diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,7406 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 1052,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0019011406844106464,
+      "grad_norm": 3.74786681060824,
+      "learning_rate": 1.886792452830189e-07,
+      "loss": 0.7265,
+      "step": 1
+    },
+    {
+      "epoch": 0.0038022813688212928,
+      "grad_norm": 3.3424396223401422,
+      "learning_rate": 3.773584905660378e-07,
+      "loss": 0.7022,
+      "step": 2
+    },
+    {
+      "epoch": 0.005703422053231939,
+      "grad_norm": 3.6138788266988686,
+      "learning_rate": 5.660377358490567e-07,
+      "loss": 0.7099,
+      "step": 3
+    },
+    {
+      "epoch": 0.0076045627376425855,
+      "grad_norm": 3.1108261115129863,
+      "learning_rate": 7.547169811320755e-07,
+      "loss": 0.6621,
+      "step": 4
+    },
+    {
+      "epoch": 0.009505703422053232,
+      "grad_norm": 3.411743805817171,
+      "learning_rate": 9.433962264150944e-07,
+      "loss": 0.7205,
+      "step": 5
+    },
+    {
+      "epoch": 0.011406844106463879,
+      "grad_norm": 3.090451007249929,
+      "learning_rate": 1.1320754716981133e-06,
+      "loss": 0.6731,
+      "step": 6
+    },
+    {
+      "epoch": 0.013307984790874524,
+      "grad_norm": 3.228378394097315,
+      "learning_rate": 1.3207547169811322e-06,
+      "loss": 0.6991,
+      "step": 7
+    },
+    {
+      "epoch": 0.015209125475285171,
+      "grad_norm": 3.0861402416352575,
+      "learning_rate": 1.509433962264151e-06,
+      "loss": 0.7209,
+      "step": 8
+    },
+    {
+      "epoch": 0.017110266159695818,
+      "grad_norm": 2.6605762774467143,
+      "learning_rate": 1.6981132075471698e-06,
+      "loss": 0.6849,
+      "step": 9
+    },
+    {
+      "epoch": 0.019011406844106463,
+      "grad_norm": 2.4648491366205376,
+      "learning_rate": 1.8867924528301889e-06,
+      "loss": 0.6569,
+      "step": 10
+    },
+    {
+      "epoch": 0.02091254752851711,
+      "grad_norm": 1.8001196565357553,
+      "learning_rate": 2.075471698113208e-06,
+      "loss": 0.6422,
+      "step": 11
+    },
+    {
+      "epoch": 0.022813688212927757,
+      "grad_norm": 1.6657227438068032,
+      "learning_rate": 2.2641509433962266e-06,
+      "loss": 0.597,
+      "step": 12
+    },
+    {
+      "epoch": 0.024714828897338403,
+      "grad_norm": 1.6344200838374798,
+      "learning_rate": 2.4528301886792453e-06,
+      "loss": 0.591,
+      "step": 13
+    },
+    {
+      "epoch": 0.026615969581749048,
+      "grad_norm": 1.4789793595290497,
+      "learning_rate": 2.6415094339622644e-06,
+      "loss": 0.5796,
+      "step": 14
+    },
+    {
+      "epoch": 0.028517110266159697,
+      "grad_norm": 1.8400513939305951,
+      "learning_rate": 2.830188679245283e-06,
+      "loss": 0.5894,
+      "step": 15
+    },
+    {
+      "epoch": 0.030418250950570342,
+      "grad_norm": 2.272871097180248,
+      "learning_rate": 3.018867924528302e-06,
+      "loss": 0.5723,
+      "step": 16
+    },
+    {
+      "epoch": 0.03231939163498099,
+      "grad_norm": 2.2580973980097263,
+      "learning_rate": 3.207547169811321e-06,
+      "loss": 0.584,
+      "step": 17
+    },
+    {
+      "epoch": 0.034220532319391636,
+      "grad_norm": 1.7989611410402615,
+      "learning_rate": 3.3962264150943395e-06,
+      "loss": 0.568,
+      "step": 18
+    },
+    {
+      "epoch": 0.03612167300380228,
+      "grad_norm": 1.573816761068462,
+      "learning_rate": 3.5849056603773586e-06,
+      "loss": 0.5474,
+      "step": 19
+    },
+    {
+      "epoch": 0.03802281368821293,
+      "grad_norm": 1.367023536605357,
+      "learning_rate": 3.7735849056603777e-06,
+      "loss": 0.5404,
+      "step": 20
+    },
+    {
+      "epoch": 0.039923954372623575,
+      "grad_norm": 1.501148818560846,
+      "learning_rate": 3.962264150943396e-06,
+      "loss": 0.5479,
+      "step": 21
+    },
+    {
+      "epoch": 0.04182509505703422,
+      "grad_norm": 1.4466000628236153,
+      "learning_rate": 4.150943396226416e-06,
+      "loss": 0.5461,
+      "step": 22
+    },
+    {
+      "epoch": 0.043726235741444866,
+      "grad_norm": 1.36885846956964,
+      "learning_rate": 4.339622641509435e-06,
+      "loss": 0.5304,
+      "step": 23
+    },
+    {
+      "epoch": 0.045627376425855515,
+      "grad_norm": 1.147711002570689,
+      "learning_rate": 4.528301886792453e-06,
+      "loss": 0.5347,
+      "step": 24
+    },
+    {
+      "epoch": 0.04752851711026616,
+      "grad_norm": 1.0863352461947726,
+      "learning_rate": 4.716981132075472e-06,
+      "loss": 0.5391,
+      "step": 25
+    },
+    {
+      "epoch": 0.049429657794676805,
+      "grad_norm": 1.081529822055453,
+      "learning_rate": 4.905660377358491e-06,
+      "loss": 0.547,
+      "step": 26
+    },
+    {
+      "epoch": 0.051330798479087454,
+      "grad_norm": 1.052886225048926,
+      "learning_rate": 5.09433962264151e-06,
+      "loss": 0.5451,
+      "step": 27
+    },
+    {
+      "epoch": 0.053231939163498096,
+      "grad_norm": 1.0698252304641873,
+      "learning_rate": 5.283018867924529e-06,
+      "loss": 0.523,
+      "step": 28
+    },
+    {
+      "epoch": 0.055133079847908745,
+      "grad_norm": 0.9955262720956431,
+      "learning_rate": 5.4716981132075475e-06,
+      "loss": 0.5082,
+      "step": 29
+    },
+    {
+      "epoch": 0.057034220532319393,
+      "grad_norm": 0.8974887518833294,
+      "learning_rate": 5.660377358490566e-06,
+      "loss": 0.5163,
+      "step": 30
+    },
+    {
+      "epoch": 0.058935361216730035,
+      "grad_norm": 1.0428704122632975,
+      "learning_rate": 5.849056603773585e-06,
+      "loss": 0.5047,
+      "step": 31
+    },
+    {
+      "epoch": 0.060836501901140684,
+      "grad_norm": 0.8990910459248553,
+      "learning_rate": 6.037735849056604e-06,
+      "loss": 0.471,
+      "step": 32
+    },
+    {
+      "epoch": 0.06273764258555133,
+      "grad_norm": 0.9640322844327028,
+      "learning_rate": 6.226415094339623e-06,
+      "loss": 0.4868,
+      "step": 33
+    },
+    {
+      "epoch": 0.06463878326996197,
+      "grad_norm": 1.0373707126656053,
+      "learning_rate": 6.415094339622642e-06,
+      "loss": 0.5014,
+      "step": 34
+    },
+    {
+      "epoch": 0.06653992395437262,
+      "grad_norm": 0.9044619419126771,
+      "learning_rate": 6.60377358490566e-06,
+      "loss": 0.5045,
+      "step": 35
+    },
+    {
+      "epoch": 0.06844106463878327,
+      "grad_norm": 0.893191048697127,
+      "learning_rate": 6.792452830188679e-06,
+      "loss": 0.4956,
+      "step": 36
+    },
+    {
+      "epoch": 0.07034220532319392,
+      "grad_norm": 0.976103743722923,
+      "learning_rate": 6.981132075471699e-06,
+      "loss": 0.4628,
+      "step": 37
+    },
+    {
+      "epoch": 0.07224334600760456,
+      "grad_norm": 0.9589310308114807,
+      "learning_rate": 7.169811320754717e-06,
+      "loss": 0.4894,
+      "step": 38
+    },
+    {
+      "epoch": 0.0741444866920152,
+      "grad_norm": 0.9351092557702548,
+      "learning_rate": 7.358490566037736e-06,
+      "loss": 0.4994,
+      "step": 39
+    },
+    {
+      "epoch": 0.07604562737642585,
+      "grad_norm": 0.9778144341210157,
+      "learning_rate": 7.5471698113207555e-06,
+      "loss": 0.4654,
+      "step": 40
+    },
+    {
+      "epoch": 0.0779467680608365,
+      "grad_norm": 0.969032252381567,
+      "learning_rate": 7.735849056603775e-06,
+      "loss": 0.4815,
+      "step": 41
+    },
+    {
+      "epoch": 0.07984790874524715,
+      "grad_norm": 0.84928283265404,
+      "learning_rate": 7.924528301886793e-06,
+      "loss": 0.465,
+      "step": 42
+    },
+    {
+      "epoch": 0.0817490494296578,
+      "grad_norm": 0.9912164350114498,
+      "learning_rate": 8.113207547169812e-06,
+      "loss": 0.4681,
+      "step": 43
+    },
+    {
+      "epoch": 0.08365019011406843,
+      "grad_norm": 0.9332364137239432,
+      "learning_rate": 8.301886792452832e-06,
+      "loss": 0.4606,
+      "step": 44
+    },
+    {
+      "epoch": 0.08555133079847908,
+      "grad_norm": 0.8813895326296612,
+      "learning_rate": 8.49056603773585e-06,
+      "loss": 0.4672,
+      "step": 45
+    },
+    {
+      "epoch": 0.08745247148288973,
+      "grad_norm": 0.9708016225540298,
+      "learning_rate": 8.67924528301887e-06,
+      "loss": 0.4703,
+      "step": 46
+    },
+    {
+      "epoch": 0.08935361216730038,
+      "grad_norm": 0.8723998697684501,
+      "learning_rate": 8.867924528301887e-06,
+      "loss": 0.4823,
+      "step": 47
+    },
+    {
+      "epoch": 0.09125475285171103,
+      "grad_norm": 0.8957511314605636,
+      "learning_rate": 9.056603773584907e-06,
+      "loss": 0.4772,
+      "step": 48
+    },
+    {
+      "epoch": 0.09315589353612168,
+      "grad_norm": 0.9132755007140755,
+      "learning_rate": 9.245283018867926e-06,
+      "loss": 0.4837,
+      "step": 49
+    },
+    {
+      "epoch": 0.09505703422053231,
+      "grad_norm": 0.832787156345999,
+      "learning_rate": 9.433962264150944e-06,
+      "loss": 0.4892,
+      "step": 50
+    },
+    {
+      "epoch": 0.09695817490494296,
+      "grad_norm": 0.8506663806064935,
+      "learning_rate": 9.622641509433963e-06,
+      "loss": 0.4705,
+      "step": 51
+    },
+    {
+      "epoch": 0.09885931558935361,
+      "grad_norm": 0.9014156196511167,
+      "learning_rate": 9.811320754716981e-06,
+      "loss": 0.4682,
+      "step": 52
+    },
+    {
+      "epoch": 0.10076045627376426,
+      "grad_norm": 0.8912951051283108,
+      "learning_rate": 1e-05,
+      "loss": 0.4833,
+      "step": 53
+    },
+    {
+      "epoch": 0.10266159695817491,
+      "grad_norm": 0.8724237196512474,
+      "learning_rate": 1.018867924528302e-05,
+      "loss": 0.4627,
+      "step": 54
+    },
+    {
+      "epoch": 0.10456273764258556,
+      "grad_norm": 0.9127194573526656,
+      "learning_rate": 1.0377358490566038e-05,
+      "loss": 0.4942,
+      "step": 55
+    },
+    {
+      "epoch": 0.10646387832699619,
+      "grad_norm": 0.8510281786281167,
+      "learning_rate": 1.0566037735849058e-05,
+      "loss": 0.4696,
+      "step": 56
+    },
+    {
+      "epoch": 0.10836501901140684,
+      "grad_norm": 0.8904440060515039,
+      "learning_rate": 1.0754716981132076e-05,
+      "loss": 0.4474,
+      "step": 57
+    },
+    {
+      "epoch": 0.11026615969581749,
+      "grad_norm": 0.9017229440485744,
+      "learning_rate": 1.0943396226415095e-05,
+      "loss": 0.4725,
+      "step": 58
+    },
+    {
+      "epoch": 0.11216730038022814,
+      "grad_norm": 0.8511345130304052,
+      "learning_rate": 1.1132075471698115e-05,
+      "loss": 0.4748,
+      "step": 59
+    },
+    {
+      "epoch": 0.11406844106463879,
+      "grad_norm": 1.0288658297945765,
+      "learning_rate": 1.1320754716981132e-05,
+      "loss": 0.4869,
+      "step": 60
+    },
+    {
+      "epoch": 0.11596958174904944,
+      "grad_norm": 0.9647446628965938,
+      "learning_rate": 1.1509433962264152e-05,
+      "loss": 0.4396,
+      "step": 61
+    },
+    {
+      "epoch": 0.11787072243346007,
+      "grad_norm": 0.8699631204750673,
+      "learning_rate": 1.169811320754717e-05,
+      "loss": 0.4679,
+      "step": 62
+    },
+    {
+      "epoch": 0.11977186311787072,
+      "grad_norm": 1.056125960232519,
+      "learning_rate": 1.188679245283019e-05,
+      "loss": 0.4526,
+      "step": 63
+    },
+    {
+      "epoch": 0.12167300380228137,
+      "grad_norm": 0.9003320928682369,
+      "learning_rate": 1.2075471698113209e-05,
+      "loss": 0.4853,
+      "step": 64
+    },
+    {
+      "epoch": 0.12357414448669202,
+      "grad_norm": 0.8738321321957669,
+      "learning_rate": 1.2264150943396227e-05,
+      "loss": 0.4751,
+      "step": 65
+    },
+    {
+      "epoch": 0.12547528517110265,
+      "grad_norm": 1.0073216402996308,
+      "learning_rate": 1.2452830188679246e-05,
+      "loss": 0.4716,
+      "step": 66
+    },
+    {
+      "epoch": 0.12737642585551331,
+      "grad_norm": 0.8887003831769286,
+      "learning_rate": 1.2641509433962264e-05,
+      "loss": 0.4569,
+      "step": 67
+    },
+    {
+      "epoch": 0.12927756653992395,
+      "grad_norm": 0.9549014707859587,
+      "learning_rate": 1.2830188679245283e-05,
+      "loss": 0.4709,
+      "step": 68
+    },
+    {
+      "epoch": 0.1311787072243346,
+      "grad_norm": 0.9743335111616435,
+      "learning_rate": 1.3018867924528303e-05,
+      "loss": 0.4815,
+      "step": 69
+    },
+    {
+      "epoch": 0.13307984790874525,
+      "grad_norm": 0.8547600310309565,
+      "learning_rate": 1.320754716981132e-05,
+      "loss": 0.4481,
+      "step": 70
+    },
+    {
+      "epoch": 0.13498098859315588,
+      "grad_norm": 0.91046608816734,
+      "learning_rate": 1.339622641509434e-05,
+      "loss": 0.4496,
+      "step": 71
+    },
+    {
+      "epoch": 0.13688212927756654,
+      "grad_norm": 1.1214645456139052,
+      "learning_rate": 1.3584905660377358e-05,
+      "loss": 0.4869,
+      "step": 72
+    },
+    {
+      "epoch": 0.13878326996197718,
+      "grad_norm": 0.9048102725737821,
+      "learning_rate": 1.3773584905660378e-05,
+      "loss": 0.4512,
+      "step": 73
+    },
+    {
+      "epoch": 0.14068441064638784,
+      "grad_norm": 1.0658303202317667,
+      "learning_rate": 1.3962264150943397e-05,
+      "loss": 0.4371,
+      "step": 74
+    },
+    {
+      "epoch": 0.14258555133079848,
+      "grad_norm": 0.9266627356242514,
+      "learning_rate": 1.4150943396226415e-05,
+      "loss": 0.4374,
+      "step": 75
+    },
+    {
+      "epoch": 0.1444866920152091,
+      "grad_norm": 1.059303770251611,
+      "learning_rate": 1.4339622641509435e-05,
+      "loss": 0.4616,
+      "step": 76
+    },
+    {
+      "epoch": 0.14638783269961977,
+      "grad_norm": 0.9421467320399862,
+      "learning_rate": 1.4528301886792452e-05,
+      "loss": 0.4713,
+      "step": 77
+    },
+    {
+      "epoch": 0.1482889733840304,
+      "grad_norm": 1.123595842472078,
+      "learning_rate": 1.4716981132075472e-05,
+      "loss": 0.4564,
+      "step": 78
+    },
+    {
+      "epoch": 0.15019011406844107,
+      "grad_norm": 1.2169438349592567,
+      "learning_rate": 1.4905660377358491e-05,
+      "loss": 0.4655,
+      "step": 79
+    },
+    {
+      "epoch": 0.1520912547528517,
+      "grad_norm": 1.0608707925225642,
+      "learning_rate": 1.5094339622641511e-05,
+      "loss": 0.4725,
+      "step": 80
+    },
+    {
+      "epoch": 0.15399239543726237,
+      "grad_norm": 1.118863040928898,
+      "learning_rate": 1.5283018867924532e-05,
+      "loss": 0.4736,
+      "step": 81
+    },
+    {
+      "epoch": 0.155893536121673,
+      "grad_norm": 0.8738681559021476,
+      "learning_rate": 1.547169811320755e-05,
+      "loss": 0.4443,
+      "step": 82
+    },
+    {
+      "epoch": 0.15779467680608364,
+      "grad_norm": 1.0416037612874256,
+      "learning_rate": 1.5660377358490568e-05,
+      "loss": 0.4628,
+      "step": 83
+    },
+    {
+      "epoch": 0.1596958174904943,
+      "grad_norm": 0.9515159299398843,
+      "learning_rate": 1.5849056603773586e-05,
+      "loss": 0.4615,
+      "step": 84
+    },
+    {
+      "epoch": 0.16159695817490494,
+      "grad_norm": 1.0627710779984239,
+      "learning_rate": 1.6037735849056607e-05,
+      "loss": 0.4608,
+      "step": 85
+    },
+    {
+      "epoch": 0.1634980988593156,
+      "grad_norm": 0.9256097055580161,
+      "learning_rate": 1.6226415094339625e-05,
+      "loss": 0.4261,
+      "step": 86
+    },
+    {
+      "epoch": 0.16539923954372623,
+      "grad_norm": 0.8584700726149466,
+      "learning_rate": 1.6415094339622643e-05,
+      "loss": 0.4651,
+      "step": 87
+    },
+    {
+      "epoch": 0.16730038022813687,
+      "grad_norm": 1.0068717817358792,
+      "learning_rate": 1.6603773584905664e-05,
+      "loss": 0.4633,
+      "step": 88
+    },
+    {
+      "epoch": 0.16920152091254753,
+      "grad_norm": 0.9879460662161074,
+      "learning_rate": 1.679245283018868e-05,
+      "loss": 0.48,
+      "step": 89
+    },
+    {
+      "epoch": 0.17110266159695817,
+      "grad_norm": 0.9778543537416002,
+      "learning_rate": 1.69811320754717e-05,
+      "loss": 0.4781,
+      "step": 90
+    },
+    {
+      "epoch": 0.17300380228136883,
+      "grad_norm": 1.0978079008157218,
+      "learning_rate": 1.716981132075472e-05,
+      "loss": 0.467,
+      "step": 91
+    },
+    {
+      "epoch": 0.17490494296577946,
+      "grad_norm": 0.9228798062434658,
+      "learning_rate": 1.735849056603774e-05,
+      "loss": 0.4777,
+      "step": 92
+    },
+    {
+      "epoch": 0.17680608365019013,
+      "grad_norm": 0.9034701469339327,
+      "learning_rate": 1.7547169811320756e-05,
+      "loss": 0.4509,
+      "step": 93
+    },
+    {
+      "epoch": 0.17870722433460076,
+      "grad_norm": 0.9757676835688418,
+      "learning_rate": 1.7735849056603774e-05,
+      "loss": 0.4664,
+      "step": 94
+    },
+    {
+      "epoch": 0.1806083650190114,
+      "grad_norm": 1.0912511192781529,
+      "learning_rate": 1.7924528301886795e-05,
+      "loss": 0.4833,
+      "step": 95
+    },
+    {
+      "epoch": 0.18250950570342206,
+      "grad_norm": 1.048921249683321,
+      "learning_rate": 1.8113207547169813e-05,
+      "loss": 0.4556,
+      "step": 96
+    },
+    {
+      "epoch": 0.1844106463878327,
+      "grad_norm": 0.8813187129058425,
+      "learning_rate": 1.830188679245283e-05,
+      "loss": 0.4737,
+      "step": 97
+    },
+    {
+      "epoch": 0.18631178707224336,
+      "grad_norm": 1.1302337711718995,
+      "learning_rate": 1.8490566037735852e-05,
+      "loss": 0.4341,
+      "step": 98
+    },
+    {
+      "epoch": 0.188212927756654,
+      "grad_norm": 0.9338776516157274,
+      "learning_rate": 1.867924528301887e-05,
+      "loss": 0.4592,
+      "step": 99
+    },
+    {
+      "epoch": 0.19011406844106463,
+      "grad_norm": 0.983508173443429,
+      "learning_rate": 1.8867924528301888e-05,
+      "loss": 0.4668,
+      "step": 100
+    },
+    {
+      "epoch": 0.1920152091254753,
+      "grad_norm": 1.0067790205062617,
+      "learning_rate": 1.905660377358491e-05,
+      "loss": 0.4803,
+      "step": 101
+    },
+    {
+      "epoch": 0.19391634980988592,
+      "grad_norm": 0.8894156066569003,
+      "learning_rate": 1.9245283018867927e-05,
+      "loss": 0.445,
+      "step": 102
+    },
+    {
+      "epoch": 0.1958174904942966,
+      "grad_norm": 1.0418288996917993,
+      "learning_rate": 1.9433962264150945e-05,
+      "loss": 0.4841,
+      "step": 103
+    },
+    {
+      "epoch": 0.19771863117870722,
+      "grad_norm": 0.9564275882768724,
+      "learning_rate": 1.9622641509433963e-05,
+      "loss": 0.4563,
+      "step": 104
+    },
+    {
+      "epoch": 0.19961977186311788,
+      "grad_norm": 0.9878994831014986,
+      "learning_rate": 1.9811320754716984e-05,
+      "loss": 0.464,
+      "step": 105
+    },
+    {
+      "epoch": 0.20152091254752852,
+      "grad_norm": 0.9933974812722013,
+      "learning_rate": 2e-05,
+      "loss": 0.4558,
+      "step": 106
+    },
+    {
+      "epoch": 0.20342205323193915,
+      "grad_norm": 0.9627766862636876,
+      "learning_rate": 1.9999944857420527e-05,
+      "loss": 0.4482,
+      "step": 107
+    },
+    {
+      "epoch": 0.20532319391634982,
+      "grad_norm": 1.0461351562569627,
+      "learning_rate": 1.9999779430290247e-05,
+      "loss": 0.4646,
+      "step": 108
+    },
+    {
+      "epoch": 0.20722433460076045,
+      "grad_norm": 0.9469004349448763,
+      "learning_rate": 1.9999503720433575e-05,
+      "loss": 0.4744,
+      "step": 109
+    },
+    {
+      "epoch": 0.20912547528517111,
+      "grad_norm": 0.954499625972231,
+      "learning_rate": 1.999911773089118e-05,
+      "loss": 0.4537,
+      "step": 110
+    },
+    {
+      "epoch": 0.21102661596958175,
+      "grad_norm": 0.8910021421190507,
+      "learning_rate": 1.999862146591996e-05,
+      "loss": 0.4421,
+      "step": 111
+    },
+    {
+      "epoch": 0.21292775665399238,
+      "grad_norm": 0.881948756838539,
+      "learning_rate": 1.9998014930992976e-05,
+      "loss": 0.4464,
+      "step": 112
+    },
+    {
+      "epoch": 0.21482889733840305,
+      "grad_norm": 0.95421425671348,
+      "learning_rate": 1.9997298132799408e-05,
+      "loss": 0.4703,
+      "step": 113
+    },
+    {
+      "epoch": 0.21673003802281368,
+      "grad_norm": 0.9295472479227239,
+      "learning_rate": 1.9996471079244477e-05,
+      "loss": 0.498,
+      "step": 114
+    },
+    {
+      "epoch": 0.21863117870722434,
+      "grad_norm": 0.9561689622245643,
+      "learning_rate": 1.999553377944936e-05,
+      "loss": 0.4618,
+      "step": 115
+    },
+    {
+      "epoch": 0.22053231939163498,
+      "grad_norm": 0.8817940872705535,
+      "learning_rate": 1.9994486243751076e-05,
+      "loss": 0.4614,
+      "step": 116
+    },
+    {
+      "epoch": 0.2224334600760456,
+      "grad_norm": 0.8776347484861887,
+      "learning_rate": 1.9993328483702393e-05,
+      "loss": 0.4586,
+      "step": 117
+    },
+    {
+      "epoch": 0.22433460076045628,
+      "grad_norm": 0.9374204740876165,
+      "learning_rate": 1.999206051207169e-05,
+      "loss": 0.454,
+      "step": 118
+    },
+    {
+      "epoch": 0.2262357414448669,
+      "grad_norm": 0.9786656671661706,
+      "learning_rate": 1.9990682342842805e-05,
+      "loss": 0.4806,
+      "step": 119
+    },
+    {
+      "epoch": 0.22813688212927757,
+      "grad_norm": 0.856870318859439,
+      "learning_rate": 1.99891939912149e-05,
+      "loss": 0.4323,
+      "step": 120
+    },
+    {
+      "epoch": 0.2300380228136882,
+      "grad_norm": 1.034947797398693,
+      "learning_rate": 1.9987595473602292e-05,
+      "loss": 0.4692,
+      "step": 121
+    },
+    {
+      "epoch": 0.23193916349809887,
+      "grad_norm": 0.878206623111752,
+      "learning_rate": 1.9985886807634246e-05,
+      "loss": 0.485,
+      "step": 122
+    },
+    {
+      "epoch": 0.2338403041825095,
+      "grad_norm": 1.1132201299094322,
+      "learning_rate": 1.9984068012154824e-05,
+      "loss": 0.4343,
+      "step": 123
+    },
+    {
+      "epoch": 0.23574144486692014,
+      "grad_norm": 0.8643937200141374,
+      "learning_rate": 1.9982139107222634e-05,
+      "loss": 0.4618,
+      "step": 124
+    },
+    {
+      "epoch": 0.2376425855513308,
+      "grad_norm": 1.0013156091599946,
+      "learning_rate": 1.9980100114110637e-05,
+      "loss": 0.4416,
+      "step": 125
+    },
+    {
+      "epoch": 0.23954372623574144,
+      "grad_norm": 0.9099456857335346,
+      "learning_rate": 1.99779510553059e-05,
+      "loss": 0.4549,
+      "step": 126
+    },
+    {
+      "epoch": 0.2414448669201521,
+      "grad_norm": 0.8588549917757148,
+      "learning_rate": 1.9975691954509347e-05,
+      "loss": 0.4669,
+      "step": 127
+    },
+    {
+      "epoch": 0.24334600760456274,
+      "grad_norm": 0.9057210459589763,
+      "learning_rate": 1.9973322836635517e-05,
+      "loss": 0.4419,
+      "step": 128
+    },
+    {
+      "epoch": 0.24524714828897337,
+      "grad_norm": 0.8583087061053347,
+      "learning_rate": 1.997084372781226e-05,
+      "loss": 0.4872,
+      "step": 129
+    },
+    {
+      "epoch": 0.24714828897338403,
+      "grad_norm": 0.9931358099526607,
+      "learning_rate": 1.9968254655380465e-05,
+      "loss": 0.4546,
+      "step": 130
+    },
+    {
+      "epoch": 0.24904942965779467,
+      "grad_norm": 0.8885187789161734,
+      "learning_rate": 1.996555564789376e-05,
+      "loss": 0.455,
+      "step": 131
+    },
+    {
+      "epoch": 0.2509505703422053,
+      "grad_norm": 1.0352484008543565,
+      "learning_rate": 1.996274673511819e-05,
+      "loss": 0.4349,
+      "step": 132
+    },
+    {
+      "epoch": 0.25285171102661597,
+      "grad_norm": 0.8829562292470634,
+      "learning_rate": 1.99598279480319e-05,
+      "loss": 0.4581,
+      "step": 133
+    },
+    {
+      "epoch": 0.25475285171102663,
+      "grad_norm": 0.9541992183731348,
+      "learning_rate": 1.9956799318824776e-05,
+      "loss": 0.4375,
+      "step": 134
+    },
+    {
+      "epoch": 0.25665399239543724,
+      "grad_norm": 0.898258420553369,
+      "learning_rate": 1.99536608808981e-05,
+      "loss": 0.4651,
+      "step": 135
+    },
+    {
+      "epoch": 0.2585551330798479,
+      "grad_norm": 0.8655564846625263,
+      "learning_rate": 1.995041266886419e-05,
+      "loss": 0.4613,
+      "step": 136
+    },
+    {
+      "epoch": 0.26045627376425856,
+      "grad_norm": 0.872133778498317,
+      "learning_rate": 1.9947054718545996e-05,
+      "loss": 0.465,
+      "step": 137
+    },
+    {
+      "epoch": 0.2623574144486692,
+      "grad_norm": 0.8195453466561612,
+      "learning_rate": 1.994358706697674e-05,
+      "loss": 0.4508,
+      "step": 138
+    },
+    {
+      "epoch": 0.26425855513307983,
+      "grad_norm": 0.8892218726218293,
+      "learning_rate": 1.9940009752399462e-05,
+      "loss": 0.4614,
+      "step": 139
+    },
+    {
+      "epoch": 0.2661596958174905,
+      "grad_norm": 0.8780507959216705,
+      "learning_rate": 1.9936322814266634e-05,
+      "loss": 0.4642,
+      "step": 140
+    },
+    {
+      "epoch": 0.26806083650190116,
+      "grad_norm": 0.903323924944171,
+      "learning_rate": 1.9932526293239713e-05,
+      "loss": 0.4699,
+      "step": 141
+    },
+    {
+      "epoch": 0.26996197718631176,
+      "grad_norm": 0.8377373498609986,
+      "learning_rate": 1.9928620231188694e-05,
+      "loss": 0.4471,
+      "step": 142
+    },
+    {
+      "epoch": 0.2718631178707224,
+      "grad_norm": 0.8735142646234093,
+      "learning_rate": 1.992460467119164e-05,
+      "loss": 0.4635,
+      "step": 143
+    },
+    {
+      "epoch": 0.2737642585551331,
+      "grad_norm": 0.8544049235394039,
+      "learning_rate": 1.992047965753422e-05,
+      "loss": 0.4687,
+      "step": 144
+    },
+    {
+      "epoch": 0.27566539923954375,
+      "grad_norm": 1.0717708305911287,
+      "learning_rate": 1.991624523570922e-05,
+      "loss": 0.4671,
+      "step": 145
+    },
+    {
+      "epoch": 0.27756653992395436,
+      "grad_norm": 0.9548291403359355,
+      "learning_rate": 1.9911901452416012e-05,
+      "loss": 0.4585,
+      "step": 146
+    },
+    {
+      "epoch": 0.279467680608365,
+      "grad_norm": 0.998989881104571,
+      "learning_rate": 1.9907448355560094e-05,
+      "loss": 0.4414,
+      "step": 147
+    },
+    {
+      "epoch": 0.2813688212927757,
+      "grad_norm": 0.9547445336828762,
+      "learning_rate": 1.9902885994252506e-05,
+      "loss": 0.4598,
+      "step": 148
+    },
+    {
+      "epoch": 0.2832699619771863,
+      "grad_norm": 0.9102699671837916,
+      "learning_rate": 1.989821441880933e-05,
+      "loss": 0.4674,
+      "step": 149
+    },
+    {
+      "epoch": 0.28517110266159695,
+      "grad_norm": 0.9106559389000519,
+      "learning_rate": 1.9893433680751105e-05,
+      "loss": 0.4466,
+      "step": 150
+    },
+    {
+      "epoch": 0.2870722433460076,
+      "grad_norm": 0.8151362937761507,
+      "learning_rate": 1.9888543832802277e-05,
+      "loss": 0.4275,
+      "step": 151
+    },
+    {
+      "epoch": 0.2889733840304182,
+      "grad_norm": 0.9719569290687768,
+      "learning_rate": 1.9883544928890612e-05,
+      "loss": 0.4482,
+      "step": 152
+    },
+    {
+      "epoch": 0.2908745247148289,
+      "grad_norm": 0.9308877035455401,
+      "learning_rate": 1.9878437024146603e-05,
+      "loss": 0.4775,
+      "step": 153
+    },
+    {
+      "epoch": 0.29277566539923955,
+      "grad_norm": 0.9160645124909511,
+      "learning_rate": 1.9873220174902857e-05,
+      "loss": 0.4589,
+      "step": 154
+    },
+    {
+      "epoch": 0.2946768060836502,
+      "grad_norm": 1.0264772587494142,
+      "learning_rate": 1.986789443869348e-05,
+      "loss": 0.4531,
+      "step": 155
+    },
+    {
+      "epoch": 0.2965779467680608,
+      "grad_norm": 0.8492645077385685,
+      "learning_rate": 1.9862459874253438e-05,
+      "loss": 0.4347,
+      "step": 156
+    },
+    {
+      "epoch": 0.2984790874524715,
+      "grad_norm": 1.049413067337185,
+      "learning_rate": 1.985691654151791e-05,
+      "loss": 0.4545,
+      "step": 157
+    },
+    {
+      "epoch": 0.30038022813688214,
+      "grad_norm": 0.8669984600562431,
+      "learning_rate": 1.9851264501621635e-05,
+      "loss": 0.4688,
+      "step": 158
+    },
+    {
+      "epoch": 0.30228136882129275,
+      "grad_norm": 1.0942795036598196,
+      "learning_rate": 1.984550381689822e-05,
+      "loss": 0.4606,
+      "step": 159
+    },
+    {
+      "epoch": 0.3041825095057034,
+      "grad_norm": 0.9916675774196649,
+      "learning_rate": 1.983963455087946e-05,
+      "loss": 0.476,
+      "step": 160
+    },
+    {
+      "epoch": 0.3060836501901141,
+      "grad_norm": 0.8659421129444885,
+      "learning_rate": 1.983365676829466e-05,
+      "loss": 0.4754,
+      "step": 161
+    },
+    {
+      "epoch": 0.30798479087452474,
+      "grad_norm": 1.0637659471663587,
+      "learning_rate": 1.982757053506989e-05,
+      "loss": 0.4272,
+      "step": 162
+    },
+    {
+      "epoch": 0.30988593155893535,
+      "grad_norm": 0.8185228080702929,
+      "learning_rate": 1.9821375918327268e-05,
+      "loss": 0.4542,
+      "step": 163
+    },
+    {
+      "epoch": 0.311787072243346,
+      "grad_norm": 0.9742812975658183,
+      "learning_rate": 1.981507298638422e-05,
+      "loss": 0.4308,
+      "step": 164
+    },
+    {
+      "epoch": 0.31368821292775667,
+      "grad_norm": 0.8773591726542533,
+      "learning_rate": 1.9808661808752735e-05,
+      "loss": 0.4519,
+      "step": 165
+    },
+    {
+      "epoch": 0.3155893536121673,
+      "grad_norm": 0.8673652144581186,
+      "learning_rate": 1.980214245613858e-05,
+      "loss": 0.4623,
+      "step": 166
+    },
+    {
+      "epoch": 0.31749049429657794,
+      "grad_norm": 0.9852151983590631,
+      "learning_rate": 1.979551500044055e-05,
+      "loss": 0.4477,
+      "step": 167
+    },
+    {
+      "epoch": 0.3193916349809886,
+      "grad_norm": 0.8134264443861172,
+      "learning_rate": 1.9788779514749635e-05,
+      "loss": 0.4666,
+      "step": 168
+    },
+    {
+      "epoch": 0.32129277566539927,
+      "grad_norm": 0.9708144153811817,
+      "learning_rate": 1.978193607334826e-05,
+      "loss": 0.4455,
+      "step": 169
+    },
+    {
+      "epoch": 0.3231939163498099,
+      "grad_norm": 0.8224223292814711,
+      "learning_rate": 1.977498475170941e-05,
+      "loss": 0.4588,
+      "step": 170
+    },
+    {
+      "epoch": 0.32509505703422054,
+      "grad_norm": 0.8109328530516361,
+      "learning_rate": 1.9767925626495857e-05,
+      "loss": 0.4544,
+      "step": 171
+    },
+    {
+      "epoch": 0.3269961977186312,
+      "grad_norm": 0.9417269949868233,
+      "learning_rate": 1.9760758775559275e-05,
+      "loss": 0.4528,
+      "step": 172
+    },
+    {
+      "epoch": 0.3288973384030418,
+      "grad_norm": 0.8419188333554508,
+      "learning_rate": 1.975348427793939e-05,
+      "loss": 0.4493,
+      "step": 173
+    },
+    {
+      "epoch": 0.33079847908745247,
+      "grad_norm": 0.9630293360485469,
+      "learning_rate": 1.9746102213863113e-05,
+      "loss": 0.4493,
+      "step": 174
+    },
+    {
+      "epoch": 0.33269961977186313,
+      "grad_norm": 0.9520386063149311,
+      "learning_rate": 1.973861266474366e-05,
+      "loss": 0.4468,
+      "step": 175
+    },
+    {
+      "epoch": 0.33460076045627374,
+      "grad_norm": 0.8769217580006831,
+      "learning_rate": 1.9731015713179643e-05,
+      "loss": 0.4502,
+      "step": 176
+    },
+    {
+      "epoch": 0.3365019011406844,
+      "grad_norm": 0.9059840699599547,
+      "learning_rate": 1.9723311442954163e-05,
+      "loss": 0.4698,
+      "step": 177
+    },
+    {
+      "epoch": 0.33840304182509506,
+      "grad_norm": 0.7975435079221801,
+      "learning_rate": 1.9715499939033883e-05,
+      "loss": 0.4271,
+      "step": 178
+    },
+    {
+      "epoch": 0.3403041825095057,
+      "grad_norm": 0.837801478572382,
+      "learning_rate": 1.9707581287568094e-05,
+      "loss": 0.4269,
+      "step": 179
+    },
+    {
+      "epoch": 0.34220532319391633,
+      "grad_norm": 0.7713580370576987,
+      "learning_rate": 1.969955557588778e-05,
+      "loss": 0.447,
+      "step": 180
+    },
+    {
+      "epoch": 0.344106463878327,
+      "grad_norm": 0.8317168804608446,
+      "learning_rate": 1.9691422892504626e-05,
+      "loss": 0.4289,
+      "step": 181
+    },
+    {
+      "epoch": 0.34600760456273766,
+      "grad_norm": 0.8866065852784972,
+      "learning_rate": 1.968318332711006e-05,
+      "loss": 0.4745,
+      "step": 182
+    },
+    {
+      "epoch": 0.34790874524714827,
+      "grad_norm": 0.8864993645896277,
+      "learning_rate": 1.9674836970574253e-05,
+      "loss": 0.4813,
+      "step": 183
+    },
+    {
+      "epoch": 0.34980988593155893,
+      "grad_norm": 0.837732419026762,
+      "learning_rate": 1.966638391494514e-05,
+      "loss": 0.4526,
+      "step": 184
+    },
+    {
+      "epoch": 0.3517110266159696,
+      "grad_norm": 0.8312073478465664,
+      "learning_rate": 1.9657824253447378e-05,
+      "loss": 0.4335,
+      "step": 185
+    },
+    {
+      "epoch": 0.35361216730038025,
+      "grad_norm": 0.8368981796860282,
+      "learning_rate": 1.9649158080481327e-05,
+      "loss": 0.4682,
+      "step": 186
+    },
+    {
+      "epoch": 0.35551330798479086,
+      "grad_norm": 0.8529165704025781,
+      "learning_rate": 1.964038549162201e-05,
+      "loss": 0.4358,
+      "step": 187
+    },
+    {
+      "epoch": 0.3574144486692015,
+      "grad_norm": 0.7986727104691629,
+      "learning_rate": 1.963150658361807e-05,
+      "loss": 0.4529,
+      "step": 188
+    },
+    {
+      "epoch": 0.3593155893536122,
+      "grad_norm": 0.7653297532232103,
+      "learning_rate": 1.962252145439068e-05,
+      "loss": 0.4166,
+      "step": 189
+    },
+    {
+      "epoch": 0.3612167300380228,
+      "grad_norm": 0.8704790398014006,
+      "learning_rate": 1.9613430203032486e-05,
+      "loss": 0.422,
+      "step": 190
+    },
+    {
+      "epoch": 0.36311787072243346,
+      "grad_norm": 0.8365928360447266,
+      "learning_rate": 1.9604232929806493e-05,
+      "loss": 0.4518,
+      "step": 191
+    },
+    {
+      "epoch": 0.3650190114068441,
+      "grad_norm": 0.8069447912135257,
+      "learning_rate": 1.9594929736144978e-05,
+      "loss": 0.4256,
+      "step": 192
+    },
+    {
+      "epoch": 0.3669201520912547,
+      "grad_norm": 0.8520223041822833,
+      "learning_rate": 1.9585520724648354e-05,
+      "loss": 0.4469,
+      "step": 193
+    },
+    {
+      "epoch": 0.3688212927756654,
+      "grad_norm": 0.8541017673131474,
+      "learning_rate": 1.957600599908406e-05,
+      "loss": 0.4388,
+      "step": 194
+    },
+    {
+      "epoch": 0.37072243346007605,
+      "grad_norm": 0.9232218320070916,
+      "learning_rate": 1.95663856643854e-05,
+      "loss": 0.4459,
+      "step": 195
+    },
+    {
+      "epoch": 0.3726235741444867,
+      "grad_norm": 0.8526035434164067,
+      "learning_rate": 1.955665982665038e-05,
+      "loss": 0.4262,
+      "step": 196
+    },
+    {
+      "epoch": 0.3745247148288973,
+      "grad_norm": 0.779760803890263,
+      "learning_rate": 1.9546828593140565e-05,
+      "loss": 0.44,
+      "step": 197
+    },
+    {
+      "epoch": 0.376425855513308,
+      "grad_norm": 0.8355206478766718,
+      "learning_rate": 1.9536892072279863e-05,
+      "loss": 0.4619,
+      "step": 198
+    },
+    {
+      "epoch": 0.37832699619771865,
+      "grad_norm": 0.9932195970421714,
+      "learning_rate": 1.9526850373653356e-05,
+      "loss": 0.474,
+      "step": 199
+    },
+    {
+      "epoch": 0.38022813688212925,
+      "grad_norm": 0.855095136433585,
+      "learning_rate": 1.9516703608006074e-05,
+      "loss": 0.4059,
+      "step": 200
+    },
+    {
+      "epoch": 0.3821292775665399,
+      "grad_norm": 0.9190716470954685,
+      "learning_rate": 1.9506451887241787e-05,
+      "loss": 0.4453,
+      "step": 201
+    },
+    {
+      "epoch": 0.3840304182509506,
+      "grad_norm": 0.8308782360983001,
+      "learning_rate": 1.949609532442176e-05,
+      "loss": 0.447,
+      "step": 202
+    },
+    {
+      "epoch": 0.38593155893536124,
+      "grad_norm": 0.88268488333017,
+      "learning_rate": 1.9485634033763507e-05,
+      "loss": 0.4259,
+      "step": 203
+    },
+    {
+      "epoch": 0.38783269961977185,
+      "grad_norm": 0.8722138579764749,
+      "learning_rate": 1.9475068130639543e-05,
+      "loss": 0.4289,
+      "step": 204
+    },
+    {
+      "epoch": 0.3897338403041825,
+      "grad_norm": 0.7884521443834249,
+      "learning_rate": 1.9464397731576093e-05,
+      "loss": 0.4508,
+      "step": 205
+    },
+    {
+      "epoch": 0.3916349809885932,
+      "grad_norm": 0.9097588323911896,
+      "learning_rate": 1.945362295425183e-05,
+      "loss": 0.4403,
+      "step": 206
+    },
+    {
+      "epoch": 0.3935361216730038,
+      "grad_norm": 0.7792637929585334,
+      "learning_rate": 1.944274391749655e-05,
+      "loss": 0.4415,
+      "step": 207
+    },
+    {
+      "epoch": 0.39543726235741444,
+      "grad_norm": 0.8838577934614718,
+      "learning_rate": 1.9431760741289886e-05,
+      "loss": 0.4252,
+      "step": 208
+    },
+    {
+      "epoch": 0.3973384030418251,
+      "grad_norm": 0.7729512760468414,
+      "learning_rate": 1.942067354675997e-05,
+      "loss": 0.4583,
+      "step": 209
+    },
+    {
+      "epoch": 0.39923954372623577,
+      "grad_norm": 0.8548205153884116,
+      "learning_rate": 1.9409482456182105e-05,
+      "loss": 0.441,
+      "step": 210
+    },
+    {
+      "epoch": 0.4011406844106464,
+      "grad_norm": 0.7808012229846085,
+      "learning_rate": 1.939818759297741e-05,
+      "loss": 0.4267,
+      "step": 211
+    },
+    {
+      "epoch": 0.40304182509505704,
+      "grad_norm": 0.7873032791037904,
+      "learning_rate": 1.9386789081711465e-05,
+      "loss": 0.4455,
+      "step": 212
+    },
+    {
+      "epoch": 0.4049429657794677,
+      "grad_norm": 0.836815122594539,
+      "learning_rate": 1.9375287048092927e-05,
+      "loss": 0.4502,
+      "step": 213
+    },
+    {
+      "epoch": 0.4068441064638783,
+      "grad_norm": 0.7323779056657036,
+      "learning_rate": 1.9363681618972166e-05,
+      "loss": 0.4359,
+      "step": 214
+    },
+    {
+      "epoch": 0.40874524714828897,
+      "grad_norm": 0.9317015009669802,
+      "learning_rate": 1.9351972922339835e-05,
+      "loss": 0.4556,
+      "step": 215
+    },
+    {
+      "epoch": 0.41064638783269963,
+      "grad_norm": 0.8156096654240479,
+      "learning_rate": 1.9340161087325483e-05,
+      "loss": 0.4425,
+      "step": 216
+    },
+    {
+      "epoch": 0.41254752851711024,
+      "grad_norm": 0.7595843907651628,
+      "learning_rate": 1.9328246244196117e-05,
+      "loss": 0.44,
+      "step": 217
+    },
+    {
+      "epoch": 0.4144486692015209,
+      "grad_norm": 0.8722011934053008,
+      "learning_rate": 1.931622852435478e-05,
+      "loss": 0.4445,
+      "step": 218
+    },
+    {
+      "epoch": 0.41634980988593157,
+      "grad_norm": 0.7657595666369396,
+      "learning_rate": 1.930410806033908e-05,
+      "loss": 0.428,
+      "step": 219
+    },
+    {
+      "epoch": 0.41825095057034223,
+      "grad_norm": 0.8909068495952155,
+      "learning_rate": 1.929188498581975e-05,
+      "loss": 0.4463,
+      "step": 220
+    },
+    {
+      "epoch": 0.42015209125475284,
+      "grad_norm": 0.7658340863163516,
+      "learning_rate": 1.9279559435599164e-05,
+      "loss": 0.4451,
+      "step": 221
+    },
+    {
+      "epoch": 0.4220532319391635,
+      "grad_norm": 0.8941867137286513,
+      "learning_rate": 1.926713154560984e-05,
+      "loss": 0.4509,
+      "step": 222
+    },
+    {
+      "epoch": 0.42395437262357416,
+      "grad_norm": 0.7812742271129935,
+      "learning_rate": 1.9254601452912972e-05,
+      "loss": 0.4632,
+      "step": 223
+    },
+    {
+      "epoch": 0.42585551330798477,
+      "grad_norm": 0.7794350992283509,
+      "learning_rate": 1.924196929569688e-05,
+      "loss": 0.4304,
+      "step": 224
+    },
+    {
+      "epoch": 0.42775665399239543,
+      "grad_norm": 0.7977586040658697,
+      "learning_rate": 1.922923521327551e-05,
+      "loss": 0.4515,
+      "step": 225
+    },
+    {
+      "epoch": 0.4296577946768061,
+      "grad_norm": 0.8149896293147532,
+      "learning_rate": 1.9216399346086893e-05,
+      "loss": 0.4351,
+      "step": 226
+    },
+    {
+      "epoch": 0.43155893536121676,
+      "grad_norm": 0.7925250416200511,
+      "learning_rate": 1.9203461835691596e-05,
+      "loss": 0.4333,
+      "step": 227
+    },
+    {
+      "epoch": 0.43346007604562736,
+      "grad_norm": 0.778975657481763,
+      "learning_rate": 1.9190422824771158e-05,
+      "loss": 0.4419,
+      "step": 228
+    },
+    {
+      "epoch": 0.435361216730038,
+      "grad_norm": 0.7121265637818893,
+      "learning_rate": 1.9177282457126515e-05,
+      "loss": 0.4335,
+      "step": 229
+    },
+    {
+      "epoch": 0.4372623574144487,
+      "grad_norm": 0.8127992463741573,
+      "learning_rate": 1.9164040877676425e-05,
+      "loss": 0.4353,
+      "step": 230
+    },
+    {
+      "epoch": 0.4391634980988593,
+      "grad_norm": 0.7426114909697389,
+      "learning_rate": 1.9150698232455853e-05,
+      "loss": 0.3969,
+      "step": 231
+    },
+    {
+      "epoch": 0.44106463878326996,
+      "grad_norm": 0.7226762563791753,
+      "learning_rate": 1.913725466861438e-05,
+      "loss": 0.42,
+      "step": 232
+    },
+    {
+      "epoch": 0.4429657794676806,
+      "grad_norm": 0.8469526644623859,
+      "learning_rate": 1.9123710334414552e-05,
+      "loss": 0.452,
+      "step": 233
+    },
+    {
+      "epoch": 0.4448669201520912,
+      "grad_norm": 0.8752496545093699,
+      "learning_rate": 1.911006537923029e-05,
+      "loss": 0.4382,
+      "step": 234
+    },
+    {
+      "epoch": 0.4467680608365019,
+      "grad_norm": 0.811469962550978,
+      "learning_rate": 1.9096319953545186e-05,
+      "loss": 0.4375,
+      "step": 235
+    },
+    {
+      "epoch": 0.44866920152091255,
+      "grad_norm": 0.8302125029201622,
+      "learning_rate": 1.908247420895089e-05,
+      "loss": 0.4389,
+      "step": 236
+    },
+    {
+      "epoch": 0.4505703422053232,
+      "grad_norm": 0.8132631388071457,
+      "learning_rate": 1.9068528298145418e-05,
+      "loss": 0.4212,
+      "step": 237
+    },
+    {
+      "epoch": 0.4524714828897338,
+      "grad_norm": 0.7884169167057629,
+      "learning_rate": 1.905448237493147e-05,
+      "loss": 0.4587,
+      "step": 238
+    },
+    {
+      "epoch": 0.4543726235741445,
+      "grad_norm": 0.742051297363074,
+      "learning_rate": 1.9040336594214727e-05,
+      "loss": 0.4188,
+      "step": 239
+    },
+    {
+      "epoch": 0.45627376425855515,
+      "grad_norm": 0.7321040605028221,
+      "learning_rate": 1.9026091112002163e-05,
+      "loss": 0.4157,
+      "step": 240
+    },
+    {
+      "epoch": 0.45817490494296575,
+      "grad_norm": 0.8444005729224642,
+      "learning_rate": 1.90117460854003e-05,
+      "loss": 0.4214,
+      "step": 241
+    },
+    {
+      "epoch": 0.4600760456273764,
+      "grad_norm": 0.84642492928624,
+      "learning_rate": 1.8997301672613496e-05,
+      "loss": 0.4655,
+      "step": 242
+    },
+    {
+      "epoch": 0.4619771863117871,
+      "grad_norm": 0.7498575289820406,
+      "learning_rate": 1.8982758032942184e-05,
+      "loss": 0.4417,
+      "step": 243
+    },
+    {
+      "epoch": 0.46387832699619774,
+      "grad_norm": 0.8298220364426889,
+      "learning_rate": 1.896811532678113e-05,
+      "loss": 0.4512,
+      "step": 244
+    },
+    {
+      "epoch": 0.46577946768060835,
+      "grad_norm": 0.7755540978916351,
+      "learning_rate": 1.8953373715617646e-05,
+      "loss": 0.4561,
+      "step": 245
+    },
+    {
+      "epoch": 0.467680608365019,
+      "grad_norm": 0.7989988137995606,
+      "learning_rate": 1.893853336202983e-05,
+      "loss": 0.4257,
+      "step": 246
+    },
+    {
+      "epoch": 0.4695817490494297,
+      "grad_norm": 0.8121893812993698,
+      "learning_rate": 1.892359442968475e-05,
+      "loss": 0.4183,
+      "step": 247
+    },
+    {
+      "epoch": 0.4714828897338403,
+      "grad_norm": 0.7224294221718089,
+      "learning_rate": 1.8908557083336668e-05,
+      "loss": 0.4353,
+      "step": 248
+    },
+    {
+      "epoch": 0.47338403041825095,
+      "grad_norm": 0.8833757832288558,
+      "learning_rate": 1.889342148882519e-05,
+      "loss": 0.4285,
+      "step": 249
+    },
+    {
+      "epoch": 0.4752851711026616,
+      "grad_norm": 0.7645832762291901,
+      "learning_rate": 1.8878187813073465e-05,
+      "loss": 0.442,
+      "step": 250
+    },
+    {
+      "epoch": 0.47718631178707227,
+      "grad_norm": 0.800748849519924,
+      "learning_rate": 1.886285622408633e-05,
+      "loss": 0.4178,
+      "step": 251
+    },
+    {
+      "epoch": 0.4790874524714829,
+      "grad_norm": 0.7428932949792251,
+      "learning_rate": 1.8847426890948447e-05,
+      "loss": 0.4273,
+      "step": 252
+    },
+    {
+      "epoch": 0.48098859315589354,
+      "grad_norm": 0.7926622713697272,
+      "learning_rate": 1.8831899983822475e-05,
+      "loss": 0.439,
+      "step": 253
+    },
+    {
+      "epoch": 0.4828897338403042,
+      "grad_norm": 0.7469028507529291,
+      "learning_rate": 1.8816275673947148e-05,
+      "loss": 0.418,
+      "step": 254
+    },
+    {
+      "epoch": 0.4847908745247148,
+      "grad_norm": 0.726549702357202,
+      "learning_rate": 1.8800554133635417e-05,
+      "loss": 0.4186,
+      "step": 255
+    },
+    {
+      "epoch": 0.4866920152091255,
+      "grad_norm": 0.8125231679705398,
+      "learning_rate": 1.8784735536272543e-05,
+      "loss": 0.4335,
+      "step": 256
+    },
+    {
+      "epoch": 0.48859315589353614,
+      "grad_norm": 0.7451793080164002,
+      "learning_rate": 1.8768820056314173e-05,
+      "loss": 0.4163,
+      "step": 257
+    },
+    {
+      "epoch": 0.49049429657794674,
+      "grad_norm": 0.7684852856647484,
+      "learning_rate": 1.875280786928444e-05,
+      "loss": 0.4313,
+      "step": 258
+    },
+    {
+      "epoch": 0.4923954372623574,
+      "grad_norm": 0.8079094757232325,
+      "learning_rate": 1.873669915177399e-05,
+      "loss": 0.4345,
+      "step": 259
+    },
+    {
+      "epoch": 0.49429657794676807,
+      "grad_norm": 0.8035667897246718,
+      "learning_rate": 1.872049408143808e-05,
+      "loss": 0.4263,
+      "step": 260
+    },
+    {
+      "epoch": 0.49619771863117873,
+      "grad_norm": 0.7375402683009233,
+      "learning_rate": 1.8704192836994578e-05,
+      "loss": 0.414,
+      "step": 261
+    },
+    {
+      "epoch": 0.49809885931558934,
+      "grad_norm": 0.7386371872437899,
+      "learning_rate": 1.8687795598222024e-05,
+      "loss": 0.4322,
+      "step": 262
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.731815657204626,
+      "learning_rate": 1.8671302545957628e-05,
+      "loss": 0.4319,
+      "step": 263
+    },
+    {
+      "epoch": 0.5019011406844106,
+      "grad_norm": 0.7469987401171589,
+      "learning_rate": 1.8654713862095272e-05,
+      "loss": 0.4259,
+      "step": 264
+    },
+    {
+      "epoch": 0.5038022813688213,
+      "grad_norm": 0.7222427963242233,
+      "learning_rate": 1.8638029729583524e-05,
+      "loss": 0.4346,
+      "step": 265
+    },
+    {
+      "epoch": 0.5057034220532319,
+      "grad_norm": 0.7201352219086354,
+      "learning_rate": 1.8621250332423603e-05,
+      "loss": 0.4233,
+      "step": 266
+    },
+    {
+      "epoch": 0.5076045627376425,
+      "grad_norm": 0.7794825817446216,
+      "learning_rate": 1.860437585566736e-05,
+      "loss": 0.4237,
+      "step": 267
+    },
+    {
+      "epoch": 0.5095057034220533,
+      "grad_norm": 0.792280272172557,
+      "learning_rate": 1.8587406485415226e-05,
+      "loss": 0.4269,
+      "step": 268
+    },
+    {
+      "epoch": 0.5114068441064639,
+      "grad_norm": 0.8567072388829465,
+      "learning_rate": 1.8570342408814173e-05,
+      "loss": 0.4418,
+      "step": 269
+    },
+    {
+      "epoch": 0.5133079847908745,
+      "grad_norm": 0.7234891722286381,
+      "learning_rate": 1.855318381405564e-05,
+      "loss": 0.444,
+      "step": 270
+    },
+    {
+      "epoch": 0.5152091254752852,
+      "grad_norm": 0.7760225742874595,
+      "learning_rate": 1.8535930890373467e-05,
+      "loss": 0.438,
+      "step": 271
+    },
+    {
+      "epoch": 0.5171102661596958,
+      "grad_norm": 0.7228428146663275,
+      "learning_rate": 1.8518583828041787e-05,
+      "loss": 0.4342,
+      "step": 272
+    },
+    {
+      "epoch": 0.5190114068441065,
+      "grad_norm": 0.7402819670722693,
+      "learning_rate": 1.8501142818372964e-05,
+      "loss": 0.4224,
+      "step": 273
+    },
+    {
+      "epoch": 0.5209125475285171,
+      "grad_norm": 0.7487405680816577,
+      "learning_rate": 1.848360805371544e-05,
+      "loss": 0.4323,
+      "step": 274
+    },
+    {
+      "epoch": 0.5228136882129277,
+      "grad_norm": 0.7220003919604745,
+      "learning_rate": 1.8465979727451653e-05,
+      "loss": 0.428,
+      "step": 275
+    },
+    {
+      "epoch": 0.5247148288973384,
+      "grad_norm": 0.7054749602308339,
+      "learning_rate": 1.8448258033995877e-05,
+      "loss": 0.4168,
+      "step": 276
+    },
+    {
+      "epoch": 0.526615969581749,
+      "grad_norm": 0.7502350521037227,
+      "learning_rate": 1.8430443168792087e-05,
+      "loss": 0.4047,
+      "step": 277
+    },
+    {
+      "epoch": 0.5285171102661597,
+      "grad_norm": 0.7686285896251883,
+      "learning_rate": 1.8412535328311813e-05,
+      "loss": 0.4443,
+      "step": 278
+    },
+    {
+      "epoch": 0.5304182509505704,
+      "grad_norm": 0.7927365995252301,
+      "learning_rate": 1.8394534710051956e-05,
+      "loss": 0.4417,
+      "step": 279
+    },
+    {
+      "epoch": 0.532319391634981,
+      "grad_norm": 0.8213495945615575,
+      "learning_rate": 1.8376441512532617e-05,
+      "loss": 0.4307,
+      "step": 280
+    },
+    {
+      "epoch": 0.5342205323193916,
+      "grad_norm": 0.7032563748938673,
+      "learning_rate": 1.835825593529492e-05,
+      "loss": 0.4069,
+      "step": 281
+    },
+    {
+      "epoch": 0.5361216730038023,
+      "grad_norm": 0.7441191117126702,
+      "learning_rate": 1.833997817889878e-05,
+      "loss": 0.4291,
+      "step": 282
+    },
+    {
+      "epoch": 0.5380228136882129,
+      "grad_norm": 0.7241783973052351,
+      "learning_rate": 1.8321608444920738e-05,
+      "loss": 0.4145,
+      "step": 283
+    },
+    {
+      "epoch": 0.5399239543726235,
+      "grad_norm": 0.7484445746296108,
+      "learning_rate": 1.830314693595169e-05,
+      "loss": 0.4135,
+      "step": 284
+    },
+    {
+      "epoch": 0.5418250950570342,
+      "grad_norm": 0.8538722690347338,
+      "learning_rate": 1.828459385559468e-05,
+      "loss": 0.4136,
+      "step": 285
+    },
+    {
+      "epoch": 0.5437262357414449,
+      "grad_norm": 0.7376394827745908,
+      "learning_rate": 1.8265949408462657e-05,
+      "loss": 0.4168,
+      "step": 286
+    },
+    {
+      "epoch": 0.5456273764258555,
+      "grad_norm": 1.0386407122446215,
+      "learning_rate": 1.8247213800176192e-05,
+      "loss": 0.4224,
+      "step": 287
+    },
+    {
+      "epoch": 0.5475285171102662,
+      "grad_norm": 0.7909110002779745,
+      "learning_rate": 1.8228387237361245e-05,
+      "loss": 0.4194,
+      "step": 288
+    },
+    {
+      "epoch": 0.5494296577946768,
+      "grad_norm": 0.7696181110928274,
+      "learning_rate": 1.8209469927646863e-05,
+      "loss": 0.4203,
+      "step": 289
+    },
+    {
+      "epoch": 0.5513307984790875,
+      "grad_norm": 0.6512287092996597,
+      "learning_rate": 1.8190462079662897e-05,
+      "loss": 0.4385,
+      "step": 290
+    },
+    {
+      "epoch": 0.5532319391634981,
+      "grad_norm": 0.924397166710462,
+      "learning_rate": 1.81713639030377e-05,
+      "loss": 0.4323,
+      "step": 291
+    },
+    {
+      "epoch": 0.5551330798479087,
+      "grad_norm": 0.6847522798183588,
+      "learning_rate": 1.8152175608395814e-05,
+      "loss": 0.4091,
+      "step": 292
+    },
+    {
+      "epoch": 0.5570342205323194,
+      "grad_norm": 0.9282390793372779,
+      "learning_rate": 1.8132897407355657e-05,
+      "loss": 0.4254,
+      "step": 293
+    },
+    {
+      "epoch": 0.55893536121673,
+      "grad_norm": 0.6922619595115697,
+      "learning_rate": 1.811352951252717e-05,
+      "loss": 0.4061,
+      "step": 294
+    },
+    {
+      "epoch": 0.5608365019011406,
+      "grad_norm": 0.7540259167363303,
+      "learning_rate": 1.809407213750949e-05,
+      "loss": 0.4188,
+      "step": 295
+    },
+    {
+      "epoch": 0.5627376425855514,
+      "grad_norm": 0.7458968624633368,
+      "learning_rate": 1.807452549688859e-05,
+      "loss": 0.4289,
+      "step": 296
+    },
+    {
+      "epoch": 0.564638783269962,
+      "grad_norm": 0.7677608540472626,
+      "learning_rate": 1.8054889806234906e-05,
+      "loss": 0.448,
+      "step": 297
+    },
+    {
+      "epoch": 0.5665399239543726,
+      "grad_norm": 0.7616409342811735,
+      "learning_rate": 1.8035165282100963e-05,
+      "loss": 0.4146,
+      "step": 298
+    },
+    {
+      "epoch": 0.5684410646387833,
+      "grad_norm": 0.7281790251630419,
+      "learning_rate": 1.8015352142018984e-05,
+      "loss": 0.4414,
+      "step": 299
+    },
+    {
+      "epoch": 0.5703422053231939,
+      "grad_norm": 0.7048740601703949,
+      "learning_rate": 1.799545060449851e-05,
+      "loss": 0.4156,
+      "step": 300
+    },
+    {
+      "epoch": 0.5722433460076045,
+      "grad_norm": 0.7259407256781026,
+      "learning_rate": 1.797546088902396e-05,
+      "loss": 0.4061,
+      "step": 301
+    },
+    {
+      "epoch": 0.5741444866920152,
+      "grad_norm": 0.7457503064059046,
+      "learning_rate": 1.7955383216052224e-05,
+      "loss": 0.4202,
+      "step": 302
+    },
+    {
+      "epoch": 0.5760456273764258,
+      "grad_norm": 0.6914787549274968,
+      "learning_rate": 1.7935217807010238e-05,
+      "loss": 0.4182,
+      "step": 303
+    },
+    {
+      "epoch": 0.5779467680608364,
+      "grad_norm": 0.7604985656488586,
+      "learning_rate": 1.7914964884292543e-05,
+      "loss": 0.4475,
+      "step": 304
+    },
+    {
+      "epoch": 0.5798479087452472,
+      "grad_norm": 0.689044727068512,
+      "learning_rate": 1.7894624671258813e-05,
+      "loss": 0.437,
+      "step": 305
+    },
+    {
+      "epoch": 0.5817490494296578,
+      "grad_norm": 0.72531642776673,
+      "learning_rate": 1.7874197392231414e-05,
+      "loss": 0.4198,
+      "step": 306
+    },
+    {
+      "epoch": 0.5836501901140685,
+      "grad_norm": 0.7328286956033915,
+      "learning_rate": 1.7853683272492913e-05,
+      "loss": 0.4251,
+      "step": 307
+    },
+    {
+      "epoch": 0.5855513307984791,
+      "grad_norm": 0.6897624363404723,
+      "learning_rate": 1.7833082538283615e-05,
+      "loss": 0.4161,
+      "step": 308
+    },
+    {
+      "epoch": 0.5874524714828897,
+      "grad_norm": 0.6900098828242334,
+      "learning_rate": 1.7812395416799034e-05,
+      "loss": 0.4107,
+      "step": 309
+    },
+    {
+      "epoch": 0.5893536121673004,
+      "grad_norm": 0.7465621825102452,
+      "learning_rate": 1.7791622136187422e-05,
+      "loss": 0.44,
+      "step": 310
+    },
+    {
+      "epoch": 0.591254752851711,
+      "grad_norm": 0.7345136183881741,
+      "learning_rate": 1.7770762925547235e-05,
+      "loss": 0.4202,
+      "step": 311
+    },
+    {
+      "epoch": 0.5931558935361216,
+      "grad_norm": 0.7253546667989501,
+      "learning_rate": 1.7749818014924612e-05,
+      "loss": 0.4163,
+      "step": 312
+    },
+    {
+      "epoch": 0.5950570342205324,
+      "grad_norm": 0.666542515874834,
+      "learning_rate": 1.7728787635310828e-05,
+      "loss": 0.4126,
+      "step": 313
+    },
+    {
+      "epoch": 0.596958174904943,
+      "grad_norm": 0.7222911458607845,
+      "learning_rate": 1.770767201863976e-05,
+      "loss": 0.4313,
+      "step": 314
+    },
+    {
+      "epoch": 0.5988593155893536,
+      "grad_norm": 0.7141059449247804,
+      "learning_rate": 1.7686471397785322e-05,
+      "loss": 0.4033,
+      "step": 315
+    },
+    {
+      "epoch": 0.6007604562737643,
+      "grad_norm": 0.7193832549823179,
+      "learning_rate": 1.76651860065589e-05,
+      "loss": 0.4169,
+      "step": 316
+    },
+    {
+      "epoch": 0.6026615969581749,
+      "grad_norm": 0.7363534273713959,
+      "learning_rate": 1.764381607970677e-05,
+      "loss": 0.4319,
+      "step": 317
+    },
+    {
+      "epoch": 0.6045627376425855,
+      "grad_norm": 0.7346847667434095,
+      "learning_rate": 1.7622361852907506e-05,
+      "loss": 0.4303,
+      "step": 318
+    },
+    {
+      "epoch": 0.6064638783269962,
+      "grad_norm": 0.8258380951944077,
+      "learning_rate": 1.760082356276939e-05,
+      "loss": 0.4382,
+      "step": 319
+    },
+    {
+      "epoch": 0.6083650190114068,
+      "grad_norm": 0.7562717110545585,
+      "learning_rate": 1.75792014468278e-05,
+      "loss": 0.4236,
+      "step": 320
+    },
+    {
+      "epoch": 0.6102661596958175,
+      "grad_norm": 0.8954664719538068,
+      "learning_rate": 1.7557495743542586e-05,
+      "loss": 0.4319,
+      "step": 321
+    },
+    {
+      "epoch": 0.6121673003802282,
+      "grad_norm": 0.7633487386695192,
+      "learning_rate": 1.7535706692295436e-05,
+      "loss": 0.4341,
+      "step": 322
+    },
+    {
+      "epoch": 0.6140684410646388,
+      "grad_norm": 0.8968917497611884,
+      "learning_rate": 1.7513834533387256e-05,
+      "loss": 0.4169,
+      "step": 323
+    },
+    {
+      "epoch": 0.6159695817490495,
+      "grad_norm": 0.7273589279653885,
+      "learning_rate": 1.7491879508035488e-05,
+      "loss": 0.4197,
+      "step": 324
+    },
+    {
+      "epoch": 0.6178707224334601,
+      "grad_norm": 0.8591211381908581,
+      "learning_rate": 1.746984185837149e-05,
+      "loss": 0.4144,
+      "step": 325
+    },
+    {
+      "epoch": 0.6197718631178707,
+      "grad_norm": 0.8061042433057314,
+      "learning_rate": 1.744772182743782e-05,
+      "loss": 0.4119,
+      "step": 326
+    },
+    {
+      "epoch": 0.6216730038022814,
+      "grad_norm": 0.7534129267419699,
+      "learning_rate": 1.7425519659185596e-05,
+      "loss": 0.41,
+      "step": 327
+    },
+    {
+      "epoch": 0.623574144486692,
+      "grad_norm": 0.9148321836671925,
+      "learning_rate": 1.740323559847179e-05,
+      "loss": 0.4342,
+      "step": 328
+    },
+    {
+      "epoch": 0.6254752851711026,
+      "grad_norm": 0.6791886686377647,
+      "learning_rate": 1.738086989105651e-05,
+      "loss": 0.4018,
+      "step": 329
+    },
+    {
+      "epoch": 0.6273764258555133,
+      "grad_norm": 1.0459542869638965,
+      "learning_rate": 1.735842278360032e-05,
+      "loss": 0.4422,
+      "step": 330
+    },
+    {
+      "epoch": 0.629277566539924,
+      "grad_norm": 0.7030132521807193,
+      "learning_rate": 1.73358945236615e-05,
+      "loss": 0.4296,
+      "step": 331
+    },
+    {
+      "epoch": 0.6311787072243346,
+      "grad_norm": 0.919498286127634,
+      "learning_rate": 1.7313285359693322e-05,
+      "loss": 0.4451,
+      "step": 332
+    },
+    {
+      "epoch": 0.6330798479087453,
+      "grad_norm": 0.7150204352803702,
+      "learning_rate": 1.7290595541041312e-05,
+      "loss": 0.4023,
+      "step": 333
+    },
+    {
+      "epoch": 0.6349809885931559,
+      "grad_norm": 0.8567506790955526,
+      "learning_rate": 1.7267825317940494e-05,
+      "loss": 0.4354,
+      "step": 334
+    },
+    {
+      "epoch": 0.6368821292775665,
+      "grad_norm": 0.8011540745154682,
+      "learning_rate": 1.724497494151264e-05,
+      "loss": 0.4187,
+      "step": 335
+    },
+    {
+      "epoch": 0.6387832699619772,
+      "grad_norm": 0.7277965923638402,
+      "learning_rate": 1.7222044663763484e-05,
+      "loss": 0.43,
+      "step": 336
+    },
+    {
+      "epoch": 0.6406844106463878,
+      "grad_norm": 0.7690874892770839,
+      "learning_rate": 1.7199034737579962e-05,
+      "loss": 0.4209,
+      "step": 337
+    },
+    {
+      "epoch": 0.6425855513307985,
+      "grad_norm": 0.7037131345046957,
+      "learning_rate": 1.7175945416727405e-05,
+      "loss": 0.428,
+      "step": 338
+    },
+    {
+      "epoch": 0.6444866920152091,
+      "grad_norm": 0.7626280803476387,
+      "learning_rate": 1.7152776955846768e-05,
+      "loss": 0.4102,
+      "step": 339
+    },
+    {
+      "epoch": 0.6463878326996197,
+      "grad_norm": 0.7385217982949507,
+      "learning_rate": 1.7129529610451775e-05,
+      "loss": 0.3994,
+      "step": 340
+    },
+    {
+      "epoch": 0.6482889733840305,
+      "grad_norm": 0.7187244242481953,
+      "learning_rate": 1.7106203636926154e-05,
+      "loss": 0.4308,
+      "step": 341
+    },
+    {
+      "epoch": 0.6501901140684411,
+      "grad_norm": 0.8606984160843195,
+      "learning_rate": 1.7082799292520767e-05,
+      "loss": 0.4198,
+      "step": 342
+    },
+    {
+      "epoch": 0.6520912547528517,
+      "grad_norm": 0.7009774797503668,
+      "learning_rate": 1.7059316835350806e-05,
+      "loss": 0.4257,
+      "step": 343
+    },
+    {
+      "epoch": 0.6539923954372624,
+      "grad_norm": 0.7461232878006195,
+      "learning_rate": 1.7035756524392924e-05,
+      "loss": 0.4249,
+      "step": 344
+    },
+    {
+      "epoch": 0.655893536121673,
+      "grad_norm": 0.7606488253448768,
+      "learning_rate": 1.7012118619482376e-05,
+      "loss": 0.4234,
+      "step": 345
+    },
+    {
+      "epoch": 0.6577946768060836,
+      "grad_norm": 0.6935989430663162,
+      "learning_rate": 1.6988403381310177e-05,
+      "loss": 0.4389,
+      "step": 346
+    },
+    {
+      "epoch": 0.6596958174904943,
+      "grad_norm": 0.8721686274410207,
+      "learning_rate": 1.696461107142021e-05,
+      "loss": 0.4363,
+      "step": 347
+    },
+    {
+      "epoch": 0.6615969581749049,
+      "grad_norm": 0.7486290672013655,
+      "learning_rate": 1.6940741952206342e-05,
+      "loss": 0.4085,
+      "step": 348
+    },
+    {
+      "epoch": 0.6634980988593155,
+      "grad_norm": 0.7924011353875043,
+      "learning_rate": 1.691679628690953e-05,
+      "loss": 0.4224,
+      "step": 349
+    },
+    {
+      "epoch": 0.6653992395437263,
+      "grad_norm": 0.8711338432182801,
+      "learning_rate": 1.6892774339614927e-05,
+      "loss": 0.4041,
+      "step": 350
+    },
+    {
+      "epoch": 0.6673003802281369,
+      "grad_norm": 0.7540681990315626,
+      "learning_rate": 1.686867637524896e-05,
+      "loss": 0.4257,
+      "step": 351
+    },
+    {
+      "epoch": 0.6692015209125475,
+      "grad_norm": 0.9765443697552242,
+      "learning_rate": 1.6844502659576414e-05,
+      "loss": 0.4235,
+      "step": 352
+    },
+    {
+      "epoch": 0.6711026615969582,
+      "grad_norm": 0.7121523644643962,
+      "learning_rate": 1.6820253459197493e-05,
+      "loss": 0.4447,
+      "step": 353
+    },
+    {
+      "epoch": 0.6730038022813688,
+      "grad_norm": 0.7507533053961629,
+      "learning_rate": 1.679592904154489e-05,
+      "loss": 0.4338,
+      "step": 354
+    },
+    {
+      "epoch": 0.6749049429657795,
+      "grad_norm": 0.9278221085939046,
+      "learning_rate": 1.677152967488084e-05,
+      "loss": 0.4233,
+      "step": 355
+    },
+    {
+      "epoch": 0.6768060836501901,
+      "grad_norm": 0.7206998423094333,
+      "learning_rate": 1.6747055628294134e-05,
+      "loss": 0.4307,
+      "step": 356
+    },
+    {
+      "epoch": 0.6787072243346007,
+      "grad_norm": 0.9440748630799297,
+      "learning_rate": 1.6722507171697184e-05,
+      "loss": 0.4248,
+      "step": 357
+    },
+    {
+      "epoch": 0.6806083650190115,
+      "grad_norm": 0.6999746396983816,
+      "learning_rate": 1.669788457582304e-05,
+      "loss": 0.4351,
+      "step": 358
+    },
+    {
+      "epoch": 0.6825095057034221,
+      "grad_norm": 0.7573791618414891,
+      "learning_rate": 1.6673188112222394e-05,
+      "loss": 0.4125,
+      "step": 359
+    },
+    {
+      "epoch": 0.6844106463878327,
+      "grad_norm": 0.9048870595823651,
+      "learning_rate": 1.6648418053260585e-05,
+      "loss": 0.4171,
+      "step": 360
+    },
+    {
+      "epoch": 0.6863117870722434,
+      "grad_norm": 0.6635862452648891,
+      "learning_rate": 1.6623574672114596e-05,
+      "loss": 0.4007,
+      "step": 361
+    },
+    {
+      "epoch": 0.688212927756654,
+      "grad_norm": 0.7521290517721314,
+      "learning_rate": 1.6598658242770054e-05,
+      "loss": 0.4253,
+      "step": 362
+    },
+    {
+      "epoch": 0.6901140684410646,
+      "grad_norm": 0.7922495734166638,
+      "learning_rate": 1.6573669040018202e-05,
+      "loss": 0.3983,
+      "step": 363
+    },
+    {
+      "epoch": 0.6920152091254753,
+      "grad_norm": 0.7561100637768147,
+      "learning_rate": 1.6548607339452853e-05,
+      "loss": 0.4031,
+      "step": 364
+    },
+    {
+      "epoch": 0.6939163498098859,
+      "grad_norm": 0.8328393439881416,
+      "learning_rate": 1.652347341746737e-05,
+      "loss": 0.421,
+      "step": 365
+    },
+    {
+      "epoch": 0.6958174904942965,
+      "grad_norm": 0.801331275005731,
+      "learning_rate": 1.6498267551251618e-05,
+      "loss": 0.4178,
+      "step": 366
+    },
+    {
+      "epoch": 0.6977186311787072,
+      "grad_norm": 0.7223213594274395,
+      "learning_rate": 1.6472990018788884e-05,
+      "loss": 0.4045,
+      "step": 367
+    },
+    {
+      "epoch": 0.6996197718631179,
+      "grad_norm": 0.7871163078855385,
+      "learning_rate": 1.644764109885284e-05,
+      "loss": 0.4406,
+      "step": 368
+    },
+    {
+      "epoch": 0.7015209125475285,
+      "grad_norm": 0.8119183562612368,
+      "learning_rate": 1.642222107100446e-05,
+      "loss": 0.4268,
+      "step": 369
+    },
+    {
+      "epoch": 0.7034220532319392,
+      "grad_norm": 0.8747372781541254,
+      "learning_rate": 1.6396730215588913e-05,
+      "loss": 0.4211,
+      "step": 370
+    },
+    {
+      "epoch": 0.7053231939163498,
+      "grad_norm": 0.7814461067155059,
+      "learning_rate": 1.6371168813732514e-05,
+      "loss": 0.4208,
+      "step": 371
+    },
+    {
+      "epoch": 0.7072243346007605,
+      "grad_norm": 0.812697926703491,
+      "learning_rate": 1.6345537147339578e-05,
+      "loss": 0.4175,
+      "step": 372
+    },
+    {
+      "epoch": 0.7091254752851711,
+      "grad_norm": 0.7623483879017262,
+      "learning_rate": 1.6319835499089358e-05,
+      "loss": 0.4025,
+      "step": 373
+    },
+    {
+      "epoch": 0.7110266159695817,
+      "grad_norm": 0.7274602955652308,
+      "learning_rate": 1.6294064152432878e-05,
+      "loss": 0.4169,
+      "step": 374
+    },
+    {
+      "epoch": 0.7129277566539924,
+      "grad_norm": 0.745438431555376,
+      "learning_rate": 1.626822339158985e-05,
+      "loss": 0.3919,
+      "step": 375
+    },
+    {
+      "epoch": 0.714828897338403,
+      "grad_norm": 0.742991672160718,
+      "learning_rate": 1.6242313501545522e-05,
+      "loss": 0.4106,
+      "step": 376
+    },
+    {
+      "epoch": 0.7167300380228137,
+      "grad_norm": 0.7085708666655373,
+      "learning_rate": 1.621633476804752e-05,
+      "loss": 0.4215,
+      "step": 377
+    },
+    {
+      "epoch": 0.7186311787072244,
+      "grad_norm": 0.8091290242475566,
+      "learning_rate": 1.6190287477602716e-05,
+      "loss": 0.4254,
+      "step": 378
+    },
+    {
+      "epoch": 0.720532319391635,
+      "grad_norm": 0.6987030721797918,
+      "learning_rate": 1.6164171917474078e-05,
+      "loss": 0.4173,
+      "step": 379
+    },
+    {
+      "epoch": 0.7224334600760456,
+      "grad_norm": 0.7608436716389589,
+      "learning_rate": 1.6137988375677466e-05,
+      "loss": 0.442,
+      "step": 380
+    },
+    {
+      "epoch": 0.7243346007604563,
+      "grad_norm": 0.7175313634173877,
+      "learning_rate": 1.6111737140978495e-05,
+      "loss": 0.4116,
+      "step": 381
+    },
+    {
+      "epoch": 0.7262357414448669,
+      "grad_norm": 0.7206709428639054,
+      "learning_rate": 1.6085418502889315e-05,
+      "loss": 0.4278,
+      "step": 382
+    },
+    {
+      "epoch": 0.7281368821292775,
+      "grad_norm": 0.7632768801437255,
+      "learning_rate": 1.6059032751665454e-05,
+      "loss": 0.4444,
+      "step": 383
+    },
+    {
+      "epoch": 0.7300380228136882,
+      "grad_norm": 0.7426501654629504,
+      "learning_rate": 1.6032580178302585e-05,
+      "loss": 0.446,
+      "step": 384
+    },
+    {
+      "epoch": 0.7319391634980988,
+      "grad_norm": 0.6794260257579373,
+      "learning_rate": 1.600606107453333e-05,
+      "loss": 0.4096,
+      "step": 385
+    },
+    {
+      "epoch": 0.7338403041825095,
+      "grad_norm": 0.7148440217345758,
+      "learning_rate": 1.597947573282405e-05,
+      "loss": 0.4163,
+      "step": 386
+    },
+    {
+      "epoch": 0.7357414448669202,
+      "grad_norm": 0.6936440826498409,
+      "learning_rate": 1.5952824446371608e-05,
+      "loss": 0.4156,
+      "step": 387
+    },
+    {
+      "epoch": 0.7376425855513308,
+      "grad_norm": 0.7174302108344963,
+      "learning_rate": 1.592610750910014e-05,
+      "loss": 0.4115,
+      "step": 388
+    },
+    {
+      "epoch": 0.7395437262357415,
+      "grad_norm": 0.6796060953523526,
+      "learning_rate": 1.589932521565781e-05,
+      "loss": 0.4061,
+      "step": 389
+    },
+    {
+      "epoch": 0.7414448669201521,
+      "grad_norm": 0.6996507854742791,
+      "learning_rate": 1.587247786141358e-05,
+      "loss": 0.4187,
+      "step": 390
+    },
+    {
+      "epoch": 0.7433460076045627,
+      "grad_norm": 0.6921394972920323,
+      "learning_rate": 1.5845565742453906e-05,
+      "loss": 0.4318,
+      "step": 391
+    },
+    {
+      "epoch": 0.7452471482889734,
+      "grad_norm": 0.7374983200834438,
+      "learning_rate": 1.581858915557953e-05,
+      "loss": 0.4157,
+      "step": 392
+    },
+    {
+      "epoch": 0.747148288973384,
+      "grad_norm": 0.7301619058569485,
+      "learning_rate": 1.5791548398302167e-05,
+      "loss": 0.4269,
+      "step": 393
+    },
+    {
+      "epoch": 0.7490494296577946,
+      "grad_norm": 0.6636788002986177,
+      "learning_rate": 1.5764443768841234e-05,
+      "loss": 0.4008,
+      "step": 394
+    },
+    {
+      "epoch": 0.7509505703422054,
+      "grad_norm": 0.6998816398818517,
+      "learning_rate": 1.5737275566120577e-05,
+      "loss": 0.4248,
+      "step": 395
+    },
+    {
+      "epoch": 0.752851711026616,
+      "grad_norm": 0.6743785470030397,
+      "learning_rate": 1.5710044089765144e-05,
+      "loss": 0.4247,
+      "step": 396
+    },
+    {
+      "epoch": 0.7547528517110266,
+      "grad_norm": 0.7100232201820885,
+      "learning_rate": 1.5682749640097708e-05,
+      "loss": 0.4129,
+      "step": 397
+    },
+    {
+      "epoch": 0.7566539923954373,
+      "grad_norm": 0.6656340209417392,
+      "learning_rate": 1.565539251813554e-05,
+      "loss": 0.398,
+      "step": 398
+    },
+    {
+      "epoch": 0.7585551330798479,
+      "grad_norm": 0.6827885731819987,
+      "learning_rate": 1.5627973025587093e-05,
+      "loss": 0.4325,
+      "step": 399
+    },
+    {
+      "epoch": 0.7604562737642585,
+      "grad_norm": 0.6852745284147198,
+      "learning_rate": 1.560049146484868e-05,
+      "loss": 0.4198,
+      "step": 400
+    },
+    {
+      "epoch": 0.7623574144486692,
+      "grad_norm": 0.6958253408778999,
+      "learning_rate": 1.5572948139001128e-05,
+      "loss": 0.4191,
+      "step": 401
+    },
+    {
+      "epoch": 0.7642585551330798,
+      "grad_norm": 0.6767669443085678,
+      "learning_rate": 1.5545343351806443e-05,
+      "loss": 0.4049,
+      "step": 402
+    },
+    {
+      "epoch": 0.7661596958174905,
+      "grad_norm": 0.6598380600671233,
+      "learning_rate": 1.551767740770446e-05,
+      "loss": 0.3999,
+      "step": 403
+    },
+    {
+      "epoch": 0.7680608365019012,
+      "grad_norm": 0.6622678972211284,
+      "learning_rate": 1.5489950611809484e-05,
+      "loss": 0.4085,
+      "step": 404
+    },
+    {
+      "epoch": 0.7699619771863118,
+      "grad_norm": 0.7008781719613915,
+      "learning_rate": 1.5462163269906928e-05,
+      "loss": 0.4181,
+      "step": 405
+    },
+    {
+      "epoch": 0.7718631178707225,
+      "grad_norm": 0.6920728970069924,
+      "learning_rate": 1.5434315688449924e-05,
+      "loss": 0.4305,
+      "step": 406
+    },
+    {
+      "epoch": 0.7737642585551331,
+      "grad_norm": 0.7228739248545227,
+      "learning_rate": 1.5406408174555978e-05,
+      "loss": 0.3987,
+      "step": 407
+    },
+    {
+      "epoch": 0.7756653992395437,
+      "grad_norm": 0.7270976967519385,
+      "learning_rate": 1.5378441036003543e-05,
+      "loss": 0.4536,
+      "step": 408
+    },
+    {
+      "epoch": 0.7775665399239544,
+      "grad_norm": 0.7289540198175145,
+      "learning_rate": 1.535041458122865e-05,
+      "loss": 0.4136,
+      "step": 409
+    },
+    {
+      "epoch": 0.779467680608365,
+      "grad_norm": 0.6914521135250499,
+      "learning_rate": 1.5322329119321508e-05,
+      "loss": 0.3964,
+      "step": 410
+    },
+    {
+      "epoch": 0.7813688212927756,
+      "grad_norm": 0.7574803102394569,
+      "learning_rate": 1.529418496002308e-05,
+      "loss": 0.4127,
+      "step": 411
+    },
+    {
+      "epoch": 0.7832699619771863,
+      "grad_norm": 0.6465490975697917,
+      "learning_rate": 1.5265982413721662e-05,
+      "loss": 0.388,
+      "step": 412
+    },
+    {
+      "epoch": 0.785171102661597,
+      "grad_norm": 0.757314571490421,
+      "learning_rate": 1.5237721791449497e-05,
+      "loss": 0.4171,
+      "step": 413
+    },
+    {
+      "epoch": 0.7870722433460076,
+      "grad_norm": 0.7374287852061289,
+      "learning_rate": 1.5209403404879305e-05,
+      "loss": 0.4186,
+      "step": 414
+    },
+    {
+      "epoch": 0.7889733840304183,
+      "grad_norm": 0.6741694742248174,
+      "learning_rate": 1.5181027566320858e-05,
+      "loss": 0.4038,
+      "step": 415
+    },
+    {
+      "epoch": 0.7908745247148289,
+      "grad_norm": 0.7164555635911194,
+      "learning_rate": 1.5152594588717544e-05,
+      "loss": 0.4338,
+      "step": 416
+    },
+    {
+      "epoch": 0.7927756653992395,
+      "grad_norm": 0.6966668078855527,
+      "learning_rate": 1.5124104785642909e-05,
+      "loss": 0.4117,
+      "step": 417
+    },
+    {
+      "epoch": 0.7946768060836502,
+      "grad_norm": 0.73777490647014,
+      "learning_rate": 1.5095558471297196e-05,
+      "loss": 0.4222,
+      "step": 418
+    },
+    {
+      "epoch": 0.7965779467680608,
+      "grad_norm": 0.7570009255108714,
+      "learning_rate": 1.5066955960503893e-05,
+      "loss": 0.4037,
+      "step": 419
+    },
+    {
+      "epoch": 0.7984790874524715,
+      "grad_norm": 0.6450027255042611,
+      "learning_rate": 1.5038297568706244e-05,
+      "loss": 0.4085,
+      "step": 420
+    },
+    {
+      "epoch": 0.8003802281368821,
+      "grad_norm": 0.7634798594062263,
+      "learning_rate": 1.5009583611963772e-05,
+      "loss": 0.4094,
+      "step": 421
+    },
+    {
+      "epoch": 0.8022813688212928,
+      "grad_norm": 0.6762732057198749,
+      "learning_rate": 1.4980814406948806e-05,
+      "loss": 0.4151,
+      "step": 422
+    },
+    {
+      "epoch": 0.8041825095057035,
+      "grad_norm": 0.7206173236703542,
+      "learning_rate": 1.4951990270942991e-05,
+      "loss": 0.4122,
+      "step": 423
+    },
+    {
+      "epoch": 0.8060836501901141,
+      "grad_norm": 0.7301585232916487,
+      "learning_rate": 1.492311152183376e-05,
+      "loss": 0.3901,
+      "step": 424
+    },
+    {
+      "epoch": 0.8079847908745247,
+      "grad_norm": 0.7102692562842765,
+      "learning_rate": 1.4894178478110856e-05,
+      "loss": 0.4147,
+      "step": 425
+    },
+    {
+      "epoch": 0.8098859315589354,
+      "grad_norm": 0.6508030223772372,
+      "learning_rate": 1.4865191458862816e-05,
+      "loss": 0.3973,
+      "step": 426
+    },
+    {
+      "epoch": 0.811787072243346,
+      "grad_norm": 0.6531832338557862,
+      "learning_rate": 1.4836150783773442e-05,
+      "loss": 0.3986,
+      "step": 427
+    },
+    {
+      "epoch": 0.8136882129277566,
+      "grad_norm": 0.709343047125379,
+      "learning_rate": 1.4807056773118276e-05,
+      "loss": 0.4313,
+      "step": 428
+    },
+    {
+      "epoch": 0.8155893536121673,
+      "grad_norm": 0.6486271729638959,
+      "learning_rate": 1.4777909747761085e-05,
+      "loss": 0.3825,
+      "step": 429
+    },
+    {
+      "epoch": 0.8174904942965779,
+      "grad_norm": 0.7379683647060636,
+      "learning_rate": 1.4748710029150296e-05,
+      "loss": 0.4242,
+      "step": 430
+    },
+    {
+      "epoch": 0.8193916349809885,
+      "grad_norm": 0.712437265440026,
+      "learning_rate": 1.4719457939315468e-05,
+      "loss": 0.415,
+      "step": 431
+    },
+    {
+      "epoch": 0.8212927756653993,
+      "grad_norm": 0.7934507968641281,
+      "learning_rate": 1.4690153800863743e-05,
+      "loss": 0.4078,
+      "step": 432
+    },
+    {
+      "epoch": 0.8231939163498099,
+      "grad_norm": 0.6669040222334286,
+      "learning_rate": 1.4660797936976278e-05,
+      "loss": 0.3983,
+      "step": 433
+    },
+    {
+      "epoch": 0.8250950570342205,
+      "grad_norm": 0.6731620889067171,
+      "learning_rate": 1.4631390671404682e-05,
+      "loss": 0.4085,
+      "step": 434
+    },
+    {
+      "epoch": 0.8269961977186312,
+      "grad_norm": 0.7250165594952145,
+      "learning_rate": 1.460193232846745e-05,
+      "loss": 0.4131,
+      "step": 435
+    },
+    {
+      "epoch": 0.8288973384030418,
+      "grad_norm": 0.7044156414406391,
+      "learning_rate": 1.4572423233046386e-05,
+      "loss": 0.4339,
+      "step": 436
+    },
+    {
+      "epoch": 0.8307984790874525,
+      "grad_norm": 0.6519695762363099,
+      "learning_rate": 1.4542863710583022e-05,
+      "loss": 0.4149,
+      "step": 437
+    },
+    {
+      "epoch": 0.8326996197718631,
+      "grad_norm": 0.73551254248906,
+      "learning_rate": 1.4513254087075015e-05,
+      "loss": 0.4078,
+      "step": 438
+    },
+    {
+      "epoch": 0.8346007604562737,
+      "grad_norm": 0.6809148466986238,
+      "learning_rate": 1.4483594689072571e-05,
+      "loss": 0.4083,
+      "step": 439
+    },
+    {
+      "epoch": 0.8365019011406845,
+      "grad_norm": 0.7040902759874987,
+      "learning_rate": 1.4453885843674837e-05,
+      "loss": 0.4032,
+      "step": 440
+    },
+    {
+      "epoch": 0.8384030418250951,
+      "grad_norm": 0.6723194011147194,
+      "learning_rate": 1.4424127878526278e-05,
+      "loss": 0.3964,
+      "step": 441
+    },
+    {
+      "epoch": 0.8403041825095057,
+      "grad_norm": 0.6892204598112396,
+      "learning_rate": 1.4394321121813093e-05,
+      "loss": 0.4087,
+      "step": 442
+    },
+    {
+      "epoch": 0.8422053231939164,
+      "grad_norm": 0.7135621526676784,
+      "learning_rate": 1.436446590225957e-05,
+      "loss": 0.4087,
+      "step": 443
+    },
+    {
+      "epoch": 0.844106463878327,
+      "grad_norm": 0.6460207727594853,
+      "learning_rate": 1.433456254912447e-05,
+      "loss": 0.3878,
+      "step": 444
+    },
+    {
+      "epoch": 0.8460076045627376,
+      "grad_norm": 0.6733725347180656,
+      "learning_rate": 1.4304611392197399e-05,
+      "loss": 0.3986,
+      "step": 445
+    },
+    {
+      "epoch": 0.8479087452471483,
+      "grad_norm": 0.7594062175890656,
+      "learning_rate": 1.427461276179517e-05,
+      "loss": 0.4076,
+      "step": 446
+    },
+    {
+      "epoch": 0.8498098859315589,
+      "grad_norm": 0.6409359384135271,
+      "learning_rate": 1.4244566988758152e-05,
+      "loss": 0.402,
+      "step": 447
+    },
+    {
+      "epoch": 0.8517110266159695,
+      "grad_norm": 0.666927060520839,
+      "learning_rate": 1.4214474404446633e-05,
+      "loss": 0.4004,
+      "step": 448
+    },
+    {
+      "epoch": 0.8536121673003803,
+      "grad_norm": 0.7710692080201269,
+      "learning_rate": 1.4184335340737158e-05,
+      "loss": 0.4189,
+      "step": 449
+    },
+    {
+      "epoch": 0.8555133079847909,
+      "grad_norm": 0.6348586421921572,
+      "learning_rate": 1.4154150130018867e-05,
+      "loss": 0.402,
+      "step": 450
+    },
+    {
+      "epoch": 0.8574144486692015,
+      "grad_norm": 0.7804723869619614,
+      "learning_rate": 1.4123919105189836e-05,
+      "loss": 0.431,
+      "step": 451
+    },
+    {
+      "epoch": 0.8593155893536122,
+      "grad_norm": 0.6397697694827792,
+      "learning_rate": 1.4093642599653406e-05,
+      "loss": 0.4035,
+      "step": 452
+    },
+    {
+      "epoch": 0.8612167300380228,
+      "grad_norm": 0.6465414286017892,
+      "learning_rate": 1.40633209473145e-05,
+      "loss": 0.4271,
+      "step": 453
+    },
+    {
+      "epoch": 0.8631178707224335,
+      "grad_norm": 0.6618688461365902,
+      "learning_rate": 1.4032954482575938e-05,
+      "loss": 0.4133,
+      "step": 454
+    },
+    {
+      "epoch": 0.8650190114068441,
+      "grad_norm": 0.6530430118680148,
+      "learning_rate": 1.4002543540334766e-05,
+      "loss": 0.414,
+      "step": 455
+    },
+    {
+      "epoch": 0.8669201520912547,
+      "grad_norm": 0.6729134290828179,
+      "learning_rate": 1.3972088455978537e-05,
+      "loss": 0.4128,
+      "step": 456
+    },
+    {
+      "epoch": 0.8688212927756654,
+      "grad_norm": 0.6617121324332526,
+      "learning_rate": 1.3941589565381635e-05,
+      "loss": 0.4092,
+      "step": 457
+    },
+    {
+      "epoch": 0.870722433460076,
+      "grad_norm": 0.7774827902611343,
+      "learning_rate": 1.391104720490156e-05,
+      "loss": 0.4213,
+      "step": 458
+    },
+    {
+      "epoch": 0.8726235741444867,
+      "grad_norm": 0.662887144311825,
+      "learning_rate": 1.3880461711375224e-05,
+      "loss": 0.3938,
+      "step": 459
+    },
+    {
+      "epoch": 0.8745247148288974,
+      "grad_norm": 0.6939628671464603,
+      "learning_rate": 1.3849833422115221e-05,
+      "loss": 0.4231,
+      "step": 460
+    },
+    {
+      "epoch": 0.876425855513308,
+      "grad_norm": 0.6962520644547294,
+      "learning_rate": 1.3819162674906134e-05,
+      "loss": 0.4051,
+      "step": 461
+    },
+    {
+      "epoch": 0.8783269961977186,
+      "grad_norm": 0.6897922342748718,
+      "learning_rate": 1.378844980800078e-05,
+      "loss": 0.3998,
+      "step": 462
+    },
+    {
+      "epoch": 0.8802281368821293,
+      "grad_norm": 0.7818650481155389,
+      "learning_rate": 1.3757695160116502e-05,
+      "loss": 0.4342,
+      "step": 463
+    },
+    {
+      "epoch": 0.8821292775665399,
+      "grad_norm": 0.6618706077770252,
+      "learning_rate": 1.3726899070431423e-05,
+      "loss": 0.3945,
+      "step": 464
+    },
+    {
+      "epoch": 0.8840304182509505,
+      "grad_norm": 0.7285841769088787,
+      "learning_rate": 1.3696061878580707e-05,
+      "loss": 0.4258,
+      "step": 465
+    },
+    {
+      "epoch": 0.8859315589353612,
+      "grad_norm": 0.6933762283944436,
+      "learning_rate": 1.3665183924652817e-05,
+      "loss": 0.416,
+      "step": 466
+    },
+    {
+      "epoch": 0.8878326996197718,
+      "grad_norm": 0.6841662038409009,
+      "learning_rate": 1.3634265549185755e-05,
+      "loss": 0.386,
+      "step": 467
+    },
+    {
+      "epoch": 0.8897338403041825,
+      "grad_norm": 0.75540325110924,
+      "learning_rate": 1.3603307093163319e-05,
+      "loss": 0.4169,
+      "step": 468
+    },
+    {
+      "epoch": 0.8916349809885932,
+      "grad_norm": 0.7249151520031727,
+      "learning_rate": 1.3572308898011328e-05,
+      "loss": 0.4153,
+      "step": 469
+    },
+    {
+      "epoch": 0.8935361216730038,
+      "grad_norm": 0.7710393467821494,
+      "learning_rate": 1.3541271305593878e-05,
+      "loss": 0.4131,
+      "step": 470
+    },
+    {
+      "epoch": 0.8954372623574145,
+      "grad_norm": 0.7682736027497585,
+      "learning_rate": 1.3510194658209547e-05,
+      "loss": 0.3958,
+      "step": 471
+    },
+    {
+      "epoch": 0.8973384030418251,
+      "grad_norm": 0.7765655888526504,
+      "learning_rate": 1.3479079298587634e-05,
+      "loss": 0.4021,
+      "step": 472
+    },
+    {
+      "epoch": 0.8992395437262357,
+      "grad_norm": 0.7284525972900238,
+      "learning_rate": 1.3447925569884374e-05,
+      "loss": 0.4164,
+      "step": 473
+    },
+    {
+      "epoch": 0.9011406844106464,
+      "grad_norm": 0.6487701615612256,
+      "learning_rate": 1.3416733815679166e-05,
+      "loss": 0.4076,
+      "step": 474
+    },
+    {
+      "epoch": 0.903041825095057,
+      "grad_norm": 0.8035797050327235,
+      "learning_rate": 1.3385504379970764e-05,
+      "loss": 0.4187,
+      "step": 475
+    },
+    {
+      "epoch": 0.9049429657794676,
+      "grad_norm": 0.7099122118956029,
+      "learning_rate": 1.3354237607173494e-05,
+      "loss": 0.4024,
+      "step": 476
+    },
+    {
+      "epoch": 0.9068441064638784,
+      "grad_norm": 0.7027438466380753,
+      "learning_rate": 1.3322933842113457e-05,
+      "loss": 0.4282,
+      "step": 477
+    },
+    {
+      "epoch": 0.908745247148289,
+      "grad_norm": 0.8366972696068998,
+      "learning_rate": 1.3291593430024727e-05,
+      "loss": 0.4108,
+      "step": 478
+    },
+    {
+      "epoch": 0.9106463878326996,
+      "grad_norm": 0.6806341742745902,
+      "learning_rate": 1.3260216716545534e-05,
+      "loss": 0.4257,
+      "step": 479
+    },
+    {
+      "epoch": 0.9125475285171103,
+      "grad_norm": 0.6861902379931203,
+      "learning_rate": 1.3228804047714462e-05,
+      "loss": 0.3886,
+      "step": 480
+    },
+    {
+      "epoch": 0.9144486692015209,
+      "grad_norm": 0.7141538355533225,
+      "learning_rate": 1.319735576996663e-05,
+      "loss": 0.441,
+      "step": 481
+    },
+    {
+      "epoch": 0.9163498098859315,
+      "grad_norm": 0.6633051209606516,
+      "learning_rate": 1.3165872230129869e-05,
+      "loss": 0.403,
+      "step": 482
+    },
+    {
+      "epoch": 0.9182509505703422,
+      "grad_norm": 0.6469727465282739,
+      "learning_rate": 1.3134353775420895e-05,
+      "loss": 0.3978,
+      "step": 483
+    },
+    {
+      "epoch": 0.9201520912547528,
+      "grad_norm": 0.6615773409265477,
+      "learning_rate": 1.3102800753441488e-05,
+      "loss": 0.4108,
+      "step": 484
+    },
+    {
+      "epoch": 0.9220532319391636,
+      "grad_norm": 0.6127907041751893,
+      "learning_rate": 1.3071213512174655e-05,
+      "loss": 0.3809,
+      "step": 485
+    },
+    {
+      "epoch": 0.9239543726235742,
+      "grad_norm": 0.6555417410803398,
+      "learning_rate": 1.3039592399980785e-05,
+      "loss": 0.3897,
+      "step": 486
+    },
+    {
+      "epoch": 0.9258555133079848,
+      "grad_norm": 0.6943149293769647,
+      "learning_rate": 1.3007937765593818e-05,
+      "loss": 0.416,
+      "step": 487
+    },
+    {
+      "epoch": 0.9277566539923955,
+      "grad_norm": 0.664476536705415,
+      "learning_rate": 1.2976249958117395e-05,
+      "loss": 0.4158,
+      "step": 488
+    },
+    {
+      "epoch": 0.9296577946768061,
+      "grad_norm": 0.7042277011634789,
+      "learning_rate": 1.2944529327021002e-05,
+      "loss": 0.3956,
+      "step": 489
+    },
+    {
+      "epoch": 0.9315589353612167,
+      "grad_norm": 0.6540997688912562,
+      "learning_rate": 1.291277622213612e-05,
+      "loss": 0.39,
+      "step": 490
+    },
+    {
+      "epoch": 0.9334600760456274,
+      "grad_norm": 0.7038043736491921,
+      "learning_rate": 1.2880990993652379e-05,
+      "loss": 0.4051,
+      "step": 491
+    },
+    {
+      "epoch": 0.935361216730038,
+      "grad_norm": 0.7142483993857893,
+      "learning_rate": 1.2849173992113669e-05,
+      "loss": 0.389,
+      "step": 492
+    },
+    {
+      "epoch": 0.9372623574144486,
+      "grad_norm": 0.6346878458379346,
+      "learning_rate": 1.2817325568414299e-05,
+      "loss": 0.4107,
+      "step": 493
+    },
+    {
+      "epoch": 0.9391634980988594,
+      "grad_norm": 0.6947354830793677,
+      "learning_rate": 1.2785446073795118e-05,
+      "loss": 0.4043,
+      "step": 494
+    },
+    {
+      "epoch": 0.94106463878327,
+      "grad_norm": 0.7734180674189518,
+      "learning_rate": 1.2753535859839638e-05,
+      "loss": 0.4164,
+      "step": 495
+    },
+    {
+      "epoch": 0.9429657794676806,
+      "grad_norm": 0.6348809628072248,
+      "learning_rate": 1.272159527847016e-05,
+      "loss": 0.3824,
+      "step": 496
+    },
+    {
+      "epoch": 0.9448669201520913,
+      "grad_norm": 0.747029494621592,
+      "learning_rate": 1.2689624681943897e-05,
+      "loss": 0.4209,
+      "step": 497
+    },
+    {
+      "epoch": 0.9467680608365019,
+      "grad_norm": 0.6981622734727043,
+      "learning_rate": 1.2657624422849077e-05,
+      "loss": 0.4191,
+      "step": 498
+    },
+    {
+      "epoch": 0.9486692015209125,
+      "grad_norm": 0.6737211070865152,
+      "learning_rate": 1.2625594854101066e-05,
+      "loss": 0.425,
+      "step": 499
+    },
+    {
+      "epoch": 0.9505703422053232,
+      "grad_norm": 0.666486572668729,
+      "learning_rate": 1.2593536328938471e-05,
+      "loss": 0.4024,
+      "step": 500
+    },
+    {
+      "epoch": 0.9524714828897338,
+      "grad_norm": 0.6377307927558606,
+      "learning_rate": 1.2561449200919253e-05,
+      "loss": 0.3852,
+      "step": 501
+    },
+    {
+      "epoch": 0.9543726235741445,
+      "grad_norm": 0.6570197297938459,
+      "learning_rate": 1.2529333823916807e-05,
+      "loss": 0.3818,
+      "step": 502
+    },
+    {
+      "epoch": 0.9562737642585551,
+      "grad_norm": 0.7169372325054635,
+      "learning_rate": 1.2497190552116082e-05,
+      "loss": 0.4104,
+      "step": 503
+    },
+    {
+      "epoch": 0.9581749049429658,
+      "grad_norm": 0.6611115802817453,
+      "learning_rate": 1.2465019740009662e-05,
+      "loss": 0.403,
+      "step": 504
+    },
+    {
+      "epoch": 0.9600760456273765,
+      "grad_norm": 0.710336894381173,
+      "learning_rate": 1.2432821742393854e-05,
+      "loss": 0.4214,
+      "step": 505
+    },
+    {
+      "epoch": 0.9619771863117871,
+      "grad_norm": 0.7457306657593686,
+      "learning_rate": 1.2400596914364792e-05,
+      "loss": 0.4035,
+      "step": 506
+    },
+    {
+      "epoch": 0.9638783269961977,
+      "grad_norm": 0.6861087047202212,
+      "learning_rate": 1.2368345611314508e-05,
+      "loss": 0.4095,
+      "step": 507
+    },
+    {
+      "epoch": 0.9657794676806084,
+      "grad_norm": 0.6792803443322428,
+      "learning_rate": 1.2336068188927002e-05,
+      "loss": 0.3773,
+      "step": 508
+    },
+    {
+      "epoch": 0.967680608365019,
+      "grad_norm": 0.7491353813378806,
+      "learning_rate": 1.2303765003174342e-05,
+      "loss": 0.3963,
+      "step": 509
+    },
+    {
+      "epoch": 0.9695817490494296,
+      "grad_norm": 0.6347006209673904,
+      "learning_rate": 1.2271436410312727e-05,
+      "loss": 0.4203,
+      "step": 510
+    },
+    {
+      "epoch": 0.9714828897338403,
+      "grad_norm": 0.7128999262214145,
+      "learning_rate": 1.2239082766878557e-05,
+      "loss": 0.4031,
+      "step": 511
+    },
+    {
+      "epoch": 0.973384030418251,
+      "grad_norm": 0.7103055629168825,
+      "learning_rate": 1.2206704429684504e-05,
+      "loss": 0.3937,
+      "step": 512
+    },
+    {
+      "epoch": 0.9752851711026616,
+      "grad_norm": 0.7149374781859501,
+      "learning_rate": 1.2174301755815572e-05,
+      "loss": 0.4163,
+      "step": 513
+    },
+    {
+      "epoch": 0.9771863117870723,
+      "grad_norm": 0.6574075647838786,
+      "learning_rate": 1.2141875102625166e-05,
+      "loss": 0.4018,
+      "step": 514
+    },
+    {
+      "epoch": 0.9790874524714829,
+      "grad_norm": 0.7397322625197243,
+      "learning_rate": 1.2109424827731144e-05,
+      "loss": 0.4111,
+      "step": 515
+    },
+    {
+      "epoch": 0.9809885931558935,
+      "grad_norm": 0.6668024828603756,
+      "learning_rate": 1.2076951289011884e-05,
+      "loss": 0.426,
+      "step": 516
+    },
+    {
+      "epoch": 0.9828897338403042,
+      "grad_norm": 0.6783027261082623,
+      "learning_rate": 1.204445484460232e-05,
+      "loss": 0.4063,
+      "step": 517
+    },
+    {
+      "epoch": 0.9847908745247148,
+      "grad_norm": 0.6759891460397746,
+      "learning_rate": 1.2011935852890004e-05,
+      "loss": 0.4127,
+      "step": 518
+    },
+    {
+      "epoch": 0.9866920152091255,
+      "grad_norm": 0.6634956019872992,
+      "learning_rate": 1.1979394672511156e-05,
+      "loss": 0.3825,
+      "step": 519
+    },
+    {
+      "epoch": 0.9885931558935361,
+      "grad_norm": 0.7725501480385996,
+      "learning_rate": 1.19468316623467e-05,
+      "loss": 0.4074,
+      "step": 520
+    },
+    {
+      "epoch": 0.9904942965779467,
+      "grad_norm": 0.6717702981744055,
+      "learning_rate": 1.1914247181518312e-05,
+      "loss": 0.4046,
+      "step": 521
+    },
+    {
+      "epoch": 0.9923954372623575,
+      "grad_norm": 0.65549078100065,
+      "learning_rate": 1.1881641589384456e-05,
+      "loss": 0.4071,
+      "step": 522
+    },
+    {
+      "epoch": 0.9942965779467681,
+      "grad_norm": 0.7214570785871159,
+      "learning_rate": 1.1849015245536424e-05,
+      "loss": 0.4112,
+      "step": 523
+    },
+    {
+      "epoch": 0.9961977186311787,
+      "grad_norm": 0.6742543347723511,
+      "learning_rate": 1.1816368509794365e-05,
+      "loss": 0.4138,
+      "step": 524
+    },
+    {
+      "epoch": 0.9980988593155894,
+      "grad_norm": 0.6679410808371169,
+      "learning_rate": 1.1783701742203326e-05,
+      "loss": 0.3967,
+      "step": 525
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6645189821815252,
+      "learning_rate": 1.1751015303029272e-05,
+      "loss": 0.3972,
+      "step": 526
+    },
+    {
+      "epoch": 1.0019011406844107,
+      "grad_norm": 0.9026768552474234,
+      "learning_rate": 1.1718309552755118e-05,
+      "loss": 0.3204,
+      "step": 527
+    },
+    {
+      "epoch": 1.0038022813688212,
+      "grad_norm": 0.7520768909268105,
+      "learning_rate": 1.1685584852076746e-05,
+      "loss": 0.3171,
+      "step": 528
+    },
+    {
+      "epoch": 1.005703422053232,
+      "grad_norm": 0.8304971161573363,
+      "learning_rate": 1.1652841561899042e-05,
+      "loss": 0.3096,
+      "step": 529
+    },
+    {
+      "epoch": 1.0076045627376427,
+      "grad_norm": 0.8317521107283936,
+      "learning_rate": 1.1620080043331901e-05,
+      "loss": 0.3058,
+      "step": 530
+    },
+    {
+      "epoch": 1.0095057034220531,
+      "grad_norm": 0.7174738036544435,
+      "learning_rate": 1.1587300657686254e-05,
+      "loss": 0.3101,
+      "step": 531
+    },
+    {
+      "epoch": 1.0114068441064639,
+      "grad_norm": 0.9253847389966938,
+      "learning_rate": 1.1554503766470069e-05,
+      "loss": 0.3098,
+      "step": 532
+    },
+    {
+      "epoch": 1.0133079847908746,
+      "grad_norm": 0.86056347375719,
+      "learning_rate": 1.1521689731384391e-05,
+      "loss": 0.3143,
+      "step": 533
+    },
+    {
+      "epoch": 1.015209125475285,
+      "grad_norm": 0.8121379657591885,
+      "learning_rate": 1.1488858914319321e-05,
+      "loss": 0.3067,
+      "step": 534
+    },
+    {
+      "epoch": 1.0171102661596958,
+      "grad_norm": 0.8050339408207571,
+      "learning_rate": 1.1456011677350052e-05,
+      "loss": 0.3019,
+      "step": 535
+    },
+    {
+      "epoch": 1.0190114068441065,
+      "grad_norm": 0.7709208767546798,
+      "learning_rate": 1.1423148382732854e-05,
+      "loss": 0.3083,
+      "step": 536
+    },
+    {
+      "epoch": 1.020912547528517,
+      "grad_norm": 0.7564699872958699,
+      "learning_rate": 1.1390269392901096e-05,
+      "loss": 0.3076,
+      "step": 537
+    },
+    {
+      "epoch": 1.0228136882129277,
+      "grad_norm": 0.7668523225177987,
+      "learning_rate": 1.1357375070461241e-05,
+      "loss": 0.2999,
+      "step": 538
+    },
+    {
+      "epoch": 1.0247148288973384,
+      "grad_norm": 0.6954113680607918,
+      "learning_rate": 1.1324465778188846e-05,
+      "loss": 0.2877,
+      "step": 539
+    },
+    {
+      "epoch": 1.026615969581749,
+      "grad_norm": 0.6873741196847571,
+      "learning_rate": 1.1291541879024568e-05,
+      "loss": 0.3054,
+      "step": 540
+    },
+    {
+      "epoch": 1.0285171102661597,
+      "grad_norm": 0.647004384029595,
+      "learning_rate": 1.1258603736070145e-05,
+      "loss": 0.2856,
+      "step": 541
+    },
+    {
+      "epoch": 1.0304182509505704,
+      "grad_norm": 0.7161940698269649,
+      "learning_rate": 1.1225651712584413e-05,
+      "loss": 0.3081,
+      "step": 542
+    },
+    {
+      "epoch": 1.032319391634981,
+      "grad_norm": 0.7032274903768369,
+      "learning_rate": 1.1192686171979288e-05,
+      "loss": 0.2948,
+      "step": 543
+    },
+    {
+      "epoch": 1.0342205323193916,
+      "grad_norm": 0.6797459514540284,
+      "learning_rate": 1.1159707477815756e-05,
+      "loss": 0.2899,
+      "step": 544
+    },
+    {
+      "epoch": 1.0361216730038023,
+      "grad_norm": 0.7574219958700112,
+      "learning_rate": 1.1126715993799875e-05,
+      "loss": 0.2925,
+      "step": 545
+    },
+    {
+      "epoch": 1.038022813688213,
+      "grad_norm": 0.7093169997518386,
+      "learning_rate": 1.1093712083778748e-05,
+      "loss": 0.2761,
+      "step": 546
+    },
+    {
+      "epoch": 1.0399239543726235,
+      "grad_norm": 0.7101116582645551,
+      "learning_rate": 1.1060696111736515e-05,
+      "loss": 0.3073,
+      "step": 547
+    },
+    {
+      "epoch": 1.0418250950570342,
+      "grad_norm": 0.7498740869570878,
+      "learning_rate": 1.1027668441790358e-05,
+      "loss": 0.2935,
+      "step": 548
+    },
+    {
+      "epoch": 1.043726235741445,
+      "grad_norm": 0.7575102500859472,
+      "learning_rate": 1.099462943818646e-05,
+      "loss": 0.3093,
+      "step": 549
+    },
+    {
+      "epoch": 1.0456273764258555,
+      "grad_norm": 0.7113782097636886,
+      "learning_rate": 1.0961579465295987e-05,
+      "loss": 0.2835,
+      "step": 550
+    },
+    {
+      "epoch": 1.0475285171102662,
+      "grad_norm": 0.8051738987650036,
+      "learning_rate": 1.0928518887611099e-05,
+      "loss": 0.3061,
+      "step": 551
+    },
+    {
+      "epoch": 1.049429657794677,
+      "grad_norm": 0.6652423225997399,
+      "learning_rate": 1.0895448069740902e-05,
+      "loss": 0.304,
+      "step": 552
+    },
+    {
+      "epoch": 1.0513307984790874,
+      "grad_norm": 0.6901150022318671,
+      "learning_rate": 1.0862367376407433e-05,
+      "loss": 0.297,
+      "step": 553
+    },
+    {
+      "epoch": 1.053231939163498,
+      "grad_norm": 0.7201231546794757,
+      "learning_rate": 1.0829277172441648e-05,
+      "loss": 0.314,
+      "step": 554
+    },
+    {
+      "epoch": 1.0551330798479088,
+      "grad_norm": 0.6702012097789832,
+      "learning_rate": 1.0796177822779384e-05,
+      "loss": 0.2872,
+      "step": 555
+    },
+    {
+      "epoch": 1.0570342205323193,
+      "grad_norm": 0.7394656929630515,
+      "learning_rate": 1.0763069692457346e-05,
+      "loss": 0.3023,
+      "step": 556
+    },
+    {
+      "epoch": 1.05893536121673,
+      "grad_norm": 0.6964411117287589,
+      "learning_rate": 1.0729953146609076e-05,
+      "loss": 0.2923,
+      "step": 557
+    },
+    {
+      "epoch": 1.0608365019011408,
+      "grad_norm": 0.768623809940278,
+      "learning_rate": 1.0696828550460928e-05,
+      "loss": 0.3032,
+      "step": 558
+    },
+    {
+      "epoch": 1.0627376425855513,
+      "grad_norm": 0.7217808776911585,
+      "learning_rate": 1.0663696269328034e-05,
+      "loss": 0.3102,
+      "step": 559
+    },
+    {
+      "epoch": 1.064638783269962,
+      "grad_norm": 0.6942163897324145,
+      "learning_rate": 1.0630556668610286e-05,
+      "loss": 0.2886,
+      "step": 560
+    },
+    {
+      "epoch": 1.0665399239543727,
+      "grad_norm": 0.7103334992859985,
+      "learning_rate": 1.059741011378829e-05,
+      "loss": 0.2944,
+      "step": 561
+    },
+    {
+      "epoch": 1.0684410646387832,
+      "grad_norm": 0.7646968277891509,
+      "learning_rate": 1.0564256970419367e-05,
+      "loss": 0.2983,
+      "step": 562
+    },
+    {
+      "epoch": 1.070342205323194,
+      "grad_norm": 0.758763238695484,
+      "learning_rate": 1.0531097604133473e-05,
+      "loss": 0.3019,
+      "step": 563
+    },
+    {
+      "epoch": 1.0722433460076046,
+      "grad_norm": 0.7481643499349109,
+      "learning_rate": 1.0497932380629207e-05,
+      "loss": 0.2973,
+      "step": 564
+    },
+    {
+      "epoch": 1.0741444866920151,
+      "grad_norm": 0.8159864254281655,
+      "learning_rate": 1.0464761665669771e-05,
+      "loss": 0.2959,
+      "step": 565
+    },
+    {
+      "epoch": 1.0760456273764258,
+      "grad_norm": 0.7299575901812068,
+      "learning_rate": 1.0431585825078916e-05,
+      "loss": 0.2943,
+      "step": 566
+    },
+    {
+      "epoch": 1.0779467680608366,
+      "grad_norm": 0.7317922011461669,
+      "learning_rate": 1.0398405224736927e-05,
+      "loss": 0.3013,
+      "step": 567
+    },
+    {
+      "epoch": 1.079847908745247,
+      "grad_norm": 0.7409554154481878,
+      "learning_rate": 1.0365220230576592e-05,
+      "loss": 0.287,
+      "step": 568
+    },
+    {
+      "epoch": 1.0817490494296578,
+      "grad_norm": 0.7729783533848208,
+      "learning_rate": 1.0332031208579133e-05,
+      "loss": 0.3026,
+      "step": 569
+    },
+    {
+      "epoch": 1.0836501901140685,
+      "grad_norm": 0.7295015532803011,
+      "learning_rate": 1.0298838524770212e-05,
+      "loss": 0.3005,
+      "step": 570
+    },
+    {
+      "epoch": 1.085551330798479,
+      "grad_norm": 0.6995206145760231,
+      "learning_rate": 1.0265642545215872e-05,
+      "loss": 0.3039,
+      "step": 571
+    },
+    {
+      "epoch": 1.0874524714828897,
+      "grad_norm": 0.7712843530329814,
+      "learning_rate": 1.0232443636018502e-05,
+      "loss": 0.3009,
+      "step": 572
+    },
+    {
+      "epoch": 1.0893536121673004,
+      "grad_norm": 0.6624052241745435,
+      "learning_rate": 1.0199242163312794e-05,
+      "loss": 0.284,
+      "step": 573
+    },
+    {
+      "epoch": 1.091254752851711,
+      "grad_norm": 0.6632714437566419,
+      "learning_rate": 1.0166038493261723e-05,
+      "loss": 0.2938,
+      "step": 574
+    },
+    {
+      "epoch": 1.0931558935361216,
+      "grad_norm": 0.8631539637769745,
+      "learning_rate": 1.013283299205249e-05,
+      "loss": 0.3001,
+      "step": 575
+    },
+    {
+      "epoch": 1.0950570342205324,
+      "grad_norm": 0.6988030829246281,
+      "learning_rate": 1.0099626025892491e-05,
+      "loss": 0.296,
+      "step": 576
+    },
+    {
+      "epoch": 1.0969581749049429,
+      "grad_norm": 0.7438947397716793,
+      "learning_rate": 1.0066417961005283e-05,
+      "loss": 0.3073,
+      "step": 577
+    },
+    {
+      "epoch": 1.0988593155893536,
+      "grad_norm": 0.7565802086669974,
+      "learning_rate": 1.0033209163626539e-05,
+      "loss": 0.2983,
+      "step": 578
+    },
+    {
+      "epoch": 1.1007604562737643,
+      "grad_norm": 0.6939946290417354,
+      "learning_rate": 1e-05,
+      "loss": 0.2884,
+      "step": 579
+    },
+    {
+      "epoch": 1.102661596958175,
+      "grad_norm": 0.6857676750762475,
+      "learning_rate": 9.966790836373465e-06,
+      "loss": 0.2919,
+      "step": 580
+    },
+    {
+      "epoch": 1.1045627376425855,
+      "grad_norm": 0.7399403043757216,
+      "learning_rate": 9.933582038994719e-06,
+      "loss": 0.305,
+      "step": 581
+    },
+    {
+      "epoch": 1.1064638783269962,
+      "grad_norm": 0.7136131408705637,
+      "learning_rate": 9.90037397410751e-06,
+      "loss": 0.2904,
+      "step": 582
+    },
+    {
+      "epoch": 1.108365019011407,
+      "grad_norm": 0.7729624879523586,
+      "learning_rate": 9.867167007947511e-06,
+      "loss": 0.2974,
+      "step": 583
+    },
+    {
+      "epoch": 1.1102661596958174,
+      "grad_norm": 0.7030126485126656,
+      "learning_rate": 9.833961506738282e-06,
+      "loss": 0.3053,
+      "step": 584
+    },
+    {
+      "epoch": 1.1121673003802282,
+      "grad_norm": 0.710601635450347,
+      "learning_rate": 9.80075783668721e-06,
+      "loss": 0.2939,
+      "step": 585
+    },
+    {
+      "epoch": 1.1140684410646389,
+      "grad_norm": 0.7722465915633965,
+      "learning_rate": 9.767556363981503e-06,
+      "loss": 0.2964,
+      "step": 586
+    },
+    {
+      "epoch": 1.1159695817490494,
+      "grad_norm": 0.6690460430563578,
+      "learning_rate": 9.734357454784131e-06,
+      "loss": 0.2943,
+      "step": 587
+    },
+    {
+      "epoch": 1.11787072243346,
+      "grad_norm": 0.6933979049012836,
+      "learning_rate": 9.701161475229791e-06,
+      "loss": 0.295,
+      "step": 588
+    },
+    {
+      "epoch": 1.1197718631178708,
+      "grad_norm": 0.7858862047602722,
+      "learning_rate": 9.66796879142087e-06,
+      "loss": 0.3,
+      "step": 589
+    },
+    {
+      "epoch": 1.1216730038022813,
+      "grad_norm": 0.6938701640343599,
+      "learning_rate": 9.634779769423412e-06,
+      "loss": 0.2957,
+      "step": 590
+    },
+    {
+      "epoch": 1.123574144486692,
+      "grad_norm": 0.6808125887475195,
+      "learning_rate": 9.601594775263073e-06,
+      "loss": 0.2946,
+      "step": 591
+    },
+    {
+      "epoch": 1.1254752851711027,
+      "grad_norm": 0.7104091069584411,
+      "learning_rate": 9.568414174921085e-06,
+      "loss": 0.3017,
+      "step": 592
+    },
+    {
+      "epoch": 1.1273764258555132,
+      "grad_norm": 0.7033203700513427,
+      "learning_rate": 9.535238334330234e-06,
+      "loss": 0.3037,
+      "step": 593
+    },
+    {
+      "epoch": 1.129277566539924,
+      "grad_norm": 0.7002757909168774,
+      "learning_rate": 9.502067619370794e-06,
+      "loss": 0.3017,
+      "step": 594
+    },
+    {
+      "epoch": 1.1311787072243347,
+      "grad_norm": 0.7136242995125125,
+      "learning_rate": 9.468902395866532e-06,
+      "loss": 0.3006,
+      "step": 595
+    },
+    {
+      "epoch": 1.1330798479087452,
+      "grad_norm": 0.7793830425937579,
+      "learning_rate": 9.435743029580638e-06,
+      "loss": 0.3004,
+      "step": 596
+    },
+    {
+      "epoch": 1.1349809885931559,
+      "grad_norm": 0.7070559669946418,
+      "learning_rate": 9.402589886211711e-06,
+      "loss": 0.2878,
+      "step": 597
+    },
+    {
+      "epoch": 1.1368821292775666,
+      "grad_norm": 0.7421621554839294,
+      "learning_rate": 9.369443331389718e-06,
+      "loss": 0.2975,
+      "step": 598
+    },
+    {
+      "epoch": 1.138783269961977,
+      "grad_norm": 0.7796168845403724,
+      "learning_rate": 9.336303730671968e-06,
+      "loss": 0.3123,
+      "step": 599
+    },
+    {
+      "epoch": 1.1406844106463878,
+      "grad_norm": 0.7382412495787148,
+      "learning_rate": 9.303171449539074e-06,
+      "loss": 0.3094,
+      "step": 600
+    },
+    {
+      "epoch": 1.1425855513307985,
+      "grad_norm": 0.6954559984071695,
+      "learning_rate": 9.270046853390924e-06,
+      "loss": 0.2951,
+      "step": 601
+    },
+    {
+      "epoch": 1.144486692015209,
+      "grad_norm": 0.7153717187603283,
+      "learning_rate": 9.236930307542654e-06,
+      "loss": 0.304,
+      "step": 602
+    },
+    {
+      "epoch": 1.1463878326996197,
+      "grad_norm": 0.6789480333297987,
+      "learning_rate": 9.203822177220621e-06,
+      "loss": 0.2788,
+      "step": 603
+    },
+    {
+      "epoch": 1.1482889733840305,
+      "grad_norm": 0.7547557419060306,
+      "learning_rate": 9.170722827558357e-06,
+      "loss": 0.291,
+      "step": 604
+    },
+    {
+      "epoch": 1.1501901140684412,
+      "grad_norm": 0.7213655281787826,
+      "learning_rate": 9.13763262359257e-06,
+      "loss": 0.3031,
+      "step": 605
+    },
+    {
+      "epoch": 1.1520912547528517,
+      "grad_norm": 0.7203926045941605,
+      "learning_rate": 9.104551930259101e-06,
+      "loss": 0.307,
+      "step": 606
+    },
+    {
+      "epoch": 1.1539923954372624,
+      "grad_norm": 0.7553345920197461,
+      "learning_rate": 9.071481112388905e-06,
+      "loss": 0.3151,
+      "step": 607
+    },
+    {
+      "epoch": 1.1558935361216731,
+      "grad_norm": 0.6596062303881036,
+      "learning_rate": 9.038420534704015e-06,
+      "loss": 0.2959,
+      "step": 608
+    },
+    {
+      "epoch": 1.1577946768060836,
+      "grad_norm": 0.7103280235369358,
+      "learning_rate": 9.005370561813545e-06,
+      "loss": 0.3029,
+      "step": 609
+    },
+    {
+      "epoch": 1.1596958174904943,
+      "grad_norm": 0.7196336086474391,
+      "learning_rate": 8.972331558209644e-06,
+      "loss": 0.3132,
+      "step": 610
+    },
+    {
+      "epoch": 1.161596958174905,
+      "grad_norm": 0.6991171957347445,
+      "learning_rate": 8.939303888263485e-06,
+      "loss": 0.2964,
+      "step": 611
+    },
+    {
+      "epoch": 1.1634980988593155,
+      "grad_norm": 0.7210346184532263,
+      "learning_rate": 8.906287916221259e-06,
+      "loss": 0.3011,
+      "step": 612
+    },
+    {
+      "epoch": 1.1653992395437263,
+      "grad_norm": 0.7147677607378655,
+      "learning_rate": 8.873284006200129e-06,
+      "loss": 0.2973,
+      "step": 613
+    },
+    {
+      "epoch": 1.167300380228137,
+      "grad_norm": 0.7051397512141437,
+      "learning_rate": 8.840292522184247e-06,
+      "loss": 0.289,
+      "step": 614
+    },
+    {
+      "epoch": 1.1692015209125475,
+      "grad_norm": 0.7746513321467852,
+      "learning_rate": 8.807313828020715e-06,
+      "loss": 0.2968,
+      "step": 615
+    },
+    {
+      "epoch": 1.1711026615969582,
+      "grad_norm": 0.7456944928208514,
+      "learning_rate": 8.774348287415589e-06,
+      "loss": 0.3086,
+      "step": 616
+    },
+    {
+      "epoch": 1.173003802281369,
+      "grad_norm": 0.7449862930791246,
+      "learning_rate": 8.74139626392986e-06,
+      "loss": 0.3068,
+      "step": 617
+    },
+    {
+      "epoch": 1.1749049429657794,
+      "grad_norm": 0.7197185940049116,
+      "learning_rate": 8.708458120975436e-06,
+      "loss": 0.289,
+      "step": 618
+    },
+    {
+      "epoch": 1.1768060836501901,
+      "grad_norm": 0.7139215783917028,
+      "learning_rate": 8.675534221811156e-06,
+      "loss": 0.291,
+      "step": 619
+    },
+    {
+      "epoch": 1.1787072243346008,
+      "grad_norm": 0.7187979109961694,
+      "learning_rate": 8.64262492953876e-06,
+      "loss": 0.2935,
+      "step": 620
+    },
+    {
+      "epoch": 1.1806083650190113,
+      "grad_norm": 0.7711579982183011,
+      "learning_rate": 8.60973060709891e-06,
+      "loss": 0.319,
+      "step": 621
+    },
+    {
+      "epoch": 1.182509505703422,
+      "grad_norm": 0.7175946906833541,
+      "learning_rate": 8.576851617267151e-06,
+      "loss": 0.3035,
+      "step": 622
+    },
+    {
+      "epoch": 1.1844106463878328,
+      "grad_norm": 0.8184089782561722,
+      "learning_rate": 8.543988322649954e-06,
+      "loss": 0.3176,
+      "step": 623
+    },
+    {
+      "epoch": 1.1863117870722433,
+      "grad_norm": 0.6782796906207955,
+      "learning_rate": 8.511141085680684e-06,
+      "loss": 0.2978,
+      "step": 624
+    },
+    {
+      "epoch": 1.188212927756654,
+      "grad_norm": 0.6849773116097128,
+      "learning_rate": 8.478310268615612e-06,
+      "loss": 0.3124,
+      "step": 625
+    },
+    {
+      "epoch": 1.1901140684410647,
+      "grad_norm": 0.7000057642841461,
+      "learning_rate": 8.445496233529934e-06,
+      "loss": 0.2941,
+      "step": 626
+    },
+    {
+      "epoch": 1.1920152091254752,
+      "grad_norm": 0.7017512598111004,
+      "learning_rate": 8.41269934231375e-06,
+      "loss": 0.3109,
+      "step": 627
+    },
+    {
+      "epoch": 1.193916349809886,
+      "grad_norm": 0.6679168915138751,
+      "learning_rate": 8.3799199566681e-06,
+      "loss": 0.2855,
+      "step": 628
+    },
+    {
+      "epoch": 1.1958174904942966,
+      "grad_norm": 0.7011394496244616,
+      "learning_rate": 8.34715843810096e-06,
+      "loss": 0.287,
+      "step": 629
+    },
+    {
+      "epoch": 1.1977186311787071,
+      "grad_norm": 0.7737328468785828,
+      "learning_rate": 8.314415147923254e-06,
+      "loss": 0.3081,
+      "step": 630
+    },
+    {
+      "epoch": 1.1996197718631179,
+      "grad_norm": 0.7291747369595318,
+      "learning_rate": 8.281690447244887e-06,
+      "loss": 0.2851,
+      "step": 631
+    },
+    {
+      "epoch": 1.2015209125475286,
+      "grad_norm": 0.7205727324732295,
+      "learning_rate": 8.248984696970732e-06,
+      "loss": 0.3043,
+      "step": 632
+    },
+    {
+      "epoch": 1.203422053231939,
+      "grad_norm": 0.7535550332956547,
+      "learning_rate": 8.216298257796677e-06,
+      "loss": 0.3065,
+      "step": 633
+    },
+    {
+      "epoch": 1.2053231939163498,
+      "grad_norm": 1.6019010067718478,
+      "learning_rate": 8.183631490205636e-06,
+      "loss": 0.2784,
+      "step": 634
+    },
+    {
+      "epoch": 1.2072243346007605,
+      "grad_norm": 0.7171795868995431,
+      "learning_rate": 8.150984754463578e-06,
+      "loss": 0.2963,
+      "step": 635
+    },
+    {
+      "epoch": 1.209125475285171,
+      "grad_norm": 0.7098322044257795,
+      "learning_rate": 8.118358410615545e-06,
+      "loss": 0.2953,
+      "step": 636
+    },
+    {
+      "epoch": 1.2110266159695817,
+      "grad_norm": 0.7056507836274983,
+      "learning_rate": 8.08575281848169e-06,
+      "loss": 0.2905,
+      "step": 637
+    },
+    {
+      "epoch": 1.2129277566539924,
+      "grad_norm": 0.7026924947713458,
+      "learning_rate": 8.0531683376533e-06,
+      "loss": 0.2786,
+      "step": 638
+    },
+    {
+      "epoch": 1.214828897338403,
+      "grad_norm": 0.7150499560999022,
+      "learning_rate": 8.020605327488846e-06,
+      "loss": 0.2963,
+      "step": 639
+    },
+    {
+      "epoch": 1.2167300380228137,
+      "grad_norm": 0.7356082612579713,
+      "learning_rate": 7.988064147110001e-06,
+      "loss": 0.3083,
+      "step": 640
+    },
+    {
+      "epoch": 1.2186311787072244,
+      "grad_norm": 0.704604582686781,
+      "learning_rate": 7.955545155397684e-06,
+      "loss": 0.293,
+      "step": 641
+    },
+    {
+      "epoch": 1.2205323193916349,
+      "grad_norm": 0.6914863032501433,
+      "learning_rate": 7.923048710988119e-06,
+      "loss": 0.2938,
+      "step": 642
+    },
+    {
+      "epoch": 1.2224334600760456,
+      "grad_norm": 0.7343680909535492,
+      "learning_rate": 7.890575172268858e-06,
+      "loss": 0.3092,
+      "step": 643
+    },
+    {
+      "epoch": 1.2243346007604563,
+      "grad_norm": 0.6695085138952159,
+      "learning_rate": 7.858124897374837e-06,
+      "loss": 0.2857,
+      "step": 644
+    },
+    {
+      "epoch": 1.2262357414448668,
+      "grad_norm": 0.7475826565229964,
+      "learning_rate": 7.825698244184432e-06,
+      "loss": 0.2967,
+      "step": 645
+    },
+    {
+      "epoch": 1.2281368821292775,
+      "grad_norm": 0.7209444151590619,
+      "learning_rate": 7.7932955703155e-06,
+      "loss": 0.2932,
+      "step": 646
+    },
+    {
+      "epoch": 1.2300380228136882,
+      "grad_norm": 0.7538032877765523,
+      "learning_rate": 7.760917233121443e-06,
+      "loss": 0.314,
+      "step": 647
+    },
+    {
+      "epoch": 1.231939163498099,
+      "grad_norm": 0.7361903726641821,
+      "learning_rate": 7.728563589687275e-06,
+      "loss": 0.2882,
+      "step": 648
+    },
+    {
+      "epoch": 1.2338403041825095,
+      "grad_norm": 0.7044802576910466,
+      "learning_rate": 7.696234996825663e-06,
+      "loss": 0.2989,
+      "step": 649
+    },
+    {
+      "epoch": 1.2357414448669202,
+      "grad_norm": 0.6903249909019692,
+      "learning_rate": 7.663931811073003e-06,
+      "loss": 0.2957,
+      "step": 650
+    },
+    {
+      "epoch": 1.2376425855513309,
+      "grad_norm": 0.7299228410624734,
+      "learning_rate": 7.631654388685496e-06,
+      "loss": 0.2903,
+      "step": 651
+    },
+    {
+      "epoch": 1.2395437262357414,
+      "grad_norm": 0.715144203464586,
+      "learning_rate": 7.599403085635208e-06,
+      "loss": 0.3039,
+      "step": 652
+    },
+    {
+      "epoch": 1.241444866920152,
+      "grad_norm": 0.7118024983968824,
+      "learning_rate": 7.567178257606147e-06,
+      "loss": 0.3025,
+      "step": 653
+    },
+    {
+      "epoch": 1.2433460076045628,
+      "grad_norm": 0.69135704861126,
+      "learning_rate": 7.534980259990341e-06,
+      "loss": 0.2859,
+      "step": 654
+    },
+    {
+      "epoch": 1.2452471482889733,
+      "grad_norm": 0.6976221519352619,
+      "learning_rate": 7.50280944788392e-06,
+      "loss": 0.2928,
+      "step": 655
+    },
+    {
+      "epoch": 1.247148288973384,
+      "grad_norm": 0.7179801796004978,
+      "learning_rate": 7.470666176083193e-06,
+      "loss": 0.306,
+      "step": 656
+    },
+    {
+      "epoch": 1.2490494296577948,
+      "grad_norm": 0.7781029245362643,
+      "learning_rate": 7.438550799080746e-06,
+      "loss": 0.3001,
+      "step": 657
+    },
+    {
+      "epoch": 1.2509505703422052,
+      "grad_norm": 0.7123691035651675,
+      "learning_rate": 7.40646367106153e-06,
+      "loss": 0.3017,
+      "step": 658
+    },
+    {
+      "epoch": 1.252851711026616,
+      "grad_norm": 0.6826218282511876,
+      "learning_rate": 7.3744051458989395e-06,
+      "loss": 0.2977,
+      "step": 659
+    },
+    {
+      "epoch": 1.2547528517110267,
+      "grad_norm": 0.7661159255775424,
+      "learning_rate": 7.342375577150928e-06,
+      "loss": 0.2894,
+      "step": 660
+    },
+    {
+      "epoch": 1.2566539923954372,
+      "grad_norm": 0.7324700867195278,
+      "learning_rate": 7.310375318056107e-06,
+      "loss": 0.3077,
+      "step": 661
+    },
+    {
+      "epoch": 1.258555133079848,
+      "grad_norm": 0.698429300204287,
+      "learning_rate": 7.278404721529843e-06,
+      "loss": 0.2988,
+      "step": 662
+    },
+    {
+      "epoch": 1.2604562737642586,
+      "grad_norm": 0.6905359290423392,
+      "learning_rate": 7.246464140160365e-06,
+      "loss": 0.2871,
+      "step": 663
+    },
+    {
+      "epoch": 1.2623574144486693,
+      "grad_norm": 0.7302545318356523,
+      "learning_rate": 7.214553926204884e-06,
+      "loss": 0.286,
+      "step": 664
+    },
+    {
+      "epoch": 1.2642585551330798,
+      "grad_norm": 0.7299874615196501,
+      "learning_rate": 7.182674431585703e-06,
+      "loss": 0.2972,
+      "step": 665
+    },
+    {
+      "epoch": 1.2661596958174905,
+      "grad_norm": 0.7584874248545316,
+      "learning_rate": 7.150826007886334e-06,
+      "loss": 0.3004,
+      "step": 666
+    },
+    {
+      "epoch": 1.2680608365019013,
+      "grad_norm": 0.721565917332782,
+      "learning_rate": 7.119009006347625e-06,
+      "loss": 0.2877,
+      "step": 667
+    },
+    {
+      "epoch": 1.2699619771863118,
+      "grad_norm": 0.7692890691323427,
+      "learning_rate": 7.087223777863883e-06,
+      "loss": 0.2936,
+      "step": 668
+    },
+    {
+      "epoch": 1.2718631178707225,
+      "grad_norm": 0.7216850656259597,
+      "learning_rate": 7.055470672979003e-06,
+      "loss": 0.3108,
+      "step": 669
+    },
+    {
+      "epoch": 1.2737642585551332,
+      "grad_norm": 0.6855487707927788,
+      "learning_rate": 7.023750041882609e-06,
+      "loss": 0.2969,
+      "step": 670
+    },
+    {
+      "epoch": 1.2756653992395437,
+      "grad_norm": 0.7018207804319794,
+      "learning_rate": 6.992062234406185e-06,
+      "loss": 0.2944,
+      "step": 671
+    },
+    {
+      "epoch": 1.2775665399239544,
+      "grad_norm": 0.7133513890448708,
+      "learning_rate": 6.960407600019217e-06,
+      "loss": 0.2857,
+      "step": 672
+    },
+    {
+      "epoch": 1.2794676806083651,
+      "grad_norm": 0.712829115171771,
+      "learning_rate": 6.9287864878253475e-06,
+      "loss": 0.2969,
+      "step": 673
+    },
+    {
+      "epoch": 1.2813688212927756,
+      "grad_norm": 0.7213526492424779,
+      "learning_rate": 6.897199246558515e-06,
+      "loss": 0.2842,
+      "step": 674
+    },
+    {
+      "epoch": 1.2832699619771863,
+      "grad_norm": 0.7133246603095769,
+      "learning_rate": 6.865646224579108e-06,
+      "loss": 0.2878,
+      "step": 675
+    },
+    {
+      "epoch": 1.285171102661597,
+      "grad_norm": 0.7630095616579328,
+      "learning_rate": 6.834127769870134e-06,
+      "loss": 0.2929,
+      "step": 676
+    },
+    {
+      "epoch": 1.2870722433460076,
+      "grad_norm": 0.7585584095215302,
+      "learning_rate": 6.802644230033373e-06,
+      "loss": 0.2976,
+      "step": 677
+    },
+    {
+      "epoch": 1.2889733840304183,
+      "grad_norm": 0.7532480721477758,
+      "learning_rate": 6.771195952285541e-06,
+      "loss": 0.3005,
+      "step": 678
+    },
+    {
+      "epoch": 1.290874524714829,
+      "grad_norm": 0.8270256329519361,
+      "learning_rate": 6.739783283454469e-06,
+      "loss": 0.2952,
+      "step": 679
+    },
+    {
+      "epoch": 1.2927756653992395,
+      "grad_norm": 0.7587301142288344,
+      "learning_rate": 6.708406569975274e-06,
+      "loss": 0.2833,
+      "step": 680
+    },
+    {
+      "epoch": 1.2946768060836502,
+      "grad_norm": 0.7033482439779688,
+      "learning_rate": 6.6770661578865444e-06,
+      "loss": 0.2912,
+      "step": 681
+    },
+    {
+      "epoch": 1.296577946768061,
+      "grad_norm": 0.6669603487950522,
+      "learning_rate": 6.645762392826509e-06,
+      "loss": 0.3021,
+      "step": 682
+    },
+    {
+      "epoch": 1.2984790874524714,
+      "grad_norm": 0.8004002738208337,
+      "learning_rate": 6.614495620029238e-06,
+      "loss": 0.3001,
+      "step": 683
+    },
+    {
+      "epoch": 1.3003802281368821,
+      "grad_norm": 0.7716813098691593,
+      "learning_rate": 6.583266184320836e-06,
+      "loss": 0.2931,
+      "step": 684
+    },
+    {
+      "epoch": 1.3022813688212929,
+      "grad_norm": 0.8057012825147489,
+      "learning_rate": 6.552074430115624e-06,
+      "loss": 0.3047,
+      "step": 685
+    },
+    {
+      "epoch": 1.3041825095057034,
+      "grad_norm": 0.6985375681980293,
+      "learning_rate": 6.520920701412371e-06,
+      "loss": 0.2847,
+      "step": 686
+    },
+    {
+      "epoch": 1.306083650190114,
+      "grad_norm": 0.75673123155883,
+      "learning_rate": 6.489805341790456e-06,
+      "loss": 0.2897,
+      "step": 687
+    },
+    {
+      "epoch": 1.3079847908745248,
+      "grad_norm": 0.8913029664087073,
+      "learning_rate": 6.458728694406124e-06,
+      "loss": 0.302,
+      "step": 688
+    },
+    {
+      "epoch": 1.3098859315589353,
+      "grad_norm": 0.7218896349426962,
+      "learning_rate": 6.427691101988673e-06,
+      "loss": 0.2985,
+      "step": 689
+    },
+    {
+      "epoch": 1.311787072243346,
+      "grad_norm": 0.7126669012820215,
+      "learning_rate": 6.396692906836686e-06,
+      "loss": 0.2973,
+      "step": 690
+    },
+    {
+      "epoch": 1.3136882129277567,
+      "grad_norm": 0.8776749967050889,
+      "learning_rate": 6.3657344508142495e-06,
+      "loss": 0.2907,
+      "step": 691
+    },
+    {
+      "epoch": 1.3155893536121672,
+      "grad_norm": 0.7267803742152971,
+      "learning_rate": 6.334816075347185e-06,
+      "loss": 0.2868,
+      "step": 692
+    },
+    {
+      "epoch": 1.317490494296578,
+      "grad_norm": 0.7104151289595657,
+      "learning_rate": 6.303938121419295e-06,
+      "loss": 0.2864,
+      "step": 693
+    },
+    {
+      "epoch": 1.3193916349809887,
+      "grad_norm": 0.723090382220675,
+      "learning_rate": 6.273100929568579e-06,
+      "loss": 0.2896,
+      "step": 694
+    },
+    {
+      "epoch": 1.3212927756653992,
+      "grad_norm": 0.7315671506293476,
+      "learning_rate": 6.242304839883502e-06,
+      "loss": 0.2941,
+      "step": 695
+    },
+    {
+      "epoch": 1.3231939163498099,
+      "grad_norm": 0.7229677017720951,
+      "learning_rate": 6.211550191999223e-06,
+      "loss": 0.3006,
+      "step": 696
+    },
+    {
+      "epoch": 1.3250950570342206,
+      "grad_norm": 0.7686691764774978,
+      "learning_rate": 6.18083732509387e-06,
+      "loss": 0.2988,
+      "step": 697
+    },
+    {
+      "epoch": 1.326996197718631,
+      "grad_norm": 0.7363604791621042,
+      "learning_rate": 6.150166577884781e-06,
+      "loss": 0.2863,
+      "step": 698
+    },
+    {
+      "epoch": 1.3288973384030418,
+      "grad_norm": 0.7449563703696024,
+      "learning_rate": 6.119538288624778e-06,
+      "loss": 0.2899,
+      "step": 699
+    },
+    {
+      "epoch": 1.3307984790874525,
+      "grad_norm": 0.767385617960065,
+      "learning_rate": 6.088952795098442e-06,
+      "loss": 0.2991,
+      "step": 700
+    },
+    {
+      "epoch": 1.332699619771863,
+      "grad_norm": 0.6867327795623182,
+      "learning_rate": 6.058410434618367e-06,
+      "loss": 0.2888,
+      "step": 701
+    },
+    {
+      "epoch": 1.3346007604562737,
+      "grad_norm": 0.6538228839866737,
+      "learning_rate": 6.027911544021465e-06,
+      "loss": 0.2861,
+      "step": 702
+    },
+    {
+      "epoch": 1.3365019011406845,
+      "grad_norm": 0.7007923724459976,
+      "learning_rate": 5.997456459665237e-06,
+      "loss": 0.2933,
+      "step": 703
+    },
+    {
+      "epoch": 1.338403041825095,
+      "grad_norm": 0.6750119268224686,
+      "learning_rate": 5.967045517424062e-06,
+      "loss": 0.2938,
+      "step": 704
+    },
+    {
+      "epoch": 1.3403041825095057,
+      "grad_norm": 0.758200582607092,
+      "learning_rate": 5.936679052685505e-06,
+      "loss": 0.3025,
+      "step": 705
+    },
+    {
+      "epoch": 1.3422053231939164,
+      "grad_norm": 0.7140089593182961,
+      "learning_rate": 5.906357400346596e-06,
+      "loss": 0.2972,
+      "step": 706
+    },
+    {
+      "epoch": 1.3441064638783269,
+      "grad_norm": 0.7147115164229634,
+      "learning_rate": 5.876080894810167e-06,
+      "loss": 0.2919,
+      "step": 707
+    },
+    {
+      "epoch": 1.3460076045627376,
+      "grad_norm": 0.7062357782223777,
+      "learning_rate": 5.845849869981137e-06,
+      "loss": 0.2883,
+      "step": 708
+    },
+    {
+      "epoch": 1.3479087452471483,
+      "grad_norm": 0.750257759175648,
+      "learning_rate": 5.815664659262845e-06,
+      "loss": 0.3108,
+      "step": 709
+    },
+    {
+      "epoch": 1.3498098859315588,
+      "grad_norm": 0.7245006496794971,
+      "learning_rate": 5.78552559555337e-06,
+      "loss": 0.2935,
+      "step": 710
+    },
+    {
+      "epoch": 1.3517110266159695,
+      "grad_norm": 0.6948142997147745,
+      "learning_rate": 5.755433011241851e-06,
+      "loss": 0.2848,
+      "step": 711
+    },
+    {
+      "epoch": 1.3536121673003803,
+      "grad_norm": 0.787716724073085,
+      "learning_rate": 5.725387238204831e-06,
+      "loss": 0.3065,
+      "step": 712
+    },
+    {
+      "epoch": 1.3555133079847907,
+      "grad_norm": 0.7880594790644672,
+      "learning_rate": 5.695388607802603e-06,
+      "loss": 0.2954,
+      "step": 713
+    },
+    {
+      "epoch": 1.3574144486692015,
+      "grad_norm": 0.691728797783722,
+      "learning_rate": 5.665437450875534e-06,
+      "loss": 0.2934,
+      "step": 714
+    },
+    {
+      "epoch": 1.3593155893536122,
+      "grad_norm": 0.7095712860656869,
+      "learning_rate": 5.635534097740435e-06,
+      "loss": 0.2941,
+      "step": 715
+    },
+    {
+      "epoch": 1.3612167300380227,
+      "grad_norm": 0.7567649999684825,
+      "learning_rate": 5.605678878186911e-06,
+      "loss": 0.2936,
+      "step": 716
+    },
+    {
+      "epoch": 1.3631178707224334,
+      "grad_norm": 0.7627593586675898,
+      "learning_rate": 5.575872121473722e-06,
+      "loss": 0.2737,
+      "step": 717
+    },
+    {
+      "epoch": 1.3650190114068441,
+      "grad_norm": 0.7648564419006864,
+      "learning_rate": 5.546114156325166e-06,
+      "loss": 0.3006,
+      "step": 718
+    },
+    {
+      "epoch": 1.3669201520912546,
+      "grad_norm": 0.7078180468770422,
+      "learning_rate": 5.516405310927431e-06,
+      "loss": 0.2892,
+      "step": 719
+    },
+    {
+      "epoch": 1.3688212927756653,
+      "grad_norm": 0.7080635970929271,
+      "learning_rate": 5.4867459129249846e-06,
+      "loss": 0.2853,
+      "step": 720
+    },
+    {
+      "epoch": 1.370722433460076,
+      "grad_norm": 0.7377054574989295,
+      "learning_rate": 5.4571362894169795e-06,
+      "loss": 0.3022,
+      "step": 721
+    },
+    {
+      "epoch": 1.3726235741444868,
+      "grad_norm": 0.7387009935184092,
+      "learning_rate": 5.427576766953615e-06,
+      "loss": 0.2916,
+      "step": 722
+    },
+    {
+      "epoch": 1.3745247148288973,
+      "grad_norm": 0.7164951514133726,
+      "learning_rate": 5.398067671532554e-06,
+      "loss": 0.2943,
+      "step": 723
+    },
+    {
+      "epoch": 1.376425855513308,
+      "grad_norm": 0.6881296761554815,
+      "learning_rate": 5.368609328595323e-06,
+      "loss": 0.2799,
+      "step": 724
+    },
+    {
+      "epoch": 1.3783269961977187,
+      "grad_norm": 0.7130801903265463,
+      "learning_rate": 5.339202063023727e-06,
+      "loss": 0.2788,
+      "step": 725
+    },
+    {
+      "epoch": 1.3802281368821292,
+      "grad_norm": 0.685403498008217,
+      "learning_rate": 5.309846199136258e-06,
+      "loss": 0.2825,
+      "step": 726
+    },
+    {
+      "epoch": 1.38212927756654,
+      "grad_norm": 0.6876497578576143,
+      "learning_rate": 5.280542060684535e-06,
+      "loss": 0.2882,
+      "step": 727
+    },
+    {
+      "epoch": 1.3840304182509506,
+      "grad_norm": 0.7055038561341793,
+      "learning_rate": 5.2512899708497086e-06,
+      "loss": 0.2957,
+      "step": 728
+    },
+    {
+      "epoch": 1.3859315589353614,
+      "grad_norm": 0.7185374167112899,
+      "learning_rate": 5.222090252238916e-06,
+      "loss": 0.2732,
+      "step": 729
+    },
+    {
+      "epoch": 1.3878326996197718,
+      "grad_norm": 0.6705556460746867,
+      "learning_rate": 5.192943226881724e-06,
+      "loss": 0.2755,
+      "step": 730
+    },
+    {
+      "epoch": 1.3897338403041826,
+      "grad_norm": 0.7437555218174322,
+      "learning_rate": 5.163849216226562e-06,
+      "loss": 0.3094,
+      "step": 731
+    },
+    {
+      "epoch": 1.3916349809885933,
+      "grad_norm": 0.6754449395688349,
+      "learning_rate": 5.134808541137183e-06,
+      "loss": 0.2854,
+      "step": 732
+    },
+    {
+      "epoch": 1.3935361216730038,
+      "grad_norm": 0.7121376813199195,
+      "learning_rate": 5.105821521889147e-06,
+      "loss": 0.3039,
+      "step": 733
+    },
+    {
+      "epoch": 1.3954372623574145,
+      "grad_norm": 0.6992265396352645,
+      "learning_rate": 5.076888478166247e-06,
+      "loss": 0.2986,
+      "step": 734
+    },
+    {
+      "epoch": 1.3973384030418252,
+      "grad_norm": 0.6993108649187968,
+      "learning_rate": 5.048009729057012e-06,
+      "loss": 0.2916,
+      "step": 735
+    },
+    {
+      "epoch": 1.3992395437262357,
+      "grad_norm": 0.7424823885578106,
+      "learning_rate": 5.0191855930511946e-06,
+      "loss": 0.3042,
+      "step": 736
+    },
+    {
+      "epoch": 1.4011406844106464,
+      "grad_norm": 0.6442190229251181,
+      "learning_rate": 4.990416388036233e-06,
+      "loss": 0.2772,
+      "step": 737
+    },
+    {
+      "epoch": 1.4030418250950571,
+      "grad_norm": 0.7867175238965394,
+      "learning_rate": 4.961702431293759e-06,
+      "loss": 0.3007,
+      "step": 738
+    },
+    {
+      "epoch": 1.4049429657794676,
+      "grad_norm": 0.7866172530477322,
+      "learning_rate": 4.933044039496107e-06,
+      "loss": 0.2833,
+      "step": 739
+    },
+    {
+      "epoch": 1.4068441064638784,
+      "grad_norm": 0.6864068831503692,
+      "learning_rate": 4.904441528702806e-06,
+      "loss": 0.279,
+      "step": 740
+    },
+    {
+      "epoch": 1.408745247148289,
+      "grad_norm": 0.7986991575216565,
+      "learning_rate": 4.875895214357093e-06,
+      "loss": 0.2865,
+      "step": 741
+    },
+    {
+      "epoch": 1.4106463878326996,
+      "grad_norm": 0.7683263423160659,
+      "learning_rate": 4.847405411282462e-06,
+      "loss": 0.3078,
+      "step": 742
+    },
+    {
+      "epoch": 1.4125475285171103,
+      "grad_norm": 0.6845841810990767,
+      "learning_rate": 4.818972433679145e-06,
+      "loss": 0.2809,
+      "step": 743
+    },
+    {
+      "epoch": 1.414448669201521,
+      "grad_norm": 0.7321332885975637,
+      "learning_rate": 4.790596595120699e-06,
+      "loss": 0.2912,
+      "step": 744
+    },
+    {
+      "epoch": 1.4163498098859315,
+      "grad_norm": 0.6415447021318383,
+      "learning_rate": 4.762278208550505e-06,
+      "loss": 0.2728,
+      "step": 745
+    },
+    {
+      "epoch": 1.4182509505703422,
+      "grad_norm": 0.7035329294177123,
+      "learning_rate": 4.734017586278337e-06,
+      "loss": 0.2868,
+      "step": 746
+    },
+    {
+      "epoch": 1.420152091254753,
+      "grad_norm": 0.7009824992959836,
+      "learning_rate": 4.7058150399769245e-06,
+      "loss": 0.2938,
+      "step": 747
+    },
+    {
+      "epoch": 1.4220532319391634,
+      "grad_norm": 0.7647853533494099,
+      "learning_rate": 4.677670880678493e-06,
+      "loss": 0.2897,
+      "step": 748
+    },
+    {
+      "epoch": 1.4239543726235742,
+      "grad_norm": 0.6835540498998305,
+      "learning_rate": 4.649585418771348e-06,
+      "loss": 0.2838,
+      "step": 749
+    },
+    {
+      "epoch": 1.4258555133079849,
+      "grad_norm": 0.7244853957373847,
+      "learning_rate": 4.621558963996458e-06,
+      "loss": 0.2916,
+      "step": 750
+    },
+    {
+      "epoch": 1.4277566539923954,
+      "grad_norm": 0.6939127653102627,
+      "learning_rate": 4.593591825444028e-06,
+      "loss": 0.2871,
+      "step": 751
+    },
+    {
+      "epoch": 1.429657794676806,
+      "grad_norm": 0.719413352240235,
+      "learning_rate": 4.565684311550077e-06,
+      "loss": 0.2948,
+      "step": 752
+    },
+    {
+      "epoch": 1.4315589353612168,
+      "grad_norm": 0.7397354503408802,
+      "learning_rate": 4.537836730093077e-06,
+      "loss": 0.3013,
+      "step": 753
+    },
+    {
+      "epoch": 1.4334600760456273,
+      "grad_norm": 0.6559875871615594,
+      "learning_rate": 4.510049388190518e-06,
+      "loss": 0.2819,
+      "step": 754
+    },
+    {
+      "epoch": 1.435361216730038,
+      "grad_norm": 0.7024784554905884,
+      "learning_rate": 4.482322592295541e-06,
+      "loss": 0.2847,
+      "step": 755
+    },
+    {
+      "epoch": 1.4372623574144487,
+      "grad_norm": 0.6855117068567005,
+      "learning_rate": 4.454656648193559e-06,
+      "loss": 0.271,
+      "step": 756
+    },
+    {
+      "epoch": 1.4391634980988592,
+      "grad_norm": 0.6875557719006284,
+      "learning_rate": 4.427051860998877e-06,
+      "loss": 0.2862,
+      "step": 757
+    },
+    {
+      "epoch": 1.44106463878327,
+      "grad_norm": 0.6368405319292746,
+      "learning_rate": 4.399508535151321e-06,
+      "loss": 0.275,
+      "step": 758
+    },
+    {
+      "epoch": 1.4429657794676807,
+      "grad_norm": 0.7070237098908724,
+      "learning_rate": 4.372026974412907e-06,
+      "loss": 0.287,
+      "step": 759
+    },
+    {
+      "epoch": 1.4448669201520912,
+      "grad_norm": 0.7214302755658806,
+      "learning_rate": 4.344607481864466e-06,
+      "loss": 0.2841,
+      "step": 760
+    },
+    {
+      "epoch": 1.446768060836502,
+      "grad_norm": 0.6805798150387056,
+      "learning_rate": 4.317250359902295e-06,
+      "loss": 0.2877,
+      "step": 761
+    },
+    {
+      "epoch": 1.4486692015209126,
+      "grad_norm": 0.7054549258410008,
+      "learning_rate": 4.2899559102348585e-06,
+      "loss": 0.2932,
+      "step": 762
+    },
+    {
+      "epoch": 1.450570342205323,
+      "grad_norm": 0.680013177910947,
+      "learning_rate": 4.262724433879427e-06,
+      "loss": 0.2863,
+      "step": 763
+    },
+    {
+      "epoch": 1.4524714828897338,
+      "grad_norm": 0.7249433516348669,
+      "learning_rate": 4.235556231158765e-06,
+      "loss": 0.2939,
+      "step": 764
+    },
+    {
+      "epoch": 1.4543726235741445,
+      "grad_norm": 0.7005572753514715,
+      "learning_rate": 4.208451601697836e-06,
+      "loss": 0.2781,
+      "step": 765
+    },
+    {
+      "epoch": 1.456273764258555,
+      "grad_norm": 0.6995960106331317,
+      "learning_rate": 4.181410844420473e-06,
+      "loss": 0.2875,
+      "step": 766
+    },
+    {
+      "epoch": 1.4581749049429658,
+      "grad_norm": 0.6921736500151292,
+      "learning_rate": 4.154434257546095e-06,
+      "loss": 0.2821,
+      "step": 767
+    },
+    {
+      "epoch": 1.4600760456273765,
+      "grad_norm": 0.6759874126813181,
+      "learning_rate": 4.127522138586424e-06,
+      "loss": 0.2829,
+      "step": 768
+    },
+    {
+      "epoch": 1.461977186311787,
+      "grad_norm": 0.6907757777847995,
+      "learning_rate": 4.10067478434219e-06,
+      "loss": 0.2887,
+      "step": 769
+    },
+    {
+      "epoch": 1.4638783269961977,
+      "grad_norm": 0.671963078553791,
+      "learning_rate": 4.073892490899865e-06,
+      "loss": 0.2824,
+      "step": 770
+    },
+    {
+      "epoch": 1.4657794676806084,
+      "grad_norm": 0.6727332010036039,
+      "learning_rate": 4.047175553628397e-06,
+      "loss": 0.2882,
+      "step": 771
+    },
+    {
+      "epoch": 1.467680608365019,
+      "grad_norm": 0.707064725708802,
+      "learning_rate": 4.020524267175954e-06,
+      "loss": 0.2822,
+      "step": 772
+    },
+    {
+      "epoch": 1.4695817490494296,
+      "grad_norm": 0.6935393091320086,
+      "learning_rate": 3.993938925466674e-06,
+      "loss": 0.2837,
+      "step": 773
+    },
+    {
+      "epoch": 1.4714828897338403,
+      "grad_norm": 0.6789549188226693,
+      "learning_rate": 3.96741982169742e-06,
+      "loss": 0.2873,
+      "step": 774
+    },
+    {
+      "epoch": 1.4733840304182508,
+      "grad_norm": 0.689529019323658,
+      "learning_rate": 3.9409672483345465e-06,
+      "loss": 0.286,
+      "step": 775
+    },
+    {
+      "epoch": 1.4752851711026616,
+      "grad_norm": 0.7410494092527087,
+      "learning_rate": 3.914581497110684e-06,
+      "loss": 0.2886,
+      "step": 776
+    },
+    {
+      "epoch": 1.4771863117870723,
+      "grad_norm": 0.7161690856614648,
+      "learning_rate": 3.888262859021508e-06,
+      "loss": 0.2827,
+      "step": 777
+    },
+    {
+      "epoch": 1.4790874524714828,
+      "grad_norm": 0.6849943061865478,
+      "learning_rate": 3.862011624322534e-06,
+      "loss": 0.2823,
+      "step": 778
+    },
+    {
+      "epoch": 1.4809885931558935,
+      "grad_norm": 0.6899930204535697,
+      "learning_rate": 3.835828082525925e-06,
+      "loss": 0.2944,
+      "step": 779
+    },
+    {
+      "epoch": 1.4828897338403042,
+      "grad_norm": 0.7179224846121838,
+      "learning_rate": 3.8097125223972864e-06,
+      "loss": 0.2978,
+      "step": 780
+    },
+    {
+      "epoch": 1.4847908745247147,
+      "grad_norm": 0.7423250574446931,
+      "learning_rate": 3.7836652319524835e-06,
+      "loss": 0.2979,
+      "step": 781
+    },
+    {
+      "epoch": 1.4866920152091254,
+      "grad_norm": 0.7182254504678957,
+      "learning_rate": 3.7576864984544814e-06,
+      "loss": 0.2933,
+      "step": 782
+    },
+    {
+      "epoch": 1.4885931558935361,
+      "grad_norm": 0.7131191437823985,
+      "learning_rate": 3.73177660841015e-06,
+      "loss": 0.31,
+      "step": 783
+    },
+    {
+      "epoch": 1.4904942965779466,
+      "grad_norm": 0.7217337463353969,
+      "learning_rate": 3.7059358475671225e-06,
+      "loss": 0.2742,
+      "step": 784
+    },
+    {
+      "epoch": 1.4923954372623573,
+      "grad_norm": 0.7070958423167292,
+      "learning_rate": 3.680164500910646e-06,
+      "loss": 0.2826,
+      "step": 785
+    },
+    {
+      "epoch": 1.494296577946768,
+      "grad_norm": 0.6924765281583363,
+      "learning_rate": 3.654462852660423e-06,
+      "loss": 0.2915,
+      "step": 786
+    },
+    {
+      "epoch": 1.4961977186311788,
+      "grad_norm": 0.6862629082439664,
+      "learning_rate": 3.6288311862674885e-06,
+      "loss": 0.2865,
+      "step": 787
+    },
+    {
+      "epoch": 1.4980988593155893,
+      "grad_norm": 0.7293413737634317,
+      "learning_rate": 3.6032697844110896e-06,
+      "loss": 0.31,
+      "step": 788
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6781245925486249,
+      "learning_rate": 3.5777789289955454e-06,
+      "loss": 0.2844,
+      "step": 789
+    },
+    {
+      "epoch": 1.5019011406844105,
+      "grad_norm": 0.655219464396975,
+      "learning_rate": 3.5523589011471592e-06,
+      "loss": 0.2678,
+      "step": 790
+    },
+    {
+      "epoch": 1.5038022813688214,
+      "grad_norm": 0.7286035108762859,
+      "learning_rate": 3.527009981211119e-06,
+      "loss": 0.2902,
+      "step": 791
+    },
+    {
+      "epoch": 1.505703422053232,
+      "grad_norm": 0.6833423358915883,
+      "learning_rate": 3.5017324487483873e-06,
+      "loss": 0.2778,
+      "step": 792
+    },
+    {
+      "epoch": 1.5076045627376424,
+      "grad_norm": 0.691380065731547,
+      "learning_rate": 3.47652658253263e-06,
+      "loss": 0.2898,
+      "step": 793
+    },
+    {
+      "epoch": 1.5095057034220534,
+      "grad_norm": 0.6946593517127586,
+      "learning_rate": 3.4513926605471504e-06,
+      "loss": 0.2828,
+      "step": 794
+    },
+    {
+      "epoch": 1.5114068441064639,
+      "grad_norm": 0.6750790323910052,
+      "learning_rate": 3.4263309599818017e-06,
+      "loss": 0.2768,
+      "step": 795
+    },
+    {
+      "epoch": 1.5133079847908744,
+      "grad_norm": 0.6931268325245681,
+      "learning_rate": 3.4013417572299446e-06,
+      "loss": 0.2783,
+      "step": 796
+    },
+    {
+      "epoch": 1.5152091254752853,
+      "grad_norm": 0.6806967499830385,
+      "learning_rate": 3.37642532788541e-06,
+      "loss": 0.2766,
+      "step": 797
+    },
+    {
+      "epoch": 1.5171102661596958,
+      "grad_norm": 0.6863882980456473,
+      "learning_rate": 3.3515819467394184e-06,
+      "loss": 0.2905,
+      "step": 798
+    },
+    {
+      "epoch": 1.5190114068441065,
+      "grad_norm": 0.6991396055600565,
+      "learning_rate": 3.326811887777607e-06,
+      "loss": 0.2903,
+      "step": 799
+    },
+    {
+      "epoch": 1.5209125475285172,
+      "grad_norm": 0.701170505097887,
+      "learning_rate": 3.3021154241769606e-06,
+      "loss": 0.2924,
+      "step": 800
+    },
+    {
+      "epoch": 1.5228136882129277,
+      "grad_norm": 0.7418493414265517,
+      "learning_rate": 3.2774928283028153e-06,
+      "loss": 0.2919,
+      "step": 801
+    },
+    {
+      "epoch": 1.5247148288973384,
+      "grad_norm": 0.7420494337571956,
+      "learning_rate": 3.2529443717058693e-06,
+      "loss": 0.2931,
+      "step": 802
+    },
+    {
+      "epoch": 1.5266159695817492,
+      "grad_norm": 0.6838582624578541,
+      "learning_rate": 3.228470325119164e-06,
+      "loss": 0.2864,
+      "step": 803
+    },
+    {
+      "epoch": 1.5285171102661597,
+      "grad_norm": 0.719400830331535,
+      "learning_rate": 3.20407095845511e-06,
+      "loss": 0.2825,
+      "step": 804
+    },
+    {
+      "epoch": 1.5304182509505704,
+      "grad_norm": 0.7667188204281035,
+      "learning_rate": 3.179746540802506e-06,
+      "loss": 0.2992,
+      "step": 805
+    },
+    {
+      "epoch": 1.532319391634981,
+      "grad_norm": 0.6746219143683031,
+      "learning_rate": 3.155497340423588e-06,
+      "loss": 0.2896,
+      "step": 806
+    },
+    {
+      "epoch": 1.5342205323193916,
+      "grad_norm": 0.6975540826815945,
+      "learning_rate": 3.1313236247510414e-06,
+      "loss": 0.2863,
+      "step": 807
+    },
+    {
+      "epoch": 1.5361216730038023,
+      "grad_norm": 0.7452492585906716,
+      "learning_rate": 3.107225660385077e-06,
+      "loss": 0.3046,
+      "step": 808
+    },
+    {
+      "epoch": 1.538022813688213,
+      "grad_norm": 0.6965225111927277,
+      "learning_rate": 3.0832037130904748e-06,
+      "loss": 0.2898,
+      "step": 809
+    },
+    {
+      "epoch": 1.5399239543726235,
+      "grad_norm": 0.7133384074035564,
+      "learning_rate": 3.0592580477936606e-06,
+      "loss": 0.2779,
+      "step": 810
+    },
+    {
+      "epoch": 1.5418250950570342,
+      "grad_norm": 0.7362459802574304,
+      "learning_rate": 3.035388928579792e-06,
+      "loss": 0.3031,
+      "step": 811
+    },
+    {
+      "epoch": 1.543726235741445,
+      "grad_norm": 0.6879664964212607,
+      "learning_rate": 3.011596618689825e-06,
+      "loss": 0.2875,
+      "step": 812
+    },
+    {
+      "epoch": 1.5456273764258555,
+      "grad_norm": 0.6741776212335533,
+      "learning_rate": 2.9878813805176252e-06,
+      "loss": 0.2737,
+      "step": 813
+    },
+    {
+      "epoch": 1.5475285171102662,
+      "grad_norm": 0.7200753520775072,
+      "learning_rate": 2.9642434756070793e-06,
+      "loss": 0.2845,
+      "step": 814
+    },
+    {
+      "epoch": 1.549429657794677,
+      "grad_norm": 0.6929446035108708,
+      "learning_rate": 2.940683164649194e-06,
+      "loss": 0.2877,
+      "step": 815
+    },
+    {
+      "epoch": 1.5513307984790874,
+      "grad_norm": 0.6976568367778189,
+      "learning_rate": 2.9172007074792342e-06,
+      "loss": 0.2842,
+      "step": 816
+    },
+    {
+      "epoch": 1.553231939163498,
+      "grad_norm": 0.6678546377317014,
+      "learning_rate": 2.8937963630738517e-06,
+      "loss": 0.2761,
+      "step": 817
+    },
+    {
+      "epoch": 1.5551330798479088,
+      "grad_norm": 0.6940759465861326,
+      "learning_rate": 2.87047038954823e-06,
+      "loss": 0.2862,
+      "step": 818
+    },
+    {
+      "epoch": 1.5570342205323193,
+      "grad_norm": 0.6690227775225014,
+      "learning_rate": 2.8472230441532365e-06,
+      "loss": 0.2779,
+      "step": 819
+    },
+    {
+      "epoch": 1.55893536121673,
+      "grad_norm": 0.6840411242272926,
+      "learning_rate": 2.8240545832725963e-06,
+      "loss": 0.2768,
+      "step": 820
+    },
+    {
+      "epoch": 1.5608365019011408,
+      "grad_norm": 0.6852816718638697,
+      "learning_rate": 2.8009652624200436e-06,
+      "loss": 0.2853,
+      "step": 821
+    },
+    {
+      "epoch": 1.5627376425855513,
+      "grad_norm": 0.6792970232410612,
+      "learning_rate": 2.7779553362365184e-06,
+      "loss": 0.2737,
+      "step": 822
+    },
+    {
+      "epoch": 1.564638783269962,
+      "grad_norm": 0.7231175105353772,
+      "learning_rate": 2.755025058487364e-06,
+      "loss": 0.2887,
+      "step": 823
+    },
+    {
+      "epoch": 1.5665399239543727,
+      "grad_norm": 0.7564121601109488,
+      "learning_rate": 2.7321746820595084e-06,
+      "loss": 0.2841,
+      "step": 824
+    },
+    {
+      "epoch": 1.5684410646387832,
+      "grad_norm": 0.7230496947374356,
+      "learning_rate": 2.709404458958693e-06,
+      "loss": 0.2945,
+      "step": 825
+    },
+    {
+      "epoch": 1.570342205323194,
+      "grad_norm": 0.6773800996473027,
+      "learning_rate": 2.6867146403066833e-06,
+      "loss": 0.2791,
+      "step": 826
+    },
+    {
+      "epoch": 1.5722433460076046,
+      "grad_norm": 0.6765266513493975,
+      "learning_rate": 2.6641054763385044e-06,
+      "loss": 0.2669,
+      "step": 827
+    },
+    {
+      "epoch": 1.5741444866920151,
+      "grad_norm": 0.7094342560094086,
+      "learning_rate": 2.6415772163996845e-06,
+      "loss": 0.2869,
+      "step": 828
+    },
+    {
+      "epoch": 1.5760456273764258,
+      "grad_norm": 0.7613174689756727,
+      "learning_rate": 2.619130108943494e-06,
+      "loss": 0.2958,
+      "step": 829
+    },
+    {
+      "epoch": 1.5779467680608366,
+      "grad_norm": 0.6503123872063992,
+      "learning_rate": 2.5967644015282146e-06,
+      "loss": 0.2659,
+      "step": 830
+    },
+    {
+      "epoch": 1.579847908745247,
+      "grad_norm": 0.6767478301446936,
+      "learning_rate": 2.5744803408144026e-06,
+      "loss": 0.2872,
+      "step": 831
+    },
+    {
+      "epoch": 1.5817490494296578,
+      "grad_norm": 0.7131535074798426,
+      "learning_rate": 2.5522781725621814e-06,
+      "loss": 0.2992,
+      "step": 832
+    },
+    {
+      "epoch": 1.5836501901140685,
+      "grad_norm": 0.6982744130002001,
+      "learning_rate": 2.530158141628515e-06,
+      "loss": 0.2838,
+      "step": 833
+    },
+    {
+      "epoch": 1.585551330798479,
+      "grad_norm": 0.7003365873705352,
+      "learning_rate": 2.508120491964512e-06,
+      "loss": 0.2926,
+      "step": 834
+    },
+    {
+      "epoch": 1.5874524714828897,
+      "grad_norm": 0.6947517264224584,
+      "learning_rate": 2.486165466612751e-06,
+      "loss": 0.2925,
+      "step": 835
+    },
+    {
+      "epoch": 1.5893536121673004,
+      "grad_norm": 0.699254005018514,
+      "learning_rate": 2.464293307704566e-06,
+      "loss": 0.2886,
+      "step": 836
+    },
+    {
+      "epoch": 1.591254752851711,
+      "grad_norm": 0.7241526375659275,
+      "learning_rate": 2.4425042564574186e-06,
+      "loss": 0.2773,
+      "step": 837
+    },
+    {
+      "epoch": 1.5931558935361216,
+      "grad_norm": 0.6871386040104367,
+      "learning_rate": 2.4207985531722034e-06,
+      "loss": 0.2875,
+      "step": 838
+    },
+    {
+      "epoch": 1.5950570342205324,
+      "grad_norm": 0.6996379375260411,
+      "learning_rate": 2.3991764372306113e-06,
+      "loss": 0.2859,
+      "step": 839
+    },
+    {
+      "epoch": 1.5969581749049429,
+      "grad_norm": 0.6380890631939141,
+      "learning_rate": 2.377638147092497e-06,
+      "loss": 0.2688,
+      "step": 840
+    },
+    {
+      "epoch": 1.5988593155893536,
+      "grad_norm": 0.7136113209293522,
+      "learning_rate": 2.3561839202932344e-06,
+      "loss": 0.2844,
+      "step": 841
+    },
+    {
+      "epoch": 1.6007604562737643,
+      "grad_norm": 0.7145125144894247,
+      "learning_rate": 2.3348139934411008e-06,
+      "loss": 0.2839,
+      "step": 842
+    },
+    {
+      "epoch": 1.6026615969581748,
+      "grad_norm": 0.6909087155238696,
+      "learning_rate": 2.3135286022146785e-06,
+      "loss": 0.2826,
+      "step": 843
+    },
+    {
+      "epoch": 1.6045627376425855,
+      "grad_norm": 0.6512420474349377,
+      "learning_rate": 2.292327981360245e-06,
+      "loss": 0.2618,
+      "step": 844
+    },
+    {
+      "epoch": 1.6064638783269962,
+      "grad_norm": 0.7194007297084652,
+      "learning_rate": 2.271212364689176e-06,
+      "loss": 0.2842,
+      "step": 845
+    },
+    {
+      "epoch": 1.6083650190114067,
+      "grad_norm": 0.695884315636097,
+      "learning_rate": 2.2501819850753925e-06,
+      "loss": 0.2929,
+      "step": 846
+    },
+    {
+      "epoch": 1.6102661596958177,
+      "grad_norm": 0.7113584741953002,
+      "learning_rate": 2.229237074452768e-06,
+      "loss": 0.2842,
+      "step": 847
+    },
+    {
+      "epoch": 1.6121673003802282,
+      "grad_norm": 0.6520484020596302,
+      "learning_rate": 2.2083778638125796e-06,
+      "loss": 0.2796,
+      "step": 848
+    },
+    {
+      "epoch": 1.6140684410646386,
+      "grad_norm": 0.7016462293815009,
+      "learning_rate": 2.1876045832009694e-06,
+      "loss": 0.2841,
+      "step": 849
+    },
+    {
+      "epoch": 1.6159695817490496,
+      "grad_norm": 0.6895203905945645,
+      "learning_rate": 2.16691746171639e-06,
+      "loss": 0.2882,
+      "step": 850
+    },
+    {
+      "epoch": 1.61787072243346,
+      "grad_norm": 0.7165264430796174,
+      "learning_rate": 2.1463167275070863e-06,
+      "loss": 0.2892,
+      "step": 851
+    },
+    {
+      "epoch": 1.6197718631178706,
+      "grad_norm": 0.6701604265430632,
+      "learning_rate": 2.125802607768588e-06,
+      "loss": 0.2873,
+      "step": 852
+    },
+    {
+      "epoch": 1.6216730038022815,
+      "grad_norm": 0.7279770691756008,
+      "learning_rate": 2.1053753287411895e-06,
+      "loss": 0.2778,
+      "step": 853
+    },
+    {
+      "epoch": 1.623574144486692,
+      "grad_norm": 0.6867028130356997,
+      "learning_rate": 2.08503511570746e-06,
+      "loss": 0.2843,
+      "step": 854
+    },
+    {
+      "epoch": 1.6254752851711025,
+      "grad_norm": 0.693036878948542,
+      "learning_rate": 2.064782192989765e-06,
+      "loss": 0.2727,
+      "step": 855
+    },
+    {
+      "epoch": 1.6273764258555135,
+      "grad_norm": 0.6899927665140845,
+      "learning_rate": 2.0446167839477815e-06,
+      "loss": 0.2785,
+      "step": 856
+    },
+    {
+      "epoch": 1.629277566539924,
+      "grad_norm": 0.6731849685395901,
+      "learning_rate": 2.0245391109760437e-06,
+      "loss": 0.2819,
+      "step": 857
+    },
+    {
+      "epoch": 1.6311787072243344,
+      "grad_norm": 0.7109608935758531,
+      "learning_rate": 2.0045493955014915e-06,
+      "loss": 0.2786,
+      "step": 858
+    },
+    {
+      "epoch": 1.6330798479087454,
+      "grad_norm": 0.7108900693932255,
+      "learning_rate": 1.984647857981017e-06,
+      "loss": 0.2856,
+      "step": 859
+    },
+    {
+      "epoch": 1.6349809885931559,
+      "grad_norm": 0.6948953383329372,
+      "learning_rate": 1.96483471789904e-06,
+      "loss": 0.2905,
+      "step": 860
+    },
+    {
+      "epoch": 1.6368821292775664,
+      "grad_norm": 0.7293218635180828,
+      "learning_rate": 1.9451101937650963e-06,
+      "loss": 0.2916,
+      "step": 861
+    },
+    {
+      "epoch": 1.6387832699619773,
+      "grad_norm": 0.7061527175076288,
+      "learning_rate": 1.925474503111412e-06,
+      "loss": 0.2915,
+      "step": 862
+    },
+    {
+      "epoch": 1.6406844106463878,
+      "grad_norm": 0.7206267033170142,
+      "learning_rate": 1.905927862490512e-06,
+      "loss": 0.2896,
+      "step": 863
+    },
+    {
+      "epoch": 1.6425855513307985,
+      "grad_norm": 0.7100577604899369,
+      "learning_rate": 1.8864704874728346e-06,
+      "loss": 0.2807,
+      "step": 864
+    },
+    {
+      "epoch": 1.6444866920152093,
+      "grad_norm": 0.7662031105722321,
+      "learning_rate": 1.8671025926443464e-06,
+      "loss": 0.2876,
+      "step": 865
+    },
+    {
+      "epoch": 1.6463878326996197,
+      "grad_norm": 0.6870846567243655,
+      "learning_rate": 1.8478243916041882e-06,
+      "loss": 0.2717,
+      "step": 866
+    },
+    {
+      "epoch": 1.6482889733840305,
+      "grad_norm": 0.7167951796337592,
+      "learning_rate": 1.828636096962304e-06,
+      "loss": 0.2725,
+      "step": 867
+    },
+    {
+      "epoch": 1.6501901140684412,
+      "grad_norm": 0.7308000282282252,
+      "learning_rate": 1.8095379203371044e-06,
+      "loss": 0.2883,
+      "step": 868
+    },
+    {
+      "epoch": 1.6520912547528517,
+      "grad_norm": 0.7166855173512088,
+      "learning_rate": 1.7905300723531393e-06,
+      "loss": 0.2791,
+      "step": 869
+    },
+    {
+      "epoch": 1.6539923954372624,
+      "grad_norm": 0.6967871186157718,
+      "learning_rate": 1.771612762638758e-06,
+      "loss": 0.2833,
+      "step": 870
+    },
+    {
+      "epoch": 1.6558935361216731,
+      "grad_norm": 0.6997749552128736,
+      "learning_rate": 1.7527861998238094e-06,
+      "loss": 0.2849,
+      "step": 871
+    },
+    {
+      "epoch": 1.6577946768060836,
+      "grad_norm": 0.7166921106065282,
+      "learning_rate": 1.7340505915373495e-06,
+      "loss": 0.2735,
+      "step": 872
+    },
+    {
+      "epoch": 1.6596958174904943,
+      "grad_norm": 0.6862718993651726,
+      "learning_rate": 1.7154061444053239e-06,
+      "loss": 0.2848,
+      "step": 873
+    },
+    {
+      "epoch": 1.661596958174905,
+      "grad_norm": 0.6953764603260656,
+      "learning_rate": 1.6968530640483126e-06,
+      "loss": 0.2829,
+      "step": 874
+    },
+    {
+      "epoch": 1.6634980988593155,
+      "grad_norm": 0.6669571251368059,
+      "learning_rate": 1.6783915550792652e-06,
+      "loss": 0.2798,
+      "step": 875
+    },
+    {
+      "epoch": 1.6653992395437263,
+      "grad_norm": 0.6683365724238115,
+      "learning_rate": 1.660021821101222e-06,
+      "loss": 0.2788,
+      "step": 876
+    },
+    {
+      "epoch": 1.667300380228137,
+      "grad_norm": 0.6712908824420041,
+      "learning_rate": 1.6417440647050853e-06,
+      "loss": 0.2728,
+      "step": 877
+    },
+    {
+      "epoch": 1.6692015209125475,
+      "grad_norm": 0.6810745396550731,
+      "learning_rate": 1.6235584874673848e-06,
+      "loss": 0.2773,
+      "step": 878
+    },
+    {
+      "epoch": 1.6711026615969582,
+      "grad_norm": 0.714676377324762,
+      "learning_rate": 1.6054652899480472e-06,
+      "loss": 0.2747,
+      "step": 879
+    },
+    {
+      "epoch": 1.673003802281369,
+      "grad_norm": 0.6240311451749357,
+      "learning_rate": 1.587464671688187e-06,
+      "loss": 0.2627,
+      "step": 880
+    },
+    {
+      "epoch": 1.6749049429657794,
+      "grad_norm": 0.672961578237575,
+      "learning_rate": 1.5695568312079156e-06,
+      "loss": 0.2764,
+      "step": 881
+    },
+    {
+      "epoch": 1.6768060836501901,
+      "grad_norm": 0.7089937637793987,
+      "learning_rate": 1.5517419660041277e-06,
+      "loss": 0.2857,
+      "step": 882
+    },
+    {
+      "epoch": 1.6787072243346008,
+      "grad_norm": 0.7303510230792566,
+      "learning_rate": 1.534020272548349e-06,
+      "loss": 0.2894,
+      "step": 883
+    },
+    {
+      "epoch": 1.6806083650190113,
+      "grad_norm": 0.7032786807341086,
+      "learning_rate": 1.5163919462845622e-06,
+      "loss": 0.2864,
+      "step": 884
+    },
+    {
+      "epoch": 1.682509505703422,
+      "grad_norm": 0.7105988575630341,
+      "learning_rate": 1.4988571816270402e-06,
+      "loss": 0.2937,
+      "step": 885
+    },
+    {
+      "epoch": 1.6844106463878328,
+      "grad_norm": 0.722904439023986,
+      "learning_rate": 1.4814161719582132e-06,
+      "loss": 0.2867,
+      "step": 886
+    },
+    {
+      "epoch": 1.6863117870722433,
+      "grad_norm": 0.6959977259103303,
+      "learning_rate": 1.4640691096265358e-06,
+      "loss": 0.2771,
+      "step": 887
+    },
+    {
+      "epoch": 1.688212927756654,
+      "grad_norm": 0.6964079371677274,
+      "learning_rate": 1.4468161859443609e-06,
+      "loss": 0.2844,
+      "step": 888
+    },
+    {
+      "epoch": 1.6901140684410647,
+      "grad_norm": 0.6710479433553769,
+      "learning_rate": 1.4296575911858268e-06,
+      "loss": 0.2636,
+      "step": 889
+    },
+    {
+      "epoch": 1.6920152091254752,
+      "grad_norm": 0.7188740812864929,
+      "learning_rate": 1.412593514584777e-06,
+      "loss": 0.2818,
+      "step": 890
+    },
+    {
+      "epoch": 1.693916349809886,
+      "grad_norm": 0.7222174060094092,
+      "learning_rate": 1.3956241443326423e-06,
+      "loss": 0.2791,
+      "step": 891
+    },
+    {
+      "epoch": 1.6958174904942966,
+      "grad_norm": 0.728543938310975,
+      "learning_rate": 1.378749667576399e-06,
+      "loss": 0.2923,
+      "step": 892
+    },
+    {
+      "epoch": 1.6977186311787071,
+      "grad_norm": 0.6973733020029113,
+      "learning_rate": 1.3619702704164783e-06,
+      "loss": 0.2876,
+      "step": 893
+    },
+    {
+      "epoch": 1.6996197718631179,
+      "grad_norm": 0.7102311566524399,
+      "learning_rate": 1.3452861379047289e-06,
+      "loss": 0.2915,
+      "step": 894
+    },
+    {
+      "epoch": 1.7015209125475286,
+      "grad_norm": 0.6734303079313543,
+      "learning_rate": 1.3286974540423747e-06,
+      "loss": 0.2777,
+      "step": 895
+    },
+    {
+      "epoch": 1.703422053231939,
+      "grad_norm": 0.7100164102742084,
+      "learning_rate": 1.3122044017779768e-06,
+      "loss": 0.2858,
+      "step": 896
+    },
+    {
+      "epoch": 1.7053231939163498,
+      "grad_norm": 0.6532094767380489,
+      "learning_rate": 1.2958071630054214e-06,
+      "loss": 0.2744,
+      "step": 897
+    },
+    {
+      "epoch": 1.7072243346007605,
+      "grad_norm": 0.7076433920958868,
+      "learning_rate": 1.279505918561923e-06,
+      "loss": 0.295,
+      "step": 898
+    },
+    {
+      "epoch": 1.709125475285171,
+      "grad_norm": 0.7275117287956319,
+      "learning_rate": 1.2633008482260146e-06,
+      "loss": 0.2903,
+      "step": 899
+    },
+    {
+      "epoch": 1.7110266159695817,
+      "grad_norm": 0.7028834265454772,
+      "learning_rate": 1.2471921307155655e-06,
+      "loss": 0.2777,
+      "step": 900
+    },
+    {
+      "epoch": 1.7129277566539924,
+      "grad_norm": 0.6841967955757563,
+      "learning_rate": 1.2311799436858275e-06,
+      "loss": 0.2856,
+      "step": 901
+    },
+    {
+      "epoch": 1.714828897338403,
+      "grad_norm": 0.6712425697449756,
+      "learning_rate": 1.2152644637274603e-06,
+      "loss": 0.2673,
+      "step": 902
+    },
+    {
+      "epoch": 1.7167300380228137,
+      "grad_norm": 0.6734584758714657,
+      "learning_rate": 1.1994458663645836e-06,
+      "loss": 0.2885,
+      "step": 903
+    },
+    {
+      "epoch": 1.7186311787072244,
+      "grad_norm": 0.6984298248061253,
+      "learning_rate": 1.1837243260528542e-06,
+      "loss": 0.2817,
+      "step": 904
+    },
+    {
+      "epoch": 1.7205323193916349,
+      "grad_norm": 0.681874046623608,
+      "learning_rate": 1.168100016177528e-06,
+      "loss": 0.2895,
+      "step": 905
+    },
+    {
+      "epoch": 1.7224334600760456,
+      "grad_norm": 0.7057399076953761,
+      "learning_rate": 1.1525731090515536e-06,
+      "loss": 0.2876,
+      "step": 906
+    },
+    {
+      "epoch": 1.7243346007604563,
+      "grad_norm": 0.7205139286988427,
+      "learning_rate": 1.137143775913675e-06,
+      "loss": 0.2897,
+      "step": 907
+    },
+    {
+      "epoch": 1.7262357414448668,
+      "grad_norm": 0.6922857366668834,
+      "learning_rate": 1.1218121869265365e-06,
+      "loss": 0.2769,
+      "step": 908
+    },
+    {
+      "epoch": 1.7281368821292775,
+      "grad_norm": 0.6586861747904412,
+      "learning_rate": 1.1065785111748117e-06,
+      "loss": 0.2602,
+      "step": 909
+    },
+    {
+      "epoch": 1.7300380228136882,
+      "grad_norm": 0.6781616851283059,
+      "learning_rate": 1.0914429166633355e-06,
+      "loss": 0.2758,
+      "step": 910
+    },
+    {
+      "epoch": 1.7319391634980987,
+      "grad_norm": 0.7169191546464649,
+      "learning_rate": 1.076405570315252e-06,
+      "loss": 0.2859,
+      "step": 911
+    },
+    {
+      "epoch": 1.7338403041825095,
+      "grad_norm": 0.7337536442644732,
+      "learning_rate": 1.0614666379701732e-06,
+      "loss": 0.2717,
+      "step": 912
+    },
+    {
+      "epoch": 1.7357414448669202,
+      "grad_norm": 0.7032912743424947,
+      "learning_rate": 1.046626284382356e-06,
+      "loss": 0.2831,
+      "step": 913
+    },
+    {
+      "epoch": 1.7376425855513307,
+      "grad_norm": 0.7302642487199171,
+      "learning_rate": 1.0318846732188737e-06,
+      "loss": 0.2764,
+      "step": 914
+    },
+    {
+      "epoch": 1.7395437262357416,
+      "grad_norm": 0.6693295937629985,
+      "learning_rate": 1.017241967057816e-06,
+      "loss": 0.2758,
+      "step": 915
+    },
+    {
+      "epoch": 1.741444866920152,
+      "grad_norm": 0.6907225510579358,
+      "learning_rate": 1.0026983273865055e-06,
+      "loss": 0.2802,
+      "step": 916
+    },
+    {
+      "epoch": 1.7433460076045626,
+      "grad_norm": 0.7088259881429065,
+      "learning_rate": 9.882539145997027e-07,
+      "loss": 0.2852,
+      "step": 917
+    },
+    {
+      "epoch": 1.7452471482889735,
+      "grad_norm": 0.6784196976575667,
+      "learning_rate": 9.739088879978409e-07,
+      "loss": 0.2866,
+      "step": 918
+    },
+    {
+      "epoch": 1.747148288973384,
+      "grad_norm": 0.6638041009329757,
+      "learning_rate": 9.59663405785277e-07,
+      "loss": 0.2664,
+      "step": 919
+    },
+    {
+      "epoch": 1.7490494296577945,
+      "grad_norm": 0.736942946808712,
+      "learning_rate": 9.455176250685338e-07,
+      "loss": 0.2846,
+      "step": 920
+    },
+    {
+      "epoch": 1.7509505703422055,
+      "grad_norm": 0.7041245018623228,
+      "learning_rate": 9.314717018545838e-07,
+      "loss": 0.2756,
+      "step": 921
+    },
+    {
+      "epoch": 1.752851711026616,
+      "grad_norm": 0.721798356350198,
+      "learning_rate": 9.17525791049112e-07,
+      "loss": 0.2694,
+      "step": 922
+    },
+    {
+      "epoch": 1.7547528517110265,
+      "grad_norm": 0.6814036007436776,
+      "learning_rate": 9.036800464548157e-07,
+      "loss": 0.275,
+      "step": 923
+    },
+    {
+      "epoch": 1.7566539923954374,
+      "grad_norm": 0.7288705451830366,
+      "learning_rate": 8.899346207697135e-07,
+      "loss": 0.2828,
+      "step": 924
+    },
+    {
+      "epoch": 1.758555133079848,
+      "grad_norm": 0.7428376062307246,
+      "learning_rate": 8.762896655854481e-07,
+      "loss": 0.2911,
+      "step": 925
+    },
+    {
+      "epoch": 1.7604562737642584,
+      "grad_norm": 0.7054928483463371,
+      "learning_rate": 8.627453313856249e-07,
+      "loss": 0.2823,
+      "step": 926
+    },
+    {
+      "epoch": 1.7623574144486693,
+      "grad_norm": 0.693606701409575,
+      "learning_rate": 8.493017675441495e-07,
+      "loss": 0.2852,
+      "step": 927
+    },
+    {
+      "epoch": 1.7642585551330798,
+      "grad_norm": 0.7187051972959767,
+      "learning_rate": 8.359591223235785e-07,
+      "loss": 0.2914,
+      "step": 928
+    },
+    {
+      "epoch": 1.7661596958174905,
+      "grad_norm": 0.7185029612498909,
+      "learning_rate": 8.227175428734868e-07,
+      "loss": 0.2904,
+      "step": 929
+    },
+    {
+      "epoch": 1.7680608365019013,
+      "grad_norm": 0.7135040106166942,
+      "learning_rate": 8.095771752288451e-07,
+      "loss": 0.2746,
+      "step": 930
+    },
+    {
+      "epoch": 1.7699619771863118,
+      "grad_norm": 0.6865698458939918,
+      "learning_rate": 7.965381643084069e-07,
+      "loss": 0.2706,
+      "step": 931
+    },
+    {
+      "epoch": 1.7718631178707225,
+      "grad_norm": 0.7120003702294833,
+      "learning_rate": 7.83600653913108e-07,
+      "loss": 0.2887,
+      "step": 932
+    },
+    {
+      "epoch": 1.7737642585551332,
+      "grad_norm": 0.6694237472931766,
+      "learning_rate": 7.707647867244927e-07,
+      "loss": 0.2798,
+      "step": 933
+    },
+    {
+      "epoch": 1.7756653992395437,
+      "grad_norm": 0.7349769076004675,
+      "learning_rate": 7.580307043031232e-07,
+      "loss": 0.29,
+      "step": 934
+    },
+    {
+      "epoch": 1.7775665399239544,
+      "grad_norm": 0.7349838772255674,
+      "learning_rate": 7.453985470870284e-07,
+      "loss": 0.2801,
+      "step": 935
+    },
+    {
+      "epoch": 1.7794676806083651,
+      "grad_norm": 0.7008528559431518,
+      "learning_rate": 7.328684543901598e-07,
+      "loss": 0.2787,
+      "step": 936
+    },
+    {
+      "epoch": 1.7813688212927756,
+      "grad_norm": 0.6524731006903562,
+      "learning_rate": 7.204405644008416e-07,
+      "loss": 0.2749,
+      "step": 937
+    },
+    {
+      "epoch": 1.7832699619771863,
+      "grad_norm": 0.6681497138680854,
+      "learning_rate": 7.081150141802518e-07,
+      "loss": 0.2829,
+      "step": 938
+    },
+    {
+      "epoch": 1.785171102661597,
+      "grad_norm": 0.6604238730498931,
+      "learning_rate": 6.958919396609231e-07,
+      "loss": 0.2775,
+      "step": 939
+    },
+    {
+      "epoch": 1.7870722433460076,
+      "grad_norm": 0.6928556213147816,
+      "learning_rate": 6.837714756452241e-07,
+      "loss": 0.2793,
+      "step": 940
+    },
+    {
+      "epoch": 1.7889733840304183,
+      "grad_norm": 0.693945957949064,
+      "learning_rate": 6.717537558038845e-07,
+      "loss": 0.2877,
+      "step": 941
+    },
+    {
+      "epoch": 1.790874524714829,
+      "grad_norm": 0.6910220671297085,
+      "learning_rate": 6.598389126745209e-07,
+      "loss": 0.2953,
+      "step": 942
+    },
+    {
+      "epoch": 1.7927756653992395,
+      "grad_norm": 0.7069711935976722,
+      "learning_rate": 6.480270776601682e-07,
+      "loss": 0.2799,
+      "step": 943
+    },
+    {
+      "epoch": 1.7946768060836502,
+      "grad_norm": 0.6897825116358226,
+      "learning_rate": 6.36318381027835e-07,
+      "loss": 0.2802,
+      "step": 944
+    },
+    {
+      "epoch": 1.796577946768061,
+      "grad_norm": 0.6724845192192414,
+      "learning_rate": 6.247129519070728e-07,
+      "loss": 0.2847,
+      "step": 945
+    },
+    {
+      "epoch": 1.7984790874524714,
+      "grad_norm": 0.718953186199371,
+      "learning_rate": 6.132109182885382e-07,
+      "loss": 0.2837,
+      "step": 946
+    },
+    {
+      "epoch": 1.8003802281368821,
+      "grad_norm": 0.6588682662087076,
+      "learning_rate": 6.018124070225928e-07,
+      "loss": 0.2754,
+      "step": 947
+    },
+    {
+      "epoch": 1.8022813688212929,
+      "grad_norm": 0.6413764809539059,
+      "learning_rate": 5.905175438178979e-07,
+      "loss": 0.2706,
+      "step": 948
+    },
+    {
+      "epoch": 1.8041825095057034,
+      "grad_norm": 0.6646032202139971,
+      "learning_rate": 5.793264532400311e-07,
+      "loss": 0.2816,
+      "step": 949
+    },
+    {
+      "epoch": 1.806083650190114,
+      "grad_norm": 0.6837645431634606,
+      "learning_rate": 5.68239258710116e-07,
+      "loss": 0.2776,
+      "step": 950
+    },
+    {
+      "epoch": 1.8079847908745248,
+      "grad_norm": 0.7003038674980157,
+      "learning_rate": 5.572560825034523e-07,
+      "loss": 0.2798,
+      "step": 951
+    },
+    {
+      "epoch": 1.8098859315589353,
+      "grad_norm": 0.6940913705850644,
+      "learning_rate": 5.463770457481732e-07,
+      "loss": 0.2912,
+      "step": 952
+    },
+    {
+      "epoch": 1.811787072243346,
+      "grad_norm": 0.6636982084604686,
+      "learning_rate": 5.35602268423906e-07,
+      "loss": 0.2833,
+      "step": 953
+    },
+    {
+      "epoch": 1.8136882129277567,
+      "grad_norm": 0.7207943181233436,
+      "learning_rate": 5.249318693604577e-07,
+      "loss": 0.295,
+      "step": 954
+    },
+    {
+      "epoch": 1.8155893536121672,
+      "grad_norm": 0.7179708419213036,
+      "learning_rate": 5.143659662364931e-07,
+      "loss": 0.3036,
+      "step": 955
+    },
+    {
+      "epoch": 1.817490494296578,
+      "grad_norm": 0.7060364197755066,
+      "learning_rate": 5.039046755782417e-07,
+      "loss": 0.2786,
+      "step": 956
+    },
+    {
+      "epoch": 1.8193916349809887,
+      "grad_norm": 0.6608672505948111,
+      "learning_rate": 4.935481127582131e-07,
+      "loss": 0.2697,
+      "step": 957
+    },
+    {
+      "epoch": 1.8212927756653992,
+      "grad_norm": 0.7472022994533832,
+      "learning_rate": 4.83296391993926e-07,
+      "loss": 0.299,
+      "step": 958
+    },
+    {
+      "epoch": 1.8231939163498099,
+      "grad_norm": 0.6720946161301063,
+      "learning_rate": 4.7314962634664616e-07,
+      "loss": 0.2613,
+      "step": 959
+    },
+    {
+      "epoch": 1.8250950570342206,
+      "grad_norm": 0.6542428189015632,
+      "learning_rate": 4.631079277201389e-07,
+      "loss": 0.2757,
+      "step": 960
+    },
+    {
+      "epoch": 1.826996197718631,
+      "grad_norm": 0.6875735554822852,
+      "learning_rate": 4.5317140685943726e-07,
+      "loss": 0.2869,
+      "step": 961
+    },
+    {
+      "epoch": 1.8288973384030418,
+      "grad_norm": 0.6766473579419423,
+      "learning_rate": 4.433401733496201e-07,
+      "loss": 0.2796,
+      "step": 962
+    },
+    {
+      "epoch": 1.8307984790874525,
+      "grad_norm": 0.7127432773866573,
+      "learning_rate": 4.3361433561460274e-07,
+      "loss": 0.2879,
+      "step": 963
+    },
+    {
+      "epoch": 1.832699619771863,
+      "grad_norm": 0.6763087251166204,
+      "learning_rate": 4.2399400091594154e-07,
+      "loss": 0.277,
+      "step": 964
+    },
+    {
+      "epoch": 1.8346007604562737,
+      "grad_norm": 0.7161252957569159,
+      "learning_rate": 4.14479275351648e-07,
+      "loss": 0.2699,
+      "step": 965
+    },
+    {
+      "epoch": 1.8365019011406845,
+      "grad_norm": 0.699620414616677,
+      "learning_rate": 4.0507026385502747e-07,
+      "loss": 0.2759,
+      "step": 966
+    },
+    {
+      "epoch": 1.838403041825095,
+      "grad_norm": 0.6675966554521812,
+      "learning_rate": 3.9576707019350903e-07,
+      "loss": 0.2924,
+      "step": 967
+    },
+    {
+      "epoch": 1.8403041825095057,
+      "grad_norm": 0.7516092483408074,
+      "learning_rate": 3.865697969675164e-07,
+      "loss": 0.2909,
+      "step": 968
+    },
+    {
+      "epoch": 1.8422053231939164,
+      "grad_norm": 0.6977971985944234,
+      "learning_rate": 3.7747854560931996e-07,
+      "loss": 0.289,
+      "step": 969
+    },
+    {
+      "epoch": 1.8441064638783269,
+      "grad_norm": 0.6803215961823417,
+      "learning_rate": 3.684934163819309e-07,
+      "loss": 0.2784,
+      "step": 970
+    },
+    {
+      "epoch": 1.8460076045627376,
+      "grad_norm": 0.6476536952915057,
+      "learning_rate": 3.596145083779912e-07,
+      "loss": 0.2679,
+      "step": 971
+    },
+    {
+      "epoch": 1.8479087452471483,
+      "grad_norm": 0.6546315321057461,
+      "learning_rate": 3.508419195186774e-07,
+      "loss": 0.2628,
+      "step": 972
+    },
+    {
+      "epoch": 1.8498098859315588,
+      "grad_norm": 0.6995610900006602,
+      "learning_rate": 3.421757465526243e-07,
+      "loss": 0.2794,
+      "step": 973
+    },
+    {
+      "epoch": 1.8517110266159695,
+      "grad_norm": 0.7229612955682959,
+      "learning_rate": 3.33616085054862e-07,
+      "loss": 0.288,
+      "step": 974
+    },
+    {
+      "epoch": 1.8536121673003803,
+      "grad_norm": 0.6919211156552034,
+      "learning_rate": 3.2516302942574794e-07,
+      "loss": 0.2776,
+      "step": 975
+    },
+    {
+      "epoch": 1.8555133079847907,
+      "grad_norm": 0.6974752381288766,
+      "learning_rate": 3.1681667288994353e-07,
+      "loss": 0.2882,
+      "step": 976
+    },
+    {
+      "epoch": 1.8574144486692015,
+      "grad_norm": 0.6759158512161695,
+      "learning_rate": 3.0857710749537585e-07,
+      "loss": 0.2715,
+      "step": 977
+    },
+    {
+      "epoch": 1.8593155893536122,
+      "grad_norm": 0.6493781419868506,
+      "learning_rate": 3.0044442411222066e-07,
+      "loss": 0.2656,
+      "step": 978
+    },
+    {
+      "epoch": 1.8612167300380227,
+      "grad_norm": 0.6684872794348289,
+      "learning_rate": 2.9241871243190555e-07,
+      "loss": 0.2753,
+      "step": 979
+    },
+    {
+      "epoch": 1.8631178707224336,
+      "grad_norm": 0.6828717589595341,
+      "learning_rate": 2.845000609661208e-07,
+      "loss": 0.2728,
+      "step": 980
+    },
+    {
+      "epoch": 1.8650190114068441,
+      "grad_norm": 0.6643948753801928,
+      "learning_rate": 2.7668855704583997e-07,
+      "loss": 0.2736,
+      "step": 981
+    },
+    {
+      "epoch": 1.8669201520912546,
+      "grad_norm": 0.6892381022323134,
+      "learning_rate": 2.689842868203563e-07,
+      "loss": 0.2691,
+      "step": 982
+    },
+    {
+      "epoch": 1.8688212927756656,
+      "grad_norm": 0.6836596852145582,
+      "learning_rate": 2.6138733525633896e-07,
+      "loss": 0.2798,
+      "step": 983
+    },
+    {
+      "epoch": 1.870722433460076,
+      "grad_norm": 0.7066162757108758,
+      "learning_rate": 2.5389778613688744e-07,
+      "loss": 0.285,
+      "step": 984
+    },
+    {
+      "epoch": 1.8726235741444865,
+      "grad_norm": 0.6514567696446193,
+      "learning_rate": 2.46515722060614e-07,
+      "loss": 0.2721,
+      "step": 985
+    },
+    {
+      "epoch": 1.8745247148288975,
+      "grad_norm": 0.6588460545574634,
+      "learning_rate": 2.392412244407294e-07,
+      "loss": 0.2585,
+      "step": 986
+    },
+    {
+      "epoch": 1.876425855513308,
+      "grad_norm": 0.6748097006697391,
+      "learning_rate": 2.3207437350414418e-07,
+      "loss": 0.2715,
+      "step": 987
+    },
+    {
+      "epoch": 1.8783269961977185,
+      "grad_norm": 0.6901613625037928,
+      "learning_rate": 2.2501524829059208e-07,
+      "loss": 0.2861,
+      "step": 988
+    },
+    {
+      "epoch": 1.8802281368821294,
+      "grad_norm": 0.7005143672667519,
+      "learning_rate": 2.180639266517448e-07,
+      "loss": 0.2826,
+      "step": 989
+    },
+    {
+      "epoch": 1.88212927756654,
+      "grad_norm": 0.7329754216042629,
+      "learning_rate": 2.1122048525036409e-07,
+      "loss": 0.284,
+      "step": 990
+    },
+    {
+      "epoch": 1.8840304182509504,
+      "grad_norm": 0.6940212345484321,
+      "learning_rate": 2.0448499955945223e-07,
+      "loss": 0.2814,
+      "step": 991
+    },
+    {
+      "epoch": 1.8859315589353614,
+      "grad_norm": 0.6950801764630833,
+      "learning_rate": 1.9785754386142164e-07,
+      "loss": 0.2832,
+      "step": 992
+    },
+    {
+      "epoch": 1.8878326996197718,
+      "grad_norm": 0.6680088695753502,
+      "learning_rate": 1.9133819124727003e-07,
+      "loss": 0.2759,
+      "step": 993
+    },
+    {
+      "epoch": 1.8897338403041823,
+      "grad_norm": 0.7179315368227291,
+      "learning_rate": 1.8492701361578326e-07,
+      "loss": 0.2956,
+      "step": 994
+    },
+    {
+      "epoch": 1.8916349809885933,
+      "grad_norm": 0.6982208698112028,
+      "learning_rate": 1.7862408167273472e-07,
+      "loss": 0.2967,
+      "step": 995
+    },
+    {
+      "epoch": 1.8935361216730038,
+      "grad_norm": 0.6857912614172924,
+      "learning_rate": 1.724294649301095e-07,
+      "loss": 0.2813,
+      "step": 996
+    },
+    {
+      "epoch": 1.8954372623574145,
+      "grad_norm": 0.7084909531916186,
+      "learning_rate": 1.6634323170533928e-07,
+      "loss": 0.2779,
+      "step": 997
+    },
+    {
+      "epoch": 1.8973384030418252,
+      "grad_norm": 0.6818216477765531,
+      "learning_rate": 1.6036544912054087e-07,
+      "loss": 0.2815,
+      "step": 998
+    },
+    {
+      "epoch": 1.8992395437262357,
+      "grad_norm": 0.7044121437532876,
+      "learning_rate": 1.544961831017855e-07,
+      "loss": 0.2899,
+      "step": 999
+    },
+    {
+      "epoch": 1.9011406844106464,
+      "grad_norm": 0.7056708146838936,
+      "learning_rate": 1.487354983783673e-07,
+      "loss": 0.2767,
+      "step": 1000
+    },
+    {
+      "epoch": 1.9030418250950571,
+      "grad_norm": 0.651100892250291,
+      "learning_rate": 1.430834584820895e-07,
+      "loss": 0.2735,
+      "step": 1001
+    },
+    {
+      "epoch": 1.9049429657794676,
+      "grad_norm": 0.6892063857611048,
+      "learning_rate": 1.375401257465625e-07,
+      "loss": 0.2816,
+      "step": 1002
+    },
+    {
+      "epoch": 1.9068441064638784,
+      "grad_norm": 0.6728345722990633,
+      "learning_rate": 1.3210556130652031e-07,
+      "loss": 0.2852,
+      "step": 1003
+    },
+    {
+      "epoch": 1.908745247148289,
+      "grad_norm": 0.6806458319606459,
+      "learning_rate": 1.2677982509714415e-07,
+      "loss": 0.2689,
+      "step": 1004
+    },
+    {
+      "epoch": 1.9106463878326996,
+      "grad_norm": 0.7039358568921888,
+      "learning_rate": 1.2156297585339872e-07,
+      "loss": 0.2751,
+      "step": 1005
+    },
+    {
+      "epoch": 1.9125475285171103,
+      "grad_norm": 0.6839482694248901,
+      "learning_rate": 1.1645507110938925e-07,
+      "loss": 0.2848,
+      "step": 1006
+    },
+    {
+      "epoch": 1.914448669201521,
+      "grad_norm": 0.683476143247811,
+      "learning_rate": 1.1145616719772545e-07,
+      "loss": 0.2833,
+      "step": 1007
+    },
+    {
+      "epoch": 1.9163498098859315,
+      "grad_norm": 0.7020211537992438,
+      "learning_rate": 1.0656631924889749e-07,
+      "loss": 0.2879,
+      "step": 1008
+    },
+    {
+      "epoch": 1.9182509505703422,
+      "grad_norm": 0.7062050122426703,
+      "learning_rate": 1.0178558119067316e-07,
+      "loss": 0.2899,
+      "step": 1009
+    },
+    {
+      "epoch": 1.920152091254753,
+      "grad_norm": 0.6827874427212586,
+      "learning_rate": 9.711400574749507e-08,
+      "loss": 0.27,
+      "step": 1010
+    },
+    {
+      "epoch": 1.9220532319391634,
+      "grad_norm": 0.6609167157619602,
+      "learning_rate": 9.255164443990994e-08,
+      "loss": 0.2715,
+      "step": 1011
+    },
+    {
+      "epoch": 1.9239543726235742,
+      "grad_norm": 0.7086696451225294,
+      "learning_rate": 8.809854758399017e-08,
+      "loss": 0.2716,
+      "step": 1012
+    },
+    {
+      "epoch": 1.9258555133079849,
+      "grad_norm": 0.6568539261635775,
+      "learning_rate": 8.375476429078543e-08,
+      "loss": 0.2745,
+      "step": 1013
+    },
+    {
+      "epoch": 1.9277566539923954,
+      "grad_norm": 0.695869205903921,
+      "learning_rate": 7.952034246577977e-08,
+      "loss": 0.2792,
+      "step": 1014
+    },
+    {
+      "epoch": 1.929657794676806,
+      "grad_norm": 0.6729183920390936,
+      "learning_rate": 7.539532880836087e-08,
+      "loss": 0.2788,
+      "step": 1015
+    },
+    {
+      "epoch": 1.9315589353612168,
+      "grad_norm": 0.64634693308591,
+      "learning_rate": 7.137976881130826e-08,
+      "loss": 0.2651,
+      "step": 1016
+    },
+    {
+      "epoch": 1.9334600760456273,
+      "grad_norm": 0.6988011124970799,
+      "learning_rate": 6.747370676028819e-08,
+      "loss": 0.2698,
+      "step": 1017
+    },
+    {
+      "epoch": 1.935361216730038,
+      "grad_norm": 0.7088301091748413,
+      "learning_rate": 6.367718573336845e-08,
+      "loss": 0.2718,
+      "step": 1018
+    },
+    {
+      "epoch": 1.9372623574144487,
+      "grad_norm": 0.7003990531376776,
+      "learning_rate": 5.999024760054095e-08,
+      "loss": 0.2738,
+      "step": 1019
+    },
+    {
+      "epoch": 1.9391634980988592,
+      "grad_norm": 0.6744196150009467,
+      "learning_rate": 5.641293302326323e-08,
+      "loss": 0.2739,
+      "step": 1020
+    },
+    {
+      "epoch": 1.94106463878327,
+      "grad_norm": 0.6764497081474102,
+      "learning_rate": 5.2945281454003236e-08,
+      "loss": 0.2763,
+      "step": 1021
+    },
+    {
+      "epoch": 1.9429657794676807,
+      "grad_norm": 0.6837817988177924,
+      "learning_rate": 4.958733113581415e-08,
+      "loss": 0.288,
+      "step": 1022
+    },
+    {
+      "epoch": 1.9448669201520912,
+      "grad_norm": 0.7147933752410155,
+      "learning_rate": 4.6339119101902475e-08,
+      "loss": 0.2772,
+      "step": 1023
+    },
+    {
+      "epoch": 1.946768060836502,
+      "grad_norm": 0.6834369534143053,
+      "learning_rate": 4.320068117522835e-08,
+      "loss": 0.288,
+      "step": 1024
+    },
+    {
+      "epoch": 1.9486692015209126,
+      "grad_norm": 0.7026389571217904,
+      "learning_rate": 4.0172051968101474e-08,
+      "loss": 0.2837,
+      "step": 1025
+    },
+    {
+      "epoch": 1.950570342205323,
+      "grad_norm": 0.6539034113068007,
+      "learning_rate": 3.7253264881809137e-08,
+      "loss": 0.2712,
+      "step": 1026
+    },
+    {
+      "epoch": 1.9524714828897338,
+      "grad_norm": 0.6787032696035695,
+      "learning_rate": 3.4444352106242086e-08,
+      "loss": 0.2729,
+      "step": 1027
+    },
+    {
+      "epoch": 1.9543726235741445,
+      "grad_norm": 0.6575865641381481,
+      "learning_rate": 3.174534461953593e-08,
+      "loss": 0.2735,
+      "step": 1028
+    },
+    {
+      "epoch": 1.956273764258555,
+      "grad_norm": 0.687189018922996,
+      "learning_rate": 2.915627218774142e-08,
+      "loss": 0.2794,
+      "step": 1029
+    },
+    {
+      "epoch": 1.9581749049429658,
+      "grad_norm": 0.7046070383040507,
+      "learning_rate": 2.667716336448356e-08,
+      "loss": 0.2834,
+      "step": 1030
+    },
+    {
+      "epoch": 1.9600760456273765,
+      "grad_norm": 0.6455522271080228,
+      "learning_rate": 2.430804549065302e-08,
+      "loss": 0.278,
+      "step": 1031
+    },
+    {
+      "epoch": 1.961977186311787,
+      "grad_norm": 0.6883479470746532,
+      "learning_rate": 2.2048944694104123e-08,
+      "loss": 0.2802,
+      "step": 1032
+    },
+    {
+      "epoch": 1.9638783269961977,
+      "grad_norm": 0.7096636250910227,
+      "learning_rate": 1.989988588936509e-08,
+      "loss": 0.2894,
+      "step": 1033
+    },
+    {
+      "epoch": 1.9657794676806084,
+      "grad_norm": 0.6847188669963606,
+      "learning_rate": 1.7860892777367133e-08,
+      "loss": 0.2809,
+      "step": 1034
+    },
+    {
+      "epoch": 1.967680608365019,
+      "grad_norm": 0.6643792512868152,
+      "learning_rate": 1.5931987845176912e-08,
+      "loss": 0.2616,
+      "step": 1035
+    },
+    {
+      "epoch": 1.9695817490494296,
+      "grad_norm": 0.6772610868459655,
+      "learning_rate": 1.411319236575337e-08,
+      "loss": 0.2732,
+      "step": 1036
+    },
+    {
+      "epoch": 1.9714828897338403,
+      "grad_norm": 0.7417987880853206,
+      "learning_rate": 1.2404526397711281e-08,
+      "loss": 0.2942,
+      "step": 1037
+    },
+    {
+      "epoch": 1.9733840304182508,
+      "grad_norm": 0.671428419018434,
+      "learning_rate": 1.0806008785100297e-08,
+      "loss": 0.2666,
+      "step": 1038
+    },
+    {
+      "epoch": 1.9752851711026616,
+      "grad_norm": 0.7672035342053988,
+      "learning_rate": 9.317657157197347e-09,
+      "loss": 0.2903,
+      "step": 1039
+    },
+    {
+      "epoch": 1.9771863117870723,
+      "grad_norm": 0.7346417240567528,
+      "learning_rate": 7.93948792831234e-09,
+      "loss": 0.2826,
+      "step": 1040
+    },
+    {
+      "epoch": 1.9790874524714828,
+      "grad_norm": 0.6711010752059788,
+      "learning_rate": 6.671516297606095e-09,
+      "loss": 0.2859,
+      "step": 1041
+    },
+    {
+      "epoch": 1.9809885931558935,
+      "grad_norm": 0.7017537465578773,
+      "learning_rate": 5.513756248924917e-09,
+      "loss": 0.2844,
+      "step": 1042
+    },
+    {
+      "epoch": 1.9828897338403042,
+      "grad_norm": 0.6803971440517318,
+      "learning_rate": 4.466220550641831e-09,
+      "loss": 0.272,
+      "step": 1043
+    },
+    {
+      "epoch": 1.9847908745247147,
+      "grad_norm": 0.6974813168149476,
+      "learning_rate": 3.528920755523357e-09,
+      "loss": 0.2873,
+      "step": 1044
+    },
+    {
+      "epoch": 1.9866920152091256,
+      "grad_norm": 0.6973336408698366,
+      "learning_rate": 2.701867200592956e-09,
+      "loss": 0.2784,
+      "step": 1045
+    },
+    {
+      "epoch": 1.9885931558935361,
+      "grad_norm": 0.6502566439347921,
+      "learning_rate": 1.9850690070266633e-09,
+      "loss": 0.2753,
+      "step": 1046
+    },
+    {
+      "epoch": 1.9904942965779466,
+      "grad_norm": 0.6914675190902179,
+      "learning_rate": 1.378534080042071e-09,
+      "loss": 0.2803,
+      "step": 1047
+    },
+    {
+      "epoch": 1.9923954372623576,
+      "grad_norm": 0.6866231706549291,
+      "learning_rate": 8.822691088195001e-10,
+      "loss": 0.2712,
+      "step": 1048
+    },
+    {
+      "epoch": 1.994296577946768,
+      "grad_norm": 0.67703437589136,
+      "learning_rate": 4.962795664265052e-10,
+      "loss": 0.2742,
+      "step": 1049
+    },
+    {
+      "epoch": 1.9961977186311786,
+      "grad_norm": 0.6690225557997433,
+      "learning_rate": 2.2056970975459223e-10,
+      "loss": 0.2709,
+      "step": 1050
+    },
+    {
+      "epoch": 1.9980988593155895,
+      "grad_norm": 0.687979234723248,
+      "learning_rate": 5.514257947369928e-11,
+      "loss": 0.3009,
+      "step": 1051
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6524667243655449,
+      "learning_rate": 0.0,
+      "loss": 0.2843,
+      "step": 1052
+    },
+    {
+      "epoch": 2.0,
+      "step": 1052,
+      "total_flos": 103569111859200.0,
+      "train_loss": 0.14411184585479728,
+      "train_runtime": 1873.7697,
+      "train_samples_per_second": 71.864,
+      "train_steps_per_second": 0.561
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1052,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "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": 103569111859200.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}