| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.0001, | |
| "eval_steps": 500, | |
| "global_step": 1000, | |
| "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 | |
| } | |
| ], | |
| "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 | |
| } | |