diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,14073 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 2001,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0014992503748125937,
+      "grad_norm": 0.06610345176976985,
+      "learning_rate": 9.950248756218907e-07,
+      "loss": 0.0231,
+      "step": 1
+    },
+    {
+      "epoch": 0.0029985007496251873,
+      "grad_norm": 0.10229378949313654,
+      "learning_rate": 1.9900497512437813e-06,
+      "loss": 0.0302,
+      "step": 2
+    },
+    {
+      "epoch": 0.004497751124437781,
+      "grad_norm": 0.09308353085324574,
+      "learning_rate": 2.9850746268656716e-06,
+      "loss": 0.0501,
+      "step": 3
+    },
+    {
+      "epoch": 0.005997001499250375,
+      "grad_norm": 0.058628771398694095,
+      "learning_rate": 3.980099502487563e-06,
+      "loss": 0.0243,
+      "step": 4
+    },
+    {
+      "epoch": 0.0074962518740629685,
+      "grad_norm": 0.0910650644048991,
+      "learning_rate": 4.975124378109453e-06,
+      "loss": 0.0383,
+      "step": 5
+    },
+    {
+      "epoch": 0.008995502248875561,
+      "grad_norm": 0.12316753445017627,
+      "learning_rate": 5.970149253731343e-06,
+      "loss": 0.0436,
+      "step": 6
+    },
+    {
+      "epoch": 0.010494752623688156,
+      "grad_norm": 0.05542672781465358,
+      "learning_rate": 6.965174129353234e-06,
+      "loss": 0.094,
+      "step": 7
+    },
+    {
+      "epoch": 0.01199400299850075,
+      "grad_norm": 0.03735550944126593,
+      "learning_rate": 7.960199004975125e-06,
+      "loss": 0.0241,
+      "step": 8
+    },
+    {
+      "epoch": 0.013493253373313344,
+      "grad_norm": 0.055574007146215375,
+      "learning_rate": 8.955223880597016e-06,
+      "loss": 0.0199,
+      "step": 9
+    },
+    {
+      "epoch": 0.014992503748125937,
+      "grad_norm": 0.08334031247034454,
+      "learning_rate": 9.950248756218906e-06,
+      "loss": 0.0309,
+      "step": 10
+    },
+    {
+      "epoch": 0.01649175412293853,
+      "grad_norm": 0.0615936672450276,
+      "learning_rate": 1.0945273631840796e-05,
+      "loss": 0.0135,
+      "step": 11
+    },
+    {
+      "epoch": 0.017991004497751123,
+      "grad_norm": 0.04494010236736683,
+      "learning_rate": 1.1940298507462686e-05,
+      "loss": 0.0232,
+      "step": 12
+    },
+    {
+      "epoch": 0.019490254872563718,
+      "grad_norm": 0.05796888012935532,
+      "learning_rate": 1.2935323383084577e-05,
+      "loss": 0.0349,
+      "step": 13
+    },
+    {
+      "epoch": 0.020989505247376312,
+      "grad_norm": 0.06339185703613222,
+      "learning_rate": 1.3930348258706468e-05,
+      "loss": 0.0309,
+      "step": 14
+    },
+    {
+      "epoch": 0.022488755622188907,
+      "grad_norm": 0.0592636851918813,
+      "learning_rate": 1.4925373134328357e-05,
+      "loss": 0.0256,
+      "step": 15
+    },
+    {
+      "epoch": 0.0239880059970015,
+      "grad_norm": 0.060507862352465004,
+      "learning_rate": 1.592039800995025e-05,
+      "loss": 0.0253,
+      "step": 16
+    },
+    {
+      "epoch": 0.025487256371814093,
+      "grad_norm": 0.06232454162862092,
+      "learning_rate": 1.691542288557214e-05,
+      "loss": 0.0341,
+      "step": 17
+    },
+    {
+      "epoch": 0.026986506746626688,
+      "grad_norm": 0.08813118060803964,
+      "learning_rate": 1.791044776119403e-05,
+      "loss": 0.0191,
+      "step": 18
+    },
+    {
+      "epoch": 0.02848575712143928,
+      "grad_norm": 0.09703603861107045,
+      "learning_rate": 1.890547263681592e-05,
+      "loss": 0.0334,
+      "step": 19
+    },
+    {
+      "epoch": 0.029985007496251874,
+      "grad_norm": 0.12217053930995624,
+      "learning_rate": 1.990049751243781e-05,
+      "loss": 0.025,
+      "step": 20
+    },
+    {
+      "epoch": 0.031484257871064465,
+      "grad_norm": 0.08708398557831935,
+      "learning_rate": 2.0895522388059702e-05,
+      "loss": 0.0385,
+      "step": 21
+    },
+    {
+      "epoch": 0.03298350824587706,
+      "grad_norm": 0.05748291151872297,
+      "learning_rate": 2.1890547263681592e-05,
+      "loss": 0.0294,
+      "step": 22
+    },
+    {
+      "epoch": 0.034482758620689655,
+      "grad_norm": 0.05267747048254378,
+      "learning_rate": 2.2885572139303486e-05,
+      "loss": 0.0236,
+      "step": 23
+    },
+    {
+      "epoch": 0.035982008995502246,
+      "grad_norm": 0.20462262185187904,
+      "learning_rate": 2.3880597014925373e-05,
+      "loss": 0.0298,
+      "step": 24
+    },
+    {
+      "epoch": 0.037481259370314844,
+      "grad_norm": 0.04730801516528991,
+      "learning_rate": 2.4875621890547266e-05,
+      "loss": 0.0231,
+      "step": 25
+    },
+    {
+      "epoch": 0.038980509745127435,
+      "grad_norm": 0.07129877545543531,
+      "learning_rate": 2.5870646766169153e-05,
+      "loss": 0.0409,
+      "step": 26
+    },
+    {
+      "epoch": 0.04047976011994003,
+      "grad_norm": 0.04787844989660117,
+      "learning_rate": 2.6865671641791047e-05,
+      "loss": 0.0244,
+      "step": 27
+    },
+    {
+      "epoch": 0.041979010494752625,
+      "grad_norm": 0.08080289031605667,
+      "learning_rate": 2.7860696517412937e-05,
+      "loss": 0.0373,
+      "step": 28
+    },
+    {
+      "epoch": 0.043478260869565216,
+      "grad_norm": 0.03127750279466034,
+      "learning_rate": 2.885572139303483e-05,
+      "loss": 0.0093,
+      "step": 29
+    },
+    {
+      "epoch": 0.044977511244377814,
+      "grad_norm": 0.0672839069697597,
+      "learning_rate": 2.9850746268656714e-05,
+      "loss": 0.0565,
+      "step": 30
+    },
+    {
+      "epoch": 0.046476761619190406,
+      "grad_norm": 0.05579809699762286,
+      "learning_rate": 3.084577114427861e-05,
+      "loss": 0.0262,
+      "step": 31
+    },
+    {
+      "epoch": 0.047976011994003,
+      "grad_norm": 0.06764953701820348,
+      "learning_rate": 3.18407960199005e-05,
+      "loss": 0.0355,
+      "step": 32
+    },
+    {
+      "epoch": 0.049475262368815595,
+      "grad_norm": 0.07766895860290135,
+      "learning_rate": 3.283582089552239e-05,
+      "loss": 0.0176,
+      "step": 33
+    },
+    {
+      "epoch": 0.050974512743628186,
+      "grad_norm": 0.0877687067355835,
+      "learning_rate": 3.383084577114428e-05,
+      "loss": 0.0228,
+      "step": 34
+    },
+    {
+      "epoch": 0.05247376311844078,
+      "grad_norm": 0.06901704426078333,
+      "learning_rate": 3.4825870646766175e-05,
+      "loss": 0.0301,
+      "step": 35
+    },
+    {
+      "epoch": 0.053973013493253376,
+      "grad_norm": 0.050365449371918836,
+      "learning_rate": 3.582089552238806e-05,
+      "loss": 0.025,
+      "step": 36
+    },
+    {
+      "epoch": 0.05547226386806597,
+      "grad_norm": 0.10183750100826967,
+      "learning_rate": 3.681592039800995e-05,
+      "loss": 0.0359,
+      "step": 37
+    },
+    {
+      "epoch": 0.05697151424287856,
+      "grad_norm": 0.05576688463550469,
+      "learning_rate": 3.781094527363184e-05,
+      "loss": 0.0263,
+      "step": 38
+    },
+    {
+      "epoch": 0.05847076461769116,
+      "grad_norm": 0.0809406763945191,
+      "learning_rate": 3.8805970149253736e-05,
+      "loss": 0.0446,
+      "step": 39
+    },
+    {
+      "epoch": 0.05997001499250375,
+      "grad_norm": 0.03977873137524195,
+      "learning_rate": 3.980099502487562e-05,
+      "loss": 0.0155,
+      "step": 40
+    },
+    {
+      "epoch": 0.06146926536731634,
+      "grad_norm": 0.04675680284402419,
+      "learning_rate": 4.079601990049751e-05,
+      "loss": 0.0213,
+      "step": 41
+    },
+    {
+      "epoch": 0.06296851574212893,
+      "grad_norm": 0.13250207008437154,
+      "learning_rate": 4.1791044776119404e-05,
+      "loss": 0.0225,
+      "step": 42
+    },
+    {
+      "epoch": 0.06446776611694154,
+      "grad_norm": 0.06061033398983015,
+      "learning_rate": 4.27860696517413e-05,
+      "loss": 0.0262,
+      "step": 43
+    },
+    {
+      "epoch": 0.06596701649175413,
+      "grad_norm": 0.08470633670479133,
+      "learning_rate": 4.3781094527363184e-05,
+      "loss": 0.0403,
+      "step": 44
+    },
+    {
+      "epoch": 0.06746626686656672,
+      "grad_norm": 0.06220515053666017,
+      "learning_rate": 4.477611940298508e-05,
+      "loss": 0.0256,
+      "step": 45
+    },
+    {
+      "epoch": 0.06896551724137931,
+      "grad_norm": 0.0421105165497031,
+      "learning_rate": 4.577114427860697e-05,
+      "loss": 0.0233,
+      "step": 46
+    },
+    {
+      "epoch": 0.0704647676161919,
+      "grad_norm": 0.056273635232726445,
+      "learning_rate": 4.676616915422886e-05,
+      "loss": 0.0271,
+      "step": 47
+    },
+    {
+      "epoch": 0.07196401799100449,
+      "grad_norm": 0.04135843131632209,
+      "learning_rate": 4.7761194029850745e-05,
+      "loss": 0.0192,
+      "step": 48
+    },
+    {
+      "epoch": 0.0734632683658171,
+      "grad_norm": 0.07545305238459443,
+      "learning_rate": 4.875621890547264e-05,
+      "loss": 0.0354,
+      "step": 49
+    },
+    {
+      "epoch": 0.07496251874062969,
+      "grad_norm": 0.03485747229470634,
+      "learning_rate": 4.975124378109453e-05,
+      "loss": 0.0194,
+      "step": 50
+    },
+    {
+      "epoch": 0.07646176911544228,
+      "grad_norm": 0.07132696830734893,
+      "learning_rate": 5.074626865671642e-05,
+      "loss": 0.0284,
+      "step": 51
+    },
+    {
+      "epoch": 0.07796101949025487,
+      "grad_norm": 0.0476562529060188,
+      "learning_rate": 5.1741293532338306e-05,
+      "loss": 0.0195,
+      "step": 52
+    },
+    {
+      "epoch": 0.07946026986506746,
+      "grad_norm": 0.04135610651103468,
+      "learning_rate": 5.2736318407960206e-05,
+      "loss": 0.0209,
+      "step": 53
+    },
+    {
+      "epoch": 0.08095952023988005,
+      "grad_norm": 0.07137198475738459,
+      "learning_rate": 5.373134328358209e-05,
+      "loss": 0.0235,
+      "step": 54
+    },
+    {
+      "epoch": 0.08245877061469266,
+      "grad_norm": 0.08351208018844225,
+      "learning_rate": 5.472636815920398e-05,
+      "loss": 0.0456,
+      "step": 55
+    },
+    {
+      "epoch": 0.08395802098950525,
+      "grad_norm": 0.0855628538810367,
+      "learning_rate": 5.5721393034825874e-05,
+      "loss": 0.0383,
+      "step": 56
+    },
+    {
+      "epoch": 0.08545727136431784,
+      "grad_norm": 0.07872354423296858,
+      "learning_rate": 5.671641791044776e-05,
+      "loss": 0.0346,
+      "step": 57
+    },
+    {
+      "epoch": 0.08695652173913043,
+      "grad_norm": 0.08549249098509387,
+      "learning_rate": 5.771144278606966e-05,
+      "loss": 0.0436,
+      "step": 58
+    },
+    {
+      "epoch": 0.08845577211394302,
+      "grad_norm": 0.08010714092690198,
+      "learning_rate": 5.870646766169154e-05,
+      "loss": 0.0278,
+      "step": 59
+    },
+    {
+      "epoch": 0.08995502248875563,
+      "grad_norm": 0.11061139734613264,
+      "learning_rate": 5.970149253731343e-05,
+      "loss": 0.0303,
+      "step": 60
+    },
+    {
+      "epoch": 0.09145427286356822,
+      "grad_norm": 0.08851009372866617,
+      "learning_rate": 6.069651741293533e-05,
+      "loss": 0.0273,
+      "step": 61
+    },
+    {
+      "epoch": 0.09295352323838081,
+      "grad_norm": 0.06431070626891758,
+      "learning_rate": 6.169154228855722e-05,
+      "loss": 0.0476,
+      "step": 62
+    },
+    {
+      "epoch": 0.0944527736131934,
+      "grad_norm": 0.04207690654899365,
+      "learning_rate": 6.268656716417911e-05,
+      "loss": 0.0187,
+      "step": 63
+    },
+    {
+      "epoch": 0.095952023988006,
+      "grad_norm": 0.0629659489389091,
+      "learning_rate": 6.3681592039801e-05,
+      "loss": 0.0294,
+      "step": 64
+    },
+    {
+      "epoch": 0.09745127436281859,
+      "grad_norm": 0.04739055753217489,
+      "learning_rate": 6.46766169154229e-05,
+      "loss": 0.0117,
+      "step": 65
+    },
+    {
+      "epoch": 0.09895052473763119,
+      "grad_norm": 0.062751616457929,
+      "learning_rate": 6.567164179104478e-05,
+      "loss": 0.0321,
+      "step": 66
+    },
+    {
+      "epoch": 0.10044977511244378,
+      "grad_norm": 0.12789865648617366,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 0.0285,
+      "step": 67
+    },
+    {
+      "epoch": 0.10194902548725637,
+      "grad_norm": 0.09087938435304468,
+      "learning_rate": 6.766169154228856e-05,
+      "loss": 0.0293,
+      "step": 68
+    },
+    {
+      "epoch": 0.10344827586206896,
+      "grad_norm": 0.07104633745977648,
+      "learning_rate": 6.865671641791044e-05,
+      "loss": 0.0248,
+      "step": 69
+    },
+    {
+      "epoch": 0.10494752623688156,
+      "grad_norm": 0.04104044865025192,
+      "learning_rate": 6.965174129353235e-05,
+      "loss": 0.0117,
+      "step": 70
+    },
+    {
+      "epoch": 0.10644677661169415,
+      "grad_norm": 0.06739350707830459,
+      "learning_rate": 7.064676616915423e-05,
+      "loss": 0.0259,
+      "step": 71
+    },
+    {
+      "epoch": 0.10794602698650675,
+      "grad_norm": 0.12081546856137051,
+      "learning_rate": 7.164179104477612e-05,
+      "loss": 0.046,
+      "step": 72
+    },
+    {
+      "epoch": 0.10944527736131934,
+      "grad_norm": 0.09650929672212062,
+      "learning_rate": 7.263681592039802e-05,
+      "loss": 0.0196,
+      "step": 73
+    },
+    {
+      "epoch": 0.11094452773613193,
+      "grad_norm": 0.03209119308258964,
+      "learning_rate": 7.36318407960199e-05,
+      "loss": 0.0096,
+      "step": 74
+    },
+    {
+      "epoch": 0.11244377811094453,
+      "grad_norm": 0.12687784195284804,
+      "learning_rate": 7.46268656716418e-05,
+      "loss": 0.0416,
+      "step": 75
+    },
+    {
+      "epoch": 0.11394302848575712,
+      "grad_norm": 0.06383164911311603,
+      "learning_rate": 7.562189054726369e-05,
+      "loss": 0.0122,
+      "step": 76
+    },
+    {
+      "epoch": 0.11544227886056972,
+      "grad_norm": 0.07876299012336277,
+      "learning_rate": 7.661691542288557e-05,
+      "loss": 0.0299,
+      "step": 77
+    },
+    {
+      "epoch": 0.11694152923538231,
+      "grad_norm": 0.07838431599591475,
+      "learning_rate": 7.761194029850747e-05,
+      "loss": 0.0272,
+      "step": 78
+    },
+    {
+      "epoch": 0.1184407796101949,
+      "grad_norm": 0.06305227657672241,
+      "learning_rate": 7.860696517412935e-05,
+      "loss": 0.0317,
+      "step": 79
+    },
+    {
+      "epoch": 0.1199400299850075,
+      "grad_norm": 0.035442799985056346,
+      "learning_rate": 7.960199004975125e-05,
+      "loss": 0.015,
+      "step": 80
+    },
+    {
+      "epoch": 0.12143928035982009,
+      "grad_norm": 0.05183095808072367,
+      "learning_rate": 8.059701492537314e-05,
+      "loss": 0.0242,
+      "step": 81
+    },
+    {
+      "epoch": 0.12293853073463268,
+      "grad_norm": 0.21288021373308544,
+      "learning_rate": 8.159203980099502e-05,
+      "loss": 0.0657,
+      "step": 82
+    },
+    {
+      "epoch": 0.12443778110944528,
+      "grad_norm": 0.14499038539812795,
+      "learning_rate": 8.258706467661693e-05,
+      "loss": 0.0539,
+      "step": 83
+    },
+    {
+      "epoch": 0.12593703148425786,
+      "grad_norm": 0.07506130007270491,
+      "learning_rate": 8.358208955223881e-05,
+      "loss": 0.0357,
+      "step": 84
+    },
+    {
+      "epoch": 0.12743628185907047,
+      "grad_norm": 0.06540286206867715,
+      "learning_rate": 8.45771144278607e-05,
+      "loss": 0.0249,
+      "step": 85
+    },
+    {
+      "epoch": 0.12893553223388307,
+      "grad_norm": 0.046324410553248214,
+      "learning_rate": 8.55721393034826e-05,
+      "loss": 0.0211,
+      "step": 86
+    },
+    {
+      "epoch": 0.13043478260869565,
+      "grad_norm": 0.04910635273005273,
+      "learning_rate": 8.656716417910447e-05,
+      "loss": 0.0174,
+      "step": 87
+    },
+    {
+      "epoch": 0.13193403298350825,
+      "grad_norm": 0.06542227021586179,
+      "learning_rate": 8.756218905472637e-05,
+      "loss": 0.0306,
+      "step": 88
+    },
+    {
+      "epoch": 0.13343328335832083,
+      "grad_norm": 0.04867631138137704,
+      "learning_rate": 8.855721393034826e-05,
+      "loss": 0.0124,
+      "step": 89
+    },
+    {
+      "epoch": 0.13493253373313344,
+      "grad_norm": 0.03955791644617872,
+      "learning_rate": 8.955223880597016e-05,
+      "loss": 0.02,
+      "step": 90
+    },
+    {
+      "epoch": 0.136431784107946,
+      "grad_norm": 0.07757595899547425,
+      "learning_rate": 9.054726368159205e-05,
+      "loss": 0.0349,
+      "step": 91
+    },
+    {
+      "epoch": 0.13793103448275862,
+      "grad_norm": 0.05833106304874586,
+      "learning_rate": 9.154228855721394e-05,
+      "loss": 0.0215,
+      "step": 92
+    },
+    {
+      "epoch": 0.13943028485757122,
+      "grad_norm": 0.07191436633738099,
+      "learning_rate": 9.253731343283582e-05,
+      "loss": 0.0351,
+      "step": 93
+    },
+    {
+      "epoch": 0.1409295352323838,
+      "grad_norm": 0.074329975473276,
+      "learning_rate": 9.353233830845772e-05,
+      "loss": 0.0334,
+      "step": 94
+    },
+    {
+      "epoch": 0.1424287856071964,
+      "grad_norm": 0.11507286561615733,
+      "learning_rate": 9.452736318407961e-05,
+      "loss": 0.039,
+      "step": 95
+    },
+    {
+      "epoch": 0.14392803598200898,
+      "grad_norm": 0.0500122604486147,
+      "learning_rate": 9.552238805970149e-05,
+      "loss": 0.027,
+      "step": 96
+    },
+    {
+      "epoch": 0.1454272863568216,
+      "grad_norm": 0.11474477698178565,
+      "learning_rate": 9.65174129353234e-05,
+      "loss": 0.0751,
+      "step": 97
+    },
+    {
+      "epoch": 0.1469265367316342,
+      "grad_norm": 0.13909793248506105,
+      "learning_rate": 9.751243781094528e-05,
+      "loss": 0.0397,
+      "step": 98
+    },
+    {
+      "epoch": 0.14842578710644677,
+      "grad_norm": 0.06883924844168375,
+      "learning_rate": 9.850746268656717e-05,
+      "loss": 0.0356,
+      "step": 99
+    },
+    {
+      "epoch": 0.14992503748125938,
+      "grad_norm": 0.03603265842985656,
+      "learning_rate": 9.950248756218906e-05,
+      "loss": 0.0132,
+      "step": 100
+    },
+    {
+      "epoch": 0.15142428785607195,
+      "grad_norm": 0.03810266042373329,
+      "learning_rate": 0.00010049751243781096,
+      "loss": 0.0142,
+      "step": 101
+    },
+    {
+      "epoch": 0.15292353823088456,
+      "grad_norm": 0.053717895257444505,
+      "learning_rate": 0.00010149253731343284,
+      "loss": 0.0277,
+      "step": 102
+    },
+    {
+      "epoch": 0.15442278860569716,
+      "grad_norm": 0.04818876674919931,
+      "learning_rate": 0.00010248756218905473,
+      "loss": 0.0195,
+      "step": 103
+    },
+    {
+      "epoch": 0.15592203898050974,
+      "grad_norm": 0.10941924254236936,
+      "learning_rate": 0.00010348258706467661,
+      "loss": 0.0384,
+      "step": 104
+    },
+    {
+      "epoch": 0.15742128935532235,
+      "grad_norm": 0.07867930264182815,
+      "learning_rate": 0.0001044776119402985,
+      "loss": 0.0162,
+      "step": 105
+    },
+    {
+      "epoch": 0.15892053973013492,
+      "grad_norm": 0.08345512376505489,
+      "learning_rate": 0.00010547263681592041,
+      "loss": 0.0328,
+      "step": 106
+    },
+    {
+      "epoch": 0.16041979010494753,
+      "grad_norm": 0.09189672805052056,
+      "learning_rate": 0.00010646766169154229,
+      "loss": 0.0277,
+      "step": 107
+    },
+    {
+      "epoch": 0.1619190404797601,
+      "grad_norm": 0.0453433254814091,
+      "learning_rate": 0.00010746268656716419,
+      "loss": 0.014,
+      "step": 108
+    },
+    {
+      "epoch": 0.1634182908545727,
+      "grad_norm": 0.05405715363413614,
+      "learning_rate": 0.00010845771144278607,
+      "loss": 0.0291,
+      "step": 109
+    },
+    {
+      "epoch": 0.16491754122938532,
+      "grad_norm": 0.03927392847370858,
+      "learning_rate": 0.00010945273631840796,
+      "loss": 0.0156,
+      "step": 110
+    },
+    {
+      "epoch": 0.1664167916041979,
+      "grad_norm": 0.08989520102953286,
+      "learning_rate": 0.00011044776119402987,
+      "loss": 0.0234,
+      "step": 111
+    },
+    {
+      "epoch": 0.1679160419790105,
+      "grad_norm": 0.1268177748837639,
+      "learning_rate": 0.00011144278606965175,
+      "loss": 0.0422,
+      "step": 112
+    },
+    {
+      "epoch": 0.16941529235382308,
+      "grad_norm": 0.03738615100934019,
+      "learning_rate": 0.00011243781094527364,
+      "loss": 0.0155,
+      "step": 113
+    },
+    {
+      "epoch": 0.17091454272863568,
+      "grad_norm": 0.04597772784671858,
+      "learning_rate": 0.00011343283582089552,
+      "loss": 0.0228,
+      "step": 114
+    },
+    {
+      "epoch": 0.1724137931034483,
+      "grad_norm": 0.059249691336683494,
+      "learning_rate": 0.00011442786069651741,
+      "loss": 0.0339,
+      "step": 115
+    },
+    {
+      "epoch": 0.17391304347826086,
+      "grad_norm": 0.08104873791292298,
+      "learning_rate": 0.00011542288557213932,
+      "loss": 0.0453,
+      "step": 116
+    },
+    {
+      "epoch": 0.17541229385307347,
+      "grad_norm": 0.06588671288622634,
+      "learning_rate": 0.0001164179104477612,
+      "loss": 0.0217,
+      "step": 117
+    },
+    {
+      "epoch": 0.17691154422788605,
+      "grad_norm": 0.03522749702212235,
+      "learning_rate": 0.00011741293532338308,
+      "loss": 0.0179,
+      "step": 118
+    },
+    {
+      "epoch": 0.17841079460269865,
+      "grad_norm": 0.0996319146350363,
+      "learning_rate": 0.00011840796019900498,
+      "loss": 0.0326,
+      "step": 119
+    },
+    {
+      "epoch": 0.17991004497751126,
+      "grad_norm": 0.034423484759631535,
+      "learning_rate": 0.00011940298507462686,
+      "loss": 0.013,
+      "step": 120
+    },
+    {
+      "epoch": 0.18140929535232383,
+      "grad_norm": 0.07738518975880311,
+      "learning_rate": 0.00012039800995024876,
+      "loss": 0.0276,
+      "step": 121
+    },
+    {
+      "epoch": 0.18290854572713644,
+      "grad_norm": 0.07113916964290797,
+      "learning_rate": 0.00012139303482587066,
+      "loss": 0.0296,
+      "step": 122
+    },
+    {
+      "epoch": 0.18440779610194902,
+      "grad_norm": 0.04801547092888486,
+      "learning_rate": 0.00012238805970149255,
+      "loss": 0.017,
+      "step": 123
+    },
+    {
+      "epoch": 0.18590704647676162,
+      "grad_norm": 0.0708304924302058,
+      "learning_rate": 0.00012338308457711443,
+      "loss": 0.0224,
+      "step": 124
+    },
+    {
+      "epoch": 0.1874062968515742,
+      "grad_norm": 0.08325794068925856,
+      "learning_rate": 0.0001243781094527363,
+      "loss": 0.0268,
+      "step": 125
+    },
+    {
+      "epoch": 0.1889055472263868,
+      "grad_norm": 0.049077714829375424,
+      "learning_rate": 0.00012537313432835822,
+      "loss": 0.0205,
+      "step": 126
+    },
+    {
+      "epoch": 0.1904047976011994,
+      "grad_norm": 0.08564942290202891,
+      "learning_rate": 0.0001263681592039801,
+      "loss": 0.0267,
+      "step": 127
+    },
+    {
+      "epoch": 0.191904047976012,
+      "grad_norm": 0.07021233023137041,
+      "learning_rate": 0.000127363184079602,
+      "loss": 0.0318,
+      "step": 128
+    },
+    {
+      "epoch": 0.1934032983508246,
+      "grad_norm": 0.09337476280465129,
+      "learning_rate": 0.00012835820895522389,
+      "loss": 0.022,
+      "step": 129
+    },
+    {
+      "epoch": 0.19490254872563717,
+      "grad_norm": 0.09811732092693733,
+      "learning_rate": 0.0001293532338308458,
+      "loss": 0.0554,
+      "step": 130
+    },
+    {
+      "epoch": 0.19640179910044978,
+      "grad_norm": 0.06961228532261528,
+      "learning_rate": 0.00013034825870646767,
+      "loss": 0.0356,
+      "step": 131
+    },
+    {
+      "epoch": 0.19790104947526238,
+      "grad_norm": 0.08084464838848786,
+      "learning_rate": 0.00013134328358208955,
+      "loss": 0.0298,
+      "step": 132
+    },
+    {
+      "epoch": 0.19940029985007496,
+      "grad_norm": 0.051768566679399344,
+      "learning_rate": 0.00013233830845771146,
+      "loss": 0.0118,
+      "step": 133
+    },
+    {
+      "epoch": 0.20089955022488756,
+      "grad_norm": 0.05361967944259781,
+      "learning_rate": 0.00013333333333333334,
+      "loss": 0.0329,
+      "step": 134
+    },
+    {
+      "epoch": 0.20239880059970014,
+      "grad_norm": 0.06705674679942888,
+      "learning_rate": 0.00013432835820895525,
+      "loss": 0.0237,
+      "step": 135
+    },
+    {
+      "epoch": 0.20389805097451275,
+      "grad_norm": 0.1588965400271315,
+      "learning_rate": 0.00013532338308457713,
+      "loss": 0.0199,
+      "step": 136
+    },
+    {
+      "epoch": 0.20539730134932535,
+      "grad_norm": 0.045155120039124384,
+      "learning_rate": 0.000136318407960199,
+      "loss": 0.0257,
+      "step": 137
+    },
+    {
+      "epoch": 0.20689655172413793,
+      "grad_norm": 0.10905612081286897,
+      "learning_rate": 0.0001373134328358209,
+      "loss": 0.0321,
+      "step": 138
+    },
+    {
+      "epoch": 0.20839580209895053,
+      "grad_norm": 0.11760170984366283,
+      "learning_rate": 0.0001383084577114428,
+      "loss": 0.0372,
+      "step": 139
+    },
+    {
+      "epoch": 0.2098950524737631,
+      "grad_norm": 0.08582443938861739,
+      "learning_rate": 0.0001393034825870647,
+      "loss": 0.0426,
+      "step": 140
+    },
+    {
+      "epoch": 0.21139430284857572,
+      "grad_norm": 0.07873025758417922,
+      "learning_rate": 0.00014029850746268658,
+      "loss": 0.025,
+      "step": 141
+    },
+    {
+      "epoch": 0.2128935532233883,
+      "grad_norm": 0.10318630643872907,
+      "learning_rate": 0.00014129353233830846,
+      "loss": 0.0333,
+      "step": 142
+    },
+    {
+      "epoch": 0.2143928035982009,
+      "grad_norm": 0.062306207851108485,
+      "learning_rate": 0.00014228855721393034,
+      "loss": 0.0174,
+      "step": 143
+    },
+    {
+      "epoch": 0.2158920539730135,
+      "grad_norm": 0.06923837226150165,
+      "learning_rate": 0.00014328358208955225,
+      "loss": 0.0236,
+      "step": 144
+    },
+    {
+      "epoch": 0.21739130434782608,
+      "grad_norm": 0.08009174706122868,
+      "learning_rate": 0.00014427860696517416,
+      "loss": 0.0377,
+      "step": 145
+    },
+    {
+      "epoch": 0.21889055472263869,
+      "grad_norm": 0.07421272166337643,
+      "learning_rate": 0.00014527363184079604,
+      "loss": 0.0285,
+      "step": 146
+    },
+    {
+      "epoch": 0.22038980509745126,
+      "grad_norm": 0.06394699190672182,
+      "learning_rate": 0.00014626865671641792,
+      "loss": 0.0201,
+      "step": 147
+    },
+    {
+      "epoch": 0.22188905547226387,
+      "grad_norm": 0.1230962560854086,
+      "learning_rate": 0.0001472636815920398,
+      "loss": 0.0418,
+      "step": 148
+    },
+    {
+      "epoch": 0.22338830584707647,
+      "grad_norm": 0.09792223962749103,
+      "learning_rate": 0.00014825870646766168,
+      "loss": 0.0428,
+      "step": 149
+    },
+    {
+      "epoch": 0.22488755622188905,
+      "grad_norm": 0.08935647805738876,
+      "learning_rate": 0.0001492537313432836,
+      "loss": 0.0286,
+      "step": 150
+    },
+    {
+      "epoch": 0.22638680659670166,
+      "grad_norm": 0.05869779630249231,
+      "learning_rate": 0.0001502487562189055,
+      "loss": 0.0266,
+      "step": 151
+    },
+    {
+      "epoch": 0.22788605697151423,
+      "grad_norm": 0.09727307974918385,
+      "learning_rate": 0.00015124378109452737,
+      "loss": 0.0251,
+      "step": 152
+    },
+    {
+      "epoch": 0.22938530734632684,
+      "grad_norm": 0.05489680833229503,
+      "learning_rate": 0.00015223880597014925,
+      "loss": 0.0256,
+      "step": 153
+    },
+    {
+      "epoch": 0.23088455772113944,
+      "grad_norm": 0.08651252294071486,
+      "learning_rate": 0.00015323383084577113,
+      "loss": 0.0324,
+      "step": 154
+    },
+    {
+      "epoch": 0.23238380809595202,
+      "grad_norm": 0.06084445807329521,
+      "learning_rate": 0.00015422885572139304,
+      "loss": 0.0181,
+      "step": 155
+    },
+    {
+      "epoch": 0.23388305847076463,
+      "grad_norm": 0.07984776396235738,
+      "learning_rate": 0.00015522388059701495,
+      "loss": 0.0293,
+      "step": 156
+    },
+    {
+      "epoch": 0.2353823088455772,
+      "grad_norm": 0.10663711823358574,
+      "learning_rate": 0.00015621890547263683,
+      "loss": 0.0224,
+      "step": 157
+    },
+    {
+      "epoch": 0.2368815592203898,
+      "grad_norm": 0.09029974452506394,
+      "learning_rate": 0.0001572139303482587,
+      "loss": 0.0391,
+      "step": 158
+    },
+    {
+      "epoch": 0.2383808095952024,
+      "grad_norm": 0.07475818991216006,
+      "learning_rate": 0.00015820895522388059,
+      "loss": 0.0325,
+      "step": 159
+    },
+    {
+      "epoch": 0.239880059970015,
+      "grad_norm": 0.10281310121414294,
+      "learning_rate": 0.0001592039800995025,
+      "loss": 0.0233,
+      "step": 160
+    },
+    {
+      "epoch": 0.2413793103448276,
+      "grad_norm": 0.07648721213716447,
+      "learning_rate": 0.0001601990049751244,
+      "loss": 0.0309,
+      "step": 161
+    },
+    {
+      "epoch": 0.24287856071964017,
+      "grad_norm": 0.1353558632906228,
+      "learning_rate": 0.00016119402985074628,
+      "loss": 0.0412,
+      "step": 162
+    },
+    {
+      "epoch": 0.24437781109445278,
+      "grad_norm": 0.11951297623629252,
+      "learning_rate": 0.00016218905472636816,
+      "loss": 0.0454,
+      "step": 163
+    },
+    {
+      "epoch": 0.24587706146926536,
+      "grad_norm": 0.09337941334405561,
+      "learning_rate": 0.00016318407960199004,
+      "loss": 0.0261,
+      "step": 164
+    },
+    {
+      "epoch": 0.24737631184407796,
+      "grad_norm": 0.07631986818436004,
+      "learning_rate": 0.00016417910447761195,
+      "loss": 0.0184,
+      "step": 165
+    },
+    {
+      "epoch": 0.24887556221889057,
+      "grad_norm": 0.06382771953073095,
+      "learning_rate": 0.00016517412935323385,
+      "loss": 0.0229,
+      "step": 166
+    },
+    {
+      "epoch": 0.25037481259370314,
+      "grad_norm": 0.0654124597455561,
+      "learning_rate": 0.00016616915422885573,
+      "loss": 0.0275,
+      "step": 167
+    },
+    {
+      "epoch": 0.2518740629685157,
+      "grad_norm": 0.07386539118384822,
+      "learning_rate": 0.00016716417910447761,
+      "loss": 0.0276,
+      "step": 168
+    },
+    {
+      "epoch": 0.25337331334332835,
+      "grad_norm": 0.10790605839437757,
+      "learning_rate": 0.0001681592039800995,
+      "loss": 0.0284,
+      "step": 169
+    },
+    {
+      "epoch": 0.25487256371814093,
+      "grad_norm": 0.07520911787134094,
+      "learning_rate": 0.0001691542288557214,
+      "loss": 0.033,
+      "step": 170
+    },
+    {
+      "epoch": 0.2563718140929535,
+      "grad_norm": 0.0437842352906429,
+      "learning_rate": 0.00017014925373134328,
+      "loss": 0.0164,
+      "step": 171
+    },
+    {
+      "epoch": 0.25787106446776614,
+      "grad_norm": 0.1241410943091543,
+      "learning_rate": 0.0001711442786069652,
+      "loss": 0.0338,
+      "step": 172
+    },
+    {
+      "epoch": 0.2593703148425787,
+      "grad_norm": 0.08555919176928421,
+      "learning_rate": 0.00017213930348258707,
+      "loss": 0.0219,
+      "step": 173
+    },
+    {
+      "epoch": 0.2608695652173913,
+      "grad_norm": 0.11079300082585909,
+      "learning_rate": 0.00017313432835820895,
+      "loss": 0.0314,
+      "step": 174
+    },
+    {
+      "epoch": 0.2623688155922039,
+      "grad_norm": 0.046045735438355655,
+      "learning_rate": 0.00017412935323383086,
+      "loss": 0.0111,
+      "step": 175
+    },
+    {
+      "epoch": 0.2638680659670165,
+      "grad_norm": 0.10739428025622968,
+      "learning_rate": 0.00017512437810945274,
+      "loss": 0.0224,
+      "step": 176
+    },
+    {
+      "epoch": 0.2653673163418291,
+      "grad_norm": 0.2449779327117588,
+      "learning_rate": 0.00017611940298507464,
+      "loss": 0.0464,
+      "step": 177
+    },
+    {
+      "epoch": 0.26686656671664166,
+      "grad_norm": 0.08360762118297338,
+      "learning_rate": 0.00017711442786069652,
+      "loss": 0.0312,
+      "step": 178
+    },
+    {
+      "epoch": 0.2683658170914543,
+      "grad_norm": 0.10593436165792847,
+      "learning_rate": 0.00017810945273631843,
+      "loss": 0.026,
+      "step": 179
+    },
+    {
+      "epoch": 0.2698650674662669,
+      "grad_norm": 0.07842081945251449,
+      "learning_rate": 0.0001791044776119403,
+      "loss": 0.029,
+      "step": 180
+    },
+    {
+      "epoch": 0.27136431784107945,
+      "grad_norm": 0.0919363194425753,
+      "learning_rate": 0.0001800995024875622,
+      "loss": 0.0246,
+      "step": 181
+    },
+    {
+      "epoch": 0.272863568215892,
+      "grad_norm": 0.07122712223445765,
+      "learning_rate": 0.0001810945273631841,
+      "loss": 0.0175,
+      "step": 182
+    },
+    {
+      "epoch": 0.27436281859070466,
+      "grad_norm": 0.12986448311978344,
+      "learning_rate": 0.00018208955223880598,
+      "loss": 0.0327,
+      "step": 183
+    },
+    {
+      "epoch": 0.27586206896551724,
+      "grad_norm": 0.08939968166772469,
+      "learning_rate": 0.00018308457711442789,
+      "loss": 0.0222,
+      "step": 184
+    },
+    {
+      "epoch": 0.2773613193403298,
+      "grad_norm": 0.12467738587334767,
+      "learning_rate": 0.00018407960199004977,
+      "loss": 0.0398,
+      "step": 185
+    },
+    {
+      "epoch": 0.27886056971514245,
+      "grad_norm": 0.08880261150192895,
+      "learning_rate": 0.00018507462686567165,
+      "loss": 0.0276,
+      "step": 186
+    },
+    {
+      "epoch": 0.280359820089955,
+      "grad_norm": 0.06274559081133477,
+      "learning_rate": 0.00018606965174129353,
+      "loss": 0.0162,
+      "step": 187
+    },
+    {
+      "epoch": 0.2818590704647676,
+      "grad_norm": 0.06826841539115386,
+      "learning_rate": 0.00018706467661691543,
+      "loss": 0.0252,
+      "step": 188
+    },
+    {
+      "epoch": 0.28335832083958024,
+      "grad_norm": 0.07615150472353269,
+      "learning_rate": 0.00018805970149253734,
+      "loss": 0.0206,
+      "step": 189
+    },
+    {
+      "epoch": 0.2848575712143928,
+      "grad_norm": 0.09408473494987521,
+      "learning_rate": 0.00018905472636815922,
+      "loss": 0.0274,
+      "step": 190
+    },
+    {
+      "epoch": 0.2863568215892054,
+      "grad_norm": 0.07069036227398094,
+      "learning_rate": 0.0001900497512437811,
+      "loss": 0.0196,
+      "step": 191
+    },
+    {
+      "epoch": 0.28785607196401797,
+      "grad_norm": 0.09160386389460257,
+      "learning_rate": 0.00019104477611940298,
+      "loss": 0.0425,
+      "step": 192
+    },
+    {
+      "epoch": 0.2893553223388306,
+      "grad_norm": 0.11814489122880895,
+      "learning_rate": 0.0001920398009950249,
+      "loss": 0.0312,
+      "step": 193
+    },
+    {
+      "epoch": 0.2908545727136432,
+      "grad_norm": 0.07267374829272864,
+      "learning_rate": 0.0001930348258706468,
+      "loss": 0.0206,
+      "step": 194
+    },
+    {
+      "epoch": 0.29235382308845576,
+      "grad_norm": 0.056069092746323,
+      "learning_rate": 0.00019402985074626867,
+      "loss": 0.0173,
+      "step": 195
+    },
+    {
+      "epoch": 0.2938530734632684,
+      "grad_norm": 0.1072633610690865,
+      "learning_rate": 0.00019502487562189055,
+      "loss": 0.0356,
+      "step": 196
+    },
+    {
+      "epoch": 0.29535232383808097,
+      "grad_norm": 0.10655067222684735,
+      "learning_rate": 0.00019601990049751244,
+      "loss": 0.0381,
+      "step": 197
+    },
+    {
+      "epoch": 0.29685157421289354,
+      "grad_norm": 0.12811931955318204,
+      "learning_rate": 0.00019701492537313434,
+      "loss": 0.0328,
+      "step": 198
+    },
+    {
+      "epoch": 0.2983508245877061,
+      "grad_norm": 0.0896744841892114,
+      "learning_rate": 0.00019800995024875625,
+      "loss": 0.0201,
+      "step": 199
+    },
+    {
+      "epoch": 0.29985007496251875,
+      "grad_norm": 0.08188692710670159,
+      "learning_rate": 0.00019900497512437813,
+      "loss": 0.026,
+      "step": 200
+    },
+    {
+      "epoch": 0.30134932533733133,
+      "grad_norm": 0.06429287812218996,
+      "learning_rate": 0.0002,
+      "loss": 0.0236,
+      "step": 201
+    },
+    {
+      "epoch": 0.3028485757121439,
+      "grad_norm": 0.0889591592714688,
+      "learning_rate": 0.00019999984769132877,
+      "loss": 0.018,
+      "step": 202
+    },
+    {
+      "epoch": 0.30434782608695654,
+      "grad_norm": 0.09099341079961838,
+      "learning_rate": 0.00019999939076577905,
+      "loss": 0.0341,
+      "step": 203
+    },
+    {
+      "epoch": 0.3058470764617691,
+      "grad_norm": 0.1056474393511896,
+      "learning_rate": 0.00019999862922474268,
+      "loss": 0.0438,
+      "step": 204
+    },
+    {
+      "epoch": 0.3073463268365817,
+      "grad_norm": 0.06324696346281937,
+      "learning_rate": 0.00019999756307053948,
+      "loss": 0.02,
+      "step": 205
+    },
+    {
+      "epoch": 0.30884557721139433,
+      "grad_norm": 0.12201171857861728,
+      "learning_rate": 0.00019999619230641713,
+      "loss": 0.0357,
+      "step": 206
+    },
+    {
+      "epoch": 0.3103448275862069,
+      "grad_norm": 0.16019354966510604,
+      "learning_rate": 0.00019999451693655123,
+      "loss": 0.0308,
+      "step": 207
+    },
+    {
+      "epoch": 0.3118440779610195,
+      "grad_norm": 0.08064191205932211,
+      "learning_rate": 0.00019999253696604522,
+      "loss": 0.0231,
+      "step": 208
+    },
+    {
+      "epoch": 0.31334332833583206,
+      "grad_norm": 0.1735406201730255,
+      "learning_rate": 0.00019999025240093044,
+      "loss": 0.0531,
+      "step": 209
+    },
+    {
+      "epoch": 0.3148425787106447,
+      "grad_norm": 0.1270042673971674,
+      "learning_rate": 0.00019998766324816607,
+      "loss": 0.0372,
+      "step": 210
+    },
+    {
+      "epoch": 0.31634182908545727,
+      "grad_norm": 0.07759625937069312,
+      "learning_rate": 0.00019998476951563915,
+      "loss": 0.0215,
+      "step": 211
+    },
+    {
+      "epoch": 0.31784107946026985,
+      "grad_norm": 0.06823751969137144,
+      "learning_rate": 0.00019998157121216444,
+      "loss": 0.0188,
+      "step": 212
+    },
+    {
+      "epoch": 0.3193403298350825,
+      "grad_norm": 0.08549856995516077,
+      "learning_rate": 0.00019997806834748456,
+      "loss": 0.0254,
+      "step": 213
+    },
+    {
+      "epoch": 0.32083958020989506,
+      "grad_norm": 0.0625354342920355,
+      "learning_rate": 0.00019997426093226986,
+      "loss": 0.014,
+      "step": 214
+    },
+    {
+      "epoch": 0.32233883058470764,
+      "grad_norm": 0.0760895159675007,
+      "learning_rate": 0.00019997014897811833,
+      "loss": 0.0136,
+      "step": 215
+    },
+    {
+      "epoch": 0.3238380809595202,
+      "grad_norm": 0.1679998356790272,
+      "learning_rate": 0.00019996573249755572,
+      "loss": 0.0546,
+      "step": 216
+    },
+    {
+      "epoch": 0.32533733133433285,
+      "grad_norm": 0.10916710981504504,
+      "learning_rate": 0.00019996101150403543,
+      "loss": 0.0275,
+      "step": 217
+    },
+    {
+      "epoch": 0.3268365817091454,
+      "grad_norm": 0.10201374094804785,
+      "learning_rate": 0.00019995598601193842,
+      "loss": 0.0289,
+      "step": 218
+    },
+    {
+      "epoch": 0.328335832083958,
+      "grad_norm": 0.08058975597656892,
+      "learning_rate": 0.00019995065603657316,
+      "loss": 0.0249,
+      "step": 219
+    },
+    {
+      "epoch": 0.32983508245877063,
+      "grad_norm": 0.15985586700330837,
+      "learning_rate": 0.00019994502159417573,
+      "loss": 0.0345,
+      "step": 220
+    },
+    {
+      "epoch": 0.3313343328335832,
+      "grad_norm": 0.14812712527123406,
+      "learning_rate": 0.0001999390827019096,
+      "loss": 0.0377,
+      "step": 221
+    },
+    {
+      "epoch": 0.3328335832083958,
+      "grad_norm": 0.1178146028681012,
+      "learning_rate": 0.00019993283937786563,
+      "loss": 0.0304,
+      "step": 222
+    },
+    {
+      "epoch": 0.3343328335832084,
+      "grad_norm": 0.10003216574781101,
+      "learning_rate": 0.0001999262916410621,
+      "loss": 0.0342,
+      "step": 223
+    },
+    {
+      "epoch": 0.335832083958021,
+      "grad_norm": 0.10739566491501087,
+      "learning_rate": 0.0001999194395114446,
+      "loss": 0.0387,
+      "step": 224
+    },
+    {
+      "epoch": 0.3373313343328336,
+      "grad_norm": 0.12471587478993325,
+      "learning_rate": 0.00019991228300988585,
+      "loss": 0.0271,
+      "step": 225
+    },
+    {
+      "epoch": 0.33883058470764615,
+      "grad_norm": 0.10534548813143461,
+      "learning_rate": 0.0001999048221581858,
+      "loss": 0.0338,
+      "step": 226
+    },
+    {
+      "epoch": 0.3403298350824588,
+      "grad_norm": 0.06743838763365886,
+      "learning_rate": 0.00019989705697907149,
+      "loss": 0.02,
+      "step": 227
+    },
+    {
+      "epoch": 0.34182908545727136,
+      "grad_norm": 0.0757865381180341,
+      "learning_rate": 0.00019988898749619702,
+      "loss": 0.0215,
+      "step": 228
+    },
+    {
+      "epoch": 0.34332833583208394,
+      "grad_norm": 0.14709355458702295,
+      "learning_rate": 0.0001998806137341434,
+      "loss": 0.0227,
+      "step": 229
+    },
+    {
+      "epoch": 0.3448275862068966,
+      "grad_norm": 0.06430710408458629,
+      "learning_rate": 0.0001998719357184186,
+      "loss": 0.0122,
+      "step": 230
+    },
+    {
+      "epoch": 0.34632683658170915,
+      "grad_norm": 0.09018751183295319,
+      "learning_rate": 0.0001998629534754574,
+      "loss": 0.0217,
+      "step": 231
+    },
+    {
+      "epoch": 0.34782608695652173,
+      "grad_norm": 0.14650415427998667,
+      "learning_rate": 0.0001998536670326212,
+      "loss": 0.0493,
+      "step": 232
+    },
+    {
+      "epoch": 0.3493253373313343,
+      "grad_norm": 0.06334988631041422,
+      "learning_rate": 0.00019984407641819812,
+      "loss": 0.0239,
+      "step": 233
+    },
+    {
+      "epoch": 0.35082458770614694,
+      "grad_norm": 0.20886697619030714,
+      "learning_rate": 0.00019983418166140285,
+      "loss": 0.0564,
+      "step": 234
+    },
+    {
+      "epoch": 0.3523238380809595,
+      "grad_norm": 0.14466113183818535,
+      "learning_rate": 0.00019982398279237655,
+      "loss": 0.0204,
+      "step": 235
+    },
+    {
+      "epoch": 0.3538230884557721,
+      "grad_norm": 0.0982852108329072,
+      "learning_rate": 0.0001998134798421867,
+      "loss": 0.0273,
+      "step": 236
+    },
+    {
+      "epoch": 0.3553223388305847,
+      "grad_norm": 0.11164221678350923,
+      "learning_rate": 0.00019980267284282717,
+      "loss": 0.0282,
+      "step": 237
+    },
+    {
+      "epoch": 0.3568215892053973,
+      "grad_norm": 0.10368498660671552,
+      "learning_rate": 0.0001997915618272179,
+      "loss": 0.0299,
+      "step": 238
+    },
+    {
+      "epoch": 0.3583208395802099,
+      "grad_norm": 0.09881979045943046,
+      "learning_rate": 0.000199780146829205,
+      "loss": 0.0255,
+      "step": 239
+    },
+    {
+      "epoch": 0.3598200899550225,
+      "grad_norm": 0.1379397238385945,
+      "learning_rate": 0.00019976842788356055,
+      "loss": 0.0217,
+      "step": 240
+    },
+    {
+      "epoch": 0.3613193403298351,
+      "grad_norm": 0.07438604732967513,
+      "learning_rate": 0.00019975640502598244,
+      "loss": 0.0243,
+      "step": 241
+    },
+    {
+      "epoch": 0.36281859070464767,
+      "grad_norm": 0.07473607637141821,
+      "learning_rate": 0.0001997440782930944,
+      "loss": 0.0258,
+      "step": 242
+    },
+    {
+      "epoch": 0.36431784107946025,
+      "grad_norm": 0.0882005313753972,
+      "learning_rate": 0.00019973144772244582,
+      "loss": 0.0268,
+      "step": 243
+    },
+    {
+      "epoch": 0.3658170914542729,
+      "grad_norm": 0.14911031648636602,
+      "learning_rate": 0.00019971851335251158,
+      "loss": 0.0378,
+      "step": 244
+    },
+    {
+      "epoch": 0.36731634182908546,
+      "grad_norm": 0.07841452122968684,
+      "learning_rate": 0.00019970527522269205,
+      "loss": 0.0174,
+      "step": 245
+    },
+    {
+      "epoch": 0.36881559220389803,
+      "grad_norm": 0.09528161070909874,
+      "learning_rate": 0.0001996917333733128,
+      "loss": 0.0295,
+      "step": 246
+    },
+    {
+      "epoch": 0.37031484257871067,
+      "grad_norm": 0.1547071536848402,
+      "learning_rate": 0.00019967788784562473,
+      "loss": 0.0352,
+      "step": 247
+    },
+    {
+      "epoch": 0.37181409295352325,
+      "grad_norm": 0.13481821743931183,
+      "learning_rate": 0.00019966373868180365,
+      "loss": 0.0375,
+      "step": 248
+    },
+    {
+      "epoch": 0.3733133433283358,
+      "grad_norm": 0.14639591924374834,
+      "learning_rate": 0.00019964928592495045,
+      "loss": 0.0408,
+      "step": 249
+    },
+    {
+      "epoch": 0.3748125937031484,
+      "grad_norm": 0.22468937977099931,
+      "learning_rate": 0.00019963452961909063,
+      "loss": 0.0487,
+      "step": 250
+    },
+    {
+      "epoch": 0.37631184407796103,
+      "grad_norm": 0.10317429082367974,
+      "learning_rate": 0.00019961946980917456,
+      "loss": 0.0353,
+      "step": 251
+    },
+    {
+      "epoch": 0.3778110944527736,
+      "grad_norm": 0.11239999568993565,
+      "learning_rate": 0.00019960410654107697,
+      "loss": 0.0198,
+      "step": 252
+    },
+    {
+      "epoch": 0.3793103448275862,
+      "grad_norm": 0.12839126097254944,
+      "learning_rate": 0.00019958843986159704,
+      "loss": 0.0271,
+      "step": 253
+    },
+    {
+      "epoch": 0.3808095952023988,
+      "grad_norm": 0.16134712680852636,
+      "learning_rate": 0.00019957246981845822,
+      "loss": 0.0939,
+      "step": 254
+    },
+    {
+      "epoch": 0.3823088455772114,
+      "grad_norm": 0.10694136315654908,
+      "learning_rate": 0.00019955619646030802,
+      "loss": 0.0191,
+      "step": 255
+    },
+    {
+      "epoch": 0.383808095952024,
+      "grad_norm": 0.13584383708557063,
+      "learning_rate": 0.00019953961983671788,
+      "loss": 0.0391,
+      "step": 256
+    },
+    {
+      "epoch": 0.3853073463268366,
+      "grad_norm": 0.07484156837632686,
+      "learning_rate": 0.0001995227399981831,
+      "loss": 0.0218,
+      "step": 257
+    },
+    {
+      "epoch": 0.3868065967016492,
+      "grad_norm": 0.11474869317330276,
+      "learning_rate": 0.00019950555699612267,
+      "loss": 0.0185,
+      "step": 258
+    },
+    {
+      "epoch": 0.38830584707646176,
+      "grad_norm": 0.08204674574283162,
+      "learning_rate": 0.00019948807088287883,
+      "loss": 0.0167,
+      "step": 259
+    },
+    {
+      "epoch": 0.38980509745127434,
+      "grad_norm": 0.08466516343662758,
+      "learning_rate": 0.00019947028171171742,
+      "loss": 0.0228,
+      "step": 260
+    },
+    {
+      "epoch": 0.391304347826087,
+      "grad_norm": 0.10295941699523375,
+      "learning_rate": 0.00019945218953682734,
+      "loss": 0.035,
+      "step": 261
+    },
+    {
+      "epoch": 0.39280359820089955,
+      "grad_norm": 0.10075799964627083,
+      "learning_rate": 0.00019943379441332047,
+      "loss": 0.0214,
+      "step": 262
+    },
+    {
+      "epoch": 0.39430284857571213,
+      "grad_norm": 0.06840271813941545,
+      "learning_rate": 0.00019941509639723155,
+      "loss": 0.0134,
+      "step": 263
+    },
+    {
+      "epoch": 0.39580209895052476,
+      "grad_norm": 0.18490087281473808,
+      "learning_rate": 0.000199396095545518,
+      "loss": 0.0315,
+      "step": 264
+    },
+    {
+      "epoch": 0.39730134932533734,
+      "grad_norm": 0.3398552153859189,
+      "learning_rate": 0.00019937679191605963,
+      "loss": 0.0447,
+      "step": 265
+    },
+    {
+      "epoch": 0.3988005997001499,
+      "grad_norm": 0.11181599880625129,
+      "learning_rate": 0.00019935718556765876,
+      "loss": 0.0252,
+      "step": 266
+    },
+    {
+      "epoch": 0.4002998500749625,
+      "grad_norm": 0.15176000894386518,
+      "learning_rate": 0.00019933727656003963,
+      "loss": 0.0439,
+      "step": 267
+    },
+    {
+      "epoch": 0.4017991004497751,
+      "grad_norm": 0.11515411944836743,
+      "learning_rate": 0.00019931706495384863,
+      "loss": 0.0317,
+      "step": 268
+    },
+    {
+      "epoch": 0.4032983508245877,
+      "grad_norm": 0.07747512816332972,
+      "learning_rate": 0.0001992965508106537,
+      "loss": 0.029,
+      "step": 269
+    },
+    {
+      "epoch": 0.4047976011994003,
+      "grad_norm": 0.16798336539466618,
+      "learning_rate": 0.00019927573419294456,
+      "loss": 0.072,
+      "step": 270
+    },
+    {
+      "epoch": 0.4062968515742129,
+      "grad_norm": 0.14842393353919936,
+      "learning_rate": 0.00019925461516413223,
+      "loss": 0.0437,
+      "step": 271
+    },
+    {
+      "epoch": 0.4077961019490255,
+      "grad_norm": 0.1450131939805502,
+      "learning_rate": 0.00019923319378854887,
+      "loss": 0.0498,
+      "step": 272
+    },
+    {
+      "epoch": 0.40929535232383807,
+      "grad_norm": 0.17363621713440097,
+      "learning_rate": 0.0001992114701314478,
+      "loss": 0.0611,
+      "step": 273
+    },
+    {
+      "epoch": 0.4107946026986507,
+      "grad_norm": 0.11435400082136321,
+      "learning_rate": 0.000199189444259003,
+      "loss": 0.0227,
+      "step": 274
+    },
+    {
+      "epoch": 0.4122938530734633,
+      "grad_norm": 0.16518050629422557,
+      "learning_rate": 0.00019916711623830903,
+      "loss": 0.0283,
+      "step": 275
+    },
+    {
+      "epoch": 0.41379310344827586,
+      "grad_norm": 0.14080617976911697,
+      "learning_rate": 0.00019914448613738106,
+      "loss": 0.0244,
+      "step": 276
+    },
+    {
+      "epoch": 0.41529235382308843,
+      "grad_norm": 0.10018538841227774,
+      "learning_rate": 0.00019912155402515417,
+      "loss": 0.0256,
+      "step": 277
+    },
+    {
+      "epoch": 0.41679160419790107,
+      "grad_norm": 0.09288507466343968,
+      "learning_rate": 0.00019909831997148362,
+      "loss": 0.0205,
+      "step": 278
+    },
+    {
+      "epoch": 0.41829085457271364,
+      "grad_norm": 0.10076396934289908,
+      "learning_rate": 0.00019907478404714436,
+      "loss": 0.0261,
+      "step": 279
+    },
+    {
+      "epoch": 0.4197901049475262,
+      "grad_norm": 0.14235107824165918,
+      "learning_rate": 0.0001990509463238309,
+      "loss": 0.0317,
+      "step": 280
+    },
+    {
+      "epoch": 0.42128935532233885,
+      "grad_norm": 0.06857782839954504,
+      "learning_rate": 0.00019902680687415705,
+      "loss": 0.016,
+      "step": 281
+    },
+    {
+      "epoch": 0.42278860569715143,
+      "grad_norm": 0.21507070087913607,
+      "learning_rate": 0.00019900236577165576,
+      "loss": 0.0411,
+      "step": 282
+    },
+    {
+      "epoch": 0.424287856071964,
+      "grad_norm": 0.0949166136382304,
+      "learning_rate": 0.0001989776230907789,
+      "loss": 0.0257,
+      "step": 283
+    },
+    {
+      "epoch": 0.4257871064467766,
+      "grad_norm": 0.1934392939287711,
+      "learning_rate": 0.00019895257890689696,
+      "loss": 0.0381,
+      "step": 284
+    },
+    {
+      "epoch": 0.4272863568215892,
+      "grad_norm": 0.17422202559565292,
+      "learning_rate": 0.00019892723329629887,
+      "loss": 0.0261,
+      "step": 285
+    },
+    {
+      "epoch": 0.4287856071964018,
+      "grad_norm": 0.07223682667659033,
+      "learning_rate": 0.0001989015863361917,
+      "loss": 0.0255,
+      "step": 286
+    },
+    {
+      "epoch": 0.4302848575712144,
+      "grad_norm": 0.12279145351334636,
+      "learning_rate": 0.0001988756381047006,
+      "loss": 0.046,
+      "step": 287
+    },
+    {
+      "epoch": 0.431784107946027,
+      "grad_norm": 0.08991953597169815,
+      "learning_rate": 0.00019884938868086835,
+      "loss": 0.0315,
+      "step": 288
+    },
+    {
+      "epoch": 0.4332833583208396,
+      "grad_norm": 0.11188039743306034,
+      "learning_rate": 0.0001988228381446553,
+      "loss": 0.0424,
+      "step": 289
+    },
+    {
+      "epoch": 0.43478260869565216,
+      "grad_norm": 0.18905162896839114,
+      "learning_rate": 0.00019879598657693891,
+      "loss": 0.0427,
+      "step": 290
+    },
+    {
+      "epoch": 0.4362818590704648,
+      "grad_norm": 0.12194462370002573,
+      "learning_rate": 0.00019876883405951377,
+      "loss": 0.0225,
+      "step": 291
+    },
+    {
+      "epoch": 0.43778110944527737,
+      "grad_norm": 0.27514043276591943,
+      "learning_rate": 0.00019874138067509117,
+      "loss": 0.0556,
+      "step": 292
+    },
+    {
+      "epoch": 0.43928035982008995,
+      "grad_norm": 0.09504669927857831,
+      "learning_rate": 0.0001987136265072988,
+      "loss": 0.0182,
+      "step": 293
+    },
+    {
+      "epoch": 0.4407796101949025,
+      "grad_norm": 0.08844611007050139,
+      "learning_rate": 0.00019868557164068074,
+      "loss": 0.018,
+      "step": 294
+    },
+    {
+      "epoch": 0.44227886056971516,
+      "grad_norm": 0.1369491212885427,
+      "learning_rate": 0.00019865721616069696,
+      "loss": 0.0492,
+      "step": 295
+    },
+    {
+      "epoch": 0.44377811094452774,
+      "grad_norm": 0.1142377954005939,
+      "learning_rate": 0.00019862856015372317,
+      "loss": 0.0334,
+      "step": 296
+    },
+    {
+      "epoch": 0.4452773613193403,
+      "grad_norm": 0.1961332653625071,
+      "learning_rate": 0.0001985996037070505,
+      "loss": 0.0481,
+      "step": 297
+    },
+    {
+      "epoch": 0.44677661169415295,
+      "grad_norm": 0.050262046246270925,
+      "learning_rate": 0.00019857034690888537,
+      "loss": 0.0154,
+      "step": 298
+    },
+    {
+      "epoch": 0.4482758620689655,
+      "grad_norm": 0.10957587831691169,
+      "learning_rate": 0.00019854078984834903,
+      "loss": 0.0302,
+      "step": 299
+    },
+    {
+      "epoch": 0.4497751124437781,
+      "grad_norm": 0.18658220857685753,
+      "learning_rate": 0.0001985109326154774,
+      "loss": 0.0317,
+      "step": 300
+    },
+    {
+      "epoch": 0.4512743628185907,
+      "grad_norm": 0.09056328641521297,
+      "learning_rate": 0.00019848077530122083,
+      "loss": 0.026,
+      "step": 301
+    },
+    {
+      "epoch": 0.4527736131934033,
+      "grad_norm": 0.16097758286803418,
+      "learning_rate": 0.00019845031799744367,
+      "loss": 0.0457,
+      "step": 302
+    },
+    {
+      "epoch": 0.4542728635682159,
+      "grad_norm": 0.07689789497731149,
+      "learning_rate": 0.0001984195607969242,
+      "loss": 0.0202,
+      "step": 303
+    },
+    {
+      "epoch": 0.45577211394302847,
+      "grad_norm": 0.10663066866505838,
+      "learning_rate": 0.00019838850379335417,
+      "loss": 0.0288,
+      "step": 304
+    },
+    {
+      "epoch": 0.4572713643178411,
+      "grad_norm": 0.10553280713408417,
+      "learning_rate": 0.00019835714708133862,
+      "loss": 0.0292,
+      "step": 305
+    },
+    {
+      "epoch": 0.4587706146926537,
+      "grad_norm": 0.15579620357045054,
+      "learning_rate": 0.0001983254907563955,
+      "loss": 0.0236,
+      "step": 306
+    },
+    {
+      "epoch": 0.46026986506746626,
+      "grad_norm": 0.09060539717469658,
+      "learning_rate": 0.00019829353491495545,
+      "loss": 0.0258,
+      "step": 307
+    },
+    {
+      "epoch": 0.4617691154422789,
+      "grad_norm": 0.12084292750432825,
+      "learning_rate": 0.00019826127965436152,
+      "loss": 0.0292,
+      "step": 308
+    },
+    {
+      "epoch": 0.46326836581709147,
+      "grad_norm": 0.20235716347011842,
+      "learning_rate": 0.0001982287250728689,
+      "loss": 0.0395,
+      "step": 309
+    },
+    {
+      "epoch": 0.46476761619190404,
+      "grad_norm": 0.09144766914600604,
+      "learning_rate": 0.00019819587126964437,
+      "loss": 0.0243,
+      "step": 310
+    },
+    {
+      "epoch": 0.4662668665667166,
+      "grad_norm": 0.1410307244411355,
+      "learning_rate": 0.00019816271834476642,
+      "loss": 0.0377,
+      "step": 311
+    },
+    {
+      "epoch": 0.46776611694152925,
+      "grad_norm": 0.12151655600598343,
+      "learning_rate": 0.0001981292663992245,
+      "loss": 0.0228,
+      "step": 312
+    },
+    {
+      "epoch": 0.46926536731634183,
+      "grad_norm": 0.13822001628609412,
+      "learning_rate": 0.00019809551553491916,
+      "loss": 0.031,
+      "step": 313
+    },
+    {
+      "epoch": 0.4707646176911544,
+      "grad_norm": 0.19923329717115837,
+      "learning_rate": 0.0001980614658546613,
+      "loss": 0.0542,
+      "step": 314
+    },
+    {
+      "epoch": 0.47226386806596704,
+      "grad_norm": 0.32163446242573723,
+      "learning_rate": 0.00019802711746217218,
+      "loss": 0.0399,
+      "step": 315
+    },
+    {
+      "epoch": 0.4737631184407796,
+      "grad_norm": 0.16137241975788913,
+      "learning_rate": 0.00019799247046208297,
+      "loss": 0.0366,
+      "step": 316
+    },
+    {
+      "epoch": 0.4752623688155922,
+      "grad_norm": 0.12091006379875421,
+      "learning_rate": 0.0001979575249599344,
+      "loss": 0.0246,
+      "step": 317
+    },
+    {
+      "epoch": 0.4767616191904048,
+      "grad_norm": 0.12233306737992392,
+      "learning_rate": 0.00019792228106217658,
+      "loss": 0.0341,
+      "step": 318
+    },
+    {
+      "epoch": 0.4782608695652174,
+      "grad_norm": 0.10780363270946153,
+      "learning_rate": 0.0001978867388761685,
+      "loss": 0.0259,
+      "step": 319
+    },
+    {
+      "epoch": 0.47976011994003,
+      "grad_norm": 0.18307588018838652,
+      "learning_rate": 0.00019785089851017787,
+      "loss": 0.0296,
+      "step": 320
+    },
+    {
+      "epoch": 0.48125937031484256,
+      "grad_norm": 0.2732266398896659,
+      "learning_rate": 0.00019781476007338058,
+      "loss": 0.0412,
+      "step": 321
+    },
+    {
+      "epoch": 0.4827586206896552,
+      "grad_norm": 0.1607986328742017,
+      "learning_rate": 0.00019777832367586063,
+      "loss": 0.044,
+      "step": 322
+    },
+    {
+      "epoch": 0.48425787106446777,
+      "grad_norm": 0.142980064491954,
+      "learning_rate": 0.0001977415894286096,
+      "loss": 0.0475,
+      "step": 323
+    },
+    {
+      "epoch": 0.48575712143928035,
+      "grad_norm": 0.06557574574115624,
+      "learning_rate": 0.0001977045574435264,
+      "loss": 0.0198,
+      "step": 324
+    },
+    {
+      "epoch": 0.487256371814093,
+      "grad_norm": 0.0839583522006225,
+      "learning_rate": 0.0001976672278334168,
+      "loss": 0.0232,
+      "step": 325
+    },
+    {
+      "epoch": 0.48875562218890556,
+      "grad_norm": 0.14621694058585588,
+      "learning_rate": 0.00019762960071199333,
+      "loss": 0.0355,
+      "step": 326
+    },
+    {
+      "epoch": 0.49025487256371814,
+      "grad_norm": 0.15402000439221136,
+      "learning_rate": 0.00019759167619387476,
+      "loss": 0.0274,
+      "step": 327
+    },
+    {
+      "epoch": 0.4917541229385307,
+      "grad_norm": 0.08261626886873759,
+      "learning_rate": 0.00019755345439458565,
+      "loss": 0.0285,
+      "step": 328
+    },
+    {
+      "epoch": 0.49325337331334335,
+      "grad_norm": 0.13478764449331881,
+      "learning_rate": 0.00019751493543055632,
+      "loss": 0.0311,
+      "step": 329
+    },
+    {
+      "epoch": 0.4947526236881559,
+      "grad_norm": 0.14610763557960335,
+      "learning_rate": 0.0001974761194191222,
+      "loss": 0.035,
+      "step": 330
+    },
+    {
+      "epoch": 0.4962518740629685,
+      "grad_norm": 0.1274425780126433,
+      "learning_rate": 0.00019743700647852354,
+      "loss": 0.0234,
+      "step": 331
+    },
+    {
+      "epoch": 0.49775112443778113,
+      "grad_norm": 0.09253891406497454,
+      "learning_rate": 0.0001973975967279052,
+      "loss": 0.0265,
+      "step": 332
+    },
+    {
+      "epoch": 0.4992503748125937,
+      "grad_norm": 0.07426081606342527,
+      "learning_rate": 0.00019735789028731604,
+      "loss": 0.0234,
+      "step": 333
+    },
+    {
+      "epoch": 0.5007496251874063,
+      "grad_norm": 0.19068637883417455,
+      "learning_rate": 0.00019731788727770885,
+      "loss": 0.0382,
+      "step": 334
+    },
+    {
+      "epoch": 0.5022488755622189,
+      "grad_norm": 0.11943906440034463,
+      "learning_rate": 0.00019727758782093967,
+      "loss": 0.0303,
+      "step": 335
+    },
+    {
+      "epoch": 0.5037481259370314,
+      "grad_norm": 0.08313883841221202,
+      "learning_rate": 0.00019723699203976766,
+      "loss": 0.0197,
+      "step": 336
+    },
+    {
+      "epoch": 0.5052473763118441,
+      "grad_norm": 0.1417916755841332,
+      "learning_rate": 0.00019719610005785465,
+      "loss": 0.0336,
+      "step": 337
+    },
+    {
+      "epoch": 0.5067466266866567,
+      "grad_norm": 0.11137030534188308,
+      "learning_rate": 0.0001971549119997646,
+      "loss": 0.03,
+      "step": 338
+    },
+    {
+      "epoch": 0.5082458770614693,
+      "grad_norm": 0.1474204159985949,
+      "learning_rate": 0.00019711342799096361,
+      "loss": 0.0305,
+      "step": 339
+    },
+    {
+      "epoch": 0.5097451274362819,
+      "grad_norm": 0.6073752811706395,
+      "learning_rate": 0.00019707164815781908,
+      "loss": 0.0512,
+      "step": 340
+    },
+    {
+      "epoch": 0.5112443778110944,
+      "grad_norm": 0.18317207850162331,
+      "learning_rate": 0.00019702957262759965,
+      "loss": 0.0244,
+      "step": 341
+    },
+    {
+      "epoch": 0.512743628185907,
+      "grad_norm": 0.14617080172691455,
+      "learning_rate": 0.00019698720152847468,
+      "loss": 0.0293,
+      "step": 342
+    },
+    {
+      "epoch": 0.5142428785607196,
+      "grad_norm": 0.21227866013123528,
+      "learning_rate": 0.0001969445349895139,
+      "loss": 0.0273,
+      "step": 343
+    },
+    {
+      "epoch": 0.5157421289355323,
+      "grad_norm": 0.11079359835152516,
+      "learning_rate": 0.00019690157314068696,
+      "loss": 0.02,
+      "step": 344
+    },
+    {
+      "epoch": 0.5172413793103449,
+      "grad_norm": 0.089040858251259,
+      "learning_rate": 0.0001968583161128631,
+      "loss": 0.0254,
+      "step": 345
+    },
+    {
+      "epoch": 0.5187406296851574,
+      "grad_norm": 0.22413723696972337,
+      "learning_rate": 0.0001968147640378108,
+      "loss": 0.0298,
+      "step": 346
+    },
+    {
+      "epoch": 0.52023988005997,
+      "grad_norm": 0.0875455049017176,
+      "learning_rate": 0.00019677091704819715,
+      "loss": 0.023,
+      "step": 347
+    },
+    {
+      "epoch": 0.5217391304347826,
+      "grad_norm": 0.10256718521877016,
+      "learning_rate": 0.0001967267752775877,
+      "loss": 0.0277,
+      "step": 348
+    },
+    {
+      "epoch": 0.5232383808095952,
+      "grad_norm": 0.17080254667328584,
+      "learning_rate": 0.00019668233886044597,
+      "loss": 0.0345,
+      "step": 349
+    },
+    {
+      "epoch": 0.5247376311844077,
+      "grad_norm": 0.1475350728620058,
+      "learning_rate": 0.00019663760793213296,
+      "loss": 0.0338,
+      "step": 350
+    },
+    {
+      "epoch": 0.5262368815592204,
+      "grad_norm": 0.10911209234569877,
+      "learning_rate": 0.00019659258262890683,
+      "loss": 0.0293,
+      "step": 351
+    },
+    {
+      "epoch": 0.527736131934033,
+      "grad_norm": 0.10321559635673279,
+      "learning_rate": 0.0001965472630879225,
+      "loss": 0.0289,
+      "step": 352
+    },
+    {
+      "epoch": 0.5292353823088456,
+      "grad_norm": 0.10787457610079623,
+      "learning_rate": 0.00019650164944723115,
+      "loss": 0.0361,
+      "step": 353
+    },
+    {
+      "epoch": 0.5307346326836582,
+      "grad_norm": 0.0837429373957743,
+      "learning_rate": 0.00019645574184577982,
+      "loss": 0.0212,
+      "step": 354
+    },
+    {
+      "epoch": 0.5322338830584707,
+      "grad_norm": 0.09945175724905725,
+      "learning_rate": 0.00019640954042341103,
+      "loss": 0.0172,
+      "step": 355
+    },
+    {
+      "epoch": 0.5337331334332833,
+      "grad_norm": 0.059734968422014076,
+      "learning_rate": 0.0001963630453208623,
+      "loss": 0.0204,
+      "step": 356
+    },
+    {
+      "epoch": 0.5352323838080959,
+      "grad_norm": 0.17974902185014147,
+      "learning_rate": 0.00019631625667976583,
+      "loss": 0.0358,
+      "step": 357
+    },
+    {
+      "epoch": 0.5367316341829086,
+      "grad_norm": 0.09827234190096189,
+      "learning_rate": 0.0001962691746426479,
+      "loss": 0.0215,
+      "step": 358
+    },
+    {
+      "epoch": 0.5382308845577212,
+      "grad_norm": 0.16005323730952734,
+      "learning_rate": 0.00019622179935292855,
+      "loss": 0.05,
+      "step": 359
+    },
+    {
+      "epoch": 0.5397301349325337,
+      "grad_norm": 0.1349661412175,
+      "learning_rate": 0.00019617413095492114,
+      "loss": 0.0352,
+      "step": 360
+    },
+    {
+      "epoch": 0.5412293853073463,
+      "grad_norm": 0.16328511581195324,
+      "learning_rate": 0.0001961261695938319,
+      "loss": 0.0397,
+      "step": 361
+    },
+    {
+      "epoch": 0.5427286356821589,
+      "grad_norm": 0.1751158166212692,
+      "learning_rate": 0.00019607791541575943,
+      "loss": 0.039,
+      "step": 362
+    },
+    {
+      "epoch": 0.5442278860569715,
+      "grad_norm": 0.14426941275516605,
+      "learning_rate": 0.0001960293685676943,
+      "loss": 0.0335,
+      "step": 363
+    },
+    {
+      "epoch": 0.545727136431784,
+      "grad_norm": 0.13144454729693233,
+      "learning_rate": 0.0001959805291975187,
+      "loss": 0.0307,
+      "step": 364
+    },
+    {
+      "epoch": 0.5472263868065967,
+      "grad_norm": 0.1025662618949345,
+      "learning_rate": 0.00019593139745400576,
+      "loss": 0.0257,
+      "step": 365
+    },
+    {
+      "epoch": 0.5487256371814093,
+      "grad_norm": 0.0980815537212099,
+      "learning_rate": 0.0001958819734868193,
+      "loss": 0.0307,
+      "step": 366
+    },
+    {
+      "epoch": 0.5502248875562219,
+      "grad_norm": 0.11890687937548734,
+      "learning_rate": 0.00019583225744651333,
+      "loss": 0.0217,
+      "step": 367
+    },
+    {
+      "epoch": 0.5517241379310345,
+      "grad_norm": 0.05424144596101746,
+      "learning_rate": 0.0001957822494845315,
+      "loss": 0.0119,
+      "step": 368
+    },
+    {
+      "epoch": 0.553223388305847,
+      "grad_norm": 0.35530516756748276,
+      "learning_rate": 0.00019573194975320673,
+      "loss": 0.0515,
+      "step": 369
+    },
+    {
+      "epoch": 0.5547226386806596,
+      "grad_norm": 0.084039953250112,
+      "learning_rate": 0.00019568135840576076,
+      "loss": 0.0222,
+      "step": 370
+    },
+    {
+      "epoch": 0.5562218890554723,
+      "grad_norm": 0.0719021205960788,
+      "learning_rate": 0.00019563047559630357,
+      "loss": 0.0129,
+      "step": 371
+    },
+    {
+      "epoch": 0.5577211394302849,
+      "grad_norm": 0.0831894222429869,
+      "learning_rate": 0.00019557930147983302,
+      "loss": 0.0203,
+      "step": 372
+    },
+    {
+      "epoch": 0.5592203898050975,
+      "grad_norm": 0.15329768574581998,
+      "learning_rate": 0.00019552783621223436,
+      "loss": 0.0395,
+      "step": 373
+    },
+    {
+      "epoch": 0.56071964017991,
+      "grad_norm": 0.2358016986495495,
+      "learning_rate": 0.00019547607995027978,
+      "loss": 0.1109,
+      "step": 374
+    },
+    {
+      "epoch": 0.5622188905547226,
+      "grad_norm": 0.12867019275964311,
+      "learning_rate": 0.0001954240328516277,
+      "loss": 0.0297,
+      "step": 375
+    },
+    {
+      "epoch": 0.5637181409295352,
+      "grad_norm": 0.29639255369886375,
+      "learning_rate": 0.0001953716950748227,
+      "loss": 0.0372,
+      "step": 376
+    },
+    {
+      "epoch": 0.5652173913043478,
+      "grad_norm": 0.11159538070632506,
+      "learning_rate": 0.0001953190667792947,
+      "loss": 0.0229,
+      "step": 377
+    },
+    {
+      "epoch": 0.5667166416791605,
+      "grad_norm": 0.21855748359403565,
+      "learning_rate": 0.00019526614812535864,
+      "loss": 0.0608,
+      "step": 378
+    },
+    {
+      "epoch": 0.568215892053973,
+      "grad_norm": 0.07792574697692431,
+      "learning_rate": 0.00019521293927421388,
+      "loss": 0.0178,
+      "step": 379
+    },
+    {
+      "epoch": 0.5697151424287856,
+      "grad_norm": 0.19263126390415305,
+      "learning_rate": 0.00019515944038794384,
+      "loss": 0.0326,
+      "step": 380
+    },
+    {
+      "epoch": 0.5712143928035982,
+      "grad_norm": 0.08384659637263117,
+      "learning_rate": 0.00019510565162951537,
+      "loss": 0.02,
+      "step": 381
+    },
+    {
+      "epoch": 0.5727136431784108,
+      "grad_norm": 0.1666776601307308,
+      "learning_rate": 0.00019505157316277837,
+      "loss": 0.0371,
+      "step": 382
+    },
+    {
+      "epoch": 0.5742128935532234,
+      "grad_norm": 0.07115309715230671,
+      "learning_rate": 0.00019499720515246525,
+      "loss": 0.0161,
+      "step": 383
+    },
+    {
+      "epoch": 0.5757121439280359,
+      "grad_norm": 0.189506645403256,
+      "learning_rate": 0.0001949425477641904,
+      "loss": 0.0385,
+      "step": 384
+    },
+    {
+      "epoch": 0.5772113943028486,
+      "grad_norm": 0.20463849444352333,
+      "learning_rate": 0.00019488760116444966,
+      "loss": 0.0334,
+      "step": 385
+    },
+    {
+      "epoch": 0.5787106446776612,
+      "grad_norm": 0.14154099501140546,
+      "learning_rate": 0.00019483236552061994,
+      "loss": 0.047,
+      "step": 386
+    },
+    {
+      "epoch": 0.5802098950524738,
+      "grad_norm": 0.14873654543331344,
+      "learning_rate": 0.0001947768410009586,
+      "loss": 0.0367,
+      "step": 387
+    },
+    {
+      "epoch": 0.5817091454272864,
+      "grad_norm": 0.12840825492818217,
+      "learning_rate": 0.0001947210277746029,
+      "loss": 0.0292,
+      "step": 388
+    },
+    {
+      "epoch": 0.5832083958020989,
+      "grad_norm": 0.23903676877323182,
+      "learning_rate": 0.00019466492601156966,
+      "loss": 0.0379,
+      "step": 389
+    },
+    {
+      "epoch": 0.5847076461769115,
+      "grad_norm": 0.15724874073609016,
+      "learning_rate": 0.00019460853588275454,
+      "loss": 0.0282,
+      "step": 390
+    },
+    {
+      "epoch": 0.5862068965517241,
+      "grad_norm": 0.13173093891442925,
+      "learning_rate": 0.0001945518575599317,
+      "loss": 0.0316,
+      "step": 391
+    },
+    {
+      "epoch": 0.5877061469265368,
+      "grad_norm": 0.28666760943594705,
+      "learning_rate": 0.0001944948912157531,
+      "loss": 0.0197,
+      "step": 392
+    },
+    {
+      "epoch": 0.5892053973013494,
+      "grad_norm": 0.1832605963942488,
+      "learning_rate": 0.00019443763702374812,
+      "loss": 0.0266,
+      "step": 393
+    },
+    {
+      "epoch": 0.5907046476761619,
+      "grad_norm": 0.13319197003396382,
+      "learning_rate": 0.00019438009515832297,
+      "loss": 0.0374,
+      "step": 394
+    },
+    {
+      "epoch": 0.5922038980509745,
+      "grad_norm": 0.0937432546642976,
+      "learning_rate": 0.0001943222657947601,
+      "loss": 0.0302,
+      "step": 395
+    },
+    {
+      "epoch": 0.5937031484257871,
+      "grad_norm": 0.0791127172795629,
+      "learning_rate": 0.00019426414910921787,
+      "loss": 0.0258,
+      "step": 396
+    },
+    {
+      "epoch": 0.5952023988005997,
+      "grad_norm": 0.11621842629982067,
+      "learning_rate": 0.00019420574527872968,
+      "loss": 0.0224,
+      "step": 397
+    },
+    {
+      "epoch": 0.5967016491754122,
+      "grad_norm": 0.18111252384023568,
+      "learning_rate": 0.0001941470544812038,
+      "loss": 0.0343,
+      "step": 398
+    },
+    {
+      "epoch": 0.5982008995502249,
+      "grad_norm": 0.12668476963938463,
+      "learning_rate": 0.00019408807689542257,
+      "loss": 0.0329,
+      "step": 399
+    },
+    {
+      "epoch": 0.5997001499250375,
+      "grad_norm": 0.08125543588426189,
+      "learning_rate": 0.0001940288127010419,
+      "loss": 0.0187,
+      "step": 400
+    },
+    {
+      "epoch": 0.6011994002998501,
+      "grad_norm": 0.12705282798340953,
+      "learning_rate": 0.00019396926207859084,
+      "loss": 0.0208,
+      "step": 401
+    },
+    {
+      "epoch": 0.6026986506746627,
+      "grad_norm": 0.09903543853284616,
+      "learning_rate": 0.0001939094252094709,
+      "loss": 0.0203,
+      "step": 402
+    },
+    {
+      "epoch": 0.6041979010494752,
+      "grad_norm": 0.09458490298044545,
+      "learning_rate": 0.0001938493022759556,
+      "loss": 0.0354,
+      "step": 403
+    },
+    {
+      "epoch": 0.6056971514242878,
+      "grad_norm": 0.13799640097610966,
+      "learning_rate": 0.0001937888934611898,
+      "loss": 0.0305,
+      "step": 404
+    },
+    {
+      "epoch": 0.6071964017991005,
+      "grad_norm": 0.10327912926042261,
+      "learning_rate": 0.00019372819894918915,
+      "loss": 0.0193,
+      "step": 405
+    },
+    {
+      "epoch": 0.6086956521739131,
+      "grad_norm": 0.12134046776724058,
+      "learning_rate": 0.00019366721892483978,
+      "loss": 0.0289,
+      "step": 406
+    },
+    {
+      "epoch": 0.6101949025487257,
+      "grad_norm": 0.3625741631498206,
+      "learning_rate": 0.00019360595357389735,
+      "loss": 0.0618,
+      "step": 407
+    },
+    {
+      "epoch": 0.6116941529235382,
+      "grad_norm": 0.16276585017367157,
+      "learning_rate": 0.00019354440308298675,
+      "loss": 0.033,
+      "step": 408
+    },
+    {
+      "epoch": 0.6131934032983508,
+      "grad_norm": 0.1337810168549044,
+      "learning_rate": 0.00019348256763960145,
+      "loss": 0.0334,
+      "step": 409
+    },
+    {
+      "epoch": 0.6146926536731634,
+      "grad_norm": 0.46538099581190734,
+      "learning_rate": 0.00019342044743210295,
+      "loss": 0.063,
+      "step": 410
+    },
+    {
+      "epoch": 0.616191904047976,
+      "grad_norm": 0.1528443799832324,
+      "learning_rate": 0.00019335804264972018,
+      "loss": 0.0419,
+      "step": 411
+    },
+    {
+      "epoch": 0.6176911544227887,
+      "grad_norm": 0.11360241608381239,
+      "learning_rate": 0.00019329535348254893,
+      "loss": 0.0279,
+      "step": 412
+    },
+    {
+      "epoch": 0.6191904047976012,
+      "grad_norm": 0.10731049141357688,
+      "learning_rate": 0.00019323238012155123,
+      "loss": 0.027,
+      "step": 413
+    },
+    {
+      "epoch": 0.6206896551724138,
+      "grad_norm": 0.1197495553442816,
+      "learning_rate": 0.0001931691227585549,
+      "loss": 0.034,
+      "step": 414
+    },
+    {
+      "epoch": 0.6221889055472264,
+      "grad_norm": 0.15775016023781924,
+      "learning_rate": 0.00019310558158625285,
+      "loss": 0.0377,
+      "step": 415
+    },
+    {
+      "epoch": 0.623688155922039,
+      "grad_norm": 0.11263624033418991,
+      "learning_rate": 0.00019304175679820247,
+      "loss": 0.0185,
+      "step": 416
+    },
+    {
+      "epoch": 0.6251874062968515,
+      "grad_norm": 0.1306077390581805,
+      "learning_rate": 0.00019297764858882514,
+      "loss": 0.0207,
+      "step": 417
+    },
+    {
+      "epoch": 0.6266866566716641,
+      "grad_norm": 0.09670200550131532,
+      "learning_rate": 0.00019291325715340563,
+      "loss": 0.0219,
+      "step": 418
+    },
+    {
+      "epoch": 0.6281859070464768,
+      "grad_norm": 0.08461352549162458,
+      "learning_rate": 0.00019284858268809137,
+      "loss": 0.0305,
+      "step": 419
+    },
+    {
+      "epoch": 0.6296851574212894,
+      "grad_norm": 0.1169472728087451,
+      "learning_rate": 0.000192783625389892,
+      "loss": 0.0351,
+      "step": 420
+    },
+    {
+      "epoch": 0.631184407796102,
+      "grad_norm": 0.3408290216906546,
+      "learning_rate": 0.00019271838545667876,
+      "loss": 0.0312,
+      "step": 421
+    },
+    {
+      "epoch": 0.6326836581709145,
+      "grad_norm": 0.09856486201168368,
+      "learning_rate": 0.00019265286308718372,
+      "loss": 0.023,
+      "step": 422
+    },
+    {
+      "epoch": 0.6341829085457271,
+      "grad_norm": 0.11853262613762018,
+      "learning_rate": 0.0001925870584809995,
+      "loss": 0.0219,
+      "step": 423
+    },
+    {
+      "epoch": 0.6356821589205397,
+      "grad_norm": 0.12277530671201359,
+      "learning_rate": 0.00019252097183857823,
+      "loss": 0.026,
+      "step": 424
+    },
+    {
+      "epoch": 0.6371814092953523,
+      "grad_norm": 0.2205369312743012,
+      "learning_rate": 0.00019245460336123134,
+      "loss": 0.046,
+      "step": 425
+    },
+    {
+      "epoch": 0.638680659670165,
+      "grad_norm": 0.10222524463832733,
+      "learning_rate": 0.0001923879532511287,
+      "loss": 0.0272,
+      "step": 426
+    },
+    {
+      "epoch": 0.6401799100449775,
+      "grad_norm": 0.1881232227964226,
+      "learning_rate": 0.00019232102171129811,
+      "loss": 0.0337,
+      "step": 427
+    },
+    {
+      "epoch": 0.6416791604197901,
+      "grad_norm": 0.2542609797286416,
+      "learning_rate": 0.00019225380894562463,
+      "loss": 0.0462,
+      "step": 428
+    },
+    {
+      "epoch": 0.6431784107946027,
+      "grad_norm": 0.15567631088633233,
+      "learning_rate": 0.00019218631515885006,
+      "loss": 0.032,
+      "step": 429
+    },
+    {
+      "epoch": 0.6446776611694153,
+      "grad_norm": 0.2672438162396934,
+      "learning_rate": 0.00019211854055657215,
+      "loss": 0.0477,
+      "step": 430
+    },
+    {
+      "epoch": 0.6461769115442278,
+      "grad_norm": 0.1926377684668028,
+      "learning_rate": 0.00019205048534524406,
+      "loss": 0.0419,
+      "step": 431
+    },
+    {
+      "epoch": 0.6476761619190404,
+      "grad_norm": 0.15284392892312818,
+      "learning_rate": 0.00019198214973217378,
+      "loss": 0.0398,
+      "step": 432
+    },
+    {
+      "epoch": 0.6491754122938531,
+      "grad_norm": 0.11044006384885063,
+      "learning_rate": 0.00019191353392552344,
+      "loss": 0.0288,
+      "step": 433
+    },
+    {
+      "epoch": 0.6506746626686657,
+      "grad_norm": 0.06779580847579911,
+      "learning_rate": 0.00019184463813430873,
+      "loss": 0.0191,
+      "step": 434
+    },
+    {
+      "epoch": 0.6521739130434783,
+      "grad_norm": 0.11624950451575107,
+      "learning_rate": 0.00019177546256839812,
+      "loss": 0.0312,
+      "step": 435
+    },
+    {
+      "epoch": 0.6536731634182908,
+      "grad_norm": 0.09177369083814481,
+      "learning_rate": 0.0001917060074385124,
+      "loss": 0.0303,
+      "step": 436
+    },
+    {
+      "epoch": 0.6551724137931034,
+      "grad_norm": 0.11556801968896689,
+      "learning_rate": 0.00019163627295622397,
+      "loss": 0.0223,
+      "step": 437
+    },
+    {
+      "epoch": 0.656671664167916,
+      "grad_norm": 0.09772753082703028,
+      "learning_rate": 0.00019156625933395614,
+      "loss": 0.0262,
+      "step": 438
+    },
+    {
+      "epoch": 0.6581709145427287,
+      "grad_norm": 0.05846912894146479,
+      "learning_rate": 0.0001914959667849825,
+      "loss": 0.0157,
+      "step": 439
+    },
+    {
+      "epoch": 0.6596701649175413,
+      "grad_norm": 0.6528373118339207,
+      "learning_rate": 0.00019142539552342638,
+      "loss": 0.0269,
+      "step": 440
+    },
+    {
+      "epoch": 0.6611694152923538,
+      "grad_norm": 0.0927370129280204,
+      "learning_rate": 0.0001913545457642601,
+      "loss": 0.0202,
+      "step": 441
+    },
+    {
+      "epoch": 0.6626686656671664,
+      "grad_norm": 0.1825329305030154,
+      "learning_rate": 0.0001912834177233043,
+      "loss": 0.0403,
+      "step": 442
+    },
+    {
+      "epoch": 0.664167916041979,
+      "grad_norm": 0.1740152075571954,
+      "learning_rate": 0.0001912120116172273,
+      "loss": 0.0491,
+      "step": 443
+    },
+    {
+      "epoch": 0.6656671664167916,
+      "grad_norm": 0.14268776420650997,
+      "learning_rate": 0.00019114032766354453,
+      "loss": 0.0375,
+      "step": 444
+    },
+    {
+      "epoch": 0.6671664167916042,
+      "grad_norm": 0.10780160091521133,
+      "learning_rate": 0.00019106836608061772,
+      "loss": 0.0327,
+      "step": 445
+    },
+    {
+      "epoch": 0.6686656671664168,
+      "grad_norm": 0.11310647498952432,
+      "learning_rate": 0.00019099612708765434,
+      "loss": 0.0299,
+      "step": 446
+    },
+    {
+      "epoch": 0.6701649175412294,
+      "grad_norm": 0.13740431187575763,
+      "learning_rate": 0.00019092361090470688,
+      "loss": 0.0358,
+      "step": 447
+    },
+    {
+      "epoch": 0.671664167916042,
+      "grad_norm": 0.18068397863290914,
+      "learning_rate": 0.0001908508177526722,
+      "loss": 0.0457,
+      "step": 448
+    },
+    {
+      "epoch": 0.6731634182908546,
+      "grad_norm": 0.10147402022124917,
+      "learning_rate": 0.00019077774785329087,
+      "loss": 0.0245,
+      "step": 449
+    },
+    {
+      "epoch": 0.6746626686656672,
+      "grad_norm": 0.17659307041157327,
+      "learning_rate": 0.0001907044014291465,
+      "loss": 0.0373,
+      "step": 450
+    },
+    {
+      "epoch": 0.6761619190404797,
+      "grad_norm": 0.21933608649614408,
+      "learning_rate": 0.000190630778703665,
+      "loss": 0.0724,
+      "step": 451
+    },
+    {
+      "epoch": 0.6776611694152923,
+      "grad_norm": 0.18695425514273356,
+      "learning_rate": 0.00019055687990111398,
+      "loss": 0.0433,
+      "step": 452
+    },
+    {
+      "epoch": 0.679160419790105,
+      "grad_norm": 0.09732659308557562,
+      "learning_rate": 0.00019048270524660196,
+      "loss": 0.0362,
+      "step": 453
+    },
+    {
+      "epoch": 0.6806596701649176,
+      "grad_norm": 0.08889244620347342,
+      "learning_rate": 0.00019040825496607786,
+      "loss": 0.0268,
+      "step": 454
+    },
+    {
+      "epoch": 0.6821589205397302,
+      "grad_norm": 0.10073399312054154,
+      "learning_rate": 0.0001903335292863301,
+      "loss": 0.0256,
+      "step": 455
+    },
+    {
+      "epoch": 0.6836581709145427,
+      "grad_norm": 0.17468190137144324,
+      "learning_rate": 0.00019025852843498607,
+      "loss": 0.0445,
+      "step": 456
+    },
+    {
+      "epoch": 0.6851574212893553,
+      "grad_norm": 0.09001197181665302,
+      "learning_rate": 0.0001901832526405114,
+      "loss": 0.0151,
+      "step": 457
+    },
+    {
+      "epoch": 0.6866566716641679,
+      "grad_norm": 0.13748748434736058,
+      "learning_rate": 0.00019010770213220916,
+      "loss": 0.0221,
+      "step": 458
+    },
+    {
+      "epoch": 0.6881559220389805,
+      "grad_norm": 0.13324313496496254,
+      "learning_rate": 0.00019003187714021938,
+      "loss": 0.0394,
+      "step": 459
+    },
+    {
+      "epoch": 0.6896551724137931,
+      "grad_norm": 0.18122876203058136,
+      "learning_rate": 0.00018995577789551803,
+      "loss": 0.0438,
+      "step": 460
+    },
+    {
+      "epoch": 0.6911544227886057,
+      "grad_norm": 0.13914037277459534,
+      "learning_rate": 0.0001898794046299167,
+      "loss": 0.0325,
+      "step": 461
+    },
+    {
+      "epoch": 0.6926536731634183,
+      "grad_norm": 0.1250826760287831,
+      "learning_rate": 0.00018980275757606157,
+      "loss": 0.0324,
+      "step": 462
+    },
+    {
+      "epoch": 0.6941529235382309,
+      "grad_norm": 0.11773650752390566,
+      "learning_rate": 0.00018972583696743285,
+      "loss": 0.0298,
+      "step": 463
+    },
+    {
+      "epoch": 0.6956521739130435,
+      "grad_norm": 0.20782604564608717,
+      "learning_rate": 0.00018964864303834406,
+      "loss": 0.0281,
+      "step": 464
+    },
+    {
+      "epoch": 0.697151424287856,
+      "grad_norm": 0.09525179474823385,
+      "learning_rate": 0.0001895711760239413,
+      "loss": 0.0252,
+      "step": 465
+    },
+    {
+      "epoch": 0.6986506746626686,
+      "grad_norm": 0.1240832012888324,
+      "learning_rate": 0.00018949343616020252,
+      "loss": 0.0231,
+      "step": 466
+    },
+    {
+      "epoch": 0.7001499250374813,
+      "grad_norm": 0.4265753911477756,
+      "learning_rate": 0.0001894154236839368,
+      "loss": 0.0669,
+      "step": 467
+    },
+    {
+      "epoch": 0.7016491754122939,
+      "grad_norm": 0.12476523044968328,
+      "learning_rate": 0.00018933713883278376,
+      "loss": 0.0273,
+      "step": 468
+    },
+    {
+      "epoch": 0.7031484257871065,
+      "grad_norm": 0.14830273846419462,
+      "learning_rate": 0.00018925858184521256,
+      "loss": 0.0303,
+      "step": 469
+    },
+    {
+      "epoch": 0.704647676161919,
+      "grad_norm": 0.12142464965135223,
+      "learning_rate": 0.00018917975296052142,
+      "loss": 0.0387,
+      "step": 470
+    },
+    {
+      "epoch": 0.7061469265367316,
+      "grad_norm": 0.12543262855683845,
+      "learning_rate": 0.0001891006524188368,
+      "loss": 0.0251,
+      "step": 471
+    },
+    {
+      "epoch": 0.7076461769115442,
+      "grad_norm": 0.19339819413209244,
+      "learning_rate": 0.00018902128046111266,
+      "loss": 0.0257,
+      "step": 472
+    },
+    {
+      "epoch": 0.7091454272863568,
+      "grad_norm": 0.15350744131356941,
+      "learning_rate": 0.00018894163732912977,
+      "loss": 0.0278,
+      "step": 473
+    },
+    {
+      "epoch": 0.7106446776611695,
+      "grad_norm": 0.23391665071790627,
+      "learning_rate": 0.0001888617232654949,
+      "loss": 0.0448,
+      "step": 474
+    },
+    {
+      "epoch": 0.712143928035982,
+      "grad_norm": 0.10994980391292404,
+      "learning_rate": 0.00018878153851364013,
+      "loss": 0.0227,
+      "step": 475
+    },
+    {
+      "epoch": 0.7136431784107946,
+      "grad_norm": 0.10665227644215766,
+      "learning_rate": 0.00018870108331782217,
+      "loss": 0.0281,
+      "step": 476
+    },
+    {
+      "epoch": 0.7151424287856072,
+      "grad_norm": 0.14531593463069997,
+      "learning_rate": 0.00018862035792312147,
+      "loss": 0.0326,
+      "step": 477
+    },
+    {
+      "epoch": 0.7166416791604198,
+      "grad_norm": 0.1182913114959111,
+      "learning_rate": 0.0001885393625754416,
+      "loss": 0.0188,
+      "step": 478
+    },
+    {
+      "epoch": 0.7181409295352323,
+      "grad_norm": 0.07106508169769966,
+      "learning_rate": 0.0001884580975215084,
+      "loss": 0.0138,
+      "step": 479
+    },
+    {
+      "epoch": 0.719640179910045,
+      "grad_norm": 0.07950185634897355,
+      "learning_rate": 0.00018837656300886937,
+      "loss": 0.0165,
+      "step": 480
+    },
+    {
+      "epoch": 0.7211394302848576,
+      "grad_norm": 0.09129244683295917,
+      "learning_rate": 0.00018829475928589271,
+      "loss": 0.0167,
+      "step": 481
+    },
+    {
+      "epoch": 0.7226386806596702,
+      "grad_norm": 0.12771843853475826,
+      "learning_rate": 0.00018821268660176678,
+      "loss": 0.022,
+      "step": 482
+    },
+    {
+      "epoch": 0.7241379310344828,
+      "grad_norm": 0.13265893909135285,
+      "learning_rate": 0.0001881303452064992,
+      "loss": 0.0345,
+      "step": 483
+    },
+    {
+      "epoch": 0.7256371814092953,
+      "grad_norm": 0.12918348541457214,
+      "learning_rate": 0.0001880477353509162,
+      "loss": 0.0372,
+      "step": 484
+    },
+    {
+      "epoch": 0.7271364317841079,
+      "grad_norm": 0.09841614043671974,
+      "learning_rate": 0.00018796485728666165,
+      "loss": 0.0214,
+      "step": 485
+    },
+    {
+      "epoch": 0.7286356821589205,
+      "grad_norm": 0.12613750913388452,
+      "learning_rate": 0.00018788171126619653,
+      "loss": 0.0272,
+      "step": 486
+    },
+    {
+      "epoch": 0.7301349325337332,
+      "grad_norm": 0.10042055414075088,
+      "learning_rate": 0.00018779829754279805,
+      "loss": 0.0286,
+      "step": 487
+    },
+    {
+      "epoch": 0.7316341829085458,
+      "grad_norm": 0.1665493001660504,
+      "learning_rate": 0.00018771461637055888,
+      "loss": 0.033,
+      "step": 488
+    },
+    {
+      "epoch": 0.7331334332833583,
+      "grad_norm": 0.1179355108298341,
+      "learning_rate": 0.00018763066800438636,
+      "loss": 0.0463,
+      "step": 489
+    },
+    {
+      "epoch": 0.7346326836581709,
+      "grad_norm": 0.10791732823530058,
+      "learning_rate": 0.0001875464527000018,
+      "loss": 0.0217,
+      "step": 490
+    },
+    {
+      "epoch": 0.7361319340329835,
+      "grad_norm": 0.3669927380669368,
+      "learning_rate": 0.00018746197071393958,
+      "loss": 0.0457,
+      "step": 491
+    },
+    {
+      "epoch": 0.7376311844077961,
+      "grad_norm": 0.1987097980139362,
+      "learning_rate": 0.00018737722230354655,
+      "loss": 0.0423,
+      "step": 492
+    },
+    {
+      "epoch": 0.7391304347826086,
+      "grad_norm": 0.09040003852732766,
+      "learning_rate": 0.00018729220772698097,
+      "loss": 0.0288,
+      "step": 493
+    },
+    {
+      "epoch": 0.7406296851574213,
+      "grad_norm": 0.20269069942355697,
+      "learning_rate": 0.00018720692724321207,
+      "loss": 0.0455,
+      "step": 494
+    },
+    {
+      "epoch": 0.7421289355322339,
+      "grad_norm": 0.1372255943190446,
+      "learning_rate": 0.00018712138111201895,
+      "loss": 0.0216,
+      "step": 495
+    },
+    {
+      "epoch": 0.7436281859070465,
+      "grad_norm": 0.23882692775642725,
+      "learning_rate": 0.00018703556959398998,
+      "loss": 0.0487,
+      "step": 496
+    },
+    {
+      "epoch": 0.7451274362818591,
+      "grad_norm": 0.10854063954832008,
+      "learning_rate": 0.0001869494929505219,
+      "loss": 0.0222,
+      "step": 497
+    },
+    {
+      "epoch": 0.7466266866566716,
+      "grad_norm": 0.09480222600585039,
+      "learning_rate": 0.00018686315144381913,
+      "loss": 0.0257,
+      "step": 498
+    },
+    {
+      "epoch": 0.7481259370314842,
+      "grad_norm": 0.16193331192920116,
+      "learning_rate": 0.00018677654533689287,
+      "loss": 0.0329,
+      "step": 499
+    },
+    {
+      "epoch": 0.7496251874062968,
+      "grad_norm": 0.08005281597137838,
+      "learning_rate": 0.00018668967489356028,
+      "loss": 0.0169,
+      "step": 500
+    },
+    {
+      "epoch": 0.7511244377811095,
+      "grad_norm": 0.12389056426209798,
+      "learning_rate": 0.00018660254037844388,
+      "loss": 0.0366,
+      "step": 501
+    },
+    {
+      "epoch": 0.7526236881559221,
+      "grad_norm": 0.11869938873623041,
+      "learning_rate": 0.00018651514205697046,
+      "loss": 0.0321,
+      "step": 502
+    },
+    {
+      "epoch": 0.7541229385307346,
+      "grad_norm": 0.0930246262254623,
+      "learning_rate": 0.0001864274801953705,
+      "loss": 0.022,
+      "step": 503
+    },
+    {
+      "epoch": 0.7556221889055472,
+      "grad_norm": 0.27274941895288046,
+      "learning_rate": 0.00018633955506067718,
+      "loss": 0.045,
+      "step": 504
+    },
+    {
+      "epoch": 0.7571214392803598,
+      "grad_norm": 0.0726365498076487,
+      "learning_rate": 0.00018625136692072575,
+      "loss": 0.0129,
+      "step": 505
+    },
+    {
+      "epoch": 0.7586206896551724,
+      "grad_norm": 0.1638000705309204,
+      "learning_rate": 0.00018616291604415258,
+      "loss": 0.0507,
+      "step": 506
+    },
+    {
+      "epoch": 0.760119940029985,
+      "grad_norm": 0.10985380636797697,
+      "learning_rate": 0.0001860742027003944,
+      "loss": 0.0172,
+      "step": 507
+    },
+    {
+      "epoch": 0.7616191904047976,
+      "grad_norm": 0.11572389017878003,
+      "learning_rate": 0.00018598522715968736,
+      "loss": 0.0359,
+      "step": 508
+    },
+    {
+      "epoch": 0.7631184407796102,
+      "grad_norm": 0.1957062355312362,
+      "learning_rate": 0.00018589598969306645,
+      "loss": 0.0286,
+      "step": 509
+    },
+    {
+      "epoch": 0.7646176911544228,
+      "grad_norm": 0.1797311903894293,
+      "learning_rate": 0.00018580649057236447,
+      "loss": 0.0422,
+      "step": 510
+    },
+    {
+      "epoch": 0.7661169415292354,
+      "grad_norm": 0.10855777727429075,
+      "learning_rate": 0.00018571673007021123,
+      "loss": 0.0348,
+      "step": 511
+    },
+    {
+      "epoch": 0.767616191904048,
+      "grad_norm": 0.09640794818380023,
+      "learning_rate": 0.00018562670846003284,
+      "loss": 0.0123,
+      "step": 512
+    },
+    {
+      "epoch": 0.7691154422788605,
+      "grad_norm": 0.21945784704110113,
+      "learning_rate": 0.00018553642601605068,
+      "loss": 0.0456,
+      "step": 513
+    },
+    {
+      "epoch": 0.7706146926536732,
+      "grad_norm": 0.11478499434793961,
+      "learning_rate": 0.00018544588301328075,
+      "loss": 0.0225,
+      "step": 514
+    },
+    {
+      "epoch": 0.7721139430284858,
+      "grad_norm": 0.10420495356687196,
+      "learning_rate": 0.00018535507972753274,
+      "loss": 0.0276,
+      "step": 515
+    },
+    {
+      "epoch": 0.7736131934032984,
+      "grad_norm": 0.06375546661931226,
+      "learning_rate": 0.00018526401643540922,
+      "loss": 0.0143,
+      "step": 516
+    },
+    {
+      "epoch": 0.775112443778111,
+      "grad_norm": 0.10013660059835842,
+      "learning_rate": 0.00018517269341430476,
+      "loss": 0.0316,
+      "step": 517
+    },
+    {
+      "epoch": 0.7766116941529235,
+      "grad_norm": 0.1525040242862086,
+      "learning_rate": 0.00018508111094240514,
+      "loss": 0.0531,
+      "step": 518
+    },
+    {
+      "epoch": 0.7781109445277361,
+      "grad_norm": 0.08246402484447012,
+      "learning_rate": 0.00018498926929868642,
+      "loss": 0.0232,
+      "step": 519
+    },
+    {
+      "epoch": 0.7796101949025487,
+      "grad_norm": 0.1147034669265538,
+      "learning_rate": 0.00018489716876291415,
+      "loss": 0.0282,
+      "step": 520
+    },
+    {
+      "epoch": 0.7811094452773614,
+      "grad_norm": 0.12214564867021378,
+      "learning_rate": 0.0001848048096156426,
+      "loss": 0.0247,
+      "step": 521
+    },
+    {
+      "epoch": 0.782608695652174,
+      "grad_norm": 0.06751036662935833,
+      "learning_rate": 0.00018471219213821375,
+      "loss": 0.0152,
+      "step": 522
+    },
+    {
+      "epoch": 0.7841079460269865,
+      "grad_norm": 0.07585774273233134,
+      "learning_rate": 0.00018461931661275643,
+      "loss": 0.0207,
+      "step": 523
+    },
+    {
+      "epoch": 0.7856071964017991,
+      "grad_norm": 0.10748840649212606,
+      "learning_rate": 0.00018452618332218563,
+      "loss": 0.0254,
+      "step": 524
+    },
+    {
+      "epoch": 0.7871064467766117,
+      "grad_norm": 0.10237085474160945,
+      "learning_rate": 0.00018443279255020152,
+      "loss": 0.0329,
+      "step": 525
+    },
+    {
+      "epoch": 0.7886056971514243,
+      "grad_norm": 0.08837009279940057,
+      "learning_rate": 0.0001843391445812886,
+      "loss": 0.0199,
+      "step": 526
+    },
+    {
+      "epoch": 0.7901049475262368,
+      "grad_norm": 0.06271891827693078,
+      "learning_rate": 0.00018424523970071477,
+      "loss": 0.0183,
+      "step": 527
+    },
+    {
+      "epoch": 0.7916041979010495,
+      "grad_norm": 0.13599180635432398,
+      "learning_rate": 0.00018415107819453062,
+      "loss": 0.0447,
+      "step": 528
+    },
+    {
+      "epoch": 0.7931034482758621,
+      "grad_norm": 0.0836743427473769,
+      "learning_rate": 0.00018405666034956844,
+      "loss": 0.0175,
+      "step": 529
+    },
+    {
+      "epoch": 0.7946026986506747,
+      "grad_norm": 0.11593648054445677,
+      "learning_rate": 0.00018396198645344135,
+      "loss": 0.0246,
+      "step": 530
+    },
+    {
+      "epoch": 0.7961019490254873,
+      "grad_norm": 0.07719300987275271,
+      "learning_rate": 0.00018386705679454242,
+      "loss": 0.0301,
+      "step": 531
+    },
+    {
+      "epoch": 0.7976011994002998,
+      "grad_norm": 0.09778571203259609,
+      "learning_rate": 0.0001837718716620439,
+      "loss": 0.0212,
+      "step": 532
+    },
+    {
+      "epoch": 0.7991004497751124,
+      "grad_norm": 0.12595595693675168,
+      "learning_rate": 0.00018367643134589617,
+      "loss": 0.0326,
+      "step": 533
+    },
+    {
+      "epoch": 0.800599700149925,
+      "grad_norm": 0.07156209728774574,
+      "learning_rate": 0.00018358073613682706,
+      "loss": 0.0165,
+      "step": 534
+    },
+    {
+      "epoch": 0.8020989505247377,
+      "grad_norm": 0.09505001360014041,
+      "learning_rate": 0.00018348478632634066,
+      "loss": 0.0187,
+      "step": 535
+    },
+    {
+      "epoch": 0.8035982008995503,
+      "grad_norm": 0.12058029362246589,
+      "learning_rate": 0.00018338858220671682,
+      "loss": 0.0282,
+      "step": 536
+    },
+    {
+      "epoch": 0.8050974512743628,
+      "grad_norm": 0.17693735284669823,
+      "learning_rate": 0.00018329212407100994,
+      "loss": 0.0324,
+      "step": 537
+    },
+    {
+      "epoch": 0.8065967016491754,
+      "grad_norm": 0.13029106214616978,
+      "learning_rate": 0.00018319541221304827,
+      "loss": 0.0209,
+      "step": 538
+    },
+    {
+      "epoch": 0.808095952023988,
+      "grad_norm": 0.1162879265464023,
+      "learning_rate": 0.00018309844692743283,
+      "loss": 0.0207,
+      "step": 539
+    },
+    {
+      "epoch": 0.8095952023988006,
+      "grad_norm": 0.20317081990377658,
+      "learning_rate": 0.00018300122850953675,
+      "loss": 0.0227,
+      "step": 540
+    },
+    {
+      "epoch": 0.8110944527736131,
+      "grad_norm": 0.19653799373191536,
+      "learning_rate": 0.00018290375725550417,
+      "loss": 0.0281,
+      "step": 541
+    },
+    {
+      "epoch": 0.8125937031484258,
+      "grad_norm": 0.11236149669329604,
+      "learning_rate": 0.00018280603346224945,
+      "loss": 0.0216,
+      "step": 542
+    },
+    {
+      "epoch": 0.8140929535232384,
+      "grad_norm": 0.1687101239767433,
+      "learning_rate": 0.00018270805742745617,
+      "loss": 0.0405,
+      "step": 543
+    },
+    {
+      "epoch": 0.815592203898051,
+      "grad_norm": 0.16013841907480525,
+      "learning_rate": 0.00018260982944957638,
+      "loss": 0.0429,
+      "step": 544
+    },
+    {
+      "epoch": 0.8170914542728636,
+      "grad_norm": 0.12453527639861811,
+      "learning_rate": 0.00018251134982782952,
+      "loss": 0.0177,
+      "step": 545
+    },
+    {
+      "epoch": 0.8185907046476761,
+      "grad_norm": 0.0913439448193188,
+      "learning_rate": 0.00018241261886220154,
+      "loss": 0.0201,
+      "step": 546
+    },
+    {
+      "epoch": 0.8200899550224887,
+      "grad_norm": 0.10694343936153222,
+      "learning_rate": 0.0001823136368534442,
+      "loss": 0.0168,
+      "step": 547
+    },
+    {
+      "epoch": 0.8215892053973014,
+      "grad_norm": 0.12873854233244145,
+      "learning_rate": 0.00018221440410307374,
+      "loss": 0.0182,
+      "step": 548
+    },
+    {
+      "epoch": 0.823088455772114,
+      "grad_norm": 0.12248324617863353,
+      "learning_rate": 0.00018211492091337042,
+      "loss": 0.0245,
+      "step": 549
+    },
+    {
+      "epoch": 0.8245877061469266,
+      "grad_norm": 0.10118859098810772,
+      "learning_rate": 0.00018201518758737724,
+      "loss": 0.031,
+      "step": 550
+    },
+    {
+      "epoch": 0.8260869565217391,
+      "grad_norm": 0.15858954636984837,
+      "learning_rate": 0.0001819152044288992,
+      "loss": 0.053,
+      "step": 551
+    },
+    {
+      "epoch": 0.8275862068965517,
+      "grad_norm": 0.099734088957312,
+      "learning_rate": 0.00018181497174250236,
+      "loss": 0.0217,
+      "step": 552
+    },
+    {
+      "epoch": 0.8290854572713643,
+      "grad_norm": 0.17676306995199245,
+      "learning_rate": 0.00018171448983351284,
+      "loss": 0.039,
+      "step": 553
+    },
+    {
+      "epoch": 0.8305847076461769,
+      "grad_norm": 0.21198277345625827,
+      "learning_rate": 0.00018161375900801604,
+      "loss": 0.0233,
+      "step": 554
+    },
+    {
+      "epoch": 0.8320839580209896,
+      "grad_norm": 0.10508077305912032,
+      "learning_rate": 0.00018151277957285543,
+      "loss": 0.0249,
+      "step": 555
+    },
+    {
+      "epoch": 0.8335832083958021,
+      "grad_norm": 0.15839124476109767,
+      "learning_rate": 0.00018141155183563193,
+      "loss": 0.019,
+      "step": 556
+    },
+    {
+      "epoch": 0.8350824587706147,
+      "grad_norm": 0.11063381503992227,
+      "learning_rate": 0.00018131007610470276,
+      "loss": 0.0284,
+      "step": 557
+    },
+    {
+      "epoch": 0.8365817091454273,
+      "grad_norm": 0.09616652397576733,
+      "learning_rate": 0.00018120835268918063,
+      "loss": 0.0234,
+      "step": 558
+    },
+    {
+      "epoch": 0.8380809595202399,
+      "grad_norm": 0.21337896061350242,
+      "learning_rate": 0.00018110638189893267,
+      "loss": 0.0221,
+      "step": 559
+    },
+    {
+      "epoch": 0.8395802098950524,
+      "grad_norm": 0.11835080605666147,
+      "learning_rate": 0.00018100416404457961,
+      "loss": 0.0162,
+      "step": 560
+    },
+    {
+      "epoch": 0.841079460269865,
+      "grad_norm": 0.0871110273154611,
+      "learning_rate": 0.00018090169943749476,
+      "loss": 0.0188,
+      "step": 561
+    },
+    {
+      "epoch": 0.8425787106446777,
+      "grad_norm": 0.15991292108921587,
+      "learning_rate": 0.00018079898838980305,
+      "loss": 0.0293,
+      "step": 562
+    },
+    {
+      "epoch": 0.8440779610194903,
+      "grad_norm": 0.10275028482175932,
+      "learning_rate": 0.00018069603121438022,
+      "loss": 0.0224,
+      "step": 563
+    },
+    {
+      "epoch": 0.8455772113943029,
+      "grad_norm": 0.2277974036113205,
+      "learning_rate": 0.00018059282822485158,
+      "loss": 0.0258,
+      "step": 564
+    },
+    {
+      "epoch": 0.8470764617691154,
+      "grad_norm": 0.10784532218963591,
+      "learning_rate": 0.0001804893797355914,
+      "loss": 0.0256,
+      "step": 565
+    },
+    {
+      "epoch": 0.848575712143928,
+      "grad_norm": 0.13579456849315263,
+      "learning_rate": 0.00018038568606172173,
+      "loss": 0.0313,
+      "step": 566
+    },
+    {
+      "epoch": 0.8500749625187406,
+      "grad_norm": 0.1139195289086778,
+      "learning_rate": 0.00018028174751911146,
+      "loss": 0.035,
+      "step": 567
+    },
+    {
+      "epoch": 0.8515742128935532,
+      "grad_norm": 0.1437361790270228,
+      "learning_rate": 0.0001801775644243754,
+      "loss": 0.0246,
+      "step": 568
+    },
+    {
+      "epoch": 0.8530734632683659,
+      "grad_norm": 0.1315744813308483,
+      "learning_rate": 0.00018007313709487334,
+      "loss": 0.0273,
+      "step": 569
+    },
+    {
+      "epoch": 0.8545727136431784,
+      "grad_norm": 0.14726220554215622,
+      "learning_rate": 0.00017996846584870908,
+      "loss": 0.0169,
+      "step": 570
+    },
+    {
+      "epoch": 0.856071964017991,
+      "grad_norm": 0.06770252618778136,
+      "learning_rate": 0.00017986355100472928,
+      "loss": 0.014,
+      "step": 571
+    },
+    {
+      "epoch": 0.8575712143928036,
+      "grad_norm": 0.19654688986822602,
+      "learning_rate": 0.00017975839288252287,
+      "loss": 0.0524,
+      "step": 572
+    },
+    {
+      "epoch": 0.8590704647676162,
+      "grad_norm": 0.11188521951103549,
+      "learning_rate": 0.00017965299180241963,
+      "loss": 0.0223,
+      "step": 573
+    },
+    {
+      "epoch": 0.8605697151424287,
+      "grad_norm": 0.16155836256152575,
+      "learning_rate": 0.00017954734808548958,
+      "loss": 0.0214,
+      "step": 574
+    },
+    {
+      "epoch": 0.8620689655172413,
+      "grad_norm": 0.09669657753074787,
+      "learning_rate": 0.00017944146205354182,
+      "loss": 0.0226,
+      "step": 575
+    },
+    {
+      "epoch": 0.863568215892054,
+      "grad_norm": 0.07035277148350057,
+      "learning_rate": 0.00017933533402912354,
+      "loss": 0.0184,
+      "step": 576
+    },
+    {
+      "epoch": 0.8650674662668666,
+      "grad_norm": 0.10734729487368178,
+      "learning_rate": 0.00017922896433551907,
+      "loss": 0.0227,
+      "step": 577
+    },
+    {
+      "epoch": 0.8665667166416792,
+      "grad_norm": 0.12745410681963906,
+      "learning_rate": 0.00017912235329674902,
+      "loss": 0.0254,
+      "step": 578
+    },
+    {
+      "epoch": 0.8680659670164917,
+      "grad_norm": 0.08587193689014985,
+      "learning_rate": 0.00017901550123756906,
+      "loss": 0.0201,
+      "step": 579
+    },
+    {
+      "epoch": 0.8695652173913043,
+      "grad_norm": 0.08573667833277826,
+      "learning_rate": 0.00017890840848346908,
+      "loss": 0.0161,
+      "step": 580
+    },
+    {
+      "epoch": 0.8710644677661169,
+      "grad_norm": 0.24128108163978015,
+      "learning_rate": 0.00017880107536067218,
+      "loss": 0.0317,
+      "step": 581
+    },
+    {
+      "epoch": 0.8725637181409296,
+      "grad_norm": 0.11234634770019401,
+      "learning_rate": 0.00017869350219613375,
+      "loss": 0.0379,
+      "step": 582
+    },
+    {
+      "epoch": 0.8740629685157422,
+      "grad_norm": 0.1786629143096894,
+      "learning_rate": 0.0001785856893175402,
+      "loss": 0.0265,
+      "step": 583
+    },
+    {
+      "epoch": 0.8755622188905547,
+      "grad_norm": 0.17384457039894063,
+      "learning_rate": 0.0001784776370533083,
+      "loss": 0.0374,
+      "step": 584
+    },
+    {
+      "epoch": 0.8770614692653673,
+      "grad_norm": 0.1507710793497251,
+      "learning_rate": 0.000178369345732584,
+      "loss": 0.027,
+      "step": 585
+    },
+    {
+      "epoch": 0.8785607196401799,
+      "grad_norm": 0.09140520812025159,
+      "learning_rate": 0.0001782608156852414,
+      "loss": 0.0211,
+      "step": 586
+    },
+    {
+      "epoch": 0.8800599700149925,
+      "grad_norm": 0.301767128796161,
+      "learning_rate": 0.00017815204724188187,
+      "loss": 0.0315,
+      "step": 587
+    },
+    {
+      "epoch": 0.881559220389805,
+      "grad_norm": 0.16101936926022856,
+      "learning_rate": 0.000178043040733833,
+      "loss": 0.0306,
+      "step": 588
+    },
+    {
+      "epoch": 0.8830584707646177,
+      "grad_norm": 0.1656349117966397,
+      "learning_rate": 0.00017793379649314744,
+      "loss": 0.045,
+      "step": 589
+    },
+    {
+      "epoch": 0.8845577211394303,
+      "grad_norm": 0.2112962853210967,
+      "learning_rate": 0.00017782431485260212,
+      "loss": 0.0579,
+      "step": 590
+    },
+    {
+      "epoch": 0.8860569715142429,
+      "grad_norm": 0.1372483465065148,
+      "learning_rate": 0.0001777145961456971,
+      "loss": 0.0376,
+      "step": 591
+    },
+    {
+      "epoch": 0.8875562218890555,
+      "grad_norm": 0.10696584922742608,
+      "learning_rate": 0.0001776046407066546,
+      "loss": 0.0213,
+      "step": 592
+    },
+    {
+      "epoch": 0.889055472263868,
+      "grad_norm": 0.1450728563008204,
+      "learning_rate": 0.00017749444887041799,
+      "loss": 0.0238,
+      "step": 593
+    },
+    {
+      "epoch": 0.8905547226386806,
+      "grad_norm": 0.11601926896640119,
+      "learning_rate": 0.00017738402097265064,
+      "loss": 0.0333,
+      "step": 594
+    },
+    {
+      "epoch": 0.8920539730134932,
+      "grad_norm": 0.10700704866001351,
+      "learning_rate": 0.00017727335734973512,
+      "loss": 0.0287,
+      "step": 595
+    },
+    {
+      "epoch": 0.8935532233883059,
+      "grad_norm": 0.13363561466349677,
+      "learning_rate": 0.00017716245833877201,
+      "loss": 0.0256,
+      "step": 596
+    },
+    {
+      "epoch": 0.8950524737631185,
+      "grad_norm": 0.0667796736223053,
+      "learning_rate": 0.00017705132427757895,
+      "loss": 0.0178,
+      "step": 597
+    },
+    {
+      "epoch": 0.896551724137931,
+      "grad_norm": 0.10641527946736849,
+      "learning_rate": 0.0001769399555046895,
+      "loss": 0.0142,
+      "step": 598
+    },
+    {
+      "epoch": 0.8980509745127436,
+      "grad_norm": 0.08884031549727572,
+      "learning_rate": 0.00017682835235935236,
+      "loss": 0.0217,
+      "step": 599
+    },
+    {
+      "epoch": 0.8995502248875562,
+      "grad_norm": 0.30418982864519956,
+      "learning_rate": 0.00017671651518153,
+      "loss": 0.0576,
+      "step": 600
+    },
+    {
+      "epoch": 0.9010494752623688,
+      "grad_norm": 0.13881219505976333,
+      "learning_rate": 0.0001766044443118978,
+      "loss": 0.0315,
+      "step": 601
+    },
+    {
+      "epoch": 0.9025487256371814,
+      "grad_norm": 0.1076877239847274,
+      "learning_rate": 0.0001764921400918432,
+      "loss": 0.0275,
+      "step": 602
+    },
+    {
+      "epoch": 0.904047976011994,
+      "grad_norm": 0.15154212814340556,
+      "learning_rate": 0.00017637960286346425,
+      "loss": 0.0334,
+      "step": 603
+    },
+    {
+      "epoch": 0.9055472263868066,
+      "grad_norm": 0.07641353287413101,
+      "learning_rate": 0.00017626683296956882,
+      "loss": 0.0153,
+      "step": 604
+    },
+    {
+      "epoch": 0.9070464767616192,
+      "grad_norm": 0.11340124769027278,
+      "learning_rate": 0.0001761538307536737,
+      "loss": 0.0252,
+      "step": 605
+    },
+    {
+      "epoch": 0.9085457271364318,
+      "grad_norm": 0.08383585004456805,
+      "learning_rate": 0.0001760405965600031,
+      "loss": 0.0181,
+      "step": 606
+    },
+    {
+      "epoch": 0.9100449775112444,
+      "grad_norm": 0.1960282699023016,
+      "learning_rate": 0.00017592713073348807,
+      "loss": 0.0236,
+      "step": 607
+    },
+    {
+      "epoch": 0.9115442278860569,
+      "grad_norm": 0.12553047330249875,
+      "learning_rate": 0.00017581343361976524,
+      "loss": 0.0193,
+      "step": 608
+    },
+    {
+      "epoch": 0.9130434782608695,
+      "grad_norm": 0.07993908623746622,
+      "learning_rate": 0.00017569950556517566,
+      "loss": 0.0118,
+      "step": 609
+    },
+    {
+      "epoch": 0.9145427286356822,
+      "grad_norm": 0.07644624754596607,
+      "learning_rate": 0.00017558534691676397,
+      "loss": 0.0178,
+      "step": 610
+    },
+    {
+      "epoch": 0.9160419790104948,
+      "grad_norm": 0.06474378879594389,
+      "learning_rate": 0.00017547095802227723,
+      "loss": 0.0134,
+      "step": 611
+    },
+    {
+      "epoch": 0.9175412293853074,
+      "grad_norm": 0.24320103822426103,
+      "learning_rate": 0.0001753563392301638,
+      "loss": 0.0358,
+      "step": 612
+    },
+    {
+      "epoch": 0.9190404797601199,
+      "grad_norm": 0.21606781623298388,
+      "learning_rate": 0.00017524149088957245,
+      "loss": 0.018,
+      "step": 613
+    },
+    {
+      "epoch": 0.9205397301349325,
+      "grad_norm": 0.13724300632851888,
+      "learning_rate": 0.00017512641335035113,
+      "loss": 0.0239,
+      "step": 614
+    },
+    {
+      "epoch": 0.9220389805097451,
+      "grad_norm": 0.24490694217869374,
+      "learning_rate": 0.00017501110696304596,
+      "loss": 0.0457,
+      "step": 615
+    },
+    {
+      "epoch": 0.9235382308845578,
+      "grad_norm": 0.09583105854859192,
+      "learning_rate": 0.00017489557207890023,
+      "loss": 0.0344,
+      "step": 616
+    },
+    {
+      "epoch": 0.9250374812593704,
+      "grad_norm": 0.11543373352251945,
+      "learning_rate": 0.0001747798090498532,
+      "loss": 0.0215,
+      "step": 617
+    },
+    {
+      "epoch": 0.9265367316341829,
+      "grad_norm": 0.2650648107879051,
+      "learning_rate": 0.00017466381822853915,
+      "loss": 0.0417,
+      "step": 618
+    },
+    {
+      "epoch": 0.9280359820089955,
+      "grad_norm": 0.15994404613414645,
+      "learning_rate": 0.00017454759996828623,
+      "loss": 0.0295,
+      "step": 619
+    },
+    {
+      "epoch": 0.9295352323838081,
+      "grad_norm": 0.12513579686861107,
+      "learning_rate": 0.0001744311546231154,
+      "loss": 0.0348,
+      "step": 620
+    },
+    {
+      "epoch": 0.9310344827586207,
+      "grad_norm": 0.09930949830951118,
+      "learning_rate": 0.00017431448254773944,
+      "loss": 0.0237,
+      "step": 621
+    },
+    {
+      "epoch": 0.9325337331334332,
+      "grad_norm": 0.2202645779778369,
+      "learning_rate": 0.00017419758409756164,
+      "loss": 0.0199,
+      "step": 622
+    },
+    {
+      "epoch": 0.9340329835082459,
+      "grad_norm": 0.12490401554820374,
+      "learning_rate": 0.000174080459628675,
+      "loss": 0.0367,
+      "step": 623
+    },
+    {
+      "epoch": 0.9355322338830585,
+      "grad_norm": 0.1425795882080626,
+      "learning_rate": 0.000173963109497861,
+      "loss": 0.0256,
+      "step": 624
+    },
+    {
+      "epoch": 0.9370314842578711,
+      "grad_norm": 0.12319971348321294,
+      "learning_rate": 0.00017384553406258842,
+      "loss": 0.029,
+      "step": 625
+    },
+    {
+      "epoch": 0.9385307346326837,
+      "grad_norm": 0.19643269683415326,
+      "learning_rate": 0.0001737277336810124,
+      "loss": 0.052,
+      "step": 626
+    },
+    {
+      "epoch": 0.9400299850074962,
+      "grad_norm": 0.12658722225852667,
+      "learning_rate": 0.00017360970871197346,
+      "loss": 0.0212,
+      "step": 627
+    },
+    {
+      "epoch": 0.9415292353823088,
+      "grad_norm": 0.19280146331288842,
+      "learning_rate": 0.000173491459514996,
+      "loss": 0.0405,
+      "step": 628
+    },
+    {
+      "epoch": 0.9430284857571214,
+      "grad_norm": 0.1201798355899327,
+      "learning_rate": 0.00017337298645028764,
+      "loss": 0.0266,
+      "step": 629
+    },
+    {
+      "epoch": 0.9445277361319341,
+      "grad_norm": 0.06066065563687962,
+      "learning_rate": 0.0001732542898787379,
+      "loss": 0.0167,
+      "step": 630
+    },
+    {
+      "epoch": 0.9460269865067467,
+      "grad_norm": 0.09652060635836711,
+      "learning_rate": 0.00017313537016191706,
+      "loss": 0.02,
+      "step": 631
+    },
+    {
+      "epoch": 0.9475262368815592,
+      "grad_norm": 0.06863279756680676,
+      "learning_rate": 0.00017301622766207527,
+      "loss": 0.0155,
+      "step": 632
+    },
+    {
+      "epoch": 0.9490254872563718,
+      "grad_norm": 0.12042412187593952,
+      "learning_rate": 0.00017289686274214118,
+      "loss": 0.0206,
+      "step": 633
+    },
+    {
+      "epoch": 0.9505247376311844,
+      "grad_norm": 0.21939496082438123,
+      "learning_rate": 0.00017277727576572107,
+      "loss": 0.0213,
+      "step": 634
+    },
+    {
+      "epoch": 0.952023988005997,
+      "grad_norm": 0.12191851364600712,
+      "learning_rate": 0.0001726574670970976,
+      "loss": 0.0305,
+      "step": 635
+    },
+    {
+      "epoch": 0.9535232383808095,
+      "grad_norm": 0.27478912515600534,
+      "learning_rate": 0.00017253743710122875,
+      "loss": 0.0283,
+      "step": 636
+    },
+    {
+      "epoch": 0.9550224887556222,
+      "grad_norm": 0.1281513047579704,
+      "learning_rate": 0.00017241718614374678,
+      "loss": 0.04,
+      "step": 637
+    },
+    {
+      "epoch": 0.9565217391304348,
+      "grad_norm": 0.1266027647091267,
+      "learning_rate": 0.00017229671459095683,
+      "loss": 0.0376,
+      "step": 638
+    },
+    {
+      "epoch": 0.9580209895052474,
+      "grad_norm": 0.16017298063734342,
+      "learning_rate": 0.00017217602280983623,
+      "loss": 0.0309,
+      "step": 639
+    },
+    {
+      "epoch": 0.95952023988006,
+      "grad_norm": 0.12020816231782688,
+      "learning_rate": 0.00017205511116803306,
+      "loss": 0.0256,
+      "step": 640
+    },
+    {
+      "epoch": 0.9610194902548725,
+      "grad_norm": 0.06023210487373309,
+      "learning_rate": 0.0001719339800338651,
+      "loss": 0.0159,
+      "step": 641
+    },
+    {
+      "epoch": 0.9625187406296851,
+      "grad_norm": 0.12858383075069013,
+      "learning_rate": 0.00017181262977631888,
+      "loss": 0.0281,
+      "step": 642
+    },
+    {
+      "epoch": 0.9640179910044977,
+      "grad_norm": 0.11429779029405647,
+      "learning_rate": 0.0001716910607650483,
+      "loss": 0.0284,
+      "step": 643
+    },
+    {
+      "epoch": 0.9655172413793104,
+      "grad_norm": 0.1072273139514921,
+      "learning_rate": 0.0001715692733703736,
+      "loss": 0.027,
+      "step": 644
+    },
+    {
+      "epoch": 0.967016491754123,
+      "grad_norm": 0.20115614173420412,
+      "learning_rate": 0.00017144726796328034,
+      "loss": 0.0259,
+      "step": 645
+    },
+    {
+      "epoch": 0.9685157421289355,
+      "grad_norm": 0.06623043636689804,
+      "learning_rate": 0.00017132504491541818,
+      "loss": 0.0185,
+      "step": 646
+    },
+    {
+      "epoch": 0.9700149925037481,
+      "grad_norm": 0.20340594306656323,
+      "learning_rate": 0.00017120260459909967,
+      "loss": 0.0286,
+      "step": 647
+    },
+    {
+      "epoch": 0.9715142428785607,
+      "grad_norm": 0.0772642579410454,
+      "learning_rate": 0.00017107994738729926,
+      "loss": 0.0132,
+      "step": 648
+    },
+    {
+      "epoch": 0.9730134932533733,
+      "grad_norm": 0.08015487685028685,
+      "learning_rate": 0.0001709570736536521,
+      "loss": 0.0233,
+      "step": 649
+    },
+    {
+      "epoch": 0.974512743628186,
+      "grad_norm": 0.1661406803223815,
+      "learning_rate": 0.0001708339837724529,
+      "loss": 0.0302,
+      "step": 650
+    },
+    {
+      "epoch": 0.9760119940029985,
+      "grad_norm": 0.1458374578095082,
+      "learning_rate": 0.00017071067811865476,
+      "loss": 0.033,
+      "step": 651
+    },
+    {
+      "epoch": 0.9775112443778111,
+      "grad_norm": 0.07798257623963753,
+      "learning_rate": 0.0001705871570678681,
+      "loss": 0.0128,
+      "step": 652
+    },
+    {
+      "epoch": 0.9790104947526237,
+      "grad_norm": 0.048418288434207646,
+      "learning_rate": 0.00017046342099635948,
+      "loss": 0.0123,
+      "step": 653
+    },
+    {
+      "epoch": 0.9805097451274363,
+      "grad_norm": 0.13431687956769212,
+      "learning_rate": 0.00017033947028105039,
+      "loss": 0.0251,
+      "step": 654
+    },
+    {
+      "epoch": 0.9820089955022488,
+      "grad_norm": 0.08226067771753236,
+      "learning_rate": 0.00017021530529951625,
+      "loss": 0.021,
+      "step": 655
+    },
+    {
+      "epoch": 0.9835082458770614,
+      "grad_norm": 0.1462577267853189,
+      "learning_rate": 0.0001700909264299851,
+      "loss": 0.0421,
+      "step": 656
+    },
+    {
+      "epoch": 0.9850074962518741,
+      "grad_norm": 0.3083259142195758,
+      "learning_rate": 0.00016996633405133655,
+      "loss": 0.0494,
+      "step": 657
+    },
+    {
+      "epoch": 0.9865067466266867,
+      "grad_norm": 0.18522845664082035,
+      "learning_rate": 0.0001698415285431006,
+      "loss": 0.0363,
+      "step": 658
+    },
+    {
+      "epoch": 0.9880059970014993,
+      "grad_norm": 0.09484537072137021,
+      "learning_rate": 0.00016971651028545648,
+      "loss": 0.0281,
+      "step": 659
+    },
+    {
+      "epoch": 0.9895052473763118,
+      "grad_norm": 0.23884894163402523,
+      "learning_rate": 0.00016959127965923142,
+      "loss": 0.0456,
+      "step": 660
+    },
+    {
+      "epoch": 0.9910044977511244,
+      "grad_norm": 0.1779125846943612,
+      "learning_rate": 0.00016946583704589973,
+      "loss": 0.0422,
+      "step": 661
+    },
+    {
+      "epoch": 0.992503748125937,
+      "grad_norm": 0.10266626269337822,
+      "learning_rate": 0.0001693401828275813,
+      "loss": 0.0203,
+      "step": 662
+    },
+    {
+      "epoch": 0.9940029985007496,
+      "grad_norm": 0.10277193051633117,
+      "learning_rate": 0.0001692143173870407,
+      "loss": 0.0213,
+      "step": 663
+    },
+    {
+      "epoch": 0.9955022488755623,
+      "grad_norm": 0.11463949080930057,
+      "learning_rate": 0.00016908824110768584,
+      "loss": 0.0339,
+      "step": 664
+    },
+    {
+      "epoch": 0.9970014992503748,
+      "grad_norm": 0.13859328494189713,
+      "learning_rate": 0.000168961954373567,
+      "loss": 0.023,
+      "step": 665
+    },
+    {
+      "epoch": 0.9985007496251874,
+      "grad_norm": 0.10931098650164782,
+      "learning_rate": 0.0001688354575693754,
+      "loss": 0.022,
+      "step": 666
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1521638053539932,
+      "learning_rate": 0.0001687087510804423,
+      "loss": 0.032,
+      "step": 667
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 0.03065001219511032,
+      "eval_runtime": 483.0726,
+      "eval_samples_per_second": 21.479,
+      "eval_steps_per_second": 1.343,
+      "step": 667
+    },
+    {
+      "epoch": 1.0014992503748126,
+      "grad_norm": 0.10472517028799198,
+      "learning_rate": 0.00016858183529273765,
+      "loss": 0.0194,
+      "step": 668
+    },
+    {
+      "epoch": 1.0029985007496252,
+      "grad_norm": 0.07086065655028383,
+      "learning_rate": 0.00016845471059286887,
+      "loss": 0.0169,
+      "step": 669
+    },
+    {
+      "epoch": 1.0044977511244377,
+      "grad_norm": 0.056470801319977317,
+      "learning_rate": 0.00016832737736807994,
+      "loss": 0.0178,
+      "step": 670
+    },
+    {
+      "epoch": 1.0059970014992503,
+      "grad_norm": 0.21736541605778517,
+      "learning_rate": 0.00016819983600624986,
+      "loss": 0.0204,
+      "step": 671
+    },
+    {
+      "epoch": 1.0074962518740629,
+      "grad_norm": 0.08048210993127422,
+      "learning_rate": 0.0001680720868958918,
+      "loss": 0.0152,
+      "step": 672
+    },
+    {
+      "epoch": 1.0089955022488755,
+      "grad_norm": 0.13724751027412635,
+      "learning_rate": 0.00016794413042615168,
+      "loss": 0.025,
+      "step": 673
+    },
+    {
+      "epoch": 1.0104947526236883,
+      "grad_norm": 0.1299998829933712,
+      "learning_rate": 0.0001678159669868071,
+      "loss": 0.0233,
+      "step": 674
+    },
+    {
+      "epoch": 1.0119940029985008,
+      "grad_norm": 0.08324573169820197,
+      "learning_rate": 0.00016768759696826608,
+      "loss": 0.0161,
+      "step": 675
+    },
+    {
+      "epoch": 1.0134932533733134,
+      "grad_norm": 0.24760800404844965,
+      "learning_rate": 0.00016755902076156604,
+      "loss": 0.0267,
+      "step": 676
+    },
+    {
+      "epoch": 1.014992503748126,
+      "grad_norm": 0.057948619616265976,
+      "learning_rate": 0.00016743023875837233,
+      "loss": 0.0083,
+      "step": 677
+    },
+    {
+      "epoch": 1.0164917541229386,
+      "grad_norm": 0.09704964114233473,
+      "learning_rate": 0.00016730125135097735,
+      "loss": 0.0111,
+      "step": 678
+    },
+    {
+      "epoch": 1.0179910044977512,
+      "grad_norm": 0.11235791780825113,
+      "learning_rate": 0.00016717205893229903,
+      "loss": 0.0196,
+      "step": 679
+    },
+    {
+      "epoch": 1.0194902548725637,
+      "grad_norm": 0.14212568674303996,
+      "learning_rate": 0.0001670426618958799,
+      "loss": 0.0179,
+      "step": 680
+    },
+    {
+      "epoch": 1.0209895052473763,
+      "grad_norm": 0.13458409942428412,
+      "learning_rate": 0.00016691306063588583,
+      "loss": 0.015,
+      "step": 681
+    },
+    {
+      "epoch": 1.0224887556221889,
+      "grad_norm": 0.10078342958992698,
+      "learning_rate": 0.00016678325554710468,
+      "loss": 0.0145,
+      "step": 682
+    },
+    {
+      "epoch": 1.0239880059970015,
+      "grad_norm": 0.2163179066668823,
+      "learning_rate": 0.00016665324702494524,
+      "loss": 0.0352,
+      "step": 683
+    },
+    {
+      "epoch": 1.025487256371814,
+      "grad_norm": 0.09314880164627648,
+      "learning_rate": 0.00016652303546543608,
+      "loss": 0.0118,
+      "step": 684
+    },
+    {
+      "epoch": 1.0269865067466266,
+      "grad_norm": 0.12711682267879498,
+      "learning_rate": 0.00016639262126522418,
+      "loss": 0.0187,
+      "step": 685
+    },
+    {
+      "epoch": 1.0284857571214392,
+      "grad_norm": 0.10899305518463998,
+      "learning_rate": 0.00016626200482157378,
+      "loss": 0.0295,
+      "step": 686
+    },
+    {
+      "epoch": 1.0299850074962518,
+      "grad_norm": 0.08551960256639587,
+      "learning_rate": 0.00016613118653236518,
+      "loss": 0.0129,
+      "step": 687
+    },
+    {
+      "epoch": 1.0314842578710646,
+      "grad_norm": 0.08800753901836143,
+      "learning_rate": 0.0001660001667960937,
+      "loss": 0.0138,
+      "step": 688
+    },
+    {
+      "epoch": 1.0329835082458771,
+      "grad_norm": 0.052385346475073355,
+      "learning_rate": 0.00016586894601186805,
+      "loss": 0.0083,
+      "step": 689
+    },
+    {
+      "epoch": 1.0344827586206897,
+      "grad_norm": 0.07054205702889728,
+      "learning_rate": 0.0001657375245794096,
+      "loss": 0.0124,
+      "step": 690
+    },
+    {
+      "epoch": 1.0359820089955023,
+      "grad_norm": 0.09975501615825792,
+      "learning_rate": 0.00016560590289905073,
+      "loss": 0.0118,
+      "step": 691
+    },
+    {
+      "epoch": 1.0374812593703149,
+      "grad_norm": 0.14047961341536686,
+      "learning_rate": 0.00016547408137173396,
+      "loss": 0.0099,
+      "step": 692
+    },
+    {
+      "epoch": 1.0389805097451275,
+      "grad_norm": 0.1189003861776989,
+      "learning_rate": 0.00016534206039901057,
+      "loss": 0.0262,
+      "step": 693
+    },
+    {
+      "epoch": 1.04047976011994,
+      "grad_norm": 0.11248270963982432,
+      "learning_rate": 0.00016520984038303924,
+      "loss": 0.0178,
+      "step": 694
+    },
+    {
+      "epoch": 1.0419790104947526,
+      "grad_norm": 0.13888415150865596,
+      "learning_rate": 0.0001650774217265851,
+      "loss": 0.0225,
+      "step": 695
+    },
+    {
+      "epoch": 1.0434782608695652,
+      "grad_norm": 0.11070223453464198,
+      "learning_rate": 0.00016494480483301836,
+      "loss": 0.0112,
+      "step": 696
+    },
+    {
+      "epoch": 1.0449775112443778,
+      "grad_norm": 0.1551674425930212,
+      "learning_rate": 0.0001648119901063131,
+      "loss": 0.0193,
+      "step": 697
+    },
+    {
+      "epoch": 1.0464767616191903,
+      "grad_norm": 0.13273313189669741,
+      "learning_rate": 0.000164678977951046,
+      "loss": 0.0283,
+      "step": 698
+    },
+    {
+      "epoch": 1.047976011994003,
+      "grad_norm": 0.11544709997166336,
+      "learning_rate": 0.00016454576877239507,
+      "loss": 0.0174,
+      "step": 699
+    },
+    {
+      "epoch": 1.0494752623688155,
+      "grad_norm": 0.1253059382700822,
+      "learning_rate": 0.00016441236297613866,
+      "loss": 0.0095,
+      "step": 700
+    },
+    {
+      "epoch": 1.050974512743628,
+      "grad_norm": 0.231346802938159,
+      "learning_rate": 0.00016427876096865394,
+      "loss": 0.0127,
+      "step": 701
+    },
+    {
+      "epoch": 1.0524737631184409,
+      "grad_norm": 0.09463184955976486,
+      "learning_rate": 0.00016414496315691581,
+      "loss": 0.0132,
+      "step": 702
+    },
+    {
+      "epoch": 1.0539730134932535,
+      "grad_norm": 0.10175851797876952,
+      "learning_rate": 0.00016401096994849557,
+      "loss": 0.0165,
+      "step": 703
+    },
+    {
+      "epoch": 1.055472263868066,
+      "grad_norm": 0.08269715474941393,
+      "learning_rate": 0.00016387678175155978,
+      "loss": 0.022,
+      "step": 704
+    },
+    {
+      "epoch": 1.0569715142428786,
+      "grad_norm": 0.0859139526446948,
+      "learning_rate": 0.000163742398974869,
+      "loss": 0.0128,
+      "step": 705
+    },
+    {
+      "epoch": 1.0584707646176912,
+      "grad_norm": 0.1135404099233105,
+      "learning_rate": 0.0001636078220277764,
+      "loss": 0.0183,
+      "step": 706
+    },
+    {
+      "epoch": 1.0599700149925038,
+      "grad_norm": 0.09275250705549713,
+      "learning_rate": 0.00016347305132022677,
+      "loss": 0.0232,
+      "step": 707
+    },
+    {
+      "epoch": 1.0614692653673163,
+      "grad_norm": 0.11572250326059627,
+      "learning_rate": 0.000163338087262755,
+      "loss": 0.0824,
+      "step": 708
+    },
+    {
+      "epoch": 1.062968515742129,
+      "grad_norm": 0.15876584912162692,
+      "learning_rate": 0.0001632029302664851,
+      "loss": 0.0239,
+      "step": 709
+    },
+    {
+      "epoch": 1.0644677661169415,
+      "grad_norm": 0.09255809526179942,
+      "learning_rate": 0.00016306758074312864,
+      "loss": 0.0154,
+      "step": 710
+    },
+    {
+      "epoch": 1.065967016491754,
+      "grad_norm": 0.11390783895932384,
+      "learning_rate": 0.00016293203910498376,
+      "loss": 0.0143,
+      "step": 711
+    },
+    {
+      "epoch": 1.0674662668665666,
+      "grad_norm": 0.23974840043852813,
+      "learning_rate": 0.00016279630576493382,
+      "loss": 0.0171,
+      "step": 712
+    },
+    {
+      "epoch": 1.0689655172413792,
+      "grad_norm": 0.17086459332122525,
+      "learning_rate": 0.00016266038113644607,
+      "loss": 0.0244,
+      "step": 713
+    },
+    {
+      "epoch": 1.0704647676161918,
+      "grad_norm": 0.05971167811590956,
+      "learning_rate": 0.00016252426563357055,
+      "loss": 0.0055,
+      "step": 714
+    },
+    {
+      "epoch": 1.0719640179910046,
+      "grad_norm": 0.0964739917335777,
+      "learning_rate": 0.00016238795967093864,
+      "loss": 0.0185,
+      "step": 715
+    },
+    {
+      "epoch": 1.0734632683658172,
+      "grad_norm": 0.10293996743254338,
+      "learning_rate": 0.00016225146366376198,
+      "loss": 0.0176,
+      "step": 716
+    },
+    {
+      "epoch": 1.0749625187406298,
+      "grad_norm": 0.08690427481589705,
+      "learning_rate": 0.00016211477802783103,
+      "loss": 0.0123,
+      "step": 717
+    },
+    {
+      "epoch": 1.0764617691154423,
+      "grad_norm": 0.08500732276952126,
+      "learning_rate": 0.00016197790317951403,
+      "loss": 0.0113,
+      "step": 718
+    },
+    {
+      "epoch": 1.077961019490255,
+      "grad_norm": 0.11613707922898552,
+      "learning_rate": 0.0001618408395357554,
+      "loss": 0.0184,
+      "step": 719
+    },
+    {
+      "epoch": 1.0794602698650675,
+      "grad_norm": 0.2060611324342325,
+      "learning_rate": 0.00016170358751407487,
+      "loss": 0.0442,
+      "step": 720
+    },
+    {
+      "epoch": 1.08095952023988,
+      "grad_norm": 0.11248886768899795,
+      "learning_rate": 0.0001615661475325658,
+      "loss": 0.0215,
+      "step": 721
+    },
+    {
+      "epoch": 1.0824587706146926,
+      "grad_norm": 0.17400298698519548,
+      "learning_rate": 0.00016142852000989433,
+      "loss": 0.0347,
+      "step": 722
+    },
+    {
+      "epoch": 1.0839580209895052,
+      "grad_norm": 0.07815925376493392,
+      "learning_rate": 0.00016129070536529766,
+      "loss": 0.0106,
+      "step": 723
+    },
+    {
+      "epoch": 1.0854572713643178,
+      "grad_norm": 0.11757645186998897,
+      "learning_rate": 0.00016115270401858314,
+      "loss": 0.0202,
+      "step": 724
+    },
+    {
+      "epoch": 1.0869565217391304,
+      "grad_norm": 0.12060617886999422,
+      "learning_rate": 0.0001610145163901268,
+      "loss": 0.0133,
+      "step": 725
+    },
+    {
+      "epoch": 1.088455772113943,
+      "grad_norm": 0.07666228061263336,
+      "learning_rate": 0.00016087614290087208,
+      "loss": 0.0114,
+      "step": 726
+    },
+    {
+      "epoch": 1.0899550224887555,
+      "grad_norm": 0.15028068516880796,
+      "learning_rate": 0.00016073758397232868,
+      "loss": 0.0224,
+      "step": 727
+    },
+    {
+      "epoch": 1.0914542728635683,
+      "grad_norm": 0.07317261491835224,
+      "learning_rate": 0.0001605988400265711,
+      "loss": 0.014,
+      "step": 728
+    },
+    {
+      "epoch": 1.092953523238381,
+      "grad_norm": 0.08354275449747522,
+      "learning_rate": 0.0001604599114862375,
+      "loss": 0.0233,
+      "step": 729
+    },
+    {
+      "epoch": 1.0944527736131935,
+      "grad_norm": 0.07542214763546543,
+      "learning_rate": 0.00016032079877452825,
+      "loss": 0.0121,
+      "step": 730
+    },
+    {
+      "epoch": 1.095952023988006,
+      "grad_norm": 0.0866470368358821,
+      "learning_rate": 0.00016018150231520486,
+      "loss": 0.0142,
+      "step": 731
+    },
+    {
+      "epoch": 1.0974512743628186,
+      "grad_norm": 0.10167617929252423,
+      "learning_rate": 0.00016004202253258842,
+      "loss": 0.0193,
+      "step": 732
+    },
+    {
+      "epoch": 1.0989505247376312,
+      "grad_norm": 0.10275133468863129,
+      "learning_rate": 0.0001599023598515586,
+      "loss": 0.013,
+      "step": 733
+    },
+    {
+      "epoch": 1.1004497751124438,
+      "grad_norm": 0.11247245630083348,
+      "learning_rate": 0.00015976251469755214,
+      "loss": 0.0171,
+      "step": 734
+    },
+    {
+      "epoch": 1.1019490254872564,
+      "grad_norm": 0.16757428374213076,
+      "learning_rate": 0.0001596224874965616,
+      "loss": 0.0238,
+      "step": 735
+    },
+    {
+      "epoch": 1.103448275862069,
+      "grad_norm": 0.09060490513425551,
+      "learning_rate": 0.00015948227867513415,
+      "loss": 0.0117,
+      "step": 736
+    },
+    {
+      "epoch": 1.1049475262368815,
+      "grad_norm": 0.0958942229372783,
+      "learning_rate": 0.00015934188866037016,
+      "loss": 0.0118,
+      "step": 737
+    },
+    {
+      "epoch": 1.106446776611694,
+      "grad_norm": 0.11408221833035843,
+      "learning_rate": 0.00015920131787992197,
+      "loss": 0.0181,
+      "step": 738
+    },
+    {
+      "epoch": 1.1079460269865067,
+      "grad_norm": 0.08779211321518267,
+      "learning_rate": 0.00015906056676199255,
+      "loss": 0.0163,
+      "step": 739
+    },
+    {
+      "epoch": 1.1094452773613193,
+      "grad_norm": 0.08339161981475392,
+      "learning_rate": 0.0001589196357353342,
+      "loss": 0.0132,
+      "step": 740
+    },
+    {
+      "epoch": 1.1109445277361318,
+      "grad_norm": 0.1766144838591971,
+      "learning_rate": 0.00015877852522924732,
+      "loss": 0.0186,
+      "step": 741
+    },
+    {
+      "epoch": 1.1124437781109444,
+      "grad_norm": 0.17475491667530327,
+      "learning_rate": 0.00015863723567357892,
+      "loss": 0.0464,
+      "step": 742
+    },
+    {
+      "epoch": 1.1139430284857572,
+      "grad_norm": 0.11280224594809592,
+      "learning_rate": 0.00015849576749872157,
+      "loss": 0.0299,
+      "step": 743
+    },
+    {
+      "epoch": 1.1154422788605698,
+      "grad_norm": 0.06496065765413261,
+      "learning_rate": 0.00015835412113561175,
+      "loss": 0.0086,
+      "step": 744
+    },
+    {
+      "epoch": 1.1169415292353824,
+      "grad_norm": 0.0898308908216899,
+      "learning_rate": 0.00015821229701572896,
+      "loss": 0.014,
+      "step": 745
+    },
+    {
+      "epoch": 1.118440779610195,
+      "grad_norm": 0.10886469850221588,
+      "learning_rate": 0.00015807029557109398,
+      "loss": 0.0195,
+      "step": 746
+    },
+    {
+      "epoch": 1.1199400299850075,
+      "grad_norm": 0.14282651466132884,
+      "learning_rate": 0.0001579281172342679,
+      "loss": 0.0135,
+      "step": 747
+    },
+    {
+      "epoch": 1.12143928035982,
+      "grad_norm": 0.10856367670748046,
+      "learning_rate": 0.00015778576243835054,
+      "loss": 0.0169,
+      "step": 748
+    },
+    {
+      "epoch": 1.1229385307346327,
+      "grad_norm": 0.06823231902579771,
+      "learning_rate": 0.00015764323161697935,
+      "loss": 0.0109,
+      "step": 749
+    },
+    {
+      "epoch": 1.1244377811094453,
+      "grad_norm": 0.08487293384058076,
+      "learning_rate": 0.00015750052520432787,
+      "loss": 0.0098,
+      "step": 750
+    },
+    {
+      "epoch": 1.1259370314842578,
+      "grad_norm": 0.07963389915287164,
+      "learning_rate": 0.0001573576436351046,
+      "loss": 0.0104,
+      "step": 751
+    },
+    {
+      "epoch": 1.1274362818590704,
+      "grad_norm": 0.159748695375408,
+      "learning_rate": 0.00015721458734455163,
+      "loss": 0.0232,
+      "step": 752
+    },
+    {
+      "epoch": 1.128935532233883,
+      "grad_norm": 0.1269515763993384,
+      "learning_rate": 0.0001570713567684432,
+      "loss": 0.0171,
+      "step": 753
+    },
+    {
+      "epoch": 1.1304347826086956,
+      "grad_norm": 0.11521833565751512,
+      "learning_rate": 0.00015692795234308445,
+      "loss": 0.0172,
+      "step": 754
+    },
+    {
+      "epoch": 1.1319340329835081,
+      "grad_norm": 0.08114505514088168,
+      "learning_rate": 0.00015678437450531013,
+      "loss": 0.0167,
+      "step": 755
+    },
+    {
+      "epoch": 1.133433283358321,
+      "grad_norm": 0.09194908508040527,
+      "learning_rate": 0.00015664062369248328,
+      "loss": 0.0129,
+      "step": 756
+    },
+    {
+      "epoch": 1.1349325337331335,
+      "grad_norm": 0.07613583096910623,
+      "learning_rate": 0.0001564967003424938,
+      "loss": 0.0159,
+      "step": 757
+    },
+    {
+      "epoch": 1.136431784107946,
+      "grad_norm": 0.10265920058076042,
+      "learning_rate": 0.00015635260489375714,
+      "loss": 0.0216,
+      "step": 758
+    },
+    {
+      "epoch": 1.1379310344827587,
+      "grad_norm": 0.0631078940073356,
+      "learning_rate": 0.00015620833778521307,
+      "loss": 0.0091,
+      "step": 759
+    },
+    {
+      "epoch": 1.1394302848575713,
+      "grad_norm": 0.11550928783152578,
+      "learning_rate": 0.0001560638994563242,
+      "loss": 0.0247,
+      "step": 760
+    },
+    {
+      "epoch": 1.1409295352323838,
+      "grad_norm": 0.05701238781802943,
+      "learning_rate": 0.0001559192903470747,
+      "loss": 0.007,
+      "step": 761
+    },
+    {
+      "epoch": 1.1424287856071964,
+      "grad_norm": 0.0863050001301802,
+      "learning_rate": 0.00015577451089796905,
+      "loss": 0.0117,
+      "step": 762
+    },
+    {
+      "epoch": 1.143928035982009,
+      "grad_norm": 0.1095952763491608,
+      "learning_rate": 0.0001556295615500305,
+      "loss": 0.0139,
+      "step": 763
+    },
+    {
+      "epoch": 1.1454272863568216,
+      "grad_norm": 0.10109899186437078,
+      "learning_rate": 0.00015548444274479995,
+      "loss": 0.0166,
+      "step": 764
+    },
+    {
+      "epoch": 1.1469265367316341,
+      "grad_norm": 0.13502630641951327,
+      "learning_rate": 0.00015533915492433443,
+      "loss": 0.0315,
+      "step": 765
+    },
+    {
+      "epoch": 1.1484257871064467,
+      "grad_norm": 0.13530559923309216,
+      "learning_rate": 0.0001551936985312058,
+      "loss": 0.0105,
+      "step": 766
+    },
+    {
+      "epoch": 1.1499250374812593,
+      "grad_norm": 0.1087257528041358,
+      "learning_rate": 0.00015504807400849958,
+      "loss": 0.0162,
+      "step": 767
+    },
+    {
+      "epoch": 1.1514242878560719,
+      "grad_norm": 0.1072652302854073,
+      "learning_rate": 0.0001549022817998132,
+      "loss": 0.0171,
+      "step": 768
+    },
+    {
+      "epoch": 1.1529235382308847,
+      "grad_norm": 0.09402648254427196,
+      "learning_rate": 0.00015475632234925504,
+      "loss": 0.0132,
+      "step": 769
+    },
+    {
+      "epoch": 1.1544227886056972,
+      "grad_norm": 0.10085528831110949,
+      "learning_rate": 0.0001546101961014429,
+      "loss": 0.0146,
+      "step": 770
+    },
+    {
+      "epoch": 1.1559220389805098,
+      "grad_norm": 0.10445329841658932,
+      "learning_rate": 0.00015446390350150273,
+      "loss": 0.0217,
+      "step": 771
+    },
+    {
+      "epoch": 1.1574212893553224,
+      "grad_norm": 0.07781814067350039,
+      "learning_rate": 0.00015431744499506706,
+      "loss": 0.0147,
+      "step": 772
+    },
+    {
+      "epoch": 1.158920539730135,
+      "grad_norm": 0.10153183985035838,
+      "learning_rate": 0.000154170821028274,
+      "loss": 0.0129,
+      "step": 773
+    },
+    {
+      "epoch": 1.1604197901049476,
+      "grad_norm": 0.1306523813414455,
+      "learning_rate": 0.0001540240320477655,
+      "loss": 0.0194,
+      "step": 774
+    },
+    {
+      "epoch": 1.1619190404797601,
+      "grad_norm": 0.07875779988064713,
+      "learning_rate": 0.0001538770785006863,
+      "loss": 0.0126,
+      "step": 775
+    },
+    {
+      "epoch": 1.1634182908545727,
+      "grad_norm": 0.06662289751984372,
+      "learning_rate": 0.0001537299608346824,
+      "loss": 0.0078,
+      "step": 776
+    },
+    {
+      "epoch": 1.1649175412293853,
+      "grad_norm": 0.11948574312781496,
+      "learning_rate": 0.00015358267949789966,
+      "loss": 0.0121,
+      "step": 777
+    },
+    {
+      "epoch": 1.1664167916041979,
+      "grad_norm": 0.0857036909342206,
+      "learning_rate": 0.00015343523493898265,
+      "loss": 0.0124,
+      "step": 778
+    },
+    {
+      "epoch": 1.1679160419790104,
+      "grad_norm": 0.050863627021771815,
+      "learning_rate": 0.000153287627607073,
+      "loss": 0.0108,
+      "step": 779
+    },
+    {
+      "epoch": 1.169415292353823,
+      "grad_norm": 0.07503552287179403,
+      "learning_rate": 0.00015313985795180828,
+      "loss": 0.0103,
+      "step": 780
+    },
+    {
+      "epoch": 1.1709145427286356,
+      "grad_norm": 0.12348759956679127,
+      "learning_rate": 0.0001529919264233205,
+      "loss": 0.0183,
+      "step": 781
+    },
+    {
+      "epoch": 1.1724137931034484,
+      "grad_norm": 0.16396605752248122,
+      "learning_rate": 0.00015284383347223472,
+      "loss": 0.0237,
+      "step": 782
+    },
+    {
+      "epoch": 1.1739130434782608,
+      "grad_norm": 0.11710104871369872,
+      "learning_rate": 0.00015269557954966778,
+      "loss": 0.0116,
+      "step": 783
+    },
+    {
+      "epoch": 1.1754122938530736,
+      "grad_norm": 0.05426089810325071,
+      "learning_rate": 0.00015254716510722677,
+      "loss": 0.0077,
+      "step": 784
+    },
+    {
+      "epoch": 1.1769115442278861,
+      "grad_norm": 0.12476718035000098,
+      "learning_rate": 0.00015239859059700794,
+      "loss": 0.0148,
+      "step": 785
+    },
+    {
+      "epoch": 1.1784107946026987,
+      "grad_norm": 0.11340575380427748,
+      "learning_rate": 0.0001522498564715949,
+      "loss": 0.0174,
+      "step": 786
+    },
+    {
+      "epoch": 1.1799100449775113,
+      "grad_norm": 0.07460825828437444,
+      "learning_rate": 0.00015210096318405767,
+      "loss": 0.0107,
+      "step": 787
+    },
+    {
+      "epoch": 1.1814092953523239,
+      "grad_norm": 0.11142555377576228,
+      "learning_rate": 0.00015195191118795096,
+      "loss": 0.0195,
+      "step": 788
+    },
+    {
+      "epoch": 1.1829085457271364,
+      "grad_norm": 0.14977712416925076,
+      "learning_rate": 0.00015180270093731303,
+      "loss": 0.0223,
+      "step": 789
+    },
+    {
+      "epoch": 1.184407796101949,
+      "grad_norm": 0.08700335413855392,
+      "learning_rate": 0.0001516533328866642,
+      "loss": 0.0123,
+      "step": 790
+    },
+    {
+      "epoch": 1.1859070464767616,
+      "grad_norm": 0.0724645987757823,
+      "learning_rate": 0.00015150380749100545,
+      "loss": 0.0086,
+      "step": 791
+    },
+    {
+      "epoch": 1.1874062968515742,
+      "grad_norm": 0.14991112523341207,
+      "learning_rate": 0.00015135412520581702,
+      "loss": 0.0116,
+      "step": 792
+    },
+    {
+      "epoch": 1.1889055472263867,
+      "grad_norm": 0.14228737821653165,
+      "learning_rate": 0.00015120428648705717,
+      "loss": 0.0189,
+      "step": 793
+    },
+    {
+      "epoch": 1.1904047976011993,
+      "grad_norm": 0.24633364927454285,
+      "learning_rate": 0.0001510542917911606,
+      "loss": 0.0327,
+      "step": 794
+    },
+    {
+      "epoch": 1.191904047976012,
+      "grad_norm": 0.09327854611652424,
+      "learning_rate": 0.00015090414157503714,
+      "loss": 0.0122,
+      "step": 795
+    },
+    {
+      "epoch": 1.1934032983508245,
+      "grad_norm": 0.23468339856981874,
+      "learning_rate": 0.00015075383629607042,
+      "loss": 0.0176,
+      "step": 796
+    },
+    {
+      "epoch": 1.1949025487256373,
+      "grad_norm": 0.05100781753856239,
+      "learning_rate": 0.00015060337641211637,
+      "loss": 0.0104,
+      "step": 797
+    },
+    {
+      "epoch": 1.1964017991004499,
+      "grad_norm": 0.19141044396688225,
+      "learning_rate": 0.00015045276238150192,
+      "loss": 0.0124,
+      "step": 798
+    },
+    {
+      "epoch": 1.1979010494752624,
+      "grad_norm": 0.08338970503080669,
+      "learning_rate": 0.00015030199466302353,
+      "loss": 0.0124,
+      "step": 799
+    },
+    {
+      "epoch": 1.199400299850075,
+      "grad_norm": 0.15672763842519447,
+      "learning_rate": 0.00015015107371594573,
+      "loss": 0.0146,
+      "step": 800
+    },
+    {
+      "epoch": 1.2008995502248876,
+      "grad_norm": 0.0879518026205462,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 0.0105,
+      "step": 801
+    },
+    {
+      "epoch": 1.2023988005997002,
+      "grad_norm": 0.049258422423769856,
+      "learning_rate": 0.00014984877397538303,
+      "loss": 0.0072,
+      "step": 802
+    },
+    {
+      "epoch": 1.2038980509745127,
+      "grad_norm": 0.10542578394208428,
+      "learning_rate": 0.00014969739610275556,
+      "loss": 0.0124,
+      "step": 803
+    },
+    {
+      "epoch": 1.2053973013493253,
+      "grad_norm": 0.09901336472239698,
+      "learning_rate": 0.00014954586684324078,
+      "loss": 0.0177,
+      "step": 804
+    },
+    {
+      "epoch": 1.206896551724138,
+      "grad_norm": 0.13921198140006943,
+      "learning_rate": 0.0001493941866584231,
+      "loss": 0.0203,
+      "step": 805
+    },
+    {
+      "epoch": 1.2083958020989505,
+      "grad_norm": 0.17504557436505902,
+      "learning_rate": 0.00014924235601034672,
+      "loss": 0.0272,
+      "step": 806
+    },
+    {
+      "epoch": 1.209895052473763,
+      "grad_norm": 0.1707709517487399,
+      "learning_rate": 0.00014909037536151409,
+      "loss": 0.0203,
+      "step": 807
+    },
+    {
+      "epoch": 1.2113943028485756,
+      "grad_norm": 0.12956420487730475,
+      "learning_rate": 0.00014893824517488464,
+      "loss": 0.0185,
+      "step": 808
+    },
+    {
+      "epoch": 1.2128935532233882,
+      "grad_norm": 0.16545558068001448,
+      "learning_rate": 0.0001487859659138733,
+      "loss": 0.017,
+      "step": 809
+    },
+    {
+      "epoch": 1.214392803598201,
+      "grad_norm": 0.10015980215539548,
+      "learning_rate": 0.00014863353804234905,
+      "loss": 0.0144,
+      "step": 810
+    },
+    {
+      "epoch": 1.2158920539730136,
+      "grad_norm": 0.07422865061562015,
+      "learning_rate": 0.00014848096202463372,
+      "loss": 0.0141,
+      "step": 811
+    },
+    {
+      "epoch": 1.2173913043478262,
+      "grad_norm": 0.16658923532851774,
+      "learning_rate": 0.00014832823832550024,
+      "loss": 0.0171,
+      "step": 812
+    },
+    {
+      "epoch": 1.2188905547226387,
+      "grad_norm": 0.2952339023440661,
+      "learning_rate": 0.00014817536741017152,
+      "loss": 0.0249,
+      "step": 813
+    },
+    {
+      "epoch": 1.2203898050974513,
+      "grad_norm": 0.343627117743269,
+      "learning_rate": 0.0001480223497443189,
+      "loss": 0.0203,
+      "step": 814
+    },
+    {
+      "epoch": 1.221889055472264,
+      "grad_norm": 0.15860364996490584,
+      "learning_rate": 0.0001478691857940607,
+      "loss": 0.0173,
+      "step": 815
+    },
+    {
+      "epoch": 1.2233883058470765,
+      "grad_norm": 0.07397939793708072,
+      "learning_rate": 0.00014771587602596084,
+      "loss": 0.0165,
+      "step": 816
+    },
+    {
+      "epoch": 1.224887556221889,
+      "grad_norm": 0.09124884086427787,
+      "learning_rate": 0.00014756242090702756,
+      "loss": 0.0189,
+      "step": 817
+    },
+    {
+      "epoch": 1.2263868065967016,
+      "grad_norm": 0.06925619214926025,
+      "learning_rate": 0.0001474088209047116,
+      "loss": 0.0198,
+      "step": 818
+    },
+    {
+      "epoch": 1.2278860569715142,
+      "grad_norm": 0.11626344421046819,
+      "learning_rate": 0.00014725507648690543,
+      "loss": 0.0219,
+      "step": 819
+    },
+    {
+      "epoch": 1.2293853073463268,
+      "grad_norm": 0.18165886590704167,
+      "learning_rate": 0.000147101188121941,
+      "loss": 0.0289,
+      "step": 820
+    },
+    {
+      "epoch": 1.2308845577211394,
+      "grad_norm": 0.08731841646113968,
+      "learning_rate": 0.00014694715627858908,
+      "loss": 0.014,
+      "step": 821
+    },
+    {
+      "epoch": 1.232383808095952,
+      "grad_norm": 0.06959854220083778,
+      "learning_rate": 0.00014679298142605734,
+      "loss": 0.0114,
+      "step": 822
+    },
+    {
+      "epoch": 1.2338830584707647,
+      "grad_norm": 0.36508436863746124,
+      "learning_rate": 0.00014663866403398913,
+      "loss": 0.0244,
+      "step": 823
+    },
+    {
+      "epoch": 1.235382308845577,
+      "grad_norm": 0.09368142163812859,
+      "learning_rate": 0.000146484204572462,
+      "loss": 0.0174,
+      "step": 824
+    },
+    {
+      "epoch": 1.23688155922039,
+      "grad_norm": 0.08193058500216474,
+      "learning_rate": 0.00014632960351198618,
+      "loss": 0.0132,
+      "step": 825
+    },
+    {
+      "epoch": 1.2383808095952025,
+      "grad_norm": 0.15614447953538005,
+      "learning_rate": 0.00014617486132350343,
+      "loss": 0.0195,
+      "step": 826
+    },
+    {
+      "epoch": 1.239880059970015,
+      "grad_norm": 0.0779958529833662,
+      "learning_rate": 0.00014601997847838518,
+      "loss": 0.0108,
+      "step": 827
+    },
+    {
+      "epoch": 1.2413793103448276,
+      "grad_norm": 0.0888114278475023,
+      "learning_rate": 0.00014586495544843152,
+      "loss": 0.0147,
+      "step": 828
+    },
+    {
+      "epoch": 1.2428785607196402,
+      "grad_norm": 0.11955268016013557,
+      "learning_rate": 0.00014570979270586945,
+      "loss": 0.0201,
+      "step": 829
+    },
+    {
+      "epoch": 1.2443778110944528,
+      "grad_norm": 0.202954135904049,
+      "learning_rate": 0.00014555449072335157,
+      "loss": 0.0301,
+      "step": 830
+    },
+    {
+      "epoch": 1.2458770614692654,
+      "grad_norm": 0.18587709894174873,
+      "learning_rate": 0.00014539904997395468,
+      "loss": 0.0368,
+      "step": 831
+    },
+    {
+      "epoch": 1.247376311844078,
+      "grad_norm": 0.07571382735258467,
+      "learning_rate": 0.00014524347093117828,
+      "loss": 0.0131,
+      "step": 832
+    },
+    {
+      "epoch": 1.2488755622188905,
+      "grad_norm": 0.1422916655610059,
+      "learning_rate": 0.00014508775406894307,
+      "loss": 0.021,
+      "step": 833
+    },
+    {
+      "epoch": 1.250374812593703,
+      "grad_norm": 0.13587129669974282,
+      "learning_rate": 0.00014493189986158965,
+      "loss": 0.0162,
+      "step": 834
+    },
+    {
+      "epoch": 1.2518740629685157,
+      "grad_norm": 0.08417093744852887,
+      "learning_rate": 0.00014477590878387696,
+      "loss": 0.0167,
+      "step": 835
+    },
+    {
+      "epoch": 1.2533733133433285,
+      "grad_norm": 0.18400438316412573,
+      "learning_rate": 0.00014461978131098088,
+      "loss": 0.0235,
+      "step": 836
+    },
+    {
+      "epoch": 1.2548725637181408,
+      "grad_norm": 0.07333689889670233,
+      "learning_rate": 0.00014446351791849276,
+      "loss": 0.0097,
+      "step": 837
+    },
+    {
+      "epoch": 1.2563718140929536,
+      "grad_norm": 0.16562856351156585,
+      "learning_rate": 0.00014430711908241798,
+      "loss": 0.0481,
+      "step": 838
+    },
+    {
+      "epoch": 1.2578710644677662,
+      "grad_norm": 0.11118124090310846,
+      "learning_rate": 0.00014415058527917452,
+      "loss": 0.0214,
+      "step": 839
+    },
+    {
+      "epoch": 1.2593703148425788,
+      "grad_norm": 0.18915002411038295,
+      "learning_rate": 0.00014399391698559152,
+      "loss": 0.0272,
+      "step": 840
+    },
+    {
+      "epoch": 1.2608695652173914,
+      "grad_norm": 0.054608561460708256,
+      "learning_rate": 0.00014383711467890774,
+      "loss": 0.0103,
+      "step": 841
+    },
+    {
+      "epoch": 1.262368815592204,
+      "grad_norm": 0.11913677742871757,
+      "learning_rate": 0.00014368017883677024,
+      "loss": 0.0264,
+      "step": 842
+    },
+    {
+      "epoch": 1.2638680659670165,
+      "grad_norm": 0.14267753773889416,
+      "learning_rate": 0.00014352310993723277,
+      "loss": 0.0257,
+      "step": 843
+    },
+    {
+      "epoch": 1.265367316341829,
+      "grad_norm": 0.20047662055314494,
+      "learning_rate": 0.00014336590845875446,
+      "loss": 0.0277,
+      "step": 844
+    },
+    {
+      "epoch": 1.2668665667166417,
+      "grad_norm": 0.13109076906839104,
+      "learning_rate": 0.00014320857488019824,
+      "loss": 0.0227,
+      "step": 845
+    },
+    {
+      "epoch": 1.2683658170914542,
+      "grad_norm": 0.12458461929595395,
+      "learning_rate": 0.00014305110968082952,
+      "loss": 0.0164,
+      "step": 846
+    },
+    {
+      "epoch": 1.2698650674662668,
+      "grad_norm": 0.09746106663140872,
+      "learning_rate": 0.0001428935133403146,
+      "loss": 0.0124,
+      "step": 847
+    },
+    {
+      "epoch": 1.2713643178410794,
+      "grad_norm": 0.12378914670534663,
+      "learning_rate": 0.00014273578633871927,
+      "loss": 0.0233,
+      "step": 848
+    },
+    {
+      "epoch": 1.272863568215892,
+      "grad_norm": 0.06854962001087733,
+      "learning_rate": 0.00014257792915650728,
+      "loss": 0.011,
+      "step": 849
+    },
+    {
+      "epoch": 1.2743628185907045,
+      "grad_norm": 0.10385595925660064,
+      "learning_rate": 0.00014241994227453901,
+      "loss": 0.0102,
+      "step": 850
+    },
+    {
+      "epoch": 1.2758620689655173,
+      "grad_norm": 0.11662093941235979,
+      "learning_rate": 0.00014226182617406996,
+      "loss": 0.0128,
+      "step": 851
+    },
+    {
+      "epoch": 1.2773613193403297,
+      "grad_norm": 0.17930864492543278,
+      "learning_rate": 0.00014210358133674912,
+      "loss": 0.0126,
+      "step": 852
+    },
+    {
+      "epoch": 1.2788605697151425,
+      "grad_norm": 0.1372485077515871,
+      "learning_rate": 0.00014194520824461771,
+      "loss": 0.012,
+      "step": 853
+    },
+    {
+      "epoch": 1.280359820089955,
+      "grad_norm": 0.16345335022814164,
+      "learning_rate": 0.0001417867073801077,
+      "loss": 0.0249,
+      "step": 854
+    },
+    {
+      "epoch": 1.2818590704647677,
+      "grad_norm": 0.0997906139397832,
+      "learning_rate": 0.00014162807922604012,
+      "loss": 0.012,
+      "step": 855
+    },
+    {
+      "epoch": 1.2833583208395802,
+      "grad_norm": 0.09163084280524973,
+      "learning_rate": 0.00014146932426562392,
+      "loss": 0.0147,
+      "step": 856
+    },
+    {
+      "epoch": 1.2848575712143928,
+      "grad_norm": 0.15802403576676588,
+      "learning_rate": 0.0001413104429824542,
+      "loss": 0.0116,
+      "step": 857
+    },
+    {
+      "epoch": 1.2863568215892054,
+      "grad_norm": 0.0854686973988408,
+      "learning_rate": 0.00014115143586051088,
+      "loss": 0.0128,
+      "step": 858
+    },
+    {
+      "epoch": 1.287856071964018,
+      "grad_norm": 0.06726116883117537,
+      "learning_rate": 0.00014099230338415728,
+      "loss": 0.0121,
+      "step": 859
+    },
+    {
+      "epoch": 1.2893553223388305,
+      "grad_norm": 0.11037334160421672,
+      "learning_rate": 0.00014083304603813848,
+      "loss": 0.0103,
+      "step": 860
+    },
+    {
+      "epoch": 1.2908545727136431,
+      "grad_norm": 0.12332181597901376,
+      "learning_rate": 0.00014067366430758004,
+      "loss": 0.0224,
+      "step": 861
+    },
+    {
+      "epoch": 1.2923538230884557,
+      "grad_norm": 0.18194326760827853,
+      "learning_rate": 0.00014051415867798628,
+      "loss": 0.0352,
+      "step": 862
+    },
+    {
+      "epoch": 1.2938530734632683,
+      "grad_norm": 0.08430952227790076,
+      "learning_rate": 0.00014035452963523902,
+      "loss": 0.0142,
+      "step": 863
+    },
+    {
+      "epoch": 1.295352323838081,
+      "grad_norm": 0.07124275034999716,
+      "learning_rate": 0.00014019477766559604,
+      "loss": 0.007,
+      "step": 864
+    },
+    {
+      "epoch": 1.2968515742128934,
+      "grad_norm": 0.09273754187553625,
+      "learning_rate": 0.00014003490325568954,
+      "loss": 0.0149,
+      "step": 865
+    },
+    {
+      "epoch": 1.2983508245877062,
+      "grad_norm": 0.24506349645358916,
+      "learning_rate": 0.00013987490689252463,
+      "loss": 0.035,
+      "step": 866
+    },
+    {
+      "epoch": 1.2998500749625188,
+      "grad_norm": 0.12312219048398249,
+      "learning_rate": 0.00013971478906347806,
+      "loss": 0.0211,
+      "step": 867
+    },
+    {
+      "epoch": 1.3013493253373314,
+      "grad_norm": 0.09854369818513668,
+      "learning_rate": 0.00013955455025629651,
+      "loss": 0.0168,
+      "step": 868
+    },
+    {
+      "epoch": 1.302848575712144,
+      "grad_norm": 0.18077070256023667,
+      "learning_rate": 0.00013939419095909512,
+      "loss": 0.0265,
+      "step": 869
+    },
+    {
+      "epoch": 1.3043478260869565,
+      "grad_norm": 0.08104201839225528,
+      "learning_rate": 0.00013923371166035616,
+      "loss": 0.0141,
+      "step": 870
+    },
+    {
+      "epoch": 1.3058470764617691,
+      "grad_norm": 0.05669225042597822,
+      "learning_rate": 0.00013907311284892736,
+      "loss": 0.0067,
+      "step": 871
+    },
+    {
+      "epoch": 1.3073463268365817,
+      "grad_norm": 0.170396059391146,
+      "learning_rate": 0.0001389123950140206,
+      "loss": 0.0144,
+      "step": 872
+    },
+    {
+      "epoch": 1.3088455772113943,
+      "grad_norm": 0.09404710563168223,
+      "learning_rate": 0.0001387515586452103,
+      "loss": 0.015,
+      "step": 873
+    },
+    {
+      "epoch": 1.3103448275862069,
+      "grad_norm": 0.07336649457841225,
+      "learning_rate": 0.00013859060423243187,
+      "loss": 0.0134,
+      "step": 874
+    },
+    {
+      "epoch": 1.3118440779610194,
+      "grad_norm": 0.15102499216151205,
+      "learning_rate": 0.00013842953226598037,
+      "loss": 0.031,
+      "step": 875
+    },
+    {
+      "epoch": 1.313343328335832,
+      "grad_norm": 0.11710688194401644,
+      "learning_rate": 0.000138268343236509,
+      "loss": 0.0228,
+      "step": 876
+    },
+    {
+      "epoch": 1.3148425787106448,
+      "grad_norm": 0.08364534811984702,
+      "learning_rate": 0.00013810703763502744,
+      "loss": 0.0129,
+      "step": 877
+    },
+    {
+      "epoch": 1.3163418290854572,
+      "grad_norm": 0.08023322549168056,
+      "learning_rate": 0.00013794561595290052,
+      "loss": 0.0198,
+      "step": 878
+    },
+    {
+      "epoch": 1.31784107946027,
+      "grad_norm": 0.10058428662726825,
+      "learning_rate": 0.00013778407868184672,
+      "loss": 0.0288,
+      "step": 879
+    },
+    {
+      "epoch": 1.3193403298350825,
+      "grad_norm": 0.10611806737953573,
+      "learning_rate": 0.00013762242631393655,
+      "loss": 0.0154,
+      "step": 880
+    },
+    {
+      "epoch": 1.3208395802098951,
+      "grad_norm": 0.07758538286432358,
+      "learning_rate": 0.00013746065934159123,
+      "loss": 0.0133,
+      "step": 881
+    },
+    {
+      "epoch": 1.3223388305847077,
+      "grad_norm": 0.11661318831041984,
+      "learning_rate": 0.0001372987782575809,
+      "loss": 0.0147,
+      "step": 882
+    },
+    {
+      "epoch": 1.3238380809595203,
+      "grad_norm": 0.10345149158644346,
+      "learning_rate": 0.00013713678355502351,
+      "loss": 0.0175,
+      "step": 883
+    },
+    {
+      "epoch": 1.3253373313343328,
+      "grad_norm": 0.17899088086579465,
+      "learning_rate": 0.00013697467572738295,
+      "loss": 0.0159,
+      "step": 884
+    },
+    {
+      "epoch": 1.3268365817091454,
+      "grad_norm": 0.10366071968769505,
+      "learning_rate": 0.00013681245526846783,
+      "loss": 0.0185,
+      "step": 885
+    },
+    {
+      "epoch": 1.328335832083958,
+      "grad_norm": 0.2320430785565242,
+      "learning_rate": 0.00013665012267242974,
+      "loss": 0.0109,
+      "step": 886
+    },
+    {
+      "epoch": 1.3298350824587706,
+      "grad_norm": 0.2165729911004686,
+      "learning_rate": 0.00013648767843376196,
+      "loss": 0.0353,
+      "step": 887
+    },
+    {
+      "epoch": 1.3313343328335832,
+      "grad_norm": 0.090948134296178,
+      "learning_rate": 0.00013632512304729785,
+      "loss": 0.0117,
+      "step": 888
+    },
+    {
+      "epoch": 1.3328335832083957,
+      "grad_norm": 0.1723357033638432,
+      "learning_rate": 0.00013616245700820922,
+      "loss": 0.0211,
+      "step": 889
+    },
+    {
+      "epoch": 1.3343328335832085,
+      "grad_norm": 0.12328155582579832,
+      "learning_rate": 0.00013599968081200514,
+      "loss": 0.0222,
+      "step": 890
+    },
+    {
+      "epoch": 1.3358320839580209,
+      "grad_norm": 0.1148672322347492,
+      "learning_rate": 0.00013583679495453,
+      "loss": 0.017,
+      "step": 891
+    },
+    {
+      "epoch": 1.3373313343328337,
+      "grad_norm": 0.08259323816085176,
+      "learning_rate": 0.00013567379993196252,
+      "loss": 0.0109,
+      "step": 892
+    },
+    {
+      "epoch": 1.338830584707646,
+      "grad_norm": 0.1149875760609526,
+      "learning_rate": 0.0001355106962408137,
+      "loss": 0.0202,
+      "step": 893
+    },
+    {
+      "epoch": 1.3403298350824588,
+      "grad_norm": 0.08902107605424671,
+      "learning_rate": 0.00013534748437792573,
+      "loss": 0.0146,
+      "step": 894
+    },
+    {
+      "epoch": 1.3418290854572714,
+      "grad_norm": 0.2630218492094372,
+      "learning_rate": 0.00013518416484047018,
+      "loss": 0.0327,
+      "step": 895
+    },
+    {
+      "epoch": 1.343328335832084,
+      "grad_norm": 0.09214551614879145,
+      "learning_rate": 0.00013502073812594675,
+      "loss": 0.0149,
+      "step": 896
+    },
+    {
+      "epoch": 1.3448275862068966,
+      "grad_norm": 0.08385569860592369,
+      "learning_rate": 0.00013485720473218154,
+      "loss": 0.0182,
+      "step": 897
+    },
+    {
+      "epoch": 1.3463268365817092,
+      "grad_norm": 0.10306630082691921,
+      "learning_rate": 0.00013469356515732558,
+      "loss": 0.0153,
+      "step": 898
+    },
+    {
+      "epoch": 1.3478260869565217,
+      "grad_norm": 0.07604456448727595,
+      "learning_rate": 0.00013452981989985348,
+      "loss": 0.0129,
+      "step": 899
+    },
+    {
+      "epoch": 1.3493253373313343,
+      "grad_norm": 0.08738991305136598,
+      "learning_rate": 0.00013436596945856164,
+      "loss": 0.0159,
+      "step": 900
+    },
+    {
+      "epoch": 1.3508245877061469,
+      "grad_norm": 0.15222706121553015,
+      "learning_rate": 0.00013420201433256689,
+      "loss": 0.021,
+      "step": 901
+    },
+    {
+      "epoch": 1.3523238380809595,
+      "grad_norm": 0.1306031031322151,
+      "learning_rate": 0.00013403795502130503,
+      "loss": 0.0167,
+      "step": 902
+    },
+    {
+      "epoch": 1.353823088455772,
+      "grad_norm": 0.08808517355752062,
+      "learning_rate": 0.00013387379202452917,
+      "loss": 0.0139,
+      "step": 903
+    },
+    {
+      "epoch": 1.3553223388305846,
+      "grad_norm": 0.11211685787913292,
+      "learning_rate": 0.00013370952584230823,
+      "loss": 0.0137,
+      "step": 904
+    },
+    {
+      "epoch": 1.3568215892053974,
+      "grad_norm": 0.08778023618067164,
+      "learning_rate": 0.00013354515697502553,
+      "loss": 0.0107,
+      "step": 905
+    },
+    {
+      "epoch": 1.3583208395802098,
+      "grad_norm": 0.06388026832690835,
+      "learning_rate": 0.0001333806859233771,
+      "loss": 0.0098,
+      "step": 906
+    },
+    {
+      "epoch": 1.3598200899550226,
+      "grad_norm": 0.1148047545564487,
+      "learning_rate": 0.00013321611318837032,
+      "loss": 0.0105,
+      "step": 907
+    },
+    {
+      "epoch": 1.3613193403298351,
+      "grad_norm": 0.10499549294742563,
+      "learning_rate": 0.0001330514392713223,
+      "loss": 0.0186,
+      "step": 908
+    },
+    {
+      "epoch": 1.3628185907046477,
+      "grad_norm": 0.07183000371069133,
+      "learning_rate": 0.00013288666467385833,
+      "loss": 0.0053,
+      "step": 909
+    },
+    {
+      "epoch": 1.3643178410794603,
+      "grad_norm": 0.09527252116065761,
+      "learning_rate": 0.0001327217898979104,
+      "loss": 0.0144,
+      "step": 910
+    },
+    {
+      "epoch": 1.3658170914542729,
+      "grad_norm": 0.17631678204685763,
+      "learning_rate": 0.00013255681544571568,
+      "loss": 0.0232,
+      "step": 911
+    },
+    {
+      "epoch": 1.3673163418290855,
+      "grad_norm": 0.09351491513727911,
+      "learning_rate": 0.00013239174181981495,
+      "loss": 0.0167,
+      "step": 912
+    },
+    {
+      "epoch": 1.368815592203898,
+      "grad_norm": 0.13425537805814766,
+      "learning_rate": 0.00013222656952305113,
+      "loss": 0.0191,
+      "step": 913
+    },
+    {
+      "epoch": 1.3703148425787106,
+      "grad_norm": 0.14230677646930492,
+      "learning_rate": 0.00013206129905856765,
+      "loss": 0.0229,
+      "step": 914
+    },
+    {
+      "epoch": 1.3718140929535232,
+      "grad_norm": 0.09153923973277442,
+      "learning_rate": 0.00013189593092980702,
+      "loss": 0.0131,
+      "step": 915
+    },
+    {
+      "epoch": 1.3733133433283358,
+      "grad_norm": 0.0578622169127209,
+      "learning_rate": 0.00013173046564050924,
+      "loss": 0.0113,
+      "step": 916
+    },
+    {
+      "epoch": 1.3748125937031483,
+      "grad_norm": 0.13891677776816042,
+      "learning_rate": 0.00013156490369471027,
+      "loss": 0.0137,
+      "step": 917
+    },
+    {
+      "epoch": 1.3763118440779611,
+      "grad_norm": 0.04903272701124958,
+      "learning_rate": 0.00013139924559674052,
+      "loss": 0.0062,
+      "step": 918
+    },
+    {
+      "epoch": 1.3778110944527735,
+      "grad_norm": 0.08762005054912386,
+      "learning_rate": 0.00013123349185122327,
+      "loss": 0.0183,
+      "step": 919
+    },
+    {
+      "epoch": 1.3793103448275863,
+      "grad_norm": 0.08308919330791038,
+      "learning_rate": 0.0001310676429630732,
+      "loss": 0.0134,
+      "step": 920
+    },
+    {
+      "epoch": 1.3808095952023989,
+      "grad_norm": 0.05992308927118659,
+      "learning_rate": 0.00013090169943749476,
+      "loss": 0.0128,
+      "step": 921
+    },
+    {
+      "epoch": 1.3823088455772115,
+      "grad_norm": 0.06850172645375537,
+      "learning_rate": 0.00013073566177998074,
+      "loss": 0.0078,
+      "step": 922
+    },
+    {
+      "epoch": 1.383808095952024,
+      "grad_norm": 0.11509963104785531,
+      "learning_rate": 0.00013056953049631057,
+      "loss": 0.0143,
+      "step": 923
+    },
+    {
+      "epoch": 1.3853073463268366,
+      "grad_norm": 0.24649040007973805,
+      "learning_rate": 0.00013040330609254903,
+      "loss": 0.0124,
+      "step": 924
+    },
+    {
+      "epoch": 1.3868065967016492,
+      "grad_norm": 0.07928186572736103,
+      "learning_rate": 0.00013023698907504446,
+      "loss": 0.0159,
+      "step": 925
+    },
+    {
+      "epoch": 1.3883058470764618,
+      "grad_norm": 0.11963166407145638,
+      "learning_rate": 0.00013007057995042732,
+      "loss": 0.033,
+      "step": 926
+    },
+    {
+      "epoch": 1.3898050974512743,
+      "grad_norm": 0.09256628228543629,
+      "learning_rate": 0.00012990407922560868,
+      "loss": 0.0168,
+      "step": 927
+    },
+    {
+      "epoch": 1.391304347826087,
+      "grad_norm": 0.12551820161013388,
+      "learning_rate": 0.0001297374874077786,
+      "loss": 0.0128,
+      "step": 928
+    },
+    {
+      "epoch": 1.3928035982008995,
+      "grad_norm": 0.09396953539790104,
+      "learning_rate": 0.00012957080500440468,
+      "loss": 0.0126,
+      "step": 929
+    },
+    {
+      "epoch": 1.394302848575712,
+      "grad_norm": 0.0814330557561346,
+      "learning_rate": 0.0001294040325232304,
+      "loss": 0.0109,
+      "step": 930
+    },
+    {
+      "epoch": 1.3958020989505249,
+      "grad_norm": 0.09170961671586021,
+      "learning_rate": 0.00012923717047227368,
+      "loss": 0.0123,
+      "step": 931
+    },
+    {
+      "epoch": 1.3973013493253372,
+      "grad_norm": 0.12429481547170583,
+      "learning_rate": 0.00012907021935982526,
+      "loss": 0.017,
+      "step": 932
+    },
+    {
+      "epoch": 1.39880059970015,
+      "grad_norm": 0.07547138071359946,
+      "learning_rate": 0.00012890317969444716,
+      "loss": 0.0101,
+      "step": 933
+    },
+    {
+      "epoch": 1.4002998500749624,
+      "grad_norm": 0.10947780038801162,
+      "learning_rate": 0.00012873605198497124,
+      "loss": 0.0122,
+      "step": 934
+    },
+    {
+      "epoch": 1.4017991004497752,
+      "grad_norm": 0.14974745222283475,
+      "learning_rate": 0.00012856883674049736,
+      "loss": 0.021,
+      "step": 935
+    },
+    {
+      "epoch": 1.4032983508245878,
+      "grad_norm": 0.06311397898354858,
+      "learning_rate": 0.00012840153447039228,
+      "loss": 0.0097,
+      "step": 936
+    },
+    {
+      "epoch": 1.4047976011994003,
+      "grad_norm": 0.18221914938913733,
+      "learning_rate": 0.00012823414568428768,
+      "loss": 0.031,
+      "step": 937
+    },
+    {
+      "epoch": 1.406296851574213,
+      "grad_norm": 0.08321813040032747,
+      "learning_rate": 0.0001280666708920788,
+      "loss": 0.0096,
+      "step": 938
+    },
+    {
+      "epoch": 1.4077961019490255,
+      "grad_norm": 0.08421041313135452,
+      "learning_rate": 0.00012789911060392294,
+      "loss": 0.0148,
+      "step": 939
+    },
+    {
+      "epoch": 1.409295352323838,
+      "grad_norm": 0.06717497940090396,
+      "learning_rate": 0.00012773146533023782,
+      "loss": 0.0148,
+      "step": 940
+    },
+    {
+      "epoch": 1.4107946026986506,
+      "grad_norm": 0.09168806606704279,
+      "learning_rate": 0.0001275637355816999,
+      "loss": 0.0171,
+      "step": 941
+    },
+    {
+      "epoch": 1.4122938530734632,
+      "grad_norm": 0.10477491574939178,
+      "learning_rate": 0.00012739592186924328,
+      "loss": 0.0134,
+      "step": 942
+    },
+    {
+      "epoch": 1.4137931034482758,
+      "grad_norm": 0.12092733743202747,
+      "learning_rate": 0.00012722802470405744,
+      "loss": 0.0167,
+      "step": 943
+    },
+    {
+      "epoch": 1.4152923538230884,
+      "grad_norm": 0.14068944332954328,
+      "learning_rate": 0.00012706004459758636,
+      "loss": 0.0296,
+      "step": 944
+    },
+    {
+      "epoch": 1.416791604197901,
+      "grad_norm": 0.14137864202064285,
+      "learning_rate": 0.00012689198206152657,
+      "loss": 0.0134,
+      "step": 945
+    },
+    {
+      "epoch": 1.4182908545727138,
+      "grad_norm": 0.0860572351607591,
+      "learning_rate": 0.00012672383760782568,
+      "loss": 0.0178,
+      "step": 946
+    },
+    {
+      "epoch": 1.419790104947526,
+      "grad_norm": 0.07404000519398271,
+      "learning_rate": 0.00012655561174868088,
+      "loss": 0.0102,
+      "step": 947
+    },
+    {
+      "epoch": 1.421289355322339,
+      "grad_norm": 0.11582588947336643,
+      "learning_rate": 0.0001263873049965373,
+      "loss": 0.0186,
+      "step": 948
+    },
+    {
+      "epoch": 1.4227886056971515,
+      "grad_norm": 0.13400349083924593,
+      "learning_rate": 0.00012621891786408648,
+      "loss": 0.011,
+      "step": 949
+    },
+    {
+      "epoch": 1.424287856071964,
+      "grad_norm": 0.20636194375809655,
+      "learning_rate": 0.00012605045086426487,
+      "loss": 0.0199,
+      "step": 950
+    },
+    {
+      "epoch": 1.4257871064467766,
+      "grad_norm": 0.03725195400505915,
+      "learning_rate": 0.00012588190451025207,
+      "loss": 0.0046,
+      "step": 951
+    },
+    {
+      "epoch": 1.4272863568215892,
+      "grad_norm": 0.05892090161081139,
+      "learning_rate": 0.00012571327931546963,
+      "loss": 0.0097,
+      "step": 952
+    },
+    {
+      "epoch": 1.4287856071964018,
+      "grad_norm": 0.12490487802223602,
+      "learning_rate": 0.00012554457579357905,
+      "loss": 0.009,
+      "step": 953
+    },
+    {
+      "epoch": 1.4302848575712144,
+      "grad_norm": 0.11179141701581286,
+      "learning_rate": 0.00012537579445848058,
+      "loss": 0.0183,
+      "step": 954
+    },
+    {
+      "epoch": 1.431784107946027,
+      "grad_norm": 0.06972546035600884,
+      "learning_rate": 0.0001252069358243114,
+      "loss": 0.0122,
+      "step": 955
+    },
+    {
+      "epoch": 1.4332833583208395,
+      "grad_norm": 0.09131261727248277,
+      "learning_rate": 0.00012503800040544416,
+      "loss": 0.0095,
+      "step": 956
+    },
+    {
+      "epoch": 1.434782608695652,
+      "grad_norm": 0.1109811036679139,
+      "learning_rate": 0.0001248689887164855,
+      "loss": 0.0105,
+      "step": 957
+    },
+    {
+      "epoch": 1.4362818590704647,
+      "grad_norm": 0.07591084965117433,
+      "learning_rate": 0.00012469990127227432,
+      "loss": 0.0138,
+      "step": 958
+    },
+    {
+      "epoch": 1.4377811094452775,
+      "grad_norm": 0.20788394817699887,
+      "learning_rate": 0.00012453073858788026,
+      "loss": 0.0384,
+      "step": 959
+    },
+    {
+      "epoch": 1.4392803598200898,
+      "grad_norm": 0.12391628024007592,
+      "learning_rate": 0.00012436150117860225,
+      "loss": 0.0136,
+      "step": 960
+    },
+    {
+      "epoch": 1.4407796101949026,
+      "grad_norm": 0.12604999899644895,
+      "learning_rate": 0.00012419218955996676,
+      "loss": 0.0211,
+      "step": 961
+    },
+    {
+      "epoch": 1.4422788605697152,
+      "grad_norm": 0.09347292600591614,
+      "learning_rate": 0.00012402280424772639,
+      "loss": 0.0056,
+      "step": 962
+    },
+    {
+      "epoch": 1.4437781109445278,
+      "grad_norm": 0.09362580278945488,
+      "learning_rate": 0.0001238533457578581,
+      "loss": 0.0149,
+      "step": 963
+    },
+    {
+      "epoch": 1.4452773613193404,
+      "grad_norm": 0.13450314184024556,
+      "learning_rate": 0.0001236838146065619,
+      "loss": 0.0183,
+      "step": 964
+    },
+    {
+      "epoch": 1.446776611694153,
+      "grad_norm": 0.06683370298885613,
+      "learning_rate": 0.000123514211310259,
+      "loss": 0.0114,
+      "step": 965
+    },
+    {
+      "epoch": 1.4482758620689655,
+      "grad_norm": 0.07533136377894585,
+      "learning_rate": 0.00012334453638559057,
+      "loss": 0.0088,
+      "step": 966
+    },
+    {
+      "epoch": 1.449775112443778,
+      "grad_norm": 0.11385337434889103,
+      "learning_rate": 0.00012317479034941573,
+      "loss": 0.0132,
+      "step": 967
+    },
+    {
+      "epoch": 1.4512743628185907,
+      "grad_norm": 0.08660001131021804,
+      "learning_rate": 0.00012300497371881046,
+      "loss": 0.0183,
+      "step": 968
+    },
+    {
+      "epoch": 1.4527736131934033,
+      "grad_norm": 0.21035359781149526,
+      "learning_rate": 0.00012283508701106557,
+      "loss": 0.015,
+      "step": 969
+    },
+    {
+      "epoch": 1.4542728635682158,
+      "grad_norm": 0.14422586173075505,
+      "learning_rate": 0.0001226651307436855,
+      "loss": 0.0205,
+      "step": 970
+    },
+    {
+      "epoch": 1.4557721139430284,
+      "grad_norm": 0.147239584775246,
+      "learning_rate": 0.0001224951054343865,
+      "loss": 0.0103,
+      "step": 971
+    },
+    {
+      "epoch": 1.4572713643178412,
+      "grad_norm": 0.11512591488947908,
+      "learning_rate": 0.00012232501160109514,
+      "loss": 0.0154,
+      "step": 972
+    },
+    {
+      "epoch": 1.4587706146926536,
+      "grad_norm": 0.16628564022136066,
+      "learning_rate": 0.00012215484976194676,
+      "loss": 0.0208,
+      "step": 973
+    },
+    {
+      "epoch": 1.4602698650674664,
+      "grad_norm": 0.1705959708232762,
+      "learning_rate": 0.00012198462043528376,
+      "loss": 0.0147,
+      "step": 974
+    },
+    {
+      "epoch": 1.461769115442279,
+      "grad_norm": 0.1385993826378694,
+      "learning_rate": 0.00012181432413965428,
+      "loss": 0.0122,
+      "step": 975
+    },
+    {
+      "epoch": 1.4632683658170915,
+      "grad_norm": 0.14300288286624654,
+      "learning_rate": 0.00012164396139381029,
+      "loss": 0.0196,
+      "step": 976
+    },
+    {
+      "epoch": 1.464767616191904,
+      "grad_norm": 0.080459269082652,
+      "learning_rate": 0.00012147353271670634,
+      "loss": 0.0085,
+      "step": 977
+    },
+    {
+      "epoch": 1.4662668665667167,
+      "grad_norm": 0.14683345902187586,
+      "learning_rate": 0.00012130303862749767,
+      "loss": 0.0337,
+      "step": 978
+    },
+    {
+      "epoch": 1.4677661169415293,
+      "grad_norm": 0.08820586327672093,
+      "learning_rate": 0.00012113247964553888,
+      "loss": 0.0091,
+      "step": 979
+    },
+    {
+      "epoch": 1.4692653673163418,
+      "grad_norm": 0.11691373222525352,
+      "learning_rate": 0.0001209618562903822,
+      "loss": 0.0138,
+      "step": 980
+    },
+    {
+      "epoch": 1.4707646176911544,
+      "grad_norm": 0.13161051562985052,
+      "learning_rate": 0.00012079116908177593,
+      "loss": 0.0087,
+      "step": 981
+    },
+    {
+      "epoch": 1.472263868065967,
+      "grad_norm": 0.09613499888195008,
+      "learning_rate": 0.00012062041853966298,
+      "loss": 0.0088,
+      "step": 982
+    },
+    {
+      "epoch": 1.4737631184407796,
+      "grad_norm": 0.09195368296362892,
+      "learning_rate": 0.00012044960518417903,
+      "loss": 0.0136,
+      "step": 983
+    },
+    {
+      "epoch": 1.4752623688155921,
+      "grad_norm": 0.17465872281858535,
+      "learning_rate": 0.00012027872953565125,
+      "loss": 0.0206,
+      "step": 984
+    },
+    {
+      "epoch": 1.4767616191904047,
+      "grad_norm": 0.07836679172273597,
+      "learning_rate": 0.00012010779211459648,
+      "loss": 0.011,
+      "step": 985
+    },
+    {
+      "epoch": 1.4782608695652173,
+      "grad_norm": 0.09365663057692926,
+      "learning_rate": 0.00011993679344171973,
+      "loss": 0.0182,
+      "step": 986
+    },
+    {
+      "epoch": 1.47976011994003,
+      "grad_norm": 0.12682030561564137,
+      "learning_rate": 0.00011976573403791262,
+      "loss": 0.0089,
+      "step": 987
+    },
+    {
+      "epoch": 1.4812593703148424,
+      "grad_norm": 0.08393483257842721,
+      "learning_rate": 0.00011959461442425177,
+      "loss": 0.0114,
+      "step": 988
+    },
+    {
+      "epoch": 1.4827586206896552,
+      "grad_norm": 0.1050372238110733,
+      "learning_rate": 0.0001194234351219972,
+      "loss": 0.0203,
+      "step": 989
+    },
+    {
+      "epoch": 1.4842578710644678,
+      "grad_norm": 0.11112289986216253,
+      "learning_rate": 0.00011925219665259075,
+      "loss": 0.016,
+      "step": 990
+    },
+    {
+      "epoch": 1.4857571214392804,
+      "grad_norm": 0.10020695139140749,
+      "learning_rate": 0.00011908089953765449,
+      "loss": 0.0081,
+      "step": 991
+    },
+    {
+      "epoch": 1.487256371814093,
+      "grad_norm": 0.11236596316087173,
+      "learning_rate": 0.00011890954429898912,
+      "loss": 0.0198,
+      "step": 992
+    },
+    {
+      "epoch": 1.4887556221889056,
+      "grad_norm": 0.07348295688717253,
+      "learning_rate": 0.00011873813145857249,
+      "loss": 0.0103,
+      "step": 993
+    },
+    {
+      "epoch": 1.4902548725637181,
+      "grad_norm": 0.07744821371483936,
+      "learning_rate": 0.00011856666153855776,
+      "loss": 0.0096,
+      "step": 994
+    },
+    {
+      "epoch": 1.4917541229385307,
+      "grad_norm": 0.09069561623382377,
+      "learning_rate": 0.00011839513506127203,
+      "loss": 0.0104,
+      "step": 995
+    },
+    {
+      "epoch": 1.4932533733133433,
+      "grad_norm": 0.05709624694520552,
+      "learning_rate": 0.00011822355254921478,
+      "loss": 0.0072,
+      "step": 996
+    },
+    {
+      "epoch": 1.4947526236881559,
+      "grad_norm": 0.11572106025581866,
+      "learning_rate": 0.00011805191452505602,
+      "loss": 0.0242,
+      "step": 997
+    },
+    {
+      "epoch": 1.4962518740629684,
+      "grad_norm": 0.09230690991056163,
+      "learning_rate": 0.00011788022151163495,
+      "loss": 0.017,
+      "step": 998
+    },
+    {
+      "epoch": 1.497751124437781,
+      "grad_norm": 0.10875800562076483,
+      "learning_rate": 0.00011770847403195834,
+      "loss": 0.0132,
+      "step": 999
+    },
+    {
+      "epoch": 1.4992503748125938,
+      "grad_norm": 0.13231704496690558,
+      "learning_rate": 0.00011753667260919872,
+      "loss": 0.0195,
+      "step": 1000
+    },
+    {
+      "epoch": 1.5007496251874062,
+      "grad_norm": 0.1281968401653238,
+      "learning_rate": 0.00011736481776669306,
+      "loss": 0.012,
+      "step": 1001
+    },
+    {
+      "epoch": 1.502248875562219,
+      "grad_norm": 0.10072623997610383,
+      "learning_rate": 0.00011719291002794096,
+      "loss": 0.0118,
+      "step": 1002
+    },
+    {
+      "epoch": 1.5037481259370313,
+      "grad_norm": 0.20173093622273117,
+      "learning_rate": 0.00011702094991660326,
+      "loss": 0.0302,
+      "step": 1003
+    },
+    {
+      "epoch": 1.5052473763118441,
+      "grad_norm": 0.24868628674117718,
+      "learning_rate": 0.00011684893795650027,
+      "loss": 0.026,
+      "step": 1004
+    },
+    {
+      "epoch": 1.5067466266866567,
+      "grad_norm": 0.20149856701683075,
+      "learning_rate": 0.00011667687467161024,
+      "loss": 0.0315,
+      "step": 1005
+    },
+    {
+      "epoch": 1.5082458770614693,
+      "grad_norm": 0.09025687720729765,
+      "learning_rate": 0.00011650476058606777,
+      "loss": 0.0128,
+      "step": 1006
+    },
+    {
+      "epoch": 1.5097451274362819,
+      "grad_norm": 0.07593193449503319,
+      "learning_rate": 0.00011633259622416224,
+      "loss": 0.0077,
+      "step": 1007
+    },
+    {
+      "epoch": 1.5112443778110944,
+      "grad_norm": 0.14453173196748317,
+      "learning_rate": 0.00011616038211033613,
+      "loss": 0.0206,
+      "step": 1008
+    },
+    {
+      "epoch": 1.512743628185907,
+      "grad_norm": 0.12215216306236011,
+      "learning_rate": 0.0001159881187691835,
+      "loss": 0.0138,
+      "step": 1009
+    },
+    {
+      "epoch": 1.5142428785607196,
+      "grad_norm": 0.1512560808287919,
+      "learning_rate": 0.00011581580672544838,
+      "loss": 0.0251,
+      "step": 1010
+    },
+    {
+      "epoch": 1.5157421289355324,
+      "grad_norm": 0.08198991344869536,
+      "learning_rate": 0.0001156434465040231,
+      "loss": 0.0116,
+      "step": 1011
+    },
+    {
+      "epoch": 1.5172413793103448,
+      "grad_norm": 0.07771015418992609,
+      "learning_rate": 0.00011547103862994684,
+      "loss": 0.0176,
+      "step": 1012
+    },
+    {
+      "epoch": 1.5187406296851576,
+      "grad_norm": 0.1523429461936693,
+      "learning_rate": 0.00011529858362840382,
+      "loss": 0.0189,
+      "step": 1013
+    },
+    {
+      "epoch": 1.52023988005997,
+      "grad_norm": 0.11328394560469286,
+      "learning_rate": 0.00011512608202472194,
+      "loss": 0.0161,
+      "step": 1014
+    },
+    {
+      "epoch": 1.5217391304347827,
+      "grad_norm": 0.11980943303869283,
+      "learning_rate": 0.00011495353434437098,
+      "loss": 0.0259,
+      "step": 1015
+    },
+    {
+      "epoch": 1.523238380809595,
+      "grad_norm": 0.1359052987718006,
+      "learning_rate": 0.00011478094111296109,
+      "loss": 0.0167,
+      "step": 1016
+    },
+    {
+      "epoch": 1.5247376311844079,
+      "grad_norm": 0.06578587863029016,
+      "learning_rate": 0.00011460830285624118,
+      "loss": 0.0104,
+      "step": 1017
+    },
+    {
+      "epoch": 1.5262368815592204,
+      "grad_norm": 0.09597998810321515,
+      "learning_rate": 0.00011443562010009731,
+      "loss": 0.0082,
+      "step": 1018
+    },
+    {
+      "epoch": 1.527736131934033,
+      "grad_norm": 0.12322964223446362,
+      "learning_rate": 0.00011426289337055119,
+      "loss": 0.009,
+      "step": 1019
+    },
+    {
+      "epoch": 1.5292353823088456,
+      "grad_norm": 0.0719193270536759,
+      "learning_rate": 0.00011409012319375827,
+      "loss": 0.0106,
+      "step": 1020
+    },
+    {
+      "epoch": 1.5307346326836582,
+      "grad_norm": 0.09940735793245362,
+      "learning_rate": 0.00011391731009600654,
+      "loss": 0.0092,
+      "step": 1021
+    },
+    {
+      "epoch": 1.5322338830584707,
+      "grad_norm": 0.07250731895990385,
+      "learning_rate": 0.00011374445460371466,
+      "loss": 0.0105,
+      "step": 1022
+    },
+    {
+      "epoch": 1.5337331334332833,
+      "grad_norm": 0.05856277224745514,
+      "learning_rate": 0.00011357155724343045,
+      "loss": 0.0101,
+      "step": 1023
+    },
+    {
+      "epoch": 1.535232383808096,
+      "grad_norm": 0.07803522533264978,
+      "learning_rate": 0.00011339861854182922,
+      "loss": 0.0157,
+      "step": 1024
+    },
+    {
+      "epoch": 1.5367316341829085,
+      "grad_norm": 0.23493948718254026,
+      "learning_rate": 0.00011322563902571226,
+      "loss": 0.027,
+      "step": 1025
+    },
+    {
+      "epoch": 1.5382308845577213,
+      "grad_norm": 0.044799835235369186,
+      "learning_rate": 0.00011305261922200519,
+      "loss": 0.0093,
+      "step": 1026
+    },
+    {
+      "epoch": 1.5397301349325336,
+      "grad_norm": 0.09126298677383313,
+      "learning_rate": 0.0001128795596577563,
+      "loss": 0.0162,
+      "step": 1027
+    },
+    {
+      "epoch": 1.5412293853073464,
+      "grad_norm": 0.1038456479026185,
+      "learning_rate": 0.00011270646086013505,
+      "loss": 0.0131,
+      "step": 1028
+    },
+    {
+      "epoch": 1.5427286356821588,
+      "grad_norm": 0.07105696856441536,
+      "learning_rate": 0.00011253332335643043,
+      "loss": 0.013,
+      "step": 1029
+    },
+    {
+      "epoch": 1.5442278860569716,
+      "grad_norm": 0.07978824589108958,
+      "learning_rate": 0.00011236014767404927,
+      "loss": 0.0119,
+      "step": 1030
+    },
+    {
+      "epoch": 1.545727136431784,
+      "grad_norm": 0.10620185121687949,
+      "learning_rate": 0.00011218693434051475,
+      "loss": 0.0194,
+      "step": 1031
+    },
+    {
+      "epoch": 1.5472263868065967,
+      "grad_norm": 0.06597817870162546,
+      "learning_rate": 0.00011201368388346471,
+      "loss": 0.0122,
+      "step": 1032
+    },
+    {
+      "epoch": 1.5487256371814093,
+      "grad_norm": 0.1391579923371684,
+      "learning_rate": 0.00011184039683065013,
+      "loss": 0.0159,
+      "step": 1033
+    },
+    {
+      "epoch": 1.550224887556222,
+      "grad_norm": 0.0583734987893039,
+      "learning_rate": 0.00011166707370993333,
+      "loss": 0.0089,
+      "step": 1034
+    },
+    {
+      "epoch": 1.5517241379310345,
+      "grad_norm": 0.05835705798534313,
+      "learning_rate": 0.00011149371504928668,
+      "loss": 0.0095,
+      "step": 1035
+    },
+    {
+      "epoch": 1.553223388305847,
+      "grad_norm": 0.1162274585062899,
+      "learning_rate": 0.0001113203213767907,
+      "loss": 0.0093,
+      "step": 1036
+    },
+    {
+      "epoch": 1.5547226386806596,
+      "grad_norm": 0.12624340143523036,
+      "learning_rate": 0.00011114689322063255,
+      "loss": 0.0133,
+      "step": 1037
+    },
+    {
+      "epoch": 1.5562218890554722,
+      "grad_norm": 0.124541965655317,
+      "learning_rate": 0.00011097343110910452,
+      "loss": 0.0128,
+      "step": 1038
+    },
+    {
+      "epoch": 1.557721139430285,
+      "grad_norm": 0.1032713740550231,
+      "learning_rate": 0.0001107999355706023,
+      "loss": 0.0137,
+      "step": 1039
+    },
+    {
+      "epoch": 1.5592203898050974,
+      "grad_norm": 0.11800022175627153,
+      "learning_rate": 0.00011062640713362333,
+      "loss": 0.0158,
+      "step": 1040
+    },
+    {
+      "epoch": 1.5607196401799102,
+      "grad_norm": 0.1302341773557835,
+      "learning_rate": 0.00011045284632676536,
+      "loss": 0.0146,
+      "step": 1041
+    },
+    {
+      "epoch": 1.5622188905547225,
+      "grad_norm": 0.1546164674351709,
+      "learning_rate": 0.00011027925367872469,
+      "loss": 0.0388,
+      "step": 1042
+    },
+    {
+      "epoch": 1.5637181409295353,
+      "grad_norm": 0.05168069492010779,
+      "learning_rate": 0.00011010562971829463,
+      "loss": 0.0084,
+      "step": 1043
+    },
+    {
+      "epoch": 1.5652173913043477,
+      "grad_norm": 0.13440988075774132,
+      "learning_rate": 0.00010993197497436391,
+      "loss": 0.0192,
+      "step": 1044
+    },
+    {
+      "epoch": 1.5667166416791605,
+      "grad_norm": 0.1296790212762601,
+      "learning_rate": 0.00010975828997591495,
+      "loss": 0.0157,
+      "step": 1045
+    },
+    {
+      "epoch": 1.568215892053973,
+      "grad_norm": 0.076210180946536,
+      "learning_rate": 0.00010958457525202241,
+      "loss": 0.0173,
+      "step": 1046
+    },
+    {
+      "epoch": 1.5697151424287856,
+      "grad_norm": 0.07970386766287546,
+      "learning_rate": 0.00010941083133185146,
+      "loss": 0.01,
+      "step": 1047
+    },
+    {
+      "epoch": 1.5712143928035982,
+      "grad_norm": 0.17171859330243994,
+      "learning_rate": 0.00010923705874465618,
+      "loss": 0.0361,
+      "step": 1048
+    },
+    {
+      "epoch": 1.5727136431784108,
+      "grad_norm": 0.18708680882057213,
+      "learning_rate": 0.00010906325801977804,
+      "loss": 0.0139,
+      "step": 1049
+    },
+    {
+      "epoch": 1.5742128935532234,
+      "grad_norm": 0.11969812919320348,
+      "learning_rate": 0.00010888942968664417,
+      "loss": 0.0175,
+      "step": 1050
+    },
+    {
+      "epoch": 1.575712143928036,
+      "grad_norm": 0.15840014306422295,
+      "learning_rate": 0.00010871557427476583,
+      "loss": 0.0169,
+      "step": 1051
+    },
+    {
+      "epoch": 1.5772113943028487,
+      "grad_norm": 0.13291084407135878,
+      "learning_rate": 0.00010854169231373676,
+      "loss": 0.0142,
+      "step": 1052
+    },
+    {
+      "epoch": 1.578710644677661,
+      "grad_norm": 0.09889951255469409,
+      "learning_rate": 0.00010836778433323158,
+      "loss": 0.0168,
+      "step": 1053
+    },
+    {
+      "epoch": 1.580209895052474,
+      "grad_norm": 0.1018450276304327,
+      "learning_rate": 0.0001081938508630041,
+      "loss": 0.0191,
+      "step": 1054
+    },
+    {
+      "epoch": 1.5817091454272862,
+      "grad_norm": 0.11305829451392016,
+      "learning_rate": 0.00010801989243288589,
+      "loss": 0.0183,
+      "step": 1055
+    },
+    {
+      "epoch": 1.583208395802099,
+      "grad_norm": 0.0819216552478812,
+      "learning_rate": 0.0001078459095727845,
+      "loss": 0.0131,
+      "step": 1056
+    },
+    {
+      "epoch": 1.5847076461769114,
+      "grad_norm": 0.12687192279885026,
+      "learning_rate": 0.00010767190281268187,
+      "loss": 0.0222,
+      "step": 1057
+    },
+    {
+      "epoch": 1.5862068965517242,
+      "grad_norm": 0.07362844017008341,
+      "learning_rate": 0.00010749787268263279,
+      "loss": 0.0101,
+      "step": 1058
+    },
+    {
+      "epoch": 1.5877061469265368,
+      "grad_norm": 0.06079600135413754,
+      "learning_rate": 0.00010732381971276318,
+      "loss": 0.0097,
+      "step": 1059
+    },
+    {
+      "epoch": 1.5892053973013494,
+      "grad_norm": 0.06661318973072765,
+      "learning_rate": 0.0001071497444332686,
+      "loss": 0.0126,
+      "step": 1060
+    },
+    {
+      "epoch": 1.590704647676162,
+      "grad_norm": 0.15022551284140978,
+      "learning_rate": 0.00010697564737441252,
+      "loss": 0.0157,
+      "step": 1061
+    },
+    {
+      "epoch": 1.5922038980509745,
+      "grad_norm": 0.12196412330052128,
+      "learning_rate": 0.00010680152906652483,
+      "loss": 0.0216,
+      "step": 1062
+    },
+    {
+      "epoch": 1.593703148425787,
+      "grad_norm": 0.10057060869798758,
+      "learning_rate": 0.00010662739004000005,
+      "loss": 0.0146,
+      "step": 1063
+    },
+    {
+      "epoch": 1.5952023988005997,
+      "grad_norm": 0.10678495246345038,
+      "learning_rate": 0.00010645323082529581,
+      "loss": 0.02,
+      "step": 1064
+    },
+    {
+      "epoch": 1.5967016491754122,
+      "grad_norm": 0.07484074937474987,
+      "learning_rate": 0.00010627905195293135,
+      "loss": 0.0115,
+      "step": 1065
+    },
+    {
+      "epoch": 1.5982008995502248,
+      "grad_norm": 0.07512912626075581,
+      "learning_rate": 0.00010610485395348571,
+      "loss": 0.0123,
+      "step": 1066
+    },
+    {
+      "epoch": 1.5997001499250376,
+      "grad_norm": 0.08792742242003913,
+      "learning_rate": 0.00010593063735759618,
+      "loss": 0.0129,
+      "step": 1067
+    },
+    {
+      "epoch": 1.60119940029985,
+      "grad_norm": 0.1622803893742513,
+      "learning_rate": 0.00010575640269595675,
+      "loss": 0.0238,
+      "step": 1068
+    },
+    {
+      "epoch": 1.6026986506746628,
+      "grad_norm": 0.13411386012680196,
+      "learning_rate": 0.00010558215049931638,
+      "loss": 0.0255,
+      "step": 1069
+    },
+    {
+      "epoch": 1.6041979010494751,
+      "grad_norm": 0.11582998216511321,
+      "learning_rate": 0.00010540788129847756,
+      "loss": 0.0089,
+      "step": 1070
+    },
+    {
+      "epoch": 1.605697151424288,
+      "grad_norm": 0.08034952754248964,
+      "learning_rate": 0.0001052335956242944,
+      "loss": 0.0142,
+      "step": 1071
+    },
+    {
+      "epoch": 1.6071964017991005,
+      "grad_norm": 0.0763174800207671,
+      "learning_rate": 0.00010505929400767134,
+      "loss": 0.0113,
+      "step": 1072
+    },
+    {
+      "epoch": 1.608695652173913,
+      "grad_norm": 0.11861926773103704,
+      "learning_rate": 0.00010488497697956135,
+      "loss": 0.0199,
+      "step": 1073
+    },
+    {
+      "epoch": 1.6101949025487257,
+      "grad_norm": 0.10574532949429762,
+      "learning_rate": 0.00010471064507096426,
+      "loss": 0.0187,
+      "step": 1074
+    },
+    {
+      "epoch": 1.6116941529235382,
+      "grad_norm": 0.13423987932488682,
+      "learning_rate": 0.00010453629881292538,
+      "loss": 0.0157,
+      "step": 1075
+    },
+    {
+      "epoch": 1.6131934032983508,
+      "grad_norm": 0.07764456401541044,
+      "learning_rate": 0.00010436193873653361,
+      "loss": 0.0088,
+      "step": 1076
+    },
+    {
+      "epoch": 1.6146926536731634,
+      "grad_norm": 0.0659718362577642,
+      "learning_rate": 0.00010418756537291996,
+      "loss": 0.0125,
+      "step": 1077
+    },
+    {
+      "epoch": 1.616191904047976,
+      "grad_norm": 0.0851649381893994,
+      "learning_rate": 0.00010401317925325598,
+      "loss": 0.0116,
+      "step": 1078
+    },
+    {
+      "epoch": 1.6176911544227885,
+      "grad_norm": 0.11170558677294636,
+      "learning_rate": 0.00010383878090875201,
+      "loss": 0.0148,
+      "step": 1079
+    },
+    {
+      "epoch": 1.6191904047976013,
+      "grad_norm": 0.2357508518298938,
+      "learning_rate": 0.00010366437087065564,
+      "loss": 0.0266,
+      "step": 1080
+    },
+    {
+      "epoch": 1.6206896551724137,
+      "grad_norm": 0.15444209087426583,
+      "learning_rate": 0.00010348994967025012,
+      "loss": 0.0215,
+      "step": 1081
+    },
+    {
+      "epoch": 1.6221889055472265,
+      "grad_norm": 0.20019270471396725,
+      "learning_rate": 0.00010331551783885263,
+      "loss": 0.0283,
+      "step": 1082
+    },
+    {
+      "epoch": 1.6236881559220389,
+      "grad_norm": 0.06468184788830145,
+      "learning_rate": 0.00010314107590781284,
+      "loss": 0.0103,
+      "step": 1083
+    },
+    {
+      "epoch": 1.6251874062968517,
+      "grad_norm": 0.06501689028923738,
+      "learning_rate": 0.00010296662440851108,
+      "loss": 0.0089,
+      "step": 1084
+    },
+    {
+      "epoch": 1.626686656671664,
+      "grad_norm": 0.07867494972789672,
+      "learning_rate": 0.0001027921638723569,
+      "loss": 0.012,
+      "step": 1085
+    },
+    {
+      "epoch": 1.6281859070464768,
+      "grad_norm": 0.053082944965822235,
+      "learning_rate": 0.00010261769483078733,
+      "loss": 0.0088,
+      "step": 1086
+    },
+    {
+      "epoch": 1.6296851574212894,
+      "grad_norm": 0.10268869576541034,
+      "learning_rate": 0.00010244321781526533,
+      "loss": 0.0107,
+      "step": 1087
+    },
+    {
+      "epoch": 1.631184407796102,
+      "grad_norm": 0.09311046814429251,
+      "learning_rate": 0.00010226873335727816,
+      "loss": 0.0145,
+      "step": 1088
+    },
+    {
+      "epoch": 1.6326836581709145,
+      "grad_norm": 0.05118955528054772,
+      "learning_rate": 0.0001020942419883357,
+      "loss": 0.0066,
+      "step": 1089
+    },
+    {
+      "epoch": 1.6341829085457271,
+      "grad_norm": 0.06776547119993927,
+      "learning_rate": 0.000101919744239969,
+      "loss": 0.0094,
+      "step": 1090
+    },
+    {
+      "epoch": 1.6356821589205397,
+      "grad_norm": 0.0735948110395489,
+      "learning_rate": 0.00010174524064372837,
+      "loss": 0.0083,
+      "step": 1091
+    },
+    {
+      "epoch": 1.6371814092953523,
+      "grad_norm": 0.16353843455702943,
+      "learning_rate": 0.00010157073173118208,
+      "loss": 0.0191,
+      "step": 1092
+    },
+    {
+      "epoch": 1.638680659670165,
+      "grad_norm": 0.07729308076766099,
+      "learning_rate": 0.00010139621803391455,
+      "loss": 0.0113,
+      "step": 1093
+    },
+    {
+      "epoch": 1.6401799100449774,
+      "grad_norm": 0.07065440168839313,
+      "learning_rate": 0.00010122170008352472,
+      "loss": 0.0111,
+      "step": 1094
+    },
+    {
+      "epoch": 1.6416791604197902,
+      "grad_norm": 0.08052200609350296,
+      "learning_rate": 0.00010104717841162458,
+      "loss": 0.0086,
+      "step": 1095
+    },
+    {
+      "epoch": 1.6431784107946026,
+      "grad_norm": 0.11666520112605404,
+      "learning_rate": 0.0001008726535498374,
+      "loss": 0.0144,
+      "step": 1096
+    },
+    {
+      "epoch": 1.6446776611694154,
+      "grad_norm": 0.0883241856124164,
+      "learning_rate": 0.00010069812602979615,
+      "loss": 0.0148,
+      "step": 1097
+    },
+    {
+      "epoch": 1.6461769115442277,
+      "grad_norm": 0.05690807601176237,
+      "learning_rate": 0.00010052359638314195,
+      "loss": 0.01,
+      "step": 1098
+    },
+    {
+      "epoch": 1.6476761619190405,
+      "grad_norm": 0.11692485410485566,
+      "learning_rate": 0.00010034906514152238,
+      "loss": 0.0204,
+      "step": 1099
+    },
+    {
+      "epoch": 1.6491754122938531,
+      "grad_norm": 0.04741744619163783,
+      "learning_rate": 0.00010017453283658984,
+      "loss": 0.0094,
+      "step": 1100
+    },
+    {
+      "epoch": 1.6506746626686657,
+      "grad_norm": 0.07678788315580089,
+      "learning_rate": 0.0001,
+      "loss": 0.0135,
+      "step": 1101
+    },
+    {
+      "epoch": 1.6521739130434783,
+      "grad_norm": 0.08617460350136832,
+      "learning_rate": 9.98254671634102e-05,
+      "loss": 0.0204,
+      "step": 1102
+    },
+    {
+      "epoch": 1.6536731634182908,
+      "grad_norm": 0.03884006559581419,
+      "learning_rate": 9.965093485847767e-05,
+      "loss": 0.0065,
+      "step": 1103
+    },
+    {
+      "epoch": 1.6551724137931034,
+      "grad_norm": 0.08381647185517602,
+      "learning_rate": 9.947640361685804e-05,
+      "loss": 0.0135,
+      "step": 1104
+    },
+    {
+      "epoch": 1.656671664167916,
+      "grad_norm": 0.08865105073519507,
+      "learning_rate": 9.930187397020386e-05,
+      "loss": 0.0195,
+      "step": 1105
+    },
+    {
+      "epoch": 1.6581709145427288,
+      "grad_norm": 0.04051930233354598,
+      "learning_rate": 9.912734645016263e-05,
+      "loss": 0.0059,
+      "step": 1106
+    },
+    {
+      "epoch": 1.6596701649175412,
+      "grad_norm": 0.08542703043188983,
+      "learning_rate": 9.895282158837545e-05,
+      "loss": 0.009,
+      "step": 1107
+    },
+    {
+      "epoch": 1.661169415292354,
+      "grad_norm": 0.09553707719347451,
+      "learning_rate": 9.877829991647528e-05,
+      "loss": 0.0069,
+      "step": 1108
+    },
+    {
+      "epoch": 1.6626686656671663,
+      "grad_norm": 0.17415049236905383,
+      "learning_rate": 9.860378196608549e-05,
+      "loss": 0.0121,
+      "step": 1109
+    },
+    {
+      "epoch": 1.6641679160419791,
+      "grad_norm": 0.12950093216787284,
+      "learning_rate": 9.842926826881796e-05,
+      "loss": 0.018,
+      "step": 1110
+    },
+    {
+      "epoch": 1.6656671664167915,
+      "grad_norm": 0.06414792177091079,
+      "learning_rate": 9.825475935627165e-05,
+      "loss": 0.0069,
+      "step": 1111
+    },
+    {
+      "epoch": 1.6671664167916043,
+      "grad_norm": 0.1994947874307293,
+      "learning_rate": 9.808025576003104e-05,
+      "loss": 0.0318,
+      "step": 1112
+    },
+    {
+      "epoch": 1.6686656671664168,
+      "grad_norm": 0.08133150905555013,
+      "learning_rate": 9.790575801166432e-05,
+      "loss": 0.0143,
+      "step": 1113
+    },
+    {
+      "epoch": 1.6701649175412294,
+      "grad_norm": 0.16043323092034265,
+      "learning_rate": 9.773126664272186e-05,
+      "loss": 0.0229,
+      "step": 1114
+    },
+    {
+      "epoch": 1.671664167916042,
+      "grad_norm": 0.13438362563029066,
+      "learning_rate": 9.755678218473469e-05,
+      "loss": 0.0136,
+      "step": 1115
+    },
+    {
+      "epoch": 1.6731634182908546,
+      "grad_norm": 0.07606777859064674,
+      "learning_rate": 9.73823051692127e-05,
+      "loss": 0.0082,
+      "step": 1116
+    },
+    {
+      "epoch": 1.6746626686656672,
+      "grad_norm": 0.04420437383237939,
+      "learning_rate": 9.720783612764314e-05,
+      "loss": 0.0085,
+      "step": 1117
+    },
+    {
+      "epoch": 1.6761619190404797,
+      "grad_norm": 0.06837921063159548,
+      "learning_rate": 9.703337559148892e-05,
+      "loss": 0.0058,
+      "step": 1118
+    },
+    {
+      "epoch": 1.6776611694152923,
+      "grad_norm": 0.10263259805785357,
+      "learning_rate": 9.685892409218717e-05,
+      "loss": 0.0146,
+      "step": 1119
+    },
+    {
+      "epoch": 1.6791604197901049,
+      "grad_norm": 0.06638028522741139,
+      "learning_rate": 9.668448216114739e-05,
+      "loss": 0.0056,
+      "step": 1120
+    },
+    {
+      "epoch": 1.6806596701649177,
+      "grad_norm": 0.07281657562169337,
+      "learning_rate": 9.651005032974994e-05,
+      "loss": 0.0089,
+      "step": 1121
+    },
+    {
+      "epoch": 1.68215892053973,
+      "grad_norm": 0.07288356305521924,
+      "learning_rate": 9.633562912934436e-05,
+      "loss": 0.0055,
+      "step": 1122
+    },
+    {
+      "epoch": 1.6836581709145428,
+      "grad_norm": 0.08467351587105704,
+      "learning_rate": 9.616121909124801e-05,
+      "loss": 0.0201,
+      "step": 1123
+    },
+    {
+      "epoch": 1.6851574212893552,
+      "grad_norm": 0.08335449765159114,
+      "learning_rate": 9.598682074674405e-05,
+      "loss": 0.0105,
+      "step": 1124
+    },
+    {
+      "epoch": 1.686656671664168,
+      "grad_norm": 0.21044561607887044,
+      "learning_rate": 9.581243462708006e-05,
+      "loss": 0.025,
+      "step": 1125
+    },
+    {
+      "epoch": 1.6881559220389803,
+      "grad_norm": 0.10557438306733595,
+      "learning_rate": 9.563806126346642e-05,
+      "loss": 0.0165,
+      "step": 1126
+    },
+    {
+      "epoch": 1.6896551724137931,
+      "grad_norm": 0.05735360112575192,
+      "learning_rate": 9.546370118707463e-05,
+      "loss": 0.0085,
+      "step": 1127
+    },
+    {
+      "epoch": 1.6911544227886057,
+      "grad_norm": 0.08984041269285588,
+      "learning_rate": 9.528935492903575e-05,
+      "loss": 0.0148,
+      "step": 1128
+    },
+    {
+      "epoch": 1.6926536731634183,
+      "grad_norm": 0.05724955939727679,
+      "learning_rate": 9.511502302043868e-05,
+      "loss": 0.0065,
+      "step": 1129
+    },
+    {
+      "epoch": 1.6941529235382309,
+      "grad_norm": 0.0879374177037415,
+      "learning_rate": 9.494070599232868e-05,
+      "loss": 0.0152,
+      "step": 1130
+    },
+    {
+      "epoch": 1.6956521739130435,
+      "grad_norm": 0.10563935529155026,
+      "learning_rate": 9.476640437570562e-05,
+      "loss": 0.0184,
+      "step": 1131
+    },
+    {
+      "epoch": 1.697151424287856,
+      "grad_norm": 0.12099493793103591,
+      "learning_rate": 9.459211870152245e-05,
+      "loss": 0.0114,
+      "step": 1132
+    },
+    {
+      "epoch": 1.6986506746626686,
+      "grad_norm": 0.1493726682077126,
+      "learning_rate": 9.441784950068362e-05,
+      "loss": 0.0179,
+      "step": 1133
+    },
+    {
+      "epoch": 1.7001499250374814,
+      "grad_norm": 0.06855320485306697,
+      "learning_rate": 9.424359730404329e-05,
+      "loss": 0.0125,
+      "step": 1134
+    },
+    {
+      "epoch": 1.7016491754122938,
+      "grad_norm": 0.14082186154079793,
+      "learning_rate": 9.406936264240386e-05,
+      "loss": 0.0117,
+      "step": 1135
+    },
+    {
+      "epoch": 1.7031484257871066,
+      "grad_norm": 0.12243397983876568,
+      "learning_rate": 9.38951460465143e-05,
+      "loss": 0.0163,
+      "step": 1136
+    },
+    {
+      "epoch": 1.704647676161919,
+      "grad_norm": 0.07295698279479494,
+      "learning_rate": 9.372094804706867e-05,
+      "loss": 0.0142,
+      "step": 1137
+    },
+    {
+      "epoch": 1.7061469265367317,
+      "grad_norm": 0.07864559697537851,
+      "learning_rate": 9.354676917470422e-05,
+      "loss": 0.0121,
+      "step": 1138
+    },
+    {
+      "epoch": 1.707646176911544,
+      "grad_norm": 0.14262164347217002,
+      "learning_rate": 9.337260996000002e-05,
+      "loss": 0.0205,
+      "step": 1139
+    },
+    {
+      "epoch": 1.7091454272863569,
+      "grad_norm": 0.1533650338506401,
+      "learning_rate": 9.319847093347522e-05,
+      "loss": 0.0154,
+      "step": 1140
+    },
+    {
+      "epoch": 1.7106446776611695,
+      "grad_norm": 0.07096001994126981,
+      "learning_rate": 9.302435262558747e-05,
+      "loss": 0.0114,
+      "step": 1141
+    },
+    {
+      "epoch": 1.712143928035982,
+      "grad_norm": 0.087774881508882,
+      "learning_rate": 9.285025556673141e-05,
+      "loss": 0.0043,
+      "step": 1142
+    },
+    {
+      "epoch": 1.7136431784107946,
+      "grad_norm": 0.12753925625272527,
+      "learning_rate": 9.267618028723686e-05,
+      "loss": 0.0108,
+      "step": 1143
+    },
+    {
+      "epoch": 1.7151424287856072,
+      "grad_norm": 0.10939037079915889,
+      "learning_rate": 9.250212731736726e-05,
+      "loss": 0.0153,
+      "step": 1144
+    },
+    {
+      "epoch": 1.7166416791604198,
+      "grad_norm": 0.10489461806582333,
+      "learning_rate": 9.232809718731814e-05,
+      "loss": 0.0161,
+      "step": 1145
+    },
+    {
+      "epoch": 1.7181409295352323,
+      "grad_norm": 0.07579596628726051,
+      "learning_rate": 9.215409042721552e-05,
+      "loss": 0.0115,
+      "step": 1146
+    },
+    {
+      "epoch": 1.7196401799100451,
+      "grad_norm": 0.06387563602644145,
+      "learning_rate": 9.198010756711412e-05,
+      "loss": 0.0111,
+      "step": 1147
+    },
+    {
+      "epoch": 1.7211394302848575,
+      "grad_norm": 0.08515512442384568,
+      "learning_rate": 9.180614913699592e-05,
+      "loss": 0.0115,
+      "step": 1148
+    },
+    {
+      "epoch": 1.7226386806596703,
+      "grad_norm": 0.08561417637376921,
+      "learning_rate": 9.163221566676847e-05,
+      "loss": 0.009,
+      "step": 1149
+    },
+    {
+      "epoch": 1.7241379310344827,
+      "grad_norm": 0.10607108021738235,
+      "learning_rate": 9.145830768626327e-05,
+      "loss": 0.0181,
+      "step": 1150
+    },
+    {
+      "epoch": 1.7256371814092955,
+      "grad_norm": 0.1611719445615411,
+      "learning_rate": 9.128442572523417e-05,
+      "loss": 0.0087,
+      "step": 1151
+    },
+    {
+      "epoch": 1.7271364317841078,
+      "grad_norm": 0.04705216120726678,
+      "learning_rate": 9.111057031335585e-05,
+      "loss": 0.0045,
+      "step": 1152
+    },
+    {
+      "epoch": 1.7286356821589206,
+      "grad_norm": 0.07828025749540468,
+      "learning_rate": 9.093674198022201e-05,
+      "loss": 0.012,
+      "step": 1153
+    },
+    {
+      "epoch": 1.7301349325337332,
+      "grad_norm": 0.08108251835589797,
+      "learning_rate": 9.076294125534383e-05,
+      "loss": 0.0049,
+      "step": 1154
+    },
+    {
+      "epoch": 1.7316341829085458,
+      "grad_norm": 0.07381251899456927,
+      "learning_rate": 9.058916866814858e-05,
+      "loss": 0.0079,
+      "step": 1155
+    },
+    {
+      "epoch": 1.7331334332833583,
+      "grad_norm": 0.05566501298258808,
+      "learning_rate": 9.04154247479776e-05,
+      "loss": 0.0074,
+      "step": 1156
+    },
+    {
+      "epoch": 1.734632683658171,
+      "grad_norm": 0.07840451984596064,
+      "learning_rate": 9.024171002408506e-05,
+      "loss": 0.0105,
+      "step": 1157
+    },
+    {
+      "epoch": 1.7361319340329835,
+      "grad_norm": 0.06648110796795341,
+      "learning_rate": 9.006802502563612e-05,
+      "loss": 0.0077,
+      "step": 1158
+    },
+    {
+      "epoch": 1.737631184407796,
+      "grad_norm": 0.04739088697599641,
+      "learning_rate": 8.989437028170537e-05,
+      "loss": 0.0034,
+      "step": 1159
+    },
+    {
+      "epoch": 1.7391304347826086,
+      "grad_norm": 0.1271185732601886,
+      "learning_rate": 8.972074632127533e-05,
+      "loss": 0.0206,
+      "step": 1160
+    },
+    {
+      "epoch": 1.7406296851574212,
+      "grad_norm": 0.15903655821127385,
+      "learning_rate": 8.954715367323468e-05,
+      "loss": 0.0426,
+      "step": 1161
+    },
+    {
+      "epoch": 1.742128935532234,
+      "grad_norm": 0.10094008224842381,
+      "learning_rate": 8.93735928663767e-05,
+      "loss": 0.0199,
+      "step": 1162
+    },
+    {
+      "epoch": 1.7436281859070464,
+      "grad_norm": 0.07079249377199838,
+      "learning_rate": 8.920006442939772e-05,
+      "loss": 0.01,
+      "step": 1163
+    },
+    {
+      "epoch": 1.7451274362818592,
+      "grad_norm": 0.05078697634845653,
+      "learning_rate": 8.902656889089548e-05,
+      "loss": 0.0062,
+      "step": 1164
+    },
+    {
+      "epoch": 1.7466266866566715,
+      "grad_norm": 0.05448053623416766,
+      "learning_rate": 8.885310677936746e-05,
+      "loss": 0.0053,
+      "step": 1165
+    },
+    {
+      "epoch": 1.7481259370314843,
+      "grad_norm": 0.05217380091305068,
+      "learning_rate": 8.867967862320934e-05,
+      "loss": 0.0067,
+      "step": 1166
+    },
+    {
+      "epoch": 1.7496251874062967,
+      "grad_norm": 0.08488127237102319,
+      "learning_rate": 8.850628495071336e-05,
+      "loss": 0.0104,
+      "step": 1167
+    },
+    {
+      "epoch": 1.7511244377811095,
+      "grad_norm": 0.12524978803940454,
+      "learning_rate": 8.833292629006668e-05,
+      "loss": 0.0108,
+      "step": 1168
+    },
+    {
+      "epoch": 1.752623688155922,
+      "grad_norm": 0.09170930081644955,
+      "learning_rate": 8.81596031693499e-05,
+      "loss": 0.0124,
+      "step": 1169
+    },
+    {
+      "epoch": 1.7541229385307346,
+      "grad_norm": 0.0646182404385978,
+      "learning_rate": 8.79863161165353e-05,
+      "loss": 0.0099,
+      "step": 1170
+    },
+    {
+      "epoch": 1.7556221889055472,
+      "grad_norm": 0.060318759018513406,
+      "learning_rate": 8.781306565948528e-05,
+      "loss": 0.0082,
+      "step": 1171
+    },
+    {
+      "epoch": 1.7571214392803598,
+      "grad_norm": 0.07161845397887222,
+      "learning_rate": 8.763985232595075e-05,
+      "loss": 0.007,
+      "step": 1172
+    },
+    {
+      "epoch": 1.7586206896551724,
+      "grad_norm": 0.103642095909335,
+      "learning_rate": 8.746667664356956e-05,
+      "loss": 0.0161,
+      "step": 1173
+    },
+    {
+      "epoch": 1.760119940029985,
+      "grad_norm": 0.08550791524475224,
+      "learning_rate": 8.729353913986496e-05,
+      "loss": 0.0147,
+      "step": 1174
+    },
+    {
+      "epoch": 1.7616191904047978,
+      "grad_norm": 0.06894588907543088,
+      "learning_rate": 8.712044034224374e-05,
+      "loss": 0.0101,
+      "step": 1175
+    },
+    {
+      "epoch": 1.76311844077961,
+      "grad_norm": 0.09780901112803982,
+      "learning_rate": 8.694738077799488e-05,
+      "loss": 0.0181,
+      "step": 1176
+    },
+    {
+      "epoch": 1.764617691154423,
+      "grad_norm": 0.07230873928187732,
+      "learning_rate": 8.677436097428775e-05,
+      "loss": 0.0061,
+      "step": 1177
+    },
+    {
+      "epoch": 1.7661169415292353,
+      "grad_norm": 0.13013689151465993,
+      "learning_rate": 8.660138145817079e-05,
+      "loss": 0.0113,
+      "step": 1178
+    },
+    {
+      "epoch": 1.767616191904048,
+      "grad_norm": 0.06939024763422029,
+      "learning_rate": 8.642844275656957e-05,
+      "loss": 0.0105,
+      "step": 1179
+    },
+    {
+      "epoch": 1.7691154422788604,
+      "grad_norm": 0.15077017041092736,
+      "learning_rate": 8.625554539628535e-05,
+      "loss": 0.031,
+      "step": 1180
+    },
+    {
+      "epoch": 1.7706146926536732,
+      "grad_norm": 0.12499087340016703,
+      "learning_rate": 8.608268990399349e-05,
+      "loss": 0.0277,
+      "step": 1181
+    },
+    {
+      "epoch": 1.7721139430284858,
+      "grad_norm": 0.11591487119439382,
+      "learning_rate": 8.590987680624174e-05,
+      "loss": 0.0157,
+      "step": 1182
+    },
+    {
+      "epoch": 1.7736131934032984,
+      "grad_norm": 0.06002291691845019,
+      "learning_rate": 8.573710662944885e-05,
+      "loss": 0.0082,
+      "step": 1183
+    },
+    {
+      "epoch": 1.775112443778111,
+      "grad_norm": 0.1038340304012269,
+      "learning_rate": 8.55643798999027e-05,
+      "loss": 0.0119,
+      "step": 1184
+    },
+    {
+      "epoch": 1.7766116941529235,
+      "grad_norm": 0.09039360537283966,
+      "learning_rate": 8.539169714375885e-05,
+      "loss": 0.0138,
+      "step": 1185
+    },
+    {
+      "epoch": 1.778110944527736,
+      "grad_norm": 0.08968147616663716,
+      "learning_rate": 8.521905888703893e-05,
+      "loss": 0.0076,
+      "step": 1186
+    },
+    {
+      "epoch": 1.7796101949025487,
+      "grad_norm": 0.06865223310681502,
+      "learning_rate": 8.504646565562906e-05,
+      "loss": 0.0119,
+      "step": 1187
+    },
+    {
+      "epoch": 1.7811094452773615,
+      "grad_norm": 0.1473984760272428,
+      "learning_rate": 8.487391797527808e-05,
+      "loss": 0.0165,
+      "step": 1188
+    },
+    {
+      "epoch": 1.7826086956521738,
+      "grad_norm": 0.06834986338018911,
+      "learning_rate": 8.47014163715962e-05,
+      "loss": 0.0072,
+      "step": 1189
+    },
+    {
+      "epoch": 1.7841079460269866,
+      "grad_norm": 0.06542032785241195,
+      "learning_rate": 8.452896137005321e-05,
+      "loss": 0.0091,
+      "step": 1190
+    },
+    {
+      "epoch": 1.785607196401799,
+      "grad_norm": 0.15656986967497505,
+      "learning_rate": 8.435655349597689e-05,
+      "loss": 0.0153,
+      "step": 1191
+    },
+    {
+      "epoch": 1.7871064467766118,
+      "grad_norm": 0.05047228873977786,
+      "learning_rate": 8.418419327455164e-05,
+      "loss": 0.0066,
+      "step": 1192
+    },
+    {
+      "epoch": 1.7886056971514241,
+      "grad_norm": 0.05650765639240601,
+      "learning_rate": 8.401188123081653e-05,
+      "loss": 0.0074,
+      "step": 1193
+    },
+    {
+      "epoch": 1.790104947526237,
+      "grad_norm": 0.09070303775142517,
+      "learning_rate": 8.383961788966391e-05,
+      "loss": 0.0138,
+      "step": 1194
+    },
+    {
+      "epoch": 1.7916041979010495,
+      "grad_norm": 0.12727798124455994,
+      "learning_rate": 8.366740377583781e-05,
+      "loss": 0.0164,
+      "step": 1195
+    },
+    {
+      "epoch": 1.793103448275862,
+      "grad_norm": 0.08435936567425999,
+      "learning_rate": 8.349523941393224e-05,
+      "loss": 0.0142,
+      "step": 1196
+    },
+    {
+      "epoch": 1.7946026986506747,
+      "grad_norm": 0.081825272446604,
+      "learning_rate": 8.332312532838978e-05,
+      "loss": 0.0228,
+      "step": 1197
+    },
+    {
+      "epoch": 1.7961019490254873,
+      "grad_norm": 0.048776431818418335,
+      "learning_rate": 8.315106204349976e-05,
+      "loss": 0.0071,
+      "step": 1198
+    },
+    {
+      "epoch": 1.7976011994002998,
+      "grad_norm": 0.10574381543468121,
+      "learning_rate": 8.297905008339677e-05,
+      "loss": 0.0132,
+      "step": 1199
+    },
+    {
+      "epoch": 1.7991004497751124,
+      "grad_norm": 0.17066204868006726,
+      "learning_rate": 8.280708997205904e-05,
+      "loss": 0.013,
+      "step": 1200
+    },
+    {
+      "epoch": 1.800599700149925,
+      "grad_norm": 0.10668783815163481,
+      "learning_rate": 8.263518223330697e-05,
+      "loss": 0.0095,
+      "step": 1201
+    },
+    {
+      "epoch": 1.8020989505247376,
+      "grad_norm": 0.16567996473627022,
+      "learning_rate": 8.24633273908013e-05,
+      "loss": 0.0211,
+      "step": 1202
+    },
+    {
+      "epoch": 1.8035982008995504,
+      "grad_norm": 0.08443187319668671,
+      "learning_rate": 8.229152596804168e-05,
+      "loss": 0.0149,
+      "step": 1203
+    },
+    {
+      "epoch": 1.8050974512743627,
+      "grad_norm": 0.08806920008710213,
+      "learning_rate": 8.211977848836506e-05,
+      "loss": 0.0088,
+      "step": 1204
+    },
+    {
+      "epoch": 1.8065967016491755,
+      "grad_norm": 0.11415903061944807,
+      "learning_rate": 8.194808547494401e-05,
+      "loss": 0.0199,
+      "step": 1205
+    },
+    {
+      "epoch": 1.8080959520239879,
+      "grad_norm": 0.09961276004274798,
+      "learning_rate": 8.177644745078526e-05,
+      "loss": 0.0074,
+      "step": 1206
+    },
+    {
+      "epoch": 1.8095952023988007,
+      "grad_norm": 0.0891187100998618,
+      "learning_rate": 8.160486493872798e-05,
+      "loss": 0.0104,
+      "step": 1207
+    },
+    {
+      "epoch": 1.811094452773613,
+      "grad_norm": 0.12230491937645052,
+      "learning_rate": 8.14333384614423e-05,
+      "loss": 0.0144,
+      "step": 1208
+    },
+    {
+      "epoch": 1.8125937031484258,
+      "grad_norm": 0.05271611218754871,
+      "learning_rate": 8.126186854142752e-05,
+      "loss": 0.0113,
+      "step": 1209
+    },
+    {
+      "epoch": 1.8140929535232384,
+      "grad_norm": 0.12062545658852573,
+      "learning_rate": 8.109045570101086e-05,
+      "loss": 0.0184,
+      "step": 1210
+    },
+    {
+      "epoch": 1.815592203898051,
+      "grad_norm": 0.13604210998453206,
+      "learning_rate": 8.091910046234552e-05,
+      "loss": 0.023,
+      "step": 1211
+    },
+    {
+      "epoch": 1.8170914542728636,
+      "grad_norm": 0.10981765208088426,
+      "learning_rate": 8.074780334740928e-05,
+      "loss": 0.0097,
+      "step": 1212
+    },
+    {
+      "epoch": 1.8185907046476761,
+      "grad_norm": 0.13794783867457727,
+      "learning_rate": 8.057656487800282e-05,
+      "loss": 0.0267,
+      "step": 1213
+    },
+    {
+      "epoch": 1.8200899550224887,
+      "grad_norm": 0.1645787389155614,
+      "learning_rate": 8.040538557574822e-05,
+      "loss": 0.0257,
+      "step": 1214
+    },
+    {
+      "epoch": 1.8215892053973013,
+      "grad_norm": 0.1349840201186409,
+      "learning_rate": 8.023426596208739e-05,
+      "loss": 0.0135,
+      "step": 1215
+    },
+    {
+      "epoch": 1.823088455772114,
+      "grad_norm": 0.09749351855460471,
+      "learning_rate": 8.00632065582803e-05,
+      "loss": 0.0118,
+      "step": 1216
+    },
+    {
+      "epoch": 1.8245877061469264,
+      "grad_norm": 0.07799943350266138,
+      "learning_rate": 7.989220788540355e-05,
+      "loss": 0.0081,
+      "step": 1217
+    },
+    {
+      "epoch": 1.8260869565217392,
+      "grad_norm": 0.09827289301737527,
+      "learning_rate": 7.972127046434878e-05,
+      "loss": 0.0093,
+      "step": 1218
+    },
+    {
+      "epoch": 1.8275862068965516,
+      "grad_norm": 0.09805007129974536,
+      "learning_rate": 7.955039481582097e-05,
+      "loss": 0.015,
+      "step": 1219
+    },
+    {
+      "epoch": 1.8290854572713644,
+      "grad_norm": 0.09422086692048044,
+      "learning_rate": 7.937958146033705e-05,
+      "loss": 0.0087,
+      "step": 1220
+    },
+    {
+      "epoch": 1.8305847076461768,
+      "grad_norm": 0.0712763271172262,
+      "learning_rate": 7.920883091822408e-05,
+      "loss": 0.0131,
+      "step": 1221
+    },
+    {
+      "epoch": 1.8320839580209896,
+      "grad_norm": 0.11306696361521645,
+      "learning_rate": 7.903814370961784e-05,
+      "loss": 0.0125,
+      "step": 1222
+    },
+    {
+      "epoch": 1.8335832083958021,
+      "grad_norm": 0.08672321400900379,
+      "learning_rate": 7.886752035446114e-05,
+      "loss": 0.0133,
+      "step": 1223
+    },
+    {
+      "epoch": 1.8350824587706147,
+      "grad_norm": 0.1672675013757812,
+      "learning_rate": 7.869696137250235e-05,
+      "loss": 0.0309,
+      "step": 1224
+    },
+    {
+      "epoch": 1.8365817091454273,
+      "grad_norm": 0.10355505496804655,
+      "learning_rate": 7.852646728329368e-05,
+      "loss": 0.0104,
+      "step": 1225
+    },
+    {
+      "epoch": 1.8380809595202399,
+      "grad_norm": 0.11892893681660466,
+      "learning_rate": 7.835603860618972e-05,
+      "loss": 0.0142,
+      "step": 1226
+    },
+    {
+      "epoch": 1.8395802098950524,
+      "grad_norm": 0.14739440025777525,
+      "learning_rate": 7.818567586034577e-05,
+      "loss": 0.0204,
+      "step": 1227
+    },
+    {
+      "epoch": 1.841079460269865,
+      "grad_norm": 0.08697589155480914,
+      "learning_rate": 7.801537956471624e-05,
+      "loss": 0.0095,
+      "step": 1228
+    },
+    {
+      "epoch": 1.8425787106446778,
+      "grad_norm": 0.0581146762167273,
+      "learning_rate": 7.784515023805328e-05,
+      "loss": 0.0078,
+      "step": 1229
+    },
+    {
+      "epoch": 1.8440779610194902,
+      "grad_norm": 0.08434253900090809,
+      "learning_rate": 7.767498839890488e-05,
+      "loss": 0.0075,
+      "step": 1230
+    },
+    {
+      "epoch": 1.845577211394303,
+      "grad_norm": 0.10900161640713057,
+      "learning_rate": 7.750489456561352e-05,
+      "loss": 0.0191,
+      "step": 1231
+    },
+    {
+      "epoch": 1.8470764617691153,
+      "grad_norm": 0.10994173678131566,
+      "learning_rate": 7.733486925631447e-05,
+      "loss": 0.0115,
+      "step": 1232
+    },
+    {
+      "epoch": 1.8485757121439281,
+      "grad_norm": 0.1171064895868367,
+      "learning_rate": 7.716491298893442e-05,
+      "loss": 0.014,
+      "step": 1233
+    },
+    {
+      "epoch": 1.8500749625187405,
+      "grad_norm": 0.10594334669279865,
+      "learning_rate": 7.699502628118958e-05,
+      "loss": 0.0144,
+      "step": 1234
+    },
+    {
+      "epoch": 1.8515742128935533,
+      "grad_norm": 0.054006355912338475,
+      "learning_rate": 7.682520965058428e-05,
+      "loss": 0.0065,
+      "step": 1235
+    },
+    {
+      "epoch": 1.8530734632683659,
+      "grad_norm": 0.05698555933294013,
+      "learning_rate": 7.66554636144095e-05,
+      "loss": 0.0099,
+      "step": 1236
+    },
+    {
+      "epoch": 1.8545727136431784,
+      "grad_norm": 0.15549806710918523,
+      "learning_rate": 7.6485788689741e-05,
+      "loss": 0.0202,
+      "step": 1237
+    },
+    {
+      "epoch": 1.856071964017991,
+      "grad_norm": 0.1409960439001676,
+      "learning_rate": 7.631618539343814e-05,
+      "loss": 0.0123,
+      "step": 1238
+    },
+    {
+      "epoch": 1.8575712143928036,
+      "grad_norm": 0.10436639288234584,
+      "learning_rate": 7.614665424214193e-05,
+      "loss": 0.0102,
+      "step": 1239
+    },
+    {
+      "epoch": 1.8590704647676162,
+      "grad_norm": 0.3067562050568254,
+      "learning_rate": 7.597719575227364e-05,
+      "loss": 0.0202,
+      "step": 1240
+    },
+    {
+      "epoch": 1.8605697151424287,
+      "grad_norm": 0.09367926022511773,
+      "learning_rate": 7.580781044003324e-05,
+      "loss": 0.0173,
+      "step": 1241
+    },
+    {
+      "epoch": 1.8620689655172413,
+      "grad_norm": 0.09079090097092132,
+      "learning_rate": 7.563849882139776e-05,
+      "loss": 0.0101,
+      "step": 1242
+    },
+    {
+      "epoch": 1.863568215892054,
+      "grad_norm": 0.10042340395367601,
+      "learning_rate": 7.546926141211974e-05,
+      "loss": 0.0101,
+      "step": 1243
+    },
+    {
+      "epoch": 1.8650674662668667,
+      "grad_norm": 0.05249586616928009,
+      "learning_rate": 7.530009872772572e-05,
+      "loss": 0.0073,
+      "step": 1244
+    },
+    {
+      "epoch": 1.866566716641679,
+      "grad_norm": 0.06628730417621559,
+      "learning_rate": 7.513101128351454e-05,
+      "loss": 0.006,
+      "step": 1245
+    },
+    {
+      "epoch": 1.8680659670164919,
+      "grad_norm": 0.09985779934896404,
+      "learning_rate": 7.496199959455584e-05,
+      "loss": 0.009,
+      "step": 1246
+    },
+    {
+      "epoch": 1.8695652173913042,
+      "grad_norm": 0.07189435277325107,
+      "learning_rate": 7.479306417568864e-05,
+      "loss": 0.0079,
+      "step": 1247
+    },
+    {
+      "epoch": 1.871064467766117,
+      "grad_norm": 0.04129148635536298,
+      "learning_rate": 7.462420554151944e-05,
+      "loss": 0.006,
+      "step": 1248
+    },
+    {
+      "epoch": 1.8725637181409296,
+      "grad_norm": 0.08500137505977622,
+      "learning_rate": 7.445542420642097e-05,
+      "loss": 0.0147,
+      "step": 1249
+    },
+    {
+      "epoch": 1.8740629685157422,
+      "grad_norm": 0.09066062675805478,
+      "learning_rate": 7.42867206845304e-05,
+      "loss": 0.0127,
+      "step": 1250
+    },
+    {
+      "epoch": 1.8755622188905547,
+      "grad_norm": 0.10120598113171132,
+      "learning_rate": 7.411809548974792e-05,
+      "loss": 0.0131,
+      "step": 1251
+    },
+    {
+      "epoch": 1.8770614692653673,
+      "grad_norm": 0.050739274353256675,
+      "learning_rate": 7.394954913573517e-05,
+      "loss": 0.0056,
+      "step": 1252
+    },
+    {
+      "epoch": 1.87856071964018,
+      "grad_norm": 0.07245322597365843,
+      "learning_rate": 7.378108213591355e-05,
+      "loss": 0.0059,
+      "step": 1253
+    },
+    {
+      "epoch": 1.8800599700149925,
+      "grad_norm": 0.15359008958070705,
+      "learning_rate": 7.361269500346274e-05,
+      "loss": 0.0164,
+      "step": 1254
+    },
+    {
+      "epoch": 1.881559220389805,
+      "grad_norm": 0.19317251107868752,
+      "learning_rate": 7.344438825131911e-05,
+      "loss": 0.0189,
+      "step": 1255
+    },
+    {
+      "epoch": 1.8830584707646176,
+      "grad_norm": 0.061858025534464414,
+      "learning_rate": 7.327616239217431e-05,
+      "loss": 0.0043,
+      "step": 1256
+    },
+    {
+      "epoch": 1.8845577211394304,
+      "grad_norm": 0.13091196479249864,
+      "learning_rate": 7.310801793847344e-05,
+      "loss": 0.0159,
+      "step": 1257
+    },
+    {
+      "epoch": 1.8860569715142428,
+      "grad_norm": 0.0544033205013549,
+      "learning_rate": 7.293995540241366e-05,
+      "loss": 0.0056,
+      "step": 1258
+    },
+    {
+      "epoch": 1.8875562218890556,
+      "grad_norm": 0.0783279329482906,
+      "learning_rate": 7.277197529594257e-05,
+      "loss": 0.0085,
+      "step": 1259
+    },
+    {
+      "epoch": 1.889055472263868,
+      "grad_norm": 0.060875813189406686,
+      "learning_rate": 7.260407813075676e-05,
+      "loss": 0.01,
+      "step": 1260
+    },
+    {
+      "epoch": 1.8905547226386807,
+      "grad_norm": 0.08212593368295347,
+      "learning_rate": 7.243626441830009e-05,
+      "loss": 0.0106,
+      "step": 1261
+    },
+    {
+      "epoch": 1.892053973013493,
+      "grad_norm": 0.14448153096893138,
+      "learning_rate": 7.226853466976222e-05,
+      "loss": 0.0316,
+      "step": 1262
+    },
+    {
+      "epoch": 1.893553223388306,
+      "grad_norm": 0.11583432656692919,
+      "learning_rate": 7.210088939607708e-05,
+      "loss": 0.0103,
+      "step": 1263
+    },
+    {
+      "epoch": 1.8950524737631185,
+      "grad_norm": 0.192579385165103,
+      "learning_rate": 7.193332910792124e-05,
+      "loss": 0.0236,
+      "step": 1264
+    },
+    {
+      "epoch": 1.896551724137931,
+      "grad_norm": 0.06822480918750315,
+      "learning_rate": 7.176585431571235e-05,
+      "loss": 0.0165,
+      "step": 1265
+    },
+    {
+      "epoch": 1.8980509745127436,
+      "grad_norm": 0.08350988856289665,
+      "learning_rate": 7.159846552960774e-05,
+      "loss": 0.0149,
+      "step": 1266
+    },
+    {
+      "epoch": 1.8995502248875562,
+      "grad_norm": 0.09708205671430116,
+      "learning_rate": 7.143116325950265e-05,
+      "loss": 0.0196,
+      "step": 1267
+    },
+    {
+      "epoch": 1.9010494752623688,
+      "grad_norm": 0.06975686550376206,
+      "learning_rate": 7.126394801502882e-05,
+      "loss": 0.008,
+      "step": 1268
+    },
+    {
+      "epoch": 1.9025487256371814,
+      "grad_norm": 0.0937817538752956,
+      "learning_rate": 7.109682030555283e-05,
+      "loss": 0.009,
+      "step": 1269
+    },
+    {
+      "epoch": 1.9040479760119942,
+      "grad_norm": 0.09751887136446406,
+      "learning_rate": 7.092978064017475e-05,
+      "loss": 0.0155,
+      "step": 1270
+    },
+    {
+      "epoch": 1.9055472263868065,
+      "grad_norm": 0.12135514343914165,
+      "learning_rate": 7.076282952772633e-05,
+      "loss": 0.0131,
+      "step": 1271
+    },
+    {
+      "epoch": 1.9070464767616193,
+      "grad_norm": 0.14474648955209451,
+      "learning_rate": 7.059596747676962e-05,
+      "loss": 0.0186,
+      "step": 1272
+    },
+    {
+      "epoch": 1.9085457271364317,
+      "grad_norm": 0.09456198132154989,
+      "learning_rate": 7.042919499559537e-05,
+      "loss": 0.0109,
+      "step": 1273
+    },
+    {
+      "epoch": 1.9100449775112445,
+      "grad_norm": 0.06368532914601598,
+      "learning_rate": 7.026251259222141e-05,
+      "loss": 0.011,
+      "step": 1274
+    },
+    {
+      "epoch": 1.9115442278860568,
+      "grad_norm": 0.07815558315112578,
+      "learning_rate": 7.009592077439134e-05,
+      "loss": 0.0096,
+      "step": 1275
+    },
+    {
+      "epoch": 1.9130434782608696,
+      "grad_norm": 0.07587411820029523,
+      "learning_rate": 6.992942004957271e-05,
+      "loss": 0.0094,
+      "step": 1276
+    },
+    {
+      "epoch": 1.9145427286356822,
+      "grad_norm": 0.14301760261232488,
+      "learning_rate": 6.976301092495556e-05,
+      "loss": 0.0195,
+      "step": 1277
+    },
+    {
+      "epoch": 1.9160419790104948,
+      "grad_norm": 0.05601598197369331,
+      "learning_rate": 6.959669390745097e-05,
+      "loss": 0.0077,
+      "step": 1278
+    },
+    {
+      "epoch": 1.9175412293853074,
+      "grad_norm": 0.17913703313927748,
+      "learning_rate": 6.943046950368944e-05,
+      "loss": 0.0207,
+      "step": 1279
+    },
+    {
+      "epoch": 1.91904047976012,
+      "grad_norm": 0.04621976604774044,
+      "learning_rate": 6.92643382200193e-05,
+      "loss": 0.0051,
+      "step": 1280
+    },
+    {
+      "epoch": 1.9205397301349325,
+      "grad_norm": 0.059822571193335644,
+      "learning_rate": 6.909830056250527e-05,
+      "loss": 0.0073,
+      "step": 1281
+    },
+    {
+      "epoch": 1.922038980509745,
+      "grad_norm": 0.050398816706857336,
+      "learning_rate": 6.893235703692685e-05,
+      "loss": 0.006,
+      "step": 1282
+    },
+    {
+      "epoch": 1.9235382308845579,
+      "grad_norm": 0.15403140307959207,
+      "learning_rate": 6.876650814877674e-05,
+      "loss": 0.0233,
+      "step": 1283
+    },
+    {
+      "epoch": 1.9250374812593702,
+      "grad_norm": 0.08002270556903888,
+      "learning_rate": 6.860075440325951e-05,
+      "loss": 0.0127,
+      "step": 1284
+    },
+    {
+      "epoch": 1.926536731634183,
+      "grad_norm": 0.09217416653554218,
+      "learning_rate": 6.843509630528977e-05,
+      "loss": 0.0078,
+      "step": 1285
+    },
+    {
+      "epoch": 1.9280359820089954,
+      "grad_norm": 0.08217684285186015,
+      "learning_rate": 6.826953435949081e-05,
+      "loss": 0.0087,
+      "step": 1286
+    },
+    {
+      "epoch": 1.9295352323838082,
+      "grad_norm": 0.1135626679946753,
+      "learning_rate": 6.8104069070193e-05,
+      "loss": 0.0106,
+      "step": 1287
+    },
+    {
+      "epoch": 1.9310344827586206,
+      "grad_norm": 0.11547720907926104,
+      "learning_rate": 6.793870094143238e-05,
+      "loss": 0.0104,
+      "step": 1288
+    },
+    {
+      "epoch": 1.9325337331334334,
+      "grad_norm": 0.06982269585114415,
+      "learning_rate": 6.77734304769489e-05,
+      "loss": 0.0078,
+      "step": 1289
+    },
+    {
+      "epoch": 1.934032983508246,
+      "grad_norm": 0.15410842004892794,
+      "learning_rate": 6.760825818018508e-05,
+      "loss": 0.0275,
+      "step": 1290
+    },
+    {
+      "epoch": 1.9355322338830585,
+      "grad_norm": 0.12670821336659963,
+      "learning_rate": 6.744318455428436e-05,
+      "loss": 0.0159,
+      "step": 1291
+    },
+    {
+      "epoch": 1.937031484257871,
+      "grad_norm": 0.0885086070508493,
+      "learning_rate": 6.727821010208962e-05,
+      "loss": 0.0111,
+      "step": 1292
+    },
+    {
+      "epoch": 1.9385307346326837,
+      "grad_norm": 0.0632145088319827,
+      "learning_rate": 6.711333532614168e-05,
+      "loss": 0.0098,
+      "step": 1293
+    },
+    {
+      "epoch": 1.9400299850074962,
+      "grad_norm": 0.10005119826138181,
+      "learning_rate": 6.69485607286777e-05,
+      "loss": 0.017,
+      "step": 1294
+    },
+    {
+      "epoch": 1.9415292353823088,
+      "grad_norm": 0.0719843361060488,
+      "learning_rate": 6.67838868116297e-05,
+      "loss": 0.0099,
+      "step": 1295
+    },
+    {
+      "epoch": 1.9430284857571214,
+      "grad_norm": 0.09905966753940625,
+      "learning_rate": 6.661931407662292e-05,
+      "loss": 0.0118,
+      "step": 1296
+    },
+    {
+      "epoch": 1.944527736131934,
+      "grad_norm": 0.07485985195447457,
+      "learning_rate": 6.64548430249745e-05,
+      "loss": 0.0095,
+      "step": 1297
+    },
+    {
+      "epoch": 1.9460269865067468,
+      "grad_norm": 0.10312354721673073,
+      "learning_rate": 6.62904741576918e-05,
+      "loss": 0.0075,
+      "step": 1298
+    },
+    {
+      "epoch": 1.9475262368815591,
+      "grad_norm": 0.0926401569025068,
+      "learning_rate": 6.612620797547087e-05,
+      "loss": 0.0106,
+      "step": 1299
+    },
+    {
+      "epoch": 1.949025487256372,
+      "grad_norm": 0.08554451492013519,
+      "learning_rate": 6.5962044978695e-05,
+      "loss": 0.0071,
+      "step": 1300
+    },
+    {
+      "epoch": 1.9505247376311843,
+      "grad_norm": 0.07530310677555721,
+      "learning_rate": 6.579798566743314e-05,
+      "loss": 0.0059,
+      "step": 1301
+    },
+    {
+      "epoch": 1.952023988005997,
+      "grad_norm": 0.09012179453580464,
+      "learning_rate": 6.56340305414384e-05,
+      "loss": 0.007,
+      "step": 1302
+    },
+    {
+      "epoch": 1.9535232383808094,
+      "grad_norm": 0.0732412409351987,
+      "learning_rate": 6.547018010014654e-05,
+      "loss": 0.009,
+      "step": 1303
+    },
+    {
+      "epoch": 1.9550224887556222,
+      "grad_norm": 0.1122090001398675,
+      "learning_rate": 6.530643484267443e-05,
+      "loss": 0.0123,
+      "step": 1304
+    },
+    {
+      "epoch": 1.9565217391304348,
+      "grad_norm": 0.05854864784619732,
+      "learning_rate": 6.51427952678185e-05,
+      "loss": 0.0088,
+      "step": 1305
+    },
+    {
+      "epoch": 1.9580209895052474,
+      "grad_norm": 0.17221127456654123,
+      "learning_rate": 6.497926187405326e-05,
+      "loss": 0.0265,
+      "step": 1306
+    },
+    {
+      "epoch": 1.95952023988006,
+      "grad_norm": 0.11007336458310958,
+      "learning_rate": 6.481583515952983e-05,
+      "loss": 0.0195,
+      "step": 1307
+    },
+    {
+      "epoch": 1.9610194902548725,
+      "grad_norm": 0.1603764107496845,
+      "learning_rate": 6.465251562207431e-05,
+      "loss": 0.0127,
+      "step": 1308
+    },
+    {
+      "epoch": 1.9625187406296851,
+      "grad_norm": 0.10303600158564385,
+      "learning_rate": 6.448930375918631e-05,
+      "loss": 0.0091,
+      "step": 1309
+    },
+    {
+      "epoch": 1.9640179910044977,
+      "grad_norm": 0.1059759977454014,
+      "learning_rate": 6.432620006803746e-05,
+      "loss": 0.0147,
+      "step": 1310
+    },
+    {
+      "epoch": 1.9655172413793105,
+      "grad_norm": 0.13503577362942953,
+      "learning_rate": 6.416320504546997e-05,
+      "loss": 0.0247,
+      "step": 1311
+    },
+    {
+      "epoch": 1.9670164917541229,
+      "grad_norm": 0.06892530634050763,
+      "learning_rate": 6.40003191879949e-05,
+      "loss": 0.01,
+      "step": 1312
+    },
+    {
+      "epoch": 1.9685157421289357,
+      "grad_norm": 0.07291043016423708,
+      "learning_rate": 6.383754299179079e-05,
+      "loss": 0.0097,
+      "step": 1313
+    },
+    {
+      "epoch": 1.970014992503748,
+      "grad_norm": 0.13914268368623947,
+      "learning_rate": 6.367487695270217e-05,
+      "loss": 0.0164,
+      "step": 1314
+    },
+    {
+      "epoch": 1.9715142428785608,
+      "grad_norm": 0.0714584621260185,
+      "learning_rate": 6.351232156623803e-05,
+      "loss": 0.01,
+      "step": 1315
+    },
+    {
+      "epoch": 1.9730134932533732,
+      "grad_norm": 0.12977132489990004,
+      "learning_rate": 6.334987732757029e-05,
+      "loss": 0.0124,
+      "step": 1316
+    },
+    {
+      "epoch": 1.974512743628186,
+      "grad_norm": 0.11389052086228867,
+      "learning_rate": 6.318754473153221e-05,
+      "loss": 0.0191,
+      "step": 1317
+    },
+    {
+      "epoch": 1.9760119940029985,
+      "grad_norm": 0.13954486292599097,
+      "learning_rate": 6.302532427261709e-05,
+      "loss": 0.0178,
+      "step": 1318
+    },
+    {
+      "epoch": 1.9775112443778111,
+      "grad_norm": 0.2039430304512413,
+      "learning_rate": 6.286321644497655e-05,
+      "loss": 0.0188,
+      "step": 1319
+    },
+    {
+      "epoch": 1.9790104947526237,
+      "grad_norm": 0.04606291429944488,
+      "learning_rate": 6.27012217424191e-05,
+      "loss": 0.003,
+      "step": 1320
+    },
+    {
+      "epoch": 1.9805097451274363,
+      "grad_norm": 0.07084029408482302,
+      "learning_rate": 6.25393406584088e-05,
+      "loss": 0.0082,
+      "step": 1321
+    },
+    {
+      "epoch": 1.9820089955022488,
+      "grad_norm": 0.06195579957195022,
+      "learning_rate": 6.237757368606345e-05,
+      "loss": 0.0056,
+      "step": 1322
+    },
+    {
+      "epoch": 1.9835082458770614,
+      "grad_norm": 0.10572839742430142,
+      "learning_rate": 6.22159213181533e-05,
+      "loss": 0.0227,
+      "step": 1323
+    },
+    {
+      "epoch": 1.9850074962518742,
+      "grad_norm": 0.0953707827939962,
+      "learning_rate": 6.205438404709947e-05,
+      "loss": 0.0135,
+      "step": 1324
+    },
+    {
+      "epoch": 1.9865067466266866,
+      "grad_norm": 0.10133284635764458,
+      "learning_rate": 6.18929623649726e-05,
+      "loss": 0.0167,
+      "step": 1325
+    },
+    {
+      "epoch": 1.9880059970014994,
+      "grad_norm": 0.1046927216676533,
+      "learning_rate": 6.173165676349103e-05,
+      "loss": 0.0153,
+      "step": 1326
+    },
+    {
+      "epoch": 1.9895052473763117,
+      "grad_norm": 0.09125433955424463,
+      "learning_rate": 6.157046773401964e-05,
+      "loss": 0.0098,
+      "step": 1327
+    },
+    {
+      "epoch": 1.9910044977511245,
+      "grad_norm": 0.07206163283122839,
+      "learning_rate": 6.140939576756817e-05,
+      "loss": 0.0088,
+      "step": 1328
+    },
+    {
+      "epoch": 1.992503748125937,
+      "grad_norm": 0.09624780783173856,
+      "learning_rate": 6.12484413547897e-05,
+      "loss": 0.0097,
+      "step": 1329
+    },
+    {
+      "epoch": 1.9940029985007497,
+      "grad_norm": 0.08193537478772897,
+      "learning_rate": 6.108760498597938e-05,
+      "loss": 0.0078,
+      "step": 1330
+    },
+    {
+      "epoch": 1.9955022488755623,
+      "grad_norm": 0.10581177346811316,
+      "learning_rate": 6.092688715107264e-05,
+      "loss": 0.0201,
+      "step": 1331
+    },
+    {
+      "epoch": 1.9970014992503748,
+      "grad_norm": 0.08582699378434917,
+      "learning_rate": 6.076628833964388e-05,
+      "loss": 0.0087,
+      "step": 1332
+    },
+    {
+      "epoch": 1.9985007496251874,
+      "grad_norm": 0.07360485364749729,
+      "learning_rate": 6.0605809040904894e-05,
+      "loss": 0.0089,
+      "step": 1333
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.10066888348562322,
+      "learning_rate": 6.0445449743703516e-05,
+      "loss": 0.0169,
+      "step": 1334
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 0.02858872152864933,
+      "eval_runtime": 481.6247,
+      "eval_samples_per_second": 21.544,
+      "eval_steps_per_second": 1.348,
+      "step": 1334
+    },
+    {
+      "epoch": 2.001499250374813,
+      "grad_norm": 0.05691507076831063,
+      "learning_rate": 6.0285210936521955e-05,
+      "loss": 0.0081,
+      "step": 1335
+    },
+    {
+      "epoch": 2.002998500749625,
+      "grad_norm": 0.047619108961378546,
+      "learning_rate": 6.012509310747538e-05,
+      "loss": 0.0059,
+      "step": 1336
+    },
+    {
+      "epoch": 2.004497751124438,
+      "grad_norm": 0.12728571721122714,
+      "learning_rate": 5.9965096744310526e-05,
+      "loss": 0.0128,
+      "step": 1337
+    },
+    {
+      "epoch": 2.0059970014992503,
+      "grad_norm": 0.05533519939522902,
+      "learning_rate": 5.9805222334404e-05,
+      "loss": 0.0063,
+      "step": 1338
+    },
+    {
+      "epoch": 2.007496251874063,
+      "grad_norm": 0.058973041277131814,
+      "learning_rate": 5.964547036476099e-05,
+      "loss": 0.0078,
+      "step": 1339
+    },
+    {
+      "epoch": 2.0089955022488755,
+      "grad_norm": 0.053801378638280425,
+      "learning_rate": 5.9485841322013755e-05,
+      "loss": 0.0045,
+      "step": 1340
+    },
+    {
+      "epoch": 2.0104947526236883,
+      "grad_norm": 0.08346927650771466,
+      "learning_rate": 5.9326335692419995e-05,
+      "loss": 0.0112,
+      "step": 1341
+    },
+    {
+      "epoch": 2.0119940029985006,
+      "grad_norm": 0.06796501482552034,
+      "learning_rate": 5.916695396186154e-05,
+      "loss": 0.0059,
+      "step": 1342
+    },
+    {
+      "epoch": 2.0134932533733134,
+      "grad_norm": 0.02845322851169814,
+      "learning_rate": 5.900769661584272e-05,
+      "loss": 0.0051,
+      "step": 1343
+    },
+    {
+      "epoch": 2.0149925037481258,
+      "grad_norm": 0.09502613453475849,
+      "learning_rate": 5.884856413948913e-05,
+      "loss": 0.0157,
+      "step": 1344
+    },
+    {
+      "epoch": 2.0164917541229386,
+      "grad_norm": 0.033394665506260045,
+      "learning_rate": 5.868955701754584e-05,
+      "loss": 0.0027,
+      "step": 1345
+    },
+    {
+      "epoch": 2.017991004497751,
+      "grad_norm": 0.08324949873495568,
+      "learning_rate": 5.853067573437612e-05,
+      "loss": 0.0101,
+      "step": 1346
+    },
+    {
+      "epoch": 2.0194902548725637,
+      "grad_norm": 0.06530137085651484,
+      "learning_rate": 5.83719207739599e-05,
+      "loss": 0.0083,
+      "step": 1347
+    },
+    {
+      "epoch": 2.0209895052473765,
+      "grad_norm": 0.057351896571536574,
+      "learning_rate": 5.82132926198923e-05,
+      "loss": 0.0057,
+      "step": 1348
+    },
+    {
+      "epoch": 2.022488755622189,
+      "grad_norm": 0.07266435520449091,
+      "learning_rate": 5.805479175538229e-05,
+      "loss": 0.004,
+      "step": 1349
+    },
+    {
+      "epoch": 2.0239880059970017,
+      "grad_norm": 0.046130085433988374,
+      "learning_rate": 5.789641866325091e-05,
+      "loss": 0.0049,
+      "step": 1350
+    },
+    {
+      "epoch": 2.025487256371814,
+      "grad_norm": 0.06091418273733409,
+      "learning_rate": 5.773817382593008e-05,
+      "loss": 0.005,
+      "step": 1351
+    },
+    {
+      "epoch": 2.026986506746627,
+      "grad_norm": 0.05620982339093061,
+      "learning_rate": 5.758005772546097e-05,
+      "loss": 0.0074,
+      "step": 1352
+    },
+    {
+      "epoch": 2.028485757121439,
+      "grad_norm": 0.07166290846746312,
+      "learning_rate": 5.7422070843492734e-05,
+      "loss": 0.0075,
+      "step": 1353
+    },
+    {
+      "epoch": 2.029985007496252,
+      "grad_norm": 0.05351548956206307,
+      "learning_rate": 5.7264213661280765e-05,
+      "loss": 0.0046,
+      "step": 1354
+    },
+    {
+      "epoch": 2.0314842578710643,
+      "grad_norm": 0.16771028867831342,
+      "learning_rate": 5.710648665968543e-05,
+      "loss": 0.0061,
+      "step": 1355
+    },
+    {
+      "epoch": 2.032983508245877,
+      "grad_norm": 0.049188943209597075,
+      "learning_rate": 5.694889031917047e-05,
+      "loss": 0.0041,
+      "step": 1356
+    },
+    {
+      "epoch": 2.0344827586206895,
+      "grad_norm": 0.04547779735623479,
+      "learning_rate": 5.679142511980175e-05,
+      "loss": 0.0027,
+      "step": 1357
+    },
+    {
+      "epoch": 2.0359820089955023,
+      "grad_norm": 0.10403833741318744,
+      "learning_rate": 5.663409154124557e-05,
+      "loss": 0.0049,
+      "step": 1358
+    },
+    {
+      "epoch": 2.0374812593703147,
+      "grad_norm": 0.04247535677566507,
+      "learning_rate": 5.647689006276726e-05,
+      "loss": 0.0018,
+      "step": 1359
+    },
+    {
+      "epoch": 2.0389805097451275,
+      "grad_norm": 0.07701782209515198,
+      "learning_rate": 5.63198211632298e-05,
+      "loss": 0.0061,
+      "step": 1360
+    },
+    {
+      "epoch": 2.04047976011994,
+      "grad_norm": 0.08747035535017979,
+      "learning_rate": 5.616288532109225e-05,
+      "loss": 0.0047,
+      "step": 1361
+    },
+    {
+      "epoch": 2.0419790104947526,
+      "grad_norm": 0.1850652979487481,
+      "learning_rate": 5.6006083014408484e-05,
+      "loss": 0.0081,
+      "step": 1362
+    },
+    {
+      "epoch": 2.0434782608695654,
+      "grad_norm": 0.03442503949688273,
+      "learning_rate": 5.584941472082549e-05,
+      "loss": 0.0028,
+      "step": 1363
+    },
+    {
+      "epoch": 2.0449775112443778,
+      "grad_norm": 0.10724209405230377,
+      "learning_rate": 5.5692880917582046e-05,
+      "loss": 0.0092,
+      "step": 1364
+    },
+    {
+      "epoch": 2.0464767616191906,
+      "grad_norm": 0.06723361064783695,
+      "learning_rate": 5.553648208150728e-05,
+      "loss": 0.0049,
+      "step": 1365
+    },
+    {
+      "epoch": 2.047976011994003,
+      "grad_norm": 0.06341433310483063,
+      "learning_rate": 5.5380218689019125e-05,
+      "loss": 0.0043,
+      "step": 1366
+    },
+    {
+      "epoch": 2.0494752623688157,
+      "grad_norm": 0.06092073707056924,
+      "learning_rate": 5.522409121612304e-05,
+      "loss": 0.0046,
+      "step": 1367
+    },
+    {
+      "epoch": 2.050974512743628,
+      "grad_norm": 0.21552813309206953,
+      "learning_rate": 5.506810013841036e-05,
+      "loss": 0.0138,
+      "step": 1368
+    },
+    {
+      "epoch": 2.052473763118441,
+      "grad_norm": 0.0655401729299929,
+      "learning_rate": 5.491224593105695e-05,
+      "loss": 0.0032,
+      "step": 1369
+    },
+    {
+      "epoch": 2.0539730134932532,
+      "grad_norm": 0.07021885435075632,
+      "learning_rate": 5.475652906882173e-05,
+      "loss": 0.0077,
+      "step": 1370
+    },
+    {
+      "epoch": 2.055472263868066,
+      "grad_norm": 0.11480398420184994,
+      "learning_rate": 5.4600950026045326e-05,
+      "loss": 0.0122,
+      "step": 1371
+    },
+    {
+      "epoch": 2.0569715142428784,
+      "grad_norm": 0.1415004918550936,
+      "learning_rate": 5.4445509276648466e-05,
+      "loss": 0.0055,
+      "step": 1372
+    },
+    {
+      "epoch": 2.058470764617691,
+      "grad_norm": 0.08525010803571845,
+      "learning_rate": 5.4290207294130615e-05,
+      "loss": 0.0043,
+      "step": 1373
+    },
+    {
+      "epoch": 2.0599700149925035,
+      "grad_norm": 0.039178570486034396,
+      "learning_rate": 5.4135044551568546e-05,
+      "loss": 0.0019,
+      "step": 1374
+    },
+    {
+      "epoch": 2.0614692653673163,
+      "grad_norm": 0.0817349771881642,
+      "learning_rate": 5.398002152161484e-05,
+      "loss": 0.0027,
+      "step": 1375
+    },
+    {
+      "epoch": 2.062968515742129,
+      "grad_norm": 0.12797938737383258,
+      "learning_rate": 5.382513867649663e-05,
+      "loss": 0.007,
+      "step": 1376
+    },
+    {
+      "epoch": 2.0644677661169415,
+      "grad_norm": 0.14457775493372024,
+      "learning_rate": 5.3670396488013854e-05,
+      "loss": 0.0166,
+      "step": 1377
+    },
+    {
+      "epoch": 2.0659670164917543,
+      "grad_norm": 0.09943983374773288,
+      "learning_rate": 5.351579542753807e-05,
+      "loss": 0.01,
+      "step": 1378
+    },
+    {
+      "epoch": 2.0674662668665666,
+      "grad_norm": 0.04088894150509561,
+      "learning_rate": 5.33613359660109e-05,
+      "loss": 0.0026,
+      "step": 1379
+    },
+    {
+      "epoch": 2.0689655172413794,
+      "grad_norm": 0.043089887974945645,
+      "learning_rate": 5.320701857394268e-05,
+      "loss": 0.0052,
+      "step": 1380
+    },
+    {
+      "epoch": 2.070464767616192,
+      "grad_norm": 0.017301891563744867,
+      "learning_rate": 5.305284372141095e-05,
+      "loss": 0.0012,
+      "step": 1381
+    },
+    {
+      "epoch": 2.0719640179910046,
+      "grad_norm": 0.05740234754230655,
+      "learning_rate": 5.2898811878059e-05,
+      "loss": 0.0027,
+      "step": 1382
+    },
+    {
+      "epoch": 2.073463268365817,
+      "grad_norm": 0.05104654315941495,
+      "learning_rate": 5.274492351309461e-05,
+      "loss": 0.0039,
+      "step": 1383
+    },
+    {
+      "epoch": 2.0749625187406298,
+      "grad_norm": 0.060428329511087744,
+      "learning_rate": 5.259117909528839e-05,
+      "loss": 0.0099,
+      "step": 1384
+    },
+    {
+      "epoch": 2.076461769115442,
+      "grad_norm": 0.05507881001018212,
+      "learning_rate": 5.243757909297247e-05,
+      "loss": 0.0059,
+      "step": 1385
+    },
+    {
+      "epoch": 2.077961019490255,
+      "grad_norm": 0.08630299613505883,
+      "learning_rate": 5.2284123974039154e-05,
+      "loss": 0.0041,
+      "step": 1386
+    },
+    {
+      "epoch": 2.0794602698650673,
+      "grad_norm": 0.14844088217359913,
+      "learning_rate": 5.213081420593933e-05,
+      "loss": 0.0085,
+      "step": 1387
+    },
+    {
+      "epoch": 2.08095952023988,
+      "grad_norm": 0.11946133785518649,
+      "learning_rate": 5.197765025568109e-05,
+      "loss": 0.0112,
+      "step": 1388
+    },
+    {
+      "epoch": 2.082458770614693,
+      "grad_norm": 0.08999841950189766,
+      "learning_rate": 5.182463258982846e-05,
+      "loss": 0.0068,
+      "step": 1389
+    },
+    {
+      "epoch": 2.0839580209895052,
+      "grad_norm": 0.08234997169720312,
+      "learning_rate": 5.1671761674499765e-05,
+      "loss": 0.0067,
+      "step": 1390
+    },
+    {
+      "epoch": 2.085457271364318,
+      "grad_norm": 0.05543198018151636,
+      "learning_rate": 5.15190379753663e-05,
+      "loss": 0.0037,
+      "step": 1391
+    },
+    {
+      "epoch": 2.0869565217391304,
+      "grad_norm": 0.1728213920668675,
+      "learning_rate": 5.1366461957650954e-05,
+      "loss": 0.014,
+      "step": 1392
+    },
+    {
+      "epoch": 2.088455772113943,
+      "grad_norm": 0.059050509744897936,
+      "learning_rate": 5.121403408612672e-05,
+      "loss": 0.0061,
+      "step": 1393
+    },
+    {
+      "epoch": 2.0899550224887555,
+      "grad_norm": 0.103095043846112,
+      "learning_rate": 5.1061754825115374e-05,
+      "loss": 0.018,
+      "step": 1394
+    },
+    {
+      "epoch": 2.0914542728635683,
+      "grad_norm": 0.03936790723429008,
+      "learning_rate": 5.090962463848592e-05,
+      "loss": 0.0025,
+      "step": 1395
+    },
+    {
+      "epoch": 2.0929535232383807,
+      "grad_norm": 0.15600675536862346,
+      "learning_rate": 5.07576439896533e-05,
+      "loss": 0.0123,
+      "step": 1396
+    },
+    {
+      "epoch": 2.0944527736131935,
+      "grad_norm": 0.13895637920809223,
+      "learning_rate": 5.0605813341576924e-05,
+      "loss": 0.0083,
+      "step": 1397
+    },
+    {
+      "epoch": 2.095952023988006,
+      "grad_norm": 0.07600005087241365,
+      "learning_rate": 5.045413315675924e-05,
+      "loss": 0.0057,
+      "step": 1398
+    },
+    {
+      "epoch": 2.0974512743628186,
+      "grad_norm": 0.07128253832259326,
+      "learning_rate": 5.0302603897244474e-05,
+      "loss": 0.0062,
+      "step": 1399
+    },
+    {
+      "epoch": 2.098950524737631,
+      "grad_norm": 0.03157517477798966,
+      "learning_rate": 5.015122602461698e-05,
+      "loss": 0.002,
+      "step": 1400
+    },
+    {
+      "epoch": 2.100449775112444,
+      "grad_norm": 0.063283575874792,
+      "learning_rate": 5.000000000000002e-05,
+      "loss": 0.0065,
+      "step": 1401
+    },
+    {
+      "epoch": 2.101949025487256,
+      "grad_norm": 0.048367470551601933,
+      "learning_rate": 4.9848926284054255e-05,
+      "loss": 0.0032,
+      "step": 1402
+    },
+    {
+      "epoch": 2.103448275862069,
+      "grad_norm": 0.03119576646559798,
+      "learning_rate": 4.969800533697649e-05,
+      "loss": 0.003,
+      "step": 1403
+    },
+    {
+      "epoch": 2.1049475262368817,
+      "grad_norm": 0.08189145443012984,
+      "learning_rate": 4.9547237618498085e-05,
+      "loss": 0.0016,
+      "step": 1404
+    },
+    {
+      "epoch": 2.106446776611694,
+      "grad_norm": 0.04950617409138366,
+      "learning_rate": 4.939662358788364e-05,
+      "loss": 0.003,
+      "step": 1405
+    },
+    {
+      "epoch": 2.107946026986507,
+      "grad_norm": 0.07768299028368211,
+      "learning_rate": 4.924616370392961e-05,
+      "loss": 0.0069,
+      "step": 1406
+    },
+    {
+      "epoch": 2.1094452773613193,
+      "grad_norm": 0.03281358107954231,
+      "learning_rate": 4.909585842496287e-05,
+      "loss": 0.0032,
+      "step": 1407
+    },
+    {
+      "epoch": 2.110944527736132,
+      "grad_norm": 0.11945720424334462,
+      "learning_rate": 4.894570820883944e-05,
+      "loss": 0.0158,
+      "step": 1408
+    },
+    {
+      "epoch": 2.1124437781109444,
+      "grad_norm": 0.07932882476750797,
+      "learning_rate": 4.8795713512942865e-05,
+      "loss": 0.0067,
+      "step": 1409
+    },
+    {
+      "epoch": 2.113943028485757,
+      "grad_norm": 0.026045115469265436,
+      "learning_rate": 4.864587479418302e-05,
+      "loss": 0.0017,
+      "step": 1410
+    },
+    {
+      "epoch": 2.1154422788605696,
+      "grad_norm": 0.08432479899132923,
+      "learning_rate": 4.8496192508994576e-05,
+      "loss": 0.0068,
+      "step": 1411
+    },
+    {
+      "epoch": 2.1169415292353824,
+      "grad_norm": 0.08090217442628098,
+      "learning_rate": 4.834666711333582e-05,
+      "loss": 0.0075,
+      "step": 1412
+    },
+    {
+      "epoch": 2.1184407796101947,
+      "grad_norm": 0.0623686180681052,
+      "learning_rate": 4.8197299062686995e-05,
+      "loss": 0.0066,
+      "step": 1413
+    },
+    {
+      "epoch": 2.1199400299850075,
+      "grad_norm": 0.053337961961056765,
+      "learning_rate": 4.8048088812049096e-05,
+      "loss": 0.0035,
+      "step": 1414
+    },
+    {
+      "epoch": 2.12143928035982,
+      "grad_norm": 0.022375264029270973,
+      "learning_rate": 4.78990368159424e-05,
+      "loss": 0.0021,
+      "step": 1415
+    },
+    {
+      "epoch": 2.1229385307346327,
+      "grad_norm": 0.11775003892209943,
+      "learning_rate": 4.7750143528405126e-05,
+      "loss": 0.01,
+      "step": 1416
+    },
+    {
+      "epoch": 2.1244377811094455,
+      "grad_norm": 0.04122303945871826,
+      "learning_rate": 4.7601409402992106e-05,
+      "loss": 0.0027,
+      "step": 1417
+    },
+    {
+      "epoch": 2.125937031484258,
+      "grad_norm": 0.047835533341228295,
+      "learning_rate": 4.745283489277325e-05,
+      "loss": 0.0022,
+      "step": 1418
+    },
+    {
+      "epoch": 2.1274362818590706,
+      "grad_norm": 0.05086786601082947,
+      "learning_rate": 4.7304420450332244e-05,
+      "loss": 0.003,
+      "step": 1419
+    },
+    {
+      "epoch": 2.128935532233883,
+      "grad_norm": 0.051518725132630144,
+      "learning_rate": 4.71561665277653e-05,
+      "loss": 0.0054,
+      "step": 1420
+    },
+    {
+      "epoch": 2.130434782608696,
+      "grad_norm": 0.05190216880471178,
+      "learning_rate": 4.700807357667952e-05,
+      "loss": 0.0023,
+      "step": 1421
+    },
+    {
+      "epoch": 2.131934032983508,
+      "grad_norm": 0.10337376557969855,
+      "learning_rate": 4.68601420481917e-05,
+      "loss": 0.0058,
+      "step": 1422
+    },
+    {
+      "epoch": 2.133433283358321,
+      "grad_norm": 0.03507599171549857,
+      "learning_rate": 4.6712372392927e-05,
+      "loss": 0.0018,
+      "step": 1423
+    },
+    {
+      "epoch": 2.1349325337331333,
+      "grad_norm": 0.0504070606518784,
+      "learning_rate": 4.656476506101737e-05,
+      "loss": 0.0041,
+      "step": 1424
+    },
+    {
+      "epoch": 2.136431784107946,
+      "grad_norm": 0.04550825149065688,
+      "learning_rate": 4.6417320502100316e-05,
+      "loss": 0.0066,
+      "step": 1425
+    },
+    {
+      "epoch": 2.1379310344827585,
+      "grad_norm": 0.08210457443940872,
+      "learning_rate": 4.6270039165317605e-05,
+      "loss": 0.0054,
+      "step": 1426
+    },
+    {
+      "epoch": 2.1394302848575713,
+      "grad_norm": 0.09426410394325767,
+      "learning_rate": 4.612292149931369e-05,
+      "loss": 0.0045,
+      "step": 1427
+    },
+    {
+      "epoch": 2.1409295352323836,
+      "grad_norm": 0.05591749148306491,
+      "learning_rate": 4.59759679522345e-05,
+      "loss": 0.0115,
+      "step": 1428
+    },
+    {
+      "epoch": 2.1424287856071964,
+      "grad_norm": 0.051607173257308396,
+      "learning_rate": 4.582917897172603e-05,
+      "loss": 0.0039,
+      "step": 1429
+    },
+    {
+      "epoch": 2.143928035982009,
+      "grad_norm": 0.08203477167290794,
+      "learning_rate": 4.568255500493292e-05,
+      "loss": 0.0019,
+      "step": 1430
+    },
+    {
+      "epoch": 2.1454272863568216,
+      "grad_norm": 0.0702400046217267,
+      "learning_rate": 4.5536096498497295e-05,
+      "loss": 0.005,
+      "step": 1431
+    },
+    {
+      "epoch": 2.1469265367316344,
+      "grad_norm": 0.04253399244012623,
+      "learning_rate": 4.5389803898557106e-05,
+      "loss": 0.0021,
+      "step": 1432
+    },
+    {
+      "epoch": 2.1484257871064467,
+      "grad_norm": 0.10867953782868013,
+      "learning_rate": 4.524367765074499e-05,
+      "loss": 0.0103,
+      "step": 1433
+    },
+    {
+      "epoch": 2.1499250374812595,
+      "grad_norm": 0.0672757026971409,
+      "learning_rate": 4.5097718200186814e-05,
+      "loss": 0.0046,
+      "step": 1434
+    },
+    {
+      "epoch": 2.151424287856072,
+      "grad_norm": 0.06915524666361397,
+      "learning_rate": 4.495192599150044e-05,
+      "loss": 0.0047,
+      "step": 1435
+    },
+    {
+      "epoch": 2.1529235382308847,
+      "grad_norm": 0.20209814933952153,
+      "learning_rate": 4.480630146879419e-05,
+      "loss": 0.0133,
+      "step": 1436
+    },
+    {
+      "epoch": 2.154422788605697,
+      "grad_norm": 0.05989447472439885,
+      "learning_rate": 4.46608450756656e-05,
+      "loss": 0.0048,
+      "step": 1437
+    },
+    {
+      "epoch": 2.15592203898051,
+      "grad_norm": 0.06402998012232496,
+      "learning_rate": 4.451555725520009e-05,
+      "loss": 0.0051,
+      "step": 1438
+    },
+    {
+      "epoch": 2.157421289355322,
+      "grad_norm": 0.054064533624966084,
+      "learning_rate": 4.437043844996952e-05,
+      "loss": 0.0037,
+      "step": 1439
+    },
+    {
+      "epoch": 2.158920539730135,
+      "grad_norm": 0.03250489649463816,
+      "learning_rate": 4.4225489102030995e-05,
+      "loss": 0.0027,
+      "step": 1440
+    },
+    {
+      "epoch": 2.1604197901049473,
+      "grad_norm": 0.07946031963668018,
+      "learning_rate": 4.4080709652925336e-05,
+      "loss": 0.0049,
+      "step": 1441
+    },
+    {
+      "epoch": 2.16191904047976,
+      "grad_norm": 0.12942691934162556,
+      "learning_rate": 4.393610054367585e-05,
+      "loss": 0.0092,
+      "step": 1442
+    },
+    {
+      "epoch": 2.1634182908545725,
+      "grad_norm": 0.10645765981635241,
+      "learning_rate": 4.379166221478697e-05,
+      "loss": 0.0102,
+      "step": 1443
+    },
+    {
+      "epoch": 2.1649175412293853,
+      "grad_norm": 0.06232769787707574,
+      "learning_rate": 4.364739510624286e-05,
+      "loss": 0.0031,
+      "step": 1444
+    },
+    {
+      "epoch": 2.166416791604198,
+      "grad_norm": 0.06283478350375783,
+      "learning_rate": 4.350329965750621e-05,
+      "loss": 0.0027,
+      "step": 1445
+    },
+    {
+      "epoch": 2.1679160419790104,
+      "grad_norm": 0.0458686612766849,
+      "learning_rate": 4.335937630751674e-05,
+      "loss": 0.0037,
+      "step": 1446
+    },
+    {
+      "epoch": 2.1694152923538232,
+      "grad_norm": 0.08933097544850213,
+      "learning_rate": 4.32156254946899e-05,
+      "loss": 0.0069,
+      "step": 1447
+    },
+    {
+      "epoch": 2.1709145427286356,
+      "grad_norm": 0.08001891739481207,
+      "learning_rate": 4.307204765691558e-05,
+      "loss": 0.0052,
+      "step": 1448
+    },
+    {
+      "epoch": 2.1724137931034484,
+      "grad_norm": 0.19361680061202965,
+      "learning_rate": 4.2928643231556844e-05,
+      "loss": 0.0076,
+      "step": 1449
+    },
+    {
+      "epoch": 2.1739130434782608,
+      "grad_norm": 0.06531644648753111,
+      "learning_rate": 4.27854126554484e-05,
+      "loss": 0.0069,
+      "step": 1450
+    },
+    {
+      "epoch": 2.1754122938530736,
+      "grad_norm": 0.11015887039069647,
+      "learning_rate": 4.264235636489542e-05,
+      "loss": 0.0055,
+      "step": 1451
+    },
+    {
+      "epoch": 2.176911544227886,
+      "grad_norm": 0.057749351044845736,
+      "learning_rate": 4.249947479567218e-05,
+      "loss": 0.003,
+      "step": 1452
+    },
+    {
+      "epoch": 2.1784107946026987,
+      "grad_norm": 0.07640922995670499,
+      "learning_rate": 4.235676838302068e-05,
+      "loss": 0.0056,
+      "step": 1453
+    },
+    {
+      "epoch": 2.179910044977511,
+      "grad_norm": 0.05338471185436684,
+      "learning_rate": 4.221423756164948e-05,
+      "loss": 0.0034,
+      "step": 1454
+    },
+    {
+      "epoch": 2.181409295352324,
+      "grad_norm": 0.11526166319637095,
+      "learning_rate": 4.207188276573214e-05,
+      "loss": 0.009,
+      "step": 1455
+    },
+    {
+      "epoch": 2.1829085457271367,
+      "grad_norm": 0.0758586393121374,
+      "learning_rate": 4.1929704428906026e-05,
+      "loss": 0.004,
+      "step": 1456
+    },
+    {
+      "epoch": 2.184407796101949,
+      "grad_norm": 0.08116550249079785,
+      "learning_rate": 4.1787702984271074e-05,
+      "loss": 0.0161,
+      "step": 1457
+    },
+    {
+      "epoch": 2.185907046476762,
+      "grad_norm": 0.08050820007995133,
+      "learning_rate": 4.1645878864388266e-05,
+      "loss": 0.009,
+      "step": 1458
+    },
+    {
+      "epoch": 2.187406296851574,
+      "grad_norm": 0.08238424313644067,
+      "learning_rate": 4.150423250127845e-05,
+      "loss": 0.0049,
+      "step": 1459
+    },
+    {
+      "epoch": 2.188905547226387,
+      "grad_norm": 0.08089595012768978,
+      "learning_rate": 4.1362764326421064e-05,
+      "loss": 0.0033,
+      "step": 1460
+    },
+    {
+      "epoch": 2.1904047976011993,
+      "grad_norm": 0.03361021628627364,
+      "learning_rate": 4.12214747707527e-05,
+      "loss": 0.0031,
+      "step": 1461
+    },
+    {
+      "epoch": 2.191904047976012,
+      "grad_norm": 0.08619764958760784,
+      "learning_rate": 4.1080364264665774e-05,
+      "loss": 0.0115,
+      "step": 1462
+    },
+    {
+      "epoch": 2.1934032983508245,
+      "grad_norm": 0.18569051233692943,
+      "learning_rate": 4.093943323800745e-05,
+      "loss": 0.0127,
+      "step": 1463
+    },
+    {
+      "epoch": 2.1949025487256373,
+      "grad_norm": 0.019196872401471405,
+      "learning_rate": 4.0798682120078044e-05,
+      "loss": 0.0007,
+      "step": 1464
+    },
+    {
+      "epoch": 2.1964017991004496,
+      "grad_norm": 0.09065601441816457,
+      "learning_rate": 4.065811133962987e-05,
+      "loss": 0.0065,
+      "step": 1465
+    },
+    {
+      "epoch": 2.1979010494752624,
+      "grad_norm": 0.04523517478416136,
+      "learning_rate": 4.0517721324865884e-05,
+      "loss": 0.0012,
+      "step": 1466
+    },
+    {
+      "epoch": 2.199400299850075,
+      "grad_norm": 0.06034653817480533,
+      "learning_rate": 4.037751250343841e-05,
+      "loss": 0.0042,
+      "step": 1467
+    },
+    {
+      "epoch": 2.2008995502248876,
+      "grad_norm": 0.025739476261626783,
+      "learning_rate": 4.023748530244789e-05,
+      "loss": 0.002,
+      "step": 1468
+    },
+    {
+      "epoch": 2.2023988005997,
+      "grad_norm": 0.09451823682114383,
+      "learning_rate": 4.009764014844143e-05,
+      "loss": 0.0032,
+      "step": 1469
+    },
+    {
+      "epoch": 2.2038980509745127,
+      "grad_norm": 0.14766234026107614,
+      "learning_rate": 3.9957977467411615e-05,
+      "loss": 0.0096,
+      "step": 1470
+    },
+    {
+      "epoch": 2.2053973013493255,
+      "grad_norm": 0.249925133129023,
+      "learning_rate": 3.981849768479517e-05,
+      "loss": 0.0056,
+      "step": 1471
+    },
+    {
+      "epoch": 2.206896551724138,
+      "grad_norm": 0.05670171893846676,
+      "learning_rate": 3.967920122547175e-05,
+      "loss": 0.0058,
+      "step": 1472
+    },
+    {
+      "epoch": 2.2083958020989507,
+      "grad_norm": 0.06840035500620087,
+      "learning_rate": 3.954008851376252e-05,
+      "loss": 0.0042,
+      "step": 1473
+    },
+    {
+      "epoch": 2.209895052473763,
+      "grad_norm": 0.045333811302793486,
+      "learning_rate": 3.940115997342891e-05,
+      "loss": 0.0029,
+      "step": 1474
+    },
+    {
+      "epoch": 2.211394302848576,
+      "grad_norm": 0.04822527848446775,
+      "learning_rate": 3.9262416027671356e-05,
+      "loss": 0.0032,
+      "step": 1475
+    },
+    {
+      "epoch": 2.212893553223388,
+      "grad_norm": 0.09769165771156645,
+      "learning_rate": 3.9123857099127936e-05,
+      "loss": 0.0066,
+      "step": 1476
+    },
+    {
+      "epoch": 2.214392803598201,
+      "grad_norm": 0.07720924648038567,
+      "learning_rate": 3.8985483609873244e-05,
+      "loss": 0.0014,
+      "step": 1477
+    },
+    {
+      "epoch": 2.2158920539730134,
+      "grad_norm": 0.04125133249769171,
+      "learning_rate": 3.8847295981416896e-05,
+      "loss": 0.003,
+      "step": 1478
+    },
+    {
+      "epoch": 2.217391304347826,
+      "grad_norm": 0.047023293986731454,
+      "learning_rate": 3.8709294634702376e-05,
+      "loss": 0.0027,
+      "step": 1479
+    },
+    {
+      "epoch": 2.2188905547226385,
+      "grad_norm": 0.03636072085533822,
+      "learning_rate": 3.857147999010567e-05,
+      "loss": 0.0028,
+      "step": 1480
+    },
+    {
+      "epoch": 2.2203898050974513,
+      "grad_norm": 0.03091756783151599,
+      "learning_rate": 3.843385246743417e-05,
+      "loss": 0.0018,
+      "step": 1481
+    },
+    {
+      "epoch": 2.2218890554722637,
+      "grad_norm": 0.06714013114185455,
+      "learning_rate": 3.829641248592515e-05,
+      "loss": 0.0049,
+      "step": 1482
+    },
+    {
+      "epoch": 2.2233883058470765,
+      "grad_norm": 0.10511179919131185,
+      "learning_rate": 3.8159160464244606e-05,
+      "loss": 0.0046,
+      "step": 1483
+    },
+    {
+      "epoch": 2.224887556221889,
+      "grad_norm": 0.03003926401709271,
+      "learning_rate": 3.802209682048602e-05,
+      "loss": 0.0023,
+      "step": 1484
+    },
+    {
+      "epoch": 2.2263868065967016,
+      "grad_norm": 0.03524357503089674,
+      "learning_rate": 3.788522197216897e-05,
+      "loss": 0.0027,
+      "step": 1485
+    },
+    {
+      "epoch": 2.2278860569715144,
+      "grad_norm": 0.11632249193591344,
+      "learning_rate": 3.774853633623806e-05,
+      "loss": 0.0047,
+      "step": 1486
+    },
+    {
+      "epoch": 2.229385307346327,
+      "grad_norm": 0.06473791327123747,
+      "learning_rate": 3.7612040329061405e-05,
+      "loss": 0.0054,
+      "step": 1487
+    },
+    {
+      "epoch": 2.2308845577211396,
+      "grad_norm": 0.049416266463761196,
+      "learning_rate": 3.747573436642951e-05,
+      "loss": 0.0034,
+      "step": 1488
+    },
+    {
+      "epoch": 2.232383808095952,
+      "grad_norm": 0.0388053166882485,
+      "learning_rate": 3.733961886355398e-05,
+      "loss": 0.004,
+      "step": 1489
+    },
+    {
+      "epoch": 2.2338830584707647,
+      "grad_norm": 0.0510239158161139,
+      "learning_rate": 3.720369423506622e-05,
+      "loss": 0.0035,
+      "step": 1490
+    },
+    {
+      "epoch": 2.235382308845577,
+      "grad_norm": 0.06027442038507764,
+      "learning_rate": 3.7067960895016275e-05,
+      "loss": 0.0038,
+      "step": 1491
+    },
+    {
+      "epoch": 2.23688155922039,
+      "grad_norm": 0.11520121661098545,
+      "learning_rate": 3.6932419256871406e-05,
+      "loss": 0.0062,
+      "step": 1492
+    },
+    {
+      "epoch": 2.2383808095952022,
+      "grad_norm": 0.06456017036528984,
+      "learning_rate": 3.679706973351491e-05,
+      "loss": 0.0027,
+      "step": 1493
+    },
+    {
+      "epoch": 2.239880059970015,
+      "grad_norm": 0.09253651512540281,
+      "learning_rate": 3.6661912737245e-05,
+      "loss": 0.0076,
+      "step": 1494
+    },
+    {
+      "epoch": 2.2413793103448274,
+      "grad_norm": 0.09671661159708796,
+      "learning_rate": 3.6526948679773257e-05,
+      "loss": 0.0033,
+      "step": 1495
+    },
+    {
+      "epoch": 2.24287856071964,
+      "grad_norm": 0.06702412437333265,
+      "learning_rate": 3.6392177972223594e-05,
+      "loss": 0.0036,
+      "step": 1496
+    },
+    {
+      "epoch": 2.244377811094453,
+      "grad_norm": 0.09090522248170291,
+      "learning_rate": 3.6257601025131026e-05,
+      "loss": 0.0062,
+      "step": 1497
+    },
+    {
+      "epoch": 2.2458770614692654,
+      "grad_norm": 0.058717896724903795,
+      "learning_rate": 3.612321824844024e-05,
+      "loss": 0.008,
+      "step": 1498
+    },
+    {
+      "epoch": 2.247376311844078,
+      "grad_norm": 0.05083383150547819,
+      "learning_rate": 3.5989030051504434e-05,
+      "loss": 0.0035,
+      "step": 1499
+    },
+    {
+      "epoch": 2.2488755622188905,
+      "grad_norm": 0.0698571999573116,
+      "learning_rate": 3.585503684308421e-05,
+      "loss": 0.004,
+      "step": 1500
+    },
+    {
+      "epoch": 2.2503748125937033,
+      "grad_norm": 0.03210517089158349,
+      "learning_rate": 3.5721239031346066e-05,
+      "loss": 0.0012,
+      "step": 1501
+    },
+    {
+      "epoch": 2.2518740629685157,
+      "grad_norm": 0.0508811347970621,
+      "learning_rate": 3.558763702386135e-05,
+      "loss": 0.0019,
+      "step": 1502
+    },
+    {
+      "epoch": 2.2533733133433285,
+      "grad_norm": 0.055664797337053686,
+      "learning_rate": 3.545423122760493e-05,
+      "loss": 0.0038,
+      "step": 1503
+    },
+    {
+      "epoch": 2.254872563718141,
+      "grad_norm": 0.08996644981999107,
+      "learning_rate": 3.5321022048954035e-05,
+      "loss": 0.0077,
+      "step": 1504
+    },
+    {
+      "epoch": 2.2563718140929536,
+      "grad_norm": 0.13609555101705403,
+      "learning_rate": 3.518800989368691e-05,
+      "loss": 0.0065,
+      "step": 1505
+    },
+    {
+      "epoch": 2.257871064467766,
+      "grad_norm": 0.050562321159216154,
+      "learning_rate": 3.5055195166981645e-05,
+      "loss": 0.0057,
+      "step": 1506
+    },
+    {
+      "epoch": 2.2593703148425788,
+      "grad_norm": 0.0728087625749308,
+      "learning_rate": 3.492257827341492e-05,
+      "loss": 0.0045,
+      "step": 1507
+    },
+    {
+      "epoch": 2.260869565217391,
+      "grad_norm": 0.09718734094804558,
+      "learning_rate": 3.479015961696077e-05,
+      "loss": 0.0071,
+      "step": 1508
+    },
+    {
+      "epoch": 2.262368815592204,
+      "grad_norm": 0.17375393173330367,
+      "learning_rate": 3.465793960098945e-05,
+      "loss": 0.002,
+      "step": 1509
+    },
+    {
+      "epoch": 2.2638680659670163,
+      "grad_norm": 0.021400919519601855,
+      "learning_rate": 3.452591862826603e-05,
+      "loss": 0.0018,
+      "step": 1510
+    },
+    {
+      "epoch": 2.265367316341829,
+      "grad_norm": 0.06996342276647459,
+      "learning_rate": 3.439409710094929e-05,
+      "loss": 0.0089,
+      "step": 1511
+    },
+    {
+      "epoch": 2.266866566716642,
+      "grad_norm": 0.061074064657634694,
+      "learning_rate": 3.426247542059041e-05,
+      "loss": 0.0011,
+      "step": 1512
+    },
+    {
+      "epoch": 2.2683658170914542,
+      "grad_norm": 0.06945067478625049,
+      "learning_rate": 3.413105398813195e-05,
+      "loss": 0.0051,
+      "step": 1513
+    },
+    {
+      "epoch": 2.269865067466267,
+      "grad_norm": 0.042079105603492865,
+      "learning_rate": 3.3999833203906326e-05,
+      "loss": 0.004,
+      "step": 1514
+    },
+    {
+      "epoch": 2.2713643178410794,
+      "grad_norm": 0.1636319307235689,
+      "learning_rate": 3.386881346763483e-05,
+      "loss": 0.0099,
+      "step": 1515
+    },
+    {
+      "epoch": 2.272863568215892,
+      "grad_norm": 0.06664106371082085,
+      "learning_rate": 3.373799517842627e-05,
+      "loss": 0.0044,
+      "step": 1516
+    },
+    {
+      "epoch": 2.2743628185907045,
+      "grad_norm": 0.08483429638899491,
+      "learning_rate": 3.360737873477584e-05,
+      "loss": 0.0034,
+      "step": 1517
+    },
+    {
+      "epoch": 2.2758620689655173,
+      "grad_norm": 0.06849123537105387,
+      "learning_rate": 3.347696453456393e-05,
+      "loss": 0.0076,
+      "step": 1518
+    },
+    {
+      "epoch": 2.2773613193403297,
+      "grad_norm": 0.18864254560141094,
+      "learning_rate": 3.334675297505476e-05,
+      "loss": 0.0068,
+      "step": 1519
+    },
+    {
+      "epoch": 2.2788605697151425,
+      "grad_norm": 0.02900241632175253,
+      "learning_rate": 3.3216744452895354e-05,
+      "loss": 0.0014,
+      "step": 1520
+    },
+    {
+      "epoch": 2.280359820089955,
+      "grad_norm": 0.06275154975100304,
+      "learning_rate": 3.308693936411421e-05,
+      "loss": 0.0055,
+      "step": 1521
+    },
+    {
+      "epoch": 2.2818590704647677,
+      "grad_norm": 0.1528582473234831,
+      "learning_rate": 3.2957338104120096e-05,
+      "loss": 0.0047,
+      "step": 1522
+    },
+    {
+      "epoch": 2.2833583208395805,
+      "grad_norm": 0.07980744631055499,
+      "learning_rate": 3.2827941067700996e-05,
+      "loss": 0.0056,
+      "step": 1523
+    },
+    {
+      "epoch": 2.284857571214393,
+      "grad_norm": 0.11944836679739665,
+      "learning_rate": 3.269874864902269e-05,
+      "loss": 0.0046,
+      "step": 1524
+    },
+    {
+      "epoch": 2.286356821589205,
+      "grad_norm": 0.023772227427949762,
+      "learning_rate": 3.2569761241627696e-05,
+      "loss": 0.0016,
+      "step": 1525
+    },
+    {
+      "epoch": 2.287856071964018,
+      "grad_norm": 0.14329800784900046,
+      "learning_rate": 3.244097923843398e-05,
+      "loss": 0.0148,
+      "step": 1526
+    },
+    {
+      "epoch": 2.2893553223388308,
+      "grad_norm": 0.1195966321417723,
+      "learning_rate": 3.231240303173394e-05,
+      "loss": 0.0057,
+      "step": 1527
+    },
+    {
+      "epoch": 2.290854572713643,
+      "grad_norm": 0.10708809110295565,
+      "learning_rate": 3.218403301319296e-05,
+      "loss": 0.0065,
+      "step": 1528
+    },
+    {
+      "epoch": 2.292353823088456,
+      "grad_norm": 0.08568289613574692,
+      "learning_rate": 3.205586957384838e-05,
+      "loss": 0.0032,
+      "step": 1529
+    },
+    {
+      "epoch": 2.2938530734632683,
+      "grad_norm": 0.035756635002319404,
+      "learning_rate": 3.192791310410822e-05,
+      "loss": 0.002,
+      "step": 1530
+    },
+    {
+      "epoch": 2.295352323838081,
+      "grad_norm": 0.06053419631054795,
+      "learning_rate": 3.1800163993750166e-05,
+      "loss": 0.0056,
+      "step": 1531
+    },
+    {
+      "epoch": 2.2968515742128934,
+      "grad_norm": 0.1039350364479123,
+      "learning_rate": 3.16726226319201e-05,
+      "loss": 0.0081,
+      "step": 1532
+    },
+    {
+      "epoch": 2.2983508245877062,
+      "grad_norm": 0.07906247819210281,
+      "learning_rate": 3.154528940713113e-05,
+      "loss": 0.0067,
+      "step": 1533
+    },
+    {
+      "epoch": 2.2998500749625186,
+      "grad_norm": 0.08407334961019255,
+      "learning_rate": 3.141816470726238e-05,
+      "loss": 0.0113,
+      "step": 1534
+    },
+    {
+      "epoch": 2.3013493253373314,
+      "grad_norm": 0.08815120231572529,
+      "learning_rate": 3.129124891955771e-05,
+      "loss": 0.0052,
+      "step": 1535
+    },
+    {
+      "epoch": 2.3028485757121437,
+      "grad_norm": 0.0666277487037083,
+      "learning_rate": 3.116454243062459e-05,
+      "loss": 0.0032,
+      "step": 1536
+    },
+    {
+      "epoch": 2.3043478260869565,
+      "grad_norm": 0.04918000761524814,
+      "learning_rate": 3.103804562643302e-05,
+      "loss": 0.0043,
+      "step": 1537
+    },
+    {
+      "epoch": 2.3058470764617693,
+      "grad_norm": 0.060187697186195384,
+      "learning_rate": 3.0911758892314166e-05,
+      "loss": 0.0035,
+      "step": 1538
+    },
+    {
+      "epoch": 2.3073463268365817,
+      "grad_norm": 0.16836993548608092,
+      "learning_rate": 3.078568261295933e-05,
+      "loss": 0.0052,
+      "step": 1539
+    },
+    {
+      "epoch": 2.3088455772113945,
+      "grad_norm": 0.06729086558094731,
+      "learning_rate": 3.0659817172418693e-05,
+      "loss": 0.0049,
+      "step": 1540
+    },
+    {
+      "epoch": 2.310344827586207,
+      "grad_norm": 0.07617128616033088,
+      "learning_rate": 3.053416295410026e-05,
+      "loss": 0.0056,
+      "step": 1541
+    },
+    {
+      "epoch": 2.3118440779610197,
+      "grad_norm": 0.114464765928607,
+      "learning_rate": 3.0408720340768572e-05,
+      "loss": 0.0029,
+      "step": 1542
+    },
+    {
+      "epoch": 2.313343328335832,
+      "grad_norm": 0.063347013144408,
+      "learning_rate": 3.0283489714543556e-05,
+      "loss": 0.0045,
+      "step": 1543
+    },
+    {
+      "epoch": 2.314842578710645,
+      "grad_norm": 0.06774768128325656,
+      "learning_rate": 3.0158471456899428e-05,
+      "loss": 0.0022,
+      "step": 1544
+    },
+    {
+      "epoch": 2.316341829085457,
+      "grad_norm": 0.048376078472133446,
+      "learning_rate": 3.0033665948663448e-05,
+      "loss": 0.0028,
+      "step": 1545
+    },
+    {
+      "epoch": 2.31784107946027,
+      "grad_norm": 0.0385182292711496,
+      "learning_rate": 2.9909073570014912e-05,
+      "loss": 0.0014,
+      "step": 1546
+    },
+    {
+      "epoch": 2.3193403298350823,
+      "grad_norm": 0.03475187689682614,
+      "learning_rate": 2.9784694700483762e-05,
+      "loss": 0.0017,
+      "step": 1547
+    },
+    {
+      "epoch": 2.320839580209895,
+      "grad_norm": 0.07347410055145828,
+      "learning_rate": 2.9660529718949627e-05,
+      "loss": 0.0049,
+      "step": 1548
+    },
+    {
+      "epoch": 2.3223388305847075,
+      "grad_norm": 0.06424491061862778,
+      "learning_rate": 2.953657900364053e-05,
+      "loss": 0.004,
+      "step": 1549
+    },
+    {
+      "epoch": 2.3238380809595203,
+      "grad_norm": 0.09419900753622726,
+      "learning_rate": 2.94128429321319e-05,
+      "loss": 0.007,
+      "step": 1550
+    },
+    {
+      "epoch": 2.3253373313343326,
+      "grad_norm": 0.08715139140623997,
+      "learning_rate": 2.9289321881345254e-05,
+      "loss": 0.0045,
+      "step": 1551
+    },
+    {
+      "epoch": 2.3268365817091454,
+      "grad_norm": 0.06268684955631437,
+      "learning_rate": 2.9166016227547133e-05,
+      "loss": 0.0041,
+      "step": 1552
+    },
+    {
+      "epoch": 2.3283358320839582,
+      "grad_norm": 0.1379363777149399,
+      "learning_rate": 2.904292634634793e-05,
+      "loss": 0.0034,
+      "step": 1553
+    },
+    {
+      "epoch": 2.3298350824587706,
+      "grad_norm": 0.10089329635825812,
+      "learning_rate": 2.8920052612700754e-05,
+      "loss": 0.0062,
+      "step": 1554
+    },
+    {
+      "epoch": 2.3313343328335834,
+      "grad_norm": 0.05387050027903078,
+      "learning_rate": 2.879739540090036e-05,
+      "loss": 0.0066,
+      "step": 1555
+    },
+    {
+      "epoch": 2.3328335832083957,
+      "grad_norm": 0.05660842464190973,
+      "learning_rate": 2.8674955084581857e-05,
+      "loss": 0.0042,
+      "step": 1556
+    },
+    {
+      "epoch": 2.3343328335832085,
+      "grad_norm": 0.05686811860047845,
+      "learning_rate": 2.8552732036719687e-05,
+      "loss": 0.0051,
+      "step": 1557
+    },
+    {
+      "epoch": 2.335832083958021,
+      "grad_norm": 0.04938712425283094,
+      "learning_rate": 2.8430726629626413e-05,
+      "loss": 0.0018,
+      "step": 1558
+    },
+    {
+      "epoch": 2.3373313343328337,
+      "grad_norm": 0.13710965858068258,
+      "learning_rate": 2.8308939234951726e-05,
+      "loss": 0.0084,
+      "step": 1559
+    },
+    {
+      "epoch": 2.338830584707646,
+      "grad_norm": 0.09065138517655111,
+      "learning_rate": 2.8187370223681132e-05,
+      "loss": 0.0052,
+      "step": 1560
+    },
+    {
+      "epoch": 2.340329835082459,
+      "grad_norm": 0.04322879635848861,
+      "learning_rate": 2.8066019966134904e-05,
+      "loss": 0.0054,
+      "step": 1561
+    },
+    {
+      "epoch": 2.341829085457271,
+      "grad_norm": 0.14140024545462096,
+      "learning_rate": 2.7944888831966987e-05,
+      "loss": 0.0069,
+      "step": 1562
+    },
+    {
+      "epoch": 2.343328335832084,
+      "grad_norm": 0.05855557490735462,
+      "learning_rate": 2.7823977190163786e-05,
+      "loss": 0.0038,
+      "step": 1563
+    },
+    {
+      "epoch": 2.344827586206897,
+      "grad_norm": 0.06774189505154501,
+      "learning_rate": 2.770328540904319e-05,
+      "loss": 0.0055,
+      "step": 1564
+    },
+    {
+      "epoch": 2.346326836581709,
+      "grad_norm": 0.06685740409439345,
+      "learning_rate": 2.7582813856253275e-05,
+      "loss": 0.0073,
+      "step": 1565
+    },
+    {
+      "epoch": 2.3478260869565215,
+      "grad_norm": 0.06597092919168687,
+      "learning_rate": 2.746256289877126e-05,
+      "loss": 0.0049,
+      "step": 1566
+    },
+    {
+      "epoch": 2.3493253373313343,
+      "grad_norm": 0.06984959414446605,
+      "learning_rate": 2.734253290290242e-05,
+      "loss": 0.0029,
+      "step": 1567
+    },
+    {
+      "epoch": 2.350824587706147,
+      "grad_norm": 0.08794480426989766,
+      "learning_rate": 2.722272423427896e-05,
+      "loss": 0.0046,
+      "step": 1568
+    },
+    {
+      "epoch": 2.3523238380809595,
+      "grad_norm": 0.08294773284079503,
+      "learning_rate": 2.7103137257858868e-05,
+      "loss": 0.0053,
+      "step": 1569
+    },
+    {
+      "epoch": 2.3538230884557723,
+      "grad_norm": 0.05311205720612156,
+      "learning_rate": 2.698377233792476e-05,
+      "loss": 0.0031,
+      "step": 1570
+    },
+    {
+      "epoch": 2.3553223388305846,
+      "grad_norm": 0.044514573918505385,
+      "learning_rate": 2.6864629838082956e-05,
+      "loss": 0.0045,
+      "step": 1571
+    },
+    {
+      "epoch": 2.3568215892053974,
+      "grad_norm": 0.08026506351168762,
+      "learning_rate": 2.6745710121262136e-05,
+      "loss": 0.0049,
+      "step": 1572
+    },
+    {
+      "epoch": 2.3583208395802098,
+      "grad_norm": 0.030283279059830456,
+      "learning_rate": 2.6627013549712355e-05,
+      "loss": 0.0017,
+      "step": 1573
+    },
+    {
+      "epoch": 2.3598200899550226,
+      "grad_norm": 0.056873656422552585,
+      "learning_rate": 2.6508540485004006e-05,
+      "loss": 0.0031,
+      "step": 1574
+    },
+    {
+      "epoch": 2.361319340329835,
+      "grad_norm": 0.0739899949993969,
+      "learning_rate": 2.639029128802657e-05,
+      "loss": 0.0048,
+      "step": 1575
+    },
+    {
+      "epoch": 2.3628185907046477,
+      "grad_norm": 0.07078189352862575,
+      "learning_rate": 2.6272266318987603e-05,
+      "loss": 0.0038,
+      "step": 1576
+    },
+    {
+      "epoch": 2.36431784107946,
+      "grad_norm": 0.09589625252399156,
+      "learning_rate": 2.615446593741161e-05,
+      "loss": 0.0061,
+      "step": 1577
+    },
+    {
+      "epoch": 2.365817091454273,
+      "grad_norm": 0.07845365529561737,
+      "learning_rate": 2.603689050213902e-05,
+      "loss": 0.0039,
+      "step": 1578
+    },
+    {
+      "epoch": 2.3673163418290857,
+      "grad_norm": 0.09914753466937914,
+      "learning_rate": 2.5919540371325e-05,
+      "loss": 0.0052,
+      "step": 1579
+    },
+    {
+      "epoch": 2.368815592203898,
+      "grad_norm": 0.12930439990410358,
+      "learning_rate": 2.580241590243837e-05,
+      "loss": 0.0081,
+      "step": 1580
+    },
+    {
+      "epoch": 2.370314842578711,
+      "grad_norm": 0.023463536942587522,
+      "learning_rate": 2.5685517452260567e-05,
+      "loss": 0.0022,
+      "step": 1581
+    },
+    {
+      "epoch": 2.371814092953523,
+      "grad_norm": 0.07218813844672958,
+      "learning_rate": 2.5568845376884587e-05,
+      "loss": 0.0027,
+      "step": 1582
+    },
+    {
+      "epoch": 2.373313343328336,
+      "grad_norm": 0.06641391796530317,
+      "learning_rate": 2.5452400031713785e-05,
+      "loss": 0.0064,
+      "step": 1583
+    },
+    {
+      "epoch": 2.3748125937031483,
+      "grad_norm": 0.03574796789731248,
+      "learning_rate": 2.5336181771460876e-05,
+      "loss": 0.0019,
+      "step": 1584
+    },
+    {
+      "epoch": 2.376311844077961,
+      "grad_norm": 0.04934069132677985,
+      "learning_rate": 2.5220190950146827e-05,
+      "loss": 0.0026,
+      "step": 1585
+    },
+    {
+      "epoch": 2.3778110944527735,
+      "grad_norm": 0.053344136812610306,
+      "learning_rate": 2.5104427921099782e-05,
+      "loss": 0.0035,
+      "step": 1586
+    },
+    {
+      "epoch": 2.3793103448275863,
+      "grad_norm": 0.05934960469965204,
+      "learning_rate": 2.4988893036954043e-05,
+      "loss": 0.0052,
+      "step": 1587
+    },
+    {
+      "epoch": 2.3808095952023987,
+      "grad_norm": 0.07363174118581005,
+      "learning_rate": 2.4873586649648894e-05,
+      "loss": 0.0021,
+      "step": 1588
+    },
+    {
+      "epoch": 2.3823088455772115,
+      "grad_norm": 0.06507057844090482,
+      "learning_rate": 2.4758509110427575e-05,
+      "loss": 0.0064,
+      "step": 1589
+    },
+    {
+      "epoch": 2.383808095952024,
+      "grad_norm": 0.04011004024330606,
+      "learning_rate": 2.464366076983623e-05,
+      "loss": 0.0016,
+      "step": 1590
+    },
+    {
+      "epoch": 2.3853073463268366,
+      "grad_norm": 0.029963463107639907,
+      "learning_rate": 2.45290419777228e-05,
+      "loss": 0.0012,
+      "step": 1591
+    },
+    {
+      "epoch": 2.386806596701649,
+      "grad_norm": 0.11065244961492637,
+      "learning_rate": 2.441465308323605e-05,
+      "loss": 0.0063,
+      "step": 1592
+    },
+    {
+      "epoch": 2.3883058470764618,
+      "grad_norm": 0.07476696771495371,
+      "learning_rate": 2.4300494434824373e-05,
+      "loss": 0.003,
+      "step": 1593
+    },
+    {
+      "epoch": 2.3898050974512746,
+      "grad_norm": 0.10626192717766343,
+      "learning_rate": 2.4186566380234798e-05,
+      "loss": 0.0101,
+      "step": 1594
+    },
+    {
+      "epoch": 2.391304347826087,
+      "grad_norm": 0.06898129289247987,
+      "learning_rate": 2.407286926651192e-05,
+      "loss": 0.0037,
+      "step": 1595
+    },
+    {
+      "epoch": 2.3928035982008997,
+      "grad_norm": 0.030965851115252685,
+      "learning_rate": 2.3959403439996907e-05,
+      "loss": 0.0023,
+      "step": 1596
+    },
+    {
+      "epoch": 2.394302848575712,
+      "grad_norm": 0.12448248521634331,
+      "learning_rate": 2.3846169246326343e-05,
+      "loss": 0.004,
+      "step": 1597
+    },
+    {
+      "epoch": 2.395802098950525,
+      "grad_norm": 0.0496245094173845,
+      "learning_rate": 2.373316703043119e-05,
+      "loss": 0.0039,
+      "step": 1598
+    },
+    {
+      "epoch": 2.3973013493253372,
+      "grad_norm": 0.06468143373332225,
+      "learning_rate": 2.362039713653581e-05,
+      "loss": 0.0039,
+      "step": 1599
+    },
+    {
+      "epoch": 2.39880059970015,
+      "grad_norm": 0.02745569957950435,
+      "learning_rate": 2.3507859908156827e-05,
+      "loss": 0.0035,
+      "step": 1600
+    },
+    {
+      "epoch": 2.4002998500749624,
+      "grad_norm": 0.0884132237063567,
+      "learning_rate": 2.339555568810221e-05,
+      "loss": 0.0073,
+      "step": 1601
+    },
+    {
+      "epoch": 2.401799100449775,
+      "grad_norm": 0.09407281067698077,
+      "learning_rate": 2.328348481847006e-05,
+      "loss": 0.0043,
+      "step": 1602
+    },
+    {
+      "epoch": 2.4032983508245875,
+      "grad_norm": 0.023867483923462753,
+      "learning_rate": 2.3171647640647687e-05,
+      "loss": 0.0009,
+      "step": 1603
+    },
+    {
+      "epoch": 2.4047976011994003,
+      "grad_norm": 0.05163847859209614,
+      "learning_rate": 2.3060044495310505e-05,
+      "loss": 0.0023,
+      "step": 1604
+    },
+    {
+      "epoch": 2.406296851574213,
+      "grad_norm": 0.047121807189413514,
+      "learning_rate": 2.2948675722421086e-05,
+      "loss": 0.003,
+      "step": 1605
+    },
+    {
+      "epoch": 2.4077961019490255,
+      "grad_norm": 0.07119567947811782,
+      "learning_rate": 2.2837541661228025e-05,
+      "loss": 0.003,
+      "step": 1606
+    },
+    {
+      "epoch": 2.409295352323838,
+      "grad_norm": 0.053849133740421475,
+      "learning_rate": 2.2726642650264895e-05,
+      "loss": 0.0019,
+      "step": 1607
+    },
+    {
+      "epoch": 2.4107946026986506,
+      "grad_norm": 0.07662771818389892,
+      "learning_rate": 2.2615979027349387e-05,
+      "loss": 0.0051,
+      "step": 1608
+    },
+    {
+      "epoch": 2.4122938530734634,
+      "grad_norm": 0.14925873267020182,
+      "learning_rate": 2.2505551129582047e-05,
+      "loss": 0.0066,
+      "step": 1609
+    },
+    {
+      "epoch": 2.413793103448276,
+      "grad_norm": 0.09592241042779556,
+      "learning_rate": 2.2395359293345396e-05,
+      "loss": 0.0061,
+      "step": 1610
+    },
+    {
+      "epoch": 2.4152923538230886,
+      "grad_norm": 0.127006071375077,
+      "learning_rate": 2.2285403854302912e-05,
+      "loss": 0.0055,
+      "step": 1611
+    },
+    {
+      "epoch": 2.416791604197901,
+      "grad_norm": 0.06229520326818543,
+      "learning_rate": 2.2175685147397906e-05,
+      "loss": 0.0054,
+      "step": 1612
+    },
+    {
+      "epoch": 2.4182908545727138,
+      "grad_norm": 0.04966340726882021,
+      "learning_rate": 2.2066203506852566e-05,
+      "loss": 0.0025,
+      "step": 1613
+    },
+    {
+      "epoch": 2.419790104947526,
+      "grad_norm": 0.052824372953982054,
+      "learning_rate": 2.195695926616702e-05,
+      "loss": 0.0024,
+      "step": 1614
+    },
+    {
+      "epoch": 2.421289355322339,
+      "grad_norm": 0.0693320471877942,
+      "learning_rate": 2.1847952758118117e-05,
+      "loss": 0.0057,
+      "step": 1615
+    },
+    {
+      "epoch": 2.4227886056971513,
+      "grad_norm": 0.1585165428374079,
+      "learning_rate": 2.173918431475861e-05,
+      "loss": 0.0079,
+      "step": 1616
+    },
+    {
+      "epoch": 2.424287856071964,
+      "grad_norm": 0.12183346017418488,
+      "learning_rate": 2.163065426741603e-05,
+      "loss": 0.0064,
+      "step": 1617
+    },
+    {
+      "epoch": 2.4257871064467764,
+      "grad_norm": 0.045242084358939956,
+      "learning_rate": 2.1522362946691698e-05,
+      "loss": 0.0054,
+      "step": 1618
+    },
+    {
+      "epoch": 2.427286356821589,
+      "grad_norm": 0.0722336789715876,
+      "learning_rate": 2.1414310682459802e-05,
+      "loss": 0.0031,
+      "step": 1619
+    },
+    {
+      "epoch": 2.428785607196402,
+      "grad_norm": 0.07131568496067937,
+      "learning_rate": 2.1306497803866277e-05,
+      "loss": 0.0058,
+      "step": 1620
+    },
+    {
+      "epoch": 2.4302848575712144,
+      "grad_norm": 0.037029961293887234,
+      "learning_rate": 2.119892463932781e-05,
+      "loss": 0.0028,
+      "step": 1621
+    },
+    {
+      "epoch": 2.431784107946027,
+      "grad_norm": 0.05980642170587757,
+      "learning_rate": 2.1091591516530952e-05,
+      "loss": 0.0089,
+      "step": 1622
+    },
+    {
+      "epoch": 2.4332833583208395,
+      "grad_norm": 0.06090138836468679,
+      "learning_rate": 2.098449876243096e-05,
+      "loss": 0.0025,
+      "step": 1623
+    },
+    {
+      "epoch": 2.4347826086956523,
+      "grad_norm": 0.05351342549456341,
+      "learning_rate": 2.0877646703250996e-05,
+      "loss": 0.0057,
+      "step": 1624
+    },
+    {
+      "epoch": 2.4362818590704647,
+      "grad_norm": 0.03433223890330924,
+      "learning_rate": 2.0771035664480942e-05,
+      "loss": 0.0013,
+      "step": 1625
+    },
+    {
+      "epoch": 2.4377811094452775,
+      "grad_norm": 0.047845862718465054,
+      "learning_rate": 2.0664665970876496e-05,
+      "loss": 0.0015,
+      "step": 1626
+    },
+    {
+      "epoch": 2.43928035982009,
+      "grad_norm": 0.02633163873788185,
+      "learning_rate": 2.0558537946458177e-05,
+      "loss": 0.0013,
+      "step": 1627
+    },
+    {
+      "epoch": 2.4407796101949026,
+      "grad_norm": 0.07332498941147272,
+      "learning_rate": 2.045265191451041e-05,
+      "loss": 0.0044,
+      "step": 1628
+    },
+    {
+      "epoch": 2.442278860569715,
+      "grad_norm": 0.16264733297325074,
+      "learning_rate": 2.0347008197580374e-05,
+      "loss": 0.0213,
+      "step": 1629
+    },
+    {
+      "epoch": 2.443778110944528,
+      "grad_norm": 0.03432376223511998,
+      "learning_rate": 2.024160711747717e-05,
+      "loss": 0.0027,
+      "step": 1630
+    },
+    {
+      "epoch": 2.44527736131934,
+      "grad_norm": 0.08444598432390558,
+      "learning_rate": 2.013644899527074e-05,
+      "loss": 0.0086,
+      "step": 1631
+    },
+    {
+      "epoch": 2.446776611694153,
+      "grad_norm": 0.09785171406288519,
+      "learning_rate": 2.0031534151290943e-05,
+      "loss": 0.0145,
+      "step": 1632
+    },
+    {
+      "epoch": 2.4482758620689653,
+      "grad_norm": 0.10686482838519329,
+      "learning_rate": 1.9926862905126665e-05,
+      "loss": 0.0106,
+      "step": 1633
+    },
+    {
+      "epoch": 2.449775112443778,
+      "grad_norm": 0.07512971821946084,
+      "learning_rate": 1.9822435575624608e-05,
+      "loss": 0.0045,
+      "step": 1634
+    },
+    {
+      "epoch": 2.451274362818591,
+      "grad_norm": 0.10390776644311271,
+      "learning_rate": 1.9718252480888566e-05,
+      "loss": 0.0067,
+      "step": 1635
+    },
+    {
+      "epoch": 2.4527736131934033,
+      "grad_norm": 0.054546501477964644,
+      "learning_rate": 1.9614313938278272e-05,
+      "loss": 0.0041,
+      "step": 1636
+    },
+    {
+      "epoch": 2.454272863568216,
+      "grad_norm": 0.1236643546327681,
+      "learning_rate": 1.9510620264408596e-05,
+      "loss": 0.0076,
+      "step": 1637
+    },
+    {
+      "epoch": 2.4557721139430284,
+      "grad_norm": 0.06368717471421793,
+      "learning_rate": 1.9407171775148436e-05,
+      "loss": 0.0046,
+      "step": 1638
+    },
+    {
+      "epoch": 2.457271364317841,
+      "grad_norm": 0.058824393187371,
+      "learning_rate": 1.930396878561983e-05,
+      "loss": 0.0056,
+      "step": 1639
+    },
+    {
+      "epoch": 2.4587706146926536,
+      "grad_norm": 0.052763276928937744,
+      "learning_rate": 1.9201011610196973e-05,
+      "loss": 0.0042,
+      "step": 1640
+    },
+    {
+      "epoch": 2.4602698650674664,
+      "grad_norm": 0.04366071201528183,
+      "learning_rate": 1.9098300562505266e-05,
+      "loss": 0.0038,
+      "step": 1641
+    },
+    {
+      "epoch": 2.4617691154422787,
+      "grad_norm": 0.060492485188534235,
+      "learning_rate": 1.8995835955420416e-05,
+      "loss": 0.0027,
+      "step": 1642
+    },
+    {
+      "epoch": 2.4632683658170915,
+      "grad_norm": 0.12378841704182127,
+      "learning_rate": 1.8893618101067355e-05,
+      "loss": 0.0098,
+      "step": 1643
+    },
+    {
+      "epoch": 2.464767616191904,
+      "grad_norm": 0.06572768630271339,
+      "learning_rate": 1.879164731081937e-05,
+      "loss": 0.0026,
+      "step": 1644
+    },
+    {
+      "epoch": 2.4662668665667167,
+      "grad_norm": 0.09310343642883542,
+      "learning_rate": 1.8689923895297245e-05,
+      "loss": 0.0047,
+      "step": 1645
+    },
+    {
+      "epoch": 2.4677661169415295,
+      "grad_norm": 0.044289718611579314,
+      "learning_rate": 1.858844816436809e-05,
+      "loss": 0.0016,
+      "step": 1646
+    },
+    {
+      "epoch": 2.469265367316342,
+      "grad_norm": 0.04572702054619638,
+      "learning_rate": 1.848722042714457e-05,
+      "loss": 0.0018,
+      "step": 1647
+    },
+    {
+      "epoch": 2.470764617691154,
+      "grad_norm": 0.04476649599527431,
+      "learning_rate": 1.838624099198397e-05,
+      "loss": 0.0048,
+      "step": 1648
+    },
+    {
+      "epoch": 2.472263868065967,
+      "grad_norm": 0.07032354780767394,
+      "learning_rate": 1.8285510166487152e-05,
+      "loss": 0.0078,
+      "step": 1649
+    },
+    {
+      "epoch": 2.47376311844078,
+      "grad_norm": 0.09134832895870344,
+      "learning_rate": 1.818502825749764e-05,
+      "loss": 0.0084,
+      "step": 1650
+    },
+    {
+      "epoch": 2.475262368815592,
+      "grad_norm": 0.10898437749116632,
+      "learning_rate": 1.808479557110081e-05,
+      "loss": 0.0026,
+      "step": 1651
+    },
+    {
+      "epoch": 2.476761619190405,
+      "grad_norm": 0.09805488182454494,
+      "learning_rate": 1.7984812412622787e-05,
+      "loss": 0.0023,
+      "step": 1652
+    },
+    {
+      "epoch": 2.4782608695652173,
+      "grad_norm": 0.08765732874500178,
+      "learning_rate": 1.78850790866296e-05,
+      "loss": 0.0035,
+      "step": 1653
+    },
+    {
+      "epoch": 2.47976011994003,
+      "grad_norm": 0.062345434719609814,
+      "learning_rate": 1.7785595896926265e-05,
+      "loss": 0.0037,
+      "step": 1654
+    },
+    {
+      "epoch": 2.4812593703148424,
+      "grad_norm": 0.05004851068742012,
+      "learning_rate": 1.7686363146555805e-05,
+      "loss": 0.0019,
+      "step": 1655
+    },
+    {
+      "epoch": 2.4827586206896552,
+      "grad_norm": 0.12233324261355612,
+      "learning_rate": 1.7587381137798432e-05,
+      "loss": 0.0084,
+      "step": 1656
+    },
+    {
+      "epoch": 2.4842578710644676,
+      "grad_norm": 0.25709397112046917,
+      "learning_rate": 1.7488650172170496e-05,
+      "loss": 0.0133,
+      "step": 1657
+    },
+    {
+      "epoch": 2.4857571214392804,
+      "grad_norm": 0.1112658667316051,
+      "learning_rate": 1.7390170550423625e-05,
+      "loss": 0.0071,
+      "step": 1658
+    },
+    {
+      "epoch": 2.4872563718140928,
+      "grad_norm": 0.033739803171623846,
+      "learning_rate": 1.7291942572543807e-05,
+      "loss": 0.0031,
+      "step": 1659
+    },
+    {
+      "epoch": 2.4887556221889056,
+      "grad_norm": 0.05031209731885149,
+      "learning_rate": 1.719396653775056e-05,
+      "loss": 0.0033,
+      "step": 1660
+    },
+    {
+      "epoch": 2.4902548725637184,
+      "grad_norm": 0.10642934597922178,
+      "learning_rate": 1.7096242744495837e-05,
+      "loss": 0.0055,
+      "step": 1661
+    },
+    {
+      "epoch": 2.4917541229385307,
+      "grad_norm": 0.07260588589200684,
+      "learning_rate": 1.6998771490463262e-05,
+      "loss": 0.0084,
+      "step": 1662
+    },
+    {
+      "epoch": 2.4932533733133435,
+      "grad_norm": 0.0583410528108744,
+      "learning_rate": 1.690155307256719e-05,
+      "loss": 0.0048,
+      "step": 1663
+    },
+    {
+      "epoch": 2.494752623688156,
+      "grad_norm": 0.09011124093388319,
+      "learning_rate": 1.680458778695174e-05,
+      "loss": 0.0043,
+      "step": 1664
+    },
+    {
+      "epoch": 2.4962518740629687,
+      "grad_norm": 0.21013483131124727,
+      "learning_rate": 1.6707875928990058e-05,
+      "loss": 0.0126,
+      "step": 1665
+    },
+    {
+      "epoch": 2.497751124437781,
+      "grad_norm": 0.044788060768008446,
+      "learning_rate": 1.661141779328319e-05,
+      "loss": 0.0045,
+      "step": 1666
+    },
+    {
+      "epoch": 2.499250374812594,
+      "grad_norm": 0.09719586856323512,
+      "learning_rate": 1.6515213673659357e-05,
+      "loss": 0.0081,
+      "step": 1667
+    },
+    {
+      "epoch": 2.500749625187406,
+      "grad_norm": 0.044720423379557714,
+      "learning_rate": 1.6419263863172997e-05,
+      "loss": 0.004,
+      "step": 1668
+    },
+    {
+      "epoch": 2.502248875562219,
+      "grad_norm": 0.052589040164308765,
+      "learning_rate": 1.632356865410384e-05,
+      "loss": 0.006,
+      "step": 1669
+    },
+    {
+      "epoch": 2.5037481259370313,
+      "grad_norm": 0.1143440138081822,
+      "learning_rate": 1.622812833795613e-05,
+      "loss": 0.008,
+      "step": 1670
+    },
+    {
+      "epoch": 2.505247376311844,
+      "grad_norm": 0.09238843455164125,
+      "learning_rate": 1.6132943205457606e-05,
+      "loss": 0.0082,
+      "step": 1671
+    },
+    {
+      "epoch": 2.506746626686657,
+      "grad_norm": 0.056736669939192,
+      "learning_rate": 1.6038013546558696e-05,
+      "loss": 0.0026,
+      "step": 1672
+    },
+    {
+      "epoch": 2.5082458770614693,
+      "grad_norm": 0.12807829910070073,
+      "learning_rate": 1.5943339650431576e-05,
+      "loss": 0.0086,
+      "step": 1673
+    },
+    {
+      "epoch": 2.5097451274362816,
+      "grad_norm": 0.050687004065841715,
+      "learning_rate": 1.5848921805469397e-05,
+      "loss": 0.0041,
+      "step": 1674
+    },
+    {
+      "epoch": 2.5112443778110944,
+      "grad_norm": 0.07876565827972132,
+      "learning_rate": 1.5754760299285252e-05,
+      "loss": 0.005,
+      "step": 1675
+    },
+    {
+      "epoch": 2.5127436281859072,
+      "grad_norm": 0.04259095718764495,
+      "learning_rate": 1.566085541871145e-05,
+      "loss": 0.0014,
+      "step": 1676
+    },
+    {
+      "epoch": 2.5142428785607196,
+      "grad_norm": 0.1209625050394713,
+      "learning_rate": 1.5567207449798515e-05,
+      "loss": 0.0108,
+      "step": 1677
+    },
+    {
+      "epoch": 2.5157421289355324,
+      "grad_norm": 0.037696859736511955,
+      "learning_rate": 1.547381667781439e-05,
+      "loss": 0.0028,
+      "step": 1678
+    },
+    {
+      "epoch": 2.5172413793103448,
+      "grad_norm": 0.03303993337115884,
+      "learning_rate": 1.538068338724361e-05,
+      "loss": 0.0024,
+      "step": 1679
+    },
+    {
+      "epoch": 2.5187406296851576,
+      "grad_norm": 0.04950974456854536,
+      "learning_rate": 1.528780786178631e-05,
+      "loss": 0.0059,
+      "step": 1680
+    },
+    {
+      "epoch": 2.52023988005997,
+      "grad_norm": 0.07067979674428795,
+      "learning_rate": 1.5195190384357404e-05,
+      "loss": 0.0055,
+      "step": 1681
+    },
+    {
+      "epoch": 2.5217391304347827,
+      "grad_norm": 0.052179735514149166,
+      "learning_rate": 1.5102831237085857e-05,
+      "loss": 0.0043,
+      "step": 1682
+    },
+    {
+      "epoch": 2.523238380809595,
+      "grad_norm": 0.08034175791414389,
+      "learning_rate": 1.5010730701313625e-05,
+      "loss": 0.0082,
+      "step": 1683
+    },
+    {
+      "epoch": 2.524737631184408,
+      "grad_norm": 0.06320598136407608,
+      "learning_rate": 1.4918889057594876e-05,
+      "loss": 0.0025,
+      "step": 1684
+    },
+    {
+      "epoch": 2.52623688155922,
+      "grad_norm": 0.1498719753089046,
+      "learning_rate": 1.4827306585695234e-05,
+      "loss": 0.0032,
+      "step": 1685
+    },
+    {
+      "epoch": 2.527736131934033,
+      "grad_norm": 0.07584597146514462,
+      "learning_rate": 1.4735983564590783e-05,
+      "loss": 0.003,
+      "step": 1686
+    },
+    {
+      "epoch": 2.529235382308846,
+      "grad_norm": 0.04125809843734668,
+      "learning_rate": 1.4644920272467244e-05,
+      "loss": 0.0015,
+      "step": 1687
+    },
+    {
+      "epoch": 2.530734632683658,
+      "grad_norm": 0.05314353996550423,
+      "learning_rate": 1.4554116986719257e-05,
+      "loss": 0.0073,
+      "step": 1688
+    },
+    {
+      "epoch": 2.5322338830584705,
+      "grad_norm": 0.010199439979138412,
+      "learning_rate": 1.4463573983949341e-05,
+      "loss": 0.0004,
+      "step": 1689
+    },
+    {
+      "epoch": 2.5337331334332833,
+      "grad_norm": 0.13381608061216652,
+      "learning_rate": 1.4373291539967182e-05,
+      "loss": 0.0059,
+      "step": 1690
+    },
+    {
+      "epoch": 2.535232383808096,
+      "grad_norm": 0.09355504087054788,
+      "learning_rate": 1.4283269929788779e-05,
+      "loss": 0.005,
+      "step": 1691
+    },
+    {
+      "epoch": 2.5367316341829085,
+      "grad_norm": 0.02939896423318598,
+      "learning_rate": 1.4193509427635543e-05,
+      "loss": 0.0027,
+      "step": 1692
+    },
+    {
+      "epoch": 2.5382308845577213,
+      "grad_norm": 0.0400275347334245,
+      "learning_rate": 1.4104010306933557e-05,
+      "loss": 0.0025,
+      "step": 1693
+    },
+    {
+      "epoch": 2.5397301349325336,
+      "grad_norm": 0.07400393800455013,
+      "learning_rate": 1.4014772840312663e-05,
+      "loss": 0.0036,
+      "step": 1694
+    },
+    {
+      "epoch": 2.5412293853073464,
+      "grad_norm": 0.07901281731584704,
+      "learning_rate": 1.3925797299605647e-05,
+      "loss": 0.0027,
+      "step": 1695
+    },
+    {
+      "epoch": 2.542728635682159,
+      "grad_norm": 0.05405597758510519,
+      "learning_rate": 1.3837083955847418e-05,
+      "loss": 0.0033,
+      "step": 1696
+    },
+    {
+      "epoch": 2.5442278860569716,
+      "grad_norm": 0.05375559377296624,
+      "learning_rate": 1.3748633079274253e-05,
+      "loss": 0.0032,
+      "step": 1697
+    },
+    {
+      "epoch": 2.545727136431784,
+      "grad_norm": 0.05039178419782183,
+      "learning_rate": 1.3660444939322836e-05,
+      "loss": 0.0035,
+      "step": 1698
+    },
+    {
+      "epoch": 2.5472263868065967,
+      "grad_norm": 0.07167457099538735,
+      "learning_rate": 1.3572519804629536e-05,
+      "loss": 0.0056,
+      "step": 1699
+    },
+    {
+      "epoch": 2.548725637181409,
+      "grad_norm": 0.07372779209688211,
+      "learning_rate": 1.3484857943029572e-05,
+      "loss": 0.0083,
+      "step": 1700
+    },
+    {
+      "epoch": 2.550224887556222,
+      "grad_norm": 0.040476404349508645,
+      "learning_rate": 1.339745962155613e-05,
+      "loss": 0.0026,
+      "step": 1701
+    },
+    {
+      "epoch": 2.5517241379310347,
+      "grad_norm": 0.08291264682769349,
+      "learning_rate": 1.3310325106439726e-05,
+      "loss": 0.0032,
+      "step": 1702
+    },
+    {
+      "epoch": 2.553223388305847,
+      "grad_norm": 0.04934727510500086,
+      "learning_rate": 1.3223454663107172e-05,
+      "loss": 0.0039,
+      "step": 1703
+    },
+    {
+      "epoch": 2.5547226386806594,
+      "grad_norm": 0.03721964515275045,
+      "learning_rate": 1.3136848556180892e-05,
+      "loss": 0.0027,
+      "step": 1704
+    },
+    {
+      "epoch": 2.556221889055472,
+      "grad_norm": 0.06989750045733815,
+      "learning_rate": 1.30505070494781e-05,
+      "loss": 0.0098,
+      "step": 1705
+    },
+    {
+      "epoch": 2.557721139430285,
+      "grad_norm": 0.038253858696665755,
+      "learning_rate": 1.296443040601003e-05,
+      "loss": 0.0023,
+      "step": 1706
+    },
+    {
+      "epoch": 2.5592203898050974,
+      "grad_norm": 0.04073764123862176,
+      "learning_rate": 1.2878618887981064e-05,
+      "loss": 0.0017,
+      "step": 1707
+    },
+    {
+      "epoch": 2.56071964017991,
+      "grad_norm": 0.08121678653741697,
+      "learning_rate": 1.279307275678795e-05,
+      "loss": 0.0042,
+      "step": 1708
+    },
+    {
+      "epoch": 2.5622188905547225,
+      "grad_norm": 0.06896884513801788,
+      "learning_rate": 1.2707792273019048e-05,
+      "loss": 0.0041,
+      "step": 1709
+    },
+    {
+      "epoch": 2.5637181409295353,
+      "grad_norm": 0.07423572460387555,
+      "learning_rate": 1.262277769645348e-05,
+      "loss": 0.004,
+      "step": 1710
+    },
+    {
+      "epoch": 2.5652173913043477,
+      "grad_norm": 0.032450974009567375,
+      "learning_rate": 1.2538029286060426e-05,
+      "loss": 0.0018,
+      "step": 1711
+    },
+    {
+      "epoch": 2.5667166416791605,
+      "grad_norm": 0.04898128979522446,
+      "learning_rate": 1.2453547299998225e-05,
+      "loss": 0.004,
+      "step": 1712
+    },
+    {
+      "epoch": 2.5682158920539733,
+      "grad_norm": 0.043498324705939055,
+      "learning_rate": 1.2369331995613665e-05,
+      "loss": 0.0017,
+      "step": 1713
+    },
+    {
+      "epoch": 2.5697151424287856,
+      "grad_norm": 0.08309623767839033,
+      "learning_rate": 1.228538362944115e-05,
+      "loss": 0.0041,
+      "step": 1714
+    },
+    {
+      "epoch": 2.571214392803598,
+      "grad_norm": 0.1282714955750836,
+      "learning_rate": 1.2201702457201947e-05,
+      "loss": 0.0044,
+      "step": 1715
+    },
+    {
+      "epoch": 2.572713643178411,
+      "grad_norm": 0.03853257098388854,
+      "learning_rate": 1.2118288733803473e-05,
+      "loss": 0.0054,
+      "step": 1716
+    },
+    {
+      "epoch": 2.5742128935532236,
+      "grad_norm": 0.13020391770943424,
+      "learning_rate": 1.2035142713338366e-05,
+      "loss": 0.0115,
+      "step": 1717
+    },
+    {
+      "epoch": 2.575712143928036,
+      "grad_norm": 0.07576542607588402,
+      "learning_rate": 1.19522646490838e-05,
+      "loss": 0.0038,
+      "step": 1718
+    },
+    {
+      "epoch": 2.5772113943028487,
+      "grad_norm": 0.025413308174999216,
+      "learning_rate": 1.1869654793500784e-05,
+      "loss": 0.0035,
+      "step": 1719
+    },
+    {
+      "epoch": 2.578710644677661,
+      "grad_norm": 0.022381286165155924,
+      "learning_rate": 1.1787313398233235e-05,
+      "loss": 0.0008,
+      "step": 1720
+    },
+    {
+      "epoch": 2.580209895052474,
+      "grad_norm": 0.07623231370859024,
+      "learning_rate": 1.1705240714107302e-05,
+      "loss": 0.0041,
+      "step": 1721
+    },
+    {
+      "epoch": 2.5817091454272862,
+      "grad_norm": 0.0827073481443242,
+      "learning_rate": 1.1623436991130654e-05,
+      "loss": 0.0063,
+      "step": 1722
+    },
+    {
+      "epoch": 2.583208395802099,
+      "grad_norm": 0.042460507585494335,
+      "learning_rate": 1.1541902478491606e-05,
+      "loss": 0.002,
+      "step": 1723
+    },
+    {
+      "epoch": 2.5847076461769114,
+      "grad_norm": 0.04491600293936866,
+      "learning_rate": 1.1460637424558407e-05,
+      "loss": 0.0037,
+      "step": 1724
+    },
+    {
+      "epoch": 2.586206896551724,
+      "grad_norm": 0.06266018554204954,
+      "learning_rate": 1.1379642076878527e-05,
+      "loss": 0.0078,
+      "step": 1725
+    },
+    {
+      "epoch": 2.5877061469265366,
+      "grad_norm": 0.1519755677743987,
+      "learning_rate": 1.129891668217783e-05,
+      "loss": 0.0092,
+      "step": 1726
+    },
+    {
+      "epoch": 2.5892053973013494,
+      "grad_norm": 0.01931708277491169,
+      "learning_rate": 1.1218461486359877e-05,
+      "loss": 0.0011,
+      "step": 1727
+    },
+    {
+      "epoch": 2.590704647676162,
+      "grad_norm": 0.13331809055097277,
+      "learning_rate": 1.1138276734505104e-05,
+      "loss": 0.0084,
+      "step": 1728
+    },
+    {
+      "epoch": 2.5922038980509745,
+      "grad_norm": 0.08602979516663713,
+      "learning_rate": 1.1058362670870249e-05,
+      "loss": 0.0047,
+      "step": 1729
+    },
+    {
+      "epoch": 2.593703148425787,
+      "grad_norm": 0.20475185102113513,
+      "learning_rate": 1.0978719538887349e-05,
+      "loss": 0.005,
+      "step": 1730
+    },
+    {
+      "epoch": 2.5952023988005997,
+      "grad_norm": 0.05396284422100616,
+      "learning_rate": 1.0899347581163221e-05,
+      "loss": 0.0033,
+      "step": 1731
+    },
+    {
+      "epoch": 2.5967016491754125,
+      "grad_norm": 0.04632926541173506,
+      "learning_rate": 1.0820247039478604e-05,
+      "loss": 0.0031,
+      "step": 1732
+    },
+    {
+      "epoch": 2.598200899550225,
+      "grad_norm": 0.06330786884284814,
+      "learning_rate": 1.0741418154787442e-05,
+      "loss": 0.0056,
+      "step": 1733
+    },
+    {
+      "epoch": 2.5997001499250376,
+      "grad_norm": 0.049581430029072754,
+      "learning_rate": 1.0662861167216243e-05,
+      "loss": 0.0048,
+      "step": 1734
+    },
+    {
+      "epoch": 2.60119940029985,
+      "grad_norm": 0.014809634178959625,
+      "learning_rate": 1.0584576316063188e-05,
+      "loss": 0.0011,
+      "step": 1735
+    },
+    {
+      "epoch": 2.6026986506746628,
+      "grad_norm": 0.06952939363160453,
+      "learning_rate": 1.0506563839797501e-05,
+      "loss": 0.0046,
+      "step": 1736
+    },
+    {
+      "epoch": 2.604197901049475,
+      "grad_norm": 0.025305092673321244,
+      "learning_rate": 1.042882397605871e-05,
+      "loss": 0.0011,
+      "step": 1737
+    },
+    {
+      "epoch": 2.605697151424288,
+      "grad_norm": 0.06048091669387347,
+      "learning_rate": 1.0351356961655945e-05,
+      "loss": 0.0037,
+      "step": 1738
+    },
+    {
+      "epoch": 2.6071964017991007,
+      "grad_norm": 0.0928376653466004,
+      "learning_rate": 1.0274163032567163e-05,
+      "loss": 0.0038,
+      "step": 1739
+    },
+    {
+      "epoch": 2.608695652173913,
+      "grad_norm": 0.05539979196115197,
+      "learning_rate": 1.0197242423938446e-05,
+      "loss": 0.0076,
+      "step": 1740
+    },
+    {
+      "epoch": 2.6101949025487254,
+      "grad_norm": 0.07917966581807787,
+      "learning_rate": 1.0120595370083318e-05,
+      "loss": 0.0042,
+      "step": 1741
+    },
+    {
+      "epoch": 2.6116941529235382,
+      "grad_norm": 0.060022482722549786,
+      "learning_rate": 1.0044222104481971e-05,
+      "loss": 0.0031,
+      "step": 1742
+    },
+    {
+      "epoch": 2.613193403298351,
+      "grad_norm": 0.17585147443485435,
+      "learning_rate": 9.968122859780648e-06,
+      "loss": 0.0082,
+      "step": 1743
+    },
+    {
+      "epoch": 2.6146926536731634,
+      "grad_norm": 0.09870264943142629,
+      "learning_rate": 9.892297867790845e-06,
+      "loss": 0.0036,
+      "step": 1744
+    },
+    {
+      "epoch": 2.6161919040479757,
+      "grad_norm": 0.06095280132546412,
+      "learning_rate": 9.816747359488632e-06,
+      "loss": 0.002,
+      "step": 1745
+    },
+    {
+      "epoch": 2.6176911544227885,
+      "grad_norm": 0.10967913707388899,
+      "learning_rate": 9.74147156501396e-06,
+      "loss": 0.0035,
+      "step": 1746
+    },
+    {
+      "epoch": 2.6191904047976013,
+      "grad_norm": 0.33902675973415286,
+      "learning_rate": 9.666470713669918e-06,
+      "loss": 0.045,
+      "step": 1747
+    },
+    {
+      "epoch": 2.6206896551724137,
+      "grad_norm": 0.04253165972530954,
+      "learning_rate": 9.591745033922173e-06,
+      "loss": 0.0024,
+      "step": 1748
+    },
+    {
+      "epoch": 2.6221889055472265,
+      "grad_norm": 0.04818383699809805,
+      "learning_rate": 9.517294753398064e-06,
+      "loss": 0.0024,
+      "step": 1749
+    },
+    {
+      "epoch": 2.623688155922039,
+      "grad_norm": 0.03809410049607696,
+      "learning_rate": 9.443120098886061e-06,
+      "loss": 0.0032,
+      "step": 1750
+    },
+    {
+      "epoch": 2.6251874062968517,
+      "grad_norm": 0.05909250477315833,
+      "learning_rate": 9.369221296335006e-06,
+      "loss": 0.0034,
+      "step": 1751
+    },
+    {
+      "epoch": 2.626686656671664,
+      "grad_norm": 0.041243274161096855,
+      "learning_rate": 9.295598570853514e-06,
+      "loss": 0.004,
+      "step": 1752
+    },
+    {
+      "epoch": 2.628185907046477,
+      "grad_norm": 0.05659045575181391,
+      "learning_rate": 9.222252146709142e-06,
+      "loss": 0.0032,
+      "step": 1753
+    },
+    {
+      "epoch": 2.6296851574212896,
+      "grad_norm": 0.0812942025512981,
+      "learning_rate": 9.149182247327837e-06,
+      "loss": 0.0047,
+      "step": 1754
+    },
+    {
+      "epoch": 2.631184407796102,
+      "grad_norm": 0.0662188120787313,
+      "learning_rate": 9.076389095293148e-06,
+      "loss": 0.0058,
+      "step": 1755
+    },
+    {
+      "epoch": 2.6326836581709143,
+      "grad_norm": 0.09566246288727459,
+      "learning_rate": 9.00387291234569e-06,
+      "loss": 0.0051,
+      "step": 1756
+    },
+    {
+      "epoch": 2.634182908545727,
+      "grad_norm": 0.08790520476527616,
+      "learning_rate": 8.931633919382298e-06,
+      "loss": 0.0092,
+      "step": 1757
+    },
+    {
+      "epoch": 2.63568215892054,
+      "grad_norm": 0.09603411271828961,
+      "learning_rate": 8.85967233645547e-06,
+      "loss": 0.0035,
+      "step": 1758
+    },
+    {
+      "epoch": 2.6371814092953523,
+      "grad_norm": 0.07679943096437322,
+      "learning_rate": 8.787988382772705e-06,
+      "loss": 0.0052,
+      "step": 1759
+    },
+    {
+      "epoch": 2.638680659670165,
+      "grad_norm": 0.06355610491952349,
+      "learning_rate": 8.716582276695728e-06,
+      "loss": 0.0033,
+      "step": 1760
+    },
+    {
+      "epoch": 2.6401799100449774,
+      "grad_norm": 0.11913399959459053,
+      "learning_rate": 8.645454235739903e-06,
+      "loss": 0.0043,
+      "step": 1761
+    },
+    {
+      "epoch": 2.6416791604197902,
+      "grad_norm": 0.05791114392247628,
+      "learning_rate": 8.574604476573621e-06,
+      "loss": 0.0047,
+      "step": 1762
+    },
+    {
+      "epoch": 2.6431784107946026,
+      "grad_norm": 0.10461299194881798,
+      "learning_rate": 8.504033215017527e-06,
+      "loss": 0.0062,
+      "step": 1763
+    },
+    {
+      "epoch": 2.6446776611694154,
+      "grad_norm": 0.10373987138158351,
+      "learning_rate": 8.433740666043898e-06,
+      "loss": 0.0096,
+      "step": 1764
+    },
+    {
+      "epoch": 2.6461769115442277,
+      "grad_norm": 0.1713331811915574,
+      "learning_rate": 8.363727043776038e-06,
+      "loss": 0.004,
+      "step": 1765
+    },
+    {
+      "epoch": 2.6476761619190405,
+      "grad_norm": 0.06786862150547672,
+      "learning_rate": 8.293992561487596e-06,
+      "loss": 0.0044,
+      "step": 1766
+    },
+    {
+      "epoch": 2.649175412293853,
+      "grad_norm": 0.07532039055497501,
+      "learning_rate": 8.224537431601886e-06,
+      "loss": 0.0033,
+      "step": 1767
+    },
+    {
+      "epoch": 2.6506746626686657,
+      "grad_norm": 0.026496825220480674,
+      "learning_rate": 8.15536186569129e-06,
+      "loss": 0.0025,
+      "step": 1768
+    },
+    {
+      "epoch": 2.6521739130434785,
+      "grad_norm": 0.04795260350131189,
+      "learning_rate": 8.086466074476563e-06,
+      "loss": 0.0043,
+      "step": 1769
+    },
+    {
+      "epoch": 2.653673163418291,
+      "grad_norm": 0.04733553100664191,
+      "learning_rate": 8.017850267826232e-06,
+      "loss": 0.0024,
+      "step": 1770
+    },
+    {
+      "epoch": 2.655172413793103,
+      "grad_norm": 0.05951843530460022,
+      "learning_rate": 7.949514654755962e-06,
+      "loss": 0.0035,
+      "step": 1771
+    },
+    {
+      "epoch": 2.656671664167916,
+      "grad_norm": 0.04050730299729002,
+      "learning_rate": 7.881459443427886e-06,
+      "loss": 0.0037,
+      "step": 1772
+    },
+    {
+      "epoch": 2.658170914542729,
+      "grad_norm": 0.06302619474576679,
+      "learning_rate": 7.81368484114996e-06,
+      "loss": 0.0035,
+      "step": 1773
+    },
+    {
+      "epoch": 2.659670164917541,
+      "grad_norm": 0.03824673473843314,
+      "learning_rate": 7.746191054375362e-06,
+      "loss": 0.0037,
+      "step": 1774
+    },
+    {
+      "epoch": 2.661169415292354,
+      "grad_norm": 0.06707823730001641,
+      "learning_rate": 7.67897828870191e-06,
+      "loss": 0.007,
+      "step": 1775
+    },
+    {
+      "epoch": 2.6626686656671663,
+      "grad_norm": 0.07806204973737949,
+      "learning_rate": 7.612046748871327e-06,
+      "loss": 0.0056,
+      "step": 1776
+    },
+    {
+      "epoch": 2.664167916041979,
+      "grad_norm": 0.043370114908618695,
+      "learning_rate": 7.545396638768698e-06,
+      "loss": 0.0035,
+      "step": 1777
+    },
+    {
+      "epoch": 2.6656671664167915,
+      "grad_norm": 0.05273280042907484,
+      "learning_rate": 7.479028161421797e-06,
+      "loss": 0.0019,
+      "step": 1778
+    },
+    {
+      "epoch": 2.6671664167916043,
+      "grad_norm": 0.059897018536616276,
+      "learning_rate": 7.412941519000527e-06,
+      "loss": 0.003,
+      "step": 1779
+    },
+    {
+      "epoch": 2.668665667166417,
+      "grad_norm": 0.07985855599457413,
+      "learning_rate": 7.347136912816277e-06,
+      "loss": 0.0051,
+      "step": 1780
+    },
+    {
+      "epoch": 2.6701649175412294,
+      "grad_norm": 0.03581758546197187,
+      "learning_rate": 7.281614543321269e-06,
+      "loss": 0.0042,
+      "step": 1781
+    },
+    {
+      "epoch": 2.6716641679160418,
+      "grad_norm": 0.04288618656047732,
+      "learning_rate": 7.216374610108012e-06,
+      "loss": 0.0035,
+      "step": 1782
+    },
+    {
+      "epoch": 2.6731634182908546,
+      "grad_norm": 0.0663922133344574,
+      "learning_rate": 7.151417311908648e-06,
+      "loss": 0.0029,
+      "step": 1783
+    },
+    {
+      "epoch": 2.6746626686656674,
+      "grad_norm": 0.1918430476451125,
+      "learning_rate": 7.086742846594385e-06,
+      "loss": 0.0135,
+      "step": 1784
+    },
+    {
+      "epoch": 2.6761619190404797,
+      "grad_norm": 0.04484775879018914,
+      "learning_rate": 7.022351411174866e-06,
+      "loss": 0.0037,
+      "step": 1785
+    },
+    {
+      "epoch": 2.677661169415292,
+      "grad_norm": 0.062248407404835907,
+      "learning_rate": 6.958243201797554e-06,
+      "loss": 0.004,
+      "step": 1786
+    },
+    {
+      "epoch": 2.679160419790105,
+      "grad_norm": 0.10446631857945263,
+      "learning_rate": 6.894418413747183e-06,
+      "loss": 0.0045,
+      "step": 1787
+    },
+    {
+      "epoch": 2.6806596701649177,
+      "grad_norm": 0.03529468151449674,
+      "learning_rate": 6.830877241445111e-06,
+      "loss": 0.0021,
+      "step": 1788
+    },
+    {
+      "epoch": 2.68215892053973,
+      "grad_norm": 0.03798589090052098,
+      "learning_rate": 6.767619878448783e-06,
+      "loss": 0.0026,
+      "step": 1789
+    },
+    {
+      "epoch": 2.683658170914543,
+      "grad_norm": 0.08669293863873424,
+      "learning_rate": 6.704646517451107e-06,
+      "loss": 0.0056,
+      "step": 1790
+    },
+    {
+      "epoch": 2.685157421289355,
+      "grad_norm": 0.046908300538182957,
+      "learning_rate": 6.6419573502798374e-06,
+      "loss": 0.0025,
+      "step": 1791
+    },
+    {
+      "epoch": 2.686656671664168,
+      "grad_norm": 0.03637779784955495,
+      "learning_rate": 6.579552567897051e-06,
+      "loss": 0.0016,
+      "step": 1792
+    },
+    {
+      "epoch": 2.6881559220389803,
+      "grad_norm": 0.04687808627810223,
+      "learning_rate": 6.517432360398556e-06,
+      "loss": 0.0053,
+      "step": 1793
+    },
+    {
+      "epoch": 2.689655172413793,
+      "grad_norm": 0.037364738444089864,
+      "learning_rate": 6.455596917013273e-06,
+      "loss": 0.0014,
+      "step": 1794
+    },
+    {
+      "epoch": 2.691154422788606,
+      "grad_norm": 0.1021041733000301,
+      "learning_rate": 6.394046426102674e-06,
+      "loss": 0.0083,
+      "step": 1795
+    },
+    {
+      "epoch": 2.6926536731634183,
+      "grad_norm": 0.018456788471741113,
+      "learning_rate": 6.332781075160243e-06,
+      "loss": 0.0009,
+      "step": 1796
+    },
+    {
+      "epoch": 2.6941529235382307,
+      "grad_norm": 0.14926815053881523,
+      "learning_rate": 6.2718010508108545e-06,
+      "loss": 0.0102,
+      "step": 1797
+    },
+    {
+      "epoch": 2.6956521739130435,
+      "grad_norm": 0.04352809658591788,
+      "learning_rate": 6.21110653881023e-06,
+      "loss": 0.0028,
+      "step": 1798
+    },
+    {
+      "epoch": 2.6971514242878563,
+      "grad_norm": 0.05279283988150011,
+      "learning_rate": 6.1506977240444074e-06,
+      "loss": 0.0017,
+      "step": 1799
+    },
+    {
+      "epoch": 2.6986506746626686,
+      "grad_norm": 0.019605598342381503,
+      "learning_rate": 6.090574790529091e-06,
+      "loss": 0.0006,
+      "step": 1800
+    },
+    {
+      "epoch": 2.7001499250374814,
+      "grad_norm": 0.0777744775067246,
+      "learning_rate": 6.030737921409169e-06,
+      "loss": 0.0029,
+      "step": 1801
+    },
+    {
+      "epoch": 2.7016491754122938,
+      "grad_norm": 0.05414552236249822,
+      "learning_rate": 5.971187298958103e-06,
+      "loss": 0.0056,
+      "step": 1802
+    },
+    {
+      "epoch": 2.7031484257871066,
+      "grad_norm": 0.04962408711837054,
+      "learning_rate": 5.911923104577455e-06,
+      "loss": 0.0039,
+      "step": 1803
+    },
+    {
+      "epoch": 2.704647676161919,
+      "grad_norm": 0.0696464944244991,
+      "learning_rate": 5.852945518796205e-06,
+      "loss": 0.0079,
+      "step": 1804
+    },
+    {
+      "epoch": 2.7061469265367317,
+      "grad_norm": 0.03714539340218805,
+      "learning_rate": 5.7942547212703315e-06,
+      "loss": 0.0019,
+      "step": 1805
+    },
+    {
+      "epoch": 2.707646176911544,
+      "grad_norm": 0.05223972716082169,
+      "learning_rate": 5.735850890782157e-06,
+      "loss": 0.0054,
+      "step": 1806
+    },
+    {
+      "epoch": 2.709145427286357,
+      "grad_norm": 0.07430860604160193,
+      "learning_rate": 5.6777342052399045e-06,
+      "loss": 0.0035,
+      "step": 1807
+    },
+    {
+      "epoch": 2.7106446776611692,
+      "grad_norm": 0.060445175053381824,
+      "learning_rate": 5.619904841677059e-06,
+      "loss": 0.0023,
+      "step": 1808
+    },
+    {
+      "epoch": 2.712143928035982,
+      "grad_norm": 0.05925366881856817,
+      "learning_rate": 5.562362976251901e-06,
+      "loss": 0.0047,
+      "step": 1809
+    },
+    {
+      "epoch": 2.713643178410795,
+      "grad_norm": 0.0516545752906272,
+      "learning_rate": 5.505108784246926e-06,
+      "loss": 0.0022,
+      "step": 1810
+    },
+    {
+      "epoch": 2.715142428785607,
+      "grad_norm": 0.10602896675857203,
+      "learning_rate": 5.448142440068316e-06,
+      "loss": 0.0042,
+      "step": 1811
+    },
+    {
+      "epoch": 2.7166416791604195,
+      "grad_norm": 0.32243857508421625,
+      "learning_rate": 5.39146411724547e-06,
+      "loss": 0.0072,
+      "step": 1812
+    },
+    {
+      "epoch": 2.7181409295352323,
+      "grad_norm": 0.05465551687606462,
+      "learning_rate": 5.335073988430372e-06,
+      "loss": 0.0036,
+      "step": 1813
+    },
+    {
+      "epoch": 2.719640179910045,
+      "grad_norm": 0.08368985646892235,
+      "learning_rate": 5.278972225397127e-06,
+      "loss": 0.0119,
+      "step": 1814
+    },
+    {
+      "epoch": 2.7211394302848575,
+      "grad_norm": 0.051321818493470575,
+      "learning_rate": 5.223158999041444e-06,
+      "loss": 0.0048,
+      "step": 1815
+    },
+    {
+      "epoch": 2.7226386806596703,
+      "grad_norm": 0.04408943737843986,
+      "learning_rate": 5.167634479380068e-06,
+      "loss": 0.0024,
+      "step": 1816
+    },
+    {
+      "epoch": 2.7241379310344827,
+      "grad_norm": 0.1202534156928614,
+      "learning_rate": 5.1123988355503475e-06,
+      "loss": 0.0078,
+      "step": 1817
+    },
+    {
+      "epoch": 2.7256371814092955,
+      "grad_norm": 0.052112172067874475,
+      "learning_rate": 5.057452235809624e-06,
+      "loss": 0.0058,
+      "step": 1818
+    },
+    {
+      "epoch": 2.727136431784108,
+      "grad_norm": 0.1431485997481884,
+      "learning_rate": 5.002794847534764e-06,
+      "loss": 0.0097,
+      "step": 1819
+    },
+    {
+      "epoch": 2.7286356821589206,
+      "grad_norm": 0.051277101130307555,
+      "learning_rate": 4.948426837221631e-06,
+      "loss": 0.0055,
+      "step": 1820
+    },
+    {
+      "epoch": 2.7301349325337334,
+      "grad_norm": 0.053350618119373285,
+      "learning_rate": 4.8943483704846475e-06,
+      "loss": 0.0038,
+      "step": 1821
+    },
+    {
+      "epoch": 2.7316341829085458,
+      "grad_norm": 0.030903608998343296,
+      "learning_rate": 4.840559612056183e-06,
+      "loss": 0.0014,
+      "step": 1822
+    },
+    {
+      "epoch": 2.733133433283358,
+      "grad_norm": 0.047055897536594594,
+      "learning_rate": 4.7870607257861415e-06,
+      "loss": 0.0051,
+      "step": 1823
+    },
+    {
+      "epoch": 2.734632683658171,
+      "grad_norm": 0.02818509921299283,
+      "learning_rate": 4.733851874641382e-06,
+      "loss": 0.0008,
+      "step": 1824
+    },
+    {
+      "epoch": 2.7361319340329837,
+      "grad_norm": 0.04570475809580388,
+      "learning_rate": 4.680933220705308e-06,
+      "loss": 0.0018,
+      "step": 1825
+    },
+    {
+      "epoch": 2.737631184407796,
+      "grad_norm": 0.0538111788348984,
+      "learning_rate": 4.628304925177318e-06,
+      "loss": 0.0037,
+      "step": 1826
+    },
+    {
+      "epoch": 2.7391304347826084,
+      "grad_norm": 0.13116424213383152,
+      "learning_rate": 4.575967148372317e-06,
+      "loss": 0.0136,
+      "step": 1827
+    },
+    {
+      "epoch": 2.7406296851574212,
+      "grad_norm": 0.04885765875397696,
+      "learning_rate": 4.523920049720265e-06,
+      "loss": 0.0025,
+      "step": 1828
+    },
+    {
+      "epoch": 2.742128935532234,
+      "grad_norm": 0.10154857801433739,
+      "learning_rate": 4.4721637877656375e-06,
+      "loss": 0.0066,
+      "step": 1829
+    },
+    {
+      "epoch": 2.7436281859070464,
+      "grad_norm": 0.047254869802745965,
+      "learning_rate": 4.420698520166988e-06,
+      "loss": 0.0034,
+      "step": 1830
+    },
+    {
+      "epoch": 2.745127436281859,
+      "grad_norm": 0.08641710628596849,
+      "learning_rate": 4.369524403696457e-06,
+      "loss": 0.0084,
+      "step": 1831
+    },
+    {
+      "epoch": 2.7466266866566715,
+      "grad_norm": 0.08831269449702958,
+      "learning_rate": 4.318641594239259e-06,
+      "loss": 0.0067,
+      "step": 1832
+    },
+    {
+      "epoch": 2.7481259370314843,
+      "grad_norm": 0.07518037530585348,
+      "learning_rate": 4.268050246793276e-06,
+      "loss": 0.0067,
+      "step": 1833
+    },
+    {
+      "epoch": 2.7496251874062967,
+      "grad_norm": 0.08886601880182082,
+      "learning_rate": 4.217750515468522e-06,
+      "loss": 0.0056,
+      "step": 1834
+    },
+    {
+      "epoch": 2.7511244377811095,
+      "grad_norm": 0.047343986827075496,
+      "learning_rate": 4.167742553486675e-06,
+      "loss": 0.0037,
+      "step": 1835
+    },
+    {
+      "epoch": 2.7526236881559223,
+      "grad_norm": 0.061256674999085256,
+      "learning_rate": 4.118026513180695e-06,
+      "loss": 0.0032,
+      "step": 1836
+    },
+    {
+      "epoch": 2.7541229385307346,
+      "grad_norm": 0.09671352746387149,
+      "learning_rate": 4.068602545994249e-06,
+      "loss": 0.0105,
+      "step": 1837
+    },
+    {
+      "epoch": 2.755622188905547,
+      "grad_norm": 0.12504578715825013,
+      "learning_rate": 4.019470802481307e-06,
+      "loss": 0.012,
+      "step": 1838
+    },
+    {
+      "epoch": 2.75712143928036,
+      "grad_norm": 0.05061259533408433,
+      "learning_rate": 3.970631432305694e-06,
+      "loss": 0.004,
+      "step": 1839
+    },
+    {
+      "epoch": 2.7586206896551726,
+      "grad_norm": 0.029406937261118465,
+      "learning_rate": 3.922084584240582e-06,
+      "loss": 0.002,
+      "step": 1840
+    },
+    {
+      "epoch": 2.760119940029985,
+      "grad_norm": 0.02532563135738694,
+      "learning_rate": 3.873830406168111e-06,
+      "loss": 0.002,
+      "step": 1841
+    },
+    {
+      "epoch": 2.7616191904047978,
+      "grad_norm": 0.10231518549891974,
+      "learning_rate": 3.825869045078867e-06,
+      "loss": 0.0048,
+      "step": 1842
+    },
+    {
+      "epoch": 2.76311844077961,
+      "grad_norm": 0.04405091545119458,
+      "learning_rate": 3.7782006470714616e-06,
+      "loss": 0.0034,
+      "step": 1843
+    },
+    {
+      "epoch": 2.764617691154423,
+      "grad_norm": 0.04214896707196821,
+      "learning_rate": 3.730825357352119e-06,
+      "loss": 0.0016,
+      "step": 1844
+    },
+    {
+      "epoch": 2.7661169415292353,
+      "grad_norm": 0.04381641722204364,
+      "learning_rate": 3.68374332023419e-06,
+      "loss": 0.0015,
+      "step": 1845
+    },
+    {
+      "epoch": 2.767616191904048,
+      "grad_norm": 0.1714819540968367,
+      "learning_rate": 3.6369546791377052e-06,
+      "loss": 0.0053,
+      "step": 1846
+    },
+    {
+      "epoch": 2.7691154422788604,
+      "grad_norm": 0.030777510297242894,
+      "learning_rate": 3.5904595765890005e-06,
+      "loss": 0.0012,
+      "step": 1847
+    },
+    {
+      "epoch": 2.770614692653673,
+      "grad_norm": 0.06278492467131958,
+      "learning_rate": 3.5442581542201923e-06,
+      "loss": 0.0038,
+      "step": 1848
+    },
+    {
+      "epoch": 2.7721139430284856,
+      "grad_norm": 0.04525524756217157,
+      "learning_rate": 3.4983505527688586e-06,
+      "loss": 0.0032,
+      "step": 1849
+    },
+    {
+      "epoch": 2.7736131934032984,
+      "grad_norm": 0.040158798347074746,
+      "learning_rate": 3.452736912077503e-06,
+      "loss": 0.0022,
+      "step": 1850
+    },
+    {
+      "epoch": 2.775112443778111,
+      "grad_norm": 0.05351156811338634,
+      "learning_rate": 3.40741737109318e-06,
+      "loss": 0.0022,
+      "step": 1851
+    },
+    {
+      "epoch": 2.7766116941529235,
+      "grad_norm": 0.08841883689534029,
+      "learning_rate": 3.3623920678670597e-06,
+      "loss": 0.003,
+      "step": 1852
+    },
+    {
+      "epoch": 2.778110944527736,
+      "grad_norm": 0.07130261183515092,
+      "learning_rate": 3.3176611395540626e-06,
+      "loss": 0.0056,
+      "step": 1853
+    },
+    {
+      "epoch": 2.7796101949025487,
+      "grad_norm": 0.09438468992037789,
+      "learning_rate": 3.273224722412327e-06,
+      "loss": 0.0046,
+      "step": 1854
+    },
+    {
+      "epoch": 2.7811094452773615,
+      "grad_norm": 0.06383128845407085,
+      "learning_rate": 3.2290829518028862e-06,
+      "loss": 0.0022,
+      "step": 1855
+    },
+    {
+      "epoch": 2.782608695652174,
+      "grad_norm": 0.07593396064006204,
+      "learning_rate": 3.1852359621892367e-06,
+      "loss": 0.0034,
+      "step": 1856
+    },
+    {
+      "epoch": 2.7841079460269866,
+      "grad_norm": 0.04086990619865844,
+      "learning_rate": 3.1416838871368924e-06,
+      "loss": 0.0023,
+      "step": 1857
+    },
+    {
+      "epoch": 2.785607196401799,
+      "grad_norm": 0.10911709604280903,
+      "learning_rate": 3.0984268593130528e-06,
+      "loss": 0.008,
+      "step": 1858
+    },
+    {
+      "epoch": 2.787106446776612,
+      "grad_norm": 0.06393379283366997,
+      "learning_rate": 3.0554650104861136e-06,
+      "loss": 0.002,
+      "step": 1859
+    },
+    {
+      "epoch": 2.788605697151424,
+      "grad_norm": 0.08071123917158006,
+      "learning_rate": 3.012798471525324e-06,
+      "loss": 0.0043,
+      "step": 1860
+    },
+    {
+      "epoch": 2.790104947526237,
+      "grad_norm": 0.03866299410166067,
+      "learning_rate": 2.970427372400353e-06,
+      "loss": 0.0023,
+      "step": 1861
+    },
+    {
+      "epoch": 2.7916041979010497,
+      "grad_norm": 0.12924488448937485,
+      "learning_rate": 2.928351842180921e-06,
+      "loss": 0.0083,
+      "step": 1862
+    },
+    {
+      "epoch": 2.793103448275862,
+      "grad_norm": 0.044056078109747175,
+      "learning_rate": 2.8865720090364034e-06,
+      "loss": 0.002,
+      "step": 1863
+    },
+    {
+      "epoch": 2.7946026986506745,
+      "grad_norm": 0.0932750403527346,
+      "learning_rate": 2.845088000235396e-06,
+      "loss": 0.0074,
+      "step": 1864
+    },
+    {
+      "epoch": 2.7961019490254873,
+      "grad_norm": 0.21546183050930776,
+      "learning_rate": 2.8038999421453826e-06,
+      "loss": 0.0067,
+      "step": 1865
+    },
+    {
+      "epoch": 2.7976011994003,
+      "grad_norm": 0.04083158743885419,
+      "learning_rate": 2.7630079602323442e-06,
+      "loss": 0.0032,
+      "step": 1866
+    },
+    {
+      "epoch": 2.7991004497751124,
+      "grad_norm": 0.026451998528825394,
+      "learning_rate": 2.7224121790603517e-06,
+      "loss": 0.0014,
+      "step": 1867
+    },
+    {
+      "epoch": 2.8005997001499248,
+      "grad_norm": 0.026543988690067504,
+      "learning_rate": 2.6821127222911857e-06,
+      "loss": 0.0012,
+      "step": 1868
+    },
+    {
+      "epoch": 2.8020989505247376,
+      "grad_norm": 0.11502358287074994,
+      "learning_rate": 2.6421097126839712e-06,
+      "loss": 0.005,
+      "step": 1869
+    },
+    {
+      "epoch": 2.8035982008995504,
+      "grad_norm": 0.06937276307112113,
+      "learning_rate": 2.6024032720948443e-06,
+      "loss": 0.0035,
+      "step": 1870
+    },
+    {
+      "epoch": 2.8050974512743627,
+      "grad_norm": 0.04889190826037233,
+      "learning_rate": 2.5629935214764865e-06,
+      "loss": 0.0053,
+      "step": 1871
+    },
+    {
+      "epoch": 2.8065967016491755,
+      "grad_norm": 0.01964804857509374,
+      "learning_rate": 2.5238805808778242e-06,
+      "loss": 0.0011,
+      "step": 1872
+    },
+    {
+      "epoch": 2.808095952023988,
+      "grad_norm": 0.08944021499669409,
+      "learning_rate": 2.4850645694436736e-06,
+      "loss": 0.0043,
+      "step": 1873
+    },
+    {
+      "epoch": 2.8095952023988007,
+      "grad_norm": 0.09138194545450047,
+      "learning_rate": 2.446545605414341e-06,
+      "loss": 0.0188,
+      "step": 1874
+    },
+    {
+      "epoch": 2.811094452773613,
+      "grad_norm": 0.06082928618151328,
+      "learning_rate": 2.4083238061252567e-06,
+      "loss": 0.0034,
+      "step": 1875
+    },
+    {
+      "epoch": 2.812593703148426,
+      "grad_norm": 0.0518451376073689,
+      "learning_rate": 2.3703992880066638e-06,
+      "loss": 0.0025,
+      "step": 1876
+    },
+    {
+      "epoch": 2.8140929535232386,
+      "grad_norm": 0.08943695133377194,
+      "learning_rate": 2.332772166583208e-06,
+      "loss": 0.0025,
+      "step": 1877
+    },
+    {
+      "epoch": 2.815592203898051,
+      "grad_norm": 0.054040088180756354,
+      "learning_rate": 2.295442556473637e-06,
+      "loss": 0.0051,
+      "step": 1878
+    },
+    {
+      "epoch": 2.8170914542728633,
+      "grad_norm": 0.07247814988589195,
+      "learning_rate": 2.2584105713904125e-06,
+      "loss": 0.0068,
+      "step": 1879
+    },
+    {
+      "epoch": 2.818590704647676,
+      "grad_norm": 0.06849834319985831,
+      "learning_rate": 2.2216763241393767e-06,
+      "loss": 0.0036,
+      "step": 1880
+    },
+    {
+      "epoch": 2.820089955022489,
+      "grad_norm": 0.07362282563904983,
+      "learning_rate": 2.1852399266194314e-06,
+      "loss": 0.0073,
+      "step": 1881
+    },
+    {
+      "epoch": 2.8215892053973013,
+      "grad_norm": 0.06944920144902611,
+      "learning_rate": 2.1491014898221582e-06,
+      "loss": 0.0024,
+      "step": 1882
+    },
+    {
+      "epoch": 2.823088455772114,
+      "grad_norm": 0.04248072490573655,
+      "learning_rate": 2.1132611238315003e-06,
+      "loss": 0.0035,
+      "step": 1883
+    },
+    {
+      "epoch": 2.8245877061469264,
+      "grad_norm": 0.0817432276248758,
+      "learning_rate": 2.0777189378234143e-06,
+      "loss": 0.0017,
+      "step": 1884
+    },
+    {
+      "epoch": 2.8260869565217392,
+      "grad_norm": 0.07240802576223083,
+      "learning_rate": 2.0424750400655947e-06,
+      "loss": 0.0098,
+      "step": 1885
+    },
+    {
+      "epoch": 2.8275862068965516,
+      "grad_norm": 0.0385971978938863,
+      "learning_rate": 2.0075295379170412e-06,
+      "loss": 0.0033,
+      "step": 1886
+    },
+    {
+      "epoch": 2.8290854572713644,
+      "grad_norm": 0.053910895946008223,
+      "learning_rate": 1.9728825378278246e-06,
+      "loss": 0.0032,
+      "step": 1887
+    },
+    {
+      "epoch": 2.8305847076461768,
+      "grad_norm": 0.09526927822276723,
+      "learning_rate": 1.9385341453386995e-06,
+      "loss": 0.0048,
+      "step": 1888
+    },
+    {
+      "epoch": 2.8320839580209896,
+      "grad_norm": 0.06975698097659441,
+      "learning_rate": 1.904484465080847e-06,
+      "loss": 0.0082,
+      "step": 1889
+    },
+    {
+      "epoch": 2.833583208395802,
+      "grad_norm": 0.17772085747797686,
+      "learning_rate": 1.8707336007754873e-06,
+      "loss": 0.0086,
+      "step": 1890
+    },
+    {
+      "epoch": 2.8350824587706147,
+      "grad_norm": 0.084062140467934,
+      "learning_rate": 1.8372816552336026e-06,
+      "loss": 0.0035,
+      "step": 1891
+    },
+    {
+      "epoch": 2.8365817091454275,
+      "grad_norm": 0.031985655169939474,
+      "learning_rate": 1.8041287303556364e-06,
+      "loss": 0.002,
+      "step": 1892
+    },
+    {
+      "epoch": 2.83808095952024,
+      "grad_norm": 0.048848294887181724,
+      "learning_rate": 1.771274927131139e-06,
+      "loss": 0.0044,
+      "step": 1893
+    },
+    {
+      "epoch": 2.839580209895052,
+      "grad_norm": 0.160521076075769,
+      "learning_rate": 1.7387203456384782e-06,
+      "loss": 0.0063,
+      "step": 1894
+    },
+    {
+      "epoch": 2.841079460269865,
+      "grad_norm": 0.0284880955110527,
+      "learning_rate": 1.706465085044584e-06,
+      "loss": 0.0011,
+      "step": 1895
+    },
+    {
+      "epoch": 2.842578710644678,
+      "grad_norm": 0.04788444160974722,
+      "learning_rate": 1.6745092436045494e-06,
+      "loss": 0.0035,
+      "step": 1896
+    },
+    {
+      "epoch": 2.84407796101949,
+      "grad_norm": 0.017049447472496787,
+      "learning_rate": 1.6428529186614195e-06,
+      "loss": 0.0005,
+      "step": 1897
+    },
+    {
+      "epoch": 2.845577211394303,
+      "grad_norm": 0.05683597326229523,
+      "learning_rate": 1.611496206645835e-06,
+      "loss": 0.0035,
+      "step": 1898
+    },
+    {
+      "epoch": 2.8470764617691153,
+      "grad_norm": 0.08486491218156729,
+      "learning_rate": 1.580439203075812e-06,
+      "loss": 0.0025,
+      "step": 1899
+    },
+    {
+      "epoch": 2.848575712143928,
+      "grad_norm": 0.03387217796464905,
+      "learning_rate": 1.5496820025563409e-06,
+      "loss": 0.0013,
+      "step": 1900
+    },
+    {
+      "epoch": 2.8500749625187405,
+      "grad_norm": 0.04816811922646314,
+      "learning_rate": 1.5192246987791981e-06,
+      "loss": 0.004,
+      "step": 1901
+    },
+    {
+      "epoch": 2.8515742128935533,
+      "grad_norm": 0.07910064579652644,
+      "learning_rate": 1.4890673845226133e-06,
+      "loss": 0.0039,
+      "step": 1902
+    },
+    {
+      "epoch": 2.853073463268366,
+      "grad_norm": 0.032934113339092096,
+      "learning_rate": 1.4592101516509914e-06,
+      "loss": 0.0016,
+      "step": 1903
+    },
+    {
+      "epoch": 2.8545727136431784,
+      "grad_norm": 0.0449861007395873,
+      "learning_rate": 1.4296530911146466e-06,
+      "loss": 0.0033,
+      "step": 1904
+    },
+    {
+      "epoch": 2.856071964017991,
+      "grad_norm": 0.0526419044525919,
+      "learning_rate": 1.400396292949513e-06,
+      "loss": 0.0025,
+      "step": 1905
+    },
+    {
+      "epoch": 2.8575712143928036,
+      "grad_norm": 0.06178628171991531,
+      "learning_rate": 1.3714398462768563e-06,
+      "loss": 0.0026,
+      "step": 1906
+    },
+    {
+      "epoch": 2.8590704647676164,
+      "grad_norm": 0.10844669704867088,
+      "learning_rate": 1.3427838393030633e-06,
+      "loss": 0.0093,
+      "step": 1907
+    },
+    {
+      "epoch": 2.8605697151424287,
+      "grad_norm": 0.08539087115108733,
+      "learning_rate": 1.3144283593192752e-06,
+      "loss": 0.0072,
+      "step": 1908
+    },
+    {
+      "epoch": 2.862068965517241,
+      "grad_norm": 0.09803222726799544,
+      "learning_rate": 1.2863734927012095e-06,
+      "loss": 0.0103,
+      "step": 1909
+    },
+    {
+      "epoch": 2.863568215892054,
+      "grad_norm": 0.05577463106669858,
+      "learning_rate": 1.2586193249088608e-06,
+      "loss": 0.0028,
+      "step": 1910
+    },
+    {
+      "epoch": 2.8650674662668667,
+      "grad_norm": 0.2114427332970624,
+      "learning_rate": 1.231165940486234e-06,
+      "loss": 0.0313,
+      "step": 1911
+    },
+    {
+      "epoch": 2.866566716641679,
+      "grad_norm": 0.14063204905909968,
+      "learning_rate": 1.20401342306109e-06,
+      "loss": 0.006,
+      "step": 1912
+    },
+    {
+      "epoch": 2.868065967016492,
+      "grad_norm": 0.08238020394722131,
+      "learning_rate": 1.1771618553447216e-06,
+      "loss": 0.0076,
+      "step": 1913
+    },
+    {
+      "epoch": 2.869565217391304,
+      "grad_norm": 0.02690643600372609,
+      "learning_rate": 1.1506113191316447e-06,
+      "loss": 0.0012,
+      "step": 1914
+    },
+    {
+      "epoch": 2.871064467766117,
+      "grad_norm": 0.08444372271644372,
+      "learning_rate": 1.1243618952994195e-06,
+      "loss": 0.0052,
+      "step": 1915
+    },
+    {
+      "epoch": 2.8725637181409294,
+      "grad_norm": 0.042791866109907366,
+      "learning_rate": 1.0984136638083177e-06,
+      "loss": 0.0018,
+      "step": 1916
+    },
+    {
+      "epoch": 2.874062968515742,
+      "grad_norm": 0.0822631678321723,
+      "learning_rate": 1.0727667037011668e-06,
+      "loss": 0.007,
+      "step": 1917
+    },
+    {
+      "epoch": 2.875562218890555,
+      "grad_norm": 0.06524093103957149,
+      "learning_rate": 1.0474210931030514e-06,
+      "loss": 0.0059,
+      "step": 1918
+    },
+    {
+      "epoch": 2.8770614692653673,
+      "grad_norm": 0.03448809415597592,
+      "learning_rate": 1.0223769092211012e-06,
+      "loss": 0.0026,
+      "step": 1919
+    },
+    {
+      "epoch": 2.8785607196401797,
+      "grad_norm": 0.046141321115156876,
+      "learning_rate": 9.976342283442463e-07,
+      "loss": 0.0022,
+      "step": 1920
+    },
+    {
+      "epoch": 2.8800599700149925,
+      "grad_norm": 0.07243493966503523,
+      "learning_rate": 9.731931258429638e-07,
+      "loss": 0.0024,
+      "step": 1921
+    },
+    {
+      "epoch": 2.8815592203898053,
+      "grad_norm": 0.056477163143498946,
+      "learning_rate": 9.490536761691204e-07,
+      "loss": 0.0033,
+      "step": 1922
+    },
+    {
+      "epoch": 2.8830584707646176,
+      "grad_norm": 0.04470477657428264,
+      "learning_rate": 9.252159528556403e-07,
+      "loss": 0.0024,
+      "step": 1923
+    },
+    {
+      "epoch": 2.8845577211394304,
+      "grad_norm": 0.027848123681753996,
+      "learning_rate": 9.016800285163718e-07,
+      "loss": 0.0024,
+      "step": 1924
+    },
+    {
+      "epoch": 2.886056971514243,
+      "grad_norm": 0.05260745768799866,
+      "learning_rate": 8.784459748458318e-07,
+      "loss": 0.0046,
+      "step": 1925
+    },
+    {
+      "epoch": 2.8875562218890556,
+      "grad_norm": 0.04282082202332062,
+      "learning_rate": 8.555138626189618e-07,
+      "loss": 0.0017,
+      "step": 1926
+    },
+    {
+      "epoch": 2.889055472263868,
+      "grad_norm": 0.02613070330147928,
+      "learning_rate": 8.328837616909613e-07,
+      "loss": 0.0019,
+      "step": 1927
+    },
+    {
+      "epoch": 2.8905547226386807,
+      "grad_norm": 0.10700496393861944,
+      "learning_rate": 8.105557409970432e-07,
+      "loss": 0.009,
+      "step": 1928
+    },
+    {
+      "epoch": 2.892053973013493,
+      "grad_norm": 0.0802537427420226,
+      "learning_rate": 7.885298685522235e-07,
+      "loss": 0.0038,
+      "step": 1929
+    },
+    {
+      "epoch": 2.893553223388306,
+      "grad_norm": 0.0505236671377372,
+      "learning_rate": 7.668062114511321e-07,
+      "loss": 0.0019,
+      "step": 1930
+    },
+    {
+      "epoch": 2.8950524737631183,
+      "grad_norm": 0.033520163658288964,
+      "learning_rate": 7.453848358678017e-07,
+      "loss": 0.0017,
+      "step": 1931
+    },
+    {
+      "epoch": 2.896551724137931,
+      "grad_norm": 0.07259240298956116,
+      "learning_rate": 7.242658070554464e-07,
+      "loss": 0.0062,
+      "step": 1932
+    },
+    {
+      "epoch": 2.898050974512744,
+      "grad_norm": 0.06383788331881589,
+      "learning_rate": 7.034491893463058e-07,
+      "loss": 0.0036,
+      "step": 1933
+    },
+    {
+      "epoch": 2.899550224887556,
+      "grad_norm": 0.05929371421132908,
+      "learning_rate": 6.829350461514006e-07,
+      "loss": 0.0038,
+      "step": 1934
+    },
+    {
+      "epoch": 2.9010494752623686,
+      "grad_norm": 0.10791565065380813,
+      "learning_rate": 6.627234399603555e-07,
+      "loss": 0.0076,
+      "step": 1935
+    },
+    {
+      "epoch": 2.9025487256371814,
+      "grad_norm": 0.03199906382024088,
+      "learning_rate": 6.428144323412544e-07,
+      "loss": 0.002,
+      "step": 1936
+    },
+    {
+      "epoch": 2.904047976011994,
+      "grad_norm": 0.044793332598965306,
+      "learning_rate": 6.232080839403631e-07,
+      "loss": 0.0031,
+      "step": 1937
+    },
+    {
+      "epoch": 2.9055472263868065,
+      "grad_norm": 0.034949113643794064,
+      "learning_rate": 6.039044544820404e-07,
+      "loss": 0.0017,
+      "step": 1938
+    },
+    {
+      "epoch": 2.9070464767616193,
+      "grad_norm": 0.061169476787731235,
+      "learning_rate": 5.849036027684606e-07,
+      "loss": 0.003,
+      "step": 1939
+    },
+    {
+      "epoch": 2.9085457271364317,
+      "grad_norm": 0.09671098321705973,
+      "learning_rate": 5.662055866795357e-07,
+      "loss": 0.0048,
+      "step": 1940
+    },
+    {
+      "epoch": 2.9100449775112445,
+      "grad_norm": 0.09483559527940648,
+      "learning_rate": 5.478104631726711e-07,
+      "loss": 0.0027,
+      "step": 1941
+    },
+    {
+      "epoch": 2.911544227886057,
+      "grad_norm": 0.03107820245195978,
+      "learning_rate": 5.297182882825879e-07,
+      "loss": 0.0015,
+      "step": 1942
+    },
+    {
+      "epoch": 2.9130434782608696,
+      "grad_norm": 0.07421783376655451,
+      "learning_rate": 5.119291171211793e-07,
+      "loss": 0.0094,
+      "step": 1943
+    },
+    {
+      "epoch": 2.9145427286356824,
+      "grad_norm": 0.09218285026776223,
+      "learning_rate": 4.944430038773762e-07,
+      "loss": 0.0085,
+      "step": 1944
+    },
+    {
+      "epoch": 2.9160419790104948,
+      "grad_norm": 0.07360416997364519,
+      "learning_rate": 4.772600018168816e-07,
+      "loss": 0.0033,
+      "step": 1945
+    },
+    {
+      "epoch": 2.917541229385307,
+      "grad_norm": 0.043634796881273585,
+      "learning_rate": 4.6038016328211476e-07,
+      "loss": 0.0034,
+      "step": 1946
+    },
+    {
+      "epoch": 2.91904047976012,
+      "grad_norm": 0.07685943518323182,
+      "learning_rate": 4.438035396920004e-07,
+      "loss": 0.005,
+      "step": 1947
+    },
+    {
+      "epoch": 2.9205397301349327,
+      "grad_norm": 0.07806672517785551,
+      "learning_rate": 4.275301815417909e-07,
+      "loss": 0.002,
+      "step": 1948
+    },
+    {
+      "epoch": 2.922038980509745,
+      "grad_norm": 0.04375018889513186,
+      "learning_rate": 4.115601384029666e-07,
+      "loss": 0.0045,
+      "step": 1949
+    },
+    {
+      "epoch": 2.923538230884558,
+      "grad_norm": 0.1430345786843898,
+      "learning_rate": 3.958934589230467e-07,
+      "loss": 0.0085,
+      "step": 1950
+    },
+    {
+      "epoch": 2.9250374812593702,
+      "grad_norm": 0.10516629111264399,
+      "learning_rate": 3.805301908254455e-07,
+      "loss": 0.0124,
+      "step": 1951
+    },
+    {
+      "epoch": 2.926536731634183,
+      "grad_norm": 0.04196769665260615,
+      "learning_rate": 3.654703809093607e-07,
+      "loss": 0.0032,
+      "step": 1952
+    },
+    {
+      "epoch": 2.9280359820089954,
+      "grad_norm": 0.12229056029404549,
+      "learning_rate": 3.50714075049563e-07,
+      "loss": 0.0044,
+      "step": 1953
+    },
+    {
+      "epoch": 2.929535232383808,
+      "grad_norm": 0.05573736381133186,
+      "learning_rate": 3.362613181963403e-07,
+      "loss": 0.0029,
+      "step": 1954
+    },
+    {
+      "epoch": 2.9310344827586206,
+      "grad_norm": 0.040779137055409795,
+      "learning_rate": 3.2211215437528694e-07,
+      "loss": 0.0021,
+      "step": 1955
+    },
+    {
+      "epoch": 2.9325337331334334,
+      "grad_norm": 0.053397472075800284,
+      "learning_rate": 3.0826662668720364e-07,
+      "loss": 0.0069,
+      "step": 1956
+    },
+    {
+      "epoch": 2.9340329835082457,
+      "grad_norm": 0.07113095453573431,
+      "learning_rate": 2.947247773079753e-07,
+      "loss": 0.0076,
+      "step": 1957
+    },
+    {
+      "epoch": 2.9355322338830585,
+      "grad_norm": 0.04310065622239369,
+      "learning_rate": 2.81486647488427e-07,
+      "loss": 0.0038,
+      "step": 1958
+    },
+    {
+      "epoch": 2.9370314842578713,
+      "grad_norm": 0.12131577298406716,
+      "learning_rate": 2.685522775541904e-07,
+      "loss": 0.0058,
+      "step": 1959
+    },
+    {
+      "epoch": 2.9385307346326837,
+      "grad_norm": 0.06607798391376911,
+      "learning_rate": 2.5592170690560414e-07,
+      "loss": 0.0033,
+      "step": 1960
+    },
+    {
+      "epoch": 2.940029985007496,
+      "grad_norm": 0.07686025540345201,
+      "learning_rate": 2.4359497401758024e-07,
+      "loss": 0.0067,
+      "step": 1961
+    },
+    {
+      "epoch": 2.941529235382309,
+      "grad_norm": 0.12480781751450243,
+      "learning_rate": 2.315721164394713e-07,
+      "loss": 0.0092,
+      "step": 1962
+    },
+    {
+      "epoch": 2.9430284857571216,
+      "grad_norm": 0.06710923415617372,
+      "learning_rate": 2.1985317079500356e-07,
+      "loss": 0.0039,
+      "step": 1963
+    },
+    {
+      "epoch": 2.944527736131934,
+      "grad_norm": 0.05328518709101643,
+      "learning_rate": 2.0843817278209942e-07,
+      "loss": 0.0041,
+      "step": 1964
+    },
+    {
+      "epoch": 2.9460269865067468,
+      "grad_norm": 0.03205720007508785,
+      "learning_rate": 1.973271571728441e-07,
+      "loss": 0.0021,
+      "step": 1965
+    },
+    {
+      "epoch": 2.947526236881559,
+      "grad_norm": 0.08713150735817307,
+      "learning_rate": 1.86520157813308e-07,
+      "loss": 0.005,
+      "step": 1966
+    },
+    {
+      "epoch": 2.949025487256372,
+      "grad_norm": 0.0897607603221516,
+      "learning_rate": 1.7601720762346897e-07,
+      "loss": 0.0109,
+      "step": 1967
+    },
+    {
+      "epoch": 2.9505247376311843,
+      "grad_norm": 0.09456767421388704,
+      "learning_rate": 1.6581833859716788e-07,
+      "loss": 0.0025,
+      "step": 1968
+    },
+    {
+      "epoch": 2.952023988005997,
+      "grad_norm": 0.02654611858782301,
+      "learning_rate": 1.5592358180189782e-07,
+      "loss": 0.0011,
+      "step": 1969
+    },
+    {
+      "epoch": 2.9535232383808094,
+      "grad_norm": 0.057986054790013425,
+      "learning_rate": 1.4633296737882607e-07,
+      "loss": 0.0051,
+      "step": 1970
+    },
+    {
+      "epoch": 2.9550224887556222,
+      "grad_norm": 0.05325409118977268,
+      "learning_rate": 1.3704652454261668e-07,
+      "loss": 0.0026,
+      "step": 1971
+    },
+    {
+      "epoch": 2.9565217391304346,
+      "grad_norm": 0.05995378335721647,
+      "learning_rate": 1.2806428158138596e-07,
+      "loss": 0.0039,
+      "step": 1972
+    },
+    {
+      "epoch": 2.9580209895052474,
+      "grad_norm": 0.11321346048305181,
+      "learning_rate": 1.193862658566025e-07,
+      "loss": 0.0046,
+      "step": 1973
+    },
+    {
+      "epoch": 2.95952023988006,
+      "grad_norm": 0.07230973808230436,
+      "learning_rate": 1.1101250380300965e-07,
+      "loss": 0.0055,
+      "step": 1974
+    },
+    {
+      "epoch": 2.9610194902548725,
+      "grad_norm": 0.08029645889349672,
+      "learning_rate": 1.0294302092853647e-07,
+      "loss": 0.0035,
+      "step": 1975
+    },
+    {
+      "epoch": 2.962518740629685,
+      "grad_norm": 0.041158986213349215,
+      "learning_rate": 9.517784181422019e-08,
+      "loss": 0.0047,
+      "step": 1976
+    },
+    {
+      "epoch": 2.9640179910044977,
+      "grad_norm": 0.05686175756545626,
+      "learning_rate": 8.771699011416168e-08,
+      "loss": 0.0038,
+      "step": 1977
+    },
+    {
+      "epoch": 2.9655172413793105,
+      "grad_norm": 0.04893183911640975,
+      "learning_rate": 8.056048855540343e-08,
+      "loss": 0.0346,
+      "step": 1978
+    },
+    {
+      "epoch": 2.967016491754123,
+      "grad_norm": 0.1379441273961598,
+      "learning_rate": 7.370835893788508e-08,
+      "loss": 0.0091,
+      "step": 1979
+    },
+    {
+      "epoch": 2.9685157421289357,
+      "grad_norm": 0.06399841406714543,
+      "learning_rate": 6.71606221343768e-08,
+      "loss": 0.0043,
+      "step": 1980
+    },
+    {
+      "epoch": 2.970014992503748,
+      "grad_norm": 0.11199174671151793,
+      "learning_rate": 6.09172980904238e-08,
+      "loss": 0.0122,
+      "step": 1981
+    },
+    {
+      "epoch": 2.971514242878561,
+      "grad_norm": 0.04482360637940532,
+      "learning_rate": 5.497840582429081e-08,
+      "loss": 0.0034,
+      "step": 1982
+    },
+    {
+      "epoch": 2.973013493253373,
+      "grad_norm": 0.17153408725973948,
+      "learning_rate": 4.934396342684e-08,
+      "loss": 0.0052,
+      "step": 1983
+    },
+    {
+      "epoch": 2.974512743628186,
+      "grad_norm": 0.04980670314050741,
+      "learning_rate": 4.401398806159751e-08,
+      "loss": 0.0029,
+      "step": 1984
+    },
+    {
+      "epoch": 2.9760119940029988,
+      "grad_norm": 0.08613536118025303,
+      "learning_rate": 3.898849596456478e-08,
+      "loss": 0.0064,
+      "step": 1985
+    },
+    {
+      "epoch": 2.977511244377811,
+      "grad_norm": 0.03045405966760291,
+      "learning_rate": 3.4267502444274015e-08,
+      "loss": 0.0016,
+      "step": 1986
+    },
+    {
+      "epoch": 2.9790104947526235,
+      "grad_norm": 0.036884946821497074,
+      "learning_rate": 2.985102188168831e-08,
+      "loss": 0.0019,
+      "step": 1987
+    },
+    {
+      "epoch": 2.9805097451274363,
+      "grad_norm": 0.0524860686083844,
+      "learning_rate": 2.573906773016832e-08,
+      "loss": 0.0034,
+      "step": 1988
+    },
+    {
+      "epoch": 2.982008995502249,
+      "grad_norm": 0.06829182858407432,
+      "learning_rate": 2.193165251545004e-08,
+      "loss": 0.0077,
+      "step": 1989
+    },
+    {
+      "epoch": 2.9835082458770614,
+      "grad_norm": 0.05804353038563701,
+      "learning_rate": 1.842878783557822e-08,
+      "loss": 0.0024,
+      "step": 1990
+    },
+    {
+      "epoch": 2.9850074962518742,
+      "grad_norm": 0.023737341328751025,
+      "learning_rate": 1.5230484360873044e-08,
+      "loss": 0.0012,
+      "step": 1991
+    },
+    {
+      "epoch": 2.9865067466266866,
+      "grad_norm": 0.04638140957693581,
+      "learning_rate": 1.2336751833941229e-08,
+      "loss": 0.0019,
+      "step": 1992
+    },
+    {
+      "epoch": 2.9880059970014994,
+      "grad_norm": 0.10493790831990461,
+      "learning_rate": 9.747599069576119e-09,
+      "loss": 0.008,
+      "step": 1993
+    },
+    {
+      "epoch": 2.9895052473763117,
+      "grad_norm": 0.06846796686090073,
+      "learning_rate": 7.463033954802078e-09,
+      "loss": 0.0064,
+      "step": 1994
+    },
+    {
+      "epoch": 2.9910044977511245,
+      "grad_norm": 0.11136491254887901,
+      "learning_rate": 5.483063448785686e-09,
+      "loss": 0.0059,
+      "step": 1995
+    },
+    {
+      "epoch": 2.992503748125937,
+      "grad_norm": 0.10745241960262616,
+      "learning_rate": 3.807693582869032e-09,
+      "loss": 0.0033,
+      "step": 1996
+    },
+    {
+      "epoch": 2.9940029985007497,
+      "grad_norm": 0.08617751521895838,
+      "learning_rate": 2.4369294605253166e-09,
+      "loss": 0.0041,
+      "step": 1997
+    },
+    {
+      "epoch": 2.995502248875562,
+      "grad_norm": 0.11414137883273419,
+      "learning_rate": 1.3707752573255405e-09,
+      "loss": 0.0066,
+      "step": 1998
+    },
+    {
+      "epoch": 2.997001499250375,
+      "grad_norm": 0.07137774757587402,
+      "learning_rate": 6.092342209607083e-10,
+      "loss": 0.0047,
+      "step": 1999
+    },
+    {
+      "epoch": 2.9985007496251876,
+      "grad_norm": 0.04455773434098678,
+      "learning_rate": 1.5230867123072756e-10,
+      "loss": 0.0032,
+      "step": 2000
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.05174225537101476,
+      "learning_rate": 0.0,
+      "loss": 0.0019,
+      "step": 2001
+    },
+    {
+      "epoch": 3.0,
+      "eval_loss": 0.03369831293821335,
+      "eval_runtime": 481.6669,
+      "eval_samples_per_second": 21.542,
+      "eval_steps_per_second": 1.347,
+      "step": 2001
+    },
+    {
+      "epoch": 3.0,
+      "step": 2001,
+      "total_flos": 2.2064687673350554e+17,
+      "train_loss": 0.016641332297180983,
+      "train_runtime": 13115.3958,
+      "train_samples_per_second": 4.882,
+      "train_steps_per_second": 0.153
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2001,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "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": 2.2064687673350554e+17,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}