| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9984, | |
| "eval_steps": 500, | |
| "global_step": 312, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0032, | |
| "grad_norm": 3.210516929626465, | |
| "learning_rate": 1.5625e-06, | |
| "loss": 2.7227, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0064, | |
| "grad_norm": 3.002136468887329, | |
| "learning_rate": 3.125e-06, | |
| "loss": 2.6726, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0096, | |
| "grad_norm": 3.242940902709961, | |
| "learning_rate": 4.6875000000000004e-06, | |
| "loss": 2.7516, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0128, | |
| "grad_norm": 3.367115020751953, | |
| "learning_rate": 6.25e-06, | |
| "loss": 2.7533, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.016, | |
| "grad_norm": 3.2682089805603027, | |
| "learning_rate": 7.8125e-06, | |
| "loss": 2.8562, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0192, | |
| "grad_norm": 2.6950130462646484, | |
| "learning_rate": 9.375000000000001e-06, | |
| "loss": 2.5765, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0224, | |
| "grad_norm": 2.6319284439086914, | |
| "learning_rate": 1.09375e-05, | |
| "loss": 2.7433, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0256, | |
| "grad_norm": 2.209996223449707, | |
| "learning_rate": 1.25e-05, | |
| "loss": 2.7297, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0288, | |
| "grad_norm": 1.8690632581710815, | |
| "learning_rate": 1.4062500000000001e-05, | |
| "loss": 2.7458, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.032, | |
| "grad_norm": 1.6140168905258179, | |
| "learning_rate": 1.5625e-05, | |
| "loss": 2.4958, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0352, | |
| "grad_norm": 1.326891303062439, | |
| "learning_rate": 1.71875e-05, | |
| "loss": 2.5878, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 1.4608954191207886, | |
| "learning_rate": 1.8750000000000002e-05, | |
| "loss": 2.743, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.0416, | |
| "grad_norm": 1.6133849620819092, | |
| "learning_rate": 2.0312500000000002e-05, | |
| "loss": 2.4727, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.0448, | |
| "grad_norm": 1.609899640083313, | |
| "learning_rate": 2.1875e-05, | |
| "loss": 2.9243, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.048, | |
| "grad_norm": 1.555704951286316, | |
| "learning_rate": 2.34375e-05, | |
| "loss": 2.5987, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.0512, | |
| "grad_norm": 1.3509330749511719, | |
| "learning_rate": 2.5e-05, | |
| "loss": 2.586, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.0544, | |
| "grad_norm": 1.2609320878982544, | |
| "learning_rate": 2.6562500000000002e-05, | |
| "loss": 2.6697, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.0576, | |
| "grad_norm": 1.2991684675216675, | |
| "learning_rate": 2.8125000000000003e-05, | |
| "loss": 2.5569, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.0608, | |
| "grad_norm": 1.432456135749817, | |
| "learning_rate": 2.96875e-05, | |
| "loss": 2.6736, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 1.418470025062561, | |
| "learning_rate": 3.125e-05, | |
| "loss": 2.4292, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0672, | |
| "grad_norm": 1.3896219730377197, | |
| "learning_rate": 3.2812500000000005e-05, | |
| "loss": 2.503, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.0704, | |
| "grad_norm": 1.2691140174865723, | |
| "learning_rate": 3.4375e-05, | |
| "loss": 2.4368, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.0736, | |
| "grad_norm": 1.4909545183181763, | |
| "learning_rate": 3.59375e-05, | |
| "loss": 2.5566, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 1.2060567140579224, | |
| "learning_rate": 3.7500000000000003e-05, | |
| "loss": 2.4947, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.225574016571045, | |
| "learning_rate": 3.90625e-05, | |
| "loss": 2.5794, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.0832, | |
| "grad_norm": 1.1829763650894165, | |
| "learning_rate": 4.0625000000000005e-05, | |
| "loss": 2.4389, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.0864, | |
| "grad_norm": 1.4608885049819946, | |
| "learning_rate": 4.21875e-05, | |
| "loss": 2.3006, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.0896, | |
| "grad_norm": 1.441902995109558, | |
| "learning_rate": 4.375e-05, | |
| "loss": 2.8311, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.0928, | |
| "grad_norm": 1.3018760681152344, | |
| "learning_rate": 4.5312500000000004e-05, | |
| "loss": 2.5907, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.096, | |
| "grad_norm": 1.089474081993103, | |
| "learning_rate": 4.6875e-05, | |
| "loss": 2.3614, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.0992, | |
| "grad_norm": 1.0250301361083984, | |
| "learning_rate": 4.8437500000000005e-05, | |
| "loss": 2.4455, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.1024, | |
| "grad_norm": 1.2610759735107422, | |
| "learning_rate": 5e-05, | |
| "loss": 2.5235, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.1056, | |
| "grad_norm": 1.1425381898880005, | |
| "learning_rate": 4.999842641886752e-05, | |
| "loss": 2.4895, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.1088, | |
| "grad_norm": 1.290959119796753, | |
| "learning_rate": 4.9993705873562665e-05, | |
| "loss": 2.3314, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.112, | |
| "grad_norm": 1.1456242799758911, | |
| "learning_rate": 4.998583895833834e-05, | |
| "loss": 2.6165, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 1.0945172309875488, | |
| "learning_rate": 4.997482666353287e-05, | |
| "loss": 2.2249, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.1184, | |
| "grad_norm": 1.024279236793518, | |
| "learning_rate": 4.996067037544542e-05, | |
| "loss": 2.4478, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.1216, | |
| "grad_norm": 1.1314380168914795, | |
| "learning_rate": 4.99433718761614e-05, | |
| "loss": 2.7661, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.1248, | |
| "grad_norm": 1.1418216228485107, | |
| "learning_rate": 4.99229333433282e-05, | |
| "loss": 2.3117, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 1.1869049072265625, | |
| "learning_rate": 4.989935734988098e-05, | |
| "loss": 2.5638, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.1312, | |
| "grad_norm": 1.2026361227035522, | |
| "learning_rate": 4.9872646863718805e-05, | |
| "loss": 2.5985, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.1344, | |
| "grad_norm": 1.1795376539230347, | |
| "learning_rate": 4.984280524733107e-05, | |
| "loss": 2.1883, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.1376, | |
| "grad_norm": 1.2123969793319702, | |
| "learning_rate": 4.980983625737411e-05, | |
| "loss": 2.6978, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.1408, | |
| "grad_norm": 1.2263802289962769, | |
| "learning_rate": 4.977374404419837e-05, | |
| "loss": 2.1447, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.144, | |
| "grad_norm": 1.304231882095337, | |
| "learning_rate": 4.973453315132592e-05, | |
| "loss": 2.4581, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.1472, | |
| "grad_norm": 1.3034032583236694, | |
| "learning_rate": 4.9692208514878444e-05, | |
| "loss": 2.568, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.1504, | |
| "grad_norm": 1.182282567024231, | |
| "learning_rate": 4.96467754629559e-05, | |
| "loss": 2.6686, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 1.1906458139419556, | |
| "learning_rate": 4.959823971496574e-05, | |
| "loss": 2.5744, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.1568, | |
| "grad_norm": 1.0540134906768799, | |
| "learning_rate": 4.954660738090297e-05, | |
| "loss": 2.4111, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 1.17970871925354, | |
| "learning_rate": 4.9491884960580894e-05, | |
| "loss": 2.3782, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.1632, | |
| "grad_norm": 1.0423855781555176, | |
| "learning_rate": 4.943407934281298e-05, | |
| "loss": 2.5053, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.1664, | |
| "grad_norm": 1.16681706905365, | |
| "learning_rate": 4.937319780454559e-05, | |
| "loss": 2.5527, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.1696, | |
| "grad_norm": 1.072268009185791, | |
| "learning_rate": 4.9309248009941914e-05, | |
| "loss": 2.3297, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.1728, | |
| "grad_norm": 1.2780967950820923, | |
| "learning_rate": 4.9242238009417175e-05, | |
| "loss": 2.5685, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.176, | |
| "grad_norm": 1.234634280204773, | |
| "learning_rate": 4.9172176238625164e-05, | |
| "loss": 2.7419, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.1792, | |
| "grad_norm": 1.080798864364624, | |
| "learning_rate": 4.909907151739633e-05, | |
| "loss": 2.4689, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.1824, | |
| "grad_norm": 1.0914878845214844, | |
| "learning_rate": 4.9022933048627496e-05, | |
| "loss": 2.0907, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.1856, | |
| "grad_norm": 1.1586689949035645, | |
| "learning_rate": 4.894377041712326e-05, | |
| "loss": 2.5898, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.1888, | |
| "grad_norm": 1.1252062320709229, | |
| "learning_rate": 4.886159358838952e-05, | |
| "loss": 2.2873, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 1.1714915037155151, | |
| "learning_rate": 4.877641290737884e-05, | |
| "loss": 2.5332, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.1952, | |
| "grad_norm": 1.1400797367095947, | |
| "learning_rate": 4.868823909718823e-05, | |
| "loss": 2.365, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.1984, | |
| "grad_norm": 1.1181775331497192, | |
| "learning_rate": 4.8597083257709194e-05, | |
| "loss": 2.5676, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.2016, | |
| "grad_norm": 1.0846837759017944, | |
| "learning_rate": 4.850295686423047e-05, | |
| "loss": 2.4235, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.2048, | |
| "grad_norm": 1.062685489654541, | |
| "learning_rate": 4.8405871765993433e-05, | |
| "loss": 2.4309, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.208, | |
| "grad_norm": 1.1622320413589478, | |
| "learning_rate": 4.8305840184700356e-05, | |
| "loss": 2.4819, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.2112, | |
| "grad_norm": 1.0995711088180542, | |
| "learning_rate": 4.820287471297598e-05, | |
| "loss": 2.5657, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.2144, | |
| "grad_norm": 1.0327019691467285, | |
| "learning_rate": 4.8096988312782174e-05, | |
| "loss": 2.5404, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.2176, | |
| "grad_norm": 1.0894691944122314, | |
| "learning_rate": 4.7988194313786275e-05, | |
| "loss": 2.4577, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.2208, | |
| "grad_norm": 1.2364932298660278, | |
| "learning_rate": 4.7876506411682994e-05, | |
| "loss": 2.4601, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.224, | |
| "grad_norm": 1.1828285455703735, | |
| "learning_rate": 4.7761938666470403e-05, | |
| "loss": 2.5505, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.2272, | |
| "grad_norm": 1.1249369382858276, | |
| "learning_rate": 4.7644505500679855e-05, | |
| "loss": 2.5093, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.2304, | |
| "grad_norm": 1.215545892715454, | |
| "learning_rate": 4.752422169756048e-05, | |
| "loss": 2.6171, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.2336, | |
| "grad_norm": 1.063791275024414, | |
| "learning_rate": 4.7401102399218136e-05, | |
| "loss": 2.7023, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.2368, | |
| "grad_norm": 1.2447466850280762, | |
| "learning_rate": 4.72751631047092e-05, | |
| "loss": 2.4954, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 1.0675621032714844, | |
| "learning_rate": 4.71464196680895e-05, | |
| "loss": 2.5718, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.2432, | |
| "grad_norm": 1.1253868341445923, | |
| "learning_rate": 4.701488829641845e-05, | |
| "loss": 2.4024, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.2464, | |
| "grad_norm": 1.0018173456192017, | |
| "learning_rate": 4.6880585547718845e-05, | |
| "loss": 2.4159, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.2496, | |
| "grad_norm": 1.1260743141174316, | |
| "learning_rate": 4.674352832889239e-05, | |
| "loss": 2.5211, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.2528, | |
| "grad_norm": 1.1455786228179932, | |
| "learning_rate": 4.660373389359137e-05, | |
| "loss": 2.4025, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 1.0881799459457397, | |
| "learning_rate": 4.6461219840046654e-05, | |
| "loss": 2.3178, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.2592, | |
| "grad_norm": 1.0626671314239502, | |
| "learning_rate": 4.6316004108852305e-05, | |
| "loss": 2.4681, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.2624, | |
| "grad_norm": 1.1673706769943237, | |
| "learning_rate": 4.6168104980707107e-05, | |
| "loss": 2.6277, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.2656, | |
| "grad_norm": 1.144708514213562, | |
| "learning_rate": 4.601754107411326e-05, | |
| "loss": 2.5948, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.2688, | |
| "grad_norm": 1.1058765649795532, | |
| "learning_rate": 4.586433134303257e-05, | |
| "loss": 2.6431, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.272, | |
| "grad_norm": 1.043436884880066, | |
| "learning_rate": 4.5708495074500414e-05, | |
| "loss": 2.4188, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.2752, | |
| "grad_norm": 1.211965560913086, | |
| "learning_rate": 4.5550051886197754e-05, | |
| "loss": 2.5749, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.2784, | |
| "grad_norm": 1.0927400588989258, | |
| "learning_rate": 4.538902172398151e-05, | |
| "loss": 2.4746, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.2816, | |
| "grad_norm": 1.1012922525405884, | |
| "learning_rate": 4.522542485937369e-05, | |
| "loss": 2.5538, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.2848, | |
| "grad_norm": 1.1138043403625488, | |
| "learning_rate": 4.505928188700945e-05, | |
| "loss": 2.4855, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.288, | |
| "grad_norm": 1.069653868675232, | |
| "learning_rate": 4.489061372204453e-05, | |
| "loss": 2.5325, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.2912, | |
| "grad_norm": 1.1098930835723877, | |
| "learning_rate": 4.4719441597522286e-05, | |
| "loss": 2.5671, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.2944, | |
| "grad_norm": 1.0448640584945679, | |
| "learning_rate": 4.454578706170075e-05, | |
| "loss": 2.6194, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.2976, | |
| "grad_norm": 1.0166351795196533, | |
| "learning_rate": 4.4369671975340025e-05, | |
| "loss": 2.2649, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.3008, | |
| "grad_norm": 1.0552963018417358, | |
| "learning_rate": 4.419111850895028e-05, | |
| "loss": 2.5805, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.304, | |
| "grad_norm": 1.093049168586731, | |
| "learning_rate": 4.401014914000078e-05, | |
| "loss": 2.5055, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.3072, | |
| "grad_norm": 1.0456770658493042, | |
| "learning_rate": 4.382678665009028e-05, | |
| "loss": 2.2681, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.3104, | |
| "grad_norm": 1.1942492723464966, | |
| "learning_rate": 4.364105412207914e-05, | |
| "loss": 2.5884, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.3136, | |
| "grad_norm": 1.0143263339996338, | |
| "learning_rate": 4.345297493718352e-05, | |
| "loss": 2.5151, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.3168, | |
| "grad_norm": 1.0093568563461304, | |
| "learning_rate": 4.326257277203194e-05, | |
| "loss": 2.2898, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 1.039635181427002, | |
| "learning_rate": 4.306987159568479e-05, | |
| "loss": 1.9802, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3232, | |
| "grad_norm": 1.183872103691101, | |
| "learning_rate": 4.2874895666616886e-05, | |
| "loss": 2.3071, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.3264, | |
| "grad_norm": 1.0931261777877808, | |
| "learning_rate": 4.267766952966369e-05, | |
| "loss": 2.5054, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.3296, | |
| "grad_norm": 1.217860460281372, | |
| "learning_rate": 4.2478218012931434e-05, | |
| "loss": 2.4392, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.3328, | |
| "grad_norm": 1.1708042621612549, | |
| "learning_rate": 4.227656622467162e-05, | |
| "loss": 2.5153, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.336, | |
| "grad_norm": 0.9975510239601135, | |
| "learning_rate": 4.207273955012018e-05, | |
| "loss": 2.5271, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.3392, | |
| "grad_norm": 1.004370093345642, | |
| "learning_rate": 4.186676364830186e-05, | |
| "loss": 2.5909, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.3424, | |
| "grad_norm": 1.0638096332550049, | |
| "learning_rate": 4.16586644488001e-05, | |
| "loss": 2.3149, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.3456, | |
| "grad_norm": 1.2675764560699463, | |
| "learning_rate": 4.144846814849282e-05, | |
| "loss": 2.3856, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.3488, | |
| "grad_norm": 1.0615671873092651, | |
| "learning_rate": 4.123620120825459e-05, | |
| "loss": 2.5105, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.352, | |
| "grad_norm": 1.2035943269729614, | |
| "learning_rate": 4.10218903496256e-05, | |
| "loss": 2.5255, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.3552, | |
| "grad_norm": 1.1443675756454468, | |
| "learning_rate": 4.0805562551447746e-05, | |
| "loss": 2.4382, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.3584, | |
| "grad_norm": 1.176398515701294, | |
| "learning_rate": 4.058724504646834e-05, | |
| "loss": 2.4046, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.3616, | |
| "grad_norm": 1.015869140625, | |
| "learning_rate": 4.036696531791193e-05, | |
| "loss": 2.4115, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.3648, | |
| "grad_norm": 1.1262589693069458, | |
| "learning_rate": 4.01447510960205e-05, | |
| "loss": 2.4621, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.368, | |
| "grad_norm": 1.190390944480896, | |
| "learning_rate": 3.992063035456259e-05, | |
| "loss": 2.6059, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.3712, | |
| "grad_norm": 1.1551072597503662, | |
| "learning_rate": 3.969463130731183e-05, | |
| "loss": 2.4153, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.3744, | |
| "grad_norm": 1.1593989133834839, | |
| "learning_rate": 3.946678240449515e-05, | |
| "loss": 2.625, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.3776, | |
| "grad_norm": 1.1310359239578247, | |
| "learning_rate": 3.92371123292113e-05, | |
| "loss": 2.7204, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.3808, | |
| "grad_norm": 0.9959138035774231, | |
| "learning_rate": 3.900564999382007e-05, | |
| "loss": 2.6826, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 0.9353814125061035, | |
| "learning_rate": 3.8772424536302564e-05, | |
| "loss": 2.5791, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.3872, | |
| "grad_norm": 1.0344676971435547, | |
| "learning_rate": 3.8537465316593146e-05, | |
| "loss": 2.8203, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.3904, | |
| "grad_norm": 0.9825398921966553, | |
| "learning_rate": 3.830080191288342e-05, | |
| "loss": 2.4996, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.3936, | |
| "grad_norm": 1.1759867668151855, | |
| "learning_rate": 3.8062464117898724e-05, | |
| "loss": 2.4976, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.3968, | |
| "grad_norm": 1.043148159980774, | |
| "learning_rate": 3.782248193514766e-05, | |
| "loss": 2.4043, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 1.0625474452972412, | |
| "learning_rate": 3.758088557514501e-05, | |
| "loss": 2.5448, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.4032, | |
| "grad_norm": 1.0381778478622437, | |
| "learning_rate": 3.7337705451608674e-05, | |
| "loss": 2.412, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.4064, | |
| "grad_norm": 1.0451267957687378, | |
| "learning_rate": 3.7092972177631e-05, | |
| "loss": 2.4634, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.4096, | |
| "grad_norm": 0.9723502397537231, | |
| "learning_rate": 3.6846716561824965e-05, | |
| "loss": 2.4095, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.4128, | |
| "grad_norm": 1.130776047706604, | |
| "learning_rate": 3.659896960444586e-05, | |
| "loss": 2.5474, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.416, | |
| "grad_norm": 1.0974534749984741, | |
| "learning_rate": 3.634976249348867e-05, | |
| "loss": 2.3988, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.4192, | |
| "grad_norm": 1.14722740650177, | |
| "learning_rate": 3.6099126600762054e-05, | |
| "loss": 2.5543, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.4224, | |
| "grad_norm": 1.1162605285644531, | |
| "learning_rate": 3.5847093477938956e-05, | |
| "loss": 2.6521, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.4256, | |
| "grad_norm": 1.118084192276001, | |
| "learning_rate": 3.559369485258472e-05, | |
| "loss": 2.4336, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.4288, | |
| "grad_norm": 1.0270711183547974, | |
| "learning_rate": 3.533896262416302e-05, | |
| "loss": 2.4505, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.432, | |
| "grad_norm": 1.1394667625427246, | |
| "learning_rate": 3.508292886002013e-05, | |
| "loss": 2.4266, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.4352, | |
| "grad_norm": 1.0506163835525513, | |
| "learning_rate": 3.4825625791348096e-05, | |
| "loss": 2.4871, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.4384, | |
| "grad_norm": 0.9986435174942017, | |
| "learning_rate": 3.456708580912725e-05, | |
| "loss": 2.2317, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.4416, | |
| "grad_norm": 1.2329692840576172, | |
| "learning_rate": 3.4307341460048633e-05, | |
| "loss": 2.6388, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.4448, | |
| "grad_norm": 1.143542766571045, | |
| "learning_rate": 3.404642544241681e-05, | |
| "loss": 2.5674, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 1.005985140800476, | |
| "learning_rate": 3.378437060203357e-05, | |
| "loss": 2.427, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.4512, | |
| "grad_norm": 1.0496375560760498, | |
| "learning_rate": 3.3521209928063126e-05, | |
| "loss": 2.2562, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.4544, | |
| "grad_norm": 1.0612940788269043, | |
| "learning_rate": 3.3256976548879184e-05, | |
| "loss": 2.4885, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.4576, | |
| "grad_norm": 1.1914249658584595, | |
| "learning_rate": 3.2991703727894544e-05, | |
| "loss": 2.4081, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.4608, | |
| "grad_norm": 1.2552766799926758, | |
| "learning_rate": 3.272542485937369e-05, | |
| "loss": 2.4895, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.464, | |
| "grad_norm": 1.1917306184768677, | |
| "learning_rate": 3.24581734642289e-05, | |
| "loss": 2.7435, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.4672, | |
| "grad_norm": 1.1121033430099487, | |
| "learning_rate": 3.218998318580043e-05, | |
| "loss": 2.3139, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.4704, | |
| "grad_norm": 1.168412446975708, | |
| "learning_rate": 3.1920887785621235e-05, | |
| "loss": 2.2553, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.4736, | |
| "grad_norm": 1.1404107809066772, | |
| "learning_rate": 3.165092113916688e-05, | |
| "loss": 2.7097, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.4768, | |
| "grad_norm": 0.9871900677680969, | |
| "learning_rate": 3.138011723159107e-05, | |
| "loss": 2.6076, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 1.1518453359603882, | |
| "learning_rate": 3.110851015344735e-05, | |
| "loss": 2.4797, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.4832, | |
| "grad_norm": 1.0925283432006836, | |
| "learning_rate": 3.083613409639764e-05, | |
| "loss": 2.3655, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.4864, | |
| "grad_norm": 1.1126621961593628, | |
| "learning_rate": 3.056302334890786e-05, | |
| "loss": 2.5094, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.4896, | |
| "grad_norm": 0.9415158033370972, | |
| "learning_rate": 3.0289212291931573e-05, | |
| "loss": 2.2806, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.4928, | |
| "grad_norm": 1.040486454963684, | |
| "learning_rate": 3.0014735394581823e-05, | |
| "loss": 2.5357, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.496, | |
| "grad_norm": 1.1474907398223877, | |
| "learning_rate": 2.9739627209791965e-05, | |
| "loss": 2.5369, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.4992, | |
| "grad_norm": 1.122162938117981, | |
| "learning_rate": 2.9463922369965917e-05, | |
| "loss": 2.4727, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.5024, | |
| "grad_norm": 1.1549993753433228, | |
| "learning_rate": 2.918765558261841e-05, | |
| "loss": 2.4274, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.5056, | |
| "grad_norm": 1.1090857982635498, | |
| "learning_rate": 2.8910861626005776e-05, | |
| "loss": 2.624, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.5088, | |
| "grad_norm": 1.0676389932632446, | |
| "learning_rate": 2.8633575344747822e-05, | |
| "loss": 2.3891, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 1.0996875762939453, | |
| "learning_rate": 2.8355831645441388e-05, | |
| "loss": 2.4476, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.5152, | |
| "grad_norm": 1.1088722944259644, | |
| "learning_rate": 2.8077665492266074e-05, | |
| "loss": 2.3327, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.5184, | |
| "grad_norm": 1.1235601902008057, | |
| "learning_rate": 2.7799111902582696e-05, | |
| "loss": 2.3989, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.5216, | |
| "grad_norm": 1.162764310836792, | |
| "learning_rate": 2.7520205942525112e-05, | |
| "loss": 2.3669, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.5248, | |
| "grad_norm": 1.1216943264007568, | |
| "learning_rate": 2.724098272258584e-05, | |
| "loss": 2.6234, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.528, | |
| "grad_norm": 1.1644459962844849, | |
| "learning_rate": 2.6961477393196126e-05, | |
| "loss": 2.4225, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.5312, | |
| "grad_norm": 1.0228266716003418, | |
| "learning_rate": 2.6681725140300997e-05, | |
| "loss": 2.6197, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.5344, | |
| "grad_norm": 1.1220487356185913, | |
| "learning_rate": 2.6401761180929797e-05, | |
| "loss": 2.5644, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.5376, | |
| "grad_norm": 1.1409152746200562, | |
| "learning_rate": 2.6121620758762877e-05, | |
| "loss": 2.2615, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.5408, | |
| "grad_norm": 1.0781875848770142, | |
| "learning_rate": 2.5841339139694855e-05, | |
| "loss": 2.451, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.544, | |
| "grad_norm": 1.1133708953857422, | |
| "learning_rate": 2.556095160739513e-05, | |
| "loss": 2.4254, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.5472, | |
| "grad_norm": 1.0333507061004639, | |
| "learning_rate": 2.528049345886615e-05, | |
| "loss": 2.3955, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.5504, | |
| "grad_norm": 1.085971474647522, | |
| "learning_rate": 2.5e-05, | |
| "loss": 2.6893, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.5536, | |
| "grad_norm": 1.1369028091430664, | |
| "learning_rate": 2.4719506541133855e-05, | |
| "loss": 2.5269, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.5568, | |
| "grad_norm": 1.037834882736206, | |
| "learning_rate": 2.443904839260488e-05, | |
| "loss": 2.5172, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 1.0019052028656006, | |
| "learning_rate": 2.415866086030516e-05, | |
| "loss": 2.1654, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.5632, | |
| "grad_norm": 0.9839959144592285, | |
| "learning_rate": 2.3878379241237136e-05, | |
| "loss": 2.4109, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.5664, | |
| "grad_norm": 1.0551410913467407, | |
| "learning_rate": 2.3598238819070202e-05, | |
| "loss": 2.4055, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.5696, | |
| "grad_norm": 1.0757516622543335, | |
| "learning_rate": 2.331827485969901e-05, | |
| "loss": 2.6649, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.5728, | |
| "grad_norm": 1.2172918319702148, | |
| "learning_rate": 2.303852260680388e-05, | |
| "loss": 2.4432, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 1.0906624794006348, | |
| "learning_rate": 2.2759017277414166e-05, | |
| "loss": 2.2428, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.5792, | |
| "grad_norm": 0.9661747813224792, | |
| "learning_rate": 2.247979405747489e-05, | |
| "loss": 2.4855, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.5824, | |
| "grad_norm": 1.0351911783218384, | |
| "learning_rate": 2.2200888097417307e-05, | |
| "loss": 2.2677, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.5856, | |
| "grad_norm": 1.0249441862106323, | |
| "learning_rate": 2.1922334507733932e-05, | |
| "loss": 2.561, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.5888, | |
| "grad_norm": 0.9759440422058105, | |
| "learning_rate": 2.164416835455862e-05, | |
| "loss": 2.2115, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.592, | |
| "grad_norm": 1.0776594877243042, | |
| "learning_rate": 2.136642465525219e-05, | |
| "loss": 2.2336, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.5952, | |
| "grad_norm": 1.052164912223816, | |
| "learning_rate": 2.1089138373994223e-05, | |
| "loss": 2.4157, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.5984, | |
| "grad_norm": 0.9809448719024658, | |
| "learning_rate": 2.0812344417381595e-05, | |
| "loss": 2.301, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.6016, | |
| "grad_norm": 1.090650200843811, | |
| "learning_rate": 2.0536077630034086e-05, | |
| "loss": 2.5767, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.6048, | |
| "grad_norm": 1.0655643939971924, | |
| "learning_rate": 2.026037279020804e-05, | |
| "loss": 2.5267, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.608, | |
| "grad_norm": 1.20234215259552, | |
| "learning_rate": 1.9985264605418183e-05, | |
| "loss": 2.5235, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.6112, | |
| "grad_norm": 1.0695698261260986, | |
| "learning_rate": 1.9710787708068433e-05, | |
| "loss": 2.4741, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.6144, | |
| "grad_norm": 0.9379785060882568, | |
| "learning_rate": 1.9436976651092144e-05, | |
| "loss": 2.5573, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.6176, | |
| "grad_norm": 1.0223705768585205, | |
| "learning_rate": 1.9163865903602374e-05, | |
| "loss": 2.4979, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.6208, | |
| "grad_norm": 1.0303515195846558, | |
| "learning_rate": 1.8891489846552646e-05, | |
| "loss": 2.5981, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.624, | |
| "grad_norm": 1.0409760475158691, | |
| "learning_rate": 1.8619882768408935e-05, | |
| "loss": 2.5319, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.6272, | |
| "grad_norm": 1.0393253564834595, | |
| "learning_rate": 1.8349078860833123e-05, | |
| "loss": 2.5587, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.6304, | |
| "grad_norm": 1.2285503149032593, | |
| "learning_rate": 1.8079112214378768e-05, | |
| "loss": 2.4699, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.6336, | |
| "grad_norm": 1.1109060049057007, | |
| "learning_rate": 1.781001681419957e-05, | |
| "loss": 2.4324, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.6368, | |
| "grad_norm": 1.032814860343933, | |
| "learning_rate": 1.75418265357711e-05, | |
| "loss": 2.7314, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 1.191022515296936, | |
| "learning_rate": 1.7274575140626318e-05, | |
| "loss": 2.3928, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.6432, | |
| "grad_norm": 1.0370298624038696, | |
| "learning_rate": 1.700829627210547e-05, | |
| "loss": 2.5094, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.6464, | |
| "grad_norm": 1.176619291305542, | |
| "learning_rate": 1.6743023451120832e-05, | |
| "loss": 2.5789, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.6496, | |
| "grad_norm": 1.0287202596664429, | |
| "learning_rate": 1.6478790071936877e-05, | |
| "loss": 2.5584, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.6528, | |
| "grad_norm": 1.0078474283218384, | |
| "learning_rate": 1.621562939796643e-05, | |
| "loss": 2.5382, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.656, | |
| "grad_norm": 0.9922024011611938, | |
| "learning_rate": 1.59535745575832e-05, | |
| "loss": 2.4357, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.6592, | |
| "grad_norm": 1.0808738470077515, | |
| "learning_rate": 1.5692658539951372e-05, | |
| "loss": 2.5151, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.6624, | |
| "grad_norm": 1.1992560625076294, | |
| "learning_rate": 1.5432914190872757e-05, | |
| "loss": 2.6586, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.6656, | |
| "grad_norm": 1.0298317670822144, | |
| "learning_rate": 1.5174374208651912e-05, | |
| "loss": 2.6879, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.6688, | |
| "grad_norm": 1.0588644742965698, | |
| "learning_rate": 1.4917071139979876e-05, | |
| "loss": 2.4113, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.672, | |
| "grad_norm": 1.0036541223526, | |
| "learning_rate": 1.466103737583699e-05, | |
| "loss": 2.3142, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.6752, | |
| "grad_norm": 0.9461038708686829, | |
| "learning_rate": 1.4406305147415283e-05, | |
| "loss": 2.3132, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.6784, | |
| "grad_norm": 1.0341013669967651, | |
| "learning_rate": 1.4152906522061048e-05, | |
| "loss": 2.3722, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.6816, | |
| "grad_norm": 1.0475574731826782, | |
| "learning_rate": 1.3900873399237952e-05, | |
| "loss": 2.5176, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.6848, | |
| "grad_norm": 1.3477550745010376, | |
| "learning_rate": 1.3650237506511331e-05, | |
| "loss": 2.4167, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.688, | |
| "grad_norm": 1.0816230773925781, | |
| "learning_rate": 1.340103039555415e-05, | |
| "loss": 2.6871, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.6912, | |
| "grad_norm": 1.0898610353469849, | |
| "learning_rate": 1.3153283438175034e-05, | |
| "loss": 2.585, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.6944, | |
| "grad_norm": 1.0709856748580933, | |
| "learning_rate": 1.2907027822369005e-05, | |
| "loss": 2.4157, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.6976, | |
| "grad_norm": 1.074513554573059, | |
| "learning_rate": 1.2662294548391328e-05, | |
| "loss": 2.5257, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.7008, | |
| "grad_norm": 1.0920336246490479, | |
| "learning_rate": 1.2419114424855e-05, | |
| "loss": 2.4891, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.704, | |
| "grad_norm": 0.8652957081794739, | |
| "learning_rate": 1.217751806485235e-05, | |
| "loss": 2.4609, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.7072, | |
| "grad_norm": 1.1371972560882568, | |
| "learning_rate": 1.1937535882101281e-05, | |
| "loss": 2.2327, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.7104, | |
| "grad_norm": 1.0960814952850342, | |
| "learning_rate": 1.1699198087116589e-05, | |
| "loss": 2.5813, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.7136, | |
| "grad_norm": 0.9509531259536743, | |
| "learning_rate": 1.1462534683406858e-05, | |
| "loss": 2.2718, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.7168, | |
| "grad_norm": 0.894752025604248, | |
| "learning_rate": 1.122757546369744e-05, | |
| "loss": 2.3841, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 1.073641061782837, | |
| "learning_rate": 1.0994350006179931e-05, | |
| "loss": 2.5524, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.7232, | |
| "grad_norm": 1.117712378501892, | |
| "learning_rate": 1.0762887670788702e-05, | |
| "loss": 2.4866, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.7264, | |
| "grad_norm": 1.120766043663025, | |
| "learning_rate": 1.0533217595504858e-05, | |
| "loss": 2.513, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.7296, | |
| "grad_norm": 1.0142333507537842, | |
| "learning_rate": 1.0305368692688174e-05, | |
| "loss": 2.448, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.7328, | |
| "grad_norm": 1.1020923852920532, | |
| "learning_rate": 1.007936964543741e-05, | |
| "loss": 2.6427, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.736, | |
| "grad_norm": 1.029630422592163, | |
| "learning_rate": 9.855248903979506e-06, | |
| "loss": 2.4163, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.7392, | |
| "grad_norm": 1.0667723417282104, | |
| "learning_rate": 9.63303468208807e-06, | |
| "loss": 2.5032, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.7424, | |
| "grad_norm": 1.1070666313171387, | |
| "learning_rate": 9.412754953531663e-06, | |
| "loss": 2.4455, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.7456, | |
| "grad_norm": 0.9969913363456726, | |
| "learning_rate": 9.19443744855226e-06, | |
| "loss": 2.3838, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.7488, | |
| "grad_norm": 0.9451704025268555, | |
| "learning_rate": 8.978109650374397e-06, | |
| "loss": 2.2581, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.752, | |
| "grad_norm": 1.1031489372253418, | |
| "learning_rate": 8.763798791745411e-06, | |
| "loss": 2.5738, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.7552, | |
| "grad_norm": 1.0689350366592407, | |
| "learning_rate": 8.551531851507186e-06, | |
| "loss": 2.0363, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.7584, | |
| "grad_norm": 1.0003691911697388, | |
| "learning_rate": 8.341335551199902e-06, | |
| "loss": 2.4822, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.7616, | |
| "grad_norm": 1.1118286848068237, | |
| "learning_rate": 8.133236351698143e-06, | |
| "loss": 2.505, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.7648, | |
| "grad_norm": 1.0897243022918701, | |
| "learning_rate": 7.927260449879829e-06, | |
| "loss": 2.5076, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 1.0602351427078247, | |
| "learning_rate": 7.723433775328384e-06, | |
| "loss": 2.503, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.7712, | |
| "grad_norm": 1.1504631042480469, | |
| "learning_rate": 7.521781987068565e-06, | |
| "loss": 2.4007, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.7744, | |
| "grad_norm": 1.1526134014129639, | |
| "learning_rate": 7.3223304703363135e-06, | |
| "loss": 2.6463, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.7776, | |
| "grad_norm": 0.9730823636054993, | |
| "learning_rate": 7.125104333383117e-06, | |
| "loss": 2.5505, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.7808, | |
| "grad_norm": 1.1611665487289429, | |
| "learning_rate": 6.930128404315214e-06, | |
| "loss": 2.3515, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.784, | |
| "grad_norm": 1.2234477996826172, | |
| "learning_rate": 6.737427227968063e-06, | |
| "loss": 2.4265, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.7872, | |
| "grad_norm": 0.9260097146034241, | |
| "learning_rate": 6.547025062816486e-06, | |
| "loss": 2.1394, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.7904, | |
| "grad_norm": 1.0301282405853271, | |
| "learning_rate": 6.358945877920861e-06, | |
| "loss": 2.3806, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.7936, | |
| "grad_norm": 0.9447835683822632, | |
| "learning_rate": 6.173213349909729e-06, | |
| "loss": 2.6367, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.7968, | |
| "grad_norm": 1.0396769046783447, | |
| "learning_rate": 5.989850859999227e-06, | |
| "loss": 2.6975, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 1.023889183998108, | |
| "learning_rate": 5.808881491049723e-06, | |
| "loss": 2.4211, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.8032, | |
| "grad_norm": 1.043057918548584, | |
| "learning_rate": 5.6303280246599786e-06, | |
| "loss": 2.5829, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.8064, | |
| "grad_norm": 1.104324460029602, | |
| "learning_rate": 5.454212938299255e-06, | |
| "loss": 2.5383, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.8096, | |
| "grad_norm": 1.1070061922073364, | |
| "learning_rate": 5.280558402477725e-06, | |
| "loss": 2.4353, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.8128, | |
| "grad_norm": 1.0229357481002808, | |
| "learning_rate": 5.1093862779554776e-06, | |
| "loss": 2.5017, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.816, | |
| "grad_norm": 1.0547471046447754, | |
| "learning_rate": 4.940718112990553e-06, | |
| "loss": 2.4777, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.8192, | |
| "grad_norm": 1.06377112865448, | |
| "learning_rate": 4.7745751406263165e-06, | |
| "loss": 2.3699, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.8224, | |
| "grad_norm": 1.158517837524414, | |
| "learning_rate": 4.610978276018496e-06, | |
| "loss": 2.4627, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.8256, | |
| "grad_norm": 1.072967290878296, | |
| "learning_rate": 4.4499481138022544e-06, | |
| "loss": 2.5287, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.8288, | |
| "grad_norm": 0.9575193524360657, | |
| "learning_rate": 4.29150492549959e-06, | |
| "loss": 2.5028, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.832, | |
| "grad_norm": 1.0296958684921265, | |
| "learning_rate": 4.135668656967434e-06, | |
| "loss": 2.6446, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.8352, | |
| "grad_norm": 0.9970160722732544, | |
| "learning_rate": 3.982458925886748e-06, | |
| "loss": 2.3393, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.8384, | |
| "grad_norm": 1.1589882373809814, | |
| "learning_rate": 3.831895019292897e-06, | |
| "loss": 2.5703, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.8416, | |
| "grad_norm": 1.0067667961120605, | |
| "learning_rate": 3.6839958911476957e-06, | |
| "loss": 2.4839, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.8448, | |
| "grad_norm": 1.046972632408142, | |
| "learning_rate": 3.5387801599533475e-06, | |
| "loss": 2.6206, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.848, | |
| "grad_norm": 1.047595500946045, | |
| "learning_rate": 3.3962661064086354e-06, | |
| "loss": 2.4652, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.8512, | |
| "grad_norm": 1.0699108839035034, | |
| "learning_rate": 3.2564716711076167e-06, | |
| "loss": 2.3963, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.8544, | |
| "grad_norm": 1.1620286703109741, | |
| "learning_rate": 3.119414452281158e-06, | |
| "loss": 2.3617, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.8576, | |
| "grad_norm": 1.1183536052703857, | |
| "learning_rate": 2.98511170358155e-06, | |
| "loss": 2.6089, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.8608, | |
| "grad_norm": 1.0337989330291748, | |
| "learning_rate": 2.853580331910505e-06, | |
| "loss": 2.3825, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.864, | |
| "grad_norm": 0.9789101481437683, | |
| "learning_rate": 2.7248368952908053e-06, | |
| "loss": 2.5147, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.8672, | |
| "grad_norm": 1.014626383781433, | |
| "learning_rate": 2.5988976007818715e-06, | |
| "loss": 2.339, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.8704, | |
| "grad_norm": 1.0315630435943604, | |
| "learning_rate": 2.475778302439524e-06, | |
| "loss": 2.4048, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.8736, | |
| "grad_norm": 1.0927337408065796, | |
| "learning_rate": 2.355494499320149e-06, | |
| "loss": 2.5846, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.8768, | |
| "grad_norm": 1.0336443185806274, | |
| "learning_rate": 2.2380613335296036e-06, | |
| "loss": 2.3522, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 1.1010684967041016, | |
| "learning_rate": 2.1234935883170047e-06, | |
| "loss": 2.5251, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.8832, | |
| "grad_norm": 1.031034231185913, | |
| "learning_rate": 2.0118056862137357e-06, | |
| "loss": 2.5216, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.8864, | |
| "grad_norm": 1.0429649353027344, | |
| "learning_rate": 1.9030116872178316e-06, | |
| "loss": 2.4645, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.8896, | |
| "grad_norm": 1.021558403968811, | |
| "learning_rate": 1.7971252870240291e-06, | |
| "loss": 2.3027, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.8928, | |
| "grad_norm": 1.1493953466415405, | |
| "learning_rate": 1.6941598152996451e-06, | |
| "loss": 2.3772, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 1.0270682573318481, | |
| "learning_rate": 1.59412823400657e-06, | |
| "loss": 2.2627, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.8992, | |
| "grad_norm": 1.0478615760803223, | |
| "learning_rate": 1.4970431357695243e-06, | |
| "loss": 2.4045, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.9024, | |
| "grad_norm": 1.0421029329299927, | |
| "learning_rate": 1.4029167422908107e-06, | |
| "loss": 2.3885, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.9056, | |
| "grad_norm": 0.8238216638565063, | |
| "learning_rate": 1.3117609028117817e-06, | |
| "loss": 2.4014, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.9088, | |
| "grad_norm": 1.1559348106384277, | |
| "learning_rate": 1.2235870926211619e-06, | |
| "loss": 2.6491, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.912, | |
| "grad_norm": 1.0746315717697144, | |
| "learning_rate": 1.138406411610482e-06, | |
| "loss": 2.2636, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.9152, | |
| "grad_norm": 1.155534267425537, | |
| "learning_rate": 1.0562295828767387e-06, | |
| "loss": 2.3605, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.9184, | |
| "grad_norm": 0.8642758727073669, | |
| "learning_rate": 9.770669513725128e-07, | |
| "loss": 2.5547, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.9216, | |
| "grad_norm": 1.0164655447006226, | |
| "learning_rate": 9.009284826036691e-07, | |
| "loss": 2.4789, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.9248, | |
| "grad_norm": 1.0909721851348877, | |
| "learning_rate": 8.278237613748407e-07, | |
| "loss": 2.3089, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.928, | |
| "grad_norm": 1.0505188703536987, | |
| "learning_rate": 7.577619905828282e-07, | |
| "loss": 2.6275, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.9312, | |
| "grad_norm": 1.0300596952438354, | |
| "learning_rate": 6.907519900580861e-07, | |
| "loss": 2.5099, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.9344, | |
| "grad_norm": 1.007779598236084, | |
| "learning_rate": 6.268021954544096e-07, | |
| "loss": 2.4924, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.9376, | |
| "grad_norm": 1.0541815757751465, | |
| "learning_rate": 5.659206571870218e-07, | |
| "loss": 2.4179, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.9408, | |
| "grad_norm": 1.1204500198364258, | |
| "learning_rate": 5.08115039419113e-07, | |
| "loss": 2.4938, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.944, | |
| "grad_norm": 0.9952044486999512, | |
| "learning_rate": 4.5339261909704e-07, | |
| "loss": 2.5642, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.9472, | |
| "grad_norm": 0.9769514799118042, | |
| "learning_rate": 4.0176028503425835e-07, | |
| "loss": 2.4987, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.9504, | |
| "grad_norm": 1.274708867073059, | |
| "learning_rate": 3.5322453704410286e-07, | |
| "loss": 2.7626, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.9536, | |
| "grad_norm": 0.93681401014328, | |
| "learning_rate": 3.077914851215585e-07, | |
| "loss": 2.3922, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.9568, | |
| "grad_norm": 0.9534211158752441, | |
| "learning_rate": 2.6546684867408413e-07, | |
| "loss": 2.4974, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 1.060401439666748, | |
| "learning_rate": 2.262559558016325e-07, | |
| "loss": 2.5511, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.9632, | |
| "grad_norm": 0.9338224530220032, | |
| "learning_rate": 1.9016374262589842e-07, | |
| "loss": 2.4446, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.9664, | |
| "grad_norm": 0.9976362586021423, | |
| "learning_rate": 1.571947526689349e-07, | |
| "loss": 2.3035, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.9696, | |
| "grad_norm": 1.1598321199417114, | |
| "learning_rate": 1.2735313628119139e-07, | |
| "loss": 2.4681, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.9728, | |
| "grad_norm": 1.0966912508010864, | |
| "learning_rate": 1.006426501190233e-07, | |
| "loss": 2.2132, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.976, | |
| "grad_norm": 1.1520273685455322, | |
| "learning_rate": 7.706665667180091e-08, | |
| "loss": 2.5282, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.9792, | |
| "grad_norm": 0.9652422666549683, | |
| "learning_rate": 5.662812383859795e-08, | |
| "loss": 2.4993, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.9824, | |
| "grad_norm": 1.2469524145126343, | |
| "learning_rate": 3.9329624554584884e-08, | |
| "loss": 2.4509, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.9856, | |
| "grad_norm": 1.0177453756332397, | |
| "learning_rate": 2.5173336467135267e-08, | |
| "loss": 2.4215, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.9888, | |
| "grad_norm": 1.1211117506027222, | |
| "learning_rate": 1.4161041661667207e-08, | |
| "loss": 2.4669, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.992, | |
| "grad_norm": 0.8371671438217163, | |
| "learning_rate": 6.294126437336734e-09, | |
| "loss": 2.3749, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.9952, | |
| "grad_norm": 1.4161032438278198, | |
| "learning_rate": 1.5735811324857352e-09, | |
| "loss": 2.426, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.9984, | |
| "grad_norm": 1.0636640787124634, | |
| "learning_rate": 0.0, | |
| "loss": 2.7309, | |
| "step": 312 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 312, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.1954372272399974e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |