rate": 3.603e-05, "loss": 8.1875, "step": 3604 }, { "epoch": 0.0003605, "grad_norm": 3.8421549797058105, "learning_rate": 3.604e-05, "loss": 8.1523, "step": 3605 }, { "epoch": 0.0003606, "grad_norm": 3.816240072250366, "learning_rate": 3.605e-05, "loss": 7.9062, "step": 3606 }, { "epoch": 0.0003607, "grad_norm": 4.049971103668213, "learning_rate": 3.606e-05, "loss": 8.3516, "step": 3607 }, { "epoch": 0.0003608, "grad_norm": 4.040900230407715, "learning_rate": 3.6070000000000005e-05, "loss": 8.0352, "step": 3608 }, { "epoch": 0.0003609, "grad_norm": 3.776742458343506, "learning_rate": 3.608e-05, "loss": 8.3164, "step": 3609 }, { "epoch": 0.000361, "grad_norm": 3.7162158489227295, "learning_rate": 3.609e-05, "loss": 8.25, "step": 3610 }, { "epoch": 0.0003611, "grad_norm": 3.9722158908843994, "learning_rate": 3.61e-05, "loss": 8.0156, "step": 3611 }, { "epoch": 0.0003612, "grad_norm": 3.803699254989624, "learning_rate": 3.611e-05, "loss": 7.9961, "step": 3612 }, { "epoch": 0.0003613, "grad_norm": 3.976182222366333, "learning_rate": 3.612e-05, "loss": 8.25, "step": 3613 }, { "epoch": 0.0003614, "grad_norm": 3.7838759422302246, "learning_rate": 3.613e-05, "loss": 8.0273, "step": 3614 }, { "epoch": 0.0003615, "grad_norm": 3.8301563262939453, "learning_rate": 3.614e-05, "loss": 8.2812, "step": 3615 }, { "epoch": 0.0003616, "grad_norm": 4.001625061035156, "learning_rate": 3.615e-05, "loss": 8.668, "step": 3616 }, { "epoch": 0.0003617, "grad_norm": 3.792609930038452, "learning_rate": 3.616e-05, "loss": 7.9883, "step": 3617 }, { "epoch": 0.0003618, "grad_norm": 4.283328533172607, "learning_rate": 3.617e-05, "loss": 8.8555, "step": 3618 }, { "epoch": 0.0003619, "grad_norm": 4.05611515045166, "learning_rate": 3.618e-05, "loss": 8.25, "step": 3619 }, { "epoch": 0.000362, "grad_norm": 4.185487270355225, "learning_rate": 3.6190000000000004e-05, "loss": 9.2422, "step": 3620 }, { "epoch": 0.0003621, "grad_norm": 3.8984696865081787, "learning_rate": 3.62e-05, "loss": 8.2812, "step": 3621 }, { "epoch": 0.0003622, "grad_norm": 4.185481071472168, "learning_rate": 3.621e-05, "loss": 9.1211, "step": 3622 }, { "epoch": 0.0003623, "grad_norm": 3.5906546115875244, "learning_rate": 3.622e-05, "loss": 7.6602, "step": 3623 }, { "epoch": 0.0003624, "grad_norm": 3.6980206966400146, "learning_rate": 3.623e-05, "loss": 8.1914, "step": 3624 }, { "epoch": 0.0003625, "grad_norm": 3.6407835483551025, "learning_rate": 3.6240000000000005e-05, "loss": 7.2578, "step": 3625 }, { "epoch": 0.0003626, "grad_norm": 4.0652079582214355, "learning_rate": 3.625e-05, "loss": 8.2617, "step": 3626 }, { "epoch": 0.0003627, "grad_norm": 3.562990188598633, "learning_rate": 3.626e-05, "loss": 7.6875, "step": 3627 }, { "epoch": 0.0003628, "grad_norm": 3.6243600845336914, "learning_rate": 3.6270000000000003e-05, "loss": 8.125, "step": 3628 }, { "epoch": 0.0003629, "grad_norm": 3.663573741912842, "learning_rate": 3.628e-05, "loss": 7.8906, "step": 3629 }, { "epoch": 0.000363, "grad_norm": 4.207284927368164, "learning_rate": 3.629e-05, "loss": 9.0977, "step": 3630 }, { "epoch": 0.0003631, "grad_norm": 3.890348196029663, "learning_rate": 3.63e-05, "loss": 9.2695, "step": 3631 }, { "epoch": 0.0003632, "grad_norm": 4.216334819793701, "learning_rate": 3.631e-05, "loss": 8.6133, "step": 3632 }, { "epoch": 0.0003633, "grad_norm": 4.130468368530273, "learning_rate": 3.632e-05, "loss": 8.3047, "step": 3633 }, { "epoch": 0.0003634, "grad_norm": 3.9328114986419678, "learning_rate": 3.633e-05, "loss": 7.9961, "step": 3634 }, { "epoch": 0.0003635, "grad_norm": 3.8146066665649414, "learning_rate": 3.634e-05, "loss": 7.6328, "step": 3635 }, { "epoch": 0.0003636, "grad_norm": 3.8578176498413086, "learning_rate": 3.635e-05, "loss": 7.9375, "step": 3636 }, { "epoch": 0.0003637, "grad_norm": 4.063955307006836, "learning_rate": 3.6360000000000004e-05, "loss": 8.0273, "step": 3637 }, { "epoch": 0.0003638, "grad_norm": 4.021401405334473, "learning_rate": 3.637e-05, "loss": 8.7617, "step": 3638 }, { "epoch": 0.0003639, "grad_norm": 3.817809820175171, "learning_rate": 3.638e-05, "loss": 8.0, "step": 3639 }, { "epoch": 0.000364, "grad_norm": 4.076965808868408, "learning_rate": 3.639e-05, "loss": 7.9258, "step": 3640 }, { "epoch": 0.0003641, "grad_norm": 3.9256434440612793, "learning_rate": 3.64e-05, "loss": 8.6875, "step": 3641 }, { "epoch": 0.0003642, "grad_norm": 4.945650100708008, "learning_rate": 3.6410000000000005e-05, "loss": 9.8086, "step": 3642 }, { "epoch": 0.0003643, "grad_norm": 3.718984842300415, "learning_rate": 3.642e-05, "loss": 8.2656, "step": 3643 }, { "epoch": 0.0003644, "grad_norm": 4.200788497924805, "learning_rate": 3.643e-05, "loss": 8.6016, "step": 3644 }, { "epoch": 0.0003645, "grad_norm": 4.26768159866333, "learning_rate": 3.6440000000000003e-05, "loss": 8.6016, "step": 3645 }, { "epoch": 0.0003646, "grad_norm": 4.211695194244385, "learning_rate": 3.645e-05, "loss": 8.7188, "step": 3646 }, { "epoch": 0.0003647, "grad_norm": 3.942966938018799, "learning_rate": 3.646e-05, "loss": 7.7617, "step": 3647 }, { "epoch": 0.0003648, "grad_norm": 4.170599460601807, "learning_rate": 3.647e-05, "loss": 8.7305, "step": 3648 }, { "epoch": 0.0003649, "grad_norm": 3.85955810546875, "learning_rate": 3.648e-05, "loss": 7.8828, "step": 3649 }, { "epoch": 0.000365, "grad_norm": 3.8474628925323486, "learning_rate": 3.649e-05, "loss": 8.0391, "step": 3650 }, { "epoch": 0.0003651, "grad_norm": 3.74211049079895, "learning_rate": 3.65e-05, "loss": 7.7383, "step": 3651 }, { "epoch": 0.0003652, "grad_norm": 3.8546195030212402, "learning_rate": 3.651e-05, "loss": 7.6914, "step": 3652 }, { "epoch": 0.0003653, "grad_norm": 4.240731716156006, "learning_rate": 3.652e-05, "loss": 8.3867, "step": 3653 }, { "epoch": 0.0003654, "grad_norm": 3.9579293727874756, "learning_rate": 3.6530000000000004e-05, "loss": 8.2227, "step": 3654 }, { "epoch": 0.0003655, "grad_norm": 3.748444080352783, "learning_rate": 3.654e-05, "loss": 7.6211, "step": 3655 }, { "epoch": 0.0003656, "grad_norm": 4.751547813415527, "learning_rate": 3.655e-05, "loss": 9.2773, "step": 3656 }, { "epoch": 0.0003657, "grad_norm": 3.8329389095306396, "learning_rate": 3.656e-05, "loss": 7.6094, "step": 3657 }, { "epoch": 0.0003658, "grad_norm": 4.126430511474609, "learning_rate": 3.657e-05, "loss": 8.5586, "step": 3658 }, { "epoch": 0.0003659, "grad_norm": 3.931842088699341, "learning_rate": 3.6580000000000006e-05, "loss": 7.6406, "step": 3659 }, { "epoch": 0.000366, "grad_norm": 3.8183586597442627, "learning_rate": 3.659e-05, "loss": 7.9609, "step": 3660 }, { "epoch": 0.0003661, "grad_norm": 3.9115753173828125, "learning_rate": 3.66e-05, "loss": 7.875, "step": 3661 }, { "epoch": 0.0003662, "grad_norm": 3.9533402919769287, "learning_rate": 3.6610000000000004e-05, "loss": 7.918, "step": 3662 }, { "epoch": 0.0003663, "grad_norm": 4.046900272369385, "learning_rate": 3.662e-05, "loss": 8.3789, "step": 3663 }, { "epoch": 0.0003664, "grad_norm": 3.671175479888916, "learning_rate": 3.663e-05, "loss": 7.1758, "step": 3664 }, { "epoch": 0.0003665, "grad_norm": 3.7850921154022217, "learning_rate": 3.664e-05, "loss": 7.4531, "step": 3665 }, { "epoch": 0.0003666, "grad_norm": 3.6717941761016846, "learning_rate": 3.665e-05, "loss": 7.6055, "step": 3666 }, { "epoch": 0.0003667, "grad_norm": 4.0944600105285645, "learning_rate": 3.666e-05, "loss": 7.8047, "step": 3667 }, { "epoch": 0.0003668, "grad_norm": 4.709226131439209, "learning_rate": 3.667e-05, "loss": 7.4805, "step": 3668 }, { "epoch": 0.0003669, "grad_norm": 4.083220958709717, "learning_rate": 3.668e-05, "loss": 8.0234, "step": 3669 }, { "epoch": 0.000367, "grad_norm": 3.77762508392334, "learning_rate": 3.669e-05, "loss": 7.1484, "step": 3670 }, { "epoch": 0.0003671, "grad_norm": 3.892266035079956, "learning_rate": 3.6700000000000004e-05, "loss": 7.3359, "step": 3671 }, { "epoch": 0.0003672, "grad_norm": 4.020049095153809, "learning_rate": 3.671e-05, "loss": 8.1992, "step": 3672 }, { "epoch": 0.0003673, "grad_norm": 4.793400287628174, "learning_rate": 3.672e-05, "loss": 8.7227, "step": 3673 }, { "epoch": 0.0003674, "grad_norm": 4.025148868560791, "learning_rate": 3.673e-05, "loss": 7.8711, "step": 3674 }, { "epoch": 0.0003675, "grad_norm": 3.608661413192749, "learning_rate": 3.674e-05, "loss": 7.4961, "step": 3675 }, { "epoch": 0.0003676, "grad_norm": 4.188398361206055, "learning_rate": 3.675e-05, "loss": 8.375, "step": 3676 }, { "epoch": 0.0003677, "grad_norm": 4.057331085205078, "learning_rate": 3.676e-05, "loss": 8.2812, "step": 3677 }, { "epoch": 0.0003678, "grad_norm": 4.1667351722717285, "learning_rate": 3.677e-05, "loss": 7.9062, "step": 3678 }, { "epoch": 0.0003679, "grad_norm": 4.138629913330078, "learning_rate": 3.6780000000000004e-05, "loss": 8.5117, "step": 3679 }, { "epoch": 0.000368, "grad_norm": 3.8195579051971436, "learning_rate": 3.679e-05, "loss": 8.0859, "step": 3680 }, { "epoch": 0.0003681, "grad_norm": 3.661707878112793, "learning_rate": 3.68e-05, "loss": 7.4688, "step": 3681 }, { "epoch": 0.0003682, "grad_norm": 6.971426486968994, "learning_rate": 3.681e-05, "loss": 8.4531, "step": 3682 }, { "epoch": 0.0003683, "grad_norm": 4.242550373077393, "learning_rate": 3.682e-05, "loss": 8.8867, "step": 3683 }, { "epoch": 0.0003684, "grad_norm": 3.850429058074951, "learning_rate": 3.683e-05, "loss": 7.6055, "step": 3684 }, { "epoch": 0.0003685, "grad_norm": 4.200206756591797, "learning_rate": 3.684e-05, "loss": 7.8633, "step": 3685 }, { "epoch": 0.0003686, "grad_norm": 3.573742389678955, "learning_rate": 3.685e-05, "loss": 7.6797, "step": 3686 }, { "epoch": 0.0003687, "grad_norm": 4.073431491851807, "learning_rate": 3.6859999999999996e-05, "loss": 7.8203, "step": 3687 }, { "epoch": 0.0003688, "grad_norm": 3.715885877609253, "learning_rate": 3.6870000000000004e-05, "loss": 7.4766, "step": 3688 }, { "epoch": 0.0003689, "grad_norm": 3.5381669998168945, "learning_rate": 3.688e-05, "loss": 7.1172, "step": 3689 }, { "epoch": 0.000369, "grad_norm": 4.377144813537598, "learning_rate": 3.689e-05, "loss": 7.6758, "step": 3690 }, { "epoch": 0.0003691, "grad_norm": 3.734431505203247, "learning_rate": 3.69e-05, "loss": 7.5352, "step": 3691 }, { "epoch": 0.0003692, "grad_norm": 3.834195137023926, "learning_rate": 3.691e-05, "loss": 7.4883, "step": 3692 }, { "epoch": 0.0003693, "grad_norm": 3.8302841186523438, "learning_rate": 3.692e-05, "loss": 7.8438, "step": 3693 }, { "epoch": 0.0003694, "grad_norm": 3.6052684783935547, "learning_rate": 3.693e-05, "loss": 7.3203, "step": 3694 }, { "epoch": 0.0003695, "grad_norm": 4.328238487243652, "learning_rate": 3.694e-05, "loss": 7.957, "step": 3695 }, { "epoch": 0.0003696, "grad_norm": 3.764779806137085, "learning_rate": 3.6950000000000004e-05, "loss": 7.5234, "step": 3696 }, { "epoch": 0.0003697, "grad_norm": 3.728281021118164, "learning_rate": 3.696e-05, "loss": 7.0273, "step": 3697 }, { "epoch": 0.0003698, "grad_norm": 3.8631412982940674, "learning_rate": 3.697e-05, "loss": 7.9883, "step": 3698 }, { "epoch": 0.0003699, "grad_norm": 4.021390438079834, "learning_rate": 3.698e-05, "loss": 7.5977, "step": 3699 }, { "epoch": 0.00037, "grad_norm": 3.6802592277526855, "learning_rate": 3.699e-05, "loss": 7.3398, "step": 3700 }, { "epoch": 0.0003701, "grad_norm": 3.6073927879333496, "learning_rate": 3.7e-05, "loss": 8.0781, "step": 3701 }, { "epoch": 0.0003702, "grad_norm": 3.721489906311035, "learning_rate": 3.701e-05, "loss": 7.8828, "step": 3702 }, { "epoch": 0.0003703, "grad_norm": 3.518146514892578, "learning_rate": 3.702e-05, "loss": 7.4141, "step": 3703 }, { "epoch": 0.0003704, "grad_norm": 3.5116593837738037, "learning_rate": 3.7029999999999996e-05, "loss": 7.0742, "step": 3704 }, { "epoch": 0.0003705, "grad_norm": 3.477902412414551, "learning_rate": 3.7040000000000005e-05, "loss": 7.2227, "step": 3705 }, { "epoch": 0.0003706, "grad_norm": 3.4165449142456055, "learning_rate": 3.705e-05, "loss": 7.2852, "step": 3706 }, { "epoch": 0.0003707, "grad_norm": 3.648942470550537, "learning_rate": 3.706e-05, "loss": 7.3516, "step": 3707 }, { "epoch": 0.0003708, "grad_norm": 4.124898433685303, "learning_rate": 3.707e-05, "loss": 7.7109, "step": 3708 }, { "epoch": 0.0003709, "grad_norm": 3.6006293296813965, "learning_rate": 3.708e-05, "loss": 7.6875, "step": 3709 }, { "epoch": 0.000371, "grad_norm": 3.820141553878784, "learning_rate": 3.709e-05, "loss": 7.5312, "step": 3710 }, { "epoch": 0.0003711, "grad_norm": 4.02821683883667, "learning_rate": 3.71e-05, "loss": 8.1055, "step": 3711 }, { "epoch": 0.0003712, "grad_norm": 3.9696545600891113, "learning_rate": 3.711e-05, "loss": 6.9727, "step": 3712 }, { "epoch": 0.0003713, "grad_norm": 3.9362540245056152, "learning_rate": 3.7120000000000004e-05, "loss": 7.5781, "step": 3713 }, { "epoch": 0.0003714, "grad_norm": 3.5901272296905518, "learning_rate": 3.713e-05, "loss": 7.5508, "step": 3714 }, { "epoch": 0.0003715, "grad_norm": 3.581475257873535, "learning_rate": 3.714e-05, "loss": 7.0234, "step": 3715 }, { "epoch": 0.0003716, "grad_norm": 3.9984290599823, "learning_rate": 3.715e-05, "loss": 8.2773, "step": 3716 }, { "epoch": 0.0003717, "grad_norm": 3.859914779663086, "learning_rate": 3.716e-05, "loss": 6.9727, "step": 3717 }, { "epoch": 0.0003718, "grad_norm": 4.044993877410889, "learning_rate": 3.717e-05, "loss": 7.5977, "step": 3718 }, { "epoch": 0.0003719, "grad_norm": 3.8882622718811035, "learning_rate": 3.718e-05, "loss": 7.4727, "step": 3719 }, { "epoch": 0.000372, "grad_norm": 3.8920814990997314, "learning_rate": 3.719e-05, "loss": 7.4062, "step": 3720 }, { "epoch": 0.0003721, "grad_norm": 4.624520301818848, "learning_rate": 3.7199999999999996e-05, "loss": 8.0156, "step": 3721 }, { "epoch": 0.0003722, "grad_norm": 3.944957971572876, "learning_rate": 3.7210000000000005e-05, "loss": 7.4727, "step": 3722 }, { "epoch": 0.0003723, "grad_norm": 3.7715561389923096, "learning_rate": 3.722e-05, "loss": 7.0625, "step": 3723 }, { "epoch": 0.0003724, "grad_norm": 4.020610332489014, "learning_rate": 3.723e-05, "loss": 8.2773, "step": 3724 }, { "epoch": 0.0003725, "grad_norm": 4.559901714324951, "learning_rate": 3.724e-05, "loss": 7.0664, "step": 3725 }, { "epoch": 0.0003726, "grad_norm": 4.080809116363525, "learning_rate": 3.725e-05, "loss": 7.2227, "step": 3726 }, { "epoch": 0.0003727, "grad_norm": 3.548936605453491, "learning_rate": 3.726e-05, "loss": 7.0625, "step": 3727 }, { "epoch": 0.0003728, "grad_norm": 3.688309669494629, "learning_rate": 3.727e-05, "loss": 7.5625, "step": 3728 }, { "epoch": 0.0003729, "grad_norm": 4.651008605957031, "learning_rate": 3.728e-05, "loss": 7.7695, "step": 3729 }, { "epoch": 0.000373, "grad_norm": 3.685234785079956, "learning_rate": 3.7290000000000004e-05, "loss": 7.4258, "step": 3730 }, { "epoch": 0.0003731, "grad_norm": 3.647007465362549, "learning_rate": 3.73e-05, "loss": 7.0898, "step": 3731 }, { "epoch": 0.0003732, "grad_norm": 3.8655905723571777, "learning_rate": 3.731e-05, "loss": 7.4336, "step": 3732 }, { "epoch": 0.0003733, "grad_norm": 3.8040971755981445, "learning_rate": 3.732e-05, "loss": 7.5859, "step": 3733 }, { "epoch": 0.0003734, "grad_norm": 3.605004072189331, "learning_rate": 3.7330000000000003e-05, "loss": 7.125, "step": 3734 }, { "epoch": 0.0003735, "grad_norm": 4.341588973999023, "learning_rate": 3.734e-05, "loss": 8.0078, "step": 3735 }, { "epoch": 0.0003736, "grad_norm": 3.5978193283081055, "learning_rate": 3.735e-05, "loss": 6.4297, "step": 3736 }, { "epoch": 0.0003737, "grad_norm": 3.9439592361450195, "learning_rate": 3.736e-05, "loss": 7.3672, "step": 3737 }, { "epoch": 0.0003738, "grad_norm": 4.321782112121582, "learning_rate": 3.7369999999999996e-05, "loss": 7.7891, "step": 3738 }, { "epoch": 0.0003739, "grad_norm": 3.7560415267944336, "learning_rate": 3.7380000000000005e-05, "loss": 7.1055, "step": 3739 }, { "epoch": 0.000374, "grad_norm": 3.4091031551361084, "learning_rate": 3.739e-05, "loss": 6.6953, "step": 3740 }, { "epoch": 0.0003741, "grad_norm": 3.6737358570098877, "learning_rate": 3.74e-05, "loss": 7.0898, "step": 3741 }, { "epoch": 0.0003742, "grad_norm": 3.517364025115967, "learning_rate": 3.741e-05, "loss": 6.7344, "step": 3742 }, { "epoch": 0.0003743, "grad_norm": 3.7535104751586914, "learning_rate": 3.742e-05, "loss": 7.5898, "step": 3743 }, { "epoch": 0.0003744, "grad_norm": 3.7284533977508545, "learning_rate": 3.743e-05, "loss": 7.3008, "step": 3744 }, { "epoch": 0.0003745, "grad_norm": 3.631471872329712, "learning_rate": 3.744e-05, "loss": 7.0547, "step": 3745 }, { "epoch": 0.0003746, "grad_norm": 3.936616897583008, "learning_rate": 3.745e-05, "loss": 7.3047, "step": 3746 }, { "epoch": 0.0003747, "grad_norm": 3.9024007320404053, "learning_rate": 3.7460000000000004e-05, "loss": 7.9375, "step": 3747 }, { "epoch": 0.0003748, "grad_norm": 3.7867276668548584, "learning_rate": 3.747e-05, "loss": 7.6328, "step": 3748 }, { "epoch": 0.0003749, "grad_norm": 4.628604412078857, "learning_rate": 3.748e-05, "loss": 8.2383, "step": 3749 }, { "epoch": 0.000375, "grad_norm": 3.8571231365203857, "learning_rate": 3.749e-05, "loss": 7.1367, "step": 3750 }, { "epoch": 0.0003751, "grad_norm": 3.697697401046753, "learning_rate": 3.7500000000000003e-05, "loss": 6.5977, "step": 3751 }, { "epoch": 0.0003752, "grad_norm": 4.110589981079102, "learning_rate": 3.751e-05, "loss": 7.0312, "step": 3752 }, { "epoch": 0.0003753, "grad_norm": 3.4988439083099365, "learning_rate": 3.752e-05, "loss": 7.1328, "step": 3753 }, { "epoch": 0.0003754, "grad_norm": 3.4654202461242676, "learning_rate": 3.753e-05, "loss": 6.4961, "step": 3754 }, { "epoch": 0.0003755, "grad_norm": 3.8919973373413086, "learning_rate": 3.7539999999999996e-05, "loss": 6.9141, "step": 3755 }, { "epoch": 0.0003756, "grad_norm": 3.753486394882202, "learning_rate": 3.7550000000000005e-05, "loss": 6.9258, "step": 3756 }, { "epoch": 0.0003757, "grad_norm": 3.737372398376465, "learning_rate": 3.756e-05, "loss": 7.1328, "step": 3757 }, { "epoch": 0.0003758, "grad_norm": 3.9871695041656494, "learning_rate": 3.757e-05, "loss": 7.8828, "step": 3758 }, { "epoch": 0.0003759, "grad_norm": 3.2618656158447266, "learning_rate": 3.758e-05, "loss": 6.2148, "step": 3759 }, { "epoch": 0.000376, "grad_norm": 3.848130464553833, "learning_rate": 3.759e-05, "loss": 6.9961, "step": 3760 }, { "epoch": 0.0003761, "grad_norm": 4.270238399505615, "learning_rate": 3.76e-05, "loss": 8.0078, "step": 3761 }, { "epoch": 0.0003762, "grad_norm": 3.5050365924835205, "learning_rate": 3.761e-05, "loss": 6.8828, "step": 3762 }, { "epoch": 0.0003763, "grad_norm": 4.014560222625732, "learning_rate": 3.762e-05, "loss": 7.5586, "step": 3763 }, { "epoch": 0.0003764, "grad_norm": 3.3177073001861572, "learning_rate": 3.7630000000000004e-05, "loss": 6.7422, "step": 3764 }, { "epoch": 0.0003765, "grad_norm": 3.7109742164611816, "learning_rate": 3.764e-05, "loss": 7.1953, "step": 3765 }, { "epoch": 0.0003766, "grad_norm": 4.2793378829956055, "learning_rate": 3.765e-05, "loss": 7.6797, "step": 3766 }, { "epoch": 0.0003767, "grad_norm": 3.625326633453369, "learning_rate": 3.766e-05, "loss": 7.6406, "step": 3767 }, { "epoch": 0.0003768, "grad_norm": 3.3608195781707764, "learning_rate": 3.7670000000000004e-05, "loss": 6.8828, "step": 3768 }, { "epoch": 0.0003769, "grad_norm": 3.692852020263672, "learning_rate": 3.768e-05, "loss": 6.7383, "step": 3769 }, { "epoch": 0.000377, "grad_norm": 3.648618698120117, "learning_rate": 3.769e-05, "loss": 6.9219, "step": 3770 }, { "epoch": 0.0003771, "grad_norm": 4.029127597808838, "learning_rate": 3.77e-05, "loss": 7.6523, "step": 3771 }, { "epoch": 0.0003772, "grad_norm": 4.059957981109619, "learning_rate": 3.7709999999999996e-05, "loss": 7.5938, "step": 3772 }, { "epoch": 0.0003773, "grad_norm": 4.275692939758301, "learning_rate": 3.7720000000000005e-05, "loss": 8.332, "step": 3773 }, { "epoch": 0.0003774, "grad_norm": 3.7841784954071045, "learning_rate": 3.773e-05, "loss": 7.0547, "step": 3774 }, { "epoch": 0.0003775, "grad_norm": 4.110271453857422, "learning_rate": 3.774e-05, "loss": 7.5938, "step": 3775 }, { "epoch": 0.0003776, "grad_norm": 3.4471139907836914, "learning_rate": 3.775e-05, "loss": 6.2305, "step": 3776 }, { "epoch": 0.0003777, "grad_norm": 3.4699087142944336, "learning_rate": 3.776e-05, "loss": 6.7148, "step": 3777 }, { "epoch": 0.0003778, "grad_norm": 3.6746304035186768, "learning_rate": 3.777e-05, "loss": 7.2305, "step": 3778 }, { "epoch": 0.0003779, "grad_norm": 3.8039770126342773, "learning_rate": 3.778e-05, "loss": 7.3398, "step": 3779 }, { "epoch": 0.000378, "grad_norm": 3.943490982055664, "learning_rate": 3.779e-05, "loss": 7.6328, "step": 3780 }, { "epoch": 0.0003781, "grad_norm": 3.8064470291137695, "learning_rate": 3.7800000000000004e-05, "loss": 6.9766, "step": 3781 }, { "epoch": 0.0003782, "grad_norm": 3.6602022647857666, "learning_rate": 3.781e-05, "loss": 6.5039, "step": 3782 }, { "epoch": 0.0003783, "grad_norm": 3.536530017852783, "learning_rate": 3.782e-05, "loss": 6.418, "step": 3783 }, { "epoch": 0.0003784, "grad_norm": 3.4223005771636963, "learning_rate": 3.783e-05, "loss": 7.0469, "step": 3784 }, { "epoch": 0.0003785, "grad_norm": 3.528369426727295, "learning_rate": 3.7840000000000004e-05, "loss": 6.8359, "step": 3785 }, { "epoch": 0.0003786, "grad_norm": 3.8635189533233643, "learning_rate": 3.785e-05, "loss": 7.2656, "step": 3786 }, { "epoch": 0.0003787, "grad_norm": 3.5053634643554688, "learning_rate": 3.786e-05, "loss": 6.8828, "step": 3787 }, { "epoch": 0.0003788, "grad_norm": 3.919947624206543, "learning_rate": 3.787e-05, "loss": 7.582, "step": 3788 }, { "epoch": 0.0003789, "grad_norm": 3.5167784690856934, "learning_rate": 3.7879999999999996e-05, "loss": 6.7969, "step": 3789 }, { "epoch": 0.000379, "grad_norm": 3.5805504322052, "learning_rate": 3.7890000000000005e-05, "loss": 6.7852, "step": 3790 }, { "epoch": 0.0003791, "grad_norm": 4.331769943237305, "learning_rate": 3.79e-05, "loss": 6.8008, "step": 3791 }, { "epoch": 0.0003792, "grad_norm": 3.6048316955566406, "learning_rate": 3.791e-05, "loss": 6.7656, "step": 3792 }, { "epoch": 0.0003793, "grad_norm": 4.002704620361328, "learning_rate": 3.792e-05, "loss": 6.9961, "step": 3793 }, { "epoch": 0.0003794, "grad_norm": 3.9942467212677, "learning_rate": 3.793e-05, "loss": 7.0312, "step": 3794 }, { "epoch": 0.0003795, "grad_norm": 4.518647193908691, "learning_rate": 3.794e-05, "loss": 8.1406, "step": 3795 }, { "epoch": 0.0003796, "grad_norm": 3.9255402088165283, "learning_rate": 3.795e-05, "loss": 6.8516, "step": 3796 }, { "epoch": 0.0003797, "grad_norm": 3.500156879425049, "learning_rate": 3.796e-05, "loss": 6.6719, "step": 3797 }, { "epoch": 0.0003798, "grad_norm": 3.6803736686706543, "learning_rate": 3.7970000000000004e-05, "loss": 7.0039, "step": 3798 }, { "epoch": 0.0003799, "grad_norm": 3.599980354309082, "learning_rate": 3.798e-05, "loss": 7.0938, "step": 3799 }, { "epoch": 0.00038, "grad_norm": 3.256418228149414, "learning_rate": 3.799e-05, "loss": 5.9805, "step": 3800 }, { "epoch": 0.0003801, "grad_norm": 4.027842044830322, "learning_rate": 3.8e-05, "loss": 7.625, "step": 3801 }, { "epoch": 0.0003802, "grad_norm": 3.7016587257385254, "learning_rate": 3.8010000000000004e-05, "loss": 6.4883, "step": 3802 }, { "epoch": 0.0003803, "grad_norm": 3.853823184967041, "learning_rate": 3.802e-05, "loss": 7.1367, "step": 3803 }, { "epoch": 0.0003804, "grad_norm": 4.212652683258057, "learning_rate": 3.803e-05, "loss": 8.0117, "step": 3804 }, { "epoch": 0.0003805, "grad_norm": 3.681837320327759, "learning_rate": 3.804e-05, "loss": 6.7344, "step": 3805 }, { "epoch": 0.0003806, "grad_norm": 3.545790195465088, "learning_rate": 3.8049999999999997e-05, "loss": 6.7891, "step": 3806 }, { "epoch": 0.0003807, "grad_norm": 3.54123592376709, "learning_rate": 3.8060000000000005e-05, "loss": 6.5156, "step": 3807 }, { "epoch": 0.0003808, "grad_norm": 3.562340021133423, "learning_rate": 3.807e-05, "loss": 6.5859, "step": 3808 }, { "epoch": 0.0003809, "grad_norm": 3.5242602825164795, "learning_rate": 3.808e-05, "loss": 6.5195, "step": 3809 }, { "epoch": 0.000381, "grad_norm": 5.35298490524292, "learning_rate": 3.809e-05, "loss": 7.5664, "step": 3810 }, { "epoch": 0.0003811, "grad_norm": 3.427039861679077, "learning_rate": 3.81e-05, "loss": 6.3242, "step": 3811 }, { "epoch": 0.0003812, "grad_norm": 4.479483127593994, "learning_rate": 3.811e-05, "loss": 7.4336, "step": 3812 }, { "epoch": 0.0003813, "grad_norm": 3.296142816543579, "learning_rate": 3.812e-05, "loss": 6.6641, "step": 3813 }, { "epoch": 0.0003814, "grad_norm": 4.045052528381348, "learning_rate": 3.813e-05, "loss": 7.8242, "step": 3814 }, { "epoch": 0.0003815, "grad_norm": 3.3801162242889404, "learning_rate": 3.8140000000000004e-05, "loss": 6.4258, "step": 3815 }, { "epoch": 0.0003816, "grad_norm": 3.656015157699585, "learning_rate": 3.815e-05, "loss": 6.8242, "step": 3816 }, { "epoch": 0.0003817, "grad_norm": 3.677790403366089, "learning_rate": 3.816e-05, "loss": 6.7969, "step": 3817 }, { "epoch": 0.0003818, "grad_norm": 3.7314274311065674, "learning_rate": 3.817e-05, "loss": 7.2383, "step": 3818 }, { "epoch": 0.0003819, "grad_norm": 3.598817825317383, "learning_rate": 3.8180000000000004e-05, "loss": 6.6523, "step": 3819 }, { "epoch": 0.000382, "grad_norm": 3.6216111183166504, "learning_rate": 3.819e-05, "loss": 6.7344, "step": 3820 }, { "epoch": 0.0003821, "grad_norm": 4.0292649269104, "learning_rate": 3.82e-05, "loss": 6.8438, "step": 3821 }, { "epoch": 0.0003822, "grad_norm": 3.4000163078308105, "learning_rate": 3.821e-05, "loss": 6.2148, "step": 3822 }, { "epoch": 0.0003823, "grad_norm": 3.5961146354675293, "learning_rate": 3.8219999999999997e-05, "loss": 6.6406, "step": 3823 }, { "epoch": 0.0003824, "grad_norm": 3.5228688716888428, "learning_rate": 3.8230000000000005e-05, "loss": 6.6758, "step": 3824 }, { "epoch": 0.0003825, "grad_norm": 3.7650704383850098, "learning_rate": 3.824e-05, "loss": 7.1055, "step": 3825 }, { "epoch": 0.0003826, "grad_norm": 3.543381452560425, "learning_rate": 3.825e-05, "loss": 6.418, "step": 3826 }, { "epoch": 0.0003827, "grad_norm": 3.725450277328491, "learning_rate": 3.826e-05, "loss": 6.8164, "step": 3827 }, { "epoch": 0.0003828, "grad_norm": 3.6379446983337402, "learning_rate": 3.827e-05, "loss": 7.0938, "step": 3828 }, { "epoch": 0.0003829, "grad_norm": 3.736476182937622, "learning_rate": 3.828e-05, "loss": 7.4375, "step": 3829 }, { "epoch": 0.000383, "grad_norm": 3.73931884765625, "learning_rate": 3.829e-05, "loss": 6.7578, "step": 3830 }, { "epoch": 0.0003831, "grad_norm": 4.645851135253906, "learning_rate": 3.83e-05, "loss": 7.707, "step": 3831 }, { "epoch": 0.0003832, "grad_norm": 3.6469063758850098, "learning_rate": 3.8310000000000004e-05, "loss": 6.6602, "step": 3832 }, { "epoch": 0.0003833, "grad_norm": 3.5975730419158936, "learning_rate": 3.832e-05, "loss": 6.4336, "step": 3833 }, { "epoch": 0.0003834, "grad_norm": 5.395107746124268, "learning_rate": 3.833e-05, "loss": 7.0078, "step": 3834 }, { "epoch": 0.0003835, "grad_norm": 3.5026297569274902, "learning_rate": 3.834e-05, "loss": 6.3945, "step": 3835 }, { "epoch": 0.0003836, "grad_norm": 3.5971925258636475, "learning_rate": 3.8350000000000004e-05, "loss": 6.875, "step": 3836 }, { "epoch": 0.0003837, "grad_norm": 4.300007343292236, "learning_rate": 3.836e-05, "loss": 6.6602, "step": 3837 }, { "epoch": 0.0003838, "grad_norm": 3.5784220695495605, "learning_rate": 3.837e-05, "loss": 7.1758, "step": 3838 }, { "epoch": 0.0003839, "grad_norm": 4.860476970672607, "learning_rate": 3.838e-05, "loss": 6.9375, "step": 3839 }, { "epoch": 0.000384, "grad_norm": 3.7582528591156006, "learning_rate": 3.839e-05, "loss": 7.0312, "step": 3840 }, { "epoch": 0.0003841, "grad_norm": 3.9537460803985596, "learning_rate": 3.8400000000000005e-05, "loss": 7.0234, "step": 3841 }, { "epoch": 0.0003842, "grad_norm": 3.4361069202423096, "learning_rate": 3.841e-05, "loss": 6.4141, "step": 3842 }, { "epoch": 0.0003843, "grad_norm": 3.455044984817505, "learning_rate": 3.842e-05, "loss": 6.5195, "step": 3843 }, { "epoch": 0.0003844, "grad_norm": 3.397707462310791, "learning_rate": 3.843e-05, "loss": 6.6562, "step": 3844 }, { "epoch": 0.0003845, "grad_norm": 3.3669912815093994, "learning_rate": 3.844e-05, "loss": 6.9023, "step": 3845 }, { "epoch": 0.0003846, "grad_norm": 3.3918566703796387, "learning_rate": 3.845e-05, "loss": 6.0859, "step": 3846 }, { "epoch": 0.0003847, "grad_norm": 3.8992884159088135, "learning_rate": 3.846e-05, "loss": 7.0039, "step": 3847 }, { "epoch": 0.0003848, "grad_norm": 3.3438005447387695, "learning_rate": 3.847e-05, "loss": 6.5547, "step": 3848 }, { "epoch": 0.0003849, "grad_norm": 3.6768224239349365, "learning_rate": 3.8480000000000004e-05, "loss": 6.8008, "step": 3849 }, { "epoch": 0.000385, "grad_norm": 3.576625347137451, "learning_rate": 3.849e-05, "loss": 6.6562, "step": 3850 }, { "epoch": 0.0003851, "grad_norm": 4.437202453613281, "learning_rate": 3.85e-05, "loss": 7.3828, "step": 3851 }, { "epoch": 0.0003852, "grad_norm": 3.645042896270752, "learning_rate": 3.851e-05, "loss": 6.6602, "step": 3852 }, { "epoch": 0.0003853, "grad_norm": 3.526789426803589, "learning_rate": 3.8520000000000004e-05, "loss": 6.8008, "step": 3853 }, { "epoch": 0.0003854, "grad_norm": 3.4846742153167725, "learning_rate": 3.853e-05, "loss": 6.5078, "step": 3854 }, { "epoch": 0.0003855, "grad_norm": 3.617932081222534, "learning_rate": 3.854e-05, "loss": 6.3398, "step": 3855 }, { "epoch": 0.0003856, "grad_norm": 3.5122201442718506, "learning_rate": 3.855e-05, "loss": 6.5273, "step": 3856 }, { "epoch": 0.0003857, "grad_norm": 3.6751372814178467, "learning_rate": 3.856e-05, "loss": 6.6523, "step": 3857 }, { "epoch": 0.0003858, "grad_norm": 3.6790711879730225, "learning_rate": 3.8570000000000005e-05, "loss": 6.3984, "step": 3858 }, { "epoch": 0.0003859, "grad_norm": 3.4605135917663574, "learning_rate": 3.858e-05, "loss": 6.2188, "step": 3859 }, { "epoch": 0.000386, "grad_norm": 3.6121532917022705, "learning_rate": 3.859e-05, "loss": 6.7148, "step": 3860 }, { "epoch": 0.0003861, "grad_norm": 3.7920093536376953, "learning_rate": 3.86e-05, "loss": 6.7188, "step": 3861 }, { "epoch": 0.0003862, "grad_norm": 3.7769675254821777, "learning_rate": 3.861e-05, "loss": 6.6406, "step": 3862 }, { "epoch": 0.0003863, "grad_norm": 3.5330495834350586, "learning_rate": 3.862e-05, "loss": 6.5039, "step": 3863 }, { "epoch": 0.0003864, "grad_norm": 4.07541561126709, "learning_rate": 3.863e-05, "loss": 6.1953, "step": 3864 }, { "epoch": 0.0003865, "grad_norm": 3.8019979000091553, "learning_rate": 3.864e-05, "loss": 6.793, "step": 3865 }, { "epoch": 0.0003866, "grad_norm": 3.8850960731506348, "learning_rate": 3.8650000000000004e-05, "loss": 6.418, "step": 3866 }, { "epoch": 0.0003867, "grad_norm": 4.231971263885498, "learning_rate": 3.866e-05, "loss": 6.9531, "step": 3867 }, { "epoch": 0.0003868, "grad_norm": 3.4800937175750732, "learning_rate": 3.867e-05, "loss": 6.7852, "step": 3868 }, { "epoch": 0.0003869, "grad_norm": 3.8651981353759766, "learning_rate": 3.868e-05, "loss": 7.0742, "step": 3869 }, { "epoch": 0.000387, "grad_norm": 3.3011200428009033, "learning_rate": 3.8690000000000004e-05, "loss": 6.0391, "step": 3870 }, { "epoch": 0.0003871, "grad_norm": 3.7749295234680176, "learning_rate": 3.87e-05, "loss": 6.832, "step": 3871 }, { "epoch": 0.0003872, "grad_norm": 3.520888328552246, "learning_rate": 3.871e-05, "loss": 6.4688, "step": 3872 }, { "epoch": 0.0003873, "grad_norm": 3.62060546875, "learning_rate": 3.872e-05, "loss": 6.457, "step": 3873 }, { "epoch": 0.0003874, "grad_norm": 3.7028284072875977, "learning_rate": 3.873e-05, "loss": 7.3125, "step": 3874 }, { "epoch": 0.0003875, "grad_norm": 3.725008249282837, "learning_rate": 3.8740000000000005e-05, "loss": 6.4922, "step": 3875 }, { "epoch": 0.0003876, "grad_norm": 6.641200065612793, "learning_rate": 3.875e-05, "loss": 8.5781, "step": 3876 }, { "epoch": 0.0003877, "grad_norm": 3.616450071334839, "learning_rate": 3.876e-05, "loss": 6.5586, "step": 3877 }, { "epoch": 0.0003878, "grad_norm": 3.6379799842834473, "learning_rate": 3.877e-05, "loss": 6.332, "step": 3878 }, { "epoch": 0.0003879, "grad_norm": 3.6136868000030518, "learning_rate": 3.878e-05, "loss": 6.4609, "step": 3879 }, { "epoch": 0.000388, "grad_norm": 3.5945732593536377, "learning_rate": 3.879e-05, "loss": 6.2539, "step": 3880 }, { "epoch": 0.0003881, "grad_norm": 3.6669209003448486, "learning_rate": 3.88e-05, "loss": 6.6445, "step": 3881 }, { "epoch": 0.0003882, "grad_norm": 3.744536876678467, "learning_rate": 3.881e-05, "loss": 6.7891, "step": 3882 }, { "epoch": 0.0003883, "grad_norm": 4.542664051055908, "learning_rate": 3.8820000000000004e-05, "loss": 7.2773, "step": 3883 }, { "epoch": 0.0003884, "grad_norm": 5.415156841278076, "learning_rate": 3.883e-05, "loss": 6.9258, "step": 3884 }, { "epoch": 0.0003885, "grad_norm": 3.834616184234619, "learning_rate": 3.884e-05, "loss": 6.5234, "step": 3885 }, { "epoch": 0.0003886, "grad_norm": 3.701594591140747, "learning_rate": 3.885e-05, "loss": 5.9297, "step": 3886 }, { "epoch": 0.0003887, "grad_norm": 3.660496473312378, "learning_rate": 3.8860000000000004e-05, "loss": 6.6914, "step": 3887 }, { "epoch": 0.0003888, "grad_norm": 3.613403558731079, "learning_rate": 3.887e-05, "loss": 6.4727, "step": 3888 }, { "epoch": 0.0003889, "grad_norm": 3.5814411640167236, "learning_rate": 3.888e-05, "loss": 6.4375, "step": 3889 }, { "epoch": 0.000389, "grad_norm": 3.558093309402466, "learning_rate": 3.889e-05, "loss": 6.3516, "step": 3890 }, { "epoch": 0.0003891, "grad_norm": 3.623140573501587, "learning_rate": 3.89e-05, "loss": 6.2734, "step": 3891 }, { "epoch": 0.0003892, "grad_norm": 3.5732738971710205, "learning_rate": 3.8910000000000005e-05, "loss": 6.75, "step": 3892 }, { "epoch": 0.0003893, "grad_norm": 3.700470209121704, "learning_rate": 3.892e-05, "loss": 6.5742, "step": 3893 }, { "epoch": 0.0003894, "grad_norm": 3.3083128929138184, "learning_rate": 3.893e-05, "loss": 6.2773, "step": 3894 }, { "epoch": 0.0003895, "grad_norm": 3.5228614807128906, "learning_rate": 3.894e-05, "loss": 6.3828, "step": 3895 }, { "epoch": 0.0003896, "grad_norm": 3.3080711364746094, "learning_rate": 3.895e-05, "loss": 5.6289, "step": 3896 }, { "epoch": 0.0003897, "grad_norm": 3.6175620555877686, "learning_rate": 3.896e-05, "loss": 6.8008, "step": 3897 }, { "epoch": 0.0003898, "grad_norm": 3.306755304336548, "learning_rate": 3.897e-05, "loss": 5.8359, "step": 3898 }, { "epoch": 0.0003899, "grad_norm": 4.066989421844482, "learning_rate": 3.898e-05, "loss": 6.7109, "step": 3899 }, { "epoch": 0.00039, "grad_norm": 4.122598171234131, "learning_rate": 3.8990000000000004e-05, "loss": 6.6562, "step": 3900 }, { "epoch": 0.0003901, "grad_norm": 3.791177988052368, "learning_rate": 3.9e-05, "loss": 6.7891, "step": 3901 }, { "epoch": 0.0003902, "grad_norm": 3.4802207946777344, "learning_rate": 3.901e-05, "loss": 6.3867, "step": 3902 }, { "epoch": 0.0003903, "grad_norm": 4.192112445831299, "learning_rate": 3.902e-05, "loss": 7.3203, "step": 3903 }, { "epoch": 0.0003904, "grad_norm": 3.5074985027313232, "learning_rate": 3.9030000000000004e-05, "loss": 6.0859, "step": 3904 }, { "epoch": 0.0003905, "grad_norm": 4.217540264129639, "learning_rate": 3.904e-05, "loss": 7.4844, "step": 3905 }, { "epoch": 0.0003906, "grad_norm": 3.637507438659668, "learning_rate": 3.905e-05, "loss": 6.2305, "step": 3906 }, { "epoch": 0.0003907, "grad_norm": 3.528379201889038, "learning_rate": 3.906e-05, "loss": 6.4961, "step": 3907 }, { "epoch": 0.0003908, "grad_norm": 4.091883182525635, "learning_rate": 3.907e-05, "loss": 6.5898, "step": 3908 }, { "epoch": 0.0003909, "grad_norm": 5.288835525512695, "learning_rate": 3.9080000000000005e-05, "loss": 6.793, "step": 3909 }, { "epoch": 0.000391, "grad_norm": 5.368162155151367, "learning_rate": 3.909e-05, "loss": 6.4297, "step": 3910 }, { "epoch": 0.0003911, "grad_norm": 3.908277750015259, "learning_rate": 3.91e-05, "loss": 6.543, "step": 3911 }, { "epoch": 0.0003912, "grad_norm": 4.534628868103027, "learning_rate": 3.911e-05, "loss": 7.3555, "step": 3912 }, { "epoch": 0.0003913, "grad_norm": 3.897803783416748, "learning_rate": 3.912e-05, "loss": 6.4297, "step": 3913 }, { "epoch": 0.0003914, "grad_norm": 3.931011438369751, "learning_rate": 3.913e-05, "loss": 6.9062, "step": 3914 }, { "epoch": 0.0003915, "grad_norm": 3.8343608379364014, "learning_rate": 3.914e-05, "loss": 6.9219, "step": 3915 }, { "epoch": 0.0003916, "grad_norm": 3.6591923236846924, "learning_rate": 3.915e-05, "loss": 6.3047, "step": 3916 }, { "epoch": 0.0003917, "grad_norm": 3.6464710235595703, "learning_rate": 3.9160000000000005e-05, "loss": 6.4453, "step": 3917 }, { "epoch": 0.0003918, "grad_norm": 3.662614345550537, "learning_rate": 3.917e-05, "loss": 6.3555, "step": 3918 }, { "epoch": 0.0003919, "grad_norm": 4.47059440612793, "learning_rate": 3.918e-05, "loss": 6.3711, "step": 3919 }, { "epoch": 0.000392, "grad_norm": 3.849656820297241, "learning_rate": 3.919e-05, "loss": 6.3633, "step": 3920 }, { "epoch": 0.0003921, "grad_norm": 3.39892315864563, "learning_rate": 3.9200000000000004e-05, "loss": 6.043, "step": 3921 }, { "epoch": 0.0003922, "grad_norm": 3.5665454864501953, "learning_rate": 3.921e-05, "loss": 6.5117, "step": 3922 }, { "epoch": 0.0003923, "grad_norm": 4.4894866943359375, "learning_rate": 3.922e-05, "loss": 7.5391, "step": 3923 }, { "epoch": 0.0003924, "grad_norm": 3.778273582458496, "learning_rate": 3.923e-05, "loss": 6.4336, "step": 3924 }, { "epoch": 0.0003925, "grad_norm": 4.10695219039917, "learning_rate": 3.924e-05, "loss": 7.3477, "step": 3925 }, { "epoch": 0.0003926, "grad_norm": 3.689099073410034, "learning_rate": 3.9250000000000005e-05, "loss": 6.4766, "step": 3926 }, { "epoch": 0.0003927, "grad_norm": 4.039351940155029, "learning_rate": 3.926e-05, "loss": 6.9531, "step": 3927 }, { "epoch": 0.0003928, "grad_norm": 3.9846694469451904, "learning_rate": 3.927e-05, "loss": 6.9336, "step": 3928 }, { "epoch": 0.0003929, "grad_norm": 4.126287937164307, "learning_rate": 3.9280000000000003e-05, "loss": 6.957, "step": 3929 }, { "epoch": 0.000393, "grad_norm": 3.5479607582092285, "learning_rate": 3.929e-05, "loss": 6.1172, "step": 3930 }, { "epoch": 0.0003931, "grad_norm": 3.5273523330688477, "learning_rate": 3.93e-05, "loss": 6.4062, "step": 3931 }, { "epoch": 0.0003932, "grad_norm": 3.694324254989624, "learning_rate": 3.931e-05, "loss": 6.7773, "step": 3932 }, { "epoch": 0.0003933, "grad_norm": 3.7769875526428223, "learning_rate": 3.932e-05, "loss": 6.4961, "step": 3933 }, { "epoch": 0.0003934, "grad_norm": 3.569777250289917, "learning_rate": 3.933e-05, "loss": 5.9336, "step": 3934 }, { "epoch": 0.0003935, "grad_norm": 4.053563117980957, "learning_rate": 3.934e-05, "loss": 6.5859, "step": 3935 }, { "epoch": 0.0003936, "grad_norm": 3.7891006469726562, "learning_rate": 3.935e-05, "loss": 6.1953, "step": 3936 }, { "epoch": 0.0003937, "grad_norm": 3.6255874633789062, "learning_rate": 3.936e-05, "loss": 6.0898, "step": 3937 }, { "epoch": 0.0003938, "grad_norm": 4.6294846534729, "learning_rate": 3.9370000000000004e-05, "loss": 6.2812, "step": 3938 }, { "epoch": 0.0003939, "grad_norm": 3.491760015487671, "learning_rate": 3.938e-05, "loss": 5.8203, "step": 3939 }, { "epoch": 0.000394, "grad_norm": 4.210688591003418, "learning_rate": 3.939e-05, "loss": 7.3477, "step": 3940 }, { "epoch": 0.0003941, "grad_norm": 3.9888815879821777, "learning_rate": 3.94e-05, "loss": 6.6484, "step": 3941 }, { "epoch": 0.0003942, "grad_norm": 3.603769063949585, "learning_rate": 3.941e-05, "loss": 6.0898, "step": 3942 }, { "epoch": 0.0003943, "grad_norm": 3.480084180831909, "learning_rate": 3.9420000000000005e-05, "loss": 6.3047, "step": 3943 }, { "epoch": 0.0003944, "grad_norm": 3.588226079940796, "learning_rate": 3.943e-05, "loss": 6.3516, "step": 3944 }, { "epoch": 0.0003945, "grad_norm": 3.6377761363983154, "learning_rate": 3.9439999999999995e-05, "loss": 6.7148, "step": 3945 }, { "epoch": 0.0003946, "grad_norm": 3.512343168258667, "learning_rate": 3.9450000000000003e-05, "loss": 5.957, "step": 3946 }, { "epoch": 0.0003947, "grad_norm": 4.237716197967529, "learning_rate": 3.946e-05, "loss": 7.0469, "step": 3947 }, { "epoch": 0.0003948, "grad_norm": 3.8926217555999756, "learning_rate": 3.947e-05, "loss": 6.9766, "step": 3948 }, { "epoch": 0.0003949, "grad_norm": 4.118444442749023, "learning_rate": 3.948e-05, "loss": 7.2109, "step": 3949 }, { "epoch": 0.000395, "grad_norm": 3.6938817501068115, "learning_rate": 3.949e-05, "loss": 6.3984, "step": 3950 }, { "epoch": 0.0003951, "grad_norm": 3.470550298690796, "learning_rate": 3.95e-05, "loss": 6.3555, "step": 3951 }, { "epoch": 0.0003952, "grad_norm": 3.5763981342315674, "learning_rate": 3.951e-05, "loss": 5.793, "step": 3952 }, { "epoch": 0.0003953, "grad_norm": 3.9389219284057617, "learning_rate": 3.952e-05, "loss": 6.7656, "step": 3953 }, { "epoch": 0.0003954, "grad_norm": 3.983100652694702, "learning_rate": 3.953e-05, "loss": 6.8477, "step": 3954 }, { "epoch": 0.0003955, "grad_norm": 3.5146944522857666, "learning_rate": 3.9540000000000004e-05, "loss": 6.0, "step": 3955 }, { "epoch": 0.0003956, "grad_norm": 3.8866071701049805, "learning_rate": 3.955e-05, "loss": 6.3125, "step": 3956 }, { "epoch": 0.0003957, "grad_norm": 3.6697378158569336, "learning_rate": 3.956e-05, "loss": 6.8594, "step": 3957 }, { "epoch": 0.0003958, "grad_norm": 3.4364497661590576, "learning_rate": 3.957e-05, "loss": 5.8008, "step": 3958 }, { "epoch": 0.0003959, "grad_norm": 3.478041648864746, "learning_rate": 3.958e-05, "loss": 5.9688, "step": 3959 }, { "epoch": 0.000396, "grad_norm": 3.8149983882904053, "learning_rate": 3.9590000000000006e-05, "loss": 5.8555, "step": 3960 }, { "epoch": 0.0003961, "grad_norm": 4.335508346557617, "learning_rate": 3.96e-05, "loss": 7.0, "step": 3961 }, { "epoch": 0.0003962, "grad_norm": 3.5021438598632812, "learning_rate": 3.9609999999999995e-05, "loss": 5.9531, "step": 3962 }, { "epoch": 0.0003963, "grad_norm": 3.6077911853790283, "learning_rate": 3.9620000000000004e-05, "loss": 6.2617, "step": 3963 }, { "epoch": 0.0003964, "grad_norm": 3.7341020107269287, "learning_rate": 3.963e-05, "loss": 6.332, "step": 3964 }, { "epoch": 0.0003965, "grad_norm": 3.5670576095581055, "learning_rate": 3.964e-05, "loss": 6.2812, "step": 3965 }, { "epoch": 0.0003966, "grad_norm": 3.5666401386260986, "learning_rate": 3.965e-05, "loss": 6.0273, "step": 3966 }, { "epoch": 0.0003967, "grad_norm": 3.370593547821045, "learning_rate": 3.966e-05, "loss": 6.3438, "step": 3967 }, { "epoch": 0.0003968, "grad_norm": 3.748072385787964, "learning_rate": 3.967e-05, "loss": 6.125, "step": 3968 }, { "epoch": 0.0003969, "grad_norm": 4.9421706199646, "learning_rate": 3.968e-05, "loss": 6.8789, "step": 3969 }, { "epoch": 0.000397, "grad_norm": 3.677942991256714, "learning_rate": 3.969e-05, "loss": 6.1562, "step": 3970 }, { "epoch": 0.0003971, "grad_norm": 3.7501626014709473, "learning_rate": 3.97e-05, "loss": 5.8984, "step": 3971 }, { "epoch": 0.0003972, "grad_norm": 3.939824104309082, "learning_rate": 3.9710000000000004e-05, "loss": 6.3633, "step": 3972 }, { "epoch": 0.0003973, "grad_norm": 3.758056402206421, "learning_rate": 3.972e-05, "loss": 6.1094, "step": 3973 }, { "epoch": 0.0003974, "grad_norm": 3.7101430892944336, "learning_rate": 3.973e-05, "loss": 6.4844, "step": 3974 }, { "epoch": 0.0003975, "grad_norm": 4.147250175476074, "learning_rate": 3.974e-05, "loss": 6.8359, "step": 3975 }, { "epoch": 0.0003976, "grad_norm": 3.4404964447021484, "learning_rate": 3.975e-05, "loss": 5.8203, "step": 3976 }, { "epoch": 0.0003977, "grad_norm": 3.9862027168273926, "learning_rate": 3.9760000000000006e-05, "loss": 6.2734, "step": 3977 }, { "epoch": 0.0003978, "grad_norm": 3.520744800567627, "learning_rate": 3.977e-05, "loss": 5.9805, "step": 3978 }, { "epoch": 0.0003979, "grad_norm": 4.299718856811523, "learning_rate": 3.9779999999999995e-05, "loss": 7.332, "step": 3979 }, { "epoch": 0.000398, "grad_norm": 3.7504000663757324, "learning_rate": 3.9790000000000004e-05, "loss": 6.4219, "step": 3980 }, { "epoch": 0.0003981, "grad_norm": 3.5925910472869873, "learning_rate": 3.98e-05, "loss": 6.5078, "step": 3981 }, { "epoch": 0.0003982, "grad_norm": 3.3286728858947754, "learning_rate": 3.981e-05, "loss": 5.8594, "step": 3982 }, { "epoch": 0.0003983, "grad_norm": 3.499295711517334, "learning_rate": 3.982e-05, "loss": 5.9961, "step": 3983 }, { "epoch": 0.0003984, "grad_norm": 3.8168468475341797, "learning_rate": 3.983e-05, "loss": 6.3516, "step": 3984 }, { "epoch": 0.0003985, "grad_norm": 3.6319031715393066, "learning_rate": 3.984e-05, "loss": 6.0977, "step": 3985 }, { "epoch": 0.0003986, "grad_norm": 4.193058013916016, "learning_rate": 3.985e-05, "loss": 7.4336, "step": 3986 }, { "epoch": 0.0003987, "grad_norm": 4.188149929046631, "learning_rate": 3.986e-05, "loss": 6.2773, "step": 3987 }, { "epoch": 0.0003988, "grad_norm": 3.7367634773254395, "learning_rate": 3.987e-05, "loss": 6.25, "step": 3988 }, { "epoch": 0.0003989, "grad_norm": 3.5641446113586426, "learning_rate": 3.9880000000000004e-05, "loss": 6.1211, "step": 3989 }, { "epoch": 0.000399, "grad_norm": 4.483705997467041, "learning_rate": 3.989e-05, "loss": 7.0742, "step": 3990 }, { "epoch": 0.0003991, "grad_norm": 3.4737515449523926, "learning_rate": 3.99e-05, "loss": 5.5312, "step": 3991 }, { "epoch": 0.0003992, "grad_norm": 4.215260028839111, "learning_rate": 3.991e-05, "loss": 6.1055, "step": 3992 }, { "epoch": 0.0003993, "grad_norm": 4.5237298011779785, "learning_rate": 3.992e-05, "loss": 7.332, "step": 3993 }, { "epoch": 0.0003994, "grad_norm": 3.8761157989501953, "learning_rate": 3.9930000000000006e-05, "loss": 6.2344, "step": 3994 }, { "epoch": 0.0003995, "grad_norm": 4.130284786224365, "learning_rate": 3.994e-05, "loss": 6.4609, "step": 3995 }, { "epoch": 0.0003996, "grad_norm": 3.541053533554077, "learning_rate": 3.9949999999999995e-05, "loss": 6.0195, "step": 3996 }, { "epoch": 0.0003997, "grad_norm": 4.248941898345947, "learning_rate": 3.9960000000000004e-05, "loss": 7.0977, "step": 3997 }, { "epoch": 0.0003998, "grad_norm": 4.139227867126465, "learning_rate": 3.997e-05, "loss": 6.957, "step": 3998 }, { "epoch": 0.0003999, "grad_norm": 3.674184799194336, "learning_rate": 3.998e-05, "loss": 6.2812, "step": 3999 }, { "epoch": 0.0004, "grad_norm": 3.7404983043670654, "learning_rate": 3.999e-05, "loss": 6.0078, "step": 4000 }, { "epoch": 0.0004, "eval_loss": 0.6988784074783325, "eval_runtime": 366.4361, "eval_samples_per_second": 27.29, "eval_steps_per_second": 1.706, "step": 4000 }, { "epoch": 0.0004001, "grad_norm": 3.4606375694274902, "learning_rate": 4e-05, "loss": 5.4453, "step": 4001 }, { "epoch": 0.0004002, "grad_norm": 3.826916456222534, "learning_rate": 4.001e-05, "loss": 6.2656, "step": 4002 }, { "epoch": 0.0004003, "grad_norm": 4.114051342010498, "learning_rate": 4.002e-05, "loss": 6.375, "step": 4003 }, { "epoch": 0.0004004, "grad_norm": 3.7272844314575195, "learning_rate": 4.003e-05, "loss": 6.3867, "step": 4004 }, { "epoch": 0.0004005, "grad_norm": 3.5052099227905273, "learning_rate": 4.004e-05, "loss": 6.0234, "step": 4005 }, { "epoch": 0.0004006, "grad_norm": 4.031330108642578, "learning_rate": 4.0050000000000004e-05, "loss": 7.0039, "step": 4006 }, { "epoch": 0.0004007, "grad_norm": 3.630326986312866, "learning_rate": 4.006e-05, "loss": 5.9414, "step": 4007 }, { "epoch": 0.0004008, "grad_norm": 4.503011226654053, "learning_rate": 4.007e-05, "loss": 6.6836, "step": 4008 }, { "epoch": 0.0004009, "grad_norm": 3.8347249031066895, "learning_rate": 4.008e-05, "loss": 6.1992, "step": 4009 }, { "epoch": 0.000401, "grad_norm": 3.586697578430176, "learning_rate": 4.009e-05, "loss": 5.7852, "step": 4010 }, { "epoch": 0.0004011, "grad_norm": 3.870250940322876, "learning_rate": 4.0100000000000006e-05, "loss": 6.5156, "step": 4011 }, { "epoch": 0.0004012, "grad_norm": 3.56308913230896, "learning_rate": 4.011e-05, "loss": 5.8672, "step": 4012 }, { "epoch": 0.0004013, "grad_norm": 3.5774266719818115, "learning_rate": 4.0119999999999995e-05, "loss": 5.8555, "step": 4013 }, { "epoch": 0.0004014, "grad_norm": 3.5455245971679688, "learning_rate": 4.0130000000000004e-05, "loss": 5.8594, "step": 4014 }, { "epoch": 0.0004015, "grad_norm": 3.949404239654541, "learning_rate": 4.014e-05, "loss": 6.2656, "step": 4015 }, { "epoch": 0.0004016, "grad_norm": 4.10886812210083, "learning_rate": 4.015e-05, "loss": 6.1797, "step": 4016 }, { "epoch": 0.0004017, "grad_norm": 3.6920166015625, "learning_rate": 4.016e-05, "loss": 6.1211, "step": 4017 }, { "epoch": 0.0004018, "grad_norm": 3.5876052379608154, "learning_rate": 4.017e-05, "loss": 5.8398, "step": 4018 }, { "epoch": 0.0004019, "grad_norm": 3.3090133666992188, "learning_rate": 4.018e-05, "loss": 5.3516, "step": 4019 }, { "epoch": 0.000402, "grad_norm": 3.4610753059387207, "learning_rate": 4.019e-05, "loss": 5.6875, "step": 4020 }, { "epoch": 0.0004021, "grad_norm": 3.2732362747192383, "learning_rate": 4.02e-05, "loss": 5.5156, "step": 4021 }, { "epoch": 0.0004022, "grad_norm": 4.044480323791504, "learning_rate": 4.021e-05, "loss": 6.4688, "step": 4022 }, { "epoch": 0.0004023, "grad_norm": 3.8909099102020264, "learning_rate": 4.0220000000000005e-05, "loss": 6.2695, "step": 4023 }, { "epoch": 0.0004024, "grad_norm": 3.6923539638519287, "learning_rate": 4.023e-05, "loss": 6.0938, "step": 4024 }, { "epoch": 0.0004025, "grad_norm": 3.8731579780578613, "learning_rate": 4.024e-05, "loss": 6.0664, "step": 4025 }, { "epoch": 0.0004026, "grad_norm": 3.885798215866089, "learning_rate": 4.025e-05, "loss": 6.6523, "step": 4026 }, { "epoch": 0.0004027, "grad_norm": 4.301191329956055, "learning_rate": 4.026e-05, "loss": 6.1953, "step": 4027 }, { "epoch": 0.0004028, "grad_norm": 3.548969268798828, "learning_rate": 4.0270000000000006e-05, "loss": 5.6875, "step": 4028 }, { "epoch": 0.0004029, "grad_norm": 3.5199947357177734, "learning_rate": 4.028e-05, "loss": 5.6211, "step": 4029 }, { "epoch": 0.000403, "grad_norm": 3.62239146232605, "learning_rate": 4.0289999999999995e-05, "loss": 6.0273, "step": 4030 }, { "epoch": 0.0004031, "grad_norm": 3.4586358070373535, "learning_rate": 4.0300000000000004e-05, "loss": 5.9414, "step": 4031 }, { "epoch": 0.0004032, "grad_norm": 3.722956657409668, "learning_rate": 4.031e-05, "loss": 6.0117, "step": 4032 }, { "epoch": 0.0004033, "grad_norm": 6.033339977264404, "learning_rate": 4.032e-05, "loss": 6.918, "step": 4033 }, { "epoch": 0.0004034, "grad_norm": 4.057127952575684, "learning_rate": 4.033e-05, "loss": 6.9023, "step": 4034 }, { "epoch": 0.0004035, "grad_norm": 4.026291370391846, "learning_rate": 4.0340000000000003e-05, "loss": 6.3789, "step": 4035 }, { "epoch": 0.0004036, "grad_norm": 3.491253614425659, "learning_rate": 4.035e-05, "loss": 5.707, "step": 4036 }, { "epoch": 0.0004037, "grad_norm": 3.1500136852264404, "learning_rate": 4.036e-05, "loss": 5.4453, "step": 4037 }, { "epoch": 0.0004038, "grad_norm": 3.66682767868042, "learning_rate": 4.037e-05, "loss": 5.8516, "step": 4038 }, { "epoch": 0.0004039, "grad_norm": 3.6704399585723877, "learning_rate": 4.038e-05, "loss": 6.3516, "step": 4039 }, { "epoch": 0.000404, "grad_norm": 3.5657052993774414, "learning_rate": 4.0390000000000005e-05, "loss": 5.375, "step": 4040 }, { "epoch": 0.0004041, "grad_norm": 3.575443744659424, "learning_rate": 4.04e-05, "loss": 6.4609, "step": 4041 }, { "epoch": 0.0004042, "grad_norm": 4.032397747039795, "learning_rate": 4.041e-05, "loss": 6.1172, "step": 4042 }, { "epoch": 0.0004043, "grad_norm": 3.8276188373565674, "learning_rate": 4.042e-05, "loss": 5.4609, "step": 4043 }, { "epoch": 0.0004044, "grad_norm": 3.7183375358581543, "learning_rate": 4.043e-05, "loss": 6.1367, "step": 4044 }, { "epoch": 0.0004045, "grad_norm": 3.5428755283355713, "learning_rate": 4.0440000000000006e-05, "loss": 6.1992, "step": 4045 }, { "epoch": 0.0004046, "grad_norm": 3.3482773303985596, "learning_rate": 4.045e-05, "loss": 5.3555, "step": 4046 }, { "epoch": 0.0004047, "grad_norm": 3.9590556621551514, "learning_rate": 4.0459999999999995e-05, "loss": 6.7812, "step": 4047 }, { "epoch": 0.0004048, "grad_norm": 3.4997849464416504, "learning_rate": 4.0470000000000004e-05, "loss": 5.5781, "step": 4048 }, { "epoch": 0.0004049, "grad_norm": 4.064661979675293, "learning_rate": 4.048e-05, "loss": 6.3477, "step": 4049 }, { "epoch": 0.000405, "grad_norm": 4.637206554412842, "learning_rate": 4.049e-05, "loss": 6.7578, "step": 4050 }, { "epoch": 0.0004051, "grad_norm": 3.969869613647461, "learning_rate": 4.05e-05, "loss": 6.3711, "step": 4051 }, { "epoch": 0.0004052, "grad_norm": 3.8579952716827393, "learning_rate": 4.051e-05, "loss": 6.6172, "step": 4052 }, { "epoch": 0.0004053, "grad_norm": 4.373425006866455, "learning_rate": 4.052e-05, "loss": 6.2539, "step": 4053 }, { "epoch": 0.0004054, "grad_norm": 4.024954319000244, "learning_rate": 4.053e-05, "loss": 6.0469, "step": 4054 }, { "epoch": 0.0004055, "grad_norm": 3.6725335121154785, "learning_rate": 4.054e-05, "loss": 5.8125, "step": 4055 }, { "epoch": 0.0004056, "grad_norm": 4.168887615203857, "learning_rate": 4.055e-05, "loss": 6.4102, "step": 4056 }, { "epoch": 0.0004057, "grad_norm": 3.6008541584014893, "learning_rate": 4.0560000000000005e-05, "loss": 5.5742, "step": 4057 }, { "epoch": 0.0004058, "grad_norm": 3.704138994216919, "learning_rate": 4.057e-05, "loss": 5.875, "step": 4058 }, { "epoch": 0.0004059, "grad_norm": 4.085643768310547, "learning_rate": 4.058e-05, "loss": 6.2422, "step": 4059 }, { "epoch": 0.000406, "grad_norm": 3.36867618560791, "learning_rate": 4.059e-05, "loss": 5.3906, "step": 4060 }, { "epoch": 0.0004061, "grad_norm": 3.8944873809814453, "learning_rate": 4.06e-05, "loss": 6.0938, "step": 4061 }, { "epoch": 0.0004062, "grad_norm": 3.608980655670166, "learning_rate": 4.0610000000000006e-05, "loss": 5.5898, "step": 4062 }, { "epoch": 0.0004063, "grad_norm": 3.9158482551574707, "learning_rate": 4.062e-05, "loss": 5.7695, "step": 4063 }, { "epoch": 0.0004064, "grad_norm": 3.6970741748809814, "learning_rate": 4.0629999999999996e-05, "loss": 5.8789, "step": 4064 }, { "epoch": 0.0004065, "grad_norm": 3.2966415882110596, "learning_rate": 4.0640000000000004e-05, "loss": 5.5195, "step": 4065 }, { "epoch": 0.0004066, "grad_norm": 3.35260272026062, "learning_rate": 4.065e-05, "loss": 5.4414, "step": 4066 }, { "epoch": 0.0004067, "grad_norm": 3.597792387008667, "learning_rate": 4.066e-05, "loss": 5.6367, "step": 4067 }, { "epoch": 0.0004068, "grad_norm": 3.889126777648926, "learning_rate": 4.067e-05, "loss": 5.9102, "step": 4068 }, { "epoch": 0.0004069, "grad_norm": 3.6573078632354736, "learning_rate": 4.068e-05, "loss": 5.6211, "step": 4069 }, { "epoch": 0.000407, "grad_norm": 3.823882579803467, "learning_rate": 4.069e-05, "loss": 5.9023, "step": 4070 }, { "epoch": 0.0004071, "grad_norm": 3.091827630996704, "learning_rate": 4.07e-05, "loss": 5.1289, "step": 4071 }, { "epoch": 0.0004072, "grad_norm": 3.5181682109832764, "learning_rate": 4.071e-05, "loss": 5.2227, "step": 4072 }, { "epoch": 0.0004073, "grad_norm": 3.8732082843780518, "learning_rate": 4.072e-05, "loss": 6.3359, "step": 4073 }, { "epoch": 0.0004074, "grad_norm": 4.9277262687683105, "learning_rate": 4.0730000000000005e-05, "loss": 6.6094, "step": 4074 }, { "epoch": 0.0004075, "grad_norm": 3.7823238372802734, "learning_rate": 4.074e-05, "loss": 5.8633, "step": 4075 }, { "epoch": 0.0004076, "grad_norm": 5.305178165435791, "learning_rate": 4.075e-05, "loss": 6.3125, "step": 4076 }, { "epoch": 0.0004077, "grad_norm": 3.9796903133392334, "learning_rate": 4.076e-05, "loss": 6.3711, "step": 4077 }, { "epoch": 0.0004078, "grad_norm": 3.5141024589538574, "learning_rate": 4.077e-05, "loss": 5.4727, "step": 4078 }, { "epoch": 0.0004079, "grad_norm": 3.2786922454833984, "learning_rate": 4.0780000000000006e-05, "loss": 5.1484, "step": 4079 }, { "epoch": 0.000408, "grad_norm": 4.347360610961914, "learning_rate": 4.079e-05, "loss": 6.168, "step": 4080 }, { "epoch": 0.0004081, "grad_norm": 3.3764026165008545, "learning_rate": 4.0799999999999996e-05, "loss": 5.6055, "step": 4081 }, { "epoch": 0.0004082, "grad_norm": 3.5384933948516846, "learning_rate": 4.0810000000000004e-05, "loss": 6.0273, "step": 4082 }, { "epoch": 0.0004083, "grad_norm": 4.531378746032715, "learning_rate": 4.082e-05, "loss": 7.0195, "step": 4083 }, { "epoch": 0.0004084, "grad_norm": 3.650240421295166, "learning_rate": 4.083e-05, "loss": 5.5234, "step": 4084 }, { "epoch": 0.0004085, "grad_norm": 3.879042625427246, "learning_rate": 4.084e-05, "loss": 6.0508, "step": 4085 }, { "epoch": 0.0004086, "grad_norm": 3.5399930477142334, "learning_rate": 4.085e-05, "loss": 5.7852, "step": 4086 }, { "epoch": 0.0004087, "grad_norm": 3.6029984951019287, "learning_rate": 4.086e-05, "loss": 6.0273, "step": 4087 }, { "epoch": 0.0004088, "grad_norm": 3.733485221862793, "learning_rate": 4.087e-05, "loss": 5.3555, "step": 4088 }, { "epoch": 0.0004089, "grad_norm": 3.5953643321990967, "learning_rate": 4.088e-05, "loss": 5.5781, "step": 4089 }, { "epoch": 0.000409, "grad_norm": 3.587406873703003, "learning_rate": 4.089e-05, "loss": 5.6875, "step": 4090 }, { "epoch": 0.0004091, "grad_norm": 3.394418478012085, "learning_rate": 4.0900000000000005e-05, "loss": 5.4453, "step": 4091 }, { "epoch": 0.0004092, "grad_norm": 3.9767611026763916, "learning_rate": 4.091e-05, "loss": 5.6758, "step": 4092 }, { "epoch": 0.0004093, "grad_norm": 3.677628517150879, "learning_rate": 4.092e-05, "loss": 5.5391, "step": 4093 }, { "epoch": 0.0004094, "grad_norm": 3.4986319541931152, "learning_rate": 4.093e-05, "loss": 5.8711, "step": 4094 }, { "epoch": 0.0004095, "grad_norm": 3.5574800968170166, "learning_rate": 4.094e-05, "loss": 5.8086, "step": 4095 }, { "epoch": 0.0004096, "grad_norm": 3.4285664558410645, "learning_rate": 4.0950000000000006e-05, "loss": 5.4648, "step": 4096 }, { "epoch": 0.0004097, "grad_norm": 3.44537091255188, "learning_rate": 4.096e-05, "loss": 5.3164, "step": 4097 }, { "epoch": 0.0004098, "grad_norm": 3.58638334274292, "learning_rate": 4.0969999999999996e-05, "loss": 5.3867, "step": 4098 }, { "epoch": 0.0004099, "grad_norm": 3.753119468688965, "learning_rate": 4.0980000000000004e-05, "loss": 5.6875, "step": 4099 }, { "epoch": 0.00041, "grad_norm": 3.629337787628174, "learning_rate": 4.099e-05, "loss": 5.6094, "step": 4100 }, { "epoch": 0.0004101, "grad_norm": 3.579216241836548, "learning_rate": 4.1e-05, "loss": 5.4531, "step": 4101 }, { "epoch": 0.0004102, "grad_norm": 3.4398245811462402, "learning_rate": 4.101e-05, "loss": 5.2773, "step": 4102 }, { "epoch": 0.0004103, "grad_norm": 3.3174123764038086, "learning_rate": 4.102e-05, "loss": 5.1211, "step": 4103 }, { "epoch": 0.0004104, "grad_norm": 3.9185245037078857, "learning_rate": 4.103e-05, "loss": 5.7891, "step": 4104 }, { "epoch": 0.0004105, "grad_norm": 3.6777477264404297, "learning_rate": 4.104e-05, "loss": 6.0469, "step": 4105 }, { "epoch": 0.0004106, "grad_norm": 3.459282398223877, "learning_rate": 4.105e-05, "loss": 5.4258, "step": 4106 }, { "epoch": 0.0004107, "grad_norm": 3.7453536987304688, "learning_rate": 4.106e-05, "loss": 5.6797, "step": 4107 }, { "epoch": 0.0004108, "grad_norm": 3.522678852081299, "learning_rate": 4.1070000000000005e-05, "loss": 5.4805, "step": 4108 }, { "epoch": 0.0004109, "grad_norm": 3.6792984008789062, "learning_rate": 4.108e-05, "loss": 5.5938, "step": 4109 }, { "epoch": 0.000411, "grad_norm": 3.6462788581848145, "learning_rate": 4.109e-05, "loss": 5.6094, "step": 4110 }, { "epoch": 0.0004111, "grad_norm": 3.2714645862579346, "learning_rate": 4.11e-05, "loss": 5.207, "step": 4111 }, { "epoch": 0.0004112, "grad_norm": 3.801706314086914, "learning_rate": 4.111e-05, "loss": 6.1211, "step": 4112 }, { "epoch": 0.0004113, "grad_norm": 3.418394088745117, "learning_rate": 4.1120000000000006e-05, "loss": 4.9414, "step": 4113 }, { "epoch": 0.0004114, "grad_norm": 3.5116078853607178, "learning_rate": 4.113e-05, "loss": 5.0938, "step": 4114 }, { "epoch": 0.0004115, "grad_norm": 3.7004990577697754, "learning_rate": 4.1139999999999996e-05, "loss": 5.9648, "step": 4115 }, { "epoch": 0.0004116, "grad_norm": 3.769848585128784, "learning_rate": 4.1150000000000004e-05, "loss": 5.6055, "step": 4116 }, { "epoch": 0.0004117, "grad_norm": 4.09637975692749, "learning_rate": 4.116e-05, "loss": 6.0625, "step": 4117 }, { "epoch": 0.0004118, "grad_norm": 3.4790797233581543, "learning_rate": 4.117e-05, "loss": 5.5078, "step": 4118 }, { "epoch": 0.0004119, "grad_norm": 3.6491708755493164, "learning_rate": 4.118e-05, "loss": 5.1797, "step": 4119 }, { "epoch": 0.000412, "grad_norm": 3.7839958667755127, "learning_rate": 4.119e-05, "loss": 5.5078, "step": 4120 }, { "epoch": 0.0004121, "grad_norm": 3.9092719554901123, "learning_rate": 4.12e-05, "loss": 5.6562, "step": 4121 }, { "epoch": 0.0004122, "grad_norm": 3.3346238136291504, "learning_rate": 4.121e-05, "loss": 5.4102, "step": 4122 }, { "epoch": 0.0004123, "grad_norm": 4.38975715637207, "learning_rate": 4.122e-05, "loss": 5.8633, "step": 4123 }, { "epoch": 0.0004124, "grad_norm": 3.710336685180664, "learning_rate": 4.123e-05, "loss": 5.4688, "step": 4124 }, { "epoch": 0.0004125, "grad_norm": 3.6123082637786865, "learning_rate": 4.1240000000000005e-05, "loss": 5.1562, "step": 4125 }, { "epoch": 0.0004126, "grad_norm": 3.5171592235565186, "learning_rate": 4.125e-05, "loss": 5.793, "step": 4126 }, { "epoch": 0.0004127, "grad_norm": 3.6877355575561523, "learning_rate": 4.126e-05, "loss": 5.4883, "step": 4127 }, { "epoch": 0.0004128, "grad_norm": 4.458221435546875, "learning_rate": 4.127e-05, "loss": 6.2422, "step": 4128 }, { "epoch": 0.0004129, "grad_norm": 3.854968547821045, "learning_rate": 4.128e-05, "loss": 5.3594, "step": 4129 }, { "epoch": 0.000413, "grad_norm": 4.326519966125488, "learning_rate": 4.1290000000000006e-05, "loss": 6.1172, "step": 4130 }, { "epoch": 0.0004131, "grad_norm": 3.3446953296661377, "learning_rate": 4.13e-05, "loss": 5.168, "step": 4131 }, { "epoch": 0.0004132, "grad_norm": 3.5881917476654053, "learning_rate": 4.1309999999999996e-05, "loss": 5.3867, "step": 4132 }, { "epoch": 0.0004133, "grad_norm": 4.611745357513428, "learning_rate": 4.1320000000000004e-05, "loss": 7.125, "step": 4133 }, { "epoch": 0.0004134, "grad_norm": 5.158962249755859, "learning_rate": 4.133e-05, "loss": 6.4805, "step": 4134 }, { "epoch": 0.0004135, "grad_norm": 3.5199623107910156, "learning_rate": 4.134e-05, "loss": 5.2383, "step": 4135 }, { "epoch": 0.0004136, "grad_norm": 5.691938877105713, "learning_rate": 4.135e-05, "loss": 6.1016, "step": 4136 }, { "epoch": 0.0004137, "grad_norm": 3.3723859786987305, "learning_rate": 4.136e-05, "loss": 5.2109, "step": 4137 }, { "epoch": 0.0004138, "grad_norm": 6.162867546081543, "learning_rate": 4.137e-05, "loss": 7.5703, "step": 4138 }, { "epoch": 0.0004139, "grad_norm": 4.326439380645752, "learning_rate": 4.138e-05, "loss": 6.2305, "step": 4139 }, { "epoch": 0.000414, "grad_norm": 3.8441545963287354, "learning_rate": 4.139e-05, "loss": 5.2812, "step": 4140 }, { "epoch": 0.0004141, "grad_norm": 4.565762042999268, "learning_rate": 4.1400000000000003e-05, "loss": 6.5039, "step": 4141 }, { "epoch": 0.0004142, "grad_norm": 3.5491926670074463, "learning_rate": 4.1410000000000005e-05, "loss": 5.1719, "step": 4142 }, { "epoch": 0.0004143, "grad_norm": 3.736703395843506, "learning_rate": 4.142e-05, "loss": 5.5312, "step": 4143 }, { "epoch": 0.0004144, "grad_norm": 4.117855548858643, "learning_rate": 4.143e-05, "loss": 6.0664, "step": 4144 }, { "epoch": 0.0004145, "grad_norm": 3.7153656482696533, "learning_rate": 4.144e-05, "loss": 5.8633, "step": 4145 }, { "epoch": 0.0004146, "grad_norm": 3.191498041152954, "learning_rate": 4.145e-05, "loss": 5.0781, "step": 4146 }, { "epoch": 0.0004147, "grad_norm": 3.89998197555542, "learning_rate": 4.1460000000000006e-05, "loss": 5.8984, "step": 4147 }, { "epoch": 0.0004148, "grad_norm": 3.8295412063598633, "learning_rate": 4.147e-05, "loss": 5.5312, "step": 4148 }, { "epoch": 0.0004149, "grad_norm": 3.934540271759033, "learning_rate": 4.1479999999999996e-05, "loss": 5.5664, "step": 4149 }, { "epoch": 0.000415, "grad_norm": 3.5701475143432617, "learning_rate": 4.1490000000000004e-05, "loss": 5.4492, "step": 4150 }, { "epoch": 0.0004151, "grad_norm": 3.2117130756378174, "learning_rate": 4.15e-05, "loss": 5.0703, "step": 4151 }, { "epoch": 0.0004152, "grad_norm": 3.373746871948242, "learning_rate": 4.151e-05, "loss": 5.0762, "step": 4152 }, { "epoch": 0.0004153, "grad_norm": 3.5973708629608154, "learning_rate": 4.152e-05, "loss": 5.457, "step": 4153 }, { "epoch": 0.0004154, "grad_norm": 3.4427895545959473, "learning_rate": 4.153e-05, "loss": 5.1758, "step": 4154 }, { "epoch": 0.0004155, "grad_norm": 4.331664085388184, "learning_rate": 4.154e-05, "loss": 5.6133, "step": 4155 }, { "epoch": 0.0004156, "grad_norm": 4.024486541748047, "learning_rate": 4.155e-05, "loss": 6.0781, "step": 4156 }, { "epoch": 0.0004157, "grad_norm": 4.06744384765625, "learning_rate": 4.156e-05, "loss": 6.0273, "step": 4157 }, { "epoch": 0.0004158, "grad_norm": 4.119234561920166, "learning_rate": 4.1570000000000003e-05, "loss": 5.5273, "step": 4158 }, { "epoch": 0.0004159, "grad_norm": 3.9706265926361084, "learning_rate": 4.1580000000000005e-05, "loss": 5.7422, "step": 4159 }, { "epoch": 0.000416, "grad_norm": 3.3748245239257812, "learning_rate": 4.159e-05, "loss": 5.0234, "step": 4160 }, { "epoch": 0.0004161, "grad_norm": 3.5364878177642822, "learning_rate": 4.16e-05, "loss": 5.3906, "step": 4161 }, { "epoch": 0.0004162, "grad_norm": 3.6886091232299805, "learning_rate": 4.161e-05, "loss": 5.5742, "step": 4162 }, { "epoch": 0.0004163, "grad_norm": 3.490872859954834, "learning_rate": 4.162e-05, "loss": 5.0078, "step": 4163 }, { "epoch": 0.0004164, "grad_norm": 4.100687503814697, "learning_rate": 4.1630000000000006e-05, "loss": 5.7031, "step": 4164 }, { "epoch": 0.0004165, "grad_norm": 3.3320984840393066, "learning_rate": 4.164e-05, "loss": 4.6602, "step": 4165 }, { "epoch": 0.0004166, "grad_norm": 3.7184898853302, "learning_rate": 4.1649999999999996e-05, "loss": 5.4766, "step": 4166 }, { "epoch": 0.0004167, "grad_norm": 3.5617237091064453, "learning_rate": 4.1660000000000004e-05, "loss": 5.2773, "step": 4167 }, { "epoch": 0.0004168, "grad_norm": 4.3061017990112305, "learning_rate": 4.167e-05, "loss": 4.9219, "step": 4168 }, { "epoch": 0.0004169, "grad_norm": 3.916656017303467, "learning_rate": 4.168e-05, "loss": 5.4023, "step": 4169 }, { "epoch": 0.000417, "grad_norm": 4.09127950668335, "learning_rate": 4.169e-05, "loss": 5.3281, "step": 4170 }, { "epoch": 0.0004171, "grad_norm": 3.869095802307129, "learning_rate": 4.17e-05, "loss": 5.1016, "step": 4171 }, { "epoch": 0.0004172, "grad_norm": 3.6654443740844727, "learning_rate": 4.171e-05, "loss": 5.332, "step": 4172 }, { "epoch": 0.0004173, "grad_norm": 3.554366111755371, "learning_rate": 4.172e-05, "loss": 4.8281, "step": 4173 }, { "epoch": 0.0004174, "grad_norm": 4.372258186340332, "learning_rate": 4.173e-05, "loss": 5.4375, "step": 4174 }, { "epoch": 0.0004175, "grad_norm": 4.621771335601807, "learning_rate": 4.1740000000000004e-05, "loss": 6.6289, "step": 4175 }, { "epoch": 0.0004176, "grad_norm": 3.802337408065796, "learning_rate": 4.1750000000000005e-05, "loss": 5.6758, "step": 4176 }, { "epoch": 0.0004177, "grad_norm": 3.4369053840637207, "learning_rate": 4.176e-05, "loss": 4.8438, "step": 4177 }, { "epoch": 0.0004178, "grad_norm": 3.4250431060791016, "learning_rate": 4.177e-05, "loss": 4.7461, "step": 4178 }, { "epoch": 0.0004179, "grad_norm": 3.225832462310791, "learning_rate": 4.178e-05, "loss": 4.7539, "step": 4179 }, { "epoch": 0.000418, "grad_norm": 3.6326236724853516, "learning_rate": 4.179e-05, "loss": 5.3477, "step": 4180 }, { "epoch": 0.0004181, "grad_norm": 3.5488100051879883, "learning_rate": 4.18e-05, "loss": 4.8223, "step": 4181 }, { "epoch": 0.0004182, "grad_norm": 3.2581963539123535, "learning_rate": 4.181e-05, "loss": 4.9414, "step": 4182 }, { "epoch": 0.0004183, "grad_norm": 4.745520114898682, "learning_rate": 4.1819999999999996e-05, "loss": 5.3613, "step": 4183 }, { "epoch": 0.0004184, "grad_norm": 3.348565101623535, "learning_rate": 4.1830000000000004e-05, "loss": 4.7812, "step": 4184 }, { "epoch": 0.0004185, "grad_norm": 3.4303088188171387, "learning_rate": 4.184e-05, "loss": 4.9746, "step": 4185 }, { "epoch": 0.0004186, "grad_norm": 3.349205255508423, "learning_rate": 4.185e-05, "loss": 4.9961, "step": 4186 }, { "epoch": 0.0004187, "grad_norm": 3.3469324111938477, "learning_rate": 4.186e-05, "loss": 4.9922, "step": 4187 }, { "epoch": 0.0004188, "grad_norm": 3.441373825073242, "learning_rate": 4.187e-05, "loss": 4.8203, "step": 4188 }, { "epoch": 0.0004189, "grad_norm": 3.6274783611297607, "learning_rate": 4.188e-05, "loss": 5.2812, "step": 4189 }, { "epoch": 0.000419, "grad_norm": 3.7490720748901367, "learning_rate": 4.189e-05, "loss": 5.1797, "step": 4190 }, { "epoch": 0.0004191, "grad_norm": 3.4138407707214355, "learning_rate": 4.19e-05, "loss": 4.9805, "step": 4191 }, { "epoch": 0.0004192, "grad_norm": 3.243234395980835, "learning_rate": 4.191e-05, "loss": 4.7109, "step": 4192 }, { "epoch": 0.0004193, "grad_norm": 3.567044734954834, "learning_rate": 4.1920000000000005e-05, "loss": 5.043, "step": 4193 }, { "epoch": 0.0004194, "grad_norm": 4.170904159545898, "learning_rate": 4.193e-05, "loss": 5.7109, "step": 4194 }, { "epoch": 0.0004195, "grad_norm": 3.589627742767334, "learning_rate": 4.194e-05, "loss": 4.9883, "step": 4195 }, { "epoch": 0.0004196, "grad_norm": 3.6273415088653564, "learning_rate": 4.195e-05, "loss": 4.7852, "step": 4196 }, { "epoch": 0.0004197, "grad_norm": 3.3961431980133057, "learning_rate": 4.196e-05, "loss": 4.8145, "step": 4197 }, { "epoch": 0.0004198, "grad_norm": 3.9560201168060303, "learning_rate": 4.197e-05, "loss": 5.3945, "step": 4198 }, { "epoch": 0.0004199, "grad_norm": 3.711775302886963, "learning_rate": 4.198e-05, "loss": 5.5078, "step": 4199 }, { "epoch": 0.00042, "grad_norm": 3.401026725769043, "learning_rate": 4.1989999999999996e-05, "loss": 4.8008, "step": 4200 }, { "epoch": 0.0004201, "grad_norm": 3.7390568256378174, "learning_rate": 4.2000000000000004e-05, "loss": 5.2969, "step": 4201 }, { "epoch": 0.0004202, "grad_norm": 3.483304738998413, "learning_rate": 4.201e-05, "loss": 5.0859, "step": 4202 }, { "epoch": 0.0004203, "grad_norm": 3.7833383083343506, "learning_rate": 4.202e-05, "loss": 5.418, "step": 4203 }, { "epoch": 0.0004204, "grad_norm": 3.1020076274871826, "learning_rate": 4.203e-05, "loss": 4.582, "step": 4204 }, { "epoch": 0.0004205, "grad_norm": 3.786668300628662, "learning_rate": 4.204e-05, "loss": 4.8281, "step": 4205 }, { "epoch": 0.0004206, "grad_norm": 3.220973491668701, "learning_rate": 4.205e-05, "loss": 4.5352, "step": 4206 }, { "epoch": 0.0004207, "grad_norm": 3.3292949199676514, "learning_rate": 4.206e-05, "loss": 4.6055, "step": 4207 }, { "epoch": 0.0004208, "grad_norm": 3.400747537612915, "learning_rate": 4.207e-05, "loss": 4.8906, "step": 4208 }, { "epoch": 0.0004209, "grad_norm": 3.537703275680542, "learning_rate": 4.208e-05, "loss": 4.7031, "step": 4209 }, { "epoch": 0.000421, "grad_norm": 3.278318166732788, "learning_rate": 4.2090000000000005e-05, "loss": 4.8711, "step": 4210 }, { "epoch": 0.0004211, "grad_norm": 3.4238805770874023, "learning_rate": 4.21e-05, "loss": 5.0547, "step": 4211 }, { "epoch": 0.0004212, "grad_norm": 3.4830808639526367, "learning_rate": 4.211e-05, "loss": 4.6797, "step": 4212 }, { "epoch": 0.0004213, "grad_norm": 4.247705936431885, "learning_rate": 4.212e-05, "loss": 6.0586, "step": 4213 }, { "epoch": 0.0004214, "grad_norm": 3.4551994800567627, "learning_rate": 4.213e-05, "loss": 4.9609, "step": 4214 }, { "epoch": 0.0004215, "grad_norm": 3.356276035308838, "learning_rate": 4.214e-05, "loss": 4.3906, "step": 4215 }, { "epoch": 0.0004216, "grad_norm": 3.396141767501831, "learning_rate": 4.215e-05, "loss": 4.8398, "step": 4216 }, { "epoch": 0.0004217, "grad_norm": 3.809727430343628, "learning_rate": 4.2159999999999996e-05, "loss": 4.9961, "step": 4217 }, { "epoch": 0.0004218, "grad_norm": 3.542297124862671, "learning_rate": 4.2170000000000005e-05, "loss": 5.3867, "step": 4218 }, { "epoch": 0.0004219, "grad_norm": 3.5642385482788086, "learning_rate": 4.218e-05, "loss": 5.0977, "step": 4219 }, { "epoch": 0.000422, "grad_norm": 3.12821364402771, "learning_rate": 4.219e-05, "loss": 4.3027, "step": 4220 }, { "epoch": 0.0004221, "grad_norm": 3.2699859142303467, "learning_rate": 4.22e-05, "loss": 4.6875, "step": 4221 }, { "epoch": 0.0004222, "grad_norm": 4.656696319580078, "learning_rate": 4.221e-05, "loss": 5.4922, "step": 4222 }, { "epoch": 0.0004223, "grad_norm": 3.6123883724212646, "learning_rate": 4.222e-05, "loss": 4.9336, "step": 4223 }, { "epoch": 0.0004224, "grad_norm": 3.2034406661987305, "learning_rate": 4.223e-05, "loss": 4.6094, "step": 4224 }, { "epoch": 0.0004225, "grad_norm": 3.629655122756958, "learning_rate": 4.224e-05, "loss": 5.3379, "step": 4225 }, { "epoch": 0.0004226, "grad_norm": 3.694410562515259, "learning_rate": 4.225e-05, "loss": 4.8555, "step": 4226 }, { "epoch": 0.0004227, "grad_norm": 3.4406235218048096, "learning_rate": 4.2260000000000005e-05, "loss": 4.7695, "step": 4227 }, { "epoch": 0.0004228, "grad_norm": 3.6419994831085205, "learning_rate": 4.227e-05, "loss": 5.0352, "step": 4228 }, { "epoch": 0.0004229, "grad_norm": 3.465791940689087, "learning_rate": 4.228e-05, "loss": 4.8359, "step": 4229 }, { "epoch": 0.000423, "grad_norm": 3.3964591026306152, "learning_rate": 4.229e-05, "loss": 4.6992, "step": 4230 }, { "epoch": 0.0004231, "grad_norm": 3.2057528495788574, "learning_rate": 4.23e-05, "loss": 4.4844, "step": 4231 }, { "epoch": 0.0004232, "grad_norm": 3.5977606773376465, "learning_rate": 4.231e-05, "loss": 5.0898, "step": 4232 }, { "epoch": 0.0004233, "grad_norm": 3.4214189052581787, "learning_rate": 4.232e-05, "loss": 4.8789, "step": 4233 }, { "epoch": 0.0004234, "grad_norm": 3.6013171672821045, "learning_rate": 4.2329999999999996e-05, "loss": 5.0195, "step": 4234 }, { "epoch": 0.0004235, "grad_norm": 4.263355255126953, "learning_rate": 4.2340000000000005e-05, "loss": 5.3633, "step": 4235 }, { "epoch": 0.0004236, "grad_norm": 3.357116222381592, "learning_rate": 4.235e-05, "loss": 4.375, "step": 4236 }, { "epoch": 0.0004237, "grad_norm": 3.553157091140747, "learning_rate": 4.236e-05, "loss": 4.5117, "step": 4237 }, { "epoch": 0.0004238, "grad_norm": 3.8400962352752686, "learning_rate": 4.237e-05, "loss": 5.0977, "step": 4238 }, { "epoch": 0.0004239, "grad_norm": 6.9885759353637695, "learning_rate": 4.238e-05, "loss": 5.2539, "step": 4239 }, { "epoch": 0.000424, "grad_norm": 3.538587808609009, "learning_rate": 4.239e-05, "loss": 4.6406, "step": 4240 }, { "epoch": 0.0004241, "grad_norm": 3.7649664878845215, "learning_rate": 4.24e-05, "loss": 4.8398, "step": 4241 }, { "epoch": 0.0004242, "grad_norm": 3.3764755725860596, "learning_rate": 4.241e-05, "loss": 4.7461, "step": 4242 }, { "epoch": 0.0004243, "grad_norm": 3.669555187225342, "learning_rate": 4.242e-05, "loss": 4.7695, "step": 4243 }, { "epoch": 0.0004244, "grad_norm": 3.5933690071105957, "learning_rate": 4.2430000000000005e-05, "loss": 4.7324, "step": 4244 }, { "epoch": 0.0004245, "grad_norm": 3.2653861045837402, "learning_rate": 4.244e-05, "loss": 4.2852, "step": 4245 }, { "epoch": 0.0004246, "grad_norm": 3.9884986877441406, "learning_rate": 4.245e-05, "loss": 4.9805, "step": 4246 }, { "epoch": 0.0004247, "grad_norm": 3.2516067028045654, "learning_rate": 4.2460000000000003e-05, "loss": 4.3203, "step": 4247 }, { "epoch": 0.0004248, "grad_norm": 3.2346811294555664, "learning_rate": 4.247e-05, "loss": 4.6484, "step": 4248 }, { "epoch": 0.0004249, "grad_norm": 3.4629881381988525, "learning_rate": 4.248e-05, "loss": 4.8984, "step": 4249 }, { "epoch": 0.000425, "grad_norm": 3.5164341926574707, "learning_rate": 4.249e-05, "loss": 4.5234, "step": 4250 }, { "epoch": 0.0004251, "grad_norm": 3.4341697692871094, "learning_rate": 4.2499999999999996e-05, "loss": 4.4492, "step": 4251 }, { "epoch": 0.0004252, "grad_norm": 3.2241580486297607, "learning_rate": 4.2510000000000005e-05, "loss": 4.1406, "step": 4252 }, { "epoch": 0.0004253, "grad_norm": 3.3622195720672607, "learning_rate": 4.252e-05, "loss": 4.3125, "step": 4253 }, { "epoch": 0.0004254, "grad_norm": 3.9681081771850586, "learning_rate": 4.253e-05, "loss": 5.168, "step": 4254 }, { "epoch": 0.0004255, "grad_norm": 4.022111892700195, "learning_rate": 4.254e-05, "loss": 5.3418, "step": 4255 }, { "epoch": 0.0004256, "grad_norm": 3.52665376663208, "learning_rate": 4.255e-05, "loss": 4.4961, "step": 4256 }, { "epoch": 0.0004257, "grad_norm": 3.554487705230713, "learning_rate": 4.256e-05, "loss": 4.5078, "step": 4257 }, { "epoch": 0.0004258, "grad_norm": 3.3719255924224854, "learning_rate": 4.257e-05, "loss": 4.5586, "step": 4258 }, { "epoch": 0.0004259, "grad_norm": 3.2635750770568848, "learning_rate": 4.258e-05, "loss": 4.3184, "step": 4259 }, { "epoch": 0.000426, "grad_norm": 3.1879544258117676, "learning_rate": 4.259e-05, "loss": 4.5879, "step": 4260 }, { "epoch": 0.0004261, "grad_norm": 3.302591323852539, "learning_rate": 4.2600000000000005e-05, "loss": 4.8828, "step": 4261 }, { "epoch": 0.0004262, "grad_norm": 3.3501760959625244, "learning_rate": 4.261e-05, "loss": 4.3789, "step": 4262 }, { "epoch": 0.0004263, "grad_norm": 3.072420835494995, "learning_rate": 4.262e-05, "loss": 4.4629, "step": 4263 }, { "epoch": 0.0004264, "grad_norm": 4.598276615142822, "learning_rate": 4.2630000000000004e-05, "loss": 5.2969, "step": 4264 }, { "epoch": 0.0004265, "grad_norm": 3.54929518699646, "learning_rate": 4.264e-05, "loss": 4.7539, "step": 4265 }, { "epoch": 0.0004266, "grad_norm": 3.756957769393921, "learning_rate": 4.265e-05, "loss": 4.582, "step": 4266 }, { "epoch": 0.0004267, "grad_norm": 3.732717990875244, "learning_rate": 4.266e-05, "loss": 4.9824, "step": 4267 }, { "epoch": 0.0004268, "grad_norm": 3.422294855117798, "learning_rate": 4.2669999999999996e-05, "loss": 5.1074, "step": 4268 }, { "epoch": 0.0004269, "grad_norm": 3.0590617656707764, "learning_rate": 4.2680000000000005e-05, "loss": 4.2539, "step": 4269 }, { "epoch": 0.000427, "grad_norm": 3.3566641807556152, "learning_rate": 4.269e-05, "loss": 4.8281, "step": 4270 }, { "epoch": 0.0004271, "grad_norm": 3.7169976234436035, "learning_rate": 4.27e-05, "loss": 4.8965, "step": 4271 }, { "epoch": 0.0004272, "grad_norm": 4.839364051818848, "learning_rate": 4.271e-05, "loss": 5.668, "step": 4272 }, { "epoch": 0.0004273, "grad_norm": 3.639251947402954, "learning_rate": 4.272e-05, "loss": 4.5117, "step": 4273 }, { "epoch": 0.0004274, "grad_norm": 3.884964942932129, "learning_rate": 4.273e-05, "loss": 5.0547, "step": 4274 }, { "epoch": 0.0004275, "grad_norm": 3.5250017642974854, "learning_rate": 4.274e-05, "loss": 4.9004, "step": 4275 }, { "epoch": 0.0004276, "grad_norm": 3.460820436477661, "learning_rate": 4.275e-05, "loss": 4.7539, "step": 4276 }, { "epoch": 0.0004277, "grad_norm": 3.3357040882110596, "learning_rate": 4.276e-05, "loss": 4.5195, "step": 4277 }, { "epoch": 0.0004278, "grad_norm": 3.386826515197754, "learning_rate": 4.2770000000000006e-05, "loss": 4.293, "step": 4278 }, { "epoch": 0.0004279, "grad_norm": 3.3886353969573975, "learning_rate": 4.278e-05, "loss": 4.3301, "step": 4279 }, { "epoch": 0.000428, "grad_norm": 3.3160009384155273, "learning_rate": 4.279e-05, "loss": 4.3086, "step": 4280 }, { "epoch": 0.0004281, "grad_norm": 3.671212911605835, "learning_rate": 4.2800000000000004e-05, "loss": 5.375, "step": 4281 }, { "epoch": 0.0004282, "grad_norm": 3.376387119293213, "learning_rate": 4.281e-05, "loss": 4.5273, "step": 4282 }, { "epoch": 0.0004283, "grad_norm": 4.032546520233154, "learning_rate": 4.282e-05, "loss": 4.8984, "step": 4283 }, { "epoch": 0.0004284, "grad_norm": 3.21494722366333, "learning_rate": 4.283e-05, "loss": 4.293, "step": 4284 }, { "epoch": 0.0004285, "grad_norm": 3.513974666595459, "learning_rate": 4.2839999999999996e-05, "loss": 4.4199, "step": 4285 }, { "epoch": 0.0004286, "grad_norm": 3.2516965866088867, "learning_rate": 4.2850000000000005e-05, "loss": 4.3652, "step": 4286 }, { "epoch": 0.0004287, "grad_norm": 3.204125165939331, "learning_rate": 4.286e-05, "loss": 4.4355, "step": 4287 }, { "epoch": 0.0004288, "grad_norm": 3.2614102363586426, "learning_rate": 4.287e-05, "loss": 4.3164, "step": 4288 }, { "epoch": 0.0004289, "grad_norm": 3.2917561531066895, "learning_rate": 4.288e-05, "loss": 4.2949, "step": 4289 }, { "epoch": 0.000429, "grad_norm": 3.2204697132110596, "learning_rate": 4.289e-05, "loss": 4.1992, "step": 4290 }, { "epoch": 0.0004291, "grad_norm": 3.86523175239563, "learning_rate": 4.29e-05, "loss": 5.0898, "step": 4291 }, { "epoch": 0.0004292, "grad_norm": 3.6371407508850098, "learning_rate": 4.291e-05, "loss": 4.7227, "step": 4292 }, { "epoch": 0.0004293, "grad_norm": 3.5192179679870605, "learning_rate": 4.292e-05, "loss": 4.3398, "step": 4293 }, { "epoch": 0.0004294, "grad_norm": 3.529805898666382, "learning_rate": 4.293e-05, "loss": 4.2578, "step": 4294 }, { "epoch": 0.0004295, "grad_norm": 3.334998607635498, "learning_rate": 4.2940000000000006e-05, "loss": 4.2969, "step": 4295 }, { "epoch": 0.0004296, "grad_norm": 3.3071017265319824, "learning_rate": 4.295e-05, "loss": 4.5586, "step": 4296 }, { "epoch": 0.0004297, "grad_norm": 3.673053503036499, "learning_rate": 4.296e-05, "loss": 4.6172, "step": 4297 }, { "epoch": 0.0004298, "grad_norm": 3.2039501667022705, "learning_rate": 4.2970000000000004e-05, "loss": 4.4297, "step": 4298 }, { "epoch": 0.0004299, "grad_norm": 3.4562950134277344, "learning_rate": 4.298e-05, "loss": 4.7676, "step": 4299 }, { "epoch": 0.00043, "grad_norm": 3.4909989833831787, "learning_rate": 4.299e-05, "loss": 4.4219, "step": 4300 }, { "epoch": 0.0004301, "grad_norm": 3.34441876411438, "learning_rate": 4.3e-05, "loss": 4.1426, "step": 4301 }, { "epoch": 0.0004302, "grad_norm": 2.9811437129974365, "learning_rate": 4.3009999999999996e-05, "loss": 4.2559, "step": 4302 }, { "epoch": 0.0004303, "grad_norm": 4.048996448516846, "learning_rate": 4.3020000000000005e-05, "loss": 5.2441, "step": 4303 }, { "epoch": 0.0004304, "grad_norm": 3.2754409313201904, "learning_rate": 4.303e-05, "loss": 4.3242, "step": 4304 }, { "epoch": 0.0004305, "grad_norm": 3.3092200756073, "learning_rate": 4.304e-05, "loss": 4.5918, "step": 4305 }, { "epoch": 0.0004306, "grad_norm": 3.3667242527008057, "learning_rate": 4.305e-05, "loss": 4.3906, "step": 4306 }, { "epoch": 0.0004307, "grad_norm": 3.7178213596343994, "learning_rate": 4.306e-05, "loss": 4.3672, "step": 4307 }, { "epoch": 0.0004308, "grad_norm": 2.8777174949645996, "learning_rate": 4.307e-05, "loss": 3.9922, "step": 4308 }, { "epoch": 0.0004309, "grad_norm": 3.315974235534668, "learning_rate": 4.308e-05, "loss": 4.3359, "step": 4309 }, { "epoch": 0.000431, "grad_norm": 7.670527935028076, "learning_rate": 4.309e-05, "loss": 5.7773, "step": 4310 }, { "epoch": 0.0004311, "grad_norm": 3.3412671089172363, "learning_rate": 4.31e-05, "loss": 4.3496, "step": 4311 }, { "epoch": 0.0004312, "grad_norm": 3.4237232208251953, "learning_rate": 4.3110000000000006e-05, "loss": 4.3711, "step": 4312 }, { "epoch": 0.0004313, "grad_norm": 3.414289951324463, "learning_rate": 4.312e-05, "loss": 4.4863, "step": 4313 }, { "epoch": 0.0004314, "grad_norm": 3.1857030391693115, "learning_rate": 4.313e-05, "loss": 4.4258, "step": 4314 }, { "epoch": 0.0004315, "grad_norm": 3.3210511207580566, "learning_rate": 4.3140000000000004e-05, "loss": 4.2617, "step": 4315 }, { "epoch": 0.0004316, "grad_norm": 3.5879602432250977, "learning_rate": 4.315e-05, "loss": 4.2891, "step": 4316 }, { "epoch": 0.0004317, "grad_norm": 3.228205442428589, "learning_rate": 4.316e-05, "loss": 4.1055, "step": 4317 }, { "epoch": 0.0004318, "grad_norm": 3.39374041557312, "learning_rate": 4.317e-05, "loss": 4.1387, "step": 4318 }, { "epoch": 0.0004319, "grad_norm": 3.4702086448669434, "learning_rate": 4.3179999999999997e-05, "loss": 4.5059, "step": 4319 }, { "epoch": 0.000432, "grad_norm": 3.2577431201934814, "learning_rate": 4.3190000000000005e-05, "loss": 4.1074, "step": 4320 }, { "epoch": 0.0004321, "grad_norm": 3.4151816368103027, "learning_rate": 4.32e-05, "loss": 4.3906, "step": 4321 }, { "epoch": 0.0004322, "grad_norm": 3.102512836456299, "learning_rate": 4.321e-05, "loss": 4.2734, "step": 4322 }, { "epoch": 0.0004323, "grad_norm": 3.176802158355713, "learning_rate": 4.322e-05, "loss": 4.2188, "step": 4323 }, { "epoch": 0.0004324, "grad_norm": 5.058439254760742, "learning_rate": 4.323e-05, "loss": 5.2637, "step": 4324 }, { "epoch": 0.0004325, "grad_norm": 3.859606981277466, "learning_rate": 4.324e-05, "loss": 4.3008, "step": 4325 }, { "epoch": 0.0004326, "grad_norm": 3.3192851543426514, "learning_rate": 4.325e-05, "loss": 4.0996, "step": 4326 }, { "epoch": 0.0004327, "grad_norm": 3.69804048538208, "learning_rate": 4.326e-05, "loss": 4.6641, "step": 4327 }, { "epoch": 0.0004328, "grad_norm": 3.4054818153381348, "learning_rate": 4.327e-05, "loss": 4.2812, "step": 4328 }, { "epoch": 0.0004329, "grad_norm": 3.723554849624634, "learning_rate": 4.3280000000000006e-05, "loss": 4.5547, "step": 4329 }, { "epoch": 0.000433, "grad_norm": 4.134061336517334, "learning_rate": 4.329e-05, "loss": 4.9082, "step": 4330 }, { "epoch": 0.0004331, "grad_norm": 3.321018695831299, "learning_rate": 4.33e-05, "loss": 4.1348, "step": 4331 }, { "epoch": 0.0004332, "grad_norm": 3.1106321811676025, "learning_rate": 4.3310000000000004e-05, "loss": 3.916, "step": 4332 }, { "epoch": 0.0004333, "grad_norm": 3.423844337463379, "learning_rate": 4.332e-05, "loss": 4.2383, "step": 4333 }, { "epoch": 0.0004334, "grad_norm": 3.9197568893432617, "learning_rate": 4.333e-05, "loss": 4.002, "step": 4334 }, { "epoch": 0.0004335, "grad_norm": 4.331193923950195, "learning_rate": 4.334e-05, "loss": 4.3047, "step": 4335 }, { "epoch": 0.0004336, "grad_norm": 3.3250527381896973, "learning_rate": 4.3349999999999997e-05, "loss": 4.293, "step": 4336 }, { "epoch": 0.0004337, "grad_norm": 3.369507312774658, "learning_rate": 4.3360000000000005e-05, "loss": 4.1445, "step": 4337 }, { "epoch": 0.0004338, "grad_norm": 3.318035364151001, "learning_rate": 4.337e-05, "loss": 4.0117, "step": 4338 }, { "epoch": 0.0004339, "grad_norm": 3.6728227138519287, "learning_rate": 4.338e-05, "loss": 4.6445, "step": 4339 }, { "epoch": 0.000434, "grad_norm": 3.2312772274017334, "learning_rate": 4.339e-05, "loss": 3.8379, "step": 4340 }, { "epoch": 0.0004341, "grad_norm": 3.167508125305176, "learning_rate": 4.34e-05, "loss": 4.0273, "step": 4341 }, { "epoch": 0.0004342, "grad_norm": 3.3273322582244873, "learning_rate": 4.341e-05, "loss": 4.2949, "step": 4342 }, { "epoch": 0.0004343, "grad_norm": 3.624297618865967, "learning_rate": 4.342e-05, "loss": 4.127, "step": 4343 }, { "epoch": 0.0004344, "grad_norm": 4.5100226402282715, "learning_rate": 4.343e-05, "loss": 4.4629, "step": 4344 }, { "epoch": 0.0004345, "grad_norm": 3.4498915672302246, "learning_rate": 4.344e-05, "loss": 4.2422, "step": 4345 }, { "epoch": 0.0004346, "grad_norm": 3.4710705280303955, "learning_rate": 4.3450000000000006e-05, "loss": 3.9629, "step": 4346 }, { "epoch": 0.0004347, "grad_norm": 4.026734828948975, "learning_rate": 4.346e-05, "loss": 4.8867, "step": 4347 }, { "epoch": 0.0004348, "grad_norm": 3.228076457977295, "learning_rate": 4.347e-05, "loss": 4.291, "step": 4348 }, { "epoch": 0.0004349, "grad_norm": 3.651885747909546, "learning_rate": 4.3480000000000004e-05, "loss": 4.123, "step": 4349 }, { "epoch": 0.000435, "grad_norm": 3.2708592414855957, "learning_rate": 4.349e-05, "loss": 4.1016, "step": 4350 }, { "epoch": 0.0004351, "grad_norm": 3.122850179672241, "learning_rate": 4.35e-05, "loss": 4.0527, "step": 4351 }, { "epoch": 0.0004352, "grad_norm": 3.489903211593628, "learning_rate": 4.351e-05, "loss": 4.207, "step": 4352 }, { "epoch": 0.0004353, "grad_norm": 3.1932284832000732, "learning_rate": 4.352e-05, "loss": 3.834, "step": 4353 }, { "epoch": 0.0004354, "grad_norm": 3.2868921756744385, "learning_rate": 4.3530000000000005e-05, "loss": 4.0488, "step": 4354 }, { "epoch": 0.0004355, "grad_norm": 3.1833255290985107, "learning_rate": 4.354e-05, "loss": 3.998, "step": 4355 }, { "epoch": 0.0004356, "grad_norm": 3.548819065093994, "learning_rate": 4.355e-05, "loss": 4.3945, "step": 4356 }, { "epoch": 0.0004357, "grad_norm": 3.3118855953216553, "learning_rate": 4.356e-05, "loss": 4.0352, "step": 4357 }, { "epoch": 0.0004358, "grad_norm": 3.427961587905884, "learning_rate": 4.357e-05, "loss": 4.6172, "step": 4358 }, { "epoch": 0.0004359, "grad_norm": 3.989943742752075, "learning_rate": 4.358e-05, "loss": 4.7988, "step": 4359 }, { "epoch": 0.000436, "grad_norm": 3.4652833938598633, "learning_rate": 4.359e-05, "loss": 4.1055, "step": 4360 }, { "epoch": 0.0004361, "grad_norm": 3.535144805908203, "learning_rate": 4.36e-05, "loss": 4.2285, "step": 4361 }, { "epoch": 0.0004362, "grad_norm": 3.407724380493164, "learning_rate": 4.361e-05, "loss": 4.2695, "step": 4362 }, { "epoch": 0.0004363, "grad_norm": 3.355693817138672, "learning_rate": 4.3620000000000006e-05, "loss": 4.2051, "step": 4363 }, { "epoch": 0.0004364, "grad_norm": 3.127380609512329, "learning_rate": 4.363e-05, "loss": 3.9297, "step": 4364 }, { "epoch": 0.0004365, "grad_norm": 3.3452811241149902, "learning_rate": 4.364e-05, "loss": 4.0156, "step": 4365 }, { "epoch": 0.0004366, "grad_norm": 3.556596517562866, "learning_rate": 4.3650000000000004e-05, "loss": 4.4004, "step": 4366 }, { "epoch": 0.0004367, "grad_norm": 2.9243733882904053, "learning_rate": 4.366e-05, "loss": 3.8086, "step": 4367 }, { "epoch": 0.0004368, "grad_norm": 3.162949562072754, "learning_rate": 4.367e-05, "loss": 3.6934, "step": 4368 }, { "epoch": 0.0004369, "grad_norm": 5.681541442871094, "learning_rate": 4.368e-05, "loss": 5.3613, "step": 4369 }, { "epoch": 0.000437, "grad_norm": 3.1346187591552734, "learning_rate": 4.369e-05, "loss": 4.1484, "step": 4370 }, { "epoch": 0.0004371, "grad_norm": 3.4794206619262695, "learning_rate": 4.3700000000000005e-05, "loss": 3.8203, "step": 4371 }, { "epoch": 0.0004372, "grad_norm": 3.1363070011138916, "learning_rate": 4.371e-05, "loss": 3.8691, "step": 4372 }, { "epoch": 0.0004373, "grad_norm": 3.3591103553771973, "learning_rate": 4.372e-05, "loss": 3.9531, "step": 4373 }, { "epoch": 0.0004374, "grad_norm": 3.8136484622955322, "learning_rate": 4.373e-05, "loss": 4.1953, "step": 4374 }, { "epoch": 0.0004375, "grad_norm": 3.3347480297088623, "learning_rate": 4.374e-05, "loss": 3.9355, "step": 4375 }, { "epoch": 0.0004376, "grad_norm": 3.1886277198791504, "learning_rate": 4.375e-05, "loss": 3.9863, "step": 4376 }, { "epoch": 0.0004377, "grad_norm": 3.171515464782715, "learning_rate": 4.376e-05, "loss": 4.2422, "step": 4377 }, { "epoch": 0.0004378, "grad_norm": 3.1116981506347656, "learning_rate": 4.377e-05, "loss": 3.8047, "step": 4378 }, { "epoch": 0.0004379, "grad_norm": 3.1574769020080566, "learning_rate": 4.378e-05, "loss": 3.9727, "step": 4379 }, { "epoch": 0.000438, "grad_norm": 4.211515426635742, "learning_rate": 4.3790000000000006e-05, "loss": 4.9746, "step": 4380 }, { "epoch": 0.0004381, "grad_norm": 3.0051376819610596, "learning_rate": 4.38e-05, "loss": 3.9395, "step": 4381 }, { "epoch": 0.0004382, "grad_norm": 3.047628402709961, "learning_rate": 4.381e-05, "loss": 3.7402, "step": 4382 }, { "epoch": 0.0004383, "grad_norm": 3.559969425201416, "learning_rate": 4.3820000000000004e-05, "loss": 4.4336, "step": 4383 }, { "epoch": 0.0004384, "grad_norm": 3.7711033821105957, "learning_rate": 4.383e-05, "loss": 4.5078, "step": 4384 }, { "epoch": 0.0004385, "grad_norm": 3.393209457397461, "learning_rate": 4.384e-05, "loss": 4.3418, "step": 4385 }, { "epoch": 0.0004386, "grad_norm": 3.179847240447998, "learning_rate": 4.385e-05, "loss": 4.1016, "step": 4386 }, { "epoch": 0.0004387, "grad_norm": 2.9167933464050293, "learning_rate": 4.386e-05, "loss": 3.7441, "step": 4387 }, { "epoch": 0.0004388, "grad_norm": 3.5401833057403564, "learning_rate": 4.3870000000000005e-05, "loss": 4.373, "step": 4388 }, { "epoch": 0.0004389, "grad_norm": 3.2971513271331787, "learning_rate": 4.388e-05, "loss": 4.123, "step": 4389 }, { "epoch": 0.000439, "grad_norm": 4.083408355712891, "learning_rate": 4.389e-05, "loss": 4.6035, "step": 4390 }, { "epoch": 0.0004391, "grad_norm": 3.2623190879821777, "learning_rate": 4.39e-05, "loss": 4.1211, "step": 4391 }, { "epoch": 0.0004392, "grad_norm": 3.393063545227051, "learning_rate": 4.391e-05, "loss": 3.8789, "step": 4392 }, { "epoch": 0.0004393, "grad_norm": 3.201707124710083, "learning_rate": 4.392e-05, "loss": 4.0879, "step": 4393 }, { "epoch": 0.0004394, "grad_norm": 3.347805976867676, "learning_rate": 4.393e-05, "loss": 4.1836, "step": 4394 }, { "epoch": 0.0004395, "grad_norm": 3.27976131439209, "learning_rate": 4.394e-05, "loss": 3.8906, "step": 4395 }, { "epoch": 0.0004396, "grad_norm": 3.8289692401885986, "learning_rate": 4.395e-05, "loss": 4.5, "step": 4396 }, { "epoch": 0.0004397, "grad_norm": 3.886551856994629, "learning_rate": 4.3960000000000006e-05, "loss": 4.3633, "step": 4397 }, { "epoch": 0.0004398, "grad_norm": 3.25569486618042, "learning_rate": 4.397e-05, "loss": 4.1211, "step": 4398 }, { "epoch": 0.0004399, "grad_norm": 3.342954635620117, "learning_rate": 4.398e-05, "loss": 3.8457, "step": 4399 }, { "epoch": 0.00044, "grad_norm": 3.376132011413574, "learning_rate": 4.3990000000000004e-05, "loss": 3.9727, "step": 4400 }, { "epoch": 0.0004401, "grad_norm": 3.4003143310546875, "learning_rate": 4.4e-05, "loss": 4.0527, "step": 4401 }, { "epoch": 0.0004402, "grad_norm": 4.261634826660156, "learning_rate": 4.401e-05, "loss": 4.1094, "step": 4402 }, { "epoch": 0.0004403, "grad_norm": 3.138030767440796, "learning_rate": 4.402e-05, "loss": 4.1758, "step": 4403 }, { "epoch": 0.0004404, "grad_norm": 3.435187339782715, "learning_rate": 4.403e-05, "loss": 3.957, "step": 4404 }, { "epoch": 0.0004405, "grad_norm": 2.9955251216888428, "learning_rate": 4.4040000000000005e-05, "loss": 3.6328, "step": 4405 }, { "epoch": 0.0004406, "grad_norm": 3.080143928527832, "learning_rate": 4.405e-05, "loss": 3.707, "step": 4406 }, { "epoch": 0.0004407, "grad_norm": 3.551837682723999, "learning_rate": 4.406e-05, "loss": 4.4199, "step": 4407 }, { "epoch": 0.0004408, "grad_norm": 4.1016764640808105, "learning_rate": 4.407e-05, "loss": 4.2988, "step": 4408 }, { "epoch": 0.0004409, "grad_norm": 3.2563037872314453, "learning_rate": 4.408e-05, "loss": 4.0391, "step": 4409 }, { "epoch": 0.000441, "grad_norm": 3.325324773788452, "learning_rate": 4.409e-05, "loss": 4.1445, "step": 4410 }, { "epoch": 0.0004411, "grad_norm": 3.2726807594299316, "learning_rate": 4.41e-05, "loss": 4.0703, "step": 4411 }, { "epoch": 0.0004412, "grad_norm": 3.113215684890747, "learning_rate": 4.411e-05, "loss": 3.9238, "step": 4412 }, { "epoch": 0.0004413, "grad_norm": 3.1880133152008057, "learning_rate": 4.412e-05, "loss": 3.8672, "step": 4413 }, { "epoch": 0.0004414, "grad_norm": 3.0497143268585205, "learning_rate": 4.4130000000000006e-05, "loss": 3.7207, "step": 4414 }, { "epoch": 0.0004415, "grad_norm": 3.1334667205810547, "learning_rate": 4.414e-05, "loss": 4.0059, "step": 4415 }, { "epoch": 0.0004416, "grad_norm": 3.052170991897583, "learning_rate": 4.415e-05, "loss": 3.6836, "step": 4416 }, { "epoch": 0.0004417, "grad_norm": 3.2291038036346436, "learning_rate": 4.4160000000000004e-05, "loss": 4.2109, "step": 4417 }, { "epoch": 0.0004418, "grad_norm": 3.56504225730896, "learning_rate": 4.417e-05, "loss": 4.7402, "step": 4418 }, { "epoch": 0.0004419, "grad_norm": 3.699209213256836, "learning_rate": 4.418e-05, "loss": 4.4062, "step": 4419 }, { "epoch": 0.000442, "grad_norm": 3.0333292484283447, "learning_rate": 4.419e-05, "loss": 3.4492, "step": 4420 }, { "epoch": 0.0004421, "grad_norm": 3.1669559478759766, "learning_rate": 4.42e-05, "loss": 4.0234, "step": 4421 }, { "epoch": 0.0004422, "grad_norm": 3.9274795055389404, "learning_rate": 4.4210000000000005e-05, "loss": 4.4688, "step": 4422 }, { "epoch": 0.0004423, "grad_norm": 3.2515194416046143, "learning_rate": 4.422e-05, "loss": 3.7754, "step": 4423 }, { "epoch": 0.0004424, "grad_norm": 3.06618332862854, "learning_rate": 4.423e-05, "loss": 3.6582, "step": 4424 }, { "epoch": 0.0004425, "grad_norm": 2.9362614154815674, "learning_rate": 4.424e-05, "loss": 3.7227, "step": 4425 }, { "epoch": 0.0004426, "grad_norm": 3.122183084487915, "learning_rate": 4.425e-05, "loss": 3.6719, "step": 4426 }, { "epoch": 0.0004427, "grad_norm": 4.1390886306762695, "learning_rate": 4.426e-05, "loss": 4.6602, "step": 4427 }, { "epoch": 0.0004428, "grad_norm": 3.402243137359619, "learning_rate": 4.427e-05, "loss": 4.125, "step": 4428 }, { "epoch": 0.0004429, "grad_norm": 3.399998188018799, "learning_rate": 4.428e-05, "loss": 3.918, "step": 4429 }, { "epoch": 0.000443, "grad_norm": 3.567279577255249, "learning_rate": 4.429e-05, "loss": 4.1953, "step": 4430 }, { "epoch": 0.0004431, "grad_norm": 3.257835865020752, "learning_rate": 4.4300000000000006e-05, "loss": 3.6914, "step": 4431 }, { "epoch": 0.0004432, "grad_norm": 5.018098831176758, "learning_rate": 4.431e-05, "loss": 5.0195, "step": 4432 }, { "epoch": 0.0004433, "grad_norm": 3.0309534072875977, "learning_rate": 4.432e-05, "loss": 3.6484, "step": 4433 }, { "epoch": 0.0004434, "grad_norm": 3.2106688022613525, "learning_rate": 4.4330000000000004e-05, "loss": 3.9336, "step": 4434 }, { "epoch": 0.0004435, "grad_norm": 3.136956214904785, "learning_rate": 4.434e-05, "loss": 3.6953, "step": 4435 }, { "epoch": 0.0004436, "grad_norm": 3.10791015625, "learning_rate": 4.435e-05, "loss": 3.875, "step": 4436 }, { "epoch": 0.0004437, "grad_norm": 3.130392074584961, "learning_rate": 4.436e-05, "loss": 4.002, "step": 4437 }, { "epoch": 0.0004438, "grad_norm": 3.51442813873291, "learning_rate": 4.437e-05, "loss": 3.8867, "step": 4438 }, { "epoch": 0.0004439, "grad_norm": 3.3942458629608154, "learning_rate": 4.438e-05, "loss": 3.8457, "step": 4439 }, { "epoch": 0.000444, "grad_norm": 3.0267603397369385, "learning_rate": 4.439e-05, "loss": 3.6562, "step": 4440 }, { "epoch": 0.0004441, "grad_norm": 2.9764490127563477, "learning_rate": 4.44e-05, "loss": 3.6367, "step": 4441 }, { "epoch": 0.0004442, "grad_norm": 2.967196464538574, "learning_rate": 4.4410000000000003e-05, "loss": 3.7344, "step": 4442 }, { "epoch": 0.0004443, "grad_norm": 3.2944579124450684, "learning_rate": 4.442e-05, "loss": 4.0449, "step": 4443 }, { "epoch": 0.0004444, "grad_norm": 3.1502978801727295, "learning_rate": 4.443e-05, "loss": 3.9453, "step": 4444 }, { "epoch": 0.0004445, "grad_norm": 3.1113669872283936, "learning_rate": 4.444e-05, "loss": 3.9844, "step": 4445 }, { "epoch": 0.0004446, "grad_norm": 3.1066434383392334, "learning_rate": 4.445e-05, "loss": 3.8262, "step": 4446 }, { "epoch": 0.0004447, "grad_norm": 3.656830310821533, "learning_rate": 4.446e-05, "loss": 4.2637, "step": 4447 }, { "epoch": 0.0004448, "grad_norm": 3.068669557571411, "learning_rate": 4.4470000000000006e-05, "loss": 3.584, "step": 4448 }, { "epoch": 0.0004449, "grad_norm": 3.044008493423462, "learning_rate": 4.448e-05, "loss": 3.6699, "step": 4449 }, { "epoch": 0.000445, "grad_norm": 3.000279664993286, "learning_rate": 4.4489999999999996e-05, "loss": 3.8613, "step": 4450 }, { "epoch": 0.0004451, "grad_norm": 3.259915590286255, "learning_rate": 4.4500000000000004e-05, "loss": 3.6367, "step": 4451 }, { "epoch": 0.0004452, "grad_norm": 3.252265691757202, "learning_rate": 4.451e-05, "loss": 3.8477, "step": 4452 }, { "epoch": 0.0004453, "grad_norm": 3.418731927871704, "learning_rate": 4.452e-05, "loss": 3.8105, "step": 4453 }, { "epoch": 0.0004454, "grad_norm": 3.133882522583008, "learning_rate": 4.453e-05, "loss": 3.4922, "step": 4454 }, { "epoch": 0.0004455, "grad_norm": 3.3016397953033447, "learning_rate": 4.454e-05, "loss": 3.8418, "step": 4455 }, { "epoch": 0.0004456, "grad_norm": 3.1321775913238525, "learning_rate": 4.455e-05, "loss": 3.5879, "step": 4456 }, { "epoch": 0.0004457, "grad_norm": 3.583219289779663, "learning_rate": 4.456e-05, "loss": 4.0254, "step": 4457 }, { "epoch": 0.0004458, "grad_norm": 3.492183208465576, "learning_rate": 4.457e-05, "loss": 4.123, "step": 4458 }, { "epoch": 0.0004459, "grad_norm": 4.111929893493652, "learning_rate": 4.4580000000000003e-05, "loss": 3.8516, "step": 4459 }, { "epoch": 0.000446, "grad_norm": 3.3053553104400635, "learning_rate": 4.459e-05, "loss": 3.8984, "step": 4460 }, { "epoch": 0.0004461, "grad_norm": 3.4948720932006836, "learning_rate": 4.46e-05, "loss": 3.9883, "step": 4461 }, { "epoch": 0.0004462, "grad_norm": 3.270261287689209, "learning_rate": 4.461e-05, "loss": 3.7676, "step": 4462 }, { "epoch": 0.0004463, "grad_norm": 3.1527910232543945, "learning_rate": 4.462e-05, "loss": 3.7227, "step": 4463 }, { "epoch": 0.0004464, "grad_norm": 3.2038075923919678, "learning_rate": 4.463e-05, "loss": 4.0039, "step": 4464 }, { "epoch": 0.0004465, "grad_norm": 3.2147271633148193, "learning_rate": 4.4640000000000006e-05, "loss": 4.0938, "step": 4465 }, { "epoch": 0.0004466, "grad_norm": 3.7212395668029785, "learning_rate": 4.465e-05, "loss": 3.8359, "step": 4466 }, { "epoch": 0.0004467, "grad_norm": 3.404538154602051, "learning_rate": 4.4659999999999996e-05, "loss": 3.6816, "step": 4467 }, { "epoch": 0.0004468, "grad_norm": 3.769585609436035, "learning_rate": 4.4670000000000004e-05, "loss": 4.1016, "step": 4468 }, { "epoch": 0.0004469, "grad_norm": 3.904418468475342, "learning_rate": 4.468e-05, "loss": 3.7363, "step": 4469 }, { "epoch": 0.000447, "grad_norm": 2.9345638751983643, "learning_rate": 4.469e-05, "loss": 3.4375, "step": 4470 }, { "epoch": 0.0004471, "grad_norm": 3.0536932945251465, "learning_rate": 4.47e-05, "loss": 3.502, "step": 4471 }, { "epoch": 0.0004472, "grad_norm": 3.2189676761627197, "learning_rate": 4.471e-05, "loss": 3.7051, "step": 4472 }, { "epoch": 0.0004473, "grad_norm": 3.041144847869873, "learning_rate": 4.472e-05, "loss": 3.5918, "step": 4473 }, { "epoch": 0.0004474, "grad_norm": 5.224328517913818, "learning_rate": 4.473e-05, "loss": 4.123, "step": 4474 }, { "epoch": 0.0004475, "grad_norm": 3.4222841262817383, "learning_rate": 4.474e-05, "loss": 3.8105, "step": 4475 }, { "epoch": 0.0004476, "grad_norm": 3.5500681400299072, "learning_rate": 4.4750000000000004e-05, "loss": 3.625, "step": 4476 }, { "epoch": 0.0004477, "grad_norm": 2.906442880630493, "learning_rate": 4.476e-05, "loss": 3.3867, "step": 4477 }, { "epoch": 0.0004478, "grad_norm": 3.48970627784729, "learning_rate": 4.477e-05, "loss": 3.8203, "step": 4478 }, { "epoch": 0.0004479, "grad_norm": 2.986858367919922, "learning_rate": 4.478e-05, "loss": 3.6133, "step": 4479 }, { "epoch": 0.000448, "grad_norm": 3.2639992237091064, "learning_rate": 4.479e-05, "loss": 3.8301, "step": 4480 }, { "epoch": 0.0004481, "grad_norm": 3.2185451984405518, "learning_rate": 4.48e-05, "loss": 3.7285, "step": 4481 }, { "epoch": 0.0004482, "grad_norm": 3.192749261856079, "learning_rate": 4.481e-05, "loss": 3.9062, "step": 4482 }, { "epoch": 0.0004483, "grad_norm": 3.0831542015075684, "learning_rate": 4.482e-05, "loss": 3.5645, "step": 4483 }, { "epoch": 0.0004484, "grad_norm": 3.139134168624878, "learning_rate": 4.4829999999999996e-05, "loss": 3.4492, "step": 4484 }, { "epoch": 0.0004485, "grad_norm": 3.0204544067382812, "learning_rate": 4.4840000000000004e-05, "loss": 3.6367, "step": 4485 }, { "epoch": 0.0004486, "grad_norm": 3.199981212615967, "learning_rate": 4.485e-05, "loss": 3.6484, "step": 4486 }, { "epoch": 0.0004487, "grad_norm": 2.892751932144165, "learning_rate": 4.486e-05, "loss": 3.2754, "step": 4487 }, { "epoch": 0.0004488, "grad_norm": 3.2739310264587402, "learning_rate": 4.487e-05, "loss": 3.5781, "step": 4488 }, { "epoch": 0.0004489, "grad_norm": 4.334787845611572, "learning_rate": 4.488e-05, "loss": 3.9316, "step": 4489 }, { "epoch": 0.000449, "grad_norm": 3.109442710876465, "learning_rate": 4.489e-05, "loss": 3.5059, "step": 4490 }, { "epoch": 0.0004491, "grad_norm": 3.3843679428100586, "learning_rate": 4.49e-05, "loss": 3.6875, "step": 4491 }, { "epoch": 0.0004492, "grad_norm": 3.297579765319824, "learning_rate": 4.491e-05, "loss": 3.8867, "step": 4492 }, { "epoch": 0.0004493, "grad_norm": 2.823868989944458, "learning_rate": 4.4920000000000004e-05, "loss": 3.3418, "step": 4493 }, { "epoch": 0.0004494, "grad_norm": 3.3911476135253906, "learning_rate": 4.493e-05, "loss": 3.8711, "step": 4494 }, { "epoch": 0.0004495, "grad_norm": 3.0930967330932617, "learning_rate": 4.494e-05, "loss": 3.3574, "step": 4495 }, { "epoch": 0.0004496, "grad_norm": 2.832514524459839, "learning_rate": 4.495e-05, "loss": 3.4941, "step": 4496 }, { "epoch": 0.0004497, "grad_norm": 3.0259110927581787, "learning_rate": 4.496e-05, "loss": 3.4414, "step": 4497 }, { "epoch": 0.0004498, "grad_norm": 3.1185426712036133, "learning_rate": 4.497e-05, "loss": 3.5137, "step": 4498 }, { "epoch": 0.0004499, "grad_norm": 2.888023614883423, "learning_rate": 4.498e-05, "loss": 3.4004, "step": 4499 }, { "epoch": 0.00045, "grad_norm": 4.178335666656494, "learning_rate": 4.499e-05, "loss": 4.3203, "step": 4500 }, { "epoch": 0.00045, "eval_loss": 0.4451688528060913, "eval_runtime": 366.6763, "eval_samples_per_second": 27.272, "eval_steps_per_second": 1.705, "step": 4500 }, { "epoch": 0.0004501, "grad_norm": 3.051055431365967, "learning_rate": 4.4999999999999996e-05, "loss": 3.5898, "step": 4501 }, { "epoch": 0.0004502, "grad_norm": 3.0838615894317627, "learning_rate": 4.5010000000000004e-05, "loss": 3.6953, "step": 4502 }, { "epoch": 0.0004503, "grad_norm": 3.187135934829712, "learning_rate": 4.502e-05, "loss": 3.3516, "step": 4503 }, { "epoch": 0.0004504, "grad_norm": 2.7712738513946533, "learning_rate": 4.503e-05, "loss": 3.4336, "step": 4504 }, { "epoch": 0.0004505, "grad_norm": 3.0009491443634033, "learning_rate": 4.504e-05, "loss": 3.7207, "step": 4505 }, { "epoch": 0.0004506, "grad_norm": 2.9187800884246826, "learning_rate": 4.505e-05, "loss": 3.4844, "step": 4506 }, { "epoch": 0.0004507, "grad_norm": 3.2779057025909424, "learning_rate": 4.506e-05, "loss": 3.5332, "step": 4507 }, { "epoch": 0.0004508, "grad_norm": 2.9774413108825684, "learning_rate": 4.507e-05, "loss": 3.5469, "step": 4508 }, { "epoch": 0.0004509, "grad_norm": 3.505708932876587, "learning_rate": 4.508e-05, "loss": 3.7676, "step": 4509 }, { "epoch": 0.000451, "grad_norm": 3.1741249561309814, "learning_rate": 4.5090000000000004e-05, "loss": 3.6777, "step": 4510 }, { "epoch": 0.0004511, "grad_norm": 4.4884419441223145, "learning_rate": 4.51e-05, "loss": 3.9199, "step": 4511 }, { "epoch": 0.0004512, "grad_norm": 3.0411839485168457, "learning_rate": 4.511e-05, "loss": 3.6016, "step": 4512 }, { "epoch": 0.0004513, "grad_norm": 3.2237861156463623, "learning_rate": 4.512e-05, "loss": 3.5723, "step": 4513 }, { "epoch": 0.0004514, "grad_norm": 3.0827009677886963, "learning_rate": 4.513e-05, "loss": 3.3809, "step": 4514 }, { "epoch": 0.0004515, "grad_norm": 3.0677378177642822, "learning_rate": 4.514e-05, "loss": 3.5801, "step": 4515 }, { "epoch": 0.0004516, "grad_norm": 3.4642274379730225, "learning_rate": 4.515e-05, "loss": 3.7129, "step": 4516 }, { "epoch": 0.0004517, "grad_norm": 4.53114652633667, "learning_rate": 4.516e-05, "loss": 4.5059, "step": 4517 }, { "epoch": 0.0004518, "grad_norm": 3.5981545448303223, "learning_rate": 4.5169999999999996e-05, "loss": 3.8613, "step": 4518 }, { "epoch": 0.0004519, "grad_norm": 3.0950846672058105, "learning_rate": 4.5180000000000004e-05, "loss": 3.5059, "step": 4519 }, { "epoch": 0.000452, "grad_norm": 3.107102632522583, "learning_rate": 4.519e-05, "loss": 3.5957, "step": 4520 }, { "epoch": 0.0004521, "grad_norm": 2.9538638591766357, "learning_rate": 4.52e-05, "loss": 3.5039, "step": 4521 }, { "epoch": 0.0004522, "grad_norm": 3.0675628185272217, "learning_rate": 4.521e-05, "loss": 3.6055, "step": 4522 }, { "epoch": 0.0004523, "grad_norm": 3.297718048095703, "learning_rate": 4.522e-05, "loss": 3.8418, "step": 4523 }, { "epoch": 0.0004524, "grad_norm": 3.3756823539733887, "learning_rate": 4.523e-05, "loss": 3.6191, "step": 4524 }, { "epoch": 0.0004525, "grad_norm": 3.175271511077881, "learning_rate": 4.524e-05, "loss": 3.4473, "step": 4525 }, { "epoch": 0.0004526, "grad_norm": 3.2302279472351074, "learning_rate": 4.525e-05, "loss": 3.7246, "step": 4526 }, { "epoch": 0.0004527, "grad_norm": 2.968247890472412, "learning_rate": 4.5260000000000004e-05, "loss": 3.1855, "step": 4527 }, { "epoch": 0.0004528, "grad_norm": 3.002324342727661, "learning_rate": 4.527e-05, "loss": 3.1426, "step": 4528 }, { "epoch": 0.0004529, "grad_norm": 3.2600293159484863, "learning_rate": 4.528e-05, "loss": 3.5098, "step": 4529 }, { "epoch": 0.000453, "grad_norm": 2.757805824279785, "learning_rate": 4.529e-05, "loss": 3.3457, "step": 4530 }, { "epoch": 0.0004531, "grad_norm": 3.031299591064453, "learning_rate": 4.53e-05, "loss": 3.3828, "step": 4531 }, { "epoch": 0.0004532, "grad_norm": 2.977142095565796, "learning_rate": 4.531e-05, "loss": 3.3711, "step": 4532 }, { "epoch": 0.0004533, "grad_norm": 3.6636605262756348, "learning_rate": 4.532e-05, "loss": 3.9902, "step": 4533 }, { "epoch": 0.0004534, "grad_norm": 2.856250762939453, "learning_rate": 4.533e-05, "loss": 3.3066, "step": 4534 }, { "epoch": 0.0004535, "grad_norm": 2.974116802215576, "learning_rate": 4.5339999999999996e-05, "loss": 3.4355, "step": 4535 }, { "epoch": 0.0004536, "grad_norm": 3.3000264167785645, "learning_rate": 4.5350000000000005e-05, "loss": 3.7715, "step": 4536 }, { "epoch": 0.0004537, "grad_norm": 2.945329427719116, "learning_rate": 4.536e-05, "loss": 3.3438, "step": 4537 }, { "epoch": 0.0004538, "grad_norm": 3.0566506385803223, "learning_rate": 4.537e-05, "loss": 3.3242, "step": 4538 }, { "epoch": 0.0004539, "grad_norm": 3.1250433921813965, "learning_rate": 4.538e-05, "loss": 3.4961, "step": 4539 }, { "epoch": 0.000454, "grad_norm": 3.106018304824829, "learning_rate": 4.539e-05, "loss": 3.3184, "step": 4540 }, { "epoch": 0.0004541, "grad_norm": 3.060356855392456, "learning_rate": 4.54e-05, "loss": 3.4199, "step": 4541 }, { "epoch": 0.0004542, "grad_norm": 2.717092752456665, "learning_rate": 4.541e-05, "loss": 3.1895, "step": 4542 }, { "epoch": 0.0004543, "grad_norm": 2.9867141246795654, "learning_rate": 4.542e-05, "loss": 3.4023, "step": 4543 }, { "epoch": 0.0004544, "grad_norm": 3.0905447006225586, "learning_rate": 4.5430000000000004e-05, "loss": 3.6895, "step": 4544 }, { "epoch": 0.0004545, "grad_norm": 3.6126909255981445, "learning_rate": 4.544e-05, "loss": 3.8359, "step": 4545 }, { "epoch": 0.0004546, "grad_norm": 3.276954412460327, "learning_rate": 4.545e-05, "loss": 3.6406, "step": 4546 }, { "epoch": 0.0004547, "grad_norm": 2.991445779800415, "learning_rate": 4.546e-05, "loss": 3.3633, "step": 4547 }, { "epoch": 0.0004548, "grad_norm": 3.145341634750366, "learning_rate": 4.5470000000000003e-05, "loss": 3.4199, "step": 4548 }, { "epoch": 0.0004549, "grad_norm": 2.9541232585906982, "learning_rate": 4.548e-05, "loss": 3.4746, "step": 4549 }, { "epoch": 0.000455, "grad_norm": 3.2786691188812256, "learning_rate": 4.549e-05, "loss": 3.6445, "step": 4550 }, { "epoch": 0.0004551, "grad_norm": 3.6114628314971924, "learning_rate": 4.55e-05, "loss": 3.5859, "step": 4551 }, { "epoch": 0.0004552, "grad_norm": 3.343808889389038, "learning_rate": 4.5509999999999996e-05, "loss": 3.375, "step": 4552 }, { "epoch": 0.0004553, "grad_norm": 2.839305877685547, "learning_rate": 4.5520000000000005e-05, "loss": 3.1699, "step": 4553 }, { "epoch": 0.0004554, "grad_norm": 3.1923487186431885, "learning_rate": 4.553e-05, "loss": 3.418, "step": 4554 }, { "epoch": 0.0004555, "grad_norm": 3.1452622413635254, "learning_rate": 4.554e-05, "loss": 3.3965, "step": 4555 }, { "epoch": 0.0004556, "grad_norm": 3.1820600032806396, "learning_rate": 4.555e-05, "loss": 3.8164, "step": 4556 }, { "epoch": 0.0004557, "grad_norm": 2.9150209426879883, "learning_rate": 4.556e-05, "loss": 3.082, "step": 4557 }, { "epoch": 0.0004558, "grad_norm": 3.2666032314300537, "learning_rate": 4.557e-05, "loss": 3.2793, "step": 4558 }, { "epoch": 0.0004559, "grad_norm": 3.010118007659912, "learning_rate": 4.558e-05, "loss": 3.0996, "step": 4559 }, { "epoch": 0.000456, "grad_norm": 3.0562148094177246, "learning_rate": 4.559e-05, "loss": 3.2695, "step": 4560 }, { "epoch": 0.0004561, "grad_norm": 2.8368160724639893, "learning_rate": 4.5600000000000004e-05, "loss": 3.1855, "step": 4561 }, { "epoch": 0.0004562, "grad_norm": 3.2904069423675537, "learning_rate": 4.561e-05, "loss": 3.4961, "step": 4562 }, { "epoch": 0.0004563, "grad_norm": 3.5901269912719727, "learning_rate": 4.562e-05, "loss": 3.6406, "step": 4563 }, { "epoch": 0.0004564, "grad_norm": 3.0019993782043457, "learning_rate": 4.563e-05, "loss": 3.084, "step": 4564 }, { "epoch": 0.0004565, "grad_norm": 3.15317964553833, "learning_rate": 4.5640000000000003e-05, "loss": 3.3828, "step": 4565 }, { "epoch": 0.0004566, "grad_norm": 3.3389527797698975, "learning_rate": 4.565e-05, "loss": 4.0977, "step": 4566 }, { "epoch": 0.0004567, "grad_norm": 3.2171781063079834, "learning_rate": 4.566e-05, "loss": 3.4141, "step": 4567 }, { "epoch": 0.0004568, "grad_norm": 3.2432215213775635, "learning_rate": 4.567e-05, "loss": 3.5156, "step": 4568 }, { "epoch": 0.0004569, "grad_norm": 3.093804359436035, "learning_rate": 4.5679999999999996e-05, "loss": 3.1309, "step": 4569 }, { "epoch": 0.000457, "grad_norm": 3.2238101959228516, "learning_rate": 4.5690000000000005e-05, "loss": 3.4277, "step": 4570 }, { "epoch": 0.0004571, "grad_norm": 4.436655521392822, "learning_rate": 4.57e-05, "loss": 4.0605, "step": 4571 }, { "epoch": 0.0004572, "grad_norm": 3.251905679702759, "learning_rate": 4.571e-05, "loss": 3.3262, "step": 4572 }, { "epoch": 0.0004573, "grad_norm": 3.4289801120758057, "learning_rate": 4.572e-05, "loss": 3.5039, "step": 4573 }, { "epoch": 0.0004574, "grad_norm": 3.4249584674835205, "learning_rate": 4.573e-05, "loss": 3.543, "step": 4574 }, { "epoch": 0.0004575, "grad_norm": 3.4442009925842285, "learning_rate": 4.574e-05, "loss": 3.5898, "step": 4575 }, { "epoch": 0.0004576, "grad_norm": 3.499016284942627, "learning_rate": 4.575e-05, "loss": 3.582, "step": 4576 }, { "epoch": 0.0004577, "grad_norm": 3.343648672103882, "learning_rate": 4.576e-05, "loss": 3.1406, "step": 4577 }, { "epoch": 0.0004578, "grad_norm": 3.316096782684326, "learning_rate": 4.5770000000000004e-05, "loss": 3.4277, "step": 4578 }, { "epoch": 0.0004579, "grad_norm": 3.2093162536621094, "learning_rate": 4.578e-05, "loss": 3.2656, "step": 4579 }, { "epoch": 0.000458, "grad_norm": 3.33439040184021, "learning_rate": 4.579e-05, "loss": 3.3242, "step": 4580 }, { "epoch": 0.0004581, "grad_norm": 2.993375539779663, "learning_rate": 4.58e-05, "loss": 3.043, "step": 4581 }, { "epoch": 0.0004582, "grad_norm": 3.385023593902588, "learning_rate": 4.5810000000000004e-05, "loss": 3.793, "step": 4582 }, { "epoch": 0.0004583, "grad_norm": 2.992016553878784, "learning_rate": 4.582e-05, "loss": 3.1074, "step": 4583 }, { "epoch": 0.0004584, "grad_norm": 3.6605024337768555, "learning_rate": 4.583e-05, "loss": 3.6094, "step": 4584 }, { "epoch": 0.0004585, "grad_norm": 3.0461432933807373, "learning_rate": 4.584e-05, "loss": 3.1934, "step": 4585 }, { "epoch": 0.0004586, "grad_norm": 3.1141693592071533, "learning_rate": 4.5849999999999996e-05, "loss": 3.168, "step": 4586 }, { "epoch": 0.0004587, "grad_norm": 3.2451374530792236, "learning_rate": 4.5860000000000005e-05, "loss": 3.2695, "step": 4587 }, { "epoch": 0.0004588, "grad_norm": 3.1429286003112793, "learning_rate": 4.587e-05, "loss": 3.2891, "step": 4588 }, { "epoch": 0.0004589, "grad_norm": 3.271604299545288, "learning_rate": 4.588e-05, "loss": 3.3359, "step": 4589 }, { "epoch": 0.000459, "grad_norm": 3.0061817169189453, "learning_rate": 4.589e-05, "loss": 3.0625, "step": 4590 }, { "epoch": 0.0004591, "grad_norm": 4.043227672576904, "learning_rate": 4.59e-05, "loss": 3.9746, "step": 4591 }, { "epoch": 0.0004592, "grad_norm": 3.712244987487793, "learning_rate": 4.591e-05, "loss": 3.6172, "step": 4592 }, { "epoch": 0.0004593, "grad_norm": 2.774465799331665, "learning_rate": 4.592e-05, "loss": 2.9844, "step": 4593 }, { "epoch": 0.0004594, "grad_norm": 4.377650737762451, "learning_rate": 4.593e-05, "loss": 4.0684, "step": 4594 }, { "epoch": 0.0004595, "grad_norm": 3.2312381267547607, "learning_rate": 4.5940000000000004e-05, "loss": 3.3301, "step": 4595 }, { "epoch": 0.0004596, "grad_norm": 3.5101516246795654, "learning_rate": 4.595e-05, "loss": 3.5469, "step": 4596 }, { "epoch": 0.0004597, "grad_norm": 3.555341958999634, "learning_rate": 4.596e-05, "loss": 3.7266, "step": 4597 }, { "epoch": 0.0004598, "grad_norm": 3.5841064453125, "learning_rate": 4.597e-05, "loss": 3.6875, "step": 4598 }, { "epoch": 0.0004599, "grad_norm": 2.876518726348877, "learning_rate": 4.5980000000000004e-05, "loss": 3.2266, "step": 4599 }, { "epoch": 0.00046, "grad_norm": 2.9207773208618164, "learning_rate": 4.599e-05, "loss": 3.0312, "step": 4600 }, { "epoch": 0.0004601, "grad_norm": 3.206237554550171, "learning_rate": 4.6e-05, "loss": 3.0566, "step": 4601 }, { "epoch": 0.0004602, "grad_norm": 2.973820447921753, "learning_rate": 4.601e-05, "loss": 3.2773, "step": 4602 }, { "epoch": 0.0004603, "grad_norm": 2.7981374263763428, "learning_rate": 4.6019999999999996e-05, "loss": 2.9805, "step": 4603 }, { "epoch": 0.0004604, "grad_norm": 3.1380207538604736, "learning_rate": 4.6030000000000005e-05, "loss": 3.1836, "step": 4604 }, { "epoch": 0.0004605, "grad_norm": 3.0024547576904297, "learning_rate": 4.604e-05, "loss": 3.0859, "step": 4605 }, { "epoch": 0.0004606, "grad_norm": 3.2511374950408936, "learning_rate": 4.605e-05, "loss": 3.3848, "step": 4606 }, { "epoch": 0.0004607, "grad_norm": 2.9090166091918945, "learning_rate": 4.606e-05, "loss": 3.25, "step": 4607 }, { "epoch": 0.0004608, "grad_norm": 4.24448299407959, "learning_rate": 4.607e-05, "loss": 3.7969, "step": 4608 }, { "epoch": 0.0004609, "grad_norm": 3.508185863494873, "learning_rate": 4.608e-05, "loss": 3.2617, "step": 4609 }, { "epoch": 0.000461, "grad_norm": 3.2529733180999756, "learning_rate": 4.609e-05, "loss": 3.5664, "step": 4610 }, { "epoch": 0.0004611, "grad_norm": 2.787061929702759, "learning_rate": 4.61e-05, "loss": 3.1211, "step": 4611 }, { "epoch": 0.0004612, "grad_norm": 2.7569823265075684, "learning_rate": 4.6110000000000004e-05, "loss": 2.998, "step": 4612 }, { "epoch": 0.0004613, "grad_norm": 2.947113275527954, "learning_rate": 4.612e-05, "loss": 3.2363, "step": 4613 }, { "epoch": 0.0004614, "grad_norm": 3.999605178833008, "learning_rate": 4.613e-05, "loss": 3.9375, "step": 4614 }, { "epoch": 0.0004615, "grad_norm": 3.12037992477417, "learning_rate": 4.614e-05, "loss": 3.2266, "step": 4615 }, { "epoch": 0.0004616, "grad_norm": 3.5467910766601562, "learning_rate": 4.6150000000000004e-05, "loss": 3.4355, "step": 4616 }, { "epoch": 0.0004617, "grad_norm": 2.7908236980438232, "learning_rate": 4.616e-05, "loss": 2.8867, "step": 4617 }, { "epoch": 0.0004618, "grad_norm": 2.961948871612549, "learning_rate": 4.617e-05, "loss": 3.1738, "step": 4618 }, { "epoch": 0.0004619, "grad_norm": 3.0067481994628906, "learning_rate": 4.618e-05, "loss": 3.2754, "step": 4619 }, { "epoch": 0.000462, "grad_norm": 3.3735923767089844, "learning_rate": 4.6189999999999997e-05, "loss": 3.7031, "step": 4620 }, { "epoch": 0.0004621, "grad_norm": 3.0538058280944824, "learning_rate": 4.6200000000000005e-05, "loss": 3.1504, "step": 4621 }, { "epoch": 0.0004622, "grad_norm": 3.0310232639312744, "learning_rate": 4.621e-05, "loss": 3.0195, "step": 4622 }, { "epoch": 0.0004623, "grad_norm": 3.4910993576049805, "learning_rate": 4.622e-05, "loss": 3.4062, "step": 4623 }, { "epoch": 0.0004624, "grad_norm": 3.0413780212402344, "learning_rate": 4.623e-05, "loss": 3.0352, "step": 4624 }, { "epoch": 0.0004625, "grad_norm": 3.210792303085327, "learning_rate": 4.624e-05, "loss": 3.3223, "step": 4625 }, { "epoch": 0.0004626, "grad_norm": 2.7651255130767822, "learning_rate": 4.625e-05, "loss": 3.0566, "step": 4626 }, { "epoch": 0.0004627, "grad_norm": 2.960127353668213, "learning_rate": 4.626e-05, "loss": 3.4531, "step": 4627 }, { "epoch": 0.0004628, "grad_norm": 3.0130105018615723, "learning_rate": 4.627e-05, "loss": 3.3535, "step": 4628 }, { "epoch": 0.0004629, "grad_norm": 3.209527015686035, "learning_rate": 4.6280000000000004e-05, "loss": 3.1309, "step": 4629 }, { "epoch": 0.000463, "grad_norm": 3.054307222366333, "learning_rate": 4.629e-05, "loss": 3.1035, "step": 4630 }, { "epoch": 0.0004631, "grad_norm": 2.9722907543182373, "learning_rate": 4.63e-05, "loss": 3.3711, "step": 4631 }, { "epoch": 0.0004632, "grad_norm": 2.9601128101348877, "learning_rate": 4.631e-05, "loss": 3.2949, "step": 4632 }, { "epoch": 0.0004633, "grad_norm": 3.375668525695801, "learning_rate": 4.6320000000000004e-05, "loss": 3.3848, "step": 4633 }, { "epoch": 0.0004634, "grad_norm": 3.0416009426116943, "learning_rate": 4.633e-05, "loss": 3.1582, "step": 4634 }, { "epoch": 0.0004635, "grad_norm": 3.2459537982940674, "learning_rate": 4.634e-05, "loss": 3.1719, "step": 4635 }, { "epoch": 0.0004636, "grad_norm": 3.058116912841797, "learning_rate": 4.635e-05, "loss": 3.2422, "step": 4636 }, { "epoch": 0.0004637, "grad_norm": 3.192560911178589, "learning_rate": 4.6359999999999997e-05, "loss": 3.168, "step": 4637 }, { "epoch": 0.0004638, "grad_norm": 3.726327657699585, "learning_rate": 4.6370000000000005e-05, "loss": 3.3906, "step": 4638 }, { "epoch": 0.0004639, "grad_norm": 2.987051248550415, "learning_rate": 4.638e-05, "loss": 3.1621, "step": 4639 }, { "epoch": 0.000464, "grad_norm": 3.231541872024536, "learning_rate": 4.639e-05, "loss": 3.082, "step": 4640 }, { "epoch": 0.0004641, "grad_norm": 5.895324230194092, "learning_rate": 4.64e-05, "loss": 4.2188, "step": 4641 }, { "epoch": 0.0004642, "grad_norm": 3.2076079845428467, "learning_rate": 4.641e-05, "loss": 3.2344, "step": 4642 }, { "epoch": 0.0004643, "grad_norm": 2.8946638107299805, "learning_rate": 4.642e-05, "loss": 2.9375, "step": 4643 }, { "epoch": 0.0004644, "grad_norm": 3.2147016525268555, "learning_rate": 4.643e-05, "loss": 3.1953, "step": 4644 }, { "epoch": 0.0004645, "grad_norm": 2.7994742393493652, "learning_rate": 4.644e-05, "loss": 2.9316, "step": 4645 }, { "epoch": 0.0004646, "grad_norm": 2.975017547607422, "learning_rate": 4.6450000000000004e-05, "loss": 3.1816, "step": 4646 }, { "epoch": 0.0004647, "grad_norm": 3.3162500858306885, "learning_rate": 4.646e-05, "loss": 3.4473, "step": 4647 }, { "epoch": 0.0004648, "grad_norm": 3.134305715560913, "learning_rate": 4.647e-05, "loss": 3.293, "step": 4648 }, { "epoch": 0.0004649, "grad_norm": 3.971130847930908, "learning_rate": 4.648e-05, "loss": 3.377, "step": 4649 }, { "epoch": 0.000465, "grad_norm": 3.0658767223358154, "learning_rate": 4.6490000000000004e-05, "loss": 3.125, "step": 4650 }, { "epoch": 0.0004651, "grad_norm": 3.0393059253692627, "learning_rate": 4.65e-05, "loss": 3.082, "step": 4651 }, { "epoch": 0.0004652, "grad_norm": 2.809842348098755, "learning_rate": 4.651e-05, "loss": 3.1699, "step": 4652 }, { "epoch": 0.0004653, "grad_norm": 2.8788135051727295, "learning_rate": 4.652e-05, "loss": 3.041, "step": 4653 }, { "epoch": 0.0004654, "grad_norm": 2.979281425476074, "learning_rate": 4.653e-05, "loss": 3.2109, "step": 4654 }, { "epoch": 0.0004655, "grad_norm": 3.1272122859954834, "learning_rate": 4.6540000000000005e-05, "loss": 3.1035, "step": 4655 }, { "epoch": 0.0004656, "grad_norm": 2.9307401180267334, "learning_rate": 4.655e-05, "loss": 3.25, "step": 4656 }, { "epoch": 0.0004657, "grad_norm": 3.7151620388031006, "learning_rate": 4.656e-05, "loss": 3.4922, "step": 4657 }, { "epoch": 0.0004658, "grad_norm": 3.6047842502593994, "learning_rate": 4.657e-05, "loss": 3.8066, "step": 4658 }, { "epoch": 0.0004659, "grad_norm": 3.0247373580932617, "learning_rate": 4.658e-05, "loss": 3.0312, "step": 4659 }, { "epoch": 0.000466, "grad_norm": 3.2321155071258545, "learning_rate": 4.659e-05, "loss": 3.4688, "step": 4660 }, { "epoch": 0.0004661, "grad_norm": 3.057260513305664, "learning_rate": 4.66e-05, "loss": 3.0059, "step": 4661 }, { "epoch": 0.0004662, "grad_norm": 3.03935170173645, "learning_rate": 4.661e-05, "loss": 3.3047, "step": 4662 }, { "epoch": 0.0004663, "grad_norm": 3.991436004638672, "learning_rate": 4.6620000000000004e-05, "loss": 3.5938, "step": 4663 }, { "epoch": 0.0004664, "grad_norm": 3.4941153526306152, "learning_rate": 4.663e-05, "loss": 3.2715, "step": 4664 }, { "epoch": 0.0004665, "grad_norm": 3.002227306365967, "learning_rate": 4.664e-05, "loss": 3.084, "step": 4665 }, { "epoch": 0.0004666, "grad_norm": 4.479686260223389, "learning_rate": 4.665e-05, "loss": 3.7383, "step": 4666 }, { "epoch": 0.0004667, "grad_norm": 3.00398588180542, "learning_rate": 4.6660000000000004e-05, "loss": 3.2207, "step": 4667 }, { "epoch": 0.0004668, "grad_norm": 3.7229092121124268, "learning_rate": 4.667e-05, "loss": 3.6133, "step": 4668 }, { "epoch": 0.0004669, "grad_norm": 2.708282709121704, "learning_rate": 4.668e-05, "loss": 2.8398, "step": 4669 }, { "epoch": 0.000467, "grad_norm": 2.930645227432251, "learning_rate": 4.669e-05, "loss": 2.8906, "step": 4670 }, { "epoch": 0.0004671, "grad_norm": 3.122342586517334, "learning_rate": 4.67e-05, "loss": 3.2539, "step": 4671 }, { "epoch": 0.0004672, "grad_norm": 3.5107548236846924, "learning_rate": 4.6710000000000005e-05, "loss": 3.8184, "step": 4672 }, { "epoch": 0.0004673, "grad_norm": 3.092520236968994, "learning_rate": 4.672e-05, "loss": 3.2734, "step": 4673 }, { "epoch": 0.0004674, "grad_norm": 3.2515995502471924, "learning_rate": 4.673e-05, "loss": 3.4473, "step": 4674 }, { "epoch": 0.0004675, "grad_norm": 3.0738046169281006, "learning_rate": 4.674e-05, "loss": 2.9746, "step": 4675 }, { "epoch": 0.0004676, "grad_norm": 2.728998899459839, "learning_rate": 4.675e-05, "loss": 2.8438, "step": 4676 }, { "epoch": 0.0004677, "grad_norm": 2.9970498085021973, "learning_rate": 4.676e-05, "loss": 3.1133, "step": 4677 }, { "epoch": 0.0004678, "grad_norm": 3.525514841079712, "learning_rate": 4.677e-05, "loss": 3.123, "step": 4678 }, { "epoch": 0.0004679, "grad_norm": 3.8426830768585205, "learning_rate": 4.678e-05, "loss": 3.7891, "step": 4679 }, { "epoch": 0.000468, "grad_norm": 4.1395392417907715, "learning_rate": 4.6790000000000004e-05, "loss": 3.9746, "step": 4680 }, { "epoch": 0.0004681, "grad_norm": 3.1620771884918213, "learning_rate": 4.68e-05, "loss": 3.2266, "step": 4681 }, { "epoch": 0.0004682, "grad_norm": 3.2321674823760986, "learning_rate": 4.681e-05, "loss": 3.4336, "step": 4682 }, { "epoch": 0.0004683, "grad_norm": 3.2428066730499268, "learning_rate": 4.682e-05, "loss": 3.0859, "step": 4683 }, { "epoch": 0.0004684, "grad_norm": 3.541463851928711, "learning_rate": 4.6830000000000004e-05, "loss": 3.4043, "step": 4684 }, { "epoch": 0.0004685, "grad_norm": 3.0401740074157715, "learning_rate": 4.684e-05, "loss": 3.3809, "step": 4685 }, { "epoch": 0.0004686, "grad_norm": 3.0577149391174316, "learning_rate": 4.685e-05, "loss": 2.9707, "step": 4686 }, { "epoch": 0.0004687, "grad_norm": 3.0174202919006348, "learning_rate": 4.686e-05, "loss": 3.0703, "step": 4687 }, { "epoch": 0.0004688, "grad_norm": 3.5271096229553223, "learning_rate": 4.687e-05, "loss": 4.0254, "step": 4688 }, { "epoch": 0.0004689, "grad_norm": 3.2100443840026855, "learning_rate": 4.6880000000000005e-05, "loss": 3.2441, "step": 4689 }, { "epoch": 0.000469, "grad_norm": 3.0281200408935547, "learning_rate": 4.689000000000001e-05, "loss": 2.9883, "step": 4690 }, { "epoch": 0.0004691, "grad_norm": 2.854539394378662, "learning_rate": 4.69e-05, "loss": 2.9863, "step": 4691 }, { "epoch": 0.0004692, "grad_norm": 3.117666006088257, "learning_rate": 4.691e-05, "loss": 3.2031, "step": 4692 }, { "epoch": 0.0004693, "grad_norm": 2.859088182449341, "learning_rate": 4.6920000000000005e-05, "loss": 3.0566, "step": 4693 }, { "epoch": 0.0004694, "grad_norm": 3.27419114112854, "learning_rate": 4.693e-05, "loss": 3.2344, "step": 4694 }, { "epoch": 0.0004695, "grad_norm": 3.1794731616973877, "learning_rate": 4.694e-05, "loss": 3.4609, "step": 4695 }, { "epoch": 0.0004696, "grad_norm": 3.5857043266296387, "learning_rate": 4.695e-05, "loss": 3.4336, "step": 4696 }, { "epoch": 0.0004697, "grad_norm": 3.222153663635254, "learning_rate": 4.696e-05, "loss": 3.0293, "step": 4697 }, { "epoch": 0.0004698, "grad_norm": 3.0103659629821777, "learning_rate": 4.697e-05, "loss": 3.0566, "step": 4698 }, { "epoch": 0.0004699, "grad_norm": 2.826457977294922, "learning_rate": 4.698e-05, "loss": 2.9824, "step": 4699 }, { "epoch": 0.00047, "grad_norm": 3.030729055404663, "learning_rate": 4.699e-05, "loss": 2.9238, "step": 4700 }, { "epoch": 0.0004701, "grad_norm": 3.197434663772583, "learning_rate": 4.7e-05, "loss": 3.166, "step": 4701 }, { "epoch": 0.0004702, "grad_norm": 2.8865232467651367, "learning_rate": 4.701e-05, "loss": 2.7832, "step": 4702 }, { "epoch": 0.0004703, "grad_norm": 3.46136212348938, "learning_rate": 4.702e-05, "loss": 3.0547, "step": 4703 }, { "epoch": 0.0004704, "grad_norm": 2.8729143142700195, "learning_rate": 4.7029999999999995e-05, "loss": 2.832, "step": 4704 }, { "epoch": 0.0004705, "grad_norm": 3.744364023208618, "learning_rate": 4.704e-05, "loss": 3.5312, "step": 4705 }, { "epoch": 0.0004706, "grad_norm": 3.224149703979492, "learning_rate": 4.7050000000000005e-05, "loss": 3.3848, "step": 4706 }, { "epoch": 0.0004707, "grad_norm": 3.337146282196045, "learning_rate": 4.706000000000001e-05, "loss": 3.3203, "step": 4707 }, { "epoch": 0.0004708, "grad_norm": 3.0690548419952393, "learning_rate": 4.7069999999999995e-05, "loss": 3.1016, "step": 4708 }, { "epoch": 0.0004709, "grad_norm": 3.608992099761963, "learning_rate": 4.708e-05, "loss": 3.3984, "step": 4709 }, { "epoch": 0.000471, "grad_norm": 3.276923179626465, "learning_rate": 4.7090000000000005e-05, "loss": 3.043, "step": 4710 }, { "epoch": 0.0004711, "grad_norm": 3.19512677192688, "learning_rate": 4.71e-05, "loss": 2.9551, "step": 4711 }, { "epoch": 0.0004712, "grad_norm": 3.2399344444274902, "learning_rate": 4.711e-05, "loss": 3.2617, "step": 4712 }, { "epoch": 0.0004713, "grad_norm": 3.0753676891326904, "learning_rate": 4.712e-05, "loss": 2.7559, "step": 4713 }, { "epoch": 0.0004714, "grad_norm": 3.1295502185821533, "learning_rate": 4.713e-05, "loss": 2.9844, "step": 4714 }, { "epoch": 0.0004715, "grad_norm": 2.8155367374420166, "learning_rate": 4.714e-05, "loss": 2.9805, "step": 4715 }, { "epoch": 0.0004716, "grad_norm": 3.0161681175231934, "learning_rate": 4.715e-05, "loss": 3.0117, "step": 4716 }, { "epoch": 0.0004717, "grad_norm": 3.40616512298584, "learning_rate": 4.716e-05, "loss": 3.1465, "step": 4717 }, { "epoch": 0.0004718, "grad_norm": 4.066618919372559, "learning_rate": 4.717e-05, "loss": 3.2051, "step": 4718 }, { "epoch": 0.0004719, "grad_norm": 3.3159821033477783, "learning_rate": 4.718e-05, "loss": 3.1953, "step": 4719 }, { "epoch": 0.000472, "grad_norm": 3.789193630218506, "learning_rate": 4.719e-05, "loss": 3.5, "step": 4720 }, { "epoch": 0.0004721, "grad_norm": 3.634530782699585, "learning_rate": 4.7199999999999995e-05, "loss": 3.5605, "step": 4721 }, { "epoch": 0.0004722, "grad_norm": 4.3847174644470215, "learning_rate": 4.721e-05, "loss": 3.6797, "step": 4722 }, { "epoch": 0.0004723, "grad_norm": 3.048835039138794, "learning_rate": 4.7220000000000005e-05, "loss": 2.7344, "step": 4723 }, { "epoch": 0.0004724, "grad_norm": 4.160604476928711, "learning_rate": 4.723000000000001e-05, "loss": 3.8555, "step": 4724 }, { "epoch": 0.0004725, "grad_norm": 2.919438362121582, "learning_rate": 4.7239999999999995e-05, "loss": 2.9492, "step": 4725 }, { "epoch": 0.0004726, "grad_norm": 3.4969019889831543, "learning_rate": 4.725e-05, "loss": 3.627, "step": 4726 }, { "epoch": 0.0004727, "grad_norm": 3.182495594024658, "learning_rate": 4.7260000000000005e-05, "loss": 3.2305, "step": 4727 }, { "epoch": 0.0004728, "grad_norm": 3.376863718032837, "learning_rate": 4.727e-05, "loss": 3.3809, "step": 4728 }, { "epoch": 0.0004729, "grad_norm": 3.3285233974456787, "learning_rate": 4.728e-05, "loss": 3.168, "step": 4729 }, { "epoch": 0.000473, "grad_norm": 3.3870933055877686, "learning_rate": 4.729e-05, "loss": 3.1699, "step": 4730 }, { "epoch": 0.0004731, "grad_norm": 3.5595250129699707, "learning_rate": 4.73e-05, "loss": 3.4785, "step": 4731 }, { "epoch": 0.0004732, "grad_norm": 3.281703472137451, "learning_rate": 4.731e-05, "loss": 3.1582, "step": 4732 }, { "epoch": 0.0004733, "grad_norm": 3.1851179599761963, "learning_rate": 4.732e-05, "loss": 3.4902, "step": 4733 }, { "epoch": 0.0004734, "grad_norm": 3.014247417449951, "learning_rate": 4.733e-05, "loss": 3.0098, "step": 4734 }, { "epoch": 0.0004735, "grad_norm": 3.262362003326416, "learning_rate": 4.734e-05, "loss": 3.0684, "step": 4735 }, { "epoch": 0.0004736, "grad_norm": 3.127680778503418, "learning_rate": 4.735e-05, "loss": 3.0156, "step": 4736 }, { "epoch": 0.0004737, "grad_norm": 3.2958037853240967, "learning_rate": 4.736e-05, "loss": 3.1875, "step": 4737 }, { "epoch": 0.0004738, "grad_norm": 3.8011887073516846, "learning_rate": 4.7369999999999995e-05, "loss": 3.8086, "step": 4738 }, { "epoch": 0.0004739, "grad_norm": 3.099651575088501, "learning_rate": 4.738e-05, "loss": 3.1074, "step": 4739 }, { "epoch": 0.000474, "grad_norm": 6.103170394897461, "learning_rate": 4.7390000000000005e-05, "loss": 3.5566, "step": 4740 }, { "epoch": 0.0004741, "grad_norm": 3.2698004245758057, "learning_rate": 4.740000000000001e-05, "loss": 3.209, "step": 4741 }, { "epoch": 0.0004742, "grad_norm": 4.685603618621826, "learning_rate": 4.7409999999999995e-05, "loss": 3.834, "step": 4742 }, { "epoch": 0.0004743, "grad_norm": 3.5504651069641113, "learning_rate": 4.742e-05, "loss": 3.0801, "step": 4743 }, { "epoch": 0.0004744, "grad_norm": 3.8041789531707764, "learning_rate": 4.7430000000000005e-05, "loss": 3.2266, "step": 4744 }, { "epoch": 0.0004745, "grad_norm": 3.056103229522705, "learning_rate": 4.744e-05, "loss": 3.1484, "step": 4745 }, { "epoch": 0.0004746, "grad_norm": 8.014861106872559, "learning_rate": 4.745e-05, "loss": 4.6133, "step": 4746 }, { "epoch": 0.0004747, "grad_norm": 3.616004705429077, "learning_rate": 4.746e-05, "loss": 3.0293, "step": 4747 }, { "epoch": 0.0004748, "grad_norm": 3.142517328262329, "learning_rate": 4.747e-05, "loss": 2.8223, "step": 4748 }, { "epoch": 0.0004749, "grad_norm": 3.005836009979248, "learning_rate": 4.748e-05, "loss": 2.875, "step": 4749 }, { "epoch": 0.000475, "grad_norm": 3.232628345489502, "learning_rate": 4.749e-05, "loss": 3.1328, "step": 4750 }, { "epoch": 0.0004751, "grad_norm": 3.698164224624634, "learning_rate": 4.75e-05, "loss": 3.4082, "step": 4751 }, { "epoch": 0.0004752, "grad_norm": 3.4597959518432617, "learning_rate": 4.751e-05, "loss": 2.9648, "step": 4752 }, { "epoch": 0.0004753, "grad_norm": 3.0004732608795166, "learning_rate": 4.752e-05, "loss": 2.9512, "step": 4753 }, { "epoch": 0.0004754, "grad_norm": 3.116643190383911, "learning_rate": 4.753e-05, "loss": 3.125, "step": 4754 }, { "epoch": 0.0004755, "grad_norm": 2.946709632873535, "learning_rate": 4.7539999999999995e-05, "loss": 2.9941, "step": 4755 }, { "epoch": 0.0004756, "grad_norm": 3.6574668884277344, "learning_rate": 4.755e-05, "loss": 2.9863, "step": 4756 }, { "epoch": 0.0004757, "grad_norm": 3.235548257827759, "learning_rate": 4.7560000000000005e-05, "loss": 3.1328, "step": 4757 }, { "epoch": 0.0004758, "grad_norm": 3.2428536415100098, "learning_rate": 4.757000000000001e-05, "loss": 3.0371, "step": 4758 }, { "epoch": 0.0004759, "grad_norm": 3.2637336254119873, "learning_rate": 4.7579999999999995e-05, "loss": 2.9824, "step": 4759 }, { "epoch": 0.000476, "grad_norm": 3.017577648162842, "learning_rate": 4.7590000000000003e-05, "loss": 2.9102, "step": 4760 }, { "epoch": 0.0004761, "grad_norm": 2.9592535495758057, "learning_rate": 4.7600000000000005e-05, "loss": 2.8984, "step": 4761 }, { "epoch": 0.0004762, "grad_norm": 3.0379645824432373, "learning_rate": 4.761e-05, "loss": 2.9883, "step": 4762 }, { "epoch": 0.0004763, "grad_norm": 2.920619487762451, "learning_rate": 4.762e-05, "loss": 2.877, "step": 4763 }, { "epoch": 0.0004764, "grad_norm": 3.3243861198425293, "learning_rate": 4.763e-05, "loss": 3.1328, "step": 4764 }, { "epoch": 0.0004765, "grad_norm": 3.1136486530303955, "learning_rate": 4.764e-05, "loss": 3.0898, "step": 4765 }, { "epoch": 0.0004766, "grad_norm": 3.096620559692383, "learning_rate": 4.765e-05, "loss": 3.0137, "step": 4766 }, { "epoch": 0.0004767, "grad_norm": 3.970536708831787, "learning_rate": 4.766e-05, "loss": 3.4961, "step": 4767 }, { "epoch": 0.0004768, "grad_norm": 3.1307883262634277, "learning_rate": 4.767e-05, "loss": 3.0723, "step": 4768 }, { "epoch": 0.0004769, "grad_norm": 3.087273597717285, "learning_rate": 4.768e-05, "loss": 2.9609, "step": 4769 }, { "epoch": 0.000477, "grad_norm": 3.0771291255950928, "learning_rate": 4.769e-05, "loss": 3.0195, "step": 4770 }, { "epoch": 0.0004771, "grad_norm": 4.540352821350098, "learning_rate": 4.77e-05, "loss": 3.5391, "step": 4771 }, { "epoch": 0.0004772, "grad_norm": 3.124443769454956, "learning_rate": 4.7709999999999996e-05, "loss": 2.9062, "step": 4772 }, { "epoch": 0.0004773, "grad_norm": 3.3153040409088135, "learning_rate": 4.772e-05, "loss": 3.1992, "step": 4773 }, { "epoch": 0.0004774, "grad_norm": 2.992619514465332, "learning_rate": 4.7730000000000005e-05, "loss": 3.0371, "step": 4774 }, { "epoch": 0.0004775, "grad_norm": 3.177959442138672, "learning_rate": 4.774000000000001e-05, "loss": 2.8359, "step": 4775 }, { "epoch": 0.0004776, "grad_norm": 3.9217638969421387, "learning_rate": 4.7749999999999995e-05, "loss": 3.7656, "step": 4776 }, { "epoch": 0.0004777, "grad_norm": 3.0828561782836914, "learning_rate": 4.7760000000000004e-05, "loss": 2.8262, "step": 4777 }, { "epoch": 0.0004778, "grad_norm": 3.270836114883423, "learning_rate": 4.7770000000000005e-05, "loss": 3.2812, "step": 4778 }, { "epoch": 0.0004779, "grad_norm": 3.1518502235412598, "learning_rate": 4.778e-05, "loss": 3.3164, "step": 4779 }, { "epoch": 0.000478, "grad_norm": 3.312922239303589, "learning_rate": 4.779e-05, "loss": 2.9551, "step": 4780 }, { "epoch": 0.0004781, "grad_norm": 3.543351650238037, "learning_rate": 4.78e-05, "loss": 3.2793, "step": 4781 }, { "epoch": 0.0004782, "grad_norm": 3.3184115886688232, "learning_rate": 4.781e-05, "loss": 3.1152, "step": 4782 }, { "epoch": 0.0004783, "grad_norm": 3.1950390338897705, "learning_rate": 4.782e-05, "loss": 3.1641, "step": 4783 }, { "epoch": 0.0004784, "grad_norm": 3.1578187942504883, "learning_rate": 4.783e-05, "loss": 2.9355, "step": 4784 }, { "epoch": 0.0004785, "grad_norm": 3.039491891860962, "learning_rate": 4.784e-05, "loss": 2.9434, "step": 4785 }, { "epoch": 0.0004786, "grad_norm": 3.837468385696411, "learning_rate": 4.785e-05, "loss": 3.4043, "step": 4786 }, { "epoch": 0.0004787, "grad_norm": 3.7416141033172607, "learning_rate": 4.786e-05, "loss": 3.4473, "step": 4787 }, { "epoch": 0.0004788, "grad_norm": 3.9664506912231445, "learning_rate": 4.787e-05, "loss": 3.5547, "step": 4788 }, { "epoch": 0.0004789, "grad_norm": 3.3748490810394287, "learning_rate": 4.7879999999999996e-05, "loss": 3.1758, "step": 4789 }, { "epoch": 0.000479, "grad_norm": 4.525698661804199, "learning_rate": 4.789e-05, "loss": 4.1914, "step": 4790 }, { "epoch": 0.0004791, "grad_norm": 4.87324857711792, "learning_rate": 4.7900000000000006e-05, "loss": 3.2266, "step": 4791 }, { "epoch": 0.0004792, "grad_norm": 3.3620924949645996, "learning_rate": 4.791000000000001e-05, "loss": 2.7246, "step": 4792 }, { "epoch": 0.0004793, "grad_norm": 3.5611748695373535, "learning_rate": 4.7919999999999995e-05, "loss": 3.0059, "step": 4793 }, { "epoch": 0.0004794, "grad_norm": 3.231464147567749, "learning_rate": 4.7930000000000004e-05, "loss": 2.873, "step": 4794 }, { "epoch": 0.0004795, "grad_norm": 3.62239146232605, "learning_rate": 4.7940000000000005e-05, "loss": 3.0742, "step": 4795 }, { "epoch": 0.0004796, "grad_norm": 4.050042629241943, "learning_rate": 4.795e-05, "loss": 3.2812, "step": 4796 }, { "epoch": 0.0004797, "grad_norm": 3.3177478313446045, "learning_rate": 4.796e-05, "loss": 3.2324, "step": 4797 }, { "epoch": 0.0004798, "grad_norm": 2.89870548248291, "learning_rate": 4.797e-05, "loss": 2.6289, "step": 4798 }, { "epoch": 0.0004799, "grad_norm": 3.0883748531341553, "learning_rate": 4.798e-05, "loss": 3.0996, "step": 4799 }, { "epoch": 0.00048, "grad_norm": 3.119446277618408, "learning_rate": 4.799e-05, "loss": 3.0547, "step": 4800 }, { "epoch": 0.0004801, "grad_norm": 3.7069942951202393, "learning_rate": 4.8e-05, "loss": 3.4316, "step": 4801 }, { "epoch": 0.0004802, "grad_norm": 3.0920073986053467, "learning_rate": 4.801e-05, "loss": 2.6758, "step": 4802 }, { "epoch": 0.0004803, "grad_norm": 3.1666815280914307, "learning_rate": 4.802e-05, "loss": 3.0488, "step": 4803 }, { "epoch": 0.0004804, "grad_norm": 3.1347286701202393, "learning_rate": 4.803e-05, "loss": 2.8145, "step": 4804 }, { "epoch": 0.0004805, "grad_norm": 3.2104568481445312, "learning_rate": 4.804e-05, "loss": 2.7539, "step": 4805 }, { "epoch": 0.0004806, "grad_norm": 3.0340535640716553, "learning_rate": 4.8049999999999996e-05, "loss": 3.3652, "step": 4806 }, { "epoch": 0.0004807, "grad_norm": 2.7663848400115967, "learning_rate": 4.806e-05, "loss": 2.6641, "step": 4807 }, { "epoch": 0.0004808, "grad_norm": 3.2019152641296387, "learning_rate": 4.8070000000000006e-05, "loss": 3.0938, "step": 4808 }, { "epoch": 0.0004809, "grad_norm": 3.152174234390259, "learning_rate": 4.808000000000001e-05, "loss": 3.3984, "step": 4809 }, { "epoch": 0.000481, "grad_norm": 2.9069466590881348, "learning_rate": 4.8089999999999995e-05, "loss": 2.8867, "step": 4810 }, { "epoch": 0.0004811, "grad_norm": 3.125547170639038, "learning_rate": 4.8100000000000004e-05, "loss": 3.1133, "step": 4811 }, { "epoch": 0.0004812, "grad_norm": 3.342869281768799, "learning_rate": 4.8110000000000005e-05, "loss": 3.6055, "step": 4812 }, { "epoch": 0.0004813, "grad_norm": 2.9983394145965576, "learning_rate": 4.812e-05, "loss": 2.709, "step": 4813 }, { "epoch": 0.0004814, "grad_norm": 3.7137179374694824, "learning_rate": 4.813e-05, "loss": 3.0508, "step": 4814 }, { "epoch": 0.0004815, "grad_norm": 2.8978734016418457, "learning_rate": 4.814e-05, "loss": 2.8105, "step": 4815 }, { "epoch": 0.0004816, "grad_norm": 4.878371238708496, "learning_rate": 4.815e-05, "loss": 3.4785, "step": 4816 }, { "epoch": 0.0004817, "grad_norm": 3.293475866317749, "learning_rate": 4.816e-05, "loss": 3.0273, "step": 4817 }, { "epoch": 0.0004818, "grad_norm": 3.1329174041748047, "learning_rate": 4.817e-05, "loss": 2.9824, "step": 4818 }, { "epoch": 0.0004819, "grad_norm": 3.2431700229644775, "learning_rate": 4.818e-05, "loss": 3.1797, "step": 4819 }, { "epoch": 0.000482, "grad_norm": 3.1588330268859863, "learning_rate": 4.819e-05, "loss": 2.959, "step": 4820 }, { "epoch": 0.0004821, "grad_norm": 3.1027257442474365, "learning_rate": 4.82e-05, "loss": 2.5977, "step": 4821 }, { "epoch": 0.0004822, "grad_norm": 3.254634380340576, "learning_rate": 4.821e-05, "loss": 2.9277, "step": 4822 }, { "epoch": 0.0004823, "grad_norm": 3.281937837600708, "learning_rate": 4.8219999999999996e-05, "loss": 3.2793, "step": 4823 }, { "epoch": 0.0004824, "grad_norm": 3.8145720958709717, "learning_rate": 4.823e-05, "loss": 3.6738, "step": 4824 }, { "epoch": 0.0004825, "grad_norm": 3.320831298828125, "learning_rate": 4.8240000000000006e-05, "loss": 3.2285, "step": 4825 }, { "epoch": 0.0004826, "grad_norm": 3.179290771484375, "learning_rate": 4.825000000000001e-05, "loss": 2.9688, "step": 4826 }, { "epoch": 0.0004827, "grad_norm": 3.494075298309326, "learning_rate": 4.8259999999999995e-05, "loss": 3.2168, "step": 4827 }, { "epoch": 0.0004828, "grad_norm": 3.5924880504608154, "learning_rate": 4.8270000000000004e-05, "loss": 3.3672, "step": 4828 }, { "epoch": 0.0004829, "grad_norm": 3.5426197052001953, "learning_rate": 4.8280000000000005e-05, "loss": 3.1777, "step": 4829 }, { "epoch": 0.000483, "grad_norm": 3.11126446723938, "learning_rate": 4.829e-05, "loss": 2.7422, "step": 4830 }, { "epoch": 0.0004831, "grad_norm": 3.393720865249634, "learning_rate": 4.83e-05, "loss": 3.2471, "step": 4831 }, { "epoch": 0.0004832, "grad_norm": 3.45908784866333, "learning_rate": 4.831e-05, "loss": 3.2266, "step": 4832 }, { "epoch": 0.0004833, "grad_norm": 4.491594314575195, "learning_rate": 4.832e-05, "loss": 3.2656, "step": 4833 }, { "epoch": 0.0004834, "grad_norm": 3.1031112670898438, "learning_rate": 4.833e-05, "loss": 2.7031, "step": 4834 }, { "epoch": 0.0004835, "grad_norm": 3.284872531890869, "learning_rate": 4.834e-05, "loss": 2.8926, "step": 4835 }, { "epoch": 0.0004836, "grad_norm": 3.9908602237701416, "learning_rate": 4.835e-05, "loss": 3.5566, "step": 4836 }, { "epoch": 0.0004837, "grad_norm": 3.1020312309265137, "learning_rate": 4.836e-05, "loss": 2.5898, "step": 4837 }, { "epoch": 0.0004838, "grad_norm": 4.201504230499268, "learning_rate": 4.837e-05, "loss": 4.2305, "step": 4838 }, { "epoch": 0.0004839, "grad_norm": 3.1145918369293213, "learning_rate": 4.838e-05, "loss": 2.6719, "step": 4839 }, { "epoch": 0.000484, "grad_norm": 2.856489658355713, "learning_rate": 4.8389999999999996e-05, "loss": 2.8555, "step": 4840 }, { "epoch": 0.0004841, "grad_norm": 4.2679572105407715, "learning_rate": 4.84e-05, "loss": 3.334, "step": 4841 }, { "epoch": 0.0004842, "grad_norm": 3.71990966796875, "learning_rate": 4.8410000000000006e-05, "loss": 2.9629, "step": 4842 }, { "epoch": 0.0004843, "grad_norm": 3.682610034942627, "learning_rate": 4.842000000000001e-05, "loss": 3.7051, "step": 4843 }, { "epoch": 0.0004844, "grad_norm": 3.2201039791107178, "learning_rate": 4.8429999999999995e-05, "loss": 3.0664, "step": 4844 }, { "epoch": 0.0004845, "grad_norm": 4.242389678955078, "learning_rate": 4.8440000000000004e-05, "loss": 3.5527, "step": 4845 }, { "epoch": 0.0004846, "grad_norm": 3.459249258041382, "learning_rate": 4.8450000000000005e-05, "loss": 2.9102, "step": 4846 }, { "epoch": 0.0004847, "grad_norm": 3.262465476989746, "learning_rate": 4.846e-05, "loss": 2.6113, "step": 4847 }, { "epoch": 0.0004848, "grad_norm": 2.7241337299346924, "learning_rate": 4.847e-05, "loss": 2.5625, "step": 4848 }, { "epoch": 0.0004849, "grad_norm": 2.937328815460205, "learning_rate": 4.8480000000000003e-05, "loss": 2.7949, "step": 4849 }, { "epoch": 0.000485, "grad_norm": 2.9463789463043213, "learning_rate": 4.849e-05, "loss": 2.6406, "step": 4850 }, { "epoch": 0.0004851, "grad_norm": 2.8056671619415283, "learning_rate": 4.85e-05, "loss": 2.9434, "step": 4851 }, { "epoch": 0.0004852, "grad_norm": 3.1404683589935303, "learning_rate": 4.851e-05, "loss": 3.1152, "step": 4852 }, { "epoch": 0.0004853, "grad_norm": 4.328506946563721, "learning_rate": 4.852e-05, "loss": 3.2559, "step": 4853 }, { "epoch": 0.0004854, "grad_norm": 4.04624605178833, "learning_rate": 4.853e-05, "loss": 3.4512, "step": 4854 }, { "epoch": 0.0004855, "grad_norm": 2.737555980682373, "learning_rate": 4.854e-05, "loss": 2.4609, "step": 4855 }, { "epoch": 0.0004856, "grad_norm": 2.8501687049865723, "learning_rate": 4.855e-05, "loss": 2.623, "step": 4856 }, { "epoch": 0.0004857, "grad_norm": 3.1903398036956787, "learning_rate": 4.8559999999999996e-05, "loss": 3.0117, "step": 4857 }, { "epoch": 0.0004858, "grad_norm": 3.0492584705352783, "learning_rate": 4.857e-05, "loss": 2.6348, "step": 4858 }, { "epoch": 0.0004859, "grad_norm": 2.9521851539611816, "learning_rate": 4.8580000000000006e-05, "loss": 2.6309, "step": 4859 }, { "epoch": 0.000486, "grad_norm": 2.6691646575927734, "learning_rate": 4.859000000000001e-05, "loss": 2.457, "step": 4860 }, { "epoch": 0.0004861, "grad_norm": 3.16009521484375, "learning_rate": 4.8599999999999995e-05, "loss": 2.8496, "step": 4861 }, { "epoch": 0.0004862, "grad_norm": 2.91507625579834, "learning_rate": 4.8610000000000004e-05, "loss": 2.5703, "step": 4862 }, { "epoch": 0.0004863, "grad_norm": 3.2791292667388916, "learning_rate": 4.8620000000000005e-05, "loss": 2.9863, "step": 4863 }, { "epoch": 0.0004864, "grad_norm": 4.775444507598877, "learning_rate": 4.863e-05, "loss": 4.0684, "step": 4864 }, { "epoch": 0.0004865, "grad_norm": 6.402577877044678, "learning_rate": 4.864e-05, "loss": 4.8594, "step": 4865 }, { "epoch": 0.0004866, "grad_norm": 3.5216941833496094, "learning_rate": 4.8650000000000003e-05, "loss": 2.5527, "step": 4866 }, { "epoch": 0.0004867, "grad_norm": 5.393802165985107, "learning_rate": 4.866e-05, "loss": 3.6602, "step": 4867 }, { "epoch": 0.0004868, "grad_norm": 4.241028785705566, "learning_rate": 4.867e-05, "loss": 3.2344, "step": 4868 }, { "epoch": 0.0004869, "grad_norm": 3.2610390186309814, "learning_rate": 4.868e-05, "loss": 2.9414, "step": 4869 }, { "epoch": 0.000487, "grad_norm": 3.3307955265045166, "learning_rate": 4.869e-05, "loss": 3.3496, "step": 4870 }, { "epoch": 0.0004871, "grad_norm": 3.6304919719696045, "learning_rate": 4.87e-05, "loss": 2.9043, "step": 4871 }, { "epoch": 0.0004872, "grad_norm": 3.736706495285034, "learning_rate": 4.871e-05, "loss": 3.0996, "step": 4872 }, { "epoch": 0.0004873, "grad_norm": 3.2173092365264893, "learning_rate": 4.872e-05, "loss": 2.6191, "step": 4873 }, { "epoch": 0.0004874, "grad_norm": 3.394976854324341, "learning_rate": 4.8729999999999996e-05, "loss": 2.9668, "step": 4874 }, { "epoch": 0.0004875, "grad_norm": 2.8851332664489746, "learning_rate": 4.874e-05, "loss": 2.8672, "step": 4875 }, { "epoch": 0.0004876, "grad_norm": 7.9235429763793945, "learning_rate": 4.8750000000000006e-05, "loss": 4.209, "step": 4876 }, { "epoch": 0.0004877, "grad_norm": 3.3067634105682373, "learning_rate": 4.876000000000001e-05, "loss": 2.8359, "step": 4877 }, { "epoch": 0.0004878, "grad_norm": 3.968092441558838, "learning_rate": 4.8769999999999996e-05, "loss": 2.9355, "step": 4878 }, { "epoch": 0.0004879, "grad_norm": 3.1470823287963867, "learning_rate": 4.8780000000000004e-05, "loss": 2.7129, "step": 4879 }, { "epoch": 0.000488, "grad_norm": 2.9188997745513916, "learning_rate": 4.8790000000000006e-05, "loss": 2.8184, "step": 4880 }, { "epoch": 0.0004881, "grad_norm": 4.0181474685668945, "learning_rate": 4.88e-05, "loss": 3.8027, "step": 4881 }, { "epoch": 0.0004882, "grad_norm": 3.9837429523468018, "learning_rate": 4.881e-05, "loss": 3.0156, "step": 4882 }, { "epoch": 0.0004883, "grad_norm": 3.4352903366088867, "learning_rate": 4.8820000000000004e-05, "loss": 3.0352, "step": 4883 }, { "epoch": 0.0004884, "grad_norm": 3.4703900814056396, "learning_rate": 4.883e-05, "loss": 3.0859, "step": 4884 }, { "epoch": 0.0004885, "grad_norm": 3.2002744674682617, "learning_rate": 4.884e-05, "loss": 2.8809, "step": 4885 }, { "epoch": 0.0004886, "grad_norm": 2.8497040271759033, "learning_rate": 4.885e-05, "loss": 2.5781, "step": 4886 }, { "epoch": 0.0004887, "grad_norm": 3.1276843547821045, "learning_rate": 4.886e-05, "loss": 2.9785, "step": 4887 }, { "epoch": 0.0004888, "grad_norm": 2.8994522094726562, "learning_rate": 4.887e-05, "loss": 2.4746, "step": 4888 }, { "epoch": 0.0004889, "grad_norm": 3.209141969680786, "learning_rate": 4.888e-05, "loss": 3.1035, "step": 4889 }, { "epoch": 0.000489, "grad_norm": 3.31471586227417, "learning_rate": 4.889e-05, "loss": 3.2188, "step": 4890 }, { "epoch": 0.0004891, "grad_norm": 3.1659867763519287, "learning_rate": 4.8899999999999996e-05, "loss": 2.6699, "step": 4891 }, { "epoch": 0.0004892, "grad_norm": 4.676686763763428, "learning_rate": 4.891e-05, "loss": 3.0254, "step": 4892 }, { "epoch": 0.0004893, "grad_norm": 3.3633439540863037, "learning_rate": 4.8920000000000006e-05, "loss": 3.041, "step": 4893 }, { "epoch": 0.0004894, "grad_norm": 3.5125536918640137, "learning_rate": 4.893000000000001e-05, "loss": 3.1152, "step": 4894 }, { "epoch": 0.0004895, "grad_norm": 3.252897262573242, "learning_rate": 4.8939999999999996e-05, "loss": 3.0098, "step": 4895 }, { "epoch": 0.0004896, "grad_norm": 2.900635004043579, "learning_rate": 4.8950000000000004e-05, "loss": 2.4727, "step": 4896 }, { "epoch": 0.0004897, "grad_norm": 3.1217844486236572, "learning_rate": 4.8960000000000006e-05, "loss": 2.7305, "step": 4897 }, { "epoch": 0.0004898, "grad_norm": 3.056335926055908, "learning_rate": 4.897e-05, "loss": 2.543, "step": 4898 }, { "epoch": 0.0004899, "grad_norm": 4.163784980773926, "learning_rate": 4.898e-05, "loss": 3.1875, "step": 4899 }, { "epoch": 0.00049, "grad_norm": 3.2485527992248535, "learning_rate": 4.8990000000000004e-05, "loss": 2.8848, "step": 4900 }, { "epoch": 0.0004901, "grad_norm": 2.869154453277588, "learning_rate": 4.9e-05, "loss": 2.4688, "step": 4901 }, { "epoch": 0.0004902, "grad_norm": 3.202078104019165, "learning_rate": 4.901e-05, "loss": 2.5879, "step": 4902 }, { "epoch": 0.0004903, "grad_norm": 3.6092686653137207, "learning_rate": 4.902e-05, "loss": 2.9707, "step": 4903 }, { "epoch": 0.0004904, "grad_norm": 3.2814273834228516, "learning_rate": 4.903e-05, "loss": 2.8066, "step": 4904 }, { "epoch": 0.0004905, "grad_norm": 3.2298812866210938, "learning_rate": 4.904e-05, "loss": 2.5898, "step": 4905 }, { "epoch": 0.0004906, "grad_norm": 3.6497714519500732, "learning_rate": 4.905e-05, "loss": 2.9609, "step": 4906 }, { "epoch": 0.0004907, "grad_norm": 3.193535327911377, "learning_rate": 4.906e-05, "loss": 2.6777, "step": 4907 }, { "epoch": 0.0004908, "grad_norm": 3.21246075630188, "learning_rate": 4.9069999999999996e-05, "loss": 2.8281, "step": 4908 }, { "epoch": 0.0004909, "grad_norm": 3.5341031551361084, "learning_rate": 4.908e-05, "loss": 2.9746, "step": 4909 }, { "epoch": 0.000491, "grad_norm": 2.7848691940307617, "learning_rate": 4.9090000000000006e-05, "loss": 2.5918, "step": 4910 }, { "epoch": 0.0004911, "grad_norm": 2.8175764083862305, "learning_rate": 4.910000000000001e-05, "loss": 2.4629, "step": 4911 }, { "epoch": 0.0004912, "grad_norm": 3.047422409057617, "learning_rate": 4.9109999999999996e-05, "loss": 2.7051, "step": 4912 }, { "epoch": 0.0004913, "grad_norm": 2.9127583503723145, "learning_rate": 4.9120000000000004e-05, "loss": 2.4893, "step": 4913 }, { "epoch": 0.0004914, "grad_norm": 2.803332567214966, "learning_rate": 4.9130000000000006e-05, "loss": 2.4219, "step": 4914 }, { "epoch": 0.0004915, "grad_norm": 2.7996559143066406, "learning_rate": 4.914e-05, "loss": 2.4629, "step": 4915 }, { "epoch": 0.0004916, "grad_norm": 5.793188571929932, "learning_rate": 4.915e-05, "loss": 3.7988, "step": 4916 }, { "epoch": 0.0004917, "grad_norm": 2.948418140411377, "learning_rate": 4.9160000000000004e-05, "loss": 2.5957, "step": 4917 }, { "epoch": 0.0004918, "grad_norm": 3.144325017929077, "learning_rate": 4.917e-05, "loss": 2.6445, "step": 4918 }, { "epoch": 0.0004919, "grad_norm": 3.0535080432891846, "learning_rate": 4.918e-05, "loss": 2.6855, "step": 4919 }, { "epoch": 0.000492, "grad_norm": 2.893406391143799, "learning_rate": 4.919e-05, "loss": 2.4395, "step": 4920 }, { "epoch": 0.0004921, "grad_norm": 3.550628662109375, "learning_rate": 4.92e-05, "loss": 3.1426, "step": 4921 }, { "epoch": 0.0004922, "grad_norm": 4.208357810974121, "learning_rate": 4.921e-05, "loss": 3.166, "step": 4922 }, { "epoch": 0.0004923, "grad_norm": 3.156026601791382, "learning_rate": 4.922e-05, "loss": 2.9043, "step": 4923 }, { "epoch": 0.0004924, "grad_norm": 3.0839414596557617, "learning_rate": 4.923e-05, "loss": 2.791, "step": 4924 }, { "epoch": 0.0004925, "grad_norm": 2.821011543273926, "learning_rate": 4.9239999999999996e-05, "loss": 2.4316, "step": 4925 }, { "epoch": 0.0004926, "grad_norm": 3.6722347736358643, "learning_rate": 4.925e-05, "loss": 2.9062, "step": 4926 }, { "epoch": 0.0004927, "grad_norm": 3.3902089595794678, "learning_rate": 4.9260000000000006e-05, "loss": 3.1934, "step": 4927 }, { "epoch": 0.0004928, "grad_norm": 5.007480144500732, "learning_rate": 4.927000000000001e-05, "loss": 3.1963, "step": 4928 }, { "epoch": 0.0004929, "grad_norm": 3.0512850284576416, "learning_rate": 4.9279999999999996e-05, "loss": 2.4824, "step": 4929 }, { "epoch": 0.000493, "grad_norm": 3.2722339630126953, "learning_rate": 4.9290000000000004e-05, "loss": 2.8516, "step": 4930 }, { "epoch": 0.0004931, "grad_norm": 3.5364317893981934, "learning_rate": 4.9300000000000006e-05, "loss": 2.873, "step": 4931 }, { "epoch": 0.0004932, "grad_norm": 3.511012554168701, "learning_rate": 4.931e-05, "loss": 2.9395, "step": 4932 }, { "epoch": 0.0004933, "grad_norm": 3.5106234550476074, "learning_rate": 4.932e-05, "loss": 2.6465, "step": 4933 }, { "epoch": 0.0004934, "grad_norm": 2.8259685039520264, "learning_rate": 4.9330000000000004e-05, "loss": 2.6719, "step": 4934 }, { "epoch": 0.0004935, "grad_norm": 3.0994811058044434, "learning_rate": 4.934e-05, "loss": 2.7559, "step": 4935 }, { "epoch": 0.0004936, "grad_norm": 3.3891987800598145, "learning_rate": 4.935e-05, "loss": 2.8223, "step": 4936 }, { "epoch": 0.0004937, "grad_norm": 3.387749195098877, "learning_rate": 4.936e-05, "loss": 2.9453, "step": 4937 }, { "epoch": 0.0004938, "grad_norm": 3.2160184383392334, "learning_rate": 4.937e-05, "loss": 2.7617, "step": 4938 }, { "epoch": 0.0004939, "grad_norm": 2.6925809383392334, "learning_rate": 4.938e-05, "loss": 2.3438, "step": 4939 }, { "epoch": 0.000494, "grad_norm": 2.8169851303100586, "learning_rate": 4.939e-05, "loss": 2.3945, "step": 4940 }, { "epoch": 0.0004941, "grad_norm": 3.3772518634796143, "learning_rate": 4.94e-05, "loss": 2.6426, "step": 4941 }, { "epoch": 0.0004942, "grad_norm": 3.1105170249938965, "learning_rate": 4.9409999999999996e-05, "loss": 2.6523, "step": 4942 }, { "epoch": 0.0004943, "grad_norm": 2.857726573944092, "learning_rate": 4.942e-05, "loss": 2.9922, "step": 4943 }, { "epoch": 0.0004944, "grad_norm": 3.4126760959625244, "learning_rate": 4.943e-05, "loss": 3.3271, "step": 4944 }, { "epoch": 0.0004945, "grad_norm": 3.1623902320861816, "learning_rate": 4.944000000000001e-05, "loss": 2.9531, "step": 4945 }, { "epoch": 0.0004946, "grad_norm": 3.628561496734619, "learning_rate": 4.9449999999999996e-05, "loss": 2.7646, "step": 4946 }, { "epoch": 0.0004947, "grad_norm": 5.002937316894531, "learning_rate": 4.9460000000000004e-05, "loss": 3.0625, "step": 4947 }, { "epoch": 0.0004948, "grad_norm": 3.109325647354126, "learning_rate": 4.9470000000000006e-05, "loss": 2.5215, "step": 4948 }, { "epoch": 0.0004949, "grad_norm": 3.6489598751068115, "learning_rate": 4.948e-05, "loss": 3.0498, "step": 4949 }, { "epoch": 0.000495, "grad_norm": 3.1119871139526367, "learning_rate": 4.949e-05, "loss": 2.5176, "step": 4950 }, { "epoch": 0.0004951, "grad_norm": 3.0596883296966553, "learning_rate": 4.9500000000000004e-05, "loss": 2.6289, "step": 4951 }, { "epoch": 0.0004952, "grad_norm": 2.7664010524749756, "learning_rate": 4.951e-05, "loss": 2.3438, "step": 4952 }, { "epoch": 0.0004953, "grad_norm": 3.4498963356018066, "learning_rate": 4.952e-05, "loss": 2.6172, "step": 4953 }, { "epoch": 0.0004954, "grad_norm": 3.1450445652008057, "learning_rate": 4.953e-05, "loss": 2.4277, "step": 4954 }, { "epoch": 0.0004955, "grad_norm": 4.105097770690918, "learning_rate": 4.9540000000000003e-05, "loss": 2.7656, "step": 4955 }, { "epoch": 0.0004956, "grad_norm": 3.547750473022461, "learning_rate": 4.955e-05, "loss": 2.9375, "step": 4956 }, { "epoch": 0.0004957, "grad_norm": 3.2620086669921875, "learning_rate": 4.956e-05, "loss": 3.0645, "step": 4957 }, { "epoch": 0.0004958, "grad_norm": 2.876377582550049, "learning_rate": 4.957e-05, "loss": 2.6328, "step": 4958 }, { "epoch": 0.0004959, "grad_norm": 2.6791906356811523, "learning_rate": 4.9579999999999996e-05, "loss": 2.293, "step": 4959 }, { "epoch": 0.000496, "grad_norm": 2.8224759101867676, "learning_rate": 4.959e-05, "loss": 2.4434, "step": 4960 }, { "epoch": 0.0004961, "grad_norm": 3.066927671432495, "learning_rate": 4.96e-05, "loss": 2.8398, "step": 4961 }, { "epoch": 0.0004962, "grad_norm": 3.9707651138305664, "learning_rate": 4.961000000000001e-05, "loss": 2.8984, "step": 4962 }, { "epoch": 0.0004963, "grad_norm": 2.862757921218872, "learning_rate": 4.9619999999999996e-05, "loss": 2.4727, "step": 4963 }, { "epoch": 0.0004964, "grad_norm": 3.317944049835205, "learning_rate": 4.9630000000000004e-05, "loss": 2.7148, "step": 4964 }, { "epoch": 0.0004965, "grad_norm": 3.1009724140167236, "learning_rate": 4.9640000000000006e-05, "loss": 2.5684, "step": 4965 }, { "epoch": 0.0004966, "grad_norm": 4.326591968536377, "learning_rate": 4.9649999999999994e-05, "loss": 2.9219, "step": 4966 }, { "epoch": 0.0004967, "grad_norm": 3.4471962451934814, "learning_rate": 4.966e-05, "loss": 2.8496, "step": 4967 }, { "epoch": 0.0004968, "grad_norm": 2.987342119216919, "learning_rate": 4.9670000000000004e-05, "loss": 2.4922, "step": 4968 }, { "epoch": 0.0004969, "grad_norm": 3.161642074584961, "learning_rate": 4.968e-05, "loss": 2.8799, "step": 4969 }, { "epoch": 0.000497, "grad_norm": 3.2874529361724854, "learning_rate": 4.969e-05, "loss": 2.6826, "step": 4970 }, { "epoch": 0.0004971, "grad_norm": 2.711935520172119, "learning_rate": 4.97e-05, "loss": 2.2695, "step": 4971 }, { "epoch": 0.0004972, "grad_norm": 3.4356820583343506, "learning_rate": 4.9710000000000003e-05, "loss": 2.8867, "step": 4972 }, { "epoch": 0.0004973, "grad_norm": 3.1066980361938477, "learning_rate": 4.972e-05, "loss": 2.5195, "step": 4973 }, { "epoch": 0.0004974, "grad_norm": 2.7720608711242676, "learning_rate": 4.973e-05, "loss": 2.3066, "step": 4974 }, { "epoch": 0.0004975, "grad_norm": 2.7561423778533936, "learning_rate": 4.974e-05, "loss": 2.4609, "step": 4975 }, { "epoch": 0.0004976, "grad_norm": 3.2306249141693115, "learning_rate": 4.9749999999999996e-05, "loss": 2.9961, "step": 4976 }, { "epoch": 0.0004977, "grad_norm": 2.7933778762817383, "learning_rate": 4.976e-05, "loss": 2.2803, "step": 4977 }, { "epoch": 0.0004978, "grad_norm": 2.905799150466919, "learning_rate": 4.977e-05, "loss": 2.5547, "step": 4978 }, { "epoch": 0.0004979, "grad_norm": 2.6143710613250732, "learning_rate": 4.978000000000001e-05, "loss": 2.2197, "step": 4979 }, { "epoch": 0.000498, "grad_norm": 5.409095287322998, "learning_rate": 4.9789999999999996e-05, "loss": 2.8125, "step": 4980 }, { "epoch": 0.0004981, "grad_norm": 2.9222419261932373, "learning_rate": 4.9800000000000004e-05, "loss": 2.6211, "step": 4981 }, { "epoch": 0.0004982, "grad_norm": 3.0503506660461426, "learning_rate": 4.9810000000000006e-05, "loss": 2.5391, "step": 4982 }, { "epoch": 0.0004983, "grad_norm": 3.278925657272339, "learning_rate": 4.9819999999999994e-05, "loss": 2.3496, "step": 4983 }, { "epoch": 0.0004984, "grad_norm": 3.5366010665893555, "learning_rate": 4.983e-05, "loss": 2.666, "step": 4984 }, { "epoch": 0.0004985, "grad_norm": 3.982550859451294, "learning_rate": 4.9840000000000004e-05, "loss": 2.9219, "step": 4985 }, { "epoch": 0.0004986, "grad_norm": 4.075919151306152, "learning_rate": 4.985e-05, "loss": 2.9336, "step": 4986 }, { "epoch": 0.0004987, "grad_norm": 2.97637677192688, "learning_rate": 4.986e-05, "loss": 2.6074, "step": 4987 }, { "epoch": 0.0004988, "grad_norm": 3.0733413696289062, "learning_rate": 4.987e-05, "loss": 2.6895, "step": 4988 }, { "epoch": 0.0004989, "grad_norm": 2.856350898742676, "learning_rate": 4.9880000000000004e-05, "loss": 2.5293, "step": 4989 }, { "epoch": 0.000499, "grad_norm": 3.8834662437438965, "learning_rate": 4.989e-05, "loss": 2.9492, "step": 4990 }, { "epoch": 0.0004991, "grad_norm": 3.028676748275757, "learning_rate": 4.99e-05, "loss": 2.2441, "step": 4991 }, { "epoch": 0.0004992, "grad_norm": 3.739980697631836, "learning_rate": 4.991e-05, "loss": 2.5547, "step": 4992 }, { "epoch": 0.0004993, "grad_norm": 3.036109685897827, "learning_rate": 4.9919999999999996e-05, "loss": 2.7402, "step": 4993 }, { "epoch": 0.0004994, "grad_norm": 3.27606463432312, "learning_rate": 4.993e-05, "loss": 2.5254, "step": 4994 }, { "epoch": 0.0004995, "grad_norm": 3.0515246391296387, "learning_rate": 4.994e-05, "loss": 2.4004, "step": 4995 }, { "epoch": 0.0004996, "grad_norm": 2.65454363822937, "learning_rate": 4.995000000000001e-05, "loss": 2.3301, "step": 4996 }, { "epoch": 0.0004997, "grad_norm": 2.9057114124298096, "learning_rate": 4.9959999999999996e-05, "loss": 2.625, "step": 4997 }, { "epoch": 0.0004998, "grad_norm": 3.1458072662353516, "learning_rate": 4.9970000000000004e-05, "loss": 2.5957, "step": 4998 }, { "epoch": 0.0004999, "grad_norm": 4.200390338897705, "learning_rate": 4.9980000000000006e-05, "loss": 2.8652, "step": 4999 }, { "epoch": 0.0005, "grad_norm": 3.235536813735962, "learning_rate": 4.9989999999999994e-05, "loss": 2.5566, "step": 5000 }, { "epoch": 0.0005, "eval_loss": 0.2768504023551941, "eval_runtime": 367.6612, "eval_samples_per_second": 27.199, "eval_steps_per_second": 1.7, "step": 5000 }, { "epoch": 0.0005001, "grad_norm": 3.067068338394165, "learning_rate": 5e-05, "loss": 2.6309, "step": 5001 }, { "epoch": 0.0005002, "grad_norm": 3.328768014907837, "learning_rate": 5.0010000000000004e-05, "loss": 2.8008, "step": 5002 }, { "epoch": 0.0005003, "grad_norm": 3.0800108909606934, "learning_rate": 5.002e-05, "loss": 2.3398, "step": 5003 }, { "epoch": 0.0005004, "grad_norm": 3.036958932876587, "learning_rate": 5.003e-05, "loss": 2.4629, "step": 5004 }, { "epoch": 0.0005005, "grad_norm": 3.164119005203247, "learning_rate": 5.004e-05, "loss": 2.6973, "step": 5005 }, { "epoch": 0.0005006, "grad_norm": 5.4457244873046875, "learning_rate": 5.0050000000000004e-05, "loss": 3.2051, "step": 5006 }, { "epoch": 0.0005007, "grad_norm": 3.8274242877960205, "learning_rate": 5.006e-05, "loss": 3.1211, "step": 5007 }, { "epoch": 0.0005008, "grad_norm": 3.2691900730133057, "learning_rate": 5.007e-05, "loss": 2.4199, "step": 5008 }, { "epoch": 0.0005009, "grad_norm": 3.4117910861968994, "learning_rate": 5.008e-05, "loss": 2.2607, "step": 5009 }, { "epoch": 0.000501, "grad_norm": 3.5231997966766357, "learning_rate": 5.0089999999999996e-05, "loss": 2.8398, "step": 5010 }, { "epoch": 0.0005011, "grad_norm": 3.054734945297241, "learning_rate": 5.01e-05, "loss": 2.4648, "step": 5011 }, { "epoch": 0.0005012, "grad_norm": 2.849677085876465, "learning_rate": 5.011e-05, "loss": 2.167, "step": 5012 }, { "epoch": 0.0005013, "grad_norm": 2.9493212699890137, "learning_rate": 5.012000000000001e-05, "loss": 2.291, "step": 5013 }, { "epoch": 0.0005014, "grad_norm": 3.1363699436187744, "learning_rate": 5.0129999999999996e-05, "loss": 2.7461, "step": 5014 }, { "epoch": 0.0005015, "grad_norm": 2.814643621444702, "learning_rate": 5.0140000000000004e-05, "loss": 2.4824, "step": 5015 }, { "epoch": 0.0005016, "grad_norm": 2.8603408336639404, "learning_rate": 5.0150000000000006e-05, "loss": 2.125, "step": 5016 }, { "epoch": 0.0005017, "grad_norm": 2.944711446762085, "learning_rate": 5.0159999999999994e-05, "loss": 2.4844, "step": 5017 }, { "epoch": 0.0005018, "grad_norm": 3.1112520694732666, "learning_rate": 5.017e-05, "loss": 2.5977, "step": 5018 }, { "epoch": 0.0005019, "grad_norm": 2.9993197917938232, "learning_rate": 5.0180000000000004e-05, "loss": 2.4922, "step": 5019 }, { "epoch": 0.000502, "grad_norm": 2.9708147048950195, "learning_rate": 5.019e-05, "loss": 2.4717, "step": 5020 }, { "epoch": 0.0005021, "grad_norm": 4.204382419586182, "learning_rate": 5.02e-05, "loss": 3.1553, "step": 5021 }, { "epoch": 0.0005022, "grad_norm": 3.1822476387023926, "learning_rate": 5.021e-05, "loss": 2.6309, "step": 5022 }, { "epoch": 0.0005023, "grad_norm": 3.3060824871063232, "learning_rate": 5.0220000000000004e-05, "loss": 2.4082, "step": 5023 }, { "epoch": 0.0005024, "grad_norm": 2.8483448028564453, "learning_rate": 5.023e-05, "loss": 2.3525, "step": 5024 }, { "epoch": 0.0005025, "grad_norm": 3.356644868850708, "learning_rate": 5.024e-05, "loss": 2.4082, "step": 5025 }, { "epoch": 0.0005026, "grad_norm": 2.91219162940979, "learning_rate": 5.025e-05, "loss": 2.2773, "step": 5026 }, { "epoch": 0.0005027, "grad_norm": 3.3623499870300293, "learning_rate": 5.0259999999999997e-05, "loss": 3.0371, "step": 5027 }, { "epoch": 0.0005028, "grad_norm": 3.377230644226074, "learning_rate": 5.027e-05, "loss": 2.8193, "step": 5028 }, { "epoch": 0.0005029, "grad_norm": 3.3483150005340576, "learning_rate": 5.028e-05, "loss": 2.7598, "step": 5029 }, { "epoch": 0.000503, "grad_norm": 3.1419856548309326, "learning_rate": 5.029000000000001e-05, "loss": 2.5557, "step": 5030 }, { "epoch": 0.0005031, "grad_norm": 2.9181790351867676, "learning_rate": 5.0299999999999996e-05, "loss": 2.2832, "step": 5031 }, { "epoch": 0.0005032, "grad_norm": 3.5577728748321533, "learning_rate": 5.0310000000000005e-05, "loss": 2.6309, "step": 5032 }, { "epoch": 0.0005033, "grad_norm": 2.787325620651245, "learning_rate": 5.0320000000000006e-05, "loss": 2.2725, "step": 5033 }, { "epoch": 0.0005034, "grad_norm": 2.802180290222168, "learning_rate": 5.0329999999999994e-05, "loss": 2.2617, "step": 5034 }, { "epoch": 0.0005035, "grad_norm": 2.8316397666931152, "learning_rate": 5.034e-05, "loss": 2.1992, "step": 5035 }, { "epoch": 0.0005036, "grad_norm": 2.9372782707214355, "learning_rate": 5.0350000000000004e-05, "loss": 2.4199, "step": 5036 }, { "epoch": 0.0005037, "grad_norm": 2.772482395172119, "learning_rate": 5.036e-05, "loss": 2.3281, "step": 5037 }, { "epoch": 0.0005038, "grad_norm": 2.5360898971557617, "learning_rate": 5.037e-05, "loss": 2.1758, "step": 5038 }, { "epoch": 0.0005039, "grad_norm": 2.7606682777404785, "learning_rate": 5.038e-05, "loss": 2.2412, "step": 5039 }, { "epoch": 0.000504, "grad_norm": 2.647916793823242, "learning_rate": 5.0390000000000004e-05, "loss": 2.001, "step": 5040 }, { "epoch": 0.0005041, "grad_norm": 2.692534923553467, "learning_rate": 5.04e-05, "loss": 2.2412, "step": 5041 }, { "epoch": 0.0005042, "grad_norm": 2.807421922683716, "learning_rate": 5.041e-05, "loss": 2.4648, "step": 5042 }, { "epoch": 0.0005043, "grad_norm": 2.594489097595215, "learning_rate": 5.042e-05, "loss": 2.1846, "step": 5043 }, { "epoch": 0.0005044, "grad_norm": 2.7565715312957764, "learning_rate": 5.0429999999999997e-05, "loss": 2.1514, "step": 5044 }, { "epoch": 0.0005045, "grad_norm": 3.1473684310913086, "learning_rate": 5.044e-05, "loss": 2.6064, "step": 5045 }, { "epoch": 0.0005046, "grad_norm": 4.937899589538574, "learning_rate": 5.045e-05, "loss": 2.7061, "step": 5046 }, { "epoch": 0.0005047, "grad_norm": 2.826744794845581, "learning_rate": 5.046000000000001e-05, "loss": 2.1299, "step": 5047 }, { "epoch": 0.0005048, "grad_norm": 3.4902610778808594, "learning_rate": 5.0469999999999996e-05, "loss": 2.6973, "step": 5048 }, { "epoch": 0.0005049, "grad_norm": 4.5450897216796875, "learning_rate": 5.0480000000000005e-05, "loss": 2.9199, "step": 5049 }, { "epoch": 0.000505, "grad_norm": 2.9845242500305176, "learning_rate": 5.0490000000000006e-05, "loss": 2.4043, "step": 5050 }, { "epoch": 0.0005051, "grad_norm": 2.936403512954712, "learning_rate": 5.0499999999999994e-05, "loss": 2.4492, "step": 5051 }, { "epoch": 0.0005052, "grad_norm": 2.697326898574829, "learning_rate": 5.051e-05, "loss": 2.127, "step": 5052 }, { "epoch": 0.0005053, "grad_norm": 2.5877013206481934, "learning_rate": 5.0520000000000004e-05, "loss": 2.0488, "step": 5053 }, { "epoch": 0.0005054, "grad_norm": 2.87369704246521, "learning_rate": 5.053e-05, "loss": 2.3418, "step": 5054 }, { "epoch": 0.0005055, "grad_norm": 2.8604819774627686, "learning_rate": 5.054e-05, "loss": 2.5771, "step": 5055 }, { "epoch": 0.0005056, "grad_norm": 3.391019821166992, "learning_rate": 5.055e-05, "loss": 2.4707, "step": 5056 }, { "epoch": 0.0005057, "grad_norm": 3.342128038406372, "learning_rate": 5.0560000000000004e-05, "loss": 2.3516, "step": 5057 }, { "epoch": 0.0005058, "grad_norm": 3.0206456184387207, "learning_rate": 5.057e-05, "loss": 2.8848, "step": 5058 }, { "epoch": 0.0005059, "grad_norm": 2.806652545928955, "learning_rate": 5.058e-05, "loss": 2.3779, "step": 5059 }, { "epoch": 0.000506, "grad_norm": 2.74412202835083, "learning_rate": 5.059e-05, "loss": 2.293, "step": 5060 }, { "epoch": 0.0005061, "grad_norm": 3.182224750518799, "learning_rate": 5.06e-05, "loss": 2.3789, "step": 5061 }, { "epoch": 0.0005062, "grad_norm": 2.848026990890503, "learning_rate": 5.061e-05, "loss": 2.2559, "step": 5062 }, { "epoch": 0.0005063, "grad_norm": 2.685464859008789, "learning_rate": 5.062e-05, "loss": 2.1797, "step": 5063 }, { "epoch": 0.0005064, "grad_norm": 3.2182085514068604, "learning_rate": 5.063000000000001e-05, "loss": 2.4307, "step": 5064 }, { "epoch": 0.0005065, "grad_norm": 3.1547040939331055, "learning_rate": 5.0639999999999996e-05, "loss": 2.2812, "step": 5065 }, { "epoch": 0.0005066, "grad_norm": 2.836766004562378, "learning_rate": 5.0650000000000005e-05, "loss": 2.1924, "step": 5066 }, { "epoch": 0.0005067, "grad_norm": 2.765444040298462, "learning_rate": 5.0660000000000006e-05, "loss": 2.252, "step": 5067 }, { "epoch": 0.0005068, "grad_norm": 2.775315284729004, "learning_rate": 5.0669999999999994e-05, "loss": 2.2285, "step": 5068 }, { "epoch": 0.0005069, "grad_norm": 3.245835065841675, "learning_rate": 5.068e-05, "loss": 2.416, "step": 5069 }, { "epoch": 0.000507, "grad_norm": 2.9470319747924805, "learning_rate": 5.0690000000000004e-05, "loss": 2.2305, "step": 5070 }, { "epoch": 0.0005071, "grad_norm": 2.7369916439056396, "learning_rate": 5.07e-05, "loss": 2.4424, "step": 5071 }, { "epoch": 0.0005072, "grad_norm": 2.5191965103149414, "learning_rate": 5.071e-05, "loss": 2.0371, "step": 5072 }, { "epoch": 0.0005073, "grad_norm": 2.6011600494384766, "learning_rate": 5.072e-05, "loss": 2.2764, "step": 5073 }, { "epoch": 0.0005074, "grad_norm": 3.0577523708343506, "learning_rate": 5.0730000000000004e-05, "loss": 2.3828, "step": 5074 }, { "epoch": 0.0005075, "grad_norm": 2.5810158252716064, "learning_rate": 5.074e-05, "loss": 2.2334, "step": 5075 }, { "epoch": 0.0005076, "grad_norm": 2.6141197681427, "learning_rate": 5.075e-05, "loss": 2.0, "step": 5076 }, { "epoch": 0.0005077, "grad_norm": 3.7614054679870605, "learning_rate": 5.076e-05, "loss": 2.459, "step": 5077 }, { "epoch": 0.0005078, "grad_norm": 3.3027162551879883, "learning_rate": 5.077e-05, "loss": 2.6387, "step": 5078 }, { "epoch": 0.0005079, "grad_norm": 3.354887008666992, "learning_rate": 5.078e-05, "loss": 2.5811, "step": 5079 }, { "epoch": 0.000508, "grad_norm": 2.9398012161254883, "learning_rate": 5.079e-05, "loss": 2.1982, "step": 5080 }, { "epoch": 0.0005081, "grad_norm": 3.3343186378479004, "learning_rate": 5.080000000000001e-05, "loss": 2.4336, "step": 5081 }, { "epoch": 0.0005082, "grad_norm": 3.029672861099243, "learning_rate": 5.0809999999999996e-05, "loss": 2.748, "step": 5082 }, { "epoch": 0.0005083, "grad_norm": 3.381859302520752, "learning_rate": 5.0820000000000005e-05, "loss": 2.4023, "step": 5083 }, { "epoch": 0.0005084, "grad_norm": 2.8518691062927246, "learning_rate": 5.0830000000000006e-05, "loss": 2.2246, "step": 5084 }, { "epoch": 0.0005085, "grad_norm": 3.2322044372558594, "learning_rate": 5.0839999999999994e-05, "loss": 2.7002, "step": 5085 }, { "epoch": 0.0005086, "grad_norm": 3.12455153465271, "learning_rate": 5.085e-05, "loss": 2.0752, "step": 5086 }, { "epoch": 0.0005087, "grad_norm": 3.0265579223632812, "learning_rate": 5.0860000000000004e-05, "loss": 2.0928, "step": 5087 }, { "epoch": 0.0005088, "grad_norm": 2.8352231979370117, "learning_rate": 5.087e-05, "loss": 2.2197, "step": 5088 }, { "epoch": 0.0005089, "grad_norm": 4.019437313079834, "learning_rate": 5.088e-05, "loss": 2.7812, "step": 5089 }, { "epoch": 0.000509, "grad_norm": 2.9904427528381348, "learning_rate": 5.089e-05, "loss": 2.126, "step": 5090 }, { "epoch": 0.0005091, "grad_norm": 2.99135684967041, "learning_rate": 5.0900000000000004e-05, "loss": 2.0059, "step": 5091 }, { "epoch": 0.0005092, "grad_norm": 3.1440083980560303, "learning_rate": 5.091e-05, "loss": 2.6836, "step": 5092 }, { "epoch": 0.0005093, "grad_norm": 4.401318073272705, "learning_rate": 5.092e-05, "loss": 2.6553, "step": 5093 }, { "epoch": 0.0005094, "grad_norm": 3.4378817081451416, "learning_rate": 5.093e-05, "loss": 2.6445, "step": 5094 }, { "epoch": 0.0005095, "grad_norm": 3.1307246685028076, "learning_rate": 5.094e-05, "loss": 2.2588, "step": 5095 }, { "epoch": 0.0005096, "grad_norm": 2.6964521408081055, "learning_rate": 5.095e-05, "loss": 1.917, "step": 5096 }, { "epoch": 0.0005097, "grad_norm": 2.801470994949341, "learning_rate": 5.096e-05, "loss": 2.4463, "step": 5097 }, { "epoch": 0.0005098, "grad_norm": 3.1953344345092773, "learning_rate": 5.097000000000001e-05, "loss": 2.2041, "step": 5098 }, { "epoch": 0.0005099, "grad_norm": 2.9455761909484863, "learning_rate": 5.0979999999999996e-05, "loss": 2.0537, "step": 5099 }, { "epoch": 0.00051, "grad_norm": 4.67949914932251, "learning_rate": 5.0990000000000005e-05, "loss": 2.4512, "step": 5100 }, { "epoch": 0.0005101, "grad_norm": 2.7000317573547363, "learning_rate": 5.1000000000000006e-05, "loss": 2.0371, "step": 5101 }, { "epoch": 0.0005102, "grad_norm": 2.77419376373291, "learning_rate": 5.1009999999999994e-05, "loss": 2.2666, "step": 5102 }, { "epoch": 0.0005103, "grad_norm": 3.120331048965454, "learning_rate": 5.102e-05, "loss": 2.3896, "step": 5103 }, { "epoch": 0.0005104, "grad_norm": 3.1404972076416016, "learning_rate": 5.1030000000000004e-05, "loss": 2.2275, "step": 5104 }, { "epoch": 0.0005105, "grad_norm": 2.6908504962921143, "learning_rate": 5.104e-05, "loss": 2.0166, "step": 5105 }, { "epoch": 0.0005106, "grad_norm": 2.7583882808685303, "learning_rate": 5.105e-05, "loss": 2.0986, "step": 5106 }, { "epoch": 0.0005107, "grad_norm": 2.566155195236206, "learning_rate": 5.106e-05, "loss": 1.9883, "step": 5107 }, { "epoch": 0.0005108, "grad_norm": 2.673900842666626, "learning_rate": 5.1070000000000004e-05, "loss": 2.1582, "step": 5108 }, { "epoch": 0.0005109, "grad_norm": 3.484880208969116, "learning_rate": 5.108e-05, "loss": 2.4209, "step": 5109 }, { "epoch": 0.000511, "grad_norm": 3.7648422718048096, "learning_rate": 5.109e-05, "loss": 2.1221, "step": 5110 }, { "epoch": 0.0005111, "grad_norm": 3.4164645671844482, "learning_rate": 5.11e-05, "loss": 2.5684, "step": 5111 }, { "epoch": 0.0005112, "grad_norm": 2.5940446853637695, "learning_rate": 5.111e-05, "loss": 1.9805, "step": 5112 }, { "epoch": 0.0005113, "grad_norm": 4.611963748931885, "learning_rate": 5.112e-05, "loss": 2.6562, "step": 5113 }, { "epoch": 0.0005114, "grad_norm": 3.521127700805664, "learning_rate": 5.113e-05, "loss": 2.2832, "step": 5114 }, { "epoch": 0.0005115, "grad_norm": 2.620272397994995, "learning_rate": 5.114000000000001e-05, "loss": 1.9424, "step": 5115 }, { "epoch": 0.0005116, "grad_norm": 2.812140703201294, "learning_rate": 5.1149999999999996e-05, "loss": 2.4863, "step": 5116 }, { "epoch": 0.0005117, "grad_norm": 2.6936028003692627, "learning_rate": 5.1160000000000005e-05, "loss": 2.0625, "step": 5117 }, { "epoch": 0.0005118, "grad_norm": 2.507796287536621, "learning_rate": 5.1170000000000006e-05, "loss": 1.873, "step": 5118 }, { "epoch": 0.0005119, "grad_norm": 3.730189085006714, "learning_rate": 5.1179999999999994e-05, "loss": 2.292, "step": 5119 }, { "epoch": 0.000512, "grad_norm": 2.818584442138672, "learning_rate": 5.119e-05, "loss": 2.082, "step": 5120 }, { "epoch": 0.0005121, "grad_norm": 2.804715633392334, "learning_rate": 5.1200000000000004e-05, "loss": 2.2432, "step": 5121 }, { "epoch": 0.0005122, "grad_norm": 2.602494716644287, "learning_rate": 5.121e-05, "loss": 1.9277, "step": 5122 }, { "epoch": 0.0005123, "grad_norm": 2.662548065185547, "learning_rate": 5.122e-05, "loss": 1.9863, "step": 5123 }, { "epoch": 0.0005124, "grad_norm": 2.7869327068328857, "learning_rate": 5.123e-05, "loss": 2.0029, "step": 5124 }, { "epoch": 0.0005125, "grad_norm": 2.6469080448150635, "learning_rate": 5.1240000000000004e-05, "loss": 2.0107, "step": 5125 }, { "epoch": 0.0005126, "grad_norm": 2.6196298599243164, "learning_rate": 5.125e-05, "loss": 1.9395, "step": 5126 }, { "epoch": 0.0005127, "grad_norm": 5.092792510986328, "learning_rate": 5.126e-05, "loss": 2.7793, "step": 5127 }, { "epoch": 0.0005128, "grad_norm": 3.1943728923797607, "learning_rate": 5.127e-05, "loss": 1.9736, "step": 5128 }, { "epoch": 0.0005129, "grad_norm": 2.6523025035858154, "learning_rate": 5.128e-05, "loss": 1.9668, "step": 5129 }, { "epoch": 0.000513, "grad_norm": 2.750728130340576, "learning_rate": 5.129e-05, "loss": 2.3555, "step": 5130 }, { "epoch": 0.0005131, "grad_norm": 2.9490647315979004, "learning_rate": 5.13e-05, "loss": 2.3262, "step": 5131 }, { "epoch": 0.0005132, "grad_norm": 2.801272392272949, "learning_rate": 5.131000000000001e-05, "loss": 2.1309, "step": 5132 }, { "epoch": 0.0005133, "grad_norm": 3.3073480129241943, "learning_rate": 5.1319999999999997e-05, "loss": 2.127, "step": 5133 }, { "epoch": 0.0005134, "grad_norm": 2.6866047382354736, "learning_rate": 5.1330000000000005e-05, "loss": 1.9834, "step": 5134 }, { "epoch": 0.0005135, "grad_norm": 3.0547852516174316, "learning_rate": 5.1340000000000007e-05, "loss": 2.5127, "step": 5135 }, { "epoch": 0.0005136, "grad_norm": 2.56762433052063, "learning_rate": 5.1349999999999995e-05, "loss": 1.9326, "step": 5136 }, { "epoch": 0.0005137, "grad_norm": 3.1184980869293213, "learning_rate": 5.136e-05, "loss": 2.2705, "step": 5137 }, { "epoch": 0.0005138, "grad_norm": 2.6578822135925293, "learning_rate": 5.1370000000000005e-05, "loss": 2.0615, "step": 5138 }, { "epoch": 0.0005139, "grad_norm": 3.1814334392547607, "learning_rate": 5.138e-05, "loss": 2.1719, "step": 5139 }, { "epoch": 0.000514, "grad_norm": 2.503666877746582, "learning_rate": 5.139e-05, "loss": 1.9023, "step": 5140 }, { "epoch": 0.0005141, "grad_norm": 2.588676929473877, "learning_rate": 5.14e-05, "loss": 2.0166, "step": 5141 }, { "epoch": 0.0005142, "grad_norm": 2.868074893951416, "learning_rate": 5.1410000000000004e-05, "loss": 2.1348, "step": 5142 }, { "epoch": 0.0005143, "grad_norm": 2.751417398452759, "learning_rate": 5.142e-05, "loss": 2.0977, "step": 5143 }, { "epoch": 0.0005144, "grad_norm": 2.7075483798980713, "learning_rate": 5.143e-05, "loss": 2.0283, "step": 5144 }, { "epoch": 0.0005145, "grad_norm": 2.984896659851074, "learning_rate": 5.144e-05, "loss": 2.1426, "step": 5145 }, { "epoch": 0.0005146, "grad_norm": 3.3068416118621826, "learning_rate": 5.145e-05, "loss": 2.3574, "step": 5146 }, { "epoch": 0.0005147, "grad_norm": 2.8091273307800293, "learning_rate": 5.146e-05, "loss": 2.1387, "step": 5147 }, { "epoch": 0.0005148, "grad_norm": 2.6160480976104736, "learning_rate": 5.147e-05, "loss": 1.9629, "step": 5148 }, { "epoch": 0.0005149, "grad_norm": 3.155529260635376, "learning_rate": 5.148000000000001e-05, "loss": 2.3984, "step": 5149 }, { "epoch": 0.000515, "grad_norm": 2.698957920074463, "learning_rate": 5.1489999999999997e-05, "loss": 1.9375, "step": 5150 }, { "epoch": 0.0005151, "grad_norm": 2.4062185287475586, "learning_rate": 5.1500000000000005e-05, "loss": 1.8457, "step": 5151 }, { "epoch": 0.0005152, "grad_norm": 2.672334671020508, "learning_rate": 5.1510000000000007e-05, "loss": 1.9473, "step": 5152 }, { "epoch": 0.0005153, "grad_norm": 2.8882555961608887, "learning_rate": 5.1519999999999995e-05, "loss": 2.1533, "step": 5153 }, { "epoch": 0.0005154, "grad_norm": 3.0568158626556396, "learning_rate": 5.153e-05, "loss": 2.0869, "step": 5154 }, { "epoch": 0.0005155, "grad_norm": 3.5276832580566406, "learning_rate": 5.1540000000000005e-05, "loss": 2.7686, "step": 5155 }, { "epoch": 0.0005156, "grad_norm": 2.8599016666412354, "learning_rate": 5.155e-05, "loss": 2.0762, "step": 5156 }, { "epoch": 0.0005157, "grad_norm": 3.1308319568634033, "learning_rate": 5.156e-05, "loss": 2.2393, "step": 5157 }, { "epoch": 0.0005158, "grad_norm": 2.6597423553466797, "learning_rate": 5.157e-05, "loss": 1.8604, "step": 5158 }, { "epoch": 0.0005159, "grad_norm": 3.0385875701904297, "learning_rate": 5.1580000000000004e-05, "loss": 2.3242, "step": 5159 }, { "epoch": 0.000516, "grad_norm": 3.123140573501587, "learning_rate": 5.159e-05, "loss": 2.2656, "step": 5160 }, { "epoch": 0.0005161, "grad_norm": 3.3444221019744873, "learning_rate": 5.16e-05, "loss": 2.1992, "step": 5161 }, { "epoch": 0.0005162, "grad_norm": 3.1859424114227295, "learning_rate": 5.161e-05, "loss": 2.6914, "step": 5162 }, { "epoch": 0.0005163, "grad_norm": 2.451913356781006, "learning_rate": 5.162e-05, "loss": 1.8691, "step": 5163 }, { "epoch": 0.0005164, "grad_norm": 2.8825628757476807, "learning_rate": 5.163e-05, "loss": 2.0938, "step": 5164 }, { "epoch": 0.0005165, "grad_norm": 2.8346612453460693, "learning_rate": 5.164e-05, "loss": 1.9463, "step": 5165 }, { "epoch": 0.0005166, "grad_norm": 2.9165539741516113, "learning_rate": 5.165000000000001e-05, "loss": 2.1533, "step": 5166 }, { "epoch": 0.0005167, "grad_norm": 2.6613125801086426, "learning_rate": 5.166e-05, "loss": 2.0225, "step": 5167 }, { "epoch": 0.0005168, "grad_norm": 2.519366502761841, "learning_rate": 5.1670000000000005e-05, "loss": 1.9023, "step": 5168 }, { "epoch": 0.0005169, "grad_norm": 2.9969735145568848, "learning_rate": 5.168000000000001e-05, "loss": 2.2861, "step": 5169 }, { "epoch": 0.000517, "grad_norm": 2.699340343475342, "learning_rate": 5.1689999999999995e-05, "loss": 1.875, "step": 5170 }, { "epoch": 0.0005171, "grad_norm": 2.7898359298706055, "learning_rate": 5.17e-05, "loss": 2.0127, "step": 5171 }, { "epoch": 0.0005172, "grad_norm": 2.7837727069854736, "learning_rate": 5.1710000000000005e-05, "loss": 2.0693, "step": 5172 }, { "epoch": 0.0005173, "grad_norm": 2.6420516967773438, "learning_rate": 5.172e-05, "loss": 1.9033, "step": 5173 }, { "epoch": 0.0005174, "grad_norm": 2.774935722351074, "learning_rate": 5.173e-05, "loss": 1.8818, "step": 5174 }, { "epoch": 0.0005175, "grad_norm": 3.116915702819824, "learning_rate": 5.174e-05, "loss": 2.1123, "step": 5175 }, { "epoch": 0.0005176, "grad_norm": 2.8182451725006104, "learning_rate": 5.1750000000000004e-05, "loss": 2.0039, "step": 5176 }, { "epoch": 0.0005177, "grad_norm": 2.9046666622161865, "learning_rate": 5.176e-05, "loss": 2.1445, "step": 5177 }, { "epoch": 0.0005178, "grad_norm": 4.111274242401123, "learning_rate": 5.177e-05, "loss": 2.5596, "step": 5178 }, { "epoch": 0.0005179, "grad_norm": 2.7475149631500244, "learning_rate": 5.178e-05, "loss": 1.8574, "step": 5179 }, { "epoch": 0.000518, "grad_norm": 3.320178747177124, "learning_rate": 5.179e-05, "loss": 2.3477, "step": 5180 }, { "epoch": 0.0005181, "grad_norm": 2.759566307067871, "learning_rate": 5.18e-05, "loss": 2.1904, "step": 5181 }, { "epoch": 0.0005182, "grad_norm": 2.4830856323242188, "learning_rate": 5.181e-05, "loss": 1.918, "step": 5182 }, { "epoch": 0.0005183, "grad_norm": 2.4625067710876465, "learning_rate": 5.182000000000001e-05, "loss": 1.8916, "step": 5183 }, { "epoch": 0.0005184, "grad_norm": 3.5898115634918213, "learning_rate": 5.183e-05, "loss": 2.4189, "step": 5184 }, { "epoch": 0.0005185, "grad_norm": 2.713151216506958, "learning_rate": 5.1840000000000005e-05, "loss": 2.0664, "step": 5185 }, { "epoch": 0.0005186, "grad_norm": 2.9408843517303467, "learning_rate": 5.185000000000001e-05, "loss": 2.0312, "step": 5186 }, { "epoch": 0.0005187, "grad_norm": 3.285951852798462, "learning_rate": 5.1859999999999995e-05, "loss": 2.2158, "step": 5187 }, { "epoch": 0.0005188, "grad_norm": 3.200726270675659, "learning_rate": 5.187e-05, "loss": 2.0479, "step": 5188 }, { "epoch": 0.0005189, "grad_norm": 2.883833408355713, "learning_rate": 5.1880000000000005e-05, "loss": 2.0918, "step": 5189 }, { "epoch": 0.000519, "grad_norm": 2.7511188983917236, "learning_rate": 5.189e-05, "loss": 2.0869, "step": 5190 }, { "epoch": 0.0005191, "grad_norm": 3.7801051139831543, "learning_rate": 5.19e-05, "loss": 2.499, "step": 5191 }, { "epoch": 0.0005192, "grad_norm": 3.4729905128479004, "learning_rate": 5.191e-05, "loss": 2.3965, "step": 5192 }, { "epoch": 0.0005193, "grad_norm": 2.659416437149048, "learning_rate": 5.1920000000000004e-05, "loss": 1.9238, "step": 5193 }, { "epoch": 0.0005194, "grad_norm": 2.6294078826904297, "learning_rate": 5.193e-05, "loss": 1.8867, "step": 5194 }, { "epoch": 0.0005195, "grad_norm": 2.982715129852295, "learning_rate": 5.194e-05, "loss": 1.9775, "step": 5195 }, { "epoch": 0.0005196, "grad_norm": 2.6217870712280273, "learning_rate": 5.195e-05, "loss": 2.0791, "step": 5196 }, { "epoch": 0.0005197, "grad_norm": 2.845123529434204, "learning_rate": 5.196e-05, "loss": 2.042, "step": 5197 }, { "epoch": 0.0005198, "grad_norm": 2.706462860107422, "learning_rate": 5.197e-05, "loss": 2.1406, "step": 5198 }, { "epoch": 0.0005199, "grad_norm": 2.5752992630004883, "learning_rate": 5.198e-05, "loss": 1.9395, "step": 5199 }, { "epoch": 0.00052, "grad_norm": 2.7343437671661377, "learning_rate": 5.199000000000001e-05, "loss": 2.1201, "step": 5200 }, { "epoch": 0.0005201, "grad_norm": 2.7991204261779785, "learning_rate": 5.2e-05, "loss": 1.9277, "step": 5201 }, { "epoch": 0.0005202, "grad_norm": 2.7249622344970703, "learning_rate": 5.201e-05, "loss": 1.9014, "step": 5202 }, { "epoch": 0.0005203, "grad_norm": 2.9683988094329834, "learning_rate": 5.202000000000001e-05, "loss": 2.3301, "step": 5203 }, { "epoch": 0.0005204, "grad_norm": 2.549675464630127, "learning_rate": 5.2029999999999995e-05, "loss": 2.0576, "step": 5204 }, { "epoch": 0.0005205, "grad_norm": 2.884549617767334, "learning_rate": 5.204e-05, "loss": 2.1162, "step": 5205 }, { "epoch": 0.0005206, "grad_norm": 3.4811131954193115, "learning_rate": 5.2050000000000005e-05, "loss": 2.1865, "step": 5206 }, { "epoch": 0.0005207, "grad_norm": 2.9320521354675293, "learning_rate": 5.206e-05, "loss": 2.1318, "step": 5207 }, { "epoch": 0.0005208, "grad_norm": 2.473130464553833, "learning_rate": 5.207e-05, "loss": 1.8008, "step": 5208 }, { "epoch": 0.0005209, "grad_norm": 2.896770715713501, "learning_rate": 5.208e-05, "loss": 2.0898, "step": 5209 }, { "epoch": 0.000521, "grad_norm": 2.777808666229248, "learning_rate": 5.2090000000000004e-05, "loss": 2.249, "step": 5210 }, { "epoch": 0.0005211, "grad_norm": 2.492285966873169, "learning_rate": 5.21e-05, "loss": 1.9082, "step": 5211 }, { "epoch": 0.0005212, "grad_norm": 3.5880565643310547, "learning_rate": 5.211e-05, "loss": 1.9844, "step": 5212 }, { "epoch": 0.0005213, "grad_norm": 2.712815999984741, "learning_rate": 5.212e-05, "loss": 1.7871, "step": 5213 }, { "epoch": 0.0005214, "grad_norm": 3.076080799102783, "learning_rate": 5.213e-05, "loss": 2.0342, "step": 5214 }, { "epoch": 0.0005215, "grad_norm": 2.8615427017211914, "learning_rate": 5.214e-05, "loss": 1.9443, "step": 5215 }, { "epoch": 0.0005216, "grad_norm": 2.6773412227630615, "learning_rate": 5.215e-05, "loss": 2.0049, "step": 5216 }, { "epoch": 0.0005217, "grad_norm": 3.6197516918182373, "learning_rate": 5.216000000000001e-05, "loss": 2.2383, "step": 5217 }, { "epoch": 0.0005218, "grad_norm": 3.092571496963501, "learning_rate": 5.217e-05, "loss": 2.4883, "step": 5218 }, { "epoch": 0.0005219, "grad_norm": 2.5264196395874023, "learning_rate": 5.218e-05, "loss": 1.7646, "step": 5219 }, { "epoch": 0.000522, "grad_norm": 2.3898637294769287, "learning_rate": 5.219000000000001e-05, "loss": 1.665, "step": 5220 }, { "epoch": 0.0005221, "grad_norm": 3.0501036643981934, "learning_rate": 5.2199999999999995e-05, "loss": 2.1123, "step": 5221 }, { "epoch": 0.0005222, "grad_norm": 2.692413568496704, "learning_rate": 5.221e-05, "loss": 1.9707, "step": 5222 }, { "epoch": 0.0005223, "grad_norm": 2.7069902420043945, "learning_rate": 5.2220000000000005e-05, "loss": 1.9648, "step": 5223 }, { "epoch": 0.0005224, "grad_norm": 2.856410503387451, "learning_rate": 5.222999999999999e-05, "loss": 1.8457, "step": 5224 }, { "epoch": 0.0005225, "grad_norm": 2.6665804386138916, "learning_rate": 5.224e-05, "loss": 1.9395, "step": 5225 }, { "epoch": 0.0005226, "grad_norm": 3.036193370819092, "learning_rate": 5.225e-05, "loss": 2.3096, "step": 5226 }, { "epoch": 0.0005227, "grad_norm": 2.7729709148406982, "learning_rate": 5.2260000000000004e-05, "loss": 1.8545, "step": 5227 }, { "epoch": 0.0005228, "grad_norm": 2.9586005210876465, "learning_rate": 5.227e-05, "loss": 1.9775, "step": 5228 }, { "epoch": 0.0005229, "grad_norm": 2.8372600078582764, "learning_rate": 5.228e-05, "loss": 1.9629, "step": 5229 }, { "epoch": 0.000523, "grad_norm": 3.4622883796691895, "learning_rate": 5.229e-05, "loss": 2.375, "step": 5230 }, { "epoch": 0.0005231, "grad_norm": 2.7652385234832764, "learning_rate": 5.23e-05, "loss": 2.0879, "step": 5231 }, { "epoch": 0.0005232, "grad_norm": 3.2243518829345703, "learning_rate": 5.231e-05, "loss": 2.1035, "step": 5232 }, { "epoch": 0.0005233, "grad_norm": 3.71882700920105, "learning_rate": 5.232e-05, "loss": 2.0879, "step": 5233 }, { "epoch": 0.0005234, "grad_norm": 2.428622007369995, "learning_rate": 5.233000000000001e-05, "loss": 1.6504, "step": 5234 }, { "epoch": 0.0005235, "grad_norm": 3.852372646331787, "learning_rate": 5.234e-05, "loss": 2.2148, "step": 5235 }, { "epoch": 0.0005236, "grad_norm": 3.200077533721924, "learning_rate": 5.235e-05, "loss": 1.9023, "step": 5236 }, { "epoch": 0.0005237, "grad_norm": 3.1853668689727783, "learning_rate": 5.236000000000001e-05, "loss": 2.2549, "step": 5237 }, { "epoch": 0.0005238, "grad_norm": 2.680206298828125, "learning_rate": 5.2369999999999995e-05, "loss": 1.833, "step": 5238 }, { "epoch": 0.0005239, "grad_norm": 2.963463544845581, "learning_rate": 5.238e-05, "loss": 2.1279, "step": 5239 }, { "epoch": 0.000524, "grad_norm": 3.481890916824341, "learning_rate": 5.2390000000000005e-05, "loss": 2.293, "step": 5240 }, { "epoch": 0.0005241, "grad_norm": 2.651437997817993, "learning_rate": 5.239999999999999e-05, "loss": 1.8799, "step": 5241 }, { "epoch": 0.0005242, "grad_norm": 2.7418296337127686, "learning_rate": 5.241e-05, "loss": 1.9316, "step": 5242 }, { "epoch": 0.0005243, "grad_norm": 2.738659381866455, "learning_rate": 5.242e-05, "loss": 1.8418, "step": 5243 }, { "epoch": 0.0005244, "grad_norm": 2.30897855758667, "learning_rate": 5.2430000000000005e-05, "loss": 1.6875, "step": 5244 }, { "epoch": 0.0005245, "grad_norm": 2.527406930923462, "learning_rate": 5.244e-05, "loss": 1.7344, "step": 5245 }, { "epoch": 0.0005246, "grad_norm": 2.5042757987976074, "learning_rate": 5.245e-05, "loss": 1.7637, "step": 5246 }, { "epoch": 0.0005247, "grad_norm": 2.7171101570129395, "learning_rate": 5.246e-05, "loss": 1.7949, "step": 5247 }, { "epoch": 0.0005248, "grad_norm": 2.630870819091797, "learning_rate": 5.247e-05, "loss": 1.9092, "step": 5248 }, { "epoch": 0.0005249, "grad_norm": 2.4991679191589355, "learning_rate": 5.248e-05, "loss": 1.6514, "step": 5249 }, { "epoch": 0.000525, "grad_norm": 2.609124183654785, "learning_rate": 5.249e-05, "loss": 1.9043, "step": 5250 }, { "epoch": 0.0005251, "grad_norm": 2.7172188758850098, "learning_rate": 5.250000000000001e-05, "loss": 1.9033, "step": 5251 }, { "epoch": 0.0005252, "grad_norm": 3.3129701614379883, "learning_rate": 5.251e-05, "loss": 2.208, "step": 5252 }, { "epoch": 0.0005253, "grad_norm": 2.5609395503997803, "learning_rate": 5.252e-05, "loss": 1.7256, "step": 5253 }, { "epoch": 0.0005254, "grad_norm": 5.410651683807373, "learning_rate": 5.253000000000001e-05, "loss": 2.1953, "step": 5254 }, { "epoch": 0.0005255, "grad_norm": 2.7116360664367676, "learning_rate": 5.2539999999999995e-05, "loss": 1.8037, "step": 5255 }, { "epoch": 0.0005256, "grad_norm": 3.9210119247436523, "learning_rate": 5.2550000000000003e-05, "loss": 2.0488, "step": 5256 }, { "epoch": 0.0005257, "grad_norm": 2.630709409713745, "learning_rate": 5.2560000000000005e-05, "loss": 1.8486, "step": 5257 }, { "epoch": 0.0005258, "grad_norm": 3.126084327697754, "learning_rate": 5.256999999999999e-05, "loss": 2.1729, "step": 5258 }, { "epoch": 0.0005259, "grad_norm": 2.6257944107055664, "learning_rate": 5.258e-05, "loss": 1.7988, "step": 5259 }, { "epoch": 0.000526, "grad_norm": 2.694988250732422, "learning_rate": 5.259e-05, "loss": 1.9502, "step": 5260 }, { "epoch": 0.0005261, "grad_norm": 2.4441659450531006, "learning_rate": 5.2600000000000005e-05, "loss": 1.7275, "step": 5261 }, { "epoch": 0.0005262, "grad_norm": 2.9759435653686523, "learning_rate": 5.261e-05, "loss": 1.9443, "step": 5262 }, { "epoch": 0.0005263, "grad_norm": 2.525765895843506, "learning_rate": 5.262e-05, "loss": 1.709, "step": 5263 }, { "epoch": 0.0005264, "grad_norm": 2.5012660026550293, "learning_rate": 5.263e-05, "loss": 1.709, "step": 5264 }, { "epoch": 0.0005265, "grad_norm": 4.091911315917969, "learning_rate": 5.264e-05, "loss": 1.8252, "step": 5265 }, { "epoch": 0.0005266, "grad_norm": 2.628835678100586, "learning_rate": 5.265e-05, "loss": 1.9697, "step": 5266 }, { "epoch": 0.0005267, "grad_norm": 2.539673089981079, "learning_rate": 5.266e-05, "loss": 1.6875, "step": 5267 }, { "epoch": 0.0005268, "grad_norm": 2.3051164150238037, "learning_rate": 5.267000000000001e-05, "loss": 1.6104, "step": 5268 }, { "epoch": 0.0005269, "grad_norm": 2.9399917125701904, "learning_rate": 5.268e-05, "loss": 2.0635, "step": 5269 }, { "epoch": 0.000527, "grad_norm": 2.5019099712371826, "learning_rate": 5.269e-05, "loss": 1.6582, "step": 5270 }, { "epoch": 0.0005271, "grad_norm": 2.471386671066284, "learning_rate": 5.270000000000001e-05, "loss": 1.6484, "step": 5271 }, { "epoch": 0.0005272, "grad_norm": 2.4753873348236084, "learning_rate": 5.2709999999999995e-05, "loss": 1.7109, "step": 5272 }, { "epoch": 0.0005273, "grad_norm": 2.5016117095947266, "learning_rate": 5.2720000000000003e-05, "loss": 1.8115, "step": 5273 }, { "epoch": 0.0005274, "grad_norm": 2.6271655559539795, "learning_rate": 5.2730000000000005e-05, "loss": 1.7959, "step": 5274 }, { "epoch": 0.0005275, "grad_norm": 2.996976375579834, "learning_rate": 5.273999999999999e-05, "loss": 2.209, "step": 5275 }, { "epoch": 0.0005276, "grad_norm": 3.049771547317505, "learning_rate": 5.275e-05, "loss": 2.1143, "step": 5276 }, { "epoch": 0.0005277, "grad_norm": 3.0326032638549805, "learning_rate": 5.276e-05, "loss": 1.792, "step": 5277 }, { "epoch": 0.0005278, "grad_norm": 2.563810348510742, "learning_rate": 5.2770000000000005e-05, "loss": 1.6973, "step": 5278 }, { "epoch": 0.0005279, "grad_norm": 2.5802128314971924, "learning_rate": 5.278e-05, "loss": 1.6865, "step": 5279 }, { "epoch": 0.000528, "grad_norm": 2.326371669769287, "learning_rate": 5.279e-05, "loss": 1.6816, "step": 5280 }, { "epoch": 0.0005281, "grad_norm": 2.368407726287842, "learning_rate": 5.28e-05, "loss": 1.6865, "step": 5281 }, { "epoch": 0.0005282, "grad_norm": 2.7306714057922363, "learning_rate": 5.281e-05, "loss": 1.7432, "step": 5282 }, { "epoch": 0.0005283, "grad_norm": 2.676426887512207, "learning_rate": 5.282e-05, "loss": 1.7002, "step": 5283 }, { "epoch": 0.0005284, "grad_norm": 2.4502899646759033, "learning_rate": 5.283e-05, "loss": 1.8301, "step": 5284 }, { "epoch": 0.0005285, "grad_norm": 2.9470925331115723, "learning_rate": 5.284000000000001e-05, "loss": 1.8936, "step": 5285 }, { "epoch": 0.0005286, "grad_norm": 2.7219319343566895, "learning_rate": 5.285e-05, "loss": 1.8193, "step": 5286 }, { "epoch": 0.0005287, "grad_norm": 2.831458806991577, "learning_rate": 5.286e-05, "loss": 1.8232, "step": 5287 }, { "epoch": 0.0005288, "grad_norm": 2.5182082653045654, "learning_rate": 5.287000000000001e-05, "loss": 1.7422, "step": 5288 }, { "epoch": 0.0005289, "grad_norm": 2.3378078937530518, "learning_rate": 5.2879999999999995e-05, "loss": 1.6768, "step": 5289 }, { "epoch": 0.000529, "grad_norm": 2.5693256855010986, "learning_rate": 5.2890000000000004e-05, "loss": 1.8086, "step": 5290 }, { "epoch": 0.0005291, "grad_norm": 2.9308154582977295, "learning_rate": 5.2900000000000005e-05, "loss": 1.7529, "step": 5291 }, { "epoch": 0.0005292, "grad_norm": 3.5053272247314453, "learning_rate": 5.290999999999999e-05, "loss": 1.8799, "step": 5292 }, { "epoch": 0.0005293, "grad_norm": 2.5281622409820557, "learning_rate": 5.292e-05, "loss": 1.5293, "step": 5293 }, { "epoch": 0.0005294, "grad_norm": 2.4280104637145996, "learning_rate": 5.293e-05, "loss": 1.5576, "step": 5294 }, { "epoch": 0.0005295, "grad_norm": 2.7755777835845947, "learning_rate": 5.2940000000000005e-05, "loss": 1.6982, "step": 5295 }, { "epoch": 0.0005296, "grad_norm": 2.875701904296875, "learning_rate": 5.295e-05, "loss": 1.8311, "step": 5296 }, { "epoch": 0.0005297, "grad_norm": 2.4332783222198486, "learning_rate": 5.296e-05, "loss": 1.5664, "step": 5297 }, { "epoch": 0.0005298, "grad_norm": 2.3971004486083984, "learning_rate": 5.297e-05, "loss": 1.709, "step": 5298 }, { "epoch": 0.0005299, "grad_norm": 2.334643840789795, "learning_rate": 5.298e-05, "loss": 1.54, "step": 5299 }, { "epoch": 0.00053, "grad_norm": 3.004922866821289, "learning_rate": 5.299e-05, "loss": 1.8135, "step": 5300 }, { "epoch": 0.0005301, "grad_norm": 2.4877896308898926, "learning_rate": 5.3e-05, "loss": 1.7168, "step": 5301 }, { "epoch": 0.0005302, "grad_norm": 2.489755630493164, "learning_rate": 5.301000000000001e-05, "loss": 1.6885, "step": 5302 }, { "epoch": 0.0005303, "grad_norm": 2.7021472454071045, "learning_rate": 5.302e-05, "loss": 1.9102, "step": 5303 }, { "epoch": 0.0005304, "grad_norm": 2.4580464363098145, "learning_rate": 5.303e-05, "loss": 1.5957, "step": 5304 }, { "epoch": 0.0005305, "grad_norm": 2.707300901412964, "learning_rate": 5.304000000000001e-05, "loss": 1.8359, "step": 5305 }, { "epoch": 0.0005306, "grad_norm": 2.624448537826538, "learning_rate": 5.3049999999999995e-05, "loss": 1.7891, "step": 5306 }, { "epoch": 0.0005307, "grad_norm": 2.1923553943634033, "learning_rate": 5.3060000000000004e-05, "loss": 1.5332, "step": 5307 }, { "epoch": 0.0005308, "grad_norm": 2.433734655380249, "learning_rate": 5.3070000000000005e-05, "loss": 1.6494, "step": 5308 }, { "epoch": 0.0005309, "grad_norm": 2.423363208770752, "learning_rate": 5.307999999999999e-05, "loss": 1.499, "step": 5309 }, { "epoch": 0.000531, "grad_norm": 2.6677303314208984, "learning_rate": 5.309e-05, "loss": 1.7861, "step": 5310 }, { "epoch": 0.0005311, "grad_norm": 3.353668451309204, "learning_rate": 5.31e-05, "loss": 1.8145, "step": 5311 }, { "epoch": 0.0005312, "grad_norm": 2.8283190727233887, "learning_rate": 5.3110000000000005e-05, "loss": 1.9912, "step": 5312 }, { "epoch": 0.0005313, "grad_norm": 3.1612722873687744, "learning_rate": 5.312e-05, "loss": 1.8818, "step": 5313 }, { "epoch": 0.0005314, "grad_norm": 2.2225265502929688, "learning_rate": 5.313e-05, "loss": 1.4727, "step": 5314 }, { "epoch": 0.0005315, "grad_norm": 2.635708808898926, "learning_rate": 5.314e-05, "loss": 1.7402, "step": 5315 }, { "epoch": 0.0005316, "grad_norm": 2.3886332511901855, "learning_rate": 5.315e-05, "loss": 1.6172, "step": 5316 }, { "epoch": 0.0005317, "grad_norm": 2.6436541080474854, "learning_rate": 5.316e-05, "loss": 1.6514, "step": 5317 }, { "epoch": 0.0005318, "grad_norm": 2.3732078075408936, "learning_rate": 5.317e-05, "loss": 1.5977, "step": 5318 }, { "epoch": 0.0005319, "grad_norm": 3.435147285461426, "learning_rate": 5.318000000000001e-05, "loss": 2.084, "step": 5319 }, { "epoch": 0.000532, "grad_norm": 2.347984552383423, "learning_rate": 5.319e-05, "loss": 1.6348, "step": 5320 }, { "epoch": 0.0005321, "grad_norm": 2.342603921890259, "learning_rate": 5.32e-05, "loss": 1.6182, "step": 5321 }, { "epoch": 0.0005322, "grad_norm": 2.3926334381103516, "learning_rate": 5.321000000000001e-05, "loss": 1.6777, "step": 5322 }, { "epoch": 0.0005323, "grad_norm": 2.4277758598327637, "learning_rate": 5.3219999999999995e-05, "loss": 1.6787, "step": 5323 }, { "epoch": 0.0005324, "grad_norm": 2.8502721786499023, "learning_rate": 5.3230000000000004e-05, "loss": 1.9268, "step": 5324 }, { "epoch": 0.0005325, "grad_norm": 3.0685720443725586, "learning_rate": 5.3240000000000005e-05, "loss": 1.873, "step": 5325 }, { "epoch": 0.0005326, "grad_norm": 2.5644328594207764, "learning_rate": 5.324999999999999e-05, "loss": 1.6611, "step": 5326 }, { "epoch": 0.0005327, "grad_norm": 2.5484654903411865, "learning_rate": 5.326e-05, "loss": 1.583, "step": 5327 }, { "epoch": 0.0005328, "grad_norm": 2.2202749252319336, "learning_rate": 5.327e-05, "loss": 1.54, "step": 5328 }, { "epoch": 0.0005329, "grad_norm": 2.693093776702881, "learning_rate": 5.3280000000000005e-05, "loss": 1.7773, "step": 5329 }, { "epoch": 0.000533, "grad_norm": 2.495924472808838, "learning_rate": 5.329e-05, "loss": 1.7275, "step": 5330 }, { "epoch": 0.0005331, "grad_norm": 2.4281322956085205, "learning_rate": 5.33e-05, "loss": 1.4795, "step": 5331 }, { "epoch": 0.0005332, "grad_norm": 2.356987714767456, "learning_rate": 5.331e-05, "loss": 1.583, "step": 5332 }, { "epoch": 0.0005333, "grad_norm": 2.6030094623565674, "learning_rate": 5.332e-05, "loss": 1.6973, "step": 5333 }, { "epoch": 0.0005334, "grad_norm": 2.5493550300598145, "learning_rate": 5.333e-05, "loss": 1.6963, "step": 5334 }, { "epoch": 0.0005335, "grad_norm": 2.1884334087371826, "learning_rate": 5.334e-05, "loss": 1.5615, "step": 5335 }, { "epoch": 0.0005336, "grad_norm": 2.7196109294891357, "learning_rate": 5.335000000000001e-05, "loss": 1.7617, "step": 5336 }, { "epoch": 0.0005337, "grad_norm": 2.413569688796997, "learning_rate": 5.336e-05, "loss": 1.5742, "step": 5337 }, { "epoch": 0.0005338, "grad_norm": 2.500955581665039, "learning_rate": 5.337e-05, "loss": 1.624, "step": 5338 }, { "epoch": 0.0005339, "grad_norm": 3.1691110134124756, "learning_rate": 5.338000000000001e-05, "loss": 1.6748, "step": 5339 }, { "epoch": 0.000534, "grad_norm": 2.33241868019104, "learning_rate": 5.3389999999999995e-05, "loss": 1.5469, "step": 5340 }, { "epoch": 0.0005341, "grad_norm": 2.445294141769409, "learning_rate": 5.3400000000000004e-05, "loss": 1.6943, "step": 5341 }, { "epoch": 0.0005342, "grad_norm": 2.7654736042022705, "learning_rate": 5.3410000000000005e-05, "loss": 1.6025, "step": 5342 }, { "epoch": 0.0005343, "grad_norm": 2.441877603530884, "learning_rate": 5.341999999999999e-05, "loss": 1.6816, "step": 5343 }, { "epoch": 0.0005344, "grad_norm": 2.439685821533203, "learning_rate": 5.343e-05, "loss": 1.9189, "step": 5344 }, { "epoch": 0.0005345, "grad_norm": 2.590050220489502, "learning_rate": 5.344e-05, "loss": 1.6152, "step": 5345 }, { "epoch": 0.0005346, "grad_norm": 2.954296827316284, "learning_rate": 5.3450000000000005e-05, "loss": 1.7334, "step": 5346 }, { "epoch": 0.0005347, "grad_norm": 2.8516154289245605, "learning_rate": 5.346e-05, "loss": 1.8711, "step": 5347 }, { "epoch": 0.0005348, "grad_norm": 2.6574857234954834, "learning_rate": 5.347e-05, "loss": 1.5918, "step": 5348 }, { "epoch": 0.0005349, "grad_norm": 2.282672166824341, "learning_rate": 5.348e-05, "loss": 1.5469, "step": 5349 }, { "epoch": 0.000535, "grad_norm": 4.296674728393555, "learning_rate": 5.349e-05, "loss": 1.6055, "step": 5350 }, { "epoch": 0.0005351, "grad_norm": 2.4820706844329834, "learning_rate": 5.35e-05, "loss": 1.6865, "step": 5351 }, { "epoch": 0.0005352, "grad_norm": 2.322924852371216, "learning_rate": 5.351e-05, "loss": 1.5039, "step": 5352 }, { "epoch": 0.0005353, "grad_norm": 2.1149885654449463, "learning_rate": 5.352000000000001e-05, "loss": 1.4775, "step": 5353 }, { "epoch": 0.0005354, "grad_norm": 2.5442628860473633, "learning_rate": 5.353e-05, "loss": 1.6719, "step": 5354 }, { "epoch": 0.0005355, "grad_norm": 3.2219278812408447, "learning_rate": 5.354e-05, "loss": 1.7324, "step": 5355 }, { "epoch": 0.0005356, "grad_norm": 4.293951511383057, "learning_rate": 5.355000000000001e-05, "loss": 1.8223, "step": 5356 }, { "epoch": 0.0005357, "grad_norm": 2.4986095428466797, "learning_rate": 5.3559999999999995e-05, "loss": 1.5166, "step": 5357 }, { "epoch": 0.0005358, "grad_norm": 2.8759100437164307, "learning_rate": 5.3570000000000004e-05, "loss": 1.8359, "step": 5358 }, { "epoch": 0.0005359, "grad_norm": 3.0930676460266113, "learning_rate": 5.3580000000000005e-05, "loss": 1.7666, "step": 5359 }, { "epoch": 0.000536, "grad_norm": 2.6368706226348877, "learning_rate": 5.3589999999999993e-05, "loss": 1.6396, "step": 5360 }, { "epoch": 0.0005361, "grad_norm": 2.3945555686950684, "learning_rate": 5.36e-05, "loss": 1.625, "step": 5361 }, { "epoch": 0.0005362, "grad_norm": 2.3714077472686768, "learning_rate": 5.3610000000000003e-05, "loss": 1.5654, "step": 5362 }, { "epoch": 0.0005363, "grad_norm": 2.5613744258880615, "learning_rate": 5.3620000000000005e-05, "loss": 1.5488, "step": 5363 }, { "epoch": 0.0005364, "grad_norm": 3.337721586227417, "learning_rate": 5.363e-05, "loss": 1.6865, "step": 5364 }, { "epoch": 0.0005365, "grad_norm": 3.2178163528442383, "learning_rate": 5.364e-05, "loss": 1.8154, "step": 5365 }, { "epoch": 0.0005366, "grad_norm": 2.482443332672119, "learning_rate": 5.365e-05, "loss": 1.5127, "step": 5366 }, { "epoch": 0.0005367, "grad_norm": 2.6042063236236572, "learning_rate": 5.366e-05, "loss": 1.5986, "step": 5367 }, { "epoch": 0.0005368, "grad_norm": 2.479278087615967, "learning_rate": 5.367e-05, "loss": 1.6309, "step": 5368 }, { "epoch": 0.0005369, "grad_norm": 2.3373095989227295, "learning_rate": 5.368e-05, "loss": 1.5391, "step": 5369 }, { "epoch": 0.000537, "grad_norm": 3.385956287384033, "learning_rate": 5.369000000000001e-05, "loss": 2.126, "step": 5370 }, { "epoch": 0.0005371, "grad_norm": 2.3657467365264893, "learning_rate": 5.37e-05, "loss": 1.4805, "step": 5371 }, { "epoch": 0.0005372, "grad_norm": 2.325706958770752, "learning_rate": 5.371e-05, "loss": 1.5176, "step": 5372 }, { "epoch": 0.0005373, "grad_norm": 2.9394378662109375, "learning_rate": 5.372000000000001e-05, "loss": 1.8643, "step": 5373 }, { "epoch": 0.0005374, "grad_norm": 2.787156820297241, "learning_rate": 5.3729999999999995e-05, "loss": 1.6826, "step": 5374 }, { "epoch": 0.0005375, "grad_norm": 2.3524041175842285, "learning_rate": 5.3740000000000004e-05, "loss": 1.6025, "step": 5375 }, { "epoch": 0.0005376, "grad_norm": 2.445774793624878, "learning_rate": 5.3750000000000005e-05, "loss": 1.5508, "step": 5376 }, { "epoch": 0.0005377, "grad_norm": 2.358213186264038, "learning_rate": 5.3759999999999994e-05, "loss": 1.5996, "step": 5377 }, { "epoch": 0.0005378, "grad_norm": 3.349168539047241, "learning_rate": 5.377e-05, "loss": 1.9619, "step": 5378 }, { "epoch": 0.0005379, "grad_norm": 2.432954788208008, "learning_rate": 5.3780000000000003e-05, "loss": 1.6221, "step": 5379 }, { "epoch": 0.000538, "grad_norm": 2.489093065261841, "learning_rate": 5.3790000000000005e-05, "loss": 1.5654, "step": 5380 }, { "epoch": 0.0005381, "grad_norm": 2.402533531188965, "learning_rate": 5.38e-05, "loss": 1.543, "step": 5381 }, { "epoch": 0.0005382, "grad_norm": 2.391444206237793, "learning_rate": 5.381e-05, "loss": 1.5576, "step": 5382 }, { "epoch": 0.0005383, "grad_norm": 2.325573205947876, "learning_rate": 5.382e-05, "loss": 1.4961, "step": 5383 }, { "epoch": 0.0005384, "grad_norm": 2.971513509750366, "learning_rate": 5.383e-05, "loss": 1.5107, "step": 5384 }, { "epoch": 0.0005385, "grad_norm": 2.556778907775879, "learning_rate": 5.384e-05, "loss": 1.5684, "step": 5385 }, { "epoch": 0.0005386, "grad_norm": 2.5303313732147217, "learning_rate": 5.385e-05, "loss": 1.5771, "step": 5386 }, { "epoch": 0.0005387, "grad_norm": 3.4144351482391357, "learning_rate": 5.386000000000001e-05, "loss": 1.9766, "step": 5387 }, { "epoch": 0.0005388, "grad_norm": 2.114643096923828, "learning_rate": 5.387e-05, "loss": 1.3662, "step": 5388 }, { "epoch": 0.0005389, "grad_norm": 2.303209066390991, "learning_rate": 5.388e-05, "loss": 1.3975, "step": 5389 }, { "epoch": 0.000539, "grad_norm": 2.6278774738311768, "learning_rate": 5.389000000000001e-05, "loss": 1.5156, "step": 5390 }, { "epoch": 0.0005391, "grad_norm": 2.4299392700195312, "learning_rate": 5.3899999999999996e-05, "loss": 1.6436, "step": 5391 }, { "epoch": 0.0005392, "grad_norm": 2.1873764991760254, "learning_rate": 5.3910000000000004e-05, "loss": 1.4434, "step": 5392 }, { "epoch": 0.0005393, "grad_norm": 2.102019786834717, "learning_rate": 5.3920000000000006e-05, "loss": 1.4492, "step": 5393 }, { "epoch": 0.0005394, "grad_norm": 3.0071775913238525, "learning_rate": 5.3929999999999994e-05, "loss": 1.8877, "step": 5394 }, { "epoch": 0.0005395, "grad_norm": 2.4146597385406494, "learning_rate": 5.394e-05, "loss": 1.4932, "step": 5395 }, { "epoch": 0.0005396, "grad_norm": 2.377025604248047, "learning_rate": 5.3950000000000004e-05, "loss": 1.375, "step": 5396 }, { "epoch": 0.0005397, "grad_norm": 2.470691442489624, "learning_rate": 5.3960000000000005e-05, "loss": 1.5371, "step": 5397 }, { "epoch": 0.0005398, "grad_norm": 2.2135932445526123, "learning_rate": 5.397e-05, "loss": 1.457, "step": 5398 }, { "epoch": 0.0005399, "grad_norm": 2.7197883129119873, "learning_rate": 5.398e-05, "loss": 1.7334, "step": 5399 }, { "epoch": 0.00054, "grad_norm": 2.4931857585906982, "learning_rate": 5.399e-05, "loss": 1.5469, "step": 5400 }, { "epoch": 0.0005401, "grad_norm": 2.3643393516540527, "learning_rate": 5.4e-05, "loss": 1.5986, "step": 5401 }, { "epoch": 0.0005402, "grad_norm": 2.2726595401763916, "learning_rate": 5.401e-05, "loss": 1.4414, "step": 5402 }, { "epoch": 0.0005403, "grad_norm": 2.2128190994262695, "learning_rate": 5.402e-05, "loss": 1.4326, "step": 5403 }, { "epoch": 0.0005404, "grad_norm": 2.3215579986572266, "learning_rate": 5.403000000000001e-05, "loss": 1.583, "step": 5404 }, { "epoch": 0.0005405, "grad_norm": 2.369455099105835, "learning_rate": 5.404e-05, "loss": 1.5449, "step": 5405 }, { "epoch": 0.0005406, "grad_norm": 2.403362989425659, "learning_rate": 5.405e-05, "loss": 1.3945, "step": 5406 }, { "epoch": 0.0005407, "grad_norm": 2.757972478866577, "learning_rate": 5.406000000000001e-05, "loss": 1.7188, "step": 5407 }, { "epoch": 0.0005408, "grad_norm": 2.3139076232910156, "learning_rate": 5.4069999999999996e-05, "loss": 1.5791, "step": 5408 }, { "epoch": 0.0005409, "grad_norm": 2.235582113265991, "learning_rate": 5.4080000000000004e-05, "loss": 1.417, "step": 5409 }, { "epoch": 0.000541, "grad_norm": 2.6220693588256836, "learning_rate": 5.4090000000000006e-05, "loss": 1.4551, "step": 5410 }, { "epoch": 0.0005411, "grad_norm": 2.510016918182373, "learning_rate": 5.4099999999999994e-05, "loss": 1.4434, "step": 5411 }, { "epoch": 0.0005412, "grad_norm": 2.7129135131835938, "learning_rate": 5.411e-05, "loss": 1.667, "step": 5412 }, { "epoch": 0.0005413, "grad_norm": 2.8475699424743652, "learning_rate": 5.4120000000000004e-05, "loss": 1.6191, "step": 5413 }, { "epoch": 0.0005414, "grad_norm": 2.5072100162506104, "learning_rate": 5.4130000000000005e-05, "loss": 1.5146, "step": 5414 }, { "epoch": 0.0005415, "grad_norm": 2.40205454826355, "learning_rate": 5.414e-05, "loss": 1.4463, "step": 5415 }, { "epoch": 0.0005416, "grad_norm": 2.296178102493286, "learning_rate": 5.415e-05, "loss": 1.4883, "step": 5416 }, { "epoch": 0.0005417, "grad_norm": 2.315936803817749, "learning_rate": 5.416e-05, "loss": 1.5801, "step": 5417 }, { "epoch": 0.0005418, "grad_norm": 2.2993931770324707, "learning_rate": 5.417e-05, "loss": 1.4648, "step": 5418 }, { "epoch": 0.0005419, "grad_norm": 2.29580020904541, "learning_rate": 5.418e-05, "loss": 1.3447, "step": 5419 }, { "epoch": 0.000542, "grad_norm": 2.726243734359741, "learning_rate": 5.419e-05, "loss": 1.5488, "step": 5420 }, { "epoch": 0.0005421, "grad_norm": 2.1043827533721924, "learning_rate": 5.420000000000001e-05, "loss": 1.3857, "step": 5421 }, { "epoch": 0.0005422, "grad_norm": 2.0589137077331543, "learning_rate": 5.421e-05, "loss": 1.3369, "step": 5422 }, { "epoch": 0.0005423, "grad_norm": 2.4117438793182373, "learning_rate": 5.422e-05, "loss": 1.418, "step": 5423 }, { "epoch": 0.0005424, "grad_norm": 2.490020751953125, "learning_rate": 5.423000000000001e-05, "loss": 1.5176, "step": 5424 }, { "epoch": 0.0005425, "grad_norm": 2.1713314056396484, "learning_rate": 5.4239999999999996e-05, "loss": 1.3359, "step": 5425 }, { "epoch": 0.0005426, "grad_norm": 2.1438803672790527, "learning_rate": 5.4250000000000004e-05, "loss": 1.4229, "step": 5426 }, { "epoch": 0.0005427, "grad_norm": 2.459476947784424, "learning_rate": 5.4260000000000006e-05, "loss": 1.5195, "step": 5427 }, { "epoch": 0.0005428, "grad_norm": 3.274837017059326, "learning_rate": 5.4269999999999994e-05, "loss": 1.9453, "step": 5428 }, { "epoch": 0.0005429, "grad_norm": 3.361891031265259, "learning_rate": 5.428e-05, "loss": 1.6865, "step": 5429 }, { "epoch": 0.000543, "grad_norm": 2.532179832458496, "learning_rate": 5.4290000000000004e-05, "loss": 1.8223, "step": 5430 }, { "epoch": 0.0005431, "grad_norm": 2.4085769653320312, "learning_rate": 5.4300000000000005e-05, "loss": 1.4863, "step": 5431 }, { "epoch": 0.0005432, "grad_norm": 2.3143832683563232, "learning_rate": 5.431e-05, "loss": 1.5264, "step": 5432 }, { "epoch": 0.0005433, "grad_norm": 2.443618059158325, "learning_rate": 5.432e-05, "loss": 1.5771, "step": 5433 }, { "epoch": 0.0005434, "grad_norm": 2.2170422077178955, "learning_rate": 5.433e-05, "loss": 1.3857, "step": 5434 }, { "epoch": 0.0005435, "grad_norm": 2.8009684085845947, "learning_rate": 5.434e-05, "loss": 1.8516, "step": 5435 }, { "epoch": 0.0005436, "grad_norm": 2.5512821674346924, "learning_rate": 5.435e-05, "loss": 1.3818, "step": 5436 }, { "epoch": 0.0005437, "grad_norm": 2.259007215499878, "learning_rate": 5.436e-05, "loss": 1.4131, "step": 5437 }, { "epoch": 0.0005438, "grad_norm": 2.3607594966888428, "learning_rate": 5.437e-05, "loss": 1.5186, "step": 5438 }, { "epoch": 0.0005439, "grad_norm": 2.7615482807159424, "learning_rate": 5.438e-05, "loss": 1.5938, "step": 5439 }, { "epoch": 0.000544, "grad_norm": 2.274744749069214, "learning_rate": 5.439e-05, "loss": 1.3643, "step": 5440 }, { "epoch": 0.0005441, "grad_norm": 2.351226329803467, "learning_rate": 5.440000000000001e-05, "loss": 1.4658, "step": 5441 }, { "epoch": 0.0005442, "grad_norm": 2.4985721111297607, "learning_rate": 5.4409999999999996e-05, "loss": 1.4678, "step": 5442 }, { "epoch": 0.0005443, "grad_norm": 2.205420732498169, "learning_rate": 5.4420000000000004e-05, "loss": 1.4141, "step": 5443 }, { "epoch": 0.0005444, "grad_norm": 2.181335210800171, "learning_rate": 5.4430000000000006e-05, "loss": 1.4707, "step": 5444 }, { "epoch": 0.0005445, "grad_norm": 2.14846134185791, "learning_rate": 5.4439999999999994e-05, "loss": 1.3838, "step": 5445 }, { "epoch": 0.0005446, "grad_norm": 2.591862916946411, "learning_rate": 5.445e-05, "loss": 1.5635, "step": 5446 }, { "epoch": 0.0005447, "grad_norm": 2.171609878540039, "learning_rate": 5.4460000000000004e-05, "loss": 1.3662, "step": 5447 }, { "epoch": 0.0005448, "grad_norm": 2.163784980773926, "learning_rate": 5.4470000000000005e-05, "loss": 1.376, "step": 5448 }, { "epoch": 0.0005449, "grad_norm": 2.5450291633605957, "learning_rate": 5.448e-05, "loss": 1.7607, "step": 5449 }, { "epoch": 0.000545, "grad_norm": 2.155191659927368, "learning_rate": 5.449e-05, "loss": 1.3467, "step": 5450 }, { "epoch": 0.0005451, "grad_norm": 2.547494649887085, "learning_rate": 5.45e-05, "loss": 1.5869, "step": 5451 }, { "epoch": 0.0005452, "grad_norm": 2.2871603965759277, "learning_rate": 5.451e-05, "loss": 1.4316, "step": 5452 }, { "epoch": 0.0005453, "grad_norm": 2.253500461578369, "learning_rate": 5.452e-05, "loss": 1.3955, "step": 5453 }, { "epoch": 0.0005454, "grad_norm": 2.349235773086548, "learning_rate": 5.453e-05, "loss": 1.4502, "step": 5454 }, { "epoch": 0.0005455, "grad_norm": 3.11142897605896, "learning_rate": 5.454e-05, "loss": 1.8438, "step": 5455 }, { "epoch": 0.0005456, "grad_norm": 2.2857236862182617, "learning_rate": 5.455e-05, "loss": 1.3721, "step": 5456 }, { "epoch": 0.0005457, "grad_norm": 2.2417335510253906, "learning_rate": 5.456e-05, "loss": 1.3975, "step": 5457 }, { "epoch": 0.0005458, "grad_norm": 3.005403518676758, "learning_rate": 5.457000000000001e-05, "loss": 1.6758, "step": 5458 }, { "epoch": 0.0005459, "grad_norm": 2.102241277694702, "learning_rate": 5.4579999999999996e-05, "loss": 1.3369, "step": 5459 }, { "epoch": 0.000546, "grad_norm": 2.5519657135009766, "learning_rate": 5.459e-05, "loss": 1.6338, "step": 5460 }, { "epoch": 0.0005461, "grad_norm": 2.071059226989746, "learning_rate": 5.4600000000000006e-05, "loss": 1.3574, "step": 5461 }, { "epoch": 0.0005462, "grad_norm": 2.160262107849121, "learning_rate": 5.4609999999999994e-05, "loss": 1.3955, "step": 5462 }, { "epoch": 0.0005463, "grad_norm": 2.2217299938201904, "learning_rate": 5.462e-05, "loss": 1.5186, "step": 5463 }, { "epoch": 0.0005464, "grad_norm": 1.9684492349624634, "learning_rate": 5.4630000000000004e-05, "loss": 1.334, "step": 5464 }, { "epoch": 0.0005465, "grad_norm": 2.551914930343628, "learning_rate": 5.4640000000000005e-05, "loss": 1.5898, "step": 5465 }, { "epoch": 0.0005466, "grad_norm": 2.3025529384613037, "learning_rate": 5.465e-05, "loss": 1.4629, "step": 5466 }, { "epoch": 0.0005467, "grad_norm": 2.422956943511963, "learning_rate": 5.466e-05, "loss": 1.4414, "step": 5467 }, { "epoch": 0.0005468, "grad_norm": 2.205836772918701, "learning_rate": 5.4670000000000003e-05, "loss": 1.457, "step": 5468 }, { "epoch": 0.0005469, "grad_norm": 2.3527743816375732, "learning_rate": 5.468e-05, "loss": 1.4775, "step": 5469 }, { "epoch": 0.000547, "grad_norm": 2.5781548023223877, "learning_rate": 5.469e-05, "loss": 1.5664, "step": 5470 }, { "epoch": 0.0005471, "grad_norm": 2.1942501068115234, "learning_rate": 5.47e-05, "loss": 1.3721, "step": 5471 }, { "epoch": 0.0005472, "grad_norm": 2.7376623153686523, "learning_rate": 5.471e-05, "loss": 1.7178, "step": 5472 }, { "epoch": 0.0005473, "grad_norm": 2.301755428314209, "learning_rate": 5.472e-05, "loss": 1.4141, "step": 5473 }, { "epoch": 0.0005474, "grad_norm": 2.1111536026000977, "learning_rate": 5.473e-05, "loss": 1.4219, "step": 5474 }, { "epoch": 0.0005475, "grad_norm": 3.3520498275756836, "learning_rate": 5.474000000000001e-05, "loss": 1.8047, "step": 5475 }, { "epoch": 0.0005476, "grad_norm": 2.1921098232269287, "learning_rate": 5.4749999999999996e-05, "loss": 1.3828, "step": 5476 }, { "epoch": 0.0005477, "grad_norm": 2.135024309158325, "learning_rate": 5.476e-05, "loss": 1.3682, "step": 5477 }, { "epoch": 0.0005478, "grad_norm": 2.2209396362304688, "learning_rate": 5.4770000000000006e-05, "loss": 1.3096, "step": 5478 }, { "epoch": 0.0005479, "grad_norm": 2.488781452178955, "learning_rate": 5.4779999999999994e-05, "loss": 1.4404, "step": 5479 }, { "epoch": 0.000548, "grad_norm": 2.290112018585205, "learning_rate": 5.479e-05, "loss": 1.4746, "step": 5480 }, { "epoch": 0.0005481, "grad_norm": 2.1676511764526367, "learning_rate": 5.4800000000000004e-05, "loss": 1.4463, "step": 5481 }, { "epoch": 0.0005482, "grad_norm": 2.507033586502075, "learning_rate": 5.4810000000000005e-05, "loss": 1.5107, "step": 5482 }, { "epoch": 0.0005483, "grad_norm": 2.31596040725708, "learning_rate": 5.482e-05, "loss": 1.375, "step": 5483 }, { "epoch": 0.0005484, "grad_norm": 3.8654398918151855, "learning_rate": 5.483e-05, "loss": 1.7744, "step": 5484 }, { "epoch": 0.0005485, "grad_norm": 2.5075442790985107, "learning_rate": 5.4840000000000003e-05, "loss": 1.3965, "step": 5485 }, { "epoch": 0.0005486, "grad_norm": 3.3337368965148926, "learning_rate": 5.485e-05, "loss": 1.5781, "step": 5486 }, { "epoch": 0.0005487, "grad_norm": 3.0748884677886963, "learning_rate": 5.486e-05, "loss": 1.4492, "step": 5487 }, { "epoch": 0.0005488, "grad_norm": 2.5058789253234863, "learning_rate": 5.487e-05, "loss": 1.3145, "step": 5488 }, { "epoch": 0.0005489, "grad_norm": 2.5538747310638428, "learning_rate": 5.488e-05, "loss": 1.3896, "step": 5489 }, { "epoch": 0.000549, "grad_norm": 2.413107395172119, "learning_rate": 5.489e-05, "loss": 1.3086, "step": 5490 }, { "epoch": 0.0005491, "grad_norm": 3.184574842453003, "learning_rate": 5.49e-05, "loss": 1.707, "step": 5491 }, { "epoch": 0.0005492, "grad_norm": 2.4939467906951904, "learning_rate": 5.491000000000001e-05, "loss": 1.4131, "step": 5492 }, { "epoch": 0.0005493, "grad_norm": 2.280385732650757, "learning_rate": 5.4919999999999996e-05, "loss": 1.2871, "step": 5493 }, { "epoch": 0.0005494, "grad_norm": 2.1914384365081787, "learning_rate": 5.493e-05, "loss": 1.375, "step": 5494 }, { "epoch": 0.0005495, "grad_norm": 2.5864098072052, "learning_rate": 5.4940000000000006e-05, "loss": 1.7012, "step": 5495 }, { "epoch": 0.0005496, "grad_norm": 3.2460954189300537, "learning_rate": 5.4949999999999994e-05, "loss": 2.0234, "step": 5496 }, { "epoch": 0.0005497, "grad_norm": 3.1469368934631348, "learning_rate": 5.496e-05, "loss": 1.6904, "step": 5497 }, { "epoch": 0.0005498, "grad_norm": 2.5989389419555664, "learning_rate": 5.4970000000000004e-05, "loss": 1.4248, "step": 5498 }, { "epoch": 0.0005499, "grad_norm": 2.6507346630096436, "learning_rate": 5.4980000000000006e-05, "loss": 1.5381, "step": 5499 }, { "epoch": 0.00055, "grad_norm": 2.250006914138794, "learning_rate": 5.499e-05, "loss": 1.4316, "step": 5500 }, { "epoch": 0.00055, "eval_loss": 0.16117964684963226, "eval_runtime": 365.4459, "eval_samples_per_second": 27.364, "eval_steps_per_second": 1.71, "step": 5500 }, { "epoch": 0.0005501, "grad_norm": 2.1657469272613525, "learning_rate": 5.5e-05, "loss": 1.375, "step": 5501 }, { "epoch": 0.0005502, "grad_norm": 2.278029680252075, "learning_rate": 5.5010000000000004e-05, "loss": 1.375, "step": 5502 }, { "epoch": 0.0005503, "grad_norm": 2.653407335281372, "learning_rate": 5.502e-05, "loss": 1.7002, "step": 5503 }, { "epoch": 0.0005504, "grad_norm": 2.2126805782318115, "learning_rate": 5.503e-05, "loss": 1.3184, "step": 5504 }, { "epoch": 0.0005505, "grad_norm": 2.2534098625183105, "learning_rate": 5.504e-05, "loss": 1.3574, "step": 5505 }, { "epoch": 0.0005506, "grad_norm": 2.7865779399871826, "learning_rate": 5.505e-05, "loss": 1.6641, "step": 5506 }, { "epoch": 0.0005507, "grad_norm": 3.0122759342193604, "learning_rate": 5.506e-05, "loss": 1.7051, "step": 5507 }, { "epoch": 0.0005508, "grad_norm": 2.089824914932251, "learning_rate": 5.507e-05, "loss": 1.2598, "step": 5508 }, { "epoch": 0.0005509, "grad_norm": 3.3949615955352783, "learning_rate": 5.508000000000001e-05, "loss": 1.8682, "step": 5509 }, { "epoch": 0.000551, "grad_norm": 5.518685817718506, "learning_rate": 5.5089999999999996e-05, "loss": 1.5371, "step": 5510 }, { "epoch": 0.0005511, "grad_norm": 3.192974328994751, "learning_rate": 5.51e-05, "loss": 1.6025, "step": 5511 }, { "epoch": 0.0005512, "grad_norm": 2.699589967727661, "learning_rate": 5.5110000000000006e-05, "loss": 1.4658, "step": 5512 }, { "epoch": 0.0005513, "grad_norm": 2.347198486328125, "learning_rate": 5.5119999999999994e-05, "loss": 1.4316, "step": 5513 }, { "epoch": 0.0005514, "grad_norm": 2.406765937805176, "learning_rate": 5.513e-05, "loss": 1.375, "step": 5514 }, { "epoch": 0.0005515, "grad_norm": 2.5467593669891357, "learning_rate": 5.5140000000000004e-05, "loss": 1.4297, "step": 5515 }, { "epoch": 0.0005516, "grad_norm": 2.204380512237549, "learning_rate": 5.5150000000000006e-05, "loss": 1.2383, "step": 5516 }, { "epoch": 0.0005517, "grad_norm": 2.2602663040161133, "learning_rate": 5.516e-05, "loss": 1.3105, "step": 5517 }, { "epoch": 0.0005518, "grad_norm": 4.868964672088623, "learning_rate": 5.517e-05, "loss": 1.9258, "step": 5518 }, { "epoch": 0.0005519, "grad_norm": 2.5829179286956787, "learning_rate": 5.5180000000000004e-05, "loss": 1.4307, "step": 5519 }, { "epoch": 0.000552, "grad_norm": 2.0237176418304443, "learning_rate": 5.519e-05, "loss": 1.2646, "step": 5520 }, { "epoch": 0.0005521, "grad_norm": 1.967319130897522, "learning_rate": 5.52e-05, "loss": 1.1914, "step": 5521 }, { "epoch": 0.0005522, "grad_norm": 2.0111095905303955, "learning_rate": 5.521e-05, "loss": 1.293, "step": 5522 }, { "epoch": 0.0005523, "grad_norm": 2.2081053256988525, "learning_rate": 5.522e-05, "loss": 1.418, "step": 5523 }, { "epoch": 0.0005524, "grad_norm": 2.613236904144287, "learning_rate": 5.523e-05, "loss": 1.4785, "step": 5524 }, { "epoch": 0.0005525, "grad_norm": 3.4059560298919678, "learning_rate": 5.524e-05, "loss": 1.6406, "step": 5525 }, { "epoch": 0.0005526, "grad_norm": 2.1261260509490967, "learning_rate": 5.525000000000001e-05, "loss": 1.3711, "step": 5526 }, { "epoch": 0.0005527, "grad_norm": 3.0109899044036865, "learning_rate": 5.5259999999999996e-05, "loss": 1.4688, "step": 5527 }, { "epoch": 0.0005528, "grad_norm": 2.294259786605835, "learning_rate": 5.527e-05, "loss": 1.4619, "step": 5528 }, { "epoch": 0.0005529, "grad_norm": 2.180101156234741, "learning_rate": 5.5280000000000006e-05, "loss": 1.4219, "step": 5529 }, { "epoch": 0.000553, "grad_norm": 2.065366268157959, "learning_rate": 5.5289999999999994e-05, "loss": 1.3633, "step": 5530 }, { "epoch": 0.0005531, "grad_norm": 2.6592462062835693, "learning_rate": 5.53e-05, "loss": 1.2856, "step": 5531 }, { "epoch": 0.0005532, "grad_norm": 2.605691432952881, "learning_rate": 5.5310000000000004e-05, "loss": 1.498, "step": 5532 }, { "epoch": 0.0005533, "grad_norm": 2.811755418777466, "learning_rate": 5.5320000000000006e-05, "loss": 1.7549, "step": 5533 }, { "epoch": 0.0005534, "grad_norm": 2.2970316410064697, "learning_rate": 5.533e-05, "loss": 1.4023, "step": 5534 }, { "epoch": 0.0005535, "grad_norm": 2.2041218280792236, "learning_rate": 5.534e-05, "loss": 1.4492, "step": 5535 }, { "epoch": 0.0005536, "grad_norm": 3.6363894939422607, "learning_rate": 5.5350000000000004e-05, "loss": 2.0703, "step": 5536 }, { "epoch": 0.0005537, "grad_norm": 2.3322548866271973, "learning_rate": 5.536e-05, "loss": 1.3018, "step": 5537 }, { "epoch": 0.0005538, "grad_norm": 2.0842840671539307, "learning_rate": 5.537e-05, "loss": 1.3398, "step": 5538 }, { "epoch": 0.0005539, "grad_norm": 2.181790828704834, "learning_rate": 5.538e-05, "loss": 1.333, "step": 5539 }, { "epoch": 0.000554, "grad_norm": 2.298056125640869, "learning_rate": 5.539e-05, "loss": 1.416, "step": 5540 }, { "epoch": 0.0005541, "grad_norm": 3.197896718978882, "learning_rate": 5.54e-05, "loss": 1.2461, "step": 5541 }, { "epoch": 0.0005542, "grad_norm": 2.2699851989746094, "learning_rate": 5.541e-05, "loss": 1.2783, "step": 5542 }, { "epoch": 0.0005543, "grad_norm": 2.420722484588623, "learning_rate": 5.542000000000001e-05, "loss": 1.5762, "step": 5543 }, { "epoch": 0.0005544, "grad_norm": 2.019577980041504, "learning_rate": 5.5429999999999996e-05, "loss": 1.2617, "step": 5544 }, { "epoch": 0.0005545, "grad_norm": 2.4632229804992676, "learning_rate": 5.544e-05, "loss": 1.396, "step": 5545 }, { "epoch": 0.0005546, "grad_norm": 2.2937636375427246, "learning_rate": 5.5450000000000006e-05, "loss": 1.3311, "step": 5546 }, { "epoch": 0.0005547, "grad_norm": 2.050861120223999, "learning_rate": 5.5459999999999994e-05, "loss": 1.2666, "step": 5547 }, { "epoch": 0.0005548, "grad_norm": 2.0719165802001953, "learning_rate": 5.547e-05, "loss": 1.3037, "step": 5548 }, { "epoch": 0.0005549, "grad_norm": 2.0907227993011475, "learning_rate": 5.5480000000000004e-05, "loss": 1.2842, "step": 5549 }, { "epoch": 0.000555, "grad_norm": 2.092015504837036, "learning_rate": 5.5490000000000006e-05, "loss": 1.2529, "step": 5550 }, { "epoch": 0.0005551, "grad_norm": 2.0380799770355225, "learning_rate": 5.55e-05, "loss": 1.2295, "step": 5551 }, { "epoch": 0.0005552, "grad_norm": 2.0194711685180664, "learning_rate": 5.551e-05, "loss": 1.1826, "step": 5552 }, { "epoch": 0.0005553, "grad_norm": 1.984717845916748, "learning_rate": 5.5520000000000004e-05, "loss": 1.2383, "step": 5553 }, { "epoch": 0.0005554, "grad_norm": 1.9688911437988281, "learning_rate": 5.553e-05, "loss": 1.2822, "step": 5554 }, { "epoch": 0.0005555, "grad_norm": 3.0318920612335205, "learning_rate": 5.554e-05, "loss": 1.8086, "step": 5555 }, { "epoch": 0.0005556, "grad_norm": 2.0116493701934814, "learning_rate": 5.555e-05, "loss": 1.1943, "step": 5556 }, { "epoch": 0.0005557, "grad_norm": 2.2734971046447754, "learning_rate": 5.5559999999999997e-05, "loss": 1.4131, "step": 5557 }, { "epoch": 0.0005558, "grad_norm": 2.3707101345062256, "learning_rate": 5.557e-05, "loss": 1.3682, "step": 5558 }, { "epoch": 0.0005559, "grad_norm": 1.9980303049087524, "learning_rate": 5.558e-05, "loss": 1.2549, "step": 5559 }, { "epoch": 0.000556, "grad_norm": 2.1238646507263184, "learning_rate": 5.559000000000001e-05, "loss": 1.2539, "step": 5560 }, { "epoch": 0.0005561, "grad_norm": 2.268760919570923, "learning_rate": 5.5599999999999996e-05, "loss": 1.4111, "step": 5561 }, { "epoch": 0.0005562, "grad_norm": 2.3243017196655273, "learning_rate": 5.561e-05, "loss": 1.373, "step": 5562 }, { "epoch": 0.0005563, "grad_norm": 2.0662007331848145, "learning_rate": 5.5620000000000006e-05, "loss": 1.3096, "step": 5563 }, { "epoch": 0.0005564, "grad_norm": 2.231890916824341, "learning_rate": 5.5629999999999994e-05, "loss": 1.373, "step": 5564 }, { "epoch": 0.0005565, "grad_norm": 2.1430857181549072, "learning_rate": 5.564e-05, "loss": 1.1699, "step": 5565 }, { "epoch": 0.0005566, "grad_norm": 1.9390007257461548, "learning_rate": 5.5650000000000004e-05, "loss": 1.2163, "step": 5566 }, { "epoch": 0.0005567, "grad_norm": 3.218642473220825, "learning_rate": 5.5660000000000006e-05, "loss": 1.9434, "step": 5567 }, { "epoch": 0.0005568, "grad_norm": 2.7889389991760254, "learning_rate": 5.567e-05, "loss": 1.7451, "step": 5568 }, { "epoch": 0.0005569, "grad_norm": 2.129279613494873, "learning_rate": 5.568e-05, "loss": 1.2842, "step": 5569 }, { "epoch": 0.000557, "grad_norm": 2.407432794570923, "learning_rate": 5.5690000000000004e-05, "loss": 1.4795, "step": 5570 }, { "epoch": 0.0005571, "grad_norm": 2.932892084121704, "learning_rate": 5.57e-05, "loss": 1.2832, "step": 5571 }, { "epoch": 0.0005572, "grad_norm": 2.2704365253448486, "learning_rate": 5.571e-05, "loss": 1.3706, "step": 5572 }, { "epoch": 0.0005573, "grad_norm": 2.6265451908111572, "learning_rate": 5.572e-05, "loss": 1.2607, "step": 5573 }, { "epoch": 0.0005574, "grad_norm": 2.1422863006591797, "learning_rate": 5.573e-05, "loss": 1.2617, "step": 5574 }, { "epoch": 0.0005575, "grad_norm": 2.9647858142852783, "learning_rate": 5.574e-05, "loss": 1.3901, "step": 5575 }, { "epoch": 0.0005576, "grad_norm": 2.489037036895752, "learning_rate": 5.575e-05, "loss": 1.3994, "step": 5576 }, { "epoch": 0.0005577, "grad_norm": 2.4596824645996094, "learning_rate": 5.576000000000001e-05, "loss": 1.3701, "step": 5577 }, { "epoch": 0.0005578, "grad_norm": 2.2265207767486572, "learning_rate": 5.5769999999999996e-05, "loss": 1.248, "step": 5578 }, { "epoch": 0.0005579, "grad_norm": 2.3037335872650146, "learning_rate": 5.578e-05, "loss": 1.333, "step": 5579 }, { "epoch": 0.000558, "grad_norm": 2.651296615600586, "learning_rate": 5.5790000000000006e-05, "loss": 1.4697, "step": 5580 }, { "epoch": 0.0005581, "grad_norm": 2.3579533100128174, "learning_rate": 5.5799999999999994e-05, "loss": 1.3623, "step": 5581 }, { "epoch": 0.0005582, "grad_norm": 2.523846387863159, "learning_rate": 5.581e-05, "loss": 1.415, "step": 5582 }, { "epoch": 0.0005583, "grad_norm": 2.5580360889434814, "learning_rate": 5.5820000000000004e-05, "loss": 1.4512, "step": 5583 }, { "epoch": 0.0005584, "grad_norm": 2.344586133956909, "learning_rate": 5.5830000000000006e-05, "loss": 1.3662, "step": 5584 }, { "epoch": 0.0005585, "grad_norm": 2.040177822113037, "learning_rate": 5.584e-05, "loss": 1.2173, "step": 5585 }, { "epoch": 0.0005586, "grad_norm": 1.9751789569854736, "learning_rate": 5.585e-05, "loss": 1.1479, "step": 5586 }, { "epoch": 0.0005587, "grad_norm": 2.2699475288391113, "learning_rate": 5.5860000000000004e-05, "loss": 1.2803, "step": 5587 }, { "epoch": 0.0005588, "grad_norm": 2.3292815685272217, "learning_rate": 5.587e-05, "loss": 1.3145, "step": 5588 }, { "epoch": 0.0005589, "grad_norm": 2.4007952213287354, "learning_rate": 5.588e-05, "loss": 1.2822, "step": 5589 }, { "epoch": 0.000559, "grad_norm": 2.313913345336914, "learning_rate": 5.589e-05, "loss": 1.2842, "step": 5590 }, { "epoch": 0.0005591, "grad_norm": 2.481891632080078, "learning_rate": 5.59e-05, "loss": 1.5146, "step": 5591 }, { "epoch": 0.0005592, "grad_norm": 2.1890406608581543, "learning_rate": 5.591e-05, "loss": 1.2607, "step": 5592 }, { "epoch": 0.0005593, "grad_norm": 2.6850531101226807, "learning_rate": 5.592e-05, "loss": 1.4482, "step": 5593 }, { "epoch": 0.0005594, "grad_norm": 2.2254996299743652, "learning_rate": 5.593000000000001e-05, "loss": 1.291, "step": 5594 }, { "epoch": 0.0005595, "grad_norm": 2.0163609981536865, "learning_rate": 5.5939999999999996e-05, "loss": 1.2124, "step": 5595 }, { "epoch": 0.0005596, "grad_norm": 2.391115188598633, "learning_rate": 5.595e-05, "loss": 1.4141, "step": 5596 }, { "epoch": 0.0005597, "grad_norm": 2.7599549293518066, "learning_rate": 5.5960000000000006e-05, "loss": 1.5264, "step": 5597 }, { "epoch": 0.0005598, "grad_norm": 2.434512138366699, "learning_rate": 5.5969999999999994e-05, "loss": 1.3662, "step": 5598 }, { "epoch": 0.0005599, "grad_norm": 2.0736303329467773, "learning_rate": 5.598e-05, "loss": 1.1714, "step": 5599 }, { "epoch": 0.00056, "grad_norm": 2.0648913383483887, "learning_rate": 5.5990000000000004e-05, "loss": 1.1797, "step": 5600 }, { "epoch": 0.0005601, "grad_norm": 4.771807670593262, "learning_rate": 5.6000000000000006e-05, "loss": 1.6016, "step": 5601 }, { "epoch": 0.0005602, "grad_norm": 2.7045605182647705, "learning_rate": 5.601e-05, "loss": 1.4512, "step": 5602 }, { "epoch": 0.0005603, "grad_norm": 2.5183730125427246, "learning_rate": 5.602e-05, "loss": 1.376, "step": 5603 }, { "epoch": 0.0005604, "grad_norm": 2.135667085647583, "learning_rate": 5.6030000000000004e-05, "loss": 1.3174, "step": 5604 }, { "epoch": 0.0005605, "grad_norm": 2.1617133617401123, "learning_rate": 5.604e-05, "loss": 1.248, "step": 5605 }, { "epoch": 0.0005606, "grad_norm": 2.7303905487060547, "learning_rate": 5.605e-05, "loss": 1.3955, "step": 5606 }, { "epoch": 0.0005607, "grad_norm": 2.1086387634277344, "learning_rate": 5.606e-05, "loss": 1.3164, "step": 5607 }, { "epoch": 0.0005608, "grad_norm": 2.3741819858551025, "learning_rate": 5.607e-05, "loss": 1.4766, "step": 5608 }, { "epoch": 0.0005609, "grad_norm": 2.0160882472991943, "learning_rate": 5.608e-05, "loss": 1.2295, "step": 5609 }, { "epoch": 0.000561, "grad_norm": 1.9973747730255127, "learning_rate": 5.609e-05, "loss": 1.1104, "step": 5610 }, { "epoch": 0.0005611, "grad_norm": 2.368208885192871, "learning_rate": 5.610000000000001e-05, "loss": 1.6182, "step": 5611 }, { "epoch": 0.0005612, "grad_norm": 3.3395214080810547, "learning_rate": 5.6109999999999996e-05, "loss": 2.1504, "step": 5612 }, { "epoch": 0.0005613, "grad_norm": 2.3721203804016113, "learning_rate": 5.612e-05, "loss": 1.3037, "step": 5613 }, { "epoch": 0.0005614, "grad_norm": 2.190359115600586, "learning_rate": 5.6130000000000006e-05, "loss": 1.167, "step": 5614 }, { "epoch": 0.0005615, "grad_norm": 2.519402503967285, "learning_rate": 5.6139999999999994e-05, "loss": 1.3716, "step": 5615 }, { "epoch": 0.0005616, "grad_norm": 2.859174966812134, "learning_rate": 5.615e-05, "loss": 1.5605, "step": 5616 }, { "epoch": 0.0005617, "grad_norm": 2.296678304672241, "learning_rate": 5.6160000000000004e-05, "loss": 1.1484, "step": 5617 }, { "epoch": 0.0005618, "grad_norm": 2.1283135414123535, "learning_rate": 5.6170000000000006e-05, "loss": 1.1885, "step": 5618 }, { "epoch": 0.0005619, "grad_norm": 2.090508222579956, "learning_rate": 5.618e-05, "loss": 1.2754, "step": 5619 }, { "epoch": 0.000562, "grad_norm": 2.2708656787872314, "learning_rate": 5.619e-05, "loss": 1.458, "step": 5620 }, { "epoch": 0.0005621, "grad_norm": 1.9705712795257568, "learning_rate": 5.6200000000000004e-05, "loss": 1.082, "step": 5621 }, { "epoch": 0.0005622, "grad_norm": 1.862807273864746, "learning_rate": 5.621e-05, "loss": 1.1211, "step": 5622 }, { "epoch": 0.0005623, "grad_norm": 2.0587210655212402, "learning_rate": 5.622e-05, "loss": 1.2495, "step": 5623 }, { "epoch": 0.0005624, "grad_norm": 1.9742335081100464, "learning_rate": 5.623e-05, "loss": 1.165, "step": 5624 }, { "epoch": 0.0005625, "grad_norm": 2.1360883712768555, "learning_rate": 5.624e-05, "loss": 1.1914, "step": 5625 }, { "epoch": 0.0005626, "grad_norm": 2.0543458461761475, "learning_rate": 5.625e-05, "loss": 1.1826, "step": 5626 }, { "epoch": 0.0005627, "grad_norm": 2.346017360687256, "learning_rate": 5.626e-05, "loss": 1.3242, "step": 5627 }, { "epoch": 0.0005628, "grad_norm": 2.727350950241089, "learning_rate": 5.627000000000001e-05, "loss": 1.417, "step": 5628 }, { "epoch": 0.0005629, "grad_norm": 2.4296886920928955, "learning_rate": 5.6279999999999996e-05, "loss": 1.188, "step": 5629 }, { "epoch": 0.000563, "grad_norm": 2.0857579708099365, "learning_rate": 5.629e-05, "loss": 1.1699, "step": 5630 }, { "epoch": 0.0005631, "grad_norm": 2.0088858604431152, "learning_rate": 5.6300000000000006e-05, "loss": 1.2188, "step": 5631 }, { "epoch": 0.0005632, "grad_norm": 2.1220197677612305, "learning_rate": 5.6309999999999994e-05, "loss": 1.2217, "step": 5632 }, { "epoch": 0.0005633, "grad_norm": 2.684288740158081, "learning_rate": 5.632e-05, "loss": 1.3774, "step": 5633 }, { "epoch": 0.0005634, "grad_norm": 2.1442840099334717, "learning_rate": 5.6330000000000004e-05, "loss": 1.2217, "step": 5634 }, { "epoch": 0.0005635, "grad_norm": 2.0995047092437744, "learning_rate": 5.6340000000000006e-05, "loss": 1.2295, "step": 5635 }, { "epoch": 0.0005636, "grad_norm": 2.410370111465454, "learning_rate": 5.635e-05, "loss": 1.2529, "step": 5636 }, { "epoch": 0.0005637, "grad_norm": 2.588259220123291, "learning_rate": 5.636e-05, "loss": 1.3447, "step": 5637 }, { "epoch": 0.0005638, "grad_norm": 2.3008322715759277, "learning_rate": 5.6370000000000004e-05, "loss": 1.2861, "step": 5638 }, { "epoch": 0.0005639, "grad_norm": 2.134218215942383, "learning_rate": 5.638e-05, "loss": 1.1191, "step": 5639 }, { "epoch": 0.000564, "grad_norm": 2.9954631328582764, "learning_rate": 5.639e-05, "loss": 1.3828, "step": 5640 }, { "epoch": 0.0005641, "grad_norm": 2.734889030456543, "learning_rate": 5.64e-05, "loss": 1.6357, "step": 5641 }, { "epoch": 0.0005642, "grad_norm": 2.154879093170166, "learning_rate": 5.641e-05, "loss": 1.2168, "step": 5642 }, { "epoch": 0.0005643, "grad_norm": 2.6774098873138428, "learning_rate": 5.642e-05, "loss": 1.5977, "step": 5643 }, { "epoch": 0.0005644, "grad_norm": 2.5120511054992676, "learning_rate": 5.643e-05, "loss": 1.3242, "step": 5644 }, { "epoch": 0.0005645, "grad_norm": 2.580930233001709, "learning_rate": 5.644000000000001e-05, "loss": 1.4297, "step": 5645 }, { "epoch": 0.0005646, "grad_norm": 2.058643341064453, "learning_rate": 5.6449999999999997e-05, "loss": 1.1025, "step": 5646 }, { "epoch": 0.0005647, "grad_norm": 1.924004077911377, "learning_rate": 5.646e-05, "loss": 1.1406, "step": 5647 }, { "epoch": 0.0005648, "grad_norm": 2.348186492919922, "learning_rate": 5.6470000000000007e-05, "loss": 1.2178, "step": 5648 }, { "epoch": 0.0005649, "grad_norm": 2.400103807449341, "learning_rate": 5.6479999999999995e-05, "loss": 1.333, "step": 5649 }, { "epoch": 0.000565, "grad_norm": 2.0180234909057617, "learning_rate": 5.649e-05, "loss": 1.0439, "step": 5650 }, { "epoch": 0.0005651, "grad_norm": 2.074004650115967, "learning_rate": 5.6500000000000005e-05, "loss": 1.1084, "step": 5651 }, { "epoch": 0.0005652, "grad_norm": 2.834411144256592, "learning_rate": 5.6510000000000006e-05, "loss": 1.3525, "step": 5652 }, { "epoch": 0.0005653, "grad_norm": 2.1253347396850586, "learning_rate": 5.652e-05, "loss": 1.1514, "step": 5653 }, { "epoch": 0.0005654, "grad_norm": 2.0319814682006836, "learning_rate": 5.653e-05, "loss": 1.1621, "step": 5654 }, { "epoch": 0.0005655, "grad_norm": 1.959982991218567, "learning_rate": 5.6540000000000004e-05, "loss": 1.1929, "step": 5655 }, { "epoch": 0.0005656, "grad_norm": 2.5499229431152344, "learning_rate": 5.655e-05, "loss": 1.4463, "step": 5656 }, { "epoch": 0.0005657, "grad_norm": 2.4173882007598877, "learning_rate": 5.656e-05, "loss": 1.2705, "step": 5657 }, { "epoch": 0.0005658, "grad_norm": 2.9077694416046143, "learning_rate": 5.657e-05, "loss": 1.418, "step": 5658 }, { "epoch": 0.0005659, "grad_norm": 2.3227171897888184, "learning_rate": 5.658e-05, "loss": 1.3213, "step": 5659 }, { "epoch": 0.000566, "grad_norm": 2.192533016204834, "learning_rate": 5.659e-05, "loss": 1.1631, "step": 5660 }, { "epoch": 0.0005661, "grad_norm": 2.154204845428467, "learning_rate": 5.66e-05, "loss": 1.2314, "step": 5661 }, { "epoch": 0.0005662, "grad_norm": 2.6472861766815186, "learning_rate": 5.661000000000001e-05, "loss": 1.4326, "step": 5662 }, { "epoch": 0.0005663, "grad_norm": 2.488671064376831, "learning_rate": 5.662e-05, "loss": 1.2031, "step": 5663 }, { "epoch": 0.0005664, "grad_norm": 2.7378642559051514, "learning_rate": 5.663e-05, "loss": 1.4014, "step": 5664 }, { "epoch": 0.0005665, "grad_norm": 2.089963436126709, "learning_rate": 5.6640000000000007e-05, "loss": 1.1357, "step": 5665 }, { "epoch": 0.0005666, "grad_norm": 2.295555353164673, "learning_rate": 5.6649999999999995e-05, "loss": 1.2432, "step": 5666 }, { "epoch": 0.0005667, "grad_norm": 3.187014102935791, "learning_rate": 5.666e-05, "loss": 1.2603, "step": 5667 }, { "epoch": 0.0005668, "grad_norm": 2.1170108318328857, "learning_rate": 5.6670000000000005e-05, "loss": 1.125, "step": 5668 }, { "epoch": 0.0005669, "grad_norm": 2.15708327293396, "learning_rate": 5.6680000000000006e-05, "loss": 1.1699, "step": 5669 }, { "epoch": 0.000567, "grad_norm": 1.9517823457717896, "learning_rate": 5.669e-05, "loss": 1.1523, "step": 5670 }, { "epoch": 0.0005671, "grad_norm": 2.6591954231262207, "learning_rate": 5.67e-05, "loss": 1.2305, "step": 5671 }, { "epoch": 0.0005672, "grad_norm": 1.9791878461837769, "learning_rate": 5.6710000000000004e-05, "loss": 1.0776, "step": 5672 }, { "epoch": 0.0005673, "grad_norm": 2.0013673305511475, "learning_rate": 5.672e-05, "loss": 1.1182, "step": 5673 }, { "epoch": 0.0005674, "grad_norm": 2.209170341491699, "learning_rate": 5.673e-05, "loss": 1.2578, "step": 5674 }, { "epoch": 0.0005675, "grad_norm": 2.586726665496826, "learning_rate": 5.674e-05, "loss": 1.3379, "step": 5675 }, { "epoch": 0.0005676, "grad_norm": 1.9457072019577026, "learning_rate": 5.675e-05, "loss": 1.0288, "step": 5676 }, { "epoch": 0.0005677, "grad_norm": 2.2997748851776123, "learning_rate": 5.676e-05, "loss": 1.2896, "step": 5677 }, { "epoch": 0.0005678, "grad_norm": 1.9310775995254517, "learning_rate": 5.677e-05, "loss": 1.2266, "step": 5678 }, { "epoch": 0.0005679, "grad_norm": 1.9865697622299194, "learning_rate": 5.678000000000001e-05, "loss": 1.1147, "step": 5679 }, { "epoch": 0.000568, "grad_norm": 1.958775281906128, "learning_rate": 5.679e-05, "loss": 1.0835, "step": 5680 }, { "epoch": 0.0005681, "grad_norm": 1.8496625423431396, "learning_rate": 5.68e-05, "loss": 1.0537, "step": 5681 }, { "epoch": 0.0005682, "grad_norm": 1.922885537147522, "learning_rate": 5.681000000000001e-05, "loss": 1.1465, "step": 5682 }, { "epoch": 0.0005683, "grad_norm": 1.9769484996795654, "learning_rate": 5.6819999999999995e-05, "loss": 1.0781, "step": 5683 }, { "epoch": 0.0005684, "grad_norm": 1.9770240783691406, "learning_rate": 5.683e-05, "loss": 1.1118, "step": 5684 }, { "epoch": 0.0005685, "grad_norm": 1.9034974575042725, "learning_rate": 5.6840000000000005e-05, "loss": 1.0576, "step": 5685 }, { "epoch": 0.0005686, "grad_norm": 2.0151827335357666, "learning_rate": 5.6850000000000006e-05, "loss": 1.0386, "step": 5686 }, { "epoch": 0.0005687, "grad_norm": 3.5754876136779785, "learning_rate": 5.686e-05, "loss": 1.6006, "step": 5687 }, { "epoch": 0.0005688, "grad_norm": 2.2252042293548584, "learning_rate": 5.687e-05, "loss": 1.1045, "step": 5688 }, { "epoch": 0.0005689, "grad_norm": 2.010793447494507, "learning_rate": 5.6880000000000004e-05, "loss": 1.0869, "step": 5689 }, { "epoch": 0.000569, "grad_norm": 1.8892912864685059, "learning_rate": 5.689e-05, "loss": 1.1279, "step": 5690 }, { "epoch": 0.0005691, "grad_norm": 1.9598063230514526, "learning_rate": 5.69e-05, "loss": 1.0938, "step": 5691 }, { "epoch": 0.0005692, "grad_norm": 1.9675894975662231, "learning_rate": 5.691e-05, "loss": 1.0811, "step": 5692 }, { "epoch": 0.0005693, "grad_norm": 1.9908654689788818, "learning_rate": 5.692e-05, "loss": 1.165, "step": 5693 }, { "epoch": 0.0005694, "grad_norm": 1.8950060606002808, "learning_rate": 5.693e-05, "loss": 1.1211, "step": 5694 }, { "epoch": 0.0005695, "grad_norm": 2.0623393058776855, "learning_rate": 5.694e-05, "loss": 1.124, "step": 5695 }, { "epoch": 0.0005696, "grad_norm": 2.491461753845215, "learning_rate": 5.695e-05, "loss": 1.3096, "step": 5696 }, { "epoch": 0.0005697, "grad_norm": 2.5069475173950195, "learning_rate": 5.696e-05, "loss": 1.3896, "step": 5697 }, { "epoch": 0.0005698, "grad_norm": 1.8650083541870117, "learning_rate": 5.697e-05, "loss": 1.0679, "step": 5698 }, { "epoch": 0.0005699, "grad_norm": 1.7204188108444214, "learning_rate": 5.698000000000001e-05, "loss": 1.0107, "step": 5699 }, { "epoch": 0.00057, "grad_norm": 2.277846336364746, "learning_rate": 5.6989999999999995e-05, "loss": 1.2178, "step": 5700 }, { "epoch": 0.0005701, "grad_norm": 2.155337333679199, "learning_rate": 5.7e-05, "loss": 1.3115, "step": 5701 }, { "epoch": 0.0005702, "grad_norm": 2.2528960704803467, "learning_rate": 5.7010000000000005e-05, "loss": 1.2407, "step": 5702 }, { "epoch": 0.0005703, "grad_norm": 1.8725662231445312, "learning_rate": 5.7020000000000006e-05, "loss": 1.1646, "step": 5703 }, { "epoch": 0.0005704, "grad_norm": 2.5592408180236816, "learning_rate": 5.703e-05, "loss": 1.3496, "step": 5704 }, { "epoch": 0.0005705, "grad_norm": 2.1637556552886963, "learning_rate": 5.704e-05, "loss": 1.249, "step": 5705 }, { "epoch": 0.0005706, "grad_norm": 2.220670461654663, "learning_rate": 5.7050000000000004e-05, "loss": 1.1992, "step": 5706 }, { "epoch": 0.0005707, "grad_norm": 1.9339892864227295, "learning_rate": 5.706e-05, "loss": 1.1338, "step": 5707 }, { "epoch": 0.0005708, "grad_norm": 2.1802027225494385, "learning_rate": 5.707e-05, "loss": 1.2393, "step": 5708 }, { "epoch": 0.0005709, "grad_norm": 1.7978614568710327, "learning_rate": 5.708e-05, "loss": 1.0391, "step": 5709 }, { "epoch": 0.000571, "grad_norm": 3.548150062561035, "learning_rate": 5.709e-05, "loss": 1.1035, "step": 5710 }, { "epoch": 0.0005711, "grad_norm": 2.054603099822998, "learning_rate": 5.71e-05, "loss": 1.0298, "step": 5711 }, { "epoch": 0.0005712, "grad_norm": 2.206544876098633, "learning_rate": 5.711e-05, "loss": 1.2344, "step": 5712 }, { "epoch": 0.0005713, "grad_norm": 4.416079521179199, "learning_rate": 5.712e-05, "loss": 1.4077, "step": 5713 }, { "epoch": 0.0005714, "grad_norm": 2.3344037532806396, "learning_rate": 5.713e-05, "loss": 1.0947, "step": 5714 }, { "epoch": 0.0005715, "grad_norm": 2.2653141021728516, "learning_rate": 5.714e-05, "loss": 1.0908, "step": 5715 }, { "epoch": 0.0005716, "grad_norm": 2.118607997894287, "learning_rate": 5.715000000000001e-05, "loss": 1.2393, "step": 5716 }, { "epoch": 0.0005717, "grad_norm": 2.184203863143921, "learning_rate": 5.7159999999999995e-05, "loss": 1.147, "step": 5717 }, { "epoch": 0.0005718, "grad_norm": 2.034911870956421, "learning_rate": 5.7169999999999996e-05, "loss": 1.1333, "step": 5718 }, { "epoch": 0.0005719, "grad_norm": 1.8440461158752441, "learning_rate": 5.7180000000000005e-05, "loss": 0.9668, "step": 5719 }, { "epoch": 0.000572, "grad_norm": 2.0628461837768555, "learning_rate": 5.7190000000000006e-05, "loss": 1.293, "step": 5720 }, { "epoch": 0.0005721, "grad_norm": 2.522508382797241, "learning_rate": 5.72e-05, "loss": 1.2866, "step": 5721 }, { "epoch": 0.0005722, "grad_norm": 2.229628324508667, "learning_rate": 5.721e-05, "loss": 1.1309, "step": 5722 }, { "epoch": 0.0005723, "grad_norm": 5.912186622619629, "learning_rate": 5.7220000000000004e-05, "loss": 1.8047, "step": 5723 }, { "epoch": 0.0005724, "grad_norm": 3.1335630416870117, "learning_rate": 5.723e-05, "loss": 1.1792, "step": 5724 }, { "epoch": 0.0005725, "grad_norm": 2.22836971282959, "learning_rate": 5.724e-05, "loss": 1.0972, "step": 5725 }, { "epoch": 0.0005726, "grad_norm": 2.203935146331787, "learning_rate": 5.725e-05, "loss": 1.186, "step": 5726 }, { "epoch": 0.0005727, "grad_norm": 1.68068528175354, "learning_rate": 5.726e-05, "loss": 1.0049, "step": 5727 }, { "epoch": 0.0005728, "grad_norm": 2.190138339996338, "learning_rate": 5.727e-05, "loss": 1.1567, "step": 5728 }, { "epoch": 0.0005729, "grad_norm": 1.9706114530563354, "learning_rate": 5.728e-05, "loss": 1.0264, "step": 5729 }, { "epoch": 0.000573, "grad_norm": 1.7482571601867676, "learning_rate": 5.729e-05, "loss": 1.0039, "step": 5730 }, { "epoch": 0.0005731, "grad_norm": 2.326173782348633, "learning_rate": 5.73e-05, "loss": 1.0903, "step": 5731 }, { "epoch": 0.0005732, "grad_norm": 1.7929534912109375, "learning_rate": 5.731e-05, "loss": 0.9854, "step": 5732 }, { "epoch": 0.0005733, "grad_norm": 1.764266848564148, "learning_rate": 5.732000000000001e-05, "loss": 0.9575, "step": 5733 }, { "epoch": 0.0005734, "grad_norm": 2.5840072631835938, "learning_rate": 5.7329999999999995e-05, "loss": 1.3408, "step": 5734 }, { "epoch": 0.0005735, "grad_norm": 2.9217374324798584, "learning_rate": 5.7339999999999996e-05, "loss": 1.3711, "step": 5735 }, { "epoch": 0.0005736, "grad_norm": 3.844555616378784, "learning_rate": 5.7350000000000005e-05, "loss": 2.0879, "step": 5736 }, { "epoch": 0.0005737, "grad_norm": 2.118861198425293, "learning_rate": 5.7360000000000006e-05, "loss": 1.0342, "step": 5737 }, { "epoch": 0.0005738, "grad_norm": 3.1677181720733643, "learning_rate": 5.737e-05, "loss": 1.3975, "step": 5738 }, { "epoch": 0.0005739, "grad_norm": 2.005305051803589, "learning_rate": 5.738e-05, "loss": 1.063, "step": 5739 }, { "epoch": 0.000574, "grad_norm": 2.0784804821014404, "learning_rate": 5.7390000000000004e-05, "loss": 1.0684, "step": 5740 }, { "epoch": 0.0005741, "grad_norm": 3.071539878845215, "learning_rate": 5.74e-05, "loss": 1.2676, "step": 5741 }, { "epoch": 0.0005742, "grad_norm": 2.1717145442962646, "learning_rate": 5.741e-05, "loss": 1.0571, "step": 5742 }, { "epoch": 0.0005743, "grad_norm": 2.0021402835845947, "learning_rate": 5.742e-05, "loss": 1.0229, "step": 5743 }, { "epoch": 0.0005744, "grad_norm": 1.7721225023269653, "learning_rate": 5.743e-05, "loss": 0.9575, "step": 5744 }, { "epoch": 0.0005745, "grad_norm": 2.241121530532837, "learning_rate": 5.744e-05, "loss": 1.2036, "step": 5745 }, { "epoch": 0.0005746, "grad_norm": 2.1638896465301514, "learning_rate": 5.745e-05, "loss": 1.064, "step": 5746 }, { "epoch": 0.0005747, "grad_norm": 2.39729642868042, "learning_rate": 5.746e-05, "loss": 1.2822, "step": 5747 }, { "epoch": 0.0005748, "grad_norm": 2.052424907684326, "learning_rate": 5.747e-05, "loss": 1.0215, "step": 5748 }, { "epoch": 0.0005749, "grad_norm": 1.981493592262268, "learning_rate": 5.748e-05, "loss": 1.0049, "step": 5749 }, { "epoch": 0.000575, "grad_norm": 1.9867534637451172, "learning_rate": 5.749000000000001e-05, "loss": 1.0015, "step": 5750 }, { "epoch": 0.0005751, "grad_norm": 3.0150108337402344, "learning_rate": 5.7499999999999995e-05, "loss": 1.2261, "step": 5751 }, { "epoch": 0.0005752, "grad_norm": 1.9989426136016846, "learning_rate": 5.7509999999999997e-05, "loss": 0.9468, "step": 5752 }, { "epoch": 0.0005753, "grad_norm": 1.9769320487976074, "learning_rate": 5.7520000000000005e-05, "loss": 0.9893, "step": 5753 }, { "epoch": 0.0005754, "grad_norm": 1.8832272291183472, "learning_rate": 5.7530000000000007e-05, "loss": 0.9966, "step": 5754 }, { "epoch": 0.0005755, "grad_norm": 1.9896249771118164, "learning_rate": 5.754e-05, "loss": 1.0957, "step": 5755 }, { "epoch": 0.0005756, "grad_norm": 2.0165281295776367, "learning_rate": 5.755e-05, "loss": 1.061, "step": 5756 }, { "epoch": 0.0005757, "grad_norm": 4.255472660064697, "learning_rate": 5.7560000000000005e-05, "loss": 1.168, "step": 5757 }, { "epoch": 0.0005758, "grad_norm": 1.8657300472259521, "learning_rate": 5.757e-05, "loss": 0.9268, "step": 5758 }, { "epoch": 0.0005759, "grad_norm": 1.934298038482666, "learning_rate": 5.758e-05, "loss": 1.0547, "step": 5759 }, { "epoch": 0.000576, "grad_norm": 1.8616300821304321, "learning_rate": 5.759e-05, "loss": 0.9517, "step": 5760 }, { "epoch": 0.0005761, "grad_norm": 1.7131308317184448, "learning_rate": 5.76e-05, "loss": 0.9521, "step": 5761 }, { "epoch": 0.0005762, "grad_norm": 1.90117609500885, "learning_rate": 5.761e-05, "loss": 1.0596, "step": 5762 }, { "epoch": 0.0005763, "grad_norm": 2.66219425201416, "learning_rate": 5.762e-05, "loss": 1.1758, "step": 5763 }, { "epoch": 0.0005764, "grad_norm": 2.0648040771484375, "learning_rate": 5.763e-05, "loss": 1.2681, "step": 5764 }, { "epoch": 0.0005765, "grad_norm": 2.287407636642456, "learning_rate": 5.764e-05, "loss": 1.0923, "step": 5765 }, { "epoch": 0.0005766, "grad_norm": 2.0532753467559814, "learning_rate": 5.765e-05, "loss": 1.0767, "step": 5766 }, { "epoch": 0.0005767, "grad_norm": 1.9127421379089355, "learning_rate": 5.766000000000001e-05, "loss": 0.9902, "step": 5767 }, { "epoch": 0.0005768, "grad_norm": 1.6706316471099854, "learning_rate": 5.7669999999999995e-05, "loss": 0.915, "step": 5768 }, { "epoch": 0.0005769, "grad_norm": 2.3771395683288574, "learning_rate": 5.768e-05, "loss": 1.0625, "step": 5769 }, { "epoch": 0.000577, "grad_norm": 1.9100340604782104, "learning_rate": 5.7690000000000005e-05, "loss": 0.9658, "step": 5770 }, { "epoch": 0.0005771, "grad_norm": 1.868527889251709, "learning_rate": 5.7700000000000007e-05, "loss": 1.0156, "step": 5771 }, { "epoch": 0.0005772, "grad_norm": 1.7585581541061401, "learning_rate": 5.771e-05, "loss": 0.9268, "step": 5772 }, { "epoch": 0.0005773, "grad_norm": 1.8959033489227295, "learning_rate": 5.772e-05, "loss": 0.9536, "step": 5773 }, { "epoch": 0.0005774, "grad_norm": 1.867139458656311, "learning_rate": 5.7730000000000005e-05, "loss": 1.0, "step": 5774 }, { "epoch": 0.0005775, "grad_norm": 1.6358270645141602, "learning_rate": 5.774e-05, "loss": 0.9214, "step": 5775 }, { "epoch": 0.0005776, "grad_norm": 2.010411024093628, "learning_rate": 5.775e-05, "loss": 0.9609, "step": 5776 }, { "epoch": 0.0005777, "grad_norm": 1.5704444646835327, "learning_rate": 5.776e-05, "loss": 0.8735, "step": 5777 }, { "epoch": 0.0005778, "grad_norm": 1.827335000038147, "learning_rate": 5.777e-05, "loss": 0.9409, "step": 5778 }, { "epoch": 0.0005779, "grad_norm": 1.7069824934005737, "learning_rate": 5.778e-05, "loss": 0.8984, "step": 5779 }, { "epoch": 0.000578, "grad_norm": 1.6686686277389526, "learning_rate": 5.779e-05, "loss": 0.8862, "step": 5780 }, { "epoch": 0.0005781, "grad_norm": 2.480184316635132, "learning_rate": 5.78e-05, "loss": 1.0791, "step": 5781 }, { "epoch": 0.0005782, "grad_norm": 1.9088640213012695, "learning_rate": 5.781e-05, "loss": 1.0312, "step": 5782 }, { "epoch": 0.0005783, "grad_norm": 1.7736129760742188, "learning_rate": 5.782e-05, "loss": 0.9204, "step": 5783 }, { "epoch": 0.0005784, "grad_norm": 1.7554395198822021, "learning_rate": 5.783000000000001e-05, "loss": 0.9634, "step": 5784 }, { "epoch": 0.0005785, "grad_norm": 2.0152337551116943, "learning_rate": 5.7839999999999995e-05, "loss": 0.874, "step": 5785 }, { "epoch": 0.0005786, "grad_norm": 1.6878819465637207, "learning_rate": 5.785e-05, "loss": 0.9248, "step": 5786 }, { "epoch": 0.0005787, "grad_norm": 1.9107030630111694, "learning_rate": 5.7860000000000005e-05, "loss": 0.9609, "step": 5787 }, { "epoch": 0.0005788, "grad_norm": 1.958961009979248, "learning_rate": 5.787000000000001e-05, "loss": 1.0776, "step": 5788 }, { "epoch": 0.0005789, "grad_norm": 1.6951417922973633, "learning_rate": 5.788e-05, "loss": 0.8779, "step": 5789 }, { "epoch": 0.000579, "grad_norm": 2.0156352519989014, "learning_rate": 5.789e-05, "loss": 1.0054, "step": 5790 }, { "epoch": 0.0005791, "grad_norm": 1.7576463222503662, "learning_rate": 5.7900000000000005e-05, "loss": 0.855, "step": 5791 }, { "epoch": 0.0005792, "grad_norm": 1.4828929901123047, "learning_rate": 5.791e-05, "loss": 0.8442, "step": 5792 }, { "epoch": 0.0005793, "grad_norm": 1.69581937789917, "learning_rate": 5.792e-05, "loss": 0.9282, "step": 5793 }, { "epoch": 0.0005794, "grad_norm": 1.922916054725647, "learning_rate": 5.793e-05, "loss": 0.9702, "step": 5794 }, { "epoch": 0.0005795, "grad_norm": 2.0185506343841553, "learning_rate": 5.794e-05, "loss": 1.0571, "step": 5795 }, { "epoch": 0.0005796, "grad_norm": 1.7874513864517212, "learning_rate": 5.795e-05, "loss": 0.9526, "step": 5796 }, { "epoch": 0.0005797, "grad_norm": 1.6726452112197876, "learning_rate": 5.796e-05, "loss": 0.8789, "step": 5797 }, { "epoch": 0.0005798, "grad_norm": 1.81720769405365, "learning_rate": 5.797e-05, "loss": 0.9414, "step": 5798 }, { "epoch": 0.0005799, "grad_norm": 1.6245219707489014, "learning_rate": 5.798e-05, "loss": 0.8921, "step": 5799 }, { "epoch": 0.00058, "grad_norm": 1.8347564935684204, "learning_rate": 5.799e-05, "loss": 0.8745, "step": 5800 }, { "epoch": 0.0005801, "grad_norm": 1.843480110168457, "learning_rate": 5.800000000000001e-05, "loss": 0.9004, "step": 5801 }, { "epoch": 0.0005802, "grad_norm": 1.781453013420105, "learning_rate": 5.8009999999999995e-05, "loss": 0.9199, "step": 5802 }, { "epoch": 0.0005803, "grad_norm": 1.734071135520935, "learning_rate": 5.802e-05, "loss": 0.9146, "step": 5803 }, { "epoch": 0.0005804, "grad_norm": 1.772889256477356, "learning_rate": 5.8030000000000005e-05, "loss": 0.8794, "step": 5804 }, { "epoch": 0.0005805, "grad_norm": 2.4896934032440186, "learning_rate": 5.804000000000001e-05, "loss": 0.9492, "step": 5805 }, { "epoch": 0.0005806, "grad_norm": 1.6063295602798462, "learning_rate": 5.805e-05, "loss": 0.9087, "step": 5806 }, { "epoch": 0.0005807, "grad_norm": 1.9447569847106934, "learning_rate": 5.806e-05, "loss": 0.9951, "step": 5807 }, { "epoch": 0.0005808, "grad_norm": 1.6404564380645752, "learning_rate": 5.8070000000000005e-05, "loss": 0.8506, "step": 5808 }, { "epoch": 0.0005809, "grad_norm": 1.6724512577056885, "learning_rate": 5.808e-05, "loss": 0.8447, "step": 5809 }, { "epoch": 0.000581, "grad_norm": 1.8658446073532104, "learning_rate": 5.809e-05, "loss": 0.8677, "step": 5810 }, { "epoch": 0.0005811, "grad_norm": 1.774506688117981, "learning_rate": 5.81e-05, "loss": 0.8779, "step": 5811 }, { "epoch": 0.0005812, "grad_norm": 1.6786613464355469, "learning_rate": 5.811e-05, "loss": 0.8633, "step": 5812 }, { "epoch": 0.0005813, "grad_norm": 1.9051158428192139, "learning_rate": 5.812e-05, "loss": 0.9424, "step": 5813 }, { "epoch": 0.0005814, "grad_norm": 1.724073052406311, "learning_rate": 5.813e-05, "loss": 0.9126, "step": 5814 }, { "epoch": 0.0005815, "grad_norm": 1.6745671033859253, "learning_rate": 5.814e-05, "loss": 0.8604, "step": 5815 }, { "epoch": 0.0005816, "grad_norm": 1.6149710416793823, "learning_rate": 5.815e-05, "loss": 0.8623, "step": 5816 }, { "epoch": 0.0005817, "grad_norm": 1.640714168548584, "learning_rate": 5.816e-05, "loss": 0.8633, "step": 5817 }, { "epoch": 0.0005818, "grad_norm": 1.8054769039154053, "learning_rate": 5.817000000000001e-05, "loss": 0.8481, "step": 5818 }, { "epoch": 0.0005819, "grad_norm": 1.6028062105178833, "learning_rate": 5.8179999999999995e-05, "loss": 0.8555, "step": 5819 }, { "epoch": 0.000582, "grad_norm": 1.8541170358657837, "learning_rate": 5.819e-05, "loss": 0.9297, "step": 5820 }, { "epoch": 0.0005821, "grad_norm": 1.7812132835388184, "learning_rate": 5.8200000000000005e-05, "loss": 0.916, "step": 5821 }, { "epoch": 0.0005822, "grad_norm": 1.6715450286865234, "learning_rate": 5.821000000000001e-05, "loss": 0.876, "step": 5822 }, { "epoch": 0.0005823, "grad_norm": 1.883423089981079, "learning_rate": 5.822e-05, "loss": 1.061, "step": 5823 }, { "epoch": 0.0005824, "grad_norm": 1.6956429481506348, "learning_rate": 5.823e-05, "loss": 0.8438, "step": 5824 }, { "epoch": 0.0005825, "grad_norm": 1.6848008632659912, "learning_rate": 5.8240000000000005e-05, "loss": 0.8887, "step": 5825 }, { "epoch": 0.0005826, "grad_norm": 1.761123776435852, "learning_rate": 5.825e-05, "loss": 0.9277, "step": 5826 }, { "epoch": 0.0005827, "grad_norm": 4.005990028381348, "learning_rate": 5.826e-05, "loss": 1.5615, "step": 5827 }, { "epoch": 0.0005828, "grad_norm": 2.3055953979492188, "learning_rate": 5.827e-05, "loss": 0.9722, "step": 5828 }, { "epoch": 0.0005829, "grad_norm": 3.2178542613983154, "learning_rate": 5.828e-05, "loss": 1.0449, "step": 5829 }, { "epoch": 0.000583, "grad_norm": 2.7492799758911133, "learning_rate": 5.829e-05, "loss": 0.9448, "step": 5830 }, { "epoch": 0.0005831, "grad_norm": 1.996016502380371, "learning_rate": 5.83e-05, "loss": 0.8779, "step": 5831 }, { "epoch": 0.0005832, "grad_norm": 1.6616058349609375, "learning_rate": 5.831e-05, "loss": 0.7661, "step": 5832 }, { "epoch": 0.0005833, "grad_norm": 1.789231300354004, "learning_rate": 5.832e-05, "loss": 0.835, "step": 5833 }, { "epoch": 0.0005834, "grad_norm": 1.9673794507980347, "learning_rate": 5.833e-05, "loss": 0.8794, "step": 5834 }, { "epoch": 0.0005835, "grad_norm": 1.7308722734451294, "learning_rate": 5.834000000000001e-05, "loss": 0.8584, "step": 5835 }, { "epoch": 0.0005836, "grad_norm": 1.6586679220199585, "learning_rate": 5.8349999999999995e-05, "loss": 0.8403, "step": 5836 }, { "epoch": 0.0005837, "grad_norm": 1.7169954776763916, "learning_rate": 5.836e-05, "loss": 0.8477, "step": 5837 }, { "epoch": 0.0005838, "grad_norm": 1.6661771535873413, "learning_rate": 5.8370000000000005e-05, "loss": 0.8574, "step": 5838 }, { "epoch": 0.0005839, "grad_norm": 1.6541742086410522, "learning_rate": 5.838000000000001e-05, "loss": 0.7949, "step": 5839 }, { "epoch": 0.000584, "grad_norm": 1.5927313566207886, "learning_rate": 5.839e-05, "loss": 0.7852, "step": 5840 }, { "epoch": 0.0005841, "grad_norm": 1.6425414085388184, "learning_rate": 5.84e-05, "loss": 0.8096, "step": 5841 }, { "epoch": 0.0005842, "grad_norm": 1.7989174127578735, "learning_rate": 5.8410000000000005e-05, "loss": 0.8882, "step": 5842 }, { "epoch": 0.0005843, "grad_norm": 1.8322278261184692, "learning_rate": 5.842e-05, "loss": 1.001, "step": 5843 }, { "epoch": 0.0005844, "grad_norm": 1.8731627464294434, "learning_rate": 5.843e-05, "loss": 0.8843, "step": 5844 }, { "epoch": 0.0005845, "grad_norm": 1.5437366962432861, "learning_rate": 5.844e-05, "loss": 0.7725, "step": 5845 }, { "epoch": 0.0005846, "grad_norm": 2.8317415714263916, "learning_rate": 5.845e-05, "loss": 1.0942, "step": 5846 }, { "epoch": 0.0005847, "grad_norm": 1.7654151916503906, "learning_rate": 5.846e-05, "loss": 0.8237, "step": 5847 }, { "epoch": 0.0005848, "grad_norm": 1.8651440143585205, "learning_rate": 5.847e-05, "loss": 0.812, "step": 5848 }, { "epoch": 0.0005849, "grad_norm": 1.6400736570358276, "learning_rate": 5.848e-05, "loss": 0.7842, "step": 5849 }, { "epoch": 0.000585, "grad_norm": 1.6788883209228516, "learning_rate": 5.849e-05, "loss": 0.8096, "step": 5850 }, { "epoch": 0.0005851, "grad_norm": 2.2326557636260986, "learning_rate": 5.85e-05, "loss": 0.96, "step": 5851 }, { "epoch": 0.0005852, "grad_norm": 1.8667100667953491, "learning_rate": 5.851000000000001e-05, "loss": 0.8208, "step": 5852 }, { "epoch": 0.0005853, "grad_norm": 1.6106916666030884, "learning_rate": 5.8519999999999995e-05, "loss": 0.7939, "step": 5853 }, { "epoch": 0.0005854, "grad_norm": 1.6850008964538574, "learning_rate": 5.853e-05, "loss": 0.8037, "step": 5854 }, { "epoch": 0.0005855, "grad_norm": 1.6820380687713623, "learning_rate": 5.8540000000000005e-05, "loss": 0.7905, "step": 5855 }, { "epoch": 0.0005856, "grad_norm": 1.6166424751281738, "learning_rate": 5.855000000000001e-05, "loss": 0.7661, "step": 5856 }, { "epoch": 0.0005857, "grad_norm": 1.606998085975647, "learning_rate": 5.856e-05, "loss": 0.8359, "step": 5857 }, { "epoch": 0.0005858, "grad_norm": 1.6939175128936768, "learning_rate": 5.857e-05, "loss": 0.8188, "step": 5858 }, { "epoch": 0.0005859, "grad_norm": 1.7543524503707886, "learning_rate": 5.8580000000000005e-05, "loss": 0.8491, "step": 5859 }, { "epoch": 0.000586, "grad_norm": 1.7315897941589355, "learning_rate": 5.859e-05, "loss": 0.8999, "step": 5860 }, { "epoch": 0.0005861, "grad_norm": 2.3954813480377197, "learning_rate": 5.86e-05, "loss": 1.1616, "step": 5861 }, { "epoch": 0.0005862, "grad_norm": 1.6844712495803833, "learning_rate": 5.861e-05, "loss": 0.772, "step": 5862 }, { "epoch": 0.0005863, "grad_norm": 1.5823957920074463, "learning_rate": 5.862e-05, "loss": 0.8071, "step": 5863 }, { "epoch": 0.0005864, "grad_norm": 1.5627663135528564, "learning_rate": 5.863e-05, "loss": 0.812, "step": 5864 }, { "epoch": 0.0005865, "grad_norm": 1.6954442262649536, "learning_rate": 5.864e-05, "loss": 0.7988, "step": 5865 }, { "epoch": 0.0005866, "grad_norm": 1.7571743726730347, "learning_rate": 5.865e-05, "loss": 0.7607, "step": 5866 }, { "epoch": 0.0005867, "grad_norm": 1.6370595693588257, "learning_rate": 5.866e-05, "loss": 0.7529, "step": 5867 }, { "epoch": 0.0005868, "grad_norm": 1.5386178493499756, "learning_rate": 5.867e-05, "loss": 0.8086, "step": 5868 }, { "epoch": 0.0005869, "grad_norm": 1.5495586395263672, "learning_rate": 5.868000000000001e-05, "loss": 0.7627, "step": 5869 }, { "epoch": 0.000587, "grad_norm": 1.72905695438385, "learning_rate": 5.8689999999999995e-05, "loss": 0.8296, "step": 5870 }, { "epoch": 0.0005871, "grad_norm": 1.6074451208114624, "learning_rate": 5.87e-05, "loss": 0.876, "step": 5871 }, { "epoch": 0.0005872, "grad_norm": 1.5799434185028076, "learning_rate": 5.8710000000000005e-05, "loss": 0.8086, "step": 5872 }, { "epoch": 0.0005873, "grad_norm": 1.7185566425323486, "learning_rate": 5.872000000000001e-05, "loss": 0.873, "step": 5873 }, { "epoch": 0.0005874, "grad_norm": 1.6175681352615356, "learning_rate": 5.873e-05, "loss": 0.7603, "step": 5874 }, { "epoch": 0.0005875, "grad_norm": 1.3885085582733154, "learning_rate": 5.8740000000000003e-05, "loss": 0.8052, "step": 5875 }, { "epoch": 0.0005876, "grad_norm": 1.5084010362625122, "learning_rate": 5.8750000000000005e-05, "loss": 0.8022, "step": 5876 }, { "epoch": 0.0005877, "grad_norm": 1.5180604457855225, "learning_rate": 5.876e-05, "loss": 0.7681, "step": 5877 }, { "epoch": 0.0005878, "grad_norm": 1.5178165435791016, "learning_rate": 5.877e-05, "loss": 0.7559, "step": 5878 }, { "epoch": 0.0005879, "grad_norm": 1.4072206020355225, "learning_rate": 5.878e-05, "loss": 0.7432, "step": 5879 }, { "epoch": 0.000588, "grad_norm": 1.6438937187194824, "learning_rate": 5.879e-05, "loss": 0.8389, "step": 5880 }, { "epoch": 0.0005881, "grad_norm": 1.6769683361053467, "learning_rate": 5.88e-05, "loss": 0.8018, "step": 5881 }, { "epoch": 0.0005882, "grad_norm": 1.623231291770935, "learning_rate": 5.881e-05, "loss": 0.7739, "step": 5882 }, { "epoch": 0.0005883, "grad_norm": 1.6477833986282349, "learning_rate": 5.882e-05, "loss": 0.7993, "step": 5883 }, { "epoch": 0.0005884, "grad_norm": 1.6617273092269897, "learning_rate": 5.883e-05, "loss": 0.8618, "step": 5884 }, { "epoch": 0.0005885, "grad_norm": 2.173892021179199, "learning_rate": 5.884e-05, "loss": 1.0137, "step": 5885 }, { "epoch": 0.0005886, "grad_norm": 1.5440952777862549, "learning_rate": 5.885000000000001e-05, "loss": 0.7686, "step": 5886 }, { "epoch": 0.0005887, "grad_norm": 1.2644119262695312, "learning_rate": 5.8859999999999995e-05, "loss": 0.7139, "step": 5887 }, { "epoch": 0.0005888, "grad_norm": 1.484767198562622, "learning_rate": 5.887e-05, "loss": 0.7417, "step": 5888 }, { "epoch": 0.0005889, "grad_norm": 1.5068488121032715, "learning_rate": 5.8880000000000005e-05, "loss": 0.7974, "step": 5889 }, { "epoch": 0.000589, "grad_norm": 3.0165534019470215, "learning_rate": 5.889000000000001e-05, "loss": 0.812, "step": 5890 }, { "epoch": 0.0005891, "grad_norm": 1.8161137104034424, "learning_rate": 5.89e-05, "loss": 0.8296, "step": 5891 }, { "epoch": 0.0005892, "grad_norm": 1.7120927572250366, "learning_rate": 5.8910000000000003e-05, "loss": 0.7227, "step": 5892 }, { "epoch": 0.0005893, "grad_norm": 1.631578803062439, "learning_rate": 5.8920000000000005e-05, "loss": 0.7925, "step": 5893 }, { "epoch": 0.0005894, "grad_norm": 1.703325629234314, "learning_rate": 5.893e-05, "loss": 0.8032, "step": 5894 }, { "epoch": 0.0005895, "grad_norm": 1.7086197137832642, "learning_rate": 5.894e-05, "loss": 0.8003, "step": 5895 }, { "epoch": 0.0005896, "grad_norm": 5.005209922790527, "learning_rate": 5.895e-05, "loss": 1.2173, "step": 5896 }, { "epoch": 0.0005897, "grad_norm": 1.6591870784759521, "learning_rate": 5.896e-05, "loss": 0.7886, "step": 5897 }, { "epoch": 0.0005898, "grad_norm": 1.8780789375305176, "learning_rate": 5.897e-05, "loss": 0.832, "step": 5898 }, { "epoch": 0.0005899, "grad_norm": 1.866897702217102, "learning_rate": 5.898e-05, "loss": 0.8823, "step": 5899 }, { "epoch": 0.00059, "grad_norm": 1.6844897270202637, "learning_rate": 5.899e-05, "loss": 0.856, "step": 5900 }, { "epoch": 0.0005901, "grad_norm": 1.6563727855682373, "learning_rate": 5.9e-05, "loss": 0.8486, "step": 5901 }, { "epoch": 0.0005902, "grad_norm": 1.4859638214111328, "learning_rate": 5.901e-05, "loss": 0.7632, "step": 5902 }, { "epoch": 0.0005903, "grad_norm": 1.5815120935440063, "learning_rate": 5.902000000000001e-05, "loss": 0.7725, "step": 5903 }, { "epoch": 0.0005904, "grad_norm": 2.134547710418701, "learning_rate": 5.9029999999999996e-05, "loss": 0.9082, "step": 5904 }, { "epoch": 0.0005905, "grad_norm": 1.654038906097412, "learning_rate": 5.904e-05, "loss": 0.7949, "step": 5905 }, { "epoch": 0.0005906, "grad_norm": 1.5547631978988647, "learning_rate": 5.9050000000000006e-05, "loss": 0.752, "step": 5906 }, { "epoch": 0.0005907, "grad_norm": 1.5662357807159424, "learning_rate": 5.906000000000001e-05, "loss": 0.8008, "step": 5907 }, { "epoch": 0.0005908, "grad_norm": 1.57384192943573, "learning_rate": 5.907e-05, "loss": 0.7471, "step": 5908 }, { "epoch": 0.0005909, "grad_norm": 1.4176363945007324, "learning_rate": 5.9080000000000004e-05, "loss": 0.7207, "step": 5909 }, { "epoch": 0.000591, "grad_norm": 2.020887851715088, "learning_rate": 5.9090000000000005e-05, "loss": 1.1572, "step": 5910 }, { "epoch": 0.0005911, "grad_norm": 1.6313410997390747, "learning_rate": 5.91e-05, "loss": 0.7993, "step": 5911 }, { "epoch": 0.0005912, "grad_norm": 1.5192641019821167, "learning_rate": 5.911e-05, "loss": 0.7412, "step": 5912 }, { "epoch": 0.0005913, "grad_norm": 1.6517242193222046, "learning_rate": 5.912e-05, "loss": 0.793, "step": 5913 }, { "epoch": 0.0005914, "grad_norm": 1.642682671546936, "learning_rate": 5.913e-05, "loss": 0.7759, "step": 5914 }, { "epoch": 0.0005915, "grad_norm": 1.656015157699585, "learning_rate": 5.914e-05, "loss": 0.7896, "step": 5915 }, { "epoch": 0.0005916, "grad_norm": 1.924726963043213, "learning_rate": 5.915e-05, "loss": 0.8882, "step": 5916 }, { "epoch": 0.0005917, "grad_norm": 1.5693986415863037, "learning_rate": 5.916e-05, "loss": 0.7817, "step": 5917 }, { "epoch": 0.0005918, "grad_norm": 1.9120745658874512, "learning_rate": 5.917e-05, "loss": 0.8506, "step": 5918 }, { "epoch": 0.0005919, "grad_norm": 1.9217987060546875, "learning_rate": 5.918e-05, "loss": 0.8892, "step": 5919 }, { "epoch": 0.000592, "grad_norm": 2.201404094696045, "learning_rate": 5.919000000000001e-05, "loss": 0.9932, "step": 5920 }, { "epoch": 0.0005921, "grad_norm": 1.560445785522461, "learning_rate": 5.9199999999999996e-05, "loss": 0.7314, "step": 5921 }, { "epoch": 0.0005922, "grad_norm": 1.4320646524429321, "learning_rate": 5.921e-05, "loss": 0.7236, "step": 5922 }, { "epoch": 0.0005923, "grad_norm": 1.3355786800384521, "learning_rate": 5.9220000000000006e-05, "loss": 0.6875, "step": 5923 }, { "epoch": 0.0005924, "grad_norm": 1.5497984886169434, "learning_rate": 5.923000000000001e-05, "loss": 0.7876, "step": 5924 }, { "epoch": 0.0005925, "grad_norm": 1.6212843656539917, "learning_rate": 5.924e-05, "loss": 0.7842, "step": 5925 }, { "epoch": 0.0005926, "grad_norm": 1.6754783391952515, "learning_rate": 5.9250000000000004e-05, "loss": 0.811, "step": 5926 }, { "epoch": 0.0005927, "grad_norm": 1.5971804857254028, "learning_rate": 5.9260000000000005e-05, "loss": 0.7827, "step": 5927 }, { "epoch": 0.0005928, "grad_norm": 1.6103317737579346, "learning_rate": 5.927e-05, "loss": 0.8335, "step": 5928 }, { "epoch": 0.0005929, "grad_norm": 1.4968230724334717, "learning_rate": 5.928e-05, "loss": 0.7837, "step": 5929 }, { "epoch": 0.000593, "grad_norm": 2.0906319618225098, "learning_rate": 5.929e-05, "loss": 0.9165, "step": 5930 }, { "epoch": 0.0005931, "grad_norm": 1.530055046081543, "learning_rate": 5.93e-05, "loss": 0.7358, "step": 5931 }, { "epoch": 0.0005932, "grad_norm": 1.9216514825820923, "learning_rate": 5.931e-05, "loss": 0.9087, "step": 5932 }, { "epoch": 0.0005933, "grad_norm": 1.5280019044876099, "learning_rate": 5.932e-05, "loss": 0.7646, "step": 5933 }, { "epoch": 0.0005934, "grad_norm": 1.504082202911377, "learning_rate": 5.933e-05, "loss": 0.7935, "step": 5934 }, { "epoch": 0.0005935, "grad_norm": 1.5964938402175903, "learning_rate": 5.934e-05, "loss": 0.8315, "step": 5935 }, { "epoch": 0.0005936, "grad_norm": 1.5614784955978394, "learning_rate": 5.935e-05, "loss": 0.8149, "step": 5936 }, { "epoch": 0.0005937, "grad_norm": 1.6300089359283447, "learning_rate": 5.936000000000001e-05, "loss": 0.7729, "step": 5937 }, { "epoch": 0.0005938, "grad_norm": 2.1969408988952637, "learning_rate": 5.9369999999999996e-05, "loss": 0.9463, "step": 5938 }, { "epoch": 0.0005939, "grad_norm": 1.637465476989746, "learning_rate": 5.938e-05, "loss": 0.7358, "step": 5939 }, { "epoch": 0.000594, "grad_norm": 2.084805488586426, "learning_rate": 5.9390000000000006e-05, "loss": 0.8447, "step": 5940 }, { "epoch": 0.0005941, "grad_norm": 1.8123067617416382, "learning_rate": 5.940000000000001e-05, "loss": 0.7827, "step": 5941 }, { "epoch": 0.0005942, "grad_norm": 2.333995819091797, "learning_rate": 5.941e-05, "loss": 0.8882, "step": 5942 }, { "epoch": 0.0005943, "grad_norm": 1.5387401580810547, "learning_rate": 5.9420000000000004e-05, "loss": 0.7437, "step": 5943 }, { "epoch": 0.0005944, "grad_norm": 1.5562912225723267, "learning_rate": 5.9430000000000005e-05, "loss": 0.7188, "step": 5944 }, { "epoch": 0.0005945, "grad_norm": 1.7533470392227173, "learning_rate": 5.944e-05, "loss": 0.8643, "step": 5945 }, { "epoch": 0.0005946, "grad_norm": 1.5385069847106934, "learning_rate": 5.945e-05, "loss": 0.7598, "step": 5946 }, { "epoch": 0.0005947, "grad_norm": 1.808990716934204, "learning_rate": 5.946e-05, "loss": 0.8101, "step": 5947 }, { "epoch": 0.0005948, "grad_norm": 1.5682218074798584, "learning_rate": 5.947e-05, "loss": 0.8096, "step": 5948 }, { "epoch": 0.0005949, "grad_norm": 1.9593143463134766, "learning_rate": 5.948e-05, "loss": 0.873, "step": 5949 }, { "epoch": 0.000595, "grad_norm": 2.1629326343536377, "learning_rate": 5.949e-05, "loss": 0.8833, "step": 5950 }, { "epoch": 0.0005951, "grad_norm": 2.159661293029785, "learning_rate": 5.95e-05, "loss": 0.8745, "step": 5951 }, { "epoch": 0.0005952, "grad_norm": 1.592955231666565, "learning_rate": 5.951e-05, "loss": 0.7007, "step": 5952 }, { "epoch": 0.0005953, "grad_norm": 1.721563696861267, "learning_rate": 5.952e-05, "loss": 0.793, "step": 5953 }, { "epoch": 0.0005954, "grad_norm": 1.742895245552063, "learning_rate": 5.953e-05, "loss": 0.7817, "step": 5954 }, { "epoch": 0.0005955, "grad_norm": 1.558279037475586, "learning_rate": 5.9539999999999996e-05, "loss": 0.8027, "step": 5955 }, { "epoch": 0.0005956, "grad_norm": 1.4903243780136108, "learning_rate": 5.955e-05, "loss": 0.7188, "step": 5956 }, { "epoch": 0.0005957, "grad_norm": 1.782842755317688, "learning_rate": 5.9560000000000006e-05, "loss": 0.7915, "step": 5957 }, { "epoch": 0.0005958, "grad_norm": 1.6400823593139648, "learning_rate": 5.957000000000001e-05, "loss": 0.7417, "step": 5958 }, { "epoch": 0.0005959, "grad_norm": 1.4021106958389282, "learning_rate": 5.958e-05, "loss": 0.6851, "step": 5959 }, { "epoch": 0.000596, "grad_norm": 1.7664570808410645, "learning_rate": 5.9590000000000004e-05, "loss": 0.8364, "step": 5960 }, { "epoch": 0.0005961, "grad_norm": 1.5353564023971558, "learning_rate": 5.9600000000000005e-05, "loss": 0.7529, "step": 5961 }, { "epoch": 0.0005962, "grad_norm": 1.3988913297653198, "learning_rate": 5.961e-05, "loss": 0.6948, "step": 5962 }, { "epoch": 0.0005963, "grad_norm": 3.4333651065826416, "learning_rate": 5.962e-05, "loss": 0.8628, "step": 5963 }, { "epoch": 0.0005964, "grad_norm": 1.5233893394470215, "learning_rate": 5.963e-05, "loss": 0.7407, "step": 5964 }, { "epoch": 0.0005965, "grad_norm": 1.7804893255233765, "learning_rate": 5.964e-05, "loss": 0.771, "step": 5965 }, { "epoch": 0.0005966, "grad_norm": 1.4234447479248047, "learning_rate": 5.965e-05, "loss": 0.6919, "step": 5966 }, { "epoch": 0.0005967, "grad_norm": 1.5506657361984253, "learning_rate": 5.966e-05, "loss": 0.7441, "step": 5967 }, { "epoch": 0.0005968, "grad_norm": 1.599525809288025, "learning_rate": 5.967e-05, "loss": 0.7598, "step": 5968 }, { "epoch": 0.0005969, "grad_norm": 2.79545259475708, "learning_rate": 5.968e-05, "loss": 0.8911, "step": 5969 }, { "epoch": 0.000597, "grad_norm": 1.5482027530670166, "learning_rate": 5.969e-05, "loss": 0.7261, "step": 5970 }, { "epoch": 0.0005971, "grad_norm": 1.4654983282089233, "learning_rate": 5.97e-05, "loss": 0.6968, "step": 5971 }, { "epoch": 0.0005972, "grad_norm": 1.8669058084487915, "learning_rate": 5.9709999999999996e-05, "loss": 0.7749, "step": 5972 }, { "epoch": 0.0005973, "grad_norm": 1.5295121669769287, "learning_rate": 5.972e-05, "loss": 0.7104, "step": 5973 }, { "epoch": 0.0005974, "grad_norm": 1.6404061317443848, "learning_rate": 5.9730000000000006e-05, "loss": 0.7319, "step": 5974 }, { "epoch": 0.0005975, "grad_norm": 1.3939292430877686, "learning_rate": 5.974000000000001e-05, "loss": 0.6626, "step": 5975 }, { "epoch": 0.0005976, "grad_norm": 1.4034295082092285, "learning_rate": 5.9749999999999995e-05, "loss": 0.6572, "step": 5976 }, { "epoch": 0.0005977, "grad_norm": 1.4362789392471313, "learning_rate": 5.9760000000000004e-05, "loss": 0.6987, "step": 5977 }, { "epoch": 0.0005978, "grad_norm": 1.4524978399276733, "learning_rate": 5.9770000000000005e-05, "loss": 0.71, "step": 5978 }, { "epoch": 0.0005979, "grad_norm": 2.0200347900390625, "learning_rate": 5.978e-05, "loss": 0.855, "step": 5979 }, { "epoch": 0.000598, "grad_norm": 1.6224732398986816, "learning_rate": 5.979e-05, "loss": 0.7871, "step": 5980 }, { "epoch": 0.0005981, "grad_norm": 1.3911391496658325, "learning_rate": 5.9800000000000003e-05, "loss": 0.6836, "step": 5981 }, { "epoch": 0.0005982, "grad_norm": 1.5443239212036133, "learning_rate": 5.981e-05, "loss": 0.7676, "step": 5982 }, { "epoch": 0.0005983, "grad_norm": 1.5114848613739014, "learning_rate": 5.982e-05, "loss": 0.73, "step": 5983 }, { "epoch": 0.0005984, "grad_norm": 1.7796193361282349, "learning_rate": 5.983e-05, "loss": 0.9609, "step": 5984 }, { "epoch": 0.0005985, "grad_norm": 1.8018146753311157, "learning_rate": 5.984e-05, "loss": 0.8877, "step": 5985 }, { "epoch": 0.0005986, "grad_norm": 1.4452275037765503, "learning_rate": 5.985e-05, "loss": 0.6768, "step": 5986 }, { "epoch": 0.0005987, "grad_norm": 1.846722960472107, "learning_rate": 5.986e-05, "loss": 0.8223, "step": 5987 }, { "epoch": 0.0005988, "grad_norm": 2.8641154766082764, "learning_rate": 5.987e-05, "loss": 0.9253, "step": 5988 }, { "epoch": 0.0005989, "grad_norm": 1.8777289390563965, "learning_rate": 5.9879999999999996e-05, "loss": 0.7803, "step": 5989 }, { "epoch": 0.000599, "grad_norm": 1.9890245199203491, "learning_rate": 5.989e-05, "loss": 0.8408, "step": 5990 }, { "epoch": 0.0005991, "grad_norm": 1.6677210330963135, "learning_rate": 5.9900000000000006e-05, "loss": 0.7793, "step": 5991 }, { "epoch": 0.0005992, "grad_norm": 1.460680365562439, "learning_rate": 5.991000000000001e-05, "loss": 0.7139, "step": 5992 }, { "epoch": 0.0005993, "grad_norm": 1.5622482299804688, "learning_rate": 5.9919999999999996e-05, "loss": 0.73, "step": 5993 }, { "epoch": 0.0005994, "grad_norm": 1.4825682640075684, "learning_rate": 5.9930000000000004e-05, "loss": 0.6875, "step": 5994 }, { "epoch": 0.0005995, "grad_norm": 1.6099538803100586, "learning_rate": 5.9940000000000005e-05, "loss": 0.792, "step": 5995 }, { "epoch": 0.0005996, "grad_norm": 1.5532127618789673, "learning_rate": 5.995e-05, "loss": 0.731, "step": 5996 }, { "epoch": 0.0005997, "grad_norm": 1.4975957870483398, "learning_rate": 5.996e-05, "loss": 0.7085, "step": 5997 }, { "epoch": 0.0005998, "grad_norm": 1.4132734537124634, "learning_rate": 5.9970000000000004e-05, "loss": 0.6748, "step": 5998 }, { "epoch": 0.0005999, "grad_norm": 1.314280390739441, "learning_rate": 5.998e-05, "loss": 0.6504, "step": 5999 }, { "epoch": 0.0006, "grad_norm": 1.5365554094314575, "learning_rate": 5.999e-05, "loss": 0.6851, "step": 6000 }, { "epoch": 0.0006, "eval_loss": 0.08720719069242477, "eval_runtime": 362.1442, "eval_samples_per_second": 27.613, "eval_steps_per_second": 1.726, "step": 6000 }, { "epoch": 0.0006001, "grad_norm": 1.5553802251815796, "learning_rate": 6e-05, "loss": 0.7573, "step": 6001 }, { "epoch": 0.0006002, "grad_norm": 1.6512336730957031, "learning_rate": 6.001e-05, "loss": 0.7559, "step": 6002 }, { "epoch": 0.0006003, "grad_norm": 1.5313918590545654, "learning_rate": 6.002e-05, "loss": 0.7388, "step": 6003 }, { "epoch": 0.0006004, "grad_norm": 1.7878886461257935, "learning_rate": 6.003e-05, "loss": 0.8364, "step": 6004 }, { "epoch": 0.0006005, "grad_norm": 1.4012184143066406, "learning_rate": 6.004e-05, "loss": 0.6924, "step": 6005 }, { "epoch": 0.0006006, "grad_norm": 1.6499422788619995, "learning_rate": 6.0049999999999996e-05, "loss": 0.7358, "step": 6006 }, { "epoch": 0.0006007, "grad_norm": 1.5204499959945679, "learning_rate": 6.006e-05, "loss": 0.6792, "step": 6007 }, { "epoch": 0.0006008, "grad_norm": 2.066147565841675, "learning_rate": 6.0070000000000006e-05, "loss": 0.8433, "step": 6008 }, { "epoch": 0.0006009, "grad_norm": 1.7936888933181763, "learning_rate": 6.008000000000001e-05, "loss": 1.0107, "step": 6009 }, { "epoch": 0.000601, "grad_norm": 1.4196463823318481, "learning_rate": 6.0089999999999996e-05, "loss": 0.7012, "step": 6010 }, { "epoch": 0.0006011, "grad_norm": 2.7417123317718506, "learning_rate": 6.0100000000000004e-05, "loss": 0.8823, "step": 6011 }, { "epoch": 0.0006012, "grad_norm": 2.072298288345337, "learning_rate": 6.0110000000000006e-05, "loss": 0.7476, "step": 6012 }, { "epoch": 0.0006013, "grad_norm": 1.7735410928726196, "learning_rate": 6.012e-05, "loss": 0.7466, "step": 6013 }, { "epoch": 0.0006014, "grad_norm": 1.6693915128707886, "learning_rate": 6.013e-05, "loss": 0.7783, "step": 6014 }, { "epoch": 0.0006015, "grad_norm": 1.4681349992752075, "learning_rate": 6.0140000000000004e-05, "loss": 0.7012, "step": 6015 }, { "epoch": 0.0006016, "grad_norm": 2.034409761428833, "learning_rate": 6.015e-05, "loss": 0.6836, "step": 6016 }, { "epoch": 0.0006017, "grad_norm": 1.5247353315353394, "learning_rate": 6.016e-05, "loss": 0.7061, "step": 6017 }, { "epoch": 0.0006018, "grad_norm": 2.1397507190704346, "learning_rate": 6.017e-05, "loss": 0.7803, "step": 6018 }, { "epoch": 0.0006019, "grad_norm": 1.5201102495193481, "learning_rate": 6.018e-05, "loss": 0.6997, "step": 6019 }, { "epoch": 0.000602, "grad_norm": 1.7340844869613647, "learning_rate": 6.019e-05, "loss": 0.8164, "step": 6020 }, { "epoch": 0.0006021, "grad_norm": 1.4697178602218628, "learning_rate": 6.02e-05, "loss": 0.6431, "step": 6021 }, { "epoch": 0.0006022, "grad_norm": 1.5053555965423584, "learning_rate": 6.021e-05, "loss": 0.6851, "step": 6022 }, { "epoch": 0.0006023, "grad_norm": 1.394316554069519, "learning_rate": 6.0219999999999996e-05, "loss": 0.6714, "step": 6023 }, { "epoch": 0.0006024, "grad_norm": 1.365586519241333, "learning_rate": 6.023e-05, "loss": 0.6343, "step": 6024 }, { "epoch": 0.0006025, "grad_norm": 1.5944674015045166, "learning_rate": 6.0240000000000006e-05, "loss": 0.7695, "step": 6025 }, { "epoch": 0.0006026, "grad_norm": 1.3342678546905518, "learning_rate": 6.025000000000001e-05, "loss": 0.6338, "step": 6026 }, { "epoch": 0.0006027, "grad_norm": 1.4751640558242798, "learning_rate": 6.0259999999999996e-05, "loss": 0.6633, "step": 6027 }, { "epoch": 0.0006028, "grad_norm": 1.5709500312805176, "learning_rate": 6.0270000000000004e-05, "loss": 0.7354, "step": 6028 }, { "epoch": 0.0006029, "grad_norm": 1.364405870437622, "learning_rate": 6.0280000000000006e-05, "loss": 0.6846, "step": 6029 }, { "epoch": 0.000603, "grad_norm": 1.280802845954895, "learning_rate": 6.029e-05, "loss": 0.645, "step": 6030 }, { "epoch": 0.0006031, "grad_norm": 1.8422123193740845, "learning_rate": 6.03e-05, "loss": 0.7554, "step": 6031 }, { "epoch": 0.0006032, "grad_norm": 1.5924108028411865, "learning_rate": 6.0310000000000004e-05, "loss": 0.6846, "step": 6032 }, { "epoch": 0.0006033, "grad_norm": 1.7534675598144531, "learning_rate": 6.032e-05, "loss": 0.8188, "step": 6033 }, { "epoch": 0.0006034, "grad_norm": 1.363912582397461, "learning_rate": 6.033e-05, "loss": 0.6841, "step": 6034 }, { "epoch": 0.0006035, "grad_norm": 1.3158562183380127, "learning_rate": 6.034e-05, "loss": 0.687, "step": 6035 }, { "epoch": 0.0006036, "grad_norm": 1.873213529586792, "learning_rate": 6.035e-05, "loss": 0.7739, "step": 6036 }, { "epoch": 0.0006037, "grad_norm": 1.4768091440200806, "learning_rate": 6.036e-05, "loss": 0.6709, "step": 6037 }, { "epoch": 0.0006038, "grad_norm": 1.7859482765197754, "learning_rate": 6.037e-05, "loss": 0.8154, "step": 6038 }, { "epoch": 0.0006039, "grad_norm": 1.961234211921692, "learning_rate": 6.038e-05, "loss": 0.6943, "step": 6039 }, { "epoch": 0.000604, "grad_norm": 1.5077617168426514, "learning_rate": 6.0389999999999996e-05, "loss": 0.6821, "step": 6040 }, { "epoch": 0.0006041, "grad_norm": 1.4009017944335938, "learning_rate": 6.04e-05, "loss": 0.6602, "step": 6041 }, { "epoch": 0.0006042, "grad_norm": 1.473061203956604, "learning_rate": 6.0410000000000006e-05, "loss": 0.75, "step": 6042 }, { "epoch": 0.0006043, "grad_norm": 1.442574143409729, "learning_rate": 6.042000000000001e-05, "loss": 0.6836, "step": 6043 }, { "epoch": 0.0006044, "grad_norm": 1.5931884050369263, "learning_rate": 6.0429999999999996e-05, "loss": 0.688, "step": 6044 }, { "epoch": 0.0006045, "grad_norm": 1.3145321607589722, "learning_rate": 6.0440000000000004e-05, "loss": 0.626, "step": 6045 }, { "epoch": 0.0006046, "grad_norm": 1.446225643157959, "learning_rate": 6.0450000000000006e-05, "loss": 0.6655, "step": 6046 }, { "epoch": 0.0006047, "grad_norm": 1.5582249164581299, "learning_rate": 6.046e-05, "loss": 0.7358, "step": 6047 }, { "epoch": 0.0006048, "grad_norm": 1.45441734790802, "learning_rate": 6.047e-05, "loss": 0.7012, "step": 6048 }, { "epoch": 0.0006049, "grad_norm": 2.6049602031707764, "learning_rate": 6.0480000000000004e-05, "loss": 0.8921, "step": 6049 }, { "epoch": 0.000605, "grad_norm": 1.6523659229278564, "learning_rate": 6.049e-05, "loss": 0.7407, "step": 6050 }, { "epoch": 0.0006051, "grad_norm": 1.8112012147903442, "learning_rate": 6.05e-05, "loss": 0.8452, "step": 6051 }, { "epoch": 0.0006052, "grad_norm": 1.4687682390213013, "learning_rate": 6.051e-05, "loss": 0.6528, "step": 6052 }, { "epoch": 0.0006053, "grad_norm": 1.4037615060806274, "learning_rate": 6.052e-05, "loss": 0.6953, "step": 6053 }, { "epoch": 0.0006054, "grad_norm": 1.3966478109359741, "learning_rate": 6.053e-05, "loss": 0.6772, "step": 6054 }, { "epoch": 0.0006055, "grad_norm": 2.193171501159668, "learning_rate": 6.054e-05, "loss": 0.9385, "step": 6055 }, { "epoch": 0.0006056, "grad_norm": 1.522186517715454, "learning_rate": 6.055e-05, "loss": 0.7227, "step": 6056 }, { "epoch": 0.0006057, "grad_norm": 1.3285267353057861, "learning_rate": 6.0559999999999996e-05, "loss": 0.6499, "step": 6057 }, { "epoch": 0.0006058, "grad_norm": 1.491991400718689, "learning_rate": 6.057e-05, "loss": 0.7163, "step": 6058 }, { "epoch": 0.0006059, "grad_norm": 1.3360713720321655, "learning_rate": 6.0580000000000006e-05, "loss": 0.688, "step": 6059 }, { "epoch": 0.000606, "grad_norm": 1.608153223991394, "learning_rate": 6.059000000000001e-05, "loss": 0.7534, "step": 6060 }, { "epoch": 0.0006061, "grad_norm": 1.4696681499481201, "learning_rate": 6.0599999999999996e-05, "loss": 0.6851, "step": 6061 }, { "epoch": 0.0006062, "grad_norm": 1.2329943180084229, "learning_rate": 6.0610000000000004e-05, "loss": 0.6392, "step": 6062 }, { "epoch": 0.0006063, "grad_norm": 1.6862188577651978, "learning_rate": 6.0620000000000006e-05, "loss": 0.7778, "step": 6063 }, { "epoch": 0.0006064, "grad_norm": 1.4250811338424683, "learning_rate": 6.063e-05, "loss": 0.6377, "step": 6064 }, { "epoch": 0.0006065, "grad_norm": 1.5216045379638672, "learning_rate": 6.064e-05, "loss": 0.7222, "step": 6065 }, { "epoch": 0.0006066, "grad_norm": 1.3850923776626587, "learning_rate": 6.0650000000000004e-05, "loss": 0.6938, "step": 6066 }, { "epoch": 0.0006067, "grad_norm": 1.2963297367095947, "learning_rate": 6.066e-05, "loss": 0.6162, "step": 6067 }, { "epoch": 0.0006068, "grad_norm": 1.714134931564331, "learning_rate": 6.067e-05, "loss": 0.8525, "step": 6068 }, { "epoch": 0.0006069, "grad_norm": 1.6639752388000488, "learning_rate": 6.068e-05, "loss": 0.7471, "step": 6069 }, { "epoch": 0.000607, "grad_norm": 1.5290673971176147, "learning_rate": 6.069e-05, "loss": 0.7007, "step": 6070 }, { "epoch": 0.0006071, "grad_norm": 1.3387386798858643, "learning_rate": 6.07e-05, "loss": 0.644, "step": 6071 }, { "epoch": 0.0006072, "grad_norm": 1.4449995756149292, "learning_rate": 6.071e-05, "loss": 0.6548, "step": 6072 }, { "epoch": 0.0006073, "grad_norm": 1.3186904191970825, "learning_rate": 6.072e-05, "loss": 0.6851, "step": 6073 }, { "epoch": 0.0006074, "grad_norm": 1.5486427545547485, "learning_rate": 6.0729999999999996e-05, "loss": 0.7241, "step": 6074 }, { "epoch": 0.0006075, "grad_norm": 2.433990716934204, "learning_rate": 6.074e-05, "loss": 0.6904, "step": 6075 }, { "epoch": 0.0006076, "grad_norm": 1.4973959922790527, "learning_rate": 6.0750000000000006e-05, "loss": 0.6118, "step": 6076 }, { "epoch": 0.0006077, "grad_norm": 1.5700976848602295, "learning_rate": 6.076000000000001e-05, "loss": 0.6963, "step": 6077 }, { "epoch": 0.0006078, "grad_norm": 1.791594386100769, "learning_rate": 6.0769999999999996e-05, "loss": 0.751, "step": 6078 }, { "epoch": 0.0006079, "grad_norm": 2.1504902839660645, "learning_rate": 6.0780000000000004e-05, "loss": 0.8877, "step": 6079 }, { "epoch": 0.000608, "grad_norm": 1.4483439922332764, "learning_rate": 6.0790000000000006e-05, "loss": 0.623, "step": 6080 }, { "epoch": 0.0006081, "grad_norm": 1.638651967048645, "learning_rate": 6.08e-05, "loss": 0.6558, "step": 6081 }, { "epoch": 0.0006082, "grad_norm": 1.3742382526397705, "learning_rate": 6.081e-05, "loss": 0.6445, "step": 6082 }, { "epoch": 0.0006083, "grad_norm": 1.690085768699646, "learning_rate": 6.0820000000000004e-05, "loss": 0.6929, "step": 6083 }, { "epoch": 0.0006084, "grad_norm": 1.4449642896652222, "learning_rate": 6.083e-05, "loss": 0.6514, "step": 6084 }, { "epoch": 0.0006085, "grad_norm": 2.428176164627075, "learning_rate": 6.084e-05, "loss": 0.7739, "step": 6085 }, { "epoch": 0.0006086, "grad_norm": 1.5959484577178955, "learning_rate": 6.085e-05, "loss": 0.6357, "step": 6086 }, { "epoch": 0.0006087, "grad_norm": 1.6490490436553955, "learning_rate": 6.0860000000000003e-05, "loss": 0.6699, "step": 6087 }, { "epoch": 0.0006088, "grad_norm": 1.5487797260284424, "learning_rate": 6.087e-05, "loss": 0.6968, "step": 6088 }, { "epoch": 0.0006089, "grad_norm": 1.2236945629119873, "learning_rate": 6.088e-05, "loss": 0.6001, "step": 6089 }, { "epoch": 0.000609, "grad_norm": 1.7166575193405151, "learning_rate": 6.089e-05, "loss": 0.7114, "step": 6090 }, { "epoch": 0.0006091, "grad_norm": 1.4670919179916382, "learning_rate": 6.0899999999999996e-05, "loss": 0.6343, "step": 6091 }, { "epoch": 0.0006092, "grad_norm": 1.4445433616638184, "learning_rate": 6.091e-05, "loss": 0.7012, "step": 6092 }, { "epoch": 0.0006093, "grad_norm": 1.329375982284546, "learning_rate": 6.0920000000000006e-05, "loss": 0.6147, "step": 6093 }, { "epoch": 0.0006094, "grad_norm": 1.2377262115478516, "learning_rate": 6.093000000000001e-05, "loss": 0.6064, "step": 6094 }, { "epoch": 0.0006095, "grad_norm": 1.3684781789779663, "learning_rate": 6.0939999999999996e-05, "loss": 0.6797, "step": 6095 }, { "epoch": 0.0006096, "grad_norm": 1.2610841989517212, "learning_rate": 6.0950000000000004e-05, "loss": 0.6279, "step": 6096 }, { "epoch": 0.0006097, "grad_norm": 1.3230469226837158, "learning_rate": 6.0960000000000006e-05, "loss": 0.6284, "step": 6097 }, { "epoch": 0.0006098, "grad_norm": 1.3344494104385376, "learning_rate": 6.097e-05, "loss": 0.6455, "step": 6098 }, { "epoch": 0.0006099, "grad_norm": 2.1403236389160156, "learning_rate": 6.098e-05, "loss": 1.0176, "step": 6099 }, { "epoch": 0.00061, "grad_norm": 1.5162384510040283, "learning_rate": 6.0990000000000004e-05, "loss": 0.7207, "step": 6100 }, { "epoch": 0.0006101, "grad_norm": 2.004774808883667, "learning_rate": 6.1e-05, "loss": 0.8926, "step": 6101 }, { "epoch": 0.0006102, "grad_norm": 1.3667763471603394, "learning_rate": 6.101e-05, "loss": 0.6323, "step": 6102 }, { "epoch": 0.0006103, "grad_norm": 1.194966435432434, "learning_rate": 6.102e-05, "loss": 0.583, "step": 6103 }, { "epoch": 0.0006104, "grad_norm": 1.3216735124588013, "learning_rate": 6.1030000000000004e-05, "loss": 0.6465, "step": 6104 }, { "epoch": 0.0006105, "grad_norm": 1.486382246017456, "learning_rate": 6.104e-05, "loss": 0.6807, "step": 6105 }, { "epoch": 0.0006106, "grad_norm": 2.2750372886657715, "learning_rate": 6.105e-05, "loss": 1.0103, "step": 6106 }, { "epoch": 0.0006107, "grad_norm": 3.482034921646118, "learning_rate": 6.106e-05, "loss": 0.7627, "step": 6107 }, { "epoch": 0.0006108, "grad_norm": 1.6870607137680054, "learning_rate": 6.107e-05, "loss": 0.6392, "step": 6108 }, { "epoch": 0.0006109, "grad_norm": 4.1532111167907715, "learning_rate": 6.108e-05, "loss": 1.0767, "step": 6109 }, { "epoch": 0.000611, "grad_norm": 1.5247021913528442, "learning_rate": 6.109e-05, "loss": 0.6348, "step": 6110 }, { "epoch": 0.0006111, "grad_norm": 1.549442172050476, "learning_rate": 6.110000000000001e-05, "loss": 0.6675, "step": 6111 }, { "epoch": 0.0006112, "grad_norm": 1.6295372247695923, "learning_rate": 6.111e-05, "loss": 0.6895, "step": 6112 }, { "epoch": 0.0006113, "grad_norm": 1.4706110954284668, "learning_rate": 6.112e-05, "loss": 0.6245, "step": 6113 }, { "epoch": 0.0006114, "grad_norm": 1.739664912223816, "learning_rate": 6.113e-05, "loss": 0.8325, "step": 6114 }, { "epoch": 0.0006115, "grad_norm": 1.614557147026062, "learning_rate": 6.114e-05, "loss": 0.7148, "step": 6115 }, { "epoch": 0.0006116, "grad_norm": 1.623680591583252, "learning_rate": 6.115e-05, "loss": 0.6597, "step": 6116 }, { "epoch": 0.0006117, "grad_norm": 1.3491045236587524, "learning_rate": 6.116e-05, "loss": 0.5996, "step": 6117 }, { "epoch": 0.0006118, "grad_norm": 1.3005375862121582, "learning_rate": 6.117e-05, "loss": 0.5903, "step": 6118 }, { "epoch": 0.0006119, "grad_norm": 1.3519699573516846, "learning_rate": 6.118e-05, "loss": 0.6133, "step": 6119 }, { "epoch": 0.000612, "grad_norm": 2.895958423614502, "learning_rate": 6.119e-05, "loss": 0.9756, "step": 6120 }, { "epoch": 0.0006121, "grad_norm": 2.477351427078247, "learning_rate": 6.120000000000001e-05, "loss": 0.8291, "step": 6121 }, { "epoch": 0.0006122, "grad_norm": 2.2054953575134277, "learning_rate": 6.120999999999999e-05, "loss": 0.7183, "step": 6122 }, { "epoch": 0.0006123, "grad_norm": 3.672135829925537, "learning_rate": 6.122e-05, "loss": 1.187, "step": 6123 }, { "epoch": 0.0006124, "grad_norm": 2.4219377040863037, "learning_rate": 6.123000000000001e-05, "loss": 0.7812, "step": 6124 }, { "epoch": 0.0006125, "grad_norm": 5.223512649536133, "learning_rate": 6.124e-05, "loss": 0.9116, "step": 6125 }, { "epoch": 0.0006126, "grad_norm": 2.252377986907959, "learning_rate": 6.125e-05, "loss": 1.0322, "step": 6126 }, { "epoch": 0.0006127, "grad_norm": 3.5240578651428223, "learning_rate": 6.126e-05, "loss": 1.2168, "step": 6127 }, { "epoch": 0.0006128, "grad_norm": 2.5054214000701904, "learning_rate": 6.127e-05, "loss": 0.8188, "step": 6128 }, { "epoch": 0.0006129, "grad_norm": 1.5566062927246094, "learning_rate": 6.128e-05, "loss": 0.7041, "step": 6129 }, { "epoch": 0.000613, "grad_norm": 1.4800119400024414, "learning_rate": 6.129e-05, "loss": 0.5854, "step": 6130 }, { "epoch": 0.0006131, "grad_norm": 1.8147279024124146, "learning_rate": 6.13e-05, "loss": 0.8003, "step": 6131 }, { "epoch": 0.0006132, "grad_norm": 1.4204871654510498, "learning_rate": 6.131e-05, "loss": 0.6201, "step": 6132 }, { "epoch": 0.0006133, "grad_norm": 1.4384766817092896, "learning_rate": 6.132e-05, "loss": 0.6345, "step": 6133 }, { "epoch": 0.0006134, "grad_norm": 1.5662084817886353, "learning_rate": 6.133e-05, "loss": 0.689, "step": 6134 }, { "epoch": 0.0006135, "grad_norm": 1.5638211965560913, "learning_rate": 6.133999999999999e-05, "loss": 0.6387, "step": 6135 }, { "epoch": 0.0006136, "grad_norm": 1.7658390998840332, "learning_rate": 6.135e-05, "loss": 0.7488, "step": 6136 }, { "epoch": 0.0006137, "grad_norm": 1.5555974245071411, "learning_rate": 6.136000000000001e-05, "loss": 0.6445, "step": 6137 }, { "epoch": 0.0006138, "grad_norm": 1.6092109680175781, "learning_rate": 6.137e-05, "loss": 0.6777, "step": 6138 }, { "epoch": 0.0006139, "grad_norm": 1.5195221900939941, "learning_rate": 6.138e-05, "loss": 0.6519, "step": 6139 }, { "epoch": 0.000614, "grad_norm": 4.042060852050781, "learning_rate": 6.139000000000001e-05, "loss": 0.9448, "step": 6140 }, { "epoch": 0.0006141, "grad_norm": 1.4042551517486572, "learning_rate": 6.14e-05, "loss": 0.603, "step": 6141 }, { "epoch": 0.0006142, "grad_norm": 1.4371025562286377, "learning_rate": 6.141e-05, "loss": 0.6411, "step": 6142 }, { "epoch": 0.0006143, "grad_norm": 1.1401251554489136, "learning_rate": 6.142e-05, "loss": 0.5542, "step": 6143 }, { "epoch": 0.0006144, "grad_norm": 1.6855483055114746, "learning_rate": 6.143e-05, "loss": 0.7407, "step": 6144 }, { "epoch": 0.0006145, "grad_norm": 1.4168471097946167, "learning_rate": 6.144000000000001e-05, "loss": 0.6001, "step": 6145 }, { "epoch": 0.0006146, "grad_norm": 2.1707684993743896, "learning_rate": 6.145e-05, "loss": 0.9136, "step": 6146 }, { "epoch": 0.0006147, "grad_norm": 1.533423662185669, "learning_rate": 6.146e-05, "loss": 0.6665, "step": 6147 }, { "epoch": 0.0006148, "grad_norm": 1.4535192251205444, "learning_rate": 6.147e-05, "loss": 0.6221, "step": 6148 }, { "epoch": 0.0006149, "grad_norm": 1.4669100046157837, "learning_rate": 6.148e-05, "loss": 0.7026, "step": 6149 }, { "epoch": 0.000615, "grad_norm": 1.4489563703536987, "learning_rate": 6.149e-05, "loss": 0.6538, "step": 6150 }, { "epoch": 0.0006151, "grad_norm": 1.4876471757888794, "learning_rate": 6.15e-05, "loss": 0.6748, "step": 6151 }, { "epoch": 0.0006152, "grad_norm": 1.736519694328308, "learning_rate": 6.151e-05, "loss": 0.8223, "step": 6152 }, { "epoch": 0.0006153, "grad_norm": 1.3515825271606445, "learning_rate": 6.152e-05, "loss": 0.5974, "step": 6153 }, { "epoch": 0.0006154, "grad_norm": 1.3721874952316284, "learning_rate": 6.153e-05, "loss": 0.5972, "step": 6154 }, { "epoch": 0.0006155, "grad_norm": 1.4386078119277954, "learning_rate": 6.154000000000001e-05, "loss": 0.6619, "step": 6155 }, { "epoch": 0.0006156, "grad_norm": 1.6986792087554932, "learning_rate": 6.154999999999999e-05, "loss": 0.6836, "step": 6156 }, { "epoch": 0.0006157, "grad_norm": 2.247748613357544, "learning_rate": 6.156e-05, "loss": 0.6572, "step": 6157 }, { "epoch": 0.0006158, "grad_norm": 1.9118703603744507, "learning_rate": 6.157000000000001e-05, "loss": 0.7832, "step": 6158 }, { "epoch": 0.0006159, "grad_norm": 1.3325278759002686, "learning_rate": 6.158e-05, "loss": 0.5972, "step": 6159 }, { "epoch": 0.000616, "grad_norm": 2.9448111057281494, "learning_rate": 6.159e-05, "loss": 0.9946, "step": 6160 }, { "epoch": 0.0006161, "grad_norm": 1.3446276187896729, "learning_rate": 6.16e-05, "loss": 0.5762, "step": 6161 }, { "epoch": 0.0006162, "grad_norm": 1.5767836570739746, "learning_rate": 6.161e-05, "loss": 0.6489, "step": 6162 }, { "epoch": 0.0006163, "grad_norm": 2.111865282058716, "learning_rate": 6.162e-05, "loss": 0.8223, "step": 6163 }, { "epoch": 0.0006164, "grad_norm": 2.025707721710205, "learning_rate": 6.163e-05, "loss": 0.7529, "step": 6164 }, { "epoch": 0.0006165, "grad_norm": 1.5718879699707031, "learning_rate": 6.164e-05, "loss": 0.6123, "step": 6165 }, { "epoch": 0.0006166, "grad_norm": 1.5455278158187866, "learning_rate": 6.165e-05, "loss": 0.6572, "step": 6166 }, { "epoch": 0.0006167, "grad_norm": 1.1608492136001587, "learning_rate": 6.166e-05, "loss": 0.5386, "step": 6167 }, { "epoch": 0.0006168, "grad_norm": 1.698362946510315, "learning_rate": 6.167e-05, "loss": 0.7178, "step": 6168 }, { "epoch": 0.0006169, "grad_norm": 1.5489366054534912, "learning_rate": 6.167999999999999e-05, "loss": 0.5869, "step": 6169 }, { "epoch": 0.000617, "grad_norm": 1.7101398706436157, "learning_rate": 6.169e-05, "loss": 0.646, "step": 6170 }, { "epoch": 0.0006171, "grad_norm": 1.4007785320281982, "learning_rate": 6.170000000000001e-05, "loss": 0.6118, "step": 6171 }, { "epoch": 0.0006172, "grad_norm": 1.5472906827926636, "learning_rate": 6.171e-05, "loss": 0.6597, "step": 6172 }, { "epoch": 0.0006173, "grad_norm": 2.241448402404785, "learning_rate": 6.172e-05, "loss": 0.6831, "step": 6173 }, { "epoch": 0.0006174, "grad_norm": 1.3851557970046997, "learning_rate": 6.173000000000001e-05, "loss": 0.5569, "step": 6174 }, { "epoch": 0.0006175, "grad_norm": 1.3902335166931152, "learning_rate": 6.174e-05, "loss": 0.5762, "step": 6175 }, { "epoch": 0.0006176, "grad_norm": 1.2751723527908325, "learning_rate": 6.175e-05, "loss": 0.5654, "step": 6176 }, { "epoch": 0.0006177, "grad_norm": 1.3091758489608765, "learning_rate": 6.176e-05, "loss": 0.5649, "step": 6177 }, { "epoch": 0.0006178, "grad_norm": 1.3855472803115845, "learning_rate": 6.177e-05, "loss": 0.5571, "step": 6178 }, { "epoch": 0.0006179, "grad_norm": 2.753174304962158, "learning_rate": 6.178000000000001e-05, "loss": 0.8467, "step": 6179 }, { "epoch": 0.000618, "grad_norm": 1.5822502374649048, "learning_rate": 6.179e-05, "loss": 0.646, "step": 6180 }, { "epoch": 0.0006181, "grad_norm": 1.5361640453338623, "learning_rate": 6.18e-05, "loss": 0.6611, "step": 6181 }, { "epoch": 0.0006182, "grad_norm": 1.59479558467865, "learning_rate": 6.181e-05, "loss": 0.6191, "step": 6182 }, { "epoch": 0.0006183, "grad_norm": 1.317305088043213, "learning_rate": 6.182e-05, "loss": 0.5845, "step": 6183 }, { "epoch": 0.0006184, "grad_norm": 1.5284531116485596, "learning_rate": 6.183e-05, "loss": 0.6328, "step": 6184 }, { "epoch": 0.0006185, "grad_norm": 1.4130276441574097, "learning_rate": 6.184e-05, "loss": 0.6101, "step": 6185 }, { "epoch": 0.0006186, "grad_norm": 1.2827122211456299, "learning_rate": 6.185e-05, "loss": 0.5884, "step": 6186 }, { "epoch": 0.0006187, "grad_norm": 1.1124165058135986, "learning_rate": 6.186e-05, "loss": 0.5239, "step": 6187 }, { "epoch": 0.0006188, "grad_norm": 1.3330076932907104, "learning_rate": 6.187e-05, "loss": 0.5854, "step": 6188 }, { "epoch": 0.0006189, "grad_norm": 2.5464541912078857, "learning_rate": 6.188000000000001e-05, "loss": 0.8247, "step": 6189 }, { "epoch": 0.000619, "grad_norm": 1.4788142442703247, "learning_rate": 6.188999999999999e-05, "loss": 0.6431, "step": 6190 }, { "epoch": 0.0006191, "grad_norm": 1.3143922090530396, "learning_rate": 6.19e-05, "loss": 0.5547, "step": 6191 }, { "epoch": 0.0006192, "grad_norm": 1.1701135635375977, "learning_rate": 6.191000000000001e-05, "loss": 0.5308, "step": 6192 }, { "epoch": 0.0006193, "grad_norm": 1.983077883720398, "learning_rate": 6.192e-05, "loss": 0.7656, "step": 6193 }, { "epoch": 0.0006194, "grad_norm": 1.4042465686798096, "learning_rate": 6.193e-05, "loss": 0.6089, "step": 6194 }, { "epoch": 0.0006195, "grad_norm": 1.273660659790039, "learning_rate": 6.194e-05, "loss": 0.605, "step": 6195 }, { "epoch": 0.0006196, "grad_norm": 1.2940797805786133, "learning_rate": 6.195e-05, "loss": 0.6072, "step": 6196 }, { "epoch": 0.0006197, "grad_norm": 1.2838053703308105, "learning_rate": 6.196e-05, "loss": 0.5681, "step": 6197 }, { "epoch": 0.0006198, "grad_norm": 1.2031972408294678, "learning_rate": 6.197e-05, "loss": 0.5559, "step": 6198 }, { "epoch": 0.0006199, "grad_norm": 2.070176362991333, "learning_rate": 6.198e-05, "loss": 0.7498, "step": 6199 }, { "epoch": 0.00062, "grad_norm": 1.229961633682251, "learning_rate": 6.199e-05, "loss": 0.5618, "step": 6200 }, { "epoch": 0.0006201, "grad_norm": 1.6335248947143555, "learning_rate": 6.2e-05, "loss": 0.6592, "step": 6201 }, { "epoch": 0.0006202, "grad_norm": 1.5217071771621704, "learning_rate": 6.201e-05, "loss": 0.6636, "step": 6202 }, { "epoch": 0.0006203, "grad_norm": 1.7696077823638916, "learning_rate": 6.201999999999999e-05, "loss": 0.6807, "step": 6203 }, { "epoch": 0.0006204, "grad_norm": 2.0146279335021973, "learning_rate": 6.203e-05, "loss": 0.9287, "step": 6204 }, { "epoch": 0.0006205, "grad_norm": 1.5057393312454224, "learning_rate": 6.204000000000001e-05, "loss": 0.6523, "step": 6205 }, { "epoch": 0.0006206, "grad_norm": 1.3235162496566772, "learning_rate": 6.205e-05, "loss": 0.5752, "step": 6206 }, { "epoch": 0.0006207, "grad_norm": 1.2800863981246948, "learning_rate": 6.206e-05, "loss": 0.5688, "step": 6207 }, { "epoch": 0.0006208, "grad_norm": 1.4795315265655518, "learning_rate": 6.207000000000001e-05, "loss": 0.6045, "step": 6208 }, { "epoch": 0.0006209, "grad_norm": 2.372642755508423, "learning_rate": 6.208e-05, "loss": 0.6807, "step": 6209 }, { "epoch": 0.000621, "grad_norm": 1.4887479543685913, "learning_rate": 6.209e-05, "loss": 0.5566, "step": 6210 }, { "epoch": 0.0006211, "grad_norm": 1.4304701089859009, "learning_rate": 6.21e-05, "loss": 0.5859, "step": 6211 }, { "epoch": 0.0006212, "grad_norm": 1.636297583580017, "learning_rate": 6.211e-05, "loss": 0.6465, "step": 6212 }, { "epoch": 0.0006213, "grad_norm": 1.3796359300613403, "learning_rate": 6.212000000000001e-05, "loss": 0.5728, "step": 6213 }, { "epoch": 0.0006214, "grad_norm": 1.2996286153793335, "learning_rate": 6.213e-05, "loss": 0.5398, "step": 6214 }, { "epoch": 0.0006215, "grad_norm": 1.234860897064209, "learning_rate": 6.214e-05, "loss": 0.5442, "step": 6215 }, { "epoch": 0.0006216, "grad_norm": 1.4250895977020264, "learning_rate": 6.215e-05, "loss": 0.6006, "step": 6216 }, { "epoch": 0.0006217, "grad_norm": 1.20131516456604, "learning_rate": 6.216e-05, "loss": 0.5239, "step": 6217 }, { "epoch": 0.0006218, "grad_norm": 1.3933985233306885, "learning_rate": 6.217e-05, "loss": 0.584, "step": 6218 }, { "epoch": 0.0006219, "grad_norm": 2.1496052742004395, "learning_rate": 6.218e-05, "loss": 0.7788, "step": 6219 }, { "epoch": 0.000622, "grad_norm": 1.558510184288025, "learning_rate": 6.219e-05, "loss": 0.6006, "step": 6220 }, { "epoch": 0.0006221, "grad_norm": 4.112026214599609, "learning_rate": 6.22e-05, "loss": 0.7517, "step": 6221 }, { "epoch": 0.0006222, "grad_norm": 1.8367027044296265, "learning_rate": 6.221e-05, "loss": 0.6196, "step": 6222 }, { "epoch": 0.0006223, "grad_norm": 2.0355629920959473, "learning_rate": 6.222000000000001e-05, "loss": 0.8071, "step": 6223 }, { "epoch": 0.0006224, "grad_norm": 1.5606682300567627, "learning_rate": 6.222999999999999e-05, "loss": 0.6716, "step": 6224 }, { "epoch": 0.0006225, "grad_norm": 1.454939365386963, "learning_rate": 6.224e-05, "loss": 0.6104, "step": 6225 }, { "epoch": 0.0006226, "grad_norm": 2.0981926918029785, "learning_rate": 6.225000000000001e-05, "loss": 0.5967, "step": 6226 }, { "epoch": 0.0006227, "grad_norm": 1.3601634502410889, "learning_rate": 6.226e-05, "loss": 0.543, "step": 6227 }, { "epoch": 0.0006228, "grad_norm": 1.3056219816207886, "learning_rate": 6.227e-05, "loss": 0.5432, "step": 6228 }, { "epoch": 0.0006229, "grad_norm": 1.2218716144561768, "learning_rate": 6.228000000000001e-05, "loss": 0.5344, "step": 6229 }, { "epoch": 0.000623, "grad_norm": 1.4955134391784668, "learning_rate": 6.229e-05, "loss": 0.6226, "step": 6230 }, { "epoch": 0.0006231, "grad_norm": 1.5450738668441772, "learning_rate": 6.23e-05, "loss": 0.5764, "step": 6231 }, { "epoch": 0.0006232, "grad_norm": 1.5347802639007568, "learning_rate": 6.231e-05, "loss": 0.6047, "step": 6232 }, { "epoch": 0.0006233, "grad_norm": 1.5982639789581299, "learning_rate": 6.232e-05, "loss": 0.6714, "step": 6233 }, { "epoch": 0.0006234, "grad_norm": 2.6113107204437256, "learning_rate": 6.233e-05, "loss": 0.7083, "step": 6234 }, { "epoch": 0.0006235, "grad_norm": 1.6445200443267822, "learning_rate": 6.234e-05, "loss": 0.6084, "step": 6235 }, { "epoch": 0.0006236, "grad_norm": 1.4309769868850708, "learning_rate": 6.235e-05, "loss": 0.5503, "step": 6236 }, { "epoch": 0.0006237, "grad_norm": 2.6508233547210693, "learning_rate": 6.235999999999999e-05, "loss": 0.7554, "step": 6237 }, { "epoch": 0.0006238, "grad_norm": 1.892777681350708, "learning_rate": 6.237e-05, "loss": 0.731, "step": 6238 }, { "epoch": 0.0006239, "grad_norm": 1.422659993171692, "learning_rate": 6.238000000000001e-05, "loss": 0.5649, "step": 6239 }, { "epoch": 0.000624, "grad_norm": 1.8252531290054321, "learning_rate": 6.239e-05, "loss": 0.6064, "step": 6240 }, { "epoch": 0.0006241, "grad_norm": 2.0503756999969482, "learning_rate": 6.24e-05, "loss": 0.9309, "step": 6241 }, { "epoch": 0.0006242, "grad_norm": 1.3754830360412598, "learning_rate": 6.241000000000001e-05, "loss": 0.5793, "step": 6242 }, { "epoch": 0.0006243, "grad_norm": 1.2413182258605957, "learning_rate": 6.242e-05, "loss": 0.5415, "step": 6243 }, { "epoch": 0.0006244, "grad_norm": 1.157693862915039, "learning_rate": 6.243e-05, "loss": 0.5063, "step": 6244 }, { "epoch": 0.0006245, "grad_norm": 2.4091522693634033, "learning_rate": 6.244e-05, "loss": 0.717, "step": 6245 }, { "epoch": 0.0006246, "grad_norm": 1.626430630683899, "learning_rate": 6.245e-05, "loss": 0.5918, "step": 6246 }, { "epoch": 0.0006247, "grad_norm": 1.3793022632598877, "learning_rate": 6.246000000000001e-05, "loss": 0.5386, "step": 6247 }, { "epoch": 0.0006248, "grad_norm": 3.696800947189331, "learning_rate": 6.247e-05, "loss": 1.2209, "step": 6248 }, { "epoch": 0.0006249, "grad_norm": 1.4566258192062378, "learning_rate": 6.248e-05, "loss": 0.6089, "step": 6249 }, { "epoch": 0.000625, "grad_norm": 1.323639154434204, "learning_rate": 6.249e-05, "loss": 0.5266, "step": 6250 }, { "epoch": 0.0006251, "grad_norm": 2.112474203109741, "learning_rate": 6.25e-05, "loss": 0.6899, "step": 6251 }, { "epoch": 0.0006252, "grad_norm": 1.2852239608764648, "learning_rate": 6.251e-05, "loss": 0.5193, "step": 6252 }, { "epoch": 0.0006253, "grad_norm": 1.354556679725647, "learning_rate": 6.252e-05, "loss": 0.6016, "step": 6253 }, { "epoch": 0.0006254, "grad_norm": 1.2461886405944824, "learning_rate": 6.253e-05, "loss": 0.5354, "step": 6254 }, { "epoch": 0.0006255, "grad_norm": 2.013084888458252, "learning_rate": 6.254e-05, "loss": 0.9143, "step": 6255 }, { "epoch": 0.0006256, "grad_norm": 1.2279975414276123, "learning_rate": 6.255e-05, "loss": 0.5317, "step": 6256 }, { "epoch": 0.0006257, "grad_norm": 1.3620824813842773, "learning_rate": 6.256000000000001e-05, "loss": 0.6245, "step": 6257 }, { "epoch": 0.0006258, "grad_norm": 1.3307284116744995, "learning_rate": 6.256999999999999e-05, "loss": 0.5559, "step": 6258 }, { "epoch": 0.0006259, "grad_norm": 1.2366081476211548, "learning_rate": 6.258e-05, "loss": 0.5493, "step": 6259 }, { "epoch": 0.000626, "grad_norm": 1.2821134328842163, "learning_rate": 6.259000000000001e-05, "loss": 0.5879, "step": 6260 }, { "epoch": 0.0006261, "grad_norm": 2.112119197845459, "learning_rate": 6.26e-05, "loss": 0.8433, "step": 6261 }, { "epoch": 0.0006262, "grad_norm": 1.6771800518035889, "learning_rate": 6.261e-05, "loss": 0.6709, "step": 6262 }, { "epoch": 0.0006263, "grad_norm": 1.531270146369934, "learning_rate": 6.262000000000001e-05, "loss": 0.5874, "step": 6263 }, { "epoch": 0.0006264, "grad_norm": 1.2914503812789917, "learning_rate": 6.263e-05, "loss": 0.5488, "step": 6264 }, { "epoch": 0.0006265, "grad_norm": 1.2997040748596191, "learning_rate": 6.264e-05, "loss": 0.5583, "step": 6265 }, { "epoch": 0.0006266, "grad_norm": 1.2204697132110596, "learning_rate": 6.265e-05, "loss": 0.499, "step": 6266 }, { "epoch": 0.0006267, "grad_norm": 1.723503828048706, "learning_rate": 6.266e-05, "loss": 0.6738, "step": 6267 }, { "epoch": 0.0006268, "grad_norm": 1.158341407775879, "learning_rate": 6.267e-05, "loss": 0.5168, "step": 6268 }, { "epoch": 0.0006269, "grad_norm": 1.4224638938903809, "learning_rate": 6.268e-05, "loss": 0.5784, "step": 6269 }, { "epoch": 0.000627, "grad_norm": 1.2845311164855957, "learning_rate": 6.269e-05, "loss": 0.5708, "step": 6270 }, { "epoch": 0.0006271, "grad_norm": 1.322809100151062, "learning_rate": 6.269999999999999e-05, "loss": 0.5571, "step": 6271 }, { "epoch": 0.0006272, "grad_norm": 1.2853106260299683, "learning_rate": 6.271e-05, "loss": 0.5225, "step": 6272 }, { "epoch": 0.0006273, "grad_norm": 1.1261441707611084, "learning_rate": 6.272000000000001e-05, "loss": 0.4937, "step": 6273 }, { "epoch": 0.0006274, "grad_norm": 1.8010724782943726, "learning_rate": 6.273e-05, "loss": 0.6426, "step": 6274 }, { "epoch": 0.0006275, "grad_norm": 2.123570203781128, "learning_rate": 6.274e-05, "loss": 0.7507, "step": 6275 }, { "epoch": 0.0006276, "grad_norm": 1.8357847929000854, "learning_rate": 6.275000000000001e-05, "loss": 0.5884, "step": 6276 }, { "epoch": 0.0006277, "grad_norm": 1.3460817337036133, "learning_rate": 6.276e-05, "loss": 0.5278, "step": 6277 }, { "epoch": 0.0006278, "grad_norm": 1.095627784729004, "learning_rate": 6.277e-05, "loss": 0.5063, "step": 6278 }, { "epoch": 0.0006279, "grad_norm": 1.7269349098205566, "learning_rate": 6.278e-05, "loss": 0.6533, "step": 6279 }, { "epoch": 0.000628, "grad_norm": 1.422385334968567, "learning_rate": 6.279e-05, "loss": 0.5439, "step": 6280 }, { "epoch": 0.0006281, "grad_norm": 1.2597748041152954, "learning_rate": 6.280000000000001e-05, "loss": 0.5471, "step": 6281 }, { "epoch": 0.0006282, "grad_norm": 1.303085446357727, "learning_rate": 6.281e-05, "loss": 0.5774, "step": 6282 }, { "epoch": 0.0006283, "grad_norm": 1.2224934101104736, "learning_rate": 6.282e-05, "loss": 0.5337, "step": 6283 }, { "epoch": 0.0006284, "grad_norm": 1.2406748533248901, "learning_rate": 6.283e-05, "loss": 0.4871, "step": 6284 }, { "epoch": 0.0006285, "grad_norm": 1.114234209060669, "learning_rate": 6.284e-05, "loss": 0.4927, "step": 6285 }, { "epoch": 0.0006286, "grad_norm": 1.326036810874939, "learning_rate": 6.285e-05, "loss": 0.6033, "step": 6286 }, { "epoch": 0.0006287, "grad_norm": 1.1210211515426636, "learning_rate": 6.286e-05, "loss": 0.4875, "step": 6287 }, { "epoch": 0.0006288, "grad_norm": 1.2281023263931274, "learning_rate": 6.287e-05, "loss": 0.5334, "step": 6288 }, { "epoch": 0.0006289, "grad_norm": 1.2000365257263184, "learning_rate": 6.288e-05, "loss": 0.4956, "step": 6289 }, { "epoch": 0.000629, "grad_norm": 1.6449037790298462, "learning_rate": 6.289e-05, "loss": 0.5977, "step": 6290 }, { "epoch": 0.0006291, "grad_norm": 2.3417065143585205, "learning_rate": 6.290000000000001e-05, "loss": 0.8713, "step": 6291 }, { "epoch": 0.0006292, "grad_norm": 1.3869773149490356, "learning_rate": 6.290999999999999e-05, "loss": 0.5161, "step": 6292 }, { "epoch": 0.0006293, "grad_norm": 1.1831398010253906, "learning_rate": 6.292e-05, "loss": 0.4766, "step": 6293 }, { "epoch": 0.0006294, "grad_norm": 1.2068742513656616, "learning_rate": 6.293000000000001e-05, "loss": 0.563, "step": 6294 }, { "epoch": 0.0006295, "grad_norm": 1.366097331047058, "learning_rate": 6.294e-05, "loss": 0.5474, "step": 6295 }, { "epoch": 0.0006296, "grad_norm": 1.2809611558914185, "learning_rate": 6.295e-05, "loss": 0.5127, "step": 6296 }, { "epoch": 0.0006297, "grad_norm": 1.66355562210083, "learning_rate": 6.296000000000001e-05, "loss": 0.5964, "step": 6297 }, { "epoch": 0.0006298, "grad_norm": 1.2659800052642822, "learning_rate": 6.297e-05, "loss": 0.5327, "step": 6298 }, { "epoch": 0.0006299, "grad_norm": 1.33529794216156, "learning_rate": 6.298e-05, "loss": 0.5581, "step": 6299 }, { "epoch": 0.00063, "grad_norm": 1.4095124006271362, "learning_rate": 6.299e-05, "loss": 0.5537, "step": 6300 }, { "epoch": 0.0006301, "grad_norm": 1.3011655807495117, "learning_rate": 6.3e-05, "loss": 0.4934, "step": 6301 }, { "epoch": 0.0006302, "grad_norm": 2.5561511516571045, "learning_rate": 6.301e-05, "loss": 0.8005, "step": 6302 }, { "epoch": 0.0006303, "grad_norm": 1.8032957315444946, "learning_rate": 6.302e-05, "loss": 0.7488, "step": 6303 }, { "epoch": 0.0006304, "grad_norm": 1.6584831476211548, "learning_rate": 6.303e-05, "loss": 0.5923, "step": 6304 }, { "epoch": 0.0006305, "grad_norm": 1.9853506088256836, "learning_rate": 6.303999999999999e-05, "loss": 0.9365, "step": 6305 }, { "epoch": 0.0006306, "grad_norm": 1.3332364559173584, "learning_rate": 6.305e-05, "loss": 0.5188, "step": 6306 }, { "epoch": 0.0006307, "grad_norm": 1.6622294187545776, "learning_rate": 6.306000000000001e-05, "loss": 0.7083, "step": 6307 }, { "epoch": 0.0006308, "grad_norm": 2.3817873001098633, "learning_rate": 6.307e-05, "loss": 0.7249, "step": 6308 }, { "epoch": 0.0006309, "grad_norm": 1.5819065570831299, "learning_rate": 6.308e-05, "loss": 0.7139, "step": 6309 }, { "epoch": 0.000631, "grad_norm": 1.358562707901001, "learning_rate": 6.309000000000001e-05, "loss": 0.5186, "step": 6310 }, { "epoch": 0.0006311, "grad_norm": 1.435044527053833, "learning_rate": 6.31e-05, "loss": 0.582, "step": 6311 }, { "epoch": 0.0006312, "grad_norm": 1.264384150505066, "learning_rate": 6.311e-05, "loss": 0.5269, "step": 6312 }, { "epoch": 0.0006313, "grad_norm": 2.0423264503479004, "learning_rate": 6.312e-05, "loss": 0.8406, "step": 6313 }, { "epoch": 0.0006314, "grad_norm": 1.5304577350616455, "learning_rate": 6.313e-05, "loss": 0.563, "step": 6314 }, { "epoch": 0.0006315, "grad_norm": 2.266411066055298, "learning_rate": 6.314000000000001e-05, "loss": 0.9119, "step": 6315 }, { "epoch": 0.0006316, "grad_norm": 1.722103476524353, "learning_rate": 6.315e-05, "loss": 0.5452, "step": 6316 }, { "epoch": 0.0006317, "grad_norm": 1.3955553770065308, "learning_rate": 6.316e-05, "loss": 0.5408, "step": 6317 }, { "epoch": 0.0006318, "grad_norm": 1.2551302909851074, "learning_rate": 6.317e-05, "loss": 0.5273, "step": 6318 }, { "epoch": 0.0006319, "grad_norm": 1.8961031436920166, "learning_rate": 6.318e-05, "loss": 0.5591, "step": 6319 }, { "epoch": 0.000632, "grad_norm": 7.158136367797852, "learning_rate": 6.319e-05, "loss": 0.8918, "step": 6320 }, { "epoch": 0.0006321, "grad_norm": 2.118948459625244, "learning_rate": 6.32e-05, "loss": 0.6646, "step": 6321 }, { "epoch": 0.0006322, "grad_norm": 1.848267674446106, "learning_rate": 6.321e-05, "loss": 0.5669, "step": 6322 }, { "epoch": 0.0006323, "grad_norm": 1.3671910762786865, "learning_rate": 6.322e-05, "loss": 0.5347, "step": 6323 }, { "epoch": 0.0006324, "grad_norm": 1.8404011726379395, "learning_rate": 6.323e-05, "loss": 0.8181, "step": 6324 }, { "epoch": 0.0006325, "grad_norm": 1.5496231317520142, "learning_rate": 6.324000000000001e-05, "loss": 0.48, "step": 6325 }, { "epoch": 0.0006326, "grad_norm": 2.027029514312744, "learning_rate": 6.324999999999999e-05, "loss": 0.7007, "step": 6326 }, { "epoch": 0.0006327, "grad_norm": 1.5602715015411377, "learning_rate": 6.326e-05, "loss": 0.5342, "step": 6327 }, { "epoch": 0.0006328, "grad_norm": 1.2844575643539429, "learning_rate": 6.327000000000001e-05, "loss": 0.4761, "step": 6328 }, { "epoch": 0.0006329, "grad_norm": 1.48810875415802, "learning_rate": 6.328e-05, "loss": 0.5132, "step": 6329 }, { "epoch": 0.000633, "grad_norm": 1.6981958150863647, "learning_rate": 6.329e-05, "loss": 0.6323, "step": 6330 }, { "epoch": 0.0006331, "grad_norm": 2.4248063564300537, "learning_rate": 6.330000000000001e-05, "loss": 0.6284, "step": 6331 }, { "epoch": 0.0006332, "grad_norm": 1.428120493888855, "learning_rate": 6.331e-05, "loss": 0.5042, "step": 6332 }, { "epoch": 0.0006333, "grad_norm": 1.2944762706756592, "learning_rate": 6.332e-05, "loss": 0.4751, "step": 6333 }, { "epoch": 0.0006334, "grad_norm": 2.560227394104004, "learning_rate": 6.333e-05, "loss": 0.7188, "step": 6334 }, { "epoch": 0.0006335, "grad_norm": 1.1113696098327637, "learning_rate": 6.334e-05, "loss": 0.4587, "step": 6335 }, { "epoch": 0.0006336, "grad_norm": 3.3463480472564697, "learning_rate": 6.335e-05, "loss": 1.0486, "step": 6336 }, { "epoch": 0.0006337, "grad_norm": 2.7146496772766113, "learning_rate": 6.336e-05, "loss": 0.7905, "step": 6337 }, { "epoch": 0.0006338, "grad_norm": 1.945244312286377, "learning_rate": 6.337e-05, "loss": 0.6143, "step": 6338 }, { "epoch": 0.0006339, "grad_norm": 1.4121110439300537, "learning_rate": 6.337999999999999e-05, "loss": 0.5398, "step": 6339 }, { "epoch": 0.000634, "grad_norm": 1.2575377225875854, "learning_rate": 6.339e-05, "loss": 0.4917, "step": 6340 }, { "epoch": 0.0006341, "grad_norm": 1.2989206314086914, "learning_rate": 6.34e-05, "loss": 0.4814, "step": 6341 }, { "epoch": 0.0006342, "grad_norm": 1.2086619138717651, "learning_rate": 6.341e-05, "loss": 0.4756, "step": 6342 }, { "epoch": 0.0006343, "grad_norm": 2.2826614379882812, "learning_rate": 6.342e-05, "loss": 0.8481, "step": 6343 }, { "epoch": 0.0006344, "grad_norm": 1.9749172925949097, "learning_rate": 6.343000000000001e-05, "loss": 0.6133, "step": 6344 }, { "epoch": 0.0006345, "grad_norm": 1.5513535737991333, "learning_rate": 6.344e-05, "loss": 0.6121, "step": 6345 }, { "epoch": 0.0006346, "grad_norm": 1.4886285066604614, "learning_rate": 6.345e-05, "loss": 0.6157, "step": 6346 }, { "epoch": 0.0006347, "grad_norm": 1.1188468933105469, "learning_rate": 6.346e-05, "loss": 0.4675, "step": 6347 }, { "epoch": 0.0006348, "grad_norm": 1.867576003074646, "learning_rate": 6.347e-05, "loss": 0.6309, "step": 6348 }, { "epoch": 0.0006349, "grad_norm": 1.3053021430969238, "learning_rate": 6.348000000000001e-05, "loss": 0.5254, "step": 6349 }, { "epoch": 0.000635, "grad_norm": 1.2170910835266113, "learning_rate": 6.349e-05, "loss": 0.4998, "step": 6350 }, { "epoch": 0.0006351, "grad_norm": 1.4668612480163574, "learning_rate": 6.35e-05, "loss": 0.5396, "step": 6351 }, { "epoch": 0.0006352, "grad_norm": 1.247548222541809, "learning_rate": 6.351000000000001e-05, "loss": 0.522, "step": 6352 }, { "epoch": 0.0006353, "grad_norm": 1.343698263168335, "learning_rate": 6.352e-05, "loss": 0.5933, "step": 6353 }, { "epoch": 0.0006354, "grad_norm": 1.2090539932250977, "learning_rate": 6.353e-05, "loss": 0.4805, "step": 6354 }, { "epoch": 0.0006355, "grad_norm": 1.5267757177352905, "learning_rate": 6.354e-05, "loss": 0.5281, "step": 6355 }, { "epoch": 0.0006356, "grad_norm": 1.6579869985580444, "learning_rate": 6.355e-05, "loss": 0.6057, "step": 6356 }, { "epoch": 0.0006357, "grad_norm": 1.513315200805664, "learning_rate": 6.356e-05, "loss": 0.5649, "step": 6357 }, { "epoch": 0.0006358, "grad_norm": 1.2920204401016235, "learning_rate": 6.357e-05, "loss": 0.5393, "step": 6358 }, { "epoch": 0.0006359, "grad_norm": 1.0864078998565674, "learning_rate": 6.358000000000001e-05, "loss": 0.4658, "step": 6359 }, { "epoch": 0.000636, "grad_norm": 1.2634496688842773, "learning_rate": 6.358999999999999e-05, "loss": 0.5022, "step": 6360 }, { "epoch": 0.0006361, "grad_norm": 1.105880856513977, "learning_rate": 6.36e-05, "loss": 0.4771, "step": 6361 }, { "epoch": 0.0006362, "grad_norm": 1.1405361890792847, "learning_rate": 6.361000000000001e-05, "loss": 0.5073, "step": 6362 }, { "epoch": 0.0006363, "grad_norm": 3.483867645263672, "learning_rate": 6.361999999999999e-05, "loss": 1.0667, "step": 6363 }, { "epoch": 0.0006364, "grad_norm": 1.7898037433624268, "learning_rate": 6.363e-05, "loss": 0.5388, "step": 6364 }, { "epoch": 0.0006365, "grad_norm": 1.2391436100006104, "learning_rate": 6.364000000000001e-05, "loss": 0.4507, "step": 6365 }, { "epoch": 0.0006366, "grad_norm": 1.1045516729354858, "learning_rate": 6.365e-05, "loss": 0.458, "step": 6366 }, { "epoch": 0.0006367, "grad_norm": 1.046285629272461, "learning_rate": 6.366e-05, "loss": 0.4648, "step": 6367 }, { "epoch": 0.0006368, "grad_norm": 1.1581072807312012, "learning_rate": 6.367e-05, "loss": 0.4712, "step": 6368 }, { "epoch": 0.0006369, "grad_norm": 1.6926162242889404, "learning_rate": 6.368e-05, "loss": 0.5728, "step": 6369 }, { "epoch": 0.000637, "grad_norm": 1.29767644405365, "learning_rate": 6.369e-05, "loss": 0.5198, "step": 6370 }, { "epoch": 0.0006371, "grad_norm": 1.136470079421997, "learning_rate": 6.37e-05, "loss": 0.4646, "step": 6371 }, { "epoch": 0.0006372, "grad_norm": 1.0796231031417847, "learning_rate": 6.371e-05, "loss": 0.429, "step": 6372 }, { "epoch": 0.0006373, "grad_norm": 1.1330997943878174, "learning_rate": 6.371999999999999e-05, "loss": 0.4595, "step": 6373 }, { "epoch": 0.0006374, "grad_norm": 1.1513479948043823, "learning_rate": 6.373e-05, "loss": 0.4875, "step": 6374 }, { "epoch": 0.0006375, "grad_norm": 1.2037591934204102, "learning_rate": 6.374e-05, "loss": 0.5459, "step": 6375 }, { "epoch": 0.0006376, "grad_norm": 1.4171353578567505, "learning_rate": 6.375e-05, "loss": 0.5518, "step": 6376 }, { "epoch": 0.0006377, "grad_norm": 1.5970433950424194, "learning_rate": 6.376e-05, "loss": 0.5161, "step": 6377 }, { "epoch": 0.0006378, "grad_norm": 1.1912060976028442, "learning_rate": 6.377000000000001e-05, "loss": 0.4429, "step": 6378 }, { "epoch": 0.0006379, "grad_norm": 1.2796026468276978, "learning_rate": 6.378e-05, "loss": 0.4963, "step": 6379 }, { "epoch": 0.000638, "grad_norm": 1.4407531023025513, "learning_rate": 6.379e-05, "loss": 0.5935, "step": 6380 }, { "epoch": 0.0006381, "grad_norm": 1.348570704460144, "learning_rate": 6.38e-05, "loss": 0.5466, "step": 6381 }, { "epoch": 0.0006382, "grad_norm": 1.1336405277252197, "learning_rate": 6.381e-05, "loss": 0.4617, "step": 6382 }, { "epoch": 0.0006383, "grad_norm": 1.2180068492889404, "learning_rate": 6.382000000000001e-05, "loss": 0.4634, "step": 6383 }, { "epoch": 0.0006384, "grad_norm": 1.0494590997695923, "learning_rate": 6.383e-05, "loss": 0.4595, "step": 6384 }, { "epoch": 0.0006385, "grad_norm": 1.3279155492782593, "learning_rate": 6.384e-05, "loss": 0.5374, "step": 6385 }, { "epoch": 0.0006386, "grad_norm": 1.906715989112854, "learning_rate": 6.385000000000001e-05, "loss": 0.5901, "step": 6386 }, { "epoch": 0.0006387, "grad_norm": 2.3137943744659424, "learning_rate": 6.386e-05, "loss": 0.6785, "step": 6387 }, { "epoch": 0.0006388, "grad_norm": 3.7279317378997803, "learning_rate": 6.387e-05, "loss": 0.7988, "step": 6388 }, { "epoch": 0.0006389, "grad_norm": 2.4148833751678467, "learning_rate": 6.388e-05, "loss": 0.7737, "step": 6389 }, { "epoch": 0.000639, "grad_norm": 1.3121864795684814, "learning_rate": 6.389e-05, "loss": 0.4556, "step": 6390 }, { "epoch": 0.0006391, "grad_norm": 1.8153197765350342, "learning_rate": 6.39e-05, "loss": 0.6904, "step": 6391 }, { "epoch": 0.0006392, "grad_norm": 1.1611392498016357, "learning_rate": 6.391e-05, "loss": 0.4839, "step": 6392 }, { "epoch": 0.0006393, "grad_norm": 1.3582239151000977, "learning_rate": 6.392000000000001e-05, "loss": 0.5037, "step": 6393 }, { "epoch": 0.0006394, "grad_norm": 1.4697034358978271, "learning_rate": 6.392999999999999e-05, "loss": 0.5293, "step": 6394 }, { "epoch": 0.0006395, "grad_norm": 1.679388165473938, "learning_rate": 6.394e-05, "loss": 0.728, "step": 6395 }, { "epoch": 0.0006396, "grad_norm": 1.0928062200546265, "learning_rate": 6.395000000000001e-05, "loss": 0.4265, "step": 6396 }, { "epoch": 0.0006397, "grad_norm": 8.773086547851562, "learning_rate": 6.395999999999999e-05, "loss": 1.2278, "step": 6397 }, { "epoch": 0.0006398, "grad_norm": 1.7383633852005005, "learning_rate": 6.397e-05, "loss": 0.5452, "step": 6398 }, { "epoch": 0.0006399, "grad_norm": 1.2912527322769165, "learning_rate": 6.398000000000001e-05, "loss": 0.4585, "step": 6399 }, { "epoch": 0.00064, "grad_norm": 1.9075850248336792, "learning_rate": 6.399e-05, "loss": 0.5771, "step": 6400 }, { "epoch": 0.0006401, "grad_norm": 2.1430504322052, "learning_rate": 6.4e-05, "loss": 0.5281, "step": 6401 }, { "epoch": 0.0006402, "grad_norm": 1.9719579219818115, "learning_rate": 6.401e-05, "loss": 0.6748, "step": 6402 }, { "epoch": 0.0006403, "grad_norm": 1.5823168754577637, "learning_rate": 6.402e-05, "loss": 0.4995, "step": 6403 }, { "epoch": 0.0006404, "grad_norm": 1.2021610736846924, "learning_rate": 6.403e-05, "loss": 0.4573, "step": 6404 }, { "epoch": 0.0006405, "grad_norm": 1.2292171716690063, "learning_rate": 6.404e-05, "loss": 0.4844, "step": 6405 }, { "epoch": 0.0006406, "grad_norm": 1.312039852142334, "learning_rate": 6.405e-05, "loss": 0.4619, "step": 6406 }, { "epoch": 0.0006407, "grad_norm": 1.3954253196716309, "learning_rate": 6.405999999999999e-05, "loss": 0.4961, "step": 6407 }, { "epoch": 0.0006408, "grad_norm": 2.8254330158233643, "learning_rate": 6.407e-05, "loss": 0.8953, "step": 6408 }, { "epoch": 0.0006409, "grad_norm": 1.5670243501663208, "learning_rate": 6.408e-05, "loss": 0.554, "step": 6409 }, { "epoch": 0.000641, "grad_norm": 1.5319308042526245, "learning_rate": 6.409e-05, "loss": 0.4963, "step": 6410 }, { "epoch": 0.0006411, "grad_norm": 1.168323040008545, "learning_rate": 6.41e-05, "loss": 0.4651, "step": 6411 }, { "epoch": 0.0006412, "grad_norm": 1.4604196548461914, "learning_rate": 6.411000000000001e-05, "loss": 0.5515, "step": 6412 }, { "epoch": 0.0006413, "grad_norm": 1.3442469835281372, "learning_rate": 6.412e-05, "loss": 0.51, "step": 6413 }, { "epoch": 0.0006414, "grad_norm": 1.2407495975494385, "learning_rate": 6.413e-05, "loss": 0.4456, "step": 6414 }, { "epoch": 0.0006415, "grad_norm": 1.054099202156067, "learning_rate": 6.414e-05, "loss": 0.4641, "step": 6415 }, { "epoch": 0.0006416, "grad_norm": 1.4334583282470703, "learning_rate": 6.415e-05, "loss": 0.5798, "step": 6416 }, { "epoch": 0.0006417, "grad_norm": 1.3165981769561768, "learning_rate": 6.416e-05, "loss": 0.4995, "step": 6417 }, { "epoch": 0.0006418, "grad_norm": 1.318405270576477, "learning_rate": 6.417e-05, "loss": 0.47, "step": 6418 }, { "epoch": 0.0006419, "grad_norm": 1.5474642515182495, "learning_rate": 6.418e-05, "loss": 0.6069, "step": 6419 }, { "epoch": 0.000642, "grad_norm": 1.4654685258865356, "learning_rate": 6.419000000000001e-05, "loss": 0.467, "step": 6420 }, { "epoch": 0.0006421, "grad_norm": 1.2679288387298584, "learning_rate": 6.42e-05, "loss": 0.4907, "step": 6421 }, { "epoch": 0.0006422, "grad_norm": 1.0179003477096558, "learning_rate": 6.421e-05, "loss": 0.4478, "step": 6422 }, { "epoch": 0.0006423, "grad_norm": 1.70924973487854, "learning_rate": 6.422e-05, "loss": 0.5159, "step": 6423 }, { "epoch": 0.0006424, "grad_norm": 1.2784104347229004, "learning_rate": 6.423e-05, "loss": 0.4739, "step": 6424 }, { "epoch": 0.0006425, "grad_norm": 3.122347831726074, "learning_rate": 6.424e-05, "loss": 0.5659, "step": 6425 }, { "epoch": 0.0006426, "grad_norm": 1.4162781238555908, "learning_rate": 6.425e-05, "loss": 0.5107, "step": 6426 }, { "epoch": 0.0006427, "grad_norm": 1.7903881072998047, "learning_rate": 6.426000000000001e-05, "loss": 0.574, "step": 6427 }, { "epoch": 0.0006428, "grad_norm": 1.3344473838806152, "learning_rate": 6.426999999999999e-05, "loss": 0.4871, "step": 6428 }, { "epoch": 0.0006429, "grad_norm": 1.2928576469421387, "learning_rate": 6.428e-05, "loss": 0.4675, "step": 6429 }, { "epoch": 0.000643, "grad_norm": 1.1451905965805054, "learning_rate": 6.429000000000001e-05, "loss": 0.4661, "step": 6430 }, { "epoch": 0.0006431, "grad_norm": 1.119084358215332, "learning_rate": 6.429999999999999e-05, "loss": 0.4546, "step": 6431 }, { "epoch": 0.0006432, "grad_norm": 1.051802158355713, "learning_rate": 6.431e-05, "loss": 0.4294, "step": 6432 }, { "epoch": 0.0006433, "grad_norm": 2.0395543575286865, "learning_rate": 6.432000000000001e-05, "loss": 0.77, "step": 6433 }, { "epoch": 0.0006434, "grad_norm": 2.3523051738739014, "learning_rate": 6.433e-05, "loss": 0.8062, "step": 6434 }, { "epoch": 0.0006435, "grad_norm": 1.3502811193466187, "learning_rate": 6.434e-05, "loss": 0.4778, "step": 6435 }, { "epoch": 0.0006436, "grad_norm": 1.2692770957946777, "learning_rate": 6.435e-05, "loss": 0.498, "step": 6436 }, { "epoch": 0.0006437, "grad_norm": 2.3005588054656982, "learning_rate": 6.436e-05, "loss": 0.668, "step": 6437 }, { "epoch": 0.0006438, "grad_norm": 2.2760753631591797, "learning_rate": 6.437e-05, "loss": 0.6902, "step": 6438 }, { "epoch": 0.0006439, "grad_norm": 1.4470828771591187, "learning_rate": 6.438e-05, "loss": 0.5142, "step": 6439 }, { "epoch": 0.000644, "grad_norm": 1.0976805686950684, "learning_rate": 6.439e-05, "loss": 0.4333, "step": 6440 }, { "epoch": 0.0006441, "grad_norm": 1.3559536933898926, "learning_rate": 6.44e-05, "loss": 0.5139, "step": 6441 }, { "epoch": 0.0006442, "grad_norm": 1.2253655195236206, "learning_rate": 6.441e-05, "loss": 0.447, "step": 6442 }, { "epoch": 0.0006443, "grad_norm": 1.7331396341323853, "learning_rate": 6.442e-05, "loss": 0.6284, "step": 6443 }, { "epoch": 0.0006444, "grad_norm": 1.949229121208191, "learning_rate": 6.443e-05, "loss": 0.7109, "step": 6444 }, { "epoch": 0.0006445, "grad_norm": 1.2023966312408447, "learning_rate": 6.444e-05, "loss": 0.4602, "step": 6445 }, { "epoch": 0.0006446, "grad_norm": 1.1662184000015259, "learning_rate": 6.445000000000001e-05, "loss": 0.4561, "step": 6446 }, { "epoch": 0.0006447, "grad_norm": 1.2656923532485962, "learning_rate": 6.446e-05, "loss": 0.498, "step": 6447 }, { "epoch": 0.0006448, "grad_norm": 1.1124790906906128, "learning_rate": 6.447e-05, "loss": 0.4751, "step": 6448 }, { "epoch": 0.0006449, "grad_norm": 1.4062652587890625, "learning_rate": 6.448e-05, "loss": 0.5117, "step": 6449 }, { "epoch": 0.000645, "grad_norm": 1.2406326532363892, "learning_rate": 6.449e-05, "loss": 0.4729, "step": 6450 }, { "epoch": 0.0006451, "grad_norm": 1.0493435859680176, "learning_rate": 6.45e-05, "loss": 0.4402, "step": 6451 }, { "epoch": 0.0006452, "grad_norm": 1.0537282228469849, "learning_rate": 6.451e-05, "loss": 0.4534, "step": 6452 }, { "epoch": 0.0006453, "grad_norm": 1.1457775831222534, "learning_rate": 6.452e-05, "loss": 0.4487, "step": 6453 }, { "epoch": 0.0006454, "grad_norm": 1.4109300374984741, "learning_rate": 6.453000000000001e-05, "loss": 0.5303, "step": 6454 }, { "epoch": 0.0006455, "grad_norm": 1.0803672075271606, "learning_rate": 6.454e-05, "loss": 0.4465, "step": 6455 }, { "epoch": 0.0006456, "grad_norm": 1.101141333580017, "learning_rate": 6.455e-05, "loss": 0.4792, "step": 6456 }, { "epoch": 0.0006457, "grad_norm": 2.4582433700561523, "learning_rate": 6.456e-05, "loss": 0.7312, "step": 6457 }, { "epoch": 0.0006458, "grad_norm": 1.2731608152389526, "learning_rate": 6.457e-05, "loss": 0.4822, "step": 6458 }, { "epoch": 0.0006459, "grad_norm": 1.3148726224899292, "learning_rate": 6.458e-05, "loss": 0.4658, "step": 6459 }, { "epoch": 0.000646, "grad_norm": 1.7852345705032349, "learning_rate": 6.459e-05, "loss": 0.5452, "step": 6460 }, { "epoch": 0.0006461, "grad_norm": 1.274753212928772, "learning_rate": 6.460000000000001e-05, "loss": 0.5164, "step": 6461 }, { "epoch": 0.0006462, "grad_norm": 1.173985481262207, "learning_rate": 6.460999999999999e-05, "loss": 0.4419, "step": 6462 }, { "epoch": 0.0006463, "grad_norm": 1.1694120168685913, "learning_rate": 6.462e-05, "loss": 0.4819, "step": 6463 }, { "epoch": 0.0006464, "grad_norm": 1.2963957786560059, "learning_rate": 6.463000000000001e-05, "loss": 0.4668, "step": 6464 }, { "epoch": 0.0006465, "grad_norm": 1.2287315130233765, "learning_rate": 6.463999999999999e-05, "loss": 0.4695, "step": 6465 }, { "epoch": 0.0006466, "grad_norm": 1.3719747066497803, "learning_rate": 6.465e-05, "loss": 0.5833, "step": 6466 }, { "epoch": 0.0006467, "grad_norm": 2.0178167819976807, "learning_rate": 6.466000000000001e-05, "loss": 0.5288, "step": 6467 }, { "epoch": 0.0006468, "grad_norm": 1.0148391723632812, "learning_rate": 6.467e-05, "loss": 0.4104, "step": 6468 }, { "epoch": 0.0006469, "grad_norm": 1.1594983339309692, "learning_rate": 6.468e-05, "loss": 0.4709, "step": 6469 }, { "epoch": 0.000647, "grad_norm": 1.2085261344909668, "learning_rate": 6.469e-05, "loss": 0.5552, "step": 6470 }, { "epoch": 0.0006471, "grad_norm": 1.0414901971817017, "learning_rate": 6.47e-05, "loss": 0.4329, "step": 6471 }, { "epoch": 0.0006472, "grad_norm": 1.1136866807937622, "learning_rate": 6.471e-05, "loss": 0.4314, "step": 6472 }, { "epoch": 0.0006473, "grad_norm": 1.0286751985549927, "learning_rate": 6.472e-05, "loss": 0.4463, "step": 6473 }, { "epoch": 0.0006474, "grad_norm": 1.04264235496521, "learning_rate": 6.473e-05, "loss": 0.4739, "step": 6474 }, { "epoch": 0.0006475, "grad_norm": 1.2967795133590698, "learning_rate": 6.474e-05, "loss": 0.4812, "step": 6475 }, { "epoch": 0.0006476, "grad_norm": 1.0789908170700073, "learning_rate": 6.475e-05, "loss": 0.4585, "step": 6476 }, { "epoch": 0.0006477, "grad_norm": 1.106982707977295, "learning_rate": 6.476e-05, "loss": 0.4736, "step": 6477 }, { "epoch": 0.0006478, "grad_norm": 1.0327823162078857, "learning_rate": 6.477e-05, "loss": 0.4155, "step": 6478 }, { "epoch": 0.0006479, "grad_norm": 1.0931192636489868, "learning_rate": 6.478e-05, "loss": 0.4358, "step": 6479 }, { "epoch": 0.000648, "grad_norm": 1.0536428689956665, "learning_rate": 6.479000000000001e-05, "loss": 0.4285, "step": 6480 }, { "epoch": 0.0006481, "grad_norm": 1.1148128509521484, "learning_rate": 6.48e-05, "loss": 0.4421, "step": 6481 }, { "epoch": 0.0006482, "grad_norm": 0.9446101784706116, "learning_rate": 6.481e-05, "loss": 0.3987, "step": 6482 }, { "epoch": 0.0006483, "grad_norm": 1.0021119117736816, "learning_rate": 6.482e-05, "loss": 0.4248, "step": 6483 }, { "epoch": 0.0006484, "grad_norm": 0.9024091958999634, "learning_rate": 6.483e-05, "loss": 0.4197, "step": 6484 }, { "epoch": 0.0006485, "grad_norm": 2.6724512577056885, "learning_rate": 6.484e-05, "loss": 0.7097, "step": 6485 }, { "epoch": 0.0006486, "grad_norm": 2.107272148132324, "learning_rate": 6.485e-05, "loss": 0.656, "step": 6486 }, { "epoch": 0.0006487, "grad_norm": 1.1949630975723267, "learning_rate": 6.486e-05, "loss": 0.4294, "step": 6487 }, { "epoch": 0.0006488, "grad_norm": 1.190617322921753, "learning_rate": 6.487000000000001e-05, "loss": 0.4436, "step": 6488 }, { "epoch": 0.0006489, "grad_norm": 0.9928703904151917, "learning_rate": 6.488e-05, "loss": 0.4324, "step": 6489 }, { "epoch": 0.000649, "grad_norm": 3.6930248737335205, "learning_rate": 6.489e-05, "loss": 0.7405, "step": 6490 }, { "epoch": 0.0006491, "grad_norm": 1.3998894691467285, "learning_rate": 6.49e-05, "loss": 0.4187, "step": 6491 }, { "epoch": 0.0006492, "grad_norm": 4.68524694442749, "learning_rate": 6.491e-05, "loss": 1.042, "step": 6492 }, { "epoch": 0.0006493, "grad_norm": 1.2329721450805664, "learning_rate": 6.492e-05, "loss": 0.407, "step": 6493 }, { "epoch": 0.0006494, "grad_norm": 1.0657070875167847, "learning_rate": 6.493e-05, "loss": 0.4189, "step": 6494 }, { "epoch": 0.0006495, "grad_norm": 1.1602028608322144, "learning_rate": 6.494000000000001e-05, "loss": 0.4355, "step": 6495 }, { "epoch": 0.0006496, "grad_norm": 1.4874273538589478, "learning_rate": 6.494999999999999e-05, "loss": 0.4944, "step": 6496 }, { "epoch": 0.0006497, "grad_norm": 4.198200702667236, "learning_rate": 6.496e-05, "loss": 0.5972, "step": 6497 }, { "epoch": 0.0006498, "grad_norm": 1.6802000999450684, "learning_rate": 6.497000000000001e-05, "loss": 0.4402, "step": 6498 }, { "epoch": 0.0006499, "grad_norm": 1.3226237297058105, "learning_rate": 6.497999999999999e-05, "loss": 0.4204, "step": 6499 }, { "epoch": 0.00065, "grad_norm": 1.4562503099441528, "learning_rate": 6.499e-05, "loss": 0.448, "step": 6500 }, { "epoch": 0.00065, "eval_loss": 0.04971218481659889, "eval_runtime": 364.9703, "eval_samples_per_second": 27.399, "eval_steps_per_second": 1.712, "step": 6500 }, { "epoch": 0.0006501, "grad_norm": 1.03931725025177, "learning_rate": 6.500000000000001e-05, "loss": 0.3943, "step": 6501 }, { "epoch": 0.0006502, "grad_norm": 1.1277257204055786, "learning_rate": 6.501e-05, "loss": 0.4182, "step": 6502 }, { "epoch": 0.0006503, "grad_norm": 1.1667401790618896, "learning_rate": 6.502e-05, "loss": 0.4065, "step": 6503 }, { "epoch": 0.0006504, "grad_norm": 1.1649590730667114, "learning_rate": 6.503e-05, "loss": 0.4451, "step": 6504 }, { "epoch": 0.0006505, "grad_norm": 1.085089087486267, "learning_rate": 6.504e-05, "loss": 0.4204, "step": 6505 }, { "epoch": 0.0006506, "grad_norm": 0.9987074136734009, "learning_rate": 6.505e-05, "loss": 0.4111, "step": 6506 }, { "epoch": 0.0006507, "grad_norm": 1.0891286134719849, "learning_rate": 6.506e-05, "loss": 0.4248, "step": 6507 }, { "epoch": 0.0006508, "grad_norm": 0.9831911325454712, "learning_rate": 6.507e-05, "loss": 0.3943, "step": 6508 }, { "epoch": 0.0006509, "grad_norm": 0.9570473432540894, "learning_rate": 6.508e-05, "loss": 0.385, "step": 6509 }, { "epoch": 0.000651, "grad_norm": 1.354292869567871, "learning_rate": 6.509e-05, "loss": 0.5371, "step": 6510 }, { "epoch": 0.0006511, "grad_norm": 1.0083353519439697, "learning_rate": 6.51e-05, "loss": 0.4094, "step": 6511 }, { "epoch": 0.0006512, "grad_norm": 0.9337421655654907, "learning_rate": 6.511e-05, "loss": 0.4092, "step": 6512 }, { "epoch": 0.0006513, "grad_norm": 0.9101032018661499, "learning_rate": 6.512e-05, "loss": 0.387, "step": 6513 }, { "epoch": 0.0006514, "grad_norm": 0.994778037071228, "learning_rate": 6.513000000000001e-05, "loss": 0.4221, "step": 6514 }, { "epoch": 0.0006515, "grad_norm": 0.8868163228034973, "learning_rate": 6.514e-05, "loss": 0.3887, "step": 6515 }, { "epoch": 0.0006516, "grad_norm": 1.0790033340454102, "learning_rate": 6.515e-05, "loss": 0.4199, "step": 6516 }, { "epoch": 0.0006517, "grad_norm": 0.9513962268829346, "learning_rate": 6.516e-05, "loss": 0.4036, "step": 6517 }, { "epoch": 0.0006518, "grad_norm": 0.966842532157898, "learning_rate": 6.517e-05, "loss": 0.4221, "step": 6518 }, { "epoch": 0.0006519, "grad_norm": 0.8695230484008789, "learning_rate": 6.518e-05, "loss": 0.3608, "step": 6519 }, { "epoch": 0.000652, "grad_norm": 0.9728794693946838, "learning_rate": 6.519e-05, "loss": 0.4062, "step": 6520 }, { "epoch": 0.0006521, "grad_norm": 1.1127318143844604, "learning_rate": 6.52e-05, "loss": 0.4353, "step": 6521 }, { "epoch": 0.0006522, "grad_norm": 0.9356803297996521, "learning_rate": 6.521000000000001e-05, "loss": 0.3857, "step": 6522 }, { "epoch": 0.0006523, "grad_norm": 0.9150017499923706, "learning_rate": 6.522e-05, "loss": 0.3821, "step": 6523 }, { "epoch": 0.0006524, "grad_norm": 1.0721808671951294, "learning_rate": 6.523e-05, "loss": 0.4216, "step": 6524 }, { "epoch": 0.0006525, "grad_norm": 0.9526981711387634, "learning_rate": 6.524e-05, "loss": 0.3779, "step": 6525 }, { "epoch": 0.0006526, "grad_norm": 1.0103404521942139, "learning_rate": 6.525e-05, "loss": 0.4048, "step": 6526 }, { "epoch": 0.0006527, "grad_norm": 0.9327437877655029, "learning_rate": 6.526e-05, "loss": 0.395, "step": 6527 }, { "epoch": 0.0006528, "grad_norm": 0.8834521174430847, "learning_rate": 6.527e-05, "loss": 0.3818, "step": 6528 }, { "epoch": 0.0006529, "grad_norm": 0.8776892423629761, "learning_rate": 6.528000000000001e-05, "loss": 0.3735, "step": 6529 }, { "epoch": 0.000653, "grad_norm": 0.9939847588539124, "learning_rate": 6.528999999999999e-05, "loss": 0.4111, "step": 6530 }, { "epoch": 0.0006531, "grad_norm": 0.998724639415741, "learning_rate": 6.53e-05, "loss": 0.4087, "step": 6531 }, { "epoch": 0.0006532, "grad_norm": 0.9717676639556885, "learning_rate": 6.531000000000001e-05, "loss": 0.3853, "step": 6532 }, { "epoch": 0.0006533, "grad_norm": 1.451299786567688, "learning_rate": 6.531999999999999e-05, "loss": 0.5186, "step": 6533 }, { "epoch": 0.0006534, "grad_norm": 0.95253586769104, "learning_rate": 6.533e-05, "loss": 0.3835, "step": 6534 }, { "epoch": 0.0006535, "grad_norm": 0.8519267439842224, "learning_rate": 6.534000000000001e-05, "loss": 0.344, "step": 6535 }, { "epoch": 0.0006536, "grad_norm": 1.0273712873458862, "learning_rate": 6.535e-05, "loss": 0.4075, "step": 6536 }, { "epoch": 0.0006537, "grad_norm": 1.1981201171875, "learning_rate": 6.536e-05, "loss": 0.4307, "step": 6537 }, { "epoch": 0.0006538, "grad_norm": 1.0395232439041138, "learning_rate": 6.537e-05, "loss": 0.3965, "step": 6538 }, { "epoch": 0.0006539, "grad_norm": 1.3438701629638672, "learning_rate": 6.538e-05, "loss": 0.405, "step": 6539 }, { "epoch": 0.000654, "grad_norm": 0.9981584548950195, "learning_rate": 6.539e-05, "loss": 0.3733, "step": 6540 }, { "epoch": 0.0006541, "grad_norm": 0.9692195057868958, "learning_rate": 6.54e-05, "loss": 0.3872, "step": 6541 }, { "epoch": 0.0006542, "grad_norm": 1.4983618259429932, "learning_rate": 6.541e-05, "loss": 0.5432, "step": 6542 }, { "epoch": 0.0006543, "grad_norm": 1.5753177404403687, "learning_rate": 6.542e-05, "loss": 0.4004, "step": 6543 }, { "epoch": 0.0006544, "grad_norm": 1.2118226289749146, "learning_rate": 6.543e-05, "loss": 0.3994, "step": 6544 }, { "epoch": 0.0006545, "grad_norm": 0.9601621627807617, "learning_rate": 6.544e-05, "loss": 0.3821, "step": 6545 }, { "epoch": 0.0006546, "grad_norm": 0.9599550366401672, "learning_rate": 6.545e-05, "loss": 0.3679, "step": 6546 }, { "epoch": 0.0006547, "grad_norm": 2.707258939743042, "learning_rate": 6.546e-05, "loss": 0.4639, "step": 6547 }, { "epoch": 0.0006548, "grad_norm": 2.5643324851989746, "learning_rate": 6.547000000000001e-05, "loss": 0.5105, "step": 6548 }, { "epoch": 0.0006549, "grad_norm": 1.3906834125518799, "learning_rate": 6.548e-05, "loss": 0.427, "step": 6549 }, { "epoch": 0.000655, "grad_norm": 1.1361576318740845, "learning_rate": 6.549e-05, "loss": 0.3708, "step": 6550 }, { "epoch": 0.0006551, "grad_norm": 0.9891266822814941, "learning_rate": 6.55e-05, "loss": 0.3682, "step": 6551 }, { "epoch": 0.0006552, "grad_norm": 1.2472726106643677, "learning_rate": 6.551e-05, "loss": 0.4463, "step": 6552 }, { "epoch": 0.0006553, "grad_norm": 1.0641355514526367, "learning_rate": 6.552e-05, "loss": 0.3435, "step": 6553 }, { "epoch": 0.0006554, "grad_norm": 1.7209123373031616, "learning_rate": 6.553e-05, "loss": 0.499, "step": 6554 }, { "epoch": 0.0006555, "grad_norm": 0.9860501289367676, "learning_rate": 6.554e-05, "loss": 0.3564, "step": 6555 }, { "epoch": 0.0006556, "grad_norm": 1.7217719554901123, "learning_rate": 6.555000000000001e-05, "loss": 0.4077, "step": 6556 }, { "epoch": 0.0006557, "grad_norm": 0.9856640696525574, "learning_rate": 6.556e-05, "loss": 0.3628, "step": 6557 }, { "epoch": 0.0006558, "grad_norm": 1.0517157316207886, "learning_rate": 6.557e-05, "loss": 0.3804, "step": 6558 }, { "epoch": 0.0006559, "grad_norm": 1.0080536603927612, "learning_rate": 6.558e-05, "loss": 0.3699, "step": 6559 }, { "epoch": 0.000656, "grad_norm": 3.1549456119537354, "learning_rate": 6.559e-05, "loss": 0.4944, "step": 6560 }, { "epoch": 0.0006561, "grad_norm": 1.1440428495407104, "learning_rate": 6.56e-05, "loss": 0.3672, "step": 6561 }, { "epoch": 0.0006562, "grad_norm": 1.6338598728179932, "learning_rate": 6.561e-05, "loss": 0.4395, "step": 6562 }, { "epoch": 0.0006563, "grad_norm": 1.1482481956481934, "learning_rate": 6.562000000000001e-05, "loss": 0.3816, "step": 6563 }, { "epoch": 0.0006564, "grad_norm": 0.991045355796814, "learning_rate": 6.563e-05, "loss": 0.3801, "step": 6564 }, { "epoch": 0.0006565, "grad_norm": 1.1286438703536987, "learning_rate": 6.564e-05, "loss": 0.3643, "step": 6565 }, { "epoch": 0.0006566, "grad_norm": 0.9734741449356079, "learning_rate": 6.565000000000001e-05, "loss": 0.3469, "step": 6566 }, { "epoch": 0.0006567, "grad_norm": 1.013515830039978, "learning_rate": 6.565999999999999e-05, "loss": 0.374, "step": 6567 }, { "epoch": 0.0006568, "grad_norm": 4.014554023742676, "learning_rate": 6.567e-05, "loss": 0.6902, "step": 6568 }, { "epoch": 0.0006569, "grad_norm": 1.896406888961792, "learning_rate": 6.568000000000001e-05, "loss": 0.3804, "step": 6569 }, { "epoch": 0.000657, "grad_norm": 1.2844294309616089, "learning_rate": 6.569e-05, "loss": 0.3657, "step": 6570 }, { "epoch": 0.0006571, "grad_norm": 1.3451746702194214, "learning_rate": 6.57e-05, "loss": 0.3879, "step": 6571 }, { "epoch": 0.0006572, "grad_norm": 0.845024824142456, "learning_rate": 6.571e-05, "loss": 0.3506, "step": 6572 }, { "epoch": 0.0006573, "grad_norm": 1.3707619905471802, "learning_rate": 6.572e-05, "loss": 0.4194, "step": 6573 }, { "epoch": 0.0006574, "grad_norm": 1.108884334564209, "learning_rate": 6.573e-05, "loss": 0.3682, "step": 6574 }, { "epoch": 0.0006575, "grad_norm": 1.4851621389389038, "learning_rate": 6.574e-05, "loss": 0.396, "step": 6575 }, { "epoch": 0.0006576, "grad_norm": 0.9890068769454956, "learning_rate": 6.575e-05, "loss": 0.3594, "step": 6576 }, { "epoch": 0.0006577, "grad_norm": 0.9809015989303589, "learning_rate": 6.576e-05, "loss": 0.3511, "step": 6577 }, { "epoch": 0.0006578, "grad_norm": 0.9301626682281494, "learning_rate": 6.577e-05, "loss": 0.3523, "step": 6578 }, { "epoch": 0.0006579, "grad_norm": 0.8530791997909546, "learning_rate": 6.578e-05, "loss": 0.3623, "step": 6579 }, { "epoch": 0.000658, "grad_norm": 0.7916417121887207, "learning_rate": 6.579e-05, "loss": 0.3401, "step": 6580 }, { "epoch": 0.0006581, "grad_norm": 0.9918129444122314, "learning_rate": 6.58e-05, "loss": 0.3806, "step": 6581 }, { "epoch": 0.0006582, "grad_norm": 0.9188030958175659, "learning_rate": 6.581000000000001e-05, "loss": 0.3701, "step": 6582 }, { "epoch": 0.0006583, "grad_norm": 0.993864893913269, "learning_rate": 6.582e-05, "loss": 0.4053, "step": 6583 }, { "epoch": 0.0006584, "grad_norm": 0.8649798631668091, "learning_rate": 6.583e-05, "loss": 0.3525, "step": 6584 }, { "epoch": 0.0006585, "grad_norm": 0.9735991358757019, "learning_rate": 6.584e-05, "loss": 0.3691, "step": 6585 }, { "epoch": 0.0006586, "grad_norm": 1.0432755947113037, "learning_rate": 6.585e-05, "loss": 0.3967, "step": 6586 }, { "epoch": 0.0006587, "grad_norm": 1.1535252332687378, "learning_rate": 6.586e-05, "loss": 0.3975, "step": 6587 }, { "epoch": 0.0006588, "grad_norm": 0.8514976501464844, "learning_rate": 6.587e-05, "loss": 0.3455, "step": 6588 }, { "epoch": 0.0006589, "grad_norm": 0.8397539258003235, "learning_rate": 6.588e-05, "loss": 0.3289, "step": 6589 }, { "epoch": 0.000659, "grad_norm": 0.8472413420677185, "learning_rate": 6.589000000000001e-05, "loss": 0.3477, "step": 6590 }, { "epoch": 0.0006591, "grad_norm": 0.8127468824386597, "learning_rate": 6.59e-05, "loss": 0.3496, "step": 6591 }, { "epoch": 0.0006592, "grad_norm": 0.8434761166572571, "learning_rate": 6.591e-05, "loss": 0.3477, "step": 6592 }, { "epoch": 0.0006593, "grad_norm": 0.9362355470657349, "learning_rate": 6.592e-05, "loss": 0.3794, "step": 6593 }, { "epoch": 0.0006594, "grad_norm": 1.5990657806396484, "learning_rate": 6.593e-05, "loss": 0.5615, "step": 6594 }, { "epoch": 0.0006595, "grad_norm": 0.7806490063667297, "learning_rate": 6.594e-05, "loss": 0.3267, "step": 6595 }, { "epoch": 0.0006596, "grad_norm": 0.9472602009773254, "learning_rate": 6.595e-05, "loss": 0.3694, "step": 6596 }, { "epoch": 0.0006597, "grad_norm": 1.820241928100586, "learning_rate": 6.596000000000001e-05, "loss": 0.3623, "step": 6597 }, { "epoch": 0.0006598, "grad_norm": 1.3111149072647095, "learning_rate": 6.597e-05, "loss": 0.4211, "step": 6598 }, { "epoch": 0.0006599, "grad_norm": 1.7332957983016968, "learning_rate": 6.598e-05, "loss": 0.5574, "step": 6599 }, { "epoch": 0.00066, "grad_norm": 1.0712653398513794, "learning_rate": 6.599000000000001e-05, "loss": 0.3293, "step": 6600 }, { "epoch": 0.0006601, "grad_norm": 6.920145034790039, "learning_rate": 6.599999999999999e-05, "loss": 0.6079, "step": 6601 }, { "epoch": 0.0006602, "grad_norm": 1.0699074268341064, "learning_rate": 6.601e-05, "loss": 0.3572, "step": 6602 }, { "epoch": 0.0006603, "grad_norm": 0.9052705764770508, "learning_rate": 6.602000000000001e-05, "loss": 0.3411, "step": 6603 }, { "epoch": 0.0006604, "grad_norm": 0.9185029864311218, "learning_rate": 6.603e-05, "loss": 0.3469, "step": 6604 }, { "epoch": 0.0006605, "grad_norm": 0.8422958254814148, "learning_rate": 6.604e-05, "loss": 0.3257, "step": 6605 }, { "epoch": 0.0006606, "grad_norm": 0.9281983971595764, "learning_rate": 6.605e-05, "loss": 0.3499, "step": 6606 }, { "epoch": 0.0006607, "grad_norm": 0.8922631144523621, "learning_rate": 6.606e-05, "loss": 0.3479, "step": 6607 }, { "epoch": 0.0006608, "grad_norm": 0.8519024848937988, "learning_rate": 6.607e-05, "loss": 0.3303, "step": 6608 }, { "epoch": 0.0006609, "grad_norm": 1.4792611598968506, "learning_rate": 6.608e-05, "loss": 0.408, "step": 6609 }, { "epoch": 0.000661, "grad_norm": 0.9307885766029358, "learning_rate": 6.609e-05, "loss": 0.3396, "step": 6610 }, { "epoch": 0.0006611, "grad_norm": 0.9870061278343201, "learning_rate": 6.61e-05, "loss": 0.3652, "step": 6611 }, { "epoch": 0.0006612, "grad_norm": 0.9354746341705322, "learning_rate": 6.611e-05, "loss": 0.3447, "step": 6612 }, { "epoch": 0.0006613, "grad_norm": 4.36238956451416, "learning_rate": 6.612e-05, "loss": 0.6912, "step": 6613 }, { "epoch": 0.0006614, "grad_norm": 1.068058729171753, "learning_rate": 6.613e-05, "loss": 0.356, "step": 6614 }, { "epoch": 0.0006615, "grad_norm": 0.9313966035842896, "learning_rate": 6.614e-05, "loss": 0.3164, "step": 6615 }, { "epoch": 0.0006616, "grad_norm": 0.8230025172233582, "learning_rate": 6.615e-05, "loss": 0.3357, "step": 6616 }, { "epoch": 0.0006617, "grad_norm": 0.7870452404022217, "learning_rate": 6.616e-05, "loss": 0.3269, "step": 6617 }, { "epoch": 0.0006618, "grad_norm": 0.9555226564407349, "learning_rate": 6.617e-05, "loss": 0.3372, "step": 6618 }, { "epoch": 0.0006619, "grad_norm": 0.7982156872749329, "learning_rate": 6.618e-05, "loss": 0.3091, "step": 6619 }, { "epoch": 0.000662, "grad_norm": 0.8481671810150146, "learning_rate": 6.619e-05, "loss": 0.3271, "step": 6620 }, { "epoch": 0.0006621, "grad_norm": 1.6612424850463867, "learning_rate": 6.62e-05, "loss": 0.4949, "step": 6621 }, { "epoch": 0.0006622, "grad_norm": 0.8787587881088257, "learning_rate": 6.621e-05, "loss": 0.3311, "step": 6622 }, { "epoch": 0.0006623, "grad_norm": 0.8238551020622253, "learning_rate": 6.622e-05, "loss": 0.3291, "step": 6623 }, { "epoch": 0.0006624, "grad_norm": 0.8153005838394165, "learning_rate": 6.623000000000001e-05, "loss": 0.3267, "step": 6624 }, { "epoch": 0.0006625, "grad_norm": 1.2538508176803589, "learning_rate": 6.624e-05, "loss": 0.3623, "step": 6625 }, { "epoch": 0.0006626, "grad_norm": 0.8147600293159485, "learning_rate": 6.625e-05, "loss": 0.3313, "step": 6626 }, { "epoch": 0.0006627, "grad_norm": 0.9060648679733276, "learning_rate": 6.626e-05, "loss": 0.353, "step": 6627 }, { "epoch": 0.0006628, "grad_norm": 0.81404048204422, "learning_rate": 6.627e-05, "loss": 0.3438, "step": 6628 }, { "epoch": 0.0006629, "grad_norm": 0.8007271885871887, "learning_rate": 6.628e-05, "loss": 0.333, "step": 6629 }, { "epoch": 0.000663, "grad_norm": 0.7897888422012329, "learning_rate": 6.629e-05, "loss": 0.3291, "step": 6630 }, { "epoch": 0.0006631, "grad_norm": 0.7997754812240601, "learning_rate": 6.630000000000001e-05, "loss": 0.3167, "step": 6631 }, { "epoch": 0.0006632, "grad_norm": 0.7762719988822937, "learning_rate": 6.631e-05, "loss": 0.3237, "step": 6632 }, { "epoch": 0.0006633, "grad_norm": 0.7300854325294495, "learning_rate": 6.632e-05, "loss": 0.3445, "step": 6633 }, { "epoch": 0.0006634, "grad_norm": 3.1596171855926514, "learning_rate": 6.633000000000001e-05, "loss": 0.6033, "step": 6634 }, { "epoch": 0.0006635, "grad_norm": 0.7889407873153687, "learning_rate": 6.633999999999999e-05, "loss": 0.3152, "step": 6635 }, { "epoch": 0.0006636, "grad_norm": 0.7632907629013062, "learning_rate": 6.635e-05, "loss": 0.3101, "step": 6636 }, { "epoch": 0.0006637, "grad_norm": 0.8322415947914124, "learning_rate": 6.636000000000001e-05, "loss": 0.324, "step": 6637 }, { "epoch": 0.0006638, "grad_norm": 0.6845356822013855, "learning_rate": 6.636999999999999e-05, "loss": 0.3118, "step": 6638 }, { "epoch": 0.0006639, "grad_norm": 1.0866153240203857, "learning_rate": 6.638e-05, "loss": 0.3462, "step": 6639 }, { "epoch": 0.000664, "grad_norm": 0.6863608956336975, "learning_rate": 6.639e-05, "loss": 0.3103, "step": 6640 }, { "epoch": 0.0006641, "grad_norm": 0.7848332524299622, "learning_rate": 6.64e-05, "loss": 0.3274, "step": 6641 }, { "epoch": 0.0006642, "grad_norm": 1.3172481060028076, "learning_rate": 6.641e-05, "loss": 0.396, "step": 6642 }, { "epoch": 0.0006643, "grad_norm": 0.7006875276565552, "learning_rate": 6.642e-05, "loss": 0.3181, "step": 6643 }, { "epoch": 0.0006644, "grad_norm": 0.8330028653144836, "learning_rate": 6.643e-05, "loss": 0.3379, "step": 6644 }, { "epoch": 0.0006645, "grad_norm": 0.6954527497291565, "learning_rate": 6.644e-05, "loss": 0.302, "step": 6645 }, { "epoch": 0.0006646, "grad_norm": 0.8154837489128113, "learning_rate": 6.645e-05, "loss": 0.3411, "step": 6646 }, { "epoch": 0.0006647, "grad_norm": 0.6617774367332458, "learning_rate": 6.646e-05, "loss": 0.3037, "step": 6647 }, { "epoch": 0.0006648, "grad_norm": 0.7139044404029846, "learning_rate": 6.647e-05, "loss": 0.3293, "step": 6648 }, { "epoch": 0.0006649, "grad_norm": 0.693368673324585, "learning_rate": 6.648e-05, "loss": 0.3127, "step": 6649 }, { "epoch": 0.000665, "grad_norm": 0.7504817843437195, "learning_rate": 6.649e-05, "loss": 0.3447, "step": 6650 }, { "epoch": 0.0006651, "grad_norm": 0.7309079766273499, "learning_rate": 6.65e-05, "loss": 0.3193, "step": 6651 }, { "epoch": 0.0006652, "grad_norm": 1.2137302160263062, "learning_rate": 6.651e-05, "loss": 0.4045, "step": 6652 }, { "epoch": 0.0006653, "grad_norm": 0.7493794560432434, "learning_rate": 6.652000000000001e-05, "loss": 0.322, "step": 6653 }, { "epoch": 0.0006654, "grad_norm": 0.8475434184074402, "learning_rate": 6.653e-05, "loss": 0.3335, "step": 6654 }, { "epoch": 0.0006655, "grad_norm": 0.7326182723045349, "learning_rate": 6.654e-05, "loss": 0.314, "step": 6655 }, { "epoch": 0.0006656, "grad_norm": 1.0516753196716309, "learning_rate": 6.655e-05, "loss": 0.3462, "step": 6656 }, { "epoch": 0.0006657, "grad_norm": 0.7796412706375122, "learning_rate": 6.656e-05, "loss": 0.3206, "step": 6657 }, { "epoch": 0.0006658, "grad_norm": 0.7046060562133789, "learning_rate": 6.657000000000001e-05, "loss": 0.304, "step": 6658 }, { "epoch": 0.0006659, "grad_norm": 0.7619063258171082, "learning_rate": 6.658e-05, "loss": 0.3269, "step": 6659 }, { "epoch": 0.000666, "grad_norm": 0.7655344009399414, "learning_rate": 6.659e-05, "loss": 0.3435, "step": 6660 }, { "epoch": 0.0006661, "grad_norm": 0.7178213000297546, "learning_rate": 6.66e-05, "loss": 0.3218, "step": 6661 }, { "epoch": 0.0006662, "grad_norm": 0.80521160364151, "learning_rate": 6.661e-05, "loss": 0.3391, "step": 6662 }, { "epoch": 0.0006663, "grad_norm": 0.7512543201446533, "learning_rate": 6.662e-05, "loss": 0.3042, "step": 6663 }, { "epoch": 0.0006664, "grad_norm": 0.7814407348632812, "learning_rate": 6.663e-05, "loss": 0.332, "step": 6664 }, { "epoch": 0.0006665, "grad_norm": 0.7427365183830261, "learning_rate": 6.664000000000001e-05, "loss": 0.323, "step": 6665 }, { "epoch": 0.0006666, "grad_norm": 0.6811539530754089, "learning_rate": 6.665e-05, "loss": 0.3052, "step": 6666 }, { "epoch": 0.0006667, "grad_norm": 0.7411268353462219, "learning_rate": 6.666e-05, "loss": 0.3005, "step": 6667 }, { "epoch": 0.0006668, "grad_norm": 0.7068269848823547, "learning_rate": 6.667000000000001e-05, "loss": 0.3284, "step": 6668 }, { "epoch": 0.0006669, "grad_norm": 0.7144364714622498, "learning_rate": 6.667999999999999e-05, "loss": 0.3391, "step": 6669 }, { "epoch": 0.000667, "grad_norm": 0.6556010246276855, "learning_rate": 6.669e-05, "loss": 0.3145, "step": 6670 }, { "epoch": 0.0006671, "grad_norm": 1.214873194694519, "learning_rate": 6.670000000000001e-05, "loss": 0.4189, "step": 6671 }, { "epoch": 0.0006672, "grad_norm": 0.7603350281715393, "learning_rate": 6.670999999999999e-05, "loss": 0.3342, "step": 6672 }, { "epoch": 0.0006673, "grad_norm": 0.7102590799331665, "learning_rate": 6.672e-05, "loss": 0.3152, "step": 6673 }, { "epoch": 0.0006674, "grad_norm": 0.7383694648742676, "learning_rate": 6.673e-05, "loss": 0.3435, "step": 6674 }, { "epoch": 0.0006675, "grad_norm": 0.6528864502906799, "learning_rate": 6.674e-05, "loss": 0.3225, "step": 6675 }, { "epoch": 0.0006676, "grad_norm": 0.8522922992706299, "learning_rate": 6.675e-05, "loss": 0.3503, "step": 6676 }, { "epoch": 0.0006677, "grad_norm": 0.7166839241981506, "learning_rate": 6.676e-05, "loss": 0.3264, "step": 6677 }, { "epoch": 0.0006678, "grad_norm": 0.7173413634300232, "learning_rate": 6.677e-05, "loss": 0.3152, "step": 6678 }, { "epoch": 0.0006679, "grad_norm": 0.7953842878341675, "learning_rate": 6.678e-05, "loss": 0.3286, "step": 6679 }, { "epoch": 0.000668, "grad_norm": 0.7262922525405884, "learning_rate": 6.679e-05, "loss": 0.3206, "step": 6680 }, { "epoch": 0.0006681, "grad_norm": 0.6505809426307678, "learning_rate": 6.68e-05, "loss": 0.3037, "step": 6681 }, { "epoch": 0.0006682, "grad_norm": 0.8676546812057495, "learning_rate": 6.681e-05, "loss": 0.3711, "step": 6682 }, { "epoch": 0.0006683, "grad_norm": 0.638432502746582, "learning_rate": 6.682e-05, "loss": 0.2893, "step": 6683 }, { "epoch": 0.0006684, "grad_norm": 0.8237349390983582, "learning_rate": 6.683e-05, "loss": 0.3335, "step": 6684 }, { "epoch": 0.0006685, "grad_norm": 0.727479100227356, "learning_rate": 6.684e-05, "loss": 0.3264, "step": 6685 }, { "epoch": 0.0006686, "grad_norm": 0.6328103542327881, "learning_rate": 6.685e-05, "loss": 0.3086, "step": 6686 }, { "epoch": 0.0006687, "grad_norm": 0.7690942883491516, "learning_rate": 6.686000000000001e-05, "loss": 0.3218, "step": 6687 }, { "epoch": 0.0006688, "grad_norm": 0.69964200258255, "learning_rate": 6.687e-05, "loss": 0.3259, "step": 6688 }, { "epoch": 0.0006689, "grad_norm": 0.6972114443778992, "learning_rate": 6.688e-05, "loss": 0.3328, "step": 6689 }, { "epoch": 0.000669, "grad_norm": 0.697297215461731, "learning_rate": 6.689e-05, "loss": 0.3267, "step": 6690 }, { "epoch": 0.0006691, "grad_norm": 1.2948123216629028, "learning_rate": 6.69e-05, "loss": 0.4321, "step": 6691 }, { "epoch": 0.0006692, "grad_norm": 0.7972717881202698, "learning_rate": 6.691000000000001e-05, "loss": 0.3184, "step": 6692 }, { "epoch": 0.0006693, "grad_norm": 0.777863085269928, "learning_rate": 6.692e-05, "loss": 0.3042, "step": 6693 }, { "epoch": 0.0006694, "grad_norm": 0.952314555644989, "learning_rate": 6.693e-05, "loss": 0.3782, "step": 6694 }, { "epoch": 0.0006695, "grad_norm": 0.667868435382843, "learning_rate": 6.694e-05, "loss": 0.2976, "step": 6695 }, { "epoch": 0.0006696, "grad_norm": 0.7156415581703186, "learning_rate": 6.695e-05, "loss": 0.3081, "step": 6696 }, { "epoch": 0.0006697, "grad_norm": 0.6847723126411438, "learning_rate": 6.696e-05, "loss": 0.3027, "step": 6697 }, { "epoch": 0.0006698, "grad_norm": 0.6780564785003662, "learning_rate": 6.697e-05, "loss": 0.3074, "step": 6698 }, { "epoch": 0.0006699, "grad_norm": 0.6628689765930176, "learning_rate": 6.698000000000001e-05, "loss": 0.3027, "step": 6699 }, { "epoch": 0.00067, "grad_norm": 0.7941015958786011, "learning_rate": 6.699e-05, "loss": 0.3264, "step": 6700 }, { "epoch": 0.0006701, "grad_norm": 0.8983973860740662, "learning_rate": 6.7e-05, "loss": 0.3368, "step": 6701 }, { "epoch": 0.0006702, "grad_norm": 0.8160014748573303, "learning_rate": 6.701000000000001e-05, "loss": 0.3267, "step": 6702 }, { "epoch": 0.0006703, "grad_norm": 1.5155754089355469, "learning_rate": 6.701999999999999e-05, "loss": 0.4702, "step": 6703 }, { "epoch": 0.0006704, "grad_norm": 0.8674948811531067, "learning_rate": 6.703e-05, "loss": 0.3071, "step": 6704 }, { "epoch": 0.0006705, "grad_norm": 1.395816445350647, "learning_rate": 6.704000000000001e-05, "loss": 0.3999, "step": 6705 }, { "epoch": 0.0006706, "grad_norm": 0.8511848449707031, "learning_rate": 6.704999999999999e-05, "loss": 0.3259, "step": 6706 }, { "epoch": 0.0006707, "grad_norm": 0.7739931344985962, "learning_rate": 6.706e-05, "loss": 0.3135, "step": 6707 }, { "epoch": 0.0006708, "grad_norm": 0.7451652884483337, "learning_rate": 6.707e-05, "loss": 0.3245, "step": 6708 }, { "epoch": 0.0006709, "grad_norm": 0.7722559571266174, "learning_rate": 6.708e-05, "loss": 0.3467, "step": 6709 }, { "epoch": 0.000671, "grad_norm": 0.8112444877624512, "learning_rate": 6.709e-05, "loss": 0.3267, "step": 6710 }, { "epoch": 0.0006711, "grad_norm": 1.377299427986145, "learning_rate": 6.71e-05, "loss": 0.4106, "step": 6711 }, { "epoch": 0.0006712, "grad_norm": 0.758207380771637, "learning_rate": 6.711e-05, "loss": 0.2981, "step": 6712 }, { "epoch": 0.0006713, "grad_norm": 0.9617581963539124, "learning_rate": 6.712e-05, "loss": 0.3169, "step": 6713 }, { "epoch": 0.0006714, "grad_norm": 0.7949542999267578, "learning_rate": 6.713e-05, "loss": 0.3223, "step": 6714 }, { "epoch": 0.0006715, "grad_norm": 0.7971898913383484, "learning_rate": 6.714e-05, "loss": 0.3206, "step": 6715 }, { "epoch": 0.0006716, "grad_norm": 0.8708210587501526, "learning_rate": 6.715e-05, "loss": 0.343, "step": 6716 }, { "epoch": 0.0006717, "grad_norm": 0.8736813068389893, "learning_rate": 6.716e-05, "loss": 0.3489, "step": 6717 }, { "epoch": 0.0006718, "grad_norm": 1.4544368982315063, "learning_rate": 6.717e-05, "loss": 0.3719, "step": 6718 }, { "epoch": 0.0006719, "grad_norm": 0.7504727244377136, "learning_rate": 6.718e-05, "loss": 0.2983, "step": 6719 }, { "epoch": 0.000672, "grad_norm": 0.7012472152709961, "learning_rate": 6.719e-05, "loss": 0.2863, "step": 6720 }, { "epoch": 0.0006721, "grad_norm": 0.7165892720222473, "learning_rate": 6.720000000000001e-05, "loss": 0.2915, "step": 6721 }, { "epoch": 0.0006722, "grad_norm": 0.687541663646698, "learning_rate": 6.721e-05, "loss": 0.2988, "step": 6722 }, { "epoch": 0.0006723, "grad_norm": 1.2955049276351929, "learning_rate": 6.722e-05, "loss": 0.3955, "step": 6723 }, { "epoch": 0.0006724, "grad_norm": 0.7190659046173096, "learning_rate": 6.723e-05, "loss": 0.2971, "step": 6724 }, { "epoch": 0.0006725, "grad_norm": 0.9322375059127808, "learning_rate": 6.724e-05, "loss": 0.3459, "step": 6725 }, { "epoch": 0.0006726, "grad_norm": 1.2733296155929565, "learning_rate": 6.725000000000001e-05, "loss": 0.4817, "step": 6726 }, { "epoch": 0.0006727, "grad_norm": 0.8756181597709656, "learning_rate": 6.726e-05, "loss": 0.3357, "step": 6727 }, { "epoch": 0.0006728, "grad_norm": 0.8180313110351562, "learning_rate": 6.727e-05, "loss": 0.3359, "step": 6728 }, { "epoch": 0.0006729, "grad_norm": 0.7750518918037415, "learning_rate": 6.728e-05, "loss": 0.3181, "step": 6729 }, { "epoch": 0.000673, "grad_norm": 0.7350186705589294, "learning_rate": 6.729e-05, "loss": 0.2993, "step": 6730 }, { "epoch": 0.0006731, "grad_norm": 1.417085886001587, "learning_rate": 6.73e-05, "loss": 0.4121, "step": 6731 }, { "epoch": 0.0006732, "grad_norm": 1.0426276922225952, "learning_rate": 6.731e-05, "loss": 0.394, "step": 6732 }, { "epoch": 0.0006733, "grad_norm": 0.8477088212966919, "learning_rate": 6.732000000000001e-05, "loss": 0.3076, "step": 6733 }, { "epoch": 0.0006734, "grad_norm": 0.811088502407074, "learning_rate": 6.733e-05, "loss": 0.3157, "step": 6734 }, { "epoch": 0.0006735, "grad_norm": 0.9746713638305664, "learning_rate": 6.734e-05, "loss": 0.312, "step": 6735 }, { "epoch": 0.0006736, "grad_norm": 0.8316006064414978, "learning_rate": 6.735000000000001e-05, "loss": 0.3254, "step": 6736 }, { "epoch": 0.0006737, "grad_norm": 0.9084553122520447, "learning_rate": 6.735999999999999e-05, "loss": 0.3184, "step": 6737 }, { "epoch": 0.0006738, "grad_norm": 0.936915934085846, "learning_rate": 6.737e-05, "loss": 0.3342, "step": 6738 }, { "epoch": 0.0006739, "grad_norm": 0.9024308323860168, "learning_rate": 6.738000000000001e-05, "loss": 0.3513, "step": 6739 }, { "epoch": 0.000674, "grad_norm": 5.580689430236816, "learning_rate": 6.738999999999999e-05, "loss": 0.7666, "step": 6740 }, { "epoch": 0.0006741, "grad_norm": 1.0180385112762451, "learning_rate": 6.74e-05, "loss": 0.3221, "step": 6741 }, { "epoch": 0.0006742, "grad_norm": 1.036532998085022, "learning_rate": 6.741000000000001e-05, "loss": 0.3491, "step": 6742 }, { "epoch": 0.0006743, "grad_norm": 0.769421398639679, "learning_rate": 6.742e-05, "loss": 0.3059, "step": 6743 }, { "epoch": 0.0006744, "grad_norm": 0.8509790897369385, "learning_rate": 6.743e-05, "loss": 0.3162, "step": 6744 }, { "epoch": 0.0006745, "grad_norm": 1.186266303062439, "learning_rate": 6.744e-05, "loss": 0.4888, "step": 6745 }, { "epoch": 0.0006746, "grad_norm": 0.9085695743560791, "learning_rate": 6.745e-05, "loss": 0.322, "step": 6746 }, { "epoch": 0.0006747, "grad_norm": 0.8790482878684998, "learning_rate": 6.746e-05, "loss": 0.3257, "step": 6747 }, { "epoch": 0.0006748, "grad_norm": 0.8217053413391113, "learning_rate": 6.747e-05, "loss": 0.2786, "step": 6748 }, { "epoch": 0.0006749, "grad_norm": 1.2313982248306274, "learning_rate": 6.748e-05, "loss": 0.3308, "step": 6749 }, { "epoch": 0.000675, "grad_norm": 1.2116960287094116, "learning_rate": 6.749e-05, "loss": 0.3868, "step": 6750 }, { "epoch": 0.0006751, "grad_norm": 0.8842405080795288, "learning_rate": 6.75e-05, "loss": 0.3301, "step": 6751 }, { "epoch": 0.0006752, "grad_norm": 0.70157790184021, "learning_rate": 6.751e-05, "loss": 0.2893, "step": 6752 }, { "epoch": 0.0006753, "grad_norm": 0.7880963087081909, "learning_rate": 6.752e-05, "loss": 0.3149, "step": 6753 }, { "epoch": 0.0006754, "grad_norm": 1.089812994003296, "learning_rate": 6.753e-05, "loss": 0.405, "step": 6754 }, { "epoch": 0.0006755, "grad_norm": 0.7792335152626038, "learning_rate": 6.754000000000001e-05, "loss": 0.3137, "step": 6755 }, { "epoch": 0.0006756, "grad_norm": 0.7259790301322937, "learning_rate": 6.755e-05, "loss": 0.3015, "step": 6756 }, { "epoch": 0.0006757, "grad_norm": 0.7795138359069824, "learning_rate": 6.756e-05, "loss": 0.313, "step": 6757 }, { "epoch": 0.0006758, "grad_norm": 0.7948536276817322, "learning_rate": 6.757e-05, "loss": 0.3145, "step": 6758 }, { "epoch": 0.0006759, "grad_norm": 0.7575865983963013, "learning_rate": 6.758e-05, "loss": 0.3254, "step": 6759 }, { "epoch": 0.000676, "grad_norm": 0.9956995248794556, "learning_rate": 6.759000000000001e-05, "loss": 0.3379, "step": 6760 }, { "epoch": 0.0006761, "grad_norm": 0.7819147706031799, "learning_rate": 6.76e-05, "loss": 0.3311, "step": 6761 }, { "epoch": 0.0006762, "grad_norm": 0.9284989237785339, "learning_rate": 6.761e-05, "loss": 0.3269, "step": 6762 }, { "epoch": 0.0006763, "grad_norm": 0.7394283413887024, "learning_rate": 6.762e-05, "loss": 0.2828, "step": 6763 }, { "epoch": 0.0006764, "grad_norm": 0.9893227219581604, "learning_rate": 6.763e-05, "loss": 0.3328, "step": 6764 }, { "epoch": 0.0006765, "grad_norm": 0.6851495504379272, "learning_rate": 6.764e-05, "loss": 0.2831, "step": 6765 }, { "epoch": 0.0006766, "grad_norm": 0.7171580791473389, "learning_rate": 6.765e-05, "loss": 0.281, "step": 6766 }, { "epoch": 0.0006767, "grad_norm": 0.7466902136802673, "learning_rate": 6.766000000000001e-05, "loss": 0.2981, "step": 6767 }, { "epoch": 0.0006768, "grad_norm": 0.7984172701835632, "learning_rate": 6.767e-05, "loss": 0.2993, "step": 6768 }, { "epoch": 0.0006769, "grad_norm": 0.7800801992416382, "learning_rate": 6.768e-05, "loss": 0.3191, "step": 6769 }, { "epoch": 0.000677, "grad_norm": 0.7060449123382568, "learning_rate": 6.769000000000001e-05, "loss": 0.2939, "step": 6770 }, { "epoch": 0.0006771, "grad_norm": 0.8864577412605286, "learning_rate": 6.769999999999999e-05, "loss": 0.3203, "step": 6771 }, { "epoch": 0.0006772, "grad_norm": 0.702970027923584, "learning_rate": 6.771e-05, "loss": 0.2996, "step": 6772 }, { "epoch": 0.0006773, "grad_norm": 0.8598361015319824, "learning_rate": 6.772000000000001e-05, "loss": 0.3125, "step": 6773 }, { "epoch": 0.0006774, "grad_norm": 0.7288208603858948, "learning_rate": 6.772999999999999e-05, "loss": 0.2969, "step": 6774 }, { "epoch": 0.0006775, "grad_norm": 0.8956363201141357, "learning_rate": 6.774e-05, "loss": 0.3125, "step": 6775 }, { "epoch": 0.0006776, "grad_norm": 0.631296694278717, "learning_rate": 6.775000000000001e-05, "loss": 0.29, "step": 6776 }, { "epoch": 0.0006777, "grad_norm": 1.2960058450698853, "learning_rate": 6.776e-05, "loss": 0.478, "step": 6777 }, { "epoch": 0.0006778, "grad_norm": 1.2035611867904663, "learning_rate": 6.777e-05, "loss": 0.3376, "step": 6778 }, { "epoch": 0.0006779, "grad_norm": 0.7155436277389526, "learning_rate": 6.778e-05, "loss": 0.2738, "step": 6779 }, { "epoch": 0.000678, "grad_norm": 0.7673712372779846, "learning_rate": 6.779e-05, "loss": 0.3064, "step": 6780 }, { "epoch": 0.0006781, "grad_norm": 0.8256517052650452, "learning_rate": 6.78e-05, "loss": 0.3408, "step": 6781 }, { "epoch": 0.0006782, "grad_norm": 0.7277462482452393, "learning_rate": 6.781e-05, "loss": 0.2986, "step": 6782 }, { "epoch": 0.0006783, "grad_norm": 0.8773739337921143, "learning_rate": 6.782e-05, "loss": 0.3452, "step": 6783 }, { "epoch": 0.0006784, "grad_norm": 0.7977916598320007, "learning_rate": 6.783e-05, "loss": 0.3228, "step": 6784 }, { "epoch": 0.0006785, "grad_norm": 0.6325885653495789, "learning_rate": 6.784e-05, "loss": 0.2753, "step": 6785 }, { "epoch": 0.0006786, "grad_norm": 0.9474706053733826, "learning_rate": 6.785e-05, "loss": 0.342, "step": 6786 }, { "epoch": 0.0006787, "grad_norm": 0.7229621410369873, "learning_rate": 6.786e-05, "loss": 0.2976, "step": 6787 }, { "epoch": 0.0006788, "grad_norm": 0.6040972471237183, "learning_rate": 6.787e-05, "loss": 0.2732, "step": 6788 }, { "epoch": 0.0006789, "grad_norm": 1.5880738496780396, "learning_rate": 6.788000000000001e-05, "loss": 0.5254, "step": 6789 }, { "epoch": 0.000679, "grad_norm": 0.766530454158783, "learning_rate": 6.789e-05, "loss": 0.3114, "step": 6790 }, { "epoch": 0.0006791, "grad_norm": 0.7376402616500854, "learning_rate": 6.79e-05, "loss": 0.3174, "step": 6791 }, { "epoch": 0.0006792, "grad_norm": 3.2554433345794678, "learning_rate": 6.791e-05, "loss": 0.4823, "step": 6792 }, { "epoch": 0.0006793, "grad_norm": 0.840853750705719, "learning_rate": 6.792e-05, "loss": 0.3218, "step": 6793 }, { "epoch": 0.0006794, "grad_norm": 0.8738373517990112, "learning_rate": 6.793000000000001e-05, "loss": 0.3186, "step": 6794 }, { "epoch": 0.0006795, "grad_norm": 0.7738059163093567, "learning_rate": 6.794e-05, "loss": 0.3142, "step": 6795 }, { "epoch": 0.0006796, "grad_norm": 0.7867991924285889, "learning_rate": 6.795e-05, "loss": 0.2955, "step": 6796 }, { "epoch": 0.0006797, "grad_norm": 0.7323338389396667, "learning_rate": 6.796e-05, "loss": 0.2834, "step": 6797 }, { "epoch": 0.0006798, "grad_norm": 0.7408937215805054, "learning_rate": 6.797e-05, "loss": 0.2966, "step": 6798 }, { "epoch": 0.0006799, "grad_norm": 1.057076096534729, "learning_rate": 6.798e-05, "loss": 0.3621, "step": 6799 }, { "epoch": 0.00068, "grad_norm": 0.7257677316665649, "learning_rate": 6.799e-05, "loss": 0.2898, "step": 6800 }, { "epoch": 0.0006801, "grad_norm": 0.6983649730682373, "learning_rate": 6.800000000000001e-05, "loss": 0.2886, "step": 6801 }, { "epoch": 0.0006802, "grad_norm": 0.7865191102027893, "learning_rate": 6.801e-05, "loss": 0.3005, "step": 6802 }, { "epoch": 0.0006803, "grad_norm": 1.1119191646575928, "learning_rate": 6.802e-05, "loss": 0.3594, "step": 6803 }, { "epoch": 0.0006804, "grad_norm": 0.7985936403274536, "learning_rate": 6.803000000000001e-05, "loss": 0.3114, "step": 6804 }, { "epoch": 0.0006805, "grad_norm": 0.6647433042526245, "learning_rate": 6.803999999999999e-05, "loss": 0.2828, "step": 6805 }, { "epoch": 0.0006806, "grad_norm": 1.1630476713180542, "learning_rate": 6.805e-05, "loss": 0.387, "step": 6806 }, { "epoch": 0.0006807, "grad_norm": 1.109668493270874, "learning_rate": 6.806000000000001e-05, "loss": 0.3226, "step": 6807 }, { "epoch": 0.0006808, "grad_norm": 0.8443289399147034, "learning_rate": 6.806999999999999e-05, "loss": 0.3088, "step": 6808 }, { "epoch": 0.0006809, "grad_norm": 1.652001976966858, "learning_rate": 6.808e-05, "loss": 0.3855, "step": 6809 }, { "epoch": 0.000681, "grad_norm": 0.8046223521232605, "learning_rate": 6.809000000000001e-05, "loss": 0.3191, "step": 6810 }, { "epoch": 0.0006811, "grad_norm": 2.053948163986206, "learning_rate": 6.81e-05, "loss": 0.4033, "step": 6811 }, { "epoch": 0.0006812, "grad_norm": 1.0023415088653564, "learning_rate": 6.811e-05, "loss": 0.3169, "step": 6812 }, { "epoch": 0.0006813, "grad_norm": 0.8542107939720154, "learning_rate": 6.812e-05, "loss": 0.2849, "step": 6813 }, { "epoch": 0.0006814, "grad_norm": 0.771658182144165, "learning_rate": 6.813e-05, "loss": 0.2908, "step": 6814 }, { "epoch": 0.0006815, "grad_norm": 0.89285808801651, "learning_rate": 6.814e-05, "loss": 0.3062, "step": 6815 }, { "epoch": 0.0006816, "grad_norm": 0.8619227409362793, "learning_rate": 6.815e-05, "loss": 0.3032, "step": 6816 }, { "epoch": 0.0006817, "grad_norm": 1.047168493270874, "learning_rate": 6.816e-05, "loss": 0.3362, "step": 6817 }, { "epoch": 0.0006818, "grad_norm": 0.9726020693778992, "learning_rate": 6.817e-05, "loss": 0.2985, "step": 6818 }, { "epoch": 0.0006819, "grad_norm": 0.7409684658050537, "learning_rate": 6.818e-05, "loss": 0.2693, "step": 6819 }, { "epoch": 0.000682, "grad_norm": 0.7690621614456177, "learning_rate": 6.819e-05, "loss": 0.2837, "step": 6820 }, { "epoch": 0.0006821, "grad_norm": 0.8043884634971619, "learning_rate": 6.82e-05, "loss": 0.3254, "step": 6821 }, { "epoch": 0.0006822, "grad_norm": 0.7533547282218933, "learning_rate": 6.821e-05, "loss": 0.2983, "step": 6822 }, { "epoch": 0.0006823, "grad_norm": 1.1310558319091797, "learning_rate": 6.822000000000001e-05, "loss": 0.363, "step": 6823 }, { "epoch": 0.0006824, "grad_norm": 1.1872040033340454, "learning_rate": 6.823e-05, "loss": 0.3219, "step": 6824 }, { "epoch": 0.0006825, "grad_norm": 0.9793376922607422, "learning_rate": 6.824e-05, "loss": 0.3489, "step": 6825 }, { "epoch": 0.0006826, "grad_norm": 0.7625957727432251, "learning_rate": 6.825e-05, "loss": 0.2771, "step": 6826 }, { "epoch": 0.0006827, "grad_norm": 0.7681121230125427, "learning_rate": 6.826e-05, "loss": 0.3005, "step": 6827 }, { "epoch": 0.0006828, "grad_norm": 0.831994354724884, "learning_rate": 6.827000000000001e-05, "loss": 0.3113, "step": 6828 }, { "epoch": 0.0006829, "grad_norm": 0.6504027247428894, "learning_rate": 6.828e-05, "loss": 0.2791, "step": 6829 }, { "epoch": 0.000683, "grad_norm": 0.855286717414856, "learning_rate": 6.829e-05, "loss": 0.3169, "step": 6830 }, { "epoch": 0.0006831, "grad_norm": 0.7506752014160156, "learning_rate": 6.83e-05, "loss": 0.2863, "step": 6831 }, { "epoch": 0.0006832, "grad_norm": 0.6989216804504395, "learning_rate": 6.831e-05, "loss": 0.2861, "step": 6832 }, { "epoch": 0.0006833, "grad_norm": 0.7412368655204773, "learning_rate": 6.832e-05, "loss": 0.3159, "step": 6833 }, { "epoch": 0.0006834, "grad_norm": 0.677550733089447, "learning_rate": 6.833e-05, "loss": 0.2936, "step": 6834 }, { "epoch": 0.0006835, "grad_norm": 1.302125334739685, "learning_rate": 6.834e-05, "loss": 0.3418, "step": 6835 }, { "epoch": 0.0006836, "grad_norm": 0.7905217409133911, "learning_rate": 6.835e-05, "loss": 0.3115, "step": 6836 }, { "epoch": 0.0006837, "grad_norm": 0.7866873145103455, "learning_rate": 6.836e-05, "loss": 0.2991, "step": 6837 }, { "epoch": 0.0006838, "grad_norm": 0.7025447487831116, "learning_rate": 6.837000000000001e-05, "loss": 0.2747, "step": 6838 }, { "epoch": 0.0006839, "grad_norm": 0.6906653642654419, "learning_rate": 6.837999999999999e-05, "loss": 0.2881, "step": 6839 }, { "epoch": 0.000684, "grad_norm": 0.6507428288459778, "learning_rate": 6.839e-05, "loss": 0.2745, "step": 6840 }, { "epoch": 0.0006841, "grad_norm": 0.7092325687408447, "learning_rate": 6.840000000000001e-05, "loss": 0.2927, "step": 6841 }, { "epoch": 0.0006842, "grad_norm": 0.79344642162323, "learning_rate": 6.840999999999999e-05, "loss": 0.3191, "step": 6842 }, { "epoch": 0.0006843, "grad_norm": 0.6421338319778442, "learning_rate": 6.842e-05, "loss": 0.2803, "step": 6843 }, { "epoch": 0.0006844, "grad_norm": 0.6311710476875305, "learning_rate": 6.843000000000001e-05, "loss": 0.2732, "step": 6844 }, { "epoch": 0.0006845, "grad_norm": 1.3130443096160889, "learning_rate": 6.844e-05, "loss": 0.3396, "step": 6845 }, { "epoch": 0.0006846, "grad_norm": 0.6774488687515259, "learning_rate": 6.845e-05, "loss": 0.2732, "step": 6846 }, { "epoch": 0.0006847, "grad_norm": 0.7677571177482605, "learning_rate": 6.846e-05, "loss": 0.291, "step": 6847 }, { "epoch": 0.0006848, "grad_norm": 0.6695631742477417, "learning_rate": 6.847e-05, "loss": 0.282, "step": 6848 }, { "epoch": 0.0006849, "grad_norm": 0.7841461896896362, "learning_rate": 6.848e-05, "loss": 0.3225, "step": 6849 }, { "epoch": 0.000685, "grad_norm": 0.7521286010742188, "learning_rate": 6.849e-05, "loss": 0.2922, "step": 6850 }, { "epoch": 0.0006851, "grad_norm": 0.6829274892807007, "learning_rate": 6.85e-05, "loss": 0.297, "step": 6851 }, { "epoch": 0.0006852, "grad_norm": 0.5868691802024841, "learning_rate": 6.851e-05, "loss": 0.264, "step": 6852 }, { "epoch": 0.0006853, "grad_norm": 0.6112059354782104, "learning_rate": 6.852e-05, "loss": 0.269, "step": 6853 }, { "epoch": 0.0006854, "grad_norm": 0.6771926283836365, "learning_rate": 6.853e-05, "loss": 0.2915, "step": 6854 }, { "epoch": 0.0006855, "grad_norm": 0.6877410411834717, "learning_rate": 6.854e-05, "loss": 0.2654, "step": 6855 }, { "epoch": 0.0006856, "grad_norm": 0.7091522216796875, "learning_rate": 6.855e-05, "loss": 0.2864, "step": 6856 }, { "epoch": 0.0006857, "grad_norm": 0.6090933084487915, "learning_rate": 6.856e-05, "loss": 0.2701, "step": 6857 }, { "epoch": 0.0006858, "grad_norm": 0.6597985625267029, "learning_rate": 6.857e-05, "loss": 0.2952, "step": 6858 }, { "epoch": 0.0006859, "grad_norm": 5.0231475830078125, "learning_rate": 6.858e-05, "loss": 0.3691, "step": 6859 }, { "epoch": 0.000686, "grad_norm": 0.829133152961731, "learning_rate": 6.859e-05, "loss": 0.2822, "step": 6860 }, { "epoch": 0.0006861, "grad_norm": 1.6139463186264038, "learning_rate": 6.86e-05, "loss": 0.3093, "step": 6861 }, { "epoch": 0.0006862, "grad_norm": 1.1528732776641846, "learning_rate": 6.861000000000001e-05, "loss": 0.3365, "step": 6862 }, { "epoch": 0.0006863, "grad_norm": 0.673590898513794, "learning_rate": 6.862e-05, "loss": 0.2788, "step": 6863 }, { "epoch": 0.0006864, "grad_norm": 0.8255468010902405, "learning_rate": 6.863e-05, "loss": 0.3047, "step": 6864 }, { "epoch": 0.0006865, "grad_norm": 0.980257511138916, "learning_rate": 6.864000000000001e-05, "loss": 0.3389, "step": 6865 }, { "epoch": 0.0006866, "grad_norm": 0.9978033900260925, "learning_rate": 6.865e-05, "loss": 0.2896, "step": 6866 }, { "epoch": 0.0006867, "grad_norm": 0.6645432114601135, "learning_rate": 6.866e-05, "loss": 0.27, "step": 6867 }, { "epoch": 0.0006868, "grad_norm": 0.7752996683120728, "learning_rate": 6.867e-05, "loss": 0.2797, "step": 6868 }, { "epoch": 0.0006869, "grad_norm": 0.846230149269104, "learning_rate": 6.868e-05, "loss": 0.3071, "step": 6869 }, { "epoch": 0.000687, "grad_norm": 0.715424656867981, "learning_rate": 6.869e-05, "loss": 0.2756, "step": 6870 }, { "epoch": 0.0006871, "grad_norm": 0.7180454134941101, "learning_rate": 6.87e-05, "loss": 0.2932, "step": 6871 }, { "epoch": 0.0006872, "grad_norm": 0.734980583190918, "learning_rate": 6.871000000000001e-05, "loss": 0.2996, "step": 6872 }, { "epoch": 0.0006873, "grad_norm": 0.7681039571762085, "learning_rate": 6.871999999999999e-05, "loss": 0.2942, "step": 6873 }, { "epoch": 0.0006874, "grad_norm": 0.7316890954971313, "learning_rate": 6.873e-05, "loss": 0.2721, "step": 6874 }, { "epoch": 0.0006875, "grad_norm": 0.6318367123603821, "learning_rate": 6.874000000000001e-05, "loss": 0.2681, "step": 6875 }, { "epoch": 0.0006876, "grad_norm": 0.6656007766723633, "learning_rate": 6.874999999999999e-05, "loss": 0.2736, "step": 6876 }, { "epoch": 0.0006877, "grad_norm": 0.6839348673820496, "learning_rate": 6.876e-05, "loss": 0.2949, "step": 6877 }, { "epoch": 0.0006878, "grad_norm": 1.1404519081115723, "learning_rate": 6.877000000000001e-05, "loss": 0.3265, "step": 6878 }, { "epoch": 0.0006879, "grad_norm": 0.6324551701545715, "learning_rate": 6.878e-05, "loss": 0.2639, "step": 6879 }, { "epoch": 0.000688, "grad_norm": 0.6577877998352051, "learning_rate": 6.879e-05, "loss": 0.2781, "step": 6880 }, { "epoch": 0.0006881, "grad_norm": 0.6746148467063904, "learning_rate": 6.88e-05, "loss": 0.2788, "step": 6881 }, { "epoch": 0.0006882, "grad_norm": 0.6653241515159607, "learning_rate": 6.881e-05, "loss": 0.2889, "step": 6882 }, { "epoch": 0.0006883, "grad_norm": 0.6338821649551392, "learning_rate": 6.882e-05, "loss": 0.2808, "step": 6883 }, { "epoch": 0.0006884, "grad_norm": 0.6709138751029968, "learning_rate": 6.883e-05, "loss": 0.2832, "step": 6884 }, { "epoch": 0.0006885, "grad_norm": 0.5759314298629761, "learning_rate": 6.884e-05, "loss": 0.2576, "step": 6885 }, { "epoch": 0.0006886, "grad_norm": 0.7626339793205261, "learning_rate": 6.885e-05, "loss": 0.2808, "step": 6886 }, { "epoch": 0.0006887, "grad_norm": 1.1623761653900146, "learning_rate": 6.886e-05, "loss": 0.4182, "step": 6887 }, { "epoch": 0.0006888, "grad_norm": 0.8023167252540588, "learning_rate": 6.887e-05, "loss": 0.301, "step": 6888 }, { "epoch": 0.0006889, "grad_norm": 1.7670682668685913, "learning_rate": 6.888e-05, "loss": 0.413, "step": 6889 }, { "epoch": 0.000689, "grad_norm": 0.7617492079734802, "learning_rate": 6.889e-05, "loss": 0.2847, "step": 6890 }, { "epoch": 0.0006891, "grad_norm": 6.765068054199219, "learning_rate": 6.89e-05, "loss": 0.5615, "step": 6891 }, { "epoch": 0.0006892, "grad_norm": 0.8808358311653137, "learning_rate": 6.891e-05, "loss": 0.2999, "step": 6892 }, { "epoch": 0.0006893, "grad_norm": 0.7774884104728699, "learning_rate": 6.892e-05, "loss": 0.2833, "step": 6893 }, { "epoch": 0.0006894, "grad_norm": 0.7922816276550293, "learning_rate": 6.893e-05, "loss": 0.2834, "step": 6894 }, { "epoch": 0.0006895, "grad_norm": 0.7029367685317993, "learning_rate": 6.894e-05, "loss": 0.2765, "step": 6895 }, { "epoch": 0.0006896, "grad_norm": 1.1723849773406982, "learning_rate": 6.895000000000001e-05, "loss": 0.3149, "step": 6896 }, { "epoch": 0.0006897, "grad_norm": 0.7313202619552612, "learning_rate": 6.896e-05, "loss": 0.2677, "step": 6897 }, { "epoch": 0.0006898, "grad_norm": 0.7777508497238159, "learning_rate": 6.897e-05, "loss": 0.2744, "step": 6898 }, { "epoch": 0.0006899, "grad_norm": 1.5544298887252808, "learning_rate": 6.898000000000001e-05, "loss": 0.3359, "step": 6899 }, { "epoch": 0.00069, "grad_norm": 2.774675130844116, "learning_rate": 6.899e-05, "loss": 0.5413, "step": 6900 }, { "epoch": 0.0006901, "grad_norm": 0.7986059784889221, "learning_rate": 6.9e-05, "loss": 0.2656, "step": 6901 }, { "epoch": 0.0006902, "grad_norm": 1.1680564880371094, "learning_rate": 6.901e-05, "loss": 0.3145, "step": 6902 }, { "epoch": 0.0006903, "grad_norm": 0.8539535999298096, "learning_rate": 6.902e-05, "loss": 0.3198, "step": 6903 }, { "epoch": 0.0006904, "grad_norm": 0.6857061982154846, "learning_rate": 6.903e-05, "loss": 0.2793, "step": 6904 }, { "epoch": 0.0006905, "grad_norm": 0.7869257926940918, "learning_rate": 6.904e-05, "loss": 0.2793, "step": 6905 }, { "epoch": 0.0006906, "grad_norm": 0.9233639240264893, "learning_rate": 6.905000000000001e-05, "loss": 0.2875, "step": 6906 }, { "epoch": 0.0006907, "grad_norm": 0.7806829214096069, "learning_rate": 6.905999999999999e-05, "loss": 0.3007, "step": 6907 }, { "epoch": 0.0006908, "grad_norm": 0.7770219445228577, "learning_rate": 6.907e-05, "loss": 0.3054, "step": 6908 }, { "epoch": 0.0006909, "grad_norm": 0.6133571267127991, "learning_rate": 6.908000000000001e-05, "loss": 0.2567, "step": 6909 }, { "epoch": 0.000691, "grad_norm": 1.9333255290985107, "learning_rate": 6.908999999999999e-05, "loss": 0.4106, "step": 6910 }, { "epoch": 0.0006911, "grad_norm": 0.9833579659461975, "learning_rate": 6.91e-05, "loss": 0.2974, "step": 6911 }, { "epoch": 0.0006912, "grad_norm": 1.0075924396514893, "learning_rate": 6.911000000000001e-05, "loss": 0.2941, "step": 6912 }, { "epoch": 0.0006913, "grad_norm": 2.9899237155914307, "learning_rate": 6.912e-05, "loss": 0.4446, "step": 6913 }, { "epoch": 0.0006914, "grad_norm": 0.8225877285003662, "learning_rate": 6.913e-05, "loss": 0.2852, "step": 6914 }, { "epoch": 0.0006915, "grad_norm": 0.7364519834518433, "learning_rate": 6.914e-05, "loss": 0.2645, "step": 6915 }, { "epoch": 0.0006916, "grad_norm": 0.6848117709159851, "learning_rate": 6.915e-05, "loss": 0.2605, "step": 6916 }, { "epoch": 0.0006917, "grad_norm": 0.8027401566505432, "learning_rate": 6.916e-05, "loss": 0.2672, "step": 6917 }, { "epoch": 0.0006918, "grad_norm": 0.6527199149131775, "learning_rate": 6.917e-05, "loss": 0.2654, "step": 6918 }, { "epoch": 0.0006919, "grad_norm": 0.6591749787330627, "learning_rate": 6.918e-05, "loss": 0.2612, "step": 6919 }, { "epoch": 0.000692, "grad_norm": 1.452710747718811, "learning_rate": 6.919e-05, "loss": 0.4023, "step": 6920 }, { "epoch": 0.0006921, "grad_norm": 0.6021394729614258, "learning_rate": 6.92e-05, "loss": 0.2585, "step": 6921 }, { "epoch": 0.0006922, "grad_norm": 4.742277145385742, "learning_rate": 6.921e-05, "loss": 0.3076, "step": 6922 }, { "epoch": 0.0006923, "grad_norm": 0.7667524814605713, "learning_rate": 6.922e-05, "loss": 0.2605, "step": 6923 }, { "epoch": 0.0006924, "grad_norm": 0.7648462653160095, "learning_rate": 6.923e-05, "loss": 0.2688, "step": 6924 }, { "epoch": 0.0006925, "grad_norm": 0.7535057663917542, "learning_rate": 6.924e-05, "loss": 0.2906, "step": 6925 }, { "epoch": 0.0006926, "grad_norm": 0.6999765634536743, "learning_rate": 6.925e-05, "loss": 0.2947, "step": 6926 }, { "epoch": 0.0006927, "grad_norm": 0.8671202659606934, "learning_rate": 6.926e-05, "loss": 0.3026, "step": 6927 }, { "epoch": 0.0006928, "grad_norm": 0.7182843685150146, "learning_rate": 6.927e-05, "loss": 0.2639, "step": 6928 }, { "epoch": 0.0006929, "grad_norm": 0.7621828317642212, "learning_rate": 6.928e-05, "loss": 0.29, "step": 6929 }, { "epoch": 0.000693, "grad_norm": 0.7141491174697876, "learning_rate": 6.929000000000001e-05, "loss": 0.2869, "step": 6930 }, { "epoch": 0.0006931, "grad_norm": 0.8775362372398376, "learning_rate": 6.93e-05, "loss": 0.2881, "step": 6931 }, { "epoch": 0.0006932, "grad_norm": 1.493428349494934, "learning_rate": 6.931e-05, "loss": 0.4263, "step": 6932 }, { "epoch": 0.0006933, "grad_norm": 0.732142984867096, "learning_rate": 6.932000000000001e-05, "loss": 0.2621, "step": 6933 }, { "epoch": 0.0006934, "grad_norm": 0.7555822730064392, "learning_rate": 6.933e-05, "loss": 0.2598, "step": 6934 }, { "epoch": 0.0006935, "grad_norm": 0.8155015110969543, "learning_rate": 6.934e-05, "loss": 0.2904, "step": 6935 }, { "epoch": 0.0006936, "grad_norm": 0.5576425194740295, "learning_rate": 6.935e-05, "loss": 0.2446, "step": 6936 }, { "epoch": 0.0006937, "grad_norm": 0.7607638239860535, "learning_rate": 6.936e-05, "loss": 0.2734, "step": 6937 }, { "epoch": 0.0006938, "grad_norm": 1.1647169589996338, "learning_rate": 6.937e-05, "loss": 0.4021, "step": 6938 }, { "epoch": 0.0006939, "grad_norm": 1.0491317510604858, "learning_rate": 6.938e-05, "loss": 0.3015, "step": 6939 }, { "epoch": 0.000694, "grad_norm": 0.8073709607124329, "learning_rate": 6.939000000000001e-05, "loss": 0.2809, "step": 6940 }, { "epoch": 0.0006941, "grad_norm": 0.72732013463974, "learning_rate": 6.939999999999999e-05, "loss": 0.2743, "step": 6941 }, { "epoch": 0.0006942, "grad_norm": 1.0625195503234863, "learning_rate": 6.941e-05, "loss": 0.4106, "step": 6942 }, { "epoch": 0.0006943, "grad_norm": 0.9433525800704956, "learning_rate": 6.942000000000001e-05, "loss": 0.3411, "step": 6943 }, { "epoch": 0.0006944, "grad_norm": 0.7006153464317322, "learning_rate": 6.942999999999999e-05, "loss": 0.2584, "step": 6944 }, { "epoch": 0.0006945, "grad_norm": 1.3899009227752686, "learning_rate": 6.944e-05, "loss": 0.3727, "step": 6945 }, { "epoch": 0.0006946, "grad_norm": 0.9198282957077026, "learning_rate": 6.945000000000001e-05, "loss": 0.3386, "step": 6946 }, { "epoch": 0.0006947, "grad_norm": 0.7298462390899658, "learning_rate": 6.946e-05, "loss": 0.2622, "step": 6947 }, { "epoch": 0.0006948, "grad_norm": 0.7429442405700684, "learning_rate": 6.947e-05, "loss": 0.261, "step": 6948 }, { "epoch": 0.0006949, "grad_norm": 0.7668522596359253, "learning_rate": 6.948e-05, "loss": 0.2881, "step": 6949 }, { "epoch": 0.000695, "grad_norm": 0.6823588013648987, "learning_rate": 6.949e-05, "loss": 0.278, "step": 6950 }, { "epoch": 0.0006951, "grad_norm": 0.6069337725639343, "learning_rate": 6.95e-05, "loss": 0.2505, "step": 6951 }, { "epoch": 0.0006952, "grad_norm": 1.1467939615249634, "learning_rate": 6.951e-05, "loss": 0.3793, "step": 6952 }, { "epoch": 0.0006953, "grad_norm": 0.8934077620506287, "learning_rate": 6.952e-05, "loss": 0.3065, "step": 6953 }, { "epoch": 0.0006954, "grad_norm": 0.7593517303466797, "learning_rate": 6.953000000000001e-05, "loss": 0.2786, "step": 6954 }, { "epoch": 0.0006955, "grad_norm": 0.587807297706604, "learning_rate": 6.954e-05, "loss": 0.2476, "step": 6955 }, { "epoch": 0.0006956, "grad_norm": 0.7230364680290222, "learning_rate": 6.955e-05, "loss": 0.2728, "step": 6956 }, { "epoch": 0.0006957, "grad_norm": 0.9407157301902771, "learning_rate": 6.956e-05, "loss": 0.3571, "step": 6957 }, { "epoch": 0.0006958, "grad_norm": 1.3212006092071533, "learning_rate": 6.957e-05, "loss": 0.312, "step": 6958 }, { "epoch": 0.0006959, "grad_norm": 0.8143905401229858, "learning_rate": 6.958e-05, "loss": 0.3319, "step": 6959 }, { "epoch": 0.000696, "grad_norm": 0.7418932318687439, "learning_rate": 6.959e-05, "loss": 0.2756, "step": 6960 }, { "epoch": 0.0006961, "grad_norm": 0.9562113285064697, "learning_rate": 6.96e-05, "loss": 0.2852, "step": 6961 }, { "epoch": 0.0006962, "grad_norm": 0.7164644598960876, "learning_rate": 6.961e-05, "loss": 0.2596, "step": 6962 }, { "epoch": 0.0006963, "grad_norm": 2.320756196975708, "learning_rate": 6.962e-05, "loss": 0.3641, "step": 6963 }, { "epoch": 0.0006964, "grad_norm": 0.7264569401741028, "learning_rate": 6.963000000000001e-05, "loss": 0.2603, "step": 6964 }, { "epoch": 0.0006965, "grad_norm": 0.8696227073669434, "learning_rate": 6.964e-05, "loss": 0.2762, "step": 6965 }, { "epoch": 0.0006966, "grad_norm": 0.7736457586288452, "learning_rate": 6.965e-05, "loss": 0.2499, "step": 6966 }, { "epoch": 0.0006967, "grad_norm": 0.7460713386535645, "learning_rate": 6.966000000000001e-05, "loss": 0.2756, "step": 6967 }, { "epoch": 0.0006968, "grad_norm": 0.7883889079093933, "learning_rate": 6.967e-05, "loss": 0.3052, "step": 6968 }, { "epoch": 0.0006969, "grad_norm": 0.7249166369438171, "learning_rate": 6.968e-05, "loss": 0.2706, "step": 6969 }, { "epoch": 0.000697, "grad_norm": 0.8029236793518066, "learning_rate": 6.969e-05, "loss": 0.2808, "step": 6970 }, { "epoch": 0.0006971, "grad_norm": 0.7636973857879639, "learning_rate": 6.97e-05, "loss": 0.2754, "step": 6971 }, { "epoch": 0.0006972, "grad_norm": 0.6970018148422241, "learning_rate": 6.971e-05, "loss": 0.2612, "step": 6972 }, { "epoch": 0.0006973, "grad_norm": 0.6350924968719482, "learning_rate": 6.972e-05, "loss": 0.2522, "step": 6973 }, { "epoch": 0.0006974, "grad_norm": 0.6835649609565735, "learning_rate": 6.973000000000001e-05, "loss": 0.2667, "step": 6974 }, { "epoch": 0.0006975, "grad_norm": 1.531586766242981, "learning_rate": 6.973999999999999e-05, "loss": 0.3433, "step": 6975 }, { "epoch": 0.0006976, "grad_norm": 0.784424364566803, "learning_rate": 6.975e-05, "loss": 0.2732, "step": 6976 }, { "epoch": 0.0006977, "grad_norm": 0.6527853012084961, "learning_rate": 6.976000000000001e-05, "loss": 0.2603, "step": 6977 }, { "epoch": 0.0006978, "grad_norm": 0.6221573948860168, "learning_rate": 6.976999999999999e-05, "loss": 0.2522, "step": 6978 }, { "epoch": 0.0006979, "grad_norm": 0.7770185470581055, "learning_rate": 6.978e-05, "loss": 0.2664, "step": 6979 }, { "epoch": 0.000698, "grad_norm": 0.5562544465065002, "learning_rate": 6.979000000000001e-05, "loss": 0.2478, "step": 6980 }, { "epoch": 0.0006981, "grad_norm": 0.6760058999061584, "learning_rate": 6.98e-05, "loss": 0.2714, "step": 6981 }, { "epoch": 0.0006982, "grad_norm": 0.675886869430542, "learning_rate": 6.981e-05, "loss": 0.2721, "step": 6982 }, { "epoch": 0.0006983, "grad_norm": 1.0800305604934692, "learning_rate": 6.982e-05, "loss": 0.3134, "step": 6983 }, { "epoch": 0.0006984, "grad_norm": 0.6709005832672119, "learning_rate": 6.983e-05, "loss": 0.2594, "step": 6984 }, { "epoch": 0.0006985, "grad_norm": 1.8012157678604126, "learning_rate": 6.984e-05, "loss": 0.3384, "step": 6985 }, { "epoch": 0.0006986, "grad_norm": 0.7090801000595093, "learning_rate": 6.985e-05, "loss": 0.2628, "step": 6986 }, { "epoch": 0.0006987, "grad_norm": 0.6993853449821472, "learning_rate": 6.986e-05, "loss": 0.2568, "step": 6987 }, { "epoch": 0.0006988, "grad_norm": 0.6504535675048828, "learning_rate": 6.987000000000001e-05, "loss": 0.2583, "step": 6988 }, { "epoch": 0.0006989, "grad_norm": 0.553688645362854, "learning_rate": 6.988e-05, "loss": 0.2521, "step": 6989 }, { "epoch": 0.000699, "grad_norm": 0.6234272122383118, "learning_rate": 6.989e-05, "loss": 0.2565, "step": 6990 }, { "epoch": 0.0006991, "grad_norm": 0.7927871942520142, "learning_rate": 6.99e-05, "loss": 0.3008, "step": 6991 }, { "epoch": 0.0006992, "grad_norm": 0.5891395211219788, "learning_rate": 6.991e-05, "loss": 0.2466, "step": 6992 }, { "epoch": 0.0006993, "grad_norm": 0.6059378981590271, "learning_rate": 6.992e-05, "loss": 0.2577, "step": 6993 }, { "epoch": 0.0006994, "grad_norm": 1.2688783407211304, "learning_rate": 6.993e-05, "loss": 0.3342, "step": 6994 }, { "epoch": 0.0006995, "grad_norm": 0.7680343389511108, "learning_rate": 6.994e-05, "loss": 0.2557, "step": 6995 }, { "epoch": 0.0006996, "grad_norm": 0.6820274591445923, "learning_rate": 6.995e-05, "loss": 0.2612, "step": 6996 }, { "epoch": 0.0006997, "grad_norm": 0.7284886837005615, "learning_rate": 6.996e-05, "loss": 0.2581, "step": 6997 }, { "epoch": 0.0006998, "grad_norm": 1.973966360092163, "learning_rate": 6.997000000000001e-05, "loss": 0.4316, "step": 6998 }, { "epoch": 0.0006999, "grad_norm": 0.5775460004806519, "learning_rate": 6.998e-05, "loss": 0.2427, "step": 6999 }, { "epoch": 0.0007, "grad_norm": 4.808237075805664, "learning_rate": 6.999e-05, "loss": 0.4492, "step": 7000 }, { "epoch": 0.0007, "eval_loss": 0.030403021723031998, "eval_runtime": 362.4086, "eval_samples_per_second": 27.593, "eval_steps_per_second": 1.725, "step": 7000 }, { "epoch": 0.0007001, "grad_norm": 0.8307585716247559, "learning_rate": 7.000000000000001e-05, "loss": 0.2679, "step": 7001 }, { "epoch": 0.0007002, "grad_norm": 0.6910313367843628, "learning_rate": 7.001e-05, "loss": 0.2467, "step": 7002 }, { "epoch": 0.0007003, "grad_norm": 0.9133024215698242, "learning_rate": 7.002e-05, "loss": 0.3354, "step": 7003 }, { "epoch": 0.0007004, "grad_norm": 0.7538177967071533, "learning_rate": 7.003e-05, "loss": 0.2761, "step": 7004 }, { "epoch": 0.0007005, "grad_norm": 0.6836936473846436, "learning_rate": 7.004e-05, "loss": 0.2589, "step": 7005 }, { "epoch": 0.0007006, "grad_norm": 0.6986979246139526, "learning_rate": 7.005e-05, "loss": 0.2457, "step": 7006 }, { "epoch": 0.0007007, "grad_norm": 0.7378904223442078, "learning_rate": 7.006e-05, "loss": 0.272, "step": 7007 }, { "epoch": 0.0007008, "grad_norm": 0.6308010816574097, "learning_rate": 7.007000000000001e-05, "loss": 0.2618, "step": 7008 }, { "epoch": 0.0007009, "grad_norm": 1.998252034187317, "learning_rate": 7.007999999999999e-05, "loss": 0.2814, "step": 7009 }, { "epoch": 0.000701, "grad_norm": 0.829488217830658, "learning_rate": 7.009e-05, "loss": 0.2583, "step": 7010 }, { "epoch": 0.0007011, "grad_norm": 1.1512203216552734, "learning_rate": 7.010000000000001e-05, "loss": 0.3521, "step": 7011 }, { "epoch": 0.0007012, "grad_norm": 0.7830899953842163, "learning_rate": 7.010999999999999e-05, "loss": 0.2651, "step": 7012 }, { "epoch": 0.0007013, "grad_norm": 0.6934369802474976, "learning_rate": 7.012e-05, "loss": 0.261, "step": 7013 }, { "epoch": 0.0007014, "grad_norm": 0.668596625328064, "learning_rate": 7.013000000000001e-05, "loss": 0.2466, "step": 7014 }, { "epoch": 0.0007015, "grad_norm": 0.756309449672699, "learning_rate": 7.014e-05, "loss": 0.2596, "step": 7015 }, { "epoch": 0.0007016, "grad_norm": 0.6385685801506042, "learning_rate": 7.015e-05, "loss": 0.2443, "step": 7016 }, { "epoch": 0.0007017, "grad_norm": 0.7099800109863281, "learning_rate": 7.016e-05, "loss": 0.2765, "step": 7017 }, { "epoch": 0.0007018, "grad_norm": 0.6158483028411865, "learning_rate": 7.017e-05, "loss": 0.2483, "step": 7018 }, { "epoch": 0.0007019, "grad_norm": 0.8103829622268677, "learning_rate": 7.018e-05, "loss": 0.2808, "step": 7019 }, { "epoch": 0.000702, "grad_norm": 0.6360248923301697, "learning_rate": 7.019e-05, "loss": 0.2416, "step": 7020 }, { "epoch": 0.0007021, "grad_norm": 0.6225180625915527, "learning_rate": 7.02e-05, "loss": 0.2439, "step": 7021 }, { "epoch": 0.0007022, "grad_norm": 0.5726187825202942, "learning_rate": 7.021000000000001e-05, "loss": 0.2388, "step": 7022 }, { "epoch": 0.0007023, "grad_norm": 0.562736988067627, "learning_rate": 7.022e-05, "loss": 0.2426, "step": 7023 }, { "epoch": 0.0007024, "grad_norm": 0.6522629261016846, "learning_rate": 7.023e-05, "loss": 0.259, "step": 7024 }, { "epoch": 0.0007025, "grad_norm": 0.5590613484382629, "learning_rate": 7.024e-05, "loss": 0.2461, "step": 7025 }, { "epoch": 0.0007026, "grad_norm": 1.0506471395492554, "learning_rate": 7.025e-05, "loss": 0.3125, "step": 7026 }, { "epoch": 0.0007027, "grad_norm": 0.7463265061378479, "learning_rate": 7.026e-05, "loss": 0.2753, "step": 7027 }, { "epoch": 0.0007028, "grad_norm": 1.6071248054504395, "learning_rate": 7.027e-05, "loss": 0.2897, "step": 7028 }, { "epoch": 0.0007029, "grad_norm": 0.69710373878479, "learning_rate": 7.028e-05, "loss": 0.2585, "step": 7029 }, { "epoch": 0.000703, "grad_norm": 0.6717625260353088, "learning_rate": 7.029e-05, "loss": 0.2501, "step": 7030 }, { "epoch": 0.0007031, "grad_norm": 0.8978825807571411, "learning_rate": 7.03e-05, "loss": 0.2637, "step": 7031 }, { "epoch": 0.0007032, "grad_norm": 0.5704598426818848, "learning_rate": 7.031000000000001e-05, "loss": 0.2299, "step": 7032 }, { "epoch": 0.0007033, "grad_norm": 0.997661828994751, "learning_rate": 7.032e-05, "loss": 0.2888, "step": 7033 }, { "epoch": 0.0007034, "grad_norm": 0.6185047030448914, "learning_rate": 7.033e-05, "loss": 0.2429, "step": 7034 }, { "epoch": 0.0007035, "grad_norm": 0.7629363536834717, "learning_rate": 7.034000000000001e-05, "loss": 0.2982, "step": 7035 }, { "epoch": 0.0007036, "grad_norm": 2.4536244869232178, "learning_rate": 7.035e-05, "loss": 0.4689, "step": 7036 }, { "epoch": 0.0007037, "grad_norm": 0.7827295064926147, "learning_rate": 7.036e-05, "loss": 0.2858, "step": 7037 }, { "epoch": 0.0007038, "grad_norm": 0.6093035936355591, "learning_rate": 7.037e-05, "loss": 0.2321, "step": 7038 }, { "epoch": 0.0007039, "grad_norm": 0.6487656831741333, "learning_rate": 7.038e-05, "loss": 0.2454, "step": 7039 }, { "epoch": 0.000704, "grad_norm": 0.9772380590438843, "learning_rate": 7.039e-05, "loss": 0.2457, "step": 7040 }, { "epoch": 0.0007041, "grad_norm": 0.6265672445297241, "learning_rate": 7.04e-05, "loss": 0.2416, "step": 7041 }, { "epoch": 0.0007042, "grad_norm": 0.7439896464347839, "learning_rate": 7.041000000000001e-05, "loss": 0.2644, "step": 7042 }, { "epoch": 0.0007043, "grad_norm": 0.7880527973175049, "learning_rate": 7.042e-05, "loss": 0.2607, "step": 7043 }, { "epoch": 0.0007044, "grad_norm": 0.5661341547966003, "learning_rate": 7.043e-05, "loss": 0.2383, "step": 7044 }, { "epoch": 0.0007045, "grad_norm": 0.5856346487998962, "learning_rate": 7.044000000000001e-05, "loss": 0.2411, "step": 7045 }, { "epoch": 0.0007046, "grad_norm": 0.5938259959220886, "learning_rate": 7.044999999999999e-05, "loss": 0.2565, "step": 7046 }, { "epoch": 0.0007047, "grad_norm": 0.5852174758911133, "learning_rate": 7.046e-05, "loss": 0.2401, "step": 7047 }, { "epoch": 0.0007048, "grad_norm": 0.6033628582954407, "learning_rate": 7.047000000000001e-05, "loss": 0.2523, "step": 7048 }, { "epoch": 0.0007049, "grad_norm": 0.6606107950210571, "learning_rate": 7.048e-05, "loss": 0.2605, "step": 7049 }, { "epoch": 0.000705, "grad_norm": 0.9846977591514587, "learning_rate": 7.049e-05, "loss": 0.2983, "step": 7050 }, { "epoch": 0.0007051, "grad_norm": 0.6883967518806458, "learning_rate": 7.05e-05, "loss": 0.2708, "step": 7051 }, { "epoch": 0.0007052, "grad_norm": 1.5532714128494263, "learning_rate": 7.051e-05, "loss": 0.3688, "step": 7052 }, { "epoch": 0.0007053, "grad_norm": 0.6477527022361755, "learning_rate": 7.052e-05, "loss": 0.25, "step": 7053 }, { "epoch": 0.0007054, "grad_norm": 2.8588967323303223, "learning_rate": 7.053e-05, "loss": 0.4503, "step": 7054 }, { "epoch": 0.0007055, "grad_norm": 0.6326694488525391, "learning_rate": 7.054e-05, "loss": 0.2188, "step": 7055 }, { "epoch": 0.0007056, "grad_norm": 0.803915798664093, "learning_rate": 7.055000000000001e-05, "loss": 0.2584, "step": 7056 }, { "epoch": 0.0007057, "grad_norm": 0.9331216216087341, "learning_rate": 7.056e-05, "loss": 0.3292, "step": 7057 }, { "epoch": 0.0007058, "grad_norm": 0.5624357461929321, "learning_rate": 7.057e-05, "loss": 0.2285, "step": 7058 }, { "epoch": 0.0007059, "grad_norm": 0.7122544646263123, "learning_rate": 7.058e-05, "loss": 0.2476, "step": 7059 }, { "epoch": 0.000706, "grad_norm": 0.6760504841804504, "learning_rate": 7.059e-05, "loss": 0.2561, "step": 7060 }, { "epoch": 0.0007061, "grad_norm": 0.6072905659675598, "learning_rate": 7.06e-05, "loss": 0.2404, "step": 7061 }, { "epoch": 0.0007062, "grad_norm": 0.7852693796157837, "learning_rate": 7.061e-05, "loss": 0.2744, "step": 7062 }, { "epoch": 0.0007063, "grad_norm": 0.6445418000221252, "learning_rate": 7.062e-05, "loss": 0.2714, "step": 7063 }, { "epoch": 0.0007064, "grad_norm": 0.9340659976005554, "learning_rate": 7.063e-05, "loss": 0.2786, "step": 7064 }, { "epoch": 0.0007065, "grad_norm": 0.596617579460144, "learning_rate": 7.064e-05, "loss": 0.2374, "step": 7065 }, { "epoch": 0.0007066, "grad_norm": 0.5787464380264282, "learning_rate": 7.065000000000001e-05, "loss": 0.2371, "step": 7066 }, { "epoch": 0.0007067, "grad_norm": 0.598484456539154, "learning_rate": 7.066e-05, "loss": 0.2352, "step": 7067 }, { "epoch": 0.0007068, "grad_norm": 0.633368730545044, "learning_rate": 7.067e-05, "loss": 0.2314, "step": 7068 }, { "epoch": 0.0007069, "grad_norm": 0.5488775968551636, "learning_rate": 7.068000000000001e-05, "loss": 0.2299, "step": 7069 }, { "epoch": 0.000707, "grad_norm": 0.5350727438926697, "learning_rate": 7.069e-05, "loss": 0.2345, "step": 7070 }, { "epoch": 0.0007071, "grad_norm": 0.6432333588600159, "learning_rate": 7.07e-05, "loss": 0.2474, "step": 7071 }, { "epoch": 0.0007072, "grad_norm": 0.60258948802948, "learning_rate": 7.071e-05, "loss": 0.2483, "step": 7072 }, { "epoch": 0.0007073, "grad_norm": 0.7571836113929749, "learning_rate": 7.072e-05, "loss": 0.2815, "step": 7073 }, { "epoch": 0.0007074, "grad_norm": 0.8476506471633911, "learning_rate": 7.073e-05, "loss": 0.3021, "step": 7074 }, { "epoch": 0.0007075, "grad_norm": 0.5610617399215698, "learning_rate": 7.074e-05, "loss": 0.2454, "step": 7075 }, { "epoch": 0.0007076, "grad_norm": 0.7464528679847717, "learning_rate": 7.075000000000001e-05, "loss": 0.2898, "step": 7076 }, { "epoch": 0.0007077, "grad_norm": 0.5852903723716736, "learning_rate": 7.076e-05, "loss": 0.2452, "step": 7077 }, { "epoch": 0.0007078, "grad_norm": 0.5367820858955383, "learning_rate": 7.077e-05, "loss": 0.2278, "step": 7078 }, { "epoch": 0.0007079, "grad_norm": 0.547356903553009, "learning_rate": 7.078000000000001e-05, "loss": 0.2323, "step": 7079 }, { "epoch": 0.000708, "grad_norm": 0.5330636501312256, "learning_rate": 7.078999999999999e-05, "loss": 0.234, "step": 7080 }, { "epoch": 0.0007081, "grad_norm": 0.5424564480781555, "learning_rate": 7.08e-05, "loss": 0.2311, "step": 7081 }, { "epoch": 0.0007082, "grad_norm": 0.5401227474212646, "learning_rate": 7.081000000000001e-05, "loss": 0.2405, "step": 7082 }, { "epoch": 0.0007083, "grad_norm": 1.2282289266586304, "learning_rate": 7.082e-05, "loss": 0.2833, "step": 7083 }, { "epoch": 0.0007084, "grad_norm": 1.706687092781067, "learning_rate": 7.083e-05, "loss": 0.3715, "step": 7084 }, { "epoch": 0.0007085, "grad_norm": 0.6320382356643677, "learning_rate": 7.084e-05, "loss": 0.2365, "step": 7085 }, { "epoch": 0.0007086, "grad_norm": 1.0006563663482666, "learning_rate": 7.085e-05, "loss": 0.3013, "step": 7086 }, { "epoch": 0.0007087, "grad_norm": 0.591720461845398, "learning_rate": 7.086e-05, "loss": 0.2504, "step": 7087 }, { "epoch": 0.0007088, "grad_norm": 0.5655443668365479, "learning_rate": 7.087e-05, "loss": 0.2433, "step": 7088 }, { "epoch": 0.0007089, "grad_norm": 0.6129018068313599, "learning_rate": 7.088e-05, "loss": 0.2415, "step": 7089 }, { "epoch": 0.000709, "grad_norm": 0.8806246519088745, "learning_rate": 7.089000000000001e-05, "loss": 0.3131, "step": 7090 }, { "epoch": 0.0007091, "grad_norm": 3.66709041595459, "learning_rate": 7.09e-05, "loss": 0.4194, "step": 7091 }, { "epoch": 0.0007092, "grad_norm": 0.6629515290260315, "learning_rate": 7.091e-05, "loss": 0.2397, "step": 7092 }, { "epoch": 0.0007093, "grad_norm": 1.096230149269104, "learning_rate": 7.092e-05, "loss": 0.2515, "step": 7093 }, { "epoch": 0.0007094, "grad_norm": 0.6950176358222961, "learning_rate": 7.093e-05, "loss": 0.2548, "step": 7094 }, { "epoch": 0.0007095, "grad_norm": 1.2441452741622925, "learning_rate": 7.094e-05, "loss": 0.3682, "step": 7095 }, { "epoch": 0.0007096, "grad_norm": 0.716272234916687, "learning_rate": 7.095e-05, "loss": 0.2465, "step": 7096 }, { "epoch": 0.0007097, "grad_norm": 1.6676087379455566, "learning_rate": 7.096e-05, "loss": 0.3446, "step": 7097 }, { "epoch": 0.0007098, "grad_norm": 0.9634347558021545, "learning_rate": 7.097e-05, "loss": 0.2997, "step": 7098 }, { "epoch": 0.0007099, "grad_norm": 0.9347872138023376, "learning_rate": 7.098e-05, "loss": 0.2888, "step": 7099 }, { "epoch": 0.00071, "grad_norm": 0.8800885677337646, "learning_rate": 7.099000000000001e-05, "loss": 0.2628, "step": 7100 }, { "epoch": 0.0007101, "grad_norm": 3.0393545627593994, "learning_rate": 7.1e-05, "loss": 0.3553, "step": 7101 }, { "epoch": 0.0007102, "grad_norm": 0.7605924606323242, "learning_rate": 7.101e-05, "loss": 0.2283, "step": 7102 }, { "epoch": 0.0007103, "grad_norm": 0.9857915043830872, "learning_rate": 7.102000000000001e-05, "loss": 0.2615, "step": 7103 }, { "epoch": 0.0007104, "grad_norm": 0.6909942626953125, "learning_rate": 7.103e-05, "loss": 0.2242, "step": 7104 }, { "epoch": 0.0007105, "grad_norm": 0.8127199411392212, "learning_rate": 7.104e-05, "loss": 0.2698, "step": 7105 }, { "epoch": 0.0007106, "grad_norm": 0.6459398865699768, "learning_rate": 7.105e-05, "loss": 0.2235, "step": 7106 }, { "epoch": 0.0007107, "grad_norm": 0.9567082524299622, "learning_rate": 7.106e-05, "loss": 0.3097, "step": 7107 }, { "epoch": 0.0007108, "grad_norm": 0.6782042980194092, "learning_rate": 7.107e-05, "loss": 0.2263, "step": 7108 }, { "epoch": 0.0007109, "grad_norm": 0.6514846086502075, "learning_rate": 7.108e-05, "loss": 0.2249, "step": 7109 }, { "epoch": 0.000711, "grad_norm": 1.0211524963378906, "learning_rate": 7.109e-05, "loss": 0.308, "step": 7110 }, { "epoch": 0.0007111, "grad_norm": 0.6992805600166321, "learning_rate": 7.11e-05, "loss": 0.244, "step": 7111 }, { "epoch": 0.0007112, "grad_norm": 0.6797493696212769, "learning_rate": 7.111e-05, "loss": 0.2457, "step": 7112 }, { "epoch": 0.0007113, "grad_norm": 0.6276497840881348, "learning_rate": 7.112000000000001e-05, "loss": 0.2314, "step": 7113 }, { "epoch": 0.0007114, "grad_norm": 0.657174825668335, "learning_rate": 7.112999999999999e-05, "loss": 0.2533, "step": 7114 }, { "epoch": 0.0007115, "grad_norm": 1.1678730249404907, "learning_rate": 7.114e-05, "loss": 0.2463, "step": 7115 }, { "epoch": 0.0007116, "grad_norm": 0.6611289978027344, "learning_rate": 7.115000000000001e-05, "loss": 0.2554, "step": 7116 }, { "epoch": 0.0007117, "grad_norm": 2.7565414905548096, "learning_rate": 7.116e-05, "loss": 0.439, "step": 7117 }, { "epoch": 0.0007118, "grad_norm": 0.9948599338531494, "learning_rate": 7.117e-05, "loss": 0.3165, "step": 7118 }, { "epoch": 0.0007119, "grad_norm": 0.6232846975326538, "learning_rate": 7.118e-05, "loss": 0.2372, "step": 7119 }, { "epoch": 0.000712, "grad_norm": 0.6444941163063049, "learning_rate": 7.119e-05, "loss": 0.2466, "step": 7120 }, { "epoch": 0.0007121, "grad_norm": 0.6959469318389893, "learning_rate": 7.12e-05, "loss": 0.2446, "step": 7121 }, { "epoch": 0.0007122, "grad_norm": 0.6532331109046936, "learning_rate": 7.121e-05, "loss": 0.2543, "step": 7122 }, { "epoch": 0.0007123, "grad_norm": 1.486360788345337, "learning_rate": 7.122e-05, "loss": 0.338, "step": 7123 }, { "epoch": 0.0007124, "grad_norm": 0.580660343170166, "learning_rate": 7.123000000000001e-05, "loss": 0.2288, "step": 7124 }, { "epoch": 0.0007125, "grad_norm": 0.7176545858383179, "learning_rate": 7.124e-05, "loss": 0.2649, "step": 7125 }, { "epoch": 0.0007126, "grad_norm": 0.7158398032188416, "learning_rate": 7.125e-05, "loss": 0.2253, "step": 7126 }, { "epoch": 0.0007127, "grad_norm": 0.5990284085273743, "learning_rate": 7.126e-05, "loss": 0.2277, "step": 7127 }, { "epoch": 0.0007128, "grad_norm": 0.6648378372192383, "learning_rate": 7.127e-05, "loss": 0.2471, "step": 7128 }, { "epoch": 0.0007129, "grad_norm": 0.6481736302375793, "learning_rate": 7.128e-05, "loss": 0.2374, "step": 7129 }, { "epoch": 0.000713, "grad_norm": 7.419454097747803, "learning_rate": 7.129e-05, "loss": 0.6338, "step": 7130 }, { "epoch": 0.0007131, "grad_norm": 0.5747849345207214, "learning_rate": 7.13e-05, "loss": 0.2126, "step": 7131 }, { "epoch": 0.0007132, "grad_norm": 0.6667564511299133, "learning_rate": 7.130999999999999e-05, "loss": 0.2419, "step": 7132 }, { "epoch": 0.0007133, "grad_norm": 0.5959503650665283, "learning_rate": 7.132e-05, "loss": 0.2371, "step": 7133 }, { "epoch": 0.0007134, "grad_norm": 0.6175190210342407, "learning_rate": 7.133000000000001e-05, "loss": 0.2394, "step": 7134 }, { "epoch": 0.0007135, "grad_norm": 0.5732832551002502, "learning_rate": 7.134e-05, "loss": 0.2313, "step": 7135 }, { "epoch": 0.0007136, "grad_norm": 0.576556384563446, "learning_rate": 7.135e-05, "loss": 0.2311, "step": 7136 }, { "epoch": 0.0007137, "grad_norm": 0.5220944881439209, "learning_rate": 7.136000000000001e-05, "loss": 0.2162, "step": 7137 }, { "epoch": 0.0007138, "grad_norm": 0.5937696695327759, "learning_rate": 7.137e-05, "loss": 0.2471, "step": 7138 }, { "epoch": 0.0007139, "grad_norm": 0.6484452486038208, "learning_rate": 7.138e-05, "loss": 0.2611, "step": 7139 }, { "epoch": 0.000714, "grad_norm": 0.6111674904823303, "learning_rate": 7.139e-05, "loss": 0.2383, "step": 7140 }, { "epoch": 0.0007141, "grad_norm": 2.137749195098877, "learning_rate": 7.14e-05, "loss": 0.4233, "step": 7141 }, { "epoch": 0.0007142, "grad_norm": 0.727592945098877, "learning_rate": 7.141e-05, "loss": 0.2676, "step": 7142 }, { "epoch": 0.0007143, "grad_norm": 0.6484252214431763, "learning_rate": 7.142e-05, "loss": 0.2339, "step": 7143 }, { "epoch": 0.0007144, "grad_norm": 1.0977895259857178, "learning_rate": 7.143e-05, "loss": 0.2665, "step": 7144 }, { "epoch": 0.0007145, "grad_norm": 1.1723270416259766, "learning_rate": 7.144e-05, "loss": 0.3542, "step": 7145 }, { "epoch": 0.0007146, "grad_norm": 1.385094404220581, "learning_rate": 7.145e-05, "loss": 0.3069, "step": 7146 }, { "epoch": 0.0007147, "grad_norm": 0.8112697601318359, "learning_rate": 7.146000000000001e-05, "loss": 0.2416, "step": 7147 }, { "epoch": 0.0007148, "grad_norm": 0.8700363039970398, "learning_rate": 7.146999999999999e-05, "loss": 0.2625, "step": 7148 }, { "epoch": 0.0007149, "grad_norm": 0.7144835591316223, "learning_rate": 7.148e-05, "loss": 0.2773, "step": 7149 }, { "epoch": 0.000715, "grad_norm": 0.6032698750495911, "learning_rate": 7.149000000000001e-05, "loss": 0.2333, "step": 7150 }, { "epoch": 0.0007151, "grad_norm": 0.7318822741508484, "learning_rate": 7.15e-05, "loss": 0.2494, "step": 7151 }, { "epoch": 0.0007152, "grad_norm": 0.5837897658348083, "learning_rate": 7.151e-05, "loss": 0.2097, "step": 7152 }, { "epoch": 0.0007153, "grad_norm": 0.6239914298057556, "learning_rate": 7.152e-05, "loss": 0.2346, "step": 7153 }, { "epoch": 0.0007154, "grad_norm": 0.5616824626922607, "learning_rate": 7.153e-05, "loss": 0.2317, "step": 7154 }, { "epoch": 0.0007155, "grad_norm": 4.966375350952148, "learning_rate": 7.154e-05, "loss": 0.4221, "step": 7155 }, { "epoch": 0.0007156, "grad_norm": 0.6621120572090149, "learning_rate": 7.155e-05, "loss": 0.2323, "step": 7156 }, { "epoch": 0.0007157, "grad_norm": 0.7091697454452515, "learning_rate": 7.156e-05, "loss": 0.2356, "step": 7157 }, { "epoch": 0.0007158, "grad_norm": 0.6042245030403137, "learning_rate": 7.157000000000001e-05, "loss": 0.2158, "step": 7158 }, { "epoch": 0.0007159, "grad_norm": 0.6950005292892456, "learning_rate": 7.158e-05, "loss": 0.2396, "step": 7159 }, { "epoch": 0.000716, "grad_norm": 0.9411141276359558, "learning_rate": 7.159e-05, "loss": 0.2965, "step": 7160 }, { "epoch": 0.0007161, "grad_norm": 2.314314603805542, "learning_rate": 7.16e-05, "loss": 0.3401, "step": 7161 }, { "epoch": 0.0007162, "grad_norm": 0.6668578386306763, "learning_rate": 7.161e-05, "loss": 0.2271, "step": 7162 }, { "epoch": 0.0007163, "grad_norm": 1.7270629405975342, "learning_rate": 7.162e-05, "loss": 0.3015, "step": 7163 }, { "epoch": 0.0007164, "grad_norm": 0.7965278029441833, "learning_rate": 7.163e-05, "loss": 0.2498, "step": 7164 }, { "epoch": 0.0007165, "grad_norm": 0.7627633213996887, "learning_rate": 7.164e-05, "loss": 0.2413, "step": 7165 }, { "epoch": 0.0007166, "grad_norm": 0.6501511931419373, "learning_rate": 7.165e-05, "loss": 0.2357, "step": 7166 }, { "epoch": 0.0007167, "grad_norm": 0.5962719917297363, "learning_rate": 7.166e-05, "loss": 0.2275, "step": 7167 }, { "epoch": 0.0007168, "grad_norm": 0.585323691368103, "learning_rate": 7.167000000000001e-05, "loss": 0.2277, "step": 7168 }, { "epoch": 0.0007169, "grad_norm": 0.6851444840431213, "learning_rate": 7.168e-05, "loss": 0.2325, "step": 7169 }, { "epoch": 0.000717, "grad_norm": 0.6078755259513855, "learning_rate": 7.169e-05, "loss": 0.2395, "step": 7170 }, { "epoch": 0.0007171, "grad_norm": 0.574161946773529, "learning_rate": 7.170000000000001e-05, "loss": 0.2308, "step": 7171 }, { "epoch": 0.0007172, "grad_norm": 0.5724591016769409, "learning_rate": 7.171e-05, "loss": 0.2325, "step": 7172 }, { "epoch": 0.0007173, "grad_norm": 0.5637148022651672, "learning_rate": 7.172e-05, "loss": 0.2264, "step": 7173 }, { "epoch": 0.0007174, "grad_norm": 0.6590225100517273, "learning_rate": 7.173e-05, "loss": 0.2529, "step": 7174 }, { "epoch": 0.0007175, "grad_norm": 0.5135878324508667, "learning_rate": 7.174e-05, "loss": 0.2188, "step": 7175 }, { "epoch": 0.0007176, "grad_norm": 0.6357454657554626, "learning_rate": 7.175e-05, "loss": 0.2395, "step": 7176 }, { "epoch": 0.0007177, "grad_norm": 1.0965118408203125, "learning_rate": 7.176e-05, "loss": 0.2622, "step": 7177 }, { "epoch": 0.0007178, "grad_norm": 0.7161939740180969, "learning_rate": 7.177e-05, "loss": 0.2527, "step": 7178 }, { "epoch": 0.0007179, "grad_norm": 2.47273588180542, "learning_rate": 7.178e-05, "loss": 0.3, "step": 7179 }, { "epoch": 0.000718, "grad_norm": 0.6954152584075928, "learning_rate": 7.179e-05, "loss": 0.2368, "step": 7180 }, { "epoch": 0.0007181, "grad_norm": 0.7640717625617981, "learning_rate": 7.180000000000001e-05, "loss": 0.2489, "step": 7181 }, { "epoch": 0.0007182, "grad_norm": 0.6073949933052063, "learning_rate": 7.180999999999999e-05, "loss": 0.21, "step": 7182 }, { "epoch": 0.0007183, "grad_norm": 0.5503464937210083, "learning_rate": 7.182e-05, "loss": 0.2146, "step": 7183 }, { "epoch": 0.0007184, "grad_norm": 0.5253496766090393, "learning_rate": 7.183000000000001e-05, "loss": 0.2119, "step": 7184 }, { "epoch": 0.0007185, "grad_norm": 0.6075765490531921, "learning_rate": 7.184e-05, "loss": 0.2233, "step": 7185 }, { "epoch": 0.0007186, "grad_norm": 0.7229404449462891, "learning_rate": 7.185e-05, "loss": 0.2375, "step": 7186 }, { "epoch": 0.0007187, "grad_norm": 0.6995594501495361, "learning_rate": 7.186e-05, "loss": 0.2399, "step": 7187 }, { "epoch": 0.0007188, "grad_norm": 0.5980567336082458, "learning_rate": 7.187e-05, "loss": 0.2341, "step": 7188 }, { "epoch": 0.0007189, "grad_norm": 0.7101425528526306, "learning_rate": 7.188e-05, "loss": 0.2684, "step": 7189 }, { "epoch": 0.000719, "grad_norm": 0.6097431778907776, "learning_rate": 7.189e-05, "loss": 0.2347, "step": 7190 }, { "epoch": 0.0007191, "grad_norm": 0.5196309685707092, "learning_rate": 7.19e-05, "loss": 0.2109, "step": 7191 }, { "epoch": 0.0007192, "grad_norm": 0.6105871796607971, "learning_rate": 7.191000000000001e-05, "loss": 0.2308, "step": 7192 }, { "epoch": 0.0007193, "grad_norm": 1.5595483779907227, "learning_rate": 7.192e-05, "loss": 0.2831, "step": 7193 }, { "epoch": 0.0007194, "grad_norm": 0.5544551014900208, "learning_rate": 7.193e-05, "loss": 0.22, "step": 7194 }, { "epoch": 0.0007195, "grad_norm": 0.5593705177307129, "learning_rate": 7.194e-05, "loss": 0.2013, "step": 7195 }, { "epoch": 0.0007196, "grad_norm": 0.6130393743515015, "learning_rate": 7.195e-05, "loss": 0.2268, "step": 7196 }, { "epoch": 0.0007197, "grad_norm": 0.5510294437408447, "learning_rate": 7.196e-05, "loss": 0.2168, "step": 7197 }, { "epoch": 0.0007198, "grad_norm": 0.5207967162132263, "learning_rate": 7.197e-05, "loss": 0.2158, "step": 7198 }, { "epoch": 0.0007199, "grad_norm": 0.6211857795715332, "learning_rate": 7.198e-05, "loss": 0.2415, "step": 7199 }, { "epoch": 0.00072, "grad_norm": 0.564426839351654, "learning_rate": 7.199e-05, "loss": 0.2104, "step": 7200 }, { "epoch": 0.0007201, "grad_norm": 0.6109751462936401, "learning_rate": 7.2e-05, "loss": 0.231, "step": 7201 }, { "epoch": 0.0007202, "grad_norm": 0.5517870783805847, "learning_rate": 7.201000000000001e-05, "loss": 0.2203, "step": 7202 }, { "epoch": 0.0007203, "grad_norm": 1.919884443283081, "learning_rate": 7.202e-05, "loss": 0.2815, "step": 7203 }, { "epoch": 0.0007204, "grad_norm": 0.5402631163597107, "learning_rate": 7.203e-05, "loss": 0.2339, "step": 7204 }, { "epoch": 0.0007205, "grad_norm": 0.6419543623924255, "learning_rate": 7.204000000000001e-05, "loss": 0.235, "step": 7205 }, { "epoch": 0.0007206, "grad_norm": 6.634181499481201, "learning_rate": 7.205e-05, "loss": 0.5928, "step": 7206 }, { "epoch": 0.0007207, "grad_norm": 1.0302337408065796, "learning_rate": 7.206e-05, "loss": 0.2416, "step": 7207 }, { "epoch": 0.0007208, "grad_norm": 0.6726213097572327, "learning_rate": 7.207e-05, "loss": 0.2227, "step": 7208 }, { "epoch": 0.0007209, "grad_norm": 0.6477869153022766, "learning_rate": 7.208e-05, "loss": 0.241, "step": 7209 }, { "epoch": 0.000721, "grad_norm": 0.5784698128700256, "learning_rate": 7.209e-05, "loss": 0.2212, "step": 7210 }, { "epoch": 0.0007211, "grad_norm": 0.47859689593315125, "learning_rate": 7.21e-05, "loss": 0.2018, "step": 7211 }, { "epoch": 0.0007212, "grad_norm": 0.646013081073761, "learning_rate": 7.211e-05, "loss": 0.2505, "step": 7212 }, { "epoch": 0.0007213, "grad_norm": 0.615487277507782, "learning_rate": 7.212e-05, "loss": 0.2482, "step": 7213 }, { "epoch": 0.0007214, "grad_norm": 0.9184458255767822, "learning_rate": 7.213e-05, "loss": 0.2767, "step": 7214 }, { "epoch": 0.0007215, "grad_norm": 0.5351276993751526, "learning_rate": 7.214000000000001e-05, "loss": 0.2139, "step": 7215 }, { "epoch": 0.0007216, "grad_norm": 0.520902693271637, "learning_rate": 7.214999999999999e-05, "loss": 0.2069, "step": 7216 }, { "epoch": 0.0007217, "grad_norm": 0.8827000856399536, "learning_rate": 7.216e-05, "loss": 0.2527, "step": 7217 }, { "epoch": 0.0007218, "grad_norm": 0.5077048540115356, "learning_rate": 7.217000000000001e-05, "loss": 0.2145, "step": 7218 }, { "epoch": 0.0007219, "grad_norm": 0.5010983347892761, "learning_rate": 7.218e-05, "loss": 0.2025, "step": 7219 }, { "epoch": 0.000722, "grad_norm": 0.5093920230865479, "learning_rate": 7.219e-05, "loss": 0.2124, "step": 7220 }, { "epoch": 0.0007221, "grad_norm": 0.5570715069770813, "learning_rate": 7.22e-05, "loss": 0.218, "step": 7221 }, { "epoch": 0.0007222, "grad_norm": 0.5382811427116394, "learning_rate": 7.221e-05, "loss": 0.2222, "step": 7222 }, { "epoch": 0.0007223, "grad_norm": 0.8405284881591797, "learning_rate": 7.222e-05, "loss": 0.2488, "step": 7223 }, { "epoch": 0.0007224, "grad_norm": 0.5121170282363892, "learning_rate": 7.223e-05, "loss": 0.2166, "step": 7224 }, { "epoch": 0.0007225, "grad_norm": 1.4797457456588745, "learning_rate": 7.224e-05, "loss": 0.2417, "step": 7225 }, { "epoch": 0.0007226, "grad_norm": 0.6273139715194702, "learning_rate": 7.225000000000001e-05, "loss": 0.2262, "step": 7226 }, { "epoch": 0.0007227, "grad_norm": 0.6362743973731995, "learning_rate": 7.226e-05, "loss": 0.2286, "step": 7227 }, { "epoch": 0.0007228, "grad_norm": 0.6842048764228821, "learning_rate": 7.227e-05, "loss": 0.224, "step": 7228 }, { "epoch": 0.0007229, "grad_norm": 0.5175364017486572, "learning_rate": 7.228e-05, "loss": 0.2073, "step": 7229 }, { "epoch": 0.000723, "grad_norm": 0.48860543966293335, "learning_rate": 7.229e-05, "loss": 0.2068, "step": 7230 }, { "epoch": 0.0007231, "grad_norm": 0.5222052335739136, "learning_rate": 7.23e-05, "loss": 0.2159, "step": 7231 }, { "epoch": 0.0007232, "grad_norm": 0.5209774971008301, "learning_rate": 7.231e-05, "loss": 0.2061, "step": 7232 }, { "epoch": 0.0007233, "grad_norm": 1.7664886713027954, "learning_rate": 7.232e-05, "loss": 0.3177, "step": 7233 }, { "epoch": 0.0007234, "grad_norm": 0.524284839630127, "learning_rate": 7.233e-05, "loss": 0.2126, "step": 7234 }, { "epoch": 0.0007235, "grad_norm": 0.5629242658615112, "learning_rate": 7.234e-05, "loss": 0.2191, "step": 7235 }, { "epoch": 0.0007236, "grad_norm": 0.5035473108291626, "learning_rate": 7.235000000000001e-05, "loss": 0.2058, "step": 7236 }, { "epoch": 0.0007237, "grad_norm": 0.5276579856872559, "learning_rate": 7.236e-05, "loss": 0.2151, "step": 7237 }, { "epoch": 0.0007238, "grad_norm": 0.48516181111335754, "learning_rate": 7.237e-05, "loss": 0.2122, "step": 7238 }, { "epoch": 0.0007239, "grad_norm": 0.5754404664039612, "learning_rate": 7.238000000000001e-05, "loss": 0.2198, "step": 7239 }, { "epoch": 0.000724, "grad_norm": 0.4860302209854126, "learning_rate": 7.239e-05, "loss": 0.2046, "step": 7240 }, { "epoch": 0.0007241, "grad_norm": 0.6111292243003845, "learning_rate": 7.24e-05, "loss": 0.2305, "step": 7241 }, { "epoch": 0.0007242, "grad_norm": 0.49134159088134766, "learning_rate": 7.241e-05, "loss": 0.2047, "step": 7242 }, { "epoch": 0.0007243, "grad_norm": 0.5437736511230469, "learning_rate": 7.242e-05, "loss": 0.2113, "step": 7243 }, { "epoch": 0.0007244, "grad_norm": 0.7377773523330688, "learning_rate": 7.243e-05, "loss": 0.2406, "step": 7244 }, { "epoch": 0.0007245, "grad_norm": 0.5279169082641602, "learning_rate": 7.244e-05, "loss": 0.2158, "step": 7245 }, { "epoch": 0.0007246, "grad_norm": 0.550542950630188, "learning_rate": 7.245e-05, "loss": 0.2227, "step": 7246 }, { "epoch": 0.0007247, "grad_norm": 0.5191932916641235, "learning_rate": 7.246e-05, "loss": 0.2205, "step": 7247 }, { "epoch": 0.0007248, "grad_norm": 1.221814751625061, "learning_rate": 7.247e-05, "loss": 0.3137, "step": 7248 }, { "epoch": 0.0007249, "grad_norm": 1.6500520706176758, "learning_rate": 7.248000000000001e-05, "loss": 0.2714, "step": 7249 }, { "epoch": 0.000725, "grad_norm": 0.680809497833252, "learning_rate": 7.248999999999999e-05, "loss": 0.2228, "step": 7250 }, { "epoch": 0.0007251, "grad_norm": 0.5752734541893005, "learning_rate": 7.25e-05, "loss": 0.203, "step": 7251 }, { "epoch": 0.0007252, "grad_norm": 0.6003434062004089, "learning_rate": 7.251000000000001e-05, "loss": 0.2244, "step": 7252 }, { "epoch": 0.0007253, "grad_norm": 0.5211772322654724, "learning_rate": 7.252e-05, "loss": 0.2092, "step": 7253 }, { "epoch": 0.0007254, "grad_norm": 0.5306843519210815, "learning_rate": 7.253e-05, "loss": 0.2216, "step": 7254 }, { "epoch": 0.0007255, "grad_norm": 0.48355337977409363, "learning_rate": 7.254000000000001e-05, "loss": 0.2002, "step": 7255 }, { "epoch": 0.0007256, "grad_norm": 0.7639783620834351, "learning_rate": 7.255e-05, "loss": 0.2357, "step": 7256 }, { "epoch": 0.0007257, "grad_norm": 0.6967151165008545, "learning_rate": 7.256e-05, "loss": 0.2445, "step": 7257 }, { "epoch": 0.0007258, "grad_norm": 0.5458248853683472, "learning_rate": 7.257e-05, "loss": 0.2205, "step": 7258 }, { "epoch": 0.0007259, "grad_norm": 0.6922258734703064, "learning_rate": 7.258e-05, "loss": 0.2216, "step": 7259 }, { "epoch": 0.000726, "grad_norm": 0.4929615557193756, "learning_rate": 7.259e-05, "loss": 0.2106, "step": 7260 }, { "epoch": 0.0007261, "grad_norm": 0.5160805583000183, "learning_rate": 7.26e-05, "loss": 0.2074, "step": 7261 }, { "epoch": 0.0007262, "grad_norm": 1.9402644634246826, "learning_rate": 7.261e-05, "loss": 0.2408, "step": 7262 }, { "epoch": 0.0007263, "grad_norm": 0.46201208233833313, "learning_rate": 7.262e-05, "loss": 0.1937, "step": 7263 }, { "epoch": 0.0007264, "grad_norm": 0.5219488143920898, "learning_rate": 7.263e-05, "loss": 0.2086, "step": 7264 }, { "epoch": 0.0007265, "grad_norm": 0.4784381687641144, "learning_rate": 7.264e-05, "loss": 0.2035, "step": 7265 }, { "epoch": 0.0007266, "grad_norm": 0.4866734743118286, "learning_rate": 7.265e-05, "loss": 0.2037, "step": 7266 }, { "epoch": 0.0007267, "grad_norm": 0.5169983506202698, "learning_rate": 7.266e-05, "loss": 0.2137, "step": 7267 }, { "epoch": 0.0007268, "grad_norm": 0.4718962013721466, "learning_rate": 7.267e-05, "loss": 0.199, "step": 7268 }, { "epoch": 0.0007269, "grad_norm": 0.521872341632843, "learning_rate": 7.268e-05, "loss": 0.2175, "step": 7269 }, { "epoch": 0.000727, "grad_norm": 0.49978336691856384, "learning_rate": 7.269000000000001e-05, "loss": 0.213, "step": 7270 }, { "epoch": 0.0007271, "grad_norm": 0.4791599214076996, "learning_rate": 7.27e-05, "loss": 0.2037, "step": 7271 }, { "epoch": 0.0007272, "grad_norm": 0.4703919589519501, "learning_rate": 7.271e-05, "loss": 0.1992, "step": 7272 }, { "epoch": 0.0007273, "grad_norm": 0.439566433429718, "learning_rate": 7.272000000000001e-05, "loss": 0.2018, "step": 7273 }, { "epoch": 0.0007274, "grad_norm": 0.47832438349723816, "learning_rate": 7.273e-05, "loss": 0.2087, "step": 7274 }, { "epoch": 0.0007275, "grad_norm": 0.9735659956932068, "learning_rate": 7.274e-05, "loss": 0.2181, "step": 7275 }, { "epoch": 0.0007276, "grad_norm": 0.4623475968837738, "learning_rate": 7.275e-05, "loss": 0.1947, "step": 7276 }, { "epoch": 0.0007277, "grad_norm": 0.4709840714931488, "learning_rate": 7.276e-05, "loss": 0.2057, "step": 7277 }, { "epoch": 0.0007278, "grad_norm": 0.49509432911872864, "learning_rate": 7.277e-05, "loss": 0.2023, "step": 7278 }, { "epoch": 0.0007279, "grad_norm": 0.5068236589431763, "learning_rate": 7.278e-05, "loss": 0.201, "step": 7279 }, { "epoch": 0.000728, "grad_norm": 0.7884330749511719, "learning_rate": 7.279e-05, "loss": 0.226, "step": 7280 }, { "epoch": 0.0007281, "grad_norm": 5.262802600860596, "learning_rate": 7.28e-05, "loss": 0.2993, "step": 7281 }, { "epoch": 0.0007282, "grad_norm": 0.5736418962478638, "learning_rate": 7.281e-05, "loss": 0.2168, "step": 7282 }, { "epoch": 0.0007283, "grad_norm": 3.2191948890686035, "learning_rate": 7.282000000000001e-05, "loss": 0.4016, "step": 7283 }, { "epoch": 0.0007284, "grad_norm": 0.6173503398895264, "learning_rate": 7.282999999999999e-05, "loss": 0.1929, "step": 7284 }, { "epoch": 0.0007285, "grad_norm": 0.8881523609161377, "learning_rate": 7.284e-05, "loss": 0.2728, "step": 7285 }, { "epoch": 0.0007286, "grad_norm": 0.500853419303894, "learning_rate": 7.285000000000001e-05, "loss": 0.1998, "step": 7286 }, { "epoch": 0.0007287, "grad_norm": 0.49271973967552185, "learning_rate": 7.286e-05, "loss": 0.1963, "step": 7287 }, { "epoch": 0.0007288, "grad_norm": 0.5352133512496948, "learning_rate": 7.287e-05, "loss": 0.208, "step": 7288 }, { "epoch": 0.0007289, "grad_norm": 0.7467477917671204, "learning_rate": 7.288000000000001e-05, "loss": 0.2314, "step": 7289 }, { "epoch": 0.000729, "grad_norm": 1.1657520532608032, "learning_rate": 7.289e-05, "loss": 0.2361, "step": 7290 }, { "epoch": 0.0007291, "grad_norm": 0.7892240881919861, "learning_rate": 7.29e-05, "loss": 0.2517, "step": 7291 }, { "epoch": 0.0007292, "grad_norm": 0.5541574358940125, "learning_rate": 7.291e-05, "loss": 0.1995, "step": 7292 }, { "epoch": 0.0007293, "grad_norm": 0.6174006462097168, "learning_rate": 7.292e-05, "loss": 0.1985, "step": 7293 }, { "epoch": 0.0007294, "grad_norm": 0.5258652567863464, "learning_rate": 7.293e-05, "loss": 0.1866, "step": 7294 }, { "epoch": 0.0007295, "grad_norm": 0.5781332850456238, "learning_rate": 7.294e-05, "loss": 0.208, "step": 7295 }, { "epoch": 0.0007296, "grad_norm": 0.5011702179908752, "learning_rate": 7.295e-05, "loss": 0.208, "step": 7296 }, { "epoch": 0.0007297, "grad_norm": 0.7021820545196533, "learning_rate": 7.296e-05, "loss": 0.2271, "step": 7297 }, { "epoch": 0.0007298, "grad_norm": 0.5188300609588623, "learning_rate": 7.297e-05, "loss": 0.2019, "step": 7298 }, { "epoch": 0.0007299, "grad_norm": 0.5135867595672607, "learning_rate": 7.298e-05, "loss": 0.1907, "step": 7299 }, { "epoch": 0.00073, "grad_norm": 0.47076326608657837, "learning_rate": 7.299e-05, "loss": 0.1943, "step": 7300 }, { "epoch": 0.0007301, "grad_norm": 0.8561511039733887, "learning_rate": 7.3e-05, "loss": 0.2634, "step": 7301 }, { "epoch": 0.0007302, "grad_norm": 0.41544100642204285, "learning_rate": 7.301e-05, "loss": 0.1854, "step": 7302 }, { "epoch": 0.0007303, "grad_norm": 0.4582894444465637, "learning_rate": 7.302e-05, "loss": 0.1927, "step": 7303 }, { "epoch": 0.0007304, "grad_norm": 0.5104141235351562, "learning_rate": 7.303000000000001e-05, "loss": 0.2021, "step": 7304 }, { "epoch": 0.0007305, "grad_norm": 0.6090421080589294, "learning_rate": 7.304e-05, "loss": 0.2289, "step": 7305 }, { "epoch": 0.0007306, "grad_norm": 0.485851526260376, "learning_rate": 7.305e-05, "loss": 0.1978, "step": 7306 }, { "epoch": 0.0007307, "grad_norm": 0.41749894618988037, "learning_rate": 7.306000000000001e-05, "loss": 0.1914, "step": 7307 }, { "epoch": 0.0007308, "grad_norm": 0.4175671935081482, "learning_rate": 7.307e-05, "loss": 0.1859, "step": 7308 }, { "epoch": 0.0007309, "grad_norm": 0.465059757232666, "learning_rate": 7.308e-05, "loss": 0.1995, "step": 7309 }, { "epoch": 0.000731, "grad_norm": 0.44294998049736023, "learning_rate": 7.309e-05, "loss": 0.198, "step": 7310 }, { "epoch": 0.0007311, "grad_norm": 0.4414581060409546, "learning_rate": 7.31e-05, "loss": 0.1803, "step": 7311 }, { "epoch": 0.0007312, "grad_norm": 0.4441189467906952, "learning_rate": 7.311e-05, "loss": 0.1949, "step": 7312 }, { "epoch": 0.0007313, "grad_norm": 0.4735046625137329, "learning_rate": 7.312e-05, "loss": 0.2164, "step": 7313 }, { "epoch": 0.0007314, "grad_norm": 0.4368090033531189, "learning_rate": 7.313e-05, "loss": 0.1993, "step": 7314 }, { "epoch": 0.0007315, "grad_norm": 0.42844894528388977, "learning_rate": 7.314e-05, "loss": 0.1925, "step": 7315 }, { "epoch": 0.0007316, "grad_norm": 0.39331182837486267, "learning_rate": 7.315e-05, "loss": 0.1896, "step": 7316 }, { "epoch": 0.0007317, "grad_norm": 0.4788910746574402, "learning_rate": 7.316000000000001e-05, "loss": 0.2067, "step": 7317 }, { "epoch": 0.0007318, "grad_norm": 0.4292828142642975, "learning_rate": 7.316999999999999e-05, "loss": 0.1964, "step": 7318 }, { "epoch": 0.0007319, "grad_norm": 0.4058398902416229, "learning_rate": 7.318e-05, "loss": 0.1982, "step": 7319 }, { "epoch": 0.000732, "grad_norm": 0.43642178177833557, "learning_rate": 7.319000000000001e-05, "loss": 0.2009, "step": 7320 }, { "epoch": 0.0007321, "grad_norm": 0.43849149346351624, "learning_rate": 7.32e-05, "loss": 0.1888, "step": 7321 }, { "epoch": 0.0007322, "grad_norm": 0.43522223830223083, "learning_rate": 7.321e-05, "loss": 0.2036, "step": 7322 }, { "epoch": 0.0007323, "grad_norm": 1.0789283514022827, "learning_rate": 7.322000000000001e-05, "loss": 0.2266, "step": 7323 }, { "epoch": 0.0007324, "grad_norm": 0.8794465065002441, "learning_rate": 7.323e-05, "loss": 0.255, "step": 7324 }, { "epoch": 0.0007325, "grad_norm": 0.5397830009460449, "learning_rate": 7.324e-05, "loss": 0.191, "step": 7325 }, { "epoch": 0.0007326, "grad_norm": 0.5121053457260132, "learning_rate": 7.325e-05, "loss": 0.2021, "step": 7326 }, { "epoch": 0.0007327, "grad_norm": 1.1845084428787231, "learning_rate": 7.326e-05, "loss": 0.2732, "step": 7327 }, { "epoch": 0.0007328, "grad_norm": 0.5448693037033081, "learning_rate": 7.327e-05, "loss": 0.2064, "step": 7328 }, { "epoch": 0.0007329, "grad_norm": 0.49568578600883484, "learning_rate": 7.328e-05, "loss": 0.1957, "step": 7329 }, { "epoch": 0.000733, "grad_norm": 0.5749213695526123, "learning_rate": 7.329e-05, "loss": 0.2073, "step": 7330 }, { "epoch": 0.0007331, "grad_norm": 0.653283953666687, "learning_rate": 7.33e-05, "loss": 0.2335, "step": 7331 }, { "epoch": 0.0007332, "grad_norm": 0.6415607929229736, "learning_rate": 7.331e-05, "loss": 0.2091, "step": 7332 }, { "epoch": 0.0007333, "grad_norm": 0.49434953927993774, "learning_rate": 7.332e-05, "loss": 0.208, "step": 7333 }, { "epoch": 0.0007334, "grad_norm": 0.5622429251670837, "learning_rate": 7.333e-05, "loss": 0.194, "step": 7334 }, { "epoch": 0.0007335, "grad_norm": 0.5433710813522339, "learning_rate": 7.334e-05, "loss": 0.196, "step": 7335 }, { "epoch": 0.0007336, "grad_norm": 0.5698725581169128, "learning_rate": 7.335e-05, "loss": 0.1993, "step": 7336 }, { "epoch": 0.0007337, "grad_norm": 0.6964457631111145, "learning_rate": 7.336e-05, "loss": 0.2349, "step": 7337 }, { "epoch": 0.0007338, "grad_norm": 0.44615158438682556, "learning_rate": 7.337000000000001e-05, "loss": 0.1914, "step": 7338 }, { "epoch": 0.0007339, "grad_norm": 0.5419948697090149, "learning_rate": 7.338e-05, "loss": 0.1934, "step": 7339 }, { "epoch": 0.000734, "grad_norm": 0.44522324204444885, "learning_rate": 7.339e-05, "loss": 0.1965, "step": 7340 }, { "epoch": 0.0007341, "grad_norm": 0.42922475934028625, "learning_rate": 7.340000000000001e-05, "loss": 0.1924, "step": 7341 }, { "epoch": 0.0007342, "grad_norm": 0.46745359897613525, "learning_rate": 7.341e-05, "loss": 0.1975, "step": 7342 }, { "epoch": 0.0007343, "grad_norm": 0.42670372128486633, "learning_rate": 7.342e-05, "loss": 0.1805, "step": 7343 }, { "epoch": 0.0007344, "grad_norm": 0.6460187435150146, "learning_rate": 7.343e-05, "loss": 0.213, "step": 7344 }, { "epoch": 0.0007345, "grad_norm": 1.687685251235962, "learning_rate": 7.344e-05, "loss": 0.2593, "step": 7345 }, { "epoch": 0.0007346, "grad_norm": 0.42223459482192993, "learning_rate": 7.345e-05, "loss": 0.1827, "step": 7346 }, { "epoch": 0.0007347, "grad_norm": 0.4287385046482086, "learning_rate": 7.346e-05, "loss": 0.1843, "step": 7347 }, { "epoch": 0.0007348, "grad_norm": 0.4938461184501648, "learning_rate": 7.347e-05, "loss": 0.1989, "step": 7348 }, { "epoch": 0.0007349, "grad_norm": 1.1912397146224976, "learning_rate": 7.348e-05, "loss": 0.2338, "step": 7349 }, { "epoch": 0.000735, "grad_norm": 0.525805652141571, "learning_rate": 7.349e-05, "loss": 0.2101, "step": 7350 }, { "epoch": 0.0007351, "grad_norm": 0.48151376843452454, "learning_rate": 7.35e-05, "loss": 0.2007, "step": 7351 }, { "epoch": 0.0007352, "grad_norm": 0.4943440556526184, "learning_rate": 7.350999999999999e-05, "loss": 0.1957, "step": 7352 }, { "epoch": 0.0007353, "grad_norm": 0.6150953769683838, "learning_rate": 7.352e-05, "loss": 0.1945, "step": 7353 }, { "epoch": 0.0007354, "grad_norm": 0.4299049377441406, "learning_rate": 7.353000000000001e-05, "loss": 0.181, "step": 7354 }, { "epoch": 0.0007355, "grad_norm": 0.46449828147888184, "learning_rate": 7.354e-05, "loss": 0.1918, "step": 7355 }, { "epoch": 0.0007356, "grad_norm": 0.4400832951068878, "learning_rate": 7.355e-05, "loss": 0.1927, "step": 7356 }, { "epoch": 0.0007357, "grad_norm": 0.5461811423301697, "learning_rate": 7.356000000000001e-05, "loss": 0.2039, "step": 7357 }, { "epoch": 0.0007358, "grad_norm": 0.48086267709732056, "learning_rate": 7.357e-05, "loss": 0.1926, "step": 7358 }, { "epoch": 0.0007359, "grad_norm": 0.47575855255126953, "learning_rate": 7.358e-05, "loss": 0.2006, "step": 7359 }, { "epoch": 0.000736, "grad_norm": 0.47889426350593567, "learning_rate": 7.359e-05, "loss": 0.2018, "step": 7360 }, { "epoch": 0.0007361, "grad_norm": 0.5811017155647278, "learning_rate": 7.36e-05, "loss": 0.2306, "step": 7361 }, { "epoch": 0.0007362, "grad_norm": 0.39591217041015625, "learning_rate": 7.361e-05, "loss": 0.1869, "step": 7362 }, { "epoch": 0.0007363, "grad_norm": 0.4870452582836151, "learning_rate": 7.362e-05, "loss": 0.2189, "step": 7363 }, { "epoch": 0.0007364, "grad_norm": 0.8045810461044312, "learning_rate": 7.363e-05, "loss": 0.2463, "step": 7364 }, { "epoch": 0.0007365, "grad_norm": 0.8989725708961487, "learning_rate": 7.364e-05, "loss": 0.2142, "step": 7365 }, { "epoch": 0.0007366, "grad_norm": 0.7122185230255127, "learning_rate": 7.365e-05, "loss": 0.2395, "step": 7366 }, { "epoch": 0.0007367, "grad_norm": 0.5073297619819641, "learning_rate": 7.366e-05, "loss": 0.1963, "step": 7367 }, { "epoch": 0.0007368, "grad_norm": 0.509922444820404, "learning_rate": 7.367e-05, "loss": 0.1998, "step": 7368 }, { "epoch": 0.0007369, "grad_norm": 0.5014863014221191, "learning_rate": 7.368e-05, "loss": 0.1973, "step": 7369 }, { "epoch": 0.000737, "grad_norm": 0.4400615692138672, "learning_rate": 7.369e-05, "loss": 0.1881, "step": 7370 }, { "epoch": 0.0007371, "grad_norm": 0.5283257961273193, "learning_rate": 7.37e-05, "loss": 0.1947, "step": 7371 }, { "epoch": 0.0007372, "grad_norm": 0.4372621774673462, "learning_rate": 7.371000000000001e-05, "loss": 0.1917, "step": 7372 }, { "epoch": 0.0007373, "grad_norm": 0.43310028314590454, "learning_rate": 7.371999999999999e-05, "loss": 0.1907, "step": 7373 }, { "epoch": 0.0007374, "grad_norm": 0.6451244950294495, "learning_rate": 7.373e-05, "loss": 0.2101, "step": 7374 }, { "epoch": 0.0007375, "grad_norm": 0.47620660066604614, "learning_rate": 7.374000000000001e-05, "loss": 0.1891, "step": 7375 }, { "epoch": 0.0007376, "grad_norm": 0.5217645168304443, "learning_rate": 7.375e-05, "loss": 0.2037, "step": 7376 }, { "epoch": 0.0007377, "grad_norm": 0.46922340989112854, "learning_rate": 7.376e-05, "loss": 0.1991, "step": 7377 }, { "epoch": 0.0007378, "grad_norm": 0.42827391624450684, "learning_rate": 7.377000000000001e-05, "loss": 0.1909, "step": 7378 }, { "epoch": 0.0007379, "grad_norm": 0.637048065662384, "learning_rate": 7.378e-05, "loss": 0.2262, "step": 7379 }, { "epoch": 0.000738, "grad_norm": 0.4661789834499359, "learning_rate": 7.379e-05, "loss": 0.1899, "step": 7380 }, { "epoch": 0.0007381, "grad_norm": 0.6267582774162292, "learning_rate": 7.38e-05, "loss": 0.1986, "step": 7381 }, { "epoch": 0.0007382, "grad_norm": 0.42628756165504456, "learning_rate": 7.381e-05, "loss": 0.1986, "step": 7382 }, { "epoch": 0.0007383, "grad_norm": 0.7806330919265747, "learning_rate": 7.382e-05, "loss": 0.219, "step": 7383 }, { "epoch": 0.0007384, "grad_norm": 0.5062658786773682, "learning_rate": 7.383e-05, "loss": 0.2126, "step": 7384 }, { "epoch": 0.0007385, "grad_norm": 0.5987305045127869, "learning_rate": 7.384e-05, "loss": 0.2074, "step": 7385 }, { "epoch": 0.0007386, "grad_norm": 0.4769977629184723, "learning_rate": 7.384999999999999e-05, "loss": 0.192, "step": 7386 }, { "epoch": 0.0007387, "grad_norm": 0.4302308261394501, "learning_rate": 7.386e-05, "loss": 0.1853, "step": 7387 }, { "epoch": 0.0007388, "grad_norm": 0.48651114106178284, "learning_rate": 7.387000000000001e-05, "loss": 0.1923, "step": 7388 }, { "epoch": 0.0007389, "grad_norm": 0.42370152473449707, "learning_rate": 7.388e-05, "loss": 0.1874, "step": 7389 }, { "epoch": 0.000739, "grad_norm": 0.3867861330509186, "learning_rate": 7.389e-05, "loss": 0.1808, "step": 7390 }, { "epoch": 0.0007391, "grad_norm": 0.4189610481262207, "learning_rate": 7.390000000000001e-05, "loss": 0.1919, "step": 7391 }, { "epoch": 0.0007392, "grad_norm": 0.5503688454627991, "learning_rate": 7.391e-05, "loss": 0.1958, "step": 7392 }, { "epoch": 0.0007393, "grad_norm": 0.48368340730667114, "learning_rate": 7.392e-05, "loss": 0.1897, "step": 7393 }, { "epoch": 0.0007394, "grad_norm": 0.404477596282959, "learning_rate": 7.393e-05, "loss": 0.1832, "step": 7394 }, { "epoch": 0.0007395, "grad_norm": 0.5074638724327087, "learning_rate": 7.394e-05, "loss": 0.1979, "step": 7395 }, { "epoch": 0.0007396, "grad_norm": 0.45375561714172363, "learning_rate": 7.395e-05, "loss": 0.1978, "step": 7396 }, { "epoch": 0.0007397, "grad_norm": 0.38874441385269165, "learning_rate": 7.396e-05, "loss": 0.1836, "step": 7397 }, { "epoch": 0.0007398, "grad_norm": 0.4363499581813812, "learning_rate": 7.397e-05, "loss": 0.1925, "step": 7398 }, { "epoch": 0.0007399, "grad_norm": 0.44526228308677673, "learning_rate": 7.398e-05, "loss": 0.1915, "step": 7399 }, { "epoch": 0.00074, "grad_norm": 0.7410265207290649, "learning_rate": 7.399e-05, "loss": 0.2211, "step": 7400 }, { "epoch": 0.0007401, "grad_norm": 0.39623939990997314, "learning_rate": 7.4e-05, "loss": 0.1835, "step": 7401 }, { "epoch": 0.0007402, "grad_norm": 0.42353275418281555, "learning_rate": 7.401e-05, "loss": 0.1909, "step": 7402 }, { "epoch": 0.0007403, "grad_norm": 0.4267234206199646, "learning_rate": 7.402e-05, "loss": 0.1969, "step": 7403 }, { "epoch": 0.0007404, "grad_norm": 0.4426077902317047, "learning_rate": 7.403e-05, "loss": 0.1924, "step": 7404 }, { "epoch": 0.0007405, "grad_norm": 0.6750543713569641, "learning_rate": 7.404e-05, "loss": 0.2429, "step": 7405 }, { "epoch": 0.0007406, "grad_norm": 0.39575839042663574, "learning_rate": 7.405000000000001e-05, "loss": 0.1823, "step": 7406 }, { "epoch": 0.0007407, "grad_norm": 0.4387655258178711, "learning_rate": 7.405999999999999e-05, "loss": 0.1881, "step": 7407 }, { "epoch": 0.0007408, "grad_norm": 0.47220295667648315, "learning_rate": 7.407e-05, "loss": 0.1996, "step": 7408 }, { "epoch": 0.0007409, "grad_norm": 0.3826233446598053, "learning_rate": 7.408000000000001e-05, "loss": 0.179, "step": 7409 }, { "epoch": 0.000741, "grad_norm": 0.5239911079406738, "learning_rate": 7.409e-05, "loss": 0.1937, "step": 7410 }, { "epoch": 0.0007411, "grad_norm": 0.37321022152900696, "learning_rate": 7.41e-05, "loss": 0.1775, "step": 7411 }, { "epoch": 0.0007412, "grad_norm": 0.7720602750778198, "learning_rate": 7.411000000000001e-05, "loss": 0.2292, "step": 7412 }, { "epoch": 0.0007413, "grad_norm": 0.5750840306282043, "learning_rate": 7.412e-05, "loss": 0.1973, "step": 7413 }, { "epoch": 0.0007414, "grad_norm": 4.924017906188965, "learning_rate": 7.413e-05, "loss": 0.3469, "step": 7414 }, { "epoch": 0.0007415, "grad_norm": 0.47601503133773804, "learning_rate": 7.414e-05, "loss": 0.1858, "step": 7415 }, { "epoch": 0.0007416, "grad_norm": 0.4628167748451233, "learning_rate": 7.415e-05, "loss": 0.1874, "step": 7416 }, { "epoch": 0.0007417, "grad_norm": 0.6044511198997498, "learning_rate": 7.416e-05, "loss": 0.2205, "step": 7417 }, { "epoch": 0.0007418, "grad_norm": 0.7673957943916321, "learning_rate": 7.417e-05, "loss": 0.1954, "step": 7418 }, { "epoch": 0.0007419, "grad_norm": 0.45705947279930115, "learning_rate": 7.418e-05, "loss": 0.1969, "step": 7419 }, { "epoch": 0.000742, "grad_norm": 0.5442333221435547, "learning_rate": 7.418999999999999e-05, "loss": 0.2039, "step": 7420 }, { "epoch": 0.0007421, "grad_norm": 0.7553530335426331, "learning_rate": 7.42e-05, "loss": 0.2479, "step": 7421 }, { "epoch": 0.0007422, "grad_norm": 1.0189530849456787, "learning_rate": 7.421000000000001e-05, "loss": 0.2329, "step": 7422 }, { "epoch": 0.0007423, "grad_norm": 0.5386321544647217, "learning_rate": 7.422e-05, "loss": 0.1906, "step": 7423 }, { "epoch": 0.0007424, "grad_norm": 1.569715976715088, "learning_rate": 7.423e-05, "loss": 0.3202, "step": 7424 }, { "epoch": 0.0007425, "grad_norm": 0.5454429388046265, "learning_rate": 7.424000000000001e-05, "loss": 0.1985, "step": 7425 }, { "epoch": 0.0007426, "grad_norm": 0.5449747443199158, "learning_rate": 7.425e-05, "loss": 0.1947, "step": 7426 }, { "epoch": 0.0007427, "grad_norm": 0.7405964136123657, "learning_rate": 7.426e-05, "loss": 0.2125, "step": 7427 }, { "epoch": 0.0007428, "grad_norm": 0.45297539234161377, "learning_rate": 7.427e-05, "loss": 0.1805, "step": 7428 }, { "epoch": 0.0007429, "grad_norm": 0.4657517671585083, "learning_rate": 7.428e-05, "loss": 0.1866, "step": 7429 }, { "epoch": 0.000743, "grad_norm": 0.45083001255989075, "learning_rate": 7.429e-05, "loss": 0.179, "step": 7430 }, { "epoch": 0.0007431, "grad_norm": 0.5598366856575012, "learning_rate": 7.43e-05, "loss": 0.1898, "step": 7431 }, { "epoch": 0.0007432, "grad_norm": 0.4710606336593628, "learning_rate": 7.431e-05, "loss": 0.1908, "step": 7432 }, { "epoch": 0.0007433, "grad_norm": 0.42426031827926636, "learning_rate": 7.432e-05, "loss": 0.1807, "step": 7433 }, { "epoch": 0.0007434, "grad_norm": 0.5185211300849915, "learning_rate": 7.433e-05, "loss": 0.1958, "step": 7434 }, { "epoch": 0.0007435, "grad_norm": 0.546903133392334, "learning_rate": 7.434e-05, "loss": 0.192, "step": 7435 }, { "epoch": 0.0007436, "grad_norm": 0.501542866230011, "learning_rate": 7.435e-05, "loss": 0.1984, "step": 7436 }, { "epoch": 0.0007437, "grad_norm": 0.4432472288608551, "learning_rate": 7.436e-05, "loss": 0.1846, "step": 7437 }, { "epoch": 0.0007438, "grad_norm": 0.4535142481327057, "learning_rate": 7.437e-05, "loss": 0.1908, "step": 7438 }, { "epoch": 0.0007439, "grad_norm": 0.49497199058532715, "learning_rate": 7.438e-05, "loss": 0.1914, "step": 7439 }, { "epoch": 0.000744, "grad_norm": 0.6516047716140747, "learning_rate": 7.439000000000001e-05, "loss": 0.2054, "step": 7440 }, { "epoch": 0.0007441, "grad_norm": 0.5227893590927124, "learning_rate": 7.439999999999999e-05, "loss": 0.2042, "step": 7441 }, { "epoch": 0.0007442, "grad_norm": 0.4323962926864624, "learning_rate": 7.441e-05, "loss": 0.1847, "step": 7442 }, { "epoch": 0.0007443, "grad_norm": 0.4462810754776001, "learning_rate": 7.442000000000001e-05, "loss": 0.186, "step": 7443 }, { "epoch": 0.0007444, "grad_norm": 0.44686514139175415, "learning_rate": 7.443e-05, "loss": 0.1833, "step": 7444 }, { "epoch": 0.0007445, "grad_norm": 0.42378124594688416, "learning_rate": 7.444e-05, "loss": 0.178, "step": 7445 }, { "epoch": 0.0007446, "grad_norm": 0.4270733594894409, "learning_rate": 7.445000000000001e-05, "loss": 0.1819, "step": 7446 }, { "epoch": 0.0007447, "grad_norm": 0.7717322111129761, "learning_rate": 7.446e-05, "loss": 0.2372, "step": 7447 }, { "epoch": 0.0007448, "grad_norm": 0.4648915231227875, "learning_rate": 7.447e-05, "loss": 0.1921, "step": 7448 }, { "epoch": 0.0007449, "grad_norm": 0.46609944105148315, "learning_rate": 7.448e-05, "loss": 0.1882, "step": 7449 }, { "epoch": 0.000745, "grad_norm": 0.589912474155426, "learning_rate": 7.449e-05, "loss": 0.2164, "step": 7450 }, { "epoch": 0.0007451, "grad_norm": 0.48248955607414246, "learning_rate": 7.45e-05, "loss": 0.1825, "step": 7451 }, { "epoch": 0.0007452, "grad_norm": 0.4037376344203949, "learning_rate": 7.451e-05, "loss": 0.1868, "step": 7452 }, { "epoch": 0.0007453, "grad_norm": 0.4312678277492523, "learning_rate": 7.452e-05, "loss": 0.1851, "step": 7453 }, { "epoch": 0.0007454, "grad_norm": 0.40998610854148865, "learning_rate": 7.452999999999999e-05, "loss": 0.1848, "step": 7454 }, { "epoch": 0.0007455, "grad_norm": 0.4464742839336395, "learning_rate": 7.454e-05, "loss": 0.1913, "step": 7455 }, { "epoch": 0.0007456, "grad_norm": 0.4490673243999481, "learning_rate": 7.455000000000001e-05, "loss": 0.1864, "step": 7456 }, { "epoch": 0.0007457, "grad_norm": 0.39138734340667725, "learning_rate": 7.456e-05, "loss": 0.1748, "step": 7457 }, { "epoch": 0.0007458, "grad_norm": 0.5053826570510864, "learning_rate": 7.457e-05, "loss": 0.2015, "step": 7458 }, { "epoch": 0.0007459, "grad_norm": 0.4277768135070801, "learning_rate": 7.458000000000001e-05, "loss": 0.1869, "step": 7459 }, { "epoch": 0.000746, "grad_norm": 0.37439417839050293, "learning_rate": 7.459e-05, "loss": 0.1818, "step": 7460 }, { "epoch": 0.0007461, "grad_norm": 0.4284929931163788, "learning_rate": 7.46e-05, "loss": 0.1865, "step": 7461 }, { "epoch": 0.0007462, "grad_norm": 7.139459609985352, "learning_rate": 7.461e-05, "loss": 0.6917, "step": 7462 }, { "epoch": 0.0007463, "grad_norm": 1.4773494005203247, "learning_rate": 7.462e-05, "loss": 0.2689, "step": 7463 }, { "epoch": 0.0007464, "grad_norm": 0.5157195329666138, "learning_rate": 7.463e-05, "loss": 0.1873, "step": 7464 }, { "epoch": 0.0007465, "grad_norm": 0.4672258496284485, "learning_rate": 7.464e-05, "loss": 0.1758, "step": 7465 }, { "epoch": 0.0007466, "grad_norm": 0.6287294626235962, "learning_rate": 7.465e-05, "loss": 0.203, "step": 7466 }, { "epoch": 0.0007467, "grad_norm": 0.5864905118942261, "learning_rate": 7.466000000000001e-05, "loss": 0.1952, "step": 7467 }, { "epoch": 0.0007468, "grad_norm": 0.5101815462112427, "learning_rate": 7.467e-05, "loss": 0.1755, "step": 7468 }, { "epoch": 0.0007469, "grad_norm": 0.9012991786003113, "learning_rate": 7.468e-05, "loss": 0.2595, "step": 7469 }, { "epoch": 0.000747, "grad_norm": 0.42554858326911926, "learning_rate": 7.469e-05, "loss": 0.1702, "step": 7470 }, { "epoch": 0.0007471, "grad_norm": 0.5329145193099976, "learning_rate": 7.47e-05, "loss": 0.1949, "step": 7471 }, { "epoch": 0.0007472, "grad_norm": 0.4959746301174164, "learning_rate": 7.471e-05, "loss": 0.179, "step": 7472 }, { "epoch": 0.0007473, "grad_norm": 1.148716688156128, "learning_rate": 7.472e-05, "loss": 0.223, "step": 7473 }, { "epoch": 0.0007474, "grad_norm": 0.5317788124084473, "learning_rate": 7.473000000000001e-05, "loss": 0.2101, "step": 7474 }, { "epoch": 0.0007475, "grad_norm": 0.5204247832298279, "learning_rate": 7.473999999999999e-05, "loss": 0.1957, "step": 7475 }, { "epoch": 0.0007476, "grad_norm": 0.6172888875007629, "learning_rate": 7.475e-05, "loss": 0.1917, "step": 7476 }, { "epoch": 0.0007477, "grad_norm": 0.6861792802810669, "learning_rate": 7.476000000000001e-05, "loss": 0.1942, "step": 7477 }, { "epoch": 0.0007478, "grad_norm": 0.4458335340023041, "learning_rate": 7.477e-05, "loss": 0.1857, "step": 7478 }, { "epoch": 0.0007479, "grad_norm": 0.43245258927345276, "learning_rate": 7.478e-05, "loss": 0.1737, "step": 7479 }, { "epoch": 0.000748, "grad_norm": 0.4693397283554077, "learning_rate": 7.479000000000001e-05, "loss": 0.1851, "step": 7480 }, { "epoch": 0.0007481, "grad_norm": 0.5341662168502808, "learning_rate": 7.48e-05, "loss": 0.1902, "step": 7481 }, { "epoch": 0.0007482, "grad_norm": 0.6064165830612183, "learning_rate": 7.481e-05, "loss": 0.2173, "step": 7482 }, { "epoch": 0.0007483, "grad_norm": 0.39724573493003845, "learning_rate": 7.482e-05, "loss": 0.1735, "step": 7483 }, { "epoch": 0.0007484, "grad_norm": 0.3705470860004425, "learning_rate": 7.483e-05, "loss": 0.1703, "step": 7484 }, { "epoch": 0.0007485, "grad_norm": 0.43255889415740967, "learning_rate": 7.484e-05, "loss": 0.1763, "step": 7485 }, { "epoch": 0.0007486, "grad_norm": 0.4192142188549042, "learning_rate": 7.485e-05, "loss": 0.1801, "step": 7486 }, { "epoch": 0.0007487, "grad_norm": 0.7527932524681091, "learning_rate": 7.486e-05, "loss": 0.2267, "step": 7487 }, { "epoch": 0.0007488, "grad_norm": 0.39235246181488037, "learning_rate": 7.486999999999999e-05, "loss": 0.1724, "step": 7488 }, { "epoch": 0.0007489, "grad_norm": 0.42596572637557983, "learning_rate": 7.488e-05, "loss": 0.1805, "step": 7489 }, { "epoch": 0.000749, "grad_norm": 0.43126609921455383, "learning_rate": 7.489000000000001e-05, "loss": 0.1859, "step": 7490 }, { "epoch": 0.0007491, "grad_norm": 0.42793038487434387, "learning_rate": 7.49e-05, "loss": 0.1788, "step": 7491 }, { "epoch": 0.0007492, "grad_norm": 0.44571396708488464, "learning_rate": 7.491e-05, "loss": 0.1943, "step": 7492 }, { "epoch": 0.0007493, "grad_norm": 0.5444081425666809, "learning_rate": 7.492000000000001e-05, "loss": 0.197, "step": 7493 }, { "epoch": 0.0007494, "grad_norm": 0.4795020520687103, "learning_rate": 7.493e-05, "loss": 0.204, "step": 7494 }, { "epoch": 0.0007495, "grad_norm": 0.4522680342197418, "learning_rate": 7.494e-05, "loss": 0.194, "step": 7495 }, { "epoch": 0.0007496, "grad_norm": 0.3937327563762665, "learning_rate": 7.495e-05, "loss": 0.1844, "step": 7496 }, { "epoch": 0.0007497, "grad_norm": 1.3814177513122559, "learning_rate": 7.496e-05, "loss": 0.2533, "step": 7497 }, { "epoch": 0.0007498, "grad_norm": 0.48238804936408997, "learning_rate": 7.497e-05, "loss": 0.1885, "step": 7498 }, { "epoch": 0.0007499, "grad_norm": 0.45935240387916565, "learning_rate": 7.498e-05, "loss": 0.1896, "step": 7499 }, { "epoch": 0.00075, "grad_norm": 0.4718035161495209, "learning_rate": 7.499e-05, "loss": 0.1913, "step": 7500 }, { "epoch": 0.00075, "eval_loss": 0.021516965702176094, "eval_runtime": 362.7437, "eval_samples_per_second": 27.568, "eval_steps_per_second": 1.723, "step": 7500 }, { "epoch": 0.0007501, "grad_norm": 0.9165650606155396, "learning_rate": 7.500000000000001e-05, "loss": 0.1976, "step": 7501 }, { "epoch": 0.0007502, "grad_norm": 0.4532630443572998, "learning_rate": 7.501e-05, "loss": 0.1943, "step": 7502 }, { "epoch": 0.0007503, "grad_norm": 0.4622610807418823, "learning_rate": 7.502e-05, "loss": 0.1764, "step": 7503 }, { "epoch": 0.0007504, "grad_norm": 0.4768509268760681, "learning_rate": 7.503e-05, "loss": 0.1836, "step": 7504 }, { "epoch": 0.0007505, "grad_norm": 0.4220264256000519, "learning_rate": 7.504e-05, "loss": 0.1754, "step": 7505 }, { "epoch": 0.0007506, "grad_norm": 0.4648164212703705, "learning_rate": 7.505e-05, "loss": 0.1864, "step": 7506 }, { "epoch": 0.0007507, "grad_norm": 0.42253103852272034, "learning_rate": 7.506e-05, "loss": 0.1842, "step": 7507 }, { "epoch": 0.0007508, "grad_norm": 0.41251248121261597, "learning_rate": 7.507000000000001e-05, "loss": 0.1821, "step": 7508 }, { "epoch": 0.0007509, "grad_norm": 0.41563406586647034, "learning_rate": 7.507999999999999e-05, "loss": 0.1869, "step": 7509 }, { "epoch": 0.000751, "grad_norm": 0.40142396092414856, "learning_rate": 7.509e-05, "loss": 0.1851, "step": 7510 }, { "epoch": 0.0007511, "grad_norm": 0.43007296323776245, "learning_rate": 7.510000000000001e-05, "loss": 0.1853, "step": 7511 }, { "epoch": 0.0007512, "grad_norm": 1.2085107564926147, "learning_rate": 7.511e-05, "loss": 0.25, "step": 7512 }, { "epoch": 0.0007513, "grad_norm": 0.6795465350151062, "learning_rate": 7.512e-05, "loss": 0.226, "step": 7513 }, { "epoch": 0.0007514, "grad_norm": 0.475106805562973, "learning_rate": 7.513000000000001e-05, "loss": 0.1952, "step": 7514 }, { "epoch": 0.0007515, "grad_norm": 0.4340527057647705, "learning_rate": 7.514e-05, "loss": 0.1792, "step": 7515 }, { "epoch": 0.0007516, "grad_norm": 0.4670300781726837, "learning_rate": 7.515e-05, "loss": 0.1843, "step": 7516 }, { "epoch": 0.0007517, "grad_norm": 0.7026162147521973, "learning_rate": 7.516e-05, "loss": 0.1896, "step": 7517 }, { "epoch": 0.0007518, "grad_norm": 0.4725876748561859, "learning_rate": 7.517e-05, "loss": 0.1976, "step": 7518 }, { "epoch": 0.0007519, "grad_norm": 0.5276523232460022, "learning_rate": 7.518e-05, "loss": 0.1975, "step": 7519 }, { "epoch": 0.000752, "grad_norm": 0.38987573981285095, "learning_rate": 7.519e-05, "loss": 0.1737, "step": 7520 }, { "epoch": 0.0007521, "grad_norm": 0.6679162383079529, "learning_rate": 7.52e-05, "loss": 0.2024, "step": 7521 }, { "epoch": 0.0007522, "grad_norm": 0.6624200940132141, "learning_rate": 7.520999999999999e-05, "loss": 0.2177, "step": 7522 }, { "epoch": 0.0007523, "grad_norm": 1.9751900434494019, "learning_rate": 7.522e-05, "loss": 0.2249, "step": 7523 }, { "epoch": 0.0007524, "grad_norm": 0.5174213647842407, "learning_rate": 7.523000000000001e-05, "loss": 0.1787, "step": 7524 }, { "epoch": 0.0007525, "grad_norm": 0.5586757659912109, "learning_rate": 7.524e-05, "loss": 0.1896, "step": 7525 }, { "epoch": 0.0007526, "grad_norm": 0.6579071879386902, "learning_rate": 7.525e-05, "loss": 0.2024, "step": 7526 }, { "epoch": 0.0007527, "grad_norm": 0.4693479835987091, "learning_rate": 7.526000000000001e-05, "loss": 0.1692, "step": 7527 }, { "epoch": 0.0007528, "grad_norm": 0.4994659423828125, "learning_rate": 7.527e-05, "loss": 0.1989, "step": 7528 }, { "epoch": 0.0007529, "grad_norm": 0.476550430059433, "learning_rate": 7.528e-05, "loss": 0.1885, "step": 7529 }, { "epoch": 0.000753, "grad_norm": 0.48724666237831116, "learning_rate": 7.529e-05, "loss": 0.1892, "step": 7530 }, { "epoch": 0.0007531, "grad_norm": 0.4618290066719055, "learning_rate": 7.53e-05, "loss": 0.1797, "step": 7531 }, { "epoch": 0.0007532, "grad_norm": 1.1197789907455444, "learning_rate": 7.531e-05, "loss": 0.2489, "step": 7532 }, { "epoch": 0.0007533, "grad_norm": 0.6072105169296265, "learning_rate": 7.532e-05, "loss": 0.1869, "step": 7533 }, { "epoch": 0.0007534, "grad_norm": 0.7256792187690735, "learning_rate": 7.533e-05, "loss": 0.2135, "step": 7534 }, { "epoch": 0.0007535, "grad_norm": 0.5450584888458252, "learning_rate": 7.534000000000001e-05, "loss": 0.1846, "step": 7535 }, { "epoch": 0.0007536, "grad_norm": 0.738315224647522, "learning_rate": 7.535e-05, "loss": 0.207, "step": 7536 }, { "epoch": 0.0007537, "grad_norm": 0.41732147336006165, "learning_rate": 7.536e-05, "loss": 0.1733, "step": 7537 }, { "epoch": 0.0007538, "grad_norm": 0.6369668841362, "learning_rate": 7.537e-05, "loss": 0.1918, "step": 7538 }, { "epoch": 0.0007539, "grad_norm": 0.6139333844184875, "learning_rate": 7.538e-05, "loss": 0.1976, "step": 7539 }, { "epoch": 0.000754, "grad_norm": 0.6071346998214722, "learning_rate": 7.539e-05, "loss": 0.2021, "step": 7540 }, { "epoch": 0.0007541, "grad_norm": 0.9224357008934021, "learning_rate": 7.54e-05, "loss": 0.2198, "step": 7541 }, { "epoch": 0.0007542, "grad_norm": 0.6226947903633118, "learning_rate": 7.541000000000001e-05, "loss": 0.2031, "step": 7542 }, { "epoch": 0.0007543, "grad_norm": 0.44756942987442017, "learning_rate": 7.541999999999999e-05, "loss": 0.1685, "step": 7543 }, { "epoch": 0.0007544, "grad_norm": 0.551895797252655, "learning_rate": 7.543e-05, "loss": 0.1893, "step": 7544 }, { "epoch": 0.0007545, "grad_norm": 2.0797204971313477, "learning_rate": 7.544000000000001e-05, "loss": 0.3489, "step": 7545 }, { "epoch": 0.0007546, "grad_norm": 0.49439364671707153, "learning_rate": 7.545e-05, "loss": 0.1821, "step": 7546 }, { "epoch": 0.0007547, "grad_norm": 0.6312146186828613, "learning_rate": 7.546e-05, "loss": 0.2114, "step": 7547 }, { "epoch": 0.0007548, "grad_norm": 0.4911887049674988, "learning_rate": 7.547000000000001e-05, "loss": 0.1764, "step": 7548 }, { "epoch": 0.0007549, "grad_norm": 0.45326417684555054, "learning_rate": 7.548e-05, "loss": 0.1736, "step": 7549 }, { "epoch": 0.000755, "grad_norm": 0.46612992882728577, "learning_rate": 7.549e-05, "loss": 0.1879, "step": 7550 }, { "epoch": 0.0007551, "grad_norm": 0.5469991564750671, "learning_rate": 7.55e-05, "loss": 0.1855, "step": 7551 }, { "epoch": 0.0007552, "grad_norm": 0.48811817169189453, "learning_rate": 7.551e-05, "loss": 0.1913, "step": 7552 }, { "epoch": 0.0007553, "grad_norm": 0.4290868639945984, "learning_rate": 7.552e-05, "loss": 0.172, "step": 7553 }, { "epoch": 0.0007554, "grad_norm": 0.4164026975631714, "learning_rate": 7.553e-05, "loss": 0.1748, "step": 7554 }, { "epoch": 0.0007555, "grad_norm": 0.5739708542823792, "learning_rate": 7.554e-05, "loss": 0.1913, "step": 7555 }, { "epoch": 0.0007556, "grad_norm": 0.584821343421936, "learning_rate": 7.555e-05, "loss": 0.204, "step": 7556 }, { "epoch": 0.0007557, "grad_norm": 0.40070706605911255, "learning_rate": 7.556e-05, "loss": 0.1747, "step": 7557 }, { "epoch": 0.0007558, "grad_norm": 0.4085039496421814, "learning_rate": 7.557000000000001e-05, "loss": 0.1798, "step": 7558 }, { "epoch": 0.0007559, "grad_norm": 0.41738244891166687, "learning_rate": 7.558e-05, "loss": 0.1814, "step": 7559 }, { "epoch": 0.000756, "grad_norm": 0.40071919560432434, "learning_rate": 7.559e-05, "loss": 0.1765, "step": 7560 }, { "epoch": 0.0007561, "grad_norm": 0.4130774736404419, "learning_rate": 7.560000000000001e-05, "loss": 0.1772, "step": 7561 }, { "epoch": 0.0007562, "grad_norm": 0.386495441198349, "learning_rate": 7.561e-05, "loss": 0.1761, "step": 7562 }, { "epoch": 0.0007563, "grad_norm": 0.3618008494377136, "learning_rate": 7.562e-05, "loss": 0.167, "step": 7563 }, { "epoch": 0.0007564, "grad_norm": 0.5081895589828491, "learning_rate": 7.563e-05, "loss": 0.1785, "step": 7564 }, { "epoch": 0.0007565, "grad_norm": 0.36260002851486206, "learning_rate": 7.564e-05, "loss": 0.1639, "step": 7565 }, { "epoch": 0.0007566, "grad_norm": 0.43156561255455017, "learning_rate": 7.565e-05, "loss": 0.1884, "step": 7566 }, { "epoch": 0.0007567, "grad_norm": 0.4242669641971588, "learning_rate": 7.566e-05, "loss": 0.1725, "step": 7567 }, { "epoch": 0.0007568, "grad_norm": 0.38370412588119507, "learning_rate": 7.567e-05, "loss": 0.1597, "step": 7568 }, { "epoch": 0.0007569, "grad_norm": 0.4324035942554474, "learning_rate": 7.568000000000001e-05, "loss": 0.1891, "step": 7569 }, { "epoch": 0.000757, "grad_norm": 1.4893518686294556, "learning_rate": 7.569e-05, "loss": 0.275, "step": 7570 }, { "epoch": 0.0007571, "grad_norm": 0.41903969645500183, "learning_rate": 7.57e-05, "loss": 0.176, "step": 7571 }, { "epoch": 0.0007572, "grad_norm": 0.6930134296417236, "learning_rate": 7.571e-05, "loss": 0.2087, "step": 7572 }, { "epoch": 0.0007573, "grad_norm": 0.4721878170967102, "learning_rate": 7.572e-05, "loss": 0.1761, "step": 7573 }, { "epoch": 0.0007574, "grad_norm": 0.5028660893440247, "learning_rate": 7.573e-05, "loss": 0.1757, "step": 7574 }, { "epoch": 0.0007575, "grad_norm": 0.45845597982406616, "learning_rate": 7.574e-05, "loss": 0.1774, "step": 7575 }, { "epoch": 0.0007576, "grad_norm": 0.45461151003837585, "learning_rate": 7.575000000000001e-05, "loss": 0.1665, "step": 7576 }, { "epoch": 0.0007577, "grad_norm": 0.6267446279525757, "learning_rate": 7.575999999999999e-05, "loss": 0.1907, "step": 7577 }, { "epoch": 0.0007578, "grad_norm": 0.39918413758277893, "learning_rate": 7.577e-05, "loss": 0.172, "step": 7578 }, { "epoch": 0.0007579, "grad_norm": 0.46049246191978455, "learning_rate": 7.578000000000001e-05, "loss": 0.1713, "step": 7579 }, { "epoch": 0.000758, "grad_norm": 0.4430519938468933, "learning_rate": 7.579e-05, "loss": 0.179, "step": 7580 }, { "epoch": 0.0007581, "grad_norm": 0.686094343662262, "learning_rate": 7.58e-05, "loss": 0.1864, "step": 7581 }, { "epoch": 0.0007582, "grad_norm": 0.5256968140602112, "learning_rate": 7.581000000000001e-05, "loss": 0.1952, "step": 7582 }, { "epoch": 0.0007583, "grad_norm": 0.4756493866443634, "learning_rate": 7.582e-05, "loss": 0.1783, "step": 7583 }, { "epoch": 0.0007584, "grad_norm": 0.5038774013519287, "learning_rate": 7.583e-05, "loss": 0.1887, "step": 7584 }, { "epoch": 0.0007585, "grad_norm": 0.3784870207309723, "learning_rate": 7.584e-05, "loss": 0.158, "step": 7585 }, { "epoch": 0.0007586, "grad_norm": 0.3814629018306732, "learning_rate": 7.585e-05, "loss": 0.1688, "step": 7586 }, { "epoch": 0.0007587, "grad_norm": 0.3632339835166931, "learning_rate": 7.586e-05, "loss": 0.1689, "step": 7587 }, { "epoch": 0.0007588, "grad_norm": 0.3821631669998169, "learning_rate": 7.587e-05, "loss": 0.1665, "step": 7588 }, { "epoch": 0.0007589, "grad_norm": 0.40399041771888733, "learning_rate": 7.588e-05, "loss": 0.1812, "step": 7589 }, { "epoch": 0.000759, "grad_norm": 0.49214446544647217, "learning_rate": 7.589e-05, "loss": 0.1893, "step": 7590 }, { "epoch": 0.0007591, "grad_norm": 0.3859829306602478, "learning_rate": 7.59e-05, "loss": 0.1659, "step": 7591 }, { "epoch": 0.0007592, "grad_norm": 0.40475544333457947, "learning_rate": 7.591000000000001e-05, "loss": 0.173, "step": 7592 }, { "epoch": 0.0007593, "grad_norm": 0.3380837142467499, "learning_rate": 7.592e-05, "loss": 0.1613, "step": 7593 }, { "epoch": 0.0007594, "grad_norm": 0.4463665783405304, "learning_rate": 7.593e-05, "loss": 0.1709, "step": 7594 }, { "epoch": 0.0007595, "grad_norm": 0.3347475826740265, "learning_rate": 7.594000000000001e-05, "loss": 0.162, "step": 7595 }, { "epoch": 0.0007596, "grad_norm": 0.39011281728744507, "learning_rate": 7.595e-05, "loss": 0.175, "step": 7596 }, { "epoch": 0.0007597, "grad_norm": 0.424954891204834, "learning_rate": 7.596e-05, "loss": 0.1672, "step": 7597 }, { "epoch": 0.0007598, "grad_norm": 0.38835087418556213, "learning_rate": 7.597e-05, "loss": 0.1725, "step": 7598 }, { "epoch": 0.0007599, "grad_norm": 0.37602484226226807, "learning_rate": 7.598e-05, "loss": 0.1693, "step": 7599 }, { "epoch": 0.00076, "grad_norm": 0.3643645942211151, "learning_rate": 7.599e-05, "loss": 0.1647, "step": 7600 }, { "epoch": 0.0007601, "grad_norm": 0.36501967906951904, "learning_rate": 7.6e-05, "loss": 0.1669, "step": 7601 }, { "epoch": 0.0007602, "grad_norm": 0.34811708331108093, "learning_rate": 7.601e-05, "loss": 0.1586, "step": 7602 }, { "epoch": 0.0007603, "grad_norm": 0.3570595383644104, "learning_rate": 7.602000000000001e-05, "loss": 0.1688, "step": 7603 }, { "epoch": 0.0007604, "grad_norm": 0.411408394575119, "learning_rate": 7.603e-05, "loss": 0.1757, "step": 7604 }, { "epoch": 0.0007605, "grad_norm": 0.45920392870903015, "learning_rate": 7.604e-05, "loss": 0.1887, "step": 7605 }, { "epoch": 0.0007606, "grad_norm": 0.39549022912979126, "learning_rate": 7.605e-05, "loss": 0.1738, "step": 7606 }, { "epoch": 0.0007607, "grad_norm": 0.3656753599643707, "learning_rate": 7.606e-05, "loss": 0.1666, "step": 7607 }, { "epoch": 0.0007608, "grad_norm": 0.32445859909057617, "learning_rate": 7.607e-05, "loss": 0.1559, "step": 7608 }, { "epoch": 0.0007609, "grad_norm": 0.5102970600128174, "learning_rate": 7.608e-05, "loss": 0.1697, "step": 7609 }, { "epoch": 0.000761, "grad_norm": 0.5840592384338379, "learning_rate": 7.609000000000001e-05, "loss": 0.192, "step": 7610 }, { "epoch": 0.0007611, "grad_norm": 0.8516383171081543, "learning_rate": 7.609999999999999e-05, "loss": 0.1871, "step": 7611 }, { "epoch": 0.0007612, "grad_norm": 0.4764192998409271, "learning_rate": 7.611e-05, "loss": 0.1598, "step": 7612 }, { "epoch": 0.0007613, "grad_norm": 0.34317582845687866, "learning_rate": 7.612000000000001e-05, "loss": 0.1567, "step": 7613 }, { "epoch": 0.0007614, "grad_norm": 0.46695294976234436, "learning_rate": 7.613e-05, "loss": 0.173, "step": 7614 }, { "epoch": 0.0007615, "grad_norm": 0.6321865320205688, "learning_rate": 7.614e-05, "loss": 0.2097, "step": 7615 }, { "epoch": 0.0007616, "grad_norm": 0.4267997145652771, "learning_rate": 7.615000000000001e-05, "loss": 0.1639, "step": 7616 }, { "epoch": 0.0007617, "grad_norm": 0.40251919627189636, "learning_rate": 7.616e-05, "loss": 0.168, "step": 7617 }, { "epoch": 0.0007618, "grad_norm": 0.5825067162513733, "learning_rate": 7.617e-05, "loss": 0.1808, "step": 7618 }, { "epoch": 0.0007619, "grad_norm": 0.37340980768203735, "learning_rate": 7.618e-05, "loss": 0.1584, "step": 7619 }, { "epoch": 0.000762, "grad_norm": 0.3515920639038086, "learning_rate": 7.619e-05, "loss": 0.1608, "step": 7620 }, { "epoch": 0.0007621, "grad_norm": 0.48071399331092834, "learning_rate": 7.62e-05, "loss": 0.1793, "step": 7621 }, { "epoch": 0.0007622, "grad_norm": 0.39304035902023315, "learning_rate": 7.621e-05, "loss": 0.1621, "step": 7622 }, { "epoch": 0.0007623, "grad_norm": 0.40765902400016785, "learning_rate": 7.622e-05, "loss": 0.1689, "step": 7623 }, { "epoch": 0.0007624, "grad_norm": 0.41402965784072876, "learning_rate": 7.623e-05, "loss": 0.1735, "step": 7624 }, { "epoch": 0.0007625, "grad_norm": 0.3803972899913788, "learning_rate": 7.624e-05, "loss": 0.1638, "step": 7625 }, { "epoch": 0.0007626, "grad_norm": 0.6322067975997925, "learning_rate": 7.625e-05, "loss": 0.2078, "step": 7626 }, { "epoch": 0.0007627, "grad_norm": 0.3621191084384918, "learning_rate": 7.626e-05, "loss": 0.1599, "step": 7627 }, { "epoch": 0.0007628, "grad_norm": 0.3360922634601593, "learning_rate": 7.627e-05, "loss": 0.1519, "step": 7628 }, { "epoch": 0.0007629, "grad_norm": 0.37620165944099426, "learning_rate": 7.628000000000001e-05, "loss": 0.1726, "step": 7629 }, { "epoch": 0.000763, "grad_norm": 0.44851481914520264, "learning_rate": 7.629e-05, "loss": 0.1827, "step": 7630 }, { "epoch": 0.0007631, "grad_norm": 0.5561351776123047, "learning_rate": 7.63e-05, "loss": 0.1719, "step": 7631 }, { "epoch": 0.0007632, "grad_norm": 0.6119112372398376, "learning_rate": 7.631e-05, "loss": 0.1886, "step": 7632 }, { "epoch": 0.0007633, "grad_norm": 1.0111428499221802, "learning_rate": 7.632e-05, "loss": 0.2168, "step": 7633 }, { "epoch": 0.0007634, "grad_norm": 0.42068493366241455, "learning_rate": 7.633e-05, "loss": 0.1674, "step": 7634 }, { "epoch": 0.0007635, "grad_norm": 0.365709513425827, "learning_rate": 7.634e-05, "loss": 0.1561, "step": 7635 }, { "epoch": 0.0007636, "grad_norm": 0.5593344569206238, "learning_rate": 7.635e-05, "loss": 0.1716, "step": 7636 }, { "epoch": 0.0007637, "grad_norm": 0.37984171509742737, "learning_rate": 7.636000000000001e-05, "loss": 0.1577, "step": 7637 }, { "epoch": 0.0007638, "grad_norm": 0.38292455673217773, "learning_rate": 7.637e-05, "loss": 0.1676, "step": 7638 }, { "epoch": 0.0007639, "grad_norm": 0.36735305190086365, "learning_rate": 7.638e-05, "loss": 0.1593, "step": 7639 }, { "epoch": 0.000764, "grad_norm": 0.47483325004577637, "learning_rate": 7.639e-05, "loss": 0.1794, "step": 7640 }, { "epoch": 0.0007641, "grad_norm": 0.4931349456310272, "learning_rate": 7.64e-05, "loss": 0.1638, "step": 7641 }, { "epoch": 0.0007642, "grad_norm": 0.5691468119621277, "learning_rate": 7.641e-05, "loss": 0.1766, "step": 7642 }, { "epoch": 0.0007643, "grad_norm": 0.39042598009109497, "learning_rate": 7.642e-05, "loss": 0.1659, "step": 7643 }, { "epoch": 0.0007644, "grad_norm": 0.8049102425575256, "learning_rate": 7.643000000000001e-05, "loss": 0.2017, "step": 7644 }, { "epoch": 0.0007645, "grad_norm": 0.34670665860176086, "learning_rate": 7.643999999999999e-05, "loss": 0.1509, "step": 7645 }, { "epoch": 0.0007646, "grad_norm": 0.3409683108329773, "learning_rate": 7.645e-05, "loss": 0.1525, "step": 7646 }, { "epoch": 0.0007647, "grad_norm": 0.38340339064598083, "learning_rate": 7.646000000000001e-05, "loss": 0.1576, "step": 7647 }, { "epoch": 0.0007648, "grad_norm": 0.8329829573631287, "learning_rate": 7.646999999999999e-05, "loss": 0.223, "step": 7648 }, { "epoch": 0.0007649, "grad_norm": 0.7235862016677856, "learning_rate": 7.648e-05, "loss": 0.2101, "step": 7649 }, { "epoch": 0.000765, "grad_norm": 0.35512569546699524, "learning_rate": 7.649000000000001e-05, "loss": 0.1611, "step": 7650 }, { "epoch": 0.0007651, "grad_norm": 0.39725664258003235, "learning_rate": 7.65e-05, "loss": 0.1606, "step": 7651 }, { "epoch": 0.0007652, "grad_norm": 0.47114476561546326, "learning_rate": 7.651e-05, "loss": 0.1812, "step": 7652 }, { "epoch": 0.0007653, "grad_norm": 0.3914787471294403, "learning_rate": 7.652e-05, "loss": 0.1664, "step": 7653 }, { "epoch": 0.0007654, "grad_norm": 0.4580937623977661, "learning_rate": 7.653e-05, "loss": 0.1665, "step": 7654 }, { "epoch": 0.0007655, "grad_norm": 0.3929605782032013, "learning_rate": 7.654e-05, "loss": 0.1534, "step": 7655 }, { "epoch": 0.0007656, "grad_norm": 0.4484695792198181, "learning_rate": 7.655e-05, "loss": 0.1812, "step": 7656 }, { "epoch": 0.0007657, "grad_norm": 0.43384209275245667, "learning_rate": 7.656e-05, "loss": 0.1854, "step": 7657 }, { "epoch": 0.0007658, "grad_norm": 0.38437792658805847, "learning_rate": 7.657e-05, "loss": 0.1582, "step": 7658 }, { "epoch": 0.0007659, "grad_norm": 0.33982738852500916, "learning_rate": 7.658e-05, "loss": 0.1604, "step": 7659 }, { "epoch": 0.000766, "grad_norm": 0.8198331594467163, "learning_rate": 7.659e-05, "loss": 0.2065, "step": 7660 }, { "epoch": 0.0007661, "grad_norm": 0.4300197660923004, "learning_rate": 7.66e-05, "loss": 0.161, "step": 7661 }, { "epoch": 0.0007662, "grad_norm": 0.40067654848098755, "learning_rate": 7.661e-05, "loss": 0.1666, "step": 7662 }, { "epoch": 0.0007663, "grad_norm": 0.36768773198127747, "learning_rate": 7.662000000000001e-05, "loss": 0.1655, "step": 7663 }, { "epoch": 0.0007664, "grad_norm": 0.3558520972728729, "learning_rate": 7.663e-05, "loss": 0.1475, "step": 7664 }, { "epoch": 0.0007665, "grad_norm": 0.5428350567817688, "learning_rate": 7.664e-05, "loss": 0.1729, "step": 7665 }, { "epoch": 0.0007666, "grad_norm": 0.48063817620277405, "learning_rate": 7.665e-05, "loss": 0.1938, "step": 7666 }, { "epoch": 0.0007667, "grad_norm": 0.3305480182170868, "learning_rate": 7.666e-05, "loss": 0.1467, "step": 7667 }, { "epoch": 0.0007668, "grad_norm": 0.7609458565711975, "learning_rate": 7.667e-05, "loss": 0.1814, "step": 7668 }, { "epoch": 0.0007669, "grad_norm": 0.3762044608592987, "learning_rate": 7.668e-05, "loss": 0.1622, "step": 7669 }, { "epoch": 0.000767, "grad_norm": 0.3550243079662323, "learning_rate": 7.669e-05, "loss": 0.1553, "step": 7670 }, { "epoch": 0.0007671, "grad_norm": 0.3320472240447998, "learning_rate": 7.670000000000001e-05, "loss": 0.1483, "step": 7671 }, { "epoch": 0.0007672, "grad_norm": 0.5858722925186157, "learning_rate": 7.671e-05, "loss": 0.1892, "step": 7672 }, { "epoch": 0.0007673, "grad_norm": 1.368833065032959, "learning_rate": 7.672e-05, "loss": 0.2322, "step": 7673 }, { "epoch": 0.0007674, "grad_norm": 0.532004714012146, "learning_rate": 7.673e-05, "loss": 0.1707, "step": 7674 }, { "epoch": 0.0007675, "grad_norm": 0.550662100315094, "learning_rate": 7.674e-05, "loss": 0.1891, "step": 7675 }, { "epoch": 0.0007676, "grad_norm": 0.8571083545684814, "learning_rate": 7.675e-05, "loss": 0.1967, "step": 7676 }, { "epoch": 0.0007677, "grad_norm": 0.5520954728126526, "learning_rate": 7.676e-05, "loss": 0.172, "step": 7677 }, { "epoch": 0.0007678, "grad_norm": 0.4285428524017334, "learning_rate": 7.677000000000001e-05, "loss": 0.1538, "step": 7678 }, { "epoch": 0.0007679, "grad_norm": 0.38580891489982605, "learning_rate": 7.678e-05, "loss": 0.147, "step": 7679 }, { "epoch": 0.000768, "grad_norm": 0.37791067361831665, "learning_rate": 7.679e-05, "loss": 0.1521, "step": 7680 }, { "epoch": 0.0007681, "grad_norm": 0.3506149351596832, "learning_rate": 7.680000000000001e-05, "loss": 0.1484, "step": 7681 }, { "epoch": 0.0007682, "grad_norm": 0.4228236973285675, "learning_rate": 7.680999999999999e-05, "loss": 0.1617, "step": 7682 }, { "epoch": 0.0007683, "grad_norm": 0.4166804850101471, "learning_rate": 7.682e-05, "loss": 0.1674, "step": 7683 }, { "epoch": 0.0007684, "grad_norm": 0.4382307529449463, "learning_rate": 7.683000000000001e-05, "loss": 0.1726, "step": 7684 }, { "epoch": 0.0007685, "grad_norm": 0.4771483838558197, "learning_rate": 7.684e-05, "loss": 0.1758, "step": 7685 }, { "epoch": 0.0007686, "grad_norm": 0.3528839945793152, "learning_rate": 7.685e-05, "loss": 0.1581, "step": 7686 }, { "epoch": 0.0007687, "grad_norm": 0.7994784116744995, "learning_rate": 7.686e-05, "loss": 0.1545, "step": 7687 }, { "epoch": 0.0007688, "grad_norm": 0.4543117582798004, "learning_rate": 7.687e-05, "loss": 0.1691, "step": 7688 }, { "epoch": 0.0007689, "grad_norm": 0.45544180274009705, "learning_rate": 7.688e-05, "loss": 0.1576, "step": 7689 }, { "epoch": 0.000769, "grad_norm": 0.432519793510437, "learning_rate": 7.689e-05, "loss": 0.1637, "step": 7690 }, { "epoch": 0.0007691, "grad_norm": 0.4242666959762573, "learning_rate": 7.69e-05, "loss": 0.163, "step": 7691 }, { "epoch": 0.0007692, "grad_norm": 0.38810381293296814, "learning_rate": 7.691e-05, "loss": 0.1603, "step": 7692 }, { "epoch": 0.0007693, "grad_norm": 0.3763027489185333, "learning_rate": 7.692e-05, "loss": 0.1527, "step": 7693 }, { "epoch": 0.0007694, "grad_norm": 0.39184290170669556, "learning_rate": 7.693e-05, "loss": 0.1674, "step": 7694 }, { "epoch": 0.0007695, "grad_norm": 0.35056430101394653, "learning_rate": 7.694e-05, "loss": 0.1562, "step": 7695 }, { "epoch": 0.0007696, "grad_norm": 0.3862243592739105, "learning_rate": 7.695e-05, "loss": 0.1631, "step": 7696 }, { "epoch": 0.0007697, "grad_norm": 0.3677322268486023, "learning_rate": 7.696000000000001e-05, "loss": 0.1675, "step": 7697 }, { "epoch": 0.0007698, "grad_norm": 0.37952372431755066, "learning_rate": 7.697e-05, "loss": 0.1625, "step": 7698 }, { "epoch": 0.0007699, "grad_norm": 0.6752346158027649, "learning_rate": 7.698e-05, "loss": 0.1904, "step": 7699 }, { "epoch": 0.00077, "grad_norm": 0.4253908097743988, "learning_rate": 7.699e-05, "loss": 0.168, "step": 7700 }, { "epoch": 0.0007701, "grad_norm": 0.3445245921611786, "learning_rate": 7.7e-05, "loss": 0.1594, "step": 7701 }, { "epoch": 0.0007702, "grad_norm": 0.37778040766716003, "learning_rate": 7.701e-05, "loss": 0.1606, "step": 7702 }, { "epoch": 0.0007703, "grad_norm": 0.3563236594200134, "learning_rate": 7.702e-05, "loss": 0.156, "step": 7703 }, { "epoch": 0.0007704, "grad_norm": 0.35258838534355164, "learning_rate": 7.703e-05, "loss": 0.1527, "step": 7704 }, { "epoch": 0.0007705, "grad_norm": 0.6172411441802979, "learning_rate": 7.704000000000001e-05, "loss": 0.1848, "step": 7705 }, { "epoch": 0.0007706, "grad_norm": 0.7439136505126953, "learning_rate": 7.705e-05, "loss": 0.2045, "step": 7706 }, { "epoch": 0.0007707, "grad_norm": 0.43097370862960815, "learning_rate": 7.706e-05, "loss": 0.1796, "step": 7707 }, { "epoch": 0.0007708, "grad_norm": 0.41762420535087585, "learning_rate": 7.707e-05, "loss": 0.1705, "step": 7708 }, { "epoch": 0.0007709, "grad_norm": 0.5003089308738708, "learning_rate": 7.708e-05, "loss": 0.1791, "step": 7709 }, { "epoch": 0.000771, "grad_norm": 0.3558039963245392, "learning_rate": 7.709e-05, "loss": 0.1578, "step": 7710 }, { "epoch": 0.0007711, "grad_norm": 0.436168372631073, "learning_rate": 7.71e-05, "loss": 0.1694, "step": 7711 }, { "epoch": 0.0007712, "grad_norm": 0.5260476469993591, "learning_rate": 7.711000000000001e-05, "loss": 0.162, "step": 7712 }, { "epoch": 0.0007713, "grad_norm": 0.45862051844596863, "learning_rate": 7.712e-05, "loss": 0.1552, "step": 7713 }, { "epoch": 0.0007714, "grad_norm": 0.3381466567516327, "learning_rate": 7.713e-05, "loss": 0.1447, "step": 7714 }, { "epoch": 0.0007715, "grad_norm": 0.3526186943054199, "learning_rate": 7.714000000000001e-05, "loss": 0.145, "step": 7715 }, { "epoch": 0.0007716, "grad_norm": 0.34365805983543396, "learning_rate": 7.714999999999999e-05, "loss": 0.1533, "step": 7716 }, { "epoch": 0.0007717, "grad_norm": 0.36505240201950073, "learning_rate": 7.716e-05, "loss": 0.1576, "step": 7717 }, { "epoch": 0.0007718, "grad_norm": 0.5150579214096069, "learning_rate": 7.717000000000001e-05, "loss": 0.1769, "step": 7718 }, { "epoch": 0.0007719, "grad_norm": 0.3398381471633911, "learning_rate": 7.718e-05, "loss": 0.1509, "step": 7719 }, { "epoch": 0.000772, "grad_norm": 0.3642643094062805, "learning_rate": 7.719e-05, "loss": 0.1587, "step": 7720 }, { "epoch": 0.0007721, "grad_norm": 1.7854269742965698, "learning_rate": 7.72e-05, "loss": 0.281, "step": 7721 }, { "epoch": 0.0007722, "grad_norm": 0.364483505487442, "learning_rate": 7.721e-05, "loss": 0.1516, "step": 7722 }, { "epoch": 0.0007723, "grad_norm": 0.6805715560913086, "learning_rate": 7.722e-05, "loss": 0.1942, "step": 7723 }, { "epoch": 0.0007724, "grad_norm": 0.39517566561698914, "learning_rate": 7.723e-05, "loss": 0.1525, "step": 7724 }, { "epoch": 0.0007725, "grad_norm": 0.37582632899284363, "learning_rate": 7.724e-05, "loss": 0.1567, "step": 7725 }, { "epoch": 0.0007726, "grad_norm": 0.4059215188026428, "learning_rate": 7.725e-05, "loss": 0.1572, "step": 7726 }, { "epoch": 0.0007727, "grad_norm": 0.4789154529571533, "learning_rate": 7.726e-05, "loss": 0.1592, "step": 7727 }, { "epoch": 0.0007728, "grad_norm": 0.6848322749137878, "learning_rate": 7.727e-05, "loss": 0.1976, "step": 7728 }, { "epoch": 0.0007729, "grad_norm": 0.3184192180633545, "learning_rate": 7.728e-05, "loss": 0.1453, "step": 7729 }, { "epoch": 0.000773, "grad_norm": 0.35173535346984863, "learning_rate": 7.729e-05, "loss": 0.1531, "step": 7730 }, { "epoch": 0.0007731, "grad_norm": 0.3543420732021332, "learning_rate": 7.730000000000001e-05, "loss": 0.1516, "step": 7731 }, { "epoch": 0.0007732, "grad_norm": 0.3434199094772339, "learning_rate": 7.731e-05, "loss": 0.1539, "step": 7732 }, { "epoch": 0.0007733, "grad_norm": 0.3729701638221741, "learning_rate": 7.732e-05, "loss": 0.156, "step": 7733 }, { "epoch": 0.0007734, "grad_norm": 0.3544479310512543, "learning_rate": 7.733e-05, "loss": 0.1488, "step": 7734 }, { "epoch": 0.0007735, "grad_norm": 0.46375659108161926, "learning_rate": 7.734e-05, "loss": 0.1637, "step": 7735 }, { "epoch": 0.0007736, "grad_norm": 0.3291948437690735, "learning_rate": 7.735e-05, "loss": 0.1565, "step": 7736 }, { "epoch": 0.0007737, "grad_norm": 0.327825665473938, "learning_rate": 7.736e-05, "loss": 0.1416, "step": 7737 }, { "epoch": 0.0007738, "grad_norm": 0.5666698217391968, "learning_rate": 7.737e-05, "loss": 0.1598, "step": 7738 }, { "epoch": 0.0007739, "grad_norm": 0.38737985491752625, "learning_rate": 7.738000000000001e-05, "loss": 0.1444, "step": 7739 }, { "epoch": 0.000774, "grad_norm": 0.38880079984664917, "learning_rate": 7.739e-05, "loss": 0.1627, "step": 7740 }, { "epoch": 0.0007741, "grad_norm": 0.4217063784599304, "learning_rate": 7.74e-05, "loss": 0.1664, "step": 7741 }, { "epoch": 0.0007742, "grad_norm": 0.32560673356056213, "learning_rate": 7.741e-05, "loss": 0.1459, "step": 7742 }, { "epoch": 0.0007743, "grad_norm": 0.5402155518531799, "learning_rate": 7.742e-05, "loss": 0.1838, "step": 7743 }, { "epoch": 0.0007744, "grad_norm": 0.3342635929584503, "learning_rate": 7.743e-05, "loss": 0.1528, "step": 7744 }, { "epoch": 0.0007745, "grad_norm": 0.9251406192779541, "learning_rate": 7.744e-05, "loss": 0.2164, "step": 7745 }, { "epoch": 0.0007746, "grad_norm": 0.34178224205970764, "learning_rate": 7.745000000000001e-05, "loss": 0.149, "step": 7746 }, { "epoch": 0.0007747, "grad_norm": 0.7903544902801514, "learning_rate": 7.746e-05, "loss": 0.1927, "step": 7747 }, { "epoch": 0.0007748, "grad_norm": 0.428143173456192, "learning_rate": 7.747e-05, "loss": 0.1549, "step": 7748 }, { "epoch": 0.0007749, "grad_norm": 0.44141685962677, "learning_rate": 7.748000000000001e-05, "loss": 0.1468, "step": 7749 }, { "epoch": 0.000775, "grad_norm": 0.5193094611167908, "learning_rate": 7.748999999999999e-05, "loss": 0.1578, "step": 7750 }, { "epoch": 0.0007751, "grad_norm": 0.41440680623054504, "learning_rate": 7.75e-05, "loss": 0.1541, "step": 7751 }, { "epoch": 0.0007752, "grad_norm": 0.5027520060539246, "learning_rate": 7.751000000000001e-05, "loss": 0.1676, "step": 7752 }, { "epoch": 0.0007753, "grad_norm": 0.4081827998161316, "learning_rate": 7.752e-05, "loss": 0.1498, "step": 7753 }, { "epoch": 0.0007754, "grad_norm": 0.316013365983963, "learning_rate": 7.753e-05, "loss": 0.142, "step": 7754 }, { "epoch": 0.0007755, "grad_norm": 0.6248079538345337, "learning_rate": 7.754e-05, "loss": 0.1661, "step": 7755 }, { "epoch": 0.0007756, "grad_norm": 0.39839357137680054, "learning_rate": 7.755e-05, "loss": 0.1584, "step": 7756 }, { "epoch": 0.0007757, "grad_norm": 0.4424418807029724, "learning_rate": 7.756e-05, "loss": 0.165, "step": 7757 }, { "epoch": 0.0007758, "grad_norm": 0.4060595631599426, "learning_rate": 7.757e-05, "loss": 0.1548, "step": 7758 }, { "epoch": 0.0007759, "grad_norm": 0.4461621046066284, "learning_rate": 7.758e-05, "loss": 0.1583, "step": 7759 }, { "epoch": 0.000776, "grad_norm": 0.41382384300231934, "learning_rate": 7.759e-05, "loss": 0.1623, "step": 7760 }, { "epoch": 0.0007761, "grad_norm": 0.608735203742981, "learning_rate": 7.76e-05, "loss": 0.1565, "step": 7761 }, { "epoch": 0.0007762, "grad_norm": 0.4113566279411316, "learning_rate": 7.761e-05, "loss": 0.1598, "step": 7762 }, { "epoch": 0.0007763, "grad_norm": 0.3585749864578247, "learning_rate": 7.762e-05, "loss": 0.1584, "step": 7763 }, { "epoch": 0.0007764, "grad_norm": 0.4358258545398712, "learning_rate": 7.763e-05, "loss": 0.1681, "step": 7764 }, { "epoch": 0.0007765, "grad_norm": 0.37854641675949097, "learning_rate": 7.764000000000001e-05, "loss": 0.1599, "step": 7765 }, { "epoch": 0.0007766, "grad_norm": 0.37787294387817383, "learning_rate": 7.765e-05, "loss": 0.1555, "step": 7766 }, { "epoch": 0.0007767, "grad_norm": 0.3425155282020569, "learning_rate": 7.766e-05, "loss": 0.158, "step": 7767 }, { "epoch": 0.0007768, "grad_norm": 0.3806110918521881, "learning_rate": 7.767000000000001e-05, "loss": 0.1537, "step": 7768 }, { "epoch": 0.0007769, "grad_norm": 0.5556537508964539, "learning_rate": 7.768e-05, "loss": 0.1769, "step": 7769 }, { "epoch": 0.000777, "grad_norm": 0.40484434366226196, "learning_rate": 7.769e-05, "loss": 0.1725, "step": 7770 }, { "epoch": 0.0007771, "grad_norm": 0.3824002146720886, "learning_rate": 7.77e-05, "loss": 0.1613, "step": 7771 }, { "epoch": 0.0007772, "grad_norm": 0.3158770203590393, "learning_rate": 7.771e-05, "loss": 0.141, "step": 7772 }, { "epoch": 0.0007773, "grad_norm": 0.3717173933982849, "learning_rate": 7.772000000000001e-05, "loss": 0.1532, "step": 7773 }, { "epoch": 0.0007774, "grad_norm": 0.4274844229221344, "learning_rate": 7.773e-05, "loss": 0.1587, "step": 7774 }, { "epoch": 0.0007775, "grad_norm": 0.4046543836593628, "learning_rate": 7.774e-05, "loss": 0.1643, "step": 7775 }, { "epoch": 0.0007776, "grad_norm": 0.3226039409637451, "learning_rate": 7.775e-05, "loss": 0.1456, "step": 7776 }, { "epoch": 0.0007777, "grad_norm": 0.5570123195648193, "learning_rate": 7.776e-05, "loss": 0.1631, "step": 7777 }, { "epoch": 0.0007778, "grad_norm": 0.3571377098560333, "learning_rate": 7.777e-05, "loss": 0.155, "step": 7778 }, { "epoch": 0.0007779, "grad_norm": 1.138600468635559, "learning_rate": 7.778e-05, "loss": 0.2061, "step": 7779 }, { "epoch": 0.000778, "grad_norm": 0.3432740867137909, "learning_rate": 7.779000000000001e-05, "loss": 0.1429, "step": 7780 }, { "epoch": 0.0007781, "grad_norm": 0.4284251928329468, "learning_rate": 7.78e-05, "loss": 0.1482, "step": 7781 }, { "epoch": 0.0007782, "grad_norm": 0.4238644540309906, "learning_rate": 7.781e-05, "loss": 0.144, "step": 7782 }, { "epoch": 0.0007783, "grad_norm": 0.44244638085365295, "learning_rate": 7.782000000000001e-05, "loss": 0.1576, "step": 7783 }, { "epoch": 0.0007784, "grad_norm": 0.37761566042900085, "learning_rate": 7.782999999999999e-05, "loss": 0.1478, "step": 7784 }, { "epoch": 0.0007785, "grad_norm": 0.4310877025127411, "learning_rate": 7.784e-05, "loss": 0.1544, "step": 7785 }, { "epoch": 0.0007786, "grad_norm": 0.3911275863647461, "learning_rate": 7.785000000000001e-05, "loss": 0.1542, "step": 7786 }, { "epoch": 0.0007787, "grad_norm": 0.3833433985710144, "learning_rate": 7.786e-05, "loss": 0.1534, "step": 7787 }, { "epoch": 0.0007788, "grad_norm": 0.3688901364803314, "learning_rate": 7.787e-05, "loss": 0.1469, "step": 7788 }, { "epoch": 0.0007789, "grad_norm": 0.39441004395484924, "learning_rate": 7.788e-05, "loss": 0.1512, "step": 7789 }, { "epoch": 0.000779, "grad_norm": 0.37613943219184875, "learning_rate": 7.789e-05, "loss": 0.1494, "step": 7790 }, { "epoch": 0.0007791, "grad_norm": 0.41287046670913696, "learning_rate": 7.79e-05, "loss": 0.1469, "step": 7791 }, { "epoch": 0.0007792, "grad_norm": 0.35042914748191833, "learning_rate": 7.791e-05, "loss": 0.1486, "step": 7792 }, { "epoch": 0.0007793, "grad_norm": 0.3768109977245331, "learning_rate": 7.792e-05, "loss": 0.152, "step": 7793 }, { "epoch": 0.0007794, "grad_norm": 0.33065614104270935, "learning_rate": 7.793e-05, "loss": 0.1479, "step": 7794 }, { "epoch": 0.0007795, "grad_norm": 0.3601537048816681, "learning_rate": 7.794e-05, "loss": 0.1597, "step": 7795 }, { "epoch": 0.0007796, "grad_norm": 0.34998899698257446, "learning_rate": 7.795e-05, "loss": 0.1503, "step": 7796 }, { "epoch": 0.0007797, "grad_norm": 0.3146880865097046, "learning_rate": 7.796e-05, "loss": 0.1412, "step": 7797 }, { "epoch": 0.0007798, "grad_norm": 0.3786034882068634, "learning_rate": 7.797e-05, "loss": 0.1472, "step": 7798 }, { "epoch": 0.0007799, "grad_norm": 0.3339554965496063, "learning_rate": 7.798000000000001e-05, "loss": 0.1521, "step": 7799 }, { "epoch": 0.00078, "grad_norm": 0.3790935277938843, "learning_rate": 7.799e-05, "loss": 0.1631, "step": 7800 }, { "epoch": 0.0007801, "grad_norm": 0.39043620228767395, "learning_rate": 7.8e-05, "loss": 0.1442, "step": 7801 }, { "epoch": 0.0007802, "grad_norm": 0.3289644420146942, "learning_rate": 7.801000000000001e-05, "loss": 0.1454, "step": 7802 }, { "epoch": 0.0007803, "grad_norm": 0.33908987045288086, "learning_rate": 7.802e-05, "loss": 0.1498, "step": 7803 }, { "epoch": 0.0007804, "grad_norm": 0.5826635360717773, "learning_rate": 7.803e-05, "loss": 0.1678, "step": 7804 }, { "epoch": 0.0007805, "grad_norm": 0.32743117213249207, "learning_rate": 7.804e-05, "loss": 0.15, "step": 7805 }, { "epoch": 0.0007806, "grad_norm": 0.35113999247550964, "learning_rate": 7.805e-05, "loss": 0.1476, "step": 7806 }, { "epoch": 0.0007807, "grad_norm": 0.37699851393699646, "learning_rate": 7.806000000000001e-05, "loss": 0.1589, "step": 7807 }, { "epoch": 0.0007808, "grad_norm": 0.3135019838809967, "learning_rate": 7.807e-05, "loss": 0.147, "step": 7808 }, { "epoch": 0.0007809, "grad_norm": 0.5410703420639038, "learning_rate": 7.808e-05, "loss": 0.1752, "step": 7809 }, { "epoch": 0.000781, "grad_norm": 0.3744533360004425, "learning_rate": 7.809e-05, "loss": 0.1543, "step": 7810 }, { "epoch": 0.0007811, "grad_norm": 0.3290867507457733, "learning_rate": 7.81e-05, "loss": 0.1504, "step": 7811 }, { "epoch": 0.0007812, "grad_norm": 0.34372928738594055, "learning_rate": 7.811e-05, "loss": 0.1526, "step": 7812 }, { "epoch": 0.0007813, "grad_norm": 0.31069982051849365, "learning_rate": 7.812e-05, "loss": 0.1418, "step": 7813 }, { "epoch": 0.0007814, "grad_norm": 0.40363672375679016, "learning_rate": 7.813000000000001e-05, "loss": 0.1586, "step": 7814 }, { "epoch": 0.0007815, "grad_norm": 0.40617135167121887, "learning_rate": 7.814e-05, "loss": 0.1495, "step": 7815 }, { "epoch": 0.0007816, "grad_norm": 0.40958961844444275, "learning_rate": 7.815e-05, "loss": 0.147, "step": 7816 }, { "epoch": 0.0007817, "grad_norm": 0.4198823869228363, "learning_rate": 7.816000000000001e-05, "loss": 0.1656, "step": 7817 }, { "epoch": 0.0007818, "grad_norm": 0.3430757522583008, "learning_rate": 7.816999999999999e-05, "loss": 0.1426, "step": 7818 }, { "epoch": 0.0007819, "grad_norm": 0.4560718834400177, "learning_rate": 7.818e-05, "loss": 0.1598, "step": 7819 }, { "epoch": 0.000782, "grad_norm": 0.3987734615802765, "learning_rate": 7.819000000000001e-05, "loss": 0.1591, "step": 7820 }, { "epoch": 0.0007821, "grad_norm": 0.33172231912612915, "learning_rate": 7.82e-05, "loss": 0.1405, "step": 7821 }, { "epoch": 0.0007822, "grad_norm": 0.33455225825309753, "learning_rate": 7.821e-05, "loss": 0.1505, "step": 7822 }, { "epoch": 0.0007823, "grad_norm": 0.33328595757484436, "learning_rate": 7.822e-05, "loss": 0.1454, "step": 7823 }, { "epoch": 0.0007824, "grad_norm": 0.5431308746337891, "learning_rate": 7.823e-05, "loss": 0.1887, "step": 7824 }, { "epoch": 0.0007825, "grad_norm": 0.3220846951007843, "learning_rate": 7.824e-05, "loss": 0.1449, "step": 7825 }, { "epoch": 0.0007826, "grad_norm": 0.37850919365882874, "learning_rate": 7.825e-05, "loss": 0.1575, "step": 7826 }, { "epoch": 0.0007827, "grad_norm": 0.3572224974632263, "learning_rate": 7.826e-05, "loss": 0.1514, "step": 7827 }, { "epoch": 0.0007828, "grad_norm": 0.3942662179470062, "learning_rate": 7.827e-05, "loss": 0.1654, "step": 7828 }, { "epoch": 0.0007829, "grad_norm": 0.3946959972381592, "learning_rate": 7.828e-05, "loss": 0.1572, "step": 7829 }, { "epoch": 0.000783, "grad_norm": 0.3043114244937897, "learning_rate": 7.829e-05, "loss": 0.1438, "step": 7830 }, { "epoch": 0.0007831, "grad_norm": 0.41395580768585205, "learning_rate": 7.83e-05, "loss": 0.1577, "step": 7831 }, { "epoch": 0.0007832, "grad_norm": 0.3440980613231659, "learning_rate": 7.831e-05, "loss": 0.156, "step": 7832 }, { "epoch": 0.0007833, "grad_norm": 0.406460165977478, "learning_rate": 7.832000000000001e-05, "loss": 0.1556, "step": 7833 }, { "epoch": 0.0007834, "grad_norm": 0.5229368209838867, "learning_rate": 7.833e-05, "loss": 0.1759, "step": 7834 }, { "epoch": 0.0007835, "grad_norm": 0.42203351855278015, "learning_rate": 7.834e-05, "loss": 0.1543, "step": 7835 }, { "epoch": 0.0007836, "grad_norm": 0.3226388394832611, "learning_rate": 7.835000000000001e-05, "loss": 0.1447, "step": 7836 }, { "epoch": 0.0007837, "grad_norm": 0.3410513997077942, "learning_rate": 7.836e-05, "loss": 0.1506, "step": 7837 }, { "epoch": 0.0007838, "grad_norm": 0.34009653329849243, "learning_rate": 7.837e-05, "loss": 0.1454, "step": 7838 }, { "epoch": 0.0007839, "grad_norm": 0.47929927706718445, "learning_rate": 7.838e-05, "loss": 0.1639, "step": 7839 }, { "epoch": 0.000784, "grad_norm": 0.5508768558502197, "learning_rate": 7.839e-05, "loss": 0.1681, "step": 7840 }, { "epoch": 0.0007841, "grad_norm": 0.8741626143455505, "learning_rate": 7.840000000000001e-05, "loss": 0.1545, "step": 7841 }, { "epoch": 0.0007842, "grad_norm": 0.7747453451156616, "learning_rate": 7.841e-05, "loss": 0.1639, "step": 7842 }, { "epoch": 0.0007843, "grad_norm": 0.3523169755935669, "learning_rate": 7.842e-05, "loss": 0.1378, "step": 7843 }, { "epoch": 0.0007844, "grad_norm": 0.36965256929397583, "learning_rate": 7.843e-05, "loss": 0.1441, "step": 7844 }, { "epoch": 0.0007845, "grad_norm": 0.49674296379089355, "learning_rate": 7.844e-05, "loss": 0.1678, "step": 7845 }, { "epoch": 0.0007846, "grad_norm": 0.40077468752861023, "learning_rate": 7.845e-05, "loss": 0.1493, "step": 7846 }, { "epoch": 0.0007847, "grad_norm": 0.5569064617156982, "learning_rate": 7.846e-05, "loss": 0.1643, "step": 7847 }, { "epoch": 0.0007848, "grad_norm": 0.7520188093185425, "learning_rate": 7.847000000000001e-05, "loss": 0.1694, "step": 7848 }, { "epoch": 0.0007849, "grad_norm": 0.3864554762840271, "learning_rate": 7.848e-05, "loss": 0.1467, "step": 7849 }, { "epoch": 0.000785, "grad_norm": 0.4093283414840698, "learning_rate": 7.849e-05, "loss": 0.1493, "step": 7850 }, { "epoch": 0.0007851, "grad_norm": 0.36037686467170715, "learning_rate": 7.850000000000001e-05, "loss": 0.1345, "step": 7851 }, { "epoch": 0.0007852, "grad_norm": 0.4377402365207672, "learning_rate": 7.850999999999999e-05, "loss": 0.147, "step": 7852 }, { "epoch": 0.0007853, "grad_norm": 0.5319004058837891, "learning_rate": 7.852e-05, "loss": 0.1862, "step": 7853 }, { "epoch": 0.0007854, "grad_norm": 0.3483973741531372, "learning_rate": 7.853000000000001e-05, "loss": 0.1357, "step": 7854 }, { "epoch": 0.0007855, "grad_norm": 0.7695083022117615, "learning_rate": 7.854e-05, "loss": 0.2, "step": 7855 }, { "epoch": 0.0007856, "grad_norm": 0.37926194071769714, "learning_rate": 7.855e-05, "loss": 0.1421, "step": 7856 }, { "epoch": 0.0007857, "grad_norm": 0.4063139259815216, "learning_rate": 7.856000000000001e-05, "loss": 0.1527, "step": 7857 }, { "epoch": 0.0007858, "grad_norm": 0.5253282785415649, "learning_rate": 7.857e-05, "loss": 0.1573, "step": 7858 }, { "epoch": 0.0007859, "grad_norm": 0.3392678499221802, "learning_rate": 7.858e-05, "loss": 0.1376, "step": 7859 }, { "epoch": 0.000786, "grad_norm": 0.41936904191970825, "learning_rate": 7.859e-05, "loss": 0.1471, "step": 7860 }, { "epoch": 0.0007861, "grad_norm": 0.6929224729537964, "learning_rate": 7.86e-05, "loss": 0.1624, "step": 7861 }, { "epoch": 0.0007862, "grad_norm": 0.35438454151153564, "learning_rate": 7.861e-05, "loss": 0.1478, "step": 7862 }, { "epoch": 0.0007863, "grad_norm": 0.3281148374080658, "learning_rate": 7.862e-05, "loss": 0.1339, "step": 7863 }, { "epoch": 0.0007864, "grad_norm": 0.3948661983013153, "learning_rate": 7.863e-05, "loss": 0.1462, "step": 7864 }, { "epoch": 0.0007865, "grad_norm": 0.5109783411026001, "learning_rate": 7.864e-05, "loss": 0.1672, "step": 7865 }, { "epoch": 0.0007866, "grad_norm": 0.48566001653671265, "learning_rate": 7.865e-05, "loss": 0.1656, "step": 7866 }, { "epoch": 0.0007867, "grad_norm": 0.3471790552139282, "learning_rate": 7.866e-05, "loss": 0.1484, "step": 7867 }, { "epoch": 0.0007868, "grad_norm": 0.322068989276886, "learning_rate": 7.867e-05, "loss": 0.1382, "step": 7868 }, { "epoch": 0.0007869, "grad_norm": 0.2925344407558441, "learning_rate": 7.868e-05, "loss": 0.1327, "step": 7869 }, { "epoch": 0.000787, "grad_norm": 0.43323126435279846, "learning_rate": 7.869000000000001e-05, "loss": 0.151, "step": 7870 }, { "epoch": 0.0007871, "grad_norm": 0.3427598476409912, "learning_rate": 7.87e-05, "loss": 0.1414, "step": 7871 }, { "epoch": 0.0007872, "grad_norm": 0.36143046617507935, "learning_rate": 7.871e-05, "loss": 0.1459, "step": 7872 }, { "epoch": 0.0007873, "grad_norm": 0.4235434830188751, "learning_rate": 7.872e-05, "loss": 0.1543, "step": 7873 }, { "epoch": 0.0007874, "grad_norm": 0.32518136501312256, "learning_rate": 7.873e-05, "loss": 0.1453, "step": 7874 }, { "epoch": 0.0007875, "grad_norm": 0.30913761258125305, "learning_rate": 7.874000000000001e-05, "loss": 0.1451, "step": 7875 }, { "epoch": 0.0007876, "grad_norm": 0.3425231873989105, "learning_rate": 7.875e-05, "loss": 0.1501, "step": 7876 }, { "epoch": 0.0007877, "grad_norm": 0.4112080931663513, "learning_rate": 7.876e-05, "loss": 0.1632, "step": 7877 }, { "epoch": 0.0007878, "grad_norm": 1.0945180654525757, "learning_rate": 7.877e-05, "loss": 0.1748, "step": 7878 }, { "epoch": 0.0007879, "grad_norm": 0.31509843468666077, "learning_rate": 7.878e-05, "loss": 0.1437, "step": 7879 }, { "epoch": 0.000788, "grad_norm": 0.3834889233112335, "learning_rate": 7.879e-05, "loss": 0.1498, "step": 7880 }, { "epoch": 0.0007881, "grad_norm": 0.40916872024536133, "learning_rate": 7.88e-05, "loss": 0.1515, "step": 7881 }, { "epoch": 0.0007882, "grad_norm": 0.34113648533821106, "learning_rate": 7.881000000000001e-05, "loss": 0.1396, "step": 7882 }, { "epoch": 0.0007883, "grad_norm": 0.41003337502479553, "learning_rate": 7.882e-05, "loss": 0.1453, "step": 7883 }, { "epoch": 0.0007884, "grad_norm": 0.38141313195228577, "learning_rate": 7.883e-05, "loss": 0.1506, "step": 7884 }, { "epoch": 0.0007885, "grad_norm": 0.6405525803565979, "learning_rate": 7.884000000000001e-05, "loss": 0.1666, "step": 7885 }, { "epoch": 0.0007886, "grad_norm": 0.5458857417106628, "learning_rate": 7.884999999999999e-05, "loss": 0.1517, "step": 7886 }, { "epoch": 0.0007887, "grad_norm": 0.4375963509082794, "learning_rate": 7.886e-05, "loss": 0.1586, "step": 7887 }, { "epoch": 0.0007888, "grad_norm": 0.3580589294433594, "learning_rate": 7.887000000000001e-05, "loss": 0.1418, "step": 7888 }, { "epoch": 0.0007889, "grad_norm": 0.4851737916469574, "learning_rate": 7.887999999999999e-05, "loss": 0.1617, "step": 7889 }, { "epoch": 0.000789, "grad_norm": 0.3856046497821808, "learning_rate": 7.889e-05, "loss": 0.1459, "step": 7890 }, { "epoch": 0.0007891, "grad_norm": 0.3718718886375427, "learning_rate": 7.890000000000001e-05, "loss": 0.1462, "step": 7891 }, { "epoch": 0.0007892, "grad_norm": 0.37324395775794983, "learning_rate": 7.891e-05, "loss": 0.1523, "step": 7892 }, { "epoch": 0.0007893, "grad_norm": 0.5335335731506348, "learning_rate": 7.892e-05, "loss": 0.1814, "step": 7893 }, { "epoch": 0.0007894, "grad_norm": 0.3738335371017456, "learning_rate": 7.893e-05, "loss": 0.1495, "step": 7894 }, { "epoch": 0.0007895, "grad_norm": 0.30939367413520813, "learning_rate": 7.894e-05, "loss": 0.1425, "step": 7895 }, { "epoch": 0.0007896, "grad_norm": 0.4040488302707672, "learning_rate": 7.895e-05, "loss": 0.1473, "step": 7896 }, { "epoch": 0.0007897, "grad_norm": 0.36211928725242615, "learning_rate": 7.896e-05, "loss": 0.1423, "step": 7897 }, { "epoch": 0.0007898, "grad_norm": 0.32226863503456116, "learning_rate": 7.897e-05, "loss": 0.1384, "step": 7898 }, { "epoch": 0.0007899, "grad_norm": 0.473587304353714, "learning_rate": 7.898e-05, "loss": 0.1644, "step": 7899 }, { "epoch": 0.00079, "grad_norm": 4.129487037658691, "learning_rate": 7.899e-05, "loss": 0.2184, "step": 7900 }, { "epoch": 0.0007901, "grad_norm": 0.33910876512527466, "learning_rate": 7.9e-05, "loss": 0.1405, "step": 7901 }, { "epoch": 0.0007902, "grad_norm": 0.6521164774894714, "learning_rate": 7.901e-05, "loss": 0.1721, "step": 7902 }, { "epoch": 0.0007903, "grad_norm": 0.38239943981170654, "learning_rate": 7.902e-05, "loss": 0.1503, "step": 7903 }, { "epoch": 0.0007904, "grad_norm": 1.1751346588134766, "learning_rate": 7.903000000000001e-05, "loss": 0.2278, "step": 7904 }, { "epoch": 0.0007905, "grad_norm": 0.48078861832618713, "learning_rate": 7.904e-05, "loss": 0.1519, "step": 7905 }, { "epoch": 0.0007906, "grad_norm": 0.4788834750652313, "learning_rate": 7.905e-05, "loss": 0.1583, "step": 7906 }, { "epoch": 0.0007907, "grad_norm": 0.9449865818023682, "learning_rate": 7.906e-05, "loss": 0.1902, "step": 7907 }, { "epoch": 0.0007908, "grad_norm": 0.4132283926010132, "learning_rate": 7.907e-05, "loss": 0.1516, "step": 7908 }, { "epoch": 0.0007909, "grad_norm": 0.43124404549598694, "learning_rate": 7.908000000000001e-05, "loss": 0.1493, "step": 7909 }, { "epoch": 0.000791, "grad_norm": 0.4902573227882385, "learning_rate": 7.909e-05, "loss": 0.1521, "step": 7910 }, { "epoch": 0.0007911, "grad_norm": 0.4065714478492737, "learning_rate": 7.91e-05, "loss": 0.1421, "step": 7911 }, { "epoch": 0.0007912, "grad_norm": 0.5426554679870605, "learning_rate": 7.911e-05, "loss": 0.1595, "step": 7912 }, { "epoch": 0.0007913, "grad_norm": 0.43201830983161926, "learning_rate": 7.912e-05, "loss": 0.1473, "step": 7913 }, { "epoch": 0.0007914, "grad_norm": 0.5732458233833313, "learning_rate": 7.913e-05, "loss": 0.181, "step": 7914 }, { "epoch": 0.0007915, "grad_norm": 0.41392019391059875, "learning_rate": 7.914e-05, "loss": 0.155, "step": 7915 }, { "epoch": 0.0007916, "grad_norm": 0.42396053671836853, "learning_rate": 7.915000000000001e-05, "loss": 0.1511, "step": 7916 }, { "epoch": 0.0007917, "grad_norm": 0.3622252345085144, "learning_rate": 7.916e-05, "loss": 0.1421, "step": 7917 }, { "epoch": 0.0007918, "grad_norm": 0.4684685468673706, "learning_rate": 7.917e-05, "loss": 0.1606, "step": 7918 }, { "epoch": 0.0007919, "grad_norm": 0.38661402463912964, "learning_rate": 7.918000000000001e-05, "loss": 0.1548, "step": 7919 }, { "epoch": 0.000792, "grad_norm": 0.3429745137691498, "learning_rate": 7.918999999999999e-05, "loss": 0.1407, "step": 7920 }, { "epoch": 0.0007921, "grad_norm": 0.30344387888908386, "learning_rate": 7.92e-05, "loss": 0.1328, "step": 7921 }, { "epoch": 0.0007922, "grad_norm": 0.3710497319698334, "learning_rate": 7.921000000000001e-05, "loss": 0.1435, "step": 7922 }, { "epoch": 0.0007923, "grad_norm": 0.37194472551345825, "learning_rate": 7.921999999999999e-05, "loss": 0.144, "step": 7923 }, { "epoch": 0.0007924, "grad_norm": 0.42788729071617126, "learning_rate": 7.923e-05, "loss": 0.1478, "step": 7924 }, { "epoch": 0.0007925, "grad_norm": 0.38140884041786194, "learning_rate": 7.924000000000001e-05, "loss": 0.1467, "step": 7925 }, { "epoch": 0.0007926, "grad_norm": 0.32990655303001404, "learning_rate": 7.925e-05, "loss": 0.1461, "step": 7926 }, { "epoch": 0.0007927, "grad_norm": 0.2901013195514679, "learning_rate": 7.926e-05, "loss": 0.1317, "step": 7927 }, { "epoch": 0.0007928, "grad_norm": 0.328251451253891, "learning_rate": 7.927e-05, "loss": 0.1436, "step": 7928 }, { "epoch": 0.0007929, "grad_norm": 0.6993154287338257, "learning_rate": 7.928e-05, "loss": 0.1659, "step": 7929 }, { "epoch": 0.000793, "grad_norm": 0.4166194498538971, "learning_rate": 7.929e-05, "loss": 0.162, "step": 7930 }, { "epoch": 0.0007931, "grad_norm": 0.3751193583011627, "learning_rate": 7.93e-05, "loss": 0.1506, "step": 7931 }, { "epoch": 0.0007932, "grad_norm": 0.3409431278705597, "learning_rate": 7.931e-05, "loss": 0.1422, "step": 7932 }, { "epoch": 0.0007933, "grad_norm": 0.4018118977546692, "learning_rate": 7.932e-05, "loss": 0.1655, "step": 7933 }, { "epoch": 0.0007934, "grad_norm": 0.5279244780540466, "learning_rate": 7.933e-05, "loss": 0.1791, "step": 7934 }, { "epoch": 0.0007935, "grad_norm": 0.31002724170684814, "learning_rate": 7.934e-05, "loss": 0.1406, "step": 7935 }, { "epoch": 0.0007936, "grad_norm": 0.30614563822746277, "learning_rate": 7.935e-05, "loss": 0.134, "step": 7936 }, { "epoch": 0.0007937, "grad_norm": 0.4066023528575897, "learning_rate": 7.936e-05, "loss": 0.1545, "step": 7937 }, { "epoch": 0.0007938, "grad_norm": 0.315294086933136, "learning_rate": 7.937000000000001e-05, "loss": 0.1436, "step": 7938 }, { "epoch": 0.0007939, "grad_norm": 0.3314513564109802, "learning_rate": 7.938e-05, "loss": 0.1362, "step": 7939 }, { "epoch": 0.000794, "grad_norm": 0.3202115297317505, "learning_rate": 7.939e-05, "loss": 0.1483, "step": 7940 }, { "epoch": 0.0007941, "grad_norm": 0.35739099979400635, "learning_rate": 7.94e-05, "loss": 0.1511, "step": 7941 }, { "epoch": 0.0007942, "grad_norm": 0.5248399972915649, "learning_rate": 7.941e-05, "loss": 0.1779, "step": 7942 }, { "epoch": 0.0007943, "grad_norm": 0.3720186948776245, "learning_rate": 7.942000000000001e-05, "loss": 0.1468, "step": 7943 }, { "epoch": 0.0007944, "grad_norm": 0.34426867961883545, "learning_rate": 7.943e-05, "loss": 0.1442, "step": 7944 }, { "epoch": 0.0007945, "grad_norm": 0.3085344135761261, "learning_rate": 7.944e-05, "loss": 0.14, "step": 7945 }, { "epoch": 0.0007946, "grad_norm": 0.30736759305000305, "learning_rate": 7.945e-05, "loss": 0.1423, "step": 7946 }, { "epoch": 0.0007947, "grad_norm": 0.4966878294944763, "learning_rate": 7.946e-05, "loss": 0.1608, "step": 7947 }, { "epoch": 0.0007948, "grad_norm": 0.320726215839386, "learning_rate": 7.947e-05, "loss": 0.1387, "step": 7948 }, { "epoch": 0.0007949, "grad_norm": 0.32221245765686035, "learning_rate": 7.948e-05, "loss": 0.1467, "step": 7949 }, { "epoch": 0.000795, "grad_norm": 0.3078095614910126, "learning_rate": 7.949000000000001e-05, "loss": 0.1463, "step": 7950 }, { "epoch": 0.0007951, "grad_norm": 0.3305199444293976, "learning_rate": 7.95e-05, "loss": 0.1338, "step": 7951 }, { "epoch": 0.0007952, "grad_norm": 0.31949639320373535, "learning_rate": 7.951e-05, "loss": 0.1353, "step": 7952 }, { "epoch": 0.0007953, "grad_norm": 0.2862573266029358, "learning_rate": 7.952000000000001e-05, "loss": 0.1323, "step": 7953 }, { "epoch": 0.0007954, "grad_norm": 0.281089723110199, "learning_rate": 7.952999999999999e-05, "loss": 0.1327, "step": 7954 }, { "epoch": 0.0007955, "grad_norm": 0.2934763729572296, "learning_rate": 7.954e-05, "loss": 0.1378, "step": 7955 }, { "epoch": 0.0007956, "grad_norm": 0.6798914670944214, "learning_rate": 7.955000000000001e-05, "loss": 0.1608, "step": 7956 }, { "epoch": 0.0007957, "grad_norm": 0.28892913460731506, "learning_rate": 7.955999999999999e-05, "loss": 0.137, "step": 7957 }, { "epoch": 0.0007958, "grad_norm": 0.3551887273788452, "learning_rate": 7.957e-05, "loss": 0.1418, "step": 7958 }, { "epoch": 0.0007959, "grad_norm": 0.3428361117839813, "learning_rate": 7.958000000000001e-05, "loss": 0.1379, "step": 7959 }, { "epoch": 0.000796, "grad_norm": 0.320889413356781, "learning_rate": 7.959e-05, "loss": 0.1445, "step": 7960 }, { "epoch": 0.0007961, "grad_norm": 0.30409809947013855, "learning_rate": 7.96e-05, "loss": 0.1387, "step": 7961 }, { "epoch": 0.0007962, "grad_norm": 0.3267022669315338, "learning_rate": 7.961e-05, "loss": 0.1296, "step": 7962 }, { "epoch": 0.0007963, "grad_norm": 0.32833611965179443, "learning_rate": 7.962e-05, "loss": 0.1385, "step": 7963 }, { "epoch": 0.0007964, "grad_norm": 0.37005800008773804, "learning_rate": 7.963e-05, "loss": 0.1472, "step": 7964 }, { "epoch": 0.0007965, "grad_norm": 0.3744034171104431, "learning_rate": 7.964e-05, "loss": 0.1439, "step": 7965 }, { "epoch": 0.0007966, "grad_norm": 0.3019333779811859, "learning_rate": 7.965e-05, "loss": 0.1337, "step": 7966 }, { "epoch": 0.0007967, "grad_norm": 0.3061756491661072, "learning_rate": 7.966e-05, "loss": 0.1347, "step": 7967 }, { "epoch": 0.0007968, "grad_norm": 0.32018759846687317, "learning_rate": 7.967e-05, "loss": 0.1421, "step": 7968 }, { "epoch": 0.0007969, "grad_norm": 0.28421124815940857, "learning_rate": 7.968e-05, "loss": 0.1273, "step": 7969 }, { "epoch": 0.000797, "grad_norm": 0.32791009545326233, "learning_rate": 7.969e-05, "loss": 0.1407, "step": 7970 }, { "epoch": 0.0007971, "grad_norm": 0.30136266350746155, "learning_rate": 7.97e-05, "loss": 0.1366, "step": 7971 }, { "epoch": 0.0007972, "grad_norm": 0.3345014750957489, "learning_rate": 7.971000000000001e-05, "loss": 0.1399, "step": 7972 }, { "epoch": 0.0007973, "grad_norm": 0.5534088015556335, "learning_rate": 7.972e-05, "loss": 0.1508, "step": 7973 }, { "epoch": 0.0007974, "grad_norm": 0.3192349076271057, "learning_rate": 7.973e-05, "loss": 0.1401, "step": 7974 }, { "epoch": 0.0007975, "grad_norm": 0.3512609004974365, "learning_rate": 7.974e-05, "loss": 0.1338, "step": 7975 }, { "epoch": 0.0007976, "grad_norm": 0.3142569363117218, "learning_rate": 7.975e-05, "loss": 0.1385, "step": 7976 }, { "epoch": 0.0007977, "grad_norm": 0.3346051275730133, "learning_rate": 7.976000000000001e-05, "loss": 0.1378, "step": 7977 }, { "epoch": 0.0007978, "grad_norm": 0.5163511037826538, "learning_rate": 7.977e-05, "loss": 0.148, "step": 7978 }, { "epoch": 0.0007979, "grad_norm": 0.322229266166687, "learning_rate": 7.978e-05, "loss": 0.1316, "step": 7979 }, { "epoch": 0.000798, "grad_norm": 0.2973017692565918, "learning_rate": 7.979000000000001e-05, "loss": 0.134, "step": 7980 }, { "epoch": 0.0007981, "grad_norm": 0.3267301023006439, "learning_rate": 7.98e-05, "loss": 0.1378, "step": 7981 }, { "epoch": 0.0007982, "grad_norm": 0.656015932559967, "learning_rate": 7.981e-05, "loss": 0.1586, "step": 7982 }, { "epoch": 0.0007983, "grad_norm": 0.2986345589160919, "learning_rate": 7.982e-05, "loss": 0.1298, "step": 7983 }, { "epoch": 0.0007984, "grad_norm": 0.33788925409317017, "learning_rate": 7.983000000000001e-05, "loss": 0.1356, "step": 7984 }, { "epoch": 0.0007985, "grad_norm": 0.28335437178611755, "learning_rate": 7.984e-05, "loss": 0.1284, "step": 7985 }, { "epoch": 0.0007986, "grad_norm": 0.2979366183280945, "learning_rate": 7.985e-05, "loss": 0.1285, "step": 7986 }, { "epoch": 0.0007987, "grad_norm": 0.28379109501838684, "learning_rate": 7.986000000000001e-05, "loss": 0.1285, "step": 7987 }, { "epoch": 0.0007988, "grad_norm": 0.4956672489643097, "learning_rate": 7.986999999999999e-05, "loss": 0.1365, "step": 7988 }, { "epoch": 0.0007989, "grad_norm": 0.29996466636657715, "learning_rate": 7.988e-05, "loss": 0.1342, "step": 7989 }, { "epoch": 0.000799, "grad_norm": 0.283013254404068, "learning_rate": 7.989000000000001e-05, "loss": 0.1217, "step": 7990 }, { "epoch": 0.0007991, "grad_norm": 0.29078409075737, "learning_rate": 7.989999999999999e-05, "loss": 0.1221, "step": 7991 }, { "epoch": 0.0007992, "grad_norm": 0.3446073830127716, "learning_rate": 7.991e-05, "loss": 0.1348, "step": 7992 }, { "epoch": 0.0007993, "grad_norm": 0.2769072949886322, "learning_rate": 7.992000000000001e-05, "loss": 0.1256, "step": 7993 }, { "epoch": 0.0007994, "grad_norm": 0.28110471367836, "learning_rate": 7.993e-05, "loss": 0.1293, "step": 7994 }, { "epoch": 0.0007995, "grad_norm": 0.3948468863964081, "learning_rate": 7.994e-05, "loss": 0.1466, "step": 7995 }, { "epoch": 0.0007996, "grad_norm": 0.2587652802467346, "learning_rate": 7.995e-05, "loss": 0.1201, "step": 7996 }, { "epoch": 0.0007997, "grad_norm": 0.2804921269416809, "learning_rate": 7.996e-05, "loss": 0.1254, "step": 7997 }, { "epoch": 0.0007998, "grad_norm": 0.31802529096603394, "learning_rate": 7.997e-05, "loss": 0.137, "step": 7998 }, { "epoch": 0.0007999, "grad_norm": 0.7691544890403748, "learning_rate": 7.998e-05, "loss": 0.1774, "step": 7999 }, { "epoch": 0.0008, "grad_norm": 3.40895938873291, "learning_rate": 7.999e-05, "loss": 0.1745, "step": 8000 }, { "epoch": 0.0008, "eval_loss": 0.016226764768362045, "eval_runtime": 389.6015, "eval_samples_per_second": 25.667, "eval_steps_per_second": 1.604, "step": 8000 }, { "epoch": 0.0008001, "grad_norm": 0.3202129602432251, "learning_rate": 8e-05, "loss": 0.1265, "step": 8001 }, { "epoch": 0.0008002, "grad_norm": 0.3307330310344696, "learning_rate": 8.001e-05, "loss": 0.1258, "step": 8002 }, { "epoch": 0.0008003, "grad_norm": 0.6703978180885315, "learning_rate": 8.002e-05, "loss": 0.1598, "step": 8003 }, { "epoch": 0.0008004, "grad_norm": 0.4847652316093445, "learning_rate": 8.003e-05, "loss": 0.1378, "step": 8004 }, { "epoch": 0.0008005, "grad_norm": 0.366886705160141, "learning_rate": 8.004e-05, "loss": 0.1316, "step": 8005 }, { "epoch": 0.0008006, "grad_norm": 0.3496139347553253, "learning_rate": 8.005000000000001e-05, "loss": 0.1282, "step": 8006 }, { "epoch": 0.0008007, "grad_norm": 0.3310615122318268, "learning_rate": 8.006e-05, "loss": 0.1295, "step": 8007 }, { "epoch": 0.0008008, "grad_norm": 0.3476350009441376, "learning_rate": 8.007e-05, "loss": 0.1304, "step": 8008 }, { "epoch": 0.0008009, "grad_norm": 0.457918643951416, "learning_rate": 8.008e-05, "loss": 0.156, "step": 8009 }, { "epoch": 0.000801, "grad_norm": 0.34123826026916504, "learning_rate": 8.009e-05, "loss": 0.135, "step": 8010 }, { "epoch": 0.0008011, "grad_norm": 0.31119588017463684, "learning_rate": 8.010000000000001e-05, "loss": 0.1221, "step": 8011 }, { "epoch": 0.0008012, "grad_norm": 0.3695564270019531, "learning_rate": 8.011e-05, "loss": 0.1353, "step": 8012 }, { "epoch": 0.0008013, "grad_norm": 0.49144652485847473, "learning_rate": 8.012e-05, "loss": 0.1431, "step": 8013 }, { "epoch": 0.0008014, "grad_norm": 0.4846400022506714, "learning_rate": 8.013000000000001e-05, "loss": 0.1354, "step": 8014 }, { "epoch": 0.0008015, "grad_norm": 0.3272921144962311, "learning_rate": 8.014e-05, "loss": 0.1292, "step": 8015 }, { "epoch": 0.0008016, "grad_norm": 0.3025423288345337, "learning_rate": 8.015e-05, "loss": 0.1259, "step": 8016 }, { "epoch": 0.0008017, "grad_norm": 0.8752557039260864, "learning_rate": 8.016e-05, "loss": 0.1584, "step": 8017 }, { "epoch": 0.0008018, "grad_norm": 0.4220777750015259, "learning_rate": 8.017000000000001e-05, "loss": 0.1422, "step": 8018 }, { "epoch": 0.0008019, "grad_norm": 0.8674742579460144, "learning_rate": 8.018e-05, "loss": 0.219, "step": 8019 }, { "epoch": 0.000802, "grad_norm": 0.5381582379341125, "learning_rate": 8.019e-05, "loss": 0.1398, "step": 8020 }, { "epoch": 0.0008021, "grad_norm": 1.6414377689361572, "learning_rate": 8.020000000000001e-05, "loss": 0.2237, "step": 8021 }, { "epoch": 0.0008022, "grad_norm": 0.4506913721561432, "learning_rate": 8.020999999999999e-05, "loss": 0.1249, "step": 8022 }, { "epoch": 0.0008023, "grad_norm": 9.942520141601562, "learning_rate": 8.022e-05, "loss": 0.7549, "step": 8023 }, { "epoch": 0.0008024, "grad_norm": 0.599538266658783, "learning_rate": 8.023000000000001e-05, "loss": 0.1387, "step": 8024 }, { "epoch": 0.0008025, "grad_norm": 0.495687872171402, "learning_rate": 8.023999999999999e-05, "loss": 0.1231, "step": 8025 }, { "epoch": 0.0008026, "grad_norm": 0.5193979144096375, "learning_rate": 8.025e-05, "loss": 0.1304, "step": 8026 }, { "epoch": 0.0008027, "grad_norm": 0.3968997299671173, "learning_rate": 8.026000000000001e-05, "loss": 0.1213, "step": 8027 }, { "epoch": 0.0008028, "grad_norm": 0.37203750014305115, "learning_rate": 8.027e-05, "loss": 0.1225, "step": 8028 }, { "epoch": 0.0008029, "grad_norm": 0.4064584970474243, "learning_rate": 8.028e-05, "loss": 0.1312, "step": 8029 }, { "epoch": 0.000803, "grad_norm": 2.3482778072357178, "learning_rate": 8.029e-05, "loss": 0.3098, "step": 8030 }, { "epoch": 0.0008031, "grad_norm": 0.4314936101436615, "learning_rate": 8.03e-05, "loss": 0.1359, "step": 8031 }, { "epoch": 0.0008032, "grad_norm": 0.4015841484069824, "learning_rate": 8.031e-05, "loss": 0.1229, "step": 8032 }, { "epoch": 0.0008033, "grad_norm": 0.4461618661880493, "learning_rate": 8.032e-05, "loss": 0.1239, "step": 8033 }, { "epoch": 0.0008034, "grad_norm": 2.072455883026123, "learning_rate": 8.033e-05, "loss": 0.3016, "step": 8034 }, { "epoch": 0.0008035, "grad_norm": 0.39215677976608276, "learning_rate": 8.034e-05, "loss": 0.1331, "step": 8035 }, { "epoch": 0.0008036, "grad_norm": 0.5516240000724792, "learning_rate": 8.035e-05, "loss": 0.1457, "step": 8036 }, { "epoch": 0.0008037, "grad_norm": 0.7387846112251282, "learning_rate": 8.036e-05, "loss": 0.1438, "step": 8037 }, { "epoch": 0.0008038, "grad_norm": 0.3824997544288635, "learning_rate": 8.037e-05, "loss": 0.1241, "step": 8038 }, { "epoch": 0.0008039, "grad_norm": 0.4071812629699707, "learning_rate": 8.038e-05, "loss": 0.1257, "step": 8039 }, { "epoch": 0.000804, "grad_norm": 0.3403361439704895, "learning_rate": 8.039000000000001e-05, "loss": 0.1212, "step": 8040 }, { "epoch": 0.0008041, "grad_norm": 0.580666720867157, "learning_rate": 8.04e-05, "loss": 0.1626, "step": 8041 }, { "epoch": 0.0008042, "grad_norm": 0.3647790253162384, "learning_rate": 8.041e-05, "loss": 0.1339, "step": 8042 }, { "epoch": 0.0008043, "grad_norm": 0.34806761145591736, "learning_rate": 8.042e-05, "loss": 0.1258, "step": 8043 }, { "epoch": 0.0008044, "grad_norm": 1.0396161079406738, "learning_rate": 8.043e-05, "loss": 0.1851, "step": 8044 }, { "epoch": 0.0008045, "grad_norm": 0.3170357346534729, "learning_rate": 8.044000000000001e-05, "loss": 0.1263, "step": 8045 }, { "epoch": 0.0008046, "grad_norm": 0.3407961428165436, "learning_rate": 8.045e-05, "loss": 0.1237, "step": 8046 }, { "epoch": 0.0008047, "grad_norm": 0.4640645384788513, "learning_rate": 8.046e-05, "loss": 0.1443, "step": 8047 }, { "epoch": 0.0008048, "grad_norm": 0.3018249571323395, "learning_rate": 8.047000000000001e-05, "loss": 0.111, "step": 8048 }, { "epoch": 0.0008049, "grad_norm": 0.7696290612220764, "learning_rate": 8.048e-05, "loss": 0.1543, "step": 8049 }, { "epoch": 0.000805, "grad_norm": 0.3650062680244446, "learning_rate": 8.049e-05, "loss": 0.1274, "step": 8050 }, { "epoch": 0.0008051, "grad_norm": 0.34273412823677063, "learning_rate": 8.05e-05, "loss": 0.1268, "step": 8051 }, { "epoch": 0.0008052, "grad_norm": 0.35924819111824036, "learning_rate": 8.051000000000001e-05, "loss": 0.1298, "step": 8052 }, { "epoch": 0.0008053, "grad_norm": 0.3455607295036316, "learning_rate": 8.052e-05, "loss": 0.1315, "step": 8053 }, { "epoch": 0.0008054, "grad_norm": 0.39464035630226135, "learning_rate": 8.053e-05, "loss": 0.1275, "step": 8054 }, { "epoch": 0.0008055, "grad_norm": 0.3353613615036011, "learning_rate": 8.054000000000001e-05, "loss": 0.1278, "step": 8055 }, { "epoch": 0.0008056, "grad_norm": 0.315790593624115, "learning_rate": 8.054999999999999e-05, "loss": 0.1274, "step": 8056 }, { "epoch": 0.0008057, "grad_norm": 0.29406729340553284, "learning_rate": 8.056e-05, "loss": 0.1207, "step": 8057 }, { "epoch": 0.0008058, "grad_norm": 0.32833918929100037, "learning_rate": 8.057000000000001e-05, "loss": 0.1249, "step": 8058 }, { "epoch": 0.0008059, "grad_norm": 0.3407994508743286, "learning_rate": 8.057999999999999e-05, "loss": 0.1265, "step": 8059 }, { "epoch": 0.000806, "grad_norm": 0.28301092982292175, "learning_rate": 8.059e-05, "loss": 0.1237, "step": 8060 }, { "epoch": 0.0008061, "grad_norm": 0.2914838194847107, "learning_rate": 8.060000000000001e-05, "loss": 0.1217, "step": 8061 }, { "epoch": 0.0008062, "grad_norm": 0.32073450088500977, "learning_rate": 8.061e-05, "loss": 0.1251, "step": 8062 }, { "epoch": 0.0008063, "grad_norm": 0.29743877053260803, "learning_rate": 8.062e-05, "loss": 0.1174, "step": 8063 }, { "epoch": 0.0008064, "grad_norm": 0.3622455298900604, "learning_rate": 8.063e-05, "loss": 0.1258, "step": 8064 }, { "epoch": 0.0008065, "grad_norm": 0.3479747772216797, "learning_rate": 8.064e-05, "loss": 0.1384, "step": 8065 }, { "epoch": 0.0008066, "grad_norm": 0.5194042921066284, "learning_rate": 8.065e-05, "loss": 0.1379, "step": 8066 }, { "epoch": 0.0008067, "grad_norm": 0.2719845771789551, "learning_rate": 8.066e-05, "loss": 0.1227, "step": 8067 }, { "epoch": 0.0008068, "grad_norm": 0.3114382028579712, "learning_rate": 8.067e-05, "loss": 0.1308, "step": 8068 }, { "epoch": 0.0008069, "grad_norm": 0.3259691894054413, "learning_rate": 8.068000000000001e-05, "loss": 0.1252, "step": 8069 }, { "epoch": 0.000807, "grad_norm": 0.32140761613845825, "learning_rate": 8.069e-05, "loss": 0.1254, "step": 8070 }, { "epoch": 0.0008071, "grad_norm": 0.2814655005931854, "learning_rate": 8.07e-05, "loss": 0.1228, "step": 8071 }, { "epoch": 0.0008072, "grad_norm": 0.3107726275920868, "learning_rate": 8.071e-05, "loss": 0.1238, "step": 8072 }, { "epoch": 0.0008073, "grad_norm": 0.3063673973083496, "learning_rate": 8.072e-05, "loss": 0.1251, "step": 8073 }, { "epoch": 0.0008074, "grad_norm": 0.28373897075653076, "learning_rate": 8.073000000000001e-05, "loss": 0.1212, "step": 8074 }, { "epoch": 0.0008075, "grad_norm": 0.5714243054389954, "learning_rate": 8.074e-05, "loss": 0.1334, "step": 8075 }, { "epoch": 0.0008076, "grad_norm": 0.26238927245140076, "learning_rate": 8.075e-05, "loss": 0.1173, "step": 8076 }, { "epoch": 0.0008077, "grad_norm": 0.3149022161960602, "learning_rate": 8.076e-05, "loss": 0.1275, "step": 8077 }, { "epoch": 0.0008078, "grad_norm": 0.29632896184921265, "learning_rate": 8.077e-05, "loss": 0.1239, "step": 8078 }, { "epoch": 0.0008079, "grad_norm": 0.27960312366485596, "learning_rate": 8.078000000000001e-05, "loss": 0.12, "step": 8079 }, { "epoch": 0.000808, "grad_norm": 0.2839134931564331, "learning_rate": 8.079e-05, "loss": 0.1249, "step": 8080 }, { "epoch": 0.0008081, "grad_norm": 0.3265821933746338, "learning_rate": 8.08e-05, "loss": 0.1254, "step": 8081 }, { "epoch": 0.0008082, "grad_norm": 0.3058502674102783, "learning_rate": 8.081000000000001e-05, "loss": 0.1233, "step": 8082 }, { "epoch": 0.0008083, "grad_norm": 0.49245432019233704, "learning_rate": 8.082e-05, "loss": 0.1345, "step": 8083 }, { "epoch": 0.0008084, "grad_norm": 0.2840079963207245, "learning_rate": 8.083e-05, "loss": 0.1202, "step": 8084 }, { "epoch": 0.0008085, "grad_norm": 0.43045926094055176, "learning_rate": 8.084e-05, "loss": 0.1366, "step": 8085 }, { "epoch": 0.0008086, "grad_norm": 0.2691253125667572, "learning_rate": 8.085000000000001e-05, "loss": 0.1226, "step": 8086 }, { "epoch": 0.0008087, "grad_norm": 0.24027742445468903, "learning_rate": 8.086e-05, "loss": 0.1117, "step": 8087 }, { "epoch": 0.0008088, "grad_norm": 0.3302721381187439, "learning_rate": 8.087e-05, "loss": 0.1266, "step": 8088 }, { "epoch": 0.0008089, "grad_norm": 0.29586073756217957, "learning_rate": 8.088000000000001e-05, "loss": 0.1246, "step": 8089 }, { "epoch": 0.000809, "grad_norm": 0.257040798664093, "learning_rate": 8.088999999999999e-05, "loss": 0.1138, "step": 8090 }, { "epoch": 0.0008091, "grad_norm": 0.35814082622528076, "learning_rate": 8.09e-05, "loss": 0.1345, "step": 8091 }, { "epoch": 0.0008092, "grad_norm": 0.377176970243454, "learning_rate": 8.091000000000001e-05, "loss": 0.1333, "step": 8092 }, { "epoch": 0.0008093, "grad_norm": 0.24571584165096283, "learning_rate": 8.091999999999999e-05, "loss": 0.1115, "step": 8093 }, { "epoch": 0.0008094, "grad_norm": 0.5715231895446777, "learning_rate": 8.093e-05, "loss": 0.1378, "step": 8094 }, { "epoch": 0.0008095, "grad_norm": 0.3336717486381531, "learning_rate": 8.094000000000001e-05, "loss": 0.1282, "step": 8095 }, { "epoch": 0.0008096, "grad_norm": 0.3536299765110016, "learning_rate": 8.095e-05, "loss": 0.1324, "step": 8096 }, { "epoch": 0.0008097, "grad_norm": 0.24772508442401886, "learning_rate": 8.096e-05, "loss": 0.1127, "step": 8097 }, { "epoch": 0.0008098, "grad_norm": 0.3084630072116852, "learning_rate": 8.097e-05, "loss": 0.1203, "step": 8098 }, { "epoch": 0.0008099, "grad_norm": 0.34206777811050415, "learning_rate": 8.098e-05, "loss": 0.1332, "step": 8099 }, { "epoch": 0.00081, "grad_norm": 0.24609732627868652, "learning_rate": 8.099e-05, "loss": 0.1144, "step": 8100 }, { "epoch": 0.0008101, "grad_norm": 0.29518646001815796, "learning_rate": 8.1e-05, "loss": 0.1219, "step": 8101 }, { "epoch": 0.0008102, "grad_norm": 0.3137916922569275, "learning_rate": 8.101e-05, "loss": 0.1224, "step": 8102 }, { "epoch": 0.0008103, "grad_norm": 0.2565911114215851, "learning_rate": 8.102e-05, "loss": 0.1164, "step": 8103 }, { "epoch": 0.0008104, "grad_norm": 0.9588075876235962, "learning_rate": 8.103e-05, "loss": 0.1422, "step": 8104 }, { "epoch": 0.0008105, "grad_norm": 0.25571489334106445, "learning_rate": 8.104e-05, "loss": 0.1109, "step": 8105 }, { "epoch": 0.0008106, "grad_norm": 0.2894692122936249, "learning_rate": 8.105e-05, "loss": 0.1214, "step": 8106 }, { "epoch": 0.0008107, "grad_norm": 0.2801630198955536, "learning_rate": 8.106e-05, "loss": 0.1125, "step": 8107 }, { "epoch": 0.0008108, "grad_norm": 0.3092218339443207, "learning_rate": 8.107000000000001e-05, "loss": 0.1234, "step": 8108 }, { "epoch": 0.0008109, "grad_norm": 0.38859060406684875, "learning_rate": 8.108e-05, "loss": 0.1315, "step": 8109 }, { "epoch": 0.000811, "grad_norm": 0.3201065957546234, "learning_rate": 8.109e-05, "loss": 0.125, "step": 8110 }, { "epoch": 0.0008111, "grad_norm": 0.27451232075691223, "learning_rate": 8.11e-05, "loss": 0.1229, "step": 8111 }, { "epoch": 0.0008112, "grad_norm": 0.30030497908592224, "learning_rate": 8.111e-05, "loss": 0.1257, "step": 8112 }, { "epoch": 0.0008113, "grad_norm": 0.32421204447746277, "learning_rate": 8.112000000000001e-05, "loss": 0.1359, "step": 8113 }, { "epoch": 0.0008114, "grad_norm": 0.26296764612197876, "learning_rate": 8.113e-05, "loss": 0.1181, "step": 8114 }, { "epoch": 0.0008115, "grad_norm": 0.24811072647571564, "learning_rate": 8.114e-05, "loss": 0.1193, "step": 8115 }, { "epoch": 0.0008116, "grad_norm": 0.2737072706222534, "learning_rate": 8.115000000000001e-05, "loss": 0.1152, "step": 8116 }, { "epoch": 0.0008117, "grad_norm": 0.3693912923336029, "learning_rate": 8.116e-05, "loss": 0.1359, "step": 8117 }, { "epoch": 0.0008118, "grad_norm": 0.3073267936706543, "learning_rate": 8.117e-05, "loss": 0.1245, "step": 8118 }, { "epoch": 0.0008119, "grad_norm": 0.2775425910949707, "learning_rate": 8.118e-05, "loss": 0.1235, "step": 8119 }, { "epoch": 0.000812, "grad_norm": 0.26199913024902344, "learning_rate": 8.119e-05, "loss": 0.12, "step": 8120 }, { "epoch": 0.0008121, "grad_norm": 0.33429381251335144, "learning_rate": 8.12e-05, "loss": 0.1306, "step": 8121 }, { "epoch": 0.0008122, "grad_norm": 0.2805701792240143, "learning_rate": 8.121e-05, "loss": 0.1198, "step": 8122 }, { "epoch": 0.0008123, "grad_norm": 0.26230496168136597, "learning_rate": 8.122000000000001e-05, "loss": 0.1206, "step": 8123 }, { "epoch": 0.0008124, "grad_norm": 0.27616167068481445, "learning_rate": 8.122999999999999e-05, "loss": 0.1215, "step": 8124 }, { "epoch": 0.0008125, "grad_norm": 0.4020824134349823, "learning_rate": 8.124e-05, "loss": 0.1347, "step": 8125 }, { "epoch": 0.0008126, "grad_norm": 0.30386677384376526, "learning_rate": 8.125000000000001e-05, "loss": 0.1313, "step": 8126 }, { "epoch": 0.0008127, "grad_norm": 0.2896801829338074, "learning_rate": 8.125999999999999e-05, "loss": 0.1203, "step": 8127 }, { "epoch": 0.0008128, "grad_norm": 0.2686094045639038, "learning_rate": 8.127e-05, "loss": 0.1224, "step": 8128 }, { "epoch": 0.0008129, "grad_norm": 0.3080318868160248, "learning_rate": 8.128000000000001e-05, "loss": 0.1224, "step": 8129 }, { "epoch": 0.000813, "grad_norm": 0.331500768661499, "learning_rate": 8.129e-05, "loss": 0.1261, "step": 8130 }, { "epoch": 0.0008131, "grad_norm": 0.25597918033599854, "learning_rate": 8.13e-05, "loss": 0.1242, "step": 8131 }, { "epoch": 0.0008132, "grad_norm": 0.2504562735557556, "learning_rate": 8.131e-05, "loss": 0.1209, "step": 8132 }, { "epoch": 0.0008133, "grad_norm": 0.27307194471359253, "learning_rate": 8.132e-05, "loss": 0.1226, "step": 8133 }, { "epoch": 0.0008134, "grad_norm": 0.25977298617362976, "learning_rate": 8.133e-05, "loss": 0.1166, "step": 8134 }, { "epoch": 0.0008135, "grad_norm": 0.403717041015625, "learning_rate": 8.134e-05, "loss": 0.1316, "step": 8135 }, { "epoch": 0.0008136, "grad_norm": 0.39485740661621094, "learning_rate": 8.135e-05, "loss": 0.1393, "step": 8136 }, { "epoch": 0.0008137, "grad_norm": 0.27854466438293457, "learning_rate": 8.136e-05, "loss": 0.1177, "step": 8137 }, { "epoch": 0.0008138, "grad_norm": 0.2546297311782837, "learning_rate": 8.137e-05, "loss": 0.1178, "step": 8138 }, { "epoch": 0.0008139, "grad_norm": 0.24863463640213013, "learning_rate": 8.138e-05, "loss": 0.1205, "step": 8139 }, { "epoch": 0.000814, "grad_norm": 0.2401416152715683, "learning_rate": 8.139e-05, "loss": 0.1115, "step": 8140 }, { "epoch": 0.0008141, "grad_norm": 0.38651880621910095, "learning_rate": 8.14e-05, "loss": 0.1298, "step": 8141 }, { "epoch": 0.0008142, "grad_norm": 0.24856111407279968, "learning_rate": 8.141e-05, "loss": 0.12, "step": 8142 }, { "epoch": 0.0008143, "grad_norm": 0.27614912390708923, "learning_rate": 8.142e-05, "loss": 0.1236, "step": 8143 }, { "epoch": 0.0008144, "grad_norm": 0.2812386155128479, "learning_rate": 8.143e-05, "loss": 0.1273, "step": 8144 }, { "epoch": 0.0008145, "grad_norm": 0.27021878957748413, "learning_rate": 8.144e-05, "loss": 0.1223, "step": 8145 }, { "epoch": 0.0008146, "grad_norm": 0.25452882051467896, "learning_rate": 8.145e-05, "loss": 0.115, "step": 8146 }, { "epoch": 0.0008147, "grad_norm": 0.2515009641647339, "learning_rate": 8.146000000000001e-05, "loss": 0.1155, "step": 8147 }, { "epoch": 0.0008148, "grad_norm": 0.27468255162239075, "learning_rate": 8.147e-05, "loss": 0.1277, "step": 8148 }, { "epoch": 0.0008149, "grad_norm": 0.30062854290008545, "learning_rate": 8.148e-05, "loss": 0.1307, "step": 8149 }, { "epoch": 0.000815, "grad_norm": 0.26882845163345337, "learning_rate": 8.149000000000001e-05, "loss": 0.1234, "step": 8150 }, { "epoch": 0.0008151, "grad_norm": 0.4284106194972992, "learning_rate": 8.15e-05, "loss": 0.1329, "step": 8151 }, { "epoch": 0.0008152, "grad_norm": 0.2600254714488983, "learning_rate": 8.151e-05, "loss": 0.1199, "step": 8152 }, { "epoch": 0.0008153, "grad_norm": 0.2593938112258911, "learning_rate": 8.152e-05, "loss": 0.1167, "step": 8153 }, { "epoch": 0.0008154, "grad_norm": 0.35058048367500305, "learning_rate": 8.153e-05, "loss": 0.1319, "step": 8154 }, { "epoch": 0.0008155, "grad_norm": 0.2629115581512451, "learning_rate": 8.154e-05, "loss": 0.1207, "step": 8155 }, { "epoch": 0.0008156, "grad_norm": 0.2404366284608841, "learning_rate": 8.155e-05, "loss": 0.1185, "step": 8156 }, { "epoch": 0.0008157, "grad_norm": 0.31744587421417236, "learning_rate": 8.156000000000001e-05, "loss": 0.1356, "step": 8157 }, { "epoch": 0.0008158, "grad_norm": 0.26406434178352356, "learning_rate": 8.156999999999999e-05, "loss": 0.1199, "step": 8158 }, { "epoch": 0.0008159, "grad_norm": 0.49710792303085327, "learning_rate": 8.158e-05, "loss": 0.1284, "step": 8159 }, { "epoch": 0.000816, "grad_norm": 0.2635711431503296, "learning_rate": 8.159000000000001e-05, "loss": 0.122, "step": 8160 }, { "epoch": 0.0008161, "grad_norm": 0.2565058171749115, "learning_rate": 8.159999999999999e-05, "loss": 0.1224, "step": 8161 }, { "epoch": 0.0008162, "grad_norm": 0.277601033449173, "learning_rate": 8.161e-05, "loss": 0.1196, "step": 8162 }, { "epoch": 0.0008163, "grad_norm": 0.35559678077697754, "learning_rate": 8.162000000000001e-05, "loss": 0.1246, "step": 8163 }, { "epoch": 0.0008164, "grad_norm": 0.32934850454330444, "learning_rate": 8.163e-05, "loss": 0.1314, "step": 8164 }, { "epoch": 0.0008165, "grad_norm": 0.45738476514816284, "learning_rate": 8.164e-05, "loss": 0.1327, "step": 8165 }, { "epoch": 0.0008166, "grad_norm": 0.3290201723575592, "learning_rate": 8.165e-05, "loss": 0.1301, "step": 8166 }, { "epoch": 0.0008167, "grad_norm": 0.3304916322231293, "learning_rate": 8.166e-05, "loss": 0.1236, "step": 8167 }, { "epoch": 0.0008168, "grad_norm": 0.4709419310092926, "learning_rate": 8.167e-05, "loss": 0.1309, "step": 8168 }, { "epoch": 0.0008169, "grad_norm": 0.2472590059041977, "learning_rate": 8.168e-05, "loss": 0.1136, "step": 8169 }, { "epoch": 0.000817, "grad_norm": 0.322237104177475, "learning_rate": 8.169e-05, "loss": 0.1248, "step": 8170 }, { "epoch": 0.0008171, "grad_norm": 0.2979406416416168, "learning_rate": 8.17e-05, "loss": 0.1334, "step": 8171 }, { "epoch": 0.0008172, "grad_norm": 0.2639550566673279, "learning_rate": 8.171e-05, "loss": 0.1192, "step": 8172 }, { "epoch": 0.0008173, "grad_norm": 0.3248070776462555, "learning_rate": 8.172e-05, "loss": 0.1211, "step": 8173 }, { "epoch": 0.0008174, "grad_norm": 0.3602343797683716, "learning_rate": 8.173e-05, "loss": 0.1223, "step": 8174 }, { "epoch": 0.0008175, "grad_norm": 0.33443495631217957, "learning_rate": 8.174e-05, "loss": 0.1254, "step": 8175 }, { "epoch": 0.0008176, "grad_norm": 0.23635318875312805, "learning_rate": 8.175e-05, "loss": 0.1149, "step": 8176 }, { "epoch": 0.0008177, "grad_norm": 0.23680636286735535, "learning_rate": 8.176e-05, "loss": 0.116, "step": 8177 }, { "epoch": 0.0008178, "grad_norm": 0.3024715483188629, "learning_rate": 8.177e-05, "loss": 0.1249, "step": 8178 }, { "epoch": 0.0008179, "grad_norm": 0.26406145095825195, "learning_rate": 8.178e-05, "loss": 0.1166, "step": 8179 }, { "epoch": 0.000818, "grad_norm": 0.3081522583961487, "learning_rate": 8.179e-05, "loss": 0.1176, "step": 8180 }, { "epoch": 0.0008181, "grad_norm": 0.2564171552658081, "learning_rate": 8.180000000000001e-05, "loss": 0.1165, "step": 8181 }, { "epoch": 0.0008182, "grad_norm": 0.25777319073677063, "learning_rate": 8.181e-05, "loss": 0.1143, "step": 8182 }, { "epoch": 0.0008183, "grad_norm": 0.2623406648635864, "learning_rate": 8.182e-05, "loss": 0.1231, "step": 8183 }, { "epoch": 0.0008184, "grad_norm": 0.3436448276042938, "learning_rate": 8.183000000000001e-05, "loss": 0.1289, "step": 8184 }, { "epoch": 0.0008185, "grad_norm": 0.39584359526634216, "learning_rate": 8.184e-05, "loss": 0.1311, "step": 8185 }, { "epoch": 0.0008186, "grad_norm": 0.2454989105463028, "learning_rate": 8.185e-05, "loss": 0.1146, "step": 8186 }, { "epoch": 0.0008187, "grad_norm": 0.444217324256897, "learning_rate": 8.186e-05, "loss": 0.1276, "step": 8187 }, { "epoch": 0.0008188, "grad_norm": 0.3582991361618042, "learning_rate": 8.187e-05, "loss": 0.1314, "step": 8188 }, { "epoch": 0.0008189, "grad_norm": 0.462783545255661, "learning_rate": 8.188e-05, "loss": 0.1328, "step": 8189 }, { "epoch": 0.000819, "grad_norm": 0.2717292308807373, "learning_rate": 8.189e-05, "loss": 0.1164, "step": 8190 }, { "epoch": 0.0008191, "grad_norm": 0.257985383272171, "learning_rate": 8.190000000000001e-05, "loss": 0.1202, "step": 8191 }, { "epoch": 0.0008192, "grad_norm": 0.2859156131744385, "learning_rate": 8.191e-05, "loss": 0.1195, "step": 8192 }, { "epoch": 0.0008193, "grad_norm": 0.27496665716171265, "learning_rate": 8.192e-05, "loss": 0.1266, "step": 8193 }, { "epoch": 0.0008194, "grad_norm": 0.2906060516834259, "learning_rate": 8.193000000000001e-05, "loss": 0.1279, "step": 8194 }, { "epoch": 0.0008195, "grad_norm": 0.25681042671203613, "learning_rate": 8.193999999999999e-05, "loss": 0.1152, "step": 8195 }, { "epoch": 0.0008196, "grad_norm": 0.28119799494743347, "learning_rate": 8.195e-05, "loss": 0.118, "step": 8196 }, { "epoch": 0.0008197, "grad_norm": 0.24184001982212067, "learning_rate": 8.196000000000001e-05, "loss": 0.1182, "step": 8197 }, { "epoch": 0.0008198, "grad_norm": 0.2804012596607208, "learning_rate": 8.197e-05, "loss": 0.1234, "step": 8198 }, { "epoch": 0.0008199, "grad_norm": 0.2748105227947235, "learning_rate": 8.198e-05, "loss": 0.1118, "step": 8199 }, { "epoch": 0.00082, "grad_norm": 0.25761836767196655, "learning_rate": 8.199e-05, "loss": 0.111, "step": 8200 }, { "epoch": 0.0008201, "grad_norm": 0.2614924907684326, "learning_rate": 8.2e-05, "loss": 0.12, "step": 8201 }, { "epoch": 0.0008202, "grad_norm": 0.27884936332702637, "learning_rate": 8.201e-05, "loss": 0.1265, "step": 8202 }, { "epoch": 0.0008203, "grad_norm": 0.37377434968948364, "learning_rate": 8.202e-05, "loss": 0.1292, "step": 8203 }, { "epoch": 0.0008204, "grad_norm": 0.23581278324127197, "learning_rate": 8.203e-05, "loss": 0.1152, "step": 8204 }, { "epoch": 0.0008205, "grad_norm": 0.2529856562614441, "learning_rate": 8.204e-05, "loss": 0.1223, "step": 8205 }, { "epoch": 0.0008206, "grad_norm": 0.3910468816757202, "learning_rate": 8.205e-05, "loss": 0.1265, "step": 8206 }, { "epoch": 0.0008207, "grad_norm": 0.3044634461402893, "learning_rate": 8.206e-05, "loss": 0.1208, "step": 8207 }, { "epoch": 0.0008208, "grad_norm": 0.29144471883773804, "learning_rate": 8.207e-05, "loss": 0.1287, "step": 8208 }, { "epoch": 0.0008209, "grad_norm": 0.5447407960891724, "learning_rate": 8.208e-05, "loss": 0.136, "step": 8209 }, { "epoch": 0.000821, "grad_norm": 0.3419993817806244, "learning_rate": 8.209e-05, "loss": 0.1263, "step": 8210 }, { "epoch": 0.0008211, "grad_norm": 0.27843335270881653, "learning_rate": 8.21e-05, "loss": 0.1265, "step": 8211 }, { "epoch": 0.0008212, "grad_norm": 0.36781027913093567, "learning_rate": 8.211e-05, "loss": 0.1213, "step": 8212 }, { "epoch": 0.0008213, "grad_norm": 0.42243266105651855, "learning_rate": 8.212e-05, "loss": 0.1275, "step": 8213 }, { "epoch": 0.0008214, "grad_norm": 0.26543208956718445, "learning_rate": 8.213e-05, "loss": 0.1194, "step": 8214 }, { "epoch": 0.0008215, "grad_norm": 0.29444581270217896, "learning_rate": 8.214000000000001e-05, "loss": 0.1199, "step": 8215 }, { "epoch": 0.0008216, "grad_norm": 0.3137739300727844, "learning_rate": 8.215e-05, "loss": 0.1231, "step": 8216 }, { "epoch": 0.0008217, "grad_norm": 0.2812126874923706, "learning_rate": 8.216e-05, "loss": 0.1256, "step": 8217 }, { "epoch": 0.0008218, "grad_norm": 0.30354052782058716, "learning_rate": 8.217000000000001e-05, "loss": 0.1254, "step": 8218 }, { "epoch": 0.0008219, "grad_norm": 0.3041916489601135, "learning_rate": 8.218e-05, "loss": 0.1278, "step": 8219 }, { "epoch": 0.000822, "grad_norm": 0.29837197065353394, "learning_rate": 8.219e-05, "loss": 0.1136, "step": 8220 }, { "epoch": 0.0008221, "grad_norm": 0.2326834797859192, "learning_rate": 8.22e-05, "loss": 0.1123, "step": 8221 }, { "epoch": 0.0008222, "grad_norm": 0.25673767924308777, "learning_rate": 8.221e-05, "loss": 0.1183, "step": 8222 }, { "epoch": 0.0008223, "grad_norm": 0.38557425141334534, "learning_rate": 8.222e-05, "loss": 0.1257, "step": 8223 }, { "epoch": 0.0008224, "grad_norm": 0.31960251927375793, "learning_rate": 8.223e-05, "loss": 0.1212, "step": 8224 }, { "epoch": 0.0008225, "grad_norm": 0.3729652166366577, "learning_rate": 8.224000000000001e-05, "loss": 0.1312, "step": 8225 }, { "epoch": 0.0008226, "grad_norm": 0.2720782160758972, "learning_rate": 8.225e-05, "loss": 0.1269, "step": 8226 }, { "epoch": 0.0008227, "grad_norm": 0.2964369058609009, "learning_rate": 8.226e-05, "loss": 0.1229, "step": 8227 }, { "epoch": 0.0008228, "grad_norm": 0.3284490406513214, "learning_rate": 8.227000000000001e-05, "loss": 0.13, "step": 8228 }, { "epoch": 0.0008229, "grad_norm": 0.25771719217300415, "learning_rate": 8.227999999999999e-05, "loss": 0.1165, "step": 8229 }, { "epoch": 0.000823, "grad_norm": 0.2636684477329254, "learning_rate": 8.229e-05, "loss": 0.1226, "step": 8230 }, { "epoch": 0.0008231, "grad_norm": 0.2904820442199707, "learning_rate": 8.230000000000001e-05, "loss": 0.1261, "step": 8231 }, { "epoch": 0.0008232, "grad_norm": 0.25932803750038147, "learning_rate": 8.231e-05, "loss": 0.1142, "step": 8232 }, { "epoch": 0.0008233, "grad_norm": 0.24145160615444183, "learning_rate": 8.232e-05, "loss": 0.1175, "step": 8233 }, { "epoch": 0.0008234, "grad_norm": 0.2582058608531952, "learning_rate": 8.233e-05, "loss": 0.1182, "step": 8234 }, { "epoch": 0.0008235, "grad_norm": 0.2467871904373169, "learning_rate": 8.234e-05, "loss": 0.1223, "step": 8235 }, { "epoch": 0.0008236, "grad_norm": 0.24092982709407806, "learning_rate": 8.235e-05, "loss": 0.1166, "step": 8236 }, { "epoch": 0.0008237, "grad_norm": 0.4118715524673462, "learning_rate": 8.236e-05, "loss": 0.1287, "step": 8237 }, { "epoch": 0.0008238, "grad_norm": 0.2566819190979004, "learning_rate": 8.237e-05, "loss": 0.1171, "step": 8238 }, { "epoch": 0.0008239, "grad_norm": 0.26521503925323486, "learning_rate": 8.238e-05, "loss": 0.118, "step": 8239 }, { "epoch": 0.000824, "grad_norm": 0.4187837243080139, "learning_rate": 8.239e-05, "loss": 0.1308, "step": 8240 }, { "epoch": 0.0008241, "grad_norm": 0.28740209341049194, "learning_rate": 8.24e-05, "loss": 0.1256, "step": 8241 }, { "epoch": 0.0008242, "grad_norm": 0.2839011251926422, "learning_rate": 8.241e-05, "loss": 0.125, "step": 8242 }, { "epoch": 0.0008243, "grad_norm": 0.8069186210632324, "learning_rate": 8.242e-05, "loss": 0.1632, "step": 8243 }, { "epoch": 0.0008244, "grad_norm": 0.24288016557693481, "learning_rate": 8.243e-05, "loss": 0.1135, "step": 8244 }, { "epoch": 0.0008245, "grad_norm": 0.3409069776535034, "learning_rate": 8.244e-05, "loss": 0.1301, "step": 8245 }, { "epoch": 0.0008246, "grad_norm": 0.24525922536849976, "learning_rate": 8.245e-05, "loss": 0.1158, "step": 8246 }, { "epoch": 0.0008247, "grad_norm": 0.2744600772857666, "learning_rate": 8.246e-05, "loss": 0.1173, "step": 8247 }, { "epoch": 0.0008248, "grad_norm": 0.4034813344478607, "learning_rate": 8.247e-05, "loss": 0.1368, "step": 8248 }, { "epoch": 0.0008249, "grad_norm": 0.319559782743454, "learning_rate": 8.248000000000001e-05, "loss": 0.1246, "step": 8249 }, { "epoch": 0.000825, "grad_norm": 0.27088743448257446, "learning_rate": 8.249e-05, "loss": 0.1181, "step": 8250 }, { "epoch": 0.0008251, "grad_norm": 0.2523137032985687, "learning_rate": 8.25e-05, "loss": 0.1176, "step": 8251 }, { "epoch": 0.0008252, "grad_norm": 0.25668174028396606, "learning_rate": 8.251000000000001e-05, "loss": 0.1184, "step": 8252 }, { "epoch": 0.0008253, "grad_norm": 0.37655991315841675, "learning_rate": 8.252e-05, "loss": 0.1366, "step": 8253 }, { "epoch": 0.0008254, "grad_norm": 0.24293242394924164, "learning_rate": 8.253e-05, "loss": 0.1133, "step": 8254 }, { "epoch": 0.0008255, "grad_norm": 0.28997358679771423, "learning_rate": 8.254e-05, "loss": 0.1235, "step": 8255 }, { "epoch": 0.0008256, "grad_norm": 0.28856682777404785, "learning_rate": 8.255e-05, "loss": 0.1293, "step": 8256 }, { "epoch": 0.0008257, "grad_norm": 0.2605460286140442, "learning_rate": 8.256e-05, "loss": 0.121, "step": 8257 }, { "epoch": 0.0008258, "grad_norm": 0.25905779004096985, "learning_rate": 8.257e-05, "loss": 0.1118, "step": 8258 }, { "epoch": 0.0008259, "grad_norm": 0.270008385181427, "learning_rate": 8.258000000000001e-05, "loss": 0.124, "step": 8259 }, { "epoch": 0.000826, "grad_norm": 0.270720899105072, "learning_rate": 8.259e-05, "loss": 0.1185, "step": 8260 }, { "epoch": 0.0008261, "grad_norm": 0.3338068425655365, "learning_rate": 8.26e-05, "loss": 0.123, "step": 8261 }, { "epoch": 0.0008262, "grad_norm": 3.9577460289001465, "learning_rate": 8.261000000000001e-05, "loss": 0.2931, "step": 8262 }, { "epoch": 0.0008263, "grad_norm": 0.2671869695186615, "learning_rate": 8.261999999999999e-05, "loss": 0.1177, "step": 8263 }, { "epoch": 0.0008264, "grad_norm": 0.32506656646728516, "learning_rate": 8.263e-05, "loss": 0.1233, "step": 8264 }, { "epoch": 0.0008265, "grad_norm": 0.2978303134441376, "learning_rate": 8.264000000000001e-05, "loss": 0.1177, "step": 8265 }, { "epoch": 0.0008266, "grad_norm": 0.33331626653671265, "learning_rate": 8.265e-05, "loss": 0.1274, "step": 8266 }, { "epoch": 0.0008267, "grad_norm": 0.3332047760486603, "learning_rate": 8.266e-05, "loss": 0.1182, "step": 8267 }, { "epoch": 0.0008268, "grad_norm": 0.38557150959968567, "learning_rate": 8.267e-05, "loss": 0.1348, "step": 8268 }, { "epoch": 0.0008269, "grad_norm": 0.3090338408946991, "learning_rate": 8.268e-05, "loss": 0.1211, "step": 8269 }, { "epoch": 0.000827, "grad_norm": 0.28389936685562134, "learning_rate": 8.269e-05, "loss": 0.1183, "step": 8270 }, { "epoch": 0.0008271, "grad_norm": 0.2615092396736145, "learning_rate": 8.27e-05, "loss": 0.1169, "step": 8271 }, { "epoch": 0.0008272, "grad_norm": 0.3285996615886688, "learning_rate": 8.271e-05, "loss": 0.1208, "step": 8272 }, { "epoch": 0.0008273, "grad_norm": 0.49238863587379456, "learning_rate": 8.272e-05, "loss": 0.1453, "step": 8273 }, { "epoch": 0.0008274, "grad_norm": 0.2479100525379181, "learning_rate": 8.273e-05, "loss": 0.1099, "step": 8274 }, { "epoch": 0.0008275, "grad_norm": 0.35445767641067505, "learning_rate": 8.274e-05, "loss": 0.1161, "step": 8275 }, { "epoch": 0.0008276, "grad_norm": 0.3050071895122528, "learning_rate": 8.275e-05, "loss": 0.1182, "step": 8276 }, { "epoch": 0.0008277, "grad_norm": 0.33559903502464294, "learning_rate": 8.276e-05, "loss": 0.1299, "step": 8277 }, { "epoch": 0.0008278, "grad_norm": 0.482623428106308, "learning_rate": 8.277e-05, "loss": 0.1367, "step": 8278 }, { "epoch": 0.0008279, "grad_norm": 0.2906891405582428, "learning_rate": 8.278e-05, "loss": 0.118, "step": 8279 }, { "epoch": 0.000828, "grad_norm": 0.2577823996543884, "learning_rate": 8.279e-05, "loss": 0.111, "step": 8280 }, { "epoch": 0.0008281, "grad_norm": 0.3549817204475403, "learning_rate": 8.280000000000001e-05, "loss": 0.1409, "step": 8281 }, { "epoch": 0.0008282, "grad_norm": 0.3377881944179535, "learning_rate": 8.281e-05, "loss": 0.1229, "step": 8282 }, { "epoch": 0.0008283, "grad_norm": 0.35716354846954346, "learning_rate": 8.282000000000001e-05, "loss": 0.1268, "step": 8283 }, { "epoch": 0.0008284, "grad_norm": 0.29833707213401794, "learning_rate": 8.283e-05, "loss": 0.1182, "step": 8284 }, { "epoch": 0.0008285, "grad_norm": 0.5133569240570068, "learning_rate": 8.284e-05, "loss": 0.1214, "step": 8285 }, { "epoch": 0.0008286, "grad_norm": 0.24109426140785217, "learning_rate": 8.285000000000001e-05, "loss": 0.1091, "step": 8286 }, { "epoch": 0.0008287, "grad_norm": 0.6600997447967529, "learning_rate": 8.286e-05, "loss": 0.1371, "step": 8287 }, { "epoch": 0.0008288, "grad_norm": 0.3192616403102875, "learning_rate": 8.287e-05, "loss": 0.1214, "step": 8288 }, { "epoch": 0.0008289, "grad_norm": 0.34313368797302246, "learning_rate": 8.288e-05, "loss": 0.1159, "step": 8289 }, { "epoch": 0.000829, "grad_norm": 0.31975916028022766, "learning_rate": 8.289e-05, "loss": 0.1293, "step": 8290 }, { "epoch": 0.0008291, "grad_norm": 0.8865758180618286, "learning_rate": 8.29e-05, "loss": 0.1255, "step": 8291 }, { "epoch": 0.0008292, "grad_norm": 0.9230272769927979, "learning_rate": 8.291e-05, "loss": 0.1318, "step": 8292 }, { "epoch": 0.0008293, "grad_norm": 0.6130544543266296, "learning_rate": 8.292000000000001e-05, "loss": 0.1463, "step": 8293 }, { "epoch": 0.0008294, "grad_norm": 0.401056170463562, "learning_rate": 8.293e-05, "loss": 0.1218, "step": 8294 }, { "epoch": 0.0008295, "grad_norm": 0.3474794924259186, "learning_rate": 8.294e-05, "loss": 0.1077, "step": 8295 }, { "epoch": 0.0008296, "grad_norm": 0.3603403866291046, "learning_rate": 8.295000000000001e-05, "loss": 0.1161, "step": 8296 }, { "epoch": 0.0008297, "grad_norm": 0.46914738416671753, "learning_rate": 8.295999999999999e-05, "loss": 0.1332, "step": 8297 }, { "epoch": 0.0008298, "grad_norm": 0.4348466992378235, "learning_rate": 8.297e-05, "loss": 0.1332, "step": 8298 }, { "epoch": 0.0008299, "grad_norm": 0.3736206889152527, "learning_rate": 8.298000000000001e-05, "loss": 0.1256, "step": 8299 }, { "epoch": 0.00083, "grad_norm": 0.3494785726070404, "learning_rate": 8.299e-05, "loss": 0.1221, "step": 8300 }, { "epoch": 0.0008301, "grad_norm": 0.2794828712940216, "learning_rate": 8.3e-05, "loss": 0.1077, "step": 8301 }, { "epoch": 0.0008302, "grad_norm": 0.3573426902294159, "learning_rate": 8.301e-05, "loss": 0.1187, "step": 8302 }, { "epoch": 0.0008303, "grad_norm": 0.2746645212173462, "learning_rate": 8.302e-05, "loss": 0.11, "step": 8303 }, { "epoch": 0.0008304, "grad_norm": 0.35572004318237305, "learning_rate": 8.303e-05, "loss": 0.1227, "step": 8304 }, { "epoch": 0.0008305, "grad_norm": 0.28891506791114807, "learning_rate": 8.304e-05, "loss": 0.1162, "step": 8305 }, { "epoch": 0.0008306, "grad_norm": 0.33817705512046814, "learning_rate": 8.305e-05, "loss": 0.124, "step": 8306 }, { "epoch": 0.0008307, "grad_norm": 0.2758975923061371, "learning_rate": 8.306e-05, "loss": 0.1147, "step": 8307 }, { "epoch": 0.0008308, "grad_norm": 0.4337676763534546, "learning_rate": 8.307e-05, "loss": 0.147, "step": 8308 }, { "epoch": 0.0008309, "grad_norm": 0.3356058895587921, "learning_rate": 8.308e-05, "loss": 0.1208, "step": 8309 }, { "epoch": 0.000831, "grad_norm": 0.2898893356323242, "learning_rate": 8.309e-05, "loss": 0.1221, "step": 8310 }, { "epoch": 0.0008311, "grad_norm": 0.2947258949279785, "learning_rate": 8.31e-05, "loss": 0.1136, "step": 8311 }, { "epoch": 0.0008312, "grad_norm": 0.4135041832923889, "learning_rate": 8.311e-05, "loss": 0.1261, "step": 8312 }, { "epoch": 0.0008313, "grad_norm": 0.3066425323486328, "learning_rate": 8.312e-05, "loss": 0.1221, "step": 8313 }, { "epoch": 0.0008314, "grad_norm": 0.2656776010990143, "learning_rate": 8.313e-05, "loss": 0.1135, "step": 8314 }, { "epoch": 0.0008315, "grad_norm": 0.2955726981163025, "learning_rate": 8.314000000000001e-05, "loss": 0.1167, "step": 8315 }, { "epoch": 0.0008316, "grad_norm": 0.2642335891723633, "learning_rate": 8.315e-05, "loss": 0.1128, "step": 8316 }, { "epoch": 0.0008317, "grad_norm": 0.23299124836921692, "learning_rate": 8.316000000000001e-05, "loss": 0.1079, "step": 8317 }, { "epoch": 0.0008318, "grad_norm": 0.2927297353744507, "learning_rate": 8.317e-05, "loss": 0.1194, "step": 8318 }, { "epoch": 0.0008319, "grad_norm": 0.3799995183944702, "learning_rate": 8.318e-05, "loss": 0.117, "step": 8319 }, { "epoch": 0.000832, "grad_norm": 0.39418503642082214, "learning_rate": 8.319000000000001e-05, "loss": 0.1209, "step": 8320 }, { "epoch": 0.0008321, "grad_norm": 0.28276681900024414, "learning_rate": 8.32e-05, "loss": 0.1157, "step": 8321 }, { "epoch": 0.0008322, "grad_norm": 0.34471064805984497, "learning_rate": 8.321e-05, "loss": 0.1181, "step": 8322 }, { "epoch": 0.0008323, "grad_norm": 0.3510381579399109, "learning_rate": 8.322e-05, "loss": 0.1277, "step": 8323 }, { "epoch": 0.0008324, "grad_norm": 0.30988457798957825, "learning_rate": 8.323e-05, "loss": 0.1222, "step": 8324 }, { "epoch": 0.0008325, "grad_norm": 0.2697628438472748, "learning_rate": 8.324e-05, "loss": 0.1149, "step": 8325 }, { "epoch": 0.0008326, "grad_norm": 0.27366572618484497, "learning_rate": 8.325e-05, "loss": 0.1159, "step": 8326 }, { "epoch": 0.0008327, "grad_norm": 0.25706392526626587, "learning_rate": 8.326000000000001e-05, "loss": 0.1133, "step": 8327 }, { "epoch": 0.0008328, "grad_norm": 0.4249774217605591, "learning_rate": 8.327e-05, "loss": 0.1223, "step": 8328 }, { "epoch": 0.0008329, "grad_norm": 0.2878964841365814, "learning_rate": 8.328e-05, "loss": 0.119, "step": 8329 }, { "epoch": 0.000833, "grad_norm": 0.34603971242904663, "learning_rate": 8.329000000000001e-05, "loss": 0.1332, "step": 8330 }, { "epoch": 0.0008331, "grad_norm": 0.2570502460002899, "learning_rate": 8.329999999999999e-05, "loss": 0.1158, "step": 8331 }, { "epoch": 0.0008332, "grad_norm": 0.25689229369163513, "learning_rate": 8.331e-05, "loss": 0.1164, "step": 8332 }, { "epoch": 0.0008333, "grad_norm": 0.4386882185935974, "learning_rate": 8.332000000000001e-05, "loss": 0.1314, "step": 8333 }, { "epoch": 0.0008334, "grad_norm": 0.26009953022003174, "learning_rate": 8.333e-05, "loss": 0.1135, "step": 8334 }, { "epoch": 0.0008335, "grad_norm": 0.2905921936035156, "learning_rate": 8.334e-05, "loss": 0.124, "step": 8335 }, { "epoch": 0.0008336, "grad_norm": 0.3394087851047516, "learning_rate": 8.335e-05, "loss": 0.1251, "step": 8336 }, { "epoch": 0.0008337, "grad_norm": 0.32263341546058655, "learning_rate": 8.336e-05, "loss": 0.1244, "step": 8337 }, { "epoch": 0.0008338, "grad_norm": 0.35376691818237305, "learning_rate": 8.337e-05, "loss": 0.1265, "step": 8338 }, { "epoch": 0.0008339, "grad_norm": 0.2781798839569092, "learning_rate": 8.338e-05, "loss": 0.1192, "step": 8339 }, { "epoch": 0.000834, "grad_norm": 0.2511618733406067, "learning_rate": 8.339e-05, "loss": 0.1119, "step": 8340 }, { "epoch": 0.0008341, "grad_norm": 0.4951576292514801, "learning_rate": 8.34e-05, "loss": 0.1381, "step": 8341 }, { "epoch": 0.0008342, "grad_norm": 0.264965683221817, "learning_rate": 8.341e-05, "loss": 0.1125, "step": 8342 }, { "epoch": 0.0008343, "grad_norm": 0.2721898853778839, "learning_rate": 8.342e-05, "loss": 0.1144, "step": 8343 }, { "epoch": 0.0008344, "grad_norm": 0.27462533116340637, "learning_rate": 8.343e-05, "loss": 0.109, "step": 8344 }, { "epoch": 0.0008345, "grad_norm": 0.2744678556919098, "learning_rate": 8.344e-05, "loss": 0.1149, "step": 8345 }, { "epoch": 0.0008346, "grad_norm": 0.5289849042892456, "learning_rate": 8.345e-05, "loss": 0.1454, "step": 8346 }, { "epoch": 0.0008347, "grad_norm": 0.2547566294670105, "learning_rate": 8.346e-05, "loss": 0.1118, "step": 8347 }, { "epoch": 0.0008348, "grad_norm": 0.3034285604953766, "learning_rate": 8.347e-05, "loss": 0.1199, "step": 8348 }, { "epoch": 0.0008349, "grad_norm": 0.30945277214050293, "learning_rate": 8.348000000000001e-05, "loss": 0.1213, "step": 8349 }, { "epoch": 0.000835, "grad_norm": 0.2395797222852707, "learning_rate": 8.349e-05, "loss": 0.1089, "step": 8350 }, { "epoch": 0.0008351, "grad_norm": 0.3820746839046478, "learning_rate": 8.350000000000001e-05, "loss": 0.1277, "step": 8351 }, { "epoch": 0.0008352, "grad_norm": 0.25428974628448486, "learning_rate": 8.351e-05, "loss": 0.1135, "step": 8352 }, { "epoch": 0.0008353, "grad_norm": 0.3403705358505249, "learning_rate": 8.352e-05, "loss": 0.1158, "step": 8353 }, { "epoch": 0.0008354, "grad_norm": 0.2928326725959778, "learning_rate": 8.353000000000001e-05, "loss": 0.1223, "step": 8354 }, { "epoch": 0.0008355, "grad_norm": 0.2385818064212799, "learning_rate": 8.354e-05, "loss": 0.1093, "step": 8355 }, { "epoch": 0.0008356, "grad_norm": 0.2374759018421173, "learning_rate": 8.355e-05, "loss": 0.1085, "step": 8356 }, { "epoch": 0.0008357, "grad_norm": 0.35778728127479553, "learning_rate": 8.356e-05, "loss": 0.1292, "step": 8357 }, { "epoch": 0.0008358, "grad_norm": 0.25890466570854187, "learning_rate": 8.357e-05, "loss": 0.1151, "step": 8358 }, { "epoch": 0.0008359, "grad_norm": 0.2757404148578644, "learning_rate": 8.358e-05, "loss": 0.1199, "step": 8359 }, { "epoch": 0.000836, "grad_norm": 0.28230440616607666, "learning_rate": 8.359e-05, "loss": 0.113, "step": 8360 }, { "epoch": 0.0008361, "grad_norm": 0.3027632534503937, "learning_rate": 8.36e-05, "loss": 0.1225, "step": 8361 }, { "epoch": 0.0008362, "grad_norm": 0.2576223611831665, "learning_rate": 8.361e-05, "loss": 0.116, "step": 8362 }, { "epoch": 0.0008363, "grad_norm": 0.392742395401001, "learning_rate": 8.362e-05, "loss": 0.127, "step": 8363 }, { "epoch": 0.0008364, "grad_norm": 0.22860021889209747, "learning_rate": 8.363000000000001e-05, "loss": 0.1018, "step": 8364 }, { "epoch": 0.0008365, "grad_norm": 0.5199876427650452, "learning_rate": 8.363999999999999e-05, "loss": 0.126, "step": 8365 }, { "epoch": 0.0008366, "grad_norm": 0.43568482995033264, "learning_rate": 8.365e-05, "loss": 0.1201, "step": 8366 }, { "epoch": 0.0008367, "grad_norm": 0.3191339373588562, "learning_rate": 8.366000000000001e-05, "loss": 0.1221, "step": 8367 }, { "epoch": 0.0008368, "grad_norm": 0.2429150938987732, "learning_rate": 8.367e-05, "loss": 0.1099, "step": 8368 }, { "epoch": 0.0008369, "grad_norm": 0.2731536030769348, "learning_rate": 8.368e-05, "loss": 0.1168, "step": 8369 }, { "epoch": 0.000837, "grad_norm": 0.2571616470813751, "learning_rate": 8.369000000000001e-05, "loss": 0.1115, "step": 8370 }, { "epoch": 0.0008371, "grad_norm": 0.2584238052368164, "learning_rate": 8.37e-05, "loss": 0.108, "step": 8371 }, { "epoch": 0.0008372, "grad_norm": 0.6482557654380798, "learning_rate": 8.371e-05, "loss": 0.1639, "step": 8372 }, { "epoch": 0.0008373, "grad_norm": 0.40680238604545593, "learning_rate": 8.372e-05, "loss": 0.1381, "step": 8373 }, { "epoch": 0.0008374, "grad_norm": 0.2401718944311142, "learning_rate": 8.373e-05, "loss": 0.1072, "step": 8374 }, { "epoch": 0.0008375, "grad_norm": 0.28217509388923645, "learning_rate": 8.374e-05, "loss": 0.1088, "step": 8375 }, { "epoch": 0.0008376, "grad_norm": 0.4751772880554199, "learning_rate": 8.375e-05, "loss": 0.1427, "step": 8376 }, { "epoch": 0.0008377, "grad_norm": 0.25133126974105835, "learning_rate": 8.376e-05, "loss": 0.1102, "step": 8377 }, { "epoch": 0.0008378, "grad_norm": 0.29090291261672974, "learning_rate": 8.377e-05, "loss": 0.1121, "step": 8378 }, { "epoch": 0.0008379, "grad_norm": 0.2963152825832367, "learning_rate": 8.378e-05, "loss": 0.1144, "step": 8379 }, { "epoch": 0.000838, "grad_norm": 0.3084557354450226, "learning_rate": 8.379e-05, "loss": 0.1188, "step": 8380 }, { "epoch": 0.0008381, "grad_norm": 0.29527875781059265, "learning_rate": 8.38e-05, "loss": 0.1141, "step": 8381 }, { "epoch": 0.0008382, "grad_norm": 0.7038413882255554, "learning_rate": 8.381e-05, "loss": 0.154, "step": 8382 }, { "epoch": 0.0008383, "grad_norm": 0.25467610359191895, "learning_rate": 8.382e-05, "loss": 0.1071, "step": 8383 }, { "epoch": 0.0008384, "grad_norm": 0.29565995931625366, "learning_rate": 8.383e-05, "loss": 0.1226, "step": 8384 }, { "epoch": 0.0008385, "grad_norm": 0.29088881611824036, "learning_rate": 8.384000000000001e-05, "loss": 0.115, "step": 8385 }, { "epoch": 0.0008386, "grad_norm": 0.24134796857833862, "learning_rate": 8.385e-05, "loss": 0.1085, "step": 8386 }, { "epoch": 0.0008387, "grad_norm": 0.46633341908454895, "learning_rate": 8.386e-05, "loss": 0.1179, "step": 8387 }, { "epoch": 0.0008388, "grad_norm": 0.2913179397583008, "learning_rate": 8.387000000000001e-05, "loss": 0.1212, "step": 8388 }, { "epoch": 0.0008389, "grad_norm": 0.3899138271808624, "learning_rate": 8.388e-05, "loss": 0.1265, "step": 8389 }, { "epoch": 0.000839, "grad_norm": 0.4130428731441498, "learning_rate": 8.389e-05, "loss": 0.1283, "step": 8390 }, { "epoch": 0.0008391, "grad_norm": 0.3082103431224823, "learning_rate": 8.39e-05, "loss": 0.1255, "step": 8391 }, { "epoch": 0.0008392, "grad_norm": 0.40410521626472473, "learning_rate": 8.391e-05, "loss": 0.122, "step": 8392 }, { "epoch": 0.0008393, "grad_norm": 0.9000693559646606, "learning_rate": 8.392e-05, "loss": 0.1403, "step": 8393 }, { "epoch": 0.0008394, "grad_norm": 0.2621038854122162, "learning_rate": 8.393e-05, "loss": 0.1055, "step": 8394 }, { "epoch": 0.0008395, "grad_norm": 0.2831513583660126, "learning_rate": 8.394e-05, "loss": 0.1067, "step": 8395 }, { "epoch": 0.0008396, "grad_norm": 0.44854921102523804, "learning_rate": 8.395e-05, "loss": 0.1294, "step": 8396 }, { "epoch": 0.0008397, "grad_norm": 0.28162941336631775, "learning_rate": 8.396e-05, "loss": 0.1076, "step": 8397 }, { "epoch": 0.0008398, "grad_norm": 0.33356034755706787, "learning_rate": 8.397000000000001e-05, "loss": 0.1154, "step": 8398 }, { "epoch": 0.0008399, "grad_norm": 0.2977070212364197, "learning_rate": 8.397999999999999e-05, "loss": 0.1146, "step": 8399 }, { "epoch": 0.00084, "grad_norm": 0.6036233305931091, "learning_rate": 8.399e-05, "loss": 0.1455, "step": 8400 }, { "epoch": 0.0008401, "grad_norm": 0.26441118121147156, "learning_rate": 8.400000000000001e-05, "loss": 0.1116, "step": 8401 }, { "epoch": 0.0008402, "grad_norm": 0.37586623430252075, "learning_rate": 8.401e-05, "loss": 0.1124, "step": 8402 }, { "epoch": 0.0008403, "grad_norm": 0.42590731382369995, "learning_rate": 8.402e-05, "loss": 0.1282, "step": 8403 }, { "epoch": 0.0008404, "grad_norm": 0.3122142553329468, "learning_rate": 8.403000000000001e-05, "loss": 0.1232, "step": 8404 }, { "epoch": 0.0008405, "grad_norm": 0.6514351963996887, "learning_rate": 8.404e-05, "loss": 0.142, "step": 8405 }, { "epoch": 0.0008406, "grad_norm": 0.2786729633808136, "learning_rate": 8.405e-05, "loss": 0.111, "step": 8406 }, { "epoch": 0.0008407, "grad_norm": 0.28742948174476624, "learning_rate": 8.406e-05, "loss": 0.1119, "step": 8407 }, { "epoch": 0.0008408, "grad_norm": 0.320432186126709, "learning_rate": 8.407e-05, "loss": 0.1128, "step": 8408 }, { "epoch": 0.0008409, "grad_norm": 0.3040125072002411, "learning_rate": 8.408e-05, "loss": 0.1146, "step": 8409 }, { "epoch": 0.000841, "grad_norm": 0.39792320132255554, "learning_rate": 8.409e-05, "loss": 0.1161, "step": 8410 }, { "epoch": 0.0008411, "grad_norm": 0.26484251022338867, "learning_rate": 8.41e-05, "loss": 0.1102, "step": 8411 }, { "epoch": 0.0008412, "grad_norm": 0.25933507084846497, "learning_rate": 8.411e-05, "loss": 0.1091, "step": 8412 }, { "epoch": 0.0008413, "grad_norm": 0.2580576241016388, "learning_rate": 8.412e-05, "loss": 0.1069, "step": 8413 }, { "epoch": 0.0008414, "grad_norm": 0.26504597067832947, "learning_rate": 8.413e-05, "loss": 0.1089, "step": 8414 }, { "epoch": 0.0008415, "grad_norm": 0.26909440755844116, "learning_rate": 8.414e-05, "loss": 0.1143, "step": 8415 }, { "epoch": 0.0008416, "grad_norm": 0.25820961594581604, "learning_rate": 8.415e-05, "loss": 0.1106, "step": 8416 }, { "epoch": 0.0008417, "grad_norm": 0.23238538205623627, "learning_rate": 8.416e-05, "loss": 0.1038, "step": 8417 }, { "epoch": 0.0008418, "grad_norm": 0.335588663816452, "learning_rate": 8.417e-05, "loss": 0.1224, "step": 8418 }, { "epoch": 0.0008419, "grad_norm": 0.5604079365730286, "learning_rate": 8.418000000000001e-05, "loss": 0.1375, "step": 8419 }, { "epoch": 0.000842, "grad_norm": 0.2831794321537018, "learning_rate": 8.419e-05, "loss": 0.1165, "step": 8420 }, { "epoch": 0.0008421, "grad_norm": 0.6355200409889221, "learning_rate": 8.42e-05, "loss": 0.127, "step": 8421 }, { "epoch": 0.0008422, "grad_norm": 0.2544653117656708, "learning_rate": 8.421000000000001e-05, "loss": 0.1092, "step": 8422 }, { "epoch": 0.0008423, "grad_norm": 0.28864893317222595, "learning_rate": 8.422e-05, "loss": 0.1159, "step": 8423 }, { "epoch": 0.0008424, "grad_norm": 0.32663410902023315, "learning_rate": 8.423e-05, "loss": 0.1195, "step": 8424 }, { "epoch": 0.0008425, "grad_norm": 0.2896595597267151, "learning_rate": 8.424e-05, "loss": 0.1182, "step": 8425 }, { "epoch": 0.0008426, "grad_norm": 0.5385744571685791, "learning_rate": 8.425e-05, "loss": 0.1254, "step": 8426 }, { "epoch": 0.0008427, "grad_norm": 0.2622899115085602, "learning_rate": 8.426e-05, "loss": 0.1069, "step": 8427 }, { "epoch": 0.0008428, "grad_norm": 0.4006927013397217, "learning_rate": 8.427e-05, "loss": 0.1242, "step": 8428 }, { "epoch": 0.0008429, "grad_norm": 0.4451708197593689, "learning_rate": 8.428e-05, "loss": 0.1318, "step": 8429 }, { "epoch": 0.000843, "grad_norm": 0.25315025448799133, "learning_rate": 8.429e-05, "loss": 0.1099, "step": 8430 }, { "epoch": 0.0008431, "grad_norm": 0.4274576008319855, "learning_rate": 8.43e-05, "loss": 0.1301, "step": 8431 }, { "epoch": 0.0008432, "grad_norm": 0.2771688997745514, "learning_rate": 8.431000000000001e-05, "loss": 0.1118, "step": 8432 }, { "epoch": 0.0008433, "grad_norm": 0.24496810138225555, "learning_rate": 8.431999999999999e-05, "loss": 0.1057, "step": 8433 }, { "epoch": 0.0008434, "grad_norm": 0.25610044598579407, "learning_rate": 8.433e-05, "loss": 0.1071, "step": 8434 }, { "epoch": 0.0008435, "grad_norm": 0.26334473490715027, "learning_rate": 8.434000000000001e-05, "loss": 0.1143, "step": 8435 }, { "epoch": 0.0008436, "grad_norm": 0.2896568477153778, "learning_rate": 8.435e-05, "loss": 0.1201, "step": 8436 }, { "epoch": 0.0008437, "grad_norm": 0.455269992351532, "learning_rate": 8.436e-05, "loss": 0.1387, "step": 8437 }, { "epoch": 0.0008438, "grad_norm": 0.33185434341430664, "learning_rate": 8.437000000000001e-05, "loss": 0.1221, "step": 8438 }, { "epoch": 0.0008439, "grad_norm": 0.2473549097776413, "learning_rate": 8.438e-05, "loss": 0.1053, "step": 8439 }, { "epoch": 0.000844, "grad_norm": 0.63054358959198, "learning_rate": 8.439e-05, "loss": 0.1376, "step": 8440 }, { "epoch": 0.0008441, "grad_norm": 0.37012726068496704, "learning_rate": 8.44e-05, "loss": 0.1287, "step": 8441 }, { "epoch": 0.0008442, "grad_norm": 0.3412618041038513, "learning_rate": 8.441e-05, "loss": 0.131, "step": 8442 }, { "epoch": 0.0008443, "grad_norm": 0.6600422263145447, "learning_rate": 8.442e-05, "loss": 0.146, "step": 8443 }, { "epoch": 0.0008444, "grad_norm": 0.8839293122291565, "learning_rate": 8.443e-05, "loss": 0.154, "step": 8444 }, { "epoch": 0.0008445, "grad_norm": 0.28500112891197205, "learning_rate": 8.444e-05, "loss": 0.1059, "step": 8445 }, { "epoch": 0.0008446, "grad_norm": 0.3337307870388031, "learning_rate": 8.445e-05, "loss": 0.1151, "step": 8446 }, { "epoch": 0.0008447, "grad_norm": 0.3037877380847931, "learning_rate": 8.446e-05, "loss": 0.1074, "step": 8447 }, { "epoch": 0.0008448, "grad_norm": 0.5050310492515564, "learning_rate": 8.447e-05, "loss": 0.1215, "step": 8448 }, { "epoch": 0.0008449, "grad_norm": 0.3303443193435669, "learning_rate": 8.448e-05, "loss": 0.1108, "step": 8449 }, { "epoch": 0.000845, "grad_norm": 0.3081909120082855, "learning_rate": 8.449e-05, "loss": 0.1127, "step": 8450 }, { "epoch": 0.0008451, "grad_norm": 0.28372883796691895, "learning_rate": 8.45e-05, "loss": 0.1074, "step": 8451 }, { "epoch": 0.0008452, "grad_norm": 0.2864672541618347, "learning_rate": 8.451e-05, "loss": 0.112, "step": 8452 }, { "epoch": 0.0008453, "grad_norm": 0.2667675018310547, "learning_rate": 8.452000000000001e-05, "loss": 0.1096, "step": 8453 }, { "epoch": 0.0008454, "grad_norm": 0.2535669803619385, "learning_rate": 8.453e-05, "loss": 0.1085, "step": 8454 }, { "epoch": 0.0008455, "grad_norm": 1.1039838790893555, "learning_rate": 8.454e-05, "loss": 0.1671, "step": 8455 }, { "epoch": 0.0008456, "grad_norm": 0.27928635478019714, "learning_rate": 8.455000000000001e-05, "loss": 0.1092, "step": 8456 }, { "epoch": 0.0008457, "grad_norm": 0.2974961996078491, "learning_rate": 8.456e-05, "loss": 0.1086, "step": 8457 }, { "epoch": 0.0008458, "grad_norm": 0.43167659640312195, "learning_rate": 8.457e-05, "loss": 0.1269, "step": 8458 }, { "epoch": 0.0008459, "grad_norm": 0.2962111830711365, "learning_rate": 8.458e-05, "loss": 0.1082, "step": 8459 }, { "epoch": 0.000846, "grad_norm": 0.28472232818603516, "learning_rate": 8.459e-05, "loss": 0.114, "step": 8460 }, { "epoch": 0.0008461, "grad_norm": 0.3290531039237976, "learning_rate": 8.46e-05, "loss": 0.1186, "step": 8461 }, { "epoch": 0.0008462, "grad_norm": 0.31834477186203003, "learning_rate": 8.461e-05, "loss": 0.1156, "step": 8462 }, { "epoch": 0.0008463, "grad_norm": 2.159715414047241, "learning_rate": 8.462e-05, "loss": 0.1547, "step": 8463 }, { "epoch": 0.0008464, "grad_norm": 0.3941461741924286, "learning_rate": 8.463e-05, "loss": 0.1165, "step": 8464 }, { "epoch": 0.0008465, "grad_norm": 0.39252033829689026, "learning_rate": 8.464e-05, "loss": 0.1136, "step": 8465 }, { "epoch": 0.0008466, "grad_norm": 0.4327937364578247, "learning_rate": 8.465000000000001e-05, "loss": 0.1255, "step": 8466 }, { "epoch": 0.0008467, "grad_norm": 0.457118421792984, "learning_rate": 8.465999999999999e-05, "loss": 0.1273, "step": 8467 }, { "epoch": 0.0008468, "grad_norm": 0.6907281279563904, "learning_rate": 8.467e-05, "loss": 0.1348, "step": 8468 }, { "epoch": 0.0008469, "grad_norm": 0.29370391368865967, "learning_rate": 8.468000000000001e-05, "loss": 0.1116, "step": 8469 }, { "epoch": 0.000847, "grad_norm": 0.5513447523117065, "learning_rate": 8.469e-05, "loss": 0.1266, "step": 8470 }, { "epoch": 0.0008471, "grad_norm": 0.29566922783851624, "learning_rate": 8.47e-05, "loss": 0.1077, "step": 8471 }, { "epoch": 0.0008472, "grad_norm": 0.3664743900299072, "learning_rate": 8.471000000000001e-05, "loss": 0.1161, "step": 8472 }, { "epoch": 0.0008473, "grad_norm": 0.3511149287223816, "learning_rate": 8.472e-05, "loss": 0.1137, "step": 8473 }, { "epoch": 0.0008474, "grad_norm": 0.39576053619384766, "learning_rate": 8.473e-05, "loss": 0.1152, "step": 8474 }, { "epoch": 0.0008475, "grad_norm": 0.31602656841278076, "learning_rate": 8.474e-05, "loss": 0.1146, "step": 8475 }, { "epoch": 0.0008476, "grad_norm": 0.28262972831726074, "learning_rate": 8.475e-05, "loss": 0.1112, "step": 8476 }, { "epoch": 0.0008477, "grad_norm": 0.3561192452907562, "learning_rate": 8.476e-05, "loss": 0.1194, "step": 8477 }, { "epoch": 0.0008478, "grad_norm": 0.30674663186073303, "learning_rate": 8.477e-05, "loss": 0.1216, "step": 8478 }, { "epoch": 0.0008479, "grad_norm": 0.5238985419273376, "learning_rate": 8.478e-05, "loss": 0.1331, "step": 8479 }, { "epoch": 0.000848, "grad_norm": 0.3820216655731201, "learning_rate": 8.479e-05, "loss": 0.1179, "step": 8480 }, { "epoch": 0.0008481, "grad_norm": 0.30187010765075684, "learning_rate": 8.48e-05, "loss": 0.1151, "step": 8481 }, { "epoch": 0.0008482, "grad_norm": 0.3691226541996002, "learning_rate": 8.481e-05, "loss": 0.1114, "step": 8482 }, { "epoch": 0.0008483, "grad_norm": 0.293366014957428, "learning_rate": 8.482e-05, "loss": 0.114, "step": 8483 }, { "epoch": 0.0008484, "grad_norm": 0.2826138138771057, "learning_rate": 8.483e-05, "loss": 0.1098, "step": 8484 }, { "epoch": 0.0008485, "grad_norm": 0.4060609042644501, "learning_rate": 8.484e-05, "loss": 0.1251, "step": 8485 }, { "epoch": 0.0008486, "grad_norm": 0.23822547495365143, "learning_rate": 8.485e-05, "loss": 0.1011, "step": 8486 }, { "epoch": 0.0008487, "grad_norm": 0.31863072514533997, "learning_rate": 8.486000000000001e-05, "loss": 0.1199, "step": 8487 }, { "epoch": 0.0008488, "grad_norm": 0.7325024604797363, "learning_rate": 8.487e-05, "loss": 0.1514, "step": 8488 }, { "epoch": 0.0008489, "grad_norm": 0.2716192305088043, "learning_rate": 8.488e-05, "loss": 0.1089, "step": 8489 }, { "epoch": 0.000849, "grad_norm": 0.3262501060962677, "learning_rate": 8.489000000000001e-05, "loss": 0.1185, "step": 8490 }, { "epoch": 0.0008491, "grad_norm": 0.27587273716926575, "learning_rate": 8.49e-05, "loss": 0.1134, "step": 8491 }, { "epoch": 0.0008492, "grad_norm": 0.31848254799842834, "learning_rate": 8.491e-05, "loss": 0.1135, "step": 8492 }, { "epoch": 0.0008493, "grad_norm": 0.3487236797809601, "learning_rate": 8.492000000000001e-05, "loss": 0.1143, "step": 8493 }, { "epoch": 0.0008494, "grad_norm": 0.2887483537197113, "learning_rate": 8.493e-05, "loss": 0.1134, "step": 8494 }, { "epoch": 0.0008495, "grad_norm": 3.056828498840332, "learning_rate": 8.494e-05, "loss": 0.2388, "step": 8495 }, { "epoch": 0.0008496, "grad_norm": 0.4552193582057953, "learning_rate": 8.495e-05, "loss": 0.1301, "step": 8496 }, { "epoch": 0.0008497, "grad_norm": 0.2789609134197235, "learning_rate": 8.496e-05, "loss": 0.1079, "step": 8497 }, { "epoch": 0.0008498, "grad_norm": 0.2965538501739502, "learning_rate": 8.497e-05, "loss": 0.1034, "step": 8498 }, { "epoch": 0.0008499, "grad_norm": 0.31228092312812805, "learning_rate": 8.498e-05, "loss": 0.1108, "step": 8499 }, { "epoch": 0.00085, "grad_norm": 0.32444027066230774, "learning_rate": 8.499000000000001e-05, "loss": 0.1105, "step": 8500 }, { "epoch": 0.00085, "eval_loss": 0.012609053403139114, "eval_runtime": 362.403, "eval_samples_per_second": 27.594, "eval_steps_per_second": 1.725, "step": 8500 } ], "logging_steps": 1, "max_steps": 10000000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "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 }