| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 2500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0008, | |
| "grad_norm": 4.92004919052124, | |
| "learning_rate": 1.6e-07, | |
| "loss": 7.328, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0016, | |
| "grad_norm": 4.8694047927856445, | |
| "learning_rate": 3.2e-07, | |
| "loss": 7.5093, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0024, | |
| "grad_norm": 5.063427448272705, | |
| "learning_rate": 4.800000000000001e-07, | |
| "loss": 7.2167, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0032, | |
| "grad_norm": 5.227718830108643, | |
| "learning_rate": 6.4e-07, | |
| "loss": 7.6708, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.004, | |
| "grad_norm": 5.120015621185303, | |
| "learning_rate": 8.000000000000001e-07, | |
| "loss": 7.493, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0048, | |
| "grad_norm": 4.556617736816406, | |
| "learning_rate": 9.600000000000001e-07, | |
| "loss": 7.114, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.0056, | |
| "grad_norm": 4.134003162384033, | |
| "learning_rate": 1.12e-06, | |
| "loss": 7.0582, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.0064, | |
| "grad_norm": 3.023643970489502, | |
| "learning_rate": 1.28e-06, | |
| "loss": 6.2808, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.0072, | |
| "grad_norm": 2.4297778606414795, | |
| "learning_rate": 1.44e-06, | |
| "loss": 6.353, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.008, | |
| "grad_norm": 2.867222785949707, | |
| "learning_rate": 1.6000000000000001e-06, | |
| "loss": 6.0667, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0088, | |
| "grad_norm": 1.387591004371643, | |
| "learning_rate": 1.76e-06, | |
| "loss": 5.6459, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.0096, | |
| "grad_norm": 2.594311475753784, | |
| "learning_rate": 1.9200000000000003e-06, | |
| "loss": 5.8506, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.0104, | |
| "grad_norm": 1.8094066381454468, | |
| "learning_rate": 2.08e-06, | |
| "loss": 5.6573, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.0112, | |
| "grad_norm": 0.9034324288368225, | |
| "learning_rate": 2.24e-06, | |
| "loss": 5.231, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.012, | |
| "grad_norm": 1.4179000854492188, | |
| "learning_rate": 2.4000000000000003e-06, | |
| "loss": 5.3384, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.0128, | |
| "grad_norm": 1.3942221403121948, | |
| "learning_rate": 2.56e-06, | |
| "loss": 5.385, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.0136, | |
| "grad_norm": 1.5524358749389648, | |
| "learning_rate": 2.7200000000000002e-06, | |
| "loss": 4.8145, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.0144, | |
| "grad_norm": 1.9171744585037231, | |
| "learning_rate": 2.88e-06, | |
| "loss": 5.5795, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.0152, | |
| "grad_norm": 2.5012264251708984, | |
| "learning_rate": 3.04e-06, | |
| "loss": 5.1458, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.016, | |
| "grad_norm": 1.894866943359375, | |
| "learning_rate": 3.2000000000000003e-06, | |
| "loss": 4.7375, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.0168, | |
| "grad_norm": 1.083099365234375, | |
| "learning_rate": 3.3600000000000004e-06, | |
| "loss": 4.659, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.0176, | |
| "grad_norm": 1.7351473569869995, | |
| "learning_rate": 3.52e-06, | |
| "loss": 4.3309, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.0184, | |
| "grad_norm": 1.2516050338745117, | |
| "learning_rate": 3.6800000000000003e-06, | |
| "loss": 4.2946, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.0192, | |
| "grad_norm": 1.946075439453125, | |
| "learning_rate": 3.8400000000000005e-06, | |
| "loss": 4.2101, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 2.922369956970215, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 4.5236, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.0208, | |
| "grad_norm": 2.0061962604522705, | |
| "learning_rate": 4.16e-06, | |
| "loss": 3.6378, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.0216, | |
| "grad_norm": 2.2109129428863525, | |
| "learning_rate": 4.32e-06, | |
| "loss": 3.1168, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.0224, | |
| "grad_norm": 2.374847173690796, | |
| "learning_rate": 4.48e-06, | |
| "loss": 2.9689, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.0232, | |
| "grad_norm": 2.2352538108825684, | |
| "learning_rate": 4.6400000000000005e-06, | |
| "loss": 2.2496, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.024, | |
| "grad_norm": 1.657749891281128, | |
| "learning_rate": 4.800000000000001e-06, | |
| "loss": 2.9017, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.0248, | |
| "grad_norm": 2.9958770275115967, | |
| "learning_rate": 4.960000000000001e-06, | |
| "loss": 2.094, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.0256, | |
| "grad_norm": 1.8515348434448242, | |
| "learning_rate": 5.12e-06, | |
| "loss": 1.9278, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.0264, | |
| "grad_norm": 2.1560873985290527, | |
| "learning_rate": 5.28e-06, | |
| "loss": 1.7917, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.0272, | |
| "grad_norm": 3.1304805278778076, | |
| "learning_rate": 5.4400000000000004e-06, | |
| "loss": 1.9759, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.028, | |
| "grad_norm": 3.0489299297332764, | |
| "learning_rate": 5.600000000000001e-06, | |
| "loss": 1.8824, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.0288, | |
| "grad_norm": 3.545546054840088, | |
| "learning_rate": 5.76e-06, | |
| "loss": 2.0749, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.0296, | |
| "grad_norm": 2.089024305343628, | |
| "learning_rate": 5.92e-06, | |
| "loss": 1.0014, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.0304, | |
| "grad_norm": 1.475943922996521, | |
| "learning_rate": 6.08e-06, | |
| "loss": 1.6105, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.0312, | |
| "grad_norm": 1.4888969659805298, | |
| "learning_rate": 6.24e-06, | |
| "loss": 1.2915, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.032, | |
| "grad_norm": 1.1156069040298462, | |
| "learning_rate": 6.4000000000000006e-06, | |
| "loss": 0.6639, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.0328, | |
| "grad_norm": 0.5972667932510376, | |
| "learning_rate": 6.560000000000001e-06, | |
| "loss": 0.4398, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.0336, | |
| "grad_norm": 1.5191650390625, | |
| "learning_rate": 6.720000000000001e-06, | |
| "loss": 1.0727, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.0344, | |
| "grad_norm": 4.311778545379639, | |
| "learning_rate": 6.88e-06, | |
| "loss": 1.6288, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.0352, | |
| "grad_norm": 2.106018543243408, | |
| "learning_rate": 7.04e-06, | |
| "loss": 0.5208, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.036, | |
| "grad_norm": 2.2589375972747803, | |
| "learning_rate": 7.2000000000000005e-06, | |
| "loss": 0.4426, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.0368, | |
| "grad_norm": 8.470863342285156, | |
| "learning_rate": 7.360000000000001e-06, | |
| "loss": 2.5704, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.0376, | |
| "grad_norm": 3.564351797103882, | |
| "learning_rate": 7.520000000000001e-06, | |
| "loss": 0.9683, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 3.827845573425293, | |
| "learning_rate": 7.680000000000001e-06, | |
| "loss": 0.5685, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.0392, | |
| "grad_norm": 3.611189842224121, | |
| "learning_rate": 7.840000000000001e-06, | |
| "loss": 1.1535, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 1.9307868480682373, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 0.6147, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.0408, | |
| "grad_norm": 3.3078556060791016, | |
| "learning_rate": 8.16e-06, | |
| "loss": 0.9248, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.0416, | |
| "grad_norm": 1.50742506980896, | |
| "learning_rate": 8.32e-06, | |
| "loss": 0.5147, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.0424, | |
| "grad_norm": 2.5260138511657715, | |
| "learning_rate": 8.48e-06, | |
| "loss": 0.5129, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.0432, | |
| "grad_norm": 0.7977801561355591, | |
| "learning_rate": 8.64e-06, | |
| "loss": 0.5833, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.044, | |
| "grad_norm": 3.389085054397583, | |
| "learning_rate": 8.8e-06, | |
| "loss": 0.4162, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.0448, | |
| "grad_norm": 3.584988832473755, | |
| "learning_rate": 8.96e-06, | |
| "loss": 1.2424, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.0456, | |
| "grad_norm": 0.4079441428184509, | |
| "learning_rate": 9.12e-06, | |
| "loss": 0.6532, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.0464, | |
| "grad_norm": 2.021636962890625, | |
| "learning_rate": 9.280000000000001e-06, | |
| "loss": 0.2637, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.0472, | |
| "grad_norm": 4.71144962310791, | |
| "learning_rate": 9.440000000000001e-06, | |
| "loss": 1.1383, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.048, | |
| "grad_norm": 1.831689715385437, | |
| "learning_rate": 9.600000000000001e-06, | |
| "loss": 0.7587, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.0488, | |
| "grad_norm": 2.6891868114471436, | |
| "learning_rate": 9.760000000000001e-06, | |
| "loss": 0.9849, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.0496, | |
| "grad_norm": 2.2791528701782227, | |
| "learning_rate": 9.920000000000002e-06, | |
| "loss": 0.74, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.0504, | |
| "grad_norm": 2.12807559967041, | |
| "learning_rate": 1.008e-05, | |
| "loss": 0.3742, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.0512, | |
| "grad_norm": 1.7115131616592407, | |
| "learning_rate": 1.024e-05, | |
| "loss": 0.8209, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.052, | |
| "grad_norm": 1.7355319261550903, | |
| "learning_rate": 1.04e-05, | |
| "loss": 1.1005, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.0528, | |
| "grad_norm": 1.7057139873504639, | |
| "learning_rate": 1.056e-05, | |
| "loss": 0.9056, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.0536, | |
| "grad_norm": 0.7381348609924316, | |
| "learning_rate": 1.072e-05, | |
| "loss": 0.4982, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.0544, | |
| "grad_norm": 3.6813762187957764, | |
| "learning_rate": 1.0880000000000001e-05, | |
| "loss": 0.8425, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.0552, | |
| "grad_norm": 0.6442511677742004, | |
| "learning_rate": 1.1040000000000001e-05, | |
| "loss": 0.4714, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.056, | |
| "grad_norm": 0.3334783613681793, | |
| "learning_rate": 1.1200000000000001e-05, | |
| "loss": 0.1425, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.0568, | |
| "grad_norm": 2.7860682010650635, | |
| "learning_rate": 1.136e-05, | |
| "loss": 1.0082, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.0576, | |
| "grad_norm": 1.0623384714126587, | |
| "learning_rate": 1.152e-05, | |
| "loss": 0.1404, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.0584, | |
| "grad_norm": 0.49509289860725403, | |
| "learning_rate": 1.168e-05, | |
| "loss": 0.3523, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.0592, | |
| "grad_norm": 0.9187251925468445, | |
| "learning_rate": 1.184e-05, | |
| "loss": 0.7012, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 2.251713514328003, | |
| "learning_rate": 1.2e-05, | |
| "loss": 0.3998, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.0608, | |
| "grad_norm": 3.6301510334014893, | |
| "learning_rate": 1.216e-05, | |
| "loss": 0.991, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.0616, | |
| "grad_norm": 2.2427806854248047, | |
| "learning_rate": 1.232e-05, | |
| "loss": 1.6949, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.0624, | |
| "grad_norm": 1.5521266460418701, | |
| "learning_rate": 1.248e-05, | |
| "loss": 0.9281, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.0632, | |
| "grad_norm": 0.9885201454162598, | |
| "learning_rate": 1.2640000000000001e-05, | |
| "loss": 0.3829, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 1.02154541015625, | |
| "learning_rate": 1.2800000000000001e-05, | |
| "loss": 0.1986, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.0648, | |
| "grad_norm": 2.9700698852539062, | |
| "learning_rate": 1.2960000000000001e-05, | |
| "loss": 0.5764, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.0656, | |
| "grad_norm": 1.9716620445251465, | |
| "learning_rate": 1.3120000000000001e-05, | |
| "loss": 0.4319, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.0664, | |
| "grad_norm": 0.23318687081336975, | |
| "learning_rate": 1.3280000000000002e-05, | |
| "loss": 0.7806, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.0672, | |
| "grad_norm": 1.2755334377288818, | |
| "learning_rate": 1.3440000000000002e-05, | |
| "loss": 0.2845, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.068, | |
| "grad_norm": 0.26743704080581665, | |
| "learning_rate": 1.3600000000000002e-05, | |
| "loss": 0.1663, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.0688, | |
| "grad_norm": 1.2172751426696777, | |
| "learning_rate": 1.376e-05, | |
| "loss": 0.3756, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.0696, | |
| "grad_norm": 3.7812864780426025, | |
| "learning_rate": 1.392e-05, | |
| "loss": 0.4978, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.0704, | |
| "grad_norm": 3.4768078327178955, | |
| "learning_rate": 1.408e-05, | |
| "loss": 0.9643, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.0712, | |
| "grad_norm": 3.9793436527252197, | |
| "learning_rate": 1.4240000000000001e-05, | |
| "loss": 0.158, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.072, | |
| "grad_norm": 2.2694666385650635, | |
| "learning_rate": 1.4400000000000001e-05, | |
| "loss": 0.881, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.0728, | |
| "grad_norm": 0.1139988824725151, | |
| "learning_rate": 1.4560000000000001e-05, | |
| "loss": 0.017, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.0736, | |
| "grad_norm": 2.4055392742156982, | |
| "learning_rate": 1.4720000000000001e-05, | |
| "loss": 0.1059, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.0744, | |
| "grad_norm": 1.4160810708999634, | |
| "learning_rate": 1.4880000000000002e-05, | |
| "loss": 0.1963, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.0752, | |
| "grad_norm": 1.9410396814346313, | |
| "learning_rate": 1.5040000000000002e-05, | |
| "loss": 0.1013, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.076, | |
| "grad_norm": 1.283490777015686, | |
| "learning_rate": 1.5200000000000002e-05, | |
| "loss": 0.2272, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 1.230212688446045, | |
| "learning_rate": 1.5360000000000002e-05, | |
| "loss": 0.7391, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.0776, | |
| "grad_norm": 9.068771362304688, | |
| "learning_rate": 1.552e-05, | |
| "loss": 0.5673, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.0784, | |
| "grad_norm": 3.8310606479644775, | |
| "learning_rate": 1.5680000000000002e-05, | |
| "loss": 1.1112, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.0792, | |
| "grad_norm": 3.0830278396606445, | |
| "learning_rate": 1.584e-05, | |
| "loss": 0.711, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 8.513973236083984, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 1.9858, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.0808, | |
| "grad_norm": 0.3588366210460663, | |
| "learning_rate": 1.616e-05, | |
| "loss": 0.149, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.0816, | |
| "grad_norm": 1.0215955972671509, | |
| "learning_rate": 1.632e-05, | |
| "loss": 0.7344, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.0824, | |
| "grad_norm": 1.8482775688171387, | |
| "learning_rate": 1.648e-05, | |
| "loss": 1.0866, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.0832, | |
| "grad_norm": 0.7888800501823425, | |
| "learning_rate": 1.664e-05, | |
| "loss": 0.5496, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.084, | |
| "grad_norm": 0.33580031991004944, | |
| "learning_rate": 1.6800000000000002e-05, | |
| "loss": 0.2747, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.0848, | |
| "grad_norm": 1.4474271535873413, | |
| "learning_rate": 1.696e-05, | |
| "loss": 0.3155, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.0856, | |
| "grad_norm": 2.6667094230651855, | |
| "learning_rate": 1.7120000000000002e-05, | |
| "loss": 1.1201, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.0864, | |
| "grad_norm": 1.0721205472946167, | |
| "learning_rate": 1.728e-05, | |
| "loss": 0.3277, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.0872, | |
| "grad_norm": 0.2696588635444641, | |
| "learning_rate": 1.7440000000000002e-05, | |
| "loss": 0.5612, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.088, | |
| "grad_norm": 2.505485773086548, | |
| "learning_rate": 1.76e-05, | |
| "loss": 0.2816, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.0888, | |
| "grad_norm": 0.6796432137489319, | |
| "learning_rate": 1.7760000000000003e-05, | |
| "loss": 0.091, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.0896, | |
| "grad_norm": 0.23082202672958374, | |
| "learning_rate": 1.792e-05, | |
| "loss": 0.7718, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.0904, | |
| "grad_norm": 1.6087664365768433, | |
| "learning_rate": 1.8080000000000003e-05, | |
| "loss": 0.8094, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.0912, | |
| "grad_norm": 1.131905198097229, | |
| "learning_rate": 1.824e-05, | |
| "loss": 0.1958, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.092, | |
| "grad_norm": 1.3417667150497437, | |
| "learning_rate": 1.8400000000000003e-05, | |
| "loss": 0.0899, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.0928, | |
| "grad_norm": 1.7206881046295166, | |
| "learning_rate": 1.8560000000000002e-05, | |
| "loss": 0.1773, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.0936, | |
| "grad_norm": 1.3308982849121094, | |
| "learning_rate": 1.8720000000000004e-05, | |
| "loss": 0.6628, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.0944, | |
| "grad_norm": 0.06212015450000763, | |
| "learning_rate": 1.8880000000000002e-05, | |
| "loss": 0.3871, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.0952, | |
| "grad_norm": 2.4711997509002686, | |
| "learning_rate": 1.904e-05, | |
| "loss": 0.2319, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.096, | |
| "grad_norm": 3.6728742122650146, | |
| "learning_rate": 1.9200000000000003e-05, | |
| "loss": 0.5643, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.0968, | |
| "grad_norm": 0.1921927034854889, | |
| "learning_rate": 1.936e-05, | |
| "loss": 0.0931, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.0976, | |
| "grad_norm": 1.9750112295150757, | |
| "learning_rate": 1.9520000000000003e-05, | |
| "loss": 0.1632, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.0984, | |
| "grad_norm": 2.1710243225097656, | |
| "learning_rate": 1.968e-05, | |
| "loss": 2.1496, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.0992, | |
| "grad_norm": 0.6705166697502136, | |
| "learning_rate": 1.9840000000000003e-05, | |
| "loss": 0.2102, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.9208804368972778, | |
| "learning_rate": 2e-05, | |
| "loss": 0.363, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.1008, | |
| "grad_norm": 0.25396543741226196, | |
| "learning_rate": 1.9999961008995607e-05, | |
| "loss": 0.9275, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.1016, | |
| "grad_norm": 3.0983469486236572, | |
| "learning_rate": 1.9999844036286483e-05, | |
| "loss": 1.6902, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.1024, | |
| "grad_norm": 0.8619891405105591, | |
| "learning_rate": 1.9999649082784807e-05, | |
| "loss": 0.9756, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.1032, | |
| "grad_norm": 1.3311007022857666, | |
| "learning_rate": 1.9999376150010868e-05, | |
| "loss": 1.0115, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.104, | |
| "grad_norm": 1.6311815977096558, | |
| "learning_rate": 1.9999025240093045e-05, | |
| "loss": 1.3444, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.1048, | |
| "grad_norm": 1.097953200340271, | |
| "learning_rate": 1.9998596355767805e-05, | |
| "loss": 0.5742, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.1056, | |
| "grad_norm": 0.12066880613565445, | |
| "learning_rate": 1.999808950037968e-05, | |
| "loss": 0.1853, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.1064, | |
| "grad_norm": 1.4088952541351318, | |
| "learning_rate": 1.9997504677881224e-05, | |
| "loss": 0.585, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.1072, | |
| "grad_norm": 0.200229674577713, | |
| "learning_rate": 1.9996841892833e-05, | |
| "loss": 0.0933, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.108, | |
| "grad_norm": 1.871922254562378, | |
| "learning_rate": 1.9996101150403543e-05, | |
| "loss": 1.4538, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.1088, | |
| "grad_norm": 0.5457477569580078, | |
| "learning_rate": 1.9995282456369313e-05, | |
| "loss": 0.3239, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.1096, | |
| "grad_norm": 0.6347384452819824, | |
| "learning_rate": 1.9994385817114644e-05, | |
| "loss": 0.8208, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.1104, | |
| "grad_norm": 0.8534179925918579, | |
| "learning_rate": 1.9993411239631713e-05, | |
| "loss": 1.9646, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.1112, | |
| "grad_norm": 0.3291590213775635, | |
| "learning_rate": 1.999235873152047e-05, | |
| "loss": 0.0857, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.112, | |
| "grad_norm": 0.5380199551582336, | |
| "learning_rate": 1.9991228300988586e-05, | |
| "loss": 0.1706, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.1128, | |
| "grad_norm": 1.0301982164382935, | |
| "learning_rate": 1.9990019956851384e-05, | |
| "loss": 0.7458, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.1136, | |
| "grad_norm": 0.8207939863204956, | |
| "learning_rate": 1.9988733708531772e-05, | |
| "loss": 0.8598, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.1144, | |
| "grad_norm": 0.11704489588737488, | |
| "learning_rate": 1.998736956606018e-05, | |
| "loss": 0.0593, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 0.8635666370391846, | |
| "learning_rate": 1.9985927540074453e-05, | |
| "loss": 0.4055, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.116, | |
| "grad_norm": 0.9759752154350281, | |
| "learning_rate": 1.9984407641819812e-05, | |
| "loss": 0.4182, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.1168, | |
| "grad_norm": 1.0212275981903076, | |
| "learning_rate": 1.998280988314872e-05, | |
| "loss": 0.3533, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.1176, | |
| "grad_norm": 0.36377808451652527, | |
| "learning_rate": 1.9981134276520828e-05, | |
| "loss": 0.1175, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.1184, | |
| "grad_norm": 1.5665260553359985, | |
| "learning_rate": 1.9979380835002846e-05, | |
| "loss": 0.2428, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.1192, | |
| "grad_norm": 1.0009765625, | |
| "learning_rate": 1.997754957226847e-05, | |
| "loss": 0.3959, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.43519535660743713, | |
| "learning_rate": 1.9975640502598243e-05, | |
| "loss": 0.3812, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.1208, | |
| "grad_norm": 1.0461794137954712, | |
| "learning_rate": 1.9973653640879486e-05, | |
| "loss": 0.5793, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.1216, | |
| "grad_norm": 1.0484498739242554, | |
| "learning_rate": 1.997158900260614e-05, | |
| "loss": 0.3245, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.1224, | |
| "grad_norm": 0.18130800127983093, | |
| "learning_rate": 1.9969446603878673e-05, | |
| "loss": 0.0716, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.1232, | |
| "grad_norm": 0.14019177854061127, | |
| "learning_rate": 1.9967226461403934e-05, | |
| "loss": 0.0946, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.124, | |
| "grad_norm": 0.48715105652809143, | |
| "learning_rate": 1.9964928592495046e-05, | |
| "loss": 0.1536, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.1248, | |
| "grad_norm": 0.4148877263069153, | |
| "learning_rate": 1.996255301507125e-05, | |
| "loss": 0.1165, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.1256, | |
| "grad_norm": 0.0783982053399086, | |
| "learning_rate": 1.9960099747657774e-05, | |
| "loss": 0.0288, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.1264, | |
| "grad_norm": 4.088832378387451, | |
| "learning_rate": 1.9957568809385693e-05, | |
| "loss": 1.1324, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.1272, | |
| "grad_norm": 0.10389735549688339, | |
| "learning_rate": 1.995496021999177e-05, | |
| "loss": 0.0405, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 0.04948841407895088, | |
| "learning_rate": 1.9952273999818312e-05, | |
| "loss": 0.0372, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.1288, | |
| "grad_norm": 0.09340821206569672, | |
| "learning_rate": 1.9949510169813006e-05, | |
| "loss": 0.5878, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.1296, | |
| "grad_norm": 1.2787668704986572, | |
| "learning_rate": 1.9946668751528745e-05, | |
| "loss": 0.3489, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.1304, | |
| "grad_norm": 1.642128586769104, | |
| "learning_rate": 1.994374976712348e-05, | |
| "loss": 0.1129, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.1312, | |
| "grad_norm": 0.1789906769990921, | |
| "learning_rate": 1.9940753239360047e-05, | |
| "loss": 0.0982, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.132, | |
| "grad_norm": 1.3861182928085327, | |
| "learning_rate": 1.9937679191605964e-05, | |
| "loss": 0.2423, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.1328, | |
| "grad_norm": 0.9052829146385193, | |
| "learning_rate": 1.9934527647833276e-05, | |
| "loss": 0.1399, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.1336, | |
| "grad_norm": 1.6789326667785645, | |
| "learning_rate": 1.9931298632618355e-05, | |
| "loss": 0.2957, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.1344, | |
| "grad_norm": 1.9615066051483154, | |
| "learning_rate": 1.9927992171141707e-05, | |
| "loss": 1.2319, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.1352, | |
| "grad_norm": 0.2719954550266266, | |
| "learning_rate": 1.9924608289187786e-05, | |
| "loss": 0.3792, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.136, | |
| "grad_norm": 0.131216362118721, | |
| "learning_rate": 1.9921147013144782e-05, | |
| "loss": 0.0956, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.1368, | |
| "grad_norm": 0.2020615190267563, | |
| "learning_rate": 1.9917608370004417e-05, | |
| "loss": 0.0629, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.1376, | |
| "grad_norm": 0.49078741669654846, | |
| "learning_rate": 1.9913992387361747e-05, | |
| "loss": 0.096, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.1384, | |
| "grad_norm": 0.09738267213106155, | |
| "learning_rate": 1.991029909341493e-05, | |
| "loss": 0.3201, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.1392, | |
| "grad_norm": 3.8677468299865723, | |
| "learning_rate": 1.990652851696501e-05, | |
| "loss": 0.1698, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.10318366438150406, | |
| "learning_rate": 1.9902680687415704e-05, | |
| "loss": 0.2468, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.1408, | |
| "grad_norm": 1.7756882905960083, | |
| "learning_rate": 1.989875563477316e-05, | |
| "loss": 0.3846, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.1416, | |
| "grad_norm": 8.583524703979492, | |
| "learning_rate": 1.9894753389645723e-05, | |
| "loss": 0.565, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.1424, | |
| "grad_norm": 0.26608502864837646, | |
| "learning_rate": 1.9890673983243708e-05, | |
| "loss": 0.059, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.1432, | |
| "grad_norm": 1.6263763904571533, | |
| "learning_rate": 1.988651744737914e-05, | |
| "loss": 0.9423, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.144, | |
| "grad_norm": 1.3715531826019287, | |
| "learning_rate": 1.988228381446553e-05, | |
| "loss": 0.2875, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.1448, | |
| "grad_norm": 1.6036182641983032, | |
| "learning_rate": 1.987797311751759e-05, | |
| "loss": 0.6963, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.1456, | |
| "grad_norm": 0.44622594118118286, | |
| "learning_rate": 1.9873585390151003e-05, | |
| "loss": 0.1357, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.1464, | |
| "grad_norm": 0.3719693720340729, | |
| "learning_rate": 1.9869120666582153e-05, | |
| "loss": 0.0775, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.1472, | |
| "grad_norm": 1.655290961265564, | |
| "learning_rate": 1.9864578981627844e-05, | |
| "loss": 0.316, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.148, | |
| "grad_norm": 0.5241470336914062, | |
| "learning_rate": 1.985996037070505e-05, | |
| "loss": 0.1641, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.1488, | |
| "grad_norm": 0.0686238631606102, | |
| "learning_rate": 1.985526486983063e-05, | |
| "loss": 0.3246, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.1496, | |
| "grad_norm": 0.04877380654215813, | |
| "learning_rate": 1.9850492515621038e-05, | |
| "loss": 0.0454, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.1504, | |
| "grad_norm": 0.033756960183382034, | |
| "learning_rate": 1.9845643345292055e-05, | |
| "loss": 0.0306, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.1512, | |
| "grad_norm": 0.050034862011671066, | |
| "learning_rate": 1.9840717396658483e-05, | |
| "loss": 0.0198, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.152, | |
| "grad_norm": 1.1677488088607788, | |
| "learning_rate": 1.983571470813386e-05, | |
| "loss": 0.9779, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.1528, | |
| "grad_norm": 2.3200299739837646, | |
| "learning_rate": 1.9830635318730155e-05, | |
| "loss": 0.2013, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 0.37079155445098877, | |
| "learning_rate": 1.982547926805747e-05, | |
| "loss": 0.0332, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.1544, | |
| "grad_norm": 0.03381960466504097, | |
| "learning_rate": 1.982024659632372e-05, | |
| "loss": 0.0156, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.1552, | |
| "grad_norm": 1.5234084129333496, | |
| "learning_rate": 1.981493734433433e-05, | |
| "loss": 0.0724, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.156, | |
| "grad_norm": 1.1328762769699097, | |
| "learning_rate": 1.9809551553491918e-05, | |
| "loss": 0.955, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.1568, | |
| "grad_norm": 4.6101975440979, | |
| "learning_rate": 1.980408926579596e-05, | |
| "loss": 1.726, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.1576, | |
| "grad_norm": 0.9627692103385925, | |
| "learning_rate": 1.979855052384247e-05, | |
| "loss": 0.4459, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.1584, | |
| "grad_norm": 0.19565686583518982, | |
| "learning_rate": 1.9792935370823676e-05, | |
| "loss": 0.1905, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.1592, | |
| "grad_norm": 1.6848326921463013, | |
| "learning_rate": 1.9787243850527663e-05, | |
| "loss": 0.7807, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 1.4652438163757324, | |
| "learning_rate": 1.9781476007338058e-05, | |
| "loss": 0.4939, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.1608, | |
| "grad_norm": 0.5617023706436157, | |
| "learning_rate": 1.9775631886233655e-05, | |
| "loss": 0.5734, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.1616, | |
| "grad_norm": 0.37801074981689453, | |
| "learning_rate": 1.9769711532788083e-05, | |
| "loss": 0.2663, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.1624, | |
| "grad_norm": 2.0784738063812256, | |
| "learning_rate": 1.976371499316945e-05, | |
| "loss": 0.7156, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.1632, | |
| "grad_norm": 0.24609781801700592, | |
| "learning_rate": 1.9757642314139977e-05, | |
| "loss": 0.1335, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.164, | |
| "grad_norm": 0.18215598165988922, | |
| "learning_rate": 1.9751493543055634e-05, | |
| "loss": 0.1607, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.1648, | |
| "grad_norm": 0.1584533452987671, | |
| "learning_rate": 1.9745268727865774e-05, | |
| "loss": 0.1947, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.1656, | |
| "grad_norm": 0.1543274223804474, | |
| "learning_rate": 1.9738967917112752e-05, | |
| "loss": 0.0954, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.1664, | |
| "grad_norm": 0.09421176463365555, | |
| "learning_rate": 1.9732591159931564e-05, | |
| "loss": 0.058, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.1672, | |
| "grad_norm": 0.037417419254779816, | |
| "learning_rate": 1.9726138506049438e-05, | |
| "loss": 0.0586, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.168, | |
| "grad_norm": 1.3091963529586792, | |
| "learning_rate": 1.9719610005785466e-05, | |
| "loss": 0.5727, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.1688, | |
| "grad_norm": 0.04180837422609329, | |
| "learning_rate": 1.9713005710050203e-05, | |
| "loss": 0.0303, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.1696, | |
| "grad_norm": 0.11628637462854385, | |
| "learning_rate": 1.9706325670345276e-05, | |
| "loss": 0.0236, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.1704, | |
| "grad_norm": 0.028453992679715157, | |
| "learning_rate": 1.9699569938762975e-05, | |
| "loss": 0.015, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.1712, | |
| "grad_norm": 1.2765767574310303, | |
| "learning_rate": 1.9692738567985853e-05, | |
| "loss": 0.6437, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.172, | |
| "grad_norm": 0.6149804592132568, | |
| "learning_rate": 1.9685831611286312e-05, | |
| "loss": 0.0557, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.1728, | |
| "grad_norm": 0.0872359424829483, | |
| "learning_rate": 1.967884912252619e-05, | |
| "loss": 0.1832, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.1736, | |
| "grad_norm": 0.031350962817668915, | |
| "learning_rate": 1.967179115615633e-05, | |
| "loss": 0.0422, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.1744, | |
| "grad_norm": 0.1513989269733429, | |
| "learning_rate": 1.9664657767216176e-05, | |
| "loss": 0.033, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.1752, | |
| "grad_norm": 0.05188186839222908, | |
| "learning_rate": 1.9657449011333328e-05, | |
| "loss": 0.0161, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.176, | |
| "grad_norm": 0.20884352922439575, | |
| "learning_rate": 1.9650164944723116e-05, | |
| "loss": 0.3403, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.1768, | |
| "grad_norm": 1.4495282173156738, | |
| "learning_rate": 1.964280562418815e-05, | |
| "loss": 0.1676, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.1776, | |
| "grad_norm": 0.01325430627912283, | |
| "learning_rate": 1.963537110711789e-05, | |
| "loss": 0.0142, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.1784, | |
| "grad_norm": 0.03731897845864296, | |
| "learning_rate": 1.962786145148819e-05, | |
| "loss": 0.013, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.1792, | |
| "grad_norm": 0.06609856337308884, | |
| "learning_rate": 1.962027671586086e-05, | |
| "loss": 0.0124, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.02560417540371418, | |
| "learning_rate": 1.961261695938319e-05, | |
| "loss": 0.0291, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.1808, | |
| "grad_norm": 0.016771750524640083, | |
| "learning_rate": 1.96048822417875e-05, | |
| "loss": 0.0089, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.1816, | |
| "grad_norm": 7.128602504730225, | |
| "learning_rate": 1.9597072623390668e-05, | |
| "loss": 1.0882, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.1824, | |
| "grad_norm": 0.03642188385128975, | |
| "learning_rate": 1.958918816509367e-05, | |
| "loss": 0.0113, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.1832, | |
| "grad_norm": 0.09811338037252426, | |
| "learning_rate": 1.95812289283811e-05, | |
| "loss": 0.016, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.184, | |
| "grad_norm": 0.04652848094701767, | |
| "learning_rate": 1.9573194975320672e-05, | |
| "loss": 0.1104, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.1848, | |
| "grad_norm": 1.6137455701828003, | |
| "learning_rate": 1.956508636856278e-05, | |
| "loss": 0.872, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.1856, | |
| "grad_norm": 1.2847460508346558, | |
| "learning_rate": 1.9556903171339963e-05, | |
| "loss": 0.9774, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.1864, | |
| "grad_norm": 0.3364383280277252, | |
| "learning_rate": 1.9548645447466433e-05, | |
| "loss": 0.0612, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.1872, | |
| "grad_norm": 1.6685212850570679, | |
| "learning_rate": 1.954031326133758e-05, | |
| "loss": 0.4306, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.188, | |
| "grad_norm": 0.12480071187019348, | |
| "learning_rate": 1.9531906677929472e-05, | |
| "loss": 0.1104, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.1888, | |
| "grad_norm": 0.5850446224212646, | |
| "learning_rate": 1.9523425762798328e-05, | |
| "loss": 0.1902, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.1896, | |
| "grad_norm": 0.14481854438781738, | |
| "learning_rate": 1.951487058208003e-05, | |
| "loss": 0.1135, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.1904, | |
| "grad_norm": 1.4755823612213135, | |
| "learning_rate": 1.95062412024896e-05, | |
| "loss": 0.9318, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.1912, | |
| "grad_norm": 0.31880441308021545, | |
| "learning_rate": 1.949753769132067e-05, | |
| "loss": 0.0736, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 0.3856407403945923, | |
| "learning_rate": 1.9488760116444966e-05, | |
| "loss": 0.4857, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.1928, | |
| "grad_norm": 1.6081260442733765, | |
| "learning_rate": 1.9479908546311783e-05, | |
| "loss": 0.2605, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.1936, | |
| "grad_norm": 1.051400065422058, | |
| "learning_rate": 1.9470983049947446e-05, | |
| "loss": 0.1468, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.1944, | |
| "grad_norm": 0.19811314344406128, | |
| "learning_rate": 1.946198369695476e-05, | |
| "loss": 0.5758, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.1952, | |
| "grad_norm": 0.7101706862449646, | |
| "learning_rate": 1.9452910557512497e-05, | |
| "loss": 0.765, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.196, | |
| "grad_norm": 1.4943866729736328, | |
| "learning_rate": 1.944376370237481e-05, | |
| "loss": 0.4381, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.1968, | |
| "grad_norm": 0.22989702224731445, | |
| "learning_rate": 1.9434543202870726e-05, | |
| "loss": 0.0615, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.1976, | |
| "grad_norm": 1.0281144380569458, | |
| "learning_rate": 1.9425249130903544e-05, | |
| "loss": 0.2893, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.1984, | |
| "grad_norm": 0.912899911403656, | |
| "learning_rate": 1.9415881558950302e-05, | |
| "loss": 0.1344, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.1992, | |
| "grad_norm": 0.4319363534450531, | |
| "learning_rate": 1.9406440560061214e-05, | |
| "loss": 0.162, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 1.5313297510147095, | |
| "learning_rate": 1.9396926207859085e-05, | |
| "loss": 0.9079, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.2008, | |
| "grad_norm": 0.1665397733449936, | |
| "learning_rate": 1.9387338576538743e-05, | |
| "loss": 0.0523, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 0.2016, | |
| "grad_norm": 0.703409731388092, | |
| "learning_rate": 1.937767774086646e-05, | |
| "loss": 0.4771, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 0.2024, | |
| "grad_norm": 0.13862833380699158, | |
| "learning_rate": 1.936794377617938e-05, | |
| "loss": 0.0484, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 0.2032, | |
| "grad_norm": 0.12500865757465363, | |
| "learning_rate": 1.935813675838491e-05, | |
| "loss": 0.033, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 0.204, | |
| "grad_norm": 0.12923550605773926, | |
| "learning_rate": 1.9348256763960146e-05, | |
| "loss": 0.0503, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.2048, | |
| "grad_norm": 0.12034843116998672, | |
| "learning_rate": 1.933830386995127e-05, | |
| "loss": 0.2762, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 0.2056, | |
| "grad_norm": 1.1866302490234375, | |
| "learning_rate": 1.9328278153972947e-05, | |
| "loss": 0.0998, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 0.2064, | |
| "grad_norm": 0.048314038664102554, | |
| "learning_rate": 1.9318179694207726e-05, | |
| "loss": 0.0296, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 0.2072, | |
| "grad_norm": 0.2854752242565155, | |
| "learning_rate": 1.9308008569405424e-05, | |
| "loss": 0.0852, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 0.208, | |
| "grad_norm": 0.16209827363491058, | |
| "learning_rate": 1.9297764858882516e-05, | |
| "loss": 0.8268, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.2088, | |
| "grad_norm": 0.01691954769194126, | |
| "learning_rate": 1.9287448642521513e-05, | |
| "loss": 0.0093, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 0.2096, | |
| "grad_norm": 0.09849057346582413, | |
| "learning_rate": 1.9277060000770342e-05, | |
| "loss": 0.0657, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 0.2104, | |
| "grad_norm": 0.05499891936779022, | |
| "learning_rate": 1.9266599014641724e-05, | |
| "loss": 0.0259, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 0.2112, | |
| "grad_norm": 0.5297953486442566, | |
| "learning_rate": 1.9256065765712524e-05, | |
| "loss": 0.0539, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 0.212, | |
| "grad_norm": 0.053190361708402634, | |
| "learning_rate": 1.9245460336123136e-05, | |
| "loss": 0.035, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.2128, | |
| "grad_norm": 0.7372400760650635, | |
| "learning_rate": 1.9234782808576823e-05, | |
| "loss": 0.186, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 0.2136, | |
| "grad_norm": 0.02464686520397663, | |
| "learning_rate": 1.9224033266339103e-05, | |
| "loss": 0.0097, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 0.2144, | |
| "grad_norm": 3.046816825866699, | |
| "learning_rate": 1.9213211793237056e-05, | |
| "loss": 0.1614, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 0.2152, | |
| "grad_norm": 2.2198116779327393, | |
| "learning_rate": 1.9202318473658707e-05, | |
| "loss": 0.6094, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 0.216, | |
| "grad_norm": 1.2823199033737183, | |
| "learning_rate": 1.9191353392552346e-05, | |
| "loss": 0.1261, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.2168, | |
| "grad_norm": 0.019597092643380165, | |
| "learning_rate": 1.9180316635425883e-05, | |
| "loss": 0.4845, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 0.2176, | |
| "grad_norm": 0.011679578572511673, | |
| "learning_rate": 1.9169208288346168e-05, | |
| "loss": 0.0162, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 0.2184, | |
| "grad_norm": 1.827360987663269, | |
| "learning_rate": 1.9158028437938316e-05, | |
| "loss": 0.8432, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 0.2192, | |
| "grad_norm": 1.7069369554519653, | |
| "learning_rate": 1.914677717138505e-05, | |
| "loss": 0.321, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.022664044052362442, | |
| "learning_rate": 1.913545457642601e-05, | |
| "loss": 0.2938, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.2208, | |
| "grad_norm": 0.5914615392684937, | |
| "learning_rate": 1.9124060741357065e-05, | |
| "loss": 0.0618, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 0.2216, | |
| "grad_norm": 0.0440104641020298, | |
| "learning_rate": 1.9112595755029625e-05, | |
| "loss": 0.0429, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 0.2224, | |
| "grad_norm": 0.263875812292099, | |
| "learning_rate": 1.9101059706849957e-05, | |
| "loss": 1.002, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 0.2232, | |
| "grad_norm": 1.789616584777832, | |
| "learning_rate": 1.908945268677849e-05, | |
| "loss": 0.3908, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 0.224, | |
| "grad_norm": 3.8418707847595215, | |
| "learning_rate": 1.907777478532909e-05, | |
| "loss": 0.4097, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.2248, | |
| "grad_norm": 0.5954731702804565, | |
| "learning_rate": 1.906602609356838e-05, | |
| "loss": 0.1258, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 0.2256, | |
| "grad_norm": 3.1581459045410156, | |
| "learning_rate": 1.905420670311502e-05, | |
| "loss": 0.3185, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 0.2264, | |
| "grad_norm": 0.37356865406036377, | |
| "learning_rate": 1.9042316706138987e-05, | |
| "loss": 0.0718, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 0.2272, | |
| "grad_norm": 0.6604268550872803, | |
| "learning_rate": 1.9030356195360875e-05, | |
| "loss": 0.1169, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 0.228, | |
| "grad_norm": 1.148862361907959, | |
| "learning_rate": 1.901832526405114e-05, | |
| "loss": 0.1075, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.2288, | |
| "grad_norm": 4.533411026000977, | |
| "learning_rate": 1.9006224006029404e-05, | |
| "loss": 0.8169, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 0.2296, | |
| "grad_norm": 0.1384696662425995, | |
| "learning_rate": 1.899405251566371e-05, | |
| "loss": 0.0574, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 0.2304, | |
| "grad_norm": 4.587916851043701, | |
| "learning_rate": 1.8981810887869784e-05, | |
| "loss": 0.3013, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 0.2312, | |
| "grad_norm": 0.07302756607532501, | |
| "learning_rate": 1.8969499218110302e-05, | |
| "loss": 0.0345, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 0.232, | |
| "grad_norm": 0.9006990194320679, | |
| "learning_rate": 1.895711760239413e-05, | |
| "loss": 0.4378, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.2328, | |
| "grad_norm": 0.07447630912065506, | |
| "learning_rate": 1.89446661372756e-05, | |
| "loss": 0.11, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 0.2336, | |
| "grad_norm": 0.10830947756767273, | |
| "learning_rate": 1.893214491985374e-05, | |
| "loss": 0.0505, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 0.2344, | |
| "grad_norm": 0.0925409272313118, | |
| "learning_rate": 1.8919554047771508e-05, | |
| "loss": 0.0449, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 0.2352, | |
| "grad_norm": 2.633633613586426, | |
| "learning_rate": 1.890689361921507e-05, | |
| "loss": 0.7599, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 0.236, | |
| "grad_norm": 0.013573708944022655, | |
| "learning_rate": 1.889416373291298e-05, | |
| "loss": 0.2034, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.2368, | |
| "grad_norm": 0.03778607025742531, | |
| "learning_rate": 1.8881364488135448e-05, | |
| "loss": 0.5998, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 0.2376, | |
| "grad_norm": 0.4637905955314636, | |
| "learning_rate": 1.886849598469356e-05, | |
| "loss": 0.0662, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 0.2384, | |
| "grad_norm": 0.058186739683151245, | |
| "learning_rate": 1.8855558322938492e-05, | |
| "loss": 0.2836, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 0.2392, | |
| "grad_norm": 0.09846732765436172, | |
| "learning_rate": 1.8842551603760725e-05, | |
| "loss": 0.1087, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.21428197622299194, | |
| "learning_rate": 1.8829475928589272e-05, | |
| "loss": 0.229, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.2408, | |
| "grad_norm": 0.3262503743171692, | |
| "learning_rate": 1.881633139939087e-05, | |
| "loss": 0.104, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 0.2416, | |
| "grad_norm": 0.9335662126541138, | |
| "learning_rate": 1.8803118118669203e-05, | |
| "loss": 0.5139, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 0.2424, | |
| "grad_norm": 0.027328329160809517, | |
| "learning_rate": 1.878983618946409e-05, | |
| "loss": 0.0222, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 0.2432, | |
| "grad_norm": 0.048500385135412216, | |
| "learning_rate": 1.8776485715350672e-05, | |
| "loss": 0.0422, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 0.244, | |
| "grad_norm": 0.3864876925945282, | |
| "learning_rate": 1.8763066800438638e-05, | |
| "loss": 0.1505, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.2448, | |
| "grad_norm": 1.1042886972427368, | |
| "learning_rate": 1.874957954937138e-05, | |
| "loss": 0.6224, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 0.2456, | |
| "grad_norm": 0.10346169769763947, | |
| "learning_rate": 1.8736024067325188e-05, | |
| "loss": 0.0375, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 0.2464, | |
| "grad_norm": 0.01699395291507244, | |
| "learning_rate": 1.8722400460008437e-05, | |
| "loss": 0.0352, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 0.2472, | |
| "grad_norm": 1.2863647937774658, | |
| "learning_rate": 1.8708708833660755e-05, | |
| "loss": 0.5494, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 0.248, | |
| "grad_norm": 2.1446785926818848, | |
| "learning_rate": 1.869494929505219e-05, | |
| "loss": 0.3612, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.2488, | |
| "grad_norm": 0.16956354677677155, | |
| "learning_rate": 1.8681121951482397e-05, | |
| "loss": 0.0355, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 0.2496, | |
| "grad_norm": 1.0704134702682495, | |
| "learning_rate": 1.8667226910779767e-05, | |
| "loss": 0.3591, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 0.2504, | |
| "grad_norm": 0.17657175660133362, | |
| "learning_rate": 1.8653264281300622e-05, | |
| "loss": 0.0436, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 0.2512, | |
| "grad_norm": 0.16666510701179504, | |
| "learning_rate": 1.8639234171928355e-05, | |
| "loss": 0.0334, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 0.252, | |
| "grad_norm": 0.14208954572677612, | |
| "learning_rate": 1.8625136692072577e-05, | |
| "loss": 0.2292, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.2528, | |
| "grad_norm": 1.2962373495101929, | |
| "learning_rate": 1.8610971951668265e-05, | |
| "loss": 0.2592, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 0.2536, | |
| "grad_norm": 0.09842484444379807, | |
| "learning_rate": 1.8596740061174912e-05, | |
| "loss": 0.0721, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 0.2544, | |
| "grad_norm": 0.07477198541164398, | |
| "learning_rate": 1.8582441131575658e-05, | |
| "loss": 0.0468, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 0.2552, | |
| "grad_norm": 1.2548961639404297, | |
| "learning_rate": 1.856807527437643e-05, | |
| "loss": 0.3335, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 0.20232835412025452, | |
| "learning_rate": 1.855364260160507e-05, | |
| "loss": 0.4149, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.2568, | |
| "grad_norm": 0.03189453110098839, | |
| "learning_rate": 1.8539143225810453e-05, | |
| "loss": 0.174, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 0.2576, | |
| "grad_norm": 0.026663975790143013, | |
| "learning_rate": 1.8524577260061628e-05, | |
| "loss": 0.0207, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 0.2584, | |
| "grad_norm": 0.03870954364538193, | |
| "learning_rate": 1.850994481794692e-05, | |
| "loss": 0.0197, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 0.2592, | |
| "grad_norm": 0.05131769925355911, | |
| "learning_rate": 1.8495246013573057e-05, | |
| "loss": 0.0187, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 0.278886079788208, | |
| "learning_rate": 1.848048096156426e-05, | |
| "loss": 0.1129, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.2608, | |
| "grad_norm": 0.030908726155757904, | |
| "learning_rate": 1.8465649777061377e-05, | |
| "loss": 0.0288, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 0.2616, | |
| "grad_norm": 1.7882804870605469, | |
| "learning_rate": 1.8450752575720967e-05, | |
| "loss": 0.3299, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 0.2624, | |
| "grad_norm": 0.9408004879951477, | |
| "learning_rate": 1.843578947371439e-05, | |
| "loss": 0.1842, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 0.2632, | |
| "grad_norm": 0.03317919746041298, | |
| "learning_rate": 1.8420760587726925e-05, | |
| "loss": 0.0249, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 0.264, | |
| "grad_norm": 0.0062853083945810795, | |
| "learning_rate": 1.8405666034956842e-05, | |
| "loss": 0.0765, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.2648, | |
| "grad_norm": 1.696451187133789, | |
| "learning_rate": 1.8390505933114503e-05, | |
| "loss": 0.4928, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 0.2656, | |
| "grad_norm": 0.11740172654390335, | |
| "learning_rate": 1.837528040042142e-05, | |
| "loss": 0.0175, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 0.2664, | |
| "grad_norm": 0.437165230512619, | |
| "learning_rate": 1.8359989555609355e-05, | |
| "loss": 0.0775, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 0.2672, | |
| "grad_norm": 0.01961176097393036, | |
| "learning_rate": 1.834463351791939e-05, | |
| "loss": 0.012, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 0.268, | |
| "grad_norm": 2.6758432388305664, | |
| "learning_rate": 1.8329212407100996e-05, | |
| "loss": 0.1287, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.2688, | |
| "grad_norm": 0.007790145929902792, | |
| "learning_rate": 1.8313726343411085e-05, | |
| "loss": 0.0071, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 0.2696, | |
| "grad_norm": 0.01610707677900791, | |
| "learning_rate": 1.82981754476131e-05, | |
| "loss": 0.004, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 0.2704, | |
| "grad_norm": 0.10419867932796478, | |
| "learning_rate": 1.8282559840976043e-05, | |
| "loss": 0.022, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 0.2712, | |
| "grad_norm": 0.020517872646450996, | |
| "learning_rate": 1.8266879645273557e-05, | |
| "loss": 0.0028, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 0.272, | |
| "grad_norm": 0.011518670246005058, | |
| "learning_rate": 1.8251134982782952e-05, | |
| "loss": 0.0121, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.2728, | |
| "grad_norm": 3.1481823921203613, | |
| "learning_rate": 1.8235325976284276e-05, | |
| "loss": 0.9131, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 0.2736, | |
| "grad_norm": 0.021789977326989174, | |
| "learning_rate": 1.8219452749059332e-05, | |
| "loss": 0.0169, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 0.2744, | |
| "grad_norm": 2.266319990158081, | |
| "learning_rate": 1.8203515424890738e-05, | |
| "loss": 0.4778, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 0.2752, | |
| "grad_norm": 2.55071759223938, | |
| "learning_rate": 1.8187514128060946e-05, | |
| "loss": 0.4595, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 0.276, | |
| "grad_norm": 1.0234826803207397, | |
| "learning_rate": 1.8171448983351284e-05, | |
| "loss": 0.6759, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.2768, | |
| "grad_norm": 0.5817314982414246, | |
| "learning_rate": 1.8155320116040983e-05, | |
| "loss": 0.2023, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 0.2776, | |
| "grad_norm": 0.07609284669160843, | |
| "learning_rate": 1.8139127651906183e-05, | |
| "loss": 0.2312, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 0.2784, | |
| "grad_norm": 0.1080310121178627, | |
| "learning_rate": 1.812287171721897e-05, | |
| "loss": 0.1537, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 0.2792, | |
| "grad_norm": 0.1361207813024521, | |
| "learning_rate": 1.81065524387464e-05, | |
| "loss": 0.0511, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.11058028042316437, | |
| "learning_rate": 1.8090169943749477e-05, | |
| "loss": 0.0885, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.2808, | |
| "grad_norm": 0.6568836569786072, | |
| "learning_rate": 1.8073724359982184e-05, | |
| "loss": 0.3678, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 0.2816, | |
| "grad_norm": 0.250731885433197, | |
| "learning_rate": 1.8057215815690494e-05, | |
| "loss": 0.5069, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 0.2824, | |
| "grad_norm": 0.0678286999464035, | |
| "learning_rate": 1.8040644439611348e-05, | |
| "loss": 0.0418, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 0.2832, | |
| "grad_norm": 1.9648799896240234, | |
| "learning_rate": 1.802401036097167e-05, | |
| "loss": 0.4493, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 0.284, | |
| "grad_norm": 0.21381056308746338, | |
| "learning_rate": 1.8007313709487334e-05, | |
| "loss": 0.1236, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.2848, | |
| "grad_norm": 0.5156528949737549, | |
| "learning_rate": 1.79905546153622e-05, | |
| "loss": 0.1819, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 0.2856, | |
| "grad_norm": 0.09592005610466003, | |
| "learning_rate": 1.7973733209287036e-05, | |
| "loss": 0.1251, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 0.2864, | |
| "grad_norm": 0.42804062366485596, | |
| "learning_rate": 1.7956849622438554e-05, | |
| "loss": 0.1349, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 0.2872, | |
| "grad_norm": 0.13698697090148926, | |
| "learning_rate": 1.7939903986478354e-05, | |
| "loss": 0.0416, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 0.288, | |
| "grad_norm": 0.01778518594801426, | |
| "learning_rate": 1.792289643355191e-05, | |
| "loss": 0.0129, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.2888, | |
| "grad_norm": 0.2408394068479538, | |
| "learning_rate": 1.7905827096287532e-05, | |
| "loss": 0.0963, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 0.2896, | |
| "grad_norm": 0.025463353842496872, | |
| "learning_rate": 1.7888696107795343e-05, | |
| "loss": 0.6419, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 0.2904, | |
| "grad_norm": 1.5628716945648193, | |
| "learning_rate": 1.7871503601666233e-05, | |
| "loss": 0.4731, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 0.2912, | |
| "grad_norm": 0.07346764206886292, | |
| "learning_rate": 1.785424971197082e-05, | |
| "loss": 0.0206, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 0.292, | |
| "grad_norm": 5.0355963706970215, | |
| "learning_rate": 1.78369345732584e-05, | |
| "loss": 0.2513, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.2928, | |
| "grad_norm": 1.0326207876205444, | |
| "learning_rate": 1.7819558320555902e-05, | |
| "loss": 1.2558, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 0.2936, | |
| "grad_norm": 0.24173852801322937, | |
| "learning_rate": 1.780212108936684e-05, | |
| "loss": 0.0767, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 0.2944, | |
| "grad_norm": 0.20275752246379852, | |
| "learning_rate": 1.7784623015670237e-05, | |
| "loss": 0.059, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 0.2952, | |
| "grad_norm": 0.1756519377231598, | |
| "learning_rate": 1.7767064235919594e-05, | |
| "loss": 0.1225, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 0.296, | |
| "grad_norm": 0.09066800028085709, | |
| "learning_rate": 1.7749444887041797e-05, | |
| "loss": 0.2309, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.2968, | |
| "grad_norm": 2.341961622238159, | |
| "learning_rate": 1.7731765106436073e-05, | |
| "loss": 0.5643, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 0.2976, | |
| "grad_norm": 0.09298836439847946, | |
| "learning_rate": 1.7714025031972904e-05, | |
| "loss": 0.272, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 0.2984, | |
| "grad_norm": 0.15143701434135437, | |
| "learning_rate": 1.7696224801992947e-05, | |
| "loss": 0.0548, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 0.2992, | |
| "grad_norm": 1.298582911491394, | |
| "learning_rate": 1.767836455530598e-05, | |
| "loss": 0.2634, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.12245524674654007, | |
| "learning_rate": 1.766044443118978e-05, | |
| "loss": 0.113, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.3008, | |
| "grad_norm": 0.08619975298643112, | |
| "learning_rate": 1.764246456938909e-05, | |
| "loss": 0.0373, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 0.3016, | |
| "grad_norm": 1.6767048835754395, | |
| "learning_rate": 1.762442511011448e-05, | |
| "loss": 0.3271, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 0.3024, | |
| "grad_norm": 0.0463847815990448, | |
| "learning_rate": 1.7606326194041274e-05, | |
| "loss": 0.3972, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 0.3032, | |
| "grad_norm": 0.018276596441864967, | |
| "learning_rate": 1.7588167962308458e-05, | |
| "loss": 0.0141, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 0.304, | |
| "grad_norm": 0.1315947324037552, | |
| "learning_rate": 1.7569950556517566e-05, | |
| "loss": 0.0556, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.3048, | |
| "grad_norm": 0.5924662947654724, | |
| "learning_rate": 1.7551674118731592e-05, | |
| "loss": 0.1275, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 0.3056, | |
| "grad_norm": 0.23044385015964508, | |
| "learning_rate": 1.7533338791473872e-05, | |
| "loss": 0.0416, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 0.3064, | |
| "grad_norm": 1.4832113981246948, | |
| "learning_rate": 1.7514944717726962e-05, | |
| "loss": 0.4879, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 0.3072, | |
| "grad_norm": 0.08857022970914841, | |
| "learning_rate": 1.749649204093155e-05, | |
| "loss": 0.042, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 0.308, | |
| "grad_norm": 0.24603064358234406, | |
| "learning_rate": 1.747798090498532e-05, | |
| "loss": 0.2623, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.3088, | |
| "grad_norm": 0.08749913424253464, | |
| "learning_rate": 1.7459411454241822e-05, | |
| "loss": 0.0188, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 0.3096, | |
| "grad_norm": 0.10140682011842728, | |
| "learning_rate": 1.7440783833509366e-05, | |
| "loss": 0.0401, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 0.3104, | |
| "grad_norm": 0.027609799057245255, | |
| "learning_rate": 1.7422098188049885e-05, | |
| "loss": 0.0173, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 0.3112, | |
| "grad_norm": 1.5163832902908325, | |
| "learning_rate": 1.7403354663577782e-05, | |
| "loss": 0.654, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 0.312, | |
| "grad_norm": 0.4432819187641144, | |
| "learning_rate": 1.7384553406258842e-05, | |
| "loss": 0.0778, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.3128, | |
| "grad_norm": 0.006748716812580824, | |
| "learning_rate": 1.7365694562709034e-05, | |
| "loss": 0.0119, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 0.3136, | |
| "grad_norm": 0.023698939010500908, | |
| "learning_rate": 1.7346778279993417e-05, | |
| "loss": 0.013, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 0.3144, | |
| "grad_norm": 0.01780467852950096, | |
| "learning_rate": 1.732780470562496e-05, | |
| "loss": 0.015, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 0.3152, | |
| "grad_norm": 1.4911342859268188, | |
| "learning_rate": 1.7308773987563406e-05, | |
| "loss": 0.1485, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 0.316, | |
| "grad_norm": 0.030542919412255287, | |
| "learning_rate": 1.7289686274214116e-05, | |
| "loss": 0.0136, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.3168, | |
| "grad_norm": 0.039021674543619156, | |
| "learning_rate": 1.727054171442692e-05, | |
| "loss": 0.0715, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 0.3176, | |
| "grad_norm": 0.24402689933776855, | |
| "learning_rate": 1.7251340457494934e-05, | |
| "loss": 0.5099, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 0.3184, | |
| "grad_norm": 2.137239456176758, | |
| "learning_rate": 1.7232082653153422e-05, | |
| "loss": 0.5133, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 0.3192, | |
| "grad_norm": 2.8100578784942627, | |
| "learning_rate": 1.721276845157861e-05, | |
| "loss": 0.428, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.1050364151597023, | |
| "learning_rate": 1.7193398003386514e-05, | |
| "loss": 0.0221, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.3208, | |
| "grad_norm": 0.02302808314561844, | |
| "learning_rate": 1.717397145963179e-05, | |
| "loss": 0.0524, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 0.3216, | |
| "grad_norm": 0.5185064077377319, | |
| "learning_rate": 1.715448897180652e-05, | |
| "loss": 0.0452, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 0.3224, | |
| "grad_norm": 0.49271076917648315, | |
| "learning_rate": 1.7134950691839063e-05, | |
| "loss": 0.1355, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 0.3232, | |
| "grad_norm": 0.2875131666660309, | |
| "learning_rate": 1.7115356772092858e-05, | |
| "loss": 0.5409, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 0.324, | |
| "grad_norm": 2.0039379596710205, | |
| "learning_rate": 1.709570736536521e-05, | |
| "loss": 0.2663, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.3248, | |
| "grad_norm": 1.235532522201538, | |
| "learning_rate": 1.7076002624886156e-05, | |
| "loss": 0.391, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 0.3256, | |
| "grad_norm": 0.9183419942855835, | |
| "learning_rate": 1.705624270431721e-05, | |
| "loss": 0.1418, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 0.3264, | |
| "grad_norm": 0.5461841821670532, | |
| "learning_rate": 1.7036427757750205e-05, | |
| "loss": 0.1603, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 0.3272, | |
| "grad_norm": 0.373760849237442, | |
| "learning_rate": 1.7016557939706075e-05, | |
| "loss": 0.0709, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 0.328, | |
| "grad_norm": 0.40094730257987976, | |
| "learning_rate": 1.6996633405133656e-05, | |
| "loss": 0.0659, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.3288, | |
| "grad_norm": 0.05968625098466873, | |
| "learning_rate": 1.6976654309408464e-05, | |
| "loss": 0.0307, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 0.3296, | |
| "grad_norm": 0.22513960301876068, | |
| "learning_rate": 1.695662080833151e-05, | |
| "loss": 0.5364, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 0.3304, | |
| "grad_norm": 0.2083749920129776, | |
| "learning_rate": 1.693653305812805e-05, | |
| "loss": 0.0446, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 0.3312, | |
| "grad_norm": 0.00946067925542593, | |
| "learning_rate": 1.6916391215446403e-05, | |
| "loss": 0.0212, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 0.332, | |
| "grad_norm": 0.08354925364255905, | |
| "learning_rate": 1.68961954373567e-05, | |
| "loss": 0.0427, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.3328, | |
| "grad_norm": 0.04629332199692726, | |
| "learning_rate": 1.6875945881349676e-05, | |
| "loss": 0.0264, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 0.3336, | |
| "grad_norm": 0.05332889407873154, | |
| "learning_rate": 1.6855642705335438e-05, | |
| "loss": 0.0174, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 0.3344, | |
| "grad_norm": 0.04525836184620857, | |
| "learning_rate": 1.6835286067642228e-05, | |
| "loss": 0.0188, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 0.3352, | |
| "grad_norm": 1.4626930952072144, | |
| "learning_rate": 1.68148761270152e-05, | |
| "loss": 0.3748, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 0.336, | |
| "grad_norm": 0.023244062438607216, | |
| "learning_rate": 1.6794413042615168e-05, | |
| "loss": 0.1176, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.3368, | |
| "grad_norm": 0.06457129120826721, | |
| "learning_rate": 1.6773896974017373e-05, | |
| "loss": 0.5326, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 0.3376, | |
| "grad_norm": 0.0086558498442173, | |
| "learning_rate": 1.6753328081210244e-05, | |
| "loss": 0.1256, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 0.3384, | |
| "grad_norm": 0.005909595172852278, | |
| "learning_rate": 1.6732706524594138e-05, | |
| "loss": 0.0242, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 0.3392, | |
| "grad_norm": 0.0692615732550621, | |
| "learning_rate": 1.6712032464980094e-05, | |
| "loss": 0.0237, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 2.266417980194092, | |
| "learning_rate": 1.6691306063588583e-05, | |
| "loss": 0.3703, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.3408, | |
| "grad_norm": 2.5376994609832764, | |
| "learning_rate": 1.6670527482048246e-05, | |
| "loss": 0.3166, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 0.3416, | |
| "grad_norm": 0.02030259743332863, | |
| "learning_rate": 1.6649696882394635e-05, | |
| "loss": 0.0218, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 0.3424, | |
| "grad_norm": 0.09955969452857971, | |
| "learning_rate": 1.6628814427068954e-05, | |
| "loss": 0.0158, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 0.3432, | |
| "grad_norm": 1.0117660760879517, | |
| "learning_rate": 1.6607880278916778e-05, | |
| "loss": 0.4415, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 0.344, | |
| "grad_norm": 0.7470207214355469, | |
| "learning_rate": 1.6586894601186804e-05, | |
| "loss": 0.0473, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.3448, | |
| "grad_norm": 2.56488299369812, | |
| "learning_rate": 1.6565857557529567e-05, | |
| "loss": 0.5677, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 0.3456, | |
| "grad_norm": 0.06736627966165543, | |
| "learning_rate": 1.654476931199615e-05, | |
| "loss": 0.0131, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 0.3464, | |
| "grad_norm": 0.2571139931678772, | |
| "learning_rate": 1.652363002903693e-05, | |
| "loss": 0.1469, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 0.3472, | |
| "grad_norm": 0.21657270193099976, | |
| "learning_rate": 1.650243987350029e-05, | |
| "loss": 0.0489, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 0.348, | |
| "grad_norm": 0.02142491564154625, | |
| "learning_rate": 1.6481199010631312e-05, | |
| "loss": 0.0332, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.3488, | |
| "grad_norm": 0.1039031520485878, | |
| "learning_rate": 1.6459907606070513e-05, | |
| "loss": 0.2827, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 0.3496, | |
| "grad_norm": 0.07285499572753906, | |
| "learning_rate": 1.643856582585254e-05, | |
| "loss": 0.0334, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 0.3504, | |
| "grad_norm": 0.0627899169921875, | |
| "learning_rate": 1.6417173836404888e-05, | |
| "loss": 0.0265, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 0.3512, | |
| "grad_norm": 0.021432699635624886, | |
| "learning_rate": 1.6395731804546582e-05, | |
| "loss": 0.0292, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 0.352, | |
| "grad_norm": 0.07570278644561768, | |
| "learning_rate": 1.63742398974869e-05, | |
| "loss": 0.0153, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.3528, | |
| "grad_norm": 0.0861375704407692, | |
| "learning_rate": 1.6352698282824045e-05, | |
| "loss": 0.0148, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 0.3536, | |
| "grad_norm": 0.12202827632427216, | |
| "learning_rate": 1.6331107128543856e-05, | |
| "loss": 0.0365, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 0.3544, | |
| "grad_norm": 0.049717921763658524, | |
| "learning_rate": 1.6309466603018497e-05, | |
| "loss": 0.0202, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 0.3552, | |
| "grad_norm": 2.4151523113250732, | |
| "learning_rate": 1.628777687500513e-05, | |
| "loss": 0.4901, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 0.356, | |
| "grad_norm": 0.03276515007019043, | |
| "learning_rate": 1.6266038113644605e-05, | |
| "loss": 0.0243, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.3568, | |
| "grad_norm": 0.09225738793611526, | |
| "learning_rate": 1.624425048846016e-05, | |
| "loss": 0.0916, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 0.3576, | |
| "grad_norm": 0.06391174346208572, | |
| "learning_rate": 1.6222414169356066e-05, | |
| "loss": 0.0105, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 0.3584, | |
| "grad_norm": 0.03721031919121742, | |
| "learning_rate": 1.620052932661633e-05, | |
| "loss": 0.0123, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 0.3592, | |
| "grad_norm": 0.2955784201622009, | |
| "learning_rate": 1.6178596130903345e-05, | |
| "loss": 1.2386, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 0.05584976449608803, | |
| "learning_rate": 1.6156614753256583e-05, | |
| "loss": 0.0926, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.3608, | |
| "grad_norm": 1.0190123319625854, | |
| "learning_rate": 1.6134585365091243e-05, | |
| "loss": 0.1162, | |
| "step": 902 | |
| }, | |
| { | |
| "epoch": 0.3616, | |
| "grad_norm": 0.09299857914447784, | |
| "learning_rate": 1.611250813819692e-05, | |
| "loss": 0.0425, | |
| "step": 904 | |
| }, | |
| { | |
| "epoch": 0.3624, | |
| "grad_norm": 0.057083725929260254, | |
| "learning_rate": 1.6090383244736256e-05, | |
| "loss": 0.0206, | |
| "step": 906 | |
| }, | |
| { | |
| "epoch": 0.3632, | |
| "grad_norm": 0.006012015510350466, | |
| "learning_rate": 1.6068210857243625e-05, | |
| "loss": 0.0242, | |
| "step": 908 | |
| }, | |
| { | |
| "epoch": 0.364, | |
| "grad_norm": 0.04129006341099739, | |
| "learning_rate": 1.6045991148623752e-05, | |
| "loss": 0.0422, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.3648, | |
| "grad_norm": 0.049417100846767426, | |
| "learning_rate": 1.6023724292150387e-05, | |
| "loss": 0.0371, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 0.3656, | |
| "grad_norm": 0.13303126394748688, | |
| "learning_rate": 1.6001410461464955e-05, | |
| "loss": 0.2441, | |
| "step": 914 | |
| }, | |
| { | |
| "epoch": 0.3664, | |
| "grad_norm": 0.1420883685350418, | |
| "learning_rate": 1.597904983057519e-05, | |
| "loss": 0.0301, | |
| "step": 916 | |
| }, | |
| { | |
| "epoch": 0.3672, | |
| "grad_norm": 0.02255750633776188, | |
| "learning_rate": 1.5956642573853784e-05, | |
| "loss": 0.1678, | |
| "step": 918 | |
| }, | |
| { | |
| "epoch": 0.368, | |
| "grad_norm": 2.711224317550659, | |
| "learning_rate": 1.5934188866037017e-05, | |
| "loss": 0.2587, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.3688, | |
| "grad_norm": 0.015457335859537125, | |
| "learning_rate": 1.591168888222342e-05, | |
| "loss": 0.0161, | |
| "step": 922 | |
| }, | |
| { | |
| "epoch": 0.3696, | |
| "grad_norm": 0.021741507574915886, | |
| "learning_rate": 1.5889142797872387e-05, | |
| "loss": 0.0163, | |
| "step": 924 | |
| }, | |
| { | |
| "epoch": 0.3704, | |
| "grad_norm": 2.3807716369628906, | |
| "learning_rate": 1.5866550788802815e-05, | |
| "loss": 0.2696, | |
| "step": 926 | |
| }, | |
| { | |
| "epoch": 0.3712, | |
| "grad_norm": 1.5936973094940186, | |
| "learning_rate": 1.5843913031191722e-05, | |
| "loss": 0.4635, | |
| "step": 928 | |
| }, | |
| { | |
| "epoch": 0.372, | |
| "grad_norm": 0.18653613328933716, | |
| "learning_rate": 1.5821229701572897e-05, | |
| "loss": 0.0413, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.3728, | |
| "grad_norm": 0.048368774354457855, | |
| "learning_rate": 1.5798500976835493e-05, | |
| "loss": 0.0504, | |
| "step": 932 | |
| }, | |
| { | |
| "epoch": 0.3736, | |
| "grad_norm": 1.1611881256103516, | |
| "learning_rate": 1.5775727034222675e-05, | |
| "loss": 0.1021, | |
| "step": 934 | |
| }, | |
| { | |
| "epoch": 0.3744, | |
| "grad_norm": 0.03345046192407608, | |
| "learning_rate": 1.575290805133023e-05, | |
| "loss": 0.0152, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 0.3752, | |
| "grad_norm": 0.035641297698020935, | |
| "learning_rate": 1.5730044206105156e-05, | |
| "loss": 0.0111, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 0.376, | |
| "grad_norm": 0.1988631784915924, | |
| "learning_rate": 1.570713567684432e-05, | |
| "loss": 0.0267, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.3768, | |
| "grad_norm": 0.10902003198862076, | |
| "learning_rate": 1.568418264219303e-05, | |
| "loss": 0.063, | |
| "step": 942 | |
| }, | |
| { | |
| "epoch": 0.3776, | |
| "grad_norm": 0.25664544105529785, | |
| "learning_rate": 1.5661185281143666e-05, | |
| "loss": 0.6095, | |
| "step": 944 | |
| }, | |
| { | |
| "epoch": 0.3784, | |
| "grad_norm": 0.02122955210506916, | |
| "learning_rate": 1.5638143773034268e-05, | |
| "loss": 0.0116, | |
| "step": 946 | |
| }, | |
| { | |
| "epoch": 0.3792, | |
| "grad_norm": 0.05375152826309204, | |
| "learning_rate": 1.5615058297547144e-05, | |
| "loss": 0.2848, | |
| "step": 948 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.5861246585845947, | |
| "learning_rate": 1.5591929034707468e-05, | |
| "loss": 0.0492, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.3808, | |
| "grad_norm": 0.2694084048271179, | |
| "learning_rate": 1.556875616488188e-05, | |
| "loss": 0.0587, | |
| "step": 952 | |
| }, | |
| { | |
| "epoch": 0.3816, | |
| "grad_norm": 0.1376236230134964, | |
| "learning_rate": 1.5545539868777075e-05, | |
| "loss": 0.094, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 0.3824, | |
| "grad_norm": 0.03570554405450821, | |
| "learning_rate": 1.5522280327438388e-05, | |
| "loss": 0.0584, | |
| "step": 956 | |
| }, | |
| { | |
| "epoch": 0.3832, | |
| "grad_norm": 0.01179492473602295, | |
| "learning_rate": 1.54989777222484e-05, | |
| "loss": 0.4026, | |
| "step": 958 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 0.4097414016723633, | |
| "learning_rate": 1.5475632234925505e-05, | |
| "loss": 0.0395, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.3848, | |
| "grad_norm": 0.021319517865777016, | |
| "learning_rate": 1.5452244047522504e-05, | |
| "loss": 0.0063, | |
| "step": 962 | |
| }, | |
| { | |
| "epoch": 0.3856, | |
| "grad_norm": 0.008077614940702915, | |
| "learning_rate": 1.5428813342425177e-05, | |
| "loss": 0.0087, | |
| "step": 964 | |
| }, | |
| { | |
| "epoch": 0.3864, | |
| "grad_norm": 0.23517374694347382, | |
| "learning_rate": 1.540534030235087e-05, | |
| "loss": 0.0319, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 0.3872, | |
| "grad_norm": 0.5178306102752686, | |
| "learning_rate": 1.5381825110347072e-05, | |
| "loss": 0.077, | |
| "step": 968 | |
| }, | |
| { | |
| "epoch": 0.388, | |
| "grad_norm": 0.15628471970558167, | |
| "learning_rate": 1.5358267949789968e-05, | |
| "loss": 0.026, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.3888, | |
| "grad_norm": 0.32474571466445923, | |
| "learning_rate": 1.533466900438303e-05, | |
| "loss": 0.049, | |
| "step": 972 | |
| }, | |
| { | |
| "epoch": 0.3896, | |
| "grad_norm": 0.024446366354823112, | |
| "learning_rate": 1.5311028458155567e-05, | |
| "loss": 0.1731, | |
| "step": 974 | |
| }, | |
| { | |
| "epoch": 0.3904, | |
| "grad_norm": 0.1003560721874237, | |
| "learning_rate": 1.528734649546132e-05, | |
| "loss": 0.0142, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 0.3912, | |
| "grad_norm": 0.05430926755070686, | |
| "learning_rate": 1.526362330097698e-05, | |
| "loss": 0.1951, | |
| "step": 978 | |
| }, | |
| { | |
| "epoch": 0.392, | |
| "grad_norm": 0.0365879088640213, | |
| "learning_rate": 1.5239859059700794e-05, | |
| "loss": 0.0093, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.3928, | |
| "grad_norm": 0.006135095842182636, | |
| "learning_rate": 1.5216053956951081e-05, | |
| "loss": 0.0098, | |
| "step": 982 | |
| }, | |
| { | |
| "epoch": 0.3936, | |
| "grad_norm": 0.25733840465545654, | |
| "learning_rate": 1.5192208178364815e-05, | |
| "loss": 0.0337, | |
| "step": 984 | |
| }, | |
| { | |
| "epoch": 0.3944, | |
| "grad_norm": 1.6840555667877197, | |
| "learning_rate": 1.5168321909896171e-05, | |
| "loss": 0.0434, | |
| "step": 986 | |
| }, | |
| { | |
| "epoch": 0.3952, | |
| "grad_norm": 0.12561841309070587, | |
| "learning_rate": 1.5144395337815066e-05, | |
| "loss": 1.176, | |
| "step": 988 | |
| }, | |
| { | |
| "epoch": 0.396, | |
| "grad_norm": 0.03084614872932434, | |
| "learning_rate": 1.5120428648705716e-05, | |
| "loss": 0.0051, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.3968, | |
| "grad_norm": 0.08889269083738327, | |
| "learning_rate": 1.5096422029465178e-05, | |
| "loss": 0.0162, | |
| "step": 992 | |
| }, | |
| { | |
| "epoch": 0.3976, | |
| "grad_norm": 0.012536651454865932, | |
| "learning_rate": 1.5072375667301893e-05, | |
| "loss": 0.0087, | |
| "step": 994 | |
| }, | |
| { | |
| "epoch": 0.3984, | |
| "grad_norm": 0.08808624744415283, | |
| "learning_rate": 1.504828974973422e-05, | |
| "loss": 0.254, | |
| "step": 996 | |
| }, | |
| { | |
| "epoch": 0.3992, | |
| "grad_norm": 0.00895402766764164, | |
| "learning_rate": 1.5024164464588982e-05, | |
| "loss": 0.0067, | |
| "step": 998 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.01615263894200325, | |
| "learning_rate": 1.5000000000000002e-05, | |
| "loss": 0.0474, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.4008, | |
| "grad_norm": 0.11077135056257248, | |
| "learning_rate": 1.4975796544406627e-05, | |
| "loss": 0.0484, | |
| "step": 1002 | |
| }, | |
| { | |
| "epoch": 0.4016, | |
| "grad_norm": 1.5227055549621582, | |
| "learning_rate": 1.4951554286552266e-05, | |
| "loss": 0.88, | |
| "step": 1004 | |
| }, | |
| { | |
| "epoch": 0.4024, | |
| "grad_norm": 0.11261726170778275, | |
| "learning_rate": 1.4927273415482916e-05, | |
| "loss": 0.0309, | |
| "step": 1006 | |
| }, | |
| { | |
| "epoch": 0.4032, | |
| "grad_norm": 3.3870084285736084, | |
| "learning_rate": 1.4902954120545687e-05, | |
| "loss": 0.2266, | |
| "step": 1008 | |
| }, | |
| { | |
| "epoch": 0.404, | |
| "grad_norm": 0.6803337335586548, | |
| "learning_rate": 1.4878596591387329e-05, | |
| "loss": 0.0631, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.4048, | |
| "grad_norm": 0.2337927520275116, | |
| "learning_rate": 1.485420101795274e-05, | |
| "loss": 0.0601, | |
| "step": 1012 | |
| }, | |
| { | |
| "epoch": 0.4056, | |
| "grad_norm": 2.0759363174438477, | |
| "learning_rate": 1.4829767590483508e-05, | |
| "loss": 0.1077, | |
| "step": 1014 | |
| }, | |
| { | |
| "epoch": 0.4064, | |
| "grad_norm": 0.16511759161949158, | |
| "learning_rate": 1.4805296499516408e-05, | |
| "loss": 0.0283, | |
| "step": 1016 | |
| }, | |
| { | |
| "epoch": 0.4072, | |
| "grad_norm": 0.013055981136858463, | |
| "learning_rate": 1.4780787935881925e-05, | |
| "loss": 0.0067, | |
| "step": 1018 | |
| }, | |
| { | |
| "epoch": 0.408, | |
| "grad_norm": 1.3341008424758911, | |
| "learning_rate": 1.4756242090702756e-05, | |
| "loss": 0.1319, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.4088, | |
| "grad_norm": 0.21409232914447784, | |
| "learning_rate": 1.4731659155392332e-05, | |
| "loss": 0.0223, | |
| "step": 1022 | |
| }, | |
| { | |
| "epoch": 0.4096, | |
| "grad_norm": 0.007527098525315523, | |
| "learning_rate": 1.470703932165333e-05, | |
| "loss": 0.008, | |
| "step": 1024 | |
| }, | |
| { | |
| "epoch": 0.4104, | |
| "grad_norm": 0.02161354571580887, | |
| "learning_rate": 1.4682382781476146e-05, | |
| "loss": 0.0093, | |
| "step": 1026 | |
| }, | |
| { | |
| "epoch": 0.4112, | |
| "grad_norm": 3.1243057250976562, | |
| "learning_rate": 1.4657689727137443e-05, | |
| "loss": 0.2774, | |
| "step": 1028 | |
| }, | |
| { | |
| "epoch": 0.412, | |
| "grad_norm": 0.7081921696662903, | |
| "learning_rate": 1.463296035119862e-05, | |
| "loss": 0.3953, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.4128, | |
| "grad_norm": 0.23352546989917755, | |
| "learning_rate": 1.4608194846504311e-05, | |
| "loss": 0.0156, | |
| "step": 1032 | |
| }, | |
| { | |
| "epoch": 0.4136, | |
| "grad_norm": 0.44725605845451355, | |
| "learning_rate": 1.4583393406180898e-05, | |
| "loss": 0.0288, | |
| "step": 1034 | |
| }, | |
| { | |
| "epoch": 0.4144, | |
| "grad_norm": 0.10920588672161102, | |
| "learning_rate": 1.4558556223635004e-05, | |
| "loss": 0.0283, | |
| "step": 1036 | |
| }, | |
| { | |
| "epoch": 0.4152, | |
| "grad_norm": 0.004958951845765114, | |
| "learning_rate": 1.4533683492551954e-05, | |
| "loss": 0.0138, | |
| "step": 1038 | |
| }, | |
| { | |
| "epoch": 0.416, | |
| "grad_norm": 0.009044544771313667, | |
| "learning_rate": 1.4508775406894308e-05, | |
| "loss": 0.0509, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.4168, | |
| "grad_norm": 4.271718502044678, | |
| "learning_rate": 1.4483832160900326e-05, | |
| "loss": 0.6421, | |
| "step": 1042 | |
| }, | |
| { | |
| "epoch": 0.4176, | |
| "grad_norm": 0.005885216407477856, | |
| "learning_rate": 1.4458853949082443e-05, | |
| "loss": 0.0097, | |
| "step": 1044 | |
| }, | |
| { | |
| "epoch": 0.4184, | |
| "grad_norm": 0.004338196478784084, | |
| "learning_rate": 1.4433840966225772e-05, | |
| "loss": 0.0071, | |
| "step": 1046 | |
| }, | |
| { | |
| "epoch": 0.4192, | |
| "grad_norm": 0.0037080624606460333, | |
| "learning_rate": 1.4408793407386587e-05, | |
| "loss": 0.012, | |
| "step": 1048 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.032303668558597565, | |
| "learning_rate": 1.4383711467890776e-05, | |
| "loss": 0.0057, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.4208, | |
| "grad_norm": 0.006096419878304005, | |
| "learning_rate": 1.4358595343332342e-05, | |
| "loss": 0.0036, | |
| "step": 1052 | |
| }, | |
| { | |
| "epoch": 0.4216, | |
| "grad_norm": 0.018376469612121582, | |
| "learning_rate": 1.4333445229571874e-05, | |
| "loss": 0.0039, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 0.4224, | |
| "grad_norm": 0.05468170344829559, | |
| "learning_rate": 1.4308261322735006e-05, | |
| "loss": 0.0417, | |
| "step": 1056 | |
| }, | |
| { | |
| "epoch": 0.4232, | |
| "grad_norm": 0.033594775944948196, | |
| "learning_rate": 1.4283043819210905e-05, | |
| "loss": 0.0169, | |
| "step": 1058 | |
| }, | |
| { | |
| "epoch": 0.424, | |
| "grad_norm": 0.006937976460903883, | |
| "learning_rate": 1.4257792915650728e-05, | |
| "loss": 0.0705, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.4248, | |
| "grad_norm": 0.01067473366856575, | |
| "learning_rate": 1.4232508808966097e-05, | |
| "loss": 0.0099, | |
| "step": 1062 | |
| }, | |
| { | |
| "epoch": 0.4256, | |
| "grad_norm": 0.032820601016283035, | |
| "learning_rate": 1.420719169632755e-05, | |
| "loss": 0.0063, | |
| "step": 1064 | |
| }, | |
| { | |
| "epoch": 0.4264, | |
| "grad_norm": 4.574073314666748, | |
| "learning_rate": 1.4181841775163014e-05, | |
| "loss": 1.646, | |
| "step": 1066 | |
| }, | |
| { | |
| "epoch": 0.4272, | |
| "grad_norm": 2.232372999191284, | |
| "learning_rate": 1.415645924315628e-05, | |
| "loss": 0.0774, | |
| "step": 1068 | |
| }, | |
| { | |
| "epoch": 0.428, | |
| "grad_norm": 0.007644816767424345, | |
| "learning_rate": 1.413104429824542e-05, | |
| "loss": 0.0042, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.4288, | |
| "grad_norm": 0.041802894324064255, | |
| "learning_rate": 1.4105597138621281e-05, | |
| "loss": 0.0084, | |
| "step": 1072 | |
| }, | |
| { | |
| "epoch": 0.4296, | |
| "grad_norm": 1.1193236112594604, | |
| "learning_rate": 1.4080117962725929e-05, | |
| "loss": 0.5387, | |
| "step": 1074 | |
| }, | |
| { | |
| "epoch": 0.4304, | |
| "grad_norm": 0.04370833560824394, | |
| "learning_rate": 1.4054606969251095e-05, | |
| "loss": 0.0382, | |
| "step": 1076 | |
| }, | |
| { | |
| "epoch": 0.4312, | |
| "grad_norm": 0.9142640233039856, | |
| "learning_rate": 1.4029064357136628e-05, | |
| "loss": 0.8189, | |
| "step": 1078 | |
| }, | |
| { | |
| "epoch": 0.432, | |
| "grad_norm": 0.30186763405799866, | |
| "learning_rate": 1.4003490325568953e-05, | |
| "loss": 0.0461, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.4328, | |
| "grad_norm": 0.05774744972586632, | |
| "learning_rate": 1.39778850739795e-05, | |
| "loss": 0.6106, | |
| "step": 1082 | |
| }, | |
| { | |
| "epoch": 0.4336, | |
| "grad_norm": 0.3426492214202881, | |
| "learning_rate": 1.3952248802043166e-05, | |
| "loss": 0.0866, | |
| "step": 1084 | |
| }, | |
| { | |
| "epoch": 0.4344, | |
| "grad_norm": 0.5425324440002441, | |
| "learning_rate": 1.3926581709676752e-05, | |
| "loss": 0.1086, | |
| "step": 1086 | |
| }, | |
| { | |
| "epoch": 0.4352, | |
| "grad_norm": 0.10301554948091507, | |
| "learning_rate": 1.3900883997037398e-05, | |
| "loss": 0.0468, | |
| "step": 1088 | |
| }, | |
| { | |
| "epoch": 0.436, | |
| "grad_norm": 1.1397227048873901, | |
| "learning_rate": 1.3875155864521031e-05, | |
| "loss": 0.2403, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.4368, | |
| "grad_norm": 0.02772480994462967, | |
| "learning_rate": 1.3849397512760797e-05, | |
| "loss": 0.1271, | |
| "step": 1092 | |
| }, | |
| { | |
| "epoch": 0.4376, | |
| "grad_norm": 0.10023550689220428, | |
| "learning_rate": 1.3823609142625492e-05, | |
| "loss": 0.0979, | |
| "step": 1094 | |
| }, | |
| { | |
| "epoch": 0.4384, | |
| "grad_norm": 1.5255858898162842, | |
| "learning_rate": 1.3797790955218014e-05, | |
| "loss": 0.7019, | |
| "step": 1096 | |
| }, | |
| { | |
| "epoch": 0.4392, | |
| "grad_norm": 0.0757753774523735, | |
| "learning_rate": 1.3771943151873768e-05, | |
| "loss": 0.0422, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 0.053413309156894684, | |
| "learning_rate": 1.3746065934159123e-05, | |
| "loss": 0.4052, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.4408, | |
| "grad_norm": 0.061455827206373215, | |
| "learning_rate": 1.3720159503869816e-05, | |
| "loss": 0.0402, | |
| "step": 1102 | |
| }, | |
| { | |
| "epoch": 0.4416, | |
| "grad_norm": 0.3662703335285187, | |
| "learning_rate": 1.3694224063029396e-05, | |
| "loss": 0.0559, | |
| "step": 1104 | |
| }, | |
| { | |
| "epoch": 0.4424, | |
| "grad_norm": 0.09608737379312515, | |
| "learning_rate": 1.3668259813887644e-05, | |
| "loss": 0.0703, | |
| "step": 1106 | |
| }, | |
| { | |
| "epoch": 0.4432, | |
| "grad_norm": 0.09362676739692688, | |
| "learning_rate": 1.3642266958918985e-05, | |
| "loss": 0.0191, | |
| "step": 1108 | |
| }, | |
| { | |
| "epoch": 0.444, | |
| "grad_norm": 0.03751781955361366, | |
| "learning_rate": 1.3616245700820922e-05, | |
| "loss": 0.1236, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.4448, | |
| "grad_norm": 2.1613173484802246, | |
| "learning_rate": 1.3590196242512463e-05, | |
| "loss": 0.6019, | |
| "step": 1112 | |
| }, | |
| { | |
| "epoch": 0.4456, | |
| "grad_norm": 0.03422972559928894, | |
| "learning_rate": 1.3564118787132507e-05, | |
| "loss": 0.0138, | |
| "step": 1114 | |
| }, | |
| { | |
| "epoch": 0.4464, | |
| "grad_norm": 0.01046049501746893, | |
| "learning_rate": 1.3538013538038295e-05, | |
| "loss": 0.0101, | |
| "step": 1116 | |
| }, | |
| { | |
| "epoch": 0.4472, | |
| "grad_norm": 0.7763115167617798, | |
| "learning_rate": 1.3511880698803801e-05, | |
| "loss": 0.1013, | |
| "step": 1118 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 0.2062385082244873, | |
| "learning_rate": 1.3485720473218153e-05, | |
| "loss": 0.0343, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.4488, | |
| "grad_norm": 0.3189225196838379, | |
| "learning_rate": 1.3459533065284049e-05, | |
| "loss": 0.0661, | |
| "step": 1122 | |
| }, | |
| { | |
| "epoch": 0.4496, | |
| "grad_norm": 0.9145419597625732, | |
| "learning_rate": 1.3433318679216154e-05, | |
| "loss": 0.555, | |
| "step": 1124 | |
| }, | |
| { | |
| "epoch": 0.4504, | |
| "grad_norm": 1.2541166543960571, | |
| "learning_rate": 1.340707751943952e-05, | |
| "loss": 0.5902, | |
| "step": 1126 | |
| }, | |
| { | |
| "epoch": 0.4512, | |
| "grad_norm": 0.02262170799076557, | |
| "learning_rate": 1.3380809790587975e-05, | |
| "loss": 0.0191, | |
| "step": 1128 | |
| }, | |
| { | |
| "epoch": 0.452, | |
| "grad_norm": 0.4805641174316406, | |
| "learning_rate": 1.3354515697502552e-05, | |
| "loss": 0.0719, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.4528, | |
| "grad_norm": 0.0759739875793457, | |
| "learning_rate": 1.3328195445229869e-05, | |
| "loss": 0.035, | |
| "step": 1132 | |
| }, | |
| { | |
| "epoch": 0.4536, | |
| "grad_norm": 0.0279012992978096, | |
| "learning_rate": 1.3301849239020537e-05, | |
| "loss": 0.0732, | |
| "step": 1134 | |
| }, | |
| { | |
| "epoch": 0.4544, | |
| "grad_norm": 0.040366217494010925, | |
| "learning_rate": 1.327547728432757e-05, | |
| "loss": 0.036, | |
| "step": 1136 | |
| }, | |
| { | |
| "epoch": 0.4552, | |
| "grad_norm": 1.5210012197494507, | |
| "learning_rate": 1.3249079786804765e-05, | |
| "loss": 0.1844, | |
| "step": 1138 | |
| }, | |
| { | |
| "epoch": 0.456, | |
| "grad_norm": 0.26963767409324646, | |
| "learning_rate": 1.3222656952305113e-05, | |
| "loss": 0.0458, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.4568, | |
| "grad_norm": 2.5474393367767334, | |
| "learning_rate": 1.319620898687918e-05, | |
| "loss": 0.4129, | |
| "step": 1142 | |
| }, | |
| { | |
| "epoch": 0.4576, | |
| "grad_norm": 0.021995197981595993, | |
| "learning_rate": 1.316973609677352e-05, | |
| "loss": 0.0168, | |
| "step": 1144 | |
| }, | |
| { | |
| "epoch": 0.4584, | |
| "grad_norm": 0.0323915109038353, | |
| "learning_rate": 1.3143238488429042e-05, | |
| "loss": 0.109, | |
| "step": 1146 | |
| }, | |
| { | |
| "epoch": 0.4592, | |
| "grad_norm": 0.050105806440114975, | |
| "learning_rate": 1.3116716368479418e-05, | |
| "loss": 0.0305, | |
| "step": 1148 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 0.13928213715553284, | |
| "learning_rate": 1.3090169943749475e-05, | |
| "loss": 0.0237, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.4608, | |
| "grad_norm": 0.06462297588586807, | |
| "learning_rate": 1.306359942125356e-05, | |
| "loss": 0.0192, | |
| "step": 1152 | |
| }, | |
| { | |
| "epoch": 0.4616, | |
| "grad_norm": 0.022758983075618744, | |
| "learning_rate": 1.3037005008193944e-05, | |
| "loss": 0.7191, | |
| "step": 1154 | |
| }, | |
| { | |
| "epoch": 0.4624, | |
| "grad_norm": 0.03799004480242729, | |
| "learning_rate": 1.3010386911959207e-05, | |
| "loss": 0.0214, | |
| "step": 1156 | |
| }, | |
| { | |
| "epoch": 0.4632, | |
| "grad_norm": 0.2102702260017395, | |
| "learning_rate": 1.2983745340122604e-05, | |
| "loss": 0.1163, | |
| "step": 1158 | |
| }, | |
| { | |
| "epoch": 0.464, | |
| "grad_norm": 0.08294696360826492, | |
| "learning_rate": 1.2957080500440469e-05, | |
| "loss": 0.0268, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.4648, | |
| "grad_norm": 0.04941265285015106, | |
| "learning_rate": 1.2930392600850574e-05, | |
| "loss": 0.0154, | |
| "step": 1162 | |
| }, | |
| { | |
| "epoch": 0.4656, | |
| "grad_norm": 0.05030106008052826, | |
| "learning_rate": 1.2903681849470528e-05, | |
| "loss": 0.1079, | |
| "step": 1164 | |
| }, | |
| { | |
| "epoch": 0.4664, | |
| "grad_norm": 0.11340274661779404, | |
| "learning_rate": 1.287694845459613e-05, | |
| "loss": 0.018, | |
| "step": 1166 | |
| }, | |
| { | |
| "epoch": 0.4672, | |
| "grad_norm": 0.06605678051710129, | |
| "learning_rate": 1.2850192624699762e-05, | |
| "loss": 0.0336, | |
| "step": 1168 | |
| }, | |
| { | |
| "epoch": 0.468, | |
| "grad_norm": 1.597208023071289, | |
| "learning_rate": 1.2823414568428767e-05, | |
| "loss": 0.7895, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.4688, | |
| "grad_norm": 0.12116717547178268, | |
| "learning_rate": 1.27966144946038e-05, | |
| "loss": 0.026, | |
| "step": 1172 | |
| }, | |
| { | |
| "epoch": 0.4696, | |
| "grad_norm": 0.04143265634775162, | |
| "learning_rate": 1.2769792612217224e-05, | |
| "loss": 0.1011, | |
| "step": 1174 | |
| }, | |
| { | |
| "epoch": 0.4704, | |
| "grad_norm": 0.09404078125953674, | |
| "learning_rate": 1.2742949130431468e-05, | |
| "loss": 0.0211, | |
| "step": 1176 | |
| }, | |
| { | |
| "epoch": 0.4712, | |
| "grad_norm": 0.024886123836040497, | |
| "learning_rate": 1.2716084258577388e-05, | |
| "loss": 0.012, | |
| "step": 1178 | |
| }, | |
| { | |
| "epoch": 0.472, | |
| "grad_norm": 0.05635695159435272, | |
| "learning_rate": 1.2689198206152657e-05, | |
| "loss": 0.0137, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.4728, | |
| "grad_norm": 0.09039817750453949, | |
| "learning_rate": 1.2662291182820115e-05, | |
| "loss": 0.0537, | |
| "step": 1182 | |
| }, | |
| { | |
| "epoch": 0.4736, | |
| "grad_norm": 0.41814690828323364, | |
| "learning_rate": 1.263536339840613e-05, | |
| "loss": 0.0814, | |
| "step": 1184 | |
| }, | |
| { | |
| "epoch": 0.4744, | |
| "grad_norm": 0.23532457649707794, | |
| "learning_rate": 1.2608415062898971e-05, | |
| "loss": 0.0322, | |
| "step": 1186 | |
| }, | |
| { | |
| "epoch": 0.4752, | |
| "grad_norm": 0.009926537983119488, | |
| "learning_rate": 1.2581446386447178e-05, | |
| "loss": 0.1281, | |
| "step": 1188 | |
| }, | |
| { | |
| "epoch": 0.476, | |
| "grad_norm": 0.4470027983188629, | |
| "learning_rate": 1.2554457579357906e-05, | |
| "loss": 0.0478, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.4768, | |
| "grad_norm": 0.05581334978342056, | |
| "learning_rate": 1.2527448852095295e-05, | |
| "loss": 0.0226, | |
| "step": 1192 | |
| }, | |
| { | |
| "epoch": 0.4776, | |
| "grad_norm": 0.06944628804922104, | |
| "learning_rate": 1.2500420415278822e-05, | |
| "loss": 0.0229, | |
| "step": 1194 | |
| }, | |
| { | |
| "epoch": 0.4784, | |
| "grad_norm": 0.11777735501527786, | |
| "learning_rate": 1.2473372479681671e-05, | |
| "loss": 0.0189, | |
| "step": 1196 | |
| }, | |
| { | |
| "epoch": 0.4792, | |
| "grad_norm": 0.37742823362350464, | |
| "learning_rate": 1.2446305256229074e-05, | |
| "loss": 0.2176, | |
| "step": 1198 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 1.1584476232528687, | |
| "learning_rate": 1.2419218955996677e-05, | |
| "loss": 0.5342, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.4808, | |
| "grad_norm": 0.21147167682647705, | |
| "learning_rate": 1.2392113790208895e-05, | |
| "loss": 0.027, | |
| "step": 1202 | |
| }, | |
| { | |
| "epoch": 0.4816, | |
| "grad_norm": 0.061478614807128906, | |
| "learning_rate": 1.236498997023725e-05, | |
| "loss": 0.027, | |
| "step": 1204 | |
| }, | |
| { | |
| "epoch": 0.4824, | |
| "grad_norm": 0.02769525721669197, | |
| "learning_rate": 1.2337847707598738e-05, | |
| "loss": 0.0325, | |
| "step": 1206 | |
| }, | |
| { | |
| "epoch": 0.4832, | |
| "grad_norm": 0.6708433628082275, | |
| "learning_rate": 1.2310687213954182e-05, | |
| "loss": 0.1057, | |
| "step": 1208 | |
| }, | |
| { | |
| "epoch": 0.484, | |
| "grad_norm": 0.08236505091190338, | |
| "learning_rate": 1.2283508701106559e-05, | |
| "loss": 0.0192, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.4848, | |
| "grad_norm": 0.01167634129524231, | |
| "learning_rate": 1.2256312380999376e-05, | |
| "loss": 0.3463, | |
| "step": 1212 | |
| }, | |
| { | |
| "epoch": 0.4856, | |
| "grad_norm": 0.03433239459991455, | |
| "learning_rate": 1.2229098465715005e-05, | |
| "loss": 0.3755, | |
| "step": 1214 | |
| }, | |
| { | |
| "epoch": 0.4864, | |
| "grad_norm": 1.2798956632614136, | |
| "learning_rate": 1.2201867167473015e-05, | |
| "loss": 0.3444, | |
| "step": 1216 | |
| }, | |
| { | |
| "epoch": 0.4872, | |
| "grad_norm": 0.003344225697219372, | |
| "learning_rate": 1.217461869862855e-05, | |
| "loss": 0.0047, | |
| "step": 1218 | |
| }, | |
| { | |
| "epoch": 0.488, | |
| "grad_norm": 0.006934499368071556, | |
| "learning_rate": 1.2147353271670634e-05, | |
| "loss": 0.0031, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.4888, | |
| "grad_norm": 0.04046230763196945, | |
| "learning_rate": 1.212007109922055e-05, | |
| "loss": 0.0566, | |
| "step": 1222 | |
| }, | |
| { | |
| "epoch": 0.4896, | |
| "grad_norm": 1.6989425420761108, | |
| "learning_rate": 1.2092772394030153e-05, | |
| "loss": 0.3364, | |
| "step": 1224 | |
| }, | |
| { | |
| "epoch": 0.4904, | |
| "grad_norm": 0.6641682982444763, | |
| "learning_rate": 1.2065457368980236e-05, | |
| "loss": 0.0774, | |
| "step": 1226 | |
| }, | |
| { | |
| "epoch": 0.4912, | |
| "grad_norm": 0.07043974846601486, | |
| "learning_rate": 1.203812623707885e-05, | |
| "loss": 0.0186, | |
| "step": 1228 | |
| }, | |
| { | |
| "epoch": 0.492, | |
| "grad_norm": 0.01750028319656849, | |
| "learning_rate": 1.2010779211459649e-05, | |
| "loss": 0.1607, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.4928, | |
| "grad_norm": 0.14301519095897675, | |
| "learning_rate": 1.1983416505380234e-05, | |
| "loss": 0.0485, | |
| "step": 1232 | |
| }, | |
| { | |
| "epoch": 0.4936, | |
| "grad_norm": 0.046129919588565826, | |
| "learning_rate": 1.1956038332220484e-05, | |
| "loss": 0.0111, | |
| "step": 1234 | |
| }, | |
| { | |
| "epoch": 0.4944, | |
| "grad_norm": 1.4883394241333008, | |
| "learning_rate": 1.192864490548089e-05, | |
| "loss": 0.5768, | |
| "step": 1236 | |
| }, | |
| { | |
| "epoch": 0.4952, | |
| "grad_norm": 1.122119665145874, | |
| "learning_rate": 1.1901236438780902e-05, | |
| "loss": 0.1481, | |
| "step": 1238 | |
| }, | |
| { | |
| "epoch": 0.496, | |
| "grad_norm": 0.19744625687599182, | |
| "learning_rate": 1.187381314585725e-05, | |
| "loss": 0.1095, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.4968, | |
| "grad_norm": 0.06524749845266342, | |
| "learning_rate": 1.184637524056227e-05, | |
| "loss": 0.1115, | |
| "step": 1242 | |
| }, | |
| { | |
| "epoch": 0.4976, | |
| "grad_norm": 0.024867022410035133, | |
| "learning_rate": 1.181892293686227e-05, | |
| "loss": 0.0095, | |
| "step": 1244 | |
| }, | |
| { | |
| "epoch": 0.4984, | |
| "grad_norm": 0.20855647325515747, | |
| "learning_rate": 1.1791456448835825e-05, | |
| "loss": 0.0315, | |
| "step": 1246 | |
| }, | |
| { | |
| "epoch": 0.4992, | |
| "grad_norm": 0.009501822292804718, | |
| "learning_rate": 1.1763975990672125e-05, | |
| "loss": 0.0212, | |
| "step": 1248 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.030127333477139473, | |
| "learning_rate": 1.1736481776669307e-05, | |
| "loss": 0.0104, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.5008, | |
| "grad_norm": 0.793460488319397, | |
| "learning_rate": 1.1708974021232768e-05, | |
| "loss": 0.1835, | |
| "step": 1252 | |
| }, | |
| { | |
| "epoch": 0.5016, | |
| "grad_norm": 0.1162368431687355, | |
| "learning_rate": 1.1681452938873516e-05, | |
| "loss": 0.0165, | |
| "step": 1254 | |
| }, | |
| { | |
| "epoch": 0.5024, | |
| "grad_norm": 0.03160820156335831, | |
| "learning_rate": 1.1653918744206478e-05, | |
| "loss": 0.0659, | |
| "step": 1256 | |
| }, | |
| { | |
| "epoch": 0.5032, | |
| "grad_norm": 2.536365509033203, | |
| "learning_rate": 1.1626371651948839e-05, | |
| "loss": 0.1661, | |
| "step": 1258 | |
| }, | |
| { | |
| "epoch": 0.504, | |
| "grad_norm": 0.06747017800807953, | |
| "learning_rate": 1.159881187691835e-05, | |
| "loss": 0.0591, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.5048, | |
| "grad_norm": 0.10115383565425873, | |
| "learning_rate": 1.157123963403168e-05, | |
| "loss": 0.0216, | |
| "step": 1262 | |
| }, | |
| { | |
| "epoch": 0.5056, | |
| "grad_norm": 0.017214270308613777, | |
| "learning_rate": 1.1543655138302714e-05, | |
| "loss": 0.0077, | |
| "step": 1264 | |
| }, | |
| { | |
| "epoch": 0.5064, | |
| "grad_norm": 0.21995119750499725, | |
| "learning_rate": 1.1516058604840891e-05, | |
| "loss": 0.0389, | |
| "step": 1266 | |
| }, | |
| { | |
| "epoch": 0.5072, | |
| "grad_norm": 0.01604562997817993, | |
| "learning_rate": 1.1488450248849523e-05, | |
| "loss": 0.0223, | |
| "step": 1268 | |
| }, | |
| { | |
| "epoch": 0.508, | |
| "grad_norm": 0.031004084274172783, | |
| "learning_rate": 1.1460830285624119e-05, | |
| "loss": 0.0159, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.5088, | |
| "grad_norm": 2.251960039138794, | |
| "learning_rate": 1.1433198930550694e-05, | |
| "loss": 0.2915, | |
| "step": 1272 | |
| }, | |
| { | |
| "epoch": 0.5096, | |
| "grad_norm": 0.03171626105904579, | |
| "learning_rate": 1.140555639910411e-05, | |
| "loss": 0.0069, | |
| "step": 1274 | |
| }, | |
| { | |
| "epoch": 0.5104, | |
| "grad_norm": 0.791187584400177, | |
| "learning_rate": 1.137790290684638e-05, | |
| "loss": 0.21, | |
| "step": 1276 | |
| }, | |
| { | |
| "epoch": 0.5112, | |
| "grad_norm": 0.016608070582151413, | |
| "learning_rate": 1.1350238669424993e-05, | |
| "loss": 0.0064, | |
| "step": 1278 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 0.0038915143813937902, | |
| "learning_rate": 1.1322563902571227e-05, | |
| "loss": 0.002, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.5128, | |
| "grad_norm": 0.2625635862350464, | |
| "learning_rate": 1.129487882209847e-05, | |
| "loss": 0.0156, | |
| "step": 1282 | |
| }, | |
| { | |
| "epoch": 0.5136, | |
| "grad_norm": 0.037254203110933304, | |
| "learning_rate": 1.1267183643900548e-05, | |
| "loss": 0.0056, | |
| "step": 1284 | |
| }, | |
| { | |
| "epoch": 0.5144, | |
| "grad_norm": 0.08309903740882874, | |
| "learning_rate": 1.1239478583950019e-05, | |
| "loss": 0.0056, | |
| "step": 1286 | |
| }, | |
| { | |
| "epoch": 0.5152, | |
| "grad_norm": 1.2419265508651733, | |
| "learning_rate": 1.1211763858296507e-05, | |
| "loss": 0.1099, | |
| "step": 1288 | |
| }, | |
| { | |
| "epoch": 0.516, | |
| "grad_norm": 0.008148876950144768, | |
| "learning_rate": 1.1184039683065014e-05, | |
| "loss": 0.0064, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.5168, | |
| "grad_norm": 0.04519444331526756, | |
| "learning_rate": 1.1156306274454218e-05, | |
| "loss": 0.0066, | |
| "step": 1292 | |
| }, | |
| { | |
| "epoch": 0.5176, | |
| "grad_norm": 0.0015114143025130033, | |
| "learning_rate": 1.1128563848734817e-05, | |
| "loss": 0.001, | |
| "step": 1294 | |
| }, | |
| { | |
| "epoch": 0.5184, | |
| "grad_norm": 0.006196278613060713, | |
| "learning_rate": 1.1100812622247823e-05, | |
| "loss": 0.0032, | |
| "step": 1296 | |
| }, | |
| { | |
| "epoch": 0.5192, | |
| "grad_norm": 0.0075135682709515095, | |
| "learning_rate": 1.1073052811402867e-05, | |
| "loss": 0.6506, | |
| "step": 1298 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 0.030847422778606415, | |
| "learning_rate": 1.1045284632676535e-05, | |
| "loss": 0.016, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.5208, | |
| "grad_norm": 0.07433111220598221, | |
| "learning_rate": 1.1017508302610665e-05, | |
| "loss": 0.0134, | |
| "step": 1302 | |
| }, | |
| { | |
| "epoch": 0.5216, | |
| "grad_norm": 0.20086438953876495, | |
| "learning_rate": 1.0989724037810651e-05, | |
| "loss": 0.0139, | |
| "step": 1304 | |
| }, | |
| { | |
| "epoch": 0.5224, | |
| "grad_norm": 0.015072687529027462, | |
| "learning_rate": 1.0961932054943778e-05, | |
| "loss": 0.0192, | |
| "step": 1306 | |
| }, | |
| { | |
| "epoch": 0.5232, | |
| "grad_norm": 0.5555530190467834, | |
| "learning_rate": 1.0934132570737508e-05, | |
| "loss": 0.0359, | |
| "step": 1308 | |
| }, | |
| { | |
| "epoch": 0.524, | |
| "grad_norm": 0.0621090903878212, | |
| "learning_rate": 1.0906325801977804e-05, | |
| "loss": 0.0126, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.5248, | |
| "grad_norm": 0.013316777534782887, | |
| "learning_rate": 1.0878511965507435e-05, | |
| "loss": 0.0085, | |
| "step": 1312 | |
| }, | |
| { | |
| "epoch": 0.5256, | |
| "grad_norm": 0.037804149091243744, | |
| "learning_rate": 1.0850691278224282e-05, | |
| "loss": 1.6496, | |
| "step": 1314 | |
| }, | |
| { | |
| "epoch": 0.5264, | |
| "grad_norm": 0.04232160374522209, | |
| "learning_rate": 1.0822863957079657e-05, | |
| "loss": 0.0068, | |
| "step": 1316 | |
| }, | |
| { | |
| "epoch": 0.5272, | |
| "grad_norm": 0.020469985902309418, | |
| "learning_rate": 1.07950302190766e-05, | |
| "loss": 0.1708, | |
| "step": 1318 | |
| }, | |
| { | |
| "epoch": 0.528, | |
| "grad_norm": 0.04445105418562889, | |
| "learning_rate": 1.0767190281268187e-05, | |
| "loss": 0.1707, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.5288, | |
| "grad_norm": 1.9354827404022217, | |
| "learning_rate": 1.0739344360755853e-05, | |
| "loss": 0.1453, | |
| "step": 1322 | |
| }, | |
| { | |
| "epoch": 0.5296, | |
| "grad_norm": 0.0484926700592041, | |
| "learning_rate": 1.071149267468767e-05, | |
| "loss": 0.0299, | |
| "step": 1324 | |
| }, | |
| { | |
| "epoch": 0.5304, | |
| "grad_norm": 0.02645016647875309, | |
| "learning_rate": 1.0683635440256689e-05, | |
| "loss": 0.0162, | |
| "step": 1326 | |
| }, | |
| { | |
| "epoch": 0.5312, | |
| "grad_norm": 0.05882829427719116, | |
| "learning_rate": 1.0655772874699217e-05, | |
| "loss": 0.0311, | |
| "step": 1328 | |
| }, | |
| { | |
| "epoch": 0.532, | |
| "grad_norm": 0.03740588575601578, | |
| "learning_rate": 1.0627905195293135e-05, | |
| "loss": 0.0254, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.5328, | |
| "grad_norm": 0.21913115680217743, | |
| "learning_rate": 1.0600032619356208e-05, | |
| "loss": 0.0732, | |
| "step": 1332 | |
| }, | |
| { | |
| "epoch": 0.5336, | |
| "grad_norm": 0.02981780469417572, | |
| "learning_rate": 1.0572155364244383e-05, | |
| "loss": 0.0603, | |
| "step": 1334 | |
| }, | |
| { | |
| "epoch": 0.5344, | |
| "grad_norm": 0.3409574627876282, | |
| "learning_rate": 1.0544273647350091e-05, | |
| "loss": 0.0482, | |
| "step": 1336 | |
| }, | |
| { | |
| "epoch": 0.5352, | |
| "grad_norm": 1.3873035907745361, | |
| "learning_rate": 1.0516387686100566e-05, | |
| "loss": 0.2645, | |
| "step": 1338 | |
| }, | |
| { | |
| "epoch": 0.536, | |
| "grad_norm": 0.9133062958717346, | |
| "learning_rate": 1.0488497697956134e-05, | |
| "loss": 0.6023, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.5368, | |
| "grad_norm": 0.014215713366866112, | |
| "learning_rate": 1.0460603900408523e-05, | |
| "loss": 0.0249, | |
| "step": 1342 | |
| }, | |
| { | |
| "epoch": 0.5376, | |
| "grad_norm": 0.013020013459026814, | |
| "learning_rate": 1.0432706510979172e-05, | |
| "loss": 0.0572, | |
| "step": 1344 | |
| }, | |
| { | |
| "epoch": 0.5384, | |
| "grad_norm": 0.07060626894235611, | |
| "learning_rate": 1.0404805747217525e-05, | |
| "loss": 0.0259, | |
| "step": 1346 | |
| }, | |
| { | |
| "epoch": 0.5392, | |
| "grad_norm": 0.16781005263328552, | |
| "learning_rate": 1.0376901826699349e-05, | |
| "loss": 0.0441, | |
| "step": 1348 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 0.0290207639336586, | |
| "learning_rate": 1.0348994967025012e-05, | |
| "loss": 0.9059, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.5408, | |
| "grad_norm": 0.03659132868051529, | |
| "learning_rate": 1.0321085385817818e-05, | |
| "loss": 0.6523, | |
| "step": 1352 | |
| }, | |
| { | |
| "epoch": 0.5416, | |
| "grad_norm": 0.10811670869588852, | |
| "learning_rate": 1.0293173300722286e-05, | |
| "loss": 0.0472, | |
| "step": 1354 | |
| }, | |
| { | |
| "epoch": 0.5424, | |
| "grad_norm": 0.04781627655029297, | |
| "learning_rate": 1.026525892940246e-05, | |
| "loss": 0.0203, | |
| "step": 1356 | |
| }, | |
| { | |
| "epoch": 0.5432, | |
| "grad_norm": 0.0950162261724472, | |
| "learning_rate": 1.0237342489540221e-05, | |
| "loss": 0.5411, | |
| "step": 1358 | |
| }, | |
| { | |
| "epoch": 0.544, | |
| "grad_norm": 0.059019722044467926, | |
| "learning_rate": 1.0209424198833571e-05, | |
| "loss": 0.049, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.5448, | |
| "grad_norm": 0.0552980937063694, | |
| "learning_rate": 1.0181504274994949e-05, | |
| "loss": 0.5069, | |
| "step": 1362 | |
| }, | |
| { | |
| "epoch": 0.5456, | |
| "grad_norm": 0.037155695259571075, | |
| "learning_rate": 1.0153582935749531e-05, | |
| "loss": 0.0366, | |
| "step": 1364 | |
| }, | |
| { | |
| "epoch": 0.5464, | |
| "grad_norm": 0.32015174627304077, | |
| "learning_rate": 1.0125660398833528e-05, | |
| "loss": 0.1187, | |
| "step": 1366 | |
| }, | |
| { | |
| "epoch": 0.5472, | |
| "grad_norm": 0.0431252084672451, | |
| "learning_rate": 1.0097736881992492e-05, | |
| "loss": 0.0293, | |
| "step": 1368 | |
| }, | |
| { | |
| "epoch": 0.548, | |
| "grad_norm": 0.05704091861844063, | |
| "learning_rate": 1.0069812602979617e-05, | |
| "loss": 0.0485, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.5488, | |
| "grad_norm": 0.0650290697813034, | |
| "learning_rate": 1.0041887779554041e-05, | |
| "loss": 0.0455, | |
| "step": 1372 | |
| }, | |
| { | |
| "epoch": 0.5496, | |
| "grad_norm": 2.1595537662506104, | |
| "learning_rate": 1.0013962629479145e-05, | |
| "loss": 0.1577, | |
| "step": 1374 | |
| }, | |
| { | |
| "epoch": 0.5504, | |
| "grad_norm": 0.05115671455860138, | |
| "learning_rate": 9.986037370520856e-06, | |
| "loss": 0.0318, | |
| "step": 1376 | |
| }, | |
| { | |
| "epoch": 0.5512, | |
| "grad_norm": 0.06018679961562157, | |
| "learning_rate": 9.958112220445964e-06, | |
| "loss": 0.0382, | |
| "step": 1378 | |
| }, | |
| { | |
| "epoch": 0.552, | |
| "grad_norm": 0.0702604129910469, | |
| "learning_rate": 9.930187397020385e-06, | |
| "loss": 0.5824, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.5528, | |
| "grad_norm": 0.038693517446517944, | |
| "learning_rate": 9.902263118007513e-06, | |
| "loss": 0.0264, | |
| "step": 1382 | |
| }, | |
| { | |
| "epoch": 0.5536, | |
| "grad_norm": 0.018939625471830368, | |
| "learning_rate": 9.874339601166474e-06, | |
| "loss": 0.0496, | |
| "step": 1384 | |
| }, | |
| { | |
| "epoch": 0.5544, | |
| "grad_norm": 0.043414074927568436, | |
| "learning_rate": 9.84641706425047e-06, | |
| "loss": 0.0181, | |
| "step": 1386 | |
| }, | |
| { | |
| "epoch": 0.5552, | |
| "grad_norm": 0.5982060432434082, | |
| "learning_rate": 9.818495725005053e-06, | |
| "loss": 0.0623, | |
| "step": 1388 | |
| }, | |
| { | |
| "epoch": 0.556, | |
| "grad_norm": 0.13789403438568115, | |
| "learning_rate": 9.790575801166432e-06, | |
| "loss": 0.3926, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.5568, | |
| "grad_norm": 0.029673095792531967, | |
| "learning_rate": 9.762657510459784e-06, | |
| "loss": 0.3696, | |
| "step": 1392 | |
| }, | |
| { | |
| "epoch": 0.5576, | |
| "grad_norm": 1.172825813293457, | |
| "learning_rate": 9.73474107059754e-06, | |
| "loss": 0.8355, | |
| "step": 1394 | |
| }, | |
| { | |
| "epoch": 0.5584, | |
| "grad_norm": 1.2738497257232666, | |
| "learning_rate": 9.706826699277719e-06, | |
| "loss": 0.2447, | |
| "step": 1396 | |
| }, | |
| { | |
| "epoch": 0.5592, | |
| "grad_norm": 1.911895751953125, | |
| "learning_rate": 9.678914614182185e-06, | |
| "loss": 0.2138, | |
| "step": 1398 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.17634525895118713, | |
| "learning_rate": 9.651005032974994e-06, | |
| "loss": 0.0485, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.5608, | |
| "grad_norm": 0.03239009529352188, | |
| "learning_rate": 9.623098173300655e-06, | |
| "loss": 0.0351, | |
| "step": 1402 | |
| }, | |
| { | |
| "epoch": 0.5616, | |
| "grad_norm": 0.12052161246538162, | |
| "learning_rate": 9.595194252782476e-06, | |
| "loss": 0.0429, | |
| "step": 1404 | |
| }, | |
| { | |
| "epoch": 0.5624, | |
| "grad_norm": 0.06105445697903633, | |
| "learning_rate": 9.567293489020831e-06, | |
| "loss": 0.0241, | |
| "step": 1406 | |
| }, | |
| { | |
| "epoch": 0.5632, | |
| "grad_norm": 0.10784406960010529, | |
| "learning_rate": 9.539396099591477e-06, | |
| "loss": 0.0291, | |
| "step": 1408 | |
| }, | |
| { | |
| "epoch": 0.564, | |
| "grad_norm": 0.5862268209457397, | |
| "learning_rate": 9.511502302043867e-06, | |
| "loss": 0.0746, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.5648, | |
| "grad_norm": 2.016383409500122, | |
| "learning_rate": 9.483612313899436e-06, | |
| "loss": 0.1858, | |
| "step": 1412 | |
| }, | |
| { | |
| "epoch": 0.5656, | |
| "grad_norm": 0.05418021231889725, | |
| "learning_rate": 9.45572635264991e-06, | |
| "loss": 0.0362, | |
| "step": 1414 | |
| }, | |
| { | |
| "epoch": 0.5664, | |
| "grad_norm": 1.7044521570205688, | |
| "learning_rate": 9.42784463575562e-06, | |
| "loss": 0.2979, | |
| "step": 1416 | |
| }, | |
| { | |
| "epoch": 0.5672, | |
| "grad_norm": 0.09087449312210083, | |
| "learning_rate": 9.399967380643795e-06, | |
| "loss": 0.0365, | |
| "step": 1418 | |
| }, | |
| { | |
| "epoch": 0.568, | |
| "grad_norm": 0.16950923204421997, | |
| "learning_rate": 9.372094804706867e-06, | |
| "loss": 0.0382, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.5688, | |
| "grad_norm": 0.1982814371585846, | |
| "learning_rate": 9.344227125300788e-06, | |
| "loss": 0.0395, | |
| "step": 1422 | |
| }, | |
| { | |
| "epoch": 0.5696, | |
| "grad_norm": 1.310996413230896, | |
| "learning_rate": 9.316364559743315e-06, | |
| "loss": 0.5231, | |
| "step": 1424 | |
| }, | |
| { | |
| "epoch": 0.5704, | |
| "grad_norm": 0.03387328237295151, | |
| "learning_rate": 9.288507325312334e-06, | |
| "loss": 0.6156, | |
| "step": 1426 | |
| }, | |
| { | |
| "epoch": 0.5712, | |
| "grad_norm": 0.009328456595540047, | |
| "learning_rate": 9.260655639244152e-06, | |
| "loss": 0.0184, | |
| "step": 1428 | |
| }, | |
| { | |
| "epoch": 0.572, | |
| "grad_norm": 0.07381106168031693, | |
| "learning_rate": 9.232809718731815e-06, | |
| "loss": 0.0261, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.5728, | |
| "grad_norm": 0.050106290727853775, | |
| "learning_rate": 9.204969780923404e-06, | |
| "loss": 0.0132, | |
| "step": 1432 | |
| }, | |
| { | |
| "epoch": 0.5736, | |
| "grad_norm": 0.010202400386333466, | |
| "learning_rate": 9.177136042920344e-06, | |
| "loss": 0.0187, | |
| "step": 1434 | |
| }, | |
| { | |
| "epoch": 0.5744, | |
| "grad_norm": 0.15420140326023102, | |
| "learning_rate": 9.14930872177572e-06, | |
| "loss": 0.062, | |
| "step": 1436 | |
| }, | |
| { | |
| "epoch": 0.5752, | |
| "grad_norm": 0.028073610737919807, | |
| "learning_rate": 9.121488034492569e-06, | |
| "loss": 0.0171, | |
| "step": 1438 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 0.018873965367674828, | |
| "learning_rate": 9.093674198022201e-06, | |
| "loss": 0.019, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.5768, | |
| "grad_norm": 0.8036087155342102, | |
| "learning_rate": 9.065867429262497e-06, | |
| "loss": 1.608, | |
| "step": 1442 | |
| }, | |
| { | |
| "epoch": 0.5776, | |
| "grad_norm": 0.19082395732402802, | |
| "learning_rate": 9.038067945056229e-06, | |
| "loss": 0.0394, | |
| "step": 1444 | |
| }, | |
| { | |
| "epoch": 0.5784, | |
| "grad_norm": 0.07714349031448364, | |
| "learning_rate": 9.01027596218935e-06, | |
| "loss": 0.0772, | |
| "step": 1446 | |
| }, | |
| { | |
| "epoch": 0.5792, | |
| "grad_norm": 0.06311635673046112, | |
| "learning_rate": 8.982491697389339e-06, | |
| "loss": 0.032, | |
| "step": 1448 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 0.13729214668273926, | |
| "learning_rate": 8.954715367323468e-06, | |
| "loss": 0.0404, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.5808, | |
| "grad_norm": 0.6665918827056885, | |
| "learning_rate": 8.926947188597133e-06, | |
| "loss": 0.2674, | |
| "step": 1452 | |
| }, | |
| { | |
| "epoch": 0.5816, | |
| "grad_norm": 0.11501887440681458, | |
| "learning_rate": 8.89918737775218e-06, | |
| "loss": 0.066, | |
| "step": 1454 | |
| }, | |
| { | |
| "epoch": 0.5824, | |
| "grad_norm": 0.04181879013776779, | |
| "learning_rate": 8.871436151265183e-06, | |
| "loss": 0.0176, | |
| "step": 1456 | |
| }, | |
| { | |
| "epoch": 0.5832, | |
| "grad_norm": 0.07718206197023392, | |
| "learning_rate": 8.843693725545787e-06, | |
| "loss": 0.3476, | |
| "step": 1458 | |
| }, | |
| { | |
| "epoch": 0.584, | |
| "grad_norm": 0.17924383282661438, | |
| "learning_rate": 8.815960316934991e-06, | |
| "loss": 0.0852, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.5848, | |
| "grad_norm": 0.016910772770643234, | |
| "learning_rate": 8.788236141703498e-06, | |
| "loss": 0.0112, | |
| "step": 1462 | |
| }, | |
| { | |
| "epoch": 0.5856, | |
| "grad_norm": 0.20013374090194702, | |
| "learning_rate": 8.760521416049983e-06, | |
| "loss": 0.0349, | |
| "step": 1464 | |
| }, | |
| { | |
| "epoch": 0.5864, | |
| "grad_norm": 0.0546087808907032, | |
| "learning_rate": 8.732816356099455e-06, | |
| "loss": 0.0229, | |
| "step": 1466 | |
| }, | |
| { | |
| "epoch": 0.5872, | |
| "grad_norm": 0.1050657406449318, | |
| "learning_rate": 8.705121177901532e-06, | |
| "loss": 0.0342, | |
| "step": 1468 | |
| }, | |
| { | |
| "epoch": 0.588, | |
| "grad_norm": 0.00835899356752634, | |
| "learning_rate": 8.677436097428775e-06, | |
| "loss": 0.2512, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.5888, | |
| "grad_norm": 0.12563996016979218, | |
| "learning_rate": 8.649761330575009e-06, | |
| "loss": 0.0289, | |
| "step": 1472 | |
| }, | |
| { | |
| "epoch": 0.5896, | |
| "grad_norm": 2.070195436477661, | |
| "learning_rate": 8.62209709315362e-06, | |
| "loss": 0.5447, | |
| "step": 1474 | |
| }, | |
| { | |
| "epoch": 0.5904, | |
| "grad_norm": 0.20926696062088013, | |
| "learning_rate": 8.594443600895892e-06, | |
| "loss": 0.0291, | |
| "step": 1476 | |
| }, | |
| { | |
| "epoch": 0.5912, | |
| "grad_norm": 0.03596784546971321, | |
| "learning_rate": 8.566801069449307e-06, | |
| "loss": 0.0246, | |
| "step": 1478 | |
| }, | |
| { | |
| "epoch": 0.592, | |
| "grad_norm": 0.03978172317147255, | |
| "learning_rate": 8.539169714375885e-06, | |
| "loss": 0.1662, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.5928, | |
| "grad_norm": 0.030970891937613487, | |
| "learning_rate": 8.511549751150478e-06, | |
| "loss": 0.1755, | |
| "step": 1482 | |
| }, | |
| { | |
| "epoch": 0.5936, | |
| "grad_norm": 0.029246820136904716, | |
| "learning_rate": 8.483941395159114e-06, | |
| "loss": 0.0162, | |
| "step": 1484 | |
| }, | |
| { | |
| "epoch": 0.5944, | |
| "grad_norm": 1.2437587976455688, | |
| "learning_rate": 8.45634486169729e-06, | |
| "loss": 0.2451, | |
| "step": 1486 | |
| }, | |
| { | |
| "epoch": 0.5952, | |
| "grad_norm": 0.01518918015062809, | |
| "learning_rate": 8.428760365968327e-06, | |
| "loss": 0.0139, | |
| "step": 1488 | |
| }, | |
| { | |
| "epoch": 0.596, | |
| "grad_norm": 0.058590278029441833, | |
| "learning_rate": 8.401188123081653e-06, | |
| "loss": 0.1158, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.5968, | |
| "grad_norm": 0.09938734769821167, | |
| "learning_rate": 8.373628348051165e-06, | |
| "loss": 0.0162, | |
| "step": 1492 | |
| }, | |
| { | |
| "epoch": 0.5976, | |
| "grad_norm": 0.221556156873703, | |
| "learning_rate": 8.346081255793524e-06, | |
| "loss": 0.0708, | |
| "step": 1494 | |
| }, | |
| { | |
| "epoch": 0.5984, | |
| "grad_norm": 0.9571221470832825, | |
| "learning_rate": 8.318547061126485e-06, | |
| "loss": 0.3115, | |
| "step": 1496 | |
| }, | |
| { | |
| "epoch": 0.5992, | |
| "grad_norm": 0.38105764985084534, | |
| "learning_rate": 8.291025978767236e-06, | |
| "loss": 0.2224, | |
| "step": 1498 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 1.5868642330169678, | |
| "learning_rate": 8.263518223330698e-06, | |
| "loss": 0.2413, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.6008, | |
| "grad_norm": 0.23962773382663727, | |
| "learning_rate": 8.236024009327879e-06, | |
| "loss": 0.0339, | |
| "step": 1502 | |
| }, | |
| { | |
| "epoch": 0.6016, | |
| "grad_norm": 0.5392745137214661, | |
| "learning_rate": 8.208543551164178e-06, | |
| "loss": 0.0817, | |
| "step": 1504 | |
| }, | |
| { | |
| "epoch": 0.6024, | |
| "grad_norm": 0.06242334842681885, | |
| "learning_rate": 8.181077063137733e-06, | |
| "loss": 0.5676, | |
| "step": 1506 | |
| }, | |
| { | |
| "epoch": 0.6032, | |
| "grad_norm": 0.05859392136335373, | |
| "learning_rate": 8.153624759437733e-06, | |
| "loss": 0.0143, | |
| "step": 1508 | |
| }, | |
| { | |
| "epoch": 0.604, | |
| "grad_norm": 0.20578180253505707, | |
| "learning_rate": 8.126186854142752e-06, | |
| "loss": 0.0336, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.6048, | |
| "grad_norm": 2.3235504627227783, | |
| "learning_rate": 8.098763561219101e-06, | |
| "loss": 0.5843, | |
| "step": 1512 | |
| }, | |
| { | |
| "epoch": 0.6056, | |
| "grad_norm": 0.014721088111400604, | |
| "learning_rate": 8.07135509451911e-06, | |
| "loss": 0.0786, | |
| "step": 1514 | |
| }, | |
| { | |
| "epoch": 0.6064, | |
| "grad_norm": 0.029024440795183182, | |
| "learning_rate": 8.04396166777952e-06, | |
| "loss": 0.0238, | |
| "step": 1516 | |
| }, | |
| { | |
| "epoch": 0.6072, | |
| "grad_norm": 0.007215852849185467, | |
| "learning_rate": 8.016583494619769e-06, | |
| "loss": 0.0082, | |
| "step": 1518 | |
| }, | |
| { | |
| "epoch": 0.608, | |
| "grad_norm": 0.8240259885787964, | |
| "learning_rate": 7.989220788540356e-06, | |
| "loss": 0.1647, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.6088, | |
| "grad_norm": 0.3796282708644867, | |
| "learning_rate": 7.961873762921153e-06, | |
| "loss": 0.0807, | |
| "step": 1522 | |
| }, | |
| { | |
| "epoch": 0.6096, | |
| "grad_norm": 0.03532170131802559, | |
| "learning_rate": 7.934542631019767e-06, | |
| "loss": 0.5003, | |
| "step": 1524 | |
| }, | |
| { | |
| "epoch": 0.6104, | |
| "grad_norm": 0.018144795671105385, | |
| "learning_rate": 7.907227605969849e-06, | |
| "loss": 0.0202, | |
| "step": 1526 | |
| }, | |
| { | |
| "epoch": 0.6112, | |
| "grad_norm": 0.06521397829055786, | |
| "learning_rate": 7.879928900779457e-06, | |
| "loss": 0.0214, | |
| "step": 1528 | |
| }, | |
| { | |
| "epoch": 0.612, | |
| "grad_norm": 0.010673885233700275, | |
| "learning_rate": 7.852646728329368e-06, | |
| "loss": 0.2308, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.6128, | |
| "grad_norm": 0.07566576451063156, | |
| "learning_rate": 7.825381301371452e-06, | |
| "loss": 0.0437, | |
| "step": 1532 | |
| }, | |
| { | |
| "epoch": 0.6136, | |
| "grad_norm": 0.01657886430621147, | |
| "learning_rate": 7.798132832526986e-06, | |
| "loss": 0.0092, | |
| "step": 1534 | |
| }, | |
| { | |
| "epoch": 0.6144, | |
| "grad_norm": 0.04421677067875862, | |
| "learning_rate": 7.770901534284996e-06, | |
| "loss": 0.2318, | |
| "step": 1536 | |
| }, | |
| { | |
| "epoch": 0.6152, | |
| "grad_norm": 0.3701765239238739, | |
| "learning_rate": 7.743687619000625e-06, | |
| "loss": 0.0437, | |
| "step": 1538 | |
| }, | |
| { | |
| "epoch": 0.616, | |
| "grad_norm": 0.006094436626881361, | |
| "learning_rate": 7.716491298893443e-06, | |
| "loss": 0.0216, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.6168, | |
| "grad_norm": 0.27293258905410767, | |
| "learning_rate": 7.689312786045823e-06, | |
| "loss": 0.0417, | |
| "step": 1542 | |
| }, | |
| { | |
| "epoch": 0.6176, | |
| "grad_norm": 0.2565436363220215, | |
| "learning_rate": 7.662152292401265e-06, | |
| "loss": 0.0669, | |
| "step": 1544 | |
| }, | |
| { | |
| "epoch": 0.6184, | |
| "grad_norm": 0.9066870212554932, | |
| "learning_rate": 7.635010029762755e-06, | |
| "loss": 0.1723, | |
| "step": 1546 | |
| }, | |
| { | |
| "epoch": 0.6192, | |
| "grad_norm": 0.040010981261730194, | |
| "learning_rate": 7.6078862097911075e-06, | |
| "loss": 0.0133, | |
| "step": 1548 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 0.7351746559143066, | |
| "learning_rate": 7.580781044003324e-06, | |
| "loss": 0.0992, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.6208, | |
| "grad_norm": 0.6215207576751709, | |
| "learning_rate": 7.553694743770928e-06, | |
| "loss": 0.6671, | |
| "step": 1552 | |
| }, | |
| { | |
| "epoch": 0.6216, | |
| "grad_norm": 0.0056650955229997635, | |
| "learning_rate": 7.526627520318329e-06, | |
| "loss": 0.0339, | |
| "step": 1554 | |
| }, | |
| { | |
| "epoch": 0.6224, | |
| "grad_norm": 0.008909706026315689, | |
| "learning_rate": 7.49957958472118e-06, | |
| "loss": 0.0059, | |
| "step": 1556 | |
| }, | |
| { | |
| "epoch": 0.6232, | |
| "grad_norm": 1.9135671854019165, | |
| "learning_rate": 7.472551147904708e-06, | |
| "loss": 0.9156, | |
| "step": 1558 | |
| }, | |
| { | |
| "epoch": 0.624, | |
| "grad_norm": 0.7585664391517639, | |
| "learning_rate": 7.445542420642097e-06, | |
| "loss": 0.103, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.6248, | |
| "grad_norm": 0.07106231898069382, | |
| "learning_rate": 7.418553613552824e-06, | |
| "loss": 0.0217, | |
| "step": 1562 | |
| }, | |
| { | |
| "epoch": 0.6256, | |
| "grad_norm": 1.5444700717926025, | |
| "learning_rate": 7.391584937101034e-06, | |
| "loss": 0.1937, | |
| "step": 1564 | |
| }, | |
| { | |
| "epoch": 0.6264, | |
| "grad_norm": 0.006505718920379877, | |
| "learning_rate": 7.364636601593875e-06, | |
| "loss": 0.014, | |
| "step": 1566 | |
| }, | |
| { | |
| "epoch": 0.6272, | |
| "grad_norm": 0.3871181309223175, | |
| "learning_rate": 7.33770881717989e-06, | |
| "loss": 0.048, | |
| "step": 1568 | |
| }, | |
| { | |
| "epoch": 0.628, | |
| "grad_norm": 0.03062368370592594, | |
| "learning_rate": 7.310801793847344e-06, | |
| "loss": 0.0133, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.6288, | |
| "grad_norm": 0.2106819897890091, | |
| "learning_rate": 7.283915741422611e-06, | |
| "loss": 0.2936, | |
| "step": 1572 | |
| }, | |
| { | |
| "epoch": 0.6296, | |
| "grad_norm": 0.9387131929397583, | |
| "learning_rate": 7.257050869568536e-06, | |
| "loss": 0.3943, | |
| "step": 1574 | |
| }, | |
| { | |
| "epoch": 0.6304, | |
| "grad_norm": 0.18858036398887634, | |
| "learning_rate": 7.2302073877827775e-06, | |
| "loss": 0.2703, | |
| "step": 1576 | |
| }, | |
| { | |
| "epoch": 0.6312, | |
| "grad_norm": 0.34716084599494934, | |
| "learning_rate": 7.203385505396203e-06, | |
| "loss": 0.0523, | |
| "step": 1578 | |
| }, | |
| { | |
| "epoch": 0.632, | |
| "grad_norm": 1.6694183349609375, | |
| "learning_rate": 7.176585431571235e-06, | |
| "loss": 0.4431, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.6328, | |
| "grad_norm": 0.07224322855472565, | |
| "learning_rate": 7.149807375300239e-06, | |
| "loss": 0.1626, | |
| "step": 1582 | |
| }, | |
| { | |
| "epoch": 0.6336, | |
| "grad_norm": 0.03537129983305931, | |
| "learning_rate": 7.123051545403874e-06, | |
| "loss": 0.0113, | |
| "step": 1584 | |
| }, | |
| { | |
| "epoch": 0.6344, | |
| "grad_norm": 0.0862552747130394, | |
| "learning_rate": 7.096318150529476e-06, | |
| "loss": 0.1374, | |
| "step": 1586 | |
| }, | |
| { | |
| "epoch": 0.6352, | |
| "grad_norm": 0.3428609073162079, | |
| "learning_rate": 7.069607399149427e-06, | |
| "loss": 0.0467, | |
| "step": 1588 | |
| }, | |
| { | |
| "epoch": 0.636, | |
| "grad_norm": 0.8518670201301575, | |
| "learning_rate": 7.042919499559538e-06, | |
| "loss": 0.179, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.6368, | |
| "grad_norm": 0.017964590340852737, | |
| "learning_rate": 7.016254659877398e-06, | |
| "loss": 0.0767, | |
| "step": 1592 | |
| }, | |
| { | |
| "epoch": 0.6376, | |
| "grad_norm": 0.042712751775979996, | |
| "learning_rate": 6.9896130880407965e-06, | |
| "loss": 0.0153, | |
| "step": 1594 | |
| }, | |
| { | |
| "epoch": 0.6384, | |
| "grad_norm": 0.3006531894207001, | |
| "learning_rate": 6.962994991806059e-06, | |
| "loss": 0.0977, | |
| "step": 1596 | |
| }, | |
| { | |
| "epoch": 0.6392, | |
| "grad_norm": 0.010738243348896503, | |
| "learning_rate": 6.9364005787464406e-06, | |
| "loss": 0.1063, | |
| "step": 1598 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.017853038385510445, | |
| "learning_rate": 6.909830056250527e-06, | |
| "loss": 0.1487, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.6408, | |
| "grad_norm": 0.38184410333633423, | |
| "learning_rate": 6.883283631520582e-06, | |
| "loss": 0.0403, | |
| "step": 1602 | |
| }, | |
| { | |
| "epoch": 0.6416, | |
| "grad_norm": 0.026657408103346825, | |
| "learning_rate": 6.856761511570963e-06, | |
| "loss": 0.0139, | |
| "step": 1604 | |
| }, | |
| { | |
| "epoch": 0.6424, | |
| "grad_norm": 0.04037747532129288, | |
| "learning_rate": 6.830263903226483e-06, | |
| "loss": 0.0218, | |
| "step": 1606 | |
| }, | |
| { | |
| "epoch": 0.6432, | |
| "grad_norm": 0.07742556184530258, | |
| "learning_rate": 6.803791013120822e-06, | |
| "loss": 0.0117, | |
| "step": 1608 | |
| }, | |
| { | |
| "epoch": 0.644, | |
| "grad_norm": 1.956659197807312, | |
| "learning_rate": 6.777343047694891e-06, | |
| "loss": 0.3377, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.6448, | |
| "grad_norm": 0.04879617318511009, | |
| "learning_rate": 6.750920213195238e-06, | |
| "loss": 0.1139, | |
| "step": 1612 | |
| }, | |
| { | |
| "epoch": 0.6456, | |
| "grad_norm": 3.4786341190338135, | |
| "learning_rate": 6.7245227156724324e-06, | |
| "loss": 0.0851, | |
| "step": 1614 | |
| }, | |
| { | |
| "epoch": 0.6464, | |
| "grad_norm": 0.018991775810718536, | |
| "learning_rate": 6.698150760979463e-06, | |
| "loss": 0.0136, | |
| "step": 1616 | |
| }, | |
| { | |
| "epoch": 0.6472, | |
| "grad_norm": 1.7258912324905396, | |
| "learning_rate": 6.671804554770135e-06, | |
| "loss": 0.3864, | |
| "step": 1618 | |
| }, | |
| { | |
| "epoch": 0.648, | |
| "grad_norm": 0.27930495142936707, | |
| "learning_rate": 6.645484302497452e-06, | |
| "loss": 0.0221, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.6488, | |
| "grad_norm": 0.023989146575331688, | |
| "learning_rate": 6.6191902094120295e-06, | |
| "loss": 0.0198, | |
| "step": 1622 | |
| }, | |
| { | |
| "epoch": 0.6496, | |
| "grad_norm": 0.026416273787617683, | |
| "learning_rate": 6.5929224805604845e-06, | |
| "loss": 0.0195, | |
| "step": 1624 | |
| }, | |
| { | |
| "epoch": 0.6504, | |
| "grad_norm": 0.021918591111898422, | |
| "learning_rate": 6.566681320783849e-06, | |
| "loss": 0.0088, | |
| "step": 1626 | |
| }, | |
| { | |
| "epoch": 0.6512, | |
| "grad_norm": 0.005978676024824381, | |
| "learning_rate": 6.540466934715953e-06, | |
| "loss": 0.0158, | |
| "step": 1628 | |
| }, | |
| { | |
| "epoch": 0.652, | |
| "grad_norm": 0.0690799206495285, | |
| "learning_rate": 6.5142795267818505e-06, | |
| "loss": 0.0426, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.6528, | |
| "grad_norm": 0.06811723858118057, | |
| "learning_rate": 6.488119301196201e-06, | |
| "loss": 0.0474, | |
| "step": 1632 | |
| }, | |
| { | |
| "epoch": 0.6536, | |
| "grad_norm": 0.3594679534435272, | |
| "learning_rate": 6.461986461961706e-06, | |
| "loss": 0.0359, | |
| "step": 1634 | |
| }, | |
| { | |
| "epoch": 0.6544, | |
| "grad_norm": 0.00533846952021122, | |
| "learning_rate": 6.435881212867494e-06, | |
| "loss": 0.0055, | |
| "step": 1636 | |
| }, | |
| { | |
| "epoch": 0.6552, | |
| "grad_norm": 0.4549589455127716, | |
| "learning_rate": 6.409803757487539e-06, | |
| "loss": 0.0631, | |
| "step": 1638 | |
| }, | |
| { | |
| "epoch": 0.656, | |
| "grad_norm": 0.02013424225151539, | |
| "learning_rate": 6.383754299179079e-06, | |
| "loss": 0.0046, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.6568, | |
| "grad_norm": 2.2123377323150635, | |
| "learning_rate": 6.357733041081018e-06, | |
| "loss": 0.482, | |
| "step": 1642 | |
| }, | |
| { | |
| "epoch": 0.6576, | |
| "grad_norm": 0.0015303940745070577, | |
| "learning_rate": 6.33174018611236e-06, | |
| "loss": 0.0028, | |
| "step": 1644 | |
| }, | |
| { | |
| "epoch": 0.6584, | |
| "grad_norm": 0.2125168740749359, | |
| "learning_rate": 6.305775936970606e-06, | |
| "loss": 0.019, | |
| "step": 1646 | |
| }, | |
| { | |
| "epoch": 0.6592, | |
| "grad_norm": 1.2157135009765625, | |
| "learning_rate": 6.27984049613019e-06, | |
| "loss": 0.4542, | |
| "step": 1648 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 0.030942708253860474, | |
| "learning_rate": 6.25393406584088e-06, | |
| "loss": 0.0049, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.6608, | |
| "grad_norm": 0.0280720554292202, | |
| "learning_rate": 6.228056848126236e-06, | |
| "loss": 0.0107, | |
| "step": 1652 | |
| }, | |
| { | |
| "epoch": 0.6616, | |
| "grad_norm": 0.6233726739883423, | |
| "learning_rate": 6.202209044781991e-06, | |
| "loss": 0.0566, | |
| "step": 1654 | |
| }, | |
| { | |
| "epoch": 0.6624, | |
| "grad_norm": 0.014730525203049183, | |
| "learning_rate": 6.176390857374508e-06, | |
| "loss": 0.0088, | |
| "step": 1656 | |
| }, | |
| { | |
| "epoch": 0.6632, | |
| "grad_norm": 1.784373164176941, | |
| "learning_rate": 6.150602487239207e-06, | |
| "loss": 0.7035, | |
| "step": 1658 | |
| }, | |
| { | |
| "epoch": 0.664, | |
| "grad_norm": 0.12886707484722137, | |
| "learning_rate": 6.124844135478971e-06, | |
| "loss": 0.0122, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.6648, | |
| "grad_norm": 0.04210919514298439, | |
| "learning_rate": 6.099116002962604e-06, | |
| "loss": 0.1507, | |
| "step": 1662 | |
| }, | |
| { | |
| "epoch": 0.6656, | |
| "grad_norm": 0.05602734535932541, | |
| "learning_rate": 6.073418290323251e-06, | |
| "loss": 0.0145, | |
| "step": 1664 | |
| }, | |
| { | |
| "epoch": 0.6664, | |
| "grad_norm": 0.04478934034705162, | |
| "learning_rate": 6.047751197956838e-06, | |
| "loss": 0.0079, | |
| "step": 1666 | |
| }, | |
| { | |
| "epoch": 0.6672, | |
| "grad_norm": 0.010290348902344704, | |
| "learning_rate": 6.022114926020504e-06, | |
| "loss": 0.0495, | |
| "step": 1668 | |
| }, | |
| { | |
| "epoch": 0.668, | |
| "grad_norm": 0.01751306839287281, | |
| "learning_rate": 5.996509674431053e-06, | |
| "loss": 0.0056, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.6688, | |
| "grad_norm": 0.1504068821668625, | |
| "learning_rate": 5.970935642863375e-06, | |
| "loss": 0.0234, | |
| "step": 1672 | |
| }, | |
| { | |
| "epoch": 0.6696, | |
| "grad_norm": 0.0935196727514267, | |
| "learning_rate": 5.94539303074891e-06, | |
| "loss": 0.0214, | |
| "step": 1674 | |
| }, | |
| { | |
| "epoch": 0.6704, | |
| "grad_norm": 1.2794703245162964, | |
| "learning_rate": 5.9198820372740726e-06, | |
| "loss": 0.8835, | |
| "step": 1676 | |
| }, | |
| { | |
| "epoch": 0.6712, | |
| "grad_norm": 0.028086047619581223, | |
| "learning_rate": 5.894402861378721e-06, | |
| "loss": 0.0076, | |
| "step": 1678 | |
| }, | |
| { | |
| "epoch": 0.672, | |
| "grad_norm": 0.2483760118484497, | |
| "learning_rate": 5.868955701754584e-06, | |
| "loss": 0.5054, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.6728, | |
| "grad_norm": 0.016343148425221443, | |
| "learning_rate": 5.843540756843722e-06, | |
| "loss": 0.1617, | |
| "step": 1682 | |
| }, | |
| { | |
| "epoch": 0.6736, | |
| "grad_norm": 0.05262403190135956, | |
| "learning_rate": 5.818158224836987e-06, | |
| "loss": 0.0523, | |
| "step": 1684 | |
| }, | |
| { | |
| "epoch": 0.6744, | |
| "grad_norm": 0.11089430004358292, | |
| "learning_rate": 5.792808303672454e-06, | |
| "loss": 0.4788, | |
| "step": 1686 | |
| }, | |
| { | |
| "epoch": 0.6752, | |
| "grad_norm": 0.0768204778432846, | |
| "learning_rate": 5.7674911910339094e-06, | |
| "loss": 0.0263, | |
| "step": 1688 | |
| }, | |
| { | |
| "epoch": 0.676, | |
| "grad_norm": 0.7316690683364868, | |
| "learning_rate": 5.742207084349274e-06, | |
| "loss": 0.064, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.6768, | |
| "grad_norm": 0.12514914572238922, | |
| "learning_rate": 5.716956180789098e-06, | |
| "loss": 0.0287, | |
| "step": 1692 | |
| }, | |
| { | |
| "epoch": 0.6776, | |
| "grad_norm": 0.12199829518795013, | |
| "learning_rate": 5.691738677265e-06, | |
| "loss": 0.1082, | |
| "step": 1694 | |
| }, | |
| { | |
| "epoch": 0.6784, | |
| "grad_norm": 0.15735796093940735, | |
| "learning_rate": 5.666554770428129e-06, | |
| "loss": 0.0281, | |
| "step": 1696 | |
| }, | |
| { | |
| "epoch": 0.6792, | |
| "grad_norm": 2.5249431133270264, | |
| "learning_rate": 5.641404656667661e-06, | |
| "loss": 0.8569, | |
| "step": 1698 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 0.12266776710748672, | |
| "learning_rate": 5.616288532109225e-06, | |
| "loss": 0.0213, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.6808, | |
| "grad_norm": 0.07756412774324417, | |
| "learning_rate": 5.591206592613416e-06, | |
| "loss": 0.02, | |
| "step": 1702 | |
| }, | |
| { | |
| "epoch": 0.6816, | |
| "grad_norm": 0.025011537596583366, | |
| "learning_rate": 5.5661590337742255e-06, | |
| "loss": 0.0081, | |
| "step": 1704 | |
| }, | |
| { | |
| "epoch": 0.6824, | |
| "grad_norm": 0.24261169135570526, | |
| "learning_rate": 5.5411460509175605e-06, | |
| "loss": 0.0702, | |
| "step": 1706 | |
| }, | |
| { | |
| "epoch": 0.6832, | |
| "grad_norm": 0.04961364343762398, | |
| "learning_rate": 5.516167839099679e-06, | |
| "loss": 0.0143, | |
| "step": 1708 | |
| }, | |
| { | |
| "epoch": 0.684, | |
| "grad_norm": 0.0382879376411438, | |
| "learning_rate": 5.491224593105695e-06, | |
| "loss": 0.0174, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.6848, | |
| "grad_norm": 0.024480927735567093, | |
| "learning_rate": 5.466316507448049e-06, | |
| "loss": 0.021, | |
| "step": 1712 | |
| }, | |
| { | |
| "epoch": 0.6856, | |
| "grad_norm": 0.355844646692276, | |
| "learning_rate": 5.441443776365003e-06, | |
| "loss": 0.0201, | |
| "step": 1714 | |
| }, | |
| { | |
| "epoch": 0.6864, | |
| "grad_norm": 0.10836853832006454, | |
| "learning_rate": 5.416606593819102e-06, | |
| "loss": 0.018, | |
| "step": 1716 | |
| }, | |
| { | |
| "epoch": 0.6872, | |
| "grad_norm": 0.1412193775177002, | |
| "learning_rate": 5.391805153495693e-06, | |
| "loss": 0.0284, | |
| "step": 1718 | |
| }, | |
| { | |
| "epoch": 0.688, | |
| "grad_norm": 0.15320326387882233, | |
| "learning_rate": 5.367039648801386e-06, | |
| "loss": 0.0345, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.6888, | |
| "grad_norm": 0.07995325326919556, | |
| "learning_rate": 5.342310272862558e-06, | |
| "loss": 0.0102, | |
| "step": 1722 | |
| }, | |
| { | |
| "epoch": 0.6896, | |
| "grad_norm": 0.04907960444688797, | |
| "learning_rate": 5.317617218523856e-06, | |
| "loss": 0.0391, | |
| "step": 1724 | |
| }, | |
| { | |
| "epoch": 0.6904, | |
| "grad_norm": 1.2691978216171265, | |
| "learning_rate": 5.292960678346674e-06, | |
| "loss": 0.402, | |
| "step": 1726 | |
| }, | |
| { | |
| "epoch": 0.6912, | |
| "grad_norm": 0.008955995552241802, | |
| "learning_rate": 5.26834084460767e-06, | |
| "loss": 0.0053, | |
| "step": 1728 | |
| }, | |
| { | |
| "epoch": 0.692, | |
| "grad_norm": 0.25324296951293945, | |
| "learning_rate": 5.243757909297247e-06, | |
| "loss": 0.035, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.6928, | |
| "grad_norm": 0.026979120448231697, | |
| "learning_rate": 5.219212064118079e-06, | |
| "loss": 0.0089, | |
| "step": 1732 | |
| }, | |
| { | |
| "epoch": 0.6936, | |
| "grad_norm": 0.015840429812669754, | |
| "learning_rate": 5.194703500483593e-06, | |
| "loss": 0.0049, | |
| "step": 1734 | |
| }, | |
| { | |
| "epoch": 0.6944, | |
| "grad_norm": 1.4288071393966675, | |
| "learning_rate": 5.1702324095164955e-06, | |
| "loss": 0.899, | |
| "step": 1736 | |
| }, | |
| { | |
| "epoch": 0.6952, | |
| "grad_norm": 1.0387349128723145, | |
| "learning_rate": 5.145798982047261e-06, | |
| "loss": 0.7296, | |
| "step": 1738 | |
| }, | |
| { | |
| "epoch": 0.696, | |
| "grad_norm": 0.4432249963283539, | |
| "learning_rate": 5.121403408612672e-06, | |
| "loss": 0.0525, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.6968, | |
| "grad_norm": 0.021967697888612747, | |
| "learning_rate": 5.0970458794543135e-06, | |
| "loss": 0.0092, | |
| "step": 1742 | |
| }, | |
| { | |
| "epoch": 0.6976, | |
| "grad_norm": 1.2726808786392212, | |
| "learning_rate": 5.072726584517086e-06, | |
| "loss": 0.1722, | |
| "step": 1744 | |
| }, | |
| { | |
| "epoch": 0.6984, | |
| "grad_norm": 0.01904178597033024, | |
| "learning_rate": 5.048445713447738e-06, | |
| "loss": 0.0209, | |
| "step": 1746 | |
| }, | |
| { | |
| "epoch": 0.6992, | |
| "grad_norm": 0.5619332194328308, | |
| "learning_rate": 5.024203455593375e-06, | |
| "loss": 0.769, | |
| "step": 1748 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 1.2965996265411377, | |
| "learning_rate": 5.000000000000003e-06, | |
| "loss": 0.1219, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.7008, | |
| "grad_norm": 0.08266191929578781, | |
| "learning_rate": 4.97583553541102e-06, | |
| "loss": 0.0208, | |
| "step": 1752 | |
| }, | |
| { | |
| "epoch": 0.7016, | |
| "grad_norm": 0.016945960000157356, | |
| "learning_rate": 4.951710250265785e-06, | |
| "loss": 0.0105, | |
| "step": 1754 | |
| }, | |
| { | |
| "epoch": 0.7024, | |
| "grad_norm": 0.06696458905935287, | |
| "learning_rate": 4.927624332698109e-06, | |
| "loss": 0.0239, | |
| "step": 1756 | |
| }, | |
| { | |
| "epoch": 0.7032, | |
| "grad_norm": 0.024396728724241257, | |
| "learning_rate": 4.903577970534823e-06, | |
| "loss": 0.0168, | |
| "step": 1758 | |
| }, | |
| { | |
| "epoch": 0.704, | |
| "grad_norm": 0.3054494857788086, | |
| "learning_rate": 4.879571351294287e-06, | |
| "loss": 0.0534, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.7048, | |
| "grad_norm": 0.038238685578107834, | |
| "learning_rate": 4.855604662184935e-06, | |
| "loss": 0.3647, | |
| "step": 1762 | |
| }, | |
| { | |
| "epoch": 0.7056, | |
| "grad_norm": 0.23634418845176697, | |
| "learning_rate": 4.831678090103832e-06, | |
| "loss": 0.0938, | |
| "step": 1764 | |
| }, | |
| { | |
| "epoch": 0.7064, | |
| "grad_norm": 0.15138259530067444, | |
| "learning_rate": 4.807791821635186e-06, | |
| "loss": 0.0674, | |
| "step": 1766 | |
| }, | |
| { | |
| "epoch": 0.7072, | |
| "grad_norm": 0.24614623188972473, | |
| "learning_rate": 4.783946043048922e-06, | |
| "loss": 0.0925, | |
| "step": 1768 | |
| }, | |
| { | |
| "epoch": 0.708, | |
| "grad_norm": 2.123792886734009, | |
| "learning_rate": 4.76014094029921e-06, | |
| "loss": 0.9384, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.7088, | |
| "grad_norm": 0.10990063846111298, | |
| "learning_rate": 4.736376699023023e-06, | |
| "loss": 0.0253, | |
| "step": 1772 | |
| }, | |
| { | |
| "epoch": 0.7096, | |
| "grad_norm": 0.3349180221557617, | |
| "learning_rate": 4.712653504538684e-06, | |
| "loss": 0.195, | |
| "step": 1774 | |
| }, | |
| { | |
| "epoch": 0.7104, | |
| "grad_norm": 0.20643967390060425, | |
| "learning_rate": 4.688971541844436e-06, | |
| "loss": 0.0572, | |
| "step": 1776 | |
| }, | |
| { | |
| "epoch": 0.7112, | |
| "grad_norm": 0.09784634411334991, | |
| "learning_rate": 4.6653309956169745e-06, | |
| "loss": 0.0524, | |
| "step": 1778 | |
| }, | |
| { | |
| "epoch": 0.712, | |
| "grad_norm": 0.34553948044776917, | |
| "learning_rate": 4.641732050210032e-06, | |
| "loss": 0.0434, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.7128, | |
| "grad_norm": 1.177024483680725, | |
| "learning_rate": 4.618174889652928e-06, | |
| "loss": 0.2077, | |
| "step": 1782 | |
| }, | |
| { | |
| "epoch": 0.7136, | |
| "grad_norm": 0.10425437986850739, | |
| "learning_rate": 4.59465969764913e-06, | |
| "loss": 0.0273, | |
| "step": 1784 | |
| }, | |
| { | |
| "epoch": 0.7144, | |
| "grad_norm": 0.09616199135780334, | |
| "learning_rate": 4.571186657574828e-06, | |
| "loss": 0.0605, | |
| "step": 1786 | |
| }, | |
| { | |
| "epoch": 0.7152, | |
| "grad_norm": 0.3832583427429199, | |
| "learning_rate": 4.5477559524775e-06, | |
| "loss": 0.0672, | |
| "step": 1788 | |
| }, | |
| { | |
| "epoch": 0.716, | |
| "grad_norm": 0.02167525887489319, | |
| "learning_rate": 4.524367765074499e-06, | |
| "loss": 0.0105, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.7168, | |
| "grad_norm": 0.04896343871951103, | |
| "learning_rate": 4.501022277751602e-06, | |
| "loss": 0.0176, | |
| "step": 1792 | |
| }, | |
| { | |
| "epoch": 0.7176, | |
| "grad_norm": 0.03762371465563774, | |
| "learning_rate": 4.477719672561615e-06, | |
| "loss": 0.0242, | |
| "step": 1794 | |
| }, | |
| { | |
| "epoch": 0.7184, | |
| "grad_norm": 3.2954201698303223, | |
| "learning_rate": 4.4544601312229295e-06, | |
| "loss": 0.3976, | |
| "step": 1796 | |
| }, | |
| { | |
| "epoch": 0.7192, | |
| "grad_norm": 2.2384305000305176, | |
| "learning_rate": 4.4312438351181246e-06, | |
| "loss": 0.2907, | |
| "step": 1798 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.017522338777780533, | |
| "learning_rate": 4.408070965292534e-06, | |
| "loss": 0.4971, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.7208, | |
| "grad_norm": 0.04784021154046059, | |
| "learning_rate": 4.384941702452856e-06, | |
| "loss": 0.6106, | |
| "step": 1802 | |
| }, | |
| { | |
| "epoch": 0.7216, | |
| "grad_norm": 1.3355712890625, | |
| "learning_rate": 4.361856226965733e-06, | |
| "loss": 0.1558, | |
| "step": 1804 | |
| }, | |
| { | |
| "epoch": 0.7224, | |
| "grad_norm": 0.026937812566757202, | |
| "learning_rate": 4.338814718856333e-06, | |
| "loss": 0.0143, | |
| "step": 1806 | |
| }, | |
| { | |
| "epoch": 0.7232, | |
| "grad_norm": 0.14190848171710968, | |
| "learning_rate": 4.315817357806974e-06, | |
| "loss": 0.4526, | |
| "step": 1808 | |
| }, | |
| { | |
| "epoch": 0.724, | |
| "grad_norm": 0.10695420950651169, | |
| "learning_rate": 4.292864323155684e-06, | |
| "loss": 0.0318, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.7248, | |
| "grad_norm": 0.988207221031189, | |
| "learning_rate": 4.26995579389485e-06, | |
| "loss": 0.1699, | |
| "step": 1812 | |
| }, | |
| { | |
| "epoch": 0.7256, | |
| "grad_norm": 0.03641045466065407, | |
| "learning_rate": 4.247091948669775e-06, | |
| "loss": 0.0399, | |
| "step": 1814 | |
| }, | |
| { | |
| "epoch": 0.7264, | |
| "grad_norm": 0.032800428569316864, | |
| "learning_rate": 4.224272965777326e-06, | |
| "loss": 0.015, | |
| "step": 1816 | |
| }, | |
| { | |
| "epoch": 0.7272, | |
| "grad_norm": 0.0268462635576725, | |
| "learning_rate": 4.201499023164508e-06, | |
| "loss": 0.0272, | |
| "step": 1818 | |
| }, | |
| { | |
| "epoch": 0.728, | |
| "grad_norm": 0.05160361900925636, | |
| "learning_rate": 4.178770298427107e-06, | |
| "loss": 0.068, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.7288, | |
| "grad_norm": 1.727393627166748, | |
| "learning_rate": 4.15608696880828e-06, | |
| "loss": 0.3799, | |
| "step": 1822 | |
| }, | |
| { | |
| "epoch": 0.7296, | |
| "grad_norm": 0.08573580533266068, | |
| "learning_rate": 4.133449211197188e-06, | |
| "loss": 0.0329, | |
| "step": 1824 | |
| }, | |
| { | |
| "epoch": 0.7304, | |
| "grad_norm": 0.025401653721928596, | |
| "learning_rate": 4.110857202127615e-06, | |
| "loss": 0.0221, | |
| "step": 1826 | |
| }, | |
| { | |
| "epoch": 0.7312, | |
| "grad_norm": 0.37113156914711, | |
| "learning_rate": 4.08831111777658e-06, | |
| "loss": 0.0818, | |
| "step": 1828 | |
| }, | |
| { | |
| "epoch": 0.732, | |
| "grad_norm": 0.06260918080806732, | |
| "learning_rate": 4.065811133962987e-06, | |
| "loss": 0.0284, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.7328, | |
| "grad_norm": 0.09040958434343338, | |
| "learning_rate": 4.04335742614622e-06, | |
| "loss": 0.0748, | |
| "step": 1832 | |
| }, | |
| { | |
| "epoch": 0.7336, | |
| "grad_norm": 0.14594818651676178, | |
| "learning_rate": 4.020950169424815e-06, | |
| "loss": 0.0702, | |
| "step": 1834 | |
| }, | |
| { | |
| "epoch": 0.7344, | |
| "grad_norm": 0.03950029984116554, | |
| "learning_rate": 3.998589538535046e-06, | |
| "loss": 0.0142, | |
| "step": 1836 | |
| }, | |
| { | |
| "epoch": 0.7352, | |
| "grad_norm": 0.07695923000574112, | |
| "learning_rate": 3.976275707849616e-06, | |
| "loss": 0.0313, | |
| "step": 1838 | |
| }, | |
| { | |
| "epoch": 0.736, | |
| "grad_norm": 0.02458553947508335, | |
| "learning_rate": 3.954008851376252e-06, | |
| "loss": 0.0769, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.7368, | |
| "grad_norm": 0.6668182611465454, | |
| "learning_rate": 3.931789142756377e-06, | |
| "loss": 0.1822, | |
| "step": 1842 | |
| }, | |
| { | |
| "epoch": 0.7376, | |
| "grad_norm": 1.2924126386642456, | |
| "learning_rate": 3.9096167552637454e-06, | |
| "loss": 0.1353, | |
| "step": 1844 | |
| }, | |
| { | |
| "epoch": 0.7384, | |
| "grad_norm": 0.021815890446305275, | |
| "learning_rate": 3.887491861803085e-06, | |
| "loss": 0.0674, | |
| "step": 1846 | |
| }, | |
| { | |
| "epoch": 0.7392, | |
| "grad_norm": 0.043568458408117294, | |
| "learning_rate": 3.86541463490876e-06, | |
| "loss": 0.0284, | |
| "step": 1848 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 0.013898522593080997, | |
| "learning_rate": 3.8433852467434175e-06, | |
| "loss": 0.0049, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.7408, | |
| "grad_norm": 0.03106304258108139, | |
| "learning_rate": 3.821403869096658e-06, | |
| "loss": 0.2808, | |
| "step": 1852 | |
| }, | |
| { | |
| "epoch": 0.7416, | |
| "grad_norm": 0.0395389050245285, | |
| "learning_rate": 3.7994706733836738e-06, | |
| "loss": 0.0115, | |
| "step": 1854 | |
| }, | |
| { | |
| "epoch": 0.7424, | |
| "grad_norm": 0.027892421931028366, | |
| "learning_rate": 3.7775858306439374e-06, | |
| "loss": 0.0126, | |
| "step": 1856 | |
| }, | |
| { | |
| "epoch": 0.7432, | |
| "grad_norm": 0.04609684646129608, | |
| "learning_rate": 3.7557495115398446e-06, | |
| "loss": 0.0117, | |
| "step": 1858 | |
| }, | |
| { | |
| "epoch": 0.744, | |
| "grad_norm": 0.218703493475914, | |
| "learning_rate": 3.7339618863553983e-06, | |
| "loss": 0.042, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 0.7448, | |
| "grad_norm": 0.03932720422744751, | |
| "learning_rate": 3.7122231249948747e-06, | |
| "loss": 0.0253, | |
| "step": 1862 | |
| }, | |
| { | |
| "epoch": 0.7456, | |
| "grad_norm": 0.04857787489891052, | |
| "learning_rate": 3.6905333969815038e-06, | |
| "loss": 0.0178, | |
| "step": 1864 | |
| }, | |
| { | |
| "epoch": 0.7464, | |
| "grad_norm": 0.12067477405071259, | |
| "learning_rate": 3.6688928714561444e-06, | |
| "loss": 0.0707, | |
| "step": 1866 | |
| }, | |
| { | |
| "epoch": 0.7472, | |
| "grad_norm": 2.37345814704895, | |
| "learning_rate": 3.6473017171759563e-06, | |
| "loss": 0.4737, | |
| "step": 1868 | |
| }, | |
| { | |
| "epoch": 0.748, | |
| "grad_norm": 0.13440009951591492, | |
| "learning_rate": 3.625760102513103e-06, | |
| "loss": 0.0347, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 0.7488, | |
| "grad_norm": 3.3364574909210205, | |
| "learning_rate": 3.604268195453421e-06, | |
| "loss": 0.1831, | |
| "step": 1872 | |
| }, | |
| { | |
| "epoch": 0.7496, | |
| "grad_norm": 0.6941204071044922, | |
| "learning_rate": 3.582826163595119e-06, | |
| "loss": 0.0828, | |
| "step": 1874 | |
| }, | |
| { | |
| "epoch": 0.7504, | |
| "grad_norm": 2.122856616973877, | |
| "learning_rate": 3.5614341741474633e-06, | |
| "loss": 0.6934, | |
| "step": 1876 | |
| }, | |
| { | |
| "epoch": 0.7512, | |
| "grad_norm": 1.272242546081543, | |
| "learning_rate": 3.540092393929494e-06, | |
| "loss": 0.1293, | |
| "step": 1878 | |
| }, | |
| { | |
| "epoch": 0.752, | |
| "grad_norm": 0.18253956735134125, | |
| "learning_rate": 3.5188009893686916e-06, | |
| "loss": 0.3014, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 0.7528, | |
| "grad_norm": 0.1802951991558075, | |
| "learning_rate": 3.4975601264997094e-06, | |
| "loss": 0.0245, | |
| "step": 1882 | |
| }, | |
| { | |
| "epoch": 0.7536, | |
| "grad_norm": 0.06385163217782974, | |
| "learning_rate": 3.476369970963072e-06, | |
| "loss": 0.3051, | |
| "step": 1884 | |
| }, | |
| { | |
| "epoch": 0.7544, | |
| "grad_norm": 0.04664032533764839, | |
| "learning_rate": 3.455230688003852e-06, | |
| "loss": 0.0243, | |
| "step": 1886 | |
| }, | |
| { | |
| "epoch": 0.7552, | |
| "grad_norm": 0.09623868763446808, | |
| "learning_rate": 3.4341424424704373e-06, | |
| "loss": 0.0285, | |
| "step": 1888 | |
| }, | |
| { | |
| "epoch": 0.756, | |
| "grad_norm": 0.12530668079853058, | |
| "learning_rate": 3.4131053988131947e-06, | |
| "loss": 0.1282, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 0.7568, | |
| "grad_norm": 0.03179166465997696, | |
| "learning_rate": 3.3921197210832235e-06, | |
| "loss": 0.0231, | |
| "step": 1892 | |
| }, | |
| { | |
| "epoch": 0.7576, | |
| "grad_norm": 0.3028177320957184, | |
| "learning_rate": 3.3711855729310482e-06, | |
| "loss": 0.0521, | |
| "step": 1894 | |
| }, | |
| { | |
| "epoch": 0.7584, | |
| "grad_norm": 0.01785534806549549, | |
| "learning_rate": 3.3503031176053657e-06, | |
| "loss": 0.0962, | |
| "step": 1896 | |
| }, | |
| { | |
| "epoch": 0.7592, | |
| "grad_norm": 0.4004635214805603, | |
| "learning_rate": 3.3294725179517573e-06, | |
| "loss": 0.0461, | |
| "step": 1898 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 0.02001468650996685, | |
| "learning_rate": 3.308693936411421e-06, | |
| "loss": 0.0133, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.7608, | |
| "grad_norm": 0.019235266372561455, | |
| "learning_rate": 3.287967535019908e-06, | |
| "loss": 0.006, | |
| "step": 1902 | |
| }, | |
| { | |
| "epoch": 0.7616, | |
| "grad_norm": 0.0368565134704113, | |
| "learning_rate": 3.2672934754058615e-06, | |
| "loss": 0.0194, | |
| "step": 1904 | |
| }, | |
| { | |
| "epoch": 0.7624, | |
| "grad_norm": 0.045688826590776443, | |
| "learning_rate": 3.2466719187897555e-06, | |
| "loss": 0.0148, | |
| "step": 1906 | |
| }, | |
| { | |
| "epoch": 0.7632, | |
| "grad_norm": 0.1634262502193451, | |
| "learning_rate": 3.2261030259826287e-06, | |
| "loss": 0.0357, | |
| "step": 1908 | |
| }, | |
| { | |
| "epoch": 0.764, | |
| "grad_norm": 0.03429366648197174, | |
| "learning_rate": 3.2055869573848374e-06, | |
| "loss": 0.1542, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 0.7648, | |
| "grad_norm": 0.04100421443581581, | |
| "learning_rate": 3.1851238729848033e-06, | |
| "loss": 0.0103, | |
| "step": 1912 | |
| }, | |
| { | |
| "epoch": 0.7656, | |
| "grad_norm": 0.07986348867416382, | |
| "learning_rate": 3.164713932357776e-06, | |
| "loss": 0.0197, | |
| "step": 1914 | |
| }, | |
| { | |
| "epoch": 0.7664, | |
| "grad_norm": 0.015995606780052185, | |
| "learning_rate": 3.144357294664565e-06, | |
| "loss": 0.0145, | |
| "step": 1916 | |
| }, | |
| { | |
| "epoch": 0.7672, | |
| "grad_norm": 1.615556240081787, | |
| "learning_rate": 3.124054118650327e-06, | |
| "loss": 0.7627, | |
| "step": 1918 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 0.1365402489900589, | |
| "learning_rate": 3.103804562643302e-06, | |
| "loss": 0.04, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 0.7688, | |
| "grad_norm": 0.2867192327976227, | |
| "learning_rate": 3.0836087845536e-06, | |
| "loss": 0.0462, | |
| "step": 1922 | |
| }, | |
| { | |
| "epoch": 0.7696, | |
| "grad_norm": 1.3241535425186157, | |
| "learning_rate": 3.063466941871952e-06, | |
| "loss": 0.6603, | |
| "step": 1924 | |
| }, | |
| { | |
| "epoch": 0.7704, | |
| "grad_norm": 0.1195574626326561, | |
| "learning_rate": 3.043379191668492e-06, | |
| "loss": 0.3676, | |
| "step": 1926 | |
| }, | |
| { | |
| "epoch": 0.7712, | |
| "grad_norm": 1.3540464639663696, | |
| "learning_rate": 3.023345690591537e-06, | |
| "loss": 0.0907, | |
| "step": 1928 | |
| }, | |
| { | |
| "epoch": 0.772, | |
| "grad_norm": 0.03964836522936821, | |
| "learning_rate": 3.003366594866345e-06, | |
| "loss": 0.1992, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 0.7728, | |
| "grad_norm": 0.16762642562389374, | |
| "learning_rate": 2.983442060293926e-06, | |
| "loss": 0.182, | |
| "step": 1932 | |
| }, | |
| { | |
| "epoch": 0.7736, | |
| "grad_norm": 0.02284320630133152, | |
| "learning_rate": 2.963572242249799e-06, | |
| "loss": 0.0122, | |
| "step": 1934 | |
| }, | |
| { | |
| "epoch": 0.7744, | |
| "grad_norm": 0.05577899515628815, | |
| "learning_rate": 2.9437572956827965e-06, | |
| "loss": 0.039, | |
| "step": 1936 | |
| }, | |
| { | |
| "epoch": 0.7752, | |
| "grad_norm": 0.08313470333814621, | |
| "learning_rate": 2.9239973751138495e-06, | |
| "loss": 0.0276, | |
| "step": 1938 | |
| }, | |
| { | |
| "epoch": 0.776, | |
| "grad_norm": 0.06292706727981567, | |
| "learning_rate": 2.9042926346347932e-06, | |
| "loss": 0.0117, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 0.7768, | |
| "grad_norm": 2.362842559814453, | |
| "learning_rate": 2.884643227907147e-06, | |
| "loss": 1.182, | |
| "step": 1942 | |
| }, | |
| { | |
| "epoch": 0.7776, | |
| "grad_norm": 1.178203821182251, | |
| "learning_rate": 2.8650493081609344e-06, | |
| "loss": 0.312, | |
| "step": 1944 | |
| }, | |
| { | |
| "epoch": 0.7784, | |
| "grad_norm": 0.3593979775905609, | |
| "learning_rate": 2.8455110281934804e-06, | |
| "loss": 0.0304, | |
| "step": 1946 | |
| }, | |
| { | |
| "epoch": 0.7792, | |
| "grad_norm": 0.2723308205604553, | |
| "learning_rate": 2.8260285403682153e-06, | |
| "loss": 0.0387, | |
| "step": 1948 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.010247028432786465, | |
| "learning_rate": 2.8066019966134907e-06, | |
| "loss": 0.0151, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.7808, | |
| "grad_norm": 1.3679031133651733, | |
| "learning_rate": 2.7872315484213954e-06, | |
| "loss": 0.1826, | |
| "step": 1952 | |
| }, | |
| { | |
| "epoch": 0.7816, | |
| "grad_norm": 0.024432742968201637, | |
| "learning_rate": 2.7679173468465813e-06, | |
| "loss": 0.0401, | |
| "step": 1954 | |
| }, | |
| { | |
| "epoch": 0.7824, | |
| "grad_norm": 0.026111546903848648, | |
| "learning_rate": 2.7486595425050667e-06, | |
| "loss": 0.2, | |
| "step": 1956 | |
| }, | |
| { | |
| "epoch": 0.7832, | |
| "grad_norm": 1.0953105688095093, | |
| "learning_rate": 2.7294582855730835e-06, | |
| "loss": 0.4462, | |
| "step": 1958 | |
| }, | |
| { | |
| "epoch": 0.784, | |
| "grad_norm": 2.070159673690796, | |
| "learning_rate": 2.7103137257858867e-06, | |
| "loss": 0.5674, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 0.7848, | |
| "grad_norm": 0.26604995131492615, | |
| "learning_rate": 2.6912260124366007e-06, | |
| "loss": 0.0452, | |
| "step": 1962 | |
| }, | |
| { | |
| "epoch": 0.7856, | |
| "grad_norm": 0.9278347492218018, | |
| "learning_rate": 2.672195294375045e-06, | |
| "loss": 0.4562, | |
| "step": 1964 | |
| }, | |
| { | |
| "epoch": 0.7864, | |
| "grad_norm": 0.44804051518440247, | |
| "learning_rate": 2.6532217200065856e-06, | |
| "loss": 0.1851, | |
| "step": 1966 | |
| }, | |
| { | |
| "epoch": 0.7872, | |
| "grad_norm": 1.6803971529006958, | |
| "learning_rate": 2.634305437290968e-06, | |
| "loss": 0.4855, | |
| "step": 1968 | |
| }, | |
| { | |
| "epoch": 0.788, | |
| "grad_norm": 0.14506347477436066, | |
| "learning_rate": 2.615446593741161e-06, | |
| "loss": 0.0346, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 0.7888, | |
| "grad_norm": 0.01994413323700428, | |
| "learning_rate": 2.596645336422219e-06, | |
| "loss": 0.0191, | |
| "step": 1972 | |
| }, | |
| { | |
| "epoch": 0.7896, | |
| "grad_norm": 1.5493515729904175, | |
| "learning_rate": 2.577901811950121e-06, | |
| "loss": 0.4661, | |
| "step": 1974 | |
| }, | |
| { | |
| "epoch": 0.7904, | |
| "grad_norm": 0.20212967693805695, | |
| "learning_rate": 2.5592161664906366e-06, | |
| "loss": 0.1535, | |
| "step": 1976 | |
| }, | |
| { | |
| "epoch": 0.7912, | |
| "grad_norm": 0.20568671822547913, | |
| "learning_rate": 2.5405885457581793e-06, | |
| "loss": 0.0685, | |
| "step": 1978 | |
| }, | |
| { | |
| "epoch": 0.792, | |
| "grad_norm": 0.2657265365123749, | |
| "learning_rate": 2.522019095014683e-06, | |
| "loss": 0.5426, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 0.7928, | |
| "grad_norm": 0.39118292927742004, | |
| "learning_rate": 2.5035079590684496e-06, | |
| "loss": 0.0688, | |
| "step": 1982 | |
| }, | |
| { | |
| "epoch": 0.7936, | |
| "grad_norm": 0.2671131193637848, | |
| "learning_rate": 2.48505528227304e-06, | |
| "loss": 0.0788, | |
| "step": 1984 | |
| }, | |
| { | |
| "epoch": 0.7944, | |
| "grad_norm": 0.055703867226839066, | |
| "learning_rate": 2.4666612085261344e-06, | |
| "loss": 0.0482, | |
| "step": 1986 | |
| }, | |
| { | |
| "epoch": 0.7952, | |
| "grad_norm": 0.03526290878653526, | |
| "learning_rate": 2.4483258812684096e-06, | |
| "loss": 0.0153, | |
| "step": 1988 | |
| }, | |
| { | |
| "epoch": 0.796, | |
| "grad_norm": 1.1341354846954346, | |
| "learning_rate": 2.4300494434824373e-06, | |
| "loss": 0.6923, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 0.7968, | |
| "grad_norm": 0.07913788408041, | |
| "learning_rate": 2.411832037691545e-06, | |
| "loss": 0.0221, | |
| "step": 1992 | |
| }, | |
| { | |
| "epoch": 0.7976, | |
| "grad_norm": 0.10583814233541489, | |
| "learning_rate": 2.3936738059587284e-06, | |
| "loss": 0.0335, | |
| "step": 1994 | |
| }, | |
| { | |
| "epoch": 0.7984, | |
| "grad_norm": 0.019925588741898537, | |
| "learning_rate": 2.37557488988552e-06, | |
| "loss": 0.0187, | |
| "step": 1996 | |
| }, | |
| { | |
| "epoch": 0.7992, | |
| "grad_norm": 0.020932232961058617, | |
| "learning_rate": 2.35753543061091e-06, | |
| "loss": 0.0373, | |
| "step": 1998 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.03722110390663147, | |
| "learning_rate": 2.339555568810221e-06, | |
| "loss": 0.0309, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.8008, | |
| "grad_norm": 0.6024539470672607, | |
| "learning_rate": 2.321635444694028e-06, | |
| "loss": 0.1057, | |
| "step": 2002 | |
| }, | |
| { | |
| "epoch": 0.8016, | |
| "grad_norm": 0.28481170535087585, | |
| "learning_rate": 2.3037751980070557e-06, | |
| "loss": 0.0379, | |
| "step": 2004 | |
| }, | |
| { | |
| "epoch": 0.8024, | |
| "grad_norm": 0.13168196380138397, | |
| "learning_rate": 2.2859749680270983e-06, | |
| "loss": 0.0436, | |
| "step": 2006 | |
| }, | |
| { | |
| "epoch": 0.8032, | |
| "grad_norm": 0.02613903023302555, | |
| "learning_rate": 2.2682348935639274e-06, | |
| "loss": 0.0622, | |
| "step": 2008 | |
| }, | |
| { | |
| "epoch": 0.804, | |
| "grad_norm": 0.3884766399860382, | |
| "learning_rate": 2.2505551129582047e-06, | |
| "loss": 0.0609, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 0.8048, | |
| "grad_norm": 0.014485559426248074, | |
| "learning_rate": 2.2329357640804118e-06, | |
| "loss": 0.024, | |
| "step": 2012 | |
| }, | |
| { | |
| "epoch": 0.8056, | |
| "grad_norm": 0.09305766969919205, | |
| "learning_rate": 2.215376984329767e-06, | |
| "loss": 0.0157, | |
| "step": 2014 | |
| }, | |
| { | |
| "epoch": 0.8064, | |
| "grad_norm": 0.23493258655071259, | |
| "learning_rate": 2.1978789106331666e-06, | |
| "loss": 0.1795, | |
| "step": 2016 | |
| }, | |
| { | |
| "epoch": 0.8072, | |
| "grad_norm": 0.2084125131368637, | |
| "learning_rate": 2.1804416794441e-06, | |
| "loss": 0.0342, | |
| "step": 2018 | |
| }, | |
| { | |
| "epoch": 0.808, | |
| "grad_norm": 0.08282342553138733, | |
| "learning_rate": 2.163065426741603e-06, | |
| "loss": 0.0217, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 0.8088, | |
| "grad_norm": 0.10017091780900955, | |
| "learning_rate": 2.1457502880291815e-06, | |
| "loss": 0.3079, | |
| "step": 2022 | |
| }, | |
| { | |
| "epoch": 0.8096, | |
| "grad_norm": 0.11336953938007355, | |
| "learning_rate": 2.128496398333768e-06, | |
| "loss": 0.0351, | |
| "step": 2024 | |
| }, | |
| { | |
| "epoch": 0.8104, | |
| "grad_norm": 2.011279344558716, | |
| "learning_rate": 2.1113038922046603e-06, | |
| "loss": 0.6162, | |
| "step": 2026 | |
| }, | |
| { | |
| "epoch": 0.8112, | |
| "grad_norm": 0.02046182006597519, | |
| "learning_rate": 2.09417290371247e-06, | |
| "loss": 0.0658, | |
| "step": 2028 | |
| }, | |
| { | |
| "epoch": 0.812, | |
| "grad_norm": 0.0351959727704525, | |
| "learning_rate": 2.0771035664480944e-06, | |
| "loss": 0.7813, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 0.8128, | |
| "grad_norm": 0.031053269281983376, | |
| "learning_rate": 2.0600960135216463e-06, | |
| "loss": 0.0579, | |
| "step": 2032 | |
| }, | |
| { | |
| "epoch": 0.8136, | |
| "grad_norm": 0.027455810457468033, | |
| "learning_rate": 2.0431503775614457e-06, | |
| "loss": 0.0735, | |
| "step": 2034 | |
| }, | |
| { | |
| "epoch": 0.8144, | |
| "grad_norm": 0.030600672587752342, | |
| "learning_rate": 2.026266790712965e-06, | |
| "loss": 0.0096, | |
| "step": 2036 | |
| }, | |
| { | |
| "epoch": 0.8152, | |
| "grad_norm": 0.11856792867183685, | |
| "learning_rate": 2.009445384637805e-06, | |
| "loss": 0.0198, | |
| "step": 2038 | |
| }, | |
| { | |
| "epoch": 0.816, | |
| "grad_norm": 0.0909661203622818, | |
| "learning_rate": 1.9926862905126663e-06, | |
| "loss": 0.0384, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 0.8168, | |
| "grad_norm": 0.008265670388936996, | |
| "learning_rate": 1.9759896390283362e-06, | |
| "loss": 0.092, | |
| "step": 2042 | |
| }, | |
| { | |
| "epoch": 0.8176, | |
| "grad_norm": 0.035349734127521515, | |
| "learning_rate": 1.959355560388654e-06, | |
| "loss": 0.0277, | |
| "step": 2044 | |
| }, | |
| { | |
| "epoch": 0.8184, | |
| "grad_norm": 0.058402419090270996, | |
| "learning_rate": 1.9427841843095063e-06, | |
| "loss": 0.0175, | |
| "step": 2046 | |
| }, | |
| { | |
| "epoch": 0.8192, | |
| "grad_norm": 0.11960252374410629, | |
| "learning_rate": 1.9262756400178163e-06, | |
| "loss": 0.4295, | |
| "step": 2048 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 0.08432687819004059, | |
| "learning_rate": 1.9098300562505266e-06, | |
| "loss": 0.0149, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.8208, | |
| "grad_norm": 1.8159247636795044, | |
| "learning_rate": 1.8934475612536019e-06, | |
| "loss": 0.2313, | |
| "step": 2052 | |
| }, | |
| { | |
| "epoch": 0.8216, | |
| "grad_norm": 0.7769438028335571, | |
| "learning_rate": 1.8771282827810278e-06, | |
| "loss": 0.066, | |
| "step": 2054 | |
| }, | |
| { | |
| "epoch": 0.8224, | |
| "grad_norm": 0.3301757276058197, | |
| "learning_rate": 1.8608723480938207e-06, | |
| "loss": 0.0571, | |
| "step": 2056 | |
| }, | |
| { | |
| "epoch": 0.8232, | |
| "grad_norm": 0.009649352170526981, | |
| "learning_rate": 1.8446798839590186e-06, | |
| "loss": 0.0781, | |
| "step": 2058 | |
| }, | |
| { | |
| "epoch": 0.824, | |
| "grad_norm": 0.014373435638844967, | |
| "learning_rate": 1.8285510166487154e-06, | |
| "loss": 0.0129, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 0.8248, | |
| "grad_norm": 0.08050563931465149, | |
| "learning_rate": 1.812485871939056e-06, | |
| "loss": 0.3065, | |
| "step": 2062 | |
| }, | |
| { | |
| "epoch": 0.8256, | |
| "grad_norm": 0.0180222075432539, | |
| "learning_rate": 1.7964845751092663e-06, | |
| "loss": 0.0089, | |
| "step": 2064 | |
| }, | |
| { | |
| "epoch": 0.8264, | |
| "grad_norm": 0.15466023981571198, | |
| "learning_rate": 1.7805472509406695e-06, | |
| "loss": 0.0321, | |
| "step": 2066 | |
| }, | |
| { | |
| "epoch": 0.8272, | |
| "grad_norm": 0.07331327348947525, | |
| "learning_rate": 1.7646740237157256e-06, | |
| "loss": 0.021, | |
| "step": 2068 | |
| }, | |
| { | |
| "epoch": 0.828, | |
| "grad_norm": 0.1734917163848877, | |
| "learning_rate": 1.7488650172170496e-06, | |
| "loss": 0.0255, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 0.8288, | |
| "grad_norm": 0.023930072784423828, | |
| "learning_rate": 1.7331203547264452e-06, | |
| "loss": 0.0887, | |
| "step": 2072 | |
| }, | |
| { | |
| "epoch": 0.8296, | |
| "grad_norm": 0.027445461601018906, | |
| "learning_rate": 1.7174401590239587e-06, | |
| "loss": 0.1292, | |
| "step": 2074 | |
| }, | |
| { | |
| "epoch": 0.8304, | |
| "grad_norm": 0.07482447475194931, | |
| "learning_rate": 1.7018245523869038e-06, | |
| "loss": 0.0384, | |
| "step": 2076 | |
| }, | |
| { | |
| "epoch": 0.8312, | |
| "grad_norm": 0.03953075036406517, | |
| "learning_rate": 1.686273656588917e-06, | |
| "loss": 0.0082, | |
| "step": 2078 | |
| }, | |
| { | |
| "epoch": 0.832, | |
| "grad_norm": 0.05215131863951683, | |
| "learning_rate": 1.6707875928990059e-06, | |
| "loss": 0.0412, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 0.8328, | |
| "grad_norm": 0.061881113797426224, | |
| "learning_rate": 1.6553664820806102e-06, | |
| "loss": 0.04, | |
| "step": 2082 | |
| }, | |
| { | |
| "epoch": 0.8336, | |
| "grad_norm": 0.1969269961118698, | |
| "learning_rate": 1.6400104443906463e-06, | |
| "loss": 0.0356, | |
| "step": 2084 | |
| }, | |
| { | |
| "epoch": 0.8344, | |
| "grad_norm": 0.0532572865486145, | |
| "learning_rate": 1.6247195995785836e-06, | |
| "loss": 0.0115, | |
| "step": 2086 | |
| }, | |
| { | |
| "epoch": 0.8352, | |
| "grad_norm": 0.07264053821563721, | |
| "learning_rate": 1.6094940668855008e-06, | |
| "loss": 0.0126, | |
| "step": 2088 | |
| }, | |
| { | |
| "epoch": 0.836, | |
| "grad_norm": 1.00359308719635, | |
| "learning_rate": 1.5943339650431578e-06, | |
| "loss": 0.1388, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 0.8368, | |
| "grad_norm": 0.0406976044178009, | |
| "learning_rate": 1.579239412273078e-06, | |
| "loss": 0.0194, | |
| "step": 2092 | |
| }, | |
| { | |
| "epoch": 0.8376, | |
| "grad_norm": 0.02532949112355709, | |
| "learning_rate": 1.5642105262856122e-06, | |
| "loss": 0.0081, | |
| "step": 2094 | |
| }, | |
| { | |
| "epoch": 0.8384, | |
| "grad_norm": 0.10549493879079819, | |
| "learning_rate": 1.5492474242790368e-06, | |
| "loss": 0.075, | |
| "step": 2096 | |
| }, | |
| { | |
| "epoch": 0.8392, | |
| "grad_norm": 0.029608091339468956, | |
| "learning_rate": 1.5343502229386209e-06, | |
| "loss": 0.022, | |
| "step": 2098 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 2.5157110691070557, | |
| "learning_rate": 1.5195190384357405e-06, | |
| "loss": 0.1403, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.8408, | |
| "grad_norm": 0.13249818980693817, | |
| "learning_rate": 1.5047539864269477e-06, | |
| "loss": 0.1111, | |
| "step": 2102 | |
| }, | |
| { | |
| "epoch": 0.8416, | |
| "grad_norm": 0.09822747856378555, | |
| "learning_rate": 1.490055182053083e-06, | |
| "loss": 0.0193, | |
| "step": 2104 | |
| }, | |
| { | |
| "epoch": 0.8424, | |
| "grad_norm": 0.7339903712272644, | |
| "learning_rate": 1.4754227399383758e-06, | |
| "loss": 0.0544, | |
| "step": 2106 | |
| }, | |
| { | |
| "epoch": 0.8432, | |
| "grad_norm": 0.03530171886086464, | |
| "learning_rate": 1.4608567741895496e-06, | |
| "loss": 0.0135, | |
| "step": 2108 | |
| }, | |
| { | |
| "epoch": 0.844, | |
| "grad_norm": 0.5551167726516724, | |
| "learning_rate": 1.446357398394934e-06, | |
| "loss": 0.1178, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 0.8448, | |
| "grad_norm": 0.15902993083000183, | |
| "learning_rate": 1.4319247256235713e-06, | |
| "loss": 0.1147, | |
| "step": 2112 | |
| }, | |
| { | |
| "epoch": 0.8456, | |
| "grad_norm": 0.7204644083976746, | |
| "learning_rate": 1.4175588684243447e-06, | |
| "loss": 0.1095, | |
| "step": 2114 | |
| }, | |
| { | |
| "epoch": 0.8464, | |
| "grad_norm": 0.05895598977804184, | |
| "learning_rate": 1.40325993882509e-06, | |
| "loss": 0.0297, | |
| "step": 2116 | |
| }, | |
| { | |
| "epoch": 0.8472, | |
| "grad_norm": 0.9869866371154785, | |
| "learning_rate": 1.3890280483317375e-06, | |
| "loss": 0.1489, | |
| "step": 2118 | |
| }, | |
| { | |
| "epoch": 0.848, | |
| "grad_norm": 0.005029854364693165, | |
| "learning_rate": 1.3748633079274254e-06, | |
| "loss": 0.1327, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 0.8488, | |
| "grad_norm": 0.14541175961494446, | |
| "learning_rate": 1.3607658280716474e-06, | |
| "loss": 0.1304, | |
| "step": 2122 | |
| }, | |
| { | |
| "epoch": 0.8496, | |
| "grad_norm": 0.051523029804229736, | |
| "learning_rate": 1.3467357186993802e-06, | |
| "loss": 0.0182, | |
| "step": 2124 | |
| }, | |
| { | |
| "epoch": 0.8504, | |
| "grad_norm": 0.13095594942569733, | |
| "learning_rate": 1.3327730892202384e-06, | |
| "loss": 0.4018, | |
| "step": 2126 | |
| }, | |
| { | |
| "epoch": 0.8512, | |
| "grad_norm": 0.07890293002128601, | |
| "learning_rate": 1.3188780485176089e-06, | |
| "loss": 0.0119, | |
| "step": 2128 | |
| }, | |
| { | |
| "epoch": 0.852, | |
| "grad_norm": 0.0385555699467659, | |
| "learning_rate": 1.30505070494781e-06, | |
| "loss": 0.0285, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 0.8528, | |
| "grad_norm": 0.022745564579963684, | |
| "learning_rate": 1.2912911663392468e-06, | |
| "loss": 0.0219, | |
| "step": 2132 | |
| }, | |
| { | |
| "epoch": 0.8536, | |
| "grad_norm": 0.04321039468050003, | |
| "learning_rate": 1.277599539991563e-06, | |
| "loss": 0.061, | |
| "step": 2134 | |
| }, | |
| { | |
| "epoch": 0.8544, | |
| "grad_norm": 0.014432685449719429, | |
| "learning_rate": 1.2639759326748136e-06, | |
| "loss": 0.0409, | |
| "step": 2136 | |
| }, | |
| { | |
| "epoch": 0.8552, | |
| "grad_norm": 0.019150329753756523, | |
| "learning_rate": 1.2504204506286244e-06, | |
| "loss": 0.0051, | |
| "step": 2138 | |
| }, | |
| { | |
| "epoch": 0.856, | |
| "grad_norm": 1.3626798391342163, | |
| "learning_rate": 1.2369331995613664e-06, | |
| "loss": 0.1418, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 0.8568, | |
| "grad_norm": 0.02268528752028942, | |
| "learning_rate": 1.223514284649331e-06, | |
| "loss": 0.2405, | |
| "step": 2142 | |
| }, | |
| { | |
| "epoch": 0.8576, | |
| "grad_norm": 0.5977320671081543, | |
| "learning_rate": 1.210163810535917e-06, | |
| "loss": 0.0792, | |
| "step": 2144 | |
| }, | |
| { | |
| "epoch": 0.8584, | |
| "grad_norm": 0.01674368605017662, | |
| "learning_rate": 1.196881881330798e-06, | |
| "loss": 0.0047, | |
| "step": 2146 | |
| }, | |
| { | |
| "epoch": 0.8592, | |
| "grad_norm": 0.22216765582561493, | |
| "learning_rate": 1.1836686006091313e-06, | |
| "loss": 0.4924, | |
| "step": 2148 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 0.04747781902551651, | |
| "learning_rate": 1.1705240714107301e-06, | |
| "loss": 0.0192, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.8608, | |
| "grad_norm": 0.12776315212249756, | |
| "learning_rate": 1.1574483962392768e-06, | |
| "loss": 0.031, | |
| "step": 2152 | |
| }, | |
| { | |
| "epoch": 0.8616, | |
| "grad_norm": 0.022700203582644463, | |
| "learning_rate": 1.1444416770615118e-06, | |
| "loss": 0.0062, | |
| "step": 2154 | |
| }, | |
| { | |
| "epoch": 0.8624, | |
| "grad_norm": 0.07675009965896606, | |
| "learning_rate": 1.1315040153064416e-06, | |
| "loss": 0.0334, | |
| "step": 2156 | |
| }, | |
| { | |
| "epoch": 0.8632, | |
| "grad_norm": 0.0223082285374403, | |
| "learning_rate": 1.1186355118645552e-06, | |
| "loss": 0.0168, | |
| "step": 2158 | |
| }, | |
| { | |
| "epoch": 0.864, | |
| "grad_norm": 0.034395378082990646, | |
| "learning_rate": 1.1058362670870248e-06, | |
| "loss": 0.0086, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 0.8648, | |
| "grad_norm": 0.015471206046640873, | |
| "learning_rate": 1.093106380784934e-06, | |
| "loss": 0.1392, | |
| "step": 2162 | |
| }, | |
| { | |
| "epoch": 0.8656, | |
| "grad_norm": 0.013013385236263275, | |
| "learning_rate": 1.0804459522284927e-06, | |
| "loss": 0.0584, | |
| "step": 2164 | |
| }, | |
| { | |
| "epoch": 0.8664, | |
| "grad_norm": 0.03496384248137474, | |
| "learning_rate": 1.0678550801462662e-06, | |
| "loss": 0.0052, | |
| "step": 2166 | |
| }, | |
| { | |
| "epoch": 0.8672, | |
| "grad_norm": 0.25604620575904846, | |
| "learning_rate": 1.0553338627244026e-06, | |
| "loss": 0.0329, | |
| "step": 2168 | |
| }, | |
| { | |
| "epoch": 0.868, | |
| "grad_norm": 1.7641575336456299, | |
| "learning_rate": 1.042882397605871e-06, | |
| "loss": 0.1352, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 0.8688, | |
| "grad_norm": 0.083366259932518, | |
| "learning_rate": 1.0305007818897006e-06, | |
| "loss": 0.0171, | |
| "step": 2172 | |
| }, | |
| { | |
| "epoch": 0.8696, | |
| "grad_norm": 1.2676173448562622, | |
| "learning_rate": 1.0181891121302145e-06, | |
| "loss": 1.594, | |
| "step": 2174 | |
| }, | |
| { | |
| "epoch": 0.8704, | |
| "grad_norm": 0.604960560798645, | |
| "learning_rate": 1.0059474843362893e-06, | |
| "loss": 0.2283, | |
| "step": 2176 | |
| }, | |
| { | |
| "epoch": 0.8712, | |
| "grad_norm": 0.16328264772891998, | |
| "learning_rate": 9.93775993970597e-07, | |
| "loss": 0.0237, | |
| "step": 2178 | |
| }, | |
| { | |
| "epoch": 0.872, | |
| "grad_norm": 0.01555766724050045, | |
| "learning_rate": 9.816747359488632e-07, | |
| "loss": 0.1311, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 0.8728, | |
| "grad_norm": 0.6382170915603638, | |
| "learning_rate": 9.696438046391288e-07, | |
| "loss": 0.0781, | |
| "step": 2182 | |
| }, | |
| { | |
| "epoch": 0.8736, | |
| "grad_norm": 0.1925554871559143, | |
| "learning_rate": 9.576832938610137e-07, | |
| "loss": 0.057, | |
| "step": 2184 | |
| }, | |
| { | |
| "epoch": 0.8744, | |
| "grad_norm": 0.08132878690958023, | |
| "learning_rate": 9.457932968849826e-07, | |
| "loss": 0.0189, | |
| "step": 2186 | |
| }, | |
| { | |
| "epoch": 0.8752, | |
| "grad_norm": 0.036015480756759644, | |
| "learning_rate": 9.339739064316233e-07, | |
| "loss": 0.1586, | |
| "step": 2188 | |
| }, | |
| { | |
| "epoch": 0.876, | |
| "grad_norm": 1.3472023010253906, | |
| "learning_rate": 9.222252146709143e-07, | |
| "loss": 0.4576, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 0.8768, | |
| "grad_norm": 2.6704301834106445, | |
| "learning_rate": 9.105473132215126e-07, | |
| "loss": 0.1486, | |
| "step": 2192 | |
| }, | |
| { | |
| "epoch": 0.8776, | |
| "grad_norm": 0.029902072623372078, | |
| "learning_rate": 8.989402931500434e-07, | |
| "loss": 0.032, | |
| "step": 2194 | |
| }, | |
| { | |
| "epoch": 0.8784, | |
| "grad_norm": 0.036093611270189285, | |
| "learning_rate": 8.874042449703779e-07, | |
| "loss": 0.0153, | |
| "step": 2196 | |
| }, | |
| { | |
| "epoch": 0.8792, | |
| "grad_norm": 0.6965168118476868, | |
| "learning_rate": 8.759392586429394e-07, | |
| "loss": 0.0854, | |
| "step": 2198 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.3597988188266754, | |
| "learning_rate": 8.645454235739903e-07, | |
| "loss": 0.0457, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.8808, | |
| "grad_norm": 0.15428324043750763, | |
| "learning_rate": 8.532228286149502e-07, | |
| "loss": 0.0202, | |
| "step": 2202 | |
| }, | |
| { | |
| "epoch": 0.8816, | |
| "grad_norm": 0.1308208853006363, | |
| "learning_rate": 8.419715620616875e-07, | |
| "loss": 0.0235, | |
| "step": 2204 | |
| }, | |
| { | |
| "epoch": 0.8824, | |
| "grad_norm": 0.8089284300804138, | |
| "learning_rate": 8.307917116538378e-07, | |
| "loss": 0.0771, | |
| "step": 2206 | |
| }, | |
| { | |
| "epoch": 0.8832, | |
| "grad_norm": 1.2626261711120605, | |
| "learning_rate": 8.196833645741187e-07, | |
| "loss": 0.6288, | |
| "step": 2208 | |
| }, | |
| { | |
| "epoch": 0.884, | |
| "grad_norm": 0.0954088419675827, | |
| "learning_rate": 8.086466074476562e-07, | |
| "loss": 0.0133, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 0.8848, | |
| "grad_norm": 0.13527199625968933, | |
| "learning_rate": 7.976815263412963e-07, | |
| "loss": 0.0211, | |
| "step": 2212 | |
| }, | |
| { | |
| "epoch": 0.8856, | |
| "grad_norm": 0.5545380115509033, | |
| "learning_rate": 7.867882067629473e-07, | |
| "loss": 0.0581, | |
| "step": 2214 | |
| }, | |
| { | |
| "epoch": 0.8864, | |
| "grad_norm": 0.06293027848005295, | |
| "learning_rate": 7.759667336609011e-07, | |
| "loss": 0.0264, | |
| "step": 2216 | |
| }, | |
| { | |
| "epoch": 0.8872, | |
| "grad_norm": 0.3431554138660431, | |
| "learning_rate": 7.652171914231777e-07, | |
| "loss": 0.0329, | |
| "step": 2218 | |
| }, | |
| { | |
| "epoch": 0.888, | |
| "grad_norm": 0.00832755770534277, | |
| "learning_rate": 7.545396638768698e-07, | |
| "loss": 0.0099, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 0.8888, | |
| "grad_norm": 2.4031519889831543, | |
| "learning_rate": 7.439342342874789e-07, | |
| "loss": 0.1471, | |
| "step": 2222 | |
| }, | |
| { | |
| "epoch": 0.8896, | |
| "grad_norm": 0.06904103606939316, | |
| "learning_rate": 7.334009853582791e-07, | |
| "loss": 0.0304, | |
| "step": 2224 | |
| }, | |
| { | |
| "epoch": 0.8904, | |
| "grad_norm": 0.24774344265460968, | |
| "learning_rate": 7.22939999229657e-07, | |
| "loss": 0.0465, | |
| "step": 2226 | |
| }, | |
| { | |
| "epoch": 0.8912, | |
| "grad_norm": 1.129273533821106, | |
| "learning_rate": 7.125513574784904e-07, | |
| "loss": 0.1176, | |
| "step": 2228 | |
| }, | |
| { | |
| "epoch": 0.892, | |
| "grad_norm": 0.13521403074264526, | |
| "learning_rate": 7.022351411174866e-07, | |
| "loss": 0.0401, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 0.8928, | |
| "grad_norm": 0.025177139788866043, | |
| "learning_rate": 6.919914305945774e-07, | |
| "loss": 0.0051, | |
| "step": 2232 | |
| }, | |
| { | |
| "epoch": 0.8936, | |
| "grad_norm": 2.7186474800109863, | |
| "learning_rate": 6.818203057922756e-07, | |
| "loss": 0.6471, | |
| "step": 2234 | |
| }, | |
| { | |
| "epoch": 0.8944, | |
| "grad_norm": 0.5871244072914124, | |
| "learning_rate": 6.717218460270536e-07, | |
| "loss": 0.0628, | |
| "step": 2236 | |
| }, | |
| { | |
| "epoch": 0.8952, | |
| "grad_norm": 0.3472304344177246, | |
| "learning_rate": 6.616961300487323e-07, | |
| "loss": 0.0305, | |
| "step": 2238 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 0.04057146608829498, | |
| "learning_rate": 6.517432360398556e-07, | |
| "loss": 0.7828, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 0.8968, | |
| "grad_norm": 0.03729462996125221, | |
| "learning_rate": 6.418632416150927e-07, | |
| "loss": 0.006, | |
| "step": 2242 | |
| }, | |
| { | |
| "epoch": 0.8976, | |
| "grad_norm": 0.022776370868086815, | |
| "learning_rate": 6.320562238206218e-07, | |
| "loss": 0.0087, | |
| "step": 2244 | |
| }, | |
| { | |
| "epoch": 0.8984, | |
| "grad_norm": 0.06814192980527878, | |
| "learning_rate": 6.223222591335409e-07, | |
| "loss": 0.0067, | |
| "step": 2246 | |
| }, | |
| { | |
| "epoch": 0.8992, | |
| "grad_norm": 0.012981448322534561, | |
| "learning_rate": 6.126614234612593e-07, | |
| "loss": 0.0273, | |
| "step": 2248 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.7651333212852478, | |
| "learning_rate": 6.030737921409169e-07, | |
| "loss": 0.1298, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.9008, | |
| "grad_norm": 0.5903016924858093, | |
| "learning_rate": 5.935594399387856e-07, | |
| "loss": 0.0611, | |
| "step": 2252 | |
| }, | |
| { | |
| "epoch": 0.9016, | |
| "grad_norm": 0.06699836999177933, | |
| "learning_rate": 5.841184410496992e-07, | |
| "loss": 0.0083, | |
| "step": 2254 | |
| }, | |
| { | |
| "epoch": 0.9024, | |
| "grad_norm": 0.124520443379879, | |
| "learning_rate": 5.747508690964599e-07, | |
| "loss": 0.3001, | |
| "step": 2256 | |
| }, | |
| { | |
| "epoch": 0.9032, | |
| "grad_norm": 0.16008125245571136, | |
| "learning_rate": 5.654567971292757e-07, | |
| "loss": 0.0201, | |
| "step": 2258 | |
| }, | |
| { | |
| "epoch": 0.904, | |
| "grad_norm": 0.07481832057237625, | |
| "learning_rate": 5.562362976251901e-07, | |
| "loss": 0.0851, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 0.9048, | |
| "grad_norm": 0.6332103610038757, | |
| "learning_rate": 5.470894424875062e-07, | |
| "loss": 0.1097, | |
| "step": 2262 | |
| }, | |
| { | |
| "epoch": 0.9056, | |
| "grad_norm": 0.014627913013100624, | |
| "learning_rate": 5.380163030452412e-07, | |
| "loss": 0.0236, | |
| "step": 2264 | |
| }, | |
| { | |
| "epoch": 0.9064, | |
| "grad_norm": 0.6561143398284912, | |
| "learning_rate": 5.290169500525577e-07, | |
| "loss": 0.0778, | |
| "step": 2266 | |
| }, | |
| { | |
| "epoch": 0.9072, | |
| "grad_norm": 0.017556993290781975, | |
| "learning_rate": 5.200914536882184e-07, | |
| "loss": 0.0088, | |
| "step": 2268 | |
| }, | |
| { | |
| "epoch": 0.908, | |
| "grad_norm": 0.11898583173751831, | |
| "learning_rate": 5.112398835550348e-07, | |
| "loss": 0.0246, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 0.9088, | |
| "grad_norm": 1.1768893003463745, | |
| "learning_rate": 5.024623086793323e-07, | |
| "loss": 0.0801, | |
| "step": 2272 | |
| }, | |
| { | |
| "epoch": 0.9096, | |
| "grad_norm": 0.29487040638923645, | |
| "learning_rate": 4.937587975103997e-07, | |
| "loss": 0.0533, | |
| "step": 2274 | |
| }, | |
| { | |
| "epoch": 0.9104, | |
| "grad_norm": 1.571694016456604, | |
| "learning_rate": 4.851294179199673e-07, | |
| "loss": 0.65, | |
| "step": 2276 | |
| }, | |
| { | |
| "epoch": 0.9112, | |
| "grad_norm": 0.01853368431329727, | |
| "learning_rate": 4.765742372016735e-07, | |
| "loss": 0.0912, | |
| "step": 2278 | |
| }, | |
| { | |
| "epoch": 0.912, | |
| "grad_norm": 1.2929413318634033, | |
| "learning_rate": 4.6809332207053083e-07, | |
| "loss": 0.0748, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 0.9128, | |
| "grad_norm": 0.019635476171970367, | |
| "learning_rate": 4.596867386624215e-07, | |
| "loss": 0.0165, | |
| "step": 2282 | |
| }, | |
| { | |
| "epoch": 0.9136, | |
| "grad_norm": 0.04413539171218872, | |
| "learning_rate": 4.5135455253357053e-07, | |
| "loss": 0.0407, | |
| "step": 2284 | |
| }, | |
| { | |
| "epoch": 0.9144, | |
| "grad_norm": 0.022891348227858543, | |
| "learning_rate": 4.4309682866004124e-07, | |
| "loss": 0.0352, | |
| "step": 2286 | |
| }, | |
| { | |
| "epoch": 0.9152, | |
| "grad_norm": 0.0767710953950882, | |
| "learning_rate": 4.349136314372204e-07, | |
| "loss": 0.0121, | |
| "step": 2288 | |
| }, | |
| { | |
| "epoch": 0.916, | |
| "grad_norm": 0.13404303789138794, | |
| "learning_rate": 4.268050246793276e-07, | |
| "loss": 0.0556, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 0.9168, | |
| "grad_norm": 0.4251805543899536, | |
| "learning_rate": 4.1877107161890416e-07, | |
| "loss": 0.1859, | |
| "step": 2292 | |
| }, | |
| { | |
| "epoch": 0.9176, | |
| "grad_norm": 1.6532477140426636, | |
| "learning_rate": 4.108118349063306e-07, | |
| "loss": 0.6478, | |
| "step": 2294 | |
| }, | |
| { | |
| "epoch": 0.9184, | |
| "grad_norm": 0.1750348061323166, | |
| "learning_rate": 4.0292737660933335e-07, | |
| "loss": 0.0384, | |
| "step": 2296 | |
| }, | |
| { | |
| "epoch": 0.9192, | |
| "grad_norm": 1.260365605354309, | |
| "learning_rate": 3.9511775821250206e-07, | |
| "loss": 0.8857, | |
| "step": 2298 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 0.05385835841298103, | |
| "learning_rate": 3.8738304061681107e-07, | |
| "loss": 0.1621, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.9208, | |
| "grad_norm": 0.4076821208000183, | |
| "learning_rate": 3.7972328413914074e-07, | |
| "loss": 0.1319, | |
| "step": 2302 | |
| }, | |
| { | |
| "epoch": 0.9216, | |
| "grad_norm": 0.42220211029052734, | |
| "learning_rate": 3.721385485118123e-07, | |
| "loss": 0.2582, | |
| "step": 2304 | |
| }, | |
| { | |
| "epoch": 0.9224, | |
| "grad_norm": 0.2544540762901306, | |
| "learning_rate": 3.646288928821151e-07, | |
| "loss": 0.0197, | |
| "step": 2306 | |
| }, | |
| { | |
| "epoch": 0.9232, | |
| "grad_norm": 1.2406009435653687, | |
| "learning_rate": 3.571943758118546e-07, | |
| "loss": 0.8589, | |
| "step": 2308 | |
| }, | |
| { | |
| "epoch": 0.924, | |
| "grad_norm": 0.11001749336719513, | |
| "learning_rate": 3.498350552768859e-07, | |
| "loss": 0.0317, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 0.9248, | |
| "grad_norm": 0.5791686177253723, | |
| "learning_rate": 3.4255098866667114e-07, | |
| "loss": 0.0695, | |
| "step": 2312 | |
| }, | |
| { | |
| "epoch": 0.9256, | |
| "grad_norm": 0.005229198839515448, | |
| "learning_rate": 3.3534223278382405e-07, | |
| "loss": 0.0093, | |
| "step": 2314 | |
| }, | |
| { | |
| "epoch": 0.9264, | |
| "grad_norm": 0.09191671758890152, | |
| "learning_rate": 3.282088438436715e-07, | |
| "loss": 0.0201, | |
| "step": 2316 | |
| }, | |
| { | |
| "epoch": 0.9272, | |
| "grad_norm": 0.32510700821876526, | |
| "learning_rate": 3.211508774738137e-07, | |
| "loss": 0.039, | |
| "step": 2318 | |
| }, | |
| { | |
| "epoch": 0.928, | |
| "grad_norm": 0.027174528688192368, | |
| "learning_rate": 3.1416838871368925e-07, | |
| "loss": 0.0191, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 0.9288, | |
| "grad_norm": 0.0871250182390213, | |
| "learning_rate": 3.072614320141487e-07, | |
| "loss": 0.0112, | |
| "step": 2322 | |
| }, | |
| { | |
| "epoch": 0.9296, | |
| "grad_norm": 0.11660738289356232, | |
| "learning_rate": 3.00430061237027e-07, | |
| "loss": 0.0281, | |
| "step": 2324 | |
| }, | |
| { | |
| "epoch": 0.9304, | |
| "grad_norm": 0.009191269055008888, | |
| "learning_rate": 2.936743296547273e-07, | |
| "loss": 0.0221, | |
| "step": 2326 | |
| }, | |
| { | |
| "epoch": 0.9312, | |
| "grad_norm": 0.05838488042354584, | |
| "learning_rate": 2.8699428994980017e-07, | |
| "loss": 0.9408, | |
| "step": 2328 | |
| }, | |
| { | |
| "epoch": 0.932, | |
| "grad_norm": 1.4220937490463257, | |
| "learning_rate": 2.8038999421453827e-07, | |
| "loss": 0.0656, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 0.9328, | |
| "grad_norm": 0.08184941858053207, | |
| "learning_rate": 2.7386149395056463e-07, | |
| "loss": 0.4858, | |
| "step": 2332 | |
| }, | |
| { | |
| "epoch": 0.9336, | |
| "grad_norm": 0.7450655102729797, | |
| "learning_rate": 2.6740884006843826e-07, | |
| "loss": 0.0379, | |
| "step": 2334 | |
| }, | |
| { | |
| "epoch": 0.9344, | |
| "grad_norm": 1.6832029819488525, | |
| "learning_rate": 2.6103208288724815e-07, | |
| "loss": 0.6098, | |
| "step": 2336 | |
| }, | |
| { | |
| "epoch": 0.9352, | |
| "grad_norm": 0.09961254149675369, | |
| "learning_rate": 2.547312721342277e-07, | |
| "loss": 0.0168, | |
| "step": 2338 | |
| }, | |
| { | |
| "epoch": 0.936, | |
| "grad_norm": 0.10063952207565308, | |
| "learning_rate": 2.4850645694436736e-07, | |
| "loss": 0.0119, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 0.9368, | |
| "grad_norm": 0.06440629810094833, | |
| "learning_rate": 2.423576858600252e-07, | |
| "loss": 0.0119, | |
| "step": 2342 | |
| }, | |
| { | |
| "epoch": 0.9376, | |
| "grad_norm": 0.01774456538259983, | |
| "learning_rate": 2.3628500683055222e-07, | |
| "loss": 0.0081, | |
| "step": 2344 | |
| }, | |
| { | |
| "epoch": 0.9384, | |
| "grad_norm": 1.3639814853668213, | |
| "learning_rate": 2.3028846721191878e-07, | |
| "loss": 0.1012, | |
| "step": 2346 | |
| }, | |
| { | |
| "epoch": 0.9392, | |
| "grad_norm": 0.010448544286191463, | |
| "learning_rate": 2.2436811376634893e-07, | |
| "loss": 0.0061, | |
| "step": 2348 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 0.05712028965353966, | |
| "learning_rate": 2.1852399266194312e-07, | |
| "loss": 0.0325, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.9408, | |
| "grad_norm": 0.027819665148854256, | |
| "learning_rate": 2.1275614947233624e-07, | |
| "loss": 0.0963, | |
| "step": 2352 | |
| }, | |
| { | |
| "epoch": 0.9416, | |
| "grad_norm": 1.7772072553634644, | |
| "learning_rate": 2.0706462917632676e-07, | |
| "loss": 0.8716, | |
| "step": 2354 | |
| }, | |
| { | |
| "epoch": 0.9424, | |
| "grad_norm": 0.47964930534362793, | |
| "learning_rate": 2.014494761575314e-07, | |
| "loss": 0.0422, | |
| "step": 2356 | |
| }, | |
| { | |
| "epoch": 0.9432, | |
| "grad_norm": 0.0034853648394346237, | |
| "learning_rate": 1.9591073420404338e-07, | |
| "loss": 0.0195, | |
| "step": 2358 | |
| }, | |
| { | |
| "epoch": 0.944, | |
| "grad_norm": 0.03955305367708206, | |
| "learning_rate": 1.9044844650808468e-07, | |
| "loss": 0.0275, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 0.9448, | |
| "grad_norm": 0.5900484919548035, | |
| "learning_rate": 1.8506265566567095e-07, | |
| "loss": 0.032, | |
| "step": 2362 | |
| }, | |
| { | |
| "epoch": 0.9456, | |
| "grad_norm": 0.006537347100675106, | |
| "learning_rate": 1.7975340367628269e-07, | |
| "loss": 0.0082, | |
| "step": 2364 | |
| }, | |
| { | |
| "epoch": 0.9464, | |
| "grad_norm": 0.03216685727238655, | |
| "learning_rate": 1.7452073194253237e-07, | |
| "loss": 0.0099, | |
| "step": 2366 | |
| }, | |
| { | |
| "epoch": 0.9472, | |
| "grad_norm": 0.06808136403560638, | |
| "learning_rate": 1.6936468126984573e-07, | |
| "loss": 0.0111, | |
| "step": 2368 | |
| }, | |
| { | |
| "epoch": 0.948, | |
| "grad_norm": 0.19587433338165283, | |
| "learning_rate": 1.6428529186614195e-07, | |
| "loss": 0.0222, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 0.9488, | |
| "grad_norm": 0.02673921547830105, | |
| "learning_rate": 1.5928260334151847e-07, | |
| "loss": 0.0362, | |
| "step": 2372 | |
| }, | |
| { | |
| "epoch": 0.9496, | |
| "grad_norm": 0.8786045908927917, | |
| "learning_rate": 1.543566547079467e-07, | |
| "loss": 0.0973, | |
| "step": 2374 | |
| }, | |
| { | |
| "epoch": 0.9504, | |
| "grad_norm": 0.16141577064990997, | |
| "learning_rate": 1.4950748437896235e-07, | |
| "loss": 0.0884, | |
| "step": 2376 | |
| }, | |
| { | |
| "epoch": 0.9512, | |
| "grad_norm": 0.010312036611139774, | |
| "learning_rate": 1.4473513016937223e-07, | |
| "loss": 0.0103, | |
| "step": 2378 | |
| }, | |
| { | |
| "epoch": 0.952, | |
| "grad_norm": 3.670624017715454, | |
| "learning_rate": 1.400396292949513e-07, | |
| "loss": 0.6124, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 0.9528, | |
| "grad_norm": 3.3416731357574463, | |
| "learning_rate": 1.3542101837215826e-07, | |
| "loss": 1.0542, | |
| "step": 2382 | |
| }, | |
| { | |
| "epoch": 0.9536, | |
| "grad_norm": 1.813645839691162, | |
| "learning_rate": 1.308793334178493e-07, | |
| "loss": 0.8749, | |
| "step": 2384 | |
| }, | |
| { | |
| "epoch": 0.9544, | |
| "grad_norm": 0.02037736214697361, | |
| "learning_rate": 1.26414609848996e-07, | |
| "loss": 0.0147, | |
| "step": 2386 | |
| }, | |
| { | |
| "epoch": 0.9552, | |
| "grad_norm": 0.29435980319976807, | |
| "learning_rate": 1.2202688248241113e-07, | |
| "loss": 0.0375, | |
| "step": 2388 | |
| }, | |
| { | |
| "epoch": 0.956, | |
| "grad_norm": 0.02400418370962143, | |
| "learning_rate": 1.1771618553447217e-07, | |
| "loss": 0.0272, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 0.9568, | |
| "grad_norm": 0.32467302680015564, | |
| "learning_rate": 1.134825526208605e-07, | |
| "loss": 0.039, | |
| "step": 2392 | |
| }, | |
| { | |
| "epoch": 0.9576, | |
| "grad_norm": 0.04526238515973091, | |
| "learning_rate": 1.0932601675629595e-07, | |
| "loss": 0.0099, | |
| "step": 2394 | |
| }, | |
| { | |
| "epoch": 0.9584, | |
| "grad_norm": 0.017281251028180122, | |
| "learning_rate": 1.052466103542793e-07, | |
| "loss": 0.0201, | |
| "step": 2396 | |
| }, | |
| { | |
| "epoch": 0.9592, | |
| "grad_norm": 0.2014533132314682, | |
| "learning_rate": 1.0124436522684244e-07, | |
| "loss": 0.232, | |
| "step": 2398 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.08100484311580658, | |
| "learning_rate": 9.731931258429638e-08, | |
| "loss": 0.1188, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.9608, | |
| "grad_norm": 0.03484244644641876, | |
| "learning_rate": 9.347148303499143e-08, | |
| "loss": 0.0144, | |
| "step": 2402 | |
| }, | |
| { | |
| "epoch": 0.9616, | |
| "grad_norm": 0.1309373676776886, | |
| "learning_rate": 8.970090658507291e-08, | |
| "loss": 0.015, | |
| "step": 2404 | |
| }, | |
| { | |
| "epoch": 0.9624, | |
| "grad_norm": 1.4030771255493164, | |
| "learning_rate": 8.600761263825475e-08, | |
| "loss": 0.5354, | |
| "step": 2406 | |
| }, | |
| { | |
| "epoch": 0.9632, | |
| "grad_norm": 0.009812920354306698, | |
| "learning_rate": 8.239162999558403e-08, | |
| "loss": 0.0114, | |
| "step": 2408 | |
| }, | |
| { | |
| "epoch": 0.964, | |
| "grad_norm": 1.2562764883041382, | |
| "learning_rate": 7.885298685522235e-08, | |
| "loss": 1.0972, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 0.9648, | |
| "grad_norm": 0.012443533167243004, | |
| "learning_rate": 7.539171081221597e-08, | |
| "loss": 0.0077, | |
| "step": 2412 | |
| }, | |
| { | |
| "epoch": 0.9656, | |
| "grad_norm": 0.011568997986614704, | |
| "learning_rate": 7.200782885829482e-08, | |
| "loss": 0.6151, | |
| "step": 2414 | |
| }, | |
| { | |
| "epoch": 0.9664, | |
| "grad_norm": 0.09277435392141342, | |
| "learning_rate": 6.870136738164612e-08, | |
| "loss": 0.0186, | |
| "step": 2416 | |
| }, | |
| { | |
| "epoch": 0.9672, | |
| "grad_norm": 0.1386393904685974, | |
| "learning_rate": 6.547235216672443e-08, | |
| "loss": 0.2091, | |
| "step": 2418 | |
| }, | |
| { | |
| "epoch": 0.968, | |
| "grad_norm": 1.1379826068878174, | |
| "learning_rate": 6.232080839403631e-08, | |
| "loss": 0.2519, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 0.9688, | |
| "grad_norm": 0.1122945249080658, | |
| "learning_rate": 5.9246760639953824e-08, | |
| "loss": 0.0173, | |
| "step": 2422 | |
| }, | |
| { | |
| "epoch": 0.9696, | |
| "grad_norm": 0.12855499982833862, | |
| "learning_rate": 5.625023287652021e-08, | |
| "loss": 0.2261, | |
| "step": 2424 | |
| }, | |
| { | |
| "epoch": 0.9704, | |
| "grad_norm": 0.007712547667324543, | |
| "learning_rate": 5.3331248471258926e-08, | |
| "loss": 0.0049, | |
| "step": 2426 | |
| }, | |
| { | |
| "epoch": 0.9712, | |
| "grad_norm": 0.033568158745765686, | |
| "learning_rate": 5.048983018699827e-08, | |
| "loss": 0.054, | |
| "step": 2428 | |
| }, | |
| { | |
| "epoch": 0.972, | |
| "grad_norm": 0.03975361958146095, | |
| "learning_rate": 4.772600018168816e-08, | |
| "loss": 0.0088, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 0.9728, | |
| "grad_norm": 0.12293253093957901, | |
| "learning_rate": 4.503978000823028e-08, | |
| "loss": 0.0239, | |
| "step": 2432 | |
| }, | |
| { | |
| "epoch": 0.9736, | |
| "grad_norm": 1.9904245138168335, | |
| "learning_rate": 4.2431190614309334e-08, | |
| "loss": 0.7805, | |
| "step": 2434 | |
| }, | |
| { | |
| "epoch": 0.9744, | |
| "grad_norm": 0.08196962624788284, | |
| "learning_rate": 3.990025234222872e-08, | |
| "loss": 0.0215, | |
| "step": 2436 | |
| }, | |
| { | |
| "epoch": 0.9752, | |
| "grad_norm": 0.015654845163226128, | |
| "learning_rate": 3.7446984928753984e-08, | |
| "loss": 0.0054, | |
| "step": 2438 | |
| }, | |
| { | |
| "epoch": 0.976, | |
| "grad_norm": 0.4907855987548828, | |
| "learning_rate": 3.50714075049563e-08, | |
| "loss": 0.0643, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 0.9768, | |
| "grad_norm": 0.11044461280107498, | |
| "learning_rate": 3.2773538596068134e-08, | |
| "loss": 0.017, | |
| "step": 2442 | |
| }, | |
| { | |
| "epoch": 0.9776, | |
| "grad_norm": 0.12533003091812134, | |
| "learning_rate": 3.0553396121330015e-08, | |
| "loss": 0.8734, | |
| "step": 2444 | |
| }, | |
| { | |
| "epoch": 0.9784, | |
| "grad_norm": 0.010446742177009583, | |
| "learning_rate": 2.8410997393860663e-08, | |
| "loss": 0.0066, | |
| "step": 2446 | |
| }, | |
| { | |
| "epoch": 0.9792, | |
| "grad_norm": 2.9740476608276367, | |
| "learning_rate": 2.6346359120514863e-08, | |
| "loss": 0.56, | |
| "step": 2448 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 0.15533113479614258, | |
| "learning_rate": 2.4359497401758026e-08, | |
| "loss": 0.0152, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.9808, | |
| "grad_norm": 0.42947232723236084, | |
| "learning_rate": 2.2450427731534052e-08, | |
| "loss": 0.1126, | |
| "step": 2452 | |
| }, | |
| { | |
| "epoch": 0.9816, | |
| "grad_norm": 0.31456366181373596, | |
| "learning_rate": 2.061916499715544e-08, | |
| "loss": 0.0455, | |
| "step": 2454 | |
| }, | |
| { | |
| "epoch": 0.9824, | |
| "grad_norm": 1.7186102867126465, | |
| "learning_rate": 1.886572347917337e-08, | |
| "loss": 0.1411, | |
| "step": 2456 | |
| }, | |
| { | |
| "epoch": 0.9832, | |
| "grad_norm": 0.3335157632827759, | |
| "learning_rate": 1.7190116851280024e-08, | |
| "loss": 0.0393, | |
| "step": 2458 | |
| }, | |
| { | |
| "epoch": 0.984, | |
| "grad_norm": 0.01097691897302866, | |
| "learning_rate": 1.5592358180189782e-08, | |
| "loss": 0.0299, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 0.9848, | |
| "grad_norm": 0.1632501780986786, | |
| "learning_rate": 1.4072459925548176e-08, | |
| "loss": 0.0287, | |
| "step": 2462 | |
| }, | |
| { | |
| "epoch": 0.9856, | |
| "grad_norm": 0.02019256353378296, | |
| "learning_rate": 1.2630433939825326e-08, | |
| "loss": 0.0658, | |
| "step": 2464 | |
| }, | |
| { | |
| "epoch": 0.9864, | |
| "grad_norm": 2.6901440620422363, | |
| "learning_rate": 1.126629146822933e-08, | |
| "loss": 0.2091, | |
| "step": 2466 | |
| }, | |
| { | |
| "epoch": 0.9872, | |
| "grad_norm": 0.05359053239226341, | |
| "learning_rate": 9.980043148619668e-09, | |
| "loss": 0.0161, | |
| "step": 2468 | |
| }, | |
| { | |
| "epoch": 0.988, | |
| "grad_norm": 1.5825371742248535, | |
| "learning_rate": 8.771699011416169e-09, | |
| "loss": 1.7133, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 0.9888, | |
| "grad_norm": 0.02139631099998951, | |
| "learning_rate": 7.641268479531283e-09, | |
| "loss": 0.0111, | |
| "step": 2472 | |
| }, | |
| { | |
| "epoch": 0.9896, | |
| "grad_norm": 0.06577183306217194, | |
| "learning_rate": 6.588760368287928e-09, | |
| "loss": 0.0126, | |
| "step": 2474 | |
| }, | |
| { | |
| "epoch": 0.9904, | |
| "grad_norm": 0.04376514256000519, | |
| "learning_rate": 5.614182885357311e-09, | |
| "loss": 0.0159, | |
| "step": 2476 | |
| }, | |
| { | |
| "epoch": 0.9912, | |
| "grad_norm": 0.02068435214459896, | |
| "learning_rate": 4.717543630688992e-09, | |
| "loss": 0.0087, | |
| "step": 2478 | |
| }, | |
| { | |
| "epoch": 0.992, | |
| "grad_norm": 2.170161724090576, | |
| "learning_rate": 3.898849596456477e-09, | |
| "loss": 0.2086, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 0.9928, | |
| "grad_norm": 1.0947741270065308, | |
| "learning_rate": 3.1581071670006013e-09, | |
| "loss": 0.0886, | |
| "step": 2482 | |
| }, | |
| { | |
| "epoch": 0.9936, | |
| "grad_norm": 0.08858716487884521, | |
| "learning_rate": 2.495322118778454e-09, | |
| "loss": 0.02, | |
| "step": 2484 | |
| }, | |
| { | |
| "epoch": 0.9944, | |
| "grad_norm": 0.011349241249263287, | |
| "learning_rate": 1.910499620322304e-09, | |
| "loss": 0.0094, | |
| "step": 2486 | |
| }, | |
| { | |
| "epoch": 0.9952, | |
| "grad_norm": 1.9726616144180298, | |
| "learning_rate": 1.4036442321962995e-09, | |
| "loss": 0.7641, | |
| "step": 2488 | |
| }, | |
| { | |
| "epoch": 0.996, | |
| "grad_norm": 0.020813768729567528, | |
| "learning_rate": 9.74759906957612e-10, | |
| "loss": 0.0056, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 0.9968, | |
| "grad_norm": 0.07518152892589569, | |
| "learning_rate": 6.238499891353389e-10, | |
| "loss": 0.0572, | |
| "step": 2492 | |
| }, | |
| { | |
| "epoch": 0.9976, | |
| "grad_norm": 0.12918703258037567, | |
| "learning_rate": 3.509172151938689e-10, | |
| "loss": 0.0209, | |
| "step": 2494 | |
| }, | |
| { | |
| "epoch": 0.9984, | |
| "grad_norm": 2.08217453956604, | |
| "learning_rate": 1.559637135173375e-10, | |
| "loss": 1.673, | |
| "step": 2496 | |
| }, | |
| { | |
| "epoch": 0.9992, | |
| "grad_norm": 0.01897619105875492, | |
| "learning_rate": 3.899100439408443e-11, | |
| "loss": 0.0482, | |
| "step": 2498 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.07024825364351273, | |
| "learning_rate": 0.0, | |
| "loss": 0.1147, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 2500, | |
| "total_flos": 5.2791220503996006e+17, | |
| "train_loss": 0.34746392381768676, | |
| "train_runtime": 41909.1346, | |
| "train_samples_per_second": 1.909, | |
| "train_steps_per_second": 0.06 | |
| } | |
| ], | |
| "logging_steps": 2, | |
| "max_steps": 2500, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "stateful_callbacks": {}, | |
| "total_flos": 5.2791220503996006e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |