diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,44373 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9987570287067182,
+  "eval_steps": 500,
+  "global_step": 6333,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00047351287363125186,
+      "grad_norm": 32.22987365722656,
+      "learning_rate": 1.0526315789473685e-07,
+      "loss": 4.8137,
+      "step": 1
+    },
+    {
+      "epoch": 0.0009470257472625037,
+      "grad_norm": 31.049623489379883,
+      "learning_rate": 2.105263157894737e-07,
+      "loss": 4.6949,
+      "step": 2
+    },
+    {
+      "epoch": 0.0014205386208937555,
+      "grad_norm": 30.38652801513672,
+      "learning_rate": 3.1578947368421055e-07,
+      "loss": 4.7183,
+      "step": 3
+    },
+    {
+      "epoch": 0.0018940514945250074,
+      "grad_norm": 30.45220184326172,
+      "learning_rate": 4.210526315789474e-07,
+      "loss": 4.6404,
+      "step": 4
+    },
+    {
+      "epoch": 0.0023675643681562593,
+      "grad_norm": 29.60318946838379,
+      "learning_rate": 5.263157894736843e-07,
+      "loss": 4.6377,
+      "step": 5
+    },
+    {
+      "epoch": 0.002841077241787511,
+      "grad_norm": 30.485383987426758,
+      "learning_rate": 6.315789473684211e-07,
+      "loss": 4.7493,
+      "step": 6
+    },
+    {
+      "epoch": 0.003314590115418763,
+      "grad_norm": 29.985885620117188,
+      "learning_rate": 7.368421052631579e-07,
+      "loss": 4.6343,
+      "step": 7
+    },
+    {
+      "epoch": 0.003788102989050015,
+      "grad_norm": 29.9451904296875,
+      "learning_rate": 8.421052631578948e-07,
+      "loss": 4.6325,
+      "step": 8
+    },
+    {
+      "epoch": 0.0042616158626812665,
+      "grad_norm": 28.334569931030273,
+      "learning_rate": 9.473684210526317e-07,
+      "loss": 4.4123,
+      "step": 9
+    },
+    {
+      "epoch": 0.004735128736312519,
+      "grad_norm": 30.083621978759766,
+      "learning_rate": 1.0526315789473685e-06,
+      "loss": 4.4909,
+      "step": 10
+    },
+    {
+      "epoch": 0.00520864160994377,
+      "grad_norm": 29.243478775024414,
+      "learning_rate": 1.1578947368421053e-06,
+      "loss": 4.4436,
+      "step": 11
+    },
+    {
+      "epoch": 0.005682154483575022,
+      "grad_norm": 27.903270721435547,
+      "learning_rate": 1.2631578947368422e-06,
+      "loss": 4.4001,
+      "step": 12
+    },
+    {
+      "epoch": 0.006155667357206274,
+      "grad_norm": 28.32964515686035,
+      "learning_rate": 1.3684210526315791e-06,
+      "loss": 4.2836,
+      "step": 13
+    },
+    {
+      "epoch": 0.006629180230837526,
+      "grad_norm": 28.63888931274414,
+      "learning_rate": 1.4736842105263159e-06,
+      "loss": 4.2648,
+      "step": 14
+    },
+    {
+      "epoch": 0.007102693104468778,
+      "grad_norm": 27.617847442626953,
+      "learning_rate": 1.5789473684210526e-06,
+      "loss": 4.1683,
+      "step": 15
+    },
+    {
+      "epoch": 0.00757620597810003,
+      "grad_norm": 27.17360496520996,
+      "learning_rate": 1.6842105263157895e-06,
+      "loss": 4.1395,
+      "step": 16
+    },
+    {
+      "epoch": 0.008049718851731282,
+      "grad_norm": 27.059186935424805,
+      "learning_rate": 1.7894736842105265e-06,
+      "loss": 3.9835,
+      "step": 17
+    },
+    {
+      "epoch": 0.008523231725362533,
+      "grad_norm": 27.937824249267578,
+      "learning_rate": 1.8947368421052634e-06,
+      "loss": 3.9632,
+      "step": 18
+    },
+    {
+      "epoch": 0.008996744598993784,
+      "grad_norm": 27.05862045288086,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 3.8811,
+      "step": 19
+    },
+    {
+      "epoch": 0.009470257472625037,
+      "grad_norm": 26.196979522705078,
+      "learning_rate": 2.105263157894737e-06,
+      "loss": 3.8312,
+      "step": 20
+    },
+    {
+      "epoch": 0.009943770346256289,
+      "grad_norm": 26.135305404663086,
+      "learning_rate": 2.2105263157894738e-06,
+      "loss": 3.7232,
+      "step": 21
+    },
+    {
+      "epoch": 0.01041728321988754,
+      "grad_norm": 25.997011184692383,
+      "learning_rate": 2.3157894736842105e-06,
+      "loss": 3.6288,
+      "step": 22
+    },
+    {
+      "epoch": 0.010890796093518793,
+      "grad_norm": 26.397438049316406,
+      "learning_rate": 2.4210526315789477e-06,
+      "loss": 3.5437,
+      "step": 23
+    },
+    {
+      "epoch": 0.011364308967150044,
+      "grad_norm": 24.695541381835938,
+      "learning_rate": 2.5263157894736844e-06,
+      "loss": 3.4521,
+      "step": 24
+    },
+    {
+      "epoch": 0.011837821840781295,
+      "grad_norm": 24.346065521240234,
+      "learning_rate": 2.631578947368421e-06,
+      "loss": 3.269,
+      "step": 25
+    },
+    {
+      "epoch": 0.012311334714412548,
+      "grad_norm": 24.483036041259766,
+      "learning_rate": 2.7368421052631583e-06,
+      "loss": 3.2028,
+      "step": 26
+    },
+    {
+      "epoch": 0.0127848475880438,
+      "grad_norm": 24.411731719970703,
+      "learning_rate": 2.842105263157895e-06,
+      "loss": 3.1983,
+      "step": 27
+    },
+    {
+      "epoch": 0.013258360461675053,
+      "grad_norm": 24.398752212524414,
+      "learning_rate": 2.9473684210526317e-06,
+      "loss": 3.0967,
+      "step": 28
+    },
+    {
+      "epoch": 0.013731873335306304,
+      "grad_norm": 24.130815505981445,
+      "learning_rate": 3.052631578947369e-06,
+      "loss": 2.9756,
+      "step": 29
+    },
+    {
+      "epoch": 0.014205386208937555,
+      "grad_norm": 22.618980407714844,
+      "learning_rate": 3.157894736842105e-06,
+      "loss": 2.7865,
+      "step": 30
+    },
+    {
+      "epoch": 0.014678899082568808,
+      "grad_norm": 22.387496948242188,
+      "learning_rate": 3.2631578947368423e-06,
+      "loss": 2.7595,
+      "step": 31
+    },
+    {
+      "epoch": 0.01515241195620006,
+      "grad_norm": 21.85611915588379,
+      "learning_rate": 3.368421052631579e-06,
+      "loss": 2.6221,
+      "step": 32
+    },
+    {
+      "epoch": 0.01562592482983131,
+      "grad_norm": 21.454317092895508,
+      "learning_rate": 3.473684210526316e-06,
+      "loss": 2.4785,
+      "step": 33
+    },
+    {
+      "epoch": 0.016099437703462564,
+      "grad_norm": 20.223222732543945,
+      "learning_rate": 3.578947368421053e-06,
+      "loss": 2.4204,
+      "step": 34
+    },
+    {
+      "epoch": 0.016572950577093813,
+      "grad_norm": 20.16985511779785,
+      "learning_rate": 3.6842105263157896e-06,
+      "loss": 2.4438,
+      "step": 35
+    },
+    {
+      "epoch": 0.017046463450725066,
+      "grad_norm": 18.412492752075195,
+      "learning_rate": 3.789473684210527e-06,
+      "loss": 2.2141,
+      "step": 36
+    },
+    {
+      "epoch": 0.01751997632435632,
+      "grad_norm": 17.698265075683594,
+      "learning_rate": 3.894736842105263e-06,
+      "loss": 2.1291,
+      "step": 37
+    },
+    {
+      "epoch": 0.01799348919798757,
+      "grad_norm": 15.409119606018066,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 2.0709,
+      "step": 38
+    },
+    {
+      "epoch": 0.01846700207161882,
+      "grad_norm": 14.957233428955078,
+      "learning_rate": 4.105263157894737e-06,
+      "loss": 1.9438,
+      "step": 39
+    },
+    {
+      "epoch": 0.018940514945250075,
+      "grad_norm": 14.200984001159668,
+      "learning_rate": 4.210526315789474e-06,
+      "loss": 1.8377,
+      "step": 40
+    },
+    {
+      "epoch": 0.019414027818881324,
+      "grad_norm": 11.768697738647461,
+      "learning_rate": 4.315789473684211e-06,
+      "loss": 1.8675,
+      "step": 41
+    },
+    {
+      "epoch": 0.019887540692512577,
+      "grad_norm": 11.194828987121582,
+      "learning_rate": 4.4210526315789476e-06,
+      "loss": 1.7639,
+      "step": 42
+    },
+    {
+      "epoch": 0.02036105356614383,
+      "grad_norm": 10.777433395385742,
+      "learning_rate": 4.526315789473685e-06,
+      "loss": 1.6869,
+      "step": 43
+    },
+    {
+      "epoch": 0.02083456643977508,
+      "grad_norm": 9.706337928771973,
+      "learning_rate": 4.631578947368421e-06,
+      "loss": 1.7424,
+      "step": 44
+    },
+    {
+      "epoch": 0.021308079313406333,
+      "grad_norm": 9.26910400390625,
+      "learning_rate": 4.736842105263158e-06,
+      "loss": 1.5467,
+      "step": 45
+    },
+    {
+      "epoch": 0.021781592187037586,
+      "grad_norm": 8.306986808776855,
+      "learning_rate": 4.842105263157895e-06,
+      "loss": 1.5561,
+      "step": 46
+    },
+    {
+      "epoch": 0.022255105060668835,
+      "grad_norm": 7.666769027709961,
+      "learning_rate": 4.947368421052632e-06,
+      "loss": 1.4493,
+      "step": 47
+    },
+    {
+      "epoch": 0.022728617934300088,
+      "grad_norm": 7.483455181121826,
+      "learning_rate": 5.052631578947369e-06,
+      "loss": 1.383,
+      "step": 48
+    },
+    {
+      "epoch": 0.02320213080793134,
+      "grad_norm": 7.80124044418335,
+      "learning_rate": 5.157894736842106e-06,
+      "loss": 1.4092,
+      "step": 49
+    },
+    {
+      "epoch": 0.02367564368156259,
+      "grad_norm": 7.231191158294678,
+      "learning_rate": 5.263157894736842e-06,
+      "loss": 1.3909,
+      "step": 50
+    },
+    {
+      "epoch": 0.024149156555193844,
+      "grad_norm": 6.258241176605225,
+      "learning_rate": 5.36842105263158e-06,
+      "loss": 1.3075,
+      "step": 51
+    },
+    {
+      "epoch": 0.024622669428825097,
+      "grad_norm": 6.165030479431152,
+      "learning_rate": 5.4736842105263165e-06,
+      "loss": 1.2565,
+      "step": 52
+    },
+    {
+      "epoch": 0.02509618230245635,
+      "grad_norm": 6.1708455085754395,
+      "learning_rate": 5.578947368421052e-06,
+      "loss": 1.3199,
+      "step": 53
+    },
+    {
+      "epoch": 0.0255696951760876,
+      "grad_norm": 5.459502696990967,
+      "learning_rate": 5.68421052631579e-06,
+      "loss": 1.1997,
+      "step": 54
+    },
+    {
+      "epoch": 0.026043208049718852,
+      "grad_norm": 5.5988640785217285,
+      "learning_rate": 5.789473684210527e-06,
+      "loss": 1.0775,
+      "step": 55
+    },
+    {
+      "epoch": 0.026516720923350105,
+      "grad_norm": 4.691328048706055,
+      "learning_rate": 5.8947368421052634e-06,
+      "loss": 1.185,
+      "step": 56
+    },
+    {
+      "epoch": 0.026990233796981355,
+      "grad_norm": 4.2112507820129395,
+      "learning_rate": 6e-06,
+      "loss": 1.099,
+      "step": 57
+    },
+    {
+      "epoch": 0.027463746670612608,
+      "grad_norm": 4.1589674949646,
+      "learning_rate": 6.105263157894738e-06,
+      "loss": 1.1201,
+      "step": 58
+    },
+    {
+      "epoch": 0.02793725954424386,
+      "grad_norm": 4.95751428604126,
+      "learning_rate": 6.2105263157894745e-06,
+      "loss": 1.1349,
+      "step": 59
+    },
+    {
+      "epoch": 0.02841077241787511,
+      "grad_norm": 3.543083906173706,
+      "learning_rate": 6.31578947368421e-06,
+      "loss": 1.0243,
+      "step": 60
+    },
+    {
+      "epoch": 0.028884285291506363,
+      "grad_norm": 2.737982749938965,
+      "learning_rate": 6.421052631578948e-06,
+      "loss": 0.9596,
+      "step": 61
+    },
+    {
+      "epoch": 0.029357798165137616,
+      "grad_norm": 3.6427838802337646,
+      "learning_rate": 6.526315789473685e-06,
+      "loss": 0.9429,
+      "step": 62
+    },
+    {
+      "epoch": 0.029831311038768866,
+      "grad_norm": 3.1644389629364014,
+      "learning_rate": 6.631578947368421e-06,
+      "loss": 1.0354,
+      "step": 63
+    },
+    {
+      "epoch": 0.03030482391240012,
+      "grad_norm": 2.886333465576172,
+      "learning_rate": 6.736842105263158e-06,
+      "loss": 0.9896,
+      "step": 64
+    },
+    {
+      "epoch": 0.030778336786031372,
+      "grad_norm": 2.797492265701294,
+      "learning_rate": 6.842105263157896e-06,
+      "loss": 0.9985,
+      "step": 65
+    },
+    {
+      "epoch": 0.03125184965966262,
+      "grad_norm": 2.594764471054077,
+      "learning_rate": 6.947368421052632e-06,
+      "loss": 0.9611,
+      "step": 66
+    },
+    {
+      "epoch": 0.031725362533293874,
+      "grad_norm": 3.788191795349121,
+      "learning_rate": 7.052631578947369e-06,
+      "loss": 0.9094,
+      "step": 67
+    },
+    {
+      "epoch": 0.03219887540692513,
+      "grad_norm": 2.9197521209716797,
+      "learning_rate": 7.157894736842106e-06,
+      "loss": 0.8584,
+      "step": 68
+    },
+    {
+      "epoch": 0.03267238828055638,
+      "grad_norm": 3.339698314666748,
+      "learning_rate": 7.263157894736843e-06,
+      "loss": 0.8872,
+      "step": 69
+    },
+    {
+      "epoch": 0.033145901154187626,
+      "grad_norm": 3.415419578552246,
+      "learning_rate": 7.368421052631579e-06,
+      "loss": 0.8177,
+      "step": 70
+    },
+    {
+      "epoch": 0.03361941402781888,
+      "grad_norm": 2.3660242557525635,
+      "learning_rate": 7.473684210526316e-06,
+      "loss": 0.9576,
+      "step": 71
+    },
+    {
+      "epoch": 0.03409292690145013,
+      "grad_norm": 2.605135202407837,
+      "learning_rate": 7.578947368421054e-06,
+      "loss": 0.8901,
+      "step": 72
+    },
+    {
+      "epoch": 0.034566439775081385,
+      "grad_norm": 3.861926317214966,
+      "learning_rate": 7.68421052631579e-06,
+      "loss": 0.8733,
+      "step": 73
+    },
+    {
+      "epoch": 0.03503995264871264,
+      "grad_norm": 3.650552272796631,
+      "learning_rate": 7.789473684210526e-06,
+      "loss": 0.9273,
+      "step": 74
+    },
+    {
+      "epoch": 0.03551346552234389,
+      "grad_norm": 3.098768949508667,
+      "learning_rate": 7.894736842105265e-06,
+      "loss": 0.9519,
+      "step": 75
+    },
+    {
+      "epoch": 0.03598697839597514,
+      "grad_norm": 3.1536178588867188,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.8491,
+      "step": 76
+    },
+    {
+      "epoch": 0.03646049126960639,
+      "grad_norm": 1.9899249076843262,
+      "learning_rate": 8.105263157894736e-06,
+      "loss": 0.8351,
+      "step": 77
+    },
+    {
+      "epoch": 0.03693400414323764,
+      "grad_norm": 2.1144354343414307,
+      "learning_rate": 8.210526315789475e-06,
+      "loss": 0.8503,
+      "step": 78
+    },
+    {
+      "epoch": 0.037407517016868896,
+      "grad_norm": 2.7250900268554688,
+      "learning_rate": 8.315789473684212e-06,
+      "loss": 0.8294,
+      "step": 79
+    },
+    {
+      "epoch": 0.03788102989050015,
+      "grad_norm": 2.3137624263763428,
+      "learning_rate": 8.421052631578948e-06,
+      "loss": 0.78,
+      "step": 80
+    },
+    {
+      "epoch": 0.0383545427641314,
+      "grad_norm": 2.3112668991088867,
+      "learning_rate": 8.526315789473685e-06,
+      "loss": 0.7696,
+      "step": 81
+    },
+    {
+      "epoch": 0.03882805563776265,
+      "grad_norm": 2.882868766784668,
+      "learning_rate": 8.631578947368422e-06,
+      "loss": 0.829,
+      "step": 82
+    },
+    {
+      "epoch": 0.0393015685113939,
+      "grad_norm": 2.7895877361297607,
+      "learning_rate": 8.736842105263158e-06,
+      "loss": 0.8318,
+      "step": 83
+    },
+    {
+      "epoch": 0.039775081385025154,
+      "grad_norm": 2.179732322692871,
+      "learning_rate": 8.842105263157895e-06,
+      "loss": 0.7615,
+      "step": 84
+    },
+    {
+      "epoch": 0.04024859425865641,
+      "grad_norm": 2.3356051445007324,
+      "learning_rate": 8.947368421052632e-06,
+      "loss": 0.6886,
+      "step": 85
+    },
+    {
+      "epoch": 0.04072210713228766,
+      "grad_norm": 2.664323568344116,
+      "learning_rate": 9.05263157894737e-06,
+      "loss": 0.7625,
+      "step": 86
+    },
+    {
+      "epoch": 0.04119562000591891,
+      "grad_norm": 2.517026424407959,
+      "learning_rate": 9.157894736842105e-06,
+      "loss": 0.6589,
+      "step": 87
+    },
+    {
+      "epoch": 0.04166913287955016,
+      "grad_norm": 1.9179812669754028,
+      "learning_rate": 9.263157894736842e-06,
+      "loss": 0.695,
+      "step": 88
+    },
+    {
+      "epoch": 0.04214264575318141,
+      "grad_norm": 2.2245266437530518,
+      "learning_rate": 9.36842105263158e-06,
+      "loss": 0.6799,
+      "step": 89
+    },
+    {
+      "epoch": 0.042616158626812665,
+      "grad_norm": 2.003324270248413,
+      "learning_rate": 9.473684210526315e-06,
+      "loss": 0.6828,
+      "step": 90
+    },
+    {
+      "epoch": 0.04308967150044392,
+      "grad_norm": 2.43106746673584,
+      "learning_rate": 9.578947368421054e-06,
+      "loss": 0.6584,
+      "step": 91
+    },
+    {
+      "epoch": 0.04356318437407517,
+      "grad_norm": 2.4158437252044678,
+      "learning_rate": 9.68421052631579e-06,
+      "loss": 0.6645,
+      "step": 92
+    },
+    {
+      "epoch": 0.044036697247706424,
+      "grad_norm": 2.5698916912078857,
+      "learning_rate": 9.789473684210527e-06,
+      "loss": 0.6451,
+      "step": 93
+    },
+    {
+      "epoch": 0.04451021012133767,
+      "grad_norm": 1.9317587614059448,
+      "learning_rate": 9.894736842105264e-06,
+      "loss": 0.6401,
+      "step": 94
+    },
+    {
+      "epoch": 0.044983722994968924,
+      "grad_norm": 1.8760066032409668,
+      "learning_rate": 1e-05,
+      "loss": 0.6576,
+      "step": 95
+    },
+    {
+      "epoch": 0.045457235868600177,
+      "grad_norm": 2.3672170639038086,
+      "learning_rate": 1.0105263157894738e-05,
+      "loss": 0.7306,
+      "step": 96
+    },
+    {
+      "epoch": 0.04593074874223143,
+      "grad_norm": 1.790613055229187,
+      "learning_rate": 1.0210526315789476e-05,
+      "loss": 0.6165,
+      "step": 97
+    },
+    {
+      "epoch": 0.04640426161586268,
+      "grad_norm": 1.7723586559295654,
+      "learning_rate": 1.0315789473684213e-05,
+      "loss": 0.5753,
+      "step": 98
+    },
+    {
+      "epoch": 0.046877774489493935,
+      "grad_norm": 2.3334202766418457,
+      "learning_rate": 1.0421052631578948e-05,
+      "loss": 0.6944,
+      "step": 99
+    },
+    {
+      "epoch": 0.04735128736312518,
+      "grad_norm": 1.6184226274490356,
+      "learning_rate": 1.0526315789473684e-05,
+      "loss": 0.5287,
+      "step": 100
+    },
+    {
+      "epoch": 0.047824800236756435,
+      "grad_norm": 1.6825703382492065,
+      "learning_rate": 1.0631578947368421e-05,
+      "loss": 0.5426,
+      "step": 101
+    },
+    {
+      "epoch": 0.04829831311038769,
+      "grad_norm": 1.4492383003234863,
+      "learning_rate": 1.073684210526316e-05,
+      "loss": 0.53,
+      "step": 102
+    },
+    {
+      "epoch": 0.04877182598401894,
+      "grad_norm": 2.225109577178955,
+      "learning_rate": 1.0842105263157896e-05,
+      "loss": 0.5375,
+      "step": 103
+    },
+    {
+      "epoch": 0.049245338857650194,
+      "grad_norm": 1.645599365234375,
+      "learning_rate": 1.0947368421052633e-05,
+      "loss": 0.5288,
+      "step": 104
+    },
+    {
+      "epoch": 0.049718851731281447,
+      "grad_norm": 1.912824034690857,
+      "learning_rate": 1.105263157894737e-05,
+      "loss": 0.5346,
+      "step": 105
+    },
+    {
+      "epoch": 0.0501923646049127,
+      "grad_norm": 2.340646505355835,
+      "learning_rate": 1.1157894736842105e-05,
+      "loss": 0.5581,
+      "step": 106
+    },
+    {
+      "epoch": 0.050665877478543946,
+      "grad_norm": 1.8115869760513306,
+      "learning_rate": 1.1263157894736843e-05,
+      "loss": 0.5986,
+      "step": 107
+    },
+    {
+      "epoch": 0.0511393903521752,
+      "grad_norm": 2.0887653827667236,
+      "learning_rate": 1.136842105263158e-05,
+      "loss": 0.5353,
+      "step": 108
+    },
+    {
+      "epoch": 0.05161290322580645,
+      "grad_norm": 1.7228904962539673,
+      "learning_rate": 1.1473684210526317e-05,
+      "loss": 0.5273,
+      "step": 109
+    },
+    {
+      "epoch": 0.052086416099437705,
+      "grad_norm": 2.6061673164367676,
+      "learning_rate": 1.1578947368421053e-05,
+      "loss": 0.5476,
+      "step": 110
+    },
+    {
+      "epoch": 0.05255992897306896,
+      "grad_norm": 1.8213406801223755,
+      "learning_rate": 1.1684210526315792e-05,
+      "loss": 0.611,
+      "step": 111
+    },
+    {
+      "epoch": 0.05303344184670021,
+      "grad_norm": 1.6525737047195435,
+      "learning_rate": 1.1789473684210527e-05,
+      "loss": 0.5687,
+      "step": 112
+    },
+    {
+      "epoch": 0.05350695472033146,
+      "grad_norm": 1.7281478643417358,
+      "learning_rate": 1.1894736842105264e-05,
+      "loss": 0.5377,
+      "step": 113
+    },
+    {
+      "epoch": 0.05398046759396271,
+      "grad_norm": 1.835943341255188,
+      "learning_rate": 1.2e-05,
+      "loss": 0.4596,
+      "step": 114
+    },
+    {
+      "epoch": 0.05445398046759396,
+      "grad_norm": 1.6497899293899536,
+      "learning_rate": 1.2105263157894737e-05,
+      "loss": 0.544,
+      "step": 115
+    },
+    {
+      "epoch": 0.054927493341225216,
+      "grad_norm": 1.4863883256912231,
+      "learning_rate": 1.2210526315789475e-05,
+      "loss": 0.4536,
+      "step": 116
+    },
+    {
+      "epoch": 0.05540100621485647,
+      "grad_norm": 1.4068655967712402,
+      "learning_rate": 1.2315789473684212e-05,
+      "loss": 0.4522,
+      "step": 117
+    },
+    {
+      "epoch": 0.05587451908848772,
+      "grad_norm": 1.8424599170684814,
+      "learning_rate": 1.2421052631578949e-05,
+      "loss": 0.5377,
+      "step": 118
+    },
+    {
+      "epoch": 0.05634803196211897,
+      "grad_norm": 1.6451683044433594,
+      "learning_rate": 1.2526315789473684e-05,
+      "loss": 0.5254,
+      "step": 119
+    },
+    {
+      "epoch": 0.05682154483575022,
+      "grad_norm": 1.8392103910446167,
+      "learning_rate": 1.263157894736842e-05,
+      "loss": 0.4758,
+      "step": 120
+    },
+    {
+      "epoch": 0.057295057709381474,
+      "grad_norm": 1.9140554666519165,
+      "learning_rate": 1.2736842105263159e-05,
+      "loss": 0.5243,
+      "step": 121
+    },
+    {
+      "epoch": 0.05776857058301273,
+      "grad_norm": 2.6478939056396484,
+      "learning_rate": 1.2842105263157896e-05,
+      "loss": 0.4688,
+      "step": 122
+    },
+    {
+      "epoch": 0.05824208345664398,
+      "grad_norm": 1.8704174757003784,
+      "learning_rate": 1.2947368421052633e-05,
+      "loss": 0.5192,
+      "step": 123
+    },
+    {
+      "epoch": 0.05871559633027523,
+      "grad_norm": 1.791478157043457,
+      "learning_rate": 1.305263157894737e-05,
+      "loss": 0.5216,
+      "step": 124
+    },
+    {
+      "epoch": 0.05918910920390648,
+      "grad_norm": 2.0871686935424805,
+      "learning_rate": 1.3157894736842108e-05,
+      "loss": 0.4937,
+      "step": 125
+    },
+    {
+      "epoch": 0.05966262207753773,
+      "grad_norm": 1.6168326139450073,
+      "learning_rate": 1.3263157894736843e-05,
+      "loss": 0.5167,
+      "step": 126
+    },
+    {
+      "epoch": 0.060136134951168985,
+      "grad_norm": 1.9244468212127686,
+      "learning_rate": 1.336842105263158e-05,
+      "loss": 0.482,
+      "step": 127
+    },
+    {
+      "epoch": 0.06060964782480024,
+      "grad_norm": 1.7854818105697632,
+      "learning_rate": 1.3473684210526316e-05,
+      "loss": 0.5376,
+      "step": 128
+    },
+    {
+      "epoch": 0.06108316069843149,
+      "grad_norm": 1.5491472482681274,
+      "learning_rate": 1.3578947368421055e-05,
+      "loss": 0.4578,
+      "step": 129
+    },
+    {
+      "epoch": 0.061556673572062744,
+      "grad_norm": 2.987952470779419,
+      "learning_rate": 1.3684210526315791e-05,
+      "loss": 0.4415,
+      "step": 130
+    },
+    {
+      "epoch": 0.06203018644569399,
+      "grad_norm": 2.423494338989258,
+      "learning_rate": 1.3789473684210528e-05,
+      "loss": 0.4908,
+      "step": 131
+    },
+    {
+      "epoch": 0.06250369931932524,
+      "grad_norm": 1.8803611993789673,
+      "learning_rate": 1.3894736842105265e-05,
+      "loss": 0.4588,
+      "step": 132
+    },
+    {
+      "epoch": 0.0629772121929565,
+      "grad_norm": 2.069321393966675,
+      "learning_rate": 1.4e-05,
+      "loss": 0.5236,
+      "step": 133
+    },
+    {
+      "epoch": 0.06345072506658775,
+      "grad_norm": 2.4028356075286865,
+      "learning_rate": 1.4105263157894738e-05,
+      "loss": 0.5207,
+      "step": 134
+    },
+    {
+      "epoch": 0.063924237940219,
+      "grad_norm": 1.9155303239822388,
+      "learning_rate": 1.4210526315789475e-05,
+      "loss": 0.4975,
+      "step": 135
+    },
+    {
+      "epoch": 0.06439775081385025,
+      "grad_norm": 3.001650810241699,
+      "learning_rate": 1.4315789473684212e-05,
+      "loss": 0.4302,
+      "step": 136
+    },
+    {
+      "epoch": 0.0648712636874815,
+      "grad_norm": 2.335148334503174,
+      "learning_rate": 1.4421052631578948e-05,
+      "loss": 0.5123,
+      "step": 137
+    },
+    {
+      "epoch": 0.06534477656111276,
+      "grad_norm": 1.7075388431549072,
+      "learning_rate": 1.4526315789473687e-05,
+      "loss": 0.456,
+      "step": 138
+    },
+    {
+      "epoch": 0.065818289434744,
+      "grad_norm": 2.3079159259796143,
+      "learning_rate": 1.4631578947368424e-05,
+      "loss": 0.486,
+      "step": 139
+    },
+    {
+      "epoch": 0.06629180230837525,
+      "grad_norm": 2.429774522781372,
+      "learning_rate": 1.4736842105263159e-05,
+      "loss": 0.4889,
+      "step": 140
+    },
+    {
+      "epoch": 0.06676531518200651,
+      "grad_norm": 2.658094644546509,
+      "learning_rate": 1.4842105263157895e-05,
+      "loss": 0.4289,
+      "step": 141
+    },
+    {
+      "epoch": 0.06723882805563776,
+      "grad_norm": 1.507424235343933,
+      "learning_rate": 1.4947368421052632e-05,
+      "loss": 0.4725,
+      "step": 142
+    },
+    {
+      "epoch": 0.06771234092926902,
+      "grad_norm": 1.9254406690597534,
+      "learning_rate": 1.505263157894737e-05,
+      "loss": 0.456,
+      "step": 143
+    },
+    {
+      "epoch": 0.06818585380290026,
+      "grad_norm": 2.3014895915985107,
+      "learning_rate": 1.5157894736842107e-05,
+      "loss": 0.4627,
+      "step": 144
+    },
+    {
+      "epoch": 0.06865936667653152,
+      "grad_norm": 2.4393584728240967,
+      "learning_rate": 1.5263157894736846e-05,
+      "loss": 0.5304,
+      "step": 145
+    },
+    {
+      "epoch": 0.06913287955016277,
+      "grad_norm": 1.627350926399231,
+      "learning_rate": 1.536842105263158e-05,
+      "loss": 0.4897,
+      "step": 146
+    },
+    {
+      "epoch": 0.06960639242379402,
+      "grad_norm": 1.4329842329025269,
+      "learning_rate": 1.5473684210526316e-05,
+      "loss": 0.4581,
+      "step": 147
+    },
+    {
+      "epoch": 0.07007990529742528,
+      "grad_norm": 1.653219223022461,
+      "learning_rate": 1.5578947368421052e-05,
+      "loss": 0.4961,
+      "step": 148
+    },
+    {
+      "epoch": 0.07055341817105652,
+      "grad_norm": 1.8321658372879028,
+      "learning_rate": 1.568421052631579e-05,
+      "loss": 0.4684,
+      "step": 149
+    },
+    {
+      "epoch": 0.07102693104468778,
+      "grad_norm": 1.8446546792984009,
+      "learning_rate": 1.578947368421053e-05,
+      "loss": 0.4637,
+      "step": 150
+    },
+    {
+      "epoch": 0.07150044391831903,
+      "grad_norm": 1.486007809638977,
+      "learning_rate": 1.5894736842105266e-05,
+      "loss": 0.3637,
+      "step": 151
+    },
+    {
+      "epoch": 0.07197395679195027,
+      "grad_norm": 1.6993294954299927,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.4323,
+      "step": 152
+    },
+    {
+      "epoch": 0.07244746966558153,
+      "grad_norm": 1.391952633857727,
+      "learning_rate": 1.6105263157894736e-05,
+      "loss": 0.4072,
+      "step": 153
+    },
+    {
+      "epoch": 0.07292098253921278,
+      "grad_norm": 2.0186688899993896,
+      "learning_rate": 1.6210526315789473e-05,
+      "loss": 0.4412,
+      "step": 154
+    },
+    {
+      "epoch": 0.07339449541284404,
+      "grad_norm": 2.2258450984954834,
+      "learning_rate": 1.6315789473684213e-05,
+      "loss": 0.4302,
+      "step": 155
+    },
+    {
+      "epoch": 0.07386800828647529,
+      "grad_norm": 1.21897554397583,
+      "learning_rate": 1.642105263157895e-05,
+      "loss": 0.4197,
+      "step": 156
+    },
+    {
+      "epoch": 0.07434152116010655,
+      "grad_norm": 2.815114974975586,
+      "learning_rate": 1.6526315789473686e-05,
+      "loss": 0.4093,
+      "step": 157
+    },
+    {
+      "epoch": 0.07481503403373779,
+      "grad_norm": 1.9448319673538208,
+      "learning_rate": 1.6631578947368423e-05,
+      "loss": 0.3686,
+      "step": 158
+    },
+    {
+      "epoch": 0.07528854690736904,
+      "grad_norm": 1.9755760431289673,
+      "learning_rate": 1.673684210526316e-05,
+      "loss": 0.4371,
+      "step": 159
+    },
+    {
+      "epoch": 0.0757620597810003,
+      "grad_norm": 2.202780246734619,
+      "learning_rate": 1.6842105263157896e-05,
+      "loss": 0.41,
+      "step": 160
+    },
+    {
+      "epoch": 0.07623557265463154,
+      "grad_norm": 1.373448133468628,
+      "learning_rate": 1.6947368421052633e-05,
+      "loss": 0.3839,
+      "step": 161
+    },
+    {
+      "epoch": 0.0767090855282628,
+      "grad_norm": 2.2185139656066895,
+      "learning_rate": 1.705263157894737e-05,
+      "loss": 0.4207,
+      "step": 162
+    },
+    {
+      "epoch": 0.07718259840189405,
+      "grad_norm": 1.905508279800415,
+      "learning_rate": 1.7157894736842107e-05,
+      "loss": 0.3555,
+      "step": 163
+    },
+    {
+      "epoch": 0.0776561112755253,
+      "grad_norm": 1.355228304862976,
+      "learning_rate": 1.7263157894736843e-05,
+      "loss": 0.4047,
+      "step": 164
+    },
+    {
+      "epoch": 0.07812962414915656,
+      "grad_norm": 1.822799563407898,
+      "learning_rate": 1.736842105263158e-05,
+      "loss": 0.4355,
+      "step": 165
+    },
+    {
+      "epoch": 0.0786031370227878,
+      "grad_norm": 1.541835069656372,
+      "learning_rate": 1.7473684210526317e-05,
+      "loss": 0.3955,
+      "step": 166
+    },
+    {
+      "epoch": 0.07907664989641906,
+      "grad_norm": 1.8151495456695557,
+      "learning_rate": 1.7578947368421054e-05,
+      "loss": 0.4162,
+      "step": 167
+    },
+    {
+      "epoch": 0.07955016277005031,
+      "grad_norm": 1.320173978805542,
+      "learning_rate": 1.768421052631579e-05,
+      "loss": 0.4257,
+      "step": 168
+    },
+    {
+      "epoch": 0.08002367564368157,
+      "grad_norm": 2.0332558155059814,
+      "learning_rate": 1.7789473684210527e-05,
+      "loss": 0.3654,
+      "step": 169
+    },
+    {
+      "epoch": 0.08049718851731281,
+      "grad_norm": 1.4273725748062134,
+      "learning_rate": 1.7894736842105264e-05,
+      "loss": 0.4193,
+      "step": 170
+    },
+    {
+      "epoch": 0.08097070139094406,
+      "grad_norm": 1.5815318822860718,
+      "learning_rate": 1.8e-05,
+      "loss": 0.4495,
+      "step": 171
+    },
+    {
+      "epoch": 0.08144421426457532,
+      "grad_norm": 1.5733940601348877,
+      "learning_rate": 1.810526315789474e-05,
+      "loss": 0.4453,
+      "step": 172
+    },
+    {
+      "epoch": 0.08191772713820657,
+      "grad_norm": 1.554313063621521,
+      "learning_rate": 1.8210526315789477e-05,
+      "loss": 0.4122,
+      "step": 173
+    },
+    {
+      "epoch": 0.08239124001183783,
+      "grad_norm": 1.6655805110931396,
+      "learning_rate": 1.831578947368421e-05,
+      "loss": 0.4201,
+      "step": 174
+    },
+    {
+      "epoch": 0.08286475288546907,
+      "grad_norm": 2.0391829013824463,
+      "learning_rate": 1.8421052631578947e-05,
+      "loss": 0.3712,
+      "step": 175
+    },
+    {
+      "epoch": 0.08333826575910032,
+      "grad_norm": 1.7739325761795044,
+      "learning_rate": 1.8526315789473684e-05,
+      "loss": 0.3718,
+      "step": 176
+    },
+    {
+      "epoch": 0.08381177863273158,
+      "grad_norm": 1.7949507236480713,
+      "learning_rate": 1.8631578947368424e-05,
+      "loss": 0.4053,
+      "step": 177
+    },
+    {
+      "epoch": 0.08428529150636282,
+      "grad_norm": 1.8814252614974976,
+      "learning_rate": 1.873684210526316e-05,
+      "loss": 0.3998,
+      "step": 178
+    },
+    {
+      "epoch": 0.08475880437999408,
+      "grad_norm": 1.8132871389389038,
+      "learning_rate": 1.8842105263157898e-05,
+      "loss": 0.4553,
+      "step": 179
+    },
+    {
+      "epoch": 0.08523231725362533,
+      "grad_norm": 1.2668806314468384,
+      "learning_rate": 1.894736842105263e-05,
+      "loss": 0.3793,
+      "step": 180
+    },
+    {
+      "epoch": 0.08570583012725659,
+      "grad_norm": 1.6622042655944824,
+      "learning_rate": 1.9052631578947368e-05,
+      "loss": 0.398,
+      "step": 181
+    },
+    {
+      "epoch": 0.08617934300088784,
+      "grad_norm": 2.258330821990967,
+      "learning_rate": 1.9157894736842108e-05,
+      "loss": 0.3751,
+      "step": 182
+    },
+    {
+      "epoch": 0.08665285587451908,
+      "grad_norm": 2.4360086917877197,
+      "learning_rate": 1.9263157894736845e-05,
+      "loss": 0.3793,
+      "step": 183
+    },
+    {
+      "epoch": 0.08712636874815034,
+      "grad_norm": 1.8272238969802856,
+      "learning_rate": 1.936842105263158e-05,
+      "loss": 0.4108,
+      "step": 184
+    },
+    {
+      "epoch": 0.08759988162178159,
+      "grad_norm": 1.8764162063598633,
+      "learning_rate": 1.9473684210526318e-05,
+      "loss": 0.3656,
+      "step": 185
+    },
+    {
+      "epoch": 0.08807339449541285,
+      "grad_norm": 1.8359413146972656,
+      "learning_rate": 1.9578947368421055e-05,
+      "loss": 0.4059,
+      "step": 186
+    },
+    {
+      "epoch": 0.0885469073690441,
+      "grad_norm": 1.6942843198776245,
+      "learning_rate": 1.968421052631579e-05,
+      "loss": 0.4086,
+      "step": 187
+    },
+    {
+      "epoch": 0.08902042024267534,
+      "grad_norm": 1.4842171669006348,
+      "learning_rate": 1.9789473684210528e-05,
+      "loss": 0.3665,
+      "step": 188
+    },
+    {
+      "epoch": 0.0894939331163066,
+      "grad_norm": 2.0384953022003174,
+      "learning_rate": 1.9894736842105265e-05,
+      "loss": 0.4014,
+      "step": 189
+    },
+    {
+      "epoch": 0.08996744598993785,
+      "grad_norm": 2.0458626747131348,
+      "learning_rate": 2e-05,
+      "loss": 0.3963,
+      "step": 190
+    },
+    {
+      "epoch": 0.09044095886356911,
+      "grad_norm": 1.6842604875564575,
+      "learning_rate": 1.999999869229824e-05,
+      "loss": 0.3383,
+      "step": 191
+    },
+    {
+      "epoch": 0.09091447173720035,
+      "grad_norm": 1.3942151069641113,
+      "learning_rate": 1.9999994769193288e-05,
+      "loss": 0.3629,
+      "step": 192
+    },
+    {
+      "epoch": 0.09138798461083161,
+      "grad_norm": 1.8772989511489868,
+      "learning_rate": 1.9999988230686176e-05,
+      "loss": 0.3963,
+      "step": 193
+    },
+    {
+      "epoch": 0.09186149748446286,
+      "grad_norm": 1.479197382926941,
+      "learning_rate": 1.9999979076778615e-05,
+      "loss": 0.409,
+      "step": 194
+    },
+    {
+      "epoch": 0.0923350103580941,
+      "grad_norm": 1.848061442375183,
+      "learning_rate": 1.9999967307473e-05,
+      "loss": 0.3716,
+      "step": 195
+    },
+    {
+      "epoch": 0.09280852323172536,
+      "grad_norm": 1.4783906936645508,
+      "learning_rate": 1.9999952922772404e-05,
+      "loss": 0.3995,
+      "step": 196
+    },
+    {
+      "epoch": 0.09328203610535661,
+      "grad_norm": 1.4269295930862427,
+      "learning_rate": 1.9999935922680593e-05,
+      "loss": 0.428,
+      "step": 197
+    },
+    {
+      "epoch": 0.09375554897898787,
+      "grad_norm": 2.0845489501953125,
+      "learning_rate": 1.9999916307202013e-05,
+      "loss": 0.3909,
+      "step": 198
+    },
+    {
+      "epoch": 0.09422906185261912,
+      "grad_norm": 1.550615668296814,
+      "learning_rate": 1.9999894076341794e-05,
+      "loss": 0.3585,
+      "step": 199
+    },
+    {
+      "epoch": 0.09470257472625036,
+      "grad_norm": 1.9279413223266602,
+      "learning_rate": 1.999986923010575e-05,
+      "loss": 0.3696,
+      "step": 200
+    },
+    {
+      "epoch": 0.09517608759988162,
+      "grad_norm": 1.6810389757156372,
+      "learning_rate": 1.999984176850038e-05,
+      "loss": 0.4183,
+      "step": 201
+    },
+    {
+      "epoch": 0.09564960047351287,
+      "grad_norm": 1.6918199062347412,
+      "learning_rate": 1.9999811691532865e-05,
+      "loss": 0.4065,
+      "step": 202
+    },
+    {
+      "epoch": 0.09612311334714413,
+      "grad_norm": 1.8900259733200073,
+      "learning_rate": 1.999977899921107e-05,
+      "loss": 0.3825,
+      "step": 203
+    },
+    {
+      "epoch": 0.09659662622077538,
+      "grad_norm": 1.9546706676483154,
+      "learning_rate": 1.999974369154355e-05,
+      "loss": 0.3864,
+      "step": 204
+    },
+    {
+      "epoch": 0.09707013909440664,
+      "grad_norm": 1.354948878288269,
+      "learning_rate": 1.9999705768539537e-05,
+      "loss": 0.3947,
+      "step": 205
+    },
+    {
+      "epoch": 0.09754365196803788,
+      "grad_norm": 1.7710624933242798,
+      "learning_rate": 1.9999665230208947e-05,
+      "loss": 0.4115,
+      "step": 206
+    },
+    {
+      "epoch": 0.09801716484166913,
+      "grad_norm": 1.6053873300552368,
+      "learning_rate": 1.9999622076562387e-05,
+      "loss": 0.3892,
+      "step": 207
+    },
+    {
+      "epoch": 0.09849067771530039,
+      "grad_norm": 1.1621034145355225,
+      "learning_rate": 1.999957630761114e-05,
+      "loss": 0.3298,
+      "step": 208
+    },
+    {
+      "epoch": 0.09896419058893163,
+      "grad_norm": 1.2299578189849854,
+      "learning_rate": 1.9999527923367175e-05,
+      "loss": 0.3176,
+      "step": 209
+    },
+    {
+      "epoch": 0.09943770346256289,
+      "grad_norm": 1.7731809616088867,
+      "learning_rate": 1.999947692384315e-05,
+      "loss": 0.3745,
+      "step": 210
+    },
+    {
+      "epoch": 0.09991121633619414,
+      "grad_norm": 1.3102498054504395,
+      "learning_rate": 1.9999423309052405e-05,
+      "loss": 0.3688,
+      "step": 211
+    },
+    {
+      "epoch": 0.1003847292098254,
+      "grad_norm": 1.3057738542556763,
+      "learning_rate": 1.9999367079008957e-05,
+      "loss": 0.3865,
+      "step": 212
+    },
+    {
+      "epoch": 0.10085824208345665,
+      "grad_norm": 1.6710134744644165,
+      "learning_rate": 1.9999308233727516e-05,
+      "loss": 0.3523,
+      "step": 213
+    },
+    {
+      "epoch": 0.10133175495708789,
+      "grad_norm": 1.6011284589767456,
+      "learning_rate": 1.9999246773223468e-05,
+      "loss": 0.3286,
+      "step": 214
+    },
+    {
+      "epoch": 0.10180526783071915,
+      "grad_norm": 1.5414258241653442,
+      "learning_rate": 1.9999182697512896e-05,
+      "loss": 0.3608,
+      "step": 215
+    },
+    {
+      "epoch": 0.1022787807043504,
+      "grad_norm": 2.0255424976348877,
+      "learning_rate": 1.9999116006612553e-05,
+      "loss": 0.3466,
+      "step": 216
+    },
+    {
+      "epoch": 0.10275229357798166,
+      "grad_norm": 1.9602148532867432,
+      "learning_rate": 1.999904670053988e-05,
+      "loss": 0.3827,
+      "step": 217
+    },
+    {
+      "epoch": 0.1032258064516129,
+      "grad_norm": 2.306156635284424,
+      "learning_rate": 1.9998974779313004e-05,
+      "loss": 0.371,
+      "step": 218
+    },
+    {
+      "epoch": 0.10369931932524415,
+      "grad_norm": 1.572983980178833,
+      "learning_rate": 1.9998900242950736e-05,
+      "loss": 0.3279,
+      "step": 219
+    },
+    {
+      "epoch": 0.10417283219887541,
+      "grad_norm": 1.3669328689575195,
+      "learning_rate": 1.9998823091472574e-05,
+      "loss": 0.3717,
+      "step": 220
+    },
+    {
+      "epoch": 0.10464634507250666,
+      "grad_norm": 1.9657361507415771,
+      "learning_rate": 1.9998743324898687e-05,
+      "loss": 0.3583,
+      "step": 221
+    },
+    {
+      "epoch": 0.10511985794613792,
+      "grad_norm": 1.8538391590118408,
+      "learning_rate": 1.9998660943249947e-05,
+      "loss": 0.3816,
+      "step": 222
+    },
+    {
+      "epoch": 0.10559337081976916,
+      "grad_norm": 1.4254624843597412,
+      "learning_rate": 1.9998575946547897e-05,
+      "loss": 0.3319,
+      "step": 223
+    },
+    {
+      "epoch": 0.10606688369340042,
+      "grad_norm": 1.865258812904358,
+      "learning_rate": 1.9998488334814766e-05,
+      "loss": 0.3527,
+      "step": 224
+    },
+    {
+      "epoch": 0.10654039656703167,
+      "grad_norm": 1.446964979171753,
+      "learning_rate": 1.9998398108073465e-05,
+      "loss": 0.4055,
+      "step": 225
+    },
+    {
+      "epoch": 0.10701390944066291,
+      "grad_norm": 1.7752127647399902,
+      "learning_rate": 1.9998305266347598e-05,
+      "loss": 0.3875,
+      "step": 226
+    },
+    {
+      "epoch": 0.10748742231429417,
+      "grad_norm": 1.8876196146011353,
+      "learning_rate": 1.9998209809661443e-05,
+      "loss": 0.3439,
+      "step": 227
+    },
+    {
+      "epoch": 0.10796093518792542,
+      "grad_norm": 1.2993323802947998,
+      "learning_rate": 1.9998111738039964e-05,
+      "loss": 0.3372,
+      "step": 228
+    },
+    {
+      "epoch": 0.10843444806155668,
+      "grad_norm": 1.4371949434280396,
+      "learning_rate": 1.9998011051508816e-05,
+      "loss": 0.3672,
+      "step": 229
+    },
+    {
+      "epoch": 0.10890796093518793,
+      "grad_norm": 1.5482254028320312,
+      "learning_rate": 1.9997907750094332e-05,
+      "loss": 0.3554,
+      "step": 230
+    },
+    {
+      "epoch": 0.10938147380881917,
+      "grad_norm": 1.4659943580627441,
+      "learning_rate": 1.9997801833823526e-05,
+      "loss": 0.3691,
+      "step": 231
+    },
+    {
+      "epoch": 0.10985498668245043,
+      "grad_norm": 1.5168931484222412,
+      "learning_rate": 1.99976933027241e-05,
+      "loss": 0.3559,
+      "step": 232
+    },
+    {
+      "epoch": 0.11032849955608168,
+      "grad_norm": 1.4778302907943726,
+      "learning_rate": 1.999758215682444e-05,
+      "loss": 0.3201,
+      "step": 233
+    },
+    {
+      "epoch": 0.11080201242971294,
+      "grad_norm": 2.5164334774017334,
+      "learning_rate": 1.9997468396153615e-05,
+      "loss": 0.3711,
+      "step": 234
+    },
+    {
+      "epoch": 0.11127552530334418,
+      "grad_norm": 1.5941665172576904,
+      "learning_rate": 1.999735202074138e-05,
+      "loss": 0.3489,
+      "step": 235
+    },
+    {
+      "epoch": 0.11174903817697544,
+      "grad_norm": 1.5780380964279175,
+      "learning_rate": 1.9997233030618167e-05,
+      "loss": 0.3282,
+      "step": 236
+    },
+    {
+      "epoch": 0.11222255105060669,
+      "grad_norm": 1.8493186235427856,
+      "learning_rate": 1.9997111425815103e-05,
+      "loss": 0.3516,
+      "step": 237
+    },
+    {
+      "epoch": 0.11269606392423794,
+      "grad_norm": 1.4640581607818604,
+      "learning_rate": 1.999698720636399e-05,
+      "loss": 0.3542,
+      "step": 238
+    },
+    {
+      "epoch": 0.1131695767978692,
+      "grad_norm": 2.8939545154571533,
+      "learning_rate": 1.9996860372297312e-05,
+      "loss": 0.3596,
+      "step": 239
+    },
+    {
+      "epoch": 0.11364308967150044,
+      "grad_norm": 1.7447127103805542,
+      "learning_rate": 1.999673092364825e-05,
+      "loss": 0.3593,
+      "step": 240
+    },
+    {
+      "epoch": 0.1141166025451317,
+      "grad_norm": 1.753167986869812,
+      "learning_rate": 1.9996598860450653e-05,
+      "loss": 0.3278,
+      "step": 241
+    },
+    {
+      "epoch": 0.11459011541876295,
+      "grad_norm": 1.8249387741088867,
+      "learning_rate": 1.9996464182739063e-05,
+      "loss": 0.3373,
+      "step": 242
+    },
+    {
+      "epoch": 0.1150636282923942,
+      "grad_norm": 1.9170994758605957,
+      "learning_rate": 1.99963268905487e-05,
+      "loss": 0.3284,
+      "step": 243
+    },
+    {
+      "epoch": 0.11553714116602545,
+      "grad_norm": 2.2071142196655273,
+      "learning_rate": 1.999618698391548e-05,
+      "loss": 0.3633,
+      "step": 244
+    },
+    {
+      "epoch": 0.1160106540396567,
+      "grad_norm": 1.6293482780456543,
+      "learning_rate": 1.9996044462875984e-05,
+      "loss": 0.3624,
+      "step": 245
+    },
+    {
+      "epoch": 0.11648416691328796,
+      "grad_norm": 2.3833389282226562,
+      "learning_rate": 1.9995899327467498e-05,
+      "loss": 0.3329,
+      "step": 246
+    },
+    {
+      "epoch": 0.1169576797869192,
+      "grad_norm": 1.9268333911895752,
+      "learning_rate": 1.999575157772797e-05,
+      "loss": 0.3451,
+      "step": 247
+    },
+    {
+      "epoch": 0.11743119266055047,
+      "grad_norm": 1.3293579816818237,
+      "learning_rate": 1.9995601213696053e-05,
+      "loss": 0.3676,
+      "step": 248
+    },
+    {
+      "epoch": 0.11790470553418171,
+      "grad_norm": 1.651096224784851,
+      "learning_rate": 1.9995448235411062e-05,
+      "loss": 0.3527,
+      "step": 249
+    },
+    {
+      "epoch": 0.11837821840781296,
+      "grad_norm": 1.949661135673523,
+      "learning_rate": 1.999529264291302e-05,
+      "loss": 0.4063,
+      "step": 250
+    },
+    {
+      "epoch": 0.11885173128144422,
+      "grad_norm": 1.3306670188903809,
+      "learning_rate": 1.9995134436242606e-05,
+      "loss": 0.3607,
+      "step": 251
+    },
+    {
+      "epoch": 0.11932524415507546,
+      "grad_norm": 1.5161584615707397,
+      "learning_rate": 1.9994973615441207e-05,
+      "loss": 0.3825,
+      "step": 252
+    },
+    {
+      "epoch": 0.11979875702870672,
+      "grad_norm": 2.071152687072754,
+      "learning_rate": 1.9994810180550884e-05,
+      "loss": 0.3306,
+      "step": 253
+    },
+    {
+      "epoch": 0.12027226990233797,
+      "grad_norm": 1.514334797859192,
+      "learning_rate": 1.9994644131614382e-05,
+      "loss": 0.3254,
+      "step": 254
+    },
+    {
+      "epoch": 0.12074578277596922,
+      "grad_norm": 1.9304691553115845,
+      "learning_rate": 1.9994475468675122e-05,
+      "loss": 0.3863,
+      "step": 255
+    },
+    {
+      "epoch": 0.12121929564960048,
+      "grad_norm": 1.6955091953277588,
+      "learning_rate": 1.9994304191777228e-05,
+      "loss": 0.3559,
+      "step": 256
+    },
+    {
+      "epoch": 0.12169280852323172,
+      "grad_norm": 1.3693172931671143,
+      "learning_rate": 1.9994130300965485e-05,
+      "loss": 0.3393,
+      "step": 257
+    },
+    {
+      "epoch": 0.12216632139686298,
+      "grad_norm": 1.4945056438446045,
+      "learning_rate": 1.9993953796285377e-05,
+      "loss": 0.3051,
+      "step": 258
+    },
+    {
+      "epoch": 0.12263983427049423,
+      "grad_norm": 1.3563364744186401,
+      "learning_rate": 1.999377467778307e-05,
+      "loss": 0.3369,
+      "step": 259
+    },
+    {
+      "epoch": 0.12311334714412549,
+      "grad_norm": 1.2931690216064453,
+      "learning_rate": 1.9993592945505402e-05,
+      "loss": 0.3297,
+      "step": 260
+    },
+    {
+      "epoch": 0.12358686001775673,
+      "grad_norm": 1.5800117254257202,
+      "learning_rate": 1.9993408599499914e-05,
+      "loss": 0.3287,
+      "step": 261
+    },
+    {
+      "epoch": 0.12406037289138798,
+      "grad_norm": 1.3716331720352173,
+      "learning_rate": 1.999322163981481e-05,
+      "loss": 0.3092,
+      "step": 262
+    },
+    {
+      "epoch": 0.12453388576501924,
+      "grad_norm": 1.6586685180664062,
+      "learning_rate": 1.9993032066499e-05,
+      "loss": 0.4093,
+      "step": 263
+    },
+    {
+      "epoch": 0.12500739863865049,
+      "grad_norm": 1.4457738399505615,
+      "learning_rate": 1.999283987960205e-05,
+      "loss": 0.3679,
+      "step": 264
+    },
+    {
+      "epoch": 0.12548091151228175,
+      "grad_norm": 1.7291022539138794,
+      "learning_rate": 1.9992645079174235e-05,
+      "loss": 0.3708,
+      "step": 265
+    },
+    {
+      "epoch": 0.125954424385913,
+      "grad_norm": 1.6154969930648804,
+      "learning_rate": 1.99924476652665e-05,
+      "loss": 0.3684,
+      "step": 266
+    },
+    {
+      "epoch": 0.12642793725954424,
+      "grad_norm": 1.2971596717834473,
+      "learning_rate": 1.999224763793048e-05,
+      "loss": 0.3378,
+      "step": 267
+    },
+    {
+      "epoch": 0.1269014501331755,
+      "grad_norm": 2.0271036624908447,
+      "learning_rate": 1.9992044997218484e-05,
+      "loss": 0.3062,
+      "step": 268
+    },
+    {
+      "epoch": 0.12737496300680676,
+      "grad_norm": 1.4624063968658447,
+      "learning_rate": 1.9991839743183514e-05,
+      "loss": 0.3105,
+      "step": 269
+    },
+    {
+      "epoch": 0.127848475880438,
+      "grad_norm": 1.4940061569213867,
+      "learning_rate": 1.9991631875879252e-05,
+      "loss": 0.3523,
+      "step": 270
+    },
+    {
+      "epoch": 0.12832198875406925,
+      "grad_norm": 1.7870043516159058,
+      "learning_rate": 1.9991421395360066e-05,
+      "loss": 0.3551,
+      "step": 271
+    },
+    {
+      "epoch": 0.1287955016277005,
+      "grad_norm": 1.2716615200042725,
+      "learning_rate": 1.9991208301681e-05,
+      "loss": 0.3073,
+      "step": 272
+    },
+    {
+      "epoch": 0.12926901450133174,
+      "grad_norm": 1.2899311780929565,
+      "learning_rate": 1.9990992594897792e-05,
+      "loss": 0.3495,
+      "step": 273
+    },
+    {
+      "epoch": 0.129742527374963,
+      "grad_norm": 1.1452953815460205,
+      "learning_rate": 1.9990774275066856e-05,
+      "loss": 0.3054,
+      "step": 274
+    },
+    {
+      "epoch": 0.13021604024859426,
+      "grad_norm": 1.4126183986663818,
+      "learning_rate": 1.9990553342245288e-05,
+      "loss": 0.3209,
+      "step": 275
+    },
+    {
+      "epoch": 0.13068955312222552,
+      "grad_norm": 1.9172788858413696,
+      "learning_rate": 1.9990329796490878e-05,
+      "loss": 0.326,
+      "step": 276
+    },
+    {
+      "epoch": 0.13116306599585675,
+      "grad_norm": 1.8034718036651611,
+      "learning_rate": 1.9990103637862086e-05,
+      "loss": 0.3302,
+      "step": 277
+    },
+    {
+      "epoch": 0.131636578869488,
+      "grad_norm": 2.3256547451019287,
+      "learning_rate": 1.998987486641806e-05,
+      "loss": 0.36,
+      "step": 278
+    },
+    {
+      "epoch": 0.13211009174311927,
+      "grad_norm": 2.554792642593384,
+      "learning_rate": 1.9989643482218642e-05,
+      "loss": 0.3144,
+      "step": 279
+    },
+    {
+      "epoch": 0.1325836046167505,
+      "grad_norm": 1.872496485710144,
+      "learning_rate": 1.9989409485324342e-05,
+      "loss": 0.3312,
+      "step": 280
+    },
+    {
+      "epoch": 0.13305711749038177,
+      "grad_norm": 1.47379469871521,
+      "learning_rate": 1.998917287579636e-05,
+      "loss": 0.359,
+      "step": 281
+    },
+    {
+      "epoch": 0.13353063036401303,
+      "grad_norm": 1.6940183639526367,
+      "learning_rate": 1.998893365369658e-05,
+      "loss": 0.3364,
+      "step": 282
+    },
+    {
+      "epoch": 0.13400414323764429,
+      "grad_norm": 2.1470634937286377,
+      "learning_rate": 1.9988691819087568e-05,
+      "loss": 0.3076,
+      "step": 283
+    },
+    {
+      "epoch": 0.13447765611127552,
+      "grad_norm": 1.484626054763794,
+      "learning_rate": 1.9988447372032573e-05,
+      "loss": 0.3061,
+      "step": 284
+    },
+    {
+      "epoch": 0.13495116898490678,
+      "grad_norm": 2.2221710681915283,
+      "learning_rate": 1.9988200312595527e-05,
+      "loss": 0.3271,
+      "step": 285
+    },
+    {
+      "epoch": 0.13542468185853804,
+      "grad_norm": 1.8547298908233643,
+      "learning_rate": 1.998795064084105e-05,
+      "loss": 0.3321,
+      "step": 286
+    },
+    {
+      "epoch": 0.13589819473216927,
+      "grad_norm": 1.4910904169082642,
+      "learning_rate": 1.9987698356834438e-05,
+      "loss": 0.3319,
+      "step": 287
+    },
+    {
+      "epoch": 0.13637170760580053,
+      "grad_norm": 1.8652129173278809,
+      "learning_rate": 1.9987443460641672e-05,
+      "loss": 0.3279,
+      "step": 288
+    },
+    {
+      "epoch": 0.1368452204794318,
+      "grad_norm": 1.7757213115692139,
+      "learning_rate": 1.998718595232942e-05,
+      "loss": 0.3662,
+      "step": 289
+    },
+    {
+      "epoch": 0.13731873335306305,
+      "grad_norm": 1.7252709865570068,
+      "learning_rate": 1.998692583196503e-05,
+      "loss": 0.3394,
+      "step": 290
+    },
+    {
+      "epoch": 0.13779224622669428,
+      "grad_norm": 1.5977404117584229,
+      "learning_rate": 1.998666309961653e-05,
+      "loss": 0.3187,
+      "step": 291
+    },
+    {
+      "epoch": 0.13826575910032554,
+      "grad_norm": 2.18762469291687,
+      "learning_rate": 1.9986397755352643e-05,
+      "loss": 0.3153,
+      "step": 292
+    },
+    {
+      "epoch": 0.1387392719739568,
+      "grad_norm": 1.6634726524353027,
+      "learning_rate": 1.9986129799242767e-05,
+      "loss": 0.3312,
+      "step": 293
+    },
+    {
+      "epoch": 0.13921278484758803,
+      "grad_norm": 1.709532618522644,
+      "learning_rate": 1.9985859231356976e-05,
+      "loss": 0.3461,
+      "step": 294
+    },
+    {
+      "epoch": 0.1396862977212193,
+      "grad_norm": 1.9468045234680176,
+      "learning_rate": 1.998558605176604e-05,
+      "loss": 0.3441,
+      "step": 295
+    },
+    {
+      "epoch": 0.14015981059485055,
+      "grad_norm": 1.3585155010223389,
+      "learning_rate": 1.9985310260541403e-05,
+      "loss": 0.332,
+      "step": 296
+    },
+    {
+      "epoch": 0.14063332346848179,
+      "grad_norm": 2.272162437438965,
+      "learning_rate": 1.99850318577552e-05,
+      "loss": 0.3187,
+      "step": 297
+    },
+    {
+      "epoch": 0.14110683634211305,
+      "grad_norm": 1.5794426202774048,
+      "learning_rate": 1.998475084348024e-05,
+      "loss": 0.2903,
+      "step": 298
+    },
+    {
+      "epoch": 0.1415803492157443,
+      "grad_norm": 1.7926369905471802,
+      "learning_rate": 1.998446721779002e-05,
+      "loss": 0.3271,
+      "step": 299
+    },
+    {
+      "epoch": 0.14205386208937557,
+      "grad_norm": 2.8953635692596436,
+      "learning_rate": 1.9984180980758724e-05,
+      "loss": 0.3276,
+      "step": 300
+    },
+    {
+      "epoch": 0.1425273749630068,
+      "grad_norm": 2.217665910720825,
+      "learning_rate": 1.998389213246121e-05,
+      "loss": 0.3039,
+      "step": 301
+    },
+    {
+      "epoch": 0.14300088783663806,
+      "grad_norm": 1.6621983051300049,
+      "learning_rate": 1.9983600672973026e-05,
+      "loss": 0.3323,
+      "step": 302
+    },
+    {
+      "epoch": 0.14347440071026932,
+      "grad_norm": 2.2490603923797607,
+      "learning_rate": 1.99833066023704e-05,
+      "loss": 0.3002,
+      "step": 303
+    },
+    {
+      "epoch": 0.14394791358390055,
+      "grad_norm": 1.3009684085845947,
+      "learning_rate": 1.9983009920730244e-05,
+      "loss": 0.3074,
+      "step": 304
+    },
+    {
+      "epoch": 0.1444214264575318,
+      "grad_norm": 1.508434534072876,
+      "learning_rate": 1.998271062813015e-05,
+      "loss": 0.3917,
+      "step": 305
+    },
+    {
+      "epoch": 0.14489493933116307,
+      "grad_norm": 1.421549677848816,
+      "learning_rate": 1.998240872464839e-05,
+      "loss": 0.3269,
+      "step": 306
+    },
+    {
+      "epoch": 0.14536845220479433,
+      "grad_norm": 1.5330595970153809,
+      "learning_rate": 1.9982104210363936e-05,
+      "loss": 0.3122,
+      "step": 307
+    },
+    {
+      "epoch": 0.14584196507842556,
+      "grad_norm": 1.7626796960830688,
+      "learning_rate": 1.9981797085356426e-05,
+      "loss": 0.3176,
+      "step": 308
+    },
+    {
+      "epoch": 0.14631547795205682,
+      "grad_norm": 1.6426533460617065,
+      "learning_rate": 1.998148734970618e-05,
+      "loss": 0.3441,
+      "step": 309
+    },
+    {
+      "epoch": 0.14678899082568808,
+      "grad_norm": 1.3211804628372192,
+      "learning_rate": 1.9981175003494217e-05,
+      "loss": 0.2941,
+      "step": 310
+    },
+    {
+      "epoch": 0.1472625036993193,
+      "grad_norm": 1.625162124633789,
+      "learning_rate": 1.9980860046802214e-05,
+      "loss": 0.3284,
+      "step": 311
+    },
+    {
+      "epoch": 0.14773601657295057,
+      "grad_norm": 1.1030138731002808,
+      "learning_rate": 1.998054247971256e-05,
+      "loss": 0.3086,
+      "step": 312
+    },
+    {
+      "epoch": 0.14820952944658183,
+      "grad_norm": 1.5524837970733643,
+      "learning_rate": 1.9980222302308297e-05,
+      "loss": 0.3069,
+      "step": 313
+    },
+    {
+      "epoch": 0.1486830423202131,
+      "grad_norm": 1.973054051399231,
+      "learning_rate": 1.997989951467318e-05,
+      "loss": 0.3185,
+      "step": 314
+    },
+    {
+      "epoch": 0.14915655519384433,
+      "grad_norm": 1.6762253046035767,
+      "learning_rate": 1.9979574116891617e-05,
+      "loss": 0.3271,
+      "step": 315
+    },
+    {
+      "epoch": 0.14963006806747559,
+      "grad_norm": 1.4578509330749512,
+      "learning_rate": 1.997924610904872e-05,
+      "loss": 0.3082,
+      "step": 316
+    },
+    {
+      "epoch": 0.15010358094110685,
+      "grad_norm": 1.9425774812698364,
+      "learning_rate": 1.9978915491230277e-05,
+      "loss": 0.3229,
+      "step": 317
+    },
+    {
+      "epoch": 0.15057709381473808,
+      "grad_norm": 1.977746605873108,
+      "learning_rate": 1.997858226352275e-05,
+      "loss": 0.3359,
+      "step": 318
+    },
+    {
+      "epoch": 0.15105060668836934,
+      "grad_norm": 1.842903971672058,
+      "learning_rate": 1.9978246426013304e-05,
+      "loss": 0.3891,
+      "step": 319
+    },
+    {
+      "epoch": 0.1515241195620006,
+      "grad_norm": 1.3752530813217163,
+      "learning_rate": 1.9977907978789762e-05,
+      "loss": 0.3273,
+      "step": 320
+    },
+    {
+      "epoch": 0.15199763243563186,
+      "grad_norm": 1.6148236989974976,
+      "learning_rate": 1.997756692194065e-05,
+      "loss": 0.3787,
+      "step": 321
+    },
+    {
+      "epoch": 0.1524711453092631,
+      "grad_norm": 1.5016567707061768,
+      "learning_rate": 1.997722325555516e-05,
+      "loss": 0.2908,
+      "step": 322
+    },
+    {
+      "epoch": 0.15294465818289435,
+      "grad_norm": 1.3146488666534424,
+      "learning_rate": 1.9976876979723185e-05,
+      "loss": 0.2753,
+      "step": 323
+    },
+    {
+      "epoch": 0.1534181710565256,
+      "grad_norm": 1.3233283758163452,
+      "learning_rate": 1.9976528094535285e-05,
+      "loss": 0.3403,
+      "step": 324
+    },
+    {
+      "epoch": 0.15389168393015684,
+      "grad_norm": 1.2820795774459839,
+      "learning_rate": 1.9976176600082702e-05,
+      "loss": 0.3536,
+      "step": 325
+    },
+    {
+      "epoch": 0.1543651968037881,
+      "grad_norm": 1.5473982095718384,
+      "learning_rate": 1.9975822496457377e-05,
+      "loss": 0.3327,
+      "step": 326
+    },
+    {
+      "epoch": 0.15483870967741936,
+      "grad_norm": 2.2213194370269775,
+      "learning_rate": 1.9975465783751908e-05,
+      "loss": 0.3157,
+      "step": 327
+    },
+    {
+      "epoch": 0.1553122225510506,
+      "grad_norm": 2.196427345275879,
+      "learning_rate": 1.9975106462059604e-05,
+      "loss": 0.2918,
+      "step": 328
+    },
+    {
+      "epoch": 0.15578573542468185,
+      "grad_norm": 1.3942145109176636,
+      "learning_rate": 1.997474453147444e-05,
+      "loss": 0.3043,
+      "step": 329
+    },
+    {
+      "epoch": 0.1562592482983131,
+      "grad_norm": 1.2132917642593384,
+      "learning_rate": 1.9974379992091065e-05,
+      "loss": 0.2957,
+      "step": 330
+    },
+    {
+      "epoch": 0.15673276117194437,
+      "grad_norm": 1.470230221748352,
+      "learning_rate": 1.997401284400483e-05,
+      "loss": 0.3122,
+      "step": 331
+    },
+    {
+      "epoch": 0.1572062740455756,
+      "grad_norm": 2.0138354301452637,
+      "learning_rate": 1.9973643087311755e-05,
+      "loss": 0.3466,
+      "step": 332
+    },
+    {
+      "epoch": 0.15767978691920687,
+      "grad_norm": 1.662995457649231,
+      "learning_rate": 1.9973270722108553e-05,
+      "loss": 0.3182,
+      "step": 333
+    },
+    {
+      "epoch": 0.15815329979283813,
+      "grad_norm": 1.7974313497543335,
+      "learning_rate": 1.9972895748492603e-05,
+      "loss": 0.3393,
+      "step": 334
+    },
+    {
+      "epoch": 0.15862681266646936,
+      "grad_norm": 1.5902550220489502,
+      "learning_rate": 1.9972518166561983e-05,
+      "loss": 0.304,
+      "step": 335
+    },
+    {
+      "epoch": 0.15910032554010062,
+      "grad_norm": 1.5947198867797852,
+      "learning_rate": 1.997213797641544e-05,
+      "loss": 0.3209,
+      "step": 336
+    },
+    {
+      "epoch": 0.15957383841373188,
+      "grad_norm": 1.3484359979629517,
+      "learning_rate": 1.9971755178152416e-05,
+      "loss": 0.3136,
+      "step": 337
+    },
+    {
+      "epoch": 0.16004735128736314,
+      "grad_norm": 1.3645367622375488,
+      "learning_rate": 1.9971369771873022e-05,
+      "loss": 0.3269,
+      "step": 338
+    },
+    {
+      "epoch": 0.16052086416099437,
+      "grad_norm": 1.8705230951309204,
+      "learning_rate": 1.997098175767806e-05,
+      "loss": 0.2756,
+      "step": 339
+    },
+    {
+      "epoch": 0.16099437703462563,
+      "grad_norm": 1.3941823244094849,
+      "learning_rate": 1.997059113566901e-05,
+      "loss": 0.291,
+      "step": 340
+    },
+    {
+      "epoch": 0.1614678899082569,
+      "grad_norm": 1.3297516107559204,
+      "learning_rate": 1.9970197905948035e-05,
+      "loss": 0.3159,
+      "step": 341
+    },
+    {
+      "epoch": 0.16194140278188812,
+      "grad_norm": 1.6793006658554077,
+      "learning_rate": 1.996980206861799e-05,
+      "loss": 0.3268,
+      "step": 342
+    },
+    {
+      "epoch": 0.16241491565551938,
+      "grad_norm": 1.1595239639282227,
+      "learning_rate": 1.9969403623782385e-05,
+      "loss": 0.2932,
+      "step": 343
+    },
+    {
+      "epoch": 0.16288842852915064,
+      "grad_norm": 2.3051133155822754,
+      "learning_rate": 1.9969002571545442e-05,
+      "loss": 0.3267,
+      "step": 344
+    },
+    {
+      "epoch": 0.1633619414027819,
+      "grad_norm": 2.491314172744751,
+      "learning_rate": 1.996859891201205e-05,
+      "loss": 0.3367,
+      "step": 345
+    },
+    {
+      "epoch": 0.16383545427641313,
+      "grad_norm": 1.5665141344070435,
+      "learning_rate": 1.996819264528778e-05,
+      "loss": 0.3012,
+      "step": 346
+    },
+    {
+      "epoch": 0.1643089671500444,
+      "grad_norm": 2.1546201705932617,
+      "learning_rate": 1.996778377147889e-05,
+      "loss": 0.3249,
+      "step": 347
+    },
+    {
+      "epoch": 0.16478248002367565,
+      "grad_norm": 2.4622466564178467,
+      "learning_rate": 1.9967372290692314e-05,
+      "loss": 0.3302,
+      "step": 348
+    },
+    {
+      "epoch": 0.16525599289730689,
+      "grad_norm": 1.5641582012176514,
+      "learning_rate": 1.9966958203035673e-05,
+      "loss": 0.3561,
+      "step": 349
+    },
+    {
+      "epoch": 0.16572950577093815,
+      "grad_norm": 1.4802383184432983,
+      "learning_rate": 1.996654150861727e-05,
+      "loss": 0.2927,
+      "step": 350
+    },
+    {
+      "epoch": 0.1662030186445694,
+      "grad_norm": 2.025378942489624,
+      "learning_rate": 1.996612220754608e-05,
+      "loss": 0.3213,
+      "step": 351
+    },
+    {
+      "epoch": 0.16667653151820064,
+      "grad_norm": 2.12949275970459,
+      "learning_rate": 1.9965700299931772e-05,
+      "loss": 0.3446,
+      "step": 352
+    },
+    {
+      "epoch": 0.1671500443918319,
+      "grad_norm": 1.340847373008728,
+      "learning_rate": 1.9965275785884692e-05,
+      "loss": 0.308,
+      "step": 353
+    },
+    {
+      "epoch": 0.16762355726546316,
+      "grad_norm": 1.11890709400177,
+      "learning_rate": 1.9964848665515867e-05,
+      "loss": 0.2796,
+      "step": 354
+    },
+    {
+      "epoch": 0.16809707013909442,
+      "grad_norm": 1.3249599933624268,
+      "learning_rate": 1.9964418938937005e-05,
+      "loss": 0.3066,
+      "step": 355
+    },
+    {
+      "epoch": 0.16857058301272565,
+      "grad_norm": 2.209138870239258,
+      "learning_rate": 1.99639866062605e-05,
+      "loss": 0.3221,
+      "step": 356
+    },
+    {
+      "epoch": 0.1690440958863569,
+      "grad_norm": 1.4605731964111328,
+      "learning_rate": 1.9963551667599425e-05,
+      "loss": 0.3104,
+      "step": 357
+    },
+    {
+      "epoch": 0.16951760875998817,
+      "grad_norm": 1.4312578439712524,
+      "learning_rate": 1.9963114123067525e-05,
+      "loss": 0.3303,
+      "step": 358
+    },
+    {
+      "epoch": 0.1699911216336194,
+      "grad_norm": 1.3438875675201416,
+      "learning_rate": 1.9962673972779244e-05,
+      "loss": 0.327,
+      "step": 359
+    },
+    {
+      "epoch": 0.17046463450725066,
+      "grad_norm": 1.717059850692749,
+      "learning_rate": 1.99622312168497e-05,
+      "loss": 0.3151,
+      "step": 360
+    },
+    {
+      "epoch": 0.17093814738088192,
+      "grad_norm": 1.7237285375595093,
+      "learning_rate": 1.9961785855394685e-05,
+      "loss": 0.288,
+      "step": 361
+    },
+    {
+      "epoch": 0.17141166025451318,
+      "grad_norm": 1.342538833618164,
+      "learning_rate": 1.9961337888530686e-05,
+      "loss": 0.2842,
+      "step": 362
+    },
+    {
+      "epoch": 0.17188517312814441,
+      "grad_norm": 1.9628273248672485,
+      "learning_rate": 1.996088731637486e-05,
+      "loss": 0.3327,
+      "step": 363
+    },
+    {
+      "epoch": 0.17235868600177567,
+      "grad_norm": 1.262096643447876,
+      "learning_rate": 1.996043413904505e-05,
+      "loss": 0.2896,
+      "step": 364
+    },
+    {
+      "epoch": 0.17283219887540693,
+      "grad_norm": 1.4459971189498901,
+      "learning_rate": 1.995997835665978e-05,
+      "loss": 0.3512,
+      "step": 365
+    },
+    {
+      "epoch": 0.17330571174903817,
+      "grad_norm": 1.829930067062378,
+      "learning_rate": 1.9959519969338257e-05,
+      "loss": 0.3222,
+      "step": 366
+    },
+    {
+      "epoch": 0.17377922462266943,
+      "grad_norm": 1.5174282789230347,
+      "learning_rate": 1.9959058977200368e-05,
+      "loss": 0.3152,
+      "step": 367
+    },
+    {
+      "epoch": 0.17425273749630069,
+      "grad_norm": 2.4811291694641113,
+      "learning_rate": 1.9958595380366683e-05,
+      "loss": 0.2855,
+      "step": 368
+    },
+    {
+      "epoch": 0.17472625036993195,
+      "grad_norm": 1.502210259437561,
+      "learning_rate": 1.995812917895844e-05,
+      "loss": 0.3019,
+      "step": 369
+    },
+    {
+      "epoch": 0.17519976324356318,
+      "grad_norm": 2.2557857036590576,
+      "learning_rate": 1.9957660373097587e-05,
+      "loss": 0.3081,
+      "step": 370
+    },
+    {
+      "epoch": 0.17567327611719444,
+      "grad_norm": 1.5600831508636475,
+      "learning_rate": 1.9957188962906722e-05,
+      "loss": 0.2764,
+      "step": 371
+    },
+    {
+      "epoch": 0.1761467889908257,
+      "grad_norm": 1.6627660989761353,
+      "learning_rate": 1.9956714948509144e-05,
+      "loss": 0.2831,
+      "step": 372
+    },
+    {
+      "epoch": 0.17662030186445693,
+      "grad_norm": 1.7966787815093994,
+      "learning_rate": 1.9956238330028825e-05,
+      "loss": 0.3213,
+      "step": 373
+    },
+    {
+      "epoch": 0.1770938147380882,
+      "grad_norm": 1.3147883415222168,
+      "learning_rate": 1.9955759107590424e-05,
+      "loss": 0.2842,
+      "step": 374
+    },
+    {
+      "epoch": 0.17756732761171945,
+      "grad_norm": 1.6301828622817993,
+      "learning_rate": 1.9955277281319265e-05,
+      "loss": 0.288,
+      "step": 375
+    },
+    {
+      "epoch": 0.17804084048535068,
+      "grad_norm": 1.4622524976730347,
+      "learning_rate": 1.995479285134138e-05,
+      "loss": 0.3361,
+      "step": 376
+    },
+    {
+      "epoch": 0.17851435335898194,
+      "grad_norm": 1.8074785470962524,
+      "learning_rate": 1.9954305817783456e-05,
+      "loss": 0.3038,
+      "step": 377
+    },
+    {
+      "epoch": 0.1789878662326132,
+      "grad_norm": 1.4103055000305176,
+      "learning_rate": 1.995381618077288e-05,
+      "loss": 0.2992,
+      "step": 378
+    },
+    {
+      "epoch": 0.17946137910624446,
+      "grad_norm": 1.4178228378295898,
+      "learning_rate": 1.9953323940437707e-05,
+      "loss": 0.3074,
+      "step": 379
+    },
+    {
+      "epoch": 0.1799348919798757,
+      "grad_norm": 2.322828769683838,
+      "learning_rate": 1.9952829096906677e-05,
+      "loss": 0.3414,
+      "step": 380
+    },
+    {
+      "epoch": 0.18040840485350695,
+      "grad_norm": 1.5784039497375488,
+      "learning_rate": 1.9952331650309217e-05,
+      "loss": 0.3,
+      "step": 381
+    },
+    {
+      "epoch": 0.18088191772713821,
+      "grad_norm": 1.6129640340805054,
+      "learning_rate": 1.9951831600775423e-05,
+      "loss": 0.3698,
+      "step": 382
+    },
+    {
+      "epoch": 0.18135543060076945,
+      "grad_norm": 1.7583578824996948,
+      "learning_rate": 1.995132894843608e-05,
+      "loss": 0.312,
+      "step": 383
+    },
+    {
+      "epoch": 0.1818289434744007,
+      "grad_norm": 1.4428211450576782,
+      "learning_rate": 1.9950823693422653e-05,
+      "loss": 0.3401,
+      "step": 384
+    },
+    {
+      "epoch": 0.18230245634803197,
+      "grad_norm": 1.8162686824798584,
+      "learning_rate": 1.995031583586729e-05,
+      "loss": 0.2884,
+      "step": 385
+    },
+    {
+      "epoch": 0.18277596922166323,
+      "grad_norm": 1.871068000793457,
+      "learning_rate": 1.9949805375902807e-05,
+      "loss": 0.2974,
+      "step": 386
+    },
+    {
+      "epoch": 0.18324948209529446,
+      "grad_norm": 1.9816604852676392,
+      "learning_rate": 1.994929231366272e-05,
+      "loss": 0.269,
+      "step": 387
+    },
+    {
+      "epoch": 0.18372299496892572,
+      "grad_norm": 1.8843307495117188,
+      "learning_rate": 1.994877664928121e-05,
+      "loss": 0.3071,
+      "step": 388
+    },
+    {
+      "epoch": 0.18419650784255698,
+      "grad_norm": 1.7556949853897095,
+      "learning_rate": 1.994825838289314e-05,
+      "loss": 0.3656,
+      "step": 389
+    },
+    {
+      "epoch": 0.1846700207161882,
+      "grad_norm": 2.0573790073394775,
+      "learning_rate": 1.9947737514634068e-05,
+      "loss": 0.3047,
+      "step": 390
+    },
+    {
+      "epoch": 0.18514353358981947,
+      "grad_norm": 1.355928897857666,
+      "learning_rate": 1.9947214044640215e-05,
+      "loss": 0.3033,
+      "step": 391
+    },
+    {
+      "epoch": 0.18561704646345073,
+      "grad_norm": 1.6918185949325562,
+      "learning_rate": 1.9946687973048493e-05,
+      "loss": 0.2985,
+      "step": 392
+    },
+    {
+      "epoch": 0.186090559337082,
+      "grad_norm": 1.477439045906067,
+      "learning_rate": 1.9946159299996485e-05,
+      "loss": 0.3319,
+      "step": 393
+    },
+    {
+      "epoch": 0.18656407221071322,
+      "grad_norm": 1.5000855922698975,
+      "learning_rate": 1.9945628025622466e-05,
+      "loss": 0.2956,
+      "step": 394
+    },
+    {
+      "epoch": 0.18703758508434448,
+      "grad_norm": 1.4173663854599,
+      "learning_rate": 1.9945094150065385e-05,
+      "loss": 0.361,
+      "step": 395
+    },
+    {
+      "epoch": 0.18751109795797574,
+      "grad_norm": 1.6704801321029663,
+      "learning_rate": 1.9944557673464873e-05,
+      "loss": 0.3124,
+      "step": 396
+    },
+    {
+      "epoch": 0.18798461083160697,
+      "grad_norm": 1.5634368658065796,
+      "learning_rate": 1.9944018595961235e-05,
+      "loss": 0.294,
+      "step": 397
+    },
+    {
+      "epoch": 0.18845812370523823,
+      "grad_norm": 1.3400732278823853,
+      "learning_rate": 1.9943476917695465e-05,
+      "loss": 0.3401,
+      "step": 398
+    },
+    {
+      "epoch": 0.1889316365788695,
+      "grad_norm": 1.7150495052337646,
+      "learning_rate": 1.9942932638809233e-05,
+      "loss": 0.319,
+      "step": 399
+    },
+    {
+      "epoch": 0.18940514945250073,
+      "grad_norm": 1.299709439277649,
+      "learning_rate": 1.9942385759444892e-05,
+      "loss": 0.3081,
+      "step": 400
+    },
+    {
+      "epoch": 0.189878662326132,
+      "grad_norm": 1.4215167760849,
+      "learning_rate": 1.9941836279745473e-05,
+      "loss": 0.3435,
+      "step": 401
+    },
+    {
+      "epoch": 0.19035217519976325,
+      "grad_norm": 1.525974154472351,
+      "learning_rate": 1.9941284199854684e-05,
+      "loss": 0.2943,
+      "step": 402
+    },
+    {
+      "epoch": 0.1908256880733945,
+      "grad_norm": 1.4975392818450928,
+      "learning_rate": 1.994072951991692e-05,
+      "loss": 0.2931,
+      "step": 403
+    },
+    {
+      "epoch": 0.19129920094702574,
+      "grad_norm": 2.3207857608795166,
+      "learning_rate": 1.9940172240077248e-05,
+      "loss": 0.2704,
+      "step": 404
+    },
+    {
+      "epoch": 0.191772713820657,
+      "grad_norm": 1.4448798894882202,
+      "learning_rate": 1.993961236048142e-05,
+      "loss": 0.3027,
+      "step": 405
+    },
+    {
+      "epoch": 0.19224622669428826,
+      "grad_norm": 1.5526810884475708,
+      "learning_rate": 1.9939049881275868e-05,
+      "loss": 0.3239,
+      "step": 406
+    },
+    {
+      "epoch": 0.1927197395679195,
+      "grad_norm": 1.3078432083129883,
+      "learning_rate": 1.9938484802607704e-05,
+      "loss": 0.2749,
+      "step": 407
+    },
+    {
+      "epoch": 0.19319325244155075,
+      "grad_norm": 1.5397154092788696,
+      "learning_rate": 1.993791712462472e-05,
+      "loss": 0.3503,
+      "step": 408
+    },
+    {
+      "epoch": 0.193666765315182,
+      "grad_norm": 1.3340824842453003,
+      "learning_rate": 1.9937346847475382e-05,
+      "loss": 0.3272,
+      "step": 409
+    },
+    {
+      "epoch": 0.19414027818881327,
+      "grad_norm": 1.463257074356079,
+      "learning_rate": 1.9936773971308847e-05,
+      "loss": 0.3026,
+      "step": 410
+    },
+    {
+      "epoch": 0.1946137910624445,
+      "grad_norm": 1.3551300764083862,
+      "learning_rate": 1.993619849627494e-05,
+      "loss": 0.2841,
+      "step": 411
+    },
+    {
+      "epoch": 0.19508730393607576,
+      "grad_norm": 1.297788143157959,
+      "learning_rate": 1.9935620422524172e-05,
+      "loss": 0.2927,
+      "step": 412
+    },
+    {
+      "epoch": 0.19556081680970702,
+      "grad_norm": 1.5475691556930542,
+      "learning_rate": 1.993503975020773e-05,
+      "loss": 0.2562,
+      "step": 413
+    },
+    {
+      "epoch": 0.19603432968333825,
+      "grad_norm": 1.5862221717834473,
+      "learning_rate": 1.993445647947749e-05,
+      "loss": 0.2706,
+      "step": 414
+    },
+    {
+      "epoch": 0.19650784255696951,
+      "grad_norm": 1.496739387512207,
+      "learning_rate": 1.9933870610486e-05,
+      "loss": 0.2786,
+      "step": 415
+    },
+    {
+      "epoch": 0.19698135543060077,
+      "grad_norm": 1.2570983171463013,
+      "learning_rate": 1.9933282143386478e-05,
+      "loss": 0.2779,
+      "step": 416
+    },
+    {
+      "epoch": 0.19745486830423203,
+      "grad_norm": 1.8604596853256226,
+      "learning_rate": 1.9932691078332843e-05,
+      "loss": 0.307,
+      "step": 417
+    },
+    {
+      "epoch": 0.19792838117786327,
+      "grad_norm": 1.6916229724884033,
+      "learning_rate": 1.9932097415479683e-05,
+      "loss": 0.2797,
+      "step": 418
+    },
+    {
+      "epoch": 0.19840189405149453,
+      "grad_norm": 1.1430197954177856,
+      "learning_rate": 1.993150115498226e-05,
+      "loss": 0.2891,
+      "step": 419
+    },
+    {
+      "epoch": 0.19887540692512579,
+      "grad_norm": 1.5780760049819946,
+      "learning_rate": 1.9930902296996516e-05,
+      "loss": 0.2891,
+      "step": 420
+    },
+    {
+      "epoch": 0.19934891979875702,
+      "grad_norm": 1.7553094625473022,
+      "learning_rate": 1.993030084167908e-05,
+      "loss": 0.3078,
+      "step": 421
+    },
+    {
+      "epoch": 0.19982243267238828,
+      "grad_norm": 1.4407962560653687,
+      "learning_rate": 1.9929696789187264e-05,
+      "loss": 0.3151,
+      "step": 422
+    },
+    {
+      "epoch": 0.20029594554601954,
+      "grad_norm": 1.4529813528060913,
+      "learning_rate": 1.9929090139679045e-05,
+      "loss": 0.3176,
+      "step": 423
+    },
+    {
+      "epoch": 0.2007694584196508,
+      "grad_norm": 1.5530153512954712,
+      "learning_rate": 1.9928480893313082e-05,
+      "loss": 0.2996,
+      "step": 424
+    },
+    {
+      "epoch": 0.20124297129328203,
+      "grad_norm": 1.2986822128295898,
+      "learning_rate": 1.992786905024873e-05,
+      "loss": 0.3124,
+      "step": 425
+    },
+    {
+      "epoch": 0.2017164841669133,
+      "grad_norm": 1.4415348768234253,
+      "learning_rate": 1.9927254610646e-05,
+      "loss": 0.2665,
+      "step": 426
+    },
+    {
+      "epoch": 0.20218999704054455,
+      "grad_norm": 1.6581053733825684,
+      "learning_rate": 1.9926637574665598e-05,
+      "loss": 0.2962,
+      "step": 427
+    },
+    {
+      "epoch": 0.20266350991417578,
+      "grad_norm": 1.351175308227539,
+      "learning_rate": 1.9926017942468903e-05,
+      "loss": 0.299,
+      "step": 428
+    },
+    {
+      "epoch": 0.20313702278780704,
+      "grad_norm": 1.107627511024475,
+      "learning_rate": 1.992539571421797e-05,
+      "loss": 0.2441,
+      "step": 429
+    },
+    {
+      "epoch": 0.2036105356614383,
+      "grad_norm": 1.7610833644866943,
+      "learning_rate": 1.9924770890075544e-05,
+      "loss": 0.2864,
+      "step": 430
+    },
+    {
+      "epoch": 0.20408404853506953,
+      "grad_norm": 1.2733440399169922,
+      "learning_rate": 1.9924143470205034e-05,
+      "loss": 0.2957,
+      "step": 431
+    },
+    {
+      "epoch": 0.2045575614087008,
+      "grad_norm": 1.452659010887146,
+      "learning_rate": 1.992351345477054e-05,
+      "loss": 0.2657,
+      "step": 432
+    },
+    {
+      "epoch": 0.20503107428233205,
+      "grad_norm": 1.7691302299499512,
+      "learning_rate": 1.992288084393684e-05,
+      "loss": 0.3139,
+      "step": 433
+    },
+    {
+      "epoch": 0.20550458715596331,
+      "grad_norm": 1.0048437118530273,
+      "learning_rate": 1.9922245637869376e-05,
+      "loss": 0.2768,
+      "step": 434
+    },
+    {
+      "epoch": 0.20597810002959455,
+      "grad_norm": 1.4264775514602661,
+      "learning_rate": 1.9921607836734292e-05,
+      "loss": 0.3009,
+      "step": 435
+    },
+    {
+      "epoch": 0.2064516129032258,
+      "grad_norm": 1.2522767782211304,
+      "learning_rate": 1.9920967440698392e-05,
+      "loss": 0.3091,
+      "step": 436
+    },
+    {
+      "epoch": 0.20692512577685707,
+      "grad_norm": 1.4616775512695312,
+      "learning_rate": 1.992032444992917e-05,
+      "loss": 0.2709,
+      "step": 437
+    },
+    {
+      "epoch": 0.2073986386504883,
+      "grad_norm": 2.048379421234131,
+      "learning_rate": 1.9919678864594788e-05,
+      "loss": 0.2922,
+      "step": 438
+    },
+    {
+      "epoch": 0.20787215152411956,
+      "grad_norm": 2.135870933532715,
+      "learning_rate": 1.9919030684864097e-05,
+      "loss": 0.3139,
+      "step": 439
+    },
+    {
+      "epoch": 0.20834566439775082,
+      "grad_norm": 1.5550521612167358,
+      "learning_rate": 1.991837991090662e-05,
+      "loss": 0.2967,
+      "step": 440
+    },
+    {
+      "epoch": 0.20881917727138208,
+      "grad_norm": 1.4404891729354858,
+      "learning_rate": 1.9917726542892562e-05,
+      "loss": 0.2625,
+      "step": 441
+    },
+    {
+      "epoch": 0.2092926901450133,
+      "grad_norm": 1.4406466484069824,
+      "learning_rate": 1.9917070580992805e-05,
+      "loss": 0.2846,
+      "step": 442
+    },
+    {
+      "epoch": 0.20976620301864457,
+      "grad_norm": 2.054635524749756,
+      "learning_rate": 1.9916412025378907e-05,
+      "loss": 0.3299,
+      "step": 443
+    },
+    {
+      "epoch": 0.21023971589227583,
+      "grad_norm": 1.5401420593261719,
+      "learning_rate": 1.9915750876223112e-05,
+      "loss": 0.2995,
+      "step": 444
+    },
+    {
+      "epoch": 0.21071322876590706,
+      "grad_norm": 1.5220952033996582,
+      "learning_rate": 1.9915087133698333e-05,
+      "loss": 0.2638,
+      "step": 445
+    },
+    {
+      "epoch": 0.21118674163953832,
+      "grad_norm": 1.1745985746383667,
+      "learning_rate": 1.9914420797978167e-05,
+      "loss": 0.2603,
+      "step": 446
+    },
+    {
+      "epoch": 0.21166025451316958,
+      "grad_norm": 1.6358566284179688,
+      "learning_rate": 1.9913751869236888e-05,
+      "loss": 0.2847,
+      "step": 447
+    },
+    {
+      "epoch": 0.21213376738680084,
+      "grad_norm": 1.0750114917755127,
+      "learning_rate": 1.9913080347649446e-05,
+      "loss": 0.2783,
+      "step": 448
+    },
+    {
+      "epoch": 0.21260728026043207,
+      "grad_norm": 2.263500928878784,
+      "learning_rate": 1.9912406233391474e-05,
+      "loss": 0.2818,
+      "step": 449
+    },
+    {
+      "epoch": 0.21308079313406333,
+      "grad_norm": 1.2337026596069336,
+      "learning_rate": 1.991172952663928e-05,
+      "loss": 0.2995,
+      "step": 450
+    },
+    {
+      "epoch": 0.2135543060076946,
+      "grad_norm": 1.5734847784042358,
+      "learning_rate": 1.9911050227569845e-05,
+      "loss": 0.2681,
+      "step": 451
+    },
+    {
+      "epoch": 0.21402781888132583,
+      "grad_norm": 1.5128898620605469,
+      "learning_rate": 1.9910368336360836e-05,
+      "loss": 0.3128,
+      "step": 452
+    },
+    {
+      "epoch": 0.2145013317549571,
+      "grad_norm": 2.963315486907959,
+      "learning_rate": 1.99096838531906e-05,
+      "loss": 0.3157,
+      "step": 453
+    },
+    {
+      "epoch": 0.21497484462858835,
+      "grad_norm": 1.357264757156372,
+      "learning_rate": 1.990899677823815e-05,
+      "loss": 0.3173,
+      "step": 454
+    },
+    {
+      "epoch": 0.21544835750221958,
+      "grad_norm": 1.2650480270385742,
+      "learning_rate": 1.9908307111683184e-05,
+      "loss": 0.2686,
+      "step": 455
+    },
+    {
+      "epoch": 0.21592187037585084,
+      "grad_norm": 1.538608193397522,
+      "learning_rate": 1.990761485370608e-05,
+      "loss": 0.3223,
+      "step": 456
+    },
+    {
+      "epoch": 0.2163953832494821,
+      "grad_norm": 1.675292730331421,
+      "learning_rate": 1.9906920004487894e-05,
+      "loss": 0.292,
+      "step": 457
+    },
+    {
+      "epoch": 0.21686889612311336,
+      "grad_norm": 1.2939106225967407,
+      "learning_rate": 1.9906222564210353e-05,
+      "loss": 0.2695,
+      "step": 458
+    },
+    {
+      "epoch": 0.2173424089967446,
+      "grad_norm": 1.1748974323272705,
+      "learning_rate": 1.990552253305587e-05,
+      "loss": 0.3129,
+      "step": 459
+    },
+    {
+      "epoch": 0.21781592187037585,
+      "grad_norm": 1.7201486825942993,
+      "learning_rate": 1.9904819911207526e-05,
+      "loss": 0.3042,
+      "step": 460
+    },
+    {
+      "epoch": 0.2182894347440071,
+      "grad_norm": 1.790076732635498,
+      "learning_rate": 1.990411469884909e-05,
+      "loss": 0.3384,
+      "step": 461
+    },
+    {
+      "epoch": 0.21876294761763834,
+      "grad_norm": 1.5797431468963623,
+      "learning_rate": 1.9903406896165e-05,
+      "loss": 0.3244,
+      "step": 462
+    },
+    {
+      "epoch": 0.2192364604912696,
+      "grad_norm": 1.390303373336792,
+      "learning_rate": 1.9902696503340378e-05,
+      "loss": 0.2862,
+      "step": 463
+    },
+    {
+      "epoch": 0.21970997336490086,
+      "grad_norm": 1.6508777141571045,
+      "learning_rate": 1.990198352056102e-05,
+      "loss": 0.313,
+      "step": 464
+    },
+    {
+      "epoch": 0.22018348623853212,
+      "grad_norm": 1.5535106658935547,
+      "learning_rate": 1.990126794801339e-05,
+      "loss": 0.313,
+      "step": 465
+    },
+    {
+      "epoch": 0.22065699911216335,
+      "grad_norm": 1.5495903491973877,
+      "learning_rate": 1.9900549785884654e-05,
+      "loss": 0.2866,
+      "step": 466
+    },
+    {
+      "epoch": 0.22113051198579461,
+      "grad_norm": 1.5182076692581177,
+      "learning_rate": 1.989982903436263e-05,
+      "loss": 0.2786,
+      "step": 467
+    },
+    {
+      "epoch": 0.22160402485942587,
+      "grad_norm": 1.5054259300231934,
+      "learning_rate": 1.989910569363583e-05,
+      "loss": 0.2599,
+      "step": 468
+    },
+    {
+      "epoch": 0.2220775377330571,
+      "grad_norm": 1.793799877166748,
+      "learning_rate": 1.989837976389344e-05,
+      "loss": 0.3061,
+      "step": 469
+    },
+    {
+      "epoch": 0.22255105060668837,
+      "grad_norm": 1.3098176717758179,
+      "learning_rate": 1.9897651245325306e-05,
+      "loss": 0.268,
+      "step": 470
+    },
+    {
+      "epoch": 0.22302456348031963,
+      "grad_norm": 1.5548096895217896,
+      "learning_rate": 1.9896920138121977e-05,
+      "loss": 0.2858,
+      "step": 471
+    },
+    {
+      "epoch": 0.2234980763539509,
+      "grad_norm": 1.5963671207427979,
+      "learning_rate": 1.989618644247466e-05,
+      "loss": 0.3073,
+      "step": 472
+    },
+    {
+      "epoch": 0.22397158922758212,
+      "grad_norm": 1.266221046447754,
+      "learning_rate": 1.989545015857525e-05,
+      "loss": 0.2709,
+      "step": 473
+    },
+    {
+      "epoch": 0.22444510210121338,
+      "grad_norm": 1.2214033603668213,
+      "learning_rate": 1.9894711286616313e-05,
+      "loss": 0.2725,
+      "step": 474
+    },
+    {
+      "epoch": 0.22491861497484464,
+      "grad_norm": 1.5835678577423096,
+      "learning_rate": 1.98939698267911e-05,
+      "loss": 0.2411,
+      "step": 475
+    },
+    {
+      "epoch": 0.22539212784847587,
+      "grad_norm": 1.4112712144851685,
+      "learning_rate": 1.989322577929352e-05,
+      "loss": 0.2913,
+      "step": 476
+    },
+    {
+      "epoch": 0.22586564072210713,
+      "grad_norm": 1.335115909576416,
+      "learning_rate": 1.9892479144318187e-05,
+      "loss": 0.2936,
+      "step": 477
+    },
+    {
+      "epoch": 0.2263391535957384,
+      "grad_norm": 1.4088594913482666,
+      "learning_rate": 1.989172992206036e-05,
+      "loss": 0.3129,
+      "step": 478
+    },
+    {
+      "epoch": 0.22681266646936962,
+      "grad_norm": 1.2447527647018433,
+      "learning_rate": 1.989097811271601e-05,
+      "loss": 0.3029,
+      "step": 479
+    },
+    {
+      "epoch": 0.22728617934300088,
+      "grad_norm": 1.2852405309677124,
+      "learning_rate": 1.9890223716481746e-05,
+      "loss": 0.2788,
+      "step": 480
+    },
+    {
+      "epoch": 0.22775969221663214,
+      "grad_norm": 2.38966703414917,
+      "learning_rate": 1.9889466733554883e-05,
+      "loss": 0.3573,
+      "step": 481
+    },
+    {
+      "epoch": 0.2282332050902634,
+      "grad_norm": 1.3226877450942993,
+      "learning_rate": 1.9888707164133403e-05,
+      "loss": 0.2997,
+      "step": 482
+    },
+    {
+      "epoch": 0.22870671796389463,
+      "grad_norm": 1.2726070880889893,
+      "learning_rate": 1.988794500841596e-05,
+      "loss": 0.27,
+      "step": 483
+    },
+    {
+      "epoch": 0.2291802308375259,
+      "grad_norm": 1.3730125427246094,
+      "learning_rate": 1.9887180266601892e-05,
+      "loss": 0.2917,
+      "step": 484
+    },
+    {
+      "epoch": 0.22965374371115715,
+      "grad_norm": 1.4895507097244263,
+      "learning_rate": 1.988641293889121e-05,
+      "loss": 0.3041,
+      "step": 485
+    },
+    {
+      "epoch": 0.2301272565847884,
+      "grad_norm": 1.5664935111999512,
+      "learning_rate": 1.9885643025484598e-05,
+      "loss": 0.2987,
+      "step": 486
+    },
+    {
+      "epoch": 0.23060076945841965,
+      "grad_norm": 1.38105046749115,
+      "learning_rate": 1.988487052658342e-05,
+      "loss": 0.2739,
+      "step": 487
+    },
+    {
+      "epoch": 0.2310742823320509,
+      "grad_norm": 1.3974148035049438,
+      "learning_rate": 1.988409544238972e-05,
+      "loss": 0.294,
+      "step": 488
+    },
+    {
+      "epoch": 0.23154779520568217,
+      "grad_norm": 1.1836129426956177,
+      "learning_rate": 1.988331777310621e-05,
+      "loss": 0.2847,
+      "step": 489
+    },
+    {
+      "epoch": 0.2320213080793134,
+      "grad_norm": 1.9731239080429077,
+      "learning_rate": 1.9882537518936283e-05,
+      "loss": 0.2773,
+      "step": 490
+    },
+    {
+      "epoch": 0.23249482095294466,
+      "grad_norm": 1.8344943523406982,
+      "learning_rate": 1.9881754680084e-05,
+      "loss": 0.3044,
+      "step": 491
+    },
+    {
+      "epoch": 0.23296833382657592,
+      "grad_norm": 1.9124189615249634,
+      "learning_rate": 1.988096925675412e-05,
+      "loss": 0.2658,
+      "step": 492
+    },
+    {
+      "epoch": 0.23344184670020715,
+      "grad_norm": 1.1877110004425049,
+      "learning_rate": 1.988018124915205e-05,
+      "loss": 0.2978,
+      "step": 493
+    },
+    {
+      "epoch": 0.2339153595738384,
+      "grad_norm": 3.0704288482666016,
+      "learning_rate": 1.987939065748389e-05,
+      "loss": 0.2754,
+      "step": 494
+    },
+    {
+      "epoch": 0.23438887244746967,
+      "grad_norm": 2.0718958377838135,
+      "learning_rate": 1.9878597481956416e-05,
+      "loss": 0.2838,
+      "step": 495
+    },
+    {
+      "epoch": 0.23486238532110093,
+      "grad_norm": 1.83311927318573,
+      "learning_rate": 1.9877801722777064e-05,
+      "loss": 0.2823,
+      "step": 496
+    },
+    {
+      "epoch": 0.23533589819473216,
+      "grad_norm": 1.3887392282485962,
+      "learning_rate": 1.9877003380153968e-05,
+      "loss": 0.2805,
+      "step": 497
+    },
+    {
+      "epoch": 0.23580941106836342,
+      "grad_norm": 1.7678110599517822,
+      "learning_rate": 1.9876202454295926e-05,
+      "loss": 0.2892,
+      "step": 498
+    },
+    {
+      "epoch": 0.23628292394199468,
+      "grad_norm": 1.2601211071014404,
+      "learning_rate": 1.9875398945412403e-05,
+      "loss": 0.3009,
+      "step": 499
+    },
+    {
+      "epoch": 0.23675643681562591,
+      "grad_norm": 1.4841456413269043,
+      "learning_rate": 1.987459285371356e-05,
+      "loss": 0.2835,
+      "step": 500
+    },
+    {
+      "epoch": 0.23722994968925717,
+      "grad_norm": 1.53849196434021,
+      "learning_rate": 1.9873784179410217e-05,
+      "loss": 0.2825,
+      "step": 501
+    },
+    {
+      "epoch": 0.23770346256288843,
+      "grad_norm": 1.1709768772125244,
+      "learning_rate": 1.9872972922713875e-05,
+      "loss": 0.2877,
+      "step": 502
+    },
+    {
+      "epoch": 0.23817697543651967,
+      "grad_norm": 1.4577317237854004,
+      "learning_rate": 1.9872159083836713e-05,
+      "loss": 0.2704,
+      "step": 503
+    },
+    {
+      "epoch": 0.23865048831015093,
+      "grad_norm": 1.207236886024475,
+      "learning_rate": 1.9871342662991582e-05,
+      "loss": 0.3042,
+      "step": 504
+    },
+    {
+      "epoch": 0.2391240011837822,
+      "grad_norm": 1.409219741821289,
+      "learning_rate": 1.9870523660392004e-05,
+      "loss": 0.2925,
+      "step": 505
+    },
+    {
+      "epoch": 0.23959751405741345,
+      "grad_norm": 2.00093412399292,
+      "learning_rate": 1.986970207625219e-05,
+      "loss": 0.3005,
+      "step": 506
+    },
+    {
+      "epoch": 0.24007102693104468,
+      "grad_norm": 1.3477380275726318,
+      "learning_rate": 1.986887791078701e-05,
+      "loss": 0.2754,
+      "step": 507
+    },
+    {
+      "epoch": 0.24054453980467594,
+      "grad_norm": 1.0551843643188477,
+      "learning_rate": 1.9868051164212017e-05,
+      "loss": 0.3066,
+      "step": 508
+    },
+    {
+      "epoch": 0.2410180526783072,
+      "grad_norm": 1.1477186679840088,
+      "learning_rate": 1.986722183674344e-05,
+      "loss": 0.2924,
+      "step": 509
+    },
+    {
+      "epoch": 0.24149156555193843,
+      "grad_norm": 1.6527049541473389,
+      "learning_rate": 1.9866389928598188e-05,
+      "loss": 0.3204,
+      "step": 510
+    },
+    {
+      "epoch": 0.2419650784255697,
+      "grad_norm": 1.1355421543121338,
+      "learning_rate": 1.986555543999383e-05,
+      "loss": 0.2904,
+      "step": 511
+    },
+    {
+      "epoch": 0.24243859129920095,
+      "grad_norm": 1.9797145128250122,
+      "learning_rate": 1.9864718371148623e-05,
+      "loss": 0.2834,
+      "step": 512
+    },
+    {
+      "epoch": 0.2429121041728322,
+      "grad_norm": 2.0345993041992188,
+      "learning_rate": 1.9863878722281492e-05,
+      "loss": 0.2688,
+      "step": 513
+    },
+    {
+      "epoch": 0.24338561704646344,
+      "grad_norm": 2.046318769454956,
+      "learning_rate": 1.986303649361204e-05,
+      "loss": 0.2846,
+      "step": 514
+    },
+    {
+      "epoch": 0.2438591299200947,
+      "grad_norm": 1.3903858661651611,
+      "learning_rate": 1.986219168536054e-05,
+      "loss": 0.2915,
+      "step": 515
+    },
+    {
+      "epoch": 0.24433264279372596,
+      "grad_norm": 1.3947668075561523,
+      "learning_rate": 1.986134429774795e-05,
+      "loss": 0.3142,
+      "step": 516
+    },
+    {
+      "epoch": 0.2448061556673572,
+      "grad_norm": 1.6972023248672485,
+      "learning_rate": 1.9860494330995892e-05,
+      "loss": 0.2862,
+      "step": 517
+    },
+    {
+      "epoch": 0.24527966854098845,
+      "grad_norm": 1.431135892868042,
+      "learning_rate": 1.9859641785326672e-05,
+      "loss": 0.2691,
+      "step": 518
+    },
+    {
+      "epoch": 0.24575318141461971,
+      "grad_norm": 1.653490424156189,
+      "learning_rate": 1.9858786660963253e-05,
+      "loss": 0.2368,
+      "step": 519
+    },
+    {
+      "epoch": 0.24622669428825097,
+      "grad_norm": 2.2051680088043213,
+      "learning_rate": 1.98579289581293e-05,
+      "loss": 0.294,
+      "step": 520
+    },
+    {
+      "epoch": 0.2467002071618822,
+      "grad_norm": 2.165895938873291,
+      "learning_rate": 1.9857068677049124e-05,
+      "loss": 0.2745,
+      "step": 521
+    },
+    {
+      "epoch": 0.24717372003551347,
+      "grad_norm": 2.177927255630493,
+      "learning_rate": 1.9856205817947728e-05,
+      "loss": 0.3092,
+      "step": 522
+    },
+    {
+      "epoch": 0.24764723290914473,
+      "grad_norm": 1.3167831897735596,
+      "learning_rate": 1.9855340381050787e-05,
+      "loss": 0.2428,
+      "step": 523
+    },
+    {
+      "epoch": 0.24812074578277596,
+      "grad_norm": 1.1702346801757812,
+      "learning_rate": 1.9854472366584646e-05,
+      "loss": 0.2834,
+      "step": 524
+    },
+    {
+      "epoch": 0.24859425865640722,
+      "grad_norm": 1.4927124977111816,
+      "learning_rate": 1.9853601774776322e-05,
+      "loss": 0.2695,
+      "step": 525
+    },
+    {
+      "epoch": 0.24906777153003848,
+      "grad_norm": 2.9464986324310303,
+      "learning_rate": 1.9852728605853516e-05,
+      "loss": 0.278,
+      "step": 526
+    },
+    {
+      "epoch": 0.24954128440366974,
+      "grad_norm": 1.3696672916412354,
+      "learning_rate": 1.9851852860044594e-05,
+      "loss": 0.2743,
+      "step": 527
+    },
+    {
+      "epoch": 0.25001479727730097,
+      "grad_norm": 1.6753240823745728,
+      "learning_rate": 1.9850974537578597e-05,
+      "loss": 0.2999,
+      "step": 528
+    },
+    {
+      "epoch": 0.2504883101509322,
+      "grad_norm": 1.2681008577346802,
+      "learning_rate": 1.9850093638685247e-05,
+      "loss": 0.2763,
+      "step": 529
+    },
+    {
+      "epoch": 0.2509618230245635,
+      "grad_norm": 1.4322853088378906,
+      "learning_rate": 1.984921016359493e-05,
+      "loss": 0.301,
+      "step": 530
+    },
+    {
+      "epoch": 0.2514353358981947,
+      "grad_norm": 1.3877912759780884,
+      "learning_rate": 1.984832411253871e-05,
+      "loss": 0.2866,
+      "step": 531
+    },
+    {
+      "epoch": 0.251908848771826,
+      "grad_norm": 1.4901669025421143,
+      "learning_rate": 1.9847435485748328e-05,
+      "loss": 0.3069,
+      "step": 532
+    },
+    {
+      "epoch": 0.25238236164545724,
+      "grad_norm": 1.5434002876281738,
+      "learning_rate": 1.9846544283456195e-05,
+      "loss": 0.2884,
+      "step": 533
+    },
+    {
+      "epoch": 0.2528558745190885,
+      "grad_norm": 1.3284276723861694,
+      "learning_rate": 1.9845650505895397e-05,
+      "loss": 0.2709,
+      "step": 534
+    },
+    {
+      "epoch": 0.25332938739271976,
+      "grad_norm": 1.1716914176940918,
+      "learning_rate": 1.9844754153299686e-05,
+      "loss": 0.273,
+      "step": 535
+    },
+    {
+      "epoch": 0.253802900266351,
+      "grad_norm": 1.208626389503479,
+      "learning_rate": 1.984385522590351e-05,
+      "loss": 0.2616,
+      "step": 536
+    },
+    {
+      "epoch": 0.2542764131399822,
+      "grad_norm": 1.6577744483947754,
+      "learning_rate": 1.9842953723941954e-05,
+      "loss": 0.2702,
+      "step": 537
+    },
+    {
+      "epoch": 0.2547499260136135,
+      "grad_norm": 1.4658805131912231,
+      "learning_rate": 1.9842049647650815e-05,
+      "loss": 0.2848,
+      "step": 538
+    },
+    {
+      "epoch": 0.25522343888724475,
+      "grad_norm": 1.4830657243728638,
+      "learning_rate": 1.984114299726654e-05,
+      "loss": 0.2536,
+      "step": 539
+    },
+    {
+      "epoch": 0.255696951760876,
+      "grad_norm": 1.5815433263778687,
+      "learning_rate": 1.984023377302625e-05,
+      "loss": 0.2853,
+      "step": 540
+    },
+    {
+      "epoch": 0.25617046463450727,
+      "grad_norm": 1.2159545421600342,
+      "learning_rate": 1.9839321975167747e-05,
+      "loss": 0.3015,
+      "step": 541
+    },
+    {
+      "epoch": 0.2566439775081385,
+      "grad_norm": 1.711661696434021,
+      "learning_rate": 1.9838407603929503e-05,
+      "loss": 0.2924,
+      "step": 542
+    },
+    {
+      "epoch": 0.25711749038176973,
+      "grad_norm": 1.3681540489196777,
+      "learning_rate": 1.9837490659550665e-05,
+      "loss": 0.2999,
+      "step": 543
+    },
+    {
+      "epoch": 0.257591003255401,
+      "grad_norm": 1.9190740585327148,
+      "learning_rate": 1.983657114227105e-05,
+      "loss": 0.3154,
+      "step": 544
+    },
+    {
+      "epoch": 0.25806451612903225,
+      "grad_norm": 1.5599863529205322,
+      "learning_rate": 1.9835649052331143e-05,
+      "loss": 0.2775,
+      "step": 545
+    },
+    {
+      "epoch": 0.2585380290026635,
+      "grad_norm": 1.3822370767593384,
+      "learning_rate": 1.9834724389972116e-05,
+      "loss": 0.2659,
+      "step": 546
+    },
+    {
+      "epoch": 0.25901154187629477,
+      "grad_norm": 1.2202012538909912,
+      "learning_rate": 1.98337971554358e-05,
+      "loss": 0.2835,
+      "step": 547
+    },
+    {
+      "epoch": 0.259485054749926,
+      "grad_norm": 1.225102186203003,
+      "learning_rate": 1.9832867348964707e-05,
+      "loss": 0.3081,
+      "step": 548
+    },
+    {
+      "epoch": 0.2599585676235573,
+      "grad_norm": 1.9328206777572632,
+      "learning_rate": 1.983193497080202e-05,
+      "loss": 0.2807,
+      "step": 549
+    },
+    {
+      "epoch": 0.2604320804971885,
+      "grad_norm": 3.2432518005371094,
+      "learning_rate": 1.9831000021191592e-05,
+      "loss": 0.3038,
+      "step": 550
+    },
+    {
+      "epoch": 0.26090559337081975,
+      "grad_norm": 1.950760841369629,
+      "learning_rate": 1.9830062500377945e-05,
+      "loss": 0.3174,
+      "step": 551
+    },
+    {
+      "epoch": 0.26137910624445104,
+      "grad_norm": 1.1726529598236084,
+      "learning_rate": 1.9829122408606288e-05,
+      "loss": 0.2757,
+      "step": 552
+    },
+    {
+      "epoch": 0.2618526191180823,
+      "grad_norm": 1.5435218811035156,
+      "learning_rate": 1.9828179746122487e-05,
+      "loss": 0.301,
+      "step": 553
+    },
+    {
+      "epoch": 0.2623261319917135,
+      "grad_norm": 2.2376794815063477,
+      "learning_rate": 1.9827234513173087e-05,
+      "loss": 0.2955,
+      "step": 554
+    },
+    {
+      "epoch": 0.2627996448653448,
+      "grad_norm": 2.062591791152954,
+      "learning_rate": 1.9826286710005305e-05,
+      "loss": 0.255,
+      "step": 555
+    },
+    {
+      "epoch": 0.263273157738976,
+      "grad_norm": 1.1453508138656616,
+      "learning_rate": 1.982533633686703e-05,
+      "loss": 0.2597,
+      "step": 556
+    },
+    {
+      "epoch": 0.26374667061260726,
+      "grad_norm": 1.465489387512207,
+      "learning_rate": 1.9824383394006825e-05,
+      "loss": 0.2905,
+      "step": 557
+    },
+    {
+      "epoch": 0.26422018348623855,
+      "grad_norm": 1.4714313745498657,
+      "learning_rate": 1.9823427881673916e-05,
+      "loss": 0.2584,
+      "step": 558
+    },
+    {
+      "epoch": 0.2646936963598698,
+      "grad_norm": 1.2892956733703613,
+      "learning_rate": 1.9822469800118215e-05,
+      "loss": 0.2792,
+      "step": 559
+    },
+    {
+      "epoch": 0.265167209233501,
+      "grad_norm": 1.3988784551620483,
+      "learning_rate": 1.98215091495903e-05,
+      "loss": 0.2749,
+      "step": 560
+    },
+    {
+      "epoch": 0.2656407221071323,
+      "grad_norm": 1.2862416505813599,
+      "learning_rate": 1.9820545930341413e-05,
+      "loss": 0.2765,
+      "step": 561
+    },
+    {
+      "epoch": 0.26611423498076353,
+      "grad_norm": 1.5326262712478638,
+      "learning_rate": 1.981958014262348e-05,
+      "loss": 0.3028,
+      "step": 562
+    },
+    {
+      "epoch": 0.2665877478543948,
+      "grad_norm": 0.968665361404419,
+      "learning_rate": 1.981861178668909e-05,
+      "loss": 0.255,
+      "step": 563
+    },
+    {
+      "epoch": 0.26706126072802605,
+      "grad_norm": 1.944764494895935,
+      "learning_rate": 1.981764086279151e-05,
+      "loss": 0.292,
+      "step": 564
+    },
+    {
+      "epoch": 0.2675347736016573,
+      "grad_norm": 1.2996660470962524,
+      "learning_rate": 1.9816667371184677e-05,
+      "loss": 0.2615,
+      "step": 565
+    },
+    {
+      "epoch": 0.26800828647528857,
+      "grad_norm": 1.362474799156189,
+      "learning_rate": 1.9815691312123194e-05,
+      "loss": 0.2889,
+      "step": 566
+    },
+    {
+      "epoch": 0.2684817993489198,
+      "grad_norm": 1.6865662336349487,
+      "learning_rate": 1.9814712685862342e-05,
+      "loss": 0.2745,
+      "step": 567
+    },
+    {
+      "epoch": 0.26895531222255104,
+      "grad_norm": 1.5344867706298828,
+      "learning_rate": 1.9813731492658073e-05,
+      "loss": 0.2901,
+      "step": 568
+    },
+    {
+      "epoch": 0.2694288250961823,
+      "grad_norm": 1.4177885055541992,
+      "learning_rate": 1.9812747732767006e-05,
+      "loss": 0.2864,
+      "step": 569
+    },
+    {
+      "epoch": 0.26990233796981355,
+      "grad_norm": 1.1646602153778076,
+      "learning_rate": 1.981176140644643e-05,
+      "loss": 0.2668,
+      "step": 570
+    },
+    {
+      "epoch": 0.2703758508434448,
+      "grad_norm": 1.417022943496704,
+      "learning_rate": 1.981077251395432e-05,
+      "loss": 0.2761,
+      "step": 571
+    },
+    {
+      "epoch": 0.2708493637170761,
+      "grad_norm": 1.7936999797821045,
+      "learning_rate": 1.9809781055549308e-05,
+      "loss": 0.2899,
+      "step": 572
+    },
+    {
+      "epoch": 0.2713228765907073,
+      "grad_norm": 2.242428779602051,
+      "learning_rate": 1.980878703149069e-05,
+      "loss": 0.2971,
+      "step": 573
+    },
+    {
+      "epoch": 0.27179638946433854,
+      "grad_norm": 1.8133538961410522,
+      "learning_rate": 1.980779044203845e-05,
+      "loss": 0.2789,
+      "step": 574
+    },
+    {
+      "epoch": 0.2722699023379698,
+      "grad_norm": 1.2497389316558838,
+      "learning_rate": 1.9806791287453245e-05,
+      "loss": 0.2885,
+      "step": 575
+    },
+    {
+      "epoch": 0.27274341521160106,
+      "grad_norm": 1.5333000421524048,
+      "learning_rate": 1.9805789567996384e-05,
+      "loss": 0.2917,
+      "step": 576
+    },
+    {
+      "epoch": 0.2732169280852323,
+      "grad_norm": 2.611602306365967,
+      "learning_rate": 1.9804785283929858e-05,
+      "loss": 0.2752,
+      "step": 577
+    },
+    {
+      "epoch": 0.2736904409588636,
+      "grad_norm": 1.3517531156539917,
+      "learning_rate": 1.9803778435516327e-05,
+      "loss": 0.2619,
+      "step": 578
+    },
+    {
+      "epoch": 0.2741639538324948,
+      "grad_norm": 1.452409267425537,
+      "learning_rate": 1.9802769023019128e-05,
+      "loss": 0.2735,
+      "step": 579
+    },
+    {
+      "epoch": 0.2746374667061261,
+      "grad_norm": 1.5018271207809448,
+      "learning_rate": 1.980175704670226e-05,
+      "loss": 0.3296,
+      "step": 580
+    },
+    {
+      "epoch": 0.27511097957975733,
+      "grad_norm": 2.2614331245422363,
+      "learning_rate": 1.9800742506830394e-05,
+      "loss": 0.2969,
+      "step": 581
+    },
+    {
+      "epoch": 0.27558449245338856,
+      "grad_norm": 1.6071243286132812,
+      "learning_rate": 1.979972540366888e-05,
+      "loss": 0.3061,
+      "step": 582
+    },
+    {
+      "epoch": 0.27605800532701985,
+      "grad_norm": 1.25821852684021,
+      "learning_rate": 1.979870573748372e-05,
+      "loss": 0.2715,
+      "step": 583
+    },
+    {
+      "epoch": 0.2765315182006511,
+      "grad_norm": 1.5205860137939453,
+      "learning_rate": 1.9797683508541606e-05,
+      "loss": 0.3016,
+      "step": 584
+    },
+    {
+      "epoch": 0.2770050310742823,
+      "grad_norm": 1.3796861171722412,
+      "learning_rate": 1.9796658717109892e-05,
+      "loss": 0.2868,
+      "step": 585
+    },
+    {
+      "epoch": 0.2774785439479136,
+      "grad_norm": 1.5328634977340698,
+      "learning_rate": 1.97956313634566e-05,
+      "loss": 0.2787,
+      "step": 586
+    },
+    {
+      "epoch": 0.27795205682154484,
+      "grad_norm": 1.6029144525527954,
+      "learning_rate": 1.9794601447850424e-05,
+      "loss": 0.2508,
+      "step": 587
+    },
+    {
+      "epoch": 0.27842556969517607,
+      "grad_norm": 1.3855677843093872,
+      "learning_rate": 1.9793568970560728e-05,
+      "loss": 0.2723,
+      "step": 588
+    },
+    {
+      "epoch": 0.27889908256880735,
+      "grad_norm": 1.2493582963943481,
+      "learning_rate": 1.979253393185755e-05,
+      "loss": 0.2589,
+      "step": 589
+    },
+    {
+      "epoch": 0.2793725954424386,
+      "grad_norm": 1.4009218215942383,
+      "learning_rate": 1.9791496332011593e-05,
+      "loss": 0.2998,
+      "step": 590
+    },
+    {
+      "epoch": 0.2798461083160698,
+      "grad_norm": 2.099388837814331,
+      "learning_rate": 1.979045617129423e-05,
+      "loss": 0.292,
+      "step": 591
+    },
+    {
+      "epoch": 0.2803196211897011,
+      "grad_norm": 1.3639984130859375,
+      "learning_rate": 1.9789413449977505e-05,
+      "loss": 0.2717,
+      "step": 592
+    },
+    {
+      "epoch": 0.28079313406333234,
+      "grad_norm": 1.2905452251434326,
+      "learning_rate": 1.9788368168334135e-05,
+      "loss": 0.245,
+      "step": 593
+    },
+    {
+      "epoch": 0.28126664693696357,
+      "grad_norm": 1.319740653038025,
+      "learning_rate": 1.9787320326637494e-05,
+      "loss": 0.259,
+      "step": 594
+    },
+    {
+      "epoch": 0.28174015981059486,
+      "grad_norm": 1.6142657995224,
+      "learning_rate": 1.9786269925161646e-05,
+      "loss": 0.2819,
+      "step": 595
+    },
+    {
+      "epoch": 0.2822136726842261,
+      "grad_norm": 1.273977279663086,
+      "learning_rate": 1.978521696418131e-05,
+      "loss": 0.2574,
+      "step": 596
+    },
+    {
+      "epoch": 0.2826871855578574,
+      "grad_norm": 1.7523685693740845,
+      "learning_rate": 1.9784161443971878e-05,
+      "loss": 0.2703,
+      "step": 597
+    },
+    {
+      "epoch": 0.2831606984314886,
+      "grad_norm": 1.5006474256515503,
+      "learning_rate": 1.9783103364809405e-05,
+      "loss": 0.3043,
+      "step": 598
+    },
+    {
+      "epoch": 0.28363421130511984,
+      "grad_norm": 1.1372324228286743,
+      "learning_rate": 1.978204272697063e-05,
+      "loss": 0.2793,
+      "step": 599
+    },
+    {
+      "epoch": 0.28410772417875113,
+      "grad_norm": 1.5428109169006348,
+      "learning_rate": 1.9780979530732947e-05,
+      "loss": 0.2455,
+      "step": 600
+    },
+    {
+      "epoch": 0.28458123705238236,
+      "grad_norm": 1.6741421222686768,
+      "learning_rate": 1.9779913776374427e-05,
+      "loss": 0.2767,
+      "step": 601
+    },
+    {
+      "epoch": 0.2850547499260136,
+      "grad_norm": 1.1833195686340332,
+      "learning_rate": 1.9778845464173805e-05,
+      "loss": 0.2639,
+      "step": 602
+    },
+    {
+      "epoch": 0.2855282627996449,
+      "grad_norm": 1.1689391136169434,
+      "learning_rate": 1.9777774594410495e-05,
+      "loss": 0.2601,
+      "step": 603
+    },
+    {
+      "epoch": 0.2860017756732761,
+      "grad_norm": 1.502760887145996,
+      "learning_rate": 1.9776701167364565e-05,
+      "loss": 0.3045,
+      "step": 604
+    },
+    {
+      "epoch": 0.28647528854690735,
+      "grad_norm": 1.974056601524353,
+      "learning_rate": 1.977562518331676e-05,
+      "loss": 0.2785,
+      "step": 605
+    },
+    {
+      "epoch": 0.28694880142053864,
+      "grad_norm": 1.409839391708374,
+      "learning_rate": 1.97745466425485e-05,
+      "loss": 0.2732,
+      "step": 606
+    },
+    {
+      "epoch": 0.28742231429416987,
+      "grad_norm": 1.0581002235412598,
+      "learning_rate": 1.9773465545341855e-05,
+      "loss": 0.2569,
+      "step": 607
+    },
+    {
+      "epoch": 0.2878958271678011,
+      "grad_norm": 1.3603756427764893,
+      "learning_rate": 1.977238189197959e-05,
+      "loss": 0.2536,
+      "step": 608
+    },
+    {
+      "epoch": 0.2883693400414324,
+      "grad_norm": 1.2487221956253052,
+      "learning_rate": 1.9771295682745115e-05,
+      "loss": 0.2998,
+      "step": 609
+    },
+    {
+      "epoch": 0.2888428529150636,
+      "grad_norm": 1.1602859497070312,
+      "learning_rate": 1.977020691792252e-05,
+      "loss": 0.2476,
+      "step": 610
+    },
+    {
+      "epoch": 0.2893163657886949,
+      "grad_norm": 1.3260509967803955,
+      "learning_rate": 1.976911559779656e-05,
+      "loss": 0.2972,
+      "step": 611
+    },
+    {
+      "epoch": 0.28978987866232614,
+      "grad_norm": 1.3510822057724,
+      "learning_rate": 1.976802172265266e-05,
+      "loss": 0.2658,
+      "step": 612
+    },
+    {
+      "epoch": 0.29026339153595737,
+      "grad_norm": 2.0700032711029053,
+      "learning_rate": 1.976692529277691e-05,
+      "loss": 0.2471,
+      "step": 613
+    },
+    {
+      "epoch": 0.29073690440958866,
+      "grad_norm": 1.5847715139389038,
+      "learning_rate": 1.9765826308456075e-05,
+      "loss": 0.2757,
+      "step": 614
+    },
+    {
+      "epoch": 0.2912104172832199,
+      "grad_norm": 1.1467682123184204,
+      "learning_rate": 1.976472476997758e-05,
+      "loss": 0.2733,
+      "step": 615
+    },
+    {
+      "epoch": 0.2916839301568511,
+      "grad_norm": 1.1180078983306885,
+      "learning_rate": 1.9763620677629525e-05,
+      "loss": 0.2494,
+      "step": 616
+    },
+    {
+      "epoch": 0.2921574430304824,
+      "grad_norm": 2.1236302852630615,
+      "learning_rate": 1.9762514031700673e-05,
+      "loss": 0.2934,
+      "step": 617
+    },
+    {
+      "epoch": 0.29263095590411364,
+      "grad_norm": 1.3528096675872803,
+      "learning_rate": 1.9761404832480455e-05,
+      "loss": 0.2811,
+      "step": 618
+    },
+    {
+      "epoch": 0.2931044687777449,
+      "grad_norm": 1.4439085721969604,
+      "learning_rate": 1.9760293080258976e-05,
+      "loss": 0.2829,
+      "step": 619
+    },
+    {
+      "epoch": 0.29357798165137616,
+      "grad_norm": 1.2082427740097046,
+      "learning_rate": 1.9759178775327e-05,
+      "loss": 0.2781,
+      "step": 620
+    },
+    {
+      "epoch": 0.2940514945250074,
+      "grad_norm": 2.3843281269073486,
+      "learning_rate": 1.975806191797596e-05,
+      "loss": 0.2968,
+      "step": 621
+    },
+    {
+      "epoch": 0.2945250073986386,
+      "grad_norm": 1.1950457096099854,
+      "learning_rate": 1.9756942508497967e-05,
+      "loss": 0.2816,
+      "step": 622
+    },
+    {
+      "epoch": 0.2949985202722699,
+      "grad_norm": 1.4412981271743774,
+      "learning_rate": 1.9755820547185787e-05,
+      "loss": 0.2838,
+      "step": 623
+    },
+    {
+      "epoch": 0.29547203314590115,
+      "grad_norm": 1.1725237369537354,
+      "learning_rate": 1.975469603433286e-05,
+      "loss": 0.2596,
+      "step": 624
+    },
+    {
+      "epoch": 0.2959455460195324,
+      "grad_norm": 1.1337695121765137,
+      "learning_rate": 1.975356897023329e-05,
+      "loss": 0.2647,
+      "step": 625
+    },
+    {
+      "epoch": 0.29641905889316367,
+      "grad_norm": 1.444916009902954,
+      "learning_rate": 1.9752439355181848e-05,
+      "loss": 0.2679,
+      "step": 626
+    },
+    {
+      "epoch": 0.2968925717667949,
+      "grad_norm": 1.7320390939712524,
+      "learning_rate": 1.975130718947398e-05,
+      "loss": 0.2317,
+      "step": 627
+    },
+    {
+      "epoch": 0.2973660846404262,
+      "grad_norm": 2.060805320739746,
+      "learning_rate": 1.9750172473405785e-05,
+      "loss": 0.269,
+      "step": 628
+    },
+    {
+      "epoch": 0.2978395975140574,
+      "grad_norm": 1.348092794418335,
+      "learning_rate": 1.9749035207274044e-05,
+      "loss": 0.2658,
+      "step": 629
+    },
+    {
+      "epoch": 0.29831311038768865,
+      "grad_norm": 1.3302925825119019,
+      "learning_rate": 1.9747895391376192e-05,
+      "loss": 0.2504,
+      "step": 630
+    },
+    {
+      "epoch": 0.29878662326131994,
+      "grad_norm": 1.1984549760818481,
+      "learning_rate": 1.9746753026010342e-05,
+      "loss": 0.2713,
+      "step": 631
+    },
+    {
+      "epoch": 0.29926013613495117,
+      "grad_norm": 1.1172891855239868,
+      "learning_rate": 1.9745608111475267e-05,
+      "loss": 0.2636,
+      "step": 632
+    },
+    {
+      "epoch": 0.2997336490085824,
+      "grad_norm": 1.5490065813064575,
+      "learning_rate": 1.9744460648070408e-05,
+      "loss": 0.2618,
+      "step": 633
+    },
+    {
+      "epoch": 0.3002071618822137,
+      "grad_norm": 1.167188048362732,
+      "learning_rate": 1.974331063609587e-05,
+      "loss": 0.2601,
+      "step": 634
+    },
+    {
+      "epoch": 0.3006806747558449,
+      "grad_norm": 1.5427215099334717,
+      "learning_rate": 1.9742158075852435e-05,
+      "loss": 0.281,
+      "step": 635
+    },
+    {
+      "epoch": 0.30115418762947616,
+      "grad_norm": 1.17001211643219,
+      "learning_rate": 1.9741002967641537e-05,
+      "loss": 0.2592,
+      "step": 636
+    },
+    {
+      "epoch": 0.30162770050310744,
+      "grad_norm": 1.1877490282058716,
+      "learning_rate": 1.9739845311765286e-05,
+      "loss": 0.3065,
+      "step": 637
+    },
+    {
+      "epoch": 0.3021012133767387,
+      "grad_norm": 1.3651347160339355,
+      "learning_rate": 1.9738685108526456e-05,
+      "loss": 0.2801,
+      "step": 638
+    },
+    {
+      "epoch": 0.3025747262503699,
+      "grad_norm": 1.283866047859192,
+      "learning_rate": 1.9737522358228487e-05,
+      "loss": 0.2939,
+      "step": 639
+    },
+    {
+      "epoch": 0.3030482391240012,
+      "grad_norm": 1.2590893507003784,
+      "learning_rate": 1.9736357061175483e-05,
+      "loss": 0.2657,
+      "step": 640
+    },
+    {
+      "epoch": 0.3035217519976324,
+      "grad_norm": 1.178700566291809,
+      "learning_rate": 1.973518921767222e-05,
+      "loss": 0.2618,
+      "step": 641
+    },
+    {
+      "epoch": 0.3039952648712637,
+      "grad_norm": 1.7493693828582764,
+      "learning_rate": 1.9734018828024136e-05,
+      "loss": 0.2763,
+      "step": 642
+    },
+    {
+      "epoch": 0.30446877774489495,
+      "grad_norm": 1.9666167497634888,
+      "learning_rate": 1.973284589253733e-05,
+      "loss": 0.2826,
+      "step": 643
+    },
+    {
+      "epoch": 0.3049422906185262,
+      "grad_norm": 1.3451874256134033,
+      "learning_rate": 1.9731670411518578e-05,
+      "loss": 0.2642,
+      "step": 644
+    },
+    {
+      "epoch": 0.30541580349215747,
+      "grad_norm": 1.3448307514190674,
+      "learning_rate": 1.973049238527531e-05,
+      "loss": 0.2672,
+      "step": 645
+    },
+    {
+      "epoch": 0.3058893163657887,
+      "grad_norm": 1.249316692352295,
+      "learning_rate": 1.9729311814115632e-05,
+      "loss": 0.2564,
+      "step": 646
+    },
+    {
+      "epoch": 0.30636282923941993,
+      "grad_norm": 1.4523890018463135,
+      "learning_rate": 1.972812869834831e-05,
+      "loss": 0.2831,
+      "step": 647
+    },
+    {
+      "epoch": 0.3068363421130512,
+      "grad_norm": 1.7998663187026978,
+      "learning_rate": 1.9726943038282772e-05,
+      "loss": 0.2962,
+      "step": 648
+    },
+    {
+      "epoch": 0.30730985498668245,
+      "grad_norm": 1.393811583518982,
+      "learning_rate": 1.9725754834229124e-05,
+      "loss": 0.2553,
+      "step": 649
+    },
+    {
+      "epoch": 0.3077833678603137,
+      "grad_norm": 1.6593152284622192,
+      "learning_rate": 1.9724564086498123e-05,
+      "loss": 0.3102,
+      "step": 650
+    },
+    {
+      "epoch": 0.30825688073394497,
+      "grad_norm": 1.6688569784164429,
+      "learning_rate": 1.97233707954012e-05,
+      "loss": 0.2706,
+      "step": 651
+    },
+    {
+      "epoch": 0.3087303936075762,
+      "grad_norm": 2.265216588973999,
+      "learning_rate": 1.972217496125045e-05,
+      "loss": 0.2884,
+      "step": 652
+    },
+    {
+      "epoch": 0.30920390648120744,
+      "grad_norm": 1.522386908531189,
+      "learning_rate": 1.972097658435863e-05,
+      "loss": 0.2754,
+      "step": 653
+    },
+    {
+      "epoch": 0.3096774193548387,
+      "grad_norm": 1.4247666597366333,
+      "learning_rate": 1.9719775665039162e-05,
+      "loss": 0.2557,
+      "step": 654
+    },
+    {
+      "epoch": 0.31015093222846996,
+      "grad_norm": 1.1649177074432373,
+      "learning_rate": 1.971857220360614e-05,
+      "loss": 0.2608,
+      "step": 655
+    },
+    {
+      "epoch": 0.3106244451021012,
+      "grad_norm": 2.187192678451538,
+      "learning_rate": 1.9717366200374313e-05,
+      "loss": 0.2989,
+      "step": 656
+    },
+    {
+      "epoch": 0.3110979579757325,
+      "grad_norm": 1.4517701864242554,
+      "learning_rate": 1.9716157655659102e-05,
+      "loss": 0.2681,
+      "step": 657
+    },
+    {
+      "epoch": 0.3115714708493637,
+      "grad_norm": 1.8214770555496216,
+      "learning_rate": 1.971494656977659e-05,
+      "loss": 0.241,
+      "step": 658
+    },
+    {
+      "epoch": 0.312044983722995,
+      "grad_norm": 1.289170742034912,
+      "learning_rate": 1.9713732943043524e-05,
+      "loss": 0.2659,
+      "step": 659
+    },
+    {
+      "epoch": 0.3125184965966262,
+      "grad_norm": 1.3986639976501465,
+      "learning_rate": 1.9712516775777315e-05,
+      "loss": 0.2649,
+      "step": 660
+    },
+    {
+      "epoch": 0.31299200947025746,
+      "grad_norm": 1.7374229431152344,
+      "learning_rate": 1.9711298068296046e-05,
+      "loss": 0.2902,
+      "step": 661
+    },
+    {
+      "epoch": 0.31346552234388875,
+      "grad_norm": 1.802095890045166,
+      "learning_rate": 1.9710076820918448e-05,
+      "loss": 0.2515,
+      "step": 662
+    },
+    {
+      "epoch": 0.31393903521752,
+      "grad_norm": 1.2375832796096802,
+      "learning_rate": 1.9708853033963936e-05,
+      "loss": 0.267,
+      "step": 663
+    },
+    {
+      "epoch": 0.3144125480911512,
+      "grad_norm": 1.2813341617584229,
+      "learning_rate": 1.9707626707752574e-05,
+      "loss": 0.2743,
+      "step": 664
+    },
+    {
+      "epoch": 0.3148860609647825,
+      "grad_norm": 1.1700478792190552,
+      "learning_rate": 1.97063978426051e-05,
+      "loss": 0.2724,
+      "step": 665
+    },
+    {
+      "epoch": 0.31535957383841373,
+      "grad_norm": 2.465041160583496,
+      "learning_rate": 1.9705166438842907e-05,
+      "loss": 0.2598,
+      "step": 666
+    },
+    {
+      "epoch": 0.31583308671204496,
+      "grad_norm": 1.6531705856323242,
+      "learning_rate": 1.9703932496788063e-05,
+      "loss": 0.2889,
+      "step": 667
+    },
+    {
+      "epoch": 0.31630659958567625,
+      "grad_norm": 1.5581059455871582,
+      "learning_rate": 1.9702696016763286e-05,
+      "loss": 0.2985,
+      "step": 668
+    },
+    {
+      "epoch": 0.3167801124593075,
+      "grad_norm": 1.4005812406539917,
+      "learning_rate": 1.9701456999091974e-05,
+      "loss": 0.2475,
+      "step": 669
+    },
+    {
+      "epoch": 0.3172536253329387,
+      "grad_norm": 1.6382842063903809,
+      "learning_rate": 1.970021544409817e-05,
+      "loss": 0.2683,
+      "step": 670
+    },
+    {
+      "epoch": 0.31772713820657,
+      "grad_norm": 1.9983477592468262,
+      "learning_rate": 1.96989713521066e-05,
+      "loss": 0.2836,
+      "step": 671
+    },
+    {
+      "epoch": 0.31820065108020124,
+      "grad_norm": 1.6255732774734497,
+      "learning_rate": 1.9697724723442643e-05,
+      "loss": 0.2627,
+      "step": 672
+    },
+    {
+      "epoch": 0.31867416395383247,
+      "grad_norm": 1.2218822240829468,
+      "learning_rate": 1.9696475558432334e-05,
+      "loss": 0.2735,
+      "step": 673
+    },
+    {
+      "epoch": 0.31914767682746376,
+      "grad_norm": 1.356476902961731,
+      "learning_rate": 1.969522385740239e-05,
+      "loss": 0.2905,
+      "step": 674
+    },
+    {
+      "epoch": 0.319621189701095,
+      "grad_norm": 1.8103471994400024,
+      "learning_rate": 1.9693969620680177e-05,
+      "loss": 0.2909,
+      "step": 675
+    },
+    {
+      "epoch": 0.3200947025747263,
+      "grad_norm": 1.908302903175354,
+      "learning_rate": 1.9692712848593726e-05,
+      "loss": 0.2738,
+      "step": 676
+    },
+    {
+      "epoch": 0.3205682154483575,
+      "grad_norm": 1.2977486848831177,
+      "learning_rate": 1.969145354147174e-05,
+      "loss": 0.2518,
+      "step": 677
+    },
+    {
+      "epoch": 0.32104172832198874,
+      "grad_norm": 1.6458895206451416,
+      "learning_rate": 1.9690191699643575e-05,
+      "loss": 0.2707,
+      "step": 678
+    },
+    {
+      "epoch": 0.32151524119562,
+      "grad_norm": 2.108949661254883,
+      "learning_rate": 1.9688927323439254e-05,
+      "loss": 0.2894,
+      "step": 679
+    },
+    {
+      "epoch": 0.32198875406925126,
+      "grad_norm": 1.1589833498001099,
+      "learning_rate": 1.9687660413189463e-05,
+      "loss": 0.255,
+      "step": 680
+    },
+    {
+      "epoch": 0.3224622669428825,
+      "grad_norm": 1.281614899635315,
+      "learning_rate": 1.9686390969225545e-05,
+      "loss": 0.2743,
+      "step": 681
+    },
+    {
+      "epoch": 0.3229357798165138,
+      "grad_norm": 1.3841043710708618,
+      "learning_rate": 1.968511899187952e-05,
+      "loss": 0.2668,
+      "step": 682
+    },
+    {
+      "epoch": 0.323409292690145,
+      "grad_norm": 1.8176034688949585,
+      "learning_rate": 1.9683844481484054e-05,
+      "loss": 0.2286,
+      "step": 683
+    },
+    {
+      "epoch": 0.32388280556377624,
+      "grad_norm": 2.2643625736236572,
+      "learning_rate": 1.9682567438372486e-05,
+      "loss": 0.2619,
+      "step": 684
+    },
+    {
+      "epoch": 0.32435631843740753,
+      "grad_norm": 2.1710567474365234,
+      "learning_rate": 1.9681287862878818e-05,
+      "loss": 0.2898,
+      "step": 685
+    },
+    {
+      "epoch": 0.32482983131103876,
+      "grad_norm": 1.4091668128967285,
+      "learning_rate": 1.9680005755337702e-05,
+      "loss": 0.2515,
+      "step": 686
+    },
+    {
+      "epoch": 0.32530334418467,
+      "grad_norm": 1.261533260345459,
+      "learning_rate": 1.9678721116084465e-05,
+      "loss": 0.2553,
+      "step": 687
+    },
+    {
+      "epoch": 0.3257768570583013,
+      "grad_norm": 1.412940263748169,
+      "learning_rate": 1.9677433945455092e-05,
+      "loss": 0.3054,
+      "step": 688
+    },
+    {
+      "epoch": 0.3262503699319325,
+      "grad_norm": 1.3808501958847046,
+      "learning_rate": 1.9676144243786236e-05,
+      "loss": 0.2611,
+      "step": 689
+    },
+    {
+      "epoch": 0.3267238828055638,
+      "grad_norm": 1.1639057397842407,
+      "learning_rate": 1.9674852011415194e-05,
+      "loss": 0.2882,
+      "step": 690
+    },
+    {
+      "epoch": 0.32719739567919504,
+      "grad_norm": 1.1630046367645264,
+      "learning_rate": 1.9673557248679945e-05,
+      "loss": 0.2231,
+      "step": 691
+    },
+    {
+      "epoch": 0.32767090855282627,
+      "grad_norm": 1.795854091644287,
+      "learning_rate": 1.9672259955919123e-05,
+      "loss": 0.3062,
+      "step": 692
+    },
+    {
+      "epoch": 0.32814442142645756,
+      "grad_norm": 1.4758808612823486,
+      "learning_rate": 1.967096013347202e-05,
+      "loss": 0.277,
+      "step": 693
+    },
+    {
+      "epoch": 0.3286179343000888,
+      "grad_norm": 1.5420145988464355,
+      "learning_rate": 1.9669657781678587e-05,
+      "loss": 0.2952,
+      "step": 694
+    },
+    {
+      "epoch": 0.32909144717372,
+      "grad_norm": 1.7470226287841797,
+      "learning_rate": 1.9668352900879447e-05,
+      "loss": 0.2645,
+      "step": 695
+    },
+    {
+      "epoch": 0.3295649600473513,
+      "grad_norm": 1.6357449293136597,
+      "learning_rate": 1.9667045491415878e-05,
+      "loss": 0.2714,
+      "step": 696
+    },
+    {
+      "epoch": 0.33003847292098254,
+      "grad_norm": 2.116548776626587,
+      "learning_rate": 1.9665735553629824e-05,
+      "loss": 0.2769,
+      "step": 697
+    },
+    {
+      "epoch": 0.33051198579461377,
+      "grad_norm": 2.229929208755493,
+      "learning_rate": 1.966442308786388e-05,
+      "loss": 0.2505,
+      "step": 698
+    },
+    {
+      "epoch": 0.33098549866824506,
+      "grad_norm": 1.3335548639297485,
+      "learning_rate": 1.966310809446131e-05,
+      "loss": 0.2516,
+      "step": 699
+    },
+    {
+      "epoch": 0.3314590115418763,
+      "grad_norm": 1.3416866064071655,
+      "learning_rate": 1.9661790573766046e-05,
+      "loss": 0.2563,
+      "step": 700
+    },
+    {
+      "epoch": 0.3319325244155075,
+      "grad_norm": 1.5719635486602783,
+      "learning_rate": 1.966047052612266e-05,
+      "loss": 0.2644,
+      "step": 701
+    },
+    {
+      "epoch": 0.3324060372891388,
+      "grad_norm": 1.4301633834838867,
+      "learning_rate": 1.9659147951876407e-05,
+      "loss": 0.2599,
+      "step": 702
+    },
+    {
+      "epoch": 0.33287955016277004,
+      "grad_norm": 1.6425291299819946,
+      "learning_rate": 1.965782285137319e-05,
+      "loss": 0.2422,
+      "step": 703
+    },
+    {
+      "epoch": 0.3333530630364013,
+      "grad_norm": 1.3709927797317505,
+      "learning_rate": 1.9656495224959578e-05,
+      "loss": 0.2653,
+      "step": 704
+    },
+    {
+      "epoch": 0.33382657591003256,
+      "grad_norm": 2.5719616413116455,
+      "learning_rate": 1.9655165072982797e-05,
+      "loss": 0.2581,
+      "step": 705
+    },
+    {
+      "epoch": 0.3343000887836638,
+      "grad_norm": 2.397643804550171,
+      "learning_rate": 1.9653832395790733e-05,
+      "loss": 0.2737,
+      "step": 706
+    },
+    {
+      "epoch": 0.3347736016572951,
+      "grad_norm": 1.7374242544174194,
+      "learning_rate": 1.965249719373194e-05,
+      "loss": 0.2835,
+      "step": 707
+    },
+    {
+      "epoch": 0.3352471145309263,
+      "grad_norm": 1.2840518951416016,
+      "learning_rate": 1.965115946715563e-05,
+      "loss": 0.2673,
+      "step": 708
+    },
+    {
+      "epoch": 0.33572062740455755,
+      "grad_norm": 1.8667082786560059,
+      "learning_rate": 1.964981921641166e-05,
+      "loss": 0.2596,
+      "step": 709
+    },
+    {
+      "epoch": 0.33619414027818884,
+      "grad_norm": 1.0804299116134644,
+      "learning_rate": 1.9648476441850574e-05,
+      "loss": 0.2432,
+      "step": 710
+    },
+    {
+      "epoch": 0.33666765315182007,
+      "grad_norm": 1.5073543787002563,
+      "learning_rate": 1.964713114382355e-05,
+      "loss": 0.2594,
+      "step": 711
+    },
+    {
+      "epoch": 0.3371411660254513,
+      "grad_norm": 1.724390983581543,
+      "learning_rate": 1.9645783322682447e-05,
+      "loss": 0.251,
+      "step": 712
+    },
+    {
+      "epoch": 0.3376146788990826,
+      "grad_norm": 1.4915387630462646,
+      "learning_rate": 1.964443297877977e-05,
+      "loss": 0.2637,
+      "step": 713
+    },
+    {
+      "epoch": 0.3380881917727138,
+      "grad_norm": 1.8173447847366333,
+      "learning_rate": 1.9643080112468683e-05,
+      "loss": 0.2878,
+      "step": 714
+    },
+    {
+      "epoch": 0.33856170464634505,
+      "grad_norm": 1.6663117408752441,
+      "learning_rate": 1.9641724724103026e-05,
+      "loss": 0.2929,
+      "step": 715
+    },
+    {
+      "epoch": 0.33903521751997634,
+      "grad_norm": 1.1947954893112183,
+      "learning_rate": 1.9640366814037283e-05,
+      "loss": 0.2699,
+      "step": 716
+    },
+    {
+      "epoch": 0.33950873039360757,
+      "grad_norm": 1.2085504531860352,
+      "learning_rate": 1.96390063826266e-05,
+      "loss": 0.258,
+      "step": 717
+    },
+    {
+      "epoch": 0.3399822432672388,
+      "grad_norm": 1.3036900758743286,
+      "learning_rate": 1.963764343022679e-05,
+      "loss": 0.283,
+      "step": 718
+    },
+    {
+      "epoch": 0.3404557561408701,
+      "grad_norm": 2.1460154056549072,
+      "learning_rate": 1.9636277957194316e-05,
+      "loss": 0.2811,
+      "step": 719
+    },
+    {
+      "epoch": 0.3409292690145013,
+      "grad_norm": 2.7816526889801025,
+      "learning_rate": 1.9634909963886304e-05,
+      "loss": 0.2436,
+      "step": 720
+    },
+    {
+      "epoch": 0.34140278188813256,
+      "grad_norm": 1.8817825317382812,
+      "learning_rate": 1.963353945066054e-05,
+      "loss": 0.2644,
+      "step": 721
+    },
+    {
+      "epoch": 0.34187629476176384,
+      "grad_norm": 1.5300520658493042,
+      "learning_rate": 1.963216641787547e-05,
+      "loss": 0.2729,
+      "step": 722
+    },
+    {
+      "epoch": 0.3423498076353951,
+      "grad_norm": 1.8480312824249268,
+      "learning_rate": 1.9630790865890196e-05,
+      "loss": 0.3109,
+      "step": 723
+    },
+    {
+      "epoch": 0.34282332050902636,
+      "grad_norm": 2.683523178100586,
+      "learning_rate": 1.9629412795064482e-05,
+      "loss": 0.2516,
+      "step": 724
+    },
+    {
+      "epoch": 0.3432968333826576,
+      "grad_norm": 1.2851934432983398,
+      "learning_rate": 1.9628032205758746e-05,
+      "loss": 0.271,
+      "step": 725
+    },
+    {
+      "epoch": 0.34377034625628883,
+      "grad_norm": 1.8185439109802246,
+      "learning_rate": 1.962664909833407e-05,
+      "loss": 0.2953,
+      "step": 726
+    },
+    {
+      "epoch": 0.3442438591299201,
+      "grad_norm": 1.4006390571594238,
+      "learning_rate": 1.9625263473152193e-05,
+      "loss": 0.2656,
+      "step": 727
+    },
+    {
+      "epoch": 0.34471737200355135,
+      "grad_norm": 1.353481650352478,
+      "learning_rate": 1.962387533057551e-05,
+      "loss": 0.2615,
+      "step": 728
+    },
+    {
+      "epoch": 0.3451908848771826,
+      "grad_norm": 1.9150645732879639,
+      "learning_rate": 1.9622484670967083e-05,
+      "loss": 0.2629,
+      "step": 729
+    },
+    {
+      "epoch": 0.34566439775081387,
+      "grad_norm": 1.6772758960723877,
+      "learning_rate": 1.9621091494690616e-05,
+      "loss": 0.2406,
+      "step": 730
+    },
+    {
+      "epoch": 0.3461379106244451,
+      "grad_norm": 1.7029880285263062,
+      "learning_rate": 1.9619695802110485e-05,
+      "loss": 0.2548,
+      "step": 731
+    },
+    {
+      "epoch": 0.34661142349807633,
+      "grad_norm": 1.2326855659484863,
+      "learning_rate": 1.961829759359172e-05,
+      "loss": 0.2976,
+      "step": 732
+    },
+    {
+      "epoch": 0.3470849363717076,
+      "grad_norm": 1.2379788160324097,
+      "learning_rate": 1.961689686950001e-05,
+      "loss": 0.2472,
+      "step": 733
+    },
+    {
+      "epoch": 0.34755844924533885,
+      "grad_norm": 2.5410702228546143,
+      "learning_rate": 1.9615493630201694e-05,
+      "loss": 0.2827,
+      "step": 734
+    },
+    {
+      "epoch": 0.3480319621189701,
+      "grad_norm": 2.657710075378418,
+      "learning_rate": 1.961408787606379e-05,
+      "loss": 0.2616,
+      "step": 735
+    },
+    {
+      "epoch": 0.34850547499260137,
+      "grad_norm": 1.4291493892669678,
+      "learning_rate": 1.9612679607453942e-05,
+      "loss": 0.2978,
+      "step": 736
+    },
+    {
+      "epoch": 0.3489789878662326,
+      "grad_norm": 1.5034074783325195,
+      "learning_rate": 1.9611268824740482e-05,
+      "loss": 0.2616,
+      "step": 737
+    },
+    {
+      "epoch": 0.3494525007398639,
+      "grad_norm": 1.3025211095809937,
+      "learning_rate": 1.9609855528292386e-05,
+      "loss": 0.2704,
+      "step": 738
+    },
+    {
+      "epoch": 0.3499260136134951,
+      "grad_norm": 2.3469293117523193,
+      "learning_rate": 1.960843971847928e-05,
+      "loss": 0.3042,
+      "step": 739
+    },
+    {
+      "epoch": 0.35039952648712636,
+      "grad_norm": 2.3904573917388916,
+      "learning_rate": 1.960702139567146e-05,
+      "loss": 0.2687,
+      "step": 740
+    },
+    {
+      "epoch": 0.35087303936075764,
+      "grad_norm": 1.4793574810028076,
+      "learning_rate": 1.9605600560239874e-05,
+      "loss": 0.2495,
+      "step": 741
+    },
+    {
+      "epoch": 0.3513465522343889,
+      "grad_norm": 1.2374844551086426,
+      "learning_rate": 1.960417721255613e-05,
+      "loss": 0.2791,
+      "step": 742
+    },
+    {
+      "epoch": 0.3518200651080201,
+      "grad_norm": 2.1964030265808105,
+      "learning_rate": 1.960275135299249e-05,
+      "loss": 0.2738,
+      "step": 743
+    },
+    {
+      "epoch": 0.3522935779816514,
+      "grad_norm": 1.7759153842926025,
+      "learning_rate": 1.9601322981921872e-05,
+      "loss": 0.2436,
+      "step": 744
+    },
+    {
+      "epoch": 0.35276709085528263,
+      "grad_norm": 1.5018988847732544,
+      "learning_rate": 1.959989209971785e-05,
+      "loss": 0.2935,
+      "step": 745
+    },
+    {
+      "epoch": 0.35324060372891386,
+      "grad_norm": 1.1463077068328857,
+      "learning_rate": 1.959845870675467e-05,
+      "loss": 0.2613,
+      "step": 746
+    },
+    {
+      "epoch": 0.35371411660254515,
+      "grad_norm": 1.4144256114959717,
+      "learning_rate": 1.9597022803407206e-05,
+      "loss": 0.2337,
+      "step": 747
+    },
+    {
+      "epoch": 0.3541876294761764,
+      "grad_norm": 1.2923469543457031,
+      "learning_rate": 1.9595584390051014e-05,
+      "loss": 0.2449,
+      "step": 748
+    },
+    {
+      "epoch": 0.3546611423498076,
+      "grad_norm": 2.338681936264038,
+      "learning_rate": 1.9594143467062295e-05,
+      "loss": 0.2694,
+      "step": 749
+    },
+    {
+      "epoch": 0.3551346552234389,
+      "grad_norm": 2.0817441940307617,
+      "learning_rate": 1.9592700034817906e-05,
+      "loss": 0.3,
+      "step": 750
+    },
+    {
+      "epoch": 0.35560816809707013,
+      "grad_norm": 1.2521568536758423,
+      "learning_rate": 1.959125409369537e-05,
+      "loss": 0.2664,
+      "step": 751
+    },
+    {
+      "epoch": 0.35608168097070136,
+      "grad_norm": 1.5073604583740234,
+      "learning_rate": 1.958980564407285e-05,
+      "loss": 0.2696,
+      "step": 752
+    },
+    {
+      "epoch": 0.35655519384433265,
+      "grad_norm": 1.10438871383667,
+      "learning_rate": 1.9588354686329182e-05,
+      "loss": 0.2712,
+      "step": 753
+    },
+    {
+      "epoch": 0.3570287067179639,
+      "grad_norm": 1.8521367311477661,
+      "learning_rate": 1.9586901220843844e-05,
+      "loss": 0.2802,
+      "step": 754
+    },
+    {
+      "epoch": 0.35750221959159517,
+      "grad_norm": 2.0266237258911133,
+      "learning_rate": 1.958544524799698e-05,
+      "loss": 0.2608,
+      "step": 755
+    },
+    {
+      "epoch": 0.3579757324652264,
+      "grad_norm": 1.1580361127853394,
+      "learning_rate": 1.958398676816938e-05,
+      "loss": 0.2582,
+      "step": 756
+    },
+    {
+      "epoch": 0.35844924533885764,
+      "grad_norm": 1.5954866409301758,
+      "learning_rate": 1.9582525781742502e-05,
+      "loss": 0.2436,
+      "step": 757
+    },
+    {
+      "epoch": 0.3589227582124889,
+      "grad_norm": 1.2600346803665161,
+      "learning_rate": 1.9581062289098448e-05,
+      "loss": 0.2658,
+      "step": 758
+    },
+    {
+      "epoch": 0.35939627108612016,
+      "grad_norm": 1.5517022609710693,
+      "learning_rate": 1.9579596290619986e-05,
+      "loss": 0.2886,
+      "step": 759
+    },
+    {
+      "epoch": 0.3598697839597514,
+      "grad_norm": 1.9510260820388794,
+      "learning_rate": 1.9578127786690532e-05,
+      "loss": 0.2839,
+      "step": 760
+    },
+    {
+      "epoch": 0.3603432968333827,
+      "grad_norm": 1.2476470470428467,
+      "learning_rate": 1.957665677769415e-05,
+      "loss": 0.2668,
+      "step": 761
+    },
+    {
+      "epoch": 0.3608168097070139,
+      "grad_norm": 1.4977624416351318,
+      "learning_rate": 1.9575183264015577e-05,
+      "loss": 0.2613,
+      "step": 762
+    },
+    {
+      "epoch": 0.36129032258064514,
+      "grad_norm": 1.4652796983718872,
+      "learning_rate": 1.95737072460402e-05,
+      "loss": 0.2766,
+      "step": 763
+    },
+    {
+      "epoch": 0.36176383545427643,
+      "grad_norm": 1.8096346855163574,
+      "learning_rate": 1.957222872415405e-05,
+      "loss": 0.2817,
+      "step": 764
+    },
+    {
+      "epoch": 0.36223734832790766,
+      "grad_norm": 1.9074580669403076,
+      "learning_rate": 1.9570747698743818e-05,
+      "loss": 0.2689,
+      "step": 765
+    },
+    {
+      "epoch": 0.3627108612015389,
+      "grad_norm": 1.3230053186416626,
+      "learning_rate": 1.956926417019686e-05,
+      "loss": 0.2504,
+      "step": 766
+    },
+    {
+      "epoch": 0.3631843740751702,
+      "grad_norm": 1.6510530710220337,
+      "learning_rate": 1.9567778138901175e-05,
+      "loss": 0.2582,
+      "step": 767
+    },
+    {
+      "epoch": 0.3636578869488014,
+      "grad_norm": 1.1170092821121216,
+      "learning_rate": 1.9566289605245416e-05,
+      "loss": 0.2573,
+      "step": 768
+    },
+    {
+      "epoch": 0.3641313998224327,
+      "grad_norm": 1.1853854656219482,
+      "learning_rate": 1.95647985696189e-05,
+      "loss": 0.2889,
+      "step": 769
+    },
+    {
+      "epoch": 0.36460491269606393,
+      "grad_norm": 1.359598159790039,
+      "learning_rate": 1.9563305032411594e-05,
+      "loss": 0.259,
+      "step": 770
+    },
+    {
+      "epoch": 0.36507842556969516,
+      "grad_norm": 1.3403595685958862,
+      "learning_rate": 1.956180899401411e-05,
+      "loss": 0.2468,
+      "step": 771
+    },
+    {
+      "epoch": 0.36555193844332645,
+      "grad_norm": 1.6090962886810303,
+      "learning_rate": 1.9560310454817736e-05,
+      "loss": 0.2536,
+      "step": 772
+    },
+    {
+      "epoch": 0.3660254513169577,
+      "grad_norm": 1.543067455291748,
+      "learning_rate": 1.9558809415214386e-05,
+      "loss": 0.2331,
+      "step": 773
+    },
+    {
+      "epoch": 0.3664989641905889,
+      "grad_norm": 1.650692343711853,
+      "learning_rate": 1.955730587559665e-05,
+      "loss": 0.2705,
+      "step": 774
+    },
+    {
+      "epoch": 0.3669724770642202,
+      "grad_norm": 1.7598607540130615,
+      "learning_rate": 1.9555799836357765e-05,
+      "loss": 0.2612,
+      "step": 775
+    },
+    {
+      "epoch": 0.36744598993785144,
+      "grad_norm": 1.1476891040802002,
+      "learning_rate": 1.955429129789162e-05,
+      "loss": 0.2401,
+      "step": 776
+    },
+    {
+      "epoch": 0.36791950281148267,
+      "grad_norm": 2.1535274982452393,
+      "learning_rate": 1.9552780260592755e-05,
+      "loss": 0.2797,
+      "step": 777
+    },
+    {
+      "epoch": 0.36839301568511396,
+      "grad_norm": 1.3766757249832153,
+      "learning_rate": 1.955126672485637e-05,
+      "loss": 0.2579,
+      "step": 778
+    },
+    {
+      "epoch": 0.3688665285587452,
+      "grad_norm": 1.8468455076217651,
+      "learning_rate": 1.954975069107832e-05,
+      "loss": 0.2515,
+      "step": 779
+    },
+    {
+      "epoch": 0.3693400414323764,
+      "grad_norm": 1.970403790473938,
+      "learning_rate": 1.95482321596551e-05,
+      "loss": 0.2513,
+      "step": 780
+    },
+    {
+      "epoch": 0.3698135543060077,
+      "grad_norm": 1.3336784839630127,
+      "learning_rate": 1.9546711130983874e-05,
+      "loss": 0.2741,
+      "step": 781
+    },
+    {
+      "epoch": 0.37028706717963894,
+      "grad_norm": 2.5626380443573,
+      "learning_rate": 1.954518760546245e-05,
+      "loss": 0.2537,
+      "step": 782
+    },
+    {
+      "epoch": 0.3707605800532702,
+      "grad_norm": 3.228623628616333,
+      "learning_rate": 1.9543661583489295e-05,
+      "loss": 0.2756,
+      "step": 783
+    },
+    {
+      "epoch": 0.37123409292690146,
+      "grad_norm": 2.226210832595825,
+      "learning_rate": 1.9542133065463518e-05,
+      "loss": 0.278,
+      "step": 784
+    },
+    {
+      "epoch": 0.3717076058005327,
+      "grad_norm": 1.0971438884735107,
+      "learning_rate": 1.9540602051784897e-05,
+      "loss": 0.2334,
+      "step": 785
+    },
+    {
+      "epoch": 0.372181118674164,
+      "grad_norm": 1.3254427909851074,
+      "learning_rate": 1.9539068542853844e-05,
+      "loss": 0.2461,
+      "step": 786
+    },
+    {
+      "epoch": 0.3726546315477952,
+      "grad_norm": 1.6858649253845215,
+      "learning_rate": 1.953753253907144e-05,
+      "loss": 0.272,
+      "step": 787
+    },
+    {
+      "epoch": 0.37312814442142644,
+      "grad_norm": 1.6952732801437378,
+      "learning_rate": 1.9535994040839413e-05,
+      "loss": 0.2775,
+      "step": 788
+    },
+    {
+      "epoch": 0.37360165729505773,
+      "grad_norm": 1.41287100315094,
+      "learning_rate": 1.9534453048560137e-05,
+      "loss": 0.2415,
+      "step": 789
+    },
+    {
+      "epoch": 0.37407517016868896,
+      "grad_norm": 1.457585096359253,
+      "learning_rate": 1.953290956263665e-05,
+      "loss": 0.2778,
+      "step": 790
+    },
+    {
+      "epoch": 0.3745486830423202,
+      "grad_norm": 1.6097171306610107,
+      "learning_rate": 1.9531363583472628e-05,
+      "loss": 0.2522,
+      "step": 791
+    },
+    {
+      "epoch": 0.3750221959159515,
+      "grad_norm": 1.8246599435806274,
+      "learning_rate": 1.9529815111472414e-05,
+      "loss": 0.27,
+      "step": 792
+    },
+    {
+      "epoch": 0.3754957087895827,
+      "grad_norm": 2.544400691986084,
+      "learning_rate": 1.9528264147040995e-05,
+      "loss": 0.2704,
+      "step": 793
+    },
+    {
+      "epoch": 0.37596922166321395,
+      "grad_norm": 1.3660106658935547,
+      "learning_rate": 1.9526710690584005e-05,
+      "loss": 0.2608,
+      "step": 794
+    },
+    {
+      "epoch": 0.37644273453684524,
+      "grad_norm": 1.1534936428070068,
+      "learning_rate": 1.9525154742507745e-05,
+      "loss": 0.2565,
+      "step": 795
+    },
+    {
+      "epoch": 0.37691624741047647,
+      "grad_norm": 1.5819586515426636,
+      "learning_rate": 1.9523596303219146e-05,
+      "loss": 0.2806,
+      "step": 796
+    },
+    {
+      "epoch": 0.3773897602841077,
+      "grad_norm": 1.9623225927352905,
+      "learning_rate": 1.9522035373125816e-05,
+      "loss": 0.2916,
+      "step": 797
+    },
+    {
+      "epoch": 0.377863273157739,
+      "grad_norm": 1.619255542755127,
+      "learning_rate": 1.9520471952635992e-05,
+      "loss": 0.2641,
+      "step": 798
+    },
+    {
+      "epoch": 0.3783367860313702,
+      "grad_norm": 1.587944746017456,
+      "learning_rate": 1.9518906042158575e-05,
+      "loss": 0.2707,
+      "step": 799
+    },
+    {
+      "epoch": 0.37881029890500145,
+      "grad_norm": 1.1555267572402954,
+      "learning_rate": 1.9517337642103116e-05,
+      "loss": 0.2526,
+      "step": 800
+    },
+    {
+      "epoch": 0.37928381177863274,
+      "grad_norm": 1.6961590051651,
+      "learning_rate": 1.9515766752879808e-05,
+      "loss": 0.2276,
+      "step": 801
+    },
+    {
+      "epoch": 0.379757324652264,
+      "grad_norm": 1.6464869976043701,
+      "learning_rate": 1.9514193374899508e-05,
+      "loss": 0.2662,
+      "step": 802
+    },
+    {
+      "epoch": 0.38023083752589526,
+      "grad_norm": 2.7278549671173096,
+      "learning_rate": 1.9512617508573713e-05,
+      "loss": 0.2696,
+      "step": 803
+    },
+    {
+      "epoch": 0.3807043503995265,
+      "grad_norm": 1.3315246105194092,
+      "learning_rate": 1.951103915431458e-05,
+      "loss": 0.2777,
+      "step": 804
+    },
+    {
+      "epoch": 0.3811778632731577,
+      "grad_norm": 1.1722429990768433,
+      "learning_rate": 1.9509458312534912e-05,
+      "loss": 0.2657,
+      "step": 805
+    },
+    {
+      "epoch": 0.381651376146789,
+      "grad_norm": 2.3142430782318115,
+      "learning_rate": 1.9507874983648163e-05,
+      "loss": 0.2829,
+      "step": 806
+    },
+    {
+      "epoch": 0.38212488902042024,
+      "grad_norm": 2.0772149562835693,
+      "learning_rate": 1.9506289168068433e-05,
+      "loss": 0.2589,
+      "step": 807
+    },
+    {
+      "epoch": 0.3825984018940515,
+      "grad_norm": 1.3622729778289795,
+      "learning_rate": 1.950470086621048e-05,
+      "loss": 0.2629,
+      "step": 808
+    },
+    {
+      "epoch": 0.38307191476768276,
+      "grad_norm": 1.4933375120162964,
+      "learning_rate": 1.9503110078489712e-05,
+      "loss": 0.2594,
+      "step": 809
+    },
+    {
+      "epoch": 0.383545427641314,
+      "grad_norm": 1.4791429042816162,
+      "learning_rate": 1.950151680532218e-05,
+      "loss": 0.2544,
+      "step": 810
+    },
+    {
+      "epoch": 0.38401894051494523,
+      "grad_norm": 1.3858115673065186,
+      "learning_rate": 1.9499921047124587e-05,
+      "loss": 0.2611,
+      "step": 811
+    },
+    {
+      "epoch": 0.3844924533885765,
+      "grad_norm": 1.273097038269043,
+      "learning_rate": 1.9498322804314297e-05,
+      "loss": 0.2752,
+      "step": 812
+    },
+    {
+      "epoch": 0.38496596626220775,
+      "grad_norm": 1.3330796957015991,
+      "learning_rate": 1.9496722077309306e-05,
+      "loss": 0.2363,
+      "step": 813
+    },
+    {
+      "epoch": 0.385439479135839,
+      "grad_norm": 1.091861605644226,
+      "learning_rate": 1.949511886652827e-05,
+      "loss": 0.2528,
+      "step": 814
+    },
+    {
+      "epoch": 0.38591299200947027,
+      "grad_norm": 1.2032009363174438,
+      "learning_rate": 1.9493513172390498e-05,
+      "loss": 0.2528,
+      "step": 815
+    },
+    {
+      "epoch": 0.3863865048831015,
+      "grad_norm": 2.2817001342773438,
+      "learning_rate": 1.949190499531594e-05,
+      "loss": 0.2587,
+      "step": 816
+    },
+    {
+      "epoch": 0.3868600177567328,
+      "grad_norm": 2.36639666557312,
+      "learning_rate": 1.9490294335725204e-05,
+      "loss": 0.2474,
+      "step": 817
+    },
+    {
+      "epoch": 0.387333530630364,
+      "grad_norm": 1.1809403896331787,
+      "learning_rate": 1.9488681194039537e-05,
+      "loss": 0.2738,
+      "step": 818
+    },
+    {
+      "epoch": 0.38780704350399525,
+      "grad_norm": 1.3689755201339722,
+      "learning_rate": 1.9487065570680845e-05,
+      "loss": 0.2547,
+      "step": 819
+    },
+    {
+      "epoch": 0.38828055637762654,
+      "grad_norm": 1.8941787481307983,
+      "learning_rate": 1.948544746607167e-05,
+      "loss": 0.2823,
+      "step": 820
+    },
+    {
+      "epoch": 0.38875406925125777,
+      "grad_norm": 1.0949586629867554,
+      "learning_rate": 1.9483826880635225e-05,
+      "loss": 0.2658,
+      "step": 821
+    },
+    {
+      "epoch": 0.389227582124889,
+      "grad_norm": 2.2636001110076904,
+      "learning_rate": 1.9482203814795344e-05,
+      "loss": 0.2608,
+      "step": 822
+    },
+    {
+      "epoch": 0.3897010949985203,
+      "grad_norm": 1.2517188787460327,
+      "learning_rate": 1.9480578268976536e-05,
+      "loss": 0.2892,
+      "step": 823
+    },
+    {
+      "epoch": 0.3901746078721515,
+      "grad_norm": 1.6371468305587769,
+      "learning_rate": 1.9478950243603946e-05,
+      "loss": 0.2999,
+      "step": 824
+    },
+    {
+      "epoch": 0.39064812074578276,
+      "grad_norm": 1.488568663597107,
+      "learning_rate": 1.947731973910336e-05,
+      "loss": 0.2885,
+      "step": 825
+    },
+    {
+      "epoch": 0.39112163361941404,
+      "grad_norm": 1.5922267436981201,
+      "learning_rate": 1.9475686755901227e-05,
+      "loss": 0.2553,
+      "step": 826
+    },
+    {
+      "epoch": 0.3915951464930453,
+      "grad_norm": 2.3070576190948486,
+      "learning_rate": 1.9474051294424634e-05,
+      "loss": 0.2591,
+      "step": 827
+    },
+    {
+      "epoch": 0.3920686593666765,
+      "grad_norm": 2.676198959350586,
+      "learning_rate": 1.9472413355101327e-05,
+      "loss": 0.2867,
+      "step": 828
+    },
+    {
+      "epoch": 0.3925421722403078,
+      "grad_norm": 1.9517337083816528,
+      "learning_rate": 1.9470772938359687e-05,
+      "loss": 0.2684,
+      "step": 829
+    },
+    {
+      "epoch": 0.39301568511393903,
+      "grad_norm": 1.6799813508987427,
+      "learning_rate": 1.946913004462875e-05,
+      "loss": 0.2631,
+      "step": 830
+    },
+    {
+      "epoch": 0.39348919798757026,
+      "grad_norm": 1.206333041191101,
+      "learning_rate": 1.9467484674338202e-05,
+      "loss": 0.2628,
+      "step": 831
+    },
+    {
+      "epoch": 0.39396271086120155,
+      "grad_norm": 1.4563113451004028,
+      "learning_rate": 1.9465836827918373e-05,
+      "loss": 0.2684,
+      "step": 832
+    },
+    {
+      "epoch": 0.3944362237348328,
+      "grad_norm": 2.2227954864501953,
+      "learning_rate": 1.9464186505800236e-05,
+      "loss": 0.2905,
+      "step": 833
+    },
+    {
+      "epoch": 0.39490973660846407,
+      "grad_norm": 1.46279776096344,
+      "learning_rate": 1.9462533708415425e-05,
+      "loss": 0.2615,
+      "step": 834
+    },
+    {
+      "epoch": 0.3953832494820953,
+      "grad_norm": 2.0220069885253906,
+      "learning_rate": 1.9460878436196206e-05,
+      "loss": 0.2427,
+      "step": 835
+    },
+    {
+      "epoch": 0.39585676235572653,
+      "grad_norm": 1.3017007112503052,
+      "learning_rate": 1.9459220689575505e-05,
+      "loss": 0.2563,
+      "step": 836
+    },
+    {
+      "epoch": 0.3963302752293578,
+      "grad_norm": 1.700585961341858,
+      "learning_rate": 1.9457560468986888e-05,
+      "loss": 0.2656,
+      "step": 837
+    },
+    {
+      "epoch": 0.39680378810298905,
+      "grad_norm": 1.8619630336761475,
+      "learning_rate": 1.9455897774864567e-05,
+      "loss": 0.2653,
+      "step": 838
+    },
+    {
+      "epoch": 0.3972773009766203,
+      "grad_norm": 1.1475433111190796,
+      "learning_rate": 1.9454232607643406e-05,
+      "loss": 0.2612,
+      "step": 839
+    },
+    {
+      "epoch": 0.39775081385025157,
+      "grad_norm": 1.736665964126587,
+      "learning_rate": 1.9452564967758912e-05,
+      "loss": 0.2655,
+      "step": 840
+    },
+    {
+      "epoch": 0.3982243267238828,
+      "grad_norm": 1.1102855205535889,
+      "learning_rate": 1.9450894855647246e-05,
+      "loss": 0.2437,
+      "step": 841
+    },
+    {
+      "epoch": 0.39869783959751404,
+      "grad_norm": 1.2545884847640991,
+      "learning_rate": 1.9449222271745202e-05,
+      "loss": 0.2401,
+      "step": 842
+    },
+    {
+      "epoch": 0.3991713524711453,
+      "grad_norm": 1.2088849544525146,
+      "learning_rate": 1.944754721649023e-05,
+      "loss": 0.2497,
+      "step": 843
+    },
+    {
+      "epoch": 0.39964486534477656,
+      "grad_norm": 1.708852767944336,
+      "learning_rate": 1.9445869690320425e-05,
+      "loss": 0.2513,
+      "step": 844
+    },
+    {
+      "epoch": 0.4001183782184078,
+      "grad_norm": 1.3181136846542358,
+      "learning_rate": 1.9444189693674528e-05,
+      "loss": 0.2814,
+      "step": 845
+    },
+    {
+      "epoch": 0.4005918910920391,
+      "grad_norm": 1.6149542331695557,
+      "learning_rate": 1.944250722699193e-05,
+      "loss": 0.2475,
+      "step": 846
+    },
+    {
+      "epoch": 0.4010654039656703,
+      "grad_norm": 1.4610974788665771,
+      "learning_rate": 1.944082229071266e-05,
+      "loss": 0.2571,
+      "step": 847
+    },
+    {
+      "epoch": 0.4015389168393016,
+      "grad_norm": 1.5411815643310547,
+      "learning_rate": 1.9439134885277394e-05,
+      "loss": 0.2255,
+      "step": 848
+    },
+    {
+      "epoch": 0.40201242971293283,
+      "grad_norm": 1.6833422183990479,
+      "learning_rate": 1.9437445011127463e-05,
+      "loss": 0.2551,
+      "step": 849
+    },
+    {
+      "epoch": 0.40248594258656406,
+      "grad_norm": 1.65785813331604,
+      "learning_rate": 1.943575266870483e-05,
+      "loss": 0.246,
+      "step": 850
+    },
+    {
+      "epoch": 0.40295945546019535,
+      "grad_norm": 1.3264704942703247,
+      "learning_rate": 1.9434057858452117e-05,
+      "loss": 0.2935,
+      "step": 851
+    },
+    {
+      "epoch": 0.4034329683338266,
+      "grad_norm": 1.166812539100647,
+      "learning_rate": 1.9432360580812583e-05,
+      "loss": 0.2528,
+      "step": 852
+    },
+    {
+      "epoch": 0.4039064812074578,
+      "grad_norm": 2.272095203399658,
+      "learning_rate": 1.9430660836230134e-05,
+      "loss": 0.2852,
+      "step": 853
+    },
+    {
+      "epoch": 0.4043799940810891,
+      "grad_norm": 1.7517335414886475,
+      "learning_rate": 1.9428958625149324e-05,
+      "loss": 0.2755,
+      "step": 854
+    },
+    {
+      "epoch": 0.40485350695472033,
+      "grad_norm": 1.2015048265457153,
+      "learning_rate": 1.942725394801535e-05,
+      "loss": 0.2628,
+      "step": 855
+    },
+    {
+      "epoch": 0.40532701982835156,
+      "grad_norm": 1.3004090785980225,
+      "learning_rate": 1.9425546805274048e-05,
+      "loss": 0.2415,
+      "step": 856
+    },
+    {
+      "epoch": 0.40580053270198285,
+      "grad_norm": 1.5480315685272217,
+      "learning_rate": 1.942383719737191e-05,
+      "loss": 0.3167,
+      "step": 857
+    },
+    {
+      "epoch": 0.4062740455756141,
+      "grad_norm": 2.4538869857788086,
+      "learning_rate": 1.9422125124756068e-05,
+      "loss": 0.2415,
+      "step": 858
+    },
+    {
+      "epoch": 0.4067475584492453,
+      "grad_norm": 1.163549542427063,
+      "learning_rate": 1.9420410587874295e-05,
+      "loss": 0.2615,
+      "step": 859
+    },
+    {
+      "epoch": 0.4072210713228766,
+      "grad_norm": 1.151456594467163,
+      "learning_rate": 1.941869358717501e-05,
+      "loss": 0.2391,
+      "step": 860
+    },
+    {
+      "epoch": 0.40769458419650784,
+      "grad_norm": 1.4595973491668701,
+      "learning_rate": 1.9416974123107287e-05,
+      "loss": 0.2871,
+      "step": 861
+    },
+    {
+      "epoch": 0.40816809707013907,
+      "grad_norm": 1.4766554832458496,
+      "learning_rate": 1.941525219612083e-05,
+      "loss": 0.2879,
+      "step": 862
+    },
+    {
+      "epoch": 0.40864160994377036,
+      "grad_norm": 1.770892858505249,
+      "learning_rate": 1.941352780666599e-05,
+      "loss": 0.2809,
+      "step": 863
+    },
+    {
+      "epoch": 0.4091151228174016,
+      "grad_norm": 1.4286603927612305,
+      "learning_rate": 1.9411800955193762e-05,
+      "loss": 0.2793,
+      "step": 864
+    },
+    {
+      "epoch": 0.4095886356910329,
+      "grad_norm": 1.2635271549224854,
+      "learning_rate": 1.9410071642155796e-05,
+      "loss": 0.267,
+      "step": 865
+    },
+    {
+      "epoch": 0.4100621485646641,
+      "grad_norm": 2.1183278560638428,
+      "learning_rate": 1.940833986800437e-05,
+      "loss": 0.2578,
+      "step": 866
+    },
+    {
+      "epoch": 0.41053566143829534,
+      "grad_norm": 1.1549620628356934,
+      "learning_rate": 1.9406605633192414e-05,
+      "loss": 0.2852,
+      "step": 867
+    },
+    {
+      "epoch": 0.41100917431192663,
+      "grad_norm": 1.3597477674484253,
+      "learning_rate": 1.9404868938173503e-05,
+      "loss": 0.2711,
+      "step": 868
+    },
+    {
+      "epoch": 0.41148268718555786,
+      "grad_norm": 1.2947059869766235,
+      "learning_rate": 1.9403129783401854e-05,
+      "loss": 0.2838,
+      "step": 869
+    },
+    {
+      "epoch": 0.4119562000591891,
+      "grad_norm": 1.0901488065719604,
+      "learning_rate": 1.9401388169332322e-05,
+      "loss": 0.2467,
+      "step": 870
+    },
+    {
+      "epoch": 0.4124297129328204,
+      "grad_norm": 1.7122608423233032,
+      "learning_rate": 1.939964409642041e-05,
+      "loss": 0.265,
+      "step": 871
+    },
+    {
+      "epoch": 0.4129032258064516,
+      "grad_norm": 1.7495481967926025,
+      "learning_rate": 1.9397897565122267e-05,
+      "loss": 0.2824,
+      "step": 872
+    },
+    {
+      "epoch": 0.41337673868008284,
+      "grad_norm": 1.2452020645141602,
+      "learning_rate": 1.939614857589468e-05,
+      "loss": 0.2695,
+      "step": 873
+    },
+    {
+      "epoch": 0.41385025155371413,
+      "grad_norm": 1.2135512828826904,
+      "learning_rate": 1.9394397129195076e-05,
+      "loss": 0.2606,
+      "step": 874
+    },
+    {
+      "epoch": 0.41432376442734536,
+      "grad_norm": 1.4891951084136963,
+      "learning_rate": 1.9392643225481535e-05,
+      "loss": 0.2587,
+      "step": 875
+    },
+    {
+      "epoch": 0.4147972773009766,
+      "grad_norm": 1.4017555713653564,
+      "learning_rate": 1.9390886865212767e-05,
+      "loss": 0.2776,
+      "step": 876
+    },
+    {
+      "epoch": 0.4152707901746079,
+      "grad_norm": 1.8024797439575195,
+      "learning_rate": 1.9389128048848136e-05,
+      "loss": 0.265,
+      "step": 877
+    },
+    {
+      "epoch": 0.4157443030482391,
+      "grad_norm": 1.2021361589431763,
+      "learning_rate": 1.9387366776847645e-05,
+      "loss": 0.2712,
+      "step": 878
+    },
+    {
+      "epoch": 0.41621781592187035,
+      "grad_norm": 1.449995756149292,
+      "learning_rate": 1.9385603049671934e-05,
+      "loss": 0.2887,
+      "step": 879
+    },
+    {
+      "epoch": 0.41669132879550164,
+      "grad_norm": 1.4844348430633545,
+      "learning_rate": 1.9383836867782287e-05,
+      "loss": 0.2384,
+      "step": 880
+    },
+    {
+      "epoch": 0.41716484166913287,
+      "grad_norm": 1.3519595861434937,
+      "learning_rate": 1.938206823164064e-05,
+      "loss": 0.2376,
+      "step": 881
+    },
+    {
+      "epoch": 0.41763835454276416,
+      "grad_norm": 1.2284427881240845,
+      "learning_rate": 1.938029714170955e-05,
+      "loss": 0.2525,
+      "step": 882
+    },
+    {
+      "epoch": 0.4181118674163954,
+      "grad_norm": 1.362317442893982,
+      "learning_rate": 1.937852359845224e-05,
+      "loss": 0.2695,
+      "step": 883
+    },
+    {
+      "epoch": 0.4185853802900266,
+      "grad_norm": 1.533306360244751,
+      "learning_rate": 1.937674760233256e-05,
+      "loss": 0.2334,
+      "step": 884
+    },
+    {
+      "epoch": 0.4190588931636579,
+      "grad_norm": 2.291046380996704,
+      "learning_rate": 1.9374969153815005e-05,
+      "loss": 0.2578,
+      "step": 885
+    },
+    {
+      "epoch": 0.41953240603728914,
+      "grad_norm": 1.2593145370483398,
+      "learning_rate": 1.937318825336471e-05,
+      "loss": 0.2854,
+      "step": 886
+    },
+    {
+      "epoch": 0.4200059189109204,
+      "grad_norm": 1.5685346126556396,
+      "learning_rate": 1.9371404901447445e-05,
+      "loss": 0.2777,
+      "step": 887
+    },
+    {
+      "epoch": 0.42047943178455166,
+      "grad_norm": 1.7853624820709229,
+      "learning_rate": 1.936961909852964e-05,
+      "loss": 0.2817,
+      "step": 888
+    },
+    {
+      "epoch": 0.4209529446581829,
+      "grad_norm": 1.8879746198654175,
+      "learning_rate": 1.9367830845078354e-05,
+      "loss": 0.2515,
+      "step": 889
+    },
+    {
+      "epoch": 0.4214264575318141,
+      "grad_norm": 1.719527244567871,
+      "learning_rate": 1.936604014156128e-05,
+      "loss": 0.2583,
+      "step": 890
+    },
+    {
+      "epoch": 0.4218999704054454,
+      "grad_norm": 1.735648274421692,
+      "learning_rate": 1.936424698844676e-05,
+      "loss": 0.2788,
+      "step": 891
+    },
+    {
+      "epoch": 0.42237348327907664,
+      "grad_norm": 1.6810283660888672,
+      "learning_rate": 1.9362451386203784e-05,
+      "loss": 0.2548,
+      "step": 892
+    },
+    {
+      "epoch": 0.4228469961527079,
+      "grad_norm": 1.0706638097763062,
+      "learning_rate": 1.9360653335301964e-05,
+      "loss": 0.2553,
+      "step": 893
+    },
+    {
+      "epoch": 0.42332050902633916,
+      "grad_norm": 1.1451069116592407,
+      "learning_rate": 1.9358852836211573e-05,
+      "loss": 0.2566,
+      "step": 894
+    },
+    {
+      "epoch": 0.4237940218999704,
+      "grad_norm": 1.3580163717269897,
+      "learning_rate": 1.9357049889403506e-05,
+      "loss": 0.2676,
+      "step": 895
+    },
+    {
+      "epoch": 0.4242675347736017,
+      "grad_norm": 1.5814894437789917,
+      "learning_rate": 1.9355244495349307e-05,
+      "loss": 0.2997,
+      "step": 896
+    },
+    {
+      "epoch": 0.4247410476472329,
+      "grad_norm": 2.3685078620910645,
+      "learning_rate": 1.9353436654521168e-05,
+      "loss": 0.261,
+      "step": 897
+    },
+    {
+      "epoch": 0.42521456052086415,
+      "grad_norm": 1.4430562257766724,
+      "learning_rate": 1.9351626367391902e-05,
+      "loss": 0.2477,
+      "step": 898
+    },
+    {
+      "epoch": 0.42568807339449544,
+      "grad_norm": 1.7693711519241333,
+      "learning_rate": 1.9349813634434977e-05,
+      "loss": 0.2577,
+      "step": 899
+    },
+    {
+      "epoch": 0.42616158626812667,
+      "grad_norm": 1.3744468688964844,
+      "learning_rate": 1.9347998456124497e-05,
+      "loss": 0.2736,
+      "step": 900
+    },
+    {
+      "epoch": 0.4266350991417579,
+      "grad_norm": 1.5893174409866333,
+      "learning_rate": 1.9346180832935202e-05,
+      "loss": 0.2518,
+      "step": 901
+    },
+    {
+      "epoch": 0.4271086120153892,
+      "grad_norm": 1.6329647302627563,
+      "learning_rate": 1.9344360765342472e-05,
+      "loss": 0.2574,
+      "step": 902
+    },
+    {
+      "epoch": 0.4275821248890204,
+      "grad_norm": 1.2065149545669556,
+      "learning_rate": 1.9342538253822334e-05,
+      "loss": 0.2522,
+      "step": 903
+    },
+    {
+      "epoch": 0.42805563776265165,
+      "grad_norm": 1.2911807298660278,
+      "learning_rate": 1.934071329885144e-05,
+      "loss": 0.2441,
+      "step": 904
+    },
+    {
+      "epoch": 0.42852915063628294,
+      "grad_norm": 1.260495901107788,
+      "learning_rate": 1.93388859009071e-05,
+      "loss": 0.2505,
+      "step": 905
+    },
+    {
+      "epoch": 0.4290026635099142,
+      "grad_norm": 1.6986874341964722,
+      "learning_rate": 1.9337056060467244e-05,
+      "loss": 0.2942,
+      "step": 906
+    },
+    {
+      "epoch": 0.4294761763835454,
+      "grad_norm": 1.6441107988357544,
+      "learning_rate": 1.933522377801045e-05,
+      "loss": 0.2577,
+      "step": 907
+    },
+    {
+      "epoch": 0.4299496892571767,
+      "grad_norm": 1.4719241857528687,
+      "learning_rate": 1.9333389054015935e-05,
+      "loss": 0.2602,
+      "step": 908
+    },
+    {
+      "epoch": 0.4304232021308079,
+      "grad_norm": 1.37874436378479,
+      "learning_rate": 1.9331551888963557e-05,
+      "loss": 0.2792,
+      "step": 909
+    },
+    {
+      "epoch": 0.43089671500443916,
+      "grad_norm": 1.7454997301101685,
+      "learning_rate": 1.93297122833338e-05,
+      "loss": 0.2456,
+      "step": 910
+    },
+    {
+      "epoch": 0.43137022787807044,
+      "grad_norm": 1.5132381916046143,
+      "learning_rate": 1.9327870237607805e-05,
+      "loss": 0.2826,
+      "step": 911
+    },
+    {
+      "epoch": 0.4318437407517017,
+      "grad_norm": 1.309891939163208,
+      "learning_rate": 1.9326025752267338e-05,
+      "loss": 0.2695,
+      "step": 912
+    },
+    {
+      "epoch": 0.43231725362533296,
+      "grad_norm": 1.375044822692871,
+      "learning_rate": 1.9324178827794803e-05,
+      "loss": 0.252,
+      "step": 913
+    },
+    {
+      "epoch": 0.4327907664989642,
+      "grad_norm": 1.6480258703231812,
+      "learning_rate": 1.9322329464673248e-05,
+      "loss": 0.2797,
+      "step": 914
+    },
+    {
+      "epoch": 0.43326427937259543,
+      "grad_norm": 1.1525005102157593,
+      "learning_rate": 1.9320477663386358e-05,
+      "loss": 0.2449,
+      "step": 915
+    },
+    {
+      "epoch": 0.4337377922462267,
+      "grad_norm": 1.945312261581421,
+      "learning_rate": 1.9318623424418446e-05,
+      "loss": 0.2355,
+      "step": 916
+    },
+    {
+      "epoch": 0.43421130511985795,
+      "grad_norm": 1.427280068397522,
+      "learning_rate": 1.9316766748254477e-05,
+      "loss": 0.2888,
+      "step": 917
+    },
+    {
+      "epoch": 0.4346848179934892,
+      "grad_norm": 1.4059946537017822,
+      "learning_rate": 1.931490763538005e-05,
+      "loss": 0.2321,
+      "step": 918
+    },
+    {
+      "epoch": 0.43515833086712047,
+      "grad_norm": 1.3967366218566895,
+      "learning_rate": 1.931304608628139e-05,
+      "loss": 0.2585,
+      "step": 919
+    },
+    {
+      "epoch": 0.4356318437407517,
+      "grad_norm": 1.3935580253601074,
+      "learning_rate": 1.931118210144537e-05,
+      "loss": 0.2663,
+      "step": 920
+    },
+    {
+      "epoch": 0.43610535661438293,
+      "grad_norm": 1.2772636413574219,
+      "learning_rate": 1.93093156813595e-05,
+      "loss": 0.2607,
+      "step": 921
+    },
+    {
+      "epoch": 0.4365788694880142,
+      "grad_norm": 1.2750356197357178,
+      "learning_rate": 1.930744682651192e-05,
+      "loss": 0.2518,
+      "step": 922
+    },
+    {
+      "epoch": 0.43705238236164545,
+      "grad_norm": 1.2921040058135986,
+      "learning_rate": 1.9305575537391416e-05,
+      "loss": 0.2623,
+      "step": 923
+    },
+    {
+      "epoch": 0.4375258952352767,
+      "grad_norm": 1.6561219692230225,
+      "learning_rate": 1.9303701814487403e-05,
+      "loss": 0.3298,
+      "step": 924
+    },
+    {
+      "epoch": 0.437999408108908,
+      "grad_norm": 1.4498628377914429,
+      "learning_rate": 1.930182565828993e-05,
+      "loss": 0.2543,
+      "step": 925
+    },
+    {
+      "epoch": 0.4384729209825392,
+      "grad_norm": 1.2270756959915161,
+      "learning_rate": 1.9299947069289694e-05,
+      "loss": 0.2553,
+      "step": 926
+    },
+    {
+      "epoch": 0.43894643385617044,
+      "grad_norm": 1.31882905960083,
+      "learning_rate": 1.9298066047978024e-05,
+      "loss": 0.2574,
+      "step": 927
+    },
+    {
+      "epoch": 0.4394199467298017,
+      "grad_norm": 1.993991494178772,
+      "learning_rate": 1.9296182594846876e-05,
+      "loss": 0.2692,
+      "step": 928
+    },
+    {
+      "epoch": 0.43989345960343296,
+      "grad_norm": 1.425070881843567,
+      "learning_rate": 1.9294296710388852e-05,
+      "loss": 0.2553,
+      "step": 929
+    },
+    {
+      "epoch": 0.44036697247706424,
+      "grad_norm": 1.2672945261001587,
+      "learning_rate": 1.9292408395097187e-05,
+      "loss": 0.2814,
+      "step": 930
+    },
+    {
+      "epoch": 0.4408404853506955,
+      "grad_norm": 1.2957855463027954,
+      "learning_rate": 1.9290517649465756e-05,
+      "loss": 0.2714,
+      "step": 931
+    },
+    {
+      "epoch": 0.4413139982243267,
+      "grad_norm": 1.4948066473007202,
+      "learning_rate": 1.9288624473989055e-05,
+      "loss": 0.2353,
+      "step": 932
+    },
+    {
+      "epoch": 0.441787511097958,
+      "grad_norm": 1.6086140871047974,
+      "learning_rate": 1.9286728869162235e-05,
+      "loss": 0.2507,
+      "step": 933
+    },
+    {
+      "epoch": 0.44226102397158923,
+      "grad_norm": 1.5332534313201904,
+      "learning_rate": 1.928483083548107e-05,
+      "loss": 0.2744,
+      "step": 934
+    },
+    {
+      "epoch": 0.44273453684522046,
+      "grad_norm": 1.8860116004943848,
+      "learning_rate": 1.928293037344197e-05,
+      "loss": 0.2706,
+      "step": 935
+    },
+    {
+      "epoch": 0.44320804971885175,
+      "grad_norm": 1.3716895580291748,
+      "learning_rate": 1.9281027483541986e-05,
+      "loss": 0.2565,
+      "step": 936
+    },
+    {
+      "epoch": 0.443681562592483,
+      "grad_norm": 2.0549087524414062,
+      "learning_rate": 1.9279122166278798e-05,
+      "loss": 0.2524,
+      "step": 937
+    },
+    {
+      "epoch": 0.4441550754661142,
+      "grad_norm": 1.379072666168213,
+      "learning_rate": 1.927721442215073e-05,
+      "loss": 0.2694,
+      "step": 938
+    },
+    {
+      "epoch": 0.4446285883397455,
+      "grad_norm": 1.1916866302490234,
+      "learning_rate": 1.9275304251656723e-05,
+      "loss": 0.2386,
+      "step": 939
+    },
+    {
+      "epoch": 0.44510210121337673,
+      "grad_norm": 2.893798589706421,
+      "learning_rate": 1.9273391655296373e-05,
+      "loss": 0.2682,
+      "step": 940
+    },
+    {
+      "epoch": 0.44557561408700797,
+      "grad_norm": 2.7330830097198486,
+      "learning_rate": 1.9271476633569895e-05,
+      "loss": 0.2632,
+      "step": 941
+    },
+    {
+      "epoch": 0.44604912696063925,
+      "grad_norm": 1.7794554233551025,
+      "learning_rate": 1.926955918697815e-05,
+      "loss": 0.2447,
+      "step": 942
+    },
+    {
+      "epoch": 0.4465226398342705,
+      "grad_norm": 1.3762879371643066,
+      "learning_rate": 1.926763931602262e-05,
+      "loss": 0.2525,
+      "step": 943
+    },
+    {
+      "epoch": 0.4469961527079018,
+      "grad_norm": 1.5030863285064697,
+      "learning_rate": 1.9265717021205437e-05,
+      "loss": 0.248,
+      "step": 944
+    },
+    {
+      "epoch": 0.447469665581533,
+      "grad_norm": 1.604885220527649,
+      "learning_rate": 1.9263792303029355e-05,
+      "loss": 0.2527,
+      "step": 945
+    },
+    {
+      "epoch": 0.44794317845516424,
+      "grad_norm": 1.0252493619918823,
+      "learning_rate": 1.9261865161997765e-05,
+      "loss": 0.2084,
+      "step": 946
+    },
+    {
+      "epoch": 0.4484166913287955,
+      "grad_norm": 1.32304048538208,
+      "learning_rate": 1.925993559861469e-05,
+      "loss": 0.2385,
+      "step": 947
+    },
+    {
+      "epoch": 0.44889020420242676,
+      "grad_norm": 1.2492711544036865,
+      "learning_rate": 1.9258003613384793e-05,
+      "loss": 0.2862,
+      "step": 948
+    },
+    {
+      "epoch": 0.449363717076058,
+      "grad_norm": 1.9659324884414673,
+      "learning_rate": 1.925606920681337e-05,
+      "loss": 0.2729,
+      "step": 949
+    },
+    {
+      "epoch": 0.4498372299496893,
+      "grad_norm": 1.930130958557129,
+      "learning_rate": 1.9254132379406335e-05,
+      "loss": 0.2429,
+      "step": 950
+    },
+    {
+      "epoch": 0.4503107428233205,
+      "grad_norm": 1.9109481573104858,
+      "learning_rate": 1.925219313167025e-05,
+      "loss": 0.2579,
+      "step": 951
+    },
+    {
+      "epoch": 0.45078425569695174,
+      "grad_norm": 1.461925983428955,
+      "learning_rate": 1.9250251464112313e-05,
+      "loss": 0.3025,
+      "step": 952
+    },
+    {
+      "epoch": 0.45125776857058303,
+      "grad_norm": 1.1421688795089722,
+      "learning_rate": 1.9248307377240346e-05,
+      "loss": 0.2611,
+      "step": 953
+    },
+    {
+      "epoch": 0.45173128144421426,
+      "grad_norm": 1.3895161151885986,
+      "learning_rate": 1.92463608715628e-05,
+      "loss": 0.2854,
+      "step": 954
+    },
+    {
+      "epoch": 0.4522047943178455,
+      "grad_norm": 1.081526517868042,
+      "learning_rate": 1.9244411947588774e-05,
+      "loss": 0.2495,
+      "step": 955
+    },
+    {
+      "epoch": 0.4526783071914768,
+      "grad_norm": 1.5755951404571533,
+      "learning_rate": 1.924246060582798e-05,
+      "loss": 0.2645,
+      "step": 956
+    },
+    {
+      "epoch": 0.453151820065108,
+      "grad_norm": 1.318973183631897,
+      "learning_rate": 1.9240506846790784e-05,
+      "loss": 0.2215,
+      "step": 957
+    },
+    {
+      "epoch": 0.45362533293873925,
+      "grad_norm": 1.0379008054733276,
+      "learning_rate": 1.9238550670988166e-05,
+      "loss": 0.2594,
+      "step": 958
+    },
+    {
+      "epoch": 0.45409884581237053,
+      "grad_norm": 1.1944754123687744,
+      "learning_rate": 1.923659207893174e-05,
+      "loss": 0.2458,
+      "step": 959
+    },
+    {
+      "epoch": 0.45457235868600177,
+      "grad_norm": 1.7871370315551758,
+      "learning_rate": 1.9234631071133768e-05,
+      "loss": 0.2622,
+      "step": 960
+    },
+    {
+      "epoch": 0.45504587155963305,
+      "grad_norm": 1.2604117393493652,
+      "learning_rate": 1.9232667648107127e-05,
+      "loss": 0.2823,
+      "step": 961
+    },
+    {
+      "epoch": 0.4555193844332643,
+      "grad_norm": 1.3337626457214355,
+      "learning_rate": 1.923070181036533e-05,
+      "loss": 0.2737,
+      "step": 962
+    },
+    {
+      "epoch": 0.4559928973068955,
+      "grad_norm": 1.0898325443267822,
+      "learning_rate": 1.9228733558422525e-05,
+      "loss": 0.2567,
+      "step": 963
+    },
+    {
+      "epoch": 0.4564664101805268,
+      "grad_norm": 1.1792700290679932,
+      "learning_rate": 1.9226762892793492e-05,
+      "loss": 0.2511,
+      "step": 964
+    },
+    {
+      "epoch": 0.45693992305415804,
+      "grad_norm": 1.4441752433776855,
+      "learning_rate": 1.922478981399363e-05,
+      "loss": 0.2526,
+      "step": 965
+    },
+    {
+      "epoch": 0.45741343592778927,
+      "grad_norm": 1.9730075597763062,
+      "learning_rate": 1.9222814322538993e-05,
+      "loss": 0.3023,
+      "step": 966
+    },
+    {
+      "epoch": 0.45788694880142056,
+      "grad_norm": 1.2090874910354614,
+      "learning_rate": 1.922083641894624e-05,
+      "loss": 0.2561,
+      "step": 967
+    },
+    {
+      "epoch": 0.4583604616750518,
+      "grad_norm": 1.5436848402023315,
+      "learning_rate": 1.9218856103732675e-05,
+      "loss": 0.2539,
+      "step": 968
+    },
+    {
+      "epoch": 0.458833974548683,
+      "grad_norm": 1.0943453311920166,
+      "learning_rate": 1.9216873377416236e-05,
+      "loss": 0.2568,
+      "step": 969
+    },
+    {
+      "epoch": 0.4593074874223143,
+      "grad_norm": 1.5270817279815674,
+      "learning_rate": 1.9214888240515478e-05,
+      "loss": 0.2381,
+      "step": 970
+    },
+    {
+      "epoch": 0.45978100029594554,
+      "grad_norm": 1.254157543182373,
+      "learning_rate": 1.9212900693549602e-05,
+      "loss": 0.2476,
+      "step": 971
+    },
+    {
+      "epoch": 0.4602545131695768,
+      "grad_norm": 1.5397264957427979,
+      "learning_rate": 1.9210910737038424e-05,
+      "loss": 0.2651,
+      "step": 972
+    },
+    {
+      "epoch": 0.46072802604320806,
+      "grad_norm": 1.4279857873916626,
+      "learning_rate": 1.9208918371502404e-05,
+      "loss": 0.249,
+      "step": 973
+    },
+    {
+      "epoch": 0.4612015389168393,
+      "grad_norm": 1.1333208084106445,
+      "learning_rate": 1.9206923597462625e-05,
+      "loss": 0.2391,
+      "step": 974
+    },
+    {
+      "epoch": 0.4616750517904706,
+      "grad_norm": 1.3642656803131104,
+      "learning_rate": 1.9204926415440798e-05,
+      "loss": 0.2462,
+      "step": 975
+    },
+    {
+      "epoch": 0.4621485646641018,
+      "grad_norm": 1.0775152444839478,
+      "learning_rate": 1.920292682595927e-05,
+      "loss": 0.2345,
+      "step": 976
+    },
+    {
+      "epoch": 0.46262207753773305,
+      "grad_norm": 1.583609938621521,
+      "learning_rate": 1.9200924829541012e-05,
+      "loss": 0.2341,
+      "step": 977
+    },
+    {
+      "epoch": 0.46309559041136433,
+      "grad_norm": 1.214303731918335,
+      "learning_rate": 1.919892042670963e-05,
+      "loss": 0.2515,
+      "step": 978
+    },
+    {
+      "epoch": 0.46356910328499557,
+      "grad_norm": 1.1225674152374268,
+      "learning_rate": 1.919691361798935e-05,
+      "loss": 0.2615,
+      "step": 979
+    },
+    {
+      "epoch": 0.4640426161586268,
+      "grad_norm": 1.8481762409210205,
+      "learning_rate": 1.9194904403905038e-05,
+      "loss": 0.2602,
+      "step": 980
+    },
+    {
+      "epoch": 0.4645161290322581,
+      "grad_norm": 1.6956995725631714,
+      "learning_rate": 1.9192892784982185e-05,
+      "loss": 0.2413,
+      "step": 981
+    },
+    {
+      "epoch": 0.4649896419058893,
+      "grad_norm": 1.4857017993927002,
+      "learning_rate": 1.919087876174691e-05,
+      "loss": 0.271,
+      "step": 982
+    },
+    {
+      "epoch": 0.46546315477952055,
+      "grad_norm": 1.868812918663025,
+      "learning_rate": 1.918886233472596e-05,
+      "loss": 0.2553,
+      "step": 983
+    },
+    {
+      "epoch": 0.46593666765315184,
+      "grad_norm": 2.422353982925415,
+      "learning_rate": 1.9186843504446716e-05,
+      "loss": 0.2641,
+      "step": 984
+    },
+    {
+      "epoch": 0.46641018052678307,
+      "grad_norm": 1.8645198345184326,
+      "learning_rate": 1.9184822271437176e-05,
+      "loss": 0.2752,
+      "step": 985
+    },
+    {
+      "epoch": 0.4668836934004143,
+      "grad_norm": 1.2485929727554321,
+      "learning_rate": 1.9182798636225983e-05,
+      "loss": 0.2439,
+      "step": 986
+    },
+    {
+      "epoch": 0.4673572062740456,
+      "grad_norm": 1.1110594272613525,
+      "learning_rate": 1.918077259934239e-05,
+      "loss": 0.2504,
+      "step": 987
+    },
+    {
+      "epoch": 0.4678307191476768,
+      "grad_norm": 1.529255747795105,
+      "learning_rate": 1.9178744161316297e-05,
+      "loss": 0.2643,
+      "step": 988
+    },
+    {
+      "epoch": 0.46830423202130805,
+      "grad_norm": 1.2772247791290283,
+      "learning_rate": 1.9176713322678212e-05,
+      "loss": 0.2207,
+      "step": 989
+    },
+    {
+      "epoch": 0.46877774489493934,
+      "grad_norm": 1.129361867904663,
+      "learning_rate": 1.917468008395929e-05,
+      "loss": 0.2468,
+      "step": 990
+    },
+    {
+      "epoch": 0.4692512577685706,
+      "grad_norm": 1.451148271560669,
+      "learning_rate": 1.9172644445691305e-05,
+      "loss": 0.2493,
+      "step": 991
+    },
+    {
+      "epoch": 0.46972477064220186,
+      "grad_norm": 1.5785322189331055,
+      "learning_rate": 1.9170606408406648e-05,
+      "loss": 0.2314,
+      "step": 992
+    },
+    {
+      "epoch": 0.4701982835158331,
+      "grad_norm": 1.7173629999160767,
+      "learning_rate": 1.916856597263836e-05,
+      "loss": 0.2629,
+      "step": 993
+    },
+    {
+      "epoch": 0.4706717963894643,
+      "grad_norm": 1.471721887588501,
+      "learning_rate": 1.916652313892009e-05,
+      "loss": 0.2752,
+      "step": 994
+    },
+    {
+      "epoch": 0.4711453092630956,
+      "grad_norm": 1.9189519882202148,
+      "learning_rate": 1.9164477907786128e-05,
+      "loss": 0.2714,
+      "step": 995
+    },
+    {
+      "epoch": 0.47161882213672685,
+      "grad_norm": 1.542122721672058,
+      "learning_rate": 1.9162430279771378e-05,
+      "loss": 0.2648,
+      "step": 996
+    },
+    {
+      "epoch": 0.4720923350103581,
+      "grad_norm": 1.2591853141784668,
+      "learning_rate": 1.916038025541138e-05,
+      "loss": 0.2792,
+      "step": 997
+    },
+    {
+      "epoch": 0.47256584788398937,
+      "grad_norm": 1.6615357398986816,
+      "learning_rate": 1.9158327835242296e-05,
+      "loss": 0.2639,
+      "step": 998
+    },
+    {
+      "epoch": 0.4730393607576206,
+      "grad_norm": 1.494132161140442,
+      "learning_rate": 1.915627301980092e-05,
+      "loss": 0.2994,
+      "step": 999
+    },
+    {
+      "epoch": 0.47351287363125183,
+      "grad_norm": 1.3454561233520508,
+      "learning_rate": 1.915421580962467e-05,
+      "loss": 0.2394,
+      "step": 1000
+    },
+    {
+      "epoch": 0.4739863865048831,
+      "grad_norm": 1.6147300004959106,
+      "learning_rate": 1.9152156205251583e-05,
+      "loss": 0.2415,
+      "step": 1001
+    },
+    {
+      "epoch": 0.47445989937851435,
+      "grad_norm": 1.5121729373931885,
+      "learning_rate": 1.9150094207220338e-05,
+      "loss": 0.2285,
+      "step": 1002
+    },
+    {
+      "epoch": 0.4749334122521456,
+      "grad_norm": 2.2810189723968506,
+      "learning_rate": 1.9148029816070223e-05,
+      "loss": 0.2352,
+      "step": 1003
+    },
+    {
+      "epoch": 0.47540692512577687,
+      "grad_norm": 1.55019211769104,
+      "learning_rate": 1.9145963032341163e-05,
+      "loss": 0.273,
+      "step": 1004
+    },
+    {
+      "epoch": 0.4758804379994081,
+      "grad_norm": 1.1222288608551025,
+      "learning_rate": 1.9143893856573702e-05,
+      "loss": 0.2655,
+      "step": 1005
+    },
+    {
+      "epoch": 0.47635395087303933,
+      "grad_norm": 2.17794132232666,
+      "learning_rate": 1.9141822289309016e-05,
+      "loss": 0.2886,
+      "step": 1006
+    },
+    {
+      "epoch": 0.4768274637466706,
+      "grad_norm": 1.7431869506835938,
+      "learning_rate": 1.9139748331088906e-05,
+      "loss": 0.2651,
+      "step": 1007
+    },
+    {
+      "epoch": 0.47730097662030185,
+      "grad_norm": 1.3544774055480957,
+      "learning_rate": 1.913767198245579e-05,
+      "loss": 0.2441,
+      "step": 1008
+    },
+    {
+      "epoch": 0.47777448949393314,
+      "grad_norm": 1.2691947221755981,
+      "learning_rate": 1.9135593243952724e-05,
+      "loss": 0.222,
+      "step": 1009
+    },
+    {
+      "epoch": 0.4782480023675644,
+      "grad_norm": 1.9088751077651978,
+      "learning_rate": 1.913351211612337e-05,
+      "loss": 0.2371,
+      "step": 1010
+    },
+    {
+      "epoch": 0.4787215152411956,
+      "grad_norm": 1.6651183366775513,
+      "learning_rate": 1.9131428599512042e-05,
+      "loss": 0.2516,
+      "step": 1011
+    },
+    {
+      "epoch": 0.4791950281148269,
+      "grad_norm": 1.7003988027572632,
+      "learning_rate": 1.9129342694663655e-05,
+      "loss": 0.2557,
+      "step": 1012
+    },
+    {
+      "epoch": 0.4796685409884581,
+      "grad_norm": 1.6492764949798584,
+      "learning_rate": 1.9127254402123755e-05,
+      "loss": 0.277,
+      "step": 1013
+    },
+    {
+      "epoch": 0.48014205386208936,
+      "grad_norm": 1.4109059572219849,
+      "learning_rate": 1.912516372243852e-05,
+      "loss": 0.2222,
+      "step": 1014
+    },
+    {
+      "epoch": 0.48061556673572065,
+      "grad_norm": 1.9010688066482544,
+      "learning_rate": 1.9123070656154748e-05,
+      "loss": 0.2614,
+      "step": 1015
+    },
+    {
+      "epoch": 0.4810890796093519,
+      "grad_norm": 1.4977048635482788,
+      "learning_rate": 1.9120975203819855e-05,
+      "loss": 0.2262,
+      "step": 1016
+    },
+    {
+      "epoch": 0.4815625924829831,
+      "grad_norm": 1.206935167312622,
+      "learning_rate": 1.9118877365981887e-05,
+      "loss": 0.2591,
+      "step": 1017
+    },
+    {
+      "epoch": 0.4820361053566144,
+      "grad_norm": 1.7453712224960327,
+      "learning_rate": 1.9116777143189517e-05,
+      "loss": 0.2233,
+      "step": 1018
+    },
+    {
+      "epoch": 0.48250961823024563,
+      "grad_norm": 1.1337833404541016,
+      "learning_rate": 1.9114674535992038e-05,
+      "loss": 0.2608,
+      "step": 1019
+    },
+    {
+      "epoch": 0.48298313110387686,
+      "grad_norm": 2.339742422103882,
+      "learning_rate": 1.9112569544939364e-05,
+      "loss": 0.2628,
+      "step": 1020
+    },
+    {
+      "epoch": 0.48345664397750815,
+      "grad_norm": 1.3370864391326904,
+      "learning_rate": 1.9110462170582036e-05,
+      "loss": 0.2664,
+      "step": 1021
+    },
+    {
+      "epoch": 0.4839301568511394,
+      "grad_norm": 1.41420578956604,
+      "learning_rate": 1.9108352413471215e-05,
+      "loss": 0.2721,
+      "step": 1022
+    },
+    {
+      "epoch": 0.48440366972477067,
+      "grad_norm": 1.787891149520874,
+      "learning_rate": 1.9106240274158693e-05,
+      "loss": 0.2511,
+      "step": 1023
+    },
+    {
+      "epoch": 0.4848771825984019,
+      "grad_norm": 1.4634735584259033,
+      "learning_rate": 1.9104125753196876e-05,
+      "loss": 0.2523,
+      "step": 1024
+    },
+    {
+      "epoch": 0.48535069547203313,
+      "grad_norm": 1.4561212062835693,
+      "learning_rate": 1.9102008851138797e-05,
+      "loss": 0.2913,
+      "step": 1025
+    },
+    {
+      "epoch": 0.4858242083456644,
+      "grad_norm": 1.1067405939102173,
+      "learning_rate": 1.9099889568538113e-05,
+      "loss": 0.2476,
+      "step": 1026
+    },
+    {
+      "epoch": 0.48629772121929565,
+      "grad_norm": 1.2803436517715454,
+      "learning_rate": 1.90977679059491e-05,
+      "loss": 0.2446,
+      "step": 1027
+    },
+    {
+      "epoch": 0.4867712340929269,
+      "grad_norm": 1.3610336780548096,
+      "learning_rate": 1.909564386392666e-05,
+      "loss": 0.2365,
+      "step": 1028
+    },
+    {
+      "epoch": 0.4872447469665582,
+      "grad_norm": 1.3339933156967163,
+      "learning_rate": 1.909351744302631e-05,
+      "loss": 0.2395,
+      "step": 1029
+    },
+    {
+      "epoch": 0.4877182598401894,
+      "grad_norm": 2.5949342250823975,
+      "learning_rate": 1.9091388643804202e-05,
+      "loss": 0.2646,
+      "step": 1030
+    },
+    {
+      "epoch": 0.48819177271382064,
+      "grad_norm": 1.3655803203582764,
+      "learning_rate": 1.9089257466817102e-05,
+      "loss": 0.2368,
+      "step": 1031
+    },
+    {
+      "epoch": 0.4886652855874519,
+      "grad_norm": 1.8703043460845947,
+      "learning_rate": 1.9087123912622397e-05,
+      "loss": 0.301,
+      "step": 1032
+    },
+    {
+      "epoch": 0.48913879846108316,
+      "grad_norm": 1.272451400756836,
+      "learning_rate": 1.9084987981778097e-05,
+      "loss": 0.2422,
+      "step": 1033
+    },
+    {
+      "epoch": 0.4896123113347144,
+      "grad_norm": 1.2157567739486694,
+      "learning_rate": 1.9082849674842835e-05,
+      "loss": 0.2515,
+      "step": 1034
+    },
+    {
+      "epoch": 0.4900858242083457,
+      "grad_norm": 1.857765555381775,
+      "learning_rate": 1.9080708992375863e-05,
+      "loss": 0.2427,
+      "step": 1035
+    },
+    {
+      "epoch": 0.4905593370819769,
+      "grad_norm": 1.2299634218215942,
+      "learning_rate": 1.907856593493706e-05,
+      "loss": 0.2659,
+      "step": 1036
+    },
+    {
+      "epoch": 0.49103284995560814,
+      "grad_norm": 1.3312522172927856,
+      "learning_rate": 1.9076420503086915e-05,
+      "loss": 0.2546,
+      "step": 1037
+    },
+    {
+      "epoch": 0.49150636282923943,
+      "grad_norm": 2.030163049697876,
+      "learning_rate": 1.9074272697386554e-05,
+      "loss": 0.2592,
+      "step": 1038
+    },
+    {
+      "epoch": 0.49197987570287066,
+      "grad_norm": 1.262022852897644,
+      "learning_rate": 1.9072122518397706e-05,
+      "loss": 0.2468,
+      "step": 1039
+    },
+    {
+      "epoch": 0.49245338857650195,
+      "grad_norm": 1.5466980934143066,
+      "learning_rate": 1.9069969966682738e-05,
+      "loss": 0.2692,
+      "step": 1040
+    },
+    {
+      "epoch": 0.4929269014501332,
+      "grad_norm": 1.136094093322754,
+      "learning_rate": 1.9067815042804622e-05,
+      "loss": 0.2811,
+      "step": 1041
+    },
+    {
+      "epoch": 0.4934004143237644,
+      "grad_norm": 3.164823532104492,
+      "learning_rate": 1.906565774732696e-05,
+      "loss": 0.252,
+      "step": 1042
+    },
+    {
+      "epoch": 0.4938739271973957,
+      "grad_norm": 1.530435562133789,
+      "learning_rate": 1.9063498080813973e-05,
+      "loss": 0.2797,
+      "step": 1043
+    },
+    {
+      "epoch": 0.49434744007102693,
+      "grad_norm": 1.6616190671920776,
+      "learning_rate": 1.9061336043830498e-05,
+      "loss": 0.2717,
+      "step": 1044
+    },
+    {
+      "epoch": 0.49482095294465817,
+      "grad_norm": 1.4890795946121216,
+      "learning_rate": 1.9059171636942e-05,
+      "loss": 0.2734,
+      "step": 1045
+    },
+    {
+      "epoch": 0.49529446581828945,
+      "grad_norm": 2.622288703918457,
+      "learning_rate": 1.905700486071455e-05,
+      "loss": 0.2473,
+      "step": 1046
+    },
+    {
+      "epoch": 0.4957679786919207,
+      "grad_norm": 1.4621379375457764,
+      "learning_rate": 1.905483571571486e-05,
+      "loss": 0.2511,
+      "step": 1047
+    },
+    {
+      "epoch": 0.4962414915655519,
+      "grad_norm": 1.3425689935684204,
+      "learning_rate": 1.905266420251024e-05,
+      "loss": 0.2709,
+      "step": 1048
+    },
+    {
+      "epoch": 0.4967150044391832,
+      "grad_norm": 1.391316294670105,
+      "learning_rate": 1.905049032166863e-05,
+      "loss": 0.2401,
+      "step": 1049
+    },
+    {
+      "epoch": 0.49718851731281444,
+      "grad_norm": 1.435872197151184,
+      "learning_rate": 1.9048314073758586e-05,
+      "loss": 0.2355,
+      "step": 1050
+    },
+    {
+      "epoch": 0.49766203018644567,
+      "grad_norm": 1.4960476160049438,
+      "learning_rate": 1.9046135459349287e-05,
+      "loss": 0.2619,
+      "step": 1051
+    },
+    {
+      "epoch": 0.49813554306007696,
+      "grad_norm": 2.056912899017334,
+      "learning_rate": 1.9043954479010532e-05,
+      "loss": 0.258,
+      "step": 1052
+    },
+    {
+      "epoch": 0.4986090559337082,
+      "grad_norm": 1.3434464931488037,
+      "learning_rate": 1.9041771133312732e-05,
+      "loss": 0.2469,
+      "step": 1053
+    },
+    {
+      "epoch": 0.4990825688073395,
+      "grad_norm": 1.5715394020080566,
+      "learning_rate": 1.9039585422826916e-05,
+      "loss": 0.2293,
+      "step": 1054
+    },
+    {
+      "epoch": 0.4995560816809707,
+      "grad_norm": 1.3329315185546875,
+      "learning_rate": 1.903739734812474e-05,
+      "loss": 0.2552,
+      "step": 1055
+    },
+    {
+      "epoch": 0.5000295945546019,
+      "grad_norm": 1.1879801750183105,
+      "learning_rate": 1.9035206909778475e-05,
+      "loss": 0.2491,
+      "step": 1056
+    },
+    {
+      "epoch": 0.5005031074282332,
+      "grad_norm": 1.3756626844406128,
+      "learning_rate": 1.9033014108361003e-05,
+      "loss": 0.2371,
+      "step": 1057
+    },
+    {
+      "epoch": 0.5009766203018644,
+      "grad_norm": 1.3430827856063843,
+      "learning_rate": 1.9030818944445836e-05,
+      "loss": 0.2535,
+      "step": 1058
+    },
+    {
+      "epoch": 0.5014501331754957,
+      "grad_norm": 1.2272851467132568,
+      "learning_rate": 1.9028621418607095e-05,
+      "loss": 0.2473,
+      "step": 1059
+    },
+    {
+      "epoch": 0.501923646049127,
+      "grad_norm": 1.979054570198059,
+      "learning_rate": 1.9026421531419522e-05,
+      "loss": 0.2675,
+      "step": 1060
+    },
+    {
+      "epoch": 0.5023971589227582,
+      "grad_norm": 1.8961822986602783,
+      "learning_rate": 1.902421928345848e-05,
+      "loss": 0.2752,
+      "step": 1061
+    },
+    {
+      "epoch": 0.5028706717963894,
+      "grad_norm": 1.5014052391052246,
+      "learning_rate": 1.902201467529994e-05,
+      "loss": 0.2348,
+      "step": 1062
+    },
+    {
+      "epoch": 0.5033441846700207,
+      "grad_norm": 1.413966178894043,
+      "learning_rate": 1.90198077075205e-05,
+      "loss": 0.2568,
+      "step": 1063
+    },
+    {
+      "epoch": 0.503817697543652,
+      "grad_norm": 1.097167730331421,
+      "learning_rate": 1.901759838069737e-05,
+      "loss": 0.2602,
+      "step": 1064
+    },
+    {
+      "epoch": 0.5042912104172832,
+      "grad_norm": 2.0718507766723633,
+      "learning_rate": 1.9015386695408377e-05,
+      "loss": 0.2524,
+      "step": 1065
+    },
+    {
+      "epoch": 0.5047647232909145,
+      "grad_norm": 1.8669925928115845,
+      "learning_rate": 1.9013172652231967e-05,
+      "loss": 0.2333,
+      "step": 1066
+    },
+    {
+      "epoch": 0.5052382361645458,
+      "grad_norm": 1.3957126140594482,
+      "learning_rate": 1.9010956251747202e-05,
+      "loss": 0.266,
+      "step": 1067
+    },
+    {
+      "epoch": 0.505711749038177,
+      "grad_norm": 2.0462698936462402,
+      "learning_rate": 1.9008737494533757e-05,
+      "loss": 0.2846,
+      "step": 1068
+    },
+    {
+      "epoch": 0.5061852619118082,
+      "grad_norm": 1.195210576057434,
+      "learning_rate": 1.9006516381171933e-05,
+      "loss": 0.2851,
+      "step": 1069
+    },
+    {
+      "epoch": 0.5066587747854395,
+      "grad_norm": 1.3141087293624878,
+      "learning_rate": 1.9004292912242634e-05,
+      "loss": 0.2393,
+      "step": 1070
+    },
+    {
+      "epoch": 0.5071322876590707,
+      "grad_norm": 1.6447049379348755,
+      "learning_rate": 1.900206708832739e-05,
+      "loss": 0.2711,
+      "step": 1071
+    },
+    {
+      "epoch": 0.507605800532702,
+      "grad_norm": 2.366652727127075,
+      "learning_rate": 1.8999838910008347e-05,
+      "loss": 0.2233,
+      "step": 1072
+    },
+    {
+      "epoch": 0.5080793134063333,
+      "grad_norm": 1.6027371883392334,
+      "learning_rate": 1.8997608377868256e-05,
+      "loss": 0.2689,
+      "step": 1073
+    },
+    {
+      "epoch": 0.5085528262799645,
+      "grad_norm": 1.4096899032592773,
+      "learning_rate": 1.8995375492490495e-05,
+      "loss": 0.2539,
+      "step": 1074
+    },
+    {
+      "epoch": 0.5090263391535957,
+      "grad_norm": 1.179916501045227,
+      "learning_rate": 1.8993140254459057e-05,
+      "loss": 0.2667,
+      "step": 1075
+    },
+    {
+      "epoch": 0.509499852027227,
+      "grad_norm": 1.141992449760437,
+      "learning_rate": 1.8990902664358542e-05,
+      "loss": 0.2461,
+      "step": 1076
+    },
+    {
+      "epoch": 0.5099733649008582,
+      "grad_norm": 1.3046503067016602,
+      "learning_rate": 1.8988662722774172e-05,
+      "loss": 0.2607,
+      "step": 1077
+    },
+    {
+      "epoch": 0.5104468777744895,
+      "grad_norm": 1.1678529977798462,
+      "learning_rate": 1.898642043029178e-05,
+      "loss": 0.2352,
+      "step": 1078
+    },
+    {
+      "epoch": 0.5109203906481208,
+      "grad_norm": 1.4803732633590698,
+      "learning_rate": 1.8984175787497822e-05,
+      "loss": 0.2612,
+      "step": 1079
+    },
+    {
+      "epoch": 0.511393903521752,
+      "grad_norm": 1.7487013339996338,
+      "learning_rate": 1.898192879497936e-05,
+      "loss": 0.2348,
+      "step": 1080
+    },
+    {
+      "epoch": 0.5118674163953832,
+      "grad_norm": 1.448327898979187,
+      "learning_rate": 1.8979679453324068e-05,
+      "loss": 0.261,
+      "step": 1081
+    },
+    {
+      "epoch": 0.5123409292690145,
+      "grad_norm": 1.1041336059570312,
+      "learning_rate": 1.8977427763120242e-05,
+      "loss": 0.2754,
+      "step": 1082
+    },
+    {
+      "epoch": 0.5128144421426457,
+      "grad_norm": 1.948614239692688,
+      "learning_rate": 1.8975173724956794e-05,
+      "loss": 0.2594,
+      "step": 1083
+    },
+    {
+      "epoch": 0.513287955016277,
+      "grad_norm": 1.1922783851623535,
+      "learning_rate": 1.897291733942324e-05,
+      "loss": 0.2457,
+      "step": 1084
+    },
+    {
+      "epoch": 0.5137614678899083,
+      "grad_norm": 1.4861289262771606,
+      "learning_rate": 1.8970658607109723e-05,
+      "loss": 0.2459,
+      "step": 1085
+    },
+    {
+      "epoch": 0.5142349807635395,
+      "grad_norm": 1.497668981552124,
+      "learning_rate": 1.896839752860699e-05,
+      "loss": 0.2828,
+      "step": 1086
+    },
+    {
+      "epoch": 0.5147084936371707,
+      "grad_norm": 1.194231390953064,
+      "learning_rate": 1.89661341045064e-05,
+      "loss": 0.2524,
+      "step": 1087
+    },
+    {
+      "epoch": 0.515182006510802,
+      "grad_norm": 1.5421984195709229,
+      "learning_rate": 1.8963868335399933e-05,
+      "loss": 0.2355,
+      "step": 1088
+    },
+    {
+      "epoch": 0.5156555193844332,
+      "grad_norm": 1.1698225736618042,
+      "learning_rate": 1.8961600221880177e-05,
+      "loss": 0.253,
+      "step": 1089
+    },
+    {
+      "epoch": 0.5161290322580645,
+      "grad_norm": 1.2847546339035034,
+      "learning_rate": 1.8959329764540343e-05,
+      "loss": 0.2463,
+      "step": 1090
+    },
+    {
+      "epoch": 0.5166025451316958,
+      "grad_norm": 1.8415346145629883,
+      "learning_rate": 1.895705696397424e-05,
+      "loss": 0.2602,
+      "step": 1091
+    },
+    {
+      "epoch": 0.517076058005327,
+      "grad_norm": 1.1953470706939697,
+      "learning_rate": 1.8954781820776293e-05,
+      "loss": 0.2544,
+      "step": 1092
+    },
+    {
+      "epoch": 0.5175495708789583,
+      "grad_norm": 1.2167423963546753,
+      "learning_rate": 1.8952504335541554e-05,
+      "loss": 0.2353,
+      "step": 1093
+    },
+    {
+      "epoch": 0.5180230837525895,
+      "grad_norm": 1.259513258934021,
+      "learning_rate": 1.8950224508865667e-05,
+      "loss": 0.2395,
+      "step": 1094
+    },
+    {
+      "epoch": 0.5184965966262208,
+      "grad_norm": 1.6243139505386353,
+      "learning_rate": 1.894794234134491e-05,
+      "loss": 0.2362,
+      "step": 1095
+    },
+    {
+      "epoch": 0.518970109499852,
+      "grad_norm": 1.0280331373214722,
+      "learning_rate": 1.8945657833576155e-05,
+      "loss": 0.2476,
+      "step": 1096
+    },
+    {
+      "epoch": 0.5194436223734833,
+      "grad_norm": 1.332411289215088,
+      "learning_rate": 1.894337098615689e-05,
+      "loss": 0.2551,
+      "step": 1097
+    },
+    {
+      "epoch": 0.5199171352471146,
+      "grad_norm": 1.2964811325073242,
+      "learning_rate": 1.8941081799685227e-05,
+      "loss": 0.2598,
+      "step": 1098
+    },
+    {
+      "epoch": 0.5203906481207458,
+      "grad_norm": 1.6705334186553955,
+      "learning_rate": 1.893879027475987e-05,
+      "loss": 0.2639,
+      "step": 1099
+    },
+    {
+      "epoch": 0.520864160994377,
+      "grad_norm": 1.3009783029556274,
+      "learning_rate": 1.8936496411980156e-05,
+      "loss": 0.2803,
+      "step": 1100
+    },
+    {
+      "epoch": 0.5213376738680083,
+      "grad_norm": 1.179648995399475,
+      "learning_rate": 1.8934200211946013e-05,
+      "loss": 0.2561,
+      "step": 1101
+    },
+    {
+      "epoch": 0.5218111867416395,
+      "grad_norm": 1.4020172357559204,
+      "learning_rate": 1.8931901675257996e-05,
+      "loss": 0.2533,
+      "step": 1102
+    },
+    {
+      "epoch": 0.5222846996152708,
+      "grad_norm": 1.20439875125885,
+      "learning_rate": 1.8929600802517262e-05,
+      "loss": 0.2684,
+      "step": 1103
+    },
+    {
+      "epoch": 0.5227582124889021,
+      "grad_norm": 1.7954668998718262,
+      "learning_rate": 1.8927297594325583e-05,
+      "loss": 0.2242,
+      "step": 1104
+    },
+    {
+      "epoch": 0.5232317253625333,
+      "grad_norm": 1.4614356756210327,
+      "learning_rate": 1.8924992051285345e-05,
+      "loss": 0.2541,
+      "step": 1105
+    },
+    {
+      "epoch": 0.5237052382361645,
+      "grad_norm": 1.306846022605896,
+      "learning_rate": 1.8922684173999538e-05,
+      "loss": 0.2382,
+      "step": 1106
+    },
+    {
+      "epoch": 0.5241787511097958,
+      "grad_norm": 1.7092262506484985,
+      "learning_rate": 1.8920373963071757e-05,
+      "loss": 0.2512,
+      "step": 1107
+    },
+    {
+      "epoch": 0.524652263983427,
+      "grad_norm": 1.5031977891921997,
+      "learning_rate": 1.8918061419106228e-05,
+      "loss": 0.2581,
+      "step": 1108
+    },
+    {
+      "epoch": 0.5251257768570583,
+      "grad_norm": 1.5629005432128906,
+      "learning_rate": 1.8915746542707767e-05,
+      "loss": 0.2439,
+      "step": 1109
+    },
+    {
+      "epoch": 0.5255992897306896,
+      "grad_norm": 1.4823302030563354,
+      "learning_rate": 1.891342933448181e-05,
+      "loss": 0.2734,
+      "step": 1110
+    },
+    {
+      "epoch": 0.5260728026043208,
+      "grad_norm": 1.880656123161316,
+      "learning_rate": 1.89111097950344e-05,
+      "loss": 0.2425,
+      "step": 1111
+    },
+    {
+      "epoch": 0.526546315477952,
+      "grad_norm": 1.1813774108886719,
+      "learning_rate": 1.8908787924972188e-05,
+      "loss": 0.2375,
+      "step": 1112
+    },
+    {
+      "epoch": 0.5270198283515833,
+      "grad_norm": 1.9654037952423096,
+      "learning_rate": 1.8906463724902443e-05,
+      "loss": 0.2817,
+      "step": 1113
+    },
+    {
+      "epoch": 0.5274933412252145,
+      "grad_norm": 1.7562179565429688,
+      "learning_rate": 1.890413719543303e-05,
+      "loss": 0.256,
+      "step": 1114
+    },
+    {
+      "epoch": 0.5279668540988458,
+      "grad_norm": 1.140051245689392,
+      "learning_rate": 1.8901808337172433e-05,
+      "loss": 0.2455,
+      "step": 1115
+    },
+    {
+      "epoch": 0.5284403669724771,
+      "grad_norm": 1.8479341268539429,
+      "learning_rate": 1.8899477150729745e-05,
+      "loss": 0.2571,
+      "step": 1116
+    },
+    {
+      "epoch": 0.5289138798461083,
+      "grad_norm": 1.7184622287750244,
+      "learning_rate": 1.889714363671466e-05,
+      "loss": 0.2463,
+      "step": 1117
+    },
+    {
+      "epoch": 0.5293873927197396,
+      "grad_norm": 1.3186061382293701,
+      "learning_rate": 1.8894807795737492e-05,
+      "loss": 0.2657,
+      "step": 1118
+    },
+    {
+      "epoch": 0.5298609055933708,
+      "grad_norm": 1.8647300004959106,
+      "learning_rate": 1.8892469628409157e-05,
+      "loss": 0.2505,
+      "step": 1119
+    },
+    {
+      "epoch": 0.530334418467002,
+      "grad_norm": 2.4725730419158936,
+      "learning_rate": 1.889012913534117e-05,
+      "loss": 0.2475,
+      "step": 1120
+    },
+    {
+      "epoch": 0.5308079313406333,
+      "grad_norm": 1.2804913520812988,
+      "learning_rate": 1.8887786317145677e-05,
+      "loss": 0.2607,
+      "step": 1121
+    },
+    {
+      "epoch": 0.5312814442142646,
+      "grad_norm": 1.6455892324447632,
+      "learning_rate": 1.8885441174435417e-05,
+      "loss": 0.2273,
+      "step": 1122
+    },
+    {
+      "epoch": 0.5317549570878958,
+      "grad_norm": 1.4230307340621948,
+      "learning_rate": 1.8883093707823733e-05,
+      "loss": 0.2571,
+      "step": 1123
+    },
+    {
+      "epoch": 0.5322284699615271,
+      "grad_norm": 2.230576992034912,
+      "learning_rate": 1.8880743917924585e-05,
+      "loss": 0.2418,
+      "step": 1124
+    },
+    {
+      "epoch": 0.5327019828351583,
+      "grad_norm": 1.6521943807601929,
+      "learning_rate": 1.8878391805352544e-05,
+      "loss": 0.2537,
+      "step": 1125
+    },
+    {
+      "epoch": 0.5331754957087896,
+      "grad_norm": 1.336004376411438,
+      "learning_rate": 1.887603737072278e-05,
+      "loss": 0.243,
+      "step": 1126
+    },
+    {
+      "epoch": 0.5336490085824208,
+      "grad_norm": 1.650602102279663,
+      "learning_rate": 1.8873680614651065e-05,
+      "loss": 0.2544,
+      "step": 1127
+    },
+    {
+      "epoch": 0.5341225214560521,
+      "grad_norm": 1.5167171955108643,
+      "learning_rate": 1.8871321537753792e-05,
+      "loss": 0.2596,
+      "step": 1128
+    },
+    {
+      "epoch": 0.5345960343296834,
+      "grad_norm": 1.3937066793441772,
+      "learning_rate": 1.8868960140647953e-05,
+      "loss": 0.2398,
+      "step": 1129
+    },
+    {
+      "epoch": 0.5350695472033146,
+      "grad_norm": 1.1855547428131104,
+      "learning_rate": 1.886659642395115e-05,
+      "loss": 0.2312,
+      "step": 1130
+    },
+    {
+      "epoch": 0.5355430600769459,
+      "grad_norm": 1.1568256616592407,
+      "learning_rate": 1.8864230388281594e-05,
+      "loss": 0.2481,
+      "step": 1131
+    },
+    {
+      "epoch": 0.5360165729505771,
+      "grad_norm": 1.2066378593444824,
+      "learning_rate": 1.886186203425809e-05,
+      "loss": 0.2306,
+      "step": 1132
+    },
+    {
+      "epoch": 0.5364900858242083,
+      "grad_norm": 1.1401264667510986,
+      "learning_rate": 1.8859491362500066e-05,
+      "loss": 0.2538,
+      "step": 1133
+    },
+    {
+      "epoch": 0.5369635986978396,
+      "grad_norm": 1.3635470867156982,
+      "learning_rate": 1.885711837362754e-05,
+      "loss": 0.247,
+      "step": 1134
+    },
+    {
+      "epoch": 0.5374371115714709,
+      "grad_norm": 1.619167685508728,
+      "learning_rate": 1.8854743068261154e-05,
+      "loss": 0.2123,
+      "step": 1135
+    },
+    {
+      "epoch": 0.5379106244451021,
+      "grad_norm": 1.432413101196289,
+      "learning_rate": 1.885236544702214e-05,
+      "loss": 0.2377,
+      "step": 1136
+    },
+    {
+      "epoch": 0.5383841373187334,
+      "grad_norm": 2.2878897190093994,
+      "learning_rate": 1.8849985510532348e-05,
+      "loss": 0.2466,
+      "step": 1137
+    },
+    {
+      "epoch": 0.5388576501923646,
+      "grad_norm": 1.848306655883789,
+      "learning_rate": 1.8847603259414215e-05,
+      "loss": 0.2458,
+      "step": 1138
+    },
+    {
+      "epoch": 0.5393311630659958,
+      "grad_norm": 1.1554219722747803,
+      "learning_rate": 1.884521869429081e-05,
+      "loss": 0.2542,
+      "step": 1139
+    },
+    {
+      "epoch": 0.5398046759396271,
+      "grad_norm": 1.239094614982605,
+      "learning_rate": 1.8842831815785783e-05,
+      "loss": 0.2569,
+      "step": 1140
+    },
+    {
+      "epoch": 0.5402781888132584,
+      "grad_norm": 1.240552544593811,
+      "learning_rate": 1.8840442624523408e-05,
+      "loss": 0.2391,
+      "step": 1141
+    },
+    {
+      "epoch": 0.5407517016868896,
+      "grad_norm": 1.8577840328216553,
+      "learning_rate": 1.8838051121128545e-05,
+      "loss": 0.224,
+      "step": 1142
+    },
+    {
+      "epoch": 0.5412252145605209,
+      "grad_norm": 1.2552670240402222,
+      "learning_rate": 1.883565730622668e-05,
+      "loss": 0.2271,
+      "step": 1143
+    },
+    {
+      "epoch": 0.5416987274341521,
+      "grad_norm": 1.31777822971344,
+      "learning_rate": 1.8833261180443877e-05,
+      "loss": 0.2579,
+      "step": 1144
+    },
+    {
+      "epoch": 0.5421722403077833,
+      "grad_norm": 1.180493950843811,
+      "learning_rate": 1.8830862744406834e-05,
+      "loss": 0.2437,
+      "step": 1145
+    },
+    {
+      "epoch": 0.5426457531814146,
+      "grad_norm": 1.399220585823059,
+      "learning_rate": 1.8828461998742827e-05,
+      "loss": 0.2548,
+      "step": 1146
+    },
+    {
+      "epoch": 0.5431192660550459,
+      "grad_norm": 1.4404245615005493,
+      "learning_rate": 1.8826058944079763e-05,
+      "loss": 0.2314,
+      "step": 1147
+    },
+    {
+      "epoch": 0.5435927789286771,
+      "grad_norm": 1.5792372226715088,
+      "learning_rate": 1.8823653581046122e-05,
+      "loss": 0.2487,
+      "step": 1148
+    },
+    {
+      "epoch": 0.5440662918023084,
+      "grad_norm": 1.5916982889175415,
+      "learning_rate": 1.8821245910271013e-05,
+      "loss": 0.2647,
+      "step": 1149
+    },
+    {
+      "epoch": 0.5445398046759397,
+      "grad_norm": 1.326357364654541,
+      "learning_rate": 1.8818835932384133e-05,
+      "loss": 0.2418,
+      "step": 1150
+    },
+    {
+      "epoch": 0.5450133175495708,
+      "grad_norm": 1.949175238609314,
+      "learning_rate": 1.8816423648015795e-05,
+      "loss": 0.252,
+      "step": 1151
+    },
+    {
+      "epoch": 0.5454868304232021,
+      "grad_norm": 1.1373693943023682,
+      "learning_rate": 1.88140090577969e-05,
+      "loss": 0.2407,
+      "step": 1152
+    },
+    {
+      "epoch": 0.5459603432968334,
+      "grad_norm": 1.1092511415481567,
+      "learning_rate": 1.8811592162358977e-05,
+      "loss": 0.2749,
+      "step": 1153
+    },
+    {
+      "epoch": 0.5464338561704646,
+      "grad_norm": 2.095144510269165,
+      "learning_rate": 1.8809172962334124e-05,
+      "loss": 0.2357,
+      "step": 1154
+    },
+    {
+      "epoch": 0.5469073690440959,
+      "grad_norm": 2.180050849914551,
+      "learning_rate": 1.8806751458355064e-05,
+      "loss": 0.2326,
+      "step": 1155
+    },
+    {
+      "epoch": 0.5473808819177272,
+      "grad_norm": 1.321502923965454,
+      "learning_rate": 1.8804327651055123e-05,
+      "loss": 0.2304,
+      "step": 1156
+    },
+    {
+      "epoch": 0.5478543947913584,
+      "grad_norm": 1.3724721670150757,
+      "learning_rate": 1.8801901541068224e-05,
+      "loss": 0.2372,
+      "step": 1157
+    },
+    {
+      "epoch": 0.5483279076649896,
+      "grad_norm": 1.1256623268127441,
+      "learning_rate": 1.8799473129028886e-05,
+      "loss": 0.2449,
+      "step": 1158
+    },
+    {
+      "epoch": 0.5488014205386209,
+      "grad_norm": 1.4590364694595337,
+      "learning_rate": 1.8797042415572244e-05,
+      "loss": 0.2559,
+      "step": 1159
+    },
+    {
+      "epoch": 0.5492749334122522,
+      "grad_norm": 1.8025602102279663,
+      "learning_rate": 1.879460940133402e-05,
+      "loss": 0.2778,
+      "step": 1160
+    },
+    {
+      "epoch": 0.5497484462858834,
+      "grad_norm": 1.1516093015670776,
+      "learning_rate": 1.879217408695056e-05,
+      "loss": 0.2249,
+      "step": 1161
+    },
+    {
+      "epoch": 0.5502219591595147,
+      "grad_norm": 1.4225125312805176,
+      "learning_rate": 1.8789736473058776e-05,
+      "loss": 0.256,
+      "step": 1162
+    },
+    {
+      "epoch": 0.550695472033146,
+      "grad_norm": 1.2126541137695312,
+      "learning_rate": 1.8787296560296223e-05,
+      "loss": 0.2269,
+      "step": 1163
+    },
+    {
+      "epoch": 0.5511689849067771,
+      "grad_norm": 1.0485234260559082,
+      "learning_rate": 1.8784854349301023e-05,
+      "loss": 0.231,
+      "step": 1164
+    },
+    {
+      "epoch": 0.5516424977804084,
+      "grad_norm": 2.125662326812744,
+      "learning_rate": 1.8782409840711916e-05,
+      "loss": 0.2419,
+      "step": 1165
+    },
+    {
+      "epoch": 0.5521160106540397,
+      "grad_norm": 2.186833620071411,
+      "learning_rate": 1.877996303516824e-05,
+      "loss": 0.2685,
+      "step": 1166
+    },
+    {
+      "epoch": 0.5525895235276709,
+      "grad_norm": 1.3703869581222534,
+      "learning_rate": 1.8777513933309938e-05,
+      "loss": 0.2663,
+      "step": 1167
+    },
+    {
+      "epoch": 0.5530630364013022,
+      "grad_norm": 1.2371742725372314,
+      "learning_rate": 1.877506253577754e-05,
+      "loss": 0.2626,
+      "step": 1168
+    },
+    {
+      "epoch": 0.5535365492749335,
+      "grad_norm": 1.1112666130065918,
+      "learning_rate": 1.877260884321219e-05,
+      "loss": 0.2318,
+      "step": 1169
+    },
+    {
+      "epoch": 0.5540100621485646,
+      "grad_norm": 1.446602463722229,
+      "learning_rate": 1.8770152856255636e-05,
+      "loss": 0.2662,
+      "step": 1170
+    },
+    {
+      "epoch": 0.5544835750221959,
+      "grad_norm": 1.32243013381958,
+      "learning_rate": 1.87676945755502e-05,
+      "loss": 0.2765,
+      "step": 1171
+    },
+    {
+      "epoch": 0.5549570878958272,
+      "grad_norm": 1.6575604677200317,
+      "learning_rate": 1.8765234001738838e-05,
+      "loss": 0.2783,
+      "step": 1172
+    },
+    {
+      "epoch": 0.5554306007694584,
+      "grad_norm": 1.1545252799987793,
+      "learning_rate": 1.8762771135465078e-05,
+      "loss": 0.2414,
+      "step": 1173
+    },
+    {
+      "epoch": 0.5559041136430897,
+      "grad_norm": 1.1764464378356934,
+      "learning_rate": 1.8760305977373067e-05,
+      "loss": 0.2403,
+      "step": 1174
+    },
+    {
+      "epoch": 0.556377626516721,
+      "grad_norm": 1.4649169445037842,
+      "learning_rate": 1.875783852810754e-05,
+      "loss": 0.2466,
+      "step": 1175
+    },
+    {
+      "epoch": 0.5568511393903521,
+      "grad_norm": 1.1785379648208618,
+      "learning_rate": 1.8755368788313834e-05,
+      "loss": 0.2363,
+      "step": 1176
+    },
+    {
+      "epoch": 0.5573246522639834,
+      "grad_norm": 1.8420625925064087,
+      "learning_rate": 1.8752896758637884e-05,
+      "loss": 0.2261,
+      "step": 1177
+    },
+    {
+      "epoch": 0.5577981651376147,
+      "grad_norm": 1.241428256034851,
+      "learning_rate": 1.8750422439726227e-05,
+      "loss": 0.2642,
+      "step": 1178
+    },
+    {
+      "epoch": 0.5582716780112459,
+      "grad_norm": 2.547609567642212,
+      "learning_rate": 1.8747945832226e-05,
+      "loss": 0.2315,
+      "step": 1179
+    },
+    {
+      "epoch": 0.5587451908848772,
+      "grad_norm": 1.8056316375732422,
+      "learning_rate": 1.874546693678493e-05,
+      "loss": 0.2672,
+      "step": 1180
+    },
+    {
+      "epoch": 0.5592187037585085,
+      "grad_norm": 1.3478612899780273,
+      "learning_rate": 1.874298575405135e-05,
+      "loss": 0.217,
+      "step": 1181
+    },
+    {
+      "epoch": 0.5596922166321396,
+      "grad_norm": 1.2503442764282227,
+      "learning_rate": 1.8740502284674197e-05,
+      "loss": 0.2321,
+      "step": 1182
+    },
+    {
+      "epoch": 0.5601657295057709,
+      "grad_norm": 2.2722413539886475,
+      "learning_rate": 1.873801652930299e-05,
+      "loss": 0.2657,
+      "step": 1183
+    },
+    {
+      "epoch": 0.5606392423794022,
+      "grad_norm": 1.1966623067855835,
+      "learning_rate": 1.873552848858786e-05,
+      "loss": 0.2579,
+      "step": 1184
+    },
+    {
+      "epoch": 0.5611127552530334,
+      "grad_norm": 1.2810187339782715,
+      "learning_rate": 1.8733038163179524e-05,
+      "loss": 0.2727,
+      "step": 1185
+    },
+    {
+      "epoch": 0.5615862681266647,
+      "grad_norm": 1.2064369916915894,
+      "learning_rate": 1.8730545553729306e-05,
+      "loss": 0.2572,
+      "step": 1186
+    },
+    {
+      "epoch": 0.562059781000296,
+      "grad_norm": 1.379677653312683,
+      "learning_rate": 1.8728050660889123e-05,
+      "loss": 0.2622,
+      "step": 1187
+    },
+    {
+      "epoch": 0.5625332938739271,
+      "grad_norm": 0.9348089098930359,
+      "learning_rate": 1.8725553485311492e-05,
+      "loss": 0.2304,
+      "step": 1188
+    },
+    {
+      "epoch": 0.5630068067475584,
+      "grad_norm": 2.2606005668640137,
+      "learning_rate": 1.872305402764952e-05,
+      "loss": 0.2673,
+      "step": 1189
+    },
+    {
+      "epoch": 0.5634803196211897,
+      "grad_norm": 1.8432506322860718,
+      "learning_rate": 1.8720552288556926e-05,
+      "loss": 0.2779,
+      "step": 1190
+    },
+    {
+      "epoch": 0.563953832494821,
+      "grad_norm": 1.0819976329803467,
+      "learning_rate": 1.8718048268688006e-05,
+      "loss": 0.2349,
+      "step": 1191
+    },
+    {
+      "epoch": 0.5644273453684522,
+      "grad_norm": 1.164947748184204,
+      "learning_rate": 1.8715541968697666e-05,
+      "loss": 0.2575,
+      "step": 1192
+    },
+    {
+      "epoch": 0.5649008582420835,
+      "grad_norm": 1.1525195837020874,
+      "learning_rate": 1.8713033389241406e-05,
+      "loss": 0.2653,
+      "step": 1193
+    },
+    {
+      "epoch": 0.5653743711157148,
+      "grad_norm": 1.0925800800323486,
+      "learning_rate": 1.8710522530975315e-05,
+      "loss": 0.2368,
+      "step": 1194
+    },
+    {
+      "epoch": 0.5658478839893459,
+      "grad_norm": 1.066985845565796,
+      "learning_rate": 1.870800939455609e-05,
+      "loss": 0.2427,
+      "step": 1195
+    },
+    {
+      "epoch": 0.5663213968629772,
+      "grad_norm": 1.2804023027420044,
+      "learning_rate": 1.8705493980641017e-05,
+      "loss": 0.2649,
+      "step": 1196
+    },
+    {
+      "epoch": 0.5667949097366085,
+      "grad_norm": 1.0042173862457275,
+      "learning_rate": 1.8702976289887978e-05,
+      "loss": 0.2408,
+      "step": 1197
+    },
+    {
+      "epoch": 0.5672684226102397,
+      "grad_norm": 1.1509785652160645,
+      "learning_rate": 1.8700456322955446e-05,
+      "loss": 0.2527,
+      "step": 1198
+    },
+    {
+      "epoch": 0.567741935483871,
+      "grad_norm": 1.66000235080719,
+      "learning_rate": 1.8697934080502498e-05,
+      "loss": 0.2614,
+      "step": 1199
+    },
+    {
+      "epoch": 0.5682154483575023,
+      "grad_norm": 1.2652623653411865,
+      "learning_rate": 1.8695409563188803e-05,
+      "loss": 0.2366,
+      "step": 1200
+    },
+    {
+      "epoch": 0.5686889612311334,
+      "grad_norm": 1.2608882188796997,
+      "learning_rate": 1.8692882771674624e-05,
+      "loss": 0.2625,
+      "step": 1201
+    },
+    {
+      "epoch": 0.5691624741047647,
+      "grad_norm": 1.0795193910598755,
+      "learning_rate": 1.8690353706620815e-05,
+      "loss": 0.2497,
+      "step": 1202
+    },
+    {
+      "epoch": 0.569635986978396,
+      "grad_norm": 1.6654939651489258,
+      "learning_rate": 1.868782236868883e-05,
+      "loss": 0.2668,
+      "step": 1203
+    },
+    {
+      "epoch": 0.5701094998520272,
+      "grad_norm": 1.3972606658935547,
+      "learning_rate": 1.8685288758540724e-05,
+      "loss": 0.2451,
+      "step": 1204
+    },
+    {
+      "epoch": 0.5705830127256585,
+      "grad_norm": 1.5168894529342651,
+      "learning_rate": 1.8682752876839127e-05,
+      "loss": 0.2625,
+      "step": 1205
+    },
+    {
+      "epoch": 0.5710565255992898,
+      "grad_norm": 1.420271873474121,
+      "learning_rate": 1.868021472424728e-05,
+      "loss": 0.2467,
+      "step": 1206
+    },
+    {
+      "epoch": 0.5715300384729209,
+      "grad_norm": 1.3395509719848633,
+      "learning_rate": 1.8677674301429012e-05,
+      "loss": 0.2397,
+      "step": 1207
+    },
+    {
+      "epoch": 0.5720035513465522,
+      "grad_norm": 1.5764024257659912,
+      "learning_rate": 1.8675131609048742e-05,
+      "loss": 0.2556,
+      "step": 1208
+    },
+    {
+      "epoch": 0.5724770642201835,
+      "grad_norm": 1.2619208097457886,
+      "learning_rate": 1.8672586647771496e-05,
+      "loss": 0.2772,
+      "step": 1209
+    },
+    {
+      "epoch": 0.5729505770938147,
+      "grad_norm": 1.2036354541778564,
+      "learning_rate": 1.8670039418262873e-05,
+      "loss": 0.2276,
+      "step": 1210
+    },
+    {
+      "epoch": 0.573424089967446,
+      "grad_norm": 1.003892183303833,
+      "learning_rate": 1.8667489921189083e-05,
+      "loss": 0.2257,
+      "step": 1211
+    },
+    {
+      "epoch": 0.5738976028410773,
+      "grad_norm": 1.2825255393981934,
+      "learning_rate": 1.8664938157216923e-05,
+      "loss": 0.26,
+      "step": 1212
+    },
+    {
+      "epoch": 0.5743711157147084,
+      "grad_norm": 1.5353515148162842,
+      "learning_rate": 1.866238412701378e-05,
+      "loss": 0.2344,
+      "step": 1213
+    },
+    {
+      "epoch": 0.5748446285883397,
+      "grad_norm": 1.1604024171829224,
+      "learning_rate": 1.8659827831247632e-05,
+      "loss": 0.2378,
+      "step": 1214
+    },
+    {
+      "epoch": 0.575318141461971,
+      "grad_norm": 1.5623371601104736,
+      "learning_rate": 1.865726927058706e-05,
+      "loss": 0.2597,
+      "step": 1215
+    },
+    {
+      "epoch": 0.5757916543356022,
+      "grad_norm": 1.4064077138900757,
+      "learning_rate": 1.8654708445701227e-05,
+      "loss": 0.2308,
+      "step": 1216
+    },
+    {
+      "epoch": 0.5762651672092335,
+      "grad_norm": 1.2463563680648804,
+      "learning_rate": 1.8652145357259897e-05,
+      "loss": 0.2289,
+      "step": 1217
+    },
+    {
+      "epoch": 0.5767386800828648,
+      "grad_norm": 1.6035746335983276,
+      "learning_rate": 1.8649580005933415e-05,
+      "loss": 0.2724,
+      "step": 1218
+    },
+    {
+      "epoch": 0.577212192956496,
+      "grad_norm": 1.4112597703933716,
+      "learning_rate": 1.8647012392392728e-05,
+      "loss": 0.255,
+      "step": 1219
+    },
+    {
+      "epoch": 0.5776857058301272,
+      "grad_norm": 1.4282629489898682,
+      "learning_rate": 1.8644442517309366e-05,
+      "loss": 0.2478,
+      "step": 1220
+    },
+    {
+      "epoch": 0.5781592187037585,
+      "grad_norm": 1.234606385231018,
+      "learning_rate": 1.8641870381355463e-05,
+      "loss": 0.2597,
+      "step": 1221
+    },
+    {
+      "epoch": 0.5786327315773898,
+      "grad_norm": 1.283451795578003,
+      "learning_rate": 1.8639295985203726e-05,
+      "loss": 0.2724,
+      "step": 1222
+    },
+    {
+      "epoch": 0.579106244451021,
+      "grad_norm": 1.5409351587295532,
+      "learning_rate": 1.8636719329527474e-05,
+      "loss": 0.267,
+      "step": 1223
+    },
+    {
+      "epoch": 0.5795797573246523,
+      "grad_norm": 1.1971688270568848,
+      "learning_rate": 1.8634140415000595e-05,
+      "loss": 0.2635,
+      "step": 1224
+    },
+    {
+      "epoch": 0.5800532701982836,
+      "grad_norm": 1.3609153032302856,
+      "learning_rate": 1.863155924229759e-05,
+      "loss": 0.2479,
+      "step": 1225
+    },
+    {
+      "epoch": 0.5805267830719147,
+      "grad_norm": 1.1887803077697754,
+      "learning_rate": 1.8628975812093535e-05,
+      "loss": 0.2453,
+      "step": 1226
+    },
+    {
+      "epoch": 0.581000295945546,
+      "grad_norm": 2.107870101928711,
+      "learning_rate": 1.86263901250641e-05,
+      "loss": 0.249,
+      "step": 1227
+    },
+    {
+      "epoch": 0.5814738088191773,
+      "grad_norm": 1.9113880395889282,
+      "learning_rate": 1.8623802181885548e-05,
+      "loss": 0.2569,
+      "step": 1228
+    },
+    {
+      "epoch": 0.5819473216928085,
+      "grad_norm": 1.2835649251937866,
+      "learning_rate": 1.862121198323473e-05,
+      "loss": 0.2462,
+      "step": 1229
+    },
+    {
+      "epoch": 0.5824208345664398,
+      "grad_norm": 1.388893723487854,
+      "learning_rate": 1.861861952978909e-05,
+      "loss": 0.2582,
+      "step": 1230
+    },
+    {
+      "epoch": 0.5828943474400711,
+      "grad_norm": 1.21442449092865,
+      "learning_rate": 1.861602482222666e-05,
+      "loss": 0.2257,
+      "step": 1231
+    },
+    {
+      "epoch": 0.5833678603137022,
+      "grad_norm": 1.585196614265442,
+      "learning_rate": 1.8613427861226056e-05,
+      "loss": 0.2475,
+      "step": 1232
+    },
+    {
+      "epoch": 0.5838413731873335,
+      "grad_norm": 2.5939364433288574,
+      "learning_rate": 1.8610828647466487e-05,
+      "loss": 0.2519,
+      "step": 1233
+    },
+    {
+      "epoch": 0.5843148860609648,
+      "grad_norm": 1.4720892906188965,
+      "learning_rate": 1.8608227181627757e-05,
+      "loss": 0.2598,
+      "step": 1234
+    },
+    {
+      "epoch": 0.584788398934596,
+      "grad_norm": 1.8123221397399902,
+      "learning_rate": 1.860562346439025e-05,
+      "loss": 0.2372,
+      "step": 1235
+    },
+    {
+      "epoch": 0.5852619118082273,
+      "grad_norm": 1.4891722202301025,
+      "learning_rate": 1.8603017496434953e-05,
+      "loss": 0.2353,
+      "step": 1236
+    },
+    {
+      "epoch": 0.5857354246818586,
+      "grad_norm": 1.0998507738113403,
+      "learning_rate": 1.860040927844342e-05,
+      "loss": 0.2437,
+      "step": 1237
+    },
+    {
+      "epoch": 0.5862089375554898,
+      "grad_norm": 1.521974802017212,
+      "learning_rate": 1.859779881109781e-05,
+      "loss": 0.2569,
+      "step": 1238
+    },
+    {
+      "epoch": 0.586682450429121,
+      "grad_norm": 1.6860134601593018,
+      "learning_rate": 1.8595186095080864e-05,
+      "loss": 0.2483,
+      "step": 1239
+    },
+    {
+      "epoch": 0.5871559633027523,
+      "grad_norm": 1.1612238883972168,
+      "learning_rate": 1.8592571131075915e-05,
+      "loss": 0.2645,
+      "step": 1240
+    },
+    {
+      "epoch": 0.5876294761763835,
+      "grad_norm": 1.3629693984985352,
+      "learning_rate": 1.8589953919766882e-05,
+      "loss": 0.2492,
+      "step": 1241
+    },
+    {
+      "epoch": 0.5881029890500148,
+      "grad_norm": 1.175735592842102,
+      "learning_rate": 1.8587334461838267e-05,
+      "loss": 0.2256,
+      "step": 1242
+    },
+    {
+      "epoch": 0.5885765019236461,
+      "grad_norm": 2.3192355632781982,
+      "learning_rate": 1.8584712757975173e-05,
+      "loss": 0.2593,
+      "step": 1243
+    },
+    {
+      "epoch": 0.5890500147972773,
+      "grad_norm": 1.1697986125946045,
+      "learning_rate": 1.858208880886327e-05,
+      "loss": 0.2484,
+      "step": 1244
+    },
+    {
+      "epoch": 0.5895235276709085,
+      "grad_norm": 1.8343950510025024,
+      "learning_rate": 1.8579462615188832e-05,
+      "loss": 0.2652,
+      "step": 1245
+    },
+    {
+      "epoch": 0.5899970405445398,
+      "grad_norm": 1.3866143226623535,
+      "learning_rate": 1.8576834177638717e-05,
+      "loss": 0.2543,
+      "step": 1246
+    },
+    {
+      "epoch": 0.590470553418171,
+      "grad_norm": 1.11716628074646,
+      "learning_rate": 1.8574203496900366e-05,
+      "loss": 0.2338,
+      "step": 1247
+    },
+    {
+      "epoch": 0.5909440662918023,
+      "grad_norm": 1.593776822090149,
+      "learning_rate": 1.8571570573661803e-05,
+      "loss": 0.2547,
+      "step": 1248
+    },
+    {
+      "epoch": 0.5914175791654336,
+      "grad_norm": 2.5242745876312256,
+      "learning_rate": 1.8568935408611652e-05,
+      "loss": 0.2697,
+      "step": 1249
+    },
+    {
+      "epoch": 0.5918910920390648,
+      "grad_norm": 1.9075709581375122,
+      "learning_rate": 1.856629800243911e-05,
+      "loss": 0.2924,
+      "step": 1250
+    },
+    {
+      "epoch": 0.592364604912696,
+      "grad_norm": 1.8078994750976562,
+      "learning_rate": 1.8563658355833965e-05,
+      "loss": 0.2521,
+      "step": 1251
+    },
+    {
+      "epoch": 0.5928381177863273,
+      "grad_norm": 1.6747920513153076,
+      "learning_rate": 1.856101646948659e-05,
+      "loss": 0.2511,
+      "step": 1252
+    },
+    {
+      "epoch": 0.5933116306599586,
+      "grad_norm": 1.425503134727478,
+      "learning_rate": 1.8558372344087953e-05,
+      "loss": 0.2487,
+      "step": 1253
+    },
+    {
+      "epoch": 0.5937851435335898,
+      "grad_norm": 2.0698137283325195,
+      "learning_rate": 1.8555725980329588e-05,
+      "loss": 0.2581,
+      "step": 1254
+    },
+    {
+      "epoch": 0.5942586564072211,
+      "grad_norm": 2.361027479171753,
+      "learning_rate": 1.8553077378903632e-05,
+      "loss": 0.2371,
+      "step": 1255
+    },
+    {
+      "epoch": 0.5947321692808524,
+      "grad_norm": 2.6887784004211426,
+      "learning_rate": 1.8550426540502802e-05,
+      "loss": 0.2425,
+      "step": 1256
+    },
+    {
+      "epoch": 0.5952056821544836,
+      "grad_norm": 1.9129482507705688,
+      "learning_rate": 1.8547773465820397e-05,
+      "loss": 0.2334,
+      "step": 1257
+    },
+    {
+      "epoch": 0.5956791950281148,
+      "grad_norm": 1.7753819227218628,
+      "learning_rate": 1.8545118155550305e-05,
+      "loss": 0.2376,
+      "step": 1258
+    },
+    {
+      "epoch": 0.5961527079017461,
+      "grad_norm": 1.9076635837554932,
+      "learning_rate": 1.8542460610386993e-05,
+      "loss": 0.2295,
+      "step": 1259
+    },
+    {
+      "epoch": 0.5966262207753773,
+      "grad_norm": 2.463299512863159,
+      "learning_rate": 1.853980083102552e-05,
+      "loss": 0.2538,
+      "step": 1260
+    },
+    {
+      "epoch": 0.5970997336490086,
+      "grad_norm": 1.5017327070236206,
+      "learning_rate": 1.8537138818161527e-05,
+      "loss": 0.2312,
+      "step": 1261
+    },
+    {
+      "epoch": 0.5975732465226399,
+      "grad_norm": 1.36251962184906,
+      "learning_rate": 1.8534474572491235e-05,
+      "loss": 0.2536,
+      "step": 1262
+    },
+    {
+      "epoch": 0.598046759396271,
+      "grad_norm": 1.2385445833206177,
+      "learning_rate": 1.853180809471145e-05,
+      "loss": 0.2439,
+      "step": 1263
+    },
+    {
+      "epoch": 0.5985202722699023,
+      "grad_norm": 1.0537524223327637,
+      "learning_rate": 1.8529139385519567e-05,
+      "loss": 0.2483,
+      "step": 1264
+    },
+    {
+      "epoch": 0.5989937851435336,
+      "grad_norm": 3.3753559589385986,
+      "learning_rate": 1.8526468445613556e-05,
+      "loss": 0.2376,
+      "step": 1265
+    },
+    {
+      "epoch": 0.5994672980171648,
+      "grad_norm": 2.3799593448638916,
+      "learning_rate": 1.8523795275691986e-05,
+      "loss": 0.2361,
+      "step": 1266
+    },
+    {
+      "epoch": 0.5999408108907961,
+      "grad_norm": 1.300310492515564,
+      "learning_rate": 1.852111987645399e-05,
+      "loss": 0.2466,
+      "step": 1267
+    },
+    {
+      "epoch": 0.6004143237644274,
+      "grad_norm": 1.4112299680709839,
+      "learning_rate": 1.851844224859929e-05,
+      "loss": 0.2418,
+      "step": 1268
+    },
+    {
+      "epoch": 0.6008878366380586,
+      "grad_norm": 1.9448109865188599,
+      "learning_rate": 1.8515762392828205e-05,
+      "loss": 0.2492,
+      "step": 1269
+    },
+    {
+      "epoch": 0.6013613495116898,
+      "grad_norm": 1.0596145391464233,
+      "learning_rate": 1.8513080309841616e-05,
+      "loss": 0.2385,
+      "step": 1270
+    },
+    {
+      "epoch": 0.6018348623853211,
+      "grad_norm": 1.4580031633377075,
+      "learning_rate": 1.8510396000341e-05,
+      "loss": 0.2615,
+      "step": 1271
+    },
+    {
+      "epoch": 0.6023083752589523,
+      "grad_norm": 1.4013538360595703,
+      "learning_rate": 1.850770946502841e-05,
+      "loss": 0.2271,
+      "step": 1272
+    },
+    {
+      "epoch": 0.6027818881325836,
+      "grad_norm": 1.3812453746795654,
+      "learning_rate": 1.8505020704606486e-05,
+      "loss": 0.2506,
+      "step": 1273
+    },
+    {
+      "epoch": 0.6032554010062149,
+      "grad_norm": 1.411999225616455,
+      "learning_rate": 1.8502329719778448e-05,
+      "loss": 0.2343,
+      "step": 1274
+    },
+    {
+      "epoch": 0.6037289138798461,
+      "grad_norm": 1.6212693452835083,
+      "learning_rate": 1.849963651124809e-05,
+      "loss": 0.2382,
+      "step": 1275
+    },
+    {
+      "epoch": 0.6042024267534774,
+      "grad_norm": 1.3826075792312622,
+      "learning_rate": 1.8496941079719805e-05,
+      "loss": 0.2541,
+      "step": 1276
+    },
+    {
+      "epoch": 0.6046759396271086,
+      "grad_norm": 2.0459675788879395,
+      "learning_rate": 1.8494243425898548e-05,
+      "loss": 0.2367,
+      "step": 1277
+    },
+    {
+      "epoch": 0.6051494525007398,
+      "grad_norm": 2.222378969192505,
+      "learning_rate": 1.8491543550489872e-05,
+      "loss": 0.2267,
+      "step": 1278
+    },
+    {
+      "epoch": 0.6056229653743711,
+      "grad_norm": 1.439396619796753,
+      "learning_rate": 1.84888414541999e-05,
+      "loss": 0.2454,
+      "step": 1279
+    },
+    {
+      "epoch": 0.6060964782480024,
+      "grad_norm": 1.3656351566314697,
+      "learning_rate": 1.8486137137735337e-05,
+      "loss": 0.2417,
+      "step": 1280
+    },
+    {
+      "epoch": 0.6065699911216336,
+      "grad_norm": 1.2693840265274048,
+      "learning_rate": 1.848343060180347e-05,
+      "loss": 0.2573,
+      "step": 1281
+    },
+    {
+      "epoch": 0.6070435039952649,
+      "grad_norm": 1.9069156646728516,
+      "learning_rate": 1.8480721847112174e-05,
+      "loss": 0.2466,
+      "step": 1282
+    },
+    {
+      "epoch": 0.6075170168688961,
+      "grad_norm": 1.1109650135040283,
+      "learning_rate": 1.8478010874369894e-05,
+      "loss": 0.2637,
+      "step": 1283
+    },
+    {
+      "epoch": 0.6079905297425274,
+      "grad_norm": 1.4270925521850586,
+      "learning_rate": 1.8475297684285657e-05,
+      "loss": 0.1981,
+      "step": 1284
+    },
+    {
+      "epoch": 0.6084640426161586,
+      "grad_norm": 1.5452510118484497,
+      "learning_rate": 1.8472582277569072e-05,
+      "loss": 0.2564,
+      "step": 1285
+    },
+    {
+      "epoch": 0.6089375554897899,
+      "grad_norm": 1.9207974672317505,
+      "learning_rate": 1.8469864654930333e-05,
+      "loss": 0.2676,
+      "step": 1286
+    },
+    {
+      "epoch": 0.6094110683634212,
+      "grad_norm": 1.18226957321167,
+      "learning_rate": 1.8467144817080204e-05,
+      "loss": 0.2474,
+      "step": 1287
+    },
+    {
+      "epoch": 0.6098845812370524,
+      "grad_norm": 1.3588470220565796,
+      "learning_rate": 1.8464422764730028e-05,
+      "loss": 0.2469,
+      "step": 1288
+    },
+    {
+      "epoch": 0.6103580941106836,
+      "grad_norm": 1.4473971128463745,
+      "learning_rate": 1.8461698498591736e-05,
+      "loss": 0.2322,
+      "step": 1289
+    },
+    {
+      "epoch": 0.6108316069843149,
+      "grad_norm": 2.050037145614624,
+      "learning_rate": 1.8458972019377834e-05,
+      "loss": 0.2611,
+      "step": 1290
+    },
+    {
+      "epoch": 0.6113051198579461,
+      "grad_norm": 1.4301786422729492,
+      "learning_rate": 1.8456243327801407e-05,
+      "loss": 0.2265,
+      "step": 1291
+    },
+    {
+      "epoch": 0.6117786327315774,
+      "grad_norm": 1.096746802330017,
+      "learning_rate": 1.8453512424576113e-05,
+      "loss": 0.2279,
+      "step": 1292
+    },
+    {
+      "epoch": 0.6122521456052087,
+      "grad_norm": 1.5921342372894287,
+      "learning_rate": 1.84507793104162e-05,
+      "loss": 0.2564,
+      "step": 1293
+    },
+    {
+      "epoch": 0.6127256584788399,
+      "grad_norm": 1.1023545265197754,
+      "learning_rate": 1.8448043986036483e-05,
+      "loss": 0.2359,
+      "step": 1294
+    },
+    {
+      "epoch": 0.6131991713524712,
+      "grad_norm": 2.039523124694824,
+      "learning_rate": 1.844530645215236e-05,
+      "loss": 0.2283,
+      "step": 1295
+    },
+    {
+      "epoch": 0.6136726842261024,
+      "grad_norm": 1.5656194686889648,
+      "learning_rate": 1.8442566709479813e-05,
+      "loss": 0.256,
+      "step": 1296
+    },
+    {
+      "epoch": 0.6141461970997336,
+      "grad_norm": 1.9712538719177246,
+      "learning_rate": 1.8439824758735385e-05,
+      "loss": 0.2526,
+      "step": 1297
+    },
+    {
+      "epoch": 0.6146197099733649,
+      "grad_norm": 1.1353607177734375,
+      "learning_rate": 1.8437080600636215e-05,
+      "loss": 0.2606,
+      "step": 1298
+    },
+    {
+      "epoch": 0.6150932228469962,
+      "grad_norm": 2.3655588626861572,
+      "learning_rate": 1.8434334235900008e-05,
+      "loss": 0.2293,
+      "step": 1299
+    },
+    {
+      "epoch": 0.6155667357206274,
+      "grad_norm": 1.5827844142913818,
+      "learning_rate": 1.8431585665245047e-05,
+      "loss": 0.2443,
+      "step": 1300
+    },
+    {
+      "epoch": 0.6160402485942587,
+      "grad_norm": 1.5247994661331177,
+      "learning_rate": 1.8428834889390194e-05,
+      "loss": 0.2444,
+      "step": 1301
+    },
+    {
+      "epoch": 0.6165137614678899,
+      "grad_norm": 1.465590476989746,
+      "learning_rate": 1.8426081909054893e-05,
+      "loss": 0.2279,
+      "step": 1302
+    },
+    {
+      "epoch": 0.6169872743415211,
+      "grad_norm": 1.2217838764190674,
+      "learning_rate": 1.8423326724959157e-05,
+      "loss": 0.2393,
+      "step": 1303
+    },
+    {
+      "epoch": 0.6174607872151524,
+      "grad_norm": 1.6876059770584106,
+      "learning_rate": 1.8420569337823576e-05,
+      "loss": 0.248,
+      "step": 1304
+    },
+    {
+      "epoch": 0.6179343000887837,
+      "grad_norm": 1.808072805404663,
+      "learning_rate": 1.841780974836932e-05,
+      "loss": 0.2678,
+      "step": 1305
+    },
+    {
+      "epoch": 0.6184078129624149,
+      "grad_norm": 2.256350517272949,
+      "learning_rate": 1.8415047957318132e-05,
+      "loss": 0.2462,
+      "step": 1306
+    },
+    {
+      "epoch": 0.6188813258360462,
+      "grad_norm": 2.597975969314575,
+      "learning_rate": 1.8412283965392334e-05,
+      "loss": 0.2504,
+      "step": 1307
+    },
+    {
+      "epoch": 0.6193548387096774,
+      "grad_norm": 1.6564836502075195,
+      "learning_rate": 1.8409517773314816e-05,
+      "loss": 0.2264,
+      "step": 1308
+    },
+    {
+      "epoch": 0.6198283515833086,
+      "grad_norm": 1.5437556505203247,
+      "learning_rate": 1.8406749381809054e-05,
+      "loss": 0.2375,
+      "step": 1309
+    },
+    {
+      "epoch": 0.6203018644569399,
+      "grad_norm": 2.6221256256103516,
+      "learning_rate": 1.8403978791599095e-05,
+      "loss": 0.2454,
+      "step": 1310
+    },
+    {
+      "epoch": 0.6207753773305712,
+      "grad_norm": 1.4474231004714966,
+      "learning_rate": 1.8401206003409554e-05,
+      "loss": 0.2489,
+      "step": 1311
+    },
+    {
+      "epoch": 0.6212488902042024,
+      "grad_norm": 2.578899621963501,
+      "learning_rate": 1.839843101796563e-05,
+      "loss": 0.2466,
+      "step": 1312
+    },
+    {
+      "epoch": 0.6217224030778337,
+      "grad_norm": 2.1055898666381836,
+      "learning_rate": 1.8395653835993098e-05,
+      "loss": 0.2741,
+      "step": 1313
+    },
+    {
+      "epoch": 0.622195915951465,
+      "grad_norm": 1.2362160682678223,
+      "learning_rate": 1.83928744582183e-05,
+      "loss": 0.2472,
+      "step": 1314
+    },
+    {
+      "epoch": 0.6226694288250961,
+      "grad_norm": 1.5515974760055542,
+      "learning_rate": 1.839009288536815e-05,
+      "loss": 0.2346,
+      "step": 1315
+    },
+    {
+      "epoch": 0.6231429416987274,
+      "grad_norm": 1.4846057891845703,
+      "learning_rate": 1.838730911817015e-05,
+      "loss": 0.2439,
+      "step": 1316
+    },
+    {
+      "epoch": 0.6236164545723587,
+      "grad_norm": 2.014627695083618,
+      "learning_rate": 1.8384523157352365e-05,
+      "loss": 0.2398,
+      "step": 1317
+    },
+    {
+      "epoch": 0.62408996744599,
+      "grad_norm": 2.682614326477051,
+      "learning_rate": 1.8381735003643434e-05,
+      "loss": 0.2567,
+      "step": 1318
+    },
+    {
+      "epoch": 0.6245634803196212,
+      "grad_norm": 1.1889935731887817,
+      "learning_rate": 1.837894465777257e-05,
+      "loss": 0.2425,
+      "step": 1319
+    },
+    {
+      "epoch": 0.6250369931932525,
+      "grad_norm": 1.283697247505188,
+      "learning_rate": 1.8376152120469567e-05,
+      "loss": 0.2576,
+      "step": 1320
+    },
+    {
+      "epoch": 0.6255105060668837,
+      "grad_norm": 1.7607721090316772,
+      "learning_rate": 1.8373357392464783e-05,
+      "loss": 0.2428,
+      "step": 1321
+    },
+    {
+      "epoch": 0.6259840189405149,
+      "grad_norm": 2.0898520946502686,
+      "learning_rate": 1.837056047448915e-05,
+      "loss": 0.246,
+      "step": 1322
+    },
+    {
+      "epoch": 0.6264575318141462,
+      "grad_norm": 1.7172597646713257,
+      "learning_rate": 1.836776136727418e-05,
+      "loss": 0.242,
+      "step": 1323
+    },
+    {
+      "epoch": 0.6269310446877775,
+      "grad_norm": 1.7374228239059448,
+      "learning_rate": 1.8364960071551948e-05,
+      "loss": 0.252,
+      "step": 1324
+    },
+    {
+      "epoch": 0.6274045575614087,
+      "grad_norm": 1.4779413938522339,
+      "learning_rate": 1.836215658805511e-05,
+      "loss": 0.2424,
+      "step": 1325
+    },
+    {
+      "epoch": 0.62787807043504,
+      "grad_norm": 1.4057080745697021,
+      "learning_rate": 1.8359350917516882e-05,
+      "loss": 0.2516,
+      "step": 1326
+    },
+    {
+      "epoch": 0.6283515833086712,
+      "grad_norm": 1.7283161878585815,
+      "learning_rate": 1.835654306067107e-05,
+      "loss": 0.2559,
+      "step": 1327
+    },
+    {
+      "epoch": 0.6288250961823024,
+      "grad_norm": 1.3666443824768066,
+      "learning_rate": 1.835373301825204e-05,
+      "loss": 0.235,
+      "step": 1328
+    },
+    {
+      "epoch": 0.6292986090559337,
+      "grad_norm": 1.4366817474365234,
+      "learning_rate": 1.8350920790994723e-05,
+      "loss": 0.2405,
+      "step": 1329
+    },
+    {
+      "epoch": 0.629772121929565,
+      "grad_norm": 1.3197062015533447,
+      "learning_rate": 1.834810637963464e-05,
+      "loss": 0.2191,
+      "step": 1330
+    },
+    {
+      "epoch": 0.6302456348031962,
+      "grad_norm": 1.0538299083709717,
+      "learning_rate": 1.8345289784907863e-05,
+      "loss": 0.2255,
+      "step": 1331
+    },
+    {
+      "epoch": 0.6307191476768275,
+      "grad_norm": 1.732539415359497,
+      "learning_rate": 1.8342471007551058e-05,
+      "loss": 0.2733,
+      "step": 1332
+    },
+    {
+      "epoch": 0.6311926605504588,
+      "grad_norm": 1.7113350629806519,
+      "learning_rate": 1.833965004830144e-05,
+      "loss": 0.2285,
+      "step": 1333
+    },
+    {
+      "epoch": 0.6316661734240899,
+      "grad_norm": 2.064622640609741,
+      "learning_rate": 1.8336826907896806e-05,
+      "loss": 0.2398,
+      "step": 1334
+    },
+    {
+      "epoch": 0.6321396862977212,
+      "grad_norm": 1.9892138242721558,
+      "learning_rate": 1.8334001587075517e-05,
+      "loss": 0.2503,
+      "step": 1335
+    },
+    {
+      "epoch": 0.6326131991713525,
+      "grad_norm": 1.079971194267273,
+      "learning_rate": 1.833117408657652e-05,
+      "loss": 0.2354,
+      "step": 1336
+    },
+    {
+      "epoch": 0.6330867120449837,
+      "grad_norm": 1.276336669921875,
+      "learning_rate": 1.8328344407139307e-05,
+      "loss": 0.2452,
+      "step": 1337
+    },
+    {
+      "epoch": 0.633560224918615,
+      "grad_norm": 1.477914571762085,
+      "learning_rate": 1.832551254950396e-05,
+      "loss": 0.2422,
+      "step": 1338
+    },
+    {
+      "epoch": 0.6340337377922463,
+      "grad_norm": 1.510981798171997,
+      "learning_rate": 1.832267851441113e-05,
+      "loss": 0.2523,
+      "step": 1339
+    },
+    {
+      "epoch": 0.6345072506658774,
+      "grad_norm": 1.4486583471298218,
+      "learning_rate": 1.8319842302602014e-05,
+      "loss": 0.2492,
+      "step": 1340
+    },
+    {
+      "epoch": 0.6349807635395087,
+      "grad_norm": 1.9352304935455322,
+      "learning_rate": 1.8317003914818415e-05,
+      "loss": 0.2373,
+      "step": 1341
+    },
+    {
+      "epoch": 0.63545427641314,
+      "grad_norm": 1.3235574960708618,
+      "learning_rate": 1.8314163351802673e-05,
+      "loss": 0.2464,
+      "step": 1342
+    },
+    {
+      "epoch": 0.6359277892867712,
+      "grad_norm": 1.9728165864944458,
+      "learning_rate": 1.831132061429772e-05,
+      "loss": 0.2387,
+      "step": 1343
+    },
+    {
+      "epoch": 0.6364013021604025,
+      "grad_norm": 1.5262460708618164,
+      "learning_rate": 1.8308475703047034e-05,
+      "loss": 0.2273,
+      "step": 1344
+    },
+    {
+      "epoch": 0.6368748150340338,
+      "grad_norm": 1.6200482845306396,
+      "learning_rate": 1.8305628618794685e-05,
+      "loss": 0.2561,
+      "step": 1345
+    },
+    {
+      "epoch": 0.6373483279076649,
+      "grad_norm": 1.5177505016326904,
+      "learning_rate": 1.8302779362285298e-05,
+      "loss": 0.2331,
+      "step": 1346
+    },
+    {
+      "epoch": 0.6378218407812962,
+      "grad_norm": 1.3834291696548462,
+      "learning_rate": 1.8299927934264064e-05,
+      "loss": 0.2117,
+      "step": 1347
+    },
+    {
+      "epoch": 0.6382953536549275,
+      "grad_norm": 1.5583209991455078,
+      "learning_rate": 1.829707433547675e-05,
+      "loss": 0.2055,
+      "step": 1348
+    },
+    {
+      "epoch": 0.6387688665285588,
+      "grad_norm": 1.7908825874328613,
+      "learning_rate": 1.8294218566669684e-05,
+      "loss": 0.264,
+      "step": 1349
+    },
+    {
+      "epoch": 0.63924237940219,
+      "grad_norm": 2.1676015853881836,
+      "learning_rate": 1.8291360628589774e-05,
+      "loss": 0.2424,
+      "step": 1350
+    },
+    {
+      "epoch": 0.6397158922758213,
+      "grad_norm": 1.7668577432632446,
+      "learning_rate": 1.8288500521984477e-05,
+      "loss": 0.2576,
+      "step": 1351
+    },
+    {
+      "epoch": 0.6401894051494526,
+      "grad_norm": 1.537380337715149,
+      "learning_rate": 1.828563824760183e-05,
+      "loss": 0.243,
+      "step": 1352
+    },
+    {
+      "epoch": 0.6406629180230837,
+      "grad_norm": 1.5818302631378174,
+      "learning_rate": 1.828277380619043e-05,
+      "loss": 0.2637,
+      "step": 1353
+    },
+    {
+      "epoch": 0.641136430896715,
+      "grad_norm": 1.2502553462982178,
+      "learning_rate": 1.827990719849945e-05,
+      "loss": 0.2299,
+      "step": 1354
+    },
+    {
+      "epoch": 0.6416099437703463,
+      "grad_norm": 1.2032712697982788,
+      "learning_rate": 1.8277038425278616e-05,
+      "loss": 0.2597,
+      "step": 1355
+    },
+    {
+      "epoch": 0.6420834566439775,
+      "grad_norm": 1.1171890497207642,
+      "learning_rate": 1.8274167487278232e-05,
+      "loss": 0.24,
+      "step": 1356
+    },
+    {
+      "epoch": 0.6425569695176088,
+      "grad_norm": 1.7673466205596924,
+      "learning_rate": 1.8271294385249168e-05,
+      "loss": 0.2274,
+      "step": 1357
+    },
+    {
+      "epoch": 0.64303048239124,
+      "grad_norm": 1.4950395822525024,
+      "learning_rate": 1.8268419119942852e-05,
+      "loss": 0.2521,
+      "step": 1358
+    },
+    {
+      "epoch": 0.6435039952648712,
+      "grad_norm": 1.344424843788147,
+      "learning_rate": 1.8265541692111277e-05,
+      "loss": 0.2201,
+      "step": 1359
+    },
+    {
+      "epoch": 0.6439775081385025,
+      "grad_norm": 1.1256235837936401,
+      "learning_rate": 1.8262662102507012e-05,
+      "loss": 0.2402,
+      "step": 1360
+    },
+    {
+      "epoch": 0.6444510210121338,
+      "grad_norm": 1.2942851781845093,
+      "learning_rate": 1.8259780351883188e-05,
+      "loss": 0.2561,
+      "step": 1361
+    },
+    {
+      "epoch": 0.644924533885765,
+      "grad_norm": 1.1802473068237305,
+      "learning_rate": 1.8256896440993498e-05,
+      "loss": 0.2516,
+      "step": 1362
+    },
+    {
+      "epoch": 0.6453980467593963,
+      "grad_norm": 1.1492762565612793,
+      "learning_rate": 1.8254010370592197e-05,
+      "loss": 0.2492,
+      "step": 1363
+    },
+    {
+      "epoch": 0.6458715596330276,
+      "grad_norm": 1.716592788696289,
+      "learning_rate": 1.825112214143411e-05,
+      "loss": 0.2208,
+      "step": 1364
+    },
+    {
+      "epoch": 0.6463450725066587,
+      "grad_norm": 1.5306360721588135,
+      "learning_rate": 1.824823175427463e-05,
+      "loss": 0.2273,
+      "step": 1365
+    },
+    {
+      "epoch": 0.64681858538029,
+      "grad_norm": 1.6594233512878418,
+      "learning_rate": 1.8245339209869705e-05,
+      "loss": 0.2514,
+      "step": 1366
+    },
+    {
+      "epoch": 0.6472920982539213,
+      "grad_norm": 1.2237942218780518,
+      "learning_rate": 1.8242444508975857e-05,
+      "loss": 0.2514,
+      "step": 1367
+    },
+    {
+      "epoch": 0.6477656111275525,
+      "grad_norm": 1.5434156656265259,
+      "learning_rate": 1.8239547652350162e-05,
+      "loss": 0.2523,
+      "step": 1368
+    },
+    {
+      "epoch": 0.6482391240011838,
+      "grad_norm": 1.3254024982452393,
+      "learning_rate": 1.8236648640750266e-05,
+      "loss": 0.2588,
+      "step": 1369
+    },
+    {
+      "epoch": 0.6487126368748151,
+      "grad_norm": 1.8970592021942139,
+      "learning_rate": 1.823374747493438e-05,
+      "loss": 0.2487,
+      "step": 1370
+    },
+    {
+      "epoch": 0.6491861497484462,
+      "grad_norm": 1.7532835006713867,
+      "learning_rate": 1.8230844155661273e-05,
+      "loss": 0.2405,
+      "step": 1371
+    },
+    {
+      "epoch": 0.6496596626220775,
+      "grad_norm": 1.1474004983901978,
+      "learning_rate": 1.822793868369028e-05,
+      "loss": 0.2221,
+      "step": 1372
+    },
+    {
+      "epoch": 0.6501331754957088,
+      "grad_norm": 1.8724579811096191,
+      "learning_rate": 1.8225031059781302e-05,
+      "loss": 0.2345,
+      "step": 1373
+    },
+    {
+      "epoch": 0.65060668836934,
+      "grad_norm": 1.337428331375122,
+      "learning_rate": 1.8222121284694798e-05,
+      "loss": 0.2737,
+      "step": 1374
+    },
+    {
+      "epoch": 0.6510802012429713,
+      "grad_norm": 1.1940720081329346,
+      "learning_rate": 1.8219209359191793e-05,
+      "loss": 0.2089,
+      "step": 1375
+    },
+    {
+      "epoch": 0.6515537141166026,
+      "grad_norm": 1.327831745147705,
+      "learning_rate": 1.821629528403387e-05,
+      "loss": 0.2629,
+      "step": 1376
+    },
+    {
+      "epoch": 0.6520272269902337,
+      "grad_norm": 1.3946506977081299,
+      "learning_rate": 1.8213379059983184e-05,
+      "loss": 0.227,
+      "step": 1377
+    },
+    {
+      "epoch": 0.652500739863865,
+      "grad_norm": 1.840470790863037,
+      "learning_rate": 1.8210460687802437e-05,
+      "loss": 0.2434,
+      "step": 1378
+    },
+    {
+      "epoch": 0.6529742527374963,
+      "grad_norm": 1.3250887393951416,
+      "learning_rate": 1.8207540168254907e-05,
+      "loss": 0.2449,
+      "step": 1379
+    },
+    {
+      "epoch": 0.6534477656111276,
+      "grad_norm": 1.3822866678237915,
+      "learning_rate": 1.8204617502104426e-05,
+      "loss": 0.2286,
+      "step": 1380
+    },
+    {
+      "epoch": 0.6539212784847588,
+      "grad_norm": 1.2406575679779053,
+      "learning_rate": 1.8201692690115385e-05,
+      "loss": 0.2551,
+      "step": 1381
+    },
+    {
+      "epoch": 0.6543947913583901,
+      "grad_norm": 1.167535662651062,
+      "learning_rate": 1.819876573305275e-05,
+      "loss": 0.2365,
+      "step": 1382
+    },
+    {
+      "epoch": 0.6548683042320214,
+      "grad_norm": 1.555350661277771,
+      "learning_rate": 1.8195836631682025e-05,
+      "loss": 0.2523,
+      "step": 1383
+    },
+    {
+      "epoch": 0.6553418171056525,
+      "grad_norm": 1.228309988975525,
+      "learning_rate": 1.8192905386769298e-05,
+      "loss": 0.2235,
+      "step": 1384
+    },
+    {
+      "epoch": 0.6558153299792838,
+      "grad_norm": 1.2090561389923096,
+      "learning_rate": 1.8189971999081207e-05,
+      "loss": 0.24,
+      "step": 1385
+    },
+    {
+      "epoch": 0.6562888428529151,
+      "grad_norm": 1.1428760290145874,
+      "learning_rate": 1.818703646938495e-05,
+      "loss": 0.2441,
+      "step": 1386
+    },
+    {
+      "epoch": 0.6567623557265463,
+      "grad_norm": 2.0970654487609863,
+      "learning_rate": 1.8184098798448285e-05,
+      "loss": 0.2685,
+      "step": 1387
+    },
+    {
+      "epoch": 0.6572358686001776,
+      "grad_norm": 1.638376235961914,
+      "learning_rate": 1.8181158987039534e-05,
+      "loss": 0.2524,
+      "step": 1388
+    },
+    {
+      "epoch": 0.6577093814738089,
+      "grad_norm": 1.7937678098678589,
+      "learning_rate": 1.8178217035927578e-05,
+      "loss": 0.2446,
+      "step": 1389
+    },
+    {
+      "epoch": 0.65818289434744,
+      "grad_norm": 1.5529017448425293,
+      "learning_rate": 1.8175272945881847e-05,
+      "loss": 0.262,
+      "step": 1390
+    },
+    {
+      "epoch": 0.6586564072210713,
+      "grad_norm": 2.30466628074646,
+      "learning_rate": 1.8172326717672348e-05,
+      "loss": 0.239,
+      "step": 1391
+    },
+    {
+      "epoch": 0.6591299200947026,
+      "grad_norm": 2.677629232406616,
+      "learning_rate": 1.8169378352069637e-05,
+      "loss": 0.2682,
+      "step": 1392
+    },
+    {
+      "epoch": 0.6596034329683338,
+      "grad_norm": 2.3182013034820557,
+      "learning_rate": 1.8166427849844826e-05,
+      "loss": 0.2425,
+      "step": 1393
+    },
+    {
+      "epoch": 0.6600769458419651,
+      "grad_norm": 1.5342023372650146,
+      "learning_rate": 1.8163475211769595e-05,
+      "loss": 0.2674,
+      "step": 1394
+    },
+    {
+      "epoch": 0.6605504587155964,
+      "grad_norm": 2.3066344261169434,
+      "learning_rate": 1.8160520438616176e-05,
+      "loss": 0.2506,
+      "step": 1395
+    },
+    {
+      "epoch": 0.6610239715892275,
+      "grad_norm": 1.210817813873291,
+      "learning_rate": 1.8157563531157366e-05,
+      "loss": 0.2399,
+      "step": 1396
+    },
+    {
+      "epoch": 0.6614974844628588,
+      "grad_norm": 1.5053683519363403,
+      "learning_rate": 1.815460449016651e-05,
+      "loss": 0.2343,
+      "step": 1397
+    },
+    {
+      "epoch": 0.6619709973364901,
+      "grad_norm": 1.1325657367706299,
+      "learning_rate": 1.8151643316417518e-05,
+      "loss": 0.2338,
+      "step": 1398
+    },
+    {
+      "epoch": 0.6624445102101213,
+      "grad_norm": 2.2156691551208496,
+      "learning_rate": 1.8148680010684856e-05,
+      "loss": 0.2328,
+      "step": 1399
+    },
+    {
+      "epoch": 0.6629180230837526,
+      "grad_norm": 1.634922742843628,
+      "learning_rate": 1.814571457374355e-05,
+      "loss": 0.2786,
+      "step": 1400
+    },
+    {
+      "epoch": 0.6633915359573839,
+      "grad_norm": 1.8771041631698608,
+      "learning_rate": 1.8142747006369176e-05,
+      "loss": 0.2626,
+      "step": 1401
+    },
+    {
+      "epoch": 0.663865048831015,
+      "grad_norm": 1.1594529151916504,
+      "learning_rate": 1.813977730933788e-05,
+      "loss": 0.2044,
+      "step": 1402
+    },
+    {
+      "epoch": 0.6643385617046463,
+      "grad_norm": 1.380021095275879,
+      "learning_rate": 1.8136805483426358e-05,
+      "loss": 0.2689,
+      "step": 1403
+    },
+    {
+      "epoch": 0.6648120745782776,
+      "grad_norm": 1.0207122564315796,
+      "learning_rate": 1.8133831529411856e-05,
+      "loss": 0.249,
+      "step": 1404
+    },
+    {
+      "epoch": 0.6652855874519088,
+      "grad_norm": 1.723662257194519,
+      "learning_rate": 1.8130855448072186e-05,
+      "loss": 0.2373,
+      "step": 1405
+    },
+    {
+      "epoch": 0.6657591003255401,
+      "grad_norm": 1.606900930404663,
+      "learning_rate": 1.8127877240185716e-05,
+      "loss": 0.2731,
+      "step": 1406
+    },
+    {
+      "epoch": 0.6662326131991714,
+      "grad_norm": 1.1863452196121216,
+      "learning_rate": 1.8124896906531363e-05,
+      "loss": 0.2379,
+      "step": 1407
+    },
+    {
+      "epoch": 0.6667061260728026,
+      "grad_norm": 1.311086893081665,
+      "learning_rate": 1.8121914447888605e-05,
+      "loss": 0.2295,
+      "step": 1408
+    },
+    {
+      "epoch": 0.6671796389464338,
+      "grad_norm": 1.2078311443328857,
+      "learning_rate": 1.811892986503748e-05,
+      "loss": 0.2209,
+      "step": 1409
+    },
+    {
+      "epoch": 0.6676531518200651,
+      "grad_norm": 2.1490330696105957,
+      "learning_rate": 1.811594315875857e-05,
+      "loss": 0.2922,
+      "step": 1410
+    },
+    {
+      "epoch": 0.6681266646936964,
+      "grad_norm": 1.3307857513427734,
+      "learning_rate": 1.8112954329833022e-05,
+      "loss": 0.229,
+      "step": 1411
+    },
+    {
+      "epoch": 0.6686001775673276,
+      "grad_norm": 1.1393177509307861,
+      "learning_rate": 1.8109963379042538e-05,
+      "loss": 0.243,
+      "step": 1412
+    },
+    {
+      "epoch": 0.6690736904409589,
+      "grad_norm": 1.3860878944396973,
+      "learning_rate": 1.810697030716937e-05,
+      "loss": 0.246,
+      "step": 1413
+    },
+    {
+      "epoch": 0.6695472033145902,
+      "grad_norm": 1.2071013450622559,
+      "learning_rate": 1.8103975114996327e-05,
+      "loss": 0.2124,
+      "step": 1414
+    },
+    {
+      "epoch": 0.6700207161882213,
+      "grad_norm": 1.5767385959625244,
+      "learning_rate": 1.8100977803306774e-05,
+      "loss": 0.2346,
+      "step": 1415
+    },
+    {
+      "epoch": 0.6704942290618526,
+      "grad_norm": 1.6426430940628052,
+      "learning_rate": 1.8097978372884627e-05,
+      "loss": 0.2258,
+      "step": 1416
+    },
+    {
+      "epoch": 0.6709677419354839,
+      "grad_norm": 1.0845011472702026,
+      "learning_rate": 1.8094976824514363e-05,
+      "loss": 0.2492,
+      "step": 1417
+    },
+    {
+      "epoch": 0.6714412548091151,
+      "grad_norm": 1.211498498916626,
+      "learning_rate": 1.8091973158981e-05,
+      "loss": 0.2457,
+      "step": 1418
+    },
+    {
+      "epoch": 0.6719147676827464,
+      "grad_norm": 1.3050991296768188,
+      "learning_rate": 1.8088967377070122e-05,
+      "loss": 0.1988,
+      "step": 1419
+    },
+    {
+      "epoch": 0.6723882805563777,
+      "grad_norm": 2.2744362354278564,
+      "learning_rate": 1.8085959479567866e-05,
+      "loss": 0.2364,
+      "step": 1420
+    },
+    {
+      "epoch": 0.6728617934300088,
+      "grad_norm": 1.9004119634628296,
+      "learning_rate": 1.808294946726091e-05,
+      "loss": 0.2387,
+      "step": 1421
+    },
+    {
+      "epoch": 0.6733353063036401,
+      "grad_norm": 1.1495134830474854,
+      "learning_rate": 1.80799373409365e-05,
+      "loss": 0.2364,
+      "step": 1422
+    },
+    {
+      "epoch": 0.6738088191772714,
+      "grad_norm": 1.161669373512268,
+      "learning_rate": 1.807692310138243e-05,
+      "loss": 0.2158,
+      "step": 1423
+    },
+    {
+      "epoch": 0.6742823320509026,
+      "grad_norm": 1.1612114906311035,
+      "learning_rate": 1.807390674938704e-05,
+      "loss": 0.2477,
+      "step": 1424
+    },
+    {
+      "epoch": 0.6747558449245339,
+      "grad_norm": 1.5632646083831787,
+      "learning_rate": 1.8070888285739227e-05,
+      "loss": 0.2696,
+      "step": 1425
+    },
+    {
+      "epoch": 0.6752293577981652,
+      "grad_norm": 1.4020699262619019,
+      "learning_rate": 1.806786771122845e-05,
+      "loss": 0.2244,
+      "step": 1426
+    },
+    {
+      "epoch": 0.6757028706717964,
+      "grad_norm": 1.7532941102981567,
+      "learning_rate": 1.80648450266447e-05,
+      "loss": 0.3041,
+      "step": 1427
+    },
+    {
+      "epoch": 0.6761763835454276,
+      "grad_norm": 2.035996675491333,
+      "learning_rate": 1.806182023277854e-05,
+      "loss": 0.2601,
+      "step": 1428
+    },
+    {
+      "epoch": 0.6766498964190589,
+      "grad_norm": 1.4259287118911743,
+      "learning_rate": 1.805879333042107e-05,
+      "loss": 0.2476,
+      "step": 1429
+    },
+    {
+      "epoch": 0.6771234092926901,
+      "grad_norm": 1.3768444061279297,
+      "learning_rate": 1.805576432036395e-05,
+      "loss": 0.2458,
+      "step": 1430
+    },
+    {
+      "epoch": 0.6775969221663214,
+      "grad_norm": 1.153491497039795,
+      "learning_rate": 1.8052733203399385e-05,
+      "loss": 0.2582,
+      "step": 1431
+    },
+    {
+      "epoch": 0.6780704350399527,
+      "grad_norm": 1.0750600099563599,
+      "learning_rate": 1.804969998032014e-05,
+      "loss": 0.2708,
+      "step": 1432
+    },
+    {
+      "epoch": 0.6785439479135839,
+      "grad_norm": 1.5494842529296875,
+      "learning_rate": 1.8046664651919517e-05,
+      "loss": 0.2314,
+      "step": 1433
+    },
+    {
+      "epoch": 0.6790174607872151,
+      "grad_norm": 1.6655199527740479,
+      "learning_rate": 1.8043627218991385e-05,
+      "loss": 0.257,
+      "step": 1434
+    },
+    {
+      "epoch": 0.6794909736608464,
+      "grad_norm": 1.0493881702423096,
+      "learning_rate": 1.8040587682330155e-05,
+      "loss": 0.2497,
+      "step": 1435
+    },
+    {
+      "epoch": 0.6799644865344776,
+      "grad_norm": 1.266782283782959,
+      "learning_rate": 1.8037546042730783e-05,
+      "loss": 0.2443,
+      "step": 1436
+    },
+    {
+      "epoch": 0.6804379994081089,
+      "grad_norm": 1.381783127784729,
+      "learning_rate": 1.8034502300988784e-05,
+      "loss": 0.2467,
+      "step": 1437
+    },
+    {
+      "epoch": 0.6809115122817402,
+      "grad_norm": 1.3426703214645386,
+      "learning_rate": 1.803145645790022e-05,
+      "loss": 0.2399,
+      "step": 1438
+    },
+    {
+      "epoch": 0.6813850251553714,
+      "grad_norm": 1.1513340473175049,
+      "learning_rate": 1.8028408514261695e-05,
+      "loss": 0.2482,
+      "step": 1439
+    },
+    {
+      "epoch": 0.6818585380290026,
+      "grad_norm": 1.5276223421096802,
+      "learning_rate": 1.8025358470870376e-05,
+      "loss": 0.2307,
+      "step": 1440
+    },
+    {
+      "epoch": 0.6823320509026339,
+      "grad_norm": 1.416548252105713,
+      "learning_rate": 1.8022306328523973e-05,
+      "loss": 0.2142,
+      "step": 1441
+    },
+    {
+      "epoch": 0.6828055637762651,
+      "grad_norm": 1.427147626876831,
+      "learning_rate": 1.801925208802074e-05,
+      "loss": 0.2139,
+      "step": 1442
+    },
+    {
+      "epoch": 0.6832790766498964,
+      "grad_norm": 1.3105965852737427,
+      "learning_rate": 1.8016195750159488e-05,
+      "loss": 0.2398,
+      "step": 1443
+    },
+    {
+      "epoch": 0.6837525895235277,
+      "grad_norm": 1.5109857320785522,
+      "learning_rate": 1.801313731573957e-05,
+      "loss": 0.2733,
+      "step": 1444
+    },
+    {
+      "epoch": 0.684226102397159,
+      "grad_norm": 1.150864601135254,
+      "learning_rate": 1.8010076785560896e-05,
+      "loss": 0.2492,
+      "step": 1445
+    },
+    {
+      "epoch": 0.6846996152707902,
+      "grad_norm": 1.219900131225586,
+      "learning_rate": 1.8007014160423907e-05,
+      "loss": 0.2619,
+      "step": 1446
+    },
+    {
+      "epoch": 0.6851731281444214,
+      "grad_norm": 1.590561032295227,
+      "learning_rate": 1.8003949441129612e-05,
+      "loss": 0.2396,
+      "step": 1447
+    },
+    {
+      "epoch": 0.6856466410180527,
+      "grad_norm": 1.0455493927001953,
+      "learning_rate": 1.8000882628479558e-05,
+      "loss": 0.234,
+      "step": 1448
+    },
+    {
+      "epoch": 0.6861201538916839,
+      "grad_norm": 1.0974806547164917,
+      "learning_rate": 1.7997813723275834e-05,
+      "loss": 0.2592,
+      "step": 1449
+    },
+    {
+      "epoch": 0.6865936667653152,
+      "grad_norm": 1.3178282976150513,
+      "learning_rate": 1.799474272632109e-05,
+      "loss": 0.2275,
+      "step": 1450
+    },
+    {
+      "epoch": 0.6870671796389465,
+      "grad_norm": 1.2756496667861938,
+      "learning_rate": 1.7991669638418515e-05,
+      "loss": 0.2165,
+      "step": 1451
+    },
+    {
+      "epoch": 0.6875406925125777,
+      "grad_norm": 1.5571712255477905,
+      "learning_rate": 1.798859446037184e-05,
+      "loss": 0.2507,
+      "step": 1452
+    },
+    {
+      "epoch": 0.6880142053862089,
+      "grad_norm": 1.1532251834869385,
+      "learning_rate": 1.798551719298535e-05,
+      "loss": 0.2464,
+      "step": 1453
+    },
+    {
+      "epoch": 0.6884877182598402,
+      "grad_norm": 1.2180898189544678,
+      "learning_rate": 1.7982437837063878e-05,
+      "loss": 0.2383,
+      "step": 1454
+    },
+    {
+      "epoch": 0.6889612311334714,
+      "grad_norm": 1.2221808433532715,
+      "learning_rate": 1.7979356393412796e-05,
+      "loss": 0.2409,
+      "step": 1455
+    },
+    {
+      "epoch": 0.6894347440071027,
+      "grad_norm": 1.3098372220993042,
+      "learning_rate": 1.797627286283803e-05,
+      "loss": 0.2462,
+      "step": 1456
+    },
+    {
+      "epoch": 0.689908256880734,
+      "grad_norm": 1.6582274436950684,
+      "learning_rate": 1.7973187246146044e-05,
+      "loss": 0.2566,
+      "step": 1457
+    },
+    {
+      "epoch": 0.6903817697543652,
+      "grad_norm": 1.5802083015441895,
+      "learning_rate": 1.7970099544143852e-05,
+      "loss": 0.2495,
+      "step": 1458
+    },
+    {
+      "epoch": 0.6908552826279964,
+      "grad_norm": 1.2195215225219727,
+      "learning_rate": 1.7967009757639008e-05,
+      "loss": 0.2477,
+      "step": 1459
+    },
+    {
+      "epoch": 0.6913287955016277,
+      "grad_norm": 1.1696422100067139,
+      "learning_rate": 1.796391788743963e-05,
+      "loss": 0.2191,
+      "step": 1460
+    },
+    {
+      "epoch": 0.6918023083752589,
+      "grad_norm": 1.1211189031600952,
+      "learning_rate": 1.7960823934354352e-05,
+      "loss": 0.2616,
+      "step": 1461
+    },
+    {
+      "epoch": 0.6922758212488902,
+      "grad_norm": 1.4870905876159668,
+      "learning_rate": 1.7957727899192375e-05,
+      "loss": 0.2307,
+      "step": 1462
+    },
+    {
+      "epoch": 0.6927493341225215,
+      "grad_norm": 0.9586571455001831,
+      "learning_rate": 1.7954629782763437e-05,
+      "loss": 0.2507,
+      "step": 1463
+    },
+    {
+      "epoch": 0.6932228469961527,
+      "grad_norm": 1.396880865097046,
+      "learning_rate": 1.7951529585877818e-05,
+      "loss": 0.2456,
+      "step": 1464
+    },
+    {
+      "epoch": 0.693696359869784,
+      "grad_norm": 1.5271177291870117,
+      "learning_rate": 1.7948427309346346e-05,
+      "loss": 0.2374,
+      "step": 1465
+    },
+    {
+      "epoch": 0.6941698727434152,
+      "grad_norm": 1.1528421640396118,
+      "learning_rate": 1.7945322953980387e-05,
+      "loss": 0.2209,
+      "step": 1466
+    },
+    {
+      "epoch": 0.6946433856170464,
+      "grad_norm": 1.230507254600525,
+      "learning_rate": 1.7942216520591867e-05,
+      "loss": 0.2414,
+      "step": 1467
+    },
+    {
+      "epoch": 0.6951168984906777,
+      "grad_norm": 1.2597784996032715,
+      "learning_rate": 1.793910800999323e-05,
+      "loss": 0.2556,
+      "step": 1468
+    },
+    {
+      "epoch": 0.695590411364309,
+      "grad_norm": 1.1090929508209229,
+      "learning_rate": 1.7935997422997484e-05,
+      "loss": 0.2575,
+      "step": 1469
+    },
+    {
+      "epoch": 0.6960639242379402,
+      "grad_norm": 1.2305208444595337,
+      "learning_rate": 1.7932884760418172e-05,
+      "loss": 0.2437,
+      "step": 1470
+    },
+    {
+      "epoch": 0.6965374371115715,
+      "grad_norm": 1.3730130195617676,
+      "learning_rate": 1.7929770023069383e-05,
+      "loss": 0.2321,
+      "step": 1471
+    },
+    {
+      "epoch": 0.6970109499852027,
+      "grad_norm": 1.4209880828857422,
+      "learning_rate": 1.7926653211765742e-05,
+      "loss": 0.2597,
+      "step": 1472
+    },
+    {
+      "epoch": 0.6974844628588339,
+      "grad_norm": 1.2759629487991333,
+      "learning_rate": 1.7923534327322427e-05,
+      "loss": 0.2424,
+      "step": 1473
+    },
+    {
+      "epoch": 0.6979579757324652,
+      "grad_norm": 1.0876637697219849,
+      "learning_rate": 1.7920413370555143e-05,
+      "loss": 0.2589,
+      "step": 1474
+    },
+    {
+      "epoch": 0.6984314886060965,
+      "grad_norm": 1.1464848518371582,
+      "learning_rate": 1.7917290342280154e-05,
+      "loss": 0.2399,
+      "step": 1475
+    },
+    {
+      "epoch": 0.6989050014797278,
+      "grad_norm": 1.3303989171981812,
+      "learning_rate": 1.7914165243314256e-05,
+      "loss": 0.2653,
+      "step": 1476
+    },
+    {
+      "epoch": 0.699378514353359,
+      "grad_norm": 1.5119075775146484,
+      "learning_rate": 1.7911038074474788e-05,
+      "loss": 0.2422,
+      "step": 1477
+    },
+    {
+      "epoch": 0.6998520272269902,
+      "grad_norm": 1.3617953062057495,
+      "learning_rate": 1.790790883657963e-05,
+      "loss": 0.253,
+      "step": 1478
+    },
+    {
+      "epoch": 0.7003255401006215,
+      "grad_norm": 1.2398014068603516,
+      "learning_rate": 1.7904777530447203e-05,
+      "loss": 0.247,
+      "step": 1479
+    },
+    {
+      "epoch": 0.7007990529742527,
+      "grad_norm": 0.9873283505439758,
+      "learning_rate": 1.7901644156896474e-05,
+      "loss": 0.2248,
+      "step": 1480
+    },
+    {
+      "epoch": 0.701272565847884,
+      "grad_norm": 1.4423537254333496,
+      "learning_rate": 1.7898508716746944e-05,
+      "loss": 0.2449,
+      "step": 1481
+    },
+    {
+      "epoch": 0.7017460787215153,
+      "grad_norm": 1.600345253944397,
+      "learning_rate": 1.7895371210818656e-05,
+      "loss": 0.252,
+      "step": 1482
+    },
+    {
+      "epoch": 0.7022195915951465,
+      "grad_norm": 1.43065345287323,
+      "learning_rate": 1.78922316399322e-05,
+      "loss": 0.247,
+      "step": 1483
+    },
+    {
+      "epoch": 0.7026931044687778,
+      "grad_norm": 1.2158089876174927,
+      "learning_rate": 1.7889090004908692e-05,
+      "loss": 0.2162,
+      "step": 1484
+    },
+    {
+      "epoch": 0.703166617342409,
+      "grad_norm": 1.3399561643600464,
+      "learning_rate": 1.78859463065698e-05,
+      "loss": 0.2449,
+      "step": 1485
+    },
+    {
+      "epoch": 0.7036401302160402,
+      "grad_norm": 2.326958179473877,
+      "learning_rate": 1.788280054573773e-05,
+      "loss": 0.2354,
+      "step": 1486
+    },
+    {
+      "epoch": 0.7041136430896715,
+      "grad_norm": 1.8712157011032104,
+      "learning_rate": 1.7879652723235223e-05,
+      "loss": 0.2409,
+      "step": 1487
+    },
+    {
+      "epoch": 0.7045871559633028,
+      "grad_norm": 1.066486120223999,
+      "learning_rate": 1.7876502839885564e-05,
+      "loss": 0.2323,
+      "step": 1488
+    },
+    {
+      "epoch": 0.705060668836934,
+      "grad_norm": 1.8184261322021484,
+      "learning_rate": 1.7873350896512574e-05,
+      "loss": 0.2201,
+      "step": 1489
+    },
+    {
+      "epoch": 0.7055341817105653,
+      "grad_norm": 2.173187255859375,
+      "learning_rate": 1.787019689394061e-05,
+      "loss": 0.2419,
+      "step": 1490
+    },
+    {
+      "epoch": 0.7060076945841965,
+      "grad_norm": 1.2974936962127686,
+      "learning_rate": 1.786704083299458e-05,
+      "loss": 0.2354,
+      "step": 1491
+    },
+    {
+      "epoch": 0.7064812074578277,
+      "grad_norm": 1.272907018661499,
+      "learning_rate": 1.786388271449991e-05,
+      "loss": 0.2416,
+      "step": 1492
+    },
+    {
+      "epoch": 0.706954720331459,
+      "grad_norm": 0.8591986894607544,
+      "learning_rate": 1.7860722539282577e-05,
+      "loss": 0.2094,
+      "step": 1493
+    },
+    {
+      "epoch": 0.7074282332050903,
+      "grad_norm": 1.1995400190353394,
+      "learning_rate": 1.7857560308169103e-05,
+      "loss": 0.2289,
+      "step": 1494
+    },
+    {
+      "epoch": 0.7079017460787215,
+      "grad_norm": 1.7277050018310547,
+      "learning_rate": 1.785439602198653e-05,
+      "loss": 0.2252,
+      "step": 1495
+    },
+    {
+      "epoch": 0.7083752589523528,
+      "grad_norm": 2.3876638412475586,
+      "learning_rate": 1.785122968156245e-05,
+      "loss": 0.2494,
+      "step": 1496
+    },
+    {
+      "epoch": 0.708848771825984,
+      "grad_norm": 2.412504196166992,
+      "learning_rate": 1.7848061287724993e-05,
+      "loss": 0.2361,
+      "step": 1497
+    },
+    {
+      "epoch": 0.7093222846996152,
+      "grad_norm": 1.439038872718811,
+      "learning_rate": 1.7844890841302815e-05,
+      "loss": 0.251,
+      "step": 1498
+    },
+    {
+      "epoch": 0.7097957975732465,
+      "grad_norm": 1.5058058500289917,
+      "learning_rate": 1.7841718343125117e-05,
+      "loss": 0.2373,
+      "step": 1499
+    },
+    {
+      "epoch": 0.7102693104468778,
+      "grad_norm": 1.5263663530349731,
+      "learning_rate": 1.7838543794021637e-05,
+      "loss": 0.2423,
+      "step": 1500
+    },
+    {
+      "epoch": 0.710742823320509,
+      "grad_norm": 1.1687594652175903,
+      "learning_rate": 1.783536719482265e-05,
+      "loss": 0.2125,
+      "step": 1501
+    },
+    {
+      "epoch": 0.7112163361941403,
+      "grad_norm": 1.1493502855300903,
+      "learning_rate": 1.783218854635896e-05,
+      "loss": 0.2257,
+      "step": 1502
+    },
+    {
+      "epoch": 0.7116898490677716,
+      "grad_norm": 2.6674232482910156,
+      "learning_rate": 1.782900784946192e-05,
+      "loss": 0.2585,
+      "step": 1503
+    },
+    {
+      "epoch": 0.7121633619414027,
+      "grad_norm": 1.5568790435791016,
+      "learning_rate": 1.7825825104963398e-05,
+      "loss": 0.2436,
+      "step": 1504
+    },
+    {
+      "epoch": 0.712636874815034,
+      "grad_norm": 1.7096036672592163,
+      "learning_rate": 1.782264031369582e-05,
+      "loss": 0.2551,
+      "step": 1505
+    },
+    {
+      "epoch": 0.7131103876886653,
+      "grad_norm": 1.8939051628112793,
+      "learning_rate": 1.7819453476492136e-05,
+      "loss": 0.2482,
+      "step": 1506
+    },
+    {
+      "epoch": 0.7135839005622966,
+      "grad_norm": 1.7977244853973389,
+      "learning_rate": 1.7816264594185826e-05,
+      "loss": 0.2449,
+      "step": 1507
+    },
+    {
+      "epoch": 0.7140574134359278,
+      "grad_norm": 1.4691609144210815,
+      "learning_rate": 1.781307366761092e-05,
+      "loss": 0.2438,
+      "step": 1508
+    },
+    {
+      "epoch": 0.7145309263095591,
+      "grad_norm": 1.6823903322219849,
+      "learning_rate": 1.7809880697601965e-05,
+      "loss": 0.2588,
+      "step": 1509
+    },
+    {
+      "epoch": 0.7150044391831903,
+      "grad_norm": 1.1935667991638184,
+      "learning_rate": 1.7806685684994063e-05,
+      "loss": 0.2381,
+      "step": 1510
+    },
+    {
+      "epoch": 0.7154779520568215,
+      "grad_norm": 1.196661114692688,
+      "learning_rate": 1.7803488630622833e-05,
+      "loss": 0.2411,
+      "step": 1511
+    },
+    {
+      "epoch": 0.7159514649304528,
+      "grad_norm": 1.9287184476852417,
+      "learning_rate": 1.7800289535324426e-05,
+      "loss": 0.2545,
+      "step": 1512
+    },
+    {
+      "epoch": 0.7164249778040841,
+      "grad_norm": 2.1813395023345947,
+      "learning_rate": 1.7797088399935547e-05,
+      "loss": 0.2496,
+      "step": 1513
+    },
+    {
+      "epoch": 0.7168984906777153,
+      "grad_norm": 1.9686847925186157,
+      "learning_rate": 1.7793885225293418e-05,
+      "loss": 0.2352,
+      "step": 1514
+    },
+    {
+      "epoch": 0.7173720035513466,
+      "grad_norm": 1.5410264730453491,
+      "learning_rate": 1.779068001223579e-05,
+      "loss": 0.2156,
+      "step": 1515
+    },
+    {
+      "epoch": 0.7178455164249778,
+      "grad_norm": 1.2590621709823608,
+      "learning_rate": 1.7787472761600973e-05,
+      "loss": 0.2278,
+      "step": 1516
+    },
+    {
+      "epoch": 0.718319029298609,
+      "grad_norm": 1.5641133785247803,
+      "learning_rate": 1.7784263474227774e-05,
+      "loss": 0.2149,
+      "step": 1517
+    },
+    {
+      "epoch": 0.7187925421722403,
+      "grad_norm": 1.5322697162628174,
+      "learning_rate": 1.7781052150955566e-05,
+      "loss": 0.2337,
+      "step": 1518
+    },
+    {
+      "epoch": 0.7192660550458716,
+      "grad_norm": 1.4889428615570068,
+      "learning_rate": 1.7777838792624228e-05,
+      "loss": 0.2447,
+      "step": 1519
+    },
+    {
+      "epoch": 0.7197395679195028,
+      "grad_norm": 2.111682176589966,
+      "learning_rate": 1.777462340007419e-05,
+      "loss": 0.2279,
+      "step": 1520
+    },
+    {
+      "epoch": 0.7202130807931341,
+      "grad_norm": 1.946251630783081,
+      "learning_rate": 1.7771405974146403e-05,
+      "loss": 0.2399,
+      "step": 1521
+    },
+    {
+      "epoch": 0.7206865936667654,
+      "grad_norm": 1.0815329551696777,
+      "learning_rate": 1.776818651568236e-05,
+      "loss": 0.2292,
+      "step": 1522
+    },
+    {
+      "epoch": 0.7211601065403965,
+      "grad_norm": 1.4069883823394775,
+      "learning_rate": 1.7764965025524072e-05,
+      "loss": 0.2677,
+      "step": 1523
+    },
+    {
+      "epoch": 0.7216336194140278,
+      "grad_norm": 1.227126121520996,
+      "learning_rate": 1.776174150451409e-05,
+      "loss": 0.2468,
+      "step": 1524
+    },
+    {
+      "epoch": 0.7221071322876591,
+      "grad_norm": 1.791418433189392,
+      "learning_rate": 1.7758515953495496e-05,
+      "loss": 0.2335,
+      "step": 1525
+    },
+    {
+      "epoch": 0.7225806451612903,
+      "grad_norm": 2.3596413135528564,
+      "learning_rate": 1.7755288373311906e-05,
+      "loss": 0.2589,
+      "step": 1526
+    },
+    {
+      "epoch": 0.7230541580349216,
+      "grad_norm": 1.5666567087173462,
+      "learning_rate": 1.7752058764807455e-05,
+      "loss": 0.2381,
+      "step": 1527
+    },
+    {
+      "epoch": 0.7235276709085529,
+      "grad_norm": 1.1688320636749268,
+      "learning_rate": 1.7748827128826822e-05,
+      "loss": 0.2469,
+      "step": 1528
+    },
+    {
+      "epoch": 0.724001183782184,
+      "grad_norm": 1.2592840194702148,
+      "learning_rate": 1.7745593466215204e-05,
+      "loss": 0.224,
+      "step": 1529
+    },
+    {
+      "epoch": 0.7244746966558153,
+      "grad_norm": 1.0283125638961792,
+      "learning_rate": 1.774235777781834e-05,
+      "loss": 0.2054,
+      "step": 1530
+    },
+    {
+      "epoch": 0.7249482095294466,
+      "grad_norm": 1.7181527614593506,
+      "learning_rate": 1.7739120064482493e-05,
+      "loss": 0.2394,
+      "step": 1531
+    },
+    {
+      "epoch": 0.7254217224030778,
+      "grad_norm": 1.6182515621185303,
+      "learning_rate": 1.7735880327054453e-05,
+      "loss": 0.2225,
+      "step": 1532
+    },
+    {
+      "epoch": 0.7258952352767091,
+      "grad_norm": 1.1025639772415161,
+      "learning_rate": 1.7732638566381544e-05,
+      "loss": 0.2413,
+      "step": 1533
+    },
+    {
+      "epoch": 0.7263687481503404,
+      "grad_norm": 1.6946557760238647,
+      "learning_rate": 1.7729394783311614e-05,
+      "loss": 0.2424,
+      "step": 1534
+    },
+    {
+      "epoch": 0.7268422610239715,
+      "grad_norm": 1.2723066806793213,
+      "learning_rate": 1.7726148978693046e-05,
+      "loss": 0.2264,
+      "step": 1535
+    },
+    {
+      "epoch": 0.7273157738976028,
+      "grad_norm": 1.2535463571548462,
+      "learning_rate": 1.7722901153374748e-05,
+      "loss": 0.2516,
+      "step": 1536
+    },
+    {
+      "epoch": 0.7277892867712341,
+      "grad_norm": 1.6399126052856445,
+      "learning_rate": 1.7719651308206157e-05,
+      "loss": 0.237,
+      "step": 1537
+    },
+    {
+      "epoch": 0.7282627996448654,
+      "grad_norm": 1.1695709228515625,
+      "learning_rate": 1.771639944403724e-05,
+      "loss": 0.2125,
+      "step": 1538
+    },
+    {
+      "epoch": 0.7287363125184966,
+      "grad_norm": 1.4084415435791016,
+      "learning_rate": 1.7713145561718486e-05,
+      "loss": 0.2153,
+      "step": 1539
+    },
+    {
+      "epoch": 0.7292098253921279,
+      "grad_norm": 1.2717535495758057,
+      "learning_rate": 1.7709889662100926e-05,
+      "loss": 0.2599,
+      "step": 1540
+    },
+    {
+      "epoch": 0.7296833382657592,
+      "grad_norm": 1.3030693531036377,
+      "learning_rate": 1.77066317460361e-05,
+      "loss": 0.2262,
+      "step": 1541
+    },
+    {
+      "epoch": 0.7301568511393903,
+      "grad_norm": 1.2973533868789673,
+      "learning_rate": 1.7703371814376088e-05,
+      "loss": 0.2526,
+      "step": 1542
+    },
+    {
+      "epoch": 0.7306303640130216,
+      "grad_norm": 1.3349380493164062,
+      "learning_rate": 1.7700109867973494e-05,
+      "loss": 0.2436,
+      "step": 1543
+    },
+    {
+      "epoch": 0.7311038768866529,
+      "grad_norm": 1.424356460571289,
+      "learning_rate": 1.769684590768145e-05,
+      "loss": 0.2335,
+      "step": 1544
+    },
+    {
+      "epoch": 0.7315773897602841,
+      "grad_norm": 1.7350075244903564,
+      "learning_rate": 1.769357993435361e-05,
+      "loss": 0.2501,
+      "step": 1545
+    },
+    {
+      "epoch": 0.7320509026339154,
+      "grad_norm": 1.5108102560043335,
+      "learning_rate": 1.7690311948844162e-05,
+      "loss": 0.2606,
+      "step": 1546
+    },
+    {
+      "epoch": 0.7325244155075467,
+      "grad_norm": 1.3136991262435913,
+      "learning_rate": 1.768704195200781e-05,
+      "loss": 0.217,
+      "step": 1547
+    },
+    {
+      "epoch": 0.7329979283811778,
+      "grad_norm": 1.9467602968215942,
+      "learning_rate": 1.7683769944699793e-05,
+      "loss": 0.245,
+      "step": 1548
+    },
+    {
+      "epoch": 0.7334714412548091,
+      "grad_norm": 1.9547033309936523,
+      "learning_rate": 1.768049592777588e-05,
+      "loss": 0.2583,
+      "step": 1549
+    },
+    {
+      "epoch": 0.7339449541284404,
+      "grad_norm": 1.5787625312805176,
+      "learning_rate": 1.7677219902092345e-05,
+      "loss": 0.2554,
+      "step": 1550
+    },
+    {
+      "epoch": 0.7344184670020716,
+      "grad_norm": 1.423520803451538,
+      "learning_rate": 1.7673941868506014e-05,
+      "loss": 0.2397,
+      "step": 1551
+    },
+    {
+      "epoch": 0.7348919798757029,
+      "grad_norm": 2.1277055740356445,
+      "learning_rate": 1.7670661827874217e-05,
+      "loss": 0.2384,
+      "step": 1552
+    },
+    {
+      "epoch": 0.7353654927493342,
+      "grad_norm": 1.7148913145065308,
+      "learning_rate": 1.7667379781054816e-05,
+      "loss": 0.2562,
+      "step": 1553
+    },
+    {
+      "epoch": 0.7358390056229653,
+      "grad_norm": 1.5137001276016235,
+      "learning_rate": 1.7664095728906202e-05,
+      "loss": 0.2452,
+      "step": 1554
+    },
+    {
+      "epoch": 0.7363125184965966,
+      "grad_norm": 1.6710152626037598,
+      "learning_rate": 1.766080967228729e-05,
+      "loss": 0.2281,
+      "step": 1555
+    },
+    {
+      "epoch": 0.7367860313702279,
+      "grad_norm": 1.5264017581939697,
+      "learning_rate": 1.7657521612057513e-05,
+      "loss": 0.2265,
+      "step": 1556
+    },
+    {
+      "epoch": 0.7372595442438591,
+      "grad_norm": 2.8512990474700928,
+      "learning_rate": 1.765423154907683e-05,
+      "loss": 0.2342,
+      "step": 1557
+    },
+    {
+      "epoch": 0.7377330571174904,
+      "grad_norm": 1.2972322702407837,
+      "learning_rate": 1.7650939484205728e-05,
+      "loss": 0.2237,
+      "step": 1558
+    },
+    {
+      "epoch": 0.7382065699911217,
+      "grad_norm": 1.2768687009811401,
+      "learning_rate": 1.7647645418305215e-05,
+      "loss": 0.2285,
+      "step": 1559
+    },
+    {
+      "epoch": 0.7386800828647528,
+      "grad_norm": 0.9737210273742676,
+      "learning_rate": 1.7644349352236822e-05,
+      "loss": 0.2494,
+      "step": 1560
+    },
+    {
+      "epoch": 0.7391535957383841,
+      "grad_norm": 1.2401894330978394,
+      "learning_rate": 1.7641051286862597e-05,
+      "loss": 0.263,
+      "step": 1561
+    },
+    {
+      "epoch": 0.7396271086120154,
+      "grad_norm": 1.2521781921386719,
+      "learning_rate": 1.763775122304513e-05,
+      "loss": 0.2624,
+      "step": 1562
+    },
+    {
+      "epoch": 0.7401006214856466,
+      "grad_norm": 1.5043491125106812,
+      "learning_rate": 1.7634449161647506e-05,
+      "loss": 0.2325,
+      "step": 1563
+    },
+    {
+      "epoch": 0.7405741343592779,
+      "grad_norm": 1.1127467155456543,
+      "learning_rate": 1.7631145103533357e-05,
+      "loss": 0.2532,
+      "step": 1564
+    },
+    {
+      "epoch": 0.7410476472329092,
+      "grad_norm": 1.3029921054840088,
+      "learning_rate": 1.7627839049566827e-05,
+      "loss": 0.2445,
+      "step": 1565
+    },
+    {
+      "epoch": 0.7415211601065403,
+      "grad_norm": 1.4428067207336426,
+      "learning_rate": 1.762453100061258e-05,
+      "loss": 0.2123,
+      "step": 1566
+    },
+    {
+      "epoch": 0.7419946729801716,
+      "grad_norm": 1.1127543449401855,
+      "learning_rate": 1.76212209575358e-05,
+      "loss": 0.2265,
+      "step": 1567
+    },
+    {
+      "epoch": 0.7424681858538029,
+      "grad_norm": 0.9954766631126404,
+      "learning_rate": 1.761790892120221e-05,
+      "loss": 0.2089,
+      "step": 1568
+    },
+    {
+      "epoch": 0.7429416987274341,
+      "grad_norm": 1.198296070098877,
+      "learning_rate": 1.7614594892478026e-05,
+      "loss": 0.2272,
+      "step": 1569
+    },
+    {
+      "epoch": 0.7434152116010654,
+      "grad_norm": 1.9979981184005737,
+      "learning_rate": 1.761127887223001e-05,
+      "loss": 0.2511,
+      "step": 1570
+    },
+    {
+      "epoch": 0.7438887244746967,
+      "grad_norm": 1.982858657836914,
+      "learning_rate": 1.7607960861325434e-05,
+      "loss": 0.2372,
+      "step": 1571
+    },
+    {
+      "epoch": 0.744362237348328,
+      "grad_norm": 1.2229499816894531,
+      "learning_rate": 1.760464086063209e-05,
+      "loss": 0.2101,
+      "step": 1572
+    },
+    {
+      "epoch": 0.7448357502219591,
+      "grad_norm": 1.25929856300354,
+      "learning_rate": 1.760131887101829e-05,
+      "loss": 0.27,
+      "step": 1573
+    },
+    {
+      "epoch": 0.7453092630955904,
+      "grad_norm": 1.0544230937957764,
+      "learning_rate": 1.7597994893352873e-05,
+      "loss": 0.2076,
+      "step": 1574
+    },
+    {
+      "epoch": 0.7457827759692217,
+      "grad_norm": 1.874146819114685,
+      "learning_rate": 1.7594668928505186e-05,
+      "loss": 0.2267,
+      "step": 1575
+    },
+    {
+      "epoch": 0.7462562888428529,
+      "grad_norm": 1.3984766006469727,
+      "learning_rate": 1.7591340977345112e-05,
+      "loss": 0.218,
+      "step": 1576
+    },
+    {
+      "epoch": 0.7467298017164842,
+      "grad_norm": 1.5700955390930176,
+      "learning_rate": 1.758801104074304e-05,
+      "loss": 0.2278,
+      "step": 1577
+    },
+    {
+      "epoch": 0.7472033145901155,
+      "grad_norm": 1.9291844367980957,
+      "learning_rate": 1.7584679119569882e-05,
+      "loss": 0.2446,
+      "step": 1578
+    },
+    {
+      "epoch": 0.7476768274637466,
+      "grad_norm": 1.6862872838974,
+      "learning_rate": 1.758134521469707e-05,
+      "loss": 0.2461,
+      "step": 1579
+    },
+    {
+      "epoch": 0.7481503403373779,
+      "grad_norm": 2.24984073638916,
+      "learning_rate": 1.7578009326996556e-05,
+      "loss": 0.2702,
+      "step": 1580
+    },
+    {
+      "epoch": 0.7486238532110092,
+      "grad_norm": 1.163398265838623,
+      "learning_rate": 1.757467145734081e-05,
+      "loss": 0.2349,
+      "step": 1581
+    },
+    {
+      "epoch": 0.7490973660846404,
+      "grad_norm": 1.2229551076889038,
+      "learning_rate": 1.757133160660282e-05,
+      "loss": 0.2439,
+      "step": 1582
+    },
+    {
+      "epoch": 0.7495708789582717,
+      "grad_norm": 1.2336183786392212,
+      "learning_rate": 1.7567989775656088e-05,
+      "loss": 0.2629,
+      "step": 1583
+    },
+    {
+      "epoch": 0.750044391831903,
+      "grad_norm": 1.5344003438949585,
+      "learning_rate": 1.756464596537464e-05,
+      "loss": 0.2362,
+      "step": 1584
+    },
+    {
+      "epoch": 0.7505179047055341,
+      "grad_norm": 1.0532424449920654,
+      "learning_rate": 1.7561300176633015e-05,
+      "loss": 0.2238,
+      "step": 1585
+    },
+    {
+      "epoch": 0.7509914175791654,
+      "grad_norm": 1.153308391571045,
+      "learning_rate": 1.755795241030628e-05,
+      "loss": 0.2429,
+      "step": 1586
+    },
+    {
+      "epoch": 0.7514649304527967,
+      "grad_norm": 1.3348098993301392,
+      "learning_rate": 1.755460266727e-05,
+      "loss": 0.2435,
+      "step": 1587
+    },
+    {
+      "epoch": 0.7519384433264279,
+      "grad_norm": 1.3655999898910522,
+      "learning_rate": 1.7551250948400273e-05,
+      "loss": 0.2197,
+      "step": 1588
+    },
+    {
+      "epoch": 0.7524119562000592,
+      "grad_norm": 2.0334396362304688,
+      "learning_rate": 1.754789725457371e-05,
+      "loss": 0.2142,
+      "step": 1589
+    },
+    {
+      "epoch": 0.7528854690736905,
+      "grad_norm": 1.0390914678573608,
+      "learning_rate": 1.754454158666744e-05,
+      "loss": 0.2386,
+      "step": 1590
+    },
+    {
+      "epoch": 0.7533589819473216,
+      "grad_norm": 1.4241347312927246,
+      "learning_rate": 1.7541183945559095e-05,
+      "loss": 0.2464,
+      "step": 1591
+    },
+    {
+      "epoch": 0.7538324948209529,
+      "grad_norm": 1.2554645538330078,
+      "learning_rate": 1.7537824332126842e-05,
+      "loss": 0.2298,
+      "step": 1592
+    },
+    {
+      "epoch": 0.7543060076945842,
+      "grad_norm": 1.4590020179748535,
+      "learning_rate": 1.753446274724935e-05,
+      "loss": 0.2441,
+      "step": 1593
+    },
+    {
+      "epoch": 0.7547795205682154,
+      "grad_norm": 1.2696107625961304,
+      "learning_rate": 1.753109919180582e-05,
+      "loss": 0.2478,
+      "step": 1594
+    },
+    {
+      "epoch": 0.7552530334418467,
+      "grad_norm": 1.4026949405670166,
+      "learning_rate": 1.7527733666675945e-05,
+      "loss": 0.256,
+      "step": 1595
+    },
+    {
+      "epoch": 0.755726546315478,
+      "grad_norm": 1.0931724309921265,
+      "learning_rate": 1.7524366172739954e-05,
+      "loss": 0.2363,
+      "step": 1596
+    },
+    {
+      "epoch": 0.7562000591891092,
+      "grad_norm": 1.0928606986999512,
+      "learning_rate": 1.7520996710878577e-05,
+      "loss": 0.2427,
+      "step": 1597
+    },
+    {
+      "epoch": 0.7566735720627404,
+      "grad_norm": 1.247383713722229,
+      "learning_rate": 1.751762528197307e-05,
+      "loss": 0.2468,
+      "step": 1598
+    },
+    {
+      "epoch": 0.7571470849363717,
+      "grad_norm": 1.1337532997131348,
+      "learning_rate": 1.7514251886905192e-05,
+      "loss": 0.2321,
+      "step": 1599
+    },
+    {
+      "epoch": 0.7576205978100029,
+      "grad_norm": 1.2659448385238647,
+      "learning_rate": 1.7510876526557225e-05,
+      "loss": 0.2439,
+      "step": 1600
+    },
+    {
+      "epoch": 0.7580941106836342,
+      "grad_norm": 1.3613624572753906,
+      "learning_rate": 1.7507499201811958e-05,
+      "loss": 0.233,
+      "step": 1601
+    },
+    {
+      "epoch": 0.7585676235572655,
+      "grad_norm": 1.2028915882110596,
+      "learning_rate": 1.7504119913552707e-05,
+      "loss": 0.2176,
+      "step": 1602
+    },
+    {
+      "epoch": 0.7590411364308968,
+      "grad_norm": 1.3185116052627563,
+      "learning_rate": 1.750073866266328e-05,
+      "loss": 0.2633,
+      "step": 1603
+    },
+    {
+      "epoch": 0.759514649304528,
+      "grad_norm": 1.093675971031189,
+      "learning_rate": 1.749735545002802e-05,
+      "loss": 0.2641,
+      "step": 1604
+    },
+    {
+      "epoch": 0.7599881621781592,
+      "grad_norm": 1.3118257522583008,
+      "learning_rate": 1.7493970276531768e-05,
+      "loss": 0.2769,
+      "step": 1605
+    },
+    {
+      "epoch": 0.7604616750517905,
+      "grad_norm": 1.8869200944900513,
+      "learning_rate": 1.7490583143059885e-05,
+      "loss": 0.247,
+      "step": 1606
+    },
+    {
+      "epoch": 0.7609351879254217,
+      "grad_norm": 1.1935513019561768,
+      "learning_rate": 1.748719405049825e-05,
+      "loss": 0.2098,
+      "step": 1607
+    },
+    {
+      "epoch": 0.761408700799053,
+      "grad_norm": 1.2408536672592163,
+      "learning_rate": 1.7483802999733237e-05,
+      "loss": 0.227,
+      "step": 1608
+    },
+    {
+      "epoch": 0.7618822136726843,
+      "grad_norm": 1.9241235256195068,
+      "learning_rate": 1.748040999165175e-05,
+      "loss": 0.2486,
+      "step": 1609
+    },
+    {
+      "epoch": 0.7623557265463154,
+      "grad_norm": 2.5321922302246094,
+      "learning_rate": 1.7477015027141192e-05,
+      "loss": 0.2458,
+      "step": 1610
+    },
+    {
+      "epoch": 0.7628292394199467,
+      "grad_norm": 1.3591722249984741,
+      "learning_rate": 1.7473618107089482e-05,
+      "loss": 0.2064,
+      "step": 1611
+    },
+    {
+      "epoch": 0.763302752293578,
+      "grad_norm": 1.9121155738830566,
+      "learning_rate": 1.747021923238506e-05,
+      "loss": 0.2565,
+      "step": 1612
+    },
+    {
+      "epoch": 0.7637762651672092,
+      "grad_norm": 1.7858729362487793,
+      "learning_rate": 1.7466818403916862e-05,
+      "loss": 0.2642,
+      "step": 1613
+    },
+    {
+      "epoch": 0.7642497780408405,
+      "grad_norm": 1.5614818334579468,
+      "learning_rate": 1.7463415622574346e-05,
+      "loss": 0.2363,
+      "step": 1614
+    },
+    {
+      "epoch": 0.7647232909144718,
+      "grad_norm": 1.199522852897644,
+      "learning_rate": 1.7460010889247473e-05,
+      "loss": 0.2525,
+      "step": 1615
+    },
+    {
+      "epoch": 0.765196803788103,
+      "grad_norm": 1.5608481168746948,
+      "learning_rate": 1.7456604204826725e-05,
+      "loss": 0.2218,
+      "step": 1616
+    },
+    {
+      "epoch": 0.7656703166617342,
+      "grad_norm": 2.3330767154693604,
+      "learning_rate": 1.7453195570203075e-05,
+      "loss": 0.2299,
+      "step": 1617
+    },
+    {
+      "epoch": 0.7661438295353655,
+      "grad_norm": 2.918964385986328,
+      "learning_rate": 1.7449784986268033e-05,
+      "loss": 0.2124,
+      "step": 1618
+    },
+    {
+      "epoch": 0.7666173424089967,
+      "grad_norm": 2.289975643157959,
+      "learning_rate": 1.7446372453913592e-05,
+      "loss": 0.2549,
+      "step": 1619
+    },
+    {
+      "epoch": 0.767090855282628,
+      "grad_norm": 1.3809151649475098,
+      "learning_rate": 1.7442957974032274e-05,
+      "loss": 0.2461,
+      "step": 1620
+    },
+    {
+      "epoch": 0.7675643681562593,
+      "grad_norm": 1.484054446220398,
+      "learning_rate": 1.74395415475171e-05,
+      "loss": 0.2427,
+      "step": 1621
+    },
+    {
+      "epoch": 0.7680378810298905,
+      "grad_norm": 1.378715991973877,
+      "learning_rate": 1.7436123175261607e-05,
+      "loss": 0.2572,
+      "step": 1622
+    },
+    {
+      "epoch": 0.7685113939035217,
+      "grad_norm": 1.168416976928711,
+      "learning_rate": 1.7432702858159835e-05,
+      "loss": 0.2395,
+      "step": 1623
+    },
+    {
+      "epoch": 0.768984906777153,
+      "grad_norm": 1.0436391830444336,
+      "learning_rate": 1.742928059710633e-05,
+      "loss": 0.2144,
+      "step": 1624
+    },
+    {
+      "epoch": 0.7694584196507842,
+      "grad_norm": 2.4827699661254883,
+      "learning_rate": 1.742585639299616e-05,
+      "loss": 0.2408,
+      "step": 1625
+    },
+    {
+      "epoch": 0.7699319325244155,
+      "grad_norm": 2.694228410720825,
+      "learning_rate": 1.7422430246724892e-05,
+      "loss": 0.2193,
+      "step": 1626
+    },
+    {
+      "epoch": 0.7704054453980468,
+      "grad_norm": 1.7454769611358643,
+      "learning_rate": 1.7419002159188593e-05,
+      "loss": 0.2328,
+      "step": 1627
+    },
+    {
+      "epoch": 0.770878958271678,
+      "grad_norm": 2.133936643600464,
+      "learning_rate": 1.7415572131283856e-05,
+      "loss": 0.2401,
+      "step": 1628
+    },
+    {
+      "epoch": 0.7713524711453092,
+      "grad_norm": 1.3482989072799683,
+      "learning_rate": 1.7412140163907765e-05,
+      "loss": 0.2338,
+      "step": 1629
+    },
+    {
+      "epoch": 0.7718259840189405,
+      "grad_norm": 1.8004233837127686,
+      "learning_rate": 1.7408706257957922e-05,
+      "loss": 0.2271,
+      "step": 1630
+    },
+    {
+      "epoch": 0.7722994968925717,
+      "grad_norm": 1.835065484046936,
+      "learning_rate": 1.740527041433243e-05,
+      "loss": 0.2661,
+      "step": 1631
+    },
+    {
+      "epoch": 0.772773009766203,
+      "grad_norm": 1.9195033311843872,
+      "learning_rate": 1.7401832633929897e-05,
+      "loss": 0.2482,
+      "step": 1632
+    },
+    {
+      "epoch": 0.7732465226398343,
+      "grad_norm": 1.346403956413269,
+      "learning_rate": 1.7398392917649448e-05,
+      "loss": 0.2437,
+      "step": 1633
+    },
+    {
+      "epoch": 0.7737200355134656,
+      "grad_norm": 1.3023998737335205,
+      "learning_rate": 1.7394951266390708e-05,
+      "loss": 0.2476,
+      "step": 1634
+    },
+    {
+      "epoch": 0.7741935483870968,
+      "grad_norm": 1.6358240842819214,
+      "learning_rate": 1.7391507681053802e-05,
+      "loss": 0.2367,
+      "step": 1635
+    },
+    {
+      "epoch": 0.774667061260728,
+      "grad_norm": 1.2242166996002197,
+      "learning_rate": 1.7388062162539368e-05,
+      "loss": 0.2527,
+      "step": 1636
+    },
+    {
+      "epoch": 0.7751405741343593,
+      "grad_norm": 1.1994009017944336,
+      "learning_rate": 1.738461471174855e-05,
+      "loss": 0.2417,
+      "step": 1637
+    },
+    {
+      "epoch": 0.7756140870079905,
+      "grad_norm": 1.6394872665405273,
+      "learning_rate": 1.7381165329582996e-05,
+      "loss": 0.2193,
+      "step": 1638
+    },
+    {
+      "epoch": 0.7760875998816218,
+      "grad_norm": 1.1021173000335693,
+      "learning_rate": 1.7377714016944856e-05,
+      "loss": 0.2529,
+      "step": 1639
+    },
+    {
+      "epoch": 0.7765611127552531,
+      "grad_norm": 0.9502620697021484,
+      "learning_rate": 1.737426077473679e-05,
+      "loss": 0.2347,
+      "step": 1640
+    },
+    {
+      "epoch": 0.7770346256288843,
+      "grad_norm": 1.5584297180175781,
+      "learning_rate": 1.737080560386196e-05,
+      "loss": 0.2619,
+      "step": 1641
+    },
+    {
+      "epoch": 0.7775081385025155,
+      "grad_norm": 1.0773392915725708,
+      "learning_rate": 1.736734850522403e-05,
+      "loss": 0.2445,
+      "step": 1642
+    },
+    {
+      "epoch": 0.7779816513761468,
+      "grad_norm": 0.9587130546569824,
+      "learning_rate": 1.736388947972717e-05,
+      "loss": 0.2168,
+      "step": 1643
+    },
+    {
+      "epoch": 0.778455164249778,
+      "grad_norm": 1.201387882232666,
+      "learning_rate": 1.7360428528276062e-05,
+      "loss": 0.2421,
+      "step": 1644
+    },
+    {
+      "epoch": 0.7789286771234093,
+      "grad_norm": 1.5380560159683228,
+      "learning_rate": 1.735696565177588e-05,
+      "loss": 0.2278,
+      "step": 1645
+    },
+    {
+      "epoch": 0.7794021899970406,
+      "grad_norm": 1.150850534439087,
+      "learning_rate": 1.7353500851132305e-05,
+      "loss": 0.2434,
+      "step": 1646
+    },
+    {
+      "epoch": 0.7798757028706718,
+      "grad_norm": 1.1250213384628296,
+      "learning_rate": 1.735003412725152e-05,
+      "loss": 0.2712,
+      "step": 1647
+    },
+    {
+      "epoch": 0.780349215744303,
+      "grad_norm": 1.428877592086792,
+      "learning_rate": 1.7346565481040218e-05,
+      "loss": 0.2397,
+      "step": 1648
+    },
+    {
+      "epoch": 0.7808227286179343,
+      "grad_norm": 1.245980978012085,
+      "learning_rate": 1.734309491340559e-05,
+      "loss": 0.2259,
+      "step": 1649
+    },
+    {
+      "epoch": 0.7812962414915655,
+      "grad_norm": 1.2803640365600586,
+      "learning_rate": 1.7339622425255323e-05,
+      "loss": 0.224,
+      "step": 1650
+    },
+    {
+      "epoch": 0.7817697543651968,
+      "grad_norm": 1.3713748455047607,
+      "learning_rate": 1.7336148017497617e-05,
+      "loss": 0.2288,
+      "step": 1651
+    },
+    {
+      "epoch": 0.7822432672388281,
+      "grad_norm": 1.4782074689865112,
+      "learning_rate": 1.7332671691041173e-05,
+      "loss": 0.2334,
+      "step": 1652
+    },
+    {
+      "epoch": 0.7827167801124593,
+      "grad_norm": 1.6904542446136475,
+      "learning_rate": 1.7329193446795186e-05,
+      "loss": 0.2539,
+      "step": 1653
+    },
+    {
+      "epoch": 0.7831902929860906,
+      "grad_norm": 2.416238784790039,
+      "learning_rate": 1.732571328566936e-05,
+      "loss": 0.2279,
+      "step": 1654
+    },
+    {
+      "epoch": 0.7836638058597218,
+      "grad_norm": 2.9311797618865967,
+      "learning_rate": 1.7322231208573897e-05,
+      "loss": 0.2518,
+      "step": 1655
+    },
+    {
+      "epoch": 0.784137318733353,
+      "grad_norm": 1.2711611986160278,
+      "learning_rate": 1.7318747216419502e-05,
+      "loss": 0.2443,
+      "step": 1656
+    },
+    {
+      "epoch": 0.7846108316069843,
+      "grad_norm": 1.742221713066101,
+      "learning_rate": 1.7315261310117376e-05,
+      "loss": 0.2431,
+      "step": 1657
+    },
+    {
+      "epoch": 0.7850843444806156,
+      "grad_norm": 1.3446333408355713,
+      "learning_rate": 1.7311773490579225e-05,
+      "loss": 0.2241,
+      "step": 1658
+    },
+    {
+      "epoch": 0.7855578573542468,
+      "grad_norm": 1.3194044828414917,
+      "learning_rate": 1.7308283758717255e-05,
+      "loss": 0.2399,
+      "step": 1659
+    },
+    {
+      "epoch": 0.7860313702278781,
+      "grad_norm": 1.2633658647537231,
+      "learning_rate": 1.7304792115444172e-05,
+      "loss": 0.2336,
+      "step": 1660
+    },
+    {
+      "epoch": 0.7865048831015093,
+      "grad_norm": 1.167573094367981,
+      "learning_rate": 1.7301298561673186e-05,
+      "loss": 0.2337,
+      "step": 1661
+    },
+    {
+      "epoch": 0.7869783959751405,
+      "grad_norm": 1.2154494524002075,
+      "learning_rate": 1.7297803098317995e-05,
+      "loss": 0.2571,
+      "step": 1662
+    },
+    {
+      "epoch": 0.7874519088487718,
+      "grad_norm": 1.3965184688568115,
+      "learning_rate": 1.7294305726292804e-05,
+      "loss": 0.2521,
+      "step": 1663
+    },
+    {
+      "epoch": 0.7879254217224031,
+      "grad_norm": 1.1656535863876343,
+      "learning_rate": 1.7290806446512324e-05,
+      "loss": 0.2134,
+      "step": 1664
+    },
+    {
+      "epoch": 0.7883989345960344,
+      "grad_norm": 1.8412251472473145,
+      "learning_rate": 1.728730525989175e-05,
+      "loss": 0.2219,
+      "step": 1665
+    },
+    {
+      "epoch": 0.7888724474696656,
+      "grad_norm": 2.7580809593200684,
+      "learning_rate": 1.7283802167346793e-05,
+      "loss": 0.2484,
+      "step": 1666
+    },
+    {
+      "epoch": 0.7893459603432968,
+      "grad_norm": 1.3492392301559448,
+      "learning_rate": 1.7280297169793643e-05,
+      "loss": 0.2292,
+      "step": 1667
+    },
+    {
+      "epoch": 0.7898194732169281,
+      "grad_norm": 1.6488065719604492,
+      "learning_rate": 1.7276790268149e-05,
+      "loss": 0.2416,
+      "step": 1668
+    },
+    {
+      "epoch": 0.7902929860905593,
+      "grad_norm": 1.1698122024536133,
+      "learning_rate": 1.7273281463330066e-05,
+      "loss": 0.2375,
+      "step": 1669
+    },
+    {
+      "epoch": 0.7907664989641906,
+      "grad_norm": 1.7447906732559204,
+      "learning_rate": 1.7269770756254532e-05,
+      "loss": 0.257,
+      "step": 1670
+    },
+    {
+      "epoch": 0.7912400118378219,
+      "grad_norm": 1.752770185470581,
+      "learning_rate": 1.726625814784059e-05,
+      "loss": 0.2087,
+      "step": 1671
+    },
+    {
+      "epoch": 0.7917135247114531,
+      "grad_norm": 1.0216140747070312,
+      "learning_rate": 1.7262743639006928e-05,
+      "loss": 0.2084,
+      "step": 1672
+    },
+    {
+      "epoch": 0.7921870375850844,
+      "grad_norm": 1.923169493675232,
+      "learning_rate": 1.7259227230672733e-05,
+      "loss": 0.231,
+      "step": 1673
+    },
+    {
+      "epoch": 0.7926605504587156,
+      "grad_norm": 1.4375102519989014,
+      "learning_rate": 1.7255708923757683e-05,
+      "loss": 0.2381,
+      "step": 1674
+    },
+    {
+      "epoch": 0.7931340633323468,
+      "grad_norm": 1.4559400081634521,
+      "learning_rate": 1.7252188719181965e-05,
+      "loss": 0.2328,
+      "step": 1675
+    },
+    {
+      "epoch": 0.7936075762059781,
+      "grad_norm": 1.6422553062438965,
+      "learning_rate": 1.724866661786625e-05,
+      "loss": 0.2457,
+      "step": 1676
+    },
+    {
+      "epoch": 0.7940810890796094,
+      "grad_norm": 1.408460259437561,
+      "learning_rate": 1.7245142620731707e-05,
+      "loss": 0.2479,
+      "step": 1677
+    },
+    {
+      "epoch": 0.7945546019532406,
+      "grad_norm": 1.476257562637329,
+      "learning_rate": 1.724161672870001e-05,
+      "loss": 0.2193,
+      "step": 1678
+    },
+    {
+      "epoch": 0.7950281148268719,
+      "grad_norm": 1.108467698097229,
+      "learning_rate": 1.7238088942693315e-05,
+      "loss": 0.2265,
+      "step": 1679
+    },
+    {
+      "epoch": 0.7955016277005031,
+      "grad_norm": 1.2234896421432495,
+      "learning_rate": 1.723455926363429e-05,
+      "loss": 0.2383,
+      "step": 1680
+    },
+    {
+      "epoch": 0.7959751405741343,
+      "grad_norm": 0.9792563915252686,
+      "learning_rate": 1.723102769244608e-05,
+      "loss": 0.2288,
+      "step": 1681
+    },
+    {
+      "epoch": 0.7964486534477656,
+      "grad_norm": 1.273815393447876,
+      "learning_rate": 1.7227494230052337e-05,
+      "loss": 0.2287,
+      "step": 1682
+    },
+    {
+      "epoch": 0.7969221663213969,
+      "grad_norm": 1.5027164220809937,
+      "learning_rate": 1.72239588773772e-05,
+      "loss": 0.2635,
+      "step": 1683
+    },
+    {
+      "epoch": 0.7973956791950281,
+      "grad_norm": 1.1177582740783691,
+      "learning_rate": 1.7220421635345312e-05,
+      "loss": 0.2351,
+      "step": 1684
+    },
+    {
+      "epoch": 0.7978691920686594,
+      "grad_norm": 1.341367483139038,
+      "learning_rate": 1.7216882504881796e-05,
+      "loss": 0.2124,
+      "step": 1685
+    },
+    {
+      "epoch": 0.7983427049422906,
+      "grad_norm": 1.2859315872192383,
+      "learning_rate": 1.721334148691229e-05,
+      "loss": 0.2186,
+      "step": 1686
+    },
+    {
+      "epoch": 0.7988162178159218,
+      "grad_norm": 1.8942023515701294,
+      "learning_rate": 1.7209798582362904e-05,
+      "loss": 0.2746,
+      "step": 1687
+    },
+    {
+      "epoch": 0.7992897306895531,
+      "grad_norm": 1.0868715047836304,
+      "learning_rate": 1.7206253792160255e-05,
+      "loss": 0.238,
+      "step": 1688
+    },
+    {
+      "epoch": 0.7997632435631844,
+      "grad_norm": 1.7027019262313843,
+      "learning_rate": 1.7202707117231443e-05,
+      "loss": 0.2381,
+      "step": 1689
+    },
+    {
+      "epoch": 0.8002367564368156,
+      "grad_norm": 1.0921725034713745,
+      "learning_rate": 1.7199158558504074e-05,
+      "loss": 0.2556,
+      "step": 1690
+    },
+    {
+      "epoch": 0.8007102693104469,
+      "grad_norm": 1.053194522857666,
+      "learning_rate": 1.7195608116906234e-05,
+      "loss": 0.2267,
+      "step": 1691
+    },
+    {
+      "epoch": 0.8011837821840782,
+      "grad_norm": 1.3293603658676147,
+      "learning_rate": 1.7192055793366506e-05,
+      "loss": 0.2312,
+      "step": 1692
+    },
+    {
+      "epoch": 0.8016572950577093,
+      "grad_norm": 1.9301906824111938,
+      "learning_rate": 1.718850158881397e-05,
+      "loss": 0.2198,
+      "step": 1693
+    },
+    {
+      "epoch": 0.8021308079313406,
+      "grad_norm": 1.5638904571533203,
+      "learning_rate": 1.7184945504178193e-05,
+      "loss": 0.2624,
+      "step": 1694
+    },
+    {
+      "epoch": 0.8026043208049719,
+      "grad_norm": 1.715846061706543,
+      "learning_rate": 1.7181387540389235e-05,
+      "loss": 0.2195,
+      "step": 1695
+    },
+    {
+      "epoch": 0.8030778336786032,
+      "grad_norm": 1.0800182819366455,
+      "learning_rate": 1.7177827698377646e-05,
+      "loss": 0.2419,
+      "step": 1696
+    },
+    {
+      "epoch": 0.8035513465522344,
+      "grad_norm": 1.205739974975586,
+      "learning_rate": 1.7174265979074464e-05,
+      "loss": 0.2211,
+      "step": 1697
+    },
+    {
+      "epoch": 0.8040248594258657,
+      "grad_norm": 1.1386871337890625,
+      "learning_rate": 1.717070238341123e-05,
+      "loss": 0.2216,
+      "step": 1698
+    },
+    {
+      "epoch": 0.8044983722994969,
+      "grad_norm": 1.242944598197937,
+      "learning_rate": 1.7167136912319962e-05,
+      "loss": 0.2197,
+      "step": 1699
+    },
+    {
+      "epoch": 0.8049718851731281,
+      "grad_norm": 1.081717610359192,
+      "learning_rate": 1.716356956673318e-05,
+      "loss": 0.2377,
+      "step": 1700
+    },
+    {
+      "epoch": 0.8054453980467594,
+      "grad_norm": 1.5992521047592163,
+      "learning_rate": 1.7160000347583885e-05,
+      "loss": 0.2345,
+      "step": 1701
+    },
+    {
+      "epoch": 0.8059189109203907,
+      "grad_norm": 1.3137495517730713,
+      "learning_rate": 1.715642925580557e-05,
+      "loss": 0.2322,
+      "step": 1702
+    },
+    {
+      "epoch": 0.8063924237940219,
+      "grad_norm": 1.910725474357605,
+      "learning_rate": 1.7152856292332225e-05,
+      "loss": 0.2487,
+      "step": 1703
+    },
+    {
+      "epoch": 0.8068659366676532,
+      "grad_norm": 1.408942699432373,
+      "learning_rate": 1.7149281458098325e-05,
+      "loss": 0.2556,
+      "step": 1704
+    },
+    {
+      "epoch": 0.8073394495412844,
+      "grad_norm": 1.5198936462402344,
+      "learning_rate": 1.7145704754038825e-05,
+      "loss": 0.2308,
+      "step": 1705
+    },
+    {
+      "epoch": 0.8078129624149156,
+      "grad_norm": 1.0505871772766113,
+      "learning_rate": 1.7142126181089184e-05,
+      "loss": 0.2366,
+      "step": 1706
+    },
+    {
+      "epoch": 0.8082864752885469,
+      "grad_norm": 1.4777036905288696,
+      "learning_rate": 1.713854574018534e-05,
+      "loss": 0.2297,
+      "step": 1707
+    },
+    {
+      "epoch": 0.8087599881621782,
+      "grad_norm": 1.320868730545044,
+      "learning_rate": 1.7134963432263725e-05,
+      "loss": 0.2222,
+      "step": 1708
+    },
+    {
+      "epoch": 0.8092335010358094,
+      "grad_norm": 1.3242131471633911,
+      "learning_rate": 1.713137925826126e-05,
+      "loss": 0.2458,
+      "step": 1709
+    },
+    {
+      "epoch": 0.8097070139094407,
+      "grad_norm": 1.8690855503082275,
+      "learning_rate": 1.7127793219115345e-05,
+      "loss": 0.2241,
+      "step": 1710
+    },
+    {
+      "epoch": 0.810180526783072,
+      "grad_norm": 1.4543685913085938,
+      "learning_rate": 1.7124205315763876e-05,
+      "loss": 0.2369,
+      "step": 1711
+    },
+    {
+      "epoch": 0.8106540396567031,
+      "grad_norm": 1.1340458393096924,
+      "learning_rate": 1.7120615549145234e-05,
+      "loss": 0.2446,
+      "step": 1712
+    },
+    {
+      "epoch": 0.8111275525303344,
+      "grad_norm": 2.3450067043304443,
+      "learning_rate": 1.711702392019829e-05,
+      "loss": 0.2507,
+      "step": 1713
+    },
+    {
+      "epoch": 0.8116010654039657,
+      "grad_norm": 1.5409314632415771,
+      "learning_rate": 1.71134304298624e-05,
+      "loss": 0.213,
+      "step": 1714
+    },
+    {
+      "epoch": 0.8120745782775969,
+      "grad_norm": 1.5783090591430664,
+      "learning_rate": 1.7109835079077406e-05,
+      "loss": 0.2278,
+      "step": 1715
+    },
+    {
+      "epoch": 0.8125480911512282,
+      "grad_norm": 1.2138385772705078,
+      "learning_rate": 1.7106237868783635e-05,
+      "loss": 0.2228,
+      "step": 1716
+    },
+    {
+      "epoch": 0.8130216040248595,
+      "grad_norm": 1.4703527688980103,
+      "learning_rate": 1.710263879992191e-05,
+      "loss": 0.2285,
+      "step": 1717
+    },
+    {
+      "epoch": 0.8134951168984906,
+      "grad_norm": 1.418495774269104,
+      "learning_rate": 1.709903787343352e-05,
+      "loss": 0.2528,
+      "step": 1718
+    },
+    {
+      "epoch": 0.8139686297721219,
+      "grad_norm": 2.0826809406280518,
+      "learning_rate": 1.7095435090260263e-05,
+      "loss": 0.2252,
+      "step": 1719
+    },
+    {
+      "epoch": 0.8144421426457532,
+      "grad_norm": 1.1822896003723145,
+      "learning_rate": 1.7091830451344406e-05,
+      "loss": 0.2336,
+      "step": 1720
+    },
+    {
+      "epoch": 0.8149156555193844,
+      "grad_norm": 2.3410282135009766,
+      "learning_rate": 1.7088223957628714e-05,
+      "loss": 0.2513,
+      "step": 1721
+    },
+    {
+      "epoch": 0.8153891683930157,
+      "grad_norm": 0.9941398501396179,
+      "learning_rate": 1.708461561005643e-05,
+      "loss": 0.2511,
+      "step": 1722
+    },
+    {
+      "epoch": 0.815862681266647,
+      "grad_norm": 1.2086396217346191,
+      "learning_rate": 1.708100540957127e-05,
+      "loss": 0.2535,
+      "step": 1723
+    },
+    {
+      "epoch": 0.8163361941402781,
+      "grad_norm": 1.3249119520187378,
+      "learning_rate": 1.7077393357117467e-05,
+      "loss": 0.252,
+      "step": 1724
+    },
+    {
+      "epoch": 0.8168097070139094,
+      "grad_norm": 1.0457037687301636,
+      "learning_rate": 1.70737794536397e-05,
+      "loss": 0.2265,
+      "step": 1725
+    },
+    {
+      "epoch": 0.8172832198875407,
+      "grad_norm": 1.7129755020141602,
+      "learning_rate": 1.7070163700083163e-05,
+      "loss": 0.2345,
+      "step": 1726
+    },
+    {
+      "epoch": 0.8177567327611719,
+      "grad_norm": 1.0255284309387207,
+      "learning_rate": 1.7066546097393518e-05,
+      "loss": 0.2303,
+      "step": 1727
+    },
+    {
+      "epoch": 0.8182302456348032,
+      "grad_norm": 1.2299504280090332,
+      "learning_rate": 1.7062926646516915e-05,
+      "loss": 0.2454,
+      "step": 1728
+    },
+    {
+      "epoch": 0.8187037585084345,
+      "grad_norm": 1.1177610158920288,
+      "learning_rate": 1.705930534839998e-05,
+      "loss": 0.2478,
+      "step": 1729
+    },
+    {
+      "epoch": 0.8191772713820658,
+      "grad_norm": 1.1740856170654297,
+      "learning_rate": 1.7055682203989838e-05,
+      "loss": 0.2253,
+      "step": 1730
+    },
+    {
+      "epoch": 0.8196507842556969,
+      "grad_norm": 2.3321869373321533,
+      "learning_rate": 1.705205721423408e-05,
+      "loss": 0.2168,
+      "step": 1731
+    },
+    {
+      "epoch": 0.8201242971293282,
+      "grad_norm": 1.5028231143951416,
+      "learning_rate": 1.7048430380080793e-05,
+      "loss": 0.2292,
+      "step": 1732
+    },
+    {
+      "epoch": 0.8205978100029595,
+      "grad_norm": 1.116110920906067,
+      "learning_rate": 1.7044801702478534e-05,
+      "loss": 0.2118,
+      "step": 1733
+    },
+    {
+      "epoch": 0.8210713228765907,
+      "grad_norm": 2.010495901107788,
+      "learning_rate": 1.7041171182376354e-05,
+      "loss": 0.2414,
+      "step": 1734
+    },
+    {
+      "epoch": 0.821544835750222,
+      "grad_norm": 1.1517298221588135,
+      "learning_rate": 1.7037538820723777e-05,
+      "loss": 0.2366,
+      "step": 1735
+    },
+    {
+      "epoch": 0.8220183486238533,
+      "grad_norm": 1.393791913986206,
+      "learning_rate": 1.7033904618470814e-05,
+      "loss": 0.2457,
+      "step": 1736
+    },
+    {
+      "epoch": 0.8224918614974844,
+      "grad_norm": 1.2726801633834839,
+      "learning_rate": 1.7030268576567956e-05,
+      "loss": 0.2476,
+      "step": 1737
+    },
+    {
+      "epoch": 0.8229653743711157,
+      "grad_norm": 1.198588490486145,
+      "learning_rate": 1.7026630695966172e-05,
+      "loss": 0.2283,
+      "step": 1738
+    },
+    {
+      "epoch": 0.823438887244747,
+      "grad_norm": 1.3464490175247192,
+      "learning_rate": 1.7022990977616922e-05,
+      "loss": 0.2221,
+      "step": 1739
+    },
+    {
+      "epoch": 0.8239124001183782,
+      "grad_norm": 1.504508137702942,
+      "learning_rate": 1.7019349422472128e-05,
+      "loss": 0.2292,
+      "step": 1740
+    },
+    {
+      "epoch": 0.8243859129920095,
+      "grad_norm": 1.8641504049301147,
+      "learning_rate": 1.7015706031484215e-05,
+      "loss": 0.2391,
+      "step": 1741
+    },
+    {
+      "epoch": 0.8248594258656408,
+      "grad_norm": 1.2077713012695312,
+      "learning_rate": 1.7012060805606067e-05,
+      "loss": 0.2267,
+      "step": 1742
+    },
+    {
+      "epoch": 0.8253329387392719,
+      "grad_norm": 1.1325595378875732,
+      "learning_rate": 1.7008413745791063e-05,
+      "loss": 0.2077,
+      "step": 1743
+    },
+    {
+      "epoch": 0.8258064516129032,
+      "grad_norm": 1.5017510652542114,
+      "learning_rate": 1.7004764852993056e-05,
+      "loss": 0.2344,
+      "step": 1744
+    },
+    {
+      "epoch": 0.8262799644865345,
+      "grad_norm": 1.8392795324325562,
+      "learning_rate": 1.7001114128166375e-05,
+      "loss": 0.2312,
+      "step": 1745
+    },
+    {
+      "epoch": 0.8267534773601657,
+      "grad_norm": 1.4507255554199219,
+      "learning_rate": 1.6997461572265838e-05,
+      "loss": 0.2629,
+      "step": 1746
+    },
+    {
+      "epoch": 0.827226990233797,
+      "grad_norm": 1.8967761993408203,
+      "learning_rate": 1.6993807186246735e-05,
+      "loss": 0.2135,
+      "step": 1747
+    },
+    {
+      "epoch": 0.8277005031074283,
+      "grad_norm": 1.259427785873413,
+      "learning_rate": 1.6990150971064827e-05,
+      "loss": 0.2199,
+      "step": 1748
+    },
+    {
+      "epoch": 0.8281740159810594,
+      "grad_norm": 1.5154216289520264,
+      "learning_rate": 1.698649292767637e-05,
+      "loss": 0.2385,
+      "step": 1749
+    },
+    {
+      "epoch": 0.8286475288546907,
+      "grad_norm": 1.33448326587677,
+      "learning_rate": 1.6982833057038087e-05,
+      "loss": 0.229,
+      "step": 1750
+    },
+    {
+      "epoch": 0.829121041728322,
+      "grad_norm": 1.2957985401153564,
+      "learning_rate": 1.6979171360107183e-05,
+      "loss": 0.2483,
+      "step": 1751
+    },
+    {
+      "epoch": 0.8295945546019532,
+      "grad_norm": 1.2026399374008179,
+      "learning_rate": 1.6975507837841338e-05,
+      "loss": 0.2452,
+      "step": 1752
+    },
+    {
+      "epoch": 0.8300680674755845,
+      "grad_norm": 1.6381677389144897,
+      "learning_rate": 1.6971842491198716e-05,
+      "loss": 0.2299,
+      "step": 1753
+    },
+    {
+      "epoch": 0.8305415803492158,
+      "grad_norm": 1.994767189025879,
+      "learning_rate": 1.6968175321137942e-05,
+      "loss": 0.2343,
+      "step": 1754
+    },
+    {
+      "epoch": 0.831015093222847,
+      "grad_norm": 1.1473991870880127,
+      "learning_rate": 1.696450632861814e-05,
+      "loss": 0.2431,
+      "step": 1755
+    },
+    {
+      "epoch": 0.8314886060964782,
+      "grad_norm": 1.2881150245666504,
+      "learning_rate": 1.6960835514598897e-05,
+      "loss": 0.245,
+      "step": 1756
+    },
+    {
+      "epoch": 0.8319621189701095,
+      "grad_norm": 1.086107850074768,
+      "learning_rate": 1.6957162880040273e-05,
+      "loss": 0.221,
+      "step": 1757
+    },
+    {
+      "epoch": 0.8324356318437407,
+      "grad_norm": 1.6053413152694702,
+      "learning_rate": 1.695348842590282e-05,
+      "loss": 0.2587,
+      "step": 1758
+    },
+    {
+      "epoch": 0.832909144717372,
+      "grad_norm": 1.3724325895309448,
+      "learning_rate": 1.694981215314755e-05,
+      "loss": 0.2508,
+      "step": 1759
+    },
+    {
+      "epoch": 0.8333826575910033,
+      "grad_norm": 1.0389055013656616,
+      "learning_rate": 1.6946134062735953e-05,
+      "loss": 0.214,
+      "step": 1760
+    },
+    {
+      "epoch": 0.8338561704646346,
+      "grad_norm": 1.638784646987915,
+      "learning_rate": 1.6942454155630005e-05,
+      "loss": 0.2257,
+      "step": 1761
+    },
+    {
+      "epoch": 0.8343296833382657,
+      "grad_norm": 1.2976818084716797,
+      "learning_rate": 1.6938772432792146e-05,
+      "loss": 0.2631,
+      "step": 1762
+    },
+    {
+      "epoch": 0.834803196211897,
+      "grad_norm": 1.4617828130722046,
+      "learning_rate": 1.69350888951853e-05,
+      "loss": 0.2217,
+      "step": 1763
+    },
+    {
+      "epoch": 0.8352767090855283,
+      "grad_norm": 1.6357537508010864,
+      "learning_rate": 1.693140354377286e-05,
+      "loss": 0.276,
+      "step": 1764
+    },
+    {
+      "epoch": 0.8357502219591595,
+      "grad_norm": 2.0060503482818604,
+      "learning_rate": 1.6927716379518683e-05,
+      "loss": 0.2226,
+      "step": 1765
+    },
+    {
+      "epoch": 0.8362237348327908,
+      "grad_norm": 1.3638052940368652,
+      "learning_rate": 1.6924027403387125e-05,
+      "loss": 0.2312,
+      "step": 1766
+    },
+    {
+      "epoch": 0.8366972477064221,
+      "grad_norm": 1.0776060819625854,
+      "learning_rate": 1.6920336616343e-05,
+      "loss": 0.2261,
+      "step": 1767
+    },
+    {
+      "epoch": 0.8371707605800532,
+      "grad_norm": 1.6560837030410767,
+      "learning_rate": 1.691664401935159e-05,
+      "loss": 0.2287,
+      "step": 1768
+    },
+    {
+      "epoch": 0.8376442734536845,
+      "grad_norm": 1.7977901697158813,
+      "learning_rate": 1.691294961337866e-05,
+      "loss": 0.2451,
+      "step": 1769
+    },
+    {
+      "epoch": 0.8381177863273158,
+      "grad_norm": 1.1947377920150757,
+      "learning_rate": 1.6909253399390453e-05,
+      "loss": 0.2327,
+      "step": 1770
+    },
+    {
+      "epoch": 0.838591299200947,
+      "grad_norm": 1.2023718357086182,
+      "learning_rate": 1.6905555378353676e-05,
+      "loss": 0.2413,
+      "step": 1771
+    },
+    {
+      "epoch": 0.8390648120745783,
+      "grad_norm": 1.2229983806610107,
+      "learning_rate": 1.6901855551235505e-05,
+      "loss": 0.2319,
+      "step": 1772
+    },
+    {
+      "epoch": 0.8395383249482096,
+      "grad_norm": 2.0484726428985596,
+      "learning_rate": 1.68981539190036e-05,
+      "loss": 0.2407,
+      "step": 1773
+    },
+    {
+      "epoch": 0.8400118378218407,
+      "grad_norm": 1.0766804218292236,
+      "learning_rate": 1.6894450482626087e-05,
+      "loss": 0.2518,
+      "step": 1774
+    },
+    {
+      "epoch": 0.840485350695472,
+      "grad_norm": 1.1905542612075806,
+      "learning_rate": 1.6890745243071558e-05,
+      "loss": 0.2279,
+      "step": 1775
+    },
+    {
+      "epoch": 0.8409588635691033,
+      "grad_norm": 1.0002597570419312,
+      "learning_rate": 1.688703820130909e-05,
+      "loss": 0.235,
+      "step": 1776
+    },
+    {
+      "epoch": 0.8414323764427345,
+      "grad_norm": 1.6852128505706787,
+      "learning_rate": 1.688332935830822e-05,
+      "loss": 0.2301,
+      "step": 1777
+    },
+    {
+      "epoch": 0.8419058893163658,
+      "grad_norm": 0.9994098544120789,
+      "learning_rate": 1.687961871503896e-05,
+      "loss": 0.2174,
+      "step": 1778
+    },
+    {
+      "epoch": 0.8423794021899971,
+      "grad_norm": 1.0002872943878174,
+      "learning_rate": 1.6875906272471795e-05,
+      "loss": 0.2541,
+      "step": 1779
+    },
+    {
+      "epoch": 0.8428529150636282,
+      "grad_norm": 1.7979168891906738,
+      "learning_rate": 1.6872192031577675e-05,
+      "loss": 0.2421,
+      "step": 1780
+    },
+    {
+      "epoch": 0.8433264279372595,
+      "grad_norm": 1.4673607349395752,
+      "learning_rate": 1.6868475993328027e-05,
+      "loss": 0.2231,
+      "step": 1781
+    },
+    {
+      "epoch": 0.8437999408108908,
+      "grad_norm": 1.4320827722549438,
+      "learning_rate": 1.6864758158694747e-05,
+      "loss": 0.2273,
+      "step": 1782
+    },
+    {
+      "epoch": 0.844273453684522,
+      "grad_norm": 1.837491750717163,
+      "learning_rate": 1.6861038528650197e-05,
+      "loss": 0.2815,
+      "step": 1783
+    },
+    {
+      "epoch": 0.8447469665581533,
+      "grad_norm": 0.963911771774292,
+      "learning_rate": 1.685731710416721e-05,
+      "loss": 0.209,
+      "step": 1784
+    },
+    {
+      "epoch": 0.8452204794317846,
+      "grad_norm": 1.31343412399292,
+      "learning_rate": 1.685359388621908e-05,
+      "loss": 0.2288,
+      "step": 1785
+    },
+    {
+      "epoch": 0.8456939923054158,
+      "grad_norm": 1.2854394912719727,
+      "learning_rate": 1.6849868875779594e-05,
+      "loss": 0.2427,
+      "step": 1786
+    },
+    {
+      "epoch": 0.846167505179047,
+      "grad_norm": 1.5275187492370605,
+      "learning_rate": 1.6846142073822983e-05,
+      "loss": 0.259,
+      "step": 1787
+    },
+    {
+      "epoch": 0.8466410180526783,
+      "grad_norm": 1.2911092042922974,
+      "learning_rate": 1.684241348132396e-05,
+      "loss": 0.258,
+      "step": 1788
+    },
+    {
+      "epoch": 0.8471145309263095,
+      "grad_norm": 1.2889440059661865,
+      "learning_rate": 1.68386830992577e-05,
+      "loss": 0.235,
+      "step": 1789
+    },
+    {
+      "epoch": 0.8475880437999408,
+      "grad_norm": 1.3690677881240845,
+      "learning_rate": 1.683495092859985e-05,
+      "loss": 0.2401,
+      "step": 1790
+    },
+    {
+      "epoch": 0.8480615566735721,
+      "grad_norm": 1.3110579252243042,
+      "learning_rate": 1.683121697032652e-05,
+      "loss": 0.2563,
+      "step": 1791
+    },
+    {
+      "epoch": 0.8485350695472034,
+      "grad_norm": 1.2487221956253052,
+      "learning_rate": 1.6827481225414298e-05,
+      "loss": 0.2234,
+      "step": 1792
+    },
+    {
+      "epoch": 0.8490085824208345,
+      "grad_norm": 1.9341635704040527,
+      "learning_rate": 1.6823743694840226e-05,
+      "loss": 0.2653,
+      "step": 1793
+    },
+    {
+      "epoch": 0.8494820952944658,
+      "grad_norm": 1.6593029499053955,
+      "learning_rate": 1.6820004379581816e-05,
+      "loss": 0.2799,
+      "step": 1794
+    },
+    {
+      "epoch": 0.8499556081680971,
+      "grad_norm": 1.5015946626663208,
+      "learning_rate": 1.6816263280617054e-05,
+      "loss": 0.2821,
+      "step": 1795
+    },
+    {
+      "epoch": 0.8504291210417283,
+      "grad_norm": 1.0924832820892334,
+      "learning_rate": 1.6812520398924393e-05,
+      "loss": 0.2256,
+      "step": 1796
+    },
+    {
+      "epoch": 0.8509026339153596,
+      "grad_norm": 3.2573161125183105,
+      "learning_rate": 1.6808775735482746e-05,
+      "loss": 0.2438,
+      "step": 1797
+    },
+    {
+      "epoch": 0.8513761467889909,
+      "grad_norm": 1.6617119312286377,
+      "learning_rate": 1.6805029291271485e-05,
+      "loss": 0.2604,
+      "step": 1798
+    },
+    {
+      "epoch": 0.851849659662622,
+      "grad_norm": 1.2218396663665771,
+      "learning_rate": 1.680128106727046e-05,
+      "loss": 0.2546,
+      "step": 1799
+    },
+    {
+      "epoch": 0.8523231725362533,
+      "grad_norm": 1.5456347465515137,
+      "learning_rate": 1.6797531064459995e-05,
+      "loss": 0.232,
+      "step": 1800
+    },
+    {
+      "epoch": 0.8527966854098846,
+      "grad_norm": 1.6752814054489136,
+      "learning_rate": 1.679377928382085e-05,
+      "loss": 0.2486,
+      "step": 1801
+    },
+    {
+      "epoch": 0.8532701982835158,
+      "grad_norm": 1.3025007247924805,
+      "learning_rate": 1.6790025726334274e-05,
+      "loss": 0.243,
+      "step": 1802
+    },
+    {
+      "epoch": 0.8537437111571471,
+      "grad_norm": 1.4086090326309204,
+      "learning_rate": 1.6786270392981976e-05,
+      "loss": 0.2331,
+      "step": 1803
+    },
+    {
+      "epoch": 0.8542172240307784,
+      "grad_norm": 1.0914981365203857,
+      "learning_rate": 1.6782513284746124e-05,
+      "loss": 0.2667,
+      "step": 1804
+    },
+    {
+      "epoch": 0.8546907369044096,
+      "grad_norm": 1.737202525138855,
+      "learning_rate": 1.6778754402609356e-05,
+      "loss": 0.2584,
+      "step": 1805
+    },
+    {
+      "epoch": 0.8551642497780408,
+      "grad_norm": 1.0595831871032715,
+      "learning_rate": 1.6774993747554767e-05,
+      "loss": 0.2386,
+      "step": 1806
+    },
+    {
+      "epoch": 0.8556377626516721,
+      "grad_norm": 1.2180547714233398,
+      "learning_rate": 1.6771231320565925e-05,
+      "loss": 0.2429,
+      "step": 1807
+    },
+    {
+      "epoch": 0.8561112755253033,
+      "grad_norm": 1.1929233074188232,
+      "learning_rate": 1.6767467122626852e-05,
+      "loss": 0.227,
+      "step": 1808
+    },
+    {
+      "epoch": 0.8565847883989346,
+      "grad_norm": 1.554271936416626,
+      "learning_rate": 1.6763701154722043e-05,
+      "loss": 0.2447,
+      "step": 1809
+    },
+    {
+      "epoch": 0.8570583012725659,
+      "grad_norm": 1.0554946660995483,
+      "learning_rate": 1.6759933417836446e-05,
+      "loss": 0.2397,
+      "step": 1810
+    },
+    {
+      "epoch": 0.8575318141461971,
+      "grad_norm": 1.350473165512085,
+      "learning_rate": 1.6756163912955478e-05,
+      "loss": 0.2526,
+      "step": 1811
+    },
+    {
+      "epoch": 0.8580053270198283,
+      "grad_norm": 1.6592644453048706,
+      "learning_rate": 1.6752392641065015e-05,
+      "loss": 0.2294,
+      "step": 1812
+    },
+    {
+      "epoch": 0.8584788398934596,
+      "grad_norm": 0.9810163974761963,
+      "learning_rate": 1.67486196031514e-05,
+      "loss": 0.2413,
+      "step": 1813
+    },
+    {
+      "epoch": 0.8589523527670908,
+      "grad_norm": 1.351130485534668,
+      "learning_rate": 1.674484480020143e-05,
+      "loss": 0.2768,
+      "step": 1814
+    },
+    {
+      "epoch": 0.8594258656407221,
+      "grad_norm": 1.1572606563568115,
+      "learning_rate": 1.6741068233202374e-05,
+      "loss": 0.233,
+      "step": 1815
+    },
+    {
+      "epoch": 0.8598993785143534,
+      "grad_norm": 1.5954844951629639,
+      "learning_rate": 1.6737289903141954e-05,
+      "loss": 0.2755,
+      "step": 1816
+    },
+    {
+      "epoch": 0.8603728913879846,
+      "grad_norm": 1.2186288833618164,
+      "learning_rate": 1.6733509811008354e-05,
+      "loss": 0.2461,
+      "step": 1817
+    },
+    {
+      "epoch": 0.8608464042616158,
+      "grad_norm": 1.1969510316848755,
+      "learning_rate": 1.6729727957790224e-05,
+      "loss": 0.2269,
+      "step": 1818
+    },
+    {
+      "epoch": 0.8613199171352471,
+      "grad_norm": 1.5450844764709473,
+      "learning_rate": 1.672594434447667e-05,
+      "loss": 0.2235,
+      "step": 1819
+    },
+    {
+      "epoch": 0.8617934300088783,
+      "grad_norm": 1.2357224225997925,
+      "learning_rate": 1.672215897205726e-05,
+      "loss": 0.2287,
+      "step": 1820
+    },
+    {
+      "epoch": 0.8622669428825096,
+      "grad_norm": 1.3368169069290161,
+      "learning_rate": 1.6718371841522015e-05,
+      "loss": 0.2543,
+      "step": 1821
+    },
+    {
+      "epoch": 0.8627404557561409,
+      "grad_norm": 1.64212167263031,
+      "learning_rate": 1.6714582953861432e-05,
+      "loss": 0.2242,
+      "step": 1822
+    },
+    {
+      "epoch": 0.8632139686297722,
+      "grad_norm": 1.6220941543579102,
+      "learning_rate": 1.6710792310066454e-05,
+      "loss": 0.2628,
+      "step": 1823
+    },
+    {
+      "epoch": 0.8636874815034034,
+      "grad_norm": 1.4587221145629883,
+      "learning_rate": 1.6706999911128488e-05,
+      "loss": 0.2459,
+      "step": 1824
+    },
+    {
+      "epoch": 0.8641609943770346,
+      "grad_norm": 1.312369704246521,
+      "learning_rate": 1.6703205758039397e-05,
+      "loss": 0.249,
+      "step": 1825
+    },
+    {
+      "epoch": 0.8646345072506659,
+      "grad_norm": 1.632554054260254,
+      "learning_rate": 1.6699409851791507e-05,
+      "loss": 0.2515,
+      "step": 1826
+    },
+    {
+      "epoch": 0.8651080201242971,
+      "grad_norm": 1.6713101863861084,
+      "learning_rate": 1.6695612193377604e-05,
+      "loss": 0.2434,
+      "step": 1827
+    },
+    {
+      "epoch": 0.8655815329979284,
+      "grad_norm": 1.163874626159668,
+      "learning_rate": 1.6691812783790924e-05,
+      "loss": 0.2101,
+      "step": 1828
+    },
+    {
+      "epoch": 0.8660550458715597,
+      "grad_norm": 1.604141116142273,
+      "learning_rate": 1.6688011624025164e-05,
+      "loss": 0.2043,
+      "step": 1829
+    },
+    {
+      "epoch": 0.8665285587451909,
+      "grad_norm": 1.5439268350601196,
+      "learning_rate": 1.6684208715074488e-05,
+      "loss": 0.2381,
+      "step": 1830
+    },
+    {
+      "epoch": 0.8670020716188221,
+      "grad_norm": 2.153048038482666,
+      "learning_rate": 1.6680404057933504e-05,
+      "loss": 0.2072,
+      "step": 1831
+    },
+    {
+      "epoch": 0.8674755844924534,
+      "grad_norm": 1.1268725395202637,
+      "learning_rate": 1.6676597653597287e-05,
+      "loss": 0.2483,
+      "step": 1832
+    },
+    {
+      "epoch": 0.8679490973660846,
+      "grad_norm": 1.470022439956665,
+      "learning_rate": 1.6672789503061362e-05,
+      "loss": 0.24,
+      "step": 1833
+    },
+    {
+      "epoch": 0.8684226102397159,
+      "grad_norm": 1.7773938179016113,
+      "learning_rate": 1.6668979607321717e-05,
+      "loss": 0.2214,
+      "step": 1834
+    },
+    {
+      "epoch": 0.8688961231133472,
+      "grad_norm": 1.1891618967056274,
+      "learning_rate": 1.6665167967374795e-05,
+      "loss": 0.2411,
+      "step": 1835
+    },
+    {
+      "epoch": 0.8693696359869784,
+      "grad_norm": 2.1788344383239746,
+      "learning_rate": 1.666135458421749e-05,
+      "loss": 0.2029,
+      "step": 1836
+    },
+    {
+      "epoch": 0.8698431488606096,
+      "grad_norm": 1.9523930549621582,
+      "learning_rate": 1.6657539458847157e-05,
+      "loss": 0.2247,
+      "step": 1837
+    },
+    {
+      "epoch": 0.8703166617342409,
+      "grad_norm": 1.6084970235824585,
+      "learning_rate": 1.6653722592261605e-05,
+      "loss": 0.2564,
+      "step": 1838
+    },
+    {
+      "epoch": 0.8707901746078721,
+      "grad_norm": 1.0069568157196045,
+      "learning_rate": 1.6649903985459093e-05,
+      "loss": 0.2296,
+      "step": 1839
+    },
+    {
+      "epoch": 0.8712636874815034,
+      "grad_norm": 1.9601439237594604,
+      "learning_rate": 1.664608363943835e-05,
+      "loss": 0.2398,
+      "step": 1840
+    },
+    {
+      "epoch": 0.8717372003551347,
+      "grad_norm": 1.4246207475662231,
+      "learning_rate": 1.664226155519855e-05,
+      "loss": 0.2336,
+      "step": 1841
+    },
+    {
+      "epoch": 0.8722107132287659,
+      "grad_norm": 1.4902780055999756,
+      "learning_rate": 1.6638437733739317e-05,
+      "loss": 0.2309,
+      "step": 1842
+    },
+    {
+      "epoch": 0.8726842261023972,
+      "grad_norm": 1.1577699184417725,
+      "learning_rate": 1.6634612176060736e-05,
+      "loss": 0.2355,
+      "step": 1843
+    },
+    {
+      "epoch": 0.8731577389760284,
+      "grad_norm": 1.28193998336792,
+      "learning_rate": 1.6630784883163347e-05,
+      "loss": 0.2353,
+      "step": 1844
+    },
+    {
+      "epoch": 0.8736312518496596,
+      "grad_norm": 1.1589275598526,
+      "learning_rate": 1.662695585604814e-05,
+      "loss": 0.2501,
+      "step": 1845
+    },
+    {
+      "epoch": 0.8741047647232909,
+      "grad_norm": 1.8391677141189575,
+      "learning_rate": 1.662312509571656e-05,
+      "loss": 0.2188,
+      "step": 1846
+    },
+    {
+      "epoch": 0.8745782775969222,
+      "grad_norm": 1.7798473834991455,
+      "learning_rate": 1.6619292603170505e-05,
+      "loss": 0.2465,
+      "step": 1847
+    },
+    {
+      "epoch": 0.8750517904705534,
+      "grad_norm": 0.9828191995620728,
+      "learning_rate": 1.6615458379412327e-05,
+      "loss": 0.247,
+      "step": 1848
+    },
+    {
+      "epoch": 0.8755253033441847,
+      "grad_norm": 1.7325772047042847,
+      "learning_rate": 1.6611622425444834e-05,
+      "loss": 0.2165,
+      "step": 1849
+    },
+    {
+      "epoch": 0.875998816217816,
+      "grad_norm": 1.395041584968567,
+      "learning_rate": 1.6607784742271275e-05,
+      "loss": 0.2143,
+      "step": 1850
+    },
+    {
+      "epoch": 0.8764723290914471,
+      "grad_norm": 1.2467516660690308,
+      "learning_rate": 1.6603945330895364e-05,
+      "loss": 0.2121,
+      "step": 1851
+    },
+    {
+      "epoch": 0.8769458419650784,
+      "grad_norm": 1.279456615447998,
+      "learning_rate": 1.660010419232126e-05,
+      "loss": 0.2339,
+      "step": 1852
+    },
+    {
+      "epoch": 0.8774193548387097,
+      "grad_norm": 1.203802227973938,
+      "learning_rate": 1.659626132755358e-05,
+      "loss": 0.2216,
+      "step": 1853
+    },
+    {
+      "epoch": 0.8778928677123409,
+      "grad_norm": 1.2188462018966675,
+      "learning_rate": 1.6592416737597382e-05,
+      "loss": 0.2611,
+      "step": 1854
+    },
+    {
+      "epoch": 0.8783663805859722,
+      "grad_norm": 1.2634485960006714,
+      "learning_rate": 1.6588570423458185e-05,
+      "loss": 0.2343,
+      "step": 1855
+    },
+    {
+      "epoch": 0.8788398934596034,
+      "grad_norm": 1.070681095123291,
+      "learning_rate": 1.6584722386141955e-05,
+      "loss": 0.2278,
+      "step": 1856
+    },
+    {
+      "epoch": 0.8793134063332347,
+      "grad_norm": 1.0444326400756836,
+      "learning_rate": 1.6580872626655113e-05,
+      "loss": 0.2236,
+      "step": 1857
+    },
+    {
+      "epoch": 0.8797869192068659,
+      "grad_norm": 1.1631354093551636,
+      "learning_rate": 1.6577021146004514e-05,
+      "loss": 0.2565,
+      "step": 1858
+    },
+    {
+      "epoch": 0.8802604320804972,
+      "grad_norm": 2.2485361099243164,
+      "learning_rate": 1.6573167945197492e-05,
+      "loss": 0.243,
+      "step": 1859
+    },
+    {
+      "epoch": 0.8807339449541285,
+      "grad_norm": 1.143649697303772,
+      "learning_rate": 1.65693130252418e-05,
+      "loss": 0.2247,
+      "step": 1860
+    },
+    {
+      "epoch": 0.8812074578277597,
+      "grad_norm": 1.7670962810516357,
+      "learning_rate": 1.6565456387145667e-05,
+      "loss": 0.2466,
+      "step": 1861
+    },
+    {
+      "epoch": 0.881680970701391,
+      "grad_norm": 1.5178930759429932,
+      "learning_rate": 1.6561598031917752e-05,
+      "loss": 0.2272,
+      "step": 1862
+    },
+    {
+      "epoch": 0.8821544835750222,
+      "grad_norm": 2.0225138664245605,
+      "learning_rate": 1.655773796056717e-05,
+      "loss": 0.2544,
+      "step": 1863
+    },
+    {
+      "epoch": 0.8826279964486534,
+      "grad_norm": 1.248473048210144,
+      "learning_rate": 1.655387617410349e-05,
+      "loss": 0.2331,
+      "step": 1864
+    },
+    {
+      "epoch": 0.8831015093222847,
+      "grad_norm": 2.6657464504241943,
+      "learning_rate": 1.6550012673536725e-05,
+      "loss": 0.2501,
+      "step": 1865
+    },
+    {
+      "epoch": 0.883575022195916,
+      "grad_norm": 1.0329293012619019,
+      "learning_rate": 1.654614745987733e-05,
+      "loss": 0.2139,
+      "step": 1866
+    },
+    {
+      "epoch": 0.8840485350695472,
+      "grad_norm": 1.3353016376495361,
+      "learning_rate": 1.6542280534136223e-05,
+      "loss": 0.2816,
+      "step": 1867
+    },
+    {
+      "epoch": 0.8845220479431785,
+      "grad_norm": 0.9897204041481018,
+      "learning_rate": 1.6538411897324757e-05,
+      "loss": 0.2147,
+      "step": 1868
+    },
+    {
+      "epoch": 0.8849955608168097,
+      "grad_norm": 1.145198941230774,
+      "learning_rate": 1.653454155045473e-05,
+      "loss": 0.228,
+      "step": 1869
+    },
+    {
+      "epoch": 0.8854690736904409,
+      "grad_norm": 1.2381608486175537,
+      "learning_rate": 1.6530669494538403e-05,
+      "loss": 0.267,
+      "step": 1870
+    },
+    {
+      "epoch": 0.8859425865640722,
+      "grad_norm": 1.4348737001419067,
+      "learning_rate": 1.6526795730588477e-05,
+      "loss": 0.2294,
+      "step": 1871
+    },
+    {
+      "epoch": 0.8864160994377035,
+      "grad_norm": 1.178130030632019,
+      "learning_rate": 1.652292025961809e-05,
+      "loss": 0.2079,
+      "step": 1872
+    },
+    {
+      "epoch": 0.8868896123113347,
+      "grad_norm": 1.6016205549240112,
+      "learning_rate": 1.6519043082640834e-05,
+      "loss": 0.22,
+      "step": 1873
+    },
+    {
+      "epoch": 0.887363125184966,
+      "grad_norm": 1.1184009313583374,
+      "learning_rate": 1.6515164200670754e-05,
+      "loss": 0.2505,
+      "step": 1874
+    },
+    {
+      "epoch": 0.8878366380585972,
+      "grad_norm": 1.3344361782073975,
+      "learning_rate": 1.651128361472233e-05,
+      "loss": 0.2063,
+      "step": 1875
+    },
+    {
+      "epoch": 0.8883101509322284,
+      "grad_norm": 1.438899278640747,
+      "learning_rate": 1.6507401325810488e-05,
+      "loss": 0.2183,
+      "step": 1876
+    },
+    {
+      "epoch": 0.8887836638058597,
+      "grad_norm": 1.2038847208023071,
+      "learning_rate": 1.650351733495061e-05,
+      "loss": 0.2361,
+      "step": 1877
+    },
+    {
+      "epoch": 0.889257176679491,
+      "grad_norm": 1.5315783023834229,
+      "learning_rate": 1.6499631643158512e-05,
+      "loss": 0.2188,
+      "step": 1878
+    },
+    {
+      "epoch": 0.8897306895531222,
+      "grad_norm": 1.375877857208252,
+      "learning_rate": 1.6495744251450464e-05,
+      "loss": 0.238,
+      "step": 1879
+    },
+    {
+      "epoch": 0.8902042024267535,
+      "grad_norm": 2.0350253582000732,
+      "learning_rate": 1.6491855160843172e-05,
+      "loss": 0.2217,
+      "step": 1880
+    },
+    {
+      "epoch": 0.8906777153003848,
+      "grad_norm": 1.5811187028884888,
+      "learning_rate": 1.648796437235379e-05,
+      "loss": 0.233,
+      "step": 1881
+    },
+    {
+      "epoch": 0.8911512281740159,
+      "grad_norm": 1.5892693996429443,
+      "learning_rate": 1.6484071886999917e-05,
+      "loss": 0.2399,
+      "step": 1882
+    },
+    {
+      "epoch": 0.8916247410476472,
+      "grad_norm": 1.705972671508789,
+      "learning_rate": 1.6480177705799594e-05,
+      "loss": 0.2467,
+      "step": 1883
+    },
+    {
+      "epoch": 0.8920982539212785,
+      "grad_norm": 1.4446234703063965,
+      "learning_rate": 1.647628182977131e-05,
+      "loss": 0.2375,
+      "step": 1884
+    },
+    {
+      "epoch": 0.8925717667949097,
+      "grad_norm": 1.0984984636306763,
+      "learning_rate": 1.6472384259933986e-05,
+      "loss": 0.2171,
+      "step": 1885
+    },
+    {
+      "epoch": 0.893045279668541,
+      "grad_norm": 1.1526954174041748,
+      "learning_rate": 1.6468484997307003e-05,
+      "loss": 0.2437,
+      "step": 1886
+    },
+    {
+      "epoch": 0.8935187925421723,
+      "grad_norm": 1.1996570825576782,
+      "learning_rate": 1.646458404291017e-05,
+      "loss": 0.2596,
+      "step": 1887
+    },
+    {
+      "epoch": 0.8939923054158035,
+      "grad_norm": 2.095825433731079,
+      "learning_rate": 1.6460681397763746e-05,
+      "loss": 0.23,
+      "step": 1888
+    },
+    {
+      "epoch": 0.8944658182894347,
+      "grad_norm": 1.0920315980911255,
+      "learning_rate": 1.645677706288843e-05,
+      "loss": 0.2274,
+      "step": 1889
+    },
+    {
+      "epoch": 0.894939331163066,
+      "grad_norm": 1.171776294708252,
+      "learning_rate": 1.6452871039305365e-05,
+      "loss": 0.2355,
+      "step": 1890
+    },
+    {
+      "epoch": 0.8954128440366973,
+      "grad_norm": 1.1109461784362793,
+      "learning_rate": 1.6448963328036125e-05,
+      "loss": 0.2205,
+      "step": 1891
+    },
+    {
+      "epoch": 0.8958863569103285,
+      "grad_norm": 1.180391550064087,
+      "learning_rate": 1.6445053930102747e-05,
+      "loss": 0.2343,
+      "step": 1892
+    },
+    {
+      "epoch": 0.8963598697839598,
+      "grad_norm": 1.4910218715667725,
+      "learning_rate": 1.6441142846527688e-05,
+      "loss": 0.2145,
+      "step": 1893
+    },
+    {
+      "epoch": 0.896833382657591,
+      "grad_norm": 1.087941288948059,
+      "learning_rate": 1.6437230078333855e-05,
+      "loss": 0.2206,
+      "step": 1894
+    },
+    {
+      "epoch": 0.8973068955312222,
+      "grad_norm": 1.3966666460037231,
+      "learning_rate": 1.6433315626544598e-05,
+      "loss": 0.2289,
+      "step": 1895
+    },
+    {
+      "epoch": 0.8977804084048535,
+      "grad_norm": 1.0643726587295532,
+      "learning_rate": 1.6429399492183703e-05,
+      "loss": 0.2301,
+      "step": 1896
+    },
+    {
+      "epoch": 0.8982539212784848,
+      "grad_norm": 1.5619611740112305,
+      "learning_rate": 1.6425481676275396e-05,
+      "loss": 0.2353,
+      "step": 1897
+    },
+    {
+      "epoch": 0.898727434152116,
+      "grad_norm": 1.496078610420227,
+      "learning_rate": 1.642156217984434e-05,
+      "loss": 0.204,
+      "step": 1898
+    },
+    {
+      "epoch": 0.8992009470257473,
+      "grad_norm": 1.6565096378326416,
+      "learning_rate": 1.6417641003915653e-05,
+      "loss": 0.2353,
+      "step": 1899
+    },
+    {
+      "epoch": 0.8996744598993786,
+      "grad_norm": 1.215949535369873,
+      "learning_rate": 1.641371814951487e-05,
+      "loss": 0.2246,
+      "step": 1900
+    },
+    {
+      "epoch": 0.9001479727730097,
+      "grad_norm": 1.6516441106796265,
+      "learning_rate": 1.6409793617667976e-05,
+      "loss": 0.2404,
+      "step": 1901
+    },
+    {
+      "epoch": 0.900621485646641,
+      "grad_norm": 1.6095350980758667,
+      "learning_rate": 1.6405867409401403e-05,
+      "loss": 0.2316,
+      "step": 1902
+    },
+    {
+      "epoch": 0.9010949985202723,
+      "grad_norm": 1.5806862115859985,
+      "learning_rate": 1.6401939525742007e-05,
+      "loss": 0.2754,
+      "step": 1903
+    },
+    {
+      "epoch": 0.9015685113939035,
+      "grad_norm": 1.4216324090957642,
+      "learning_rate": 1.6398009967717086e-05,
+      "loss": 0.2322,
+      "step": 1904
+    },
+    {
+      "epoch": 0.9020420242675348,
+      "grad_norm": 1.7498546838760376,
+      "learning_rate": 1.639407873635438e-05,
+      "loss": 0.2352,
+      "step": 1905
+    },
+    {
+      "epoch": 0.9025155371411661,
+      "grad_norm": 0.969447135925293,
+      "learning_rate": 1.639014583268207e-05,
+      "loss": 0.2338,
+      "step": 1906
+    },
+    {
+      "epoch": 0.9029890500147972,
+      "grad_norm": 2.0034878253936768,
+      "learning_rate": 1.638621125772876e-05,
+      "loss": 0.2381,
+      "step": 1907
+    },
+    {
+      "epoch": 0.9034625628884285,
+      "grad_norm": 1.6915373802185059,
+      "learning_rate": 1.6382275012523503e-05,
+      "loss": 0.25,
+      "step": 1908
+    },
+    {
+      "epoch": 0.9039360757620598,
+      "grad_norm": 1.4558416604995728,
+      "learning_rate": 1.637833709809579e-05,
+      "loss": 0.2475,
+      "step": 1909
+    },
+    {
+      "epoch": 0.904409588635691,
+      "grad_norm": 1.660110354423523,
+      "learning_rate": 1.6374397515475543e-05,
+      "loss": 0.2226,
+      "step": 1910
+    },
+    {
+      "epoch": 0.9048831015093223,
+      "grad_norm": 1.2827781438827515,
+      "learning_rate": 1.637045626569312e-05,
+      "loss": 0.2301,
+      "step": 1911
+    },
+    {
+      "epoch": 0.9053566143829536,
+      "grad_norm": 2.6431972980499268,
+      "learning_rate": 1.6366513349779313e-05,
+      "loss": 0.2198,
+      "step": 1912
+    },
+    {
+      "epoch": 0.9058301272565847,
+      "grad_norm": 2.985435962677002,
+      "learning_rate": 1.6362568768765362e-05,
+      "loss": 0.2571,
+      "step": 1913
+    },
+    {
+      "epoch": 0.906303640130216,
+      "grad_norm": 2.624217987060547,
+      "learning_rate": 1.635862252368293e-05,
+      "loss": 0.2407,
+      "step": 1914
+    },
+    {
+      "epoch": 0.9067771530038473,
+      "grad_norm": 2.1122894287109375,
+      "learning_rate": 1.635467461556412e-05,
+      "loss": 0.229,
+      "step": 1915
+    },
+    {
+      "epoch": 0.9072506658774785,
+      "grad_norm": 1.5032678842544556,
+      "learning_rate": 1.6350725045441472e-05,
+      "loss": 0.2417,
+      "step": 1916
+    },
+    {
+      "epoch": 0.9077241787511098,
+      "grad_norm": 1.1959266662597656,
+      "learning_rate": 1.6346773814347952e-05,
+      "loss": 0.2241,
+      "step": 1917
+    },
+    {
+      "epoch": 0.9081976916247411,
+      "grad_norm": 2.0419347286224365,
+      "learning_rate": 1.634282092331697e-05,
+      "loss": 0.2398,
+      "step": 1918
+    },
+    {
+      "epoch": 0.9086712044983724,
+      "grad_norm": 2.168184280395508,
+      "learning_rate": 1.6338866373382366e-05,
+      "loss": 0.2275,
+      "step": 1919
+    },
+    {
+      "epoch": 0.9091447173720035,
+      "grad_norm": 2.695232629776001,
+      "learning_rate": 1.6334910165578413e-05,
+      "loss": 0.2307,
+      "step": 1920
+    },
+    {
+      "epoch": 0.9096182302456348,
+      "grad_norm": 2.0876898765563965,
+      "learning_rate": 1.6330952300939817e-05,
+      "loss": 0.2517,
+      "step": 1921
+    },
+    {
+      "epoch": 0.9100917431192661,
+      "grad_norm": 1.803197979927063,
+      "learning_rate": 1.6326992780501727e-05,
+      "loss": 0.2283,
+      "step": 1922
+    },
+    {
+      "epoch": 0.9105652559928973,
+      "grad_norm": 1.1686948537826538,
+      "learning_rate": 1.632303160529971e-05,
+      "loss": 0.2201,
+      "step": 1923
+    },
+    {
+      "epoch": 0.9110387688665286,
+      "grad_norm": 1.152266025543213,
+      "learning_rate": 1.6319068776369783e-05,
+      "loss": 0.2007,
+      "step": 1924
+    },
+    {
+      "epoch": 0.9115122817401599,
+      "grad_norm": 2.0857367515563965,
+      "learning_rate": 1.631510429474837e-05,
+      "loss": 0.2101,
+      "step": 1925
+    },
+    {
+      "epoch": 0.911985794613791,
+      "grad_norm": 1.8467085361480713,
+      "learning_rate": 1.6311138161472355e-05,
+      "loss": 0.2554,
+      "step": 1926
+    },
+    {
+      "epoch": 0.9124593074874223,
+      "grad_norm": 2.0132124423980713,
+      "learning_rate": 1.6307170377579038e-05,
+      "loss": 0.2226,
+      "step": 1927
+    },
+    {
+      "epoch": 0.9129328203610536,
+      "grad_norm": 2.2759344577789307,
+      "learning_rate": 1.6303200944106155e-05,
+      "loss": 0.2046,
+      "step": 1928
+    },
+    {
+      "epoch": 0.9134063332346848,
+      "grad_norm": 1.024839162826538,
+      "learning_rate": 1.6299229862091876e-05,
+      "loss": 0.2192,
+      "step": 1929
+    },
+    {
+      "epoch": 0.9138798461083161,
+      "grad_norm": 1.2846004962921143,
+      "learning_rate": 1.629525713257479e-05,
+      "loss": 0.2037,
+      "step": 1930
+    },
+    {
+      "epoch": 0.9143533589819474,
+      "grad_norm": 1.3948590755462646,
+      "learning_rate": 1.6291282756593937e-05,
+      "loss": 0.2416,
+      "step": 1931
+    },
+    {
+      "epoch": 0.9148268718555785,
+      "grad_norm": 2.4310286045074463,
+      "learning_rate": 1.628730673518877e-05,
+      "loss": 0.2268,
+      "step": 1932
+    },
+    {
+      "epoch": 0.9153003847292098,
+      "grad_norm": 1.4620105028152466,
+      "learning_rate": 1.6283329069399188e-05,
+      "loss": 0.2336,
+      "step": 1933
+    },
+    {
+      "epoch": 0.9157738976028411,
+      "grad_norm": 1.0024254322052002,
+      "learning_rate": 1.6279349760265497e-05,
+      "loss": 0.2386,
+      "step": 1934
+    },
+    {
+      "epoch": 0.9162474104764723,
+      "grad_norm": 1.0868014097213745,
+      "learning_rate": 1.6275368808828457e-05,
+      "loss": 0.2486,
+      "step": 1935
+    },
+    {
+      "epoch": 0.9167209233501036,
+      "grad_norm": 2.2207415103912354,
+      "learning_rate": 1.6271386216129245e-05,
+      "loss": 0.2333,
+      "step": 1936
+    },
+    {
+      "epoch": 0.9171944362237349,
+      "grad_norm": 1.3191596269607544,
+      "learning_rate": 1.6267401983209464e-05,
+      "loss": 0.2305,
+      "step": 1937
+    },
+    {
+      "epoch": 0.917667949097366,
+      "grad_norm": 1.842523217201233,
+      "learning_rate": 1.626341611111116e-05,
+      "loss": 0.2442,
+      "step": 1938
+    },
+    {
+      "epoch": 0.9181414619709973,
+      "grad_norm": 1.3315726518630981,
+      "learning_rate": 1.62594286008768e-05,
+      "loss": 0.24,
+      "step": 1939
+    },
+    {
+      "epoch": 0.9186149748446286,
+      "grad_norm": 2.0374701023101807,
+      "learning_rate": 1.6255439453549274e-05,
+      "loss": 0.2319,
+      "step": 1940
+    },
+    {
+      "epoch": 0.9190884877182598,
+      "grad_norm": 1.2693907022476196,
+      "learning_rate": 1.62514486701719e-05,
+      "loss": 0.2409,
+      "step": 1941
+    },
+    {
+      "epoch": 0.9195620005918911,
+      "grad_norm": 1.0667670965194702,
+      "learning_rate": 1.6247456251788444e-05,
+      "loss": 0.2414,
+      "step": 1942
+    },
+    {
+      "epoch": 0.9200355134655224,
+      "grad_norm": 1.2325665950775146,
+      "learning_rate": 1.624346219944307e-05,
+      "loss": 0.2227,
+      "step": 1943
+    },
+    {
+      "epoch": 0.9205090263391535,
+      "grad_norm": 2.379380941390991,
+      "learning_rate": 1.6239466514180393e-05,
+      "loss": 0.2169,
+      "step": 1944
+    },
+    {
+      "epoch": 0.9209825392127848,
+      "grad_norm": 1.3998842239379883,
+      "learning_rate": 1.623546919704544e-05,
+      "loss": 0.2434,
+      "step": 1945
+    },
+    {
+      "epoch": 0.9214560520864161,
+      "grad_norm": 1.3410049676895142,
+      "learning_rate": 1.6231470249083675e-05,
+      "loss": 0.2471,
+      "step": 1946
+    },
+    {
+      "epoch": 0.9219295649600473,
+      "grad_norm": 1.5513018369674683,
+      "learning_rate": 1.622746967134098e-05,
+      "loss": 0.2519,
+      "step": 1947
+    },
+    {
+      "epoch": 0.9224030778336786,
+      "grad_norm": 1.0884742736816406,
+      "learning_rate": 1.622346746486367e-05,
+      "loss": 0.2354,
+      "step": 1948
+    },
+    {
+      "epoch": 0.9228765907073099,
+      "grad_norm": 1.212095856666565,
+      "learning_rate": 1.6219463630698484e-05,
+      "loss": 0.2266,
+      "step": 1949
+    },
+    {
+      "epoch": 0.9233501035809412,
+      "grad_norm": 1.2227774858474731,
+      "learning_rate": 1.6215458169892582e-05,
+      "loss": 0.2474,
+      "step": 1950
+    },
+    {
+      "epoch": 0.9238236164545723,
+      "grad_norm": 1.3644037246704102,
+      "learning_rate": 1.6211451083493564e-05,
+      "loss": 0.2658,
+      "step": 1951
+    },
+    {
+      "epoch": 0.9242971293282036,
+      "grad_norm": 1.7697309255599976,
+      "learning_rate": 1.6207442372549436e-05,
+      "loss": 0.2118,
+      "step": 1952
+    },
+    {
+      "epoch": 0.9247706422018349,
+      "grad_norm": 1.1578645706176758,
+      "learning_rate": 1.6203432038108638e-05,
+      "loss": 0.2359,
+      "step": 1953
+    },
+    {
+      "epoch": 0.9252441550754661,
+      "grad_norm": 2.0185935497283936,
+      "learning_rate": 1.6199420081220035e-05,
+      "loss": 0.2172,
+      "step": 1954
+    },
+    {
+      "epoch": 0.9257176679490974,
+      "grad_norm": 1.3043760061264038,
+      "learning_rate": 1.619540650293292e-05,
+      "loss": 0.2327,
+      "step": 1955
+    },
+    {
+      "epoch": 0.9261911808227287,
+      "grad_norm": 0.9914813041687012,
+      "learning_rate": 1.6191391304297e-05,
+      "loss": 0.2352,
+      "step": 1956
+    },
+    {
+      "epoch": 0.9266646936963598,
+      "grad_norm": 2.466811180114746,
+      "learning_rate": 1.6187374486362414e-05,
+      "loss": 0.2604,
+      "step": 1957
+    },
+    {
+      "epoch": 0.9271382065699911,
+      "grad_norm": 1.7497360706329346,
+      "learning_rate": 1.6183356050179724e-05,
+      "loss": 0.2298,
+      "step": 1958
+    },
+    {
+      "epoch": 0.9276117194436224,
+      "grad_norm": 1.3502507209777832,
+      "learning_rate": 1.617933599679991e-05,
+      "loss": 0.2325,
+      "step": 1959
+    },
+    {
+      "epoch": 0.9280852323172536,
+      "grad_norm": 1.0206633806228638,
+      "learning_rate": 1.6175314327274377e-05,
+      "loss": 0.2066,
+      "step": 1960
+    },
+    {
+      "epoch": 0.9285587451908849,
+      "grad_norm": 1.507358193397522,
+      "learning_rate": 1.6171291042654957e-05,
+      "loss": 0.2335,
+      "step": 1961
+    },
+    {
+      "epoch": 0.9290322580645162,
+      "grad_norm": 1.3536046743392944,
+      "learning_rate": 1.6167266143993904e-05,
+      "loss": 0.2261,
+      "step": 1962
+    },
+    {
+      "epoch": 0.9295057709381473,
+      "grad_norm": 0.9721176028251648,
+      "learning_rate": 1.6163239632343883e-05,
+      "loss": 0.2134,
+      "step": 1963
+    },
+    {
+      "epoch": 0.9299792838117786,
+      "grad_norm": 1.2623369693756104,
+      "learning_rate": 1.6159211508757996e-05,
+      "loss": 0.2231,
+      "step": 1964
+    },
+    {
+      "epoch": 0.9304527966854099,
+      "grad_norm": 1.866391658782959,
+      "learning_rate": 1.615518177428976e-05,
+      "loss": 0.2441,
+      "step": 1965
+    },
+    {
+      "epoch": 0.9309263095590411,
+      "grad_norm": 1.2802680730819702,
+      "learning_rate": 1.6151150429993106e-05,
+      "loss": 0.2501,
+      "step": 1966
+    },
+    {
+      "epoch": 0.9313998224326724,
+      "grad_norm": 1.4352434873580933,
+      "learning_rate": 1.61471174769224e-05,
+      "loss": 0.2254,
+      "step": 1967
+    },
+    {
+      "epoch": 0.9318733353063037,
+      "grad_norm": 1.0232847929000854,
+      "learning_rate": 1.614308291613242e-05,
+      "loss": 0.2456,
+      "step": 1968
+    },
+    {
+      "epoch": 0.9323468481799349,
+      "grad_norm": 1.138548493385315,
+      "learning_rate": 1.6139046748678366e-05,
+      "loss": 0.2217,
+      "step": 1969
+    },
+    {
+      "epoch": 0.9328203610535661,
+      "grad_norm": 1.6218969821929932,
+      "learning_rate": 1.613500897561586e-05,
+      "loss": 0.2379,
+      "step": 1970
+    },
+    {
+      "epoch": 0.9332938739271974,
+      "grad_norm": 1.2417789697647095,
+      "learning_rate": 1.6130969598000945e-05,
+      "loss": 0.2304,
+      "step": 1971
+    },
+    {
+      "epoch": 0.9337673868008286,
+      "grad_norm": 1.4800796508789062,
+      "learning_rate": 1.6126928616890077e-05,
+      "loss": 0.2745,
+      "step": 1972
+    },
+    {
+      "epoch": 0.9342408996744599,
+      "grad_norm": 2.1304409503936768,
+      "learning_rate": 1.6122886033340134e-05,
+      "loss": 0.2395,
+      "step": 1973
+    },
+    {
+      "epoch": 0.9347144125480912,
+      "grad_norm": 1.1594027280807495,
+      "learning_rate": 1.6118841848408418e-05,
+      "loss": 0.2544,
+      "step": 1974
+    },
+    {
+      "epoch": 0.9351879254217224,
+      "grad_norm": 1.1337738037109375,
+      "learning_rate": 1.6114796063152648e-05,
+      "loss": 0.2067,
+      "step": 1975
+    },
+    {
+      "epoch": 0.9356614382953536,
+      "grad_norm": 1.8558512926101685,
+      "learning_rate": 1.611074867863096e-05,
+      "loss": 0.2369,
+      "step": 1976
+    },
+    {
+      "epoch": 0.9361349511689849,
+      "grad_norm": 1.4643030166625977,
+      "learning_rate": 1.61066996959019e-05,
+      "loss": 0.207,
+      "step": 1977
+    },
+    {
+      "epoch": 0.9366084640426161,
+      "grad_norm": 1.9388593435287476,
+      "learning_rate": 1.610264911602445e-05,
+      "loss": 0.2065,
+      "step": 1978
+    },
+    {
+      "epoch": 0.9370819769162474,
+      "grad_norm": 1.1505235433578491,
+      "learning_rate": 1.6098596940058e-05,
+      "loss": 0.2367,
+      "step": 1979
+    },
+    {
+      "epoch": 0.9375554897898787,
+      "grad_norm": 1.003483533859253,
+      "learning_rate": 1.6094543169062353e-05,
+      "loss": 0.2095,
+      "step": 1980
+    },
+    {
+      "epoch": 0.9380290026635099,
+      "grad_norm": 1.2500495910644531,
+      "learning_rate": 1.6090487804097734e-05,
+      "loss": 0.2377,
+      "step": 1981
+    },
+    {
+      "epoch": 0.9385025155371411,
+      "grad_norm": 1.7271959781646729,
+      "learning_rate": 1.6086430846224787e-05,
+      "loss": 0.2288,
+      "step": 1982
+    },
+    {
+      "epoch": 0.9389760284107724,
+      "grad_norm": 1.2142095565795898,
+      "learning_rate": 1.6082372296504568e-05,
+      "loss": 0.2158,
+      "step": 1983
+    },
+    {
+      "epoch": 0.9394495412844037,
+      "grad_norm": 0.9605250954627991,
+      "learning_rate": 1.6078312155998554e-05,
+      "loss": 0.205,
+      "step": 1984
+    },
+    {
+      "epoch": 0.9399230541580349,
+      "grad_norm": 1.0248587131500244,
+      "learning_rate": 1.6074250425768632e-05,
+      "loss": 0.2326,
+      "step": 1985
+    },
+    {
+      "epoch": 0.9403965670316662,
+      "grad_norm": 1.8944402933120728,
+      "learning_rate": 1.607018710687711e-05,
+      "loss": 0.227,
+      "step": 1986
+    },
+    {
+      "epoch": 0.9408700799052975,
+      "grad_norm": 0.8332754373550415,
+      "learning_rate": 1.6066122200386713e-05,
+      "loss": 0.2075,
+      "step": 1987
+    },
+    {
+      "epoch": 0.9413435927789287,
+      "grad_norm": 1.4765225648880005,
+      "learning_rate": 1.6062055707360575e-05,
+      "loss": 0.2085,
+      "step": 1988
+    },
+    {
+      "epoch": 0.9418171056525599,
+      "grad_norm": 1.3047866821289062,
+      "learning_rate": 1.6057987628862246e-05,
+      "loss": 0.2134,
+      "step": 1989
+    },
+    {
+      "epoch": 0.9422906185261912,
+      "grad_norm": 1.0208147764205933,
+      "learning_rate": 1.6053917965955698e-05,
+      "loss": 0.2091,
+      "step": 1990
+    },
+    {
+      "epoch": 0.9427641313998224,
+      "grad_norm": 1.4134602546691895,
+      "learning_rate": 1.6049846719705307e-05,
+      "loss": 0.2519,
+      "step": 1991
+    },
+    {
+      "epoch": 0.9432376442734537,
+      "grad_norm": 1.1737221479415894,
+      "learning_rate": 1.604577389117587e-05,
+      "loss": 0.2391,
+      "step": 1992
+    },
+    {
+      "epoch": 0.943711157147085,
+      "grad_norm": 1.1827245950698853,
+      "learning_rate": 1.60416994814326e-05,
+      "loss": 0.2178,
+      "step": 1993
+    },
+    {
+      "epoch": 0.9441846700207162,
+      "grad_norm": 1.5996443033218384,
+      "learning_rate": 1.6037623491541114e-05,
+      "loss": 0.2106,
+      "step": 1994
+    },
+    {
+      "epoch": 0.9446581828943474,
+      "grad_norm": 1.233630895614624,
+      "learning_rate": 1.6033545922567447e-05,
+      "loss": 0.2299,
+      "step": 1995
+    },
+    {
+      "epoch": 0.9451316957679787,
+      "grad_norm": 0.9926683306694031,
+      "learning_rate": 1.6029466775578054e-05,
+      "loss": 0.2464,
+      "step": 1996
+    },
+    {
+      "epoch": 0.9456052086416099,
+      "grad_norm": 1.6486265659332275,
+      "learning_rate": 1.602538605163979e-05,
+      "loss": 0.23,
+      "step": 1997
+    },
+    {
+      "epoch": 0.9460787215152412,
+      "grad_norm": 1.1196324825286865,
+      "learning_rate": 1.602130375181994e-05,
+      "loss": 0.2234,
+      "step": 1998
+    },
+    {
+      "epoch": 0.9465522343888725,
+      "grad_norm": 1.0825940370559692,
+      "learning_rate": 1.6017219877186173e-05,
+      "loss": 0.207,
+      "step": 1999
+    },
+    {
+      "epoch": 0.9470257472625037,
+      "grad_norm": 1.6445330381393433,
+      "learning_rate": 1.60131344288066e-05,
+      "loss": 0.2202,
+      "step": 2000
+    },
+    {
+      "epoch": 0.947499260136135,
+      "grad_norm": 1.1196541786193848,
+      "learning_rate": 1.600904740774973e-05,
+      "loss": 0.2332,
+      "step": 2001
+    },
+    {
+      "epoch": 0.9479727730097662,
+      "grad_norm": 1.173668622970581,
+      "learning_rate": 1.6004958815084476e-05,
+      "loss": 0.25,
+      "step": 2002
+    },
+    {
+      "epoch": 0.9484462858833974,
+      "grad_norm": 1.4854817390441895,
+      "learning_rate": 1.6000868651880175e-05,
+      "loss": 0.2444,
+      "step": 2003
+    },
+    {
+      "epoch": 0.9489197987570287,
+      "grad_norm": 1.8547112941741943,
+      "learning_rate": 1.599677691920657e-05,
+      "loss": 0.2711,
+      "step": 2004
+    },
+    {
+      "epoch": 0.94939331163066,
+      "grad_norm": 1.0332943201065063,
+      "learning_rate": 1.5992683618133817e-05,
+      "loss": 0.231,
+      "step": 2005
+    },
+    {
+      "epoch": 0.9498668245042912,
+      "grad_norm": 1.8548601865768433,
+      "learning_rate": 1.5988588749732472e-05,
+      "loss": 0.2198,
+      "step": 2006
+    },
+    {
+      "epoch": 0.9503403373779225,
+      "grad_norm": 1.8581714630126953,
+      "learning_rate": 1.5984492315073512e-05,
+      "loss": 0.2327,
+      "step": 2007
+    },
+    {
+      "epoch": 0.9508138502515537,
+      "grad_norm": 1.062828540802002,
+      "learning_rate": 1.5980394315228323e-05,
+      "loss": 0.224,
+      "step": 2008
+    },
+    {
+      "epoch": 0.9512873631251849,
+      "grad_norm": 1.4137210845947266,
+      "learning_rate": 1.5976294751268695e-05,
+      "loss": 0.2407,
+      "step": 2009
+    },
+    {
+      "epoch": 0.9517608759988162,
+      "grad_norm": 0.9517819285392761,
+      "learning_rate": 1.597219362426683e-05,
+      "loss": 0.2383,
+      "step": 2010
+    },
+    {
+      "epoch": 0.9522343888724475,
+      "grad_norm": 1.2080825567245483,
+      "learning_rate": 1.5968090935295335e-05,
+      "loss": 0.2174,
+      "step": 2011
+    },
+    {
+      "epoch": 0.9527079017460787,
+      "grad_norm": 1.1361576318740845,
+      "learning_rate": 1.596398668542723e-05,
+      "loss": 0.2225,
+      "step": 2012
+    },
+    {
+      "epoch": 0.95318141461971,
+      "grad_norm": 1.490893840789795,
+      "learning_rate": 1.5959880875735944e-05,
+      "loss": 0.2416,
+      "step": 2013
+    },
+    {
+      "epoch": 0.9536549274933412,
+      "grad_norm": 1.1593446731567383,
+      "learning_rate": 1.5955773507295313e-05,
+      "loss": 0.2363,
+      "step": 2014
+    },
+    {
+      "epoch": 0.9541284403669725,
+      "grad_norm": 1.0832358598709106,
+      "learning_rate": 1.5951664581179578e-05,
+      "loss": 0.2241,
+      "step": 2015
+    },
+    {
+      "epoch": 0.9546019532406037,
+      "grad_norm": 1.426303505897522,
+      "learning_rate": 1.5947554098463386e-05,
+      "loss": 0.2233,
+      "step": 2016
+    },
+    {
+      "epoch": 0.955075466114235,
+      "grad_norm": 1.7878800630569458,
+      "learning_rate": 1.5943442060221795e-05,
+      "loss": 0.2577,
+      "step": 2017
+    },
+    {
+      "epoch": 0.9555489789878663,
+      "grad_norm": 1.071764349937439,
+      "learning_rate": 1.5939328467530276e-05,
+      "loss": 0.2123,
+      "step": 2018
+    },
+    {
+      "epoch": 0.9560224918614975,
+      "grad_norm": 1.2568780183792114,
+      "learning_rate": 1.593521332146469e-05,
+      "loss": 0.2506,
+      "step": 2019
+    },
+    {
+      "epoch": 0.9564960047351287,
+      "grad_norm": 1.9318242073059082,
+      "learning_rate": 1.593109662310132e-05,
+      "loss": 0.2185,
+      "step": 2020
+    },
+    {
+      "epoch": 0.95696951760876,
+      "grad_norm": 1.3242648839950562,
+      "learning_rate": 1.5926978373516842e-05,
+      "loss": 0.2374,
+      "step": 2021
+    },
+    {
+      "epoch": 0.9574430304823912,
+      "grad_norm": 1.1390552520751953,
+      "learning_rate": 1.5922858573788356e-05,
+      "loss": 0.2206,
+      "step": 2022
+    },
+    {
+      "epoch": 0.9579165433560225,
+      "grad_norm": 1.5206451416015625,
+      "learning_rate": 1.5918737224993345e-05,
+      "loss": 0.2218,
+      "step": 2023
+    },
+    {
+      "epoch": 0.9583900562296538,
+      "grad_norm": 1.1097302436828613,
+      "learning_rate": 1.591461432820971e-05,
+      "loss": 0.1886,
+      "step": 2024
+    },
+    {
+      "epoch": 0.958863569103285,
+      "grad_norm": 1.2331650257110596,
+      "learning_rate": 1.591048988451576e-05,
+      "loss": 0.2343,
+      "step": 2025
+    },
+    {
+      "epoch": 0.9593370819769163,
+      "grad_norm": 2.753371000289917,
+      "learning_rate": 1.5906363894990197e-05,
+      "loss": 0.2412,
+      "step": 2026
+    },
+    {
+      "epoch": 0.9598105948505475,
+      "grad_norm": 0.9969640374183655,
+      "learning_rate": 1.590223636071214e-05,
+      "loss": 0.2389,
+      "step": 2027
+    },
+    {
+      "epoch": 0.9602841077241787,
+      "grad_norm": 2.2440433502197266,
+      "learning_rate": 1.58981072827611e-05,
+      "loss": 0.1999,
+      "step": 2028
+    },
+    {
+      "epoch": 0.96075762059781,
+      "grad_norm": 1.106385350227356,
+      "learning_rate": 1.5893976662217e-05,
+      "loss": 0.2168,
+      "step": 2029
+    },
+    {
+      "epoch": 0.9612311334714413,
+      "grad_norm": 2.1727712154388428,
+      "learning_rate": 1.588984450016017e-05,
+      "loss": 0.2357,
+      "step": 2030
+    },
+    {
+      "epoch": 0.9617046463450725,
+      "grad_norm": 1.639740228652954,
+      "learning_rate": 1.5885710797671326e-05,
+      "loss": 0.2533,
+      "step": 2031
+    },
+    {
+      "epoch": 0.9621781592187038,
+      "grad_norm": 0.9460485577583313,
+      "learning_rate": 1.5881575555831604e-05,
+      "loss": 0.2454,
+      "step": 2032
+    },
+    {
+      "epoch": 0.962651672092335,
+      "grad_norm": 1.232308268547058,
+      "learning_rate": 1.5877438775722536e-05,
+      "loss": 0.2304,
+      "step": 2033
+    },
+    {
+      "epoch": 0.9631251849659662,
+      "grad_norm": 1.2091732025146484,
+      "learning_rate": 1.587330045842606e-05,
+      "loss": 0.2175,
+      "step": 2034
+    },
+    {
+      "epoch": 0.9635986978395975,
+      "grad_norm": 1.2049965858459473,
+      "learning_rate": 1.586916060502451e-05,
+      "loss": 0.2437,
+      "step": 2035
+    },
+    {
+      "epoch": 0.9640722107132288,
+      "grad_norm": 1.1409215927124023,
+      "learning_rate": 1.586501921660062e-05,
+      "loss": 0.2253,
+      "step": 2036
+    },
+    {
+      "epoch": 0.96454572358686,
+      "grad_norm": 1.720682978630066,
+      "learning_rate": 1.5860876294237535e-05,
+      "loss": 0.2218,
+      "step": 2037
+    },
+    {
+      "epoch": 0.9650192364604913,
+      "grad_norm": 1.6841931343078613,
+      "learning_rate": 1.58567318390188e-05,
+      "loss": 0.2172,
+      "step": 2038
+    },
+    {
+      "epoch": 0.9654927493341225,
+      "grad_norm": 0.894889235496521,
+      "learning_rate": 1.5852585852028348e-05,
+      "loss": 0.232,
+      "step": 2039
+    },
+    {
+      "epoch": 0.9659662622077537,
+      "grad_norm": 1.6859184503555298,
+      "learning_rate": 1.584843833435053e-05,
+      "loss": 0.2161,
+      "step": 2040
+    },
+    {
+      "epoch": 0.966439775081385,
+      "grad_norm": 1.403887391090393,
+      "learning_rate": 1.5844289287070088e-05,
+      "loss": 0.2235,
+      "step": 2041
+    },
+    {
+      "epoch": 0.9669132879550163,
+      "grad_norm": 1.843876600265503,
+      "learning_rate": 1.5840138711272165e-05,
+      "loss": 0.201,
+      "step": 2042
+    },
+    {
+      "epoch": 0.9673868008286475,
+      "grad_norm": 0.9713191986083984,
+      "learning_rate": 1.58359866080423e-05,
+      "loss": 0.2385,
+      "step": 2043
+    },
+    {
+      "epoch": 0.9678603137022788,
+      "grad_norm": 1.014763593673706,
+      "learning_rate": 1.583183297846644e-05,
+      "loss": 0.231,
+      "step": 2044
+    },
+    {
+      "epoch": 0.96833382657591,
+      "grad_norm": 1.2346670627593994,
+      "learning_rate": 1.5827677823630922e-05,
+      "loss": 0.2111,
+      "step": 2045
+    },
+    {
+      "epoch": 0.9688073394495413,
+      "grad_norm": 1.3078184127807617,
+      "learning_rate": 1.5823521144622493e-05,
+      "loss": 0.2232,
+      "step": 2046
+    },
+    {
+      "epoch": 0.9692808523231725,
+      "grad_norm": 1.7039018869400024,
+      "learning_rate": 1.5819362942528288e-05,
+      "loss": 0.2271,
+      "step": 2047
+    },
+    {
+      "epoch": 0.9697543651968038,
+      "grad_norm": 2.7722675800323486,
+      "learning_rate": 1.5815203218435847e-05,
+      "loss": 0.1962,
+      "step": 2048
+    },
+    {
+      "epoch": 0.9702278780704351,
+      "grad_norm": 1.9730979204177856,
+      "learning_rate": 1.5811041973433103e-05,
+      "loss": 0.2382,
+      "step": 2049
+    },
+    {
+      "epoch": 0.9707013909440663,
+      "grad_norm": 0.8822180032730103,
+      "learning_rate": 1.580687920860839e-05,
+      "loss": 0.2026,
+      "step": 2050
+    },
+    {
+      "epoch": 0.9711749038176976,
+      "grad_norm": 1.793717861175537,
+      "learning_rate": 1.5802714925050444e-05,
+      "loss": 0.2262,
+      "step": 2051
+    },
+    {
+      "epoch": 0.9716484166913288,
+      "grad_norm": 1.0498217344284058,
+      "learning_rate": 1.5798549123848386e-05,
+      "loss": 0.2067,
+      "step": 2052
+    },
+    {
+      "epoch": 0.97212192956496,
+      "grad_norm": 1.1422361135482788,
+      "learning_rate": 1.5794381806091742e-05,
+      "loss": 0.2108,
+      "step": 2053
+    },
+    {
+      "epoch": 0.9725954424385913,
+      "grad_norm": 1.9571365118026733,
+      "learning_rate": 1.579021297287044e-05,
+      "loss": 0.2208,
+      "step": 2054
+    },
+    {
+      "epoch": 0.9730689553122226,
+      "grad_norm": 1.8578828573226929,
+      "learning_rate": 1.5786042625274795e-05,
+      "loss": 0.242,
+      "step": 2055
+    },
+    {
+      "epoch": 0.9735424681858538,
+      "grad_norm": 1.3278276920318604,
+      "learning_rate": 1.5781870764395515e-05,
+      "loss": 0.2039,
+      "step": 2056
+    },
+    {
+      "epoch": 0.9740159810594851,
+      "grad_norm": 1.0172450542449951,
+      "learning_rate": 1.5777697391323717e-05,
+      "loss": 0.2138,
+      "step": 2057
+    },
+    {
+      "epoch": 0.9744894939331163,
+      "grad_norm": 1.599645733833313,
+      "learning_rate": 1.577352250715091e-05,
+      "loss": 0.2476,
+      "step": 2058
+    },
+    {
+      "epoch": 0.9749630068067475,
+      "grad_norm": 1.6316616535186768,
+      "learning_rate": 1.5769346112968985e-05,
+      "loss": 0.2329,
+      "step": 2059
+    },
+    {
+      "epoch": 0.9754365196803788,
+      "grad_norm": 1.9370191097259521,
+      "learning_rate": 1.5765168209870243e-05,
+      "loss": 0.2382,
+      "step": 2060
+    },
+    {
+      "epoch": 0.9759100325540101,
+      "grad_norm": 1.4778871536254883,
+      "learning_rate": 1.5760988798947372e-05,
+      "loss": 0.2407,
+      "step": 2061
+    },
+    {
+      "epoch": 0.9763835454276413,
+      "grad_norm": 1.1680418252944946,
+      "learning_rate": 1.575680788129346e-05,
+      "loss": 0.2301,
+      "step": 2062
+    },
+    {
+      "epoch": 0.9768570583012726,
+      "grad_norm": 1.1255524158477783,
+      "learning_rate": 1.575262545800198e-05,
+      "loss": 0.2359,
+      "step": 2063
+    },
+    {
+      "epoch": 0.9773305711749039,
+      "grad_norm": 1.652750849723816,
+      "learning_rate": 1.5748441530166814e-05,
+      "loss": 0.2416,
+      "step": 2064
+    },
+    {
+      "epoch": 0.977804084048535,
+      "grad_norm": 1.3256607055664062,
+      "learning_rate": 1.5744256098882217e-05,
+      "loss": 0.2516,
+      "step": 2065
+    },
+    {
+      "epoch": 0.9782775969221663,
+      "grad_norm": 1.3550041913986206,
+      "learning_rate": 1.5740069165242854e-05,
+      "loss": 0.2473,
+      "step": 2066
+    },
+    {
+      "epoch": 0.9787511097957976,
+      "grad_norm": 1.5172563791275024,
+      "learning_rate": 1.5735880730343776e-05,
+      "loss": 0.2229,
+      "step": 2067
+    },
+    {
+      "epoch": 0.9792246226694288,
+      "grad_norm": 1.0142927169799805,
+      "learning_rate": 1.573169079528043e-05,
+      "loss": 0.2393,
+      "step": 2068
+    },
+    {
+      "epoch": 0.9796981355430601,
+      "grad_norm": 0.9681066274642944,
+      "learning_rate": 1.5727499361148647e-05,
+      "loss": 0.2281,
+      "step": 2069
+    },
+    {
+      "epoch": 0.9801716484166914,
+      "grad_norm": 1.673020362854004,
+      "learning_rate": 1.5723306429044663e-05,
+      "loss": 0.2345,
+      "step": 2070
+    },
+    {
+      "epoch": 0.9806451612903225,
+      "grad_norm": 1.8111623525619507,
+      "learning_rate": 1.57191120000651e-05,
+      "loss": 0.207,
+      "step": 2071
+    },
+    {
+      "epoch": 0.9811186741639538,
+      "grad_norm": 2.4038047790527344,
+      "learning_rate": 1.571491607530696e-05,
+      "loss": 0.222,
+      "step": 2072
+    },
+    {
+      "epoch": 0.9815921870375851,
+      "grad_norm": 2.7947847843170166,
+      "learning_rate": 1.5710718655867658e-05,
+      "loss": 0.2338,
+      "step": 2073
+    },
+    {
+      "epoch": 0.9820656999112163,
+      "grad_norm": 2.207833766937256,
+      "learning_rate": 1.5706519742844982e-05,
+      "loss": 0.2417,
+      "step": 2074
+    },
+    {
+      "epoch": 0.9825392127848476,
+      "grad_norm": 1.7318761348724365,
+      "learning_rate": 1.5702319337337118e-05,
+      "loss": 0.2323,
+      "step": 2075
+    },
+    {
+      "epoch": 0.9830127256584789,
+      "grad_norm": 1.349669098854065,
+      "learning_rate": 1.5698117440442643e-05,
+      "loss": 0.2377,
+      "step": 2076
+    },
+    {
+      "epoch": 0.9834862385321101,
+      "grad_norm": 1.8161946535110474,
+      "learning_rate": 1.5693914053260524e-05,
+      "loss": 0.2341,
+      "step": 2077
+    },
+    {
+      "epoch": 0.9839597514057413,
+      "grad_norm": 2.296698570251465,
+      "learning_rate": 1.5689709176890113e-05,
+      "loss": 0.226,
+      "step": 2078
+    },
+    {
+      "epoch": 0.9844332642793726,
+      "grad_norm": 1.6288074254989624,
+      "learning_rate": 1.5685502812431156e-05,
+      "loss": 0.2476,
+      "step": 2079
+    },
+    {
+      "epoch": 0.9849067771530039,
+      "grad_norm": 1.528737187385559,
+      "learning_rate": 1.5681294960983787e-05,
+      "loss": 0.2455,
+      "step": 2080
+    },
+    {
+      "epoch": 0.9853802900266351,
+      "grad_norm": 1.2330217361450195,
+      "learning_rate": 1.567708562364853e-05,
+      "loss": 0.2377,
+      "step": 2081
+    },
+    {
+      "epoch": 0.9858538029002664,
+      "grad_norm": 1.067199945449829,
+      "learning_rate": 1.56728748015263e-05,
+      "loss": 0.2244,
+      "step": 2082
+    },
+    {
+      "epoch": 0.9863273157738977,
+      "grad_norm": 1.0939373970031738,
+      "learning_rate": 1.566866249571839e-05,
+      "loss": 0.2275,
+      "step": 2083
+    },
+    {
+      "epoch": 0.9868008286475288,
+      "grad_norm": 1.1805354356765747,
+      "learning_rate": 1.566444870732649e-05,
+      "loss": 0.2211,
+      "step": 2084
+    },
+    {
+      "epoch": 0.9872743415211601,
+      "grad_norm": 2.558940887451172,
+      "learning_rate": 1.5660233437452676e-05,
+      "loss": 0.2368,
+      "step": 2085
+    },
+    {
+      "epoch": 0.9877478543947914,
+      "grad_norm": 1.4615000486373901,
+      "learning_rate": 1.565601668719941e-05,
+      "loss": 0.2194,
+      "step": 2086
+    },
+    {
+      "epoch": 0.9882213672684226,
+      "grad_norm": 1.341464638710022,
+      "learning_rate": 1.565179845766955e-05,
+      "loss": 0.2447,
+      "step": 2087
+    },
+    {
+      "epoch": 0.9886948801420539,
+      "grad_norm": 1.341123104095459,
+      "learning_rate": 1.564757874996632e-05,
+      "loss": 0.2241,
+      "step": 2088
+    },
+    {
+      "epoch": 0.9891683930156852,
+      "grad_norm": 0.9674537181854248,
+      "learning_rate": 1.5643357565193355e-05,
+      "loss": 0.2312,
+      "step": 2089
+    },
+    {
+      "epoch": 0.9896419058893163,
+      "grad_norm": 1.1203300952911377,
+      "learning_rate": 1.5639134904454663e-05,
+      "loss": 0.2513,
+      "step": 2090
+    },
+    {
+      "epoch": 0.9901154187629476,
+      "grad_norm": 1.7176257371902466,
+      "learning_rate": 1.5634910768854634e-05,
+      "loss": 0.2076,
+      "step": 2091
+    },
+    {
+      "epoch": 0.9905889316365789,
+      "grad_norm": 2.126793146133423,
+      "learning_rate": 1.5630685159498057e-05,
+      "loss": 0.2445,
+      "step": 2092
+    },
+    {
+      "epoch": 0.9910624445102101,
+      "grad_norm": 1.8809027671813965,
+      "learning_rate": 1.56264580774901e-05,
+      "loss": 0.2538,
+      "step": 2093
+    },
+    {
+      "epoch": 0.9915359573838414,
+      "grad_norm": 1.7825748920440674,
+      "learning_rate": 1.562222952393631e-05,
+      "loss": 0.2337,
+      "step": 2094
+    },
+    {
+      "epoch": 0.9920094702574727,
+      "grad_norm": 1.1806340217590332,
+      "learning_rate": 1.5617999499942623e-05,
+      "loss": 0.2319,
+      "step": 2095
+    },
+    {
+      "epoch": 0.9924829831311038,
+      "grad_norm": 1.4441994428634644,
+      "learning_rate": 1.5613768006615367e-05,
+      "loss": 0.2383,
+      "step": 2096
+    },
+    {
+      "epoch": 0.9929564960047351,
+      "grad_norm": 0.8531612157821655,
+      "learning_rate": 1.5609535045061247e-05,
+      "loss": 0.2205,
+      "step": 2097
+    },
+    {
+      "epoch": 0.9934300088783664,
+      "grad_norm": 1.457108497619629,
+      "learning_rate": 1.5605300616387347e-05,
+      "loss": 0.2266,
+      "step": 2098
+    },
+    {
+      "epoch": 0.9939035217519976,
+      "grad_norm": 1.050585389137268,
+      "learning_rate": 1.5601064721701155e-05,
+      "loss": 0.2373,
+      "step": 2099
+    },
+    {
+      "epoch": 0.9943770346256289,
+      "grad_norm": 0.9880205392837524,
+      "learning_rate": 1.5596827362110512e-05,
+      "loss": 0.2344,
+      "step": 2100
+    },
+    {
+      "epoch": 0.9948505474992602,
+      "grad_norm": 1.1303571462631226,
+      "learning_rate": 1.559258853872367e-05,
+      "loss": 0.2044,
+      "step": 2101
+    },
+    {
+      "epoch": 0.9953240603728913,
+      "grad_norm": 1.6071451902389526,
+      "learning_rate": 1.5588348252649246e-05,
+      "loss": 0.2096,
+      "step": 2102
+    },
+    {
+      "epoch": 0.9957975732465226,
+      "grad_norm": 1.32492995262146,
+      "learning_rate": 1.5584106504996247e-05,
+      "loss": 0.2312,
+      "step": 2103
+    },
+    {
+      "epoch": 0.9962710861201539,
+      "grad_norm": 1.9682942628860474,
+      "learning_rate": 1.5579863296874066e-05,
+      "loss": 0.2378,
+      "step": 2104
+    },
+    {
+      "epoch": 0.9967445989937851,
+      "grad_norm": 1.4948111772537231,
+      "learning_rate": 1.5575618629392466e-05,
+      "loss": 0.2195,
+      "step": 2105
+    },
+    {
+      "epoch": 0.9972181118674164,
+      "grad_norm": 1.1894011497497559,
+      "learning_rate": 1.5571372503661604e-05,
+      "loss": 0.2456,
+      "step": 2106
+    },
+    {
+      "epoch": 0.9976916247410477,
+      "grad_norm": 1.0907820463180542,
+      "learning_rate": 1.556712492079201e-05,
+      "loss": 0.2328,
+      "step": 2107
+    },
+    {
+      "epoch": 0.998165137614679,
+      "grad_norm": 1.6044102907180786,
+      "learning_rate": 1.5562875881894605e-05,
+      "loss": 0.2159,
+      "step": 2108
+    },
+    {
+      "epoch": 0.9986386504883101,
+      "grad_norm": 2.430710554122925,
+      "learning_rate": 1.5558625388080676e-05,
+      "loss": 0.2135,
+      "step": 2109
+    },
+    {
+      "epoch": 0.9991121633619414,
+      "grad_norm": 1.4718191623687744,
+      "learning_rate": 1.5554373440461904e-05,
+      "loss": 0.2448,
+      "step": 2110
+    },
+    {
+      "epoch": 0.9995856762355727,
+      "grad_norm": 1.1206008195877075,
+      "learning_rate": 1.5550120040150338e-05,
+      "loss": 0.2055,
+      "step": 2111
+    },
+    {
+      "epoch": 1.0000591891092039,
+      "grad_norm": 1.4358619451522827,
+      "learning_rate": 1.5545865188258423e-05,
+      "loss": 0.2423,
+      "step": 2112
+    },
+    {
+      "epoch": 1.0005327019828352,
+      "grad_norm": 1.060018539428711,
+      "learning_rate": 1.5541608885898968e-05,
+      "loss": 0.2124,
+      "step": 2113
+    },
+    {
+      "epoch": 1.0010062148564665,
+      "grad_norm": 1.6677261590957642,
+      "learning_rate": 1.553735113418517e-05,
+      "loss": 0.2323,
+      "step": 2114
+    },
+    {
+      "epoch": 1.0014797277300977,
+      "grad_norm": 1.8145389556884766,
+      "learning_rate": 1.55330919342306e-05,
+      "loss": 0.2383,
+      "step": 2115
+    },
+    {
+      "epoch": 1.0019532406037288,
+      "grad_norm": 1.2736999988555908,
+      "learning_rate": 1.552883128714922e-05,
+      "loss": 0.2094,
+      "step": 2116
+    },
+    {
+      "epoch": 1.00242675347736,
+      "grad_norm": 1.4934360980987549,
+      "learning_rate": 1.552456919405535e-05,
+      "loss": 0.2103,
+      "step": 2117
+    },
+    {
+      "epoch": 1.0029002663509914,
+      "grad_norm": 1.330871343612671,
+      "learning_rate": 1.5520305656063702e-05,
+      "loss": 0.2176,
+      "step": 2118
+    },
+    {
+      "epoch": 1.0033737792246227,
+      "grad_norm": 1.226676344871521,
+      "learning_rate": 1.5516040674289364e-05,
+      "loss": 0.2181,
+      "step": 2119
+    },
+    {
+      "epoch": 1.003847292098254,
+      "grad_norm": 0.9475164413452148,
+      "learning_rate": 1.5511774249847806e-05,
+      "loss": 0.2092,
+      "step": 2120
+    },
+    {
+      "epoch": 1.0043208049718853,
+      "grad_norm": 2.017439126968384,
+      "learning_rate": 1.5507506383854867e-05,
+      "loss": 0.2404,
+      "step": 2121
+    },
+    {
+      "epoch": 1.0047943178455163,
+      "grad_norm": 2.2948458194732666,
+      "learning_rate": 1.5503237077426762e-05,
+      "loss": 0.2225,
+      "step": 2122
+    },
+    {
+      "epoch": 1.0052678307191476,
+      "grad_norm": 1.144311547279358,
+      "learning_rate": 1.5498966331680093e-05,
+      "loss": 0.2437,
+      "step": 2123
+    },
+    {
+      "epoch": 1.005741343592779,
+      "grad_norm": 1.512420892715454,
+      "learning_rate": 1.5494694147731822e-05,
+      "loss": 0.2337,
+      "step": 2124
+    },
+    {
+      "epoch": 1.0062148564664102,
+      "grad_norm": 1.3547443151474,
+      "learning_rate": 1.549042052669931e-05,
+      "loss": 0.2101,
+      "step": 2125
+    },
+    {
+      "epoch": 1.0066883693400415,
+      "grad_norm": 1.429449200630188,
+      "learning_rate": 1.5486145469700278e-05,
+      "loss": 0.2054,
+      "step": 2126
+    },
+    {
+      "epoch": 1.0071618822136728,
+      "grad_norm": 2.169478178024292,
+      "learning_rate": 1.5481868977852823e-05,
+      "loss": 0.2465,
+      "step": 2127
+    },
+    {
+      "epoch": 1.007635395087304,
+      "grad_norm": 1.4005268812179565,
+      "learning_rate": 1.547759105227542e-05,
+      "loss": 0.2212,
+      "step": 2128
+    },
+    {
+      "epoch": 1.008108907960935,
+      "grad_norm": 1.521437644958496,
+      "learning_rate": 1.547331169408692e-05,
+      "loss": 0.2458,
+      "step": 2129
+    },
+    {
+      "epoch": 1.0085824208345664,
+      "grad_norm": 1.0772839784622192,
+      "learning_rate": 1.5469030904406554e-05,
+      "loss": 0.2265,
+      "step": 2130
+    },
+    {
+      "epoch": 1.0090559337081977,
+      "grad_norm": 1.4369512796401978,
+      "learning_rate": 1.546474868435391e-05,
+      "loss": 0.2037,
+      "step": 2131
+    },
+    {
+      "epoch": 1.009529446581829,
+      "grad_norm": 1.2946951389312744,
+      "learning_rate": 1.546046503504897e-05,
+      "loss": 0.2261,
+      "step": 2132
+    },
+    {
+      "epoch": 1.0100029594554603,
+      "grad_norm": 1.3359379768371582,
+      "learning_rate": 1.5456179957612074e-05,
+      "loss": 0.2487,
+      "step": 2133
+    },
+    {
+      "epoch": 1.0104764723290915,
+      "grad_norm": 2.1459126472473145,
+      "learning_rate": 1.545189345316395e-05,
+      "loss": 0.2004,
+      "step": 2134
+    },
+    {
+      "epoch": 1.0109499852027226,
+      "grad_norm": 2.0224199295043945,
+      "learning_rate": 1.5447605522825687e-05,
+      "loss": 0.2237,
+      "step": 2135
+    },
+    {
+      "epoch": 1.011423498076354,
+      "grad_norm": 1.4307156801223755,
+      "learning_rate": 1.5443316167718756e-05,
+      "loss": 0.2076,
+      "step": 2136
+    },
+    {
+      "epoch": 1.0118970109499852,
+      "grad_norm": 1.0832078456878662,
+      "learning_rate": 1.543902538896499e-05,
+      "loss": 0.2175,
+      "step": 2137
+    },
+    {
+      "epoch": 1.0123705238236165,
+      "grad_norm": 1.5245671272277832,
+      "learning_rate": 1.543473318768661e-05,
+      "loss": 0.245,
+      "step": 2138
+    },
+    {
+      "epoch": 1.0128440366972478,
+      "grad_norm": 1.020795226097107,
+      "learning_rate": 1.5430439565006193e-05,
+      "loss": 0.2282,
+      "step": 2139
+    },
+    {
+      "epoch": 1.013317549570879,
+      "grad_norm": 2.5934853553771973,
+      "learning_rate": 1.5426144522046692e-05,
+      "loss": 0.2264,
+      "step": 2140
+    },
+    {
+      "epoch": 1.0137910624445101,
+      "grad_norm": 1.0803589820861816,
+      "learning_rate": 1.5421848059931443e-05,
+      "loss": 0.2158,
+      "step": 2141
+    },
+    {
+      "epoch": 1.0142645753181414,
+      "grad_norm": 1.115559697151184,
+      "learning_rate": 1.541755017978414e-05,
+      "loss": 0.2259,
+      "step": 2142
+    },
+    {
+      "epoch": 1.0147380881917727,
+      "grad_norm": 1.1411402225494385,
+      "learning_rate": 1.5413250882728847e-05,
+      "loss": 0.2196,
+      "step": 2143
+    },
+    {
+      "epoch": 1.015211601065404,
+      "grad_norm": 1.0891308784484863,
+      "learning_rate": 1.540895016989001e-05,
+      "loss": 0.2257,
+      "step": 2144
+    },
+    {
+      "epoch": 1.0156851139390353,
+      "grad_norm": 1.2236292362213135,
+      "learning_rate": 1.5404648042392437e-05,
+      "loss": 0.2483,
+      "step": 2145
+    },
+    {
+      "epoch": 1.0161586268126666,
+      "grad_norm": 1.3101778030395508,
+      "learning_rate": 1.5400344501361305e-05,
+      "loss": 0.2389,
+      "step": 2146
+    },
+    {
+      "epoch": 1.0166321396862976,
+      "grad_norm": 1.1559494733810425,
+      "learning_rate": 1.5396039547922174e-05,
+      "loss": 0.2229,
+      "step": 2147
+    },
+    {
+      "epoch": 1.017105652559929,
+      "grad_norm": 1.1943864822387695,
+      "learning_rate": 1.5391733183200952e-05,
+      "loss": 0.2181,
+      "step": 2148
+    },
+    {
+      "epoch": 1.0175791654335602,
+      "grad_norm": 1.2304420471191406,
+      "learning_rate": 1.5387425408323934e-05,
+      "loss": 0.2105,
+      "step": 2149
+    },
+    {
+      "epoch": 1.0180526783071915,
+      "grad_norm": 1.0440901517868042,
+      "learning_rate": 1.5383116224417767e-05,
+      "loss": 0.2332,
+      "step": 2150
+    },
+    {
+      "epoch": 1.0185261911808228,
+      "grad_norm": 1.2044522762298584,
+      "learning_rate": 1.5378805632609487e-05,
+      "loss": 0.2375,
+      "step": 2151
+    },
+    {
+      "epoch": 1.018999704054454,
+      "grad_norm": 1.4749743938446045,
+      "learning_rate": 1.5374493634026486e-05,
+      "loss": 0.1913,
+      "step": 2152
+    },
+    {
+      "epoch": 1.0194732169280851,
+      "grad_norm": 1.4272947311401367,
+      "learning_rate": 1.537018022979652e-05,
+      "loss": 0.22,
+      "step": 2153
+    },
+    {
+      "epoch": 1.0199467298017164,
+      "grad_norm": 1.7914042472839355,
+      "learning_rate": 1.5365865421047724e-05,
+      "loss": 0.2344,
+      "step": 2154
+    },
+    {
+      "epoch": 1.0204202426753477,
+      "grad_norm": 1.1792020797729492,
+      "learning_rate": 1.5361549208908594e-05,
+      "loss": 0.2023,
+      "step": 2155
+    },
+    {
+      "epoch": 1.020893755548979,
+      "grad_norm": 1.5299314260482788,
+      "learning_rate": 1.5357231594507988e-05,
+      "loss": 0.2242,
+      "step": 2156
+    },
+    {
+      "epoch": 1.0213672684226103,
+      "grad_norm": 1.1294841766357422,
+      "learning_rate": 1.5352912578975144e-05,
+      "loss": 0.2355,
+      "step": 2157
+    },
+    {
+      "epoch": 1.0218407812962416,
+      "grad_norm": 1.5787099599838257,
+      "learning_rate": 1.5348592163439655e-05,
+      "loss": 0.2163,
+      "step": 2158
+    },
+    {
+      "epoch": 1.0223142941698729,
+      "grad_norm": 1.2531263828277588,
+      "learning_rate": 1.5344270349031486e-05,
+      "loss": 0.2435,
+      "step": 2159
+    },
+    {
+      "epoch": 1.022787807043504,
+      "grad_norm": 1.2263544797897339,
+      "learning_rate": 1.5339947136880962e-05,
+      "loss": 0.213,
+      "step": 2160
+    },
+    {
+      "epoch": 1.0232613199171352,
+      "grad_norm": 1.8795762062072754,
+      "learning_rate": 1.5335622528118777e-05,
+      "loss": 0.2309,
+      "step": 2161
+    },
+    {
+      "epoch": 1.0237348327907665,
+      "grad_norm": 1.558018684387207,
+      "learning_rate": 1.5331296523876e-05,
+      "loss": 0.231,
+      "step": 2162
+    },
+    {
+      "epoch": 1.0242083456643978,
+      "grad_norm": 1.8025420904159546,
+      "learning_rate": 1.5326969125284043e-05,
+      "loss": 0.202,
+      "step": 2163
+    },
+    {
+      "epoch": 1.024681858538029,
+      "grad_norm": 1.190397024154663,
+      "learning_rate": 1.5322640333474704e-05,
+      "loss": 0.2224,
+      "step": 2164
+    },
+    {
+      "epoch": 1.0251553714116604,
+      "grad_norm": 1.9141662120819092,
+      "learning_rate": 1.5318310149580133e-05,
+      "loss": 0.2584,
+      "step": 2165
+    },
+    {
+      "epoch": 1.0256288842852914,
+      "grad_norm": 1.1253691911697388,
+      "learning_rate": 1.531397857473285e-05,
+      "loss": 0.2169,
+      "step": 2166
+    },
+    {
+      "epoch": 1.0261023971589227,
+      "grad_norm": 1.264083981513977,
+      "learning_rate": 1.530964561006574e-05,
+      "loss": 0.2143,
+      "step": 2167
+    },
+    {
+      "epoch": 1.026575910032554,
+      "grad_norm": 1.5587284564971924,
+      "learning_rate": 1.5305311256712038e-05,
+      "loss": 0.2312,
+      "step": 2168
+    },
+    {
+      "epoch": 1.0270494229061853,
+      "grad_norm": 1.0813679695129395,
+      "learning_rate": 1.5300975515805358e-05,
+      "loss": 0.2241,
+      "step": 2169
+    },
+    {
+      "epoch": 1.0275229357798166,
+      "grad_norm": 1.4243192672729492,
+      "learning_rate": 1.5296638388479673e-05,
+      "loss": 0.2121,
+      "step": 2170
+    },
+    {
+      "epoch": 1.0279964486534479,
+      "grad_norm": 1.1746392250061035,
+      "learning_rate": 1.5292299875869313e-05,
+      "loss": 0.238,
+      "step": 2171
+    },
+    {
+      "epoch": 1.028469961527079,
+      "grad_norm": 1.360106110572815,
+      "learning_rate": 1.528795997910898e-05,
+      "loss": 0.2336,
+      "step": 2172
+    },
+    {
+      "epoch": 1.0289434744007102,
+      "grad_norm": 1.4826574325561523,
+      "learning_rate": 1.5283618699333725e-05,
+      "loss": 0.2278,
+      "step": 2173
+    },
+    {
+      "epoch": 1.0294169872743415,
+      "grad_norm": 1.1893174648284912,
+      "learning_rate": 1.5279276037678972e-05,
+      "loss": 0.2175,
+      "step": 2174
+    },
+    {
+      "epoch": 1.0298905001479728,
+      "grad_norm": 1.326848030090332,
+      "learning_rate": 1.5274931995280504e-05,
+      "loss": 0.2246,
+      "step": 2175
+    },
+    {
+      "epoch": 1.030364013021604,
+      "grad_norm": 1.2376220226287842,
+      "learning_rate": 1.527058657327446e-05,
+      "loss": 0.2405,
+      "step": 2176
+    },
+    {
+      "epoch": 1.0308375258952354,
+      "grad_norm": 0.9731859564781189,
+      "learning_rate": 1.5266239772797343e-05,
+      "loss": 0.2302,
+      "step": 2177
+    },
+    {
+      "epoch": 1.0313110387688664,
+      "grad_norm": 1.755089521408081,
+      "learning_rate": 1.5261891594986014e-05,
+      "loss": 0.1949,
+      "step": 2178
+    },
+    {
+      "epoch": 1.0317845516424977,
+      "grad_norm": 2.296330451965332,
+      "learning_rate": 1.5257542040977706e-05,
+      "loss": 0.2192,
+      "step": 2179
+    },
+    {
+      "epoch": 1.032258064516129,
+      "grad_norm": 2.243488311767578,
+      "learning_rate": 1.5253191111909997e-05,
+      "loss": 0.2227,
+      "step": 2180
+    },
+    {
+      "epoch": 1.0327315773897603,
+      "grad_norm": 1.5681140422821045,
+      "learning_rate": 1.524883880892083e-05,
+      "loss": 0.2159,
+      "step": 2181
+    },
+    {
+      "epoch": 1.0332050902633916,
+      "grad_norm": 1.9723291397094727,
+      "learning_rate": 1.5244485133148507e-05,
+      "loss": 0.232,
+      "step": 2182
+    },
+    {
+      "epoch": 1.0336786031370229,
+      "grad_norm": 1.862541913986206,
+      "learning_rate": 1.5240130085731695e-05,
+      "loss": 0.2463,
+      "step": 2183
+    },
+    {
+      "epoch": 1.034152116010654,
+      "grad_norm": 1.2561465501785278,
+      "learning_rate": 1.5235773667809409e-05,
+      "loss": 0.2121,
+      "step": 2184
+    },
+    {
+      "epoch": 1.0346256288842852,
+      "grad_norm": 1.7062082290649414,
+      "learning_rate": 1.5231415880521032e-05,
+      "loss": 0.2215,
+      "step": 2185
+    },
+    {
+      "epoch": 1.0350991417579165,
+      "grad_norm": 1.24093496799469,
+      "learning_rate": 1.5227056725006301e-05,
+      "loss": 0.2052,
+      "step": 2186
+    },
+    {
+      "epoch": 1.0355726546315478,
+      "grad_norm": 1.7129935026168823,
+      "learning_rate": 1.5222696202405307e-05,
+      "loss": 0.1988,
+      "step": 2187
+    },
+    {
+      "epoch": 1.036046167505179,
+      "grad_norm": 1.3242411613464355,
+      "learning_rate": 1.5218334313858507e-05,
+      "loss": 0.2089,
+      "step": 2188
+    },
+    {
+      "epoch": 1.0365196803788104,
+      "grad_norm": 2.027369737625122,
+      "learning_rate": 1.5213971060506709e-05,
+      "loss": 0.2572,
+      "step": 2189
+    },
+    {
+      "epoch": 1.0369931932524414,
+      "grad_norm": 1.3638339042663574,
+      "learning_rate": 1.520960644349108e-05,
+      "loss": 0.2261,
+      "step": 2190
+    },
+    {
+      "epoch": 1.0374667061260727,
+      "grad_norm": 1.4748543500900269,
+      "learning_rate": 1.5205240463953146e-05,
+      "loss": 0.2149,
+      "step": 2191
+    },
+    {
+      "epoch": 1.037940218999704,
+      "grad_norm": 1.1796306371688843,
+      "learning_rate": 1.5200873123034783e-05,
+      "loss": 0.237,
+      "step": 2192
+    },
+    {
+      "epoch": 1.0384137318733353,
+      "grad_norm": 1.1525838375091553,
+      "learning_rate": 1.5196504421878229e-05,
+      "loss": 0.2385,
+      "step": 2193
+    },
+    {
+      "epoch": 1.0388872447469666,
+      "grad_norm": 2.3929810523986816,
+      "learning_rate": 1.5192134361626074e-05,
+      "loss": 0.2005,
+      "step": 2194
+    },
+    {
+      "epoch": 1.0393607576205979,
+      "grad_norm": 1.697789192199707,
+      "learning_rate": 1.5187762943421266e-05,
+      "loss": 0.2422,
+      "step": 2195
+    },
+    {
+      "epoch": 1.0398342704942292,
+      "grad_norm": 1.2148844003677368,
+      "learning_rate": 1.5183390168407108e-05,
+      "loss": 0.2188,
+      "step": 2196
+    },
+    {
+      "epoch": 1.0403077833678602,
+      "grad_norm": 1.2797152996063232,
+      "learning_rate": 1.5179016037727256e-05,
+      "loss": 0.215,
+      "step": 2197
+    },
+    {
+      "epoch": 1.0407812962414915,
+      "grad_norm": 1.7191182374954224,
+      "learning_rate": 1.5174640552525724e-05,
+      "loss": 0.2422,
+      "step": 2198
+    },
+    {
+      "epoch": 1.0412548091151228,
+      "grad_norm": 1.3166422843933105,
+      "learning_rate": 1.5170263713946873e-05,
+      "loss": 0.22,
+      "step": 2199
+    },
+    {
+      "epoch": 1.041728321988754,
+      "grad_norm": 1.1417112350463867,
+      "learning_rate": 1.5165885523135426e-05,
+      "loss": 0.2137,
+      "step": 2200
+    },
+    {
+      "epoch": 1.0422018348623854,
+      "grad_norm": 1.1441370248794556,
+      "learning_rate": 1.516150598123646e-05,
+      "loss": 0.2146,
+      "step": 2201
+    },
+    {
+      "epoch": 1.0426753477360167,
+      "grad_norm": 1.5537104606628418,
+      "learning_rate": 1.5157125089395397e-05,
+      "loss": 0.21,
+      "step": 2202
+    },
+    {
+      "epoch": 1.0431488606096477,
+      "grad_norm": 1.3530882596969604,
+      "learning_rate": 1.5152742848758018e-05,
+      "loss": 0.2085,
+      "step": 2203
+    },
+    {
+      "epoch": 1.043622373483279,
+      "grad_norm": 1.541207194328308,
+      "learning_rate": 1.5148359260470456e-05,
+      "loss": 0.2315,
+      "step": 2204
+    },
+    {
+      "epoch": 1.0440958863569103,
+      "grad_norm": 1.0132625102996826,
+      "learning_rate": 1.5143974325679196e-05,
+      "loss": 0.2008,
+      "step": 2205
+    },
+    {
+      "epoch": 1.0445693992305416,
+      "grad_norm": 1.1668498516082764,
+      "learning_rate": 1.5139588045531077e-05,
+      "loss": 0.2239,
+      "step": 2206
+    },
+    {
+      "epoch": 1.0450429121041729,
+      "grad_norm": 1.6192450523376465,
+      "learning_rate": 1.5135200421173288e-05,
+      "loss": 0.2306,
+      "step": 2207
+    },
+    {
+      "epoch": 1.0455164249778042,
+      "grad_norm": 1.6233320236206055,
+      "learning_rate": 1.5130811453753369e-05,
+      "loss": 0.2243,
+      "step": 2208
+    },
+    {
+      "epoch": 1.0459899378514352,
+      "grad_norm": 1.1391911506652832,
+      "learning_rate": 1.512642114441921e-05,
+      "loss": 0.1966,
+      "step": 2209
+    },
+    {
+      "epoch": 1.0464634507250665,
+      "grad_norm": 1.7349532842636108,
+      "learning_rate": 1.512202949431906e-05,
+      "loss": 0.2256,
+      "step": 2210
+    },
+    {
+      "epoch": 1.0469369635986978,
+      "grad_norm": 1.411200761795044,
+      "learning_rate": 1.5117636504601505e-05,
+      "loss": 0.2225,
+      "step": 2211
+    },
+    {
+      "epoch": 1.047410476472329,
+      "grad_norm": 1.5413966178894043,
+      "learning_rate": 1.5113242176415495e-05,
+      "loss": 0.2301,
+      "step": 2212
+    },
+    {
+      "epoch": 1.0478839893459604,
+      "grad_norm": 1.4936213493347168,
+      "learning_rate": 1.5108846510910322e-05,
+      "loss": 0.2299,
+      "step": 2213
+    },
+    {
+      "epoch": 1.0483575022195917,
+      "grad_norm": 2.126307964324951,
+      "learning_rate": 1.5104449509235628e-05,
+      "loss": 0.2195,
+      "step": 2214
+    },
+    {
+      "epoch": 1.0488310150932227,
+      "grad_norm": 1.6920348405838013,
+      "learning_rate": 1.5100051172541408e-05,
+      "loss": 0.2081,
+      "step": 2215
+    },
+    {
+      "epoch": 1.049304527966854,
+      "grad_norm": 1.1273527145385742,
+      "learning_rate": 1.5095651501978007e-05,
+      "loss": 0.2157,
+      "step": 2216
+    },
+    {
+      "epoch": 1.0497780408404853,
+      "grad_norm": 0.9604322910308838,
+      "learning_rate": 1.5091250498696113e-05,
+      "loss": 0.2404,
+      "step": 2217
+    },
+    {
+      "epoch": 1.0502515537141166,
+      "grad_norm": 1.639671802520752,
+      "learning_rate": 1.508684816384677e-05,
+      "loss": 0.2384,
+      "step": 2218
+    },
+    {
+      "epoch": 1.050725066587748,
+      "grad_norm": 1.559471845626831,
+      "learning_rate": 1.5082444498581362e-05,
+      "loss": 0.2078,
+      "step": 2219
+    },
+    {
+      "epoch": 1.0511985794613792,
+      "grad_norm": 1.1831077337265015,
+      "learning_rate": 1.5078039504051626e-05,
+      "loss": 0.2396,
+      "step": 2220
+    },
+    {
+      "epoch": 1.0516720923350102,
+      "grad_norm": 1.6754920482635498,
+      "learning_rate": 1.5073633181409645e-05,
+      "loss": 0.2161,
+      "step": 2221
+    },
+    {
+      "epoch": 1.0521456052086415,
+      "grad_norm": 1.6285377740859985,
+      "learning_rate": 1.5069225531807852e-05,
+      "loss": 0.2252,
+      "step": 2222
+    },
+    {
+      "epoch": 1.0526191180822728,
+      "grad_norm": 0.9481242895126343,
+      "learning_rate": 1.5064816556399027e-05,
+      "loss": 0.2061,
+      "step": 2223
+    },
+    {
+      "epoch": 1.053092630955904,
+      "grad_norm": 1.5364612340927124,
+      "learning_rate": 1.506040625633629e-05,
+      "loss": 0.229,
+      "step": 2224
+    },
+    {
+      "epoch": 1.0535661438295354,
+      "grad_norm": 1.2100945711135864,
+      "learning_rate": 1.5055994632773119e-05,
+      "loss": 0.2278,
+      "step": 2225
+    },
+    {
+      "epoch": 1.0540396567031667,
+      "grad_norm": 1.6874003410339355,
+      "learning_rate": 1.5051581686863323e-05,
+      "loss": 0.2117,
+      "step": 2226
+    },
+    {
+      "epoch": 1.054513169576798,
+      "grad_norm": 1.3330941200256348,
+      "learning_rate": 1.5047167419761075e-05,
+      "loss": 0.2267,
+      "step": 2227
+    },
+    {
+      "epoch": 1.054986682450429,
+      "grad_norm": 1.1632763147354126,
+      "learning_rate": 1.5042751832620879e-05,
+      "loss": 0.234,
+      "step": 2228
+    },
+    {
+      "epoch": 1.0554601953240603,
+      "grad_norm": 1.2275227308273315,
+      "learning_rate": 1.5038334926597587e-05,
+      "loss": 0.2079,
+      "step": 2229
+    },
+    {
+      "epoch": 1.0559337081976916,
+      "grad_norm": 1.616982102394104,
+      "learning_rate": 1.50339167028464e-05,
+      "loss": 0.2224,
+      "step": 2230
+    },
+    {
+      "epoch": 1.056407221071323,
+      "grad_norm": 1.067596197128296,
+      "learning_rate": 1.5029497162522865e-05,
+      "loss": 0.229,
+      "step": 2231
+    },
+    {
+      "epoch": 1.0568807339449542,
+      "grad_norm": 1.3072737455368042,
+      "learning_rate": 1.5025076306782866e-05,
+      "loss": 0.205,
+      "step": 2232
+    },
+    {
+      "epoch": 1.0573542468185855,
+      "grad_norm": 1.5054287910461426,
+      "learning_rate": 1.5020654136782637e-05,
+      "loss": 0.2159,
+      "step": 2233
+    },
+    {
+      "epoch": 1.0578277596922165,
+      "grad_norm": 1.238438606262207,
+      "learning_rate": 1.5016230653678757e-05,
+      "loss": 0.2145,
+      "step": 2234
+    },
+    {
+      "epoch": 1.0583012725658478,
+      "grad_norm": 1.200659155845642,
+      "learning_rate": 1.5011805858628137e-05,
+      "loss": 0.26,
+      "step": 2235
+    },
+    {
+      "epoch": 1.0587747854394791,
+      "grad_norm": 1.4251799583435059,
+      "learning_rate": 1.5007379752788045e-05,
+      "loss": 0.2004,
+      "step": 2236
+    },
+    {
+      "epoch": 1.0592482983131104,
+      "grad_norm": 1.2545171976089478,
+      "learning_rate": 1.5002952337316088e-05,
+      "loss": 0.2127,
+      "step": 2237
+    },
+    {
+      "epoch": 1.0597218111867417,
+      "grad_norm": 1.177280306816101,
+      "learning_rate": 1.499852361337021e-05,
+      "loss": 0.212,
+      "step": 2238
+    },
+    {
+      "epoch": 1.060195324060373,
+      "grad_norm": 1.241126537322998,
+      "learning_rate": 1.4994093582108704e-05,
+      "loss": 0.2193,
+      "step": 2239
+    },
+    {
+      "epoch": 1.060668836934004,
+      "grad_norm": 1.5119160413742065,
+      "learning_rate": 1.49896622446902e-05,
+      "loss": 0.2198,
+      "step": 2240
+    },
+    {
+      "epoch": 1.0611423498076353,
+      "grad_norm": 1.6282936334609985,
+      "learning_rate": 1.498522960227367e-05,
+      "loss": 0.2185,
+      "step": 2241
+    },
+    {
+      "epoch": 1.0616158626812666,
+      "grad_norm": 0.9974349737167358,
+      "learning_rate": 1.4980795656018432e-05,
+      "loss": 0.2323,
+      "step": 2242
+    },
+    {
+      "epoch": 1.062089375554898,
+      "grad_norm": 1.2673429250717163,
+      "learning_rate": 1.4976360407084141e-05,
+      "loss": 0.2032,
+      "step": 2243
+    },
+    {
+      "epoch": 1.0625628884285292,
+      "grad_norm": 1.0515581369400024,
+      "learning_rate": 1.4971923856630792e-05,
+      "loss": 0.201,
+      "step": 2244
+    },
+    {
+      "epoch": 1.0630364013021605,
+      "grad_norm": 1.0648077726364136,
+      "learning_rate": 1.4967486005818727e-05,
+      "loss": 0.2237,
+      "step": 2245
+    },
+    {
+      "epoch": 1.0635099141757918,
+      "grad_norm": 1.3801108598709106,
+      "learning_rate": 1.4963046855808618e-05,
+      "loss": 0.2375,
+      "step": 2246
+    },
+    {
+      "epoch": 1.0639834270494228,
+      "grad_norm": 1.4944959878921509,
+      "learning_rate": 1.4958606407761482e-05,
+      "loss": 0.2441,
+      "step": 2247
+    },
+    {
+      "epoch": 1.0644569399230541,
+      "grad_norm": 2.0347981452941895,
+      "learning_rate": 1.4954164662838677e-05,
+      "loss": 0.2568,
+      "step": 2248
+    },
+    {
+      "epoch": 1.0649304527966854,
+      "grad_norm": 1.357796311378479,
+      "learning_rate": 1.4949721622201896e-05,
+      "loss": 0.2161,
+      "step": 2249
+    },
+    {
+      "epoch": 1.0654039656703167,
+      "grad_norm": 1.0744569301605225,
+      "learning_rate": 1.4945277287013178e-05,
+      "loss": 0.2269,
+      "step": 2250
+    },
+    {
+      "epoch": 1.065877478543948,
+      "grad_norm": 1.2349190711975098,
+      "learning_rate": 1.4940831658434893e-05,
+      "loss": 0.2213,
+      "step": 2251
+    },
+    {
+      "epoch": 1.066350991417579,
+      "grad_norm": 1.181551456451416,
+      "learning_rate": 1.4936384737629753e-05,
+      "loss": 0.218,
+      "step": 2252
+    },
+    {
+      "epoch": 1.0668245042912103,
+      "grad_norm": 1.792569637298584,
+      "learning_rate": 1.4931936525760806e-05,
+      "loss": 0.2103,
+      "step": 2253
+    },
+    {
+      "epoch": 1.0672980171648416,
+      "grad_norm": 1.04933500289917,
+      "learning_rate": 1.4927487023991441e-05,
+      "loss": 0.2085,
+      "step": 2254
+    },
+    {
+      "epoch": 1.067771530038473,
+      "grad_norm": 1.1748439073562622,
+      "learning_rate": 1.4923036233485383e-05,
+      "loss": 0.2093,
+      "step": 2255
+    },
+    {
+      "epoch": 1.0682450429121042,
+      "grad_norm": 1.149575114250183,
+      "learning_rate": 1.4918584155406688e-05,
+      "loss": 0.251,
+      "step": 2256
+    },
+    {
+      "epoch": 1.0687185557857355,
+      "grad_norm": 1.4834145307540894,
+      "learning_rate": 1.4914130790919761e-05,
+      "loss": 0.2117,
+      "step": 2257
+    },
+    {
+      "epoch": 1.0691920686593668,
+      "grad_norm": 1.1242401599884033,
+      "learning_rate": 1.4909676141189332e-05,
+      "loss": 0.2525,
+      "step": 2258
+    },
+    {
+      "epoch": 1.0696655815329978,
+      "grad_norm": 2.0916712284088135,
+      "learning_rate": 1.4905220207380468e-05,
+      "loss": 0.2408,
+      "step": 2259
+    },
+    {
+      "epoch": 1.0701390944066291,
+      "grad_norm": 1.4777238368988037,
+      "learning_rate": 1.4900762990658585e-05,
+      "loss": 0.2203,
+      "step": 2260
+    },
+    {
+      "epoch": 1.0706126072802604,
+      "grad_norm": 1.251544713973999,
+      "learning_rate": 1.4896304492189417e-05,
+      "loss": 0.2283,
+      "step": 2261
+    },
+    {
+      "epoch": 1.0710861201538917,
+      "grad_norm": 1.3505147695541382,
+      "learning_rate": 1.489184471313905e-05,
+      "loss": 0.2255,
+      "step": 2262
+    },
+    {
+      "epoch": 1.071559633027523,
+      "grad_norm": 1.4952011108398438,
+      "learning_rate": 1.4887383654673889e-05,
+      "loss": 0.2214,
+      "step": 2263
+    },
+    {
+      "epoch": 1.0720331459011543,
+      "grad_norm": 1.3948256969451904,
+      "learning_rate": 1.488292131796068e-05,
+      "loss": 0.2502,
+      "step": 2264
+    },
+    {
+      "epoch": 1.0725066587747853,
+      "grad_norm": 1.5061980485916138,
+      "learning_rate": 1.4878457704166506e-05,
+      "loss": 0.1799,
+      "step": 2265
+    },
+    {
+      "epoch": 1.0729801716484166,
+      "grad_norm": 1.324911117553711,
+      "learning_rate": 1.4873992814458786e-05,
+      "loss": 0.2278,
+      "step": 2266
+    },
+    {
+      "epoch": 1.073453684522048,
+      "grad_norm": 1.1821624040603638,
+      "learning_rate": 1.4869526650005264e-05,
+      "loss": 0.2209,
+      "step": 2267
+    },
+    {
+      "epoch": 1.0739271973956792,
+      "grad_norm": 1.988358736038208,
+      "learning_rate": 1.4865059211974024e-05,
+      "loss": 0.2366,
+      "step": 2268
+    },
+    {
+      "epoch": 1.0744007102693105,
+      "grad_norm": 1.212149739265442,
+      "learning_rate": 1.4860590501533482e-05,
+      "loss": 0.2386,
+      "step": 2269
+    },
+    {
+      "epoch": 1.0748742231429418,
+      "grad_norm": 1.2160612344741821,
+      "learning_rate": 1.4856120519852383e-05,
+      "loss": 0.2227,
+      "step": 2270
+    },
+    {
+      "epoch": 1.0753477360165729,
+      "grad_norm": 1.1903105974197388,
+      "learning_rate": 1.4851649268099813e-05,
+      "loss": 0.2442,
+      "step": 2271
+    },
+    {
+      "epoch": 1.0758212488902041,
+      "grad_norm": 1.6250642538070679,
+      "learning_rate": 1.484717674744518e-05,
+      "loss": 0.23,
+      "step": 2272
+    },
+    {
+      "epoch": 1.0762947617638354,
+      "grad_norm": 1.3602406978607178,
+      "learning_rate": 1.484270295905823e-05,
+      "loss": 0.2328,
+      "step": 2273
+    },
+    {
+      "epoch": 1.0767682746374667,
+      "grad_norm": 1.4769501686096191,
+      "learning_rate": 1.4838227904109041e-05,
+      "loss": 0.2319,
+      "step": 2274
+    },
+    {
+      "epoch": 1.077241787511098,
+      "grad_norm": 1.1678149700164795,
+      "learning_rate": 1.4833751583768017e-05,
+      "loss": 0.221,
+      "step": 2275
+    },
+    {
+      "epoch": 1.0777153003847293,
+      "grad_norm": 2.216442823410034,
+      "learning_rate": 1.48292739992059e-05,
+      "loss": 0.2286,
+      "step": 2276
+    },
+    {
+      "epoch": 1.0781888132583606,
+      "grad_norm": 2.1490554809570312,
+      "learning_rate": 1.4824795151593756e-05,
+      "loss": 0.2138,
+      "step": 2277
+    },
+    {
+      "epoch": 1.0786623261319916,
+      "grad_norm": 1.926507830619812,
+      "learning_rate": 1.4820315042102986e-05,
+      "loss": 0.2501,
+      "step": 2278
+    },
+    {
+      "epoch": 1.079135839005623,
+      "grad_norm": 1.7893540859222412,
+      "learning_rate": 1.481583367190532e-05,
+      "loss": 0.2324,
+      "step": 2279
+    },
+    {
+      "epoch": 1.0796093518792542,
+      "grad_norm": 1.2722523212432861,
+      "learning_rate": 1.4811351042172813e-05,
+      "loss": 0.2112,
+      "step": 2280
+    },
+    {
+      "epoch": 1.0800828647528855,
+      "grad_norm": 2.427342414855957,
+      "learning_rate": 1.480686715407786e-05,
+      "loss": 0.2359,
+      "step": 2281
+    },
+    {
+      "epoch": 1.0805563776265168,
+      "grad_norm": 1.4813814163208008,
+      "learning_rate": 1.4802382008793174e-05,
+      "loss": 0.2157,
+      "step": 2282
+    },
+    {
+      "epoch": 1.0810298905001479,
+      "grad_norm": 1.7409425973892212,
+      "learning_rate": 1.4797895607491803e-05,
+      "loss": 0.2105,
+      "step": 2283
+    },
+    {
+      "epoch": 1.0815034033737791,
+      "grad_norm": 1.6621959209442139,
+      "learning_rate": 1.4793407951347125e-05,
+      "loss": 0.2328,
+      "step": 2284
+    },
+    {
+      "epoch": 1.0819769162474104,
+      "grad_norm": 1.1202408075332642,
+      "learning_rate": 1.4788919041532836e-05,
+      "loss": 0.2422,
+      "step": 2285
+    },
+    {
+      "epoch": 1.0824504291210417,
+      "grad_norm": 1.863373041152954,
+      "learning_rate": 1.4784428879222974e-05,
+      "loss": 0.2561,
+      "step": 2286
+    },
+    {
+      "epoch": 1.082923941994673,
+      "grad_norm": 1.392762541770935,
+      "learning_rate": 1.4779937465591893e-05,
+      "loss": 0.2376,
+      "step": 2287
+    },
+    {
+      "epoch": 1.0833974548683043,
+      "grad_norm": 1.4913394451141357,
+      "learning_rate": 1.4775444801814283e-05,
+      "loss": 0.2238,
+      "step": 2288
+    },
+    {
+      "epoch": 1.0838709677419356,
+      "grad_norm": 0.916836142539978,
+      "learning_rate": 1.4770950889065154e-05,
+      "loss": 0.2274,
+      "step": 2289
+    },
+    {
+      "epoch": 1.0843444806155667,
+      "grad_norm": 1.3088781833648682,
+      "learning_rate": 1.4766455728519846e-05,
+      "loss": 0.2168,
+      "step": 2290
+    },
+    {
+      "epoch": 1.084817993489198,
+      "grad_norm": 1.986661672592163,
+      "learning_rate": 1.4761959321354025e-05,
+      "loss": 0.2137,
+      "step": 2291
+    },
+    {
+      "epoch": 1.0852915063628292,
+      "grad_norm": 2.061962604522705,
+      "learning_rate": 1.4757461668743681e-05,
+      "loss": 0.1949,
+      "step": 2292
+    },
+    {
+      "epoch": 1.0857650192364605,
+      "grad_norm": 1.4279181957244873,
+      "learning_rate": 1.4752962771865136e-05,
+      "loss": 0.2108,
+      "step": 2293
+    },
+    {
+      "epoch": 1.0862385321100918,
+      "grad_norm": 1.3412773609161377,
+      "learning_rate": 1.474846263189503e-05,
+      "loss": 0.2322,
+      "step": 2294
+    },
+    {
+      "epoch": 1.086712044983723,
+      "grad_norm": 2.3214855194091797,
+      "learning_rate": 1.4743961250010331e-05,
+      "loss": 0.198,
+      "step": 2295
+    },
+    {
+      "epoch": 1.0871855578573542,
+      "grad_norm": 1.6193276643753052,
+      "learning_rate": 1.4739458627388332e-05,
+      "loss": 0.2224,
+      "step": 2296
+    },
+    {
+      "epoch": 1.0876590707309854,
+      "grad_norm": 1.793225646018982,
+      "learning_rate": 1.4734954765206652e-05,
+      "loss": 0.2259,
+      "step": 2297
+    },
+    {
+      "epoch": 1.0881325836046167,
+      "grad_norm": 1.7120318412780762,
+      "learning_rate": 1.4730449664643234e-05,
+      "loss": 0.2271,
+      "step": 2298
+    },
+    {
+      "epoch": 1.088606096478248,
+      "grad_norm": 1.1128010749816895,
+      "learning_rate": 1.4725943326876342e-05,
+      "loss": 0.2003,
+      "step": 2299
+    },
+    {
+      "epoch": 1.0890796093518793,
+      "grad_norm": 1.3492709398269653,
+      "learning_rate": 1.472143575308456e-05,
+      "loss": 0.2332,
+      "step": 2300
+    },
+    {
+      "epoch": 1.0895531222255106,
+      "grad_norm": 1.4029978513717651,
+      "learning_rate": 1.471692694444681e-05,
+      "loss": 0.2185,
+      "step": 2301
+    },
+    {
+      "epoch": 1.0900266350991417,
+      "grad_norm": 1.254938006401062,
+      "learning_rate": 1.471241690214232e-05,
+      "loss": 0.223,
+      "step": 2302
+    },
+    {
+      "epoch": 1.090500147972773,
+      "grad_norm": 1.5678329467773438,
+      "learning_rate": 1.4707905627350653e-05,
+      "loss": 0.2186,
+      "step": 2303
+    },
+    {
+      "epoch": 1.0909736608464042,
+      "grad_norm": 1.8724418878555298,
+      "learning_rate": 1.4703393121251685e-05,
+      "loss": 0.2135,
+      "step": 2304
+    },
+    {
+      "epoch": 1.0914471737200355,
+      "grad_norm": 1.3198801279067993,
+      "learning_rate": 1.4698879385025625e-05,
+      "loss": 0.2076,
+      "step": 2305
+    },
+    {
+      "epoch": 1.0919206865936668,
+      "grad_norm": 1.733998417854309,
+      "learning_rate": 1.4694364419852986e-05,
+      "loss": 0.2184,
+      "step": 2306
+    },
+    {
+      "epoch": 1.092394199467298,
+      "grad_norm": 0.94035804271698,
+      "learning_rate": 1.4689848226914627e-05,
+      "loss": 0.2124,
+      "step": 2307
+    },
+    {
+      "epoch": 1.0928677123409292,
+      "grad_norm": 0.9416120052337646,
+      "learning_rate": 1.4685330807391704e-05,
+      "loss": 0.2126,
+      "step": 2308
+    },
+    {
+      "epoch": 1.0933412252145605,
+      "grad_norm": 1.1188304424285889,
+      "learning_rate": 1.468081216246571e-05,
+      "loss": 0.2299,
+      "step": 2309
+    },
+    {
+      "epoch": 1.0938147380881917,
+      "grad_norm": 1.0400240421295166,
+      "learning_rate": 1.467629229331845e-05,
+      "loss": 0.2196,
+      "step": 2310
+    },
+    {
+      "epoch": 1.094288250961823,
+      "grad_norm": 1.3964438438415527,
+      "learning_rate": 1.4671771201132054e-05,
+      "loss": 0.2477,
+      "step": 2311
+    },
+    {
+      "epoch": 1.0947617638354543,
+      "grad_norm": 1.3592251539230347,
+      "learning_rate": 1.4667248887088971e-05,
+      "loss": 0.2073,
+      "step": 2312
+    },
+    {
+      "epoch": 1.0952352767090856,
+      "grad_norm": 3.0954136848449707,
+      "learning_rate": 1.4662725352371967e-05,
+      "loss": 0.2125,
+      "step": 2313
+    },
+    {
+      "epoch": 1.0957087895827167,
+      "grad_norm": 1.2312672138214111,
+      "learning_rate": 1.4658200598164127e-05,
+      "loss": 0.2282,
+      "step": 2314
+    },
+    {
+      "epoch": 1.096182302456348,
+      "grad_norm": 1.0197677612304688,
+      "learning_rate": 1.4653674625648861e-05,
+      "loss": 0.2203,
+      "step": 2315
+    },
+    {
+      "epoch": 1.0966558153299792,
+      "grad_norm": 1.2136882543563843,
+      "learning_rate": 1.4649147436009895e-05,
+      "loss": 0.2358,
+      "step": 2316
+    },
+    {
+      "epoch": 1.0971293282036105,
+      "grad_norm": 1.1311389207839966,
+      "learning_rate": 1.4644619030431264e-05,
+      "loss": 0.2187,
+      "step": 2317
+    },
+    {
+      "epoch": 1.0976028410772418,
+      "grad_norm": 1.0771368741989136,
+      "learning_rate": 1.4640089410097333e-05,
+      "loss": 0.2128,
+      "step": 2318
+    },
+    {
+      "epoch": 1.098076353950873,
+      "grad_norm": 1.247301459312439,
+      "learning_rate": 1.463555857619278e-05,
+      "loss": 0.2158,
+      "step": 2319
+    },
+    {
+      "epoch": 1.0985498668245044,
+      "grad_norm": 1.5069024562835693,
+      "learning_rate": 1.4631026529902601e-05,
+      "loss": 0.2248,
+      "step": 2320
+    },
+    {
+      "epoch": 1.0990233796981355,
+      "grad_norm": 1.451413869857788,
+      "learning_rate": 1.462649327241211e-05,
+      "loss": 0.2121,
+      "step": 2321
+    },
+    {
+      "epoch": 1.0994968925717667,
+      "grad_norm": 1.4309951066970825,
+      "learning_rate": 1.4621958804906938e-05,
+      "loss": 0.2378,
+      "step": 2322
+    },
+    {
+      "epoch": 1.099970405445398,
+      "grad_norm": 1.5729237794876099,
+      "learning_rate": 1.4617423128573028e-05,
+      "loss": 0.2183,
+      "step": 2323
+    },
+    {
+      "epoch": 1.1004439183190293,
+      "grad_norm": 2.354729413986206,
+      "learning_rate": 1.4612886244596647e-05,
+      "loss": 0.223,
+      "step": 2324
+    },
+    {
+      "epoch": 1.1009174311926606,
+      "grad_norm": 1.197218894958496,
+      "learning_rate": 1.4608348154164367e-05,
+      "loss": 0.2095,
+      "step": 2325
+    },
+    {
+      "epoch": 1.101390944066292,
+      "grad_norm": 2.5518031120300293,
+      "learning_rate": 1.4603808858463085e-05,
+      "loss": 0.2002,
+      "step": 2326
+    },
+    {
+      "epoch": 1.101864456939923,
+      "grad_norm": 1.1937416791915894,
+      "learning_rate": 1.459926835868001e-05,
+      "loss": 0.2171,
+      "step": 2327
+    },
+    {
+      "epoch": 1.1023379698135543,
+      "grad_norm": 1.7216546535491943,
+      "learning_rate": 1.4594726656002664e-05,
+      "loss": 0.2304,
+      "step": 2328
+    },
+    {
+      "epoch": 1.1028114826871855,
+      "grad_norm": 1.408703327178955,
+      "learning_rate": 1.4590183751618889e-05,
+      "loss": 0.2391,
+      "step": 2329
+    },
+    {
+      "epoch": 1.1032849955608168,
+      "grad_norm": 1.8831274509429932,
+      "learning_rate": 1.4585639646716836e-05,
+      "loss": 0.2414,
+      "step": 2330
+    },
+    {
+      "epoch": 1.1037585084344481,
+      "grad_norm": 1.7065414190292358,
+      "learning_rate": 1.4581094342484972e-05,
+      "loss": 0.2365,
+      "step": 2331
+    },
+    {
+      "epoch": 1.1042320213080794,
+      "grad_norm": 1.767787218093872,
+      "learning_rate": 1.4576547840112077e-05,
+      "loss": 0.2229,
+      "step": 2332
+    },
+    {
+      "epoch": 1.1047055341817105,
+      "grad_norm": 1.4647932052612305,
+      "learning_rate": 1.4572000140787244e-05,
+      "loss": 0.2404,
+      "step": 2333
+    },
+    {
+      "epoch": 1.1051790470553418,
+      "grad_norm": 1.9528988599777222,
+      "learning_rate": 1.4567451245699884e-05,
+      "loss": 0.2326,
+      "step": 2334
+    },
+    {
+      "epoch": 1.105652559928973,
+      "grad_norm": 1.284548282623291,
+      "learning_rate": 1.456290115603971e-05,
+      "loss": 0.2263,
+      "step": 2335
+    },
+    {
+      "epoch": 1.1061260728026043,
+      "grad_norm": 1.5598032474517822,
+      "learning_rate": 1.4558349872996762e-05,
+      "loss": 0.2211,
+      "step": 2336
+    },
+    {
+      "epoch": 1.1065995856762356,
+      "grad_norm": 1.3689323663711548,
+      "learning_rate": 1.4553797397761376e-05,
+      "loss": 0.2168,
+      "step": 2337
+    },
+    {
+      "epoch": 1.107073098549867,
+      "grad_norm": 0.9562427401542664,
+      "learning_rate": 1.4549243731524213e-05,
+      "loss": 0.213,
+      "step": 2338
+    },
+    {
+      "epoch": 1.107546611423498,
+      "grad_norm": 1.2045793533325195,
+      "learning_rate": 1.4544688875476238e-05,
+      "loss": 0.2287,
+      "step": 2339
+    },
+    {
+      "epoch": 1.1080201242971293,
+      "grad_norm": 1.2727464437484741,
+      "learning_rate": 1.4540132830808733e-05,
+      "loss": 0.2112,
+      "step": 2340
+    },
+    {
+      "epoch": 1.1084936371707605,
+      "grad_norm": 1.1461354494094849,
+      "learning_rate": 1.4535575598713287e-05,
+      "loss": 0.2019,
+      "step": 2341
+    },
+    {
+      "epoch": 1.1089671500443918,
+      "grad_norm": 1.330062985420227,
+      "learning_rate": 1.4531017180381797e-05,
+      "loss": 0.212,
+      "step": 2342
+    },
+    {
+      "epoch": 1.1094406629180231,
+      "grad_norm": 1.4322924613952637,
+      "learning_rate": 1.4526457577006473e-05,
+      "loss": 0.2197,
+      "step": 2343
+    },
+    {
+      "epoch": 1.1099141757916544,
+      "grad_norm": 1.5362545251846313,
+      "learning_rate": 1.4521896789779842e-05,
+      "loss": 0.2195,
+      "step": 2344
+    },
+    {
+      "epoch": 1.1103876886652855,
+      "grad_norm": 1.0807383060455322,
+      "learning_rate": 1.4517334819894724e-05,
+      "loss": 0.2394,
+      "step": 2345
+    },
+    {
+      "epoch": 1.1108612015389168,
+      "grad_norm": 1.3042844533920288,
+      "learning_rate": 1.4512771668544266e-05,
+      "loss": 0.2393,
+      "step": 2346
+    },
+    {
+      "epoch": 1.111334714412548,
+      "grad_norm": 1.3964802026748657,
+      "learning_rate": 1.4508207336921914e-05,
+      "loss": 0.2293,
+      "step": 2347
+    },
+    {
+      "epoch": 1.1118082272861793,
+      "grad_norm": 0.9700664281845093,
+      "learning_rate": 1.4503641826221424e-05,
+      "loss": 0.2346,
+      "step": 2348
+    },
+    {
+      "epoch": 1.1122817401598106,
+      "grad_norm": 1.142829418182373,
+      "learning_rate": 1.449907513763686e-05,
+      "loss": 0.2185,
+      "step": 2349
+    },
+    {
+      "epoch": 1.112755253033442,
+      "grad_norm": 1.3726601600646973,
+      "learning_rate": 1.44945072723626e-05,
+      "loss": 0.2247,
+      "step": 2350
+    },
+    {
+      "epoch": 1.1132287659070732,
+      "grad_norm": 1.0523629188537598,
+      "learning_rate": 1.4489938231593321e-05,
+      "loss": 0.2263,
+      "step": 2351
+    },
+    {
+      "epoch": 1.1137022787807043,
+      "grad_norm": 1.1638647317886353,
+      "learning_rate": 1.4485368016524013e-05,
+      "loss": 0.2094,
+      "step": 2352
+    },
+    {
+      "epoch": 1.1141757916543356,
+      "grad_norm": 1.121314525604248,
+      "learning_rate": 1.4480796628349972e-05,
+      "loss": 0.2036,
+      "step": 2353
+    },
+    {
+      "epoch": 1.1146493045279668,
+      "grad_norm": 1.4763176441192627,
+      "learning_rate": 1.4476224068266798e-05,
+      "loss": 0.2385,
+      "step": 2354
+    },
+    {
+      "epoch": 1.1151228174015981,
+      "grad_norm": 1.7545726299285889,
+      "learning_rate": 1.4471650337470402e-05,
+      "loss": 0.2234,
+      "step": 2355
+    },
+    {
+      "epoch": 1.1155963302752294,
+      "grad_norm": 1.8628343343734741,
+      "learning_rate": 1.4467075437156998e-05,
+      "loss": 0.2142,
+      "step": 2356
+    },
+    {
+      "epoch": 1.1160698431488607,
+      "grad_norm": 1.5996010303497314,
+      "learning_rate": 1.446249936852311e-05,
+      "loss": 0.2413,
+      "step": 2357
+    },
+    {
+      "epoch": 1.1165433560224918,
+      "grad_norm": 1.7511857748031616,
+      "learning_rate": 1.4457922132765563e-05,
+      "loss": 0.2044,
+      "step": 2358
+    },
+    {
+      "epoch": 1.117016868896123,
+      "grad_norm": 0.8672329783439636,
+      "learning_rate": 1.4453343731081488e-05,
+      "loss": 0.2006,
+      "step": 2359
+    },
+    {
+      "epoch": 1.1174903817697543,
+      "grad_norm": 1.460821509361267,
+      "learning_rate": 1.444876416466832e-05,
+      "loss": 0.2255,
+      "step": 2360
+    },
+    {
+      "epoch": 1.1179638946433856,
+      "grad_norm": 1.3242175579071045,
+      "learning_rate": 1.4444183434723807e-05,
+      "loss": 0.2342,
+      "step": 2361
+    },
+    {
+      "epoch": 1.118437407517017,
+      "grad_norm": 1.3085192441940308,
+      "learning_rate": 1.4439601542445987e-05,
+      "loss": 0.2688,
+      "step": 2362
+    },
+    {
+      "epoch": 1.1189109203906482,
+      "grad_norm": 1.3582148551940918,
+      "learning_rate": 1.4435018489033214e-05,
+      "loss": 0.2452,
+      "step": 2363
+    },
+    {
+      "epoch": 1.1193844332642793,
+      "grad_norm": 1.1882847547531128,
+      "learning_rate": 1.443043427568414e-05,
+      "loss": 0.2238,
+      "step": 2364
+    },
+    {
+      "epoch": 1.1198579461379106,
+      "grad_norm": 1.574207067489624,
+      "learning_rate": 1.4425848903597724e-05,
+      "loss": 0.2382,
+      "step": 2365
+    },
+    {
+      "epoch": 1.1203314590115419,
+      "grad_norm": 1.0986261367797852,
+      "learning_rate": 1.4421262373973223e-05,
+      "loss": 0.2102,
+      "step": 2366
+    },
+    {
+      "epoch": 1.1208049718851731,
+      "grad_norm": 1.3533656597137451,
+      "learning_rate": 1.4416674688010202e-05,
+      "loss": 0.2158,
+      "step": 2367
+    },
+    {
+      "epoch": 1.1212784847588044,
+      "grad_norm": 1.6296641826629639,
+      "learning_rate": 1.4412085846908526e-05,
+      "loss": 0.2101,
+      "step": 2368
+    },
+    {
+      "epoch": 1.1217519976324357,
+      "grad_norm": 1.0967878103256226,
+      "learning_rate": 1.4407495851868359e-05,
+      "loss": 0.2318,
+      "step": 2369
+    },
+    {
+      "epoch": 1.1222255105060668,
+      "grad_norm": 1.2239967584609985,
+      "learning_rate": 1.440290470409017e-05,
+      "loss": 0.2045,
+      "step": 2370
+    },
+    {
+      "epoch": 1.122699023379698,
+      "grad_norm": 1.5295300483703613,
+      "learning_rate": 1.4398312404774735e-05,
+      "loss": 0.2554,
+      "step": 2371
+    },
+    {
+      "epoch": 1.1231725362533294,
+      "grad_norm": 1.227840542793274,
+      "learning_rate": 1.439371895512312e-05,
+      "loss": 0.2367,
+      "step": 2372
+    },
+    {
+      "epoch": 1.1236460491269606,
+      "grad_norm": 1.9119755029678345,
+      "learning_rate": 1.4389124356336696e-05,
+      "loss": 0.2372,
+      "step": 2373
+    },
+    {
+      "epoch": 1.124119562000592,
+      "grad_norm": 1.034201979637146,
+      "learning_rate": 1.4384528609617143e-05,
+      "loss": 0.199,
+      "step": 2374
+    },
+    {
+      "epoch": 1.1245930748742232,
+      "grad_norm": 1.1491769552230835,
+      "learning_rate": 1.4379931716166429e-05,
+      "loss": 0.2293,
+      "step": 2375
+    },
+    {
+      "epoch": 1.1250665877478543,
+      "grad_norm": 1.0914734601974487,
+      "learning_rate": 1.4375333677186829e-05,
+      "loss": 0.2179,
+      "step": 2376
+    },
+    {
+      "epoch": 1.1255401006214856,
+      "grad_norm": 2.126800775527954,
+      "learning_rate": 1.4370734493880916e-05,
+      "loss": 0.2166,
+      "step": 2377
+    },
+    {
+      "epoch": 1.1260136134951169,
+      "grad_norm": 2.5790648460388184,
+      "learning_rate": 1.4366134167451562e-05,
+      "loss": 0.256,
+      "step": 2378
+    },
+    {
+      "epoch": 1.1264871263687481,
+      "grad_norm": 1.1876254081726074,
+      "learning_rate": 1.4361532699101935e-05,
+      "loss": 0.223,
+      "step": 2379
+    },
+    {
+      "epoch": 1.1269606392423794,
+      "grad_norm": 1.856112003326416,
+      "learning_rate": 1.4356930090035508e-05,
+      "loss": 0.2211,
+      "step": 2380
+    },
+    {
+      "epoch": 1.1274341521160107,
+      "grad_norm": 1.3353132009506226,
+      "learning_rate": 1.4352326341456048e-05,
+      "loss": 0.2256,
+      "step": 2381
+    },
+    {
+      "epoch": 1.127907664989642,
+      "grad_norm": 1.3428007364273071,
+      "learning_rate": 1.4347721454567623e-05,
+      "loss": 0.2296,
+      "step": 2382
+    },
+    {
+      "epoch": 1.128381177863273,
+      "grad_norm": 1.5457638502120972,
+      "learning_rate": 1.4343115430574592e-05,
+      "loss": 0.2188,
+      "step": 2383
+    },
+    {
+      "epoch": 1.1288546907369044,
+      "grad_norm": 1.9435369968414307,
+      "learning_rate": 1.433850827068162e-05,
+      "loss": 0.2184,
+      "step": 2384
+    },
+    {
+      "epoch": 1.1293282036105357,
+      "grad_norm": 1.060576319694519,
+      "learning_rate": 1.4333899976093664e-05,
+      "loss": 0.2332,
+      "step": 2385
+    },
+    {
+      "epoch": 1.129801716484167,
+      "grad_norm": 2.13041090965271,
+      "learning_rate": 1.4329290548015981e-05,
+      "loss": 0.231,
+      "step": 2386
+    },
+    {
+      "epoch": 1.1302752293577982,
+      "grad_norm": 1.1683634519577026,
+      "learning_rate": 1.4324679987654118e-05,
+      "loss": 0.211,
+      "step": 2387
+    },
+    {
+      "epoch": 1.1307487422314293,
+      "grad_norm": 1.2120347023010254,
+      "learning_rate": 1.4320068296213926e-05,
+      "loss": 0.2094,
+      "step": 2388
+    },
+    {
+      "epoch": 1.1312222551050606,
+      "grad_norm": 1.454931616783142,
+      "learning_rate": 1.4315455474901547e-05,
+      "loss": 0.2472,
+      "step": 2389
+    },
+    {
+      "epoch": 1.1316957679786919,
+      "grad_norm": 1.3851542472839355,
+      "learning_rate": 1.4310841524923421e-05,
+      "loss": 0.2272,
+      "step": 2390
+    },
+    {
+      "epoch": 1.1321692808523232,
+      "grad_norm": 1.0476808547973633,
+      "learning_rate": 1.4306226447486283e-05,
+      "loss": 0.2176,
+      "step": 2391
+    },
+    {
+      "epoch": 1.1326427937259544,
+      "grad_norm": 1.5204083919525146,
+      "learning_rate": 1.4301610243797158e-05,
+      "loss": 0.2392,
+      "step": 2392
+    },
+    {
+      "epoch": 1.1331163065995857,
+      "grad_norm": 1.792072057723999,
+      "learning_rate": 1.4296992915063373e-05,
+      "loss": 0.2097,
+      "step": 2393
+    },
+    {
+      "epoch": 1.133589819473217,
+      "grad_norm": 1.989022135734558,
+      "learning_rate": 1.4292374462492547e-05,
+      "loss": 0.2483,
+      "step": 2394
+    },
+    {
+      "epoch": 1.134063332346848,
+      "grad_norm": 1.7426658868789673,
+      "learning_rate": 1.4287754887292589e-05,
+      "loss": 0.2133,
+      "step": 2395
+    },
+    {
+      "epoch": 1.1345368452204794,
+      "grad_norm": 1.5938878059387207,
+      "learning_rate": 1.42831341906717e-05,
+      "loss": 0.2123,
+      "step": 2396
+    },
+    {
+      "epoch": 1.1350103580941107,
+      "grad_norm": 1.6211594343185425,
+      "learning_rate": 1.4278512373838386e-05,
+      "loss": 0.2364,
+      "step": 2397
+    },
+    {
+      "epoch": 1.135483870967742,
+      "grad_norm": 0.9879429936408997,
+      "learning_rate": 1.4273889438001439e-05,
+      "loss": 0.2109,
+      "step": 2398
+    },
+    {
+      "epoch": 1.1359573838413732,
+      "grad_norm": 1.1694254875183105,
+      "learning_rate": 1.426926538436994e-05,
+      "loss": 0.2156,
+      "step": 2399
+    },
+    {
+      "epoch": 1.1364308967150045,
+      "grad_norm": 1.0623806715011597,
+      "learning_rate": 1.4264640214153263e-05,
+      "loss": 0.1991,
+      "step": 2400
+    },
+    {
+      "epoch": 1.1369044095886358,
+      "grad_norm": 1.6476950645446777,
+      "learning_rate": 1.4260013928561081e-05,
+      "loss": 0.242,
+      "step": 2401
+    },
+    {
+      "epoch": 1.1373779224622669,
+      "grad_norm": 1.9095466136932373,
+      "learning_rate": 1.4255386528803354e-05,
+      "loss": 0.2194,
+      "step": 2402
+    },
+    {
+      "epoch": 1.1378514353358982,
+      "grad_norm": 1.2309021949768066,
+      "learning_rate": 1.4250758016090335e-05,
+      "loss": 0.2141,
+      "step": 2403
+    },
+    {
+      "epoch": 1.1383249482095295,
+      "grad_norm": 1.0731322765350342,
+      "learning_rate": 1.4246128391632562e-05,
+      "loss": 0.2075,
+      "step": 2404
+    },
+    {
+      "epoch": 1.1387984610831607,
+      "grad_norm": 1.4842568635940552,
+      "learning_rate": 1.4241497656640872e-05,
+      "loss": 0.2145,
+      "step": 2405
+    },
+    {
+      "epoch": 1.139271973956792,
+      "grad_norm": 1.6689938306808472,
+      "learning_rate": 1.4236865812326386e-05,
+      "loss": 0.2281,
+      "step": 2406
+    },
+    {
+      "epoch": 1.139745486830423,
+      "grad_norm": 1.186214566230774,
+      "learning_rate": 1.4232232859900523e-05,
+      "loss": 0.2296,
+      "step": 2407
+    },
+    {
+      "epoch": 1.1402189997040544,
+      "grad_norm": 2.036839246749878,
+      "learning_rate": 1.4227598800574984e-05,
+      "loss": 0.1975,
+      "step": 2408
+    },
+    {
+      "epoch": 1.1406925125776857,
+      "grad_norm": 1.4921070337295532,
+      "learning_rate": 1.4222963635561761e-05,
+      "loss": 0.2342,
+      "step": 2409
+    },
+    {
+      "epoch": 1.141166025451317,
+      "grad_norm": 1.0753675699234009,
+      "learning_rate": 1.4218327366073142e-05,
+      "loss": 0.2251,
+      "step": 2410
+    },
+    {
+      "epoch": 1.1416395383249482,
+      "grad_norm": 1.8981343507766724,
+      "learning_rate": 1.4213689993321693e-05,
+      "loss": 0.2247,
+      "step": 2411
+    },
+    {
+      "epoch": 1.1421130511985795,
+      "grad_norm": 1.5621330738067627,
+      "learning_rate": 1.4209051518520279e-05,
+      "loss": 0.2106,
+      "step": 2412
+    },
+    {
+      "epoch": 1.1425865640722108,
+      "grad_norm": 1.145002841949463,
+      "learning_rate": 1.4204411942882046e-05,
+      "loss": 0.2496,
+      "step": 2413
+    },
+    {
+      "epoch": 1.1430600769458419,
+      "grad_norm": 1.893953561782837,
+      "learning_rate": 1.4199771267620429e-05,
+      "loss": 0.223,
+      "step": 2414
+    },
+    {
+      "epoch": 1.1435335898194732,
+      "grad_norm": 2.6021175384521484,
+      "learning_rate": 1.4195129493949152e-05,
+      "loss": 0.2389,
+      "step": 2415
+    },
+    {
+      "epoch": 1.1440071026931045,
+      "grad_norm": 1.2552067041397095,
+      "learning_rate": 1.4190486623082224e-05,
+      "loss": 0.2072,
+      "step": 2416
+    },
+    {
+      "epoch": 1.1444806155667357,
+      "grad_norm": 1.4053682088851929,
+      "learning_rate": 1.418584265623395e-05,
+      "loss": 0.2115,
+      "step": 2417
+    },
+    {
+      "epoch": 1.144954128440367,
+      "grad_norm": 1.3213428258895874,
+      "learning_rate": 1.418119759461891e-05,
+      "loss": 0.2304,
+      "step": 2418
+    },
+    {
+      "epoch": 1.145427641313998,
+      "grad_norm": 1.09917151927948,
+      "learning_rate": 1.4176551439451976e-05,
+      "loss": 0.2189,
+      "step": 2419
+    },
+    {
+      "epoch": 1.1459011541876294,
+      "grad_norm": 3.225097417831421,
+      "learning_rate": 1.4171904191948306e-05,
+      "loss": 0.215,
+      "step": 2420
+    },
+    {
+      "epoch": 1.1463746670612607,
+      "grad_norm": 1.6562659740447998,
+      "learning_rate": 1.416725585332334e-05,
+      "loss": 0.2157,
+      "step": 2421
+    },
+    {
+      "epoch": 1.146848179934892,
+      "grad_norm": 1.3881036043167114,
+      "learning_rate": 1.4162606424792809e-05,
+      "loss": 0.2283,
+      "step": 2422
+    },
+    {
+      "epoch": 1.1473216928085233,
+      "grad_norm": 1.0177631378173828,
+      "learning_rate": 1.4157955907572722e-05,
+      "loss": 0.1981,
+      "step": 2423
+    },
+    {
+      "epoch": 1.1477952056821545,
+      "grad_norm": 1.4494411945343018,
+      "learning_rate": 1.4153304302879383e-05,
+      "loss": 0.2367,
+      "step": 2424
+    },
+    {
+      "epoch": 1.1482687185557858,
+      "grad_norm": 1.3473752737045288,
+      "learning_rate": 1.4148651611929371e-05,
+      "loss": 0.236,
+      "step": 2425
+    },
+    {
+      "epoch": 1.148742231429417,
+      "grad_norm": 1.4328070878982544,
+      "learning_rate": 1.4143997835939552e-05,
+      "loss": 0.2182,
+      "step": 2426
+    },
+    {
+      "epoch": 1.1492157443030482,
+      "grad_norm": 1.770027995109558,
+      "learning_rate": 1.4139342976127077e-05,
+      "loss": 0.2246,
+      "step": 2427
+    },
+    {
+      "epoch": 1.1496892571766795,
+      "grad_norm": 1.1986833810806274,
+      "learning_rate": 1.413468703370938e-05,
+      "loss": 0.2309,
+      "step": 2428
+    },
+    {
+      "epoch": 1.1501627700503108,
+      "grad_norm": 1.74411940574646,
+      "learning_rate": 1.4130030009904174e-05,
+      "loss": 0.2027,
+      "step": 2429
+    },
+    {
+      "epoch": 1.150636282923942,
+      "grad_norm": 1.3250808715820312,
+      "learning_rate": 1.4125371905929468e-05,
+      "loss": 0.2219,
+      "step": 2430
+    },
+    {
+      "epoch": 1.1511097957975733,
+      "grad_norm": 1.336143136024475,
+      "learning_rate": 1.4120712723003535e-05,
+      "loss": 0.2279,
+      "step": 2431
+    },
+    {
+      "epoch": 1.1515833086712046,
+      "grad_norm": 1.1617181301116943,
+      "learning_rate": 1.4116052462344942e-05,
+      "loss": 0.2143,
+      "step": 2432
+    },
+    {
+      "epoch": 1.1520568215448357,
+      "grad_norm": 1.7668710947036743,
+      "learning_rate": 1.4111391125172537e-05,
+      "loss": 0.2268,
+      "step": 2433
+    },
+    {
+      "epoch": 1.152530334418467,
+      "grad_norm": 1.455324649810791,
+      "learning_rate": 1.4106728712705446e-05,
+      "loss": 0.2289,
+      "step": 2434
+    },
+    {
+      "epoch": 1.1530038472920983,
+      "grad_norm": 1.4046434164047241,
+      "learning_rate": 1.4102065226163078e-05,
+      "loss": 0.2297,
+      "step": 2435
+    },
+    {
+      "epoch": 1.1534773601657295,
+      "grad_norm": 1.2692493200302124,
+      "learning_rate": 1.4097400666765122e-05,
+      "loss": 0.2339,
+      "step": 2436
+    },
+    {
+      "epoch": 1.1539508730393608,
+      "grad_norm": 1.7210558652877808,
+      "learning_rate": 1.4092735035731553e-05,
+      "loss": 0.2129,
+      "step": 2437
+    },
+    {
+      "epoch": 1.154424385912992,
+      "grad_norm": 1.2296080589294434,
+      "learning_rate": 1.4088068334282617e-05,
+      "loss": 0.2163,
+      "step": 2438
+    },
+    {
+      "epoch": 1.1548978987866232,
+      "grad_norm": 1.0044761896133423,
+      "learning_rate": 1.4083400563638847e-05,
+      "loss": 0.2193,
+      "step": 2439
+    },
+    {
+      "epoch": 1.1553714116602545,
+      "grad_norm": 2.0132339000701904,
+      "learning_rate": 1.407873172502105e-05,
+      "loss": 0.2281,
+      "step": 2440
+    },
+    {
+      "epoch": 1.1558449245338858,
+      "grad_norm": 2.1369824409484863,
+      "learning_rate": 1.4074061819650322e-05,
+      "loss": 0.2122,
+      "step": 2441
+    },
+    {
+      "epoch": 1.156318437407517,
+      "grad_norm": 1.4052802324295044,
+      "learning_rate": 1.4069390848748029e-05,
+      "loss": 0.2225,
+      "step": 2442
+    },
+    {
+      "epoch": 1.1567919502811483,
+      "grad_norm": 1.6282340288162231,
+      "learning_rate": 1.4064718813535817e-05,
+      "loss": 0.2247,
+      "step": 2443
+    },
+    {
+      "epoch": 1.1572654631547796,
+      "grad_norm": 1.5721560716629028,
+      "learning_rate": 1.4060045715235609e-05,
+      "loss": 0.2201,
+      "step": 2444
+    },
+    {
+      "epoch": 1.1577389760284107,
+      "grad_norm": 1.0424257516860962,
+      "learning_rate": 1.4055371555069615e-05,
+      "loss": 0.2363,
+      "step": 2445
+    },
+    {
+      "epoch": 1.158212488902042,
+      "grad_norm": 1.8262851238250732,
+      "learning_rate": 1.4050696334260312e-05,
+      "loss": 0.2222,
+      "step": 2446
+    },
+    {
+      "epoch": 1.1586860017756733,
+      "grad_norm": 2.008397340774536,
+      "learning_rate": 1.4046020054030465e-05,
+      "loss": 0.242,
+      "step": 2447
+    },
+    {
+      "epoch": 1.1591595146493046,
+      "grad_norm": 3.8094611167907715,
+      "learning_rate": 1.40413427156031e-05,
+      "loss": 0.2242,
+      "step": 2448
+    },
+    {
+      "epoch": 1.1596330275229358,
+      "grad_norm": 1.1817164421081543,
+      "learning_rate": 1.4036664320201538e-05,
+      "loss": 0.2312,
+      "step": 2449
+    },
+    {
+      "epoch": 1.160106540396567,
+      "grad_norm": 1.771918535232544,
+      "learning_rate": 1.4031984869049366e-05,
+      "loss": 0.209,
+      "step": 2450
+    },
+    {
+      "epoch": 1.1605800532701982,
+      "grad_norm": 1.6216905117034912,
+      "learning_rate": 1.4027304363370446e-05,
+      "loss": 0.2296,
+      "step": 2451
+    },
+    {
+      "epoch": 1.1610535661438295,
+      "grad_norm": 1.4540348052978516,
+      "learning_rate": 1.4022622804388923e-05,
+      "loss": 0.2135,
+      "step": 2452
+    },
+    {
+      "epoch": 1.1615270790174608,
+      "grad_norm": 1.7169173955917358,
+      "learning_rate": 1.4017940193329213e-05,
+      "loss": 0.2417,
+      "step": 2453
+    },
+    {
+      "epoch": 1.162000591891092,
+      "grad_norm": 1.2987480163574219,
+      "learning_rate": 1.401325653141601e-05,
+      "loss": 0.2275,
+      "step": 2454
+    },
+    {
+      "epoch": 1.1624741047647233,
+      "grad_norm": 1.303578495979309,
+      "learning_rate": 1.4008571819874273e-05,
+      "loss": 0.2323,
+      "step": 2455
+    },
+    {
+      "epoch": 1.1629476176383546,
+      "grad_norm": 1.2520272731781006,
+      "learning_rate": 1.4003886059929248e-05,
+      "loss": 0.2025,
+      "step": 2456
+    },
+    {
+      "epoch": 1.1634211305119857,
+      "grad_norm": 2.3889143466949463,
+      "learning_rate": 1.3999199252806452e-05,
+      "loss": 0.22,
+      "step": 2457
+    },
+    {
+      "epoch": 1.163894643385617,
+      "grad_norm": 1.60451340675354,
+      "learning_rate": 1.3994511399731675e-05,
+      "loss": 0.2413,
+      "step": 2458
+    },
+    {
+      "epoch": 1.1643681562592483,
+      "grad_norm": 1.227243185043335,
+      "learning_rate": 1.3989822501930972e-05,
+      "loss": 0.1986,
+      "step": 2459
+    },
+    {
+      "epoch": 1.1648416691328796,
+      "grad_norm": 1.0574538707733154,
+      "learning_rate": 1.3985132560630688e-05,
+      "loss": 0.2294,
+      "step": 2460
+    },
+    {
+      "epoch": 1.1653151820065109,
+      "grad_norm": 1.0879645347595215,
+      "learning_rate": 1.3980441577057426e-05,
+      "loss": 0.2376,
+      "step": 2461
+    },
+    {
+      "epoch": 1.1657886948801421,
+      "grad_norm": 2.3020236492156982,
+      "learning_rate": 1.397574955243807e-05,
+      "loss": 0.2257,
+      "step": 2462
+    },
+    {
+      "epoch": 1.1662622077537734,
+      "grad_norm": 1.1904938220977783,
+      "learning_rate": 1.3971056487999773e-05,
+      "loss": 0.2329,
+      "step": 2463
+    },
+    {
+      "epoch": 1.1667357206274045,
+      "grad_norm": 2.3944849967956543,
+      "learning_rate": 1.3966362384969963e-05,
+      "loss": 0.2004,
+      "step": 2464
+    },
+    {
+      "epoch": 1.1672092335010358,
+      "grad_norm": 1.1166476011276245,
+      "learning_rate": 1.3961667244576335e-05,
+      "loss": 0.2009,
+      "step": 2465
+    },
+    {
+      "epoch": 1.167682746374667,
+      "grad_norm": 1.1066415309906006,
+      "learning_rate": 1.3956971068046855e-05,
+      "loss": 0.2289,
+      "step": 2466
+    },
+    {
+      "epoch": 1.1681562592482984,
+      "grad_norm": 1.3403149843215942,
+      "learning_rate": 1.3952273856609767e-05,
+      "loss": 0.2034,
+      "step": 2467
+    },
+    {
+      "epoch": 1.1686297721219296,
+      "grad_norm": 1.3216232061386108,
+      "learning_rate": 1.3947575611493583e-05,
+      "loss": 0.2223,
+      "step": 2468
+    },
+    {
+      "epoch": 1.1691032849955607,
+      "grad_norm": 2.135986328125,
+      "learning_rate": 1.3942876333927077e-05,
+      "loss": 0.2079,
+      "step": 2469
+    },
+    {
+      "epoch": 1.169576797869192,
+      "grad_norm": 1.8347281217575073,
+      "learning_rate": 1.3938176025139305e-05,
+      "loss": 0.1986,
+      "step": 2470
+    },
+    {
+      "epoch": 1.1700503107428233,
+      "grad_norm": 1.1069951057434082,
+      "learning_rate": 1.3933474686359588e-05,
+      "loss": 0.2261,
+      "step": 2471
+    },
+    {
+      "epoch": 1.1705238236164546,
+      "grad_norm": 1.534125804901123,
+      "learning_rate": 1.3928772318817509e-05,
+      "loss": 0.23,
+      "step": 2472
+    },
+    {
+      "epoch": 1.1709973364900859,
+      "grad_norm": 0.9962480664253235,
+      "learning_rate": 1.3924068923742935e-05,
+      "loss": 0.2145,
+      "step": 2473
+    },
+    {
+      "epoch": 1.1714708493637171,
+      "grad_norm": 1.4505356550216675,
+      "learning_rate": 1.391936450236599e-05,
+      "loss": 0.2107,
+      "step": 2474
+    },
+    {
+      "epoch": 1.1719443622373484,
+      "grad_norm": 1.0452574491500854,
+      "learning_rate": 1.3914659055917068e-05,
+      "loss": 0.2004,
+      "step": 2475
+    },
+    {
+      "epoch": 1.1724178751109795,
+      "grad_norm": 1.055141568183899,
+      "learning_rate": 1.3909952585626836e-05,
+      "loss": 0.2231,
+      "step": 2476
+    },
+    {
+      "epoch": 1.1728913879846108,
+      "grad_norm": 1.314717411994934,
+      "learning_rate": 1.3905245092726225e-05,
+      "loss": 0.2346,
+      "step": 2477
+    },
+    {
+      "epoch": 1.173364900858242,
+      "grad_norm": 1.5005841255187988,
+      "learning_rate": 1.3900536578446437e-05,
+      "loss": 0.2334,
+      "step": 2478
+    },
+    {
+      "epoch": 1.1738384137318734,
+      "grad_norm": 1.6286835670471191,
+      "learning_rate": 1.3895827044018934e-05,
+      "loss": 0.2517,
+      "step": 2479
+    },
+    {
+      "epoch": 1.1743119266055047,
+      "grad_norm": 1.1182934045791626,
+      "learning_rate": 1.3891116490675451e-05,
+      "loss": 0.2388,
+      "step": 2480
+    },
+    {
+      "epoch": 1.1747854394791357,
+      "grad_norm": 1.1708893775939941,
+      "learning_rate": 1.3886404919647988e-05,
+      "loss": 0.2131,
+      "step": 2481
+    },
+    {
+      "epoch": 1.175258952352767,
+      "grad_norm": 1.1157783269882202,
+      "learning_rate": 1.3881692332168815e-05,
+      "loss": 0.2378,
+      "step": 2482
+    },
+    {
+      "epoch": 1.1757324652263983,
+      "grad_norm": 1.3915766477584839,
+      "learning_rate": 1.3876978729470459e-05,
+      "loss": 0.2189,
+      "step": 2483
+    },
+    {
+      "epoch": 1.1762059781000296,
+      "grad_norm": 1.132906198501587,
+      "learning_rate": 1.3872264112785714e-05,
+      "loss": 0.2082,
+      "step": 2484
+    },
+    {
+      "epoch": 1.1766794909736609,
+      "grad_norm": 1.12389075756073,
+      "learning_rate": 1.386754848334765e-05,
+      "loss": 0.1907,
+      "step": 2485
+    },
+    {
+      "epoch": 1.1771530038472922,
+      "grad_norm": 1.498472809791565,
+      "learning_rate": 1.3862831842389591e-05,
+      "loss": 0.2291,
+      "step": 2486
+    },
+    {
+      "epoch": 1.1776265167209234,
+      "grad_norm": 1.148629903793335,
+      "learning_rate": 1.3858114191145126e-05,
+      "loss": 0.2093,
+      "step": 2487
+    },
+    {
+      "epoch": 1.1781000295945545,
+      "grad_norm": 1.5816822052001953,
+      "learning_rate": 1.3853395530848114e-05,
+      "loss": 0.225,
+      "step": 2488
+    },
+    {
+      "epoch": 1.1785735424681858,
+      "grad_norm": 1.4523855447769165,
+      "learning_rate": 1.384867586273268e-05,
+      "loss": 0.2078,
+      "step": 2489
+    },
+    {
+      "epoch": 1.179047055341817,
+      "grad_norm": 1.0670571327209473,
+      "learning_rate": 1.38439551880332e-05,
+      "loss": 0.2214,
+      "step": 2490
+    },
+    {
+      "epoch": 1.1795205682154484,
+      "grad_norm": 1.1799019575119019,
+      "learning_rate": 1.3839233507984324e-05,
+      "loss": 0.2186,
+      "step": 2491
+    },
+    {
+      "epoch": 1.1799940810890797,
+      "grad_norm": 1.5703297853469849,
+      "learning_rate": 1.3834510823820963e-05,
+      "loss": 0.2402,
+      "step": 2492
+    },
+    {
+      "epoch": 1.180467593962711,
+      "grad_norm": 1.9375343322753906,
+      "learning_rate": 1.3829787136778288e-05,
+      "loss": 0.2241,
+      "step": 2493
+    },
+    {
+      "epoch": 1.1809411068363422,
+      "grad_norm": 1.1045565605163574,
+      "learning_rate": 1.382506244809173e-05,
+      "loss": 0.2247,
+      "step": 2494
+    },
+    {
+      "epoch": 1.1814146197099733,
+      "grad_norm": 1.2954577207565308,
+      "learning_rate": 1.3820336758996994e-05,
+      "loss": 0.2266,
+      "step": 2495
+    },
+    {
+      "epoch": 1.1818881325836046,
+      "grad_norm": 1.3069260120391846,
+      "learning_rate": 1.3815610070730032e-05,
+      "loss": 0.2192,
+      "step": 2496
+    },
+    {
+      "epoch": 1.1823616454572359,
+      "grad_norm": 0.9491212368011475,
+      "learning_rate": 1.3810882384527067e-05,
+      "loss": 0.2241,
+      "step": 2497
+    },
+    {
+      "epoch": 1.1828351583308672,
+      "grad_norm": 1.5114809274673462,
+      "learning_rate": 1.3806153701624578e-05,
+      "loss": 0.2567,
+      "step": 2498
+    },
+    {
+      "epoch": 1.1833086712044985,
+      "grad_norm": 1.5127360820770264,
+      "learning_rate": 1.3801424023259308e-05,
+      "loss": 0.2152,
+      "step": 2499
+    },
+    {
+      "epoch": 1.1837821840781295,
+      "grad_norm": 1.0083001852035522,
+      "learning_rate": 1.3796693350668258e-05,
+      "loss": 0.2371,
+      "step": 2500
+    },
+    {
+      "epoch": 1.1842556969517608,
+      "grad_norm": 1.140708088874817,
+      "learning_rate": 1.3791961685088689e-05,
+      "loss": 0.231,
+      "step": 2501
+    },
+    {
+      "epoch": 1.184729209825392,
+      "grad_norm": 1.298230528831482,
+      "learning_rate": 1.3787229027758122e-05,
+      "loss": 0.2102,
+      "step": 2502
+    },
+    {
+      "epoch": 1.1852027226990234,
+      "grad_norm": 1.1305428743362427,
+      "learning_rate": 1.378249537991434e-05,
+      "loss": 0.2321,
+      "step": 2503
+    },
+    {
+      "epoch": 1.1856762355726547,
+      "grad_norm": 1.2196705341339111,
+      "learning_rate": 1.3777760742795381e-05,
+      "loss": 0.2243,
+      "step": 2504
+    },
+    {
+      "epoch": 1.186149748446286,
+      "grad_norm": 1.1564288139343262,
+      "learning_rate": 1.3773025117639543e-05,
+      "loss": 0.2223,
+      "step": 2505
+    },
+    {
+      "epoch": 1.1866232613199172,
+      "grad_norm": 1.8472498655319214,
+      "learning_rate": 1.3768288505685385e-05,
+      "loss": 0.199,
+      "step": 2506
+    },
+    {
+      "epoch": 1.1870967741935483,
+      "grad_norm": 1.086516261100769,
+      "learning_rate": 1.3763550908171724e-05,
+      "loss": 0.2352,
+      "step": 2507
+    },
+    {
+      "epoch": 1.1875702870671796,
+      "grad_norm": 1.063267707824707,
+      "learning_rate": 1.375881232633763e-05,
+      "loss": 0.2142,
+      "step": 2508
+    },
+    {
+      "epoch": 1.1880437999408109,
+      "grad_norm": 1.4095113277435303,
+      "learning_rate": 1.3754072761422434e-05,
+      "loss": 0.192,
+      "step": 2509
+    },
+    {
+      "epoch": 1.1885173128144422,
+      "grad_norm": 1.2273898124694824,
+      "learning_rate": 1.3749332214665723e-05,
+      "loss": 0.237,
+      "step": 2510
+    },
+    {
+      "epoch": 1.1889908256880735,
+      "grad_norm": 1.414565920829773,
+      "learning_rate": 1.374459068730734e-05,
+      "loss": 0.2165,
+      "step": 2511
+    },
+    {
+      "epoch": 1.1894643385617045,
+      "grad_norm": 1.405671238899231,
+      "learning_rate": 1.373984818058739e-05,
+      "loss": 0.2649,
+      "step": 2512
+    },
+    {
+      "epoch": 1.1899378514353358,
+      "grad_norm": 1.4026716947555542,
+      "learning_rate": 1.3735104695746225e-05,
+      "loss": 0.2252,
+      "step": 2513
+    },
+    {
+      "epoch": 1.190411364308967,
+      "grad_norm": 1.2324182987213135,
+      "learning_rate": 1.373036023402446e-05,
+      "loss": 0.2129,
+      "step": 2514
+    },
+    {
+      "epoch": 1.1908848771825984,
+      "grad_norm": 1.2538621425628662,
+      "learning_rate": 1.3725614796662962e-05,
+      "loss": 0.2241,
+      "step": 2515
+    },
+    {
+      "epoch": 1.1913583900562297,
+      "grad_norm": 1.2133036851882935,
+      "learning_rate": 1.372086838490286e-05,
+      "loss": 0.2213,
+      "step": 2516
+    },
+    {
+      "epoch": 1.191831902929861,
+      "grad_norm": 1.0370423793792725,
+      "learning_rate": 1.3716120999985527e-05,
+      "loss": 0.216,
+      "step": 2517
+    },
+    {
+      "epoch": 1.1923054158034923,
+      "grad_norm": 1.403160810470581,
+      "learning_rate": 1.3711372643152597e-05,
+      "loss": 0.2511,
+      "step": 2518
+    },
+    {
+      "epoch": 1.1927789286771233,
+      "grad_norm": 1.4306610822677612,
+      "learning_rate": 1.3706623315645953e-05,
+      "loss": 0.2194,
+      "step": 2519
+    },
+    {
+      "epoch": 1.1932524415507546,
+      "grad_norm": 0.9785909056663513,
+      "learning_rate": 1.370187301870774e-05,
+      "loss": 0.2138,
+      "step": 2520
+    },
+    {
+      "epoch": 1.193725954424386,
+      "grad_norm": 1.315924882888794,
+      "learning_rate": 1.3697121753580353e-05,
+      "loss": 0.2212,
+      "step": 2521
+    },
+    {
+      "epoch": 1.1941994672980172,
+      "grad_norm": 1.1268949508666992,
+      "learning_rate": 1.3692369521506437e-05,
+      "loss": 0.221,
+      "step": 2522
+    },
+    {
+      "epoch": 1.1946729801716485,
+      "grad_norm": 2.0519661903381348,
+      "learning_rate": 1.3687616323728894e-05,
+      "loss": 0.2417,
+      "step": 2523
+    },
+    {
+      "epoch": 1.1951464930452798,
+      "grad_norm": 1.7941192388534546,
+      "learning_rate": 1.3682862161490877e-05,
+      "loss": 0.2387,
+      "step": 2524
+    },
+    {
+      "epoch": 1.195620005918911,
+      "grad_norm": 1.0652010440826416,
+      "learning_rate": 1.367810703603579e-05,
+      "loss": 0.2142,
+      "step": 2525
+    },
+    {
+      "epoch": 1.196093518792542,
+      "grad_norm": 1.2762157917022705,
+      "learning_rate": 1.3673350948607294e-05,
+      "loss": 0.2268,
+      "step": 2526
+    },
+    {
+      "epoch": 1.1965670316661734,
+      "grad_norm": 1.4592194557189941,
+      "learning_rate": 1.3668593900449292e-05,
+      "loss": 0.2157,
+      "step": 2527
+    },
+    {
+      "epoch": 1.1970405445398047,
+      "grad_norm": 1.3134995698928833,
+      "learning_rate": 1.3663835892805947e-05,
+      "loss": 0.2369,
+      "step": 2528
+    },
+    {
+      "epoch": 1.197514057413436,
+      "grad_norm": 2.157960891723633,
+      "learning_rate": 1.365907692692167e-05,
+      "loss": 0.1831,
+      "step": 2529
+    },
+    {
+      "epoch": 1.1979875702870673,
+      "grad_norm": 1.0059032440185547,
+      "learning_rate": 1.3654317004041122e-05,
+      "loss": 0.2266,
+      "step": 2530
+    },
+    {
+      "epoch": 1.1984610831606983,
+      "grad_norm": 1.2563735246658325,
+      "learning_rate": 1.3649556125409218e-05,
+      "loss": 0.2189,
+      "step": 2531
+    },
+    {
+      "epoch": 1.1989345960343296,
+      "grad_norm": 0.9727022051811218,
+      "learning_rate": 1.3644794292271116e-05,
+      "loss": 0.2255,
+      "step": 2532
+    },
+    {
+      "epoch": 1.199408108907961,
+      "grad_norm": 0.8149449825286865,
+      "learning_rate": 1.3640031505872225e-05,
+      "loss": 0.2023,
+      "step": 2533
+    },
+    {
+      "epoch": 1.1998816217815922,
+      "grad_norm": 1.0001932382583618,
+      "learning_rate": 1.3635267767458213e-05,
+      "loss": 0.2004,
+      "step": 2534
+    },
+    {
+      "epoch": 1.2003551346552235,
+      "grad_norm": 1.593064546585083,
+      "learning_rate": 1.3630503078274988e-05,
+      "loss": 0.2215,
+      "step": 2535
+    },
+    {
+      "epoch": 1.2008286475288548,
+      "grad_norm": 1.3267443180084229,
+      "learning_rate": 1.3625737439568705e-05,
+      "loss": 0.2162,
+      "step": 2536
+    },
+    {
+      "epoch": 1.201302160402486,
+      "grad_norm": 1.1851780414581299,
+      "learning_rate": 1.3620970852585774e-05,
+      "loss": 0.2121,
+      "step": 2537
+    },
+    {
+      "epoch": 1.2017756732761171,
+      "grad_norm": 0.9846872687339783,
+      "learning_rate": 1.3616203318572847e-05,
+      "loss": 0.2268,
+      "step": 2538
+    },
+    {
+      "epoch": 1.2022491861497484,
+      "grad_norm": 1.268949031829834,
+      "learning_rate": 1.3611434838776827e-05,
+      "loss": 0.2392,
+      "step": 2539
+    },
+    {
+      "epoch": 1.2027226990233797,
+      "grad_norm": 1.274588704109192,
+      "learning_rate": 1.3606665414444868e-05,
+      "loss": 0.2315,
+      "step": 2540
+    },
+    {
+      "epoch": 1.203196211897011,
+      "grad_norm": 2.332651376724243,
+      "learning_rate": 1.3601895046824363e-05,
+      "loss": 0.2311,
+      "step": 2541
+    },
+    {
+      "epoch": 1.2036697247706423,
+      "grad_norm": 2.8526484966278076,
+      "learning_rate": 1.3597123737162955e-05,
+      "loss": 0.2059,
+      "step": 2542
+    },
+    {
+      "epoch": 1.2041432376442733,
+      "grad_norm": 1.6713618040084839,
+      "learning_rate": 1.359235148670854e-05,
+      "loss": 0.2153,
+      "step": 2543
+    },
+    {
+      "epoch": 1.2046167505179046,
+      "grad_norm": 1.6479053497314453,
+      "learning_rate": 1.3587578296709248e-05,
+      "loss": 0.2273,
+      "step": 2544
+    },
+    {
+      "epoch": 1.205090263391536,
+      "grad_norm": 1.376947283744812,
+      "learning_rate": 1.358280416841346e-05,
+      "loss": 0.2047,
+      "step": 2545
+    },
+    {
+      "epoch": 1.2055637762651672,
+      "grad_norm": 1.5050996541976929,
+      "learning_rate": 1.3578029103069805e-05,
+      "loss": 0.211,
+      "step": 2546
+    },
+    {
+      "epoch": 1.2060372891387985,
+      "grad_norm": 1.4506083726882935,
+      "learning_rate": 1.357325310192716e-05,
+      "loss": 0.2179,
+      "step": 2547
+    },
+    {
+      "epoch": 1.2065108020124298,
+      "grad_norm": 1.4101526737213135,
+      "learning_rate": 1.3568476166234634e-05,
+      "loss": 0.2305,
+      "step": 2548
+    },
+    {
+      "epoch": 1.206984314886061,
+      "grad_norm": 2.1561319828033447,
+      "learning_rate": 1.3563698297241596e-05,
+      "loss": 0.2227,
+      "step": 2549
+    },
+    {
+      "epoch": 1.2074578277596921,
+      "grad_norm": 1.0047372579574585,
+      "learning_rate": 1.3558919496197645e-05,
+      "loss": 0.2211,
+      "step": 2550
+    },
+    {
+      "epoch": 1.2079313406333234,
+      "grad_norm": 1.103215217590332,
+      "learning_rate": 1.3554139764352634e-05,
+      "loss": 0.2091,
+      "step": 2551
+    },
+    {
+      "epoch": 1.2084048535069547,
+      "grad_norm": 2.47819185256958,
+      "learning_rate": 1.3549359102956655e-05,
+      "loss": 0.231,
+      "step": 2552
+    },
+    {
+      "epoch": 1.208878366380586,
+      "grad_norm": 0.9107365608215332,
+      "learning_rate": 1.3544577513260046e-05,
+      "loss": 0.2048,
+      "step": 2553
+    },
+    {
+      "epoch": 1.2093518792542173,
+      "grad_norm": 1.520993947982788,
+      "learning_rate": 1.353979499651338e-05,
+      "loss": 0.2213,
+      "step": 2554
+    },
+    {
+      "epoch": 1.2098253921278486,
+      "grad_norm": 0.9904034733772278,
+      "learning_rate": 1.3535011553967486e-05,
+      "loss": 0.2281,
+      "step": 2555
+    },
+    {
+      "epoch": 1.2102989050014799,
+      "grad_norm": 1.3577964305877686,
+      "learning_rate": 1.3530227186873419e-05,
+      "loss": 0.2299,
+      "step": 2556
+    },
+    {
+      "epoch": 1.210772417875111,
+      "grad_norm": 1.2297004461288452,
+      "learning_rate": 1.352544189648249e-05,
+      "loss": 0.1907,
+      "step": 2557
+    },
+    {
+      "epoch": 1.2112459307487422,
+      "grad_norm": 1.2518497705459595,
+      "learning_rate": 1.3520655684046242e-05,
+      "loss": 0.2217,
+      "step": 2558
+    },
+    {
+      "epoch": 1.2117194436223735,
+      "grad_norm": 1.2009005546569824,
+      "learning_rate": 1.3515868550816467e-05,
+      "loss": 0.1946,
+      "step": 2559
+    },
+    {
+      "epoch": 1.2121929564960048,
+      "grad_norm": 1.5080748796463013,
+      "learning_rate": 1.3511080498045189e-05,
+      "loss": 0.2365,
+      "step": 2560
+    },
+    {
+      "epoch": 1.212666469369636,
+      "grad_norm": 1.2145692110061646,
+      "learning_rate": 1.3506291526984679e-05,
+      "loss": 0.2444,
+      "step": 2561
+    },
+    {
+      "epoch": 1.2131399822432671,
+      "grad_norm": 1.701706886291504,
+      "learning_rate": 1.3501501638887447e-05,
+      "loss": 0.2327,
+      "step": 2562
+    },
+    {
+      "epoch": 1.2136134951168984,
+      "grad_norm": 3.05845308303833,
+      "learning_rate": 1.349671083500624e-05,
+      "loss": 0.2352,
+      "step": 2563
+    },
+    {
+      "epoch": 1.2140870079905297,
+      "grad_norm": 1.9371395111083984,
+      "learning_rate": 1.3491919116594045e-05,
+      "loss": 0.2098,
+      "step": 2564
+    },
+    {
+      "epoch": 1.214560520864161,
+      "grad_norm": 1.3178801536560059,
+      "learning_rate": 1.3487126484904093e-05,
+      "loss": 0.2176,
+      "step": 2565
+    },
+    {
+      "epoch": 1.2150340337377923,
+      "grad_norm": 1.2978839874267578,
+      "learning_rate": 1.3482332941189854e-05,
+      "loss": 0.2286,
+      "step": 2566
+    },
+    {
+      "epoch": 1.2155075466114236,
+      "grad_norm": 1.7363958358764648,
+      "learning_rate": 1.3477538486705025e-05,
+      "loss": 0.2034,
+      "step": 2567
+    },
+    {
+      "epoch": 1.2159810594850549,
+      "grad_norm": 2.09183669090271,
+      "learning_rate": 1.3472743122703552e-05,
+      "loss": 0.2326,
+      "step": 2568
+    },
+    {
+      "epoch": 1.216454572358686,
+      "grad_norm": 0.9742308855056763,
+      "learning_rate": 1.3467946850439622e-05,
+      "loss": 0.1946,
+      "step": 2569
+    },
+    {
+      "epoch": 1.2169280852323172,
+      "grad_norm": 1.709175705909729,
+      "learning_rate": 1.3463149671167646e-05,
+      "loss": 0.2307,
+      "step": 2570
+    },
+    {
+      "epoch": 1.2174015981059485,
+      "grad_norm": 1.1259140968322754,
+      "learning_rate": 1.3458351586142284e-05,
+      "loss": 0.2385,
+      "step": 2571
+    },
+    {
+      "epoch": 1.2178751109795798,
+      "grad_norm": 1.0130417346954346,
+      "learning_rate": 1.3453552596618427e-05,
+      "loss": 0.2264,
+      "step": 2572
+    },
+    {
+      "epoch": 1.218348623853211,
+      "grad_norm": 1.774648904800415,
+      "learning_rate": 1.3448752703851207e-05,
+      "loss": 0.2396,
+      "step": 2573
+    },
+    {
+      "epoch": 1.2188221367268421,
+      "grad_norm": 2.2298903465270996,
+      "learning_rate": 1.3443951909095984e-05,
+      "loss": 0.2075,
+      "step": 2574
+    },
+    {
+      "epoch": 1.2192956496004734,
+      "grad_norm": 1.3856366872787476,
+      "learning_rate": 1.3439150213608367e-05,
+      "loss": 0.2104,
+      "step": 2575
+    },
+    {
+      "epoch": 1.2197691624741047,
+      "grad_norm": 1.630552887916565,
+      "learning_rate": 1.343434761864419e-05,
+      "loss": 0.2092,
+      "step": 2576
+    },
+    {
+      "epoch": 1.220242675347736,
+      "grad_norm": 1.0852291584014893,
+      "learning_rate": 1.3429544125459524e-05,
+      "loss": 0.2098,
+      "step": 2577
+    },
+    {
+      "epoch": 1.2207161882213673,
+      "grad_norm": 1.034583568572998,
+      "learning_rate": 1.342473973531068e-05,
+      "loss": 0.2085,
+      "step": 2578
+    },
+    {
+      "epoch": 1.2211897010949986,
+      "grad_norm": 1.044459581375122,
+      "learning_rate": 1.3419934449454194e-05,
+      "loss": 0.2173,
+      "step": 2579
+    },
+    {
+      "epoch": 1.2216632139686299,
+      "grad_norm": 1.3877681493759155,
+      "learning_rate": 1.3415128269146846e-05,
+      "loss": 0.2386,
+      "step": 2580
+    },
+    {
+      "epoch": 1.222136726842261,
+      "grad_norm": 1.530917763710022,
+      "learning_rate": 1.3410321195645648e-05,
+      "loss": 0.2139,
+      "step": 2581
+    },
+    {
+      "epoch": 1.2226102397158922,
+      "grad_norm": 1.024557113647461,
+      "learning_rate": 1.3405513230207839e-05,
+      "loss": 0.236,
+      "step": 2582
+    },
+    {
+      "epoch": 1.2230837525895235,
+      "grad_norm": 1.3001816272735596,
+      "learning_rate": 1.3400704374090898e-05,
+      "loss": 0.2451,
+      "step": 2583
+    },
+    {
+      "epoch": 1.2235572654631548,
+      "grad_norm": 1.623937964439392,
+      "learning_rate": 1.3395894628552535e-05,
+      "loss": 0.2361,
+      "step": 2584
+    },
+    {
+      "epoch": 1.224030778336786,
+      "grad_norm": 1.2357534170150757,
+      "learning_rate": 1.3391083994850696e-05,
+      "loss": 0.2363,
+      "step": 2585
+    },
+    {
+      "epoch": 1.2245042912104174,
+      "grad_norm": 1.3607426881790161,
+      "learning_rate": 1.3386272474243546e-05,
+      "loss": 0.2231,
+      "step": 2586
+    },
+    {
+      "epoch": 1.2249778040840484,
+      "grad_norm": 1.3958909511566162,
+      "learning_rate": 1.3381460067989505e-05,
+      "loss": 0.2305,
+      "step": 2587
+    },
+    {
+      "epoch": 1.2254513169576797,
+      "grad_norm": 1.3651123046875,
+      "learning_rate": 1.3376646777347202e-05,
+      "loss": 0.233,
+      "step": 2588
+    },
+    {
+      "epoch": 1.225924829831311,
+      "grad_norm": 1.5809210538864136,
+      "learning_rate": 1.3371832603575509e-05,
+      "loss": 0.2273,
+      "step": 2589
+    },
+    {
+      "epoch": 1.2263983427049423,
+      "grad_norm": 1.3318142890930176,
+      "learning_rate": 1.3367017547933529e-05,
+      "loss": 0.2202,
+      "step": 2590
+    },
+    {
+      "epoch": 1.2268718555785736,
+      "grad_norm": 2.5720956325531006,
+      "learning_rate": 1.3362201611680587e-05,
+      "loss": 0.2073,
+      "step": 2591
+    },
+    {
+      "epoch": 1.2273453684522049,
+      "grad_norm": 1.4822810888290405,
+      "learning_rate": 1.3357384796076253e-05,
+      "loss": 0.2111,
+      "step": 2592
+    },
+    {
+      "epoch": 1.227818881325836,
+      "grad_norm": 1.8587872982025146,
+      "learning_rate": 1.3352567102380315e-05,
+      "loss": 0.2369,
+      "step": 2593
+    },
+    {
+      "epoch": 1.2282923941994672,
+      "grad_norm": 0.9261167645454407,
+      "learning_rate": 1.3347748531852791e-05,
+      "loss": 0.2318,
+      "step": 2594
+    },
+    {
+      "epoch": 1.2287659070730985,
+      "grad_norm": 1.5869637727737427,
+      "learning_rate": 1.3342929085753939e-05,
+      "loss": 0.2472,
+      "step": 2595
+    },
+    {
+      "epoch": 1.2292394199467298,
+      "grad_norm": 0.9304991364479065,
+      "learning_rate": 1.3338108765344233e-05,
+      "loss": 0.2046,
+      "step": 2596
+    },
+    {
+      "epoch": 1.229712932820361,
+      "grad_norm": 1.8789327144622803,
+      "learning_rate": 1.3333287571884382e-05,
+      "loss": 0.2467,
+      "step": 2597
+    },
+    {
+      "epoch": 1.2301864456939924,
+      "grad_norm": 3.1216373443603516,
+      "learning_rate": 1.3328465506635324e-05,
+      "loss": 0.2007,
+      "step": 2598
+    },
+    {
+      "epoch": 1.2306599585676237,
+      "grad_norm": 1.2716069221496582,
+      "learning_rate": 1.3323642570858221e-05,
+      "loss": 0.2083,
+      "step": 2599
+    },
+    {
+      "epoch": 1.2311334714412547,
+      "grad_norm": 1.1587867736816406,
+      "learning_rate": 1.331881876581447e-05,
+      "loss": 0.1932,
+      "step": 2600
+    },
+    {
+      "epoch": 1.231606984314886,
+      "grad_norm": 1.1460086107254028,
+      "learning_rate": 1.3313994092765688e-05,
+      "loss": 0.2137,
+      "step": 2601
+    },
+    {
+      "epoch": 1.2320804971885173,
+      "grad_norm": 1.066895604133606,
+      "learning_rate": 1.3309168552973718e-05,
+      "loss": 0.2257,
+      "step": 2602
+    },
+    {
+      "epoch": 1.2325540100621486,
+      "grad_norm": 1.1454565525054932,
+      "learning_rate": 1.3304342147700642e-05,
+      "loss": 0.2358,
+      "step": 2603
+    },
+    {
+      "epoch": 1.2330275229357799,
+      "grad_norm": 1.030795693397522,
+      "learning_rate": 1.3299514878208752e-05,
+      "loss": 0.2263,
+      "step": 2604
+    },
+    {
+      "epoch": 1.233501035809411,
+      "grad_norm": 1.6368428468704224,
+      "learning_rate": 1.3294686745760578e-05,
+      "loss": 0.2432,
+      "step": 2605
+    },
+    {
+      "epoch": 1.2339745486830422,
+      "grad_norm": 1.6315193176269531,
+      "learning_rate": 1.328985775161887e-05,
+      "loss": 0.2042,
+      "step": 2606
+    },
+    {
+      "epoch": 1.2344480615566735,
+      "grad_norm": 1.179657220840454,
+      "learning_rate": 1.3285027897046603e-05,
+      "loss": 0.2091,
+      "step": 2607
+    },
+    {
+      "epoch": 1.2349215744303048,
+      "grad_norm": 1.292107105255127,
+      "learning_rate": 1.3280197183306982e-05,
+      "loss": 0.2323,
+      "step": 2608
+    },
+    {
+      "epoch": 1.235395087303936,
+      "grad_norm": 1.0263645648956299,
+      "learning_rate": 1.3275365611663432e-05,
+      "loss": 0.2211,
+      "step": 2609
+    },
+    {
+      "epoch": 1.2358686001775674,
+      "grad_norm": 1.5201125144958496,
+      "learning_rate": 1.3270533183379604e-05,
+      "loss": 0.2254,
+      "step": 2610
+    },
+    {
+      "epoch": 1.2363421130511987,
+      "grad_norm": 1.2638511657714844,
+      "learning_rate": 1.3265699899719374e-05,
+      "loss": 0.2249,
+      "step": 2611
+    },
+    {
+      "epoch": 1.2368156259248297,
+      "grad_norm": 1.429612398147583,
+      "learning_rate": 1.3260865761946837e-05,
+      "loss": 0.2316,
+      "step": 2612
+    },
+    {
+      "epoch": 1.237289138798461,
+      "grad_norm": 1.2139184474945068,
+      "learning_rate": 1.3256030771326325e-05,
+      "loss": 0.2282,
+      "step": 2613
+    },
+    {
+      "epoch": 1.2377626516720923,
+      "grad_norm": 2.151301383972168,
+      "learning_rate": 1.325119492912237e-05,
+      "loss": 0.2462,
+      "step": 2614
+    },
+    {
+      "epoch": 1.2382361645457236,
+      "grad_norm": 1.421278715133667,
+      "learning_rate": 1.3246358236599747e-05,
+      "loss": 0.2238,
+      "step": 2615
+    },
+    {
+      "epoch": 1.238709677419355,
+      "grad_norm": 1.212721824645996,
+      "learning_rate": 1.3241520695023449e-05,
+      "loss": 0.2312,
+      "step": 2616
+    },
+    {
+      "epoch": 1.2391831902929862,
+      "grad_norm": 1.3026739358901978,
+      "learning_rate": 1.3236682305658682e-05,
+      "loss": 0.2466,
+      "step": 2617
+    },
+    {
+      "epoch": 1.2396567031666172,
+      "grad_norm": 1.0858564376831055,
+      "learning_rate": 1.3231843069770882e-05,
+      "loss": 0.2156,
+      "step": 2618
+    },
+    {
+      "epoch": 1.2401302160402485,
+      "grad_norm": 1.2696760892868042,
+      "learning_rate": 1.3227002988625705e-05,
+      "loss": 0.2361,
+      "step": 2619
+    },
+    {
+      "epoch": 1.2406037289138798,
+      "grad_norm": 1.2995176315307617,
+      "learning_rate": 1.322216206348903e-05,
+      "loss": 0.2306,
+      "step": 2620
+    },
+    {
+      "epoch": 1.241077241787511,
+      "grad_norm": 1.480022668838501,
+      "learning_rate": 1.3217320295626953e-05,
+      "loss": 0.2266,
+      "step": 2621
+    },
+    {
+      "epoch": 1.2415507546611424,
+      "grad_norm": 1.1644608974456787,
+      "learning_rate": 1.3212477686305789e-05,
+      "loss": 0.2018,
+      "step": 2622
+    },
+    {
+      "epoch": 1.2420242675347737,
+      "grad_norm": 0.9980354309082031,
+      "learning_rate": 1.3207634236792077e-05,
+      "loss": 0.2019,
+      "step": 2623
+    },
+    {
+      "epoch": 1.2424977804084048,
+      "grad_norm": 1.6177034378051758,
+      "learning_rate": 1.3202789948352577e-05,
+      "loss": 0.2148,
+      "step": 2624
+    },
+    {
+      "epoch": 1.242971293282036,
+      "grad_norm": 1.0906509160995483,
+      "learning_rate": 1.319794482225426e-05,
+      "loss": 0.2256,
+      "step": 2625
+    },
+    {
+      "epoch": 1.2434448061556673,
+      "grad_norm": 1.3726736307144165,
+      "learning_rate": 1.3193098859764329e-05,
+      "loss": 0.2089,
+      "step": 2626
+    },
+    {
+      "epoch": 1.2439183190292986,
+      "grad_norm": 1.480689525604248,
+      "learning_rate": 1.3188252062150195e-05,
+      "loss": 0.2264,
+      "step": 2627
+    },
+    {
+      "epoch": 1.24439183190293,
+      "grad_norm": 1.417734980583191,
+      "learning_rate": 1.3183404430679492e-05,
+      "loss": 0.2317,
+      "step": 2628
+    },
+    {
+      "epoch": 1.2448653447765612,
+      "grad_norm": 1.1104220151901245,
+      "learning_rate": 1.3178555966620067e-05,
+      "loss": 0.2314,
+      "step": 2629
+    },
+    {
+      "epoch": 1.2453388576501925,
+      "grad_norm": 1.307004690170288,
+      "learning_rate": 1.3173706671239999e-05,
+      "loss": 0.23,
+      "step": 2630
+    },
+    {
+      "epoch": 1.2458123705238235,
+      "grad_norm": 1.3435912132263184,
+      "learning_rate": 1.3168856545807565e-05,
+      "loss": 0.2255,
+      "step": 2631
+    },
+    {
+      "epoch": 1.2462858833974548,
+      "grad_norm": 1.3050034046173096,
+      "learning_rate": 1.3164005591591273e-05,
+      "loss": 0.2399,
+      "step": 2632
+    },
+    {
+      "epoch": 1.2467593962710861,
+      "grad_norm": 1.0008726119995117,
+      "learning_rate": 1.3159153809859842e-05,
+      "loss": 0.2318,
+      "step": 2633
+    },
+    {
+      "epoch": 1.2472329091447174,
+      "grad_norm": 1.289227843284607,
+      "learning_rate": 1.3154301201882209e-05,
+      "loss": 0.2061,
+      "step": 2634
+    },
+    {
+      "epoch": 1.2477064220183487,
+      "grad_norm": 1.4629969596862793,
+      "learning_rate": 1.3149447768927526e-05,
+      "loss": 0.2023,
+      "step": 2635
+    },
+    {
+      "epoch": 1.2481799348919798,
+      "grad_norm": 1.1824212074279785,
+      "learning_rate": 1.3144593512265162e-05,
+      "loss": 0.2296,
+      "step": 2636
+    },
+    {
+      "epoch": 1.248653447765611,
+      "grad_norm": 1.5664751529693604,
+      "learning_rate": 1.3139738433164704e-05,
+      "loss": 0.2159,
+      "step": 2637
+    },
+    {
+      "epoch": 1.2491269606392423,
+      "grad_norm": 1.077606201171875,
+      "learning_rate": 1.3134882532895945e-05,
+      "loss": 0.2186,
+      "step": 2638
+    },
+    {
+      "epoch": 1.2496004735128736,
+      "grad_norm": 1.1679469347000122,
+      "learning_rate": 1.3130025812728904e-05,
+      "loss": 0.2338,
+      "step": 2639
+    },
+    {
+      "epoch": 1.250073986386505,
+      "grad_norm": 1.1365487575531006,
+      "learning_rate": 1.3125168273933811e-05,
+      "loss": 0.235,
+      "step": 2640
+    },
+    {
+      "epoch": 1.2505474992601362,
+      "grad_norm": 0.9639025330543518,
+      "learning_rate": 1.31203099177811e-05,
+      "loss": 0.2164,
+      "step": 2641
+    },
+    {
+      "epoch": 1.2510210121337675,
+      "grad_norm": 1.8010168075561523,
+      "learning_rate": 1.3115450745541434e-05,
+      "loss": 0.2062,
+      "step": 2642
+    },
+    {
+      "epoch": 1.2514945250073986,
+      "grad_norm": 1.1365344524383545,
+      "learning_rate": 1.311059075848568e-05,
+      "loss": 0.1904,
+      "step": 2643
+    },
+    {
+      "epoch": 1.2519680378810298,
+      "grad_norm": 2.163564682006836,
+      "learning_rate": 1.3105729957884923e-05,
+      "loss": 0.2219,
+      "step": 2644
+    },
+    {
+      "epoch": 1.2524415507546611,
+      "grad_norm": 1.4166170358657837,
+      "learning_rate": 1.3100868345010454e-05,
+      "loss": 0.2137,
+      "step": 2645
+    },
+    {
+      "epoch": 1.2529150636282924,
+      "grad_norm": 1.539779543876648,
+      "learning_rate": 1.3096005921133785e-05,
+      "loss": 0.2185,
+      "step": 2646
+    },
+    {
+      "epoch": 1.2533885765019237,
+      "grad_norm": 2.17313289642334,
+      "learning_rate": 1.3091142687526637e-05,
+      "loss": 0.2425,
+      "step": 2647
+    },
+    {
+      "epoch": 1.2538620893755548,
+      "grad_norm": 2.488083600997925,
+      "learning_rate": 1.3086278645460939e-05,
+      "loss": 0.2248,
+      "step": 2648
+    },
+    {
+      "epoch": 1.2543356022491863,
+      "grad_norm": 1.1161004304885864,
+      "learning_rate": 1.3081413796208835e-05,
+      "loss": 0.2153,
+      "step": 2649
+    },
+    {
+      "epoch": 1.2548091151228173,
+      "grad_norm": 1.0800951719284058,
+      "learning_rate": 1.307654814104268e-05,
+      "loss": 0.2141,
+      "step": 2650
+    },
+    {
+      "epoch": 1.2552826279964486,
+      "grad_norm": 1.1490174531936646,
+      "learning_rate": 1.3071681681235039e-05,
+      "loss": 0.2563,
+      "step": 2651
+    },
+    {
+      "epoch": 1.25575614087008,
+      "grad_norm": 1.0029891729354858,
+      "learning_rate": 1.3066814418058685e-05,
+      "loss": 0.2291,
+      "step": 2652
+    },
+    {
+      "epoch": 1.2562296537437112,
+      "grad_norm": 1.9102951288223267,
+      "learning_rate": 1.3061946352786607e-05,
+      "loss": 0.2197,
+      "step": 2653
+    },
+    {
+      "epoch": 1.2567031666173425,
+      "grad_norm": 1.408528208732605,
+      "learning_rate": 1.3057077486692e-05,
+      "loss": 0.2255,
+      "step": 2654
+    },
+    {
+      "epoch": 1.2571766794909736,
+      "grad_norm": 1.8932093381881714,
+      "learning_rate": 1.3052207821048268e-05,
+      "loss": 0.2058,
+      "step": 2655
+    },
+    {
+      "epoch": 1.2576501923646048,
+      "grad_norm": 1.3944228887557983,
+      "learning_rate": 1.3047337357129025e-05,
+      "loss": 0.2077,
+      "step": 2656
+    },
+    {
+      "epoch": 1.2581237052382361,
+      "grad_norm": 1.0805628299713135,
+      "learning_rate": 1.3042466096208099e-05,
+      "loss": 0.2436,
+      "step": 2657
+    },
+    {
+      "epoch": 1.2585972181118674,
+      "grad_norm": 1.1540179252624512,
+      "learning_rate": 1.3037594039559514e-05,
+      "loss": 0.2202,
+      "step": 2658
+    },
+    {
+      "epoch": 1.2590707309854987,
+      "grad_norm": 1.449994683265686,
+      "learning_rate": 1.303272118845751e-05,
+      "loss": 0.227,
+      "step": 2659
+    },
+    {
+      "epoch": 1.25954424385913,
+      "grad_norm": 1.6480882167816162,
+      "learning_rate": 1.3027847544176537e-05,
+      "loss": 0.206,
+      "step": 2660
+    },
+    {
+      "epoch": 1.2600177567327613,
+      "grad_norm": 1.056485891342163,
+      "learning_rate": 1.3022973107991251e-05,
+      "loss": 0.2193,
+      "step": 2661
+    },
+    {
+      "epoch": 1.2604912696063924,
+      "grad_norm": 1.3506393432617188,
+      "learning_rate": 1.301809788117651e-05,
+      "loss": 0.2331,
+      "step": 2662
+    },
+    {
+      "epoch": 1.2609647824800236,
+      "grad_norm": 1.8271267414093018,
+      "learning_rate": 1.3013221865007385e-05,
+      "loss": 0.2206,
+      "step": 2663
+    },
+    {
+      "epoch": 1.261438295353655,
+      "grad_norm": 1.0800082683563232,
+      "learning_rate": 1.3008345060759149e-05,
+      "loss": 0.2127,
+      "step": 2664
+    },
+    {
+      "epoch": 1.2619118082272862,
+      "grad_norm": 1.0898518562316895,
+      "learning_rate": 1.3003467469707287e-05,
+      "loss": 0.2179,
+      "step": 2665
+    },
+    {
+      "epoch": 1.2623853211009175,
+      "grad_norm": 1.098704218864441,
+      "learning_rate": 1.299858909312748e-05,
+      "loss": 0.2281,
+      "step": 2666
+    },
+    {
+      "epoch": 1.2628588339745486,
+      "grad_norm": 1.6652623414993286,
+      "learning_rate": 1.2993709932295628e-05,
+      "loss": 0.2241,
+      "step": 2667
+    },
+    {
+      "epoch": 1.2633323468481799,
+      "grad_norm": 1.2261773347854614,
+      "learning_rate": 1.2988829988487822e-05,
+      "loss": 0.2483,
+      "step": 2668
+    },
+    {
+      "epoch": 1.2638058597218111,
+      "grad_norm": 0.9692365527153015,
+      "learning_rate": 1.2983949262980362e-05,
+      "loss": 0.2104,
+      "step": 2669
+    },
+    {
+      "epoch": 1.2642793725954424,
+      "grad_norm": 1.218557357788086,
+      "learning_rate": 1.2979067757049763e-05,
+      "loss": 0.241,
+      "step": 2670
+    },
+    {
+      "epoch": 1.2647528854690737,
+      "grad_norm": 1.0587778091430664,
+      "learning_rate": 1.297418547197273e-05,
+      "loss": 0.2213,
+      "step": 2671
+    },
+    {
+      "epoch": 1.265226398342705,
+      "grad_norm": 2.7871766090393066,
+      "learning_rate": 1.2969302409026181e-05,
+      "loss": 0.22,
+      "step": 2672
+    },
+    {
+      "epoch": 1.2656999112163363,
+      "grad_norm": 1.0976502895355225,
+      "learning_rate": 1.2964418569487232e-05,
+      "loss": 0.2031,
+      "step": 2673
+    },
+    {
+      "epoch": 1.2661734240899674,
+      "grad_norm": 1.201554298400879,
+      "learning_rate": 1.2959533954633205e-05,
+      "loss": 0.2362,
+      "step": 2674
+    },
+    {
+      "epoch": 1.2666469369635986,
+      "grad_norm": 1.15542733669281,
+      "learning_rate": 1.2954648565741623e-05,
+      "loss": 0.2212,
+      "step": 2675
+    },
+    {
+      "epoch": 1.26712044983723,
+      "grad_norm": 1.641326665878296,
+      "learning_rate": 1.2949762404090214e-05,
+      "loss": 0.2476,
+      "step": 2676
+    },
+    {
+      "epoch": 1.2675939627108612,
+      "grad_norm": 1.6406174898147583,
+      "learning_rate": 1.2944875470956902e-05,
+      "loss": 0.2303,
+      "step": 2677
+    },
+    {
+      "epoch": 1.2680674755844925,
+      "grad_norm": 1.359363079071045,
+      "learning_rate": 1.2939987767619821e-05,
+      "loss": 0.2267,
+      "step": 2678
+    },
+    {
+      "epoch": 1.2685409884581236,
+      "grad_norm": 1.6398720741271973,
+      "learning_rate": 1.2935099295357304e-05,
+      "loss": 0.2117,
+      "step": 2679
+    },
+    {
+      "epoch": 1.269014501331755,
+      "grad_norm": 1.0096100568771362,
+      "learning_rate": 1.2930210055447879e-05,
+      "loss": 0.2299,
+      "step": 2680
+    },
+    {
+      "epoch": 1.2694880142053862,
+      "grad_norm": 1.2073965072631836,
+      "learning_rate": 1.2925320049170282e-05,
+      "loss": 0.2236,
+      "step": 2681
+    },
+    {
+      "epoch": 1.2699615270790174,
+      "grad_norm": 2.227213144302368,
+      "learning_rate": 1.292042927780345e-05,
+      "loss": 0.2152,
+      "step": 2682
+    },
+    {
+      "epoch": 1.2704350399526487,
+      "grad_norm": 2.0689918994903564,
+      "learning_rate": 1.2915537742626512e-05,
+      "loss": 0.2129,
+      "step": 2683
+    },
+    {
+      "epoch": 1.27090855282628,
+      "grad_norm": 2.3078198432922363,
+      "learning_rate": 1.2910645444918809e-05,
+      "loss": 0.2196,
+      "step": 2684
+    },
+    {
+      "epoch": 1.2713820656999113,
+      "grad_norm": 1.536636233329773,
+      "learning_rate": 1.2905752385959863e-05,
+      "loss": 0.2474,
+      "step": 2685
+    },
+    {
+      "epoch": 1.2718555785735424,
+      "grad_norm": 1.6504733562469482,
+      "learning_rate": 1.2900858567029416e-05,
+      "loss": 0.2419,
+      "step": 2686
+    },
+    {
+      "epoch": 1.2723290914471737,
+      "grad_norm": 2.3625659942626953,
+      "learning_rate": 1.2895963989407394e-05,
+      "loss": 0.2319,
+      "step": 2687
+    },
+    {
+      "epoch": 1.272802604320805,
+      "grad_norm": 1.3434407711029053,
+      "learning_rate": 1.2891068654373928e-05,
+      "loss": 0.2157,
+      "step": 2688
+    },
+    {
+      "epoch": 1.2732761171944362,
+      "grad_norm": 0.922859251499176,
+      "learning_rate": 1.2886172563209348e-05,
+      "loss": 0.1955,
+      "step": 2689
+    },
+    {
+      "epoch": 1.2737496300680675,
+      "grad_norm": 2.3007314205169678,
+      "learning_rate": 1.2881275717194175e-05,
+      "loss": 0.2455,
+      "step": 2690
+    },
+    {
+      "epoch": 1.2742231429416988,
+      "grad_norm": 2.0500521659851074,
+      "learning_rate": 1.2876378117609136e-05,
+      "loss": 0.2188,
+      "step": 2691
+    },
+    {
+      "epoch": 1.27469665581533,
+      "grad_norm": 1.2446473836898804,
+      "learning_rate": 1.2871479765735151e-05,
+      "loss": 0.2171,
+      "step": 2692
+    },
+    {
+      "epoch": 1.2751701686889612,
+      "grad_norm": 1.333303451538086,
+      "learning_rate": 1.2866580662853334e-05,
+      "loss": 0.2597,
+      "step": 2693
+    },
+    {
+      "epoch": 1.2756436815625924,
+      "grad_norm": 1.047584891319275,
+      "learning_rate": 1.2861680810244998e-05,
+      "loss": 0.2296,
+      "step": 2694
+    },
+    {
+      "epoch": 1.2761171944362237,
+      "grad_norm": 1.1778916120529175,
+      "learning_rate": 1.2856780209191655e-05,
+      "loss": 0.2263,
+      "step": 2695
+    },
+    {
+      "epoch": 1.276590707309855,
+      "grad_norm": 2.103959083557129,
+      "learning_rate": 1.2851878860975007e-05,
+      "loss": 0.223,
+      "step": 2696
+    },
+    {
+      "epoch": 1.2770642201834863,
+      "grad_norm": 2.1192874908447266,
+      "learning_rate": 1.2846976766876956e-05,
+      "loss": 0.2031,
+      "step": 2697
+    },
+    {
+      "epoch": 1.2775377330571174,
+      "grad_norm": 1.320399522781372,
+      "learning_rate": 1.2842073928179594e-05,
+      "loss": 0.224,
+      "step": 2698
+    },
+    {
+      "epoch": 1.2780112459307487,
+      "grad_norm": 2.2415990829467773,
+      "learning_rate": 1.2837170346165216e-05,
+      "loss": 0.2253,
+      "step": 2699
+    },
+    {
+      "epoch": 1.27848475880438,
+      "grad_norm": 1.6020069122314453,
+      "learning_rate": 1.2832266022116304e-05,
+      "loss": 0.2348,
+      "step": 2700
+    },
+    {
+      "epoch": 1.2789582716780112,
+      "grad_norm": 1.4526188373565674,
+      "learning_rate": 1.2827360957315542e-05,
+      "loss": 0.2167,
+      "step": 2701
+    },
+    {
+      "epoch": 1.2794317845516425,
+      "grad_norm": 1.26724374294281,
+      "learning_rate": 1.2822455153045792e-05,
+      "loss": 0.2219,
+      "step": 2702
+    },
+    {
+      "epoch": 1.2799052974252738,
+      "grad_norm": 2.3599040508270264,
+      "learning_rate": 1.281754861059013e-05,
+      "loss": 0.2317,
+      "step": 2703
+    },
+    {
+      "epoch": 1.280378810298905,
+      "grad_norm": 1.8642935752868652,
+      "learning_rate": 1.2812641331231806e-05,
+      "loss": 0.2083,
+      "step": 2704
+    },
+    {
+      "epoch": 1.2808523231725362,
+      "grad_norm": 1.1180553436279297,
+      "learning_rate": 1.2807733316254275e-05,
+      "loss": 0.1894,
+      "step": 2705
+    },
+    {
+      "epoch": 1.2813258360461675,
+      "grad_norm": 1.712058663368225,
+      "learning_rate": 1.2802824566941186e-05,
+      "loss": 0.2353,
+      "step": 2706
+    },
+    {
+      "epoch": 1.2817993489197987,
+      "grad_norm": 1.1192132234573364,
+      "learning_rate": 1.279791508457637e-05,
+      "loss": 0.2343,
+      "step": 2707
+    },
+    {
+      "epoch": 1.28227286179343,
+      "grad_norm": 1.7516214847564697,
+      "learning_rate": 1.279300487044385e-05,
+      "loss": 0.2302,
+      "step": 2708
+    },
+    {
+      "epoch": 1.2827463746670613,
+      "grad_norm": 1.5895251035690308,
+      "learning_rate": 1.2788093925827858e-05,
+      "loss": 0.2144,
+      "step": 2709
+    },
+    {
+      "epoch": 1.2832198875406924,
+      "grad_norm": 1.3747384548187256,
+      "learning_rate": 1.2783182252012796e-05,
+      "loss": 0.2168,
+      "step": 2710
+    },
+    {
+      "epoch": 1.283693400414324,
+      "grad_norm": 1.1060492992401123,
+      "learning_rate": 1.2778269850283263e-05,
+      "loss": 0.2158,
+      "step": 2711
+    },
+    {
+      "epoch": 1.284166913287955,
+      "grad_norm": 0.8876561522483826,
+      "learning_rate": 1.2773356721924053e-05,
+      "loss": 0.1834,
+      "step": 2712
+    },
+    {
+      "epoch": 1.2846404261615862,
+      "grad_norm": 1.2856721878051758,
+      "learning_rate": 1.276844286822015e-05,
+      "loss": 0.2477,
+      "step": 2713
+    },
+    {
+      "epoch": 1.2851139390352175,
+      "grad_norm": 2.298548460006714,
+      "learning_rate": 1.2763528290456719e-05,
+      "loss": 0.2009,
+      "step": 2714
+    },
+    {
+      "epoch": 1.2855874519088488,
+      "grad_norm": 1.167225956916809,
+      "learning_rate": 1.2758612989919126e-05,
+      "loss": 0.2077,
+      "step": 2715
+    },
+    {
+      "epoch": 1.28606096478248,
+      "grad_norm": 1.04922354221344,
+      "learning_rate": 1.275369696789292e-05,
+      "loss": 0.2088,
+      "step": 2716
+    },
+    {
+      "epoch": 1.2865344776561112,
+      "grad_norm": 1.2038118839263916,
+      "learning_rate": 1.2748780225663835e-05,
+      "loss": 0.2267,
+      "step": 2717
+    },
+    {
+      "epoch": 1.2870079905297425,
+      "grad_norm": 1.241967797279358,
+      "learning_rate": 1.27438627645178e-05,
+      "loss": 0.2171,
+      "step": 2718
+    },
+    {
+      "epoch": 1.2874815034033738,
+      "grad_norm": 1.1782163381576538,
+      "learning_rate": 1.2738944585740933e-05,
+      "loss": 0.2194,
+      "step": 2719
+    },
+    {
+      "epoch": 1.287955016277005,
+      "grad_norm": 0.991797685623169,
+      "learning_rate": 1.2734025690619529e-05,
+      "loss": 0.2126,
+      "step": 2720
+    },
+    {
+      "epoch": 1.2884285291506363,
+      "grad_norm": 1.2841802835464478,
+      "learning_rate": 1.2729106080440081e-05,
+      "loss": 0.2145,
+      "step": 2721
+    },
+    {
+      "epoch": 1.2889020420242676,
+      "grad_norm": 1.4147202968597412,
+      "learning_rate": 1.2724185756489267e-05,
+      "loss": 0.2368,
+      "step": 2722
+    },
+    {
+      "epoch": 1.289375554897899,
+      "grad_norm": 1.1640384197235107,
+      "learning_rate": 1.271926472005395e-05,
+      "loss": 0.2179,
+      "step": 2723
+    },
+    {
+      "epoch": 1.28984906777153,
+      "grad_norm": 1.5157102346420288,
+      "learning_rate": 1.2714342972421177e-05,
+      "loss": 0.2185,
+      "step": 2724
+    },
+    {
+      "epoch": 1.2903225806451613,
+      "grad_norm": 1.4633888006210327,
+      "learning_rate": 1.2709420514878183e-05,
+      "loss": 0.2216,
+      "step": 2725
+    },
+    {
+      "epoch": 1.2907960935187925,
+      "grad_norm": 1.7542057037353516,
+      "learning_rate": 1.2704497348712397e-05,
+      "loss": 0.2188,
+      "step": 2726
+    },
+    {
+      "epoch": 1.2912696063924238,
+      "grad_norm": 1.3297491073608398,
+      "learning_rate": 1.2699573475211415e-05,
+      "loss": 0.2083,
+      "step": 2727
+    },
+    {
+      "epoch": 1.2917431192660551,
+      "grad_norm": 1.0260534286499023,
+      "learning_rate": 1.2694648895663038e-05,
+      "loss": 0.2324,
+      "step": 2728
+    },
+    {
+      "epoch": 1.2922166321396862,
+      "grad_norm": 1.0116316080093384,
+      "learning_rate": 1.2689723611355236e-05,
+      "loss": 0.2188,
+      "step": 2729
+    },
+    {
+      "epoch": 1.2926901450133175,
+      "grad_norm": 1.2791035175323486,
+      "learning_rate": 1.2684797623576173e-05,
+      "loss": 0.2286,
+      "step": 2730
+    },
+    {
+      "epoch": 1.2931636578869488,
+      "grad_norm": 2.320425033569336,
+      "learning_rate": 1.2679870933614189e-05,
+      "loss": 0.2229,
+      "step": 2731
+    },
+    {
+      "epoch": 1.29363717076058,
+      "grad_norm": 1.2323137521743774,
+      "learning_rate": 1.2674943542757819e-05,
+      "loss": 0.2404,
+      "step": 2732
+    },
+    {
+      "epoch": 1.2941106836342113,
+      "grad_norm": 1.2795385122299194,
+      "learning_rate": 1.267001545229577e-05,
+      "loss": 0.2178,
+      "step": 2733
+    },
+    {
+      "epoch": 1.2945841965078426,
+      "grad_norm": 1.4273737668991089,
+      "learning_rate": 1.2665086663516937e-05,
+      "loss": 0.2046,
+      "step": 2734
+    },
+    {
+      "epoch": 1.295057709381474,
+      "grad_norm": 1.6153572797775269,
+      "learning_rate": 1.26601571777104e-05,
+      "loss": 0.2214,
+      "step": 2735
+    },
+    {
+      "epoch": 1.295531222255105,
+      "grad_norm": 1.3819822072982788,
+      "learning_rate": 1.2655226996165415e-05,
+      "loss": 0.2056,
+      "step": 2736
+    },
+    {
+      "epoch": 1.2960047351287363,
+      "grad_norm": 1.2994166612625122,
+      "learning_rate": 1.2650296120171424e-05,
+      "loss": 0.1976,
+      "step": 2737
+    },
+    {
+      "epoch": 1.2964782480023676,
+      "grad_norm": 2.0569915771484375,
+      "learning_rate": 1.2645364551018049e-05,
+      "loss": 0.2245,
+      "step": 2738
+    },
+    {
+      "epoch": 1.2969517608759988,
+      "grad_norm": 1.7465271949768066,
+      "learning_rate": 1.2640432289995097e-05,
+      "loss": 0.2217,
+      "step": 2739
+    },
+    {
+      "epoch": 1.2974252737496301,
+      "grad_norm": 1.355583667755127,
+      "learning_rate": 1.2635499338392554e-05,
+      "loss": 0.206,
+      "step": 2740
+    },
+    {
+      "epoch": 1.2978987866232612,
+      "grad_norm": 1.9201940298080444,
+      "learning_rate": 1.2630565697500583e-05,
+      "loss": 0.2159,
+      "step": 2741
+    },
+    {
+      "epoch": 1.2983722994968927,
+      "grad_norm": 2.0526649951934814,
+      "learning_rate": 1.262563136860953e-05,
+      "loss": 0.204,
+      "step": 2742
+    },
+    {
+      "epoch": 1.2988458123705238,
+      "grad_norm": 1.1159857511520386,
+      "learning_rate": 1.2620696353009925e-05,
+      "loss": 0.2238,
+      "step": 2743
+    },
+    {
+      "epoch": 1.299319325244155,
+      "grad_norm": 1.1484869718551636,
+      "learning_rate": 1.2615760651992469e-05,
+      "loss": 0.2084,
+      "step": 2744
+    },
+    {
+      "epoch": 1.2997928381177863,
+      "grad_norm": 2.430534839630127,
+      "learning_rate": 1.261082426684805e-05,
+      "loss": 0.2258,
+      "step": 2745
+    },
+    {
+      "epoch": 1.3002663509914176,
+      "grad_norm": 1.5022352933883667,
+      "learning_rate": 1.2605887198867732e-05,
+      "loss": 0.2052,
+      "step": 2746
+    },
+    {
+      "epoch": 1.300739863865049,
+      "grad_norm": 1.0015721321105957,
+      "learning_rate": 1.2600949449342754e-05,
+      "loss": 0.2288,
+      "step": 2747
+    },
+    {
+      "epoch": 1.30121337673868,
+      "grad_norm": 1.3668203353881836,
+      "learning_rate": 1.259601101956454e-05,
+      "loss": 0.2349,
+      "step": 2748
+    },
+    {
+      "epoch": 1.3016868896123113,
+      "grad_norm": 2.162443161010742,
+      "learning_rate": 1.259107191082469e-05,
+      "loss": 0.237,
+      "step": 2749
+    },
+    {
+      "epoch": 1.3021604024859426,
+      "grad_norm": 1.806618094444275,
+      "learning_rate": 1.2586132124414978e-05,
+      "loss": 0.2307,
+      "step": 2750
+    },
+    {
+      "epoch": 1.3026339153595738,
+      "grad_norm": 1.345781922340393,
+      "learning_rate": 1.2581191661627355e-05,
+      "loss": 0.2118,
+      "step": 2751
+    },
+    {
+      "epoch": 1.3031074282332051,
+      "grad_norm": 1.1584277153015137,
+      "learning_rate": 1.2576250523753956e-05,
+      "loss": 0.2287,
+      "step": 2752
+    },
+    {
+      "epoch": 1.3035809411068364,
+      "grad_norm": 1.5245954990386963,
+      "learning_rate": 1.257130871208709e-05,
+      "loss": 0.2424,
+      "step": 2753
+    },
+    {
+      "epoch": 1.3040544539804677,
+      "grad_norm": 1.4683634042739868,
+      "learning_rate": 1.2566366227919232e-05,
+      "loss": 0.2133,
+      "step": 2754
+    },
+    {
+      "epoch": 1.3045279668540988,
+      "grad_norm": 2.778150796890259,
+      "learning_rate": 1.2561423072543043e-05,
+      "loss": 0.2346,
+      "step": 2755
+    },
+    {
+      "epoch": 1.30500147972773,
+      "grad_norm": 2.2693684101104736,
+      "learning_rate": 1.2556479247251364e-05,
+      "loss": 0.221,
+      "step": 2756
+    },
+    {
+      "epoch": 1.3054749926013614,
+      "grad_norm": 2.705554962158203,
+      "learning_rate": 1.2551534753337198e-05,
+      "loss": 0.2178,
+      "step": 2757
+    },
+    {
+      "epoch": 1.3059485054749926,
+      "grad_norm": 1.8546373844146729,
+      "learning_rate": 1.2546589592093734e-05,
+      "loss": 0.2251,
+      "step": 2758
+    },
+    {
+      "epoch": 1.306422018348624,
+      "grad_norm": 1.1411086320877075,
+      "learning_rate": 1.2541643764814328e-05,
+      "loss": 0.2077,
+      "step": 2759
+    },
+    {
+      "epoch": 1.306895531222255,
+      "grad_norm": 1.2965822219848633,
+      "learning_rate": 1.2536697272792517e-05,
+      "loss": 0.2308,
+      "step": 2760
+    },
+    {
+      "epoch": 1.3073690440958863,
+      "grad_norm": 1.3362536430358887,
+      "learning_rate": 1.2531750117322004e-05,
+      "loss": 0.2004,
+      "step": 2761
+    },
+    {
+      "epoch": 1.3078425569695176,
+      "grad_norm": 1.4326553344726562,
+      "learning_rate": 1.2526802299696674e-05,
+      "loss": 0.223,
+      "step": 2762
+    },
+    {
+      "epoch": 1.3083160698431489,
+      "grad_norm": 2.0484044551849365,
+      "learning_rate": 1.252185382121058e-05,
+      "loss": 0.2205,
+      "step": 2763
+    },
+    {
+      "epoch": 1.3087895827167801,
+      "grad_norm": 1.2032407522201538,
+      "learning_rate": 1.2516904683157947e-05,
+      "loss": 0.237,
+      "step": 2764
+    },
+    {
+      "epoch": 1.3092630955904114,
+      "grad_norm": 1.9354078769683838,
+      "learning_rate": 1.2511954886833173e-05,
+      "loss": 0.2233,
+      "step": 2765
+    },
+    {
+      "epoch": 1.3097366084640427,
+      "grad_norm": 1.3299360275268555,
+      "learning_rate": 1.2507004433530832e-05,
+      "loss": 0.2189,
+      "step": 2766
+    },
+    {
+      "epoch": 1.3102101213376738,
+      "grad_norm": 1.600842833518982,
+      "learning_rate": 1.2502053324545666e-05,
+      "loss": 0.2371,
+      "step": 2767
+    },
+    {
+      "epoch": 1.310683634211305,
+      "grad_norm": 2.048098564147949,
+      "learning_rate": 1.2497101561172593e-05,
+      "loss": 0.224,
+      "step": 2768
+    },
+    {
+      "epoch": 1.3111571470849364,
+      "grad_norm": 1.361019492149353,
+      "learning_rate": 1.2492149144706696e-05,
+      "loss": 0.2266,
+      "step": 2769
+    },
+    {
+      "epoch": 1.3116306599585676,
+      "grad_norm": 2.096219778060913,
+      "learning_rate": 1.2487196076443233e-05,
+      "loss": 0.2166,
+      "step": 2770
+    },
+    {
+      "epoch": 1.312104172832199,
+      "grad_norm": 1.8196600675582886,
+      "learning_rate": 1.2482242357677631e-05,
+      "loss": 0.1956,
+      "step": 2771
+    },
+    {
+      "epoch": 1.31257768570583,
+      "grad_norm": 0.9747202396392822,
+      "learning_rate": 1.2477287989705487e-05,
+      "loss": 0.2051,
+      "step": 2772
+    },
+    {
+      "epoch": 1.3130511985794615,
+      "grad_norm": 1.0282551050186157,
+      "learning_rate": 1.2472332973822568e-05,
+      "loss": 0.2144,
+      "step": 2773
+    },
+    {
+      "epoch": 1.3135247114530926,
+      "grad_norm": 0.8776658773422241,
+      "learning_rate": 1.2467377311324809e-05,
+      "loss": 0.2119,
+      "step": 2774
+    },
+    {
+      "epoch": 1.3139982243267239,
+      "grad_norm": 1.1889859437942505,
+      "learning_rate": 1.2462421003508318e-05,
+      "loss": 0.2086,
+      "step": 2775
+    },
+    {
+      "epoch": 1.3144717372003552,
+      "grad_norm": 2.8763959407806396,
+      "learning_rate": 1.2457464051669368e-05,
+      "loss": 0.2256,
+      "step": 2776
+    },
+    {
+      "epoch": 1.3149452500739864,
+      "grad_norm": 1.6497411727905273,
+      "learning_rate": 1.2452506457104406e-05,
+      "loss": 0.2192,
+      "step": 2777
+    },
+    {
+      "epoch": 1.3154187629476177,
+      "grad_norm": 1.5440013408660889,
+      "learning_rate": 1.2447548221110037e-05,
+      "loss": 0.2078,
+      "step": 2778
+    },
+    {
+      "epoch": 1.3158922758212488,
+      "grad_norm": 1.3232966661453247,
+      "learning_rate": 1.2442589344983043e-05,
+      "loss": 0.2108,
+      "step": 2779
+    },
+    {
+      "epoch": 1.31636578869488,
+      "grad_norm": 0.9742394685745239,
+      "learning_rate": 1.2437629830020372e-05,
+      "loss": 0.2167,
+      "step": 2780
+    },
+    {
+      "epoch": 1.3168393015685114,
+      "grad_norm": 1.2739520072937012,
+      "learning_rate": 1.2432669677519134e-05,
+      "loss": 0.2363,
+      "step": 2781
+    },
+    {
+      "epoch": 1.3173128144421427,
+      "grad_norm": 1.014618158340454,
+      "learning_rate": 1.2427708888776611e-05,
+      "loss": 0.2226,
+      "step": 2782
+    },
+    {
+      "epoch": 1.317786327315774,
+      "grad_norm": 2.0462191104888916,
+      "learning_rate": 1.2422747465090246e-05,
+      "loss": 0.2393,
+      "step": 2783
+    },
+    {
+      "epoch": 1.3182598401894052,
+      "grad_norm": 1.8126726150512695,
+      "learning_rate": 1.2417785407757657e-05,
+      "loss": 0.2352,
+      "step": 2784
+    },
+    {
+      "epoch": 1.3187333530630365,
+      "grad_norm": 1.3824222087860107,
+      "learning_rate": 1.2412822718076619e-05,
+      "loss": 0.2277,
+      "step": 2785
+    },
+    {
+      "epoch": 1.3192068659366676,
+      "grad_norm": 1.3327887058258057,
+      "learning_rate": 1.2407859397345073e-05,
+      "loss": 0.2179,
+      "step": 2786
+    },
+    {
+      "epoch": 1.3196803788102989,
+      "grad_norm": 1.1972007751464844,
+      "learning_rate": 1.2402895446861131e-05,
+      "loss": 0.2516,
+      "step": 2787
+    },
+    {
+      "epoch": 1.3201538916839302,
+      "grad_norm": 1.0115021467208862,
+      "learning_rate": 1.239793086792307e-05,
+      "loss": 0.2129,
+      "step": 2788
+    },
+    {
+      "epoch": 1.3206274045575614,
+      "grad_norm": 2.7716517448425293,
+      "learning_rate": 1.2392965661829321e-05,
+      "loss": 0.2244,
+      "step": 2789
+    },
+    {
+      "epoch": 1.3211009174311927,
+      "grad_norm": 2.3599562644958496,
+      "learning_rate": 1.238799982987849e-05,
+      "loss": 0.2132,
+      "step": 2790
+    },
+    {
+      "epoch": 1.3215744303048238,
+      "grad_norm": 2.0140914916992188,
+      "learning_rate": 1.238303337336934e-05,
+      "loss": 0.1975,
+      "step": 2791
+    },
+    {
+      "epoch": 1.322047943178455,
+      "grad_norm": 1.8659518957138062,
+      "learning_rate": 1.2378066293600801e-05,
+      "loss": 0.2089,
+      "step": 2792
+    },
+    {
+      "epoch": 1.3225214560520864,
+      "grad_norm": 2.2264151573181152,
+      "learning_rate": 1.2373098591871964e-05,
+      "loss": 0.1899,
+      "step": 2793
+    },
+    {
+      "epoch": 1.3229949689257177,
+      "grad_norm": 2.3644020557403564,
+      "learning_rate": 1.2368130269482084e-05,
+      "loss": 0.2203,
+      "step": 2794
+    },
+    {
+      "epoch": 1.323468481799349,
+      "grad_norm": 1.8323869705200195,
+      "learning_rate": 1.2363161327730577e-05,
+      "loss": 0.2109,
+      "step": 2795
+    },
+    {
+      "epoch": 1.3239419946729802,
+      "grad_norm": 1.3173034191131592,
+      "learning_rate": 1.2358191767917024e-05,
+      "loss": 0.2372,
+      "step": 2796
+    },
+    {
+      "epoch": 1.3244155075466115,
+      "grad_norm": 1.2270554304122925,
+      "learning_rate": 1.2353221591341163e-05,
+      "loss": 0.211,
+      "step": 2797
+    },
+    {
+      "epoch": 1.3248890204202426,
+      "grad_norm": 2.137927770614624,
+      "learning_rate": 1.2348250799302898e-05,
+      "loss": 0.2115,
+      "step": 2798
+    },
+    {
+      "epoch": 1.3253625332938739,
+      "grad_norm": 2.619537591934204,
+      "learning_rate": 1.2343279393102292e-05,
+      "loss": 0.2316,
+      "step": 2799
+    },
+    {
+      "epoch": 1.3258360461675052,
+      "grad_norm": 1.1064987182617188,
+      "learning_rate": 1.2338307374039564e-05,
+      "loss": 0.2106,
+      "step": 2800
+    },
+    {
+      "epoch": 1.3263095590411365,
+      "grad_norm": 1.347868800163269,
+      "learning_rate": 1.2333334743415103e-05,
+      "loss": 0.2111,
+      "step": 2801
+    },
+    {
+      "epoch": 1.3267830719147677,
+      "grad_norm": 1.0354684591293335,
+      "learning_rate": 1.2328361502529444e-05,
+      "loss": 0.221,
+      "step": 2802
+    },
+    {
+      "epoch": 1.3272565847883988,
+      "grad_norm": 1.5998934507369995,
+      "learning_rate": 1.23233876526833e-05,
+      "loss": 0.2271,
+      "step": 2803
+    },
+    {
+      "epoch": 1.3277300976620303,
+      "grad_norm": 1.4590028524398804,
+      "learning_rate": 1.231841319517753e-05,
+      "loss": 0.2234,
+      "step": 2804
+    },
+    {
+      "epoch": 1.3282036105356614,
+      "grad_norm": 2.0347890853881836,
+      "learning_rate": 1.2313438131313155e-05,
+      "loss": 0.2375,
+      "step": 2805
+    },
+    {
+      "epoch": 1.3286771234092927,
+      "grad_norm": 1.5571720600128174,
+      "learning_rate": 1.2308462462391356e-05,
+      "loss": 0.2286,
+      "step": 2806
+    },
+    {
+      "epoch": 1.329150636282924,
+      "grad_norm": 1.1657594442367554,
+      "learning_rate": 1.2303486189713466e-05,
+      "loss": 0.2066,
+      "step": 2807
+    },
+    {
+      "epoch": 1.3296241491565552,
+      "grad_norm": 1.2159301042556763,
+      "learning_rate": 1.2298509314580986e-05,
+      "loss": 0.2515,
+      "step": 2808
+    },
+    {
+      "epoch": 1.3300976620301865,
+      "grad_norm": 2.0197834968566895,
+      "learning_rate": 1.2293531838295572e-05,
+      "loss": 0.2153,
+      "step": 2809
+    },
+    {
+      "epoch": 1.3305711749038176,
+      "grad_norm": 1.6019450426101685,
+      "learning_rate": 1.228855376215903e-05,
+      "loss": 0.2245,
+      "step": 2810
+    },
+    {
+      "epoch": 1.3310446877774489,
+      "grad_norm": 1.1272090673446655,
+      "learning_rate": 1.228357508747333e-05,
+      "loss": 0.2135,
+      "step": 2811
+    },
+    {
+      "epoch": 1.3315182006510802,
+      "grad_norm": 1.550205111503601,
+      "learning_rate": 1.2278595815540595e-05,
+      "loss": 0.2498,
+      "step": 2812
+    },
+    {
+      "epoch": 1.3319917135247115,
+      "grad_norm": 1.4885501861572266,
+      "learning_rate": 1.2273615947663107e-05,
+      "loss": 0.2371,
+      "step": 2813
+    },
+    {
+      "epoch": 1.3324652263983428,
+      "grad_norm": 1.3265790939331055,
+      "learning_rate": 1.2268635485143303e-05,
+      "loss": 0.2216,
+      "step": 2814
+    },
+    {
+      "epoch": 1.332938739271974,
+      "grad_norm": 1.121368646621704,
+      "learning_rate": 1.2263654429283774e-05,
+      "loss": 0.2311,
+      "step": 2815
+    },
+    {
+      "epoch": 1.3334122521456053,
+      "grad_norm": 1.9394561052322388,
+      "learning_rate": 1.2258672781387267e-05,
+      "loss": 0.2228,
+      "step": 2816
+    },
+    {
+      "epoch": 1.3338857650192364,
+      "grad_norm": 1.1712671518325806,
+      "learning_rate": 1.2253690542756682e-05,
+      "loss": 0.2071,
+      "step": 2817
+    },
+    {
+      "epoch": 1.3343592778928677,
+      "grad_norm": 1.167645812034607,
+      "learning_rate": 1.2248707714695077e-05,
+      "loss": 0.2191,
+      "step": 2818
+    },
+    {
+      "epoch": 1.334832790766499,
+      "grad_norm": 1.0730623006820679,
+      "learning_rate": 1.2243724298505664e-05,
+      "loss": 0.2258,
+      "step": 2819
+    },
+    {
+      "epoch": 1.3353063036401303,
+      "grad_norm": 1.0903040170669556,
+      "learning_rate": 1.2238740295491804e-05,
+      "loss": 0.2109,
+      "step": 2820
+    },
+    {
+      "epoch": 1.3357798165137615,
+      "grad_norm": 0.9441567063331604,
+      "learning_rate": 1.223375570695702e-05,
+      "loss": 0.2366,
+      "step": 2821
+    },
+    {
+      "epoch": 1.3362533293873926,
+      "grad_norm": 0.9626446962356567,
+      "learning_rate": 1.222877053420498e-05,
+      "loss": 0.1965,
+      "step": 2822
+    },
+    {
+      "epoch": 1.336726842261024,
+      "grad_norm": 1.2279541492462158,
+      "learning_rate": 1.2223784778539508e-05,
+      "loss": 0.2155,
+      "step": 2823
+    },
+    {
+      "epoch": 1.3372003551346552,
+      "grad_norm": 1.0313448905944824,
+      "learning_rate": 1.2218798441264579e-05,
+      "loss": 0.2165,
+      "step": 2824
+    },
+    {
+      "epoch": 1.3376738680082865,
+      "grad_norm": 1.0775771141052246,
+      "learning_rate": 1.2213811523684325e-05,
+      "loss": 0.2141,
+      "step": 2825
+    },
+    {
+      "epoch": 1.3381473808819178,
+      "grad_norm": 0.9873069524765015,
+      "learning_rate": 1.2208824027103021e-05,
+      "loss": 0.1984,
+      "step": 2826
+    },
+    {
+      "epoch": 1.338620893755549,
+      "grad_norm": 1.219462513923645,
+      "learning_rate": 1.2203835952825105e-05,
+      "loss": 0.2372,
+      "step": 2827
+    },
+    {
+      "epoch": 1.3390944066291803,
+      "grad_norm": 1.2963061332702637,
+      "learning_rate": 1.2198847302155154e-05,
+      "loss": 0.2206,
+      "step": 2828
+    },
+    {
+      "epoch": 1.3395679195028114,
+      "grad_norm": 1.0308513641357422,
+      "learning_rate": 1.2193858076397905e-05,
+      "loss": 0.2288,
+      "step": 2829
+    },
+    {
+      "epoch": 1.3400414323764427,
+      "grad_norm": 1.5890069007873535,
+      "learning_rate": 1.2188868276858238e-05,
+      "loss": 0.2171,
+      "step": 2830
+    },
+    {
+      "epoch": 1.340514945250074,
+      "grad_norm": 2.1345863342285156,
+      "learning_rate": 1.2183877904841193e-05,
+      "loss": 0.2138,
+      "step": 2831
+    },
+    {
+      "epoch": 1.3409884581237053,
+      "grad_norm": 1.1476088762283325,
+      "learning_rate": 1.217888696165195e-05,
+      "loss": 0.2144,
+      "step": 2832
+    },
+    {
+      "epoch": 1.3414619709973366,
+      "grad_norm": 1.109824776649475,
+      "learning_rate": 1.2173895448595842e-05,
+      "loss": 0.2018,
+      "step": 2833
+    },
+    {
+      "epoch": 1.3419354838709676,
+      "grad_norm": 1.1557180881500244,
+      "learning_rate": 1.216890336697835e-05,
+      "loss": 0.2178,
+      "step": 2834
+    },
+    {
+      "epoch": 1.3424089967445991,
+      "grad_norm": 1.1168832778930664,
+      "learning_rate": 1.2163910718105108e-05,
+      "loss": 0.2174,
+      "step": 2835
+    },
+    {
+      "epoch": 1.3428825096182302,
+      "grad_norm": 1.25518000125885,
+      "learning_rate": 1.2158917503281891e-05,
+      "loss": 0.2305,
+      "step": 2836
+    },
+    {
+      "epoch": 1.3433560224918615,
+      "grad_norm": 1.8499336242675781,
+      "learning_rate": 1.215392372381463e-05,
+      "loss": 0.202,
+      "step": 2837
+    },
+    {
+      "epoch": 1.3438295353654928,
+      "grad_norm": 1.5814791917800903,
+      "learning_rate": 1.2148929381009398e-05,
+      "loss": 0.2027,
+      "step": 2838
+    },
+    {
+      "epoch": 1.344303048239124,
+      "grad_norm": 1.3987447023391724,
+      "learning_rate": 1.2143934476172416e-05,
+      "loss": 0.2271,
+      "step": 2839
+    },
+    {
+      "epoch": 1.3447765611127553,
+      "grad_norm": 1.075582504272461,
+      "learning_rate": 1.2138939010610055e-05,
+      "loss": 0.1957,
+      "step": 2840
+    },
+    {
+      "epoch": 1.3452500739863864,
+      "grad_norm": 1.5337646007537842,
+      "learning_rate": 1.2133942985628833e-05,
+      "loss": 0.2177,
+      "step": 2841
+    },
+    {
+      "epoch": 1.3457235868600177,
+      "grad_norm": 2.3917672634124756,
+      "learning_rate": 1.2128946402535409e-05,
+      "loss": 0.2185,
+      "step": 2842
+    },
+    {
+      "epoch": 1.346197099733649,
+      "grad_norm": 2.155311346054077,
+      "learning_rate": 1.2123949262636592e-05,
+      "loss": 0.221,
+      "step": 2843
+    },
+    {
+      "epoch": 1.3466706126072803,
+      "grad_norm": 0.9766375422477722,
+      "learning_rate": 1.2118951567239331e-05,
+      "loss": 0.2271,
+      "step": 2844
+    },
+    {
+      "epoch": 1.3471441254809116,
+      "grad_norm": 1.0675561428070068,
+      "learning_rate": 1.2113953317650733e-05,
+      "loss": 0.2214,
+      "step": 2845
+    },
+    {
+      "epoch": 1.3476176383545428,
+      "grad_norm": 1.6465510129928589,
+      "learning_rate": 1.2108954515178037e-05,
+      "loss": 0.2025,
+      "step": 2846
+    },
+    {
+      "epoch": 1.3480911512281741,
+      "grad_norm": 1.7982463836669922,
+      "learning_rate": 1.2103955161128635e-05,
+      "loss": 0.203,
+      "step": 2847
+    },
+    {
+      "epoch": 1.3485646641018052,
+      "grad_norm": 1.2066468000411987,
+      "learning_rate": 1.2098955256810057e-05,
+      "loss": 0.215,
+      "step": 2848
+    },
+    {
+      "epoch": 1.3490381769754365,
+      "grad_norm": 1.1641322374343872,
+      "learning_rate": 1.2093954803529981e-05,
+      "loss": 0.2276,
+      "step": 2849
+    },
+    {
+      "epoch": 1.3495116898490678,
+      "grad_norm": 1.1573890447616577,
+      "learning_rate": 1.2088953802596229e-05,
+      "loss": 0.2076,
+      "step": 2850
+    },
+    {
+      "epoch": 1.349985202722699,
+      "grad_norm": 0.9064158201217651,
+      "learning_rate": 1.208395225531676e-05,
+      "loss": 0.2124,
+      "step": 2851
+    },
+    {
+      "epoch": 1.3504587155963304,
+      "grad_norm": 1.221864104270935,
+      "learning_rate": 1.207895016299968e-05,
+      "loss": 0.203,
+      "step": 2852
+    },
+    {
+      "epoch": 1.3509322284699614,
+      "grad_norm": 1.3113806247711182,
+      "learning_rate": 1.2073947526953245e-05,
+      "loss": 0.2201,
+      "step": 2853
+    },
+    {
+      "epoch": 1.3514057413435927,
+      "grad_norm": 1.3473477363586426,
+      "learning_rate": 1.206894434848584e-05,
+      "loss": 0.2346,
+      "step": 2854
+    },
+    {
+      "epoch": 1.351879254217224,
+      "grad_norm": 1.0617995262145996,
+      "learning_rate": 1.2063940628906001e-05,
+      "loss": 0.2009,
+      "step": 2855
+    },
+    {
+      "epoch": 1.3523527670908553,
+      "grad_norm": 1.075543761253357,
+      "learning_rate": 1.20589363695224e-05,
+      "loss": 0.2371,
+      "step": 2856
+    },
+    {
+      "epoch": 1.3528262799644866,
+      "grad_norm": 1.068983554840088,
+      "learning_rate": 1.2053931571643857e-05,
+      "loss": 0.2256,
+      "step": 2857
+    },
+    {
+      "epoch": 1.3532997928381179,
+      "grad_norm": 0.9813727736473083,
+      "learning_rate": 1.2048926236579326e-05,
+      "loss": 0.2171,
+      "step": 2858
+    },
+    {
+      "epoch": 1.3537733057117491,
+      "grad_norm": 1.6709295511245728,
+      "learning_rate": 1.2043920365637904e-05,
+      "loss": 0.2091,
+      "step": 2859
+    },
+    {
+      "epoch": 1.3542468185853802,
+      "grad_norm": 0.9895414710044861,
+      "learning_rate": 1.2038913960128828e-05,
+      "loss": 0.2049,
+      "step": 2860
+    },
+    {
+      "epoch": 1.3547203314590115,
+      "grad_norm": 1.3172953128814697,
+      "learning_rate": 1.2033907021361476e-05,
+      "loss": 0.2308,
+      "step": 2861
+    },
+    {
+      "epoch": 1.3551938443326428,
+      "grad_norm": 1.719438076019287,
+      "learning_rate": 1.2028899550645362e-05,
+      "loss": 0.2239,
+      "step": 2862
+    },
+    {
+      "epoch": 1.355667357206274,
+      "grad_norm": 1.0997658967971802,
+      "learning_rate": 1.2023891549290143e-05,
+      "loss": 0.2123,
+      "step": 2863
+    },
+    {
+      "epoch": 1.3561408700799054,
+      "grad_norm": 1.2592804431915283,
+      "learning_rate": 1.2018883018605614e-05,
+      "loss": 0.2157,
+      "step": 2864
+    },
+    {
+      "epoch": 1.3566143829535364,
+      "grad_norm": 1.2449421882629395,
+      "learning_rate": 1.201387395990171e-05,
+      "loss": 0.2011,
+      "step": 2865
+    },
+    {
+      "epoch": 1.357087895827168,
+      "grad_norm": 1.2251133918762207,
+      "learning_rate": 1.20088643744885e-05,
+      "loss": 0.2227,
+      "step": 2866
+    },
+    {
+      "epoch": 1.357561408700799,
+      "grad_norm": 1.0823248624801636,
+      "learning_rate": 1.2003854263676196e-05,
+      "loss": 0.2384,
+      "step": 2867
+    },
+    {
+      "epoch": 1.3580349215744303,
+      "grad_norm": 1.333550214767456,
+      "learning_rate": 1.1998843628775136e-05,
+      "loss": 0.2299,
+      "step": 2868
+    },
+    {
+      "epoch": 1.3585084344480616,
+      "grad_norm": 1.7682206630706787,
+      "learning_rate": 1.199383247109581e-05,
+      "loss": 0.2277,
+      "step": 2869
+    },
+    {
+      "epoch": 1.3589819473216929,
+      "grad_norm": 1.3045170307159424,
+      "learning_rate": 1.1988820791948834e-05,
+      "loss": 0.2093,
+      "step": 2870
+    },
+    {
+      "epoch": 1.3594554601953242,
+      "grad_norm": 2.6720004081726074,
+      "learning_rate": 1.1983808592644967e-05,
+      "loss": 0.2414,
+      "step": 2871
+    },
+    {
+      "epoch": 1.3599289730689552,
+      "grad_norm": 0.9892565608024597,
+      "learning_rate": 1.1978795874495103e-05,
+      "loss": 0.206,
+      "step": 2872
+    },
+    {
+      "epoch": 1.3604024859425865,
+      "grad_norm": 1.4673343896865845,
+      "learning_rate": 1.1973782638810264e-05,
+      "loss": 0.2143,
+      "step": 2873
+    },
+    {
+      "epoch": 1.3608759988162178,
+      "grad_norm": 1.2298039197921753,
+      "learning_rate": 1.1968768886901621e-05,
+      "loss": 0.227,
+      "step": 2874
+    },
+    {
+      "epoch": 1.361349511689849,
+      "grad_norm": 1.48917555809021,
+      "learning_rate": 1.1963754620080467e-05,
+      "loss": 0.2148,
+      "step": 2875
+    },
+    {
+      "epoch": 1.3618230245634804,
+      "grad_norm": 1.349953532218933,
+      "learning_rate": 1.1958739839658238e-05,
+      "loss": 0.1988,
+      "step": 2876
+    },
+    {
+      "epoch": 1.3622965374371117,
+      "grad_norm": 1.3024402856826782,
+      "learning_rate": 1.1953724546946502e-05,
+      "loss": 0.2332,
+      "step": 2877
+    },
+    {
+      "epoch": 1.362770050310743,
+      "grad_norm": 1.4080761671066284,
+      "learning_rate": 1.1948708743256954e-05,
+      "loss": 0.221,
+      "step": 2878
+    },
+    {
+      "epoch": 1.363243563184374,
+      "grad_norm": 1.1572239398956299,
+      "learning_rate": 1.1943692429901437e-05,
+      "loss": 0.2094,
+      "step": 2879
+    },
+    {
+      "epoch": 1.3637170760580053,
+      "grad_norm": 1.7980104684829712,
+      "learning_rate": 1.1938675608191914e-05,
+      "loss": 0.2282,
+      "step": 2880
+    },
+    {
+      "epoch": 1.3641905889316366,
+      "grad_norm": 1.5248477458953857,
+      "learning_rate": 1.193365827944049e-05,
+      "loss": 0.2235,
+      "step": 2881
+    },
+    {
+      "epoch": 1.3646641018052679,
+      "grad_norm": 1.2481404542922974,
+      "learning_rate": 1.1928640444959396e-05,
+      "loss": 0.2013,
+      "step": 2882
+    },
+    {
+      "epoch": 1.3651376146788992,
+      "grad_norm": 2.0264244079589844,
+      "learning_rate": 1.1923622106060997e-05,
+      "loss": 0.2341,
+      "step": 2883
+    },
+    {
+      "epoch": 1.3656111275525302,
+      "grad_norm": 1.1340547800064087,
+      "learning_rate": 1.19186032640578e-05,
+      "loss": 0.2295,
+      "step": 2884
+    },
+    {
+      "epoch": 1.3660846404261615,
+      "grad_norm": 1.4470584392547607,
+      "learning_rate": 1.1913583920262424e-05,
+      "loss": 0.2207,
+      "step": 2885
+    },
+    {
+      "epoch": 1.3665581532997928,
+      "grad_norm": 1.103227138519287,
+      "learning_rate": 1.1908564075987637e-05,
+      "loss": 0.2065,
+      "step": 2886
+    },
+    {
+      "epoch": 1.367031666173424,
+      "grad_norm": 1.558107852935791,
+      "learning_rate": 1.1903543732546326e-05,
+      "loss": 0.2203,
+      "step": 2887
+    },
+    {
+      "epoch": 1.3675051790470554,
+      "grad_norm": 1.080611228942871,
+      "learning_rate": 1.1898522891251516e-05,
+      "loss": 0.2328,
+      "step": 2888
+    },
+    {
+      "epoch": 1.3679786919206867,
+      "grad_norm": 1.09610915184021,
+      "learning_rate": 1.1893501553416357e-05,
+      "loss": 0.2292,
+      "step": 2889
+    },
+    {
+      "epoch": 1.368452204794318,
+      "grad_norm": 1.4019349813461304,
+      "learning_rate": 1.1888479720354138e-05,
+      "loss": 0.2118,
+      "step": 2890
+    },
+    {
+      "epoch": 1.368925717667949,
+      "grad_norm": 1.5220277309417725,
+      "learning_rate": 1.1883457393378263e-05,
+      "loss": 0.2051,
+      "step": 2891
+    },
+    {
+      "epoch": 1.3693992305415803,
+      "grad_norm": 0.9200198650360107,
+      "learning_rate": 1.187843457380228e-05,
+      "loss": 0.214,
+      "step": 2892
+    },
+    {
+      "epoch": 1.3698727434152116,
+      "grad_norm": 1.2134068012237549,
+      "learning_rate": 1.1873411262939854e-05,
+      "loss": 0.2068,
+      "step": 2893
+    },
+    {
+      "epoch": 1.3703462562888429,
+      "grad_norm": 1.125510334968567,
+      "learning_rate": 1.1868387462104787e-05,
+      "loss": 0.2247,
+      "step": 2894
+    },
+    {
+      "epoch": 1.3708197691624742,
+      "grad_norm": 0.9510089159011841,
+      "learning_rate": 1.1863363172611003e-05,
+      "loss": 0.2077,
+      "step": 2895
+    },
+    {
+      "epoch": 1.3712932820361052,
+      "grad_norm": 1.2528704404830933,
+      "learning_rate": 1.1858338395772555e-05,
+      "loss": 0.234,
+      "step": 2896
+    },
+    {
+      "epoch": 1.3717667949097367,
+      "grad_norm": 1.718252182006836,
+      "learning_rate": 1.185331313290363e-05,
+      "loss": 0.2241,
+      "step": 2897
+    },
+    {
+      "epoch": 1.3722403077833678,
+      "grad_norm": 1.3889955282211304,
+      "learning_rate": 1.1848287385318533e-05,
+      "loss": 0.239,
+      "step": 2898
+    },
+    {
+      "epoch": 1.372713820656999,
+      "grad_norm": 1.056844711303711,
+      "learning_rate": 1.1843261154331702e-05,
+      "loss": 0.1979,
+      "step": 2899
+    },
+    {
+      "epoch": 1.3731873335306304,
+      "grad_norm": 1.1087300777435303,
+      "learning_rate": 1.1838234441257698e-05,
+      "loss": 0.2094,
+      "step": 2900
+    },
+    {
+      "epoch": 1.3736608464042617,
+      "grad_norm": 1.014634609222412,
+      "learning_rate": 1.1833207247411208e-05,
+      "loss": 0.2052,
+      "step": 2901
+    },
+    {
+      "epoch": 1.374134359277893,
+      "grad_norm": 1.0202566385269165,
+      "learning_rate": 1.182817957410705e-05,
+      "loss": 0.2064,
+      "step": 2902
+    },
+    {
+      "epoch": 1.374607872151524,
+      "grad_norm": 1.1275465488433838,
+      "learning_rate": 1.1823151422660162e-05,
+      "loss": 0.2161,
+      "step": 2903
+    },
+    {
+      "epoch": 1.3750813850251553,
+      "grad_norm": 1.6918755769729614,
+      "learning_rate": 1.1818122794385604e-05,
+      "loss": 0.2127,
+      "step": 2904
+    },
+    {
+      "epoch": 1.3755548978987866,
+      "grad_norm": 1.5377037525177002,
+      "learning_rate": 1.1813093690598572e-05,
+      "loss": 0.1938,
+      "step": 2905
+    },
+    {
+      "epoch": 1.3760284107724179,
+      "grad_norm": 1.1492862701416016,
+      "learning_rate": 1.1808064112614375e-05,
+      "loss": 0.2127,
+      "step": 2906
+    },
+    {
+      "epoch": 1.3765019236460492,
+      "grad_norm": 1.8122957944869995,
+      "learning_rate": 1.1803034061748453e-05,
+      "loss": 0.1997,
+      "step": 2907
+    },
+    {
+      "epoch": 1.3769754365196802,
+      "grad_norm": 1.2996703386306763,
+      "learning_rate": 1.1798003539316365e-05,
+      "loss": 0.226,
+      "step": 2908
+    },
+    {
+      "epoch": 1.3774489493933118,
+      "grad_norm": 1.1555538177490234,
+      "learning_rate": 1.1792972546633799e-05,
+      "loss": 0.2417,
+      "step": 2909
+    },
+    {
+      "epoch": 1.3779224622669428,
+      "grad_norm": 1.0537813901901245,
+      "learning_rate": 1.178794108501656e-05,
+      "loss": 0.221,
+      "step": 2910
+    },
+    {
+      "epoch": 1.378395975140574,
+      "grad_norm": 1.2504982948303223,
+      "learning_rate": 1.1782909155780577e-05,
+      "loss": 0.207,
+      "step": 2911
+    },
+    {
+      "epoch": 1.3788694880142054,
+      "grad_norm": 1.0087714195251465,
+      "learning_rate": 1.1777876760241907e-05,
+      "loss": 0.2108,
+      "step": 2912
+    },
+    {
+      "epoch": 1.3793430008878367,
+      "grad_norm": 0.9194630980491638,
+      "learning_rate": 1.1772843899716719e-05,
+      "loss": 0.2024,
+      "step": 2913
+    },
+    {
+      "epoch": 1.379816513761468,
+      "grad_norm": 1.1061056852340698,
+      "learning_rate": 1.1767810575521312e-05,
+      "loss": 0.2317,
+      "step": 2914
+    },
+    {
+      "epoch": 1.380290026635099,
+      "grad_norm": 1.6991933584213257,
+      "learning_rate": 1.1762776788972106e-05,
+      "loss": 0.2297,
+      "step": 2915
+    },
+    {
+      "epoch": 1.3807635395087303,
+      "grad_norm": 1.4516056776046753,
+      "learning_rate": 1.1757742541385636e-05,
+      "loss": 0.2268,
+      "step": 2916
+    },
+    {
+      "epoch": 1.3812370523823616,
+      "grad_norm": 1.5717034339904785,
+      "learning_rate": 1.1752707834078558e-05,
+      "loss": 0.2298,
+      "step": 2917
+    },
+    {
+      "epoch": 1.381710565255993,
+      "grad_norm": 1.2218979597091675,
+      "learning_rate": 1.1747672668367659e-05,
+      "loss": 0.2222,
+      "step": 2918
+    },
+    {
+      "epoch": 1.3821840781296242,
+      "grad_norm": 1.1709834337234497,
+      "learning_rate": 1.1742637045569832e-05,
+      "loss": 0.2199,
+      "step": 2919
+    },
+    {
+      "epoch": 1.3826575910032555,
+      "grad_norm": 1.1610695123672485,
+      "learning_rate": 1.1737600967002095e-05,
+      "loss": 0.2203,
+      "step": 2920
+    },
+    {
+      "epoch": 1.3831311038768868,
+      "grad_norm": 1.1131560802459717,
+      "learning_rate": 1.1732564433981594e-05,
+      "loss": 0.205,
+      "step": 2921
+    },
+    {
+      "epoch": 1.3836046167505178,
+      "grad_norm": 1.2609494924545288,
+      "learning_rate": 1.1727527447825575e-05,
+      "loss": 0.2129,
+      "step": 2922
+    },
+    {
+      "epoch": 1.3840781296241491,
+      "grad_norm": 1.1561344861984253,
+      "learning_rate": 1.1722490009851418e-05,
+      "loss": 0.2299,
+      "step": 2923
+    },
+    {
+      "epoch": 1.3845516424977804,
+      "grad_norm": 1.0187231302261353,
+      "learning_rate": 1.1717452121376616e-05,
+      "loss": 0.1932,
+      "step": 2924
+    },
+    {
+      "epoch": 1.3850251553714117,
+      "grad_norm": 1.7099744081497192,
+      "learning_rate": 1.1712413783718782e-05,
+      "loss": 0.2273,
+      "step": 2925
+    },
+    {
+      "epoch": 1.385498668245043,
+      "grad_norm": 1.1517730951309204,
+      "learning_rate": 1.1707374998195643e-05,
+      "loss": 0.2322,
+      "step": 2926
+    },
+    {
+      "epoch": 1.385972181118674,
+      "grad_norm": 1.6895732879638672,
+      "learning_rate": 1.1702335766125042e-05,
+      "loss": 0.2198,
+      "step": 2927
+    },
+    {
+      "epoch": 1.3864456939923056,
+      "grad_norm": 1.1857640743255615,
+      "learning_rate": 1.1697296088824945e-05,
+      "loss": 0.2639,
+      "step": 2928
+    },
+    {
+      "epoch": 1.3869192068659366,
+      "grad_norm": 1.5664935111999512,
+      "learning_rate": 1.1692255967613432e-05,
+      "loss": 0.2146,
+      "step": 2929
+    },
+    {
+      "epoch": 1.387392719739568,
+      "grad_norm": 1.2540571689605713,
+      "learning_rate": 1.1687215403808697e-05,
+      "loss": 0.2288,
+      "step": 2930
+    },
+    {
+      "epoch": 1.3878662326131992,
+      "grad_norm": 1.223510980606079,
+      "learning_rate": 1.1682174398729044e-05,
+      "loss": 0.2205,
+      "step": 2931
+    },
+    {
+      "epoch": 1.3883397454868305,
+      "grad_norm": 0.9939408302307129,
+      "learning_rate": 1.1677132953692911e-05,
+      "loss": 0.2321,
+      "step": 2932
+    },
+    {
+      "epoch": 1.3888132583604618,
+      "grad_norm": 1.3630565404891968,
+      "learning_rate": 1.1672091070018832e-05,
+      "loss": 0.2274,
+      "step": 2933
+    },
+    {
+      "epoch": 1.3892867712340928,
+      "grad_norm": 1.4759294986724854,
+      "learning_rate": 1.1667048749025462e-05,
+      "loss": 0.2319,
+      "step": 2934
+    },
+    {
+      "epoch": 1.3897602841077241,
+      "grad_norm": 0.9733860492706299,
+      "learning_rate": 1.1662005992031577e-05,
+      "loss": 0.2087,
+      "step": 2935
+    },
+    {
+      "epoch": 1.3902337969813554,
+      "grad_norm": 1.1344149112701416,
+      "learning_rate": 1.1656962800356058e-05,
+      "loss": 0.2226,
+      "step": 2936
+    },
+    {
+      "epoch": 1.3907073098549867,
+      "grad_norm": 1.0230250358581543,
+      "learning_rate": 1.1651919175317903e-05,
+      "loss": 0.2076,
+      "step": 2937
+    },
+    {
+      "epoch": 1.391180822728618,
+      "grad_norm": 1.2815759181976318,
+      "learning_rate": 1.1646875118236225e-05,
+      "loss": 0.209,
+      "step": 2938
+    },
+    {
+      "epoch": 1.391654335602249,
+      "grad_norm": 1.4215916395187378,
+      "learning_rate": 1.1641830630430246e-05,
+      "loss": 0.2402,
+      "step": 2939
+    },
+    {
+      "epoch": 1.3921278484758806,
+      "grad_norm": 1.8291083574295044,
+      "learning_rate": 1.1636785713219305e-05,
+      "loss": 0.2072,
+      "step": 2940
+    },
+    {
+      "epoch": 1.3926013613495116,
+      "grad_norm": 1.3542451858520508,
+      "learning_rate": 1.163174036792285e-05,
+      "loss": 0.2279,
+      "step": 2941
+    },
+    {
+      "epoch": 1.393074874223143,
+      "grad_norm": 1.9127904176712036,
+      "learning_rate": 1.1626694595860443e-05,
+      "loss": 0.2264,
+      "step": 2942
+    },
+    {
+      "epoch": 1.3935483870967742,
+      "grad_norm": 1.4430345296859741,
+      "learning_rate": 1.1621648398351762e-05,
+      "loss": 0.2149,
+      "step": 2943
+    },
+    {
+      "epoch": 1.3940218999704055,
+      "grad_norm": 2.1805191040039062,
+      "learning_rate": 1.1616601776716583e-05,
+      "loss": 0.2259,
+      "step": 2944
+    },
+    {
+      "epoch": 1.3944954128440368,
+      "grad_norm": 1.4834105968475342,
+      "learning_rate": 1.1611554732274806e-05,
+      "loss": 0.2054,
+      "step": 2945
+    },
+    {
+      "epoch": 1.3949689257176678,
+      "grad_norm": 1.532042145729065,
+      "learning_rate": 1.1606507266346436e-05,
+      "loss": 0.2169,
+      "step": 2946
+    },
+    {
+      "epoch": 1.3954424385912991,
+      "grad_norm": 1.2459840774536133,
+      "learning_rate": 1.160145938025159e-05,
+      "loss": 0.2091,
+      "step": 2947
+    },
+    {
+      "epoch": 1.3959159514649304,
+      "grad_norm": 1.190146565437317,
+      "learning_rate": 1.159641107531049e-05,
+      "loss": 0.1977,
+      "step": 2948
+    },
+    {
+      "epoch": 1.3963894643385617,
+      "grad_norm": 1.033531665802002,
+      "learning_rate": 1.1591362352843477e-05,
+      "loss": 0.1989,
+      "step": 2949
+    },
+    {
+      "epoch": 1.396862977212193,
+      "grad_norm": 1.877530813217163,
+      "learning_rate": 1.158631321417099e-05,
+      "loss": 0.2075,
+      "step": 2950
+    },
+    {
+      "epoch": 1.3973364900858243,
+      "grad_norm": 1.1932871341705322,
+      "learning_rate": 1.1581263660613585e-05,
+      "loss": 0.2094,
+      "step": 2951
+    },
+    {
+      "epoch": 1.3978100029594556,
+      "grad_norm": 1.5696194171905518,
+      "learning_rate": 1.1576213693491925e-05,
+      "loss": 0.207,
+      "step": 2952
+    },
+    {
+      "epoch": 1.3982835158330866,
+      "grad_norm": 1.6218510866165161,
+      "learning_rate": 1.157116331412678e-05,
+      "loss": 0.1988,
+      "step": 2953
+    },
+    {
+      "epoch": 1.398757028706718,
+      "grad_norm": 1.69428551197052,
+      "learning_rate": 1.1566112523839028e-05,
+      "loss": 0.2118,
+      "step": 2954
+    },
+    {
+      "epoch": 1.3992305415803492,
+      "grad_norm": 0.9720800518989563,
+      "learning_rate": 1.1561061323949652e-05,
+      "loss": 0.1867,
+      "step": 2955
+    },
+    {
+      "epoch": 1.3997040544539805,
+      "grad_norm": 1.5280438661575317,
+      "learning_rate": 1.155600971577975e-05,
+      "loss": 0.2288,
+      "step": 2956
+    },
+    {
+      "epoch": 1.4001775673276118,
+      "grad_norm": 1.5449702739715576,
+      "learning_rate": 1.1550957700650517e-05,
+      "loss": 0.2242,
+      "step": 2957
+    },
+    {
+      "epoch": 1.4006510802012428,
+      "grad_norm": 1.8118822574615479,
+      "learning_rate": 1.1545905279883258e-05,
+      "loss": 0.2122,
+      "step": 2958
+    },
+    {
+      "epoch": 1.4011245930748744,
+      "grad_norm": 1.7148773670196533,
+      "learning_rate": 1.1540852454799388e-05,
+      "loss": 0.2237,
+      "step": 2959
+    },
+    {
+      "epoch": 1.4015981059485054,
+      "grad_norm": 1.2663835287094116,
+      "learning_rate": 1.1535799226720421e-05,
+      "loss": 0.2229,
+      "step": 2960
+    },
+    {
+      "epoch": 1.4020716188221367,
+      "grad_norm": 1.3801897764205933,
+      "learning_rate": 1.1530745596967982e-05,
+      "loss": 0.2263,
+      "step": 2961
+    },
+    {
+      "epoch": 1.402545131695768,
+      "grad_norm": 1.4142658710479736,
+      "learning_rate": 1.1525691566863802e-05,
+      "loss": 0.2169,
+      "step": 2962
+    },
+    {
+      "epoch": 1.4030186445693993,
+      "grad_norm": 0.985097348690033,
+      "learning_rate": 1.1520637137729711e-05,
+      "loss": 0.2025,
+      "step": 2963
+    },
+    {
+      "epoch": 1.4034921574430306,
+      "grad_norm": 1.7682098150253296,
+      "learning_rate": 1.1515582310887647e-05,
+      "loss": 0.2098,
+      "step": 2964
+    },
+    {
+      "epoch": 1.4039656703166616,
+      "grad_norm": 1.2826495170593262,
+      "learning_rate": 1.1510527087659648e-05,
+      "loss": 0.2185,
+      "step": 2965
+    },
+    {
+      "epoch": 1.404439183190293,
+      "grad_norm": 0.9184038639068604,
+      "learning_rate": 1.1505471469367864e-05,
+      "loss": 0.2079,
+      "step": 2966
+    },
+    {
+      "epoch": 1.4049126960639242,
+      "grad_norm": 1.2887505292892456,
+      "learning_rate": 1.1500415457334539e-05,
+      "loss": 0.1951,
+      "step": 2967
+    },
+    {
+      "epoch": 1.4053862089375555,
+      "grad_norm": 1.2137596607208252,
+      "learning_rate": 1.1495359052882028e-05,
+      "loss": 0.2093,
+      "step": 2968
+    },
+    {
+      "epoch": 1.4058597218111868,
+      "grad_norm": 1.2313615083694458,
+      "learning_rate": 1.1490302257332781e-05,
+      "loss": 0.2015,
+      "step": 2969
+    },
+    {
+      "epoch": 1.4063332346848179,
+      "grad_norm": 1.5044950246810913,
+      "learning_rate": 1.1485245072009357e-05,
+      "loss": 0.2187,
+      "step": 2970
+    },
+    {
+      "epoch": 1.4068067475584494,
+      "grad_norm": 1.2070937156677246,
+      "learning_rate": 1.1480187498234412e-05,
+      "loss": 0.2067,
+      "step": 2971
+    },
+    {
+      "epoch": 1.4072802604320804,
+      "grad_norm": 1.3706823587417603,
+      "learning_rate": 1.1475129537330707e-05,
+      "loss": 0.2081,
+      "step": 2972
+    },
+    {
+      "epoch": 1.4077537733057117,
+      "grad_norm": 1.1687662601470947,
+      "learning_rate": 1.1470071190621103e-05,
+      "loss": 0.2081,
+      "step": 2973
+    },
+    {
+      "epoch": 1.408227286179343,
+      "grad_norm": 1.102828025817871,
+      "learning_rate": 1.1465012459428562e-05,
+      "loss": 0.2129,
+      "step": 2974
+    },
+    {
+      "epoch": 1.4087007990529743,
+      "grad_norm": 1.8313281536102295,
+      "learning_rate": 1.1459953345076142e-05,
+      "loss": 0.2111,
+      "step": 2975
+    },
+    {
+      "epoch": 1.4091743119266056,
+      "grad_norm": 1.3744657039642334,
+      "learning_rate": 1.1454893848887013e-05,
+      "loss": 0.2099,
+      "step": 2976
+    },
+    {
+      "epoch": 1.4096478248002366,
+      "grad_norm": 0.9316054582595825,
+      "learning_rate": 1.144983397218443e-05,
+      "loss": 0.2132,
+      "step": 2977
+    },
+    {
+      "epoch": 1.410121337673868,
+      "grad_norm": 1.1082512140274048,
+      "learning_rate": 1.1444773716291759e-05,
+      "loss": 0.1941,
+      "step": 2978
+    },
+    {
+      "epoch": 1.4105948505474992,
+      "grad_norm": 1.3401765823364258,
+      "learning_rate": 1.143971308253246e-05,
+      "loss": 0.2226,
+      "step": 2979
+    },
+    {
+      "epoch": 1.4110683634211305,
+      "grad_norm": 1.1040364503860474,
+      "learning_rate": 1.1434652072230092e-05,
+      "loss": 0.1832,
+      "step": 2980
+    },
+    {
+      "epoch": 1.4115418762947618,
+      "grad_norm": 1.007563829421997,
+      "learning_rate": 1.142959068670832e-05,
+      "loss": 0.2122,
+      "step": 2981
+    },
+    {
+      "epoch": 1.412015389168393,
+      "grad_norm": 1.306701421737671,
+      "learning_rate": 1.1424528927290892e-05,
+      "loss": 0.2162,
+      "step": 2982
+    },
+    {
+      "epoch": 1.4124889020420244,
+      "grad_norm": 0.9834753274917603,
+      "learning_rate": 1.1419466795301665e-05,
+      "loss": 0.201,
+      "step": 2983
+    },
+    {
+      "epoch": 1.4129624149156554,
+      "grad_norm": 1.4253736734390259,
+      "learning_rate": 1.1414404292064593e-05,
+      "loss": 0.2153,
+      "step": 2984
+    },
+    {
+      "epoch": 1.4134359277892867,
+      "grad_norm": 1.133834719657898,
+      "learning_rate": 1.1409341418903725e-05,
+      "loss": 0.2283,
+      "step": 2985
+    },
+    {
+      "epoch": 1.413909440662918,
+      "grad_norm": 1.4796968698501587,
+      "learning_rate": 1.1404278177143202e-05,
+      "loss": 0.2251,
+      "step": 2986
+    },
+    {
+      "epoch": 1.4143829535365493,
+      "grad_norm": 1.3336002826690674,
+      "learning_rate": 1.139921456810727e-05,
+      "loss": 0.206,
+      "step": 2987
+    },
+    {
+      "epoch": 1.4148564664101806,
+      "grad_norm": 1.420716404914856,
+      "learning_rate": 1.1394150593120268e-05,
+      "loss": 0.2226,
+      "step": 2988
+    },
+    {
+      "epoch": 1.4153299792838117,
+      "grad_norm": 1.5567362308502197,
+      "learning_rate": 1.1389086253506626e-05,
+      "loss": 0.2219,
+      "step": 2989
+    },
+    {
+      "epoch": 1.4158034921574432,
+      "grad_norm": 0.9303296804428101,
+      "learning_rate": 1.1384021550590878e-05,
+      "loss": 0.2277,
+      "step": 2990
+    },
+    {
+      "epoch": 1.4162770050310742,
+      "grad_norm": 1.0575077533721924,
+      "learning_rate": 1.1378956485697644e-05,
+      "loss": 0.2224,
+      "step": 2991
+    },
+    {
+      "epoch": 1.4167505179047055,
+      "grad_norm": 1.0305496454238892,
+      "learning_rate": 1.1373891060151643e-05,
+      "loss": 0.2341,
+      "step": 2992
+    },
+    {
+      "epoch": 1.4172240307783368,
+      "grad_norm": 1.0244580507278442,
+      "learning_rate": 1.1368825275277689e-05,
+      "loss": 0.2011,
+      "step": 2993
+    },
+    {
+      "epoch": 1.417697543651968,
+      "grad_norm": 1.31196928024292,
+      "learning_rate": 1.136375913240069e-05,
+      "loss": 0.2253,
+      "step": 2994
+    },
+    {
+      "epoch": 1.4181710565255994,
+      "grad_norm": 1.118857502937317,
+      "learning_rate": 1.1358692632845645e-05,
+      "loss": 0.2278,
+      "step": 2995
+    },
+    {
+      "epoch": 1.4186445693992304,
+      "grad_norm": 1.114423155784607,
+      "learning_rate": 1.1353625777937652e-05,
+      "loss": 0.2004,
+      "step": 2996
+    },
+    {
+      "epoch": 1.4191180822728617,
+      "grad_norm": 1.2260422706604004,
+      "learning_rate": 1.1348558569001896e-05,
+      "loss": 0.2056,
+      "step": 2997
+    },
+    {
+      "epoch": 1.419591595146493,
+      "grad_norm": 1.8472795486450195,
+      "learning_rate": 1.1343491007363652e-05,
+      "loss": 0.2181,
+      "step": 2998
+    },
+    {
+      "epoch": 1.4200651080201243,
+      "grad_norm": 1.8274016380310059,
+      "learning_rate": 1.1338423094348299e-05,
+      "loss": 0.2138,
+      "step": 2999
+    },
+    {
+      "epoch": 1.4205386208937556,
+      "grad_norm": 1.1858888864517212,
+      "learning_rate": 1.1333354831281295e-05,
+      "loss": 0.2022,
+      "step": 3000
+    },
+    {
+      "epoch": 1.4210121337673867,
+      "grad_norm": 1.1788923740386963,
+      "learning_rate": 1.13282862194882e-05,
+      "loss": 0.1814,
+      "step": 3001
+    },
+    {
+      "epoch": 1.4214856466410182,
+      "grad_norm": 1.1308039426803589,
+      "learning_rate": 1.132321726029466e-05,
+      "loss": 0.2167,
+      "step": 3002
+    },
+    {
+      "epoch": 1.4219591595146492,
+      "grad_norm": 1.0506905317306519,
+      "learning_rate": 1.1318147955026408e-05,
+      "loss": 0.2011,
+      "step": 3003
+    },
+    {
+      "epoch": 1.4224326723882805,
+      "grad_norm": 1.2325364351272583,
+      "learning_rate": 1.1313078305009278e-05,
+      "loss": 0.1922,
+      "step": 3004
+    },
+    {
+      "epoch": 1.4229061852619118,
+      "grad_norm": 1.2299723625183105,
+      "learning_rate": 1.130800831156918e-05,
+      "loss": 0.2292,
+      "step": 3005
+    },
+    {
+      "epoch": 1.423379698135543,
+      "grad_norm": 1.5796177387237549,
+      "learning_rate": 1.1302937976032131e-05,
+      "loss": 0.2311,
+      "step": 3006
+    },
+    {
+      "epoch": 1.4238532110091744,
+      "grad_norm": 1.015859603881836,
+      "learning_rate": 1.1297867299724223e-05,
+      "loss": 0.2398,
+      "step": 3007
+    },
+    {
+      "epoch": 1.4243267238828055,
+      "grad_norm": 1.0847722291946411,
+      "learning_rate": 1.1292796283971645e-05,
+      "loss": 0.2105,
+      "step": 3008
+    },
+    {
+      "epoch": 1.4248002367564367,
+      "grad_norm": 1.5696632862091064,
+      "learning_rate": 1.1287724930100668e-05,
+      "loss": 0.22,
+      "step": 3009
+    },
+    {
+      "epoch": 1.425273749630068,
+      "grad_norm": 1.0364569425582886,
+      "learning_rate": 1.128265323943766e-05,
+      "loss": 0.2086,
+      "step": 3010
+    },
+    {
+      "epoch": 1.4257472625036993,
+      "grad_norm": 1.0130901336669922,
+      "learning_rate": 1.127758121330907e-05,
+      "loss": 0.1888,
+      "step": 3011
+    },
+    {
+      "epoch": 1.4262207753773306,
+      "grad_norm": 1.5478230714797974,
+      "learning_rate": 1.127250885304144e-05,
+      "loss": 0.2325,
+      "step": 3012
+    },
+    {
+      "epoch": 1.426694288250962,
+      "grad_norm": 1.2723689079284668,
+      "learning_rate": 1.1267436159961393e-05,
+      "loss": 0.2187,
+      "step": 3013
+    },
+    {
+      "epoch": 1.4271678011245932,
+      "grad_norm": 1.111171007156372,
+      "learning_rate": 1.1262363135395648e-05,
+      "loss": 0.215,
+      "step": 3014
+    },
+    {
+      "epoch": 1.4276413139982242,
+      "grad_norm": 1.3690720796585083,
+      "learning_rate": 1.1257289780671004e-05,
+      "loss": 0.2206,
+      "step": 3015
+    },
+    {
+      "epoch": 1.4281148268718555,
+      "grad_norm": 1.2229105234146118,
+      "learning_rate": 1.1252216097114344e-05,
+      "loss": 0.2297,
+      "step": 3016
+    },
+    {
+      "epoch": 1.4285883397454868,
+      "grad_norm": 1.8208054304122925,
+      "learning_rate": 1.124714208605265e-05,
+      "loss": 0.2429,
+      "step": 3017
+    },
+    {
+      "epoch": 1.4290618526191181,
+      "grad_norm": 1.015411615371704,
+      "learning_rate": 1.1242067748812968e-05,
+      "loss": 0.2145,
+      "step": 3018
+    },
+    {
+      "epoch": 1.4295353654927494,
+      "grad_norm": 1.8651050329208374,
+      "learning_rate": 1.123699308672245e-05,
+      "loss": 0.2222,
+      "step": 3019
+    },
+    {
+      "epoch": 1.4300088783663805,
+      "grad_norm": 2.2374584674835205,
+      "learning_rate": 1.1231918101108325e-05,
+      "loss": 0.2141,
+      "step": 3020
+    },
+    {
+      "epoch": 1.430482391240012,
+      "grad_norm": 1.1194039583206177,
+      "learning_rate": 1.1226842793297904e-05,
+      "loss": 0.212,
+      "step": 3021
+    },
+    {
+      "epoch": 1.430955904113643,
+      "grad_norm": 1.2247196435928345,
+      "learning_rate": 1.122176716461859e-05,
+      "loss": 0.2164,
+      "step": 3022
+    },
+    {
+      "epoch": 1.4314294169872743,
+      "grad_norm": 1.0762709379196167,
+      "learning_rate": 1.1216691216397854e-05,
+      "loss": 0.2328,
+      "step": 3023
+    },
+    {
+      "epoch": 1.4319029298609056,
+      "grad_norm": 1.0583688020706177,
+      "learning_rate": 1.1211614949963274e-05,
+      "loss": 0.1974,
+      "step": 3024
+    },
+    {
+      "epoch": 1.432376442734537,
+      "grad_norm": 1.0842574834823608,
+      "learning_rate": 1.1206538366642491e-05,
+      "loss": 0.1934,
+      "step": 3025
+    },
+    {
+      "epoch": 1.4328499556081682,
+      "grad_norm": 1.6864724159240723,
+      "learning_rate": 1.1201461467763238e-05,
+      "loss": 0.227,
+      "step": 3026
+    },
+    {
+      "epoch": 1.4333234684817993,
+      "grad_norm": 1.0597320795059204,
+      "learning_rate": 1.1196384254653328e-05,
+      "loss": 0.2115,
+      "step": 3027
+    },
+    {
+      "epoch": 1.4337969813554305,
+      "grad_norm": 1.4735287427902222,
+      "learning_rate": 1.1191306728640659e-05,
+      "loss": 0.2357,
+      "step": 3028
+    },
+    {
+      "epoch": 1.4342704942290618,
+      "grad_norm": 1.1243826150894165,
+      "learning_rate": 1.1186228891053205e-05,
+      "loss": 0.2247,
+      "step": 3029
+    },
+    {
+      "epoch": 1.4347440071026931,
+      "grad_norm": 1.8924311399459839,
+      "learning_rate": 1.1181150743219031e-05,
+      "loss": 0.2168,
+      "step": 3030
+    },
+    {
+      "epoch": 1.4352175199763244,
+      "grad_norm": 1.0397160053253174,
+      "learning_rate": 1.1176072286466273e-05,
+      "loss": 0.2292,
+      "step": 3031
+    },
+    {
+      "epoch": 1.4356910328499555,
+      "grad_norm": 1.515010118484497,
+      "learning_rate": 1.1170993522123155e-05,
+      "loss": 0.236,
+      "step": 3032
+    },
+    {
+      "epoch": 1.436164545723587,
+      "grad_norm": 1.094772458076477,
+      "learning_rate": 1.1165914451517978e-05,
+      "loss": 0.1897,
+      "step": 3033
+    },
+    {
+      "epoch": 1.436638058597218,
+      "grad_norm": 1.71537446975708,
+      "learning_rate": 1.1160835075979124e-05,
+      "loss": 0.1892,
+      "step": 3034
+    },
+    {
+      "epoch": 1.4371115714708493,
+      "grad_norm": 1.3985260725021362,
+      "learning_rate": 1.1155755396835057e-05,
+      "loss": 0.2327,
+      "step": 3035
+    },
+    {
+      "epoch": 1.4375850843444806,
+      "grad_norm": 0.8228690028190613,
+      "learning_rate": 1.1150675415414314e-05,
+      "loss": 0.2021,
+      "step": 3036
+    },
+    {
+      "epoch": 1.438058597218112,
+      "grad_norm": 0.931536078453064,
+      "learning_rate": 1.1145595133045517e-05,
+      "loss": 0.2083,
+      "step": 3037
+    },
+    {
+      "epoch": 1.4385321100917432,
+      "grad_norm": 1.1511605978012085,
+      "learning_rate": 1.1140514551057361e-05,
+      "loss": 0.201,
+      "step": 3038
+    },
+    {
+      "epoch": 1.4390056229653743,
+      "grad_norm": 1.2221344709396362,
+      "learning_rate": 1.113543367077863e-05,
+      "loss": 0.2049,
+      "step": 3039
+    },
+    {
+      "epoch": 1.4394791358390056,
+      "grad_norm": 1.1700915098190308,
+      "learning_rate": 1.1130352493538176e-05,
+      "loss": 0.208,
+      "step": 3040
+    },
+    {
+      "epoch": 1.4399526487126368,
+      "grad_norm": 0.9786930084228516,
+      "learning_rate": 1.1125271020664931e-05,
+      "loss": 0.2006,
+      "step": 3041
+    },
+    {
+      "epoch": 1.4404261615862681,
+      "grad_norm": 1.0784828662872314,
+      "learning_rate": 1.1120189253487912e-05,
+      "loss": 0.2204,
+      "step": 3042
+    },
+    {
+      "epoch": 1.4408996744598994,
+      "grad_norm": 1.8664531707763672,
+      "learning_rate": 1.1115107193336194e-05,
+      "loss": 0.1987,
+      "step": 3043
+    },
+    {
+      "epoch": 1.4413731873335307,
+      "grad_norm": 1.3436121940612793,
+      "learning_rate": 1.111002484153895e-05,
+      "loss": 0.2301,
+      "step": 3044
+    },
+    {
+      "epoch": 1.441846700207162,
+      "grad_norm": 1.8887630701065063,
+      "learning_rate": 1.1104942199425418e-05,
+      "loss": 0.2189,
+      "step": 3045
+    },
+    {
+      "epoch": 1.442320213080793,
+      "grad_norm": 1.4706549644470215,
+      "learning_rate": 1.1099859268324911e-05,
+      "loss": 0.2382,
+      "step": 3046
+    },
+    {
+      "epoch": 1.4427937259544243,
+      "grad_norm": 1.3858911991119385,
+      "learning_rate": 1.1094776049566822e-05,
+      "loss": 0.2085,
+      "step": 3047
+    },
+    {
+      "epoch": 1.4432672388280556,
+      "grad_norm": 1.5077974796295166,
+      "learning_rate": 1.1089692544480622e-05,
+      "loss": 0.2127,
+      "step": 3048
+    },
+    {
+      "epoch": 1.443740751701687,
+      "grad_norm": 1.5535578727722168,
+      "learning_rate": 1.1084608754395846e-05,
+      "loss": 0.2164,
+      "step": 3049
+    },
+    {
+      "epoch": 1.4442142645753182,
+      "grad_norm": 1.3303495645523071,
+      "learning_rate": 1.1079524680642115e-05,
+      "loss": 0.2167,
+      "step": 3050
+    },
+    {
+      "epoch": 1.4446877774489493,
+      "grad_norm": 1.0496814250946045,
+      "learning_rate": 1.1074440324549118e-05,
+      "loss": 0.2073,
+      "step": 3051
+    },
+    {
+      "epoch": 1.4451612903225808,
+      "grad_norm": 2.0565896034240723,
+      "learning_rate": 1.1069355687446624e-05,
+      "loss": 0.2019,
+      "step": 3052
+    },
+    {
+      "epoch": 1.4456348031962118,
+      "grad_norm": 1.2639986276626587,
+      "learning_rate": 1.1064270770664463e-05,
+      "loss": 0.2108,
+      "step": 3053
+    },
+    {
+      "epoch": 1.4461083160698431,
+      "grad_norm": 1.359134554862976,
+      "learning_rate": 1.1059185575532548e-05,
+      "loss": 0.2302,
+      "step": 3054
+    },
+    {
+      "epoch": 1.4465818289434744,
+      "grad_norm": 1.306850552558899,
+      "learning_rate": 1.1054100103380862e-05,
+      "loss": 0.228,
+      "step": 3055
+    },
+    {
+      "epoch": 1.4470553418171057,
+      "grad_norm": 1.1254212856292725,
+      "learning_rate": 1.1049014355539466e-05,
+      "loss": 0.2271,
+      "step": 3056
+    },
+    {
+      "epoch": 1.447528854690737,
+      "grad_norm": 1.1044193506240845,
+      "learning_rate": 1.1043928333338486e-05,
+      "loss": 0.2116,
+      "step": 3057
+    },
+    {
+      "epoch": 1.448002367564368,
+      "grad_norm": 1.277251958847046,
+      "learning_rate": 1.1038842038108117e-05,
+      "loss": 0.2246,
+      "step": 3058
+    },
+    {
+      "epoch": 1.4484758804379994,
+      "grad_norm": 1.157020092010498,
+      "learning_rate": 1.1033755471178638e-05,
+      "loss": 0.2379,
+      "step": 3059
+    },
+    {
+      "epoch": 1.4489493933116306,
+      "grad_norm": 1.0917078256607056,
+      "learning_rate": 1.102866863388039e-05,
+      "loss": 0.2357,
+      "step": 3060
+    },
+    {
+      "epoch": 1.449422906185262,
+      "grad_norm": 1.5090773105621338,
+      "learning_rate": 1.1023581527543784e-05,
+      "loss": 0.2085,
+      "step": 3061
+    },
+    {
+      "epoch": 1.4498964190588932,
+      "grad_norm": 1.000868320465088,
+      "learning_rate": 1.1018494153499302e-05,
+      "loss": 0.2221,
+      "step": 3062
+    },
+    {
+      "epoch": 1.4503699319325243,
+      "grad_norm": 1.2444463968276978,
+      "learning_rate": 1.10134065130775e-05,
+      "loss": 0.2355,
+      "step": 3063
+    },
+    {
+      "epoch": 1.4508434448061558,
+      "grad_norm": 1.5682604312896729,
+      "learning_rate": 1.1008318607609e-05,
+      "loss": 0.2147,
+      "step": 3064
+    },
+    {
+      "epoch": 1.4513169576797869,
+      "grad_norm": 1.108665108680725,
+      "learning_rate": 1.1003230438424498e-05,
+      "loss": 0.205,
+      "step": 3065
+    },
+    {
+      "epoch": 1.4517904705534181,
+      "grad_norm": 1.2267842292785645,
+      "learning_rate": 1.0998142006854754e-05,
+      "loss": 0.2354,
+      "step": 3066
+    },
+    {
+      "epoch": 1.4522639834270494,
+      "grad_norm": 1.2378571033477783,
+      "learning_rate": 1.0993053314230593e-05,
+      "loss": 0.2122,
+      "step": 3067
+    },
+    {
+      "epoch": 1.4527374963006807,
+      "grad_norm": 1.5879398584365845,
+      "learning_rate": 1.0987964361882921e-05,
+      "loss": 0.2425,
+      "step": 3068
+    },
+    {
+      "epoch": 1.453211009174312,
+      "grad_norm": 0.9649191498756409,
+      "learning_rate": 1.0982875151142702e-05,
+      "loss": 0.2217,
+      "step": 3069
+    },
+    {
+      "epoch": 1.453684522047943,
+      "grad_norm": 1.1818419694900513,
+      "learning_rate": 1.097778568334097e-05,
+      "loss": 0.1961,
+      "step": 3070
+    },
+    {
+      "epoch": 1.4541580349215744,
+      "grad_norm": 1.4016741514205933,
+      "learning_rate": 1.0972695959808822e-05,
+      "loss": 0.1939,
+      "step": 3071
+    },
+    {
+      "epoch": 1.4546315477952056,
+      "grad_norm": 1.1574840545654297,
+      "learning_rate": 1.096760598187743e-05,
+      "loss": 0.2055,
+      "step": 3072
+    },
+    {
+      "epoch": 1.455105060668837,
+      "grad_norm": 2.2353291511535645,
+      "learning_rate": 1.096251575087803e-05,
+      "loss": 0.2041,
+      "step": 3073
+    },
+    {
+      "epoch": 1.4555785735424682,
+      "grad_norm": 1.6579928398132324,
+      "learning_rate": 1.095742526814192e-05,
+      "loss": 0.1982,
+      "step": 3074
+    },
+    {
+      "epoch": 1.4560520864160995,
+      "grad_norm": 1.2612777948379517,
+      "learning_rate": 1.0952334535000466e-05,
+      "loss": 0.2086,
+      "step": 3075
+    },
+    {
+      "epoch": 1.4565255992897308,
+      "grad_norm": 1.2440173625946045,
+      "learning_rate": 1.0947243552785103e-05,
+      "loss": 0.2239,
+      "step": 3076
+    },
+    {
+      "epoch": 1.4569991121633619,
+      "grad_norm": 1.0874110460281372,
+      "learning_rate": 1.0942152322827326e-05,
+      "loss": 0.2099,
+      "step": 3077
+    },
+    {
+      "epoch": 1.4574726250369932,
+      "grad_norm": 1.3692322969436646,
+      "learning_rate": 1.0937060846458701e-05,
+      "loss": 0.2069,
+      "step": 3078
+    },
+    {
+      "epoch": 1.4579461379106244,
+      "grad_norm": 1.1790695190429688,
+      "learning_rate": 1.0931969125010848e-05,
+      "loss": 0.2467,
+      "step": 3079
+    },
+    {
+      "epoch": 1.4584196507842557,
+      "grad_norm": 1.1012235879898071,
+      "learning_rate": 1.0926877159815463e-05,
+      "loss": 0.2214,
+      "step": 3080
+    },
+    {
+      "epoch": 1.458893163657887,
+      "grad_norm": 1.178139328956604,
+      "learning_rate": 1.0921784952204299e-05,
+      "loss": 0.2138,
+      "step": 3081
+    },
+    {
+      "epoch": 1.459366676531518,
+      "grad_norm": 1.5324974060058594,
+      "learning_rate": 1.0916692503509172e-05,
+      "loss": 0.234,
+      "step": 3082
+    },
+    {
+      "epoch": 1.4598401894051496,
+      "grad_norm": 1.391147255897522,
+      "learning_rate": 1.0911599815061966e-05,
+      "loss": 0.2091,
+      "step": 3083
+    },
+    {
+      "epoch": 1.4603137022787807,
+      "grad_norm": 1.076026439666748,
+      "learning_rate": 1.0906506888194621e-05,
+      "loss": 0.2073,
+      "step": 3084
+    },
+    {
+      "epoch": 1.460787215152412,
+      "grad_norm": 1.4098458290100098,
+      "learning_rate": 1.0901413724239144e-05,
+      "loss": 0.2358,
+      "step": 3085
+    },
+    {
+      "epoch": 1.4612607280260432,
+      "grad_norm": 1.8523685932159424,
+      "learning_rate": 1.0896320324527606e-05,
+      "loss": 0.2471,
+      "step": 3086
+    },
+    {
+      "epoch": 1.4617342408996745,
+      "grad_norm": 1.161373496055603,
+      "learning_rate": 1.0891226690392136e-05,
+      "loss": 0.2168,
+      "step": 3087
+    },
+    {
+      "epoch": 1.4622077537733058,
+      "grad_norm": 1.116976261138916,
+      "learning_rate": 1.0886132823164919e-05,
+      "loss": 0.2141,
+      "step": 3088
+    },
+    {
+      "epoch": 1.4626812666469369,
+      "grad_norm": 0.9809169769287109,
+      "learning_rate": 1.0881038724178211e-05,
+      "loss": 0.2297,
+      "step": 3089
+    },
+    {
+      "epoch": 1.4631547795205682,
+      "grad_norm": 1.2216544151306152,
+      "learning_rate": 1.0875944394764325e-05,
+      "loss": 0.2317,
+      "step": 3090
+    },
+    {
+      "epoch": 1.4636282923941994,
+      "grad_norm": 1.036920189857483,
+      "learning_rate": 1.0870849836255632e-05,
+      "loss": 0.2307,
+      "step": 3091
+    },
+    {
+      "epoch": 1.4641018052678307,
+      "grad_norm": 1.1497174501419067,
+      "learning_rate": 1.0865755049984568e-05,
+      "loss": 0.2114,
+      "step": 3092
+    },
+    {
+      "epoch": 1.464575318141462,
+      "grad_norm": 1.21004056930542,
+      "learning_rate": 1.0860660037283622e-05,
+      "loss": 0.2054,
+      "step": 3093
+    },
+    {
+      "epoch": 1.465048831015093,
+      "grad_norm": 2.018868923187256,
+      "learning_rate": 1.0855564799485345e-05,
+      "loss": 0.2074,
+      "step": 3094
+    },
+    {
+      "epoch": 1.4655223438887246,
+      "grad_norm": 1.4326393604278564,
+      "learning_rate": 1.0850469337922348e-05,
+      "loss": 0.2179,
+      "step": 3095
+    },
+    {
+      "epoch": 1.4659958567623557,
+      "grad_norm": 1.4479272365570068,
+      "learning_rate": 1.0845373653927303e-05,
+      "loss": 0.2191,
+      "step": 3096
+    },
+    {
+      "epoch": 1.466469369635987,
+      "grad_norm": 1.525227665901184,
+      "learning_rate": 1.084027774883293e-05,
+      "loss": 0.2227,
+      "step": 3097
+    },
+    {
+      "epoch": 1.4669428825096182,
+      "grad_norm": 1.1272501945495605,
+      "learning_rate": 1.0835181623972022e-05,
+      "loss": 0.2156,
+      "step": 3098
+    },
+    {
+      "epoch": 1.4674163953832495,
+      "grad_norm": 1.327453851699829,
+      "learning_rate": 1.0830085280677415e-05,
+      "loss": 0.2185,
+      "step": 3099
+    },
+    {
+      "epoch": 1.4678899082568808,
+      "grad_norm": 1.1839863061904907,
+      "learning_rate": 1.082498872028201e-05,
+      "loss": 0.2343,
+      "step": 3100
+    },
+    {
+      "epoch": 1.4683634211305119,
+      "grad_norm": 0.9597224593162537,
+      "learning_rate": 1.0819891944118768e-05,
+      "loss": 0.193,
+      "step": 3101
+    },
+    {
+      "epoch": 1.4688369340041432,
+      "grad_norm": 1.457085371017456,
+      "learning_rate": 1.0814794953520695e-05,
+      "loss": 0.2102,
+      "step": 3102
+    },
+    {
+      "epoch": 1.4693104468777745,
+      "grad_norm": 1.054195761680603,
+      "learning_rate": 1.0809697749820862e-05,
+      "loss": 0.2354,
+      "step": 3103
+    },
+    {
+      "epoch": 1.4697839597514057,
+      "grad_norm": 1.249380111694336,
+      "learning_rate": 1.0804600334352398e-05,
+      "loss": 0.2031,
+      "step": 3104
+    },
+    {
+      "epoch": 1.470257472625037,
+      "grad_norm": 1.4922364950180054,
+      "learning_rate": 1.0799502708448475e-05,
+      "loss": 0.2107,
+      "step": 3105
+    },
+    {
+      "epoch": 1.4707309854986683,
+      "grad_norm": 1.830451488494873,
+      "learning_rate": 1.079440487344233e-05,
+      "loss": 0.2185,
+      "step": 3106
+    },
+    {
+      "epoch": 1.4712044983722996,
+      "grad_norm": 1.072477102279663,
+      "learning_rate": 1.0789306830667256e-05,
+      "loss": 0.2233,
+      "step": 3107
+    },
+    {
+      "epoch": 1.4716780112459307,
+      "grad_norm": 1.220098853111267,
+      "learning_rate": 1.0784208581456595e-05,
+      "loss": 0.2268,
+      "step": 3108
+    },
+    {
+      "epoch": 1.472151524119562,
+      "grad_norm": 1.263378381729126,
+      "learning_rate": 1.0779110127143746e-05,
+      "loss": 0.2277,
+      "step": 3109
+    },
+    {
+      "epoch": 1.4726250369931932,
+      "grad_norm": 1.1181433200836182,
+      "learning_rate": 1.0774011469062157e-05,
+      "loss": 0.2124,
+      "step": 3110
+    },
+    {
+      "epoch": 1.4730985498668245,
+      "grad_norm": 0.9625453352928162,
+      "learning_rate": 1.0768912608545336e-05,
+      "loss": 0.2112,
+      "step": 3111
+    },
+    {
+      "epoch": 1.4735720627404558,
+      "grad_norm": 1.250014066696167,
+      "learning_rate": 1.0763813546926842e-05,
+      "loss": 0.2388,
+      "step": 3112
+    },
+    {
+      "epoch": 1.474045575614087,
+      "grad_norm": 1.1710668802261353,
+      "learning_rate": 1.0758714285540281e-05,
+      "loss": 0.2245,
+      "step": 3113
+    },
+    {
+      "epoch": 1.4745190884877182,
+      "grad_norm": 1.6824264526367188,
+      "learning_rate": 1.0753614825719321e-05,
+      "loss": 0.1949,
+      "step": 3114
+    },
+    {
+      "epoch": 1.4749926013613495,
+      "grad_norm": 1.1846730709075928,
+      "learning_rate": 1.0748515168797673e-05,
+      "loss": 0.2139,
+      "step": 3115
+    },
+    {
+      "epoch": 1.4754661142349808,
+      "grad_norm": 1.0218571424484253,
+      "learning_rate": 1.07434153161091e-05,
+      "loss": 0.1977,
+      "step": 3116
+    },
+    {
+      "epoch": 1.475939627108612,
+      "grad_norm": 1.459490418434143,
+      "learning_rate": 1.0738315268987424e-05,
+      "loss": 0.2223,
+      "step": 3117
+    },
+    {
+      "epoch": 1.4764131399822433,
+      "grad_norm": 1.3154277801513672,
+      "learning_rate": 1.0733215028766515e-05,
+      "loss": 0.1916,
+      "step": 3118
+    },
+    {
+      "epoch": 1.4768866528558746,
+      "grad_norm": 1.44888436794281,
+      "learning_rate": 1.0728114596780287e-05,
+      "loss": 0.1832,
+      "step": 3119
+    },
+    {
+      "epoch": 1.4773601657295057,
+      "grad_norm": 1.1630711555480957,
+      "learning_rate": 1.072301397436271e-05,
+      "loss": 0.2025,
+      "step": 3120
+    },
+    {
+      "epoch": 1.477833678603137,
+      "grad_norm": 1.9314302206039429,
+      "learning_rate": 1.0717913162847803e-05,
+      "loss": 0.1954,
+      "step": 3121
+    },
+    {
+      "epoch": 1.4783071914767683,
+      "grad_norm": 2.154477119445801,
+      "learning_rate": 1.0712812163569638e-05,
+      "loss": 0.2119,
+      "step": 3122
+    },
+    {
+      "epoch": 1.4787807043503995,
+      "grad_norm": 1.4404516220092773,
+      "learning_rate": 1.0707710977862322e-05,
+      "loss": 0.2379,
+      "step": 3123
+    },
+    {
+      "epoch": 1.4792542172240308,
+      "grad_norm": 1.926571249961853,
+      "learning_rate": 1.070260960706003e-05,
+      "loss": 0.2268,
+      "step": 3124
+    },
+    {
+      "epoch": 1.479727730097662,
+      "grad_norm": 1.082242488861084,
+      "learning_rate": 1.0697508052496975e-05,
+      "loss": 0.2129,
+      "step": 3125
+    },
+    {
+      "epoch": 1.4802012429712934,
+      "grad_norm": 1.2488141059875488,
+      "learning_rate": 1.0692406315507416e-05,
+      "loss": 0.225,
+      "step": 3126
+    },
+    {
+      "epoch": 1.4806747558449245,
+      "grad_norm": 1.4362695217132568,
+      "learning_rate": 1.0687304397425666e-05,
+      "loss": 0.2294,
+      "step": 3127
+    },
+    {
+      "epoch": 1.4811482687185558,
+      "grad_norm": 1.3312585353851318,
+      "learning_rate": 1.068220229958608e-05,
+      "loss": 0.2222,
+      "step": 3128
+    },
+    {
+      "epoch": 1.481621781592187,
+      "grad_norm": 1.7034131288528442,
+      "learning_rate": 1.0677100023323064e-05,
+      "loss": 0.2519,
+      "step": 3129
+    },
+    {
+      "epoch": 1.4820952944658183,
+      "grad_norm": 1.3612613677978516,
+      "learning_rate": 1.067199756997107e-05,
+      "loss": 0.2048,
+      "step": 3130
+    },
+    {
+      "epoch": 1.4825688073394496,
+      "grad_norm": 1.0770941972732544,
+      "learning_rate": 1.0666894940864595e-05,
+      "loss": 0.2233,
+      "step": 3131
+    },
+    {
+      "epoch": 1.4830423202130807,
+      "grad_norm": 1.6613733768463135,
+      "learning_rate": 1.0661792137338183e-05,
+      "loss": 0.2034,
+      "step": 3132
+    },
+    {
+      "epoch": 1.483515833086712,
+      "grad_norm": 1.7866359949111938,
+      "learning_rate": 1.0656689160726418e-05,
+      "loss": 0.2307,
+      "step": 3133
+    },
+    {
+      "epoch": 1.4839893459603433,
+      "grad_norm": 0.9103512167930603,
+      "learning_rate": 1.065158601236394e-05,
+      "loss": 0.2262,
+      "step": 3134
+    },
+    {
+      "epoch": 1.4844628588339746,
+      "grad_norm": 1.5666887760162354,
+      "learning_rate": 1.0646482693585427e-05,
+      "loss": 0.1796,
+      "step": 3135
+    },
+    {
+      "epoch": 1.4849363717076058,
+      "grad_norm": 1.1646493673324585,
+      "learning_rate": 1.0641379205725599e-05,
+      "loss": 0.2278,
+      "step": 3136
+    },
+    {
+      "epoch": 1.4854098845812371,
+      "grad_norm": 1.1764169931411743,
+      "learning_rate": 1.063627555011923e-05,
+      "loss": 0.1983,
+      "step": 3137
+    },
+    {
+      "epoch": 1.4858833974548684,
+      "grad_norm": 1.3643893003463745,
+      "learning_rate": 1.0631171728101129e-05,
+      "loss": 0.1969,
+      "step": 3138
+    },
+    {
+      "epoch": 1.4863569103284995,
+      "grad_norm": 1.2476388216018677,
+      "learning_rate": 1.0626067741006155e-05,
+      "loss": 0.1941,
+      "step": 3139
+    },
+    {
+      "epoch": 1.4868304232021308,
+      "grad_norm": 1.1567429304122925,
+      "learning_rate": 1.0620963590169197e-05,
+      "loss": 0.2146,
+      "step": 3140
+    },
+    {
+      "epoch": 1.487303936075762,
+      "grad_norm": 1.037000060081482,
+      "learning_rate": 1.0615859276925207e-05,
+      "loss": 0.2062,
+      "step": 3141
+    },
+    {
+      "epoch": 1.4877774489493933,
+      "grad_norm": 1.5151640176773071,
+      "learning_rate": 1.0610754802609162e-05,
+      "loss": 0.2161,
+      "step": 3142
+    },
+    {
+      "epoch": 1.4882509618230246,
+      "grad_norm": 1.4791309833526611,
+      "learning_rate": 1.060565016855609e-05,
+      "loss": 0.2273,
+      "step": 3143
+    },
+    {
+      "epoch": 1.4887244746966557,
+      "grad_norm": 1.1170471906661987,
+      "learning_rate": 1.0600545376101061e-05,
+      "loss": 0.2279,
+      "step": 3144
+    },
+    {
+      "epoch": 1.489197987570287,
+      "grad_norm": 1.664336085319519,
+      "learning_rate": 1.0595440426579182e-05,
+      "loss": 0.2168,
+      "step": 3145
+    },
+    {
+      "epoch": 1.4896715004439183,
+      "grad_norm": 1.1722487211227417,
+      "learning_rate": 1.0590335321325603e-05,
+      "loss": 0.2215,
+      "step": 3146
+    },
+    {
+      "epoch": 1.4901450133175496,
+      "grad_norm": 1.0749174356460571,
+      "learning_rate": 1.0585230061675515e-05,
+      "loss": 0.2348,
+      "step": 3147
+    },
+    {
+      "epoch": 1.4906185261911808,
+      "grad_norm": 1.5461552143096924,
+      "learning_rate": 1.0580124648964153e-05,
+      "loss": 0.2247,
+      "step": 3148
+    },
+    {
+      "epoch": 1.4910920390648121,
+      "grad_norm": 0.9260118007659912,
+      "learning_rate": 1.0575019084526785e-05,
+      "loss": 0.2142,
+      "step": 3149
+    },
+    {
+      "epoch": 1.4915655519384434,
+      "grad_norm": 1.4796782732009888,
+      "learning_rate": 1.0569913369698722e-05,
+      "loss": 0.2071,
+      "step": 3150
+    },
+    {
+      "epoch": 1.4920390648120745,
+      "grad_norm": 1.0759540796279907,
+      "learning_rate": 1.0564807505815316e-05,
+      "loss": 0.1932,
+      "step": 3151
+    },
+    {
+      "epoch": 1.4925125776857058,
+      "grad_norm": 1.082331895828247,
+      "learning_rate": 1.0559701494211953e-05,
+      "loss": 0.212,
+      "step": 3152
+    },
+    {
+      "epoch": 1.492986090559337,
+      "grad_norm": 1.0994702577590942,
+      "learning_rate": 1.0554595336224065e-05,
+      "loss": 0.2176,
+      "step": 3153
+    },
+    {
+      "epoch": 1.4934596034329684,
+      "grad_norm": 0.989517092704773,
+      "learning_rate": 1.0549489033187116e-05,
+      "loss": 0.2085,
+      "step": 3154
+    },
+    {
+      "epoch": 1.4939331163065996,
+      "grad_norm": 1.2326009273529053,
+      "learning_rate": 1.0544382586436613e-05,
+      "loss": 0.1911,
+      "step": 3155
+    },
+    {
+      "epoch": 1.4944066291802307,
+      "grad_norm": 1.4476752281188965,
+      "learning_rate": 1.0539275997308099e-05,
+      "loss": 0.2365,
+      "step": 3156
+    },
+    {
+      "epoch": 1.4948801420538622,
+      "grad_norm": 1.07553231716156,
+      "learning_rate": 1.0534169267137146e-05,
+      "loss": 0.2231,
+      "step": 3157
+    },
+    {
+      "epoch": 1.4953536549274933,
+      "grad_norm": 1.2448638677597046,
+      "learning_rate": 1.0529062397259375e-05,
+      "loss": 0.2211,
+      "step": 3158
+    },
+    {
+      "epoch": 1.4958271678011246,
+      "grad_norm": 1.1470022201538086,
+      "learning_rate": 1.0523955389010438e-05,
+      "loss": 0.2268,
+      "step": 3159
+    },
+    {
+      "epoch": 1.4963006806747559,
+      "grad_norm": 1.2596864700317383,
+      "learning_rate": 1.0518848243726026e-05,
+      "loss": 0.2212,
+      "step": 3160
+    },
+    {
+      "epoch": 1.4967741935483871,
+      "grad_norm": 1.1401036977767944,
+      "learning_rate": 1.0513740962741861e-05,
+      "loss": 0.2301,
+      "step": 3161
+    },
+    {
+      "epoch": 1.4972477064220184,
+      "grad_norm": 1.7496001720428467,
+      "learning_rate": 1.05086335473937e-05,
+      "loss": 0.2162,
+      "step": 3162
+    },
+    {
+      "epoch": 1.4977212192956495,
+      "grad_norm": 1.3748828172683716,
+      "learning_rate": 1.0503525999017343e-05,
+      "loss": 0.2101,
+      "step": 3163
+    },
+    {
+      "epoch": 1.4981947321692808,
+      "grad_norm": 1.002260446548462,
+      "learning_rate": 1.049841831894862e-05,
+      "loss": 0.2274,
+      "step": 3164
+    },
+    {
+      "epoch": 1.498668245042912,
+      "grad_norm": 1.6265779733657837,
+      "learning_rate": 1.0493310508523393e-05,
+      "loss": 0.2283,
+      "step": 3165
+    },
+    {
+      "epoch": 1.4991417579165434,
+      "grad_norm": 1.0258605480194092,
+      "learning_rate": 1.0488202569077564e-05,
+      "loss": 0.199,
+      "step": 3166
+    },
+    {
+      "epoch": 1.4996152707901746,
+      "grad_norm": 1.2962850332260132,
+      "learning_rate": 1.0483094501947062e-05,
+      "loss": 0.2447,
+      "step": 3167
+    },
+    {
+      "epoch": 1.5000887836638057,
+      "grad_norm": 0.9795529246330261,
+      "learning_rate": 1.0477986308467851e-05,
+      "loss": 0.2269,
+      "step": 3168
+    },
+    {
+      "epoch": 1.5005622965374372,
+      "grad_norm": 1.3065211772918701,
+      "learning_rate": 1.0472877989975933e-05,
+      "loss": 0.2304,
+      "step": 3169
+    },
+    {
+      "epoch": 1.5010358094110683,
+      "grad_norm": 2.2391700744628906,
+      "learning_rate": 1.046776954780734e-05,
+      "loss": 0.2273,
+      "step": 3170
+    },
+    {
+      "epoch": 1.5015093222846996,
+      "grad_norm": 1.249174952507019,
+      "learning_rate": 1.0462660983298132e-05,
+      "loss": 0.1968,
+      "step": 3171
+    },
+    {
+      "epoch": 1.5019828351583309,
+      "grad_norm": 1.0384050607681274,
+      "learning_rate": 1.045755229778441e-05,
+      "loss": 0.2087,
+      "step": 3172
+    },
+    {
+      "epoch": 1.5024563480319622,
+      "grad_norm": 1.1867655515670776,
+      "learning_rate": 1.0452443492602296e-05,
+      "loss": 0.2292,
+      "step": 3173
+    },
+    {
+      "epoch": 1.5029298609055934,
+      "grad_norm": 1.0175690650939941,
+      "learning_rate": 1.0447334569087953e-05,
+      "loss": 0.2039,
+      "step": 3174
+    },
+    {
+      "epoch": 1.5034033737792245,
+      "grad_norm": 1.395185947418213,
+      "learning_rate": 1.0442225528577568e-05,
+      "loss": 0.2102,
+      "step": 3175
+    },
+    {
+      "epoch": 1.503876886652856,
+      "grad_norm": 1.4154939651489258,
+      "learning_rate": 1.0437116372407364e-05,
+      "loss": 0.2346,
+      "step": 3176
+    },
+    {
+      "epoch": 1.504350399526487,
+      "grad_norm": 1.1730825901031494,
+      "learning_rate": 1.0432007101913588e-05,
+      "loss": 0.1973,
+      "step": 3177
+    },
+    {
+      "epoch": 1.5048239124001184,
+      "grad_norm": 1.7297412157058716,
+      "learning_rate": 1.0426897718432523e-05,
+      "loss": 0.216,
+      "step": 3178
+    },
+    {
+      "epoch": 1.5052974252737497,
+      "grad_norm": 0.9606209993362427,
+      "learning_rate": 1.0421788223300478e-05,
+      "loss": 0.2313,
+      "step": 3179
+    },
+    {
+      "epoch": 1.5057709381473807,
+      "grad_norm": 1.1701611280441284,
+      "learning_rate": 1.041667861785379e-05,
+      "loss": 0.208,
+      "step": 3180
+    },
+    {
+      "epoch": 1.5062444510210122,
+      "grad_norm": 1.2683926820755005,
+      "learning_rate": 1.0411568903428832e-05,
+      "loss": 0.229,
+      "step": 3181
+    },
+    {
+      "epoch": 1.5067179638946433,
+      "grad_norm": 1.734331727027893,
+      "learning_rate": 1.0406459081361998e-05,
+      "loss": 0.2352,
+      "step": 3182
+    },
+    {
+      "epoch": 1.5071914767682746,
+      "grad_norm": 1.5925458669662476,
+      "learning_rate": 1.0401349152989713e-05,
+      "loss": 0.2294,
+      "step": 3183
+    },
+    {
+      "epoch": 1.5076649896419059,
+      "grad_norm": 1.7187014818191528,
+      "learning_rate": 1.0396239119648426e-05,
+      "loss": 0.2182,
+      "step": 3184
+    },
+    {
+      "epoch": 1.5081385025155372,
+      "grad_norm": 1.6393986940383911,
+      "learning_rate": 1.0391128982674621e-05,
+      "loss": 0.2139,
+      "step": 3185
+    },
+    {
+      "epoch": 1.5086120153891684,
+      "grad_norm": 1.1025725603103638,
+      "learning_rate": 1.0386018743404805e-05,
+      "loss": 0.2337,
+      "step": 3186
+    },
+    {
+      "epoch": 1.5090855282627995,
+      "grad_norm": 1.29452383518219,
+      "learning_rate": 1.0380908403175509e-05,
+      "loss": 0.1877,
+      "step": 3187
+    },
+    {
+      "epoch": 1.509559041136431,
+      "grad_norm": 2.088343858718872,
+      "learning_rate": 1.0375797963323295e-05,
+      "loss": 0.2016,
+      "step": 3188
+    },
+    {
+      "epoch": 1.510032554010062,
+      "grad_norm": 1.1434667110443115,
+      "learning_rate": 1.037068742518475e-05,
+      "loss": 0.2022,
+      "step": 3189
+    },
+    {
+      "epoch": 1.5105060668836934,
+      "grad_norm": 1.0598808526992798,
+      "learning_rate": 1.0365576790096486e-05,
+      "loss": 0.2173,
+      "step": 3190
+    },
+    {
+      "epoch": 1.5109795797573247,
+      "grad_norm": 1.190727710723877,
+      "learning_rate": 1.0360466059395138e-05,
+      "loss": 0.2441,
+      "step": 3191
+    },
+    {
+      "epoch": 1.511453092630956,
+      "grad_norm": 1.5847951173782349,
+      "learning_rate": 1.0355355234417369e-05,
+      "loss": 0.2008,
+      "step": 3192
+    },
+    {
+      "epoch": 1.5119266055045872,
+      "grad_norm": 1.251030445098877,
+      "learning_rate": 1.035024431649987e-05,
+      "loss": 0.2403,
+      "step": 3193
+    },
+    {
+      "epoch": 1.5124001183782183,
+      "grad_norm": 1.6456767320632935,
+      "learning_rate": 1.0345133306979342e-05,
+      "loss": 0.2249,
+      "step": 3194
+    },
+    {
+      "epoch": 1.5128736312518498,
+      "grad_norm": 1.1858950853347778,
+      "learning_rate": 1.0340022207192532e-05,
+      "loss": 0.2287,
+      "step": 3195
+    },
+    {
+      "epoch": 1.5133471441254809,
+      "grad_norm": 1.4092812538146973,
+      "learning_rate": 1.0334911018476194e-05,
+      "loss": 0.2057,
+      "step": 3196
+    },
+    {
+      "epoch": 1.5138206569991122,
+      "grad_norm": 1.0549933910369873,
+      "learning_rate": 1.0329799742167108e-05,
+      "loss": 0.1973,
+      "step": 3197
+    },
+    {
+      "epoch": 1.5142941698727435,
+      "grad_norm": 1.5164347887039185,
+      "learning_rate": 1.0324688379602083e-05,
+      "loss": 0.2488,
+      "step": 3198
+    },
+    {
+      "epoch": 1.5147676827463745,
+      "grad_norm": 1.8977534770965576,
+      "learning_rate": 1.0319576932117945e-05,
+      "loss": 0.2014,
+      "step": 3199
+    },
+    {
+      "epoch": 1.515241195620006,
+      "grad_norm": 1.5914242267608643,
+      "learning_rate": 1.0314465401051544e-05,
+      "loss": 0.2275,
+      "step": 3200
+    },
+    {
+      "epoch": 1.515714708493637,
+      "grad_norm": 1.1438359022140503,
+      "learning_rate": 1.030935378773975e-05,
+      "loss": 0.2144,
+      "step": 3201
+    },
+    {
+      "epoch": 1.5161882213672684,
+      "grad_norm": 1.1040819883346558,
+      "learning_rate": 1.0304242093519456e-05,
+      "loss": 0.218,
+      "step": 3202
+    },
+    {
+      "epoch": 1.5166617342408997,
+      "grad_norm": 1.5802520513534546,
+      "learning_rate": 1.0299130319727576e-05,
+      "loss": 0.2124,
+      "step": 3203
+    },
+    {
+      "epoch": 1.517135247114531,
+      "grad_norm": 1.4523437023162842,
+      "learning_rate": 1.029401846770105e-05,
+      "loss": 0.2154,
+      "step": 3204
+    },
+    {
+      "epoch": 1.5176087599881622,
+      "grad_norm": 1.1811856031417847,
+      "learning_rate": 1.0288906538776831e-05,
+      "loss": 0.22,
+      "step": 3205
+    },
+    {
+      "epoch": 1.5180822728617933,
+      "grad_norm": 1.3491801023483276,
+      "learning_rate": 1.0283794534291891e-05,
+      "loss": 0.2113,
+      "step": 3206
+    },
+    {
+      "epoch": 1.5185557857354248,
+      "grad_norm": 1.7651821374893188,
+      "learning_rate": 1.027868245558323e-05,
+      "loss": 0.2177,
+      "step": 3207
+    },
+    {
+      "epoch": 1.519029298609056,
+      "grad_norm": 1.385327696800232,
+      "learning_rate": 1.0273570303987859e-05,
+      "loss": 0.2135,
+      "step": 3208
+    },
+    {
+      "epoch": 1.5195028114826872,
+      "grad_norm": 1.7392898797988892,
+      "learning_rate": 1.0268458080842815e-05,
+      "loss": 0.2174,
+      "step": 3209
+    },
+    {
+      "epoch": 1.5199763243563185,
+      "grad_norm": 1.3350859880447388,
+      "learning_rate": 1.0263345787485156e-05,
+      "loss": 0.2226,
+      "step": 3210
+    },
+    {
+      "epoch": 1.5204498372299495,
+      "grad_norm": 1.0405269861221313,
+      "learning_rate": 1.025823342525194e-05,
+      "loss": 0.2163,
+      "step": 3211
+    },
+    {
+      "epoch": 1.520923350103581,
+      "grad_norm": 1.7307740449905396,
+      "learning_rate": 1.0253120995480264e-05,
+      "loss": 0.2056,
+      "step": 3212
+    },
+    {
+      "epoch": 1.521396862977212,
+      "grad_norm": 1.300768494606018,
+      "learning_rate": 1.024800849950723e-05,
+      "loss": 0.1799,
+      "step": 3213
+    },
+    {
+      "epoch": 1.5218703758508434,
+      "grad_norm": 1.4394781589508057,
+      "learning_rate": 1.024289593866997e-05,
+      "loss": 0.2101,
+      "step": 3214
+    },
+    {
+      "epoch": 1.5223438887244747,
+      "grad_norm": 1.2956507205963135,
+      "learning_rate": 1.0237783314305621e-05,
+      "loss": 0.2143,
+      "step": 3215
+    },
+    {
+      "epoch": 1.522817401598106,
+      "grad_norm": 1.18267822265625,
+      "learning_rate": 1.0232670627751337e-05,
+      "loss": 0.2064,
+      "step": 3216
+    },
+    {
+      "epoch": 1.5232909144717373,
+      "grad_norm": 1.1501325368881226,
+      "learning_rate": 1.02275578803443e-05,
+      "loss": 0.2249,
+      "step": 3217
+    },
+    {
+      "epoch": 1.5237644273453683,
+      "grad_norm": 1.0747531652450562,
+      "learning_rate": 1.0222445073421692e-05,
+      "loss": 0.2215,
+      "step": 3218
+    },
+    {
+      "epoch": 1.5242379402189998,
+      "grad_norm": 1.061420202255249,
+      "learning_rate": 1.021733220832072e-05,
+      "loss": 0.1964,
+      "step": 3219
+    },
+    {
+      "epoch": 1.524711453092631,
+      "grad_norm": 1.1536870002746582,
+      "learning_rate": 1.0212219286378606e-05,
+      "loss": 0.2291,
+      "step": 3220
+    },
+    {
+      "epoch": 1.5251849659662622,
+      "grad_norm": 2.1566736698150635,
+      "learning_rate": 1.0207106308932585e-05,
+      "loss": 0.2479,
+      "step": 3221
+    },
+    {
+      "epoch": 1.5256584788398935,
+      "grad_norm": 1.3876272439956665,
+      "learning_rate": 1.0201993277319906e-05,
+      "loss": 0.2181,
+      "step": 3222
+    },
+    {
+      "epoch": 1.5261319917135248,
+      "grad_norm": 1.3905116319656372,
+      "learning_rate": 1.0196880192877836e-05,
+      "loss": 0.2275,
+      "step": 3223
+    },
+    {
+      "epoch": 1.526605504587156,
+      "grad_norm": 1.5620601177215576,
+      "learning_rate": 1.019176705694365e-05,
+      "loss": 0.222,
+      "step": 3224
+    },
+    {
+      "epoch": 1.5270790174607871,
+      "grad_norm": 1.212353229522705,
+      "learning_rate": 1.018665387085464e-05,
+      "loss": 0.196,
+      "step": 3225
+    },
+    {
+      "epoch": 1.5275525303344186,
+      "grad_norm": 1.0069432258605957,
+      "learning_rate": 1.018154063594811e-05,
+      "loss": 0.2234,
+      "step": 3226
+    },
+    {
+      "epoch": 1.5280260432080497,
+      "grad_norm": 1.6228710412979126,
+      "learning_rate": 1.017642735356138e-05,
+      "loss": 0.2348,
+      "step": 3227
+    },
+    {
+      "epoch": 1.528499556081681,
+      "grad_norm": 1.7563085556030273,
+      "learning_rate": 1.0171314025031777e-05,
+      "loss": 0.2149,
+      "step": 3228
+    },
+    {
+      "epoch": 1.5289730689553123,
+      "grad_norm": 1.539036750793457,
+      "learning_rate": 1.0166200651696642e-05,
+      "loss": 0.2116,
+      "step": 3229
+    },
+    {
+      "epoch": 1.5294465818289433,
+      "grad_norm": 2.3316597938537598,
+      "learning_rate": 1.0161087234893332e-05,
+      "loss": 0.2123,
+      "step": 3230
+    },
+    {
+      "epoch": 1.5299200947025748,
+      "grad_norm": 1.3801292181015015,
+      "learning_rate": 1.0155973775959209e-05,
+      "loss": 0.2363,
+      "step": 3231
+    },
+    {
+      "epoch": 1.530393607576206,
+      "grad_norm": 1.6556155681610107,
+      "learning_rate": 1.0150860276231649e-05,
+      "loss": 0.2092,
+      "step": 3232
+    },
+    {
+      "epoch": 1.5308671204498372,
+      "grad_norm": 1.1947085857391357,
+      "learning_rate": 1.0145746737048037e-05,
+      "loss": 0.2128,
+      "step": 3233
+    },
+    {
+      "epoch": 1.5313406333234685,
+      "grad_norm": 1.491898775100708,
+      "learning_rate": 1.0140633159745775e-05,
+      "loss": 0.2195,
+      "step": 3234
+    },
+    {
+      "epoch": 1.5318141461970998,
+      "grad_norm": 1.2863268852233887,
+      "learning_rate": 1.0135519545662267e-05,
+      "loss": 0.2168,
+      "step": 3235
+    },
+    {
+      "epoch": 1.532287659070731,
+      "grad_norm": 2.18508243560791,
+      "learning_rate": 1.013040589613493e-05,
+      "loss": 0.2423,
+      "step": 3236
+    },
+    {
+      "epoch": 1.5327611719443621,
+      "grad_norm": 0.9997796416282654,
+      "learning_rate": 1.0125292212501186e-05,
+      "loss": 0.2101,
+      "step": 3237
+    },
+    {
+      "epoch": 1.5332346848179936,
+      "grad_norm": 3.1009128093719482,
+      "learning_rate": 1.0120178496098474e-05,
+      "loss": 0.2044,
+      "step": 3238
+    },
+    {
+      "epoch": 1.5337081976916247,
+      "grad_norm": 2.1549055576324463,
+      "learning_rate": 1.0115064748264236e-05,
+      "loss": 0.2169,
+      "step": 3239
+    },
+    {
+      "epoch": 1.534181710565256,
+      "grad_norm": 1.5560920238494873,
+      "learning_rate": 1.0109950970335923e-05,
+      "loss": 0.2285,
+      "step": 3240
+    },
+    {
+      "epoch": 1.5346552234388873,
+      "grad_norm": 2.1575727462768555,
+      "learning_rate": 1.0104837163650994e-05,
+      "loss": 0.2077,
+      "step": 3241
+    },
+    {
+      "epoch": 1.5351287363125183,
+      "grad_norm": 2.3875033855438232,
+      "learning_rate": 1.0099723329546915e-05,
+      "loss": 0.2247,
+      "step": 3242
+    },
+    {
+      "epoch": 1.5356022491861498,
+      "grad_norm": 1.0770213603973389,
+      "learning_rate": 1.0094609469361162e-05,
+      "loss": 0.215,
+      "step": 3243
+    },
+    {
+      "epoch": 1.536075762059781,
+      "grad_norm": 1.1401655673980713,
+      "learning_rate": 1.0089495584431217e-05,
+      "loss": 0.2126,
+      "step": 3244
+    },
+    {
+      "epoch": 1.5365492749334122,
+      "grad_norm": 1.0763803720474243,
+      "learning_rate": 1.0084381676094566e-05,
+      "loss": 0.2448,
+      "step": 3245
+    },
+    {
+      "epoch": 1.5370227878070435,
+      "grad_norm": 1.8991236686706543,
+      "learning_rate": 1.00792677456887e-05,
+      "loss": 0.2243,
+      "step": 3246
+    },
+    {
+      "epoch": 1.5374963006806748,
+      "grad_norm": 2.337759256362915,
+      "learning_rate": 1.0074153794551119e-05,
+      "loss": 0.2146,
+      "step": 3247
+    },
+    {
+      "epoch": 1.537969813554306,
+      "grad_norm": 2.1561119556427,
+      "learning_rate": 1.0069039824019326e-05,
+      "loss": 0.195,
+      "step": 3248
+    },
+    {
+      "epoch": 1.5384433264279371,
+      "grad_norm": 2.016658306121826,
+      "learning_rate": 1.0063925835430838e-05,
+      "loss": 0.2334,
+      "step": 3249
+    },
+    {
+      "epoch": 1.5389168393015686,
+      "grad_norm": 2.295039653778076,
+      "learning_rate": 1.005881183012316e-05,
+      "loss": 0.2266,
+      "step": 3250
+    },
+    {
+      "epoch": 1.5393903521751997,
+      "grad_norm": 2.0591042041778564,
+      "learning_rate": 1.0053697809433817e-05,
+      "loss": 0.2052,
+      "step": 3251
+    },
+    {
+      "epoch": 1.539863865048831,
+      "grad_norm": 1.625144362449646,
+      "learning_rate": 1.004858377470033e-05,
+      "loss": 0.2076,
+      "step": 3252
+    },
+    {
+      "epoch": 1.5403373779224623,
+      "grad_norm": 0.9673202633857727,
+      "learning_rate": 1.0043469727260228e-05,
+      "loss": 0.2103,
+      "step": 3253
+    },
+    {
+      "epoch": 1.5408108907960936,
+      "grad_norm": 1.1474332809448242,
+      "learning_rate": 1.0038355668451037e-05,
+      "loss": 0.214,
+      "step": 3254
+    },
+    {
+      "epoch": 1.5412844036697249,
+      "grad_norm": 1.004038691520691,
+      "learning_rate": 1.0033241599610288e-05,
+      "loss": 0.1771,
+      "step": 3255
+    },
+    {
+      "epoch": 1.541757916543356,
+      "grad_norm": 1.4131745100021362,
+      "learning_rate": 1.0028127522075522e-05,
+      "loss": 0.2241,
+      "step": 3256
+    },
+    {
+      "epoch": 1.5422314294169874,
+      "grad_norm": 2.0983598232269287,
+      "learning_rate": 1.0023013437184273e-05,
+      "loss": 0.2309,
+      "step": 3257
+    },
+    {
+      "epoch": 1.5427049422906185,
+      "grad_norm": 1.784072995185852,
+      "learning_rate": 1.0017899346274082e-05,
+      "loss": 0.2212,
+      "step": 3258
+    },
+    {
+      "epoch": 1.5431784551642498,
+      "grad_norm": 2.016923427581787,
+      "learning_rate": 1.0012785250682488e-05,
+      "loss": 0.2223,
+      "step": 3259
+    },
+    {
+      "epoch": 1.543651968037881,
+      "grad_norm": 1.3807404041290283,
+      "learning_rate": 1.0007671151747038e-05,
+      "loss": 0.2082,
+      "step": 3260
+    },
+    {
+      "epoch": 1.5441254809115121,
+      "grad_norm": 1.9663400650024414,
+      "learning_rate": 1.000255705080527e-05,
+      "loss": 0.2303,
+      "step": 3261
+    },
+    {
+      "epoch": 1.5445989937851436,
+      "grad_norm": 1.2780094146728516,
+      "learning_rate": 9.997442949194733e-06,
+      "loss": 0.2188,
+      "step": 3262
+    },
+    {
+      "epoch": 1.5450725066587747,
+      "grad_norm": 1.079207181930542,
+      "learning_rate": 9.992328848252965e-06,
+      "loss": 0.2033,
+      "step": 3263
+    },
+    {
+      "epoch": 1.545546019532406,
+      "grad_norm": 1.1068617105484009,
+      "learning_rate": 9.987214749317514e-06,
+      "loss": 0.2441,
+      "step": 3264
+    },
+    {
+      "epoch": 1.5460195324060373,
+      "grad_norm": 1.965084195137024,
+      "learning_rate": 9.982100653725921e-06,
+      "loss": 0.1883,
+      "step": 3265
+    },
+    {
+      "epoch": 1.5464930452796686,
+      "grad_norm": 2.6802639961242676,
+      "learning_rate": 9.97698656281573e-06,
+      "loss": 0.2319,
+      "step": 3266
+    },
+    {
+      "epoch": 1.5469665581532999,
+      "grad_norm": 1.9177919626235962,
+      "learning_rate": 9.971872477924482e-06,
+      "loss": 0.2197,
+      "step": 3267
+    },
+    {
+      "epoch": 1.547440071026931,
+      "grad_norm": 2.636075735092163,
+      "learning_rate": 9.966758400389714e-06,
+      "loss": 0.2109,
+      "step": 3268
+    },
+    {
+      "epoch": 1.5479135839005624,
+      "grad_norm": 2.6643176078796387,
+      "learning_rate": 9.961644331548967e-06,
+      "loss": 0.2137,
+      "step": 3269
+    },
+    {
+      "epoch": 1.5483870967741935,
+      "grad_norm": 2.1698949337005615,
+      "learning_rate": 9.956530272739775e-06,
+      "loss": 0.2371,
+      "step": 3270
+    },
+    {
+      "epoch": 1.5488606096478248,
+      "grad_norm": 1.0527149438858032,
+      "learning_rate": 9.951416225299671e-06,
+      "loss": 0.2027,
+      "step": 3271
+    },
+    {
+      "epoch": 1.549334122521456,
+      "grad_norm": 1.1439855098724365,
+      "learning_rate": 9.946302190566186e-06,
+      "loss": 0.2106,
+      "step": 3272
+    },
+    {
+      "epoch": 1.5498076353950871,
+      "grad_norm": 1.49906325340271,
+      "learning_rate": 9.941188169876843e-06,
+      "loss": 0.2192,
+      "step": 3273
+    },
+    {
+      "epoch": 1.5502811482687187,
+      "grad_norm": 1.464437484741211,
+      "learning_rate": 9.936074164569168e-06,
+      "loss": 0.2061,
+      "step": 3274
+    },
+    {
+      "epoch": 1.5507546611423497,
+      "grad_norm": 2.1014366149902344,
+      "learning_rate": 9.930960175980677e-06,
+      "loss": 0.2131,
+      "step": 3275
+    },
+    {
+      "epoch": 1.551228174015981,
+      "grad_norm": 1.1291313171386719,
+      "learning_rate": 9.925846205448886e-06,
+      "loss": 0.2004,
+      "step": 3276
+    },
+    {
+      "epoch": 1.5517016868896123,
+      "grad_norm": 2.514810800552368,
+      "learning_rate": 9.920732254311306e-06,
+      "loss": 0.1979,
+      "step": 3277
+    },
+    {
+      "epoch": 1.5521751997632436,
+      "grad_norm": 1.7324765920639038,
+      "learning_rate": 9.91561832390544e-06,
+      "loss": 0.1977,
+      "step": 3278
+    },
+    {
+      "epoch": 1.5526487126368749,
+      "grad_norm": 1.4034252166748047,
+      "learning_rate": 9.910504415568788e-06,
+      "loss": 0.2243,
+      "step": 3279
+    },
+    {
+      "epoch": 1.553122225510506,
+      "grad_norm": 1.4264874458312988,
+      "learning_rate": 9.90539053063884e-06,
+      "loss": 0.2129,
+      "step": 3280
+    },
+    {
+      "epoch": 1.5535957383841374,
+      "grad_norm": 1.2019026279449463,
+      "learning_rate": 9.900276670453085e-06,
+      "loss": 0.2052,
+      "step": 3281
+    },
+    {
+      "epoch": 1.5540692512577685,
+      "grad_norm": 1.222839593887329,
+      "learning_rate": 9.895162836349006e-06,
+      "loss": 0.2145,
+      "step": 3282
+    },
+    {
+      "epoch": 1.5545427641313998,
+      "grad_norm": 1.3310260772705078,
+      "learning_rate": 9.890049029664079e-06,
+      "loss": 0.2224,
+      "step": 3283
+    },
+    {
+      "epoch": 1.555016277005031,
+      "grad_norm": 1.5830497741699219,
+      "learning_rate": 9.884935251735766e-06,
+      "loss": 0.2102,
+      "step": 3284
+    },
+    {
+      "epoch": 1.5554897898786624,
+      "grad_norm": 1.1304932832717896,
+      "learning_rate": 9.879821503901527e-06,
+      "loss": 0.2128,
+      "step": 3285
+    },
+    {
+      "epoch": 1.5559633027522937,
+      "grad_norm": 1.5791419744491577,
+      "learning_rate": 9.874707787498814e-06,
+      "loss": 0.2253,
+      "step": 3286
+    },
+    {
+      "epoch": 1.5564368156259247,
+      "grad_norm": 2.089458465576172,
+      "learning_rate": 9.869594103865074e-06,
+      "loss": 0.209,
+      "step": 3287
+    },
+    {
+      "epoch": 1.5569103284995562,
+      "grad_norm": 1.063925862312317,
+      "learning_rate": 9.864480454337735e-06,
+      "loss": 0.2186,
+      "step": 3288
+    },
+    {
+      "epoch": 1.5573838413731873,
+      "grad_norm": 1.1000468730926514,
+      "learning_rate": 9.859366840254227e-06,
+      "loss": 0.212,
+      "step": 3289
+    },
+    {
+      "epoch": 1.5578573542468186,
+      "grad_norm": 1.3365740776062012,
+      "learning_rate": 9.854253262951964e-06,
+      "loss": 0.1934,
+      "step": 3290
+    },
+    {
+      "epoch": 1.5583308671204499,
+      "grad_norm": 1.6963863372802734,
+      "learning_rate": 9.849139723768354e-06,
+      "loss": 0.2115,
+      "step": 3291
+    },
+    {
+      "epoch": 1.558804379994081,
+      "grad_norm": 1.2676887512207031,
+      "learning_rate": 9.844026224040794e-06,
+      "loss": 0.2232,
+      "step": 3292
+    },
+    {
+      "epoch": 1.5592778928677125,
+      "grad_norm": 1.4118210077285767,
+      "learning_rate": 9.838912765106671e-06,
+      "loss": 0.2005,
+      "step": 3293
+    },
+    {
+      "epoch": 1.5597514057413435,
+      "grad_norm": 1.1821280717849731,
+      "learning_rate": 9.83379934830336e-06,
+      "loss": 0.2133,
+      "step": 3294
+    },
+    {
+      "epoch": 1.5602249186149748,
+      "grad_norm": 1.2033153772354126,
+      "learning_rate": 9.828685974968224e-06,
+      "loss": 0.2165,
+      "step": 3295
+    },
+    {
+      "epoch": 1.560698431488606,
+      "grad_norm": 1.9517135620117188,
+      "learning_rate": 9.823572646438622e-06,
+      "loss": 0.2394,
+      "step": 3296
+    },
+    {
+      "epoch": 1.5611719443622374,
+      "grad_norm": 1.292248249053955,
+      "learning_rate": 9.818459364051893e-06,
+      "loss": 0.2039,
+      "step": 3297
+    },
+    {
+      "epoch": 1.5616454572358687,
+      "grad_norm": 1.03531813621521,
+      "learning_rate": 9.813346129145364e-06,
+      "loss": 0.2124,
+      "step": 3298
+    },
+    {
+      "epoch": 1.5621189701094997,
+      "grad_norm": 1.404657006263733,
+      "learning_rate": 9.808232943056354e-06,
+      "loss": 0.2242,
+      "step": 3299
+    },
+    {
+      "epoch": 1.5625924829831312,
+      "grad_norm": 1.1680656671524048,
+      "learning_rate": 9.803119807122167e-06,
+      "loss": 0.193,
+      "step": 3300
+    },
+    {
+      "epoch": 1.5630659958567623,
+      "grad_norm": 1.476773738861084,
+      "learning_rate": 9.798006722680096e-06,
+      "loss": 0.2155,
+      "step": 3301
+    },
+    {
+      "epoch": 1.5635395087303936,
+      "grad_norm": 1.3005181550979614,
+      "learning_rate": 9.792893691067417e-06,
+      "loss": 0.1809,
+      "step": 3302
+    },
+    {
+      "epoch": 1.5640130216040249,
+      "grad_norm": 1.2317800521850586,
+      "learning_rate": 9.787780713621397e-06,
+      "loss": 0.2186,
+      "step": 3303
+    },
+    {
+      "epoch": 1.564486534477656,
+      "grad_norm": 1.0465822219848633,
+      "learning_rate": 9.782667791679283e-06,
+      "loss": 0.2259,
+      "step": 3304
+    },
+    {
+      "epoch": 1.5649600473512875,
+      "grad_norm": 1.5197325944900513,
+      "learning_rate": 9.777554926578311e-06,
+      "loss": 0.237,
+      "step": 3305
+    },
+    {
+      "epoch": 1.5654335602249185,
+      "grad_norm": 1.2708994150161743,
+      "learning_rate": 9.772442119655706e-06,
+      "loss": 0.2116,
+      "step": 3306
+    },
+    {
+      "epoch": 1.5659070730985498,
+      "grad_norm": 1.392824411392212,
+      "learning_rate": 9.767329372248666e-06,
+      "loss": 0.2197,
+      "step": 3307
+    },
+    {
+      "epoch": 1.566380585972181,
+      "grad_norm": 2.88095760345459,
+      "learning_rate": 9.762216685694382e-06,
+      "loss": 0.2157,
+      "step": 3308
+    },
+    {
+      "epoch": 1.5668540988458124,
+      "grad_norm": 1.6576155424118042,
+      "learning_rate": 9.757104061330033e-06,
+      "loss": 0.1998,
+      "step": 3309
+    },
+    {
+      "epoch": 1.5673276117194437,
+      "grad_norm": 1.031028151512146,
+      "learning_rate": 9.751991500492772e-06,
+      "loss": 0.2168,
+      "step": 3310
+    },
+    {
+      "epoch": 1.5678011245930747,
+      "grad_norm": 2.0515482425689697,
+      "learning_rate": 9.746879004519741e-06,
+      "loss": 0.1973,
+      "step": 3311
+    },
+    {
+      "epoch": 1.5682746374667063,
+      "grad_norm": 1.582655668258667,
+      "learning_rate": 9.741766574748066e-06,
+      "loss": 0.2349,
+      "step": 3312
+    },
+    {
+      "epoch": 1.5687481503403373,
+      "grad_norm": 0.992260754108429,
+      "learning_rate": 9.736654212514851e-06,
+      "loss": 0.1884,
+      "step": 3313
+    },
+    {
+      "epoch": 1.5692216632139686,
+      "grad_norm": 0.8423192501068115,
+      "learning_rate": 9.731541919157186e-06,
+      "loss": 0.2027,
+      "step": 3314
+    },
+    {
+      "epoch": 1.5696951760876,
+      "grad_norm": 1.082379937171936,
+      "learning_rate": 9.726429696012143e-06,
+      "loss": 0.207,
+      "step": 3315
+    },
+    {
+      "epoch": 1.5701686889612312,
+      "grad_norm": 1.25356125831604,
+      "learning_rate": 9.721317544416775e-06,
+      "loss": 0.2285,
+      "step": 3316
+    },
+    {
+      "epoch": 1.5706422018348625,
+      "grad_norm": 1.3357824087142944,
+      "learning_rate": 9.716205465708114e-06,
+      "loss": 0.1885,
+      "step": 3317
+    },
+    {
+      "epoch": 1.5711157147084935,
+      "grad_norm": 1.4782239198684692,
+      "learning_rate": 9.711093461223175e-06,
+      "loss": 0.1942,
+      "step": 3318
+    },
+    {
+      "epoch": 1.571589227582125,
+      "grad_norm": 1.226721167564392,
+      "learning_rate": 9.705981532298955e-06,
+      "loss": 0.2043,
+      "step": 3319
+    },
+    {
+      "epoch": 1.5720627404557561,
+      "grad_norm": 1.5207408666610718,
+      "learning_rate": 9.700869680272422e-06,
+      "loss": 0.2314,
+      "step": 3320
+    },
+    {
+      "epoch": 1.5725362533293874,
+      "grad_norm": 1.0883424282073975,
+      "learning_rate": 9.695757906480545e-06,
+      "loss": 0.2212,
+      "step": 3321
+    },
+    {
+      "epoch": 1.5730097662030187,
+      "grad_norm": 1.4195133447647095,
+      "learning_rate": 9.690646212260254e-06,
+      "loss": 0.2189,
+      "step": 3322
+    },
+    {
+      "epoch": 1.5734832790766498,
+      "grad_norm": 2.070651054382324,
+      "learning_rate": 9.68553459894846e-06,
+      "loss": 0.2199,
+      "step": 3323
+    },
+    {
+      "epoch": 1.5739567919502813,
+      "grad_norm": 1.0643229484558105,
+      "learning_rate": 9.680423067882057e-06,
+      "loss": 0.1874,
+      "step": 3324
+    },
+    {
+      "epoch": 1.5744303048239123,
+      "grad_norm": 2.199005365371704,
+      "learning_rate": 9.675311620397917e-06,
+      "loss": 0.2141,
+      "step": 3325
+    },
+    {
+      "epoch": 1.5749038176975436,
+      "grad_norm": 1.1408580541610718,
+      "learning_rate": 9.670200257832891e-06,
+      "loss": 0.1889,
+      "step": 3326
+    },
+    {
+      "epoch": 1.575377330571175,
+      "grad_norm": 1.1677594184875488,
+      "learning_rate": 9.665088981523807e-06,
+      "loss": 0.2221,
+      "step": 3327
+    },
+    {
+      "epoch": 1.5758508434448062,
+      "grad_norm": 1.3932559490203857,
+      "learning_rate": 9.659977792807468e-06,
+      "loss": 0.2027,
+      "step": 3328
+    },
+    {
+      "epoch": 1.5763243563184375,
+      "grad_norm": 1.122646689414978,
+      "learning_rate": 9.654866693020656e-06,
+      "loss": 0.1988,
+      "step": 3329
+    },
+    {
+      "epoch": 1.5767978691920685,
+      "grad_norm": 1.0918054580688477,
+      "learning_rate": 9.649755683500134e-06,
+      "loss": 0.2109,
+      "step": 3330
+    },
+    {
+      "epoch": 1.5772713820657,
+      "grad_norm": 1.0284793376922607,
+      "learning_rate": 9.644644765582633e-06,
+      "loss": 0.1926,
+      "step": 3331
+    },
+    {
+      "epoch": 1.5777448949393311,
+      "grad_norm": 1.006220817565918,
+      "learning_rate": 9.639533940604867e-06,
+      "loss": 0.23,
+      "step": 3332
+    },
+    {
+      "epoch": 1.5782184078129624,
+      "grad_norm": 1.0561009645462036,
+      "learning_rate": 9.634423209903518e-06,
+      "loss": 0.2246,
+      "step": 3333
+    },
+    {
+      "epoch": 1.5786919206865937,
+      "grad_norm": 1.0437111854553223,
+      "learning_rate": 9.629312574815251e-06,
+      "loss": 0.2267,
+      "step": 3334
+    },
+    {
+      "epoch": 1.5791654335602248,
+      "grad_norm": 2.093397617340088,
+      "learning_rate": 9.624202036676707e-06,
+      "loss": 0.2099,
+      "step": 3335
+    },
+    {
+      "epoch": 1.5796389464338563,
+      "grad_norm": 2.2291595935821533,
+      "learning_rate": 9.619091596824493e-06,
+      "loss": 0.2153,
+      "step": 3336
+    },
+    {
+      "epoch": 1.5801124593074873,
+      "grad_norm": 1.3327558040618896,
+      "learning_rate": 9.613981256595199e-06,
+      "loss": 0.1965,
+      "step": 3337
+    },
+    {
+      "epoch": 1.5805859721811186,
+      "grad_norm": 0.9187760353088379,
+      "learning_rate": 9.60887101732538e-06,
+      "loss": 0.2022,
+      "step": 3338
+    },
+    {
+      "epoch": 1.58105948505475,
+      "grad_norm": 1.6713463068008423,
+      "learning_rate": 9.603760880351576e-06,
+      "loss": 0.2052,
+      "step": 3339
+    },
+    {
+      "epoch": 1.5815329979283812,
+      "grad_norm": 1.2200876474380493,
+      "learning_rate": 9.59865084701029e-06,
+      "loss": 0.2185,
+      "step": 3340
+    },
+    {
+      "epoch": 1.5820065108020125,
+      "grad_norm": 1.1830116510391235,
+      "learning_rate": 9.593540918638006e-06,
+      "loss": 0.1816,
+      "step": 3341
+    },
+    {
+      "epoch": 1.5824800236756436,
+      "grad_norm": 1.3431400060653687,
+      "learning_rate": 9.588431096571171e-06,
+      "loss": 0.2059,
+      "step": 3342
+    },
+    {
+      "epoch": 1.582953536549275,
+      "grad_norm": 1.1804959774017334,
+      "learning_rate": 9.583321382146212e-06,
+      "loss": 0.2229,
+      "step": 3343
+    },
+    {
+      "epoch": 1.5834270494229061,
+      "grad_norm": 1.8426966667175293,
+      "learning_rate": 9.578211776699527e-06,
+      "loss": 0.2133,
+      "step": 3344
+    },
+    {
+      "epoch": 1.5839005622965374,
+      "grad_norm": 1.2378004789352417,
+      "learning_rate": 9.57310228156748e-06,
+      "loss": 0.236,
+      "step": 3345
+    },
+    {
+      "epoch": 1.5843740751701687,
+      "grad_norm": 1.0932939052581787,
+      "learning_rate": 9.567992898086415e-06,
+      "loss": 0.2039,
+      "step": 3346
+    },
+    {
+      "epoch": 1.5848475880438,
+      "grad_norm": 1.0567042827606201,
+      "learning_rate": 9.56288362759264e-06,
+      "loss": 0.1963,
+      "step": 3347
+    },
+    {
+      "epoch": 1.5853211009174313,
+      "grad_norm": 1.230892300605774,
+      "learning_rate": 9.557774471422434e-06,
+      "loss": 0.2288,
+      "step": 3348
+    },
+    {
+      "epoch": 1.5857946137910623,
+      "grad_norm": 1.4192466735839844,
+      "learning_rate": 9.552665430912049e-06,
+      "loss": 0.2143,
+      "step": 3349
+    },
+    {
+      "epoch": 1.5862681266646939,
+      "grad_norm": 1.3479020595550537,
+      "learning_rate": 9.547556507397705e-06,
+      "loss": 0.2058,
+      "step": 3350
+    },
+    {
+      "epoch": 1.586741639538325,
+      "grad_norm": 1.4939135313034058,
+      "learning_rate": 9.542447702215596e-06,
+      "loss": 0.2394,
+      "step": 3351
+    },
+    {
+      "epoch": 1.5872151524119562,
+      "grad_norm": 1.3415539264678955,
+      "learning_rate": 9.537339016701871e-06,
+      "loss": 0.2256,
+      "step": 3352
+    },
+    {
+      "epoch": 1.5876886652855875,
+      "grad_norm": 1.2717926502227783,
+      "learning_rate": 9.532230452192666e-06,
+      "loss": 0.2284,
+      "step": 3353
+    },
+    {
+      "epoch": 1.5881621781592186,
+      "grad_norm": 1.1318105459213257,
+      "learning_rate": 9.527122010024072e-06,
+      "loss": 0.2142,
+      "step": 3354
+    },
+    {
+      "epoch": 1.58863569103285,
+      "grad_norm": 1.1898558139801025,
+      "learning_rate": 9.522013691532154e-06,
+      "loss": 0.2224,
+      "step": 3355
+    },
+    {
+      "epoch": 1.5891092039064811,
+      "grad_norm": 1.3306890726089478,
+      "learning_rate": 9.516905498052944e-06,
+      "loss": 0.2193,
+      "step": 3356
+    },
+    {
+      "epoch": 1.5895827167801124,
+      "grad_norm": 1.2777800559997559,
+      "learning_rate": 9.51179743092244e-06,
+      "loss": 0.2109,
+      "step": 3357
+    },
+    {
+      "epoch": 1.5900562296537437,
+      "grad_norm": 1.1809998750686646,
+      "learning_rate": 9.50668949147661e-06,
+      "loss": 0.2307,
+      "step": 3358
+    },
+    {
+      "epoch": 1.590529742527375,
+      "grad_norm": 1.5275858640670776,
+      "learning_rate": 9.50158168105138e-06,
+      "loss": 0.2074,
+      "step": 3359
+    },
+    {
+      "epoch": 1.5910032554010063,
+      "grad_norm": 1.153929352760315,
+      "learning_rate": 9.496474000982657e-06,
+      "loss": 0.1873,
+      "step": 3360
+    },
+    {
+      "epoch": 1.5914767682746374,
+      "grad_norm": 1.1171983480453491,
+      "learning_rate": 9.4913664526063e-06,
+      "loss": 0.2069,
+      "step": 3361
+    },
+    {
+      "epoch": 1.5919502811482689,
+      "grad_norm": 1.2041152715682983,
+      "learning_rate": 9.48625903725814e-06,
+      "loss": 0.2075,
+      "step": 3362
+    },
+    {
+      "epoch": 1.5924237940219,
+      "grad_norm": 2.3762893676757812,
+      "learning_rate": 9.481151756273976e-06,
+      "loss": 0.2169,
+      "step": 3363
+    },
+    {
+      "epoch": 1.5928973068955312,
+      "grad_norm": 3.4273829460144043,
+      "learning_rate": 9.476044610989562e-06,
+      "loss": 0.2338,
+      "step": 3364
+    },
+    {
+      "epoch": 1.5933708197691625,
+      "grad_norm": 0.9142179489135742,
+      "learning_rate": 9.470937602740624e-06,
+      "loss": 0.216,
+      "step": 3365
+    },
+    {
+      "epoch": 1.5938443326427936,
+      "grad_norm": 1.2260160446166992,
+      "learning_rate": 9.465830732862857e-06,
+      "loss": 0.2242,
+      "step": 3366
+    },
+    {
+      "epoch": 1.594317845516425,
+      "grad_norm": 1.3920302391052246,
+      "learning_rate": 9.460724002691906e-06,
+      "loss": 0.216,
+      "step": 3367
+    },
+    {
+      "epoch": 1.5947913583900561,
+      "grad_norm": 0.9747047424316406,
+      "learning_rate": 9.455617413563389e-06,
+      "loss": 0.2445,
+      "step": 3368
+    },
+    {
+      "epoch": 1.5952648712636874,
+      "grad_norm": 1.1095210313796997,
+      "learning_rate": 9.450510966812885e-06,
+      "loss": 0.1813,
+      "step": 3369
+    },
+    {
+      "epoch": 1.5957383841373187,
+      "grad_norm": 1.0806443691253662,
+      "learning_rate": 9.445404663775938e-06,
+      "loss": 0.2429,
+      "step": 3370
+    },
+    {
+      "epoch": 1.59621189701095,
+      "grad_norm": 1.3196625709533691,
+      "learning_rate": 9.44029850578805e-06,
+      "loss": 0.244,
+      "step": 3371
+    },
+    {
+      "epoch": 1.5966854098845813,
+      "grad_norm": 2.6262784004211426,
+      "learning_rate": 9.435192494184689e-06,
+      "loss": 0.2116,
+      "step": 3372
+    },
+    {
+      "epoch": 1.5971589227582124,
+      "grad_norm": 1.1327464580535889,
+      "learning_rate": 9.43008663030128e-06,
+      "loss": 0.1857,
+      "step": 3373
+    },
+    {
+      "epoch": 1.5976324356318439,
+      "grad_norm": 0.9777623414993286,
+      "learning_rate": 9.424980915473217e-06,
+      "loss": 0.2091,
+      "step": 3374
+    },
+    {
+      "epoch": 1.598105948505475,
+      "grad_norm": 1.3106697797775269,
+      "learning_rate": 9.419875351035848e-06,
+      "loss": 0.213,
+      "step": 3375
+    },
+    {
+      "epoch": 1.5985794613791062,
+      "grad_norm": 1.0573484897613525,
+      "learning_rate": 9.414769938324487e-06,
+      "loss": 0.2028,
+      "step": 3376
+    },
+    {
+      "epoch": 1.5990529742527375,
+      "grad_norm": 1.3696736097335815,
+      "learning_rate": 9.4096646786744e-06,
+      "loss": 0.2101,
+      "step": 3377
+    },
+    {
+      "epoch": 1.5995264871263688,
+      "grad_norm": 1.2642654180526733,
+      "learning_rate": 9.404559573420822e-06,
+      "loss": 0.2352,
+      "step": 3378
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.4823346138000488,
+      "learning_rate": 9.399454623898942e-06,
+      "loss": 0.2035,
+      "step": 3379
+    },
+    {
+      "epoch": 1.6004735128736312,
+      "grad_norm": 1.1810314655303955,
+      "learning_rate": 9.394349831443912e-06,
+      "loss": 0.2161,
+      "step": 3380
+    },
+    {
+      "epoch": 1.6009470257472627,
+      "grad_norm": 1.377890706062317,
+      "learning_rate": 9.389245197390842e-06,
+      "loss": 0.2498,
+      "step": 3381
+    },
+    {
+      "epoch": 1.6014205386208937,
+      "grad_norm": 1.6414611339569092,
+      "learning_rate": 9.384140723074796e-06,
+      "loss": 0.2296,
+      "step": 3382
+    },
+    {
+      "epoch": 1.601894051494525,
+      "grad_norm": 1.2565416097640991,
+      "learning_rate": 9.379036409830804e-06,
+      "loss": 0.2056,
+      "step": 3383
+    },
+    {
+      "epoch": 1.6023675643681563,
+      "grad_norm": 1.1565498113632202,
+      "learning_rate": 9.37393225899385e-06,
+      "loss": 0.2115,
+      "step": 3384
+    },
+    {
+      "epoch": 1.6028410772417874,
+      "grad_norm": 1.0674059391021729,
+      "learning_rate": 9.368828271898874e-06,
+      "loss": 0.2103,
+      "step": 3385
+    },
+    {
+      "epoch": 1.6033145901154189,
+      "grad_norm": 0.9259766340255737,
+      "learning_rate": 9.363724449880773e-06,
+      "loss": 0.1965,
+      "step": 3386
+    },
+    {
+      "epoch": 1.60378810298905,
+      "grad_norm": 1.08977472782135,
+      "learning_rate": 9.358620794274404e-06,
+      "loss": 0.1933,
+      "step": 3387
+    },
+    {
+      "epoch": 1.6042616158626812,
+      "grad_norm": 1.5023025274276733,
+      "learning_rate": 9.35351730641458e-06,
+      "loss": 0.2293,
+      "step": 3388
+    },
+    {
+      "epoch": 1.6047351287363125,
+      "grad_norm": 1.0194486379623413,
+      "learning_rate": 9.348413987636065e-06,
+      "loss": 0.2344,
+      "step": 3389
+    },
+    {
+      "epoch": 1.6052086416099438,
+      "grad_norm": 2.0349748134613037,
+      "learning_rate": 9.343310839273587e-06,
+      "loss": 0.2097,
+      "step": 3390
+    },
+    {
+      "epoch": 1.605682154483575,
+      "grad_norm": 0.9003639221191406,
+      "learning_rate": 9.338207862661824e-06,
+      "loss": 0.2193,
+      "step": 3391
+    },
+    {
+      "epoch": 1.6061556673572062,
+      "grad_norm": 1.3183727264404297,
+      "learning_rate": 9.33310505913541e-06,
+      "loss": 0.2418,
+      "step": 3392
+    },
+    {
+      "epoch": 1.6066291802308377,
+      "grad_norm": 1.593930721282959,
+      "learning_rate": 9.328002430028932e-06,
+      "loss": 0.1988,
+      "step": 3393
+    },
+    {
+      "epoch": 1.6071026931044687,
+      "grad_norm": 1.2510972023010254,
+      "learning_rate": 9.322899976676938e-06,
+      "loss": 0.201,
+      "step": 3394
+    },
+    {
+      "epoch": 1.6075762059781,
+      "grad_norm": 1.210264801979065,
+      "learning_rate": 9.317797700413925e-06,
+      "loss": 0.2076,
+      "step": 3395
+    },
+    {
+      "epoch": 1.6080497188517313,
+      "grad_norm": 1.0849723815917969,
+      "learning_rate": 9.31269560257434e-06,
+      "loss": 0.2217,
+      "step": 3396
+    },
+    {
+      "epoch": 1.6085232317253624,
+      "grad_norm": 1.2907907962799072,
+      "learning_rate": 9.307593684492588e-06,
+      "loss": 0.2071,
+      "step": 3397
+    },
+    {
+      "epoch": 1.6089967445989939,
+      "grad_norm": 1.6149160861968994,
+      "learning_rate": 9.302491947503027e-06,
+      "loss": 0.2145,
+      "step": 3398
+    },
+    {
+      "epoch": 1.609470257472625,
+      "grad_norm": 1.4612150192260742,
+      "learning_rate": 9.29739039293997e-06,
+      "loss": 0.2184,
+      "step": 3399
+    },
+    {
+      "epoch": 1.6099437703462562,
+      "grad_norm": 1.7195252180099487,
+      "learning_rate": 9.292289022137678e-06,
+      "loss": 0.1968,
+      "step": 3400
+    },
+    {
+      "epoch": 1.6104172832198875,
+      "grad_norm": 2.2150564193725586,
+      "learning_rate": 9.287187836430366e-06,
+      "loss": 0.2549,
+      "step": 3401
+    },
+    {
+      "epoch": 1.6108907960935188,
+      "grad_norm": 1.1368675231933594,
+      "learning_rate": 9.282086837152198e-06,
+      "loss": 0.2028,
+      "step": 3402
+    },
+    {
+      "epoch": 1.61136430896715,
+      "grad_norm": 1.2357794046401978,
+      "learning_rate": 9.276986025637291e-06,
+      "loss": 0.2165,
+      "step": 3403
+    },
+    {
+      "epoch": 1.6118378218407812,
+      "grad_norm": 1.7708053588867188,
+      "learning_rate": 9.271885403219715e-06,
+      "loss": 0.1969,
+      "step": 3404
+    },
+    {
+      "epoch": 1.6123113347144127,
+      "grad_norm": 1.0533392429351807,
+      "learning_rate": 9.266784971233487e-06,
+      "loss": 0.2193,
+      "step": 3405
+    },
+    {
+      "epoch": 1.6127848475880437,
+      "grad_norm": 2.6355881690979004,
+      "learning_rate": 9.261684731012575e-06,
+      "loss": 0.201,
+      "step": 3406
+    },
+    {
+      "epoch": 1.613258360461675,
+      "grad_norm": 1.196489930152893,
+      "learning_rate": 9.256584683890902e-06,
+      "loss": 0.2151,
+      "step": 3407
+    },
+    {
+      "epoch": 1.6137318733353063,
+      "grad_norm": 1.8537029027938843,
+      "learning_rate": 9.25148483120233e-06,
+      "loss": 0.2343,
+      "step": 3408
+    },
+    {
+      "epoch": 1.6142053862089376,
+      "grad_norm": 1.2622405290603638,
+      "learning_rate": 9.24638517428068e-06,
+      "loss": 0.2154,
+      "step": 3409
+    },
+    {
+      "epoch": 1.614678899082569,
+      "grad_norm": 1.1913985013961792,
+      "learning_rate": 9.24128571445972e-06,
+      "loss": 0.2275,
+      "step": 3410
+    },
+    {
+      "epoch": 1.6151524119562,
+      "grad_norm": 1.129180669784546,
+      "learning_rate": 9.236186453073161e-06,
+      "loss": 0.208,
+      "step": 3411
+    },
+    {
+      "epoch": 1.6156259248298315,
+      "grad_norm": 1.8293282985687256,
+      "learning_rate": 9.231087391454665e-06,
+      "loss": 0.2184,
+      "step": 3412
+    },
+    {
+      "epoch": 1.6160994377034625,
+      "grad_norm": 1.8663169145584106,
+      "learning_rate": 9.225988530937846e-06,
+      "loss": 0.2305,
+      "step": 3413
+    },
+    {
+      "epoch": 1.6165729505770938,
+      "grad_norm": 1.2646862268447876,
+      "learning_rate": 9.220889872856258e-06,
+      "loss": 0.2168,
+      "step": 3414
+    },
+    {
+      "epoch": 1.6170464634507251,
+      "grad_norm": 1.4086544513702393,
+      "learning_rate": 9.215791418543407e-06,
+      "loss": 0.2109,
+      "step": 3415
+    },
+    {
+      "epoch": 1.6175199763243562,
+      "grad_norm": 1.7981756925582886,
+      "learning_rate": 9.210693169332746e-06,
+      "loss": 0.2098,
+      "step": 3416
+    },
+    {
+      "epoch": 1.6179934891979877,
+      "grad_norm": 1.776749849319458,
+      "learning_rate": 9.205595126557673e-06,
+      "loss": 0.2017,
+      "step": 3417
+    },
+    {
+      "epoch": 1.6184670020716188,
+      "grad_norm": 1.2278921604156494,
+      "learning_rate": 9.200497291551528e-06,
+      "loss": 0.2068,
+      "step": 3418
+    },
+    {
+      "epoch": 1.61894051494525,
+      "grad_norm": 0.9366362690925598,
+      "learning_rate": 9.195399665647607e-06,
+      "loss": 0.1798,
+      "step": 3419
+    },
+    {
+      "epoch": 1.6194140278188813,
+      "grad_norm": 1.8072518110275269,
+      "learning_rate": 9.190302250179141e-06,
+      "loss": 0.2245,
+      "step": 3420
+    },
+    {
+      "epoch": 1.6198875406925126,
+      "grad_norm": 1.4255504608154297,
+      "learning_rate": 9.185205046479308e-06,
+      "loss": 0.2082,
+      "step": 3421
+    },
+    {
+      "epoch": 1.620361053566144,
+      "grad_norm": 1.2994076013565063,
+      "learning_rate": 9.180108055881236e-06,
+      "loss": 0.2107,
+      "step": 3422
+    },
+    {
+      "epoch": 1.620834566439775,
+      "grad_norm": 1.6966938972473145,
+      "learning_rate": 9.175011279717992e-06,
+      "loss": 0.2548,
+      "step": 3423
+    },
+    {
+      "epoch": 1.6213080793134065,
+      "grad_norm": 1.2594822645187378,
+      "learning_rate": 9.169914719322588e-06,
+      "loss": 0.2031,
+      "step": 3424
+    },
+    {
+      "epoch": 1.6217815921870375,
+      "grad_norm": 1.0965392589569092,
+      "learning_rate": 9.164818376027981e-06,
+      "loss": 0.2233,
+      "step": 3425
+    },
+    {
+      "epoch": 1.6222551050606688,
+      "grad_norm": 1.0597667694091797,
+      "learning_rate": 9.159722251167073e-06,
+      "loss": 0.2034,
+      "step": 3426
+    },
+    {
+      "epoch": 1.6227286179343001,
+      "grad_norm": 1.3505845069885254,
+      "learning_rate": 9.154626346072702e-06,
+      "loss": 0.2138,
+      "step": 3427
+    },
+    {
+      "epoch": 1.6232021308079312,
+      "grad_norm": 0.9662488102912903,
+      "learning_rate": 9.149530662077655e-06,
+      "loss": 0.1981,
+      "step": 3428
+    },
+    {
+      "epoch": 1.6236756436815627,
+      "grad_norm": 0.9503706693649292,
+      "learning_rate": 9.144435200514658e-06,
+      "loss": 0.2161,
+      "step": 3429
+    },
+    {
+      "epoch": 1.6241491565551938,
+      "grad_norm": 1.316527247428894,
+      "learning_rate": 9.139339962716383e-06,
+      "loss": 0.188,
+      "step": 3430
+    },
+    {
+      "epoch": 1.624622669428825,
+      "grad_norm": 1.0049852132797241,
+      "learning_rate": 9.134244950015437e-06,
+      "loss": 0.2057,
+      "step": 3431
+    },
+    {
+      "epoch": 1.6250961823024563,
+      "grad_norm": 1.619187355041504,
+      "learning_rate": 9.129150163744371e-06,
+      "loss": 0.2018,
+      "step": 3432
+    },
+    {
+      "epoch": 1.6255696951760876,
+      "grad_norm": 1.270595908164978,
+      "learning_rate": 9.12405560523568e-06,
+      "loss": 0.1814,
+      "step": 3433
+    },
+    {
+      "epoch": 1.626043208049719,
+      "grad_norm": 1.6350579261779785,
+      "learning_rate": 9.118961275821792e-06,
+      "loss": 0.2457,
+      "step": 3434
+    },
+    {
+      "epoch": 1.62651672092335,
+      "grad_norm": 2.1566319465637207,
+      "learning_rate": 9.113867176835086e-06,
+      "loss": 0.1939,
+      "step": 3435
+    },
+    {
+      "epoch": 1.6269902337969815,
+      "grad_norm": 1.6590288877487183,
+      "learning_rate": 9.10877330960787e-06,
+      "loss": 0.2094,
+      "step": 3436
+    },
+    {
+      "epoch": 1.6274637466706126,
+      "grad_norm": 1.642244577407837,
+      "learning_rate": 9.103679675472395e-06,
+      "loss": 0.2082,
+      "step": 3437
+    },
+    {
+      "epoch": 1.6279372595442438,
+      "grad_norm": 1.484878659248352,
+      "learning_rate": 9.098586275760854e-06,
+      "loss": 0.2094,
+      "step": 3438
+    },
+    {
+      "epoch": 1.6284107724178751,
+      "grad_norm": 1.3303133249282837,
+      "learning_rate": 9.093493111805379e-06,
+      "loss": 0.2161,
+      "step": 3439
+    },
+    {
+      "epoch": 1.6288842852915064,
+      "grad_norm": 1.818798542022705,
+      "learning_rate": 9.088400184938036e-06,
+      "loss": 0.2261,
+      "step": 3440
+    },
+    {
+      "epoch": 1.6293577981651377,
+      "grad_norm": 1.2942332029342651,
+      "learning_rate": 9.08330749649083e-06,
+      "loss": 0.2272,
+      "step": 3441
+    },
+    {
+      "epoch": 1.6298313110387688,
+      "grad_norm": 1.2421215772628784,
+      "learning_rate": 9.078215047795703e-06,
+      "loss": 0.2111,
+      "step": 3442
+    },
+    {
+      "epoch": 1.6303048239124003,
+      "grad_norm": 1.215605616569519,
+      "learning_rate": 9.073122840184537e-06,
+      "loss": 0.2487,
+      "step": 3443
+    },
+    {
+      "epoch": 1.6307783367860313,
+      "grad_norm": 1.1501661539077759,
+      "learning_rate": 9.068030874989152e-06,
+      "loss": 0.2214,
+      "step": 3444
+    },
+    {
+      "epoch": 1.6312518496596626,
+      "grad_norm": 1.0269571542739868,
+      "learning_rate": 9.062939153541302e-06,
+      "loss": 0.2149,
+      "step": 3445
+    },
+    {
+      "epoch": 1.631725362533294,
+      "grad_norm": 1.087510108947754,
+      "learning_rate": 9.057847677172675e-06,
+      "loss": 0.2047,
+      "step": 3446
+    },
+    {
+      "epoch": 1.632198875406925,
+      "grad_norm": 1.1288443803787231,
+      "learning_rate": 9.052756447214899e-06,
+      "loss": 0.2051,
+      "step": 3447
+    },
+    {
+      "epoch": 1.6326723882805565,
+      "grad_norm": 0.9681136608123779,
+      "learning_rate": 9.047665464999537e-06,
+      "loss": 0.2058,
+      "step": 3448
+    },
+    {
+      "epoch": 1.6331459011541876,
+      "grad_norm": 1.3318946361541748,
+      "learning_rate": 9.042574731858084e-06,
+      "loss": 0.224,
+      "step": 3449
+    },
+    {
+      "epoch": 1.6336194140278189,
+      "grad_norm": 1.4043301343917847,
+      "learning_rate": 9.037484249121974e-06,
+      "loss": 0.2254,
+      "step": 3450
+    },
+    {
+      "epoch": 1.6340929269014501,
+      "grad_norm": 1.0534493923187256,
+      "learning_rate": 9.032394018122572e-06,
+      "loss": 0.217,
+      "step": 3451
+    },
+    {
+      "epoch": 1.6345664397750814,
+      "grad_norm": 1.055869221687317,
+      "learning_rate": 9.027304040191181e-06,
+      "loss": 0.2174,
+      "step": 3452
+    },
+    {
+      "epoch": 1.6350399526487127,
+      "grad_norm": 0.9580625891685486,
+      "learning_rate": 9.022214316659035e-06,
+      "loss": 0.2125,
+      "step": 3453
+    },
+    {
+      "epoch": 1.6355134655223438,
+      "grad_norm": 1.0676701068878174,
+      "learning_rate": 9.0171248488573e-06,
+      "loss": 0.2094,
+      "step": 3454
+    },
+    {
+      "epoch": 1.6359869783959753,
+      "grad_norm": 1.6258785724639893,
+      "learning_rate": 9.012035638117082e-06,
+      "loss": 0.2406,
+      "step": 3455
+    },
+    {
+      "epoch": 1.6364604912696064,
+      "grad_norm": 1.652103304862976,
+      "learning_rate": 9.006946685769408e-06,
+      "loss": 0.1958,
+      "step": 3456
+    },
+    {
+      "epoch": 1.6369340041432376,
+      "grad_norm": 1.625702977180481,
+      "learning_rate": 9.001857993145251e-06,
+      "loss": 0.2426,
+      "step": 3457
+    },
+    {
+      "epoch": 1.637407517016869,
+      "grad_norm": 1.5374822616577148,
+      "learning_rate": 8.996769561575504e-06,
+      "loss": 0.2269,
+      "step": 3458
+    },
+    {
+      "epoch": 1.6378810298905,
+      "grad_norm": 1.0075321197509766,
+      "learning_rate": 8.991681392391001e-06,
+      "loss": 0.1967,
+      "step": 3459
+    },
+    {
+      "epoch": 1.6383545427641315,
+      "grad_norm": 1.2052757740020752,
+      "learning_rate": 8.986593486922504e-06,
+      "loss": 0.1997,
+      "step": 3460
+    },
+    {
+      "epoch": 1.6388280556377626,
+      "grad_norm": 1.0683456659317017,
+      "learning_rate": 8.981505846500703e-06,
+      "loss": 0.2165,
+      "step": 3461
+    },
+    {
+      "epoch": 1.6393015685113939,
+      "grad_norm": 1.3364335298538208,
+      "learning_rate": 8.976418472456222e-06,
+      "loss": 0.2098,
+      "step": 3462
+    },
+    {
+      "epoch": 1.6397750813850251,
+      "grad_norm": 1.8314017057418823,
+      "learning_rate": 8.971331366119613e-06,
+      "loss": 0.2215,
+      "step": 3463
+    },
+    {
+      "epoch": 1.6402485942586564,
+      "grad_norm": 1.2816479206085205,
+      "learning_rate": 8.966244528821366e-06,
+      "loss": 0.2137,
+      "step": 3464
+    },
+    {
+      "epoch": 1.6407221071322877,
+      "grad_norm": 1.2321319580078125,
+      "learning_rate": 8.961157961891886e-06,
+      "loss": 0.2345,
+      "step": 3465
+    },
+    {
+      "epoch": 1.6411956200059188,
+      "grad_norm": 1.0520602464675903,
+      "learning_rate": 8.95607166666152e-06,
+      "loss": 0.2141,
+      "step": 3466
+    },
+    {
+      "epoch": 1.6416691328795503,
+      "grad_norm": 1.592511773109436,
+      "learning_rate": 8.950985644460539e-06,
+      "loss": 0.2279,
+      "step": 3467
+    },
+    {
+      "epoch": 1.6421426457531814,
+      "grad_norm": 1.1186596155166626,
+      "learning_rate": 8.945899896619143e-06,
+      "loss": 0.2285,
+      "step": 3468
+    },
+    {
+      "epoch": 1.6426161586268127,
+      "grad_norm": 1.4300451278686523,
+      "learning_rate": 8.940814424467457e-06,
+      "loss": 0.2494,
+      "step": 3469
+    },
+    {
+      "epoch": 1.643089671500444,
+      "grad_norm": 1.1781500577926636,
+      "learning_rate": 8.935729229335544e-06,
+      "loss": 0.2253,
+      "step": 3470
+    },
+    {
+      "epoch": 1.6435631843740752,
+      "grad_norm": 1.3187671899795532,
+      "learning_rate": 8.930644312553381e-06,
+      "loss": 0.2244,
+      "step": 3471
+    },
+    {
+      "epoch": 1.6440366972477065,
+      "grad_norm": 1.1596239805221558,
+      "learning_rate": 8.925559675450883e-06,
+      "loss": 0.2193,
+      "step": 3472
+    },
+    {
+      "epoch": 1.6445102101213376,
+      "grad_norm": 2.0229430198669434,
+      "learning_rate": 8.920475319357886e-06,
+      "loss": 0.2182,
+      "step": 3473
+    },
+    {
+      "epoch": 1.644983722994969,
+      "grad_norm": 1.2594224214553833,
+      "learning_rate": 8.915391245604159e-06,
+      "loss": 0.2311,
+      "step": 3474
+    },
+    {
+      "epoch": 1.6454572358686002,
+      "grad_norm": 1.5540090799331665,
+      "learning_rate": 8.910307455519385e-06,
+      "loss": 0.2034,
+      "step": 3475
+    },
+    {
+      "epoch": 1.6459307487422314,
+      "grad_norm": 1.0850417613983154,
+      "learning_rate": 8.905223950433178e-06,
+      "loss": 0.1913,
+      "step": 3476
+    },
+    {
+      "epoch": 1.6464042616158627,
+      "grad_norm": 1.3703140020370483,
+      "learning_rate": 8.90014073167509e-06,
+      "loss": 0.2108,
+      "step": 3477
+    },
+    {
+      "epoch": 1.6468777744894938,
+      "grad_norm": 1.050058364868164,
+      "learning_rate": 8.895057800574584e-06,
+      "loss": 0.2306,
+      "step": 3478
+    },
+    {
+      "epoch": 1.6473512873631253,
+      "grad_norm": 1.3480294942855835,
+      "learning_rate": 8.889975158461051e-06,
+      "loss": 0.202,
+      "step": 3479
+    },
+    {
+      "epoch": 1.6478248002367564,
+      "grad_norm": 1.2146999835968018,
+      "learning_rate": 8.884892806663808e-06,
+      "loss": 0.198,
+      "step": 3480
+    },
+    {
+      "epoch": 1.6482983131103877,
+      "grad_norm": 1.6478931903839111,
+      "learning_rate": 8.879810746512091e-06,
+      "loss": 0.2188,
+      "step": 3481
+    },
+    {
+      "epoch": 1.648771825984019,
+      "grad_norm": 1.0835014581680298,
+      "learning_rate": 8.874728979335069e-06,
+      "loss": 0.2076,
+      "step": 3482
+    },
+    {
+      "epoch": 1.6492453388576502,
+      "grad_norm": 1.0451796054840088,
+      "learning_rate": 8.869647506461823e-06,
+      "loss": 0.221,
+      "step": 3483
+    },
+    {
+      "epoch": 1.6497188517312815,
+      "grad_norm": 1.0795484781265259,
+      "learning_rate": 8.86456632922137e-06,
+      "loss": 0.2048,
+      "step": 3484
+    },
+    {
+      "epoch": 1.6501923646049126,
+      "grad_norm": 1.3780536651611328,
+      "learning_rate": 8.85948544894264e-06,
+      "loss": 0.2048,
+      "step": 3485
+    },
+    {
+      "epoch": 1.650665877478544,
+      "grad_norm": 1.2751306295394897,
+      "learning_rate": 8.854404866954485e-06,
+      "loss": 0.2238,
+      "step": 3486
+    },
+    {
+      "epoch": 1.6511393903521752,
+      "grad_norm": 1.0825250148773193,
+      "learning_rate": 8.84932458458569e-06,
+      "loss": 0.2076,
+      "step": 3487
+    },
+    {
+      "epoch": 1.6516129032258065,
+      "grad_norm": 1.1926277875900269,
+      "learning_rate": 8.844244603164946e-06,
+      "loss": 0.2038,
+      "step": 3488
+    },
+    {
+      "epoch": 1.6520864160994377,
+      "grad_norm": 1.0996829271316528,
+      "learning_rate": 8.839164924020878e-06,
+      "loss": 0.1927,
+      "step": 3489
+    },
+    {
+      "epoch": 1.6525599289730688,
+      "grad_norm": 0.9266497492790222,
+      "learning_rate": 8.834085548482024e-06,
+      "loss": 0.2048,
+      "step": 3490
+    },
+    {
+      "epoch": 1.6530334418467003,
+      "grad_norm": 1.086111068725586,
+      "learning_rate": 8.829006477876847e-06,
+      "loss": 0.2219,
+      "step": 3491
+    },
+    {
+      "epoch": 1.6535069547203314,
+      "grad_norm": 1.0157992839813232,
+      "learning_rate": 8.82392771353373e-06,
+      "loss": 0.1998,
+      "step": 3492
+    },
+    {
+      "epoch": 1.6539804675939627,
+      "grad_norm": 1.2996865510940552,
+      "learning_rate": 8.818849256780972e-06,
+      "loss": 0.2155,
+      "step": 3493
+    },
+    {
+      "epoch": 1.654453980467594,
+      "grad_norm": 1.0794087648391724,
+      "learning_rate": 8.813771108946798e-06,
+      "loss": 0.208,
+      "step": 3494
+    },
+    {
+      "epoch": 1.6549274933412252,
+      "grad_norm": 1.0981913805007935,
+      "learning_rate": 8.808693271359346e-06,
+      "loss": 0.2074,
+      "step": 3495
+    },
+    {
+      "epoch": 1.6554010062148565,
+      "grad_norm": 1.7508763074874878,
+      "learning_rate": 8.803615745346675e-06,
+      "loss": 0.238,
+      "step": 3496
+    },
+    {
+      "epoch": 1.6558745190884876,
+      "grad_norm": 1.0915553569793701,
+      "learning_rate": 8.798538532236764e-06,
+      "loss": 0.2166,
+      "step": 3497
+    },
+    {
+      "epoch": 1.656348031962119,
+      "grad_norm": 0.9217873811721802,
+      "learning_rate": 8.79346163335751e-06,
+      "loss": 0.1957,
+      "step": 3498
+    },
+    {
+      "epoch": 1.6568215448357502,
+      "grad_norm": 2.03841233253479,
+      "learning_rate": 8.78838505003673e-06,
+      "loss": 0.2089,
+      "step": 3499
+    },
+    {
+      "epoch": 1.6572950577093815,
+      "grad_norm": 1.8650535345077515,
+      "learning_rate": 8.783308783602148e-06,
+      "loss": 0.188,
+      "step": 3500
+    },
+    {
+      "epoch": 1.6577685705830127,
+      "grad_norm": 0.9951745867729187,
+      "learning_rate": 8.778232835381415e-06,
+      "loss": 0.1941,
+      "step": 3501
+    },
+    {
+      "epoch": 1.6582420834566438,
+      "grad_norm": 1.168714165687561,
+      "learning_rate": 8.773157206702097e-06,
+      "loss": 0.2112,
+      "step": 3502
+    },
+    {
+      "epoch": 1.6587155963302753,
+      "grad_norm": 1.0129159688949585,
+      "learning_rate": 8.768081898891679e-06,
+      "loss": 0.2112,
+      "step": 3503
+    },
+    {
+      "epoch": 1.6591891092039064,
+      "grad_norm": 1.5583324432373047,
+      "learning_rate": 8.763006913277553e-06,
+      "loss": 0.2165,
+      "step": 3504
+    },
+    {
+      "epoch": 1.659662622077538,
+      "grad_norm": 1.3734880685806274,
+      "learning_rate": 8.757932251187037e-06,
+      "loss": 0.2241,
+      "step": 3505
+    },
+    {
+      "epoch": 1.660136134951169,
+      "grad_norm": 1.3286396265029907,
+      "learning_rate": 8.752857913947357e-06,
+      "loss": 0.2349,
+      "step": 3506
+    },
+    {
+      "epoch": 1.6606096478248003,
+      "grad_norm": 1.3131154775619507,
+      "learning_rate": 8.747783902885657e-06,
+      "loss": 0.2454,
+      "step": 3507
+    },
+    {
+      "epoch": 1.6610831606984315,
+      "grad_norm": 1.0709985494613647,
+      "learning_rate": 8.742710219329e-06,
+      "loss": 0.225,
+      "step": 3508
+    },
+    {
+      "epoch": 1.6615566735720626,
+      "grad_norm": 1.4955830574035645,
+      "learning_rate": 8.737636864604357e-06,
+      "loss": 0.1989,
+      "step": 3509
+    },
+    {
+      "epoch": 1.6620301864456941,
+      "grad_norm": 1.0613009929656982,
+      "learning_rate": 8.73256384003861e-06,
+      "loss": 0.2329,
+      "step": 3510
+    },
+    {
+      "epoch": 1.6625036993193252,
+      "grad_norm": 1.2073135375976562,
+      "learning_rate": 8.727491146958566e-06,
+      "loss": 0.2259,
+      "step": 3511
+    },
+    {
+      "epoch": 1.6629772121929565,
+      "grad_norm": 1.160805583000183,
+      "learning_rate": 8.722418786690936e-06,
+      "loss": 0.2198,
+      "step": 3512
+    },
+    {
+      "epoch": 1.6634507250665878,
+      "grad_norm": 1.5360057353973389,
+      "learning_rate": 8.717346760562345e-06,
+      "loss": 0.2242,
+      "step": 3513
+    },
+    {
+      "epoch": 1.663924237940219,
+      "grad_norm": 1.0685076713562012,
+      "learning_rate": 8.712275069899337e-06,
+      "loss": 0.2136,
+      "step": 3514
+    },
+    {
+      "epoch": 1.6643977508138503,
+      "grad_norm": 1.6072028875350952,
+      "learning_rate": 8.707203716028358e-06,
+      "loss": 0.2215,
+      "step": 3515
+    },
+    {
+      "epoch": 1.6648712636874814,
+      "grad_norm": 1.209231972694397,
+      "learning_rate": 8.702132700275778e-06,
+      "loss": 0.2225,
+      "step": 3516
+    },
+    {
+      "epoch": 1.665344776561113,
+      "grad_norm": 1.0955963134765625,
+      "learning_rate": 8.697062023967869e-06,
+      "loss": 0.2164,
+      "step": 3517
+    },
+    {
+      "epoch": 1.665818289434744,
+      "grad_norm": 0.9355467557907104,
+      "learning_rate": 8.691991688430818e-06,
+      "loss": 0.2106,
+      "step": 3518
+    },
+    {
+      "epoch": 1.6662918023083753,
+      "grad_norm": 0.9584832787513733,
+      "learning_rate": 8.686921694990724e-06,
+      "loss": 0.2093,
+      "step": 3519
+    },
+    {
+      "epoch": 1.6667653151820065,
+      "grad_norm": 1.7122430801391602,
+      "learning_rate": 8.681852044973591e-06,
+      "loss": 0.2166,
+      "step": 3520
+    },
+    {
+      "epoch": 1.6672388280556376,
+      "grad_norm": 1.5137004852294922,
+      "learning_rate": 8.67678273970534e-06,
+      "loss": 0.2199,
+      "step": 3521
+    },
+    {
+      "epoch": 1.6677123409292691,
+      "grad_norm": 1.2796556949615479,
+      "learning_rate": 8.671713780511798e-06,
+      "loss": 0.2053,
+      "step": 3522
+    },
+    {
+      "epoch": 1.6681858538029002,
+      "grad_norm": 1.1300054788589478,
+      "learning_rate": 8.666645168718705e-06,
+      "loss": 0.2256,
+      "step": 3523
+    },
+    {
+      "epoch": 1.6686593666765315,
+      "grad_norm": 1.094119906425476,
+      "learning_rate": 8.661576905651704e-06,
+      "loss": 0.2042,
+      "step": 3524
+    },
+    {
+      "epoch": 1.6691328795501628,
+      "grad_norm": 1.1889249086380005,
+      "learning_rate": 8.656508992636352e-06,
+      "loss": 0.2144,
+      "step": 3525
+    },
+    {
+      "epoch": 1.669606392423794,
+      "grad_norm": 0.9747273325920105,
+      "learning_rate": 8.65144143099811e-06,
+      "loss": 0.1979,
+      "step": 3526
+    },
+    {
+      "epoch": 1.6700799052974253,
+      "grad_norm": 1.1570683717727661,
+      "learning_rate": 8.646374222062352e-06,
+      "loss": 0.2429,
+      "step": 3527
+    },
+    {
+      "epoch": 1.6705534181710564,
+      "grad_norm": 1.3277277946472168,
+      "learning_rate": 8.641307367154356e-06,
+      "loss": 0.2064,
+      "step": 3528
+    },
+    {
+      "epoch": 1.671026931044688,
+      "grad_norm": 1.3713091611862183,
+      "learning_rate": 8.636240867599314e-06,
+      "loss": 0.2051,
+      "step": 3529
+    },
+    {
+      "epoch": 1.671500443918319,
+      "grad_norm": 1.2886407375335693,
+      "learning_rate": 8.631174724722315e-06,
+      "loss": 0.2237,
+      "step": 3530
+    },
+    {
+      "epoch": 1.6719739567919503,
+      "grad_norm": 1.255431056022644,
+      "learning_rate": 8.626108939848362e-06,
+      "loss": 0.2136,
+      "step": 3531
+    },
+    {
+      "epoch": 1.6724474696655816,
+      "grad_norm": 1.0875186920166016,
+      "learning_rate": 8.621043514302361e-06,
+      "loss": 0.1884,
+      "step": 3532
+    },
+    {
+      "epoch": 1.6729209825392126,
+      "grad_norm": 1.1179817914962769,
+      "learning_rate": 8.615978449409124e-06,
+      "loss": 0.2243,
+      "step": 3533
+    },
+    {
+      "epoch": 1.6733944954128441,
+      "grad_norm": 0.9758694171905518,
+      "learning_rate": 8.610913746493377e-06,
+      "loss": 0.2246,
+      "step": 3534
+    },
+    {
+      "epoch": 1.6738680082864752,
+      "grad_norm": 1.0491384267807007,
+      "learning_rate": 8.605849406879736e-06,
+      "loss": 0.241,
+      "step": 3535
+    },
+    {
+      "epoch": 1.6743415211601067,
+      "grad_norm": 1.1519798040390015,
+      "learning_rate": 8.600785431892731e-06,
+      "loss": 0.2076,
+      "step": 3536
+    },
+    {
+      "epoch": 1.6748150340337378,
+      "grad_norm": 1.3320951461791992,
+      "learning_rate": 8.5957218228568e-06,
+      "loss": 0.2107,
+      "step": 3537
+    },
+    {
+      "epoch": 1.675288546907369,
+      "grad_norm": 1.127048134803772,
+      "learning_rate": 8.590658581096278e-06,
+      "loss": 0.2077,
+      "step": 3538
+    },
+    {
+      "epoch": 1.6757620597810003,
+      "grad_norm": 1.4201406240463257,
+      "learning_rate": 8.585595707935408e-06,
+      "loss": 0.2116,
+      "step": 3539
+    },
+    {
+      "epoch": 1.6762355726546314,
+      "grad_norm": 1.774510145187378,
+      "learning_rate": 8.580533204698336e-06,
+      "loss": 0.2003,
+      "step": 3540
+    },
+    {
+      "epoch": 1.676709085528263,
+      "grad_norm": 1.9603132009506226,
+      "learning_rate": 8.57547107270911e-06,
+      "loss": 0.2353,
+      "step": 3541
+    },
+    {
+      "epoch": 1.677182598401894,
+      "grad_norm": 1.5556761026382446,
+      "learning_rate": 8.570409313291683e-06,
+      "loss": 0.2191,
+      "step": 3542
+    },
+    {
+      "epoch": 1.6776561112755253,
+      "grad_norm": 1.459722638130188,
+      "learning_rate": 8.56534792776991e-06,
+      "loss": 0.1919,
+      "step": 3543
+    },
+    {
+      "epoch": 1.6781296241491566,
+      "grad_norm": 1.822821021080017,
+      "learning_rate": 8.560286917467543e-06,
+      "loss": 0.2449,
+      "step": 3544
+    },
+    {
+      "epoch": 1.6786031370227879,
+      "grad_norm": 1.4285695552825928,
+      "learning_rate": 8.555226283708246e-06,
+      "loss": 0.1897,
+      "step": 3545
+    },
+    {
+      "epoch": 1.6790766498964191,
+      "grad_norm": 1.1042506694793701,
+      "learning_rate": 8.550166027815576e-06,
+      "loss": 0.2184,
+      "step": 3546
+    },
+    {
+      "epoch": 1.6795501627700502,
+      "grad_norm": 0.9724579453468323,
+      "learning_rate": 8.545106151112994e-06,
+      "loss": 0.2061,
+      "step": 3547
+    },
+    {
+      "epoch": 1.6800236756436817,
+      "grad_norm": 1.1628752946853638,
+      "learning_rate": 8.540046654923863e-06,
+      "loss": 0.1937,
+      "step": 3548
+    },
+    {
+      "epoch": 1.6804971885173128,
+      "grad_norm": 1.408586025238037,
+      "learning_rate": 8.534987540571445e-06,
+      "loss": 0.2233,
+      "step": 3549
+    },
+    {
+      "epoch": 1.680970701390944,
+      "grad_norm": 1.0293395519256592,
+      "learning_rate": 8.529928809378902e-06,
+      "loss": 0.2141,
+      "step": 3550
+    },
+    {
+      "epoch": 1.6814442142645754,
+      "grad_norm": 2.2602131366729736,
+      "learning_rate": 8.524870462669296e-06,
+      "loss": 0.2243,
+      "step": 3551
+    },
+    {
+      "epoch": 1.6819177271382064,
+      "grad_norm": 1.1277772188186646,
+      "learning_rate": 8.519812501765591e-06,
+      "loss": 0.2004,
+      "step": 3552
+    },
+    {
+      "epoch": 1.682391240011838,
+      "grad_norm": 1.7212578058242798,
+      "learning_rate": 8.514754927990646e-06,
+      "loss": 0.1934,
+      "step": 3553
+    },
+    {
+      "epoch": 1.682864752885469,
+      "grad_norm": 1.1089184284210205,
+      "learning_rate": 8.509697742667219e-06,
+      "loss": 0.2354,
+      "step": 3554
+    },
+    {
+      "epoch": 1.6833382657591003,
+      "grad_norm": 1.5018812417984009,
+      "learning_rate": 8.504640947117973e-06,
+      "loss": 0.2144,
+      "step": 3555
+    },
+    {
+      "epoch": 1.6838117786327316,
+      "grad_norm": 1.6116001605987549,
+      "learning_rate": 8.49958454266546e-06,
+      "loss": 0.2174,
+      "step": 3556
+    },
+    {
+      "epoch": 1.6842852915063629,
+      "grad_norm": 1.9937829971313477,
+      "learning_rate": 8.494528530632136e-06,
+      "loss": 0.2132,
+      "step": 3557
+    },
+    {
+      "epoch": 1.6847588043799941,
+      "grad_norm": 1.3421846628189087,
+      "learning_rate": 8.48947291234035e-06,
+      "loss": 0.2168,
+      "step": 3558
+    },
+    {
+      "epoch": 1.6852323172536252,
+      "grad_norm": 1.112237572669983,
+      "learning_rate": 8.484417689112356e-06,
+      "loss": 0.2025,
+      "step": 3559
+    },
+    {
+      "epoch": 1.6857058301272567,
+      "grad_norm": 1.0491182804107666,
+      "learning_rate": 8.47936286227029e-06,
+      "loss": 0.2236,
+      "step": 3560
+    },
+    {
+      "epoch": 1.6861793430008878,
+      "grad_norm": 1.0888773202896118,
+      "learning_rate": 8.4743084331362e-06,
+      "loss": 0.2186,
+      "step": 3561
+    },
+    {
+      "epoch": 1.686652855874519,
+      "grad_norm": 0.9247356653213501,
+      "learning_rate": 8.46925440303202e-06,
+      "loss": 0.2121,
+      "step": 3562
+    },
+    {
+      "epoch": 1.6871263687481504,
+      "grad_norm": 0.986939013004303,
+      "learning_rate": 8.464200773279582e-06,
+      "loss": 0.2269,
+      "step": 3563
+    },
+    {
+      "epoch": 1.6875998816217814,
+      "grad_norm": 1.2144672870635986,
+      "learning_rate": 8.459147545200617e-06,
+      "loss": 0.219,
+      "step": 3564
+    },
+    {
+      "epoch": 1.688073394495413,
+      "grad_norm": 1.2081797122955322,
+      "learning_rate": 8.454094720116745e-06,
+      "loss": 0.1891,
+      "step": 3565
+    },
+    {
+      "epoch": 1.688546907369044,
+      "grad_norm": 1.819753646850586,
+      "learning_rate": 8.449042299349487e-06,
+      "loss": 0.2095,
+      "step": 3566
+    },
+    {
+      "epoch": 1.6890204202426755,
+      "grad_norm": 0.952063262462616,
+      "learning_rate": 8.443990284220252e-06,
+      "loss": 0.2111,
+      "step": 3567
+    },
+    {
+      "epoch": 1.6894939331163066,
+      "grad_norm": 1.0713047981262207,
+      "learning_rate": 8.43893867605035e-06,
+      "loss": 0.2046,
+      "step": 3568
+    },
+    {
+      "epoch": 1.6899674459899379,
+      "grad_norm": 1.6247305870056152,
+      "learning_rate": 8.433887476160976e-06,
+      "loss": 0.1997,
+      "step": 3569
+    },
+    {
+      "epoch": 1.6904409588635692,
+      "grad_norm": 0.9922040104866028,
+      "learning_rate": 8.428836685873223e-06,
+      "loss": 0.1791,
+      "step": 3570
+    },
+    {
+      "epoch": 1.6909144717372002,
+      "grad_norm": 1.1413791179656982,
+      "learning_rate": 8.423786306508076e-06,
+      "loss": 0.2028,
+      "step": 3571
+    },
+    {
+      "epoch": 1.6913879846108317,
+      "grad_norm": 1.1221214532852173,
+      "learning_rate": 8.418736339386417e-06,
+      "loss": 0.2044,
+      "step": 3572
+    },
+    {
+      "epoch": 1.6918614974844628,
+      "grad_norm": 1.0321074724197388,
+      "learning_rate": 8.413686785829013e-06,
+      "loss": 0.2371,
+      "step": 3573
+    },
+    {
+      "epoch": 1.692335010358094,
+      "grad_norm": 1.458929181098938,
+      "learning_rate": 8.408637647156528e-06,
+      "loss": 0.228,
+      "step": 3574
+    },
+    {
+      "epoch": 1.6928085232317254,
+      "grad_norm": 1.4914942979812622,
+      "learning_rate": 8.403588924689511e-06,
+      "loss": 0.1998,
+      "step": 3575
+    },
+    {
+      "epoch": 1.6932820361053567,
+      "grad_norm": 1.992150902748108,
+      "learning_rate": 8.398540619748414e-06,
+      "loss": 0.2095,
+      "step": 3576
+    },
+    {
+      "epoch": 1.693755548978988,
+      "grad_norm": 0.9819469451904297,
+      "learning_rate": 8.393492733653566e-06,
+      "loss": 0.2025,
+      "step": 3577
+    },
+    {
+      "epoch": 1.694229061852619,
+      "grad_norm": 0.8754822611808777,
+      "learning_rate": 8.388445267725197e-06,
+      "loss": 0.1947,
+      "step": 3578
+    },
+    {
+      "epoch": 1.6947025747262505,
+      "grad_norm": 1.3592606782913208,
+      "learning_rate": 8.38339822328342e-06,
+      "loss": 0.2065,
+      "step": 3579
+    },
+    {
+      "epoch": 1.6951760875998816,
+      "grad_norm": 1.2748628854751587,
+      "learning_rate": 8.378351601648243e-06,
+      "loss": 0.2091,
+      "step": 3580
+    },
+    {
+      "epoch": 1.6956496004735129,
+      "grad_norm": 1.1652089357376099,
+      "learning_rate": 8.373305404139558e-06,
+      "loss": 0.1885,
+      "step": 3581
+    },
+    {
+      "epoch": 1.6961231133471442,
+      "grad_norm": 1.0622856616973877,
+      "learning_rate": 8.368259632077153e-06,
+      "loss": 0.2115,
+      "step": 3582
+    },
+    {
+      "epoch": 1.6965966262207752,
+      "grad_norm": 1.0490455627441406,
+      "learning_rate": 8.363214286780699e-06,
+      "loss": 0.2059,
+      "step": 3583
+    },
+    {
+      "epoch": 1.6970701390944067,
+      "grad_norm": 1.0844651460647583,
+      "learning_rate": 8.35816936956976e-06,
+      "loss": 0.223,
+      "step": 3584
+    },
+    {
+      "epoch": 1.6975436519680378,
+      "grad_norm": 1.131882905960083,
+      "learning_rate": 8.35312488176378e-06,
+      "loss": 0.2312,
+      "step": 3585
+    },
+    {
+      "epoch": 1.698017164841669,
+      "grad_norm": 1.6975009441375732,
+      "learning_rate": 8.348080824682102e-06,
+      "loss": 0.2397,
+      "step": 3586
+    },
+    {
+      "epoch": 1.6984906777153004,
+      "grad_norm": 1.1549729108810425,
+      "learning_rate": 8.343037199643947e-06,
+      "loss": 0.2051,
+      "step": 3587
+    },
+    {
+      "epoch": 1.6989641905889317,
+      "grad_norm": 1.2071688175201416,
+      "learning_rate": 8.33799400796843e-06,
+      "loss": 0.2313,
+      "step": 3588
+    },
+    {
+      "epoch": 1.699437703462563,
+      "grad_norm": 1.1195539236068726,
+      "learning_rate": 8.332951250974543e-06,
+      "loss": 0.2112,
+      "step": 3589
+    },
+    {
+      "epoch": 1.699911216336194,
+      "grad_norm": 1.7075669765472412,
+      "learning_rate": 8.327908929981175e-06,
+      "loss": 0.2271,
+      "step": 3590
+    },
+    {
+      "epoch": 1.7003847292098255,
+      "grad_norm": 1.0217007398605347,
+      "learning_rate": 8.322867046307096e-06,
+      "loss": 0.1973,
+      "step": 3591
+    },
+    {
+      "epoch": 1.7008582420834566,
+      "grad_norm": 1.1293320655822754,
+      "learning_rate": 8.317825601270954e-06,
+      "loss": 0.2094,
+      "step": 3592
+    },
+    {
+      "epoch": 1.7013317549570879,
+      "grad_norm": 1.1042124032974243,
+      "learning_rate": 8.312784596191308e-06,
+      "loss": 0.2178,
+      "step": 3593
+    },
+    {
+      "epoch": 1.7018052678307192,
+      "grad_norm": 1.4056512117385864,
+      "learning_rate": 8.307744032386571e-06,
+      "loss": 0.2054,
+      "step": 3594
+    },
+    {
+      "epoch": 1.7022787807043502,
+      "grad_norm": 1.0147428512573242,
+      "learning_rate": 8.302703911175057e-06,
+      "loss": 0.1927,
+      "step": 3595
+    },
+    {
+      "epoch": 1.7027522935779817,
+      "grad_norm": 1.2608839273452759,
+      "learning_rate": 8.297664233874958e-06,
+      "loss": 0.2028,
+      "step": 3596
+    },
+    {
+      "epoch": 1.7032258064516128,
+      "grad_norm": 1.170110821723938,
+      "learning_rate": 8.292625001804359e-06,
+      "loss": 0.21,
+      "step": 3597
+    },
+    {
+      "epoch": 1.7036993193252443,
+      "grad_norm": 1.0504069328308105,
+      "learning_rate": 8.28758621628122e-06,
+      "loss": 0.2135,
+      "step": 3598
+    },
+    {
+      "epoch": 1.7041728321988754,
+      "grad_norm": 0.971572995185852,
+      "learning_rate": 8.282547878623384e-06,
+      "loss": 0.2039,
+      "step": 3599
+    },
+    {
+      "epoch": 1.7046463450725067,
+      "grad_norm": 1.2856515645980835,
+      "learning_rate": 8.277509990148584e-06,
+      "loss": 0.2159,
+      "step": 3600
+    },
+    {
+      "epoch": 1.705119857946138,
+      "grad_norm": 0.9921664595603943,
+      "learning_rate": 8.272472552174426e-06,
+      "loss": 0.2166,
+      "step": 3601
+    },
+    {
+      "epoch": 1.705593370819769,
+      "grad_norm": 1.2325758934020996,
+      "learning_rate": 8.267435566018409e-06,
+      "loss": 0.1934,
+      "step": 3602
+    },
+    {
+      "epoch": 1.7060668836934005,
+      "grad_norm": 1.430541753768921,
+      "learning_rate": 8.262399032997906e-06,
+      "loss": 0.2026,
+      "step": 3603
+    },
+    {
+      "epoch": 1.7065403965670316,
+      "grad_norm": 1.1874326467514038,
+      "learning_rate": 8.257362954430172e-06,
+      "loss": 0.1966,
+      "step": 3604
+    },
+    {
+      "epoch": 1.707013909440663,
+      "grad_norm": 1.9422804117202759,
+      "learning_rate": 8.252327331632343e-06,
+      "loss": 0.2389,
+      "step": 3605
+    },
+    {
+      "epoch": 1.7074874223142942,
+      "grad_norm": 1.1945825815200806,
+      "learning_rate": 8.247292165921443e-06,
+      "loss": 0.2184,
+      "step": 3606
+    },
+    {
+      "epoch": 1.7079609351879255,
+      "grad_norm": 1.6779309511184692,
+      "learning_rate": 8.242257458614368e-06,
+      "loss": 0.1963,
+      "step": 3607
+    },
+    {
+      "epoch": 1.7084344480615568,
+      "grad_norm": 1.2465808391571045,
+      "learning_rate": 8.237223211027897e-06,
+      "loss": 0.2136,
+      "step": 3608
+    },
+    {
+      "epoch": 1.7089079609351878,
+      "grad_norm": 2.2780213356018066,
+      "learning_rate": 8.23218942447869e-06,
+      "loss": 0.22,
+      "step": 3609
+    },
+    {
+      "epoch": 1.7093814738088193,
+      "grad_norm": 1.2657932043075562,
+      "learning_rate": 8.227156100283283e-06,
+      "loss": 0.2034,
+      "step": 3610
+    },
+    {
+      "epoch": 1.7098549866824504,
+      "grad_norm": 1.1345802545547485,
+      "learning_rate": 8.222123239758097e-06,
+      "loss": 0.1913,
+      "step": 3611
+    },
+    {
+      "epoch": 1.7103284995560817,
+      "grad_norm": 1.0455700159072876,
+      "learning_rate": 8.217090844219425e-06,
+      "loss": 0.2135,
+      "step": 3612
+    },
+    {
+      "epoch": 1.710802012429713,
+      "grad_norm": 1.8622885942459106,
+      "learning_rate": 8.212058914983445e-06,
+      "loss": 0.2242,
+      "step": 3613
+    },
+    {
+      "epoch": 1.711275525303344,
+      "grad_norm": 1.0153396129608154,
+      "learning_rate": 8.207027453366206e-06,
+      "loss": 0.2115,
+      "step": 3614
+    },
+    {
+      "epoch": 1.7117490381769755,
+      "grad_norm": 1.053398609161377,
+      "learning_rate": 8.201996460683638e-06,
+      "loss": 0.2055,
+      "step": 3615
+    },
+    {
+      "epoch": 1.7122225510506066,
+      "grad_norm": 1.2141062021255493,
+      "learning_rate": 8.19696593825155e-06,
+      "loss": 0.1882,
+      "step": 3616
+    },
+    {
+      "epoch": 1.712696063924238,
+      "grad_norm": 1.2081797122955322,
+      "learning_rate": 8.191935887385628e-06,
+      "loss": 0.2217,
+      "step": 3617
+    },
+    {
+      "epoch": 1.7131695767978692,
+      "grad_norm": 1.0772361755371094,
+      "learning_rate": 8.186906309401431e-06,
+      "loss": 0.2318,
+      "step": 3618
+    },
+    {
+      "epoch": 1.7136430896715005,
+      "grad_norm": 1.2851800918579102,
+      "learning_rate": 8.181877205614398e-06,
+      "loss": 0.2083,
+      "step": 3619
+    },
+    {
+      "epoch": 1.7141166025451318,
+      "grad_norm": 1.463058590888977,
+      "learning_rate": 8.176848577339843e-06,
+      "loss": 0.2274,
+      "step": 3620
+    },
+    {
+      "epoch": 1.7145901154187628,
+      "grad_norm": 1.5855882167816162,
+      "learning_rate": 8.171820425892952e-06,
+      "loss": 0.2146,
+      "step": 3621
+    },
+    {
+      "epoch": 1.7150636282923943,
+      "grad_norm": 1.2082139253616333,
+      "learning_rate": 8.166792752588797e-06,
+      "loss": 0.2106,
+      "step": 3622
+    },
+    {
+      "epoch": 1.7155371411660254,
+      "grad_norm": 0.9224141240119934,
+      "learning_rate": 8.161765558742307e-06,
+      "loss": 0.1913,
+      "step": 3623
+    },
+    {
+      "epoch": 1.7160106540396567,
+      "grad_norm": 1.1134727001190186,
+      "learning_rate": 8.156738845668303e-06,
+      "loss": 0.2006,
+      "step": 3624
+    },
+    {
+      "epoch": 1.716484166913288,
+      "grad_norm": 1.2312144041061401,
+      "learning_rate": 8.15171261468147e-06,
+      "loss": 0.2164,
+      "step": 3625
+    },
+    {
+      "epoch": 1.716957679786919,
+      "grad_norm": 1.2112536430358887,
+      "learning_rate": 8.146686867096376e-06,
+      "loss": 0.2118,
+      "step": 3626
+    },
+    {
+      "epoch": 1.7174311926605506,
+      "grad_norm": 0.9137608408927917,
+      "learning_rate": 8.141661604227448e-06,
+      "loss": 0.1946,
+      "step": 3627
+    },
+    {
+      "epoch": 1.7179047055341816,
+      "grad_norm": 1.5561513900756836,
+      "learning_rate": 8.136636827389002e-06,
+      "loss": 0.2256,
+      "step": 3628
+    },
+    {
+      "epoch": 1.718378218407813,
+      "grad_norm": 0.8672730922698975,
+      "learning_rate": 8.131612537895218e-06,
+      "loss": 0.1872,
+      "step": 3629
+    },
+    {
+      "epoch": 1.7188517312814442,
+      "grad_norm": 1.000693917274475,
+      "learning_rate": 8.126588737060149e-06,
+      "loss": 0.2139,
+      "step": 3630
+    },
+    {
+      "epoch": 1.7193252441550755,
+      "grad_norm": 0.9694979190826416,
+      "learning_rate": 8.121565426197722e-06,
+      "loss": 0.1927,
+      "step": 3631
+    },
+    {
+      "epoch": 1.7197987570287068,
+      "grad_norm": 0.90753573179245,
+      "learning_rate": 8.116542606621736e-06,
+      "loss": 0.2048,
+      "step": 3632
+    },
+    {
+      "epoch": 1.7202722699023378,
+      "grad_norm": 1.230448842048645,
+      "learning_rate": 8.111520279645864e-06,
+      "loss": 0.2298,
+      "step": 3633
+    },
+    {
+      "epoch": 1.7207457827759693,
+      "grad_norm": 1.0563310384750366,
+      "learning_rate": 8.106498446583641e-06,
+      "loss": 0.2158,
+      "step": 3634
+    },
+    {
+      "epoch": 1.7212192956496004,
+      "grad_norm": 1.2370612621307373,
+      "learning_rate": 8.101477108748486e-06,
+      "loss": 0.214,
+      "step": 3635
+    },
+    {
+      "epoch": 1.7216928085232317,
+      "grad_norm": 0.9708116054534912,
+      "learning_rate": 8.096456267453674e-06,
+      "loss": 0.1969,
+      "step": 3636
+    },
+    {
+      "epoch": 1.722166321396863,
+      "grad_norm": 1.1120837926864624,
+      "learning_rate": 8.091435924012365e-06,
+      "loss": 0.2115,
+      "step": 3637
+    },
+    {
+      "epoch": 1.7226398342704943,
+      "grad_norm": 0.838356077671051,
+      "learning_rate": 8.086416079737577e-06,
+      "loss": 0.2015,
+      "step": 3638
+    },
+    {
+      "epoch": 1.7231133471441256,
+      "grad_norm": 1.4962152242660522,
+      "learning_rate": 8.081396735942204e-06,
+      "loss": 0.2272,
+      "step": 3639
+    },
+    {
+      "epoch": 1.7235868600177566,
+      "grad_norm": 1.4194129705429077,
+      "learning_rate": 8.076377893939003e-06,
+      "loss": 0.2158,
+      "step": 3640
+    },
+    {
+      "epoch": 1.7240603728913881,
+      "grad_norm": 1.0114027261734009,
+      "learning_rate": 8.071359555040607e-06,
+      "loss": 0.218,
+      "step": 3641
+    },
+    {
+      "epoch": 1.7245338857650192,
+      "grad_norm": 1.3933831453323364,
+      "learning_rate": 8.066341720559513e-06,
+      "loss": 0.1817,
+      "step": 3642
+    },
+    {
+      "epoch": 1.7250073986386505,
+      "grad_norm": 1.2769237756729126,
+      "learning_rate": 8.06132439180809e-06,
+      "loss": 0.208,
+      "step": 3643
+    },
+    {
+      "epoch": 1.7254809115122818,
+      "grad_norm": 1.4070305824279785,
+      "learning_rate": 8.056307570098566e-06,
+      "loss": 0.2166,
+      "step": 3644
+    },
+    {
+      "epoch": 1.7259544243859128,
+      "grad_norm": 1.0970323085784912,
+      "learning_rate": 8.051291256743048e-06,
+      "loss": 0.1809,
+      "step": 3645
+    },
+    {
+      "epoch": 1.7264279372595444,
+      "grad_norm": 1.1892532110214233,
+      "learning_rate": 8.046275453053501e-06,
+      "loss": 0.2006,
+      "step": 3646
+    },
+    {
+      "epoch": 1.7269014501331754,
+      "grad_norm": 1.8722046613693237,
+      "learning_rate": 8.041260160341766e-06,
+      "loss": 0.2364,
+      "step": 3647
+    },
+    {
+      "epoch": 1.7273749630068067,
+      "grad_norm": 1.7485390901565552,
+      "learning_rate": 8.036245379919535e-06,
+      "loss": 0.1906,
+      "step": 3648
+    },
+    {
+      "epoch": 1.727848475880438,
+      "grad_norm": 2.5090813636779785,
+      "learning_rate": 8.03123111309838e-06,
+      "loss": 0.2261,
+      "step": 3649
+    },
+    {
+      "epoch": 1.7283219887540693,
+      "grad_norm": 1.2415854930877686,
+      "learning_rate": 8.026217361189737e-06,
+      "loss": 0.1927,
+      "step": 3650
+    },
+    {
+      "epoch": 1.7287955016277006,
+      "grad_norm": 1.012861728668213,
+      "learning_rate": 8.0212041255049e-06,
+      "loss": 0.2023,
+      "step": 3651
+    },
+    {
+      "epoch": 1.7292690145013316,
+      "grad_norm": 0.9463083148002625,
+      "learning_rate": 8.016191407355034e-06,
+      "loss": 0.2002,
+      "step": 3652
+    },
+    {
+      "epoch": 1.7297425273749631,
+      "grad_norm": 1.2277804613113403,
+      "learning_rate": 8.011179208051168e-06,
+      "loss": 0.1973,
+      "step": 3653
+    },
+    {
+      "epoch": 1.7302160402485942,
+      "grad_norm": 0.9926773309707642,
+      "learning_rate": 8.006167528904194e-06,
+      "loss": 0.1912,
+      "step": 3654
+    },
+    {
+      "epoch": 1.7306895531222255,
+      "grad_norm": 2.3296539783477783,
+      "learning_rate": 8.001156371224868e-06,
+      "loss": 0.2088,
+      "step": 3655
+    },
+    {
+      "epoch": 1.7311630659958568,
+      "grad_norm": 2.002706289291382,
+      "learning_rate": 7.996145736323807e-06,
+      "loss": 0.2359,
+      "step": 3656
+    },
+    {
+      "epoch": 1.7316365788694879,
+      "grad_norm": 1.5044262409210205,
+      "learning_rate": 7.991135625511503e-06,
+      "loss": 0.2037,
+      "step": 3657
+    },
+    {
+      "epoch": 1.7321100917431194,
+      "grad_norm": 1.2724112272262573,
+      "learning_rate": 7.986126040098291e-06,
+      "loss": 0.2139,
+      "step": 3658
+    },
+    {
+      "epoch": 1.7325836046167504,
+      "grad_norm": 1.1246263980865479,
+      "learning_rate": 7.981116981394388e-06,
+      "loss": 0.2028,
+      "step": 3659
+    },
+    {
+      "epoch": 1.7330571174903817,
+      "grad_norm": 0.9579681158065796,
+      "learning_rate": 7.97610845070986e-06,
+      "loss": 0.1982,
+      "step": 3660
+    },
+    {
+      "epoch": 1.733530630364013,
+      "grad_norm": 1.939510703086853,
+      "learning_rate": 7.971100449354643e-06,
+      "loss": 0.2115,
+      "step": 3661
+    },
+    {
+      "epoch": 1.7340041432376443,
+      "grad_norm": 1.2159603834152222,
+      "learning_rate": 7.96609297863853e-06,
+      "loss": 0.2167,
+      "step": 3662
+    },
+    {
+      "epoch": 1.7344776561112756,
+      "grad_norm": 1.5756498575210571,
+      "learning_rate": 7.961086039871178e-06,
+      "loss": 0.2003,
+      "step": 3663
+    },
+    {
+      "epoch": 1.7349511689849066,
+      "grad_norm": 1.5483087301254272,
+      "learning_rate": 7.956079634362101e-06,
+      "loss": 0.1893,
+      "step": 3664
+    },
+    {
+      "epoch": 1.7354246818585382,
+      "grad_norm": 0.9675044417381287,
+      "learning_rate": 7.951073763420679e-06,
+      "loss": 0.2116,
+      "step": 3665
+    },
+    {
+      "epoch": 1.7358981947321692,
+      "grad_norm": 1.4665745496749878,
+      "learning_rate": 7.946068428356146e-06,
+      "loss": 0.2102,
+      "step": 3666
+    },
+    {
+      "epoch": 1.7363717076058005,
+      "grad_norm": 1.4936654567718506,
+      "learning_rate": 7.941063630477603e-06,
+      "loss": 0.2233,
+      "step": 3667
+    },
+    {
+      "epoch": 1.7368452204794318,
+      "grad_norm": 1.196628212928772,
+      "learning_rate": 7.936059371094004e-06,
+      "loss": 0.205,
+      "step": 3668
+    },
+    {
+      "epoch": 1.737318733353063,
+      "grad_norm": 1.1913931369781494,
+      "learning_rate": 7.931055651514165e-06,
+      "loss": 0.2089,
+      "step": 3669
+    },
+    {
+      "epoch": 1.7377922462266944,
+      "grad_norm": 1.4205213785171509,
+      "learning_rate": 7.926052473046756e-06,
+      "loss": 0.233,
+      "step": 3670
+    },
+    {
+      "epoch": 1.7382657591003254,
+      "grad_norm": 1.5477283000946045,
+      "learning_rate": 7.921049837000318e-06,
+      "loss": 0.2258,
+      "step": 3671
+    },
+    {
+      "epoch": 1.738739271973957,
+      "grad_norm": 2.6856656074523926,
+      "learning_rate": 7.916047744683244e-06,
+      "loss": 0.216,
+      "step": 3672
+    },
+    {
+      "epoch": 1.739212784847588,
+      "grad_norm": 1.0358319282531738,
+      "learning_rate": 7.911046197403775e-06,
+      "loss": 0.2089,
+      "step": 3673
+    },
+    {
+      "epoch": 1.7396862977212193,
+      "grad_norm": 1.1069475412368774,
+      "learning_rate": 7.90604519647002e-06,
+      "loss": 0.2109,
+      "step": 3674
+    },
+    {
+      "epoch": 1.7401598105948506,
+      "grad_norm": 1.5351327657699585,
+      "learning_rate": 7.901044743189943e-06,
+      "loss": 0.2081,
+      "step": 3675
+    },
+    {
+      "epoch": 1.7406333234684817,
+      "grad_norm": 1.5493526458740234,
+      "learning_rate": 7.896044838871365e-06,
+      "loss": 0.2187,
+      "step": 3676
+    },
+    {
+      "epoch": 1.7411068363421132,
+      "grad_norm": 1.2532929182052612,
+      "learning_rate": 7.891045484821961e-06,
+      "loss": 0.2365,
+      "step": 3677
+    },
+    {
+      "epoch": 1.7415803492157442,
+      "grad_norm": 1.0188684463500977,
+      "learning_rate": 7.886046682349267e-06,
+      "loss": 0.2313,
+      "step": 3678
+    },
+    {
+      "epoch": 1.7420538620893755,
+      "grad_norm": 1.1516002416610718,
+      "learning_rate": 7.88104843276067e-06,
+      "loss": 0.2071,
+      "step": 3679
+    },
+    {
+      "epoch": 1.7425273749630068,
+      "grad_norm": 1.1517740488052368,
+      "learning_rate": 7.876050737363411e-06,
+      "loss": 0.2039,
+      "step": 3680
+    },
+    {
+      "epoch": 1.743000887836638,
+      "grad_norm": 1.396052598953247,
+      "learning_rate": 7.871053597464593e-06,
+      "loss": 0.2138,
+      "step": 3681
+    },
+    {
+      "epoch": 1.7434744007102694,
+      "grad_norm": 1.08030104637146,
+      "learning_rate": 7.86605701437117e-06,
+      "loss": 0.2316,
+      "step": 3682
+    },
+    {
+      "epoch": 1.7439479135839004,
+      "grad_norm": 2.2446389198303223,
+      "learning_rate": 7.861060989389947e-06,
+      "loss": 0.2267,
+      "step": 3683
+    },
+    {
+      "epoch": 1.744421426457532,
+      "grad_norm": 1.3251670598983765,
+      "learning_rate": 7.856065523827586e-06,
+      "loss": 0.1961,
+      "step": 3684
+    },
+    {
+      "epoch": 1.744894939331163,
+      "grad_norm": 1.188880205154419,
+      "learning_rate": 7.851070618990607e-06,
+      "loss": 0.2184,
+      "step": 3685
+    },
+    {
+      "epoch": 1.7453684522047943,
+      "grad_norm": 1.126604676246643,
+      "learning_rate": 7.846076276185372e-06,
+      "loss": 0.24,
+      "step": 3686
+    },
+    {
+      "epoch": 1.7458419650784256,
+      "grad_norm": 1.0912270545959473,
+      "learning_rate": 7.841082496718112e-06,
+      "loss": 0.2049,
+      "step": 3687
+    },
+    {
+      "epoch": 1.7463154779520567,
+      "grad_norm": 1.5733311176300049,
+      "learning_rate": 7.836089281894895e-06,
+      "loss": 0.2068,
+      "step": 3688
+    },
+    {
+      "epoch": 1.7467889908256882,
+      "grad_norm": 0.9812442064285278,
+      "learning_rate": 7.831096633021651e-06,
+      "loss": 0.2069,
+      "step": 3689
+    },
+    {
+      "epoch": 1.7472625036993192,
+      "grad_norm": 1.4512606859207153,
+      "learning_rate": 7.826104551404161e-06,
+      "loss": 0.2207,
+      "step": 3690
+    },
+    {
+      "epoch": 1.7477360165729505,
+      "grad_norm": 2.106611967086792,
+      "learning_rate": 7.821113038348052e-06,
+      "loss": 0.2208,
+      "step": 3691
+    },
+    {
+      "epoch": 1.7482095294465818,
+      "grad_norm": 1.069126009941101,
+      "learning_rate": 7.81612209515881e-06,
+      "loss": 0.2355,
+      "step": 3692
+    },
+    {
+      "epoch": 1.748683042320213,
+      "grad_norm": 1.088439702987671,
+      "learning_rate": 7.811131723141763e-06,
+      "loss": 0.2171,
+      "step": 3693
+    },
+    {
+      "epoch": 1.7491565551938444,
+      "grad_norm": 1.1019070148468018,
+      "learning_rate": 7.806141923602098e-06,
+      "loss": 0.2021,
+      "step": 3694
+    },
+    {
+      "epoch": 1.7496300680674755,
+      "grad_norm": 1.3674668073654175,
+      "learning_rate": 7.801152697844849e-06,
+      "loss": 0.2084,
+      "step": 3695
+    },
+    {
+      "epoch": 1.750103580941107,
+      "grad_norm": 1.3417162895202637,
+      "learning_rate": 7.796164047174898e-06,
+      "loss": 0.2,
+      "step": 3696
+    },
+    {
+      "epoch": 1.750577093814738,
+      "grad_norm": 1.5355737209320068,
+      "learning_rate": 7.79117597289698e-06,
+      "loss": 0.2159,
+      "step": 3697
+    },
+    {
+      "epoch": 1.7510506066883693,
+      "grad_norm": 1.4953505992889404,
+      "learning_rate": 7.786188476315678e-06,
+      "loss": 0.2188,
+      "step": 3698
+    },
+    {
+      "epoch": 1.7515241195620006,
+      "grad_norm": 1.4030543565750122,
+      "learning_rate": 7.781201558735423e-06,
+      "loss": 0.2026,
+      "step": 3699
+    },
+    {
+      "epoch": 1.751997632435632,
+      "grad_norm": 1.5488836765289307,
+      "learning_rate": 7.776215221460496e-06,
+      "loss": 0.2155,
+      "step": 3700
+    },
+    {
+      "epoch": 1.7524711453092632,
+      "grad_norm": 1.2541826963424683,
+      "learning_rate": 7.771229465795024e-06,
+      "loss": 0.2176,
+      "step": 3701
+    },
+    {
+      "epoch": 1.7529446581828942,
+      "grad_norm": 0.9930739998817444,
+      "learning_rate": 7.766244293042983e-06,
+      "loss": 0.2305,
+      "step": 3702
+    },
+    {
+      "epoch": 1.7534181710565258,
+      "grad_norm": 1.1047271490097046,
+      "learning_rate": 7.7612597045082e-06,
+      "loss": 0.2213,
+      "step": 3703
+    },
+    {
+      "epoch": 1.7538916839301568,
+      "grad_norm": 1.2397290468215942,
+      "learning_rate": 7.75627570149434e-06,
+      "loss": 0.2186,
+      "step": 3704
+    },
+    {
+      "epoch": 1.754365196803788,
+      "grad_norm": 1.353489637374878,
+      "learning_rate": 7.751292285304928e-06,
+      "loss": 0.2263,
+      "step": 3705
+    },
+    {
+      "epoch": 1.7548387096774194,
+      "grad_norm": 1.1363734006881714,
+      "learning_rate": 7.746309457243324e-06,
+      "loss": 0.2071,
+      "step": 3706
+    },
+    {
+      "epoch": 1.7553122225510505,
+      "grad_norm": 1.537649393081665,
+      "learning_rate": 7.74132721861274e-06,
+      "loss": 0.2063,
+      "step": 3707
+    },
+    {
+      "epoch": 1.755785735424682,
+      "grad_norm": 1.076953649520874,
+      "learning_rate": 7.73634557071623e-06,
+      "loss": 0.2097,
+      "step": 3708
+    },
+    {
+      "epoch": 1.756259248298313,
+      "grad_norm": 1.2859452962875366,
+      "learning_rate": 7.731364514856698e-06,
+      "loss": 0.2212,
+      "step": 3709
+    },
+    {
+      "epoch": 1.7567327611719443,
+      "grad_norm": 1.1581248044967651,
+      "learning_rate": 7.726384052336893e-06,
+      "loss": 0.2189,
+      "step": 3710
+    },
+    {
+      "epoch": 1.7572062740455756,
+      "grad_norm": 1.3928226232528687,
+      "learning_rate": 7.721404184459405e-06,
+      "loss": 0.2324,
+      "step": 3711
+    },
+    {
+      "epoch": 1.757679786919207,
+      "grad_norm": 1.322218894958496,
+      "learning_rate": 7.716424912526672e-06,
+      "loss": 0.201,
+      "step": 3712
+    },
+    {
+      "epoch": 1.7581532997928382,
+      "grad_norm": 1.102550745010376,
+      "learning_rate": 7.711446237840971e-06,
+      "loss": 0.2371,
+      "step": 3713
+    },
+    {
+      "epoch": 1.7586268126664693,
+      "grad_norm": 1.2192142009735107,
+      "learning_rate": 7.70646816170443e-06,
+      "loss": 0.2255,
+      "step": 3714
+    },
+    {
+      "epoch": 1.7591003255401008,
+      "grad_norm": 1.6167590618133545,
+      "learning_rate": 7.701490685419014e-06,
+      "loss": 0.2051,
+      "step": 3715
+    },
+    {
+      "epoch": 1.7595738384137318,
+      "grad_norm": 1.1392605304718018,
+      "learning_rate": 7.696513810286534e-06,
+      "loss": 0.2178,
+      "step": 3716
+    },
+    {
+      "epoch": 1.7600473512873631,
+      "grad_norm": 0.9116237759590149,
+      "learning_rate": 7.69153753760865e-06,
+      "loss": 0.2093,
+      "step": 3717
+    },
+    {
+      "epoch": 1.7605208641609944,
+      "grad_norm": 1.1358370780944824,
+      "learning_rate": 7.686561868686848e-06,
+      "loss": 0.2036,
+      "step": 3718
+    },
+    {
+      "epoch": 1.7609943770346255,
+      "grad_norm": 1.4042670726776123,
+      "learning_rate": 7.681586804822471e-06,
+      "loss": 0.2023,
+      "step": 3719
+    },
+    {
+      "epoch": 1.761467889908257,
+      "grad_norm": 1.1344070434570312,
+      "learning_rate": 7.676612347316702e-06,
+      "loss": 0.225,
+      "step": 3720
+    },
+    {
+      "epoch": 1.761941402781888,
+      "grad_norm": 1.1140799522399902,
+      "learning_rate": 7.671638497470558e-06,
+      "loss": 0.1941,
+      "step": 3721
+    },
+    {
+      "epoch": 1.7624149156555193,
+      "grad_norm": 0.9498646259307861,
+      "learning_rate": 7.666665256584902e-06,
+      "loss": 0.2175,
+      "step": 3722
+    },
+    {
+      "epoch": 1.7628884285291506,
+      "grad_norm": 1.3666812181472778,
+      "learning_rate": 7.66169262596044e-06,
+      "loss": 0.2125,
+      "step": 3723
+    },
+    {
+      "epoch": 1.763361941402782,
+      "grad_norm": 1.0839076042175293,
+      "learning_rate": 7.656720606897711e-06,
+      "loss": 0.2253,
+      "step": 3724
+    },
+    {
+      "epoch": 1.7638354542764132,
+      "grad_norm": 0.9082208871841431,
+      "learning_rate": 7.651749200697104e-06,
+      "loss": 0.1788,
+      "step": 3725
+    },
+    {
+      "epoch": 1.7643089671500443,
+      "grad_norm": 0.9702308773994446,
+      "learning_rate": 7.646778408658839e-06,
+      "loss": 0.2188,
+      "step": 3726
+    },
+    {
+      "epoch": 1.7647824800236758,
+      "grad_norm": 0.9975780844688416,
+      "learning_rate": 7.64180823208298e-06,
+      "loss": 0.2132,
+      "step": 3727
+    },
+    {
+      "epoch": 1.7652559928973068,
+      "grad_norm": 2.509084463119507,
+      "learning_rate": 7.636838672269425e-06,
+      "loss": 0.2237,
+      "step": 3728
+    },
+    {
+      "epoch": 1.7657295057709381,
+      "grad_norm": 1.6179232597351074,
+      "learning_rate": 7.63186973051792e-06,
+      "loss": 0.2145,
+      "step": 3729
+    },
+    {
+      "epoch": 1.7662030186445694,
+      "grad_norm": 1.5962454080581665,
+      "learning_rate": 7.626901408128039e-06,
+      "loss": 0.1912,
+      "step": 3730
+    },
+    {
+      "epoch": 1.7666765315182007,
+      "grad_norm": 0.9649477005004883,
+      "learning_rate": 7.621933706399202e-06,
+      "loss": 0.2245,
+      "step": 3731
+    },
+    {
+      "epoch": 1.767150044391832,
+      "grad_norm": 1.120977520942688,
+      "learning_rate": 7.616966626630663e-06,
+      "loss": 0.2079,
+      "step": 3732
+    },
+    {
+      "epoch": 1.767623557265463,
+      "grad_norm": 1.2438452243804932,
+      "learning_rate": 7.612000170121513e-06,
+      "loss": 0.2257,
+      "step": 3733
+    },
+    {
+      "epoch": 1.7680970701390946,
+      "grad_norm": 0.9499486684799194,
+      "learning_rate": 7.607034338170681e-06,
+      "loss": 0.2267,
+      "step": 3734
+    },
+    {
+      "epoch": 1.7685705830127256,
+      "grad_norm": 1.2596803903579712,
+      "learning_rate": 7.6020691320769325e-06,
+      "loss": 0.2202,
+      "step": 3735
+    },
+    {
+      "epoch": 1.769044095886357,
+      "grad_norm": 2.5322797298431396,
+      "learning_rate": 7.597104553138872e-06,
+      "loss": 0.2052,
+      "step": 3736
+    },
+    {
+      "epoch": 1.7695176087599882,
+      "grad_norm": 1.7047964334487915,
+      "learning_rate": 7.592140602654931e-06,
+      "loss": 0.2073,
+      "step": 3737
+    },
+    {
+      "epoch": 1.7699911216336193,
+      "grad_norm": 0.9767723083496094,
+      "learning_rate": 7.587177281923388e-06,
+      "loss": 0.2141,
+      "step": 3738
+    },
+    {
+      "epoch": 1.7704646345072508,
+      "grad_norm": 1.1681851148605347,
+      "learning_rate": 7.582214592242348e-06,
+      "loss": 0.2123,
+      "step": 3739
+    },
+    {
+      "epoch": 1.7709381473808818,
+      "grad_norm": 1.0731382369995117,
+      "learning_rate": 7.577252534909758e-06,
+      "loss": 0.1867,
+      "step": 3740
+    },
+    {
+      "epoch": 1.7714116602545131,
+      "grad_norm": 1.2089591026306152,
+      "learning_rate": 7.5722911112233956e-06,
+      "loss": 0.1974,
+      "step": 3741
+    },
+    {
+      "epoch": 1.7718851731281444,
+      "grad_norm": 1.1869101524353027,
+      "learning_rate": 7.56733032248087e-06,
+      "loss": 0.2027,
+      "step": 3742
+    },
+    {
+      "epoch": 1.7723586860017757,
+      "grad_norm": 1.4566733837127686,
+      "learning_rate": 7.562370169979633e-06,
+      "loss": 0.2055,
+      "step": 3743
+    },
+    {
+      "epoch": 1.772832198875407,
+      "grad_norm": 1.1467299461364746,
+      "learning_rate": 7.55741065501696e-06,
+      "loss": 0.2077,
+      "step": 3744
+    },
+    {
+      "epoch": 1.773305711749038,
+      "grad_norm": 1.3079264163970947,
+      "learning_rate": 7.552451778889967e-06,
+      "loss": 0.1882,
+      "step": 3745
+    },
+    {
+      "epoch": 1.7737792246226696,
+      "grad_norm": 1.1860123872756958,
+      "learning_rate": 7.547493542895601e-06,
+      "loss": 0.1968,
+      "step": 3746
+    },
+    {
+      "epoch": 1.7742527374963006,
+      "grad_norm": 1.007750391960144,
+      "learning_rate": 7.542535948330636e-06,
+      "loss": 0.2064,
+      "step": 3747
+    },
+    {
+      "epoch": 1.774726250369932,
+      "grad_norm": 1.3178592920303345,
+      "learning_rate": 7.537578996491683e-06,
+      "loss": 0.2012,
+      "step": 3748
+    },
+    {
+      "epoch": 1.7751997632435632,
+      "grad_norm": 1.3056480884552002,
+      "learning_rate": 7.532622688675193e-06,
+      "loss": 0.2228,
+      "step": 3749
+    },
+    {
+      "epoch": 1.7756732761171943,
+      "grad_norm": 1.8601652383804321,
+      "learning_rate": 7.527667026177434e-06,
+      "loss": 0.186,
+      "step": 3750
+    },
+    {
+      "epoch": 1.7761467889908258,
+      "grad_norm": 2.2548739910125732,
+      "learning_rate": 7.522712010294516e-06,
+      "loss": 0.2139,
+      "step": 3751
+    },
+    {
+      "epoch": 1.7766203018644569,
+      "grad_norm": 1.1233947277069092,
+      "learning_rate": 7.517757642322372e-06,
+      "loss": 0.2151,
+      "step": 3752
+    },
+    {
+      "epoch": 1.7770938147380881,
+      "grad_norm": 1.173248291015625,
+      "learning_rate": 7.5128039235567686e-06,
+      "loss": 0.2143,
+      "step": 3753
+    },
+    {
+      "epoch": 1.7775673276117194,
+      "grad_norm": 1.1922003030776978,
+      "learning_rate": 7.507850855293305e-06,
+      "loss": 0.2198,
+      "step": 3754
+    },
+    {
+      "epoch": 1.7780408404853507,
+      "grad_norm": 1.0876096487045288,
+      "learning_rate": 7.502898438827408e-06,
+      "loss": 0.2106,
+      "step": 3755
+    },
+    {
+      "epoch": 1.778514353358982,
+      "grad_norm": 1.1909540891647339,
+      "learning_rate": 7.497946675454334e-06,
+      "loss": 0.2067,
+      "step": 3756
+    },
+    {
+      "epoch": 1.778987866232613,
+      "grad_norm": 1.084885597229004,
+      "learning_rate": 7.49299556646917e-06,
+      "loss": 0.1632,
+      "step": 3757
+    },
+    {
+      "epoch": 1.7794613791062446,
+      "grad_norm": 0.9690635800361633,
+      "learning_rate": 7.48804511316683e-06,
+      "loss": 0.2316,
+      "step": 3758
+    },
+    {
+      "epoch": 1.7799348919798756,
+      "grad_norm": 1.4600977897644043,
+      "learning_rate": 7.483095316842057e-06,
+      "loss": 0.2089,
+      "step": 3759
+    },
+    {
+      "epoch": 1.780408404853507,
+      "grad_norm": 1.4429023265838623,
+      "learning_rate": 7.478146178789423e-06,
+      "loss": 0.2065,
+      "step": 3760
+    },
+    {
+      "epoch": 1.7808819177271382,
+      "grad_norm": 1.128029227256775,
+      "learning_rate": 7.47319770030333e-06,
+      "loss": 0.2118,
+      "step": 3761
+    },
+    {
+      "epoch": 1.7813554306007695,
+      "grad_norm": 1.9069571495056152,
+      "learning_rate": 7.4682498826779984e-06,
+      "loss": 0.225,
+      "step": 3762
+    },
+    {
+      "epoch": 1.7818289434744008,
+      "grad_norm": 1.254188895225525,
+      "learning_rate": 7.463302727207486e-06,
+      "loss": 0.204,
+      "step": 3763
+    },
+    {
+      "epoch": 1.7823024563480319,
+      "grad_norm": 1.4239470958709717,
+      "learning_rate": 7.458356235185674e-06,
+      "loss": 0.2052,
+      "step": 3764
+    },
+    {
+      "epoch": 1.7827759692216634,
+      "grad_norm": 0.9776591658592224,
+      "learning_rate": 7.45341040790627e-06,
+      "loss": 0.2175,
+      "step": 3765
+    },
+    {
+      "epoch": 1.7832494820952944,
+      "grad_norm": 1.543384075164795,
+      "learning_rate": 7.4484652466628036e-06,
+      "loss": 0.2158,
+      "step": 3766
+    },
+    {
+      "epoch": 1.7837229949689257,
+      "grad_norm": 1.3612316846847534,
+      "learning_rate": 7.443520752748639e-06,
+      "loss": 0.2223,
+      "step": 3767
+    },
+    {
+      "epoch": 1.784196507842557,
+      "grad_norm": 1.331251859664917,
+      "learning_rate": 7.438576927456958e-06,
+      "loss": 0.2304,
+      "step": 3768
+    },
+    {
+      "epoch": 1.784670020716188,
+      "grad_norm": 1.209085464477539,
+      "learning_rate": 7.433633772080772e-06,
+      "loss": 0.2219,
+      "step": 3769
+    },
+    {
+      "epoch": 1.7851435335898196,
+      "grad_norm": 2.226287364959717,
+      "learning_rate": 7.428691287912915e-06,
+      "loss": 0.2201,
+      "step": 3770
+    },
+    {
+      "epoch": 1.7856170464634507,
+      "grad_norm": 1.0809687376022339,
+      "learning_rate": 7.423749476246046e-06,
+      "loss": 0.2028,
+      "step": 3771
+    },
+    {
+      "epoch": 1.786090559337082,
+      "grad_norm": 1.1204088926315308,
+      "learning_rate": 7.4188083383726475e-06,
+      "loss": 0.2243,
+      "step": 3772
+    },
+    {
+      "epoch": 1.7865640722107132,
+      "grad_norm": 1.2815765142440796,
+      "learning_rate": 7.413867875585026e-06,
+      "loss": 0.2129,
+      "step": 3773
+    },
+    {
+      "epoch": 1.7870375850843445,
+      "grad_norm": 1.0807607173919678,
+      "learning_rate": 7.408928089175314e-06,
+      "loss": 0.2257,
+      "step": 3774
+    },
+    {
+      "epoch": 1.7875110979579758,
+      "grad_norm": 1.4838930368423462,
+      "learning_rate": 7.403988980435461e-06,
+      "loss": 0.2331,
+      "step": 3775
+    },
+    {
+      "epoch": 1.7879846108316069,
+      "grad_norm": 0.974097490310669,
+      "learning_rate": 7.399050550657249e-06,
+      "loss": 0.2171,
+      "step": 3776
+    },
+    {
+      "epoch": 1.7884581237052384,
+      "grad_norm": 1.6029471158981323,
+      "learning_rate": 7.394112801132271e-06,
+      "loss": 0.1972,
+      "step": 3777
+    },
+    {
+      "epoch": 1.7889316365788694,
+      "grad_norm": 1.0069618225097656,
+      "learning_rate": 7.389175733151953e-06,
+      "loss": 0.202,
+      "step": 3778
+    },
+    {
+      "epoch": 1.7894051494525007,
+      "grad_norm": 1.8258870840072632,
+      "learning_rate": 7.384239348007534e-06,
+      "loss": 0.1895,
+      "step": 3779
+    },
+    {
+      "epoch": 1.789878662326132,
+      "grad_norm": 1.2527416944503784,
+      "learning_rate": 7.379303646990081e-06,
+      "loss": 0.2081,
+      "step": 3780
+    },
+    {
+      "epoch": 1.790352175199763,
+      "grad_norm": 1.7757433652877808,
+      "learning_rate": 7.374368631390474e-06,
+      "loss": 0.2195,
+      "step": 3781
+    },
+    {
+      "epoch": 1.7908256880733946,
+      "grad_norm": 0.9863547086715698,
+      "learning_rate": 7.369434302499423e-06,
+      "loss": 0.2148,
+      "step": 3782
+    },
+    {
+      "epoch": 1.7912992009470257,
+      "grad_norm": 1.3424357175827026,
+      "learning_rate": 7.364500661607452e-06,
+      "loss": 0.2289,
+      "step": 3783
+    },
+    {
+      "epoch": 1.791772713820657,
+      "grad_norm": 1.5163919925689697,
+      "learning_rate": 7.359567710004907e-06,
+      "loss": 0.2079,
+      "step": 3784
+    },
+    {
+      "epoch": 1.7922462266942882,
+      "grad_norm": 1.132380723953247,
+      "learning_rate": 7.354635448981955e-06,
+      "loss": 0.2004,
+      "step": 3785
+    },
+    {
+      "epoch": 1.7927197395679195,
+      "grad_norm": 1.174932599067688,
+      "learning_rate": 7.349703879828582e-06,
+      "loss": 0.1928,
+      "step": 3786
+    },
+    {
+      "epoch": 1.7931932524415508,
+      "grad_norm": 1.1085500717163086,
+      "learning_rate": 7.344773003834589e-06,
+      "loss": 0.1902,
+      "step": 3787
+    },
+    {
+      "epoch": 1.7936667653151819,
+      "grad_norm": 1.5855284929275513,
+      "learning_rate": 7.339842822289602e-06,
+      "loss": 0.2048,
+      "step": 3788
+    },
+    {
+      "epoch": 1.7941402781888134,
+      "grad_norm": 1.047294020652771,
+      "learning_rate": 7.334913336483063e-06,
+      "loss": 0.1961,
+      "step": 3789
+    },
+    {
+      "epoch": 1.7946137910624445,
+      "grad_norm": 1.2375208139419556,
+      "learning_rate": 7.329984547704231e-06,
+      "loss": 0.1988,
+      "step": 3790
+    },
+    {
+      "epoch": 1.7950873039360757,
+      "grad_norm": 1.0275269746780396,
+      "learning_rate": 7.3250564572421814e-06,
+      "loss": 0.239,
+      "step": 3791
+    },
+    {
+      "epoch": 1.795560816809707,
+      "grad_norm": 1.0698773860931396,
+      "learning_rate": 7.320129066385811e-06,
+      "loss": 0.2176,
+      "step": 3792
+    },
+    {
+      "epoch": 1.7960343296833383,
+      "grad_norm": 1.3771909475326538,
+      "learning_rate": 7.315202376423829e-06,
+      "loss": 0.2128,
+      "step": 3793
+    },
+    {
+      "epoch": 1.7965078425569696,
+      "grad_norm": 1.1806304454803467,
+      "learning_rate": 7.3102763886447645e-06,
+      "loss": 0.1979,
+      "step": 3794
+    },
+    {
+      "epoch": 1.7969813554306007,
+      "grad_norm": 1.011577844619751,
+      "learning_rate": 7.305351104336963e-06,
+      "loss": 0.1838,
+      "step": 3795
+    },
+    {
+      "epoch": 1.7974548683042322,
+      "grad_norm": 2.0910136699676514,
+      "learning_rate": 7.3004265247885865e-06,
+      "loss": 0.1845,
+      "step": 3796
+    },
+    {
+      "epoch": 1.7979283811778632,
+      "grad_norm": 0.9872161746025085,
+      "learning_rate": 7.295502651287607e-06,
+      "loss": 0.2177,
+      "step": 3797
+    },
+    {
+      "epoch": 1.7984018940514945,
+      "grad_norm": 1.0551612377166748,
+      "learning_rate": 7.290579485121818e-06,
+      "loss": 0.2106,
+      "step": 3798
+    },
+    {
+      "epoch": 1.7988754069251258,
+      "grad_norm": 0.9977513551712036,
+      "learning_rate": 7.285657027578827e-06,
+      "loss": 0.2187,
+      "step": 3799
+    },
+    {
+      "epoch": 1.7993489197987569,
+      "grad_norm": 1.0159823894500732,
+      "learning_rate": 7.280735279946054e-06,
+      "loss": 0.22,
+      "step": 3800
+    },
+    {
+      "epoch": 1.7998224326723884,
+      "grad_norm": 1.2964645624160767,
+      "learning_rate": 7.275814243510736e-06,
+      "loss": 0.2177,
+      "step": 3801
+    },
+    {
+      "epoch": 1.8002959455460195,
+      "grad_norm": 1.3088321685791016,
+      "learning_rate": 7.270893919559922e-06,
+      "loss": 0.2346,
+      "step": 3802
+    },
+    {
+      "epoch": 1.8007694584196507,
+      "grad_norm": 1.1884821653366089,
+      "learning_rate": 7.265974309380475e-06,
+      "loss": 0.1938,
+      "step": 3803
+    },
+    {
+      "epoch": 1.801242971293282,
+      "grad_norm": 1.544429063796997,
+      "learning_rate": 7.26105541425907e-06,
+      "loss": 0.2023,
+      "step": 3804
+    },
+    {
+      "epoch": 1.8017164841669133,
+      "grad_norm": 0.9471104741096497,
+      "learning_rate": 7.2561372354822035e-06,
+      "loss": 0.2057,
+      "step": 3805
+    },
+    {
+      "epoch": 1.8021899970405446,
+      "grad_norm": 1.2937753200531006,
+      "learning_rate": 7.251219774336169e-06,
+      "loss": 0.2245,
+      "step": 3806
+    },
+    {
+      "epoch": 1.8026635099141757,
+      "grad_norm": 1.004191517829895,
+      "learning_rate": 7.246303032107084e-06,
+      "loss": 0.1868,
+      "step": 3807
+    },
+    {
+      "epoch": 1.8031370227878072,
+      "grad_norm": 1.476674199104309,
+      "learning_rate": 7.2413870100808755e-06,
+      "loss": 0.2203,
+      "step": 3808
+    },
+    {
+      "epoch": 1.8036105356614383,
+      "grad_norm": 0.9113790988922119,
+      "learning_rate": 7.2364717095432825e-06,
+      "loss": 0.2114,
+      "step": 3809
+    },
+    {
+      "epoch": 1.8040840485350695,
+      "grad_norm": 1.3018831014633179,
+      "learning_rate": 7.231557131779854e-06,
+      "loss": 0.2191,
+      "step": 3810
+    },
+    {
+      "epoch": 1.8045575614087008,
+      "grad_norm": 1.301070213317871,
+      "learning_rate": 7.226643278075948e-06,
+      "loss": 0.1999,
+      "step": 3811
+    },
+    {
+      "epoch": 1.805031074282332,
+      "grad_norm": 1.776887059211731,
+      "learning_rate": 7.2217301497167405e-06,
+      "loss": 0.2103,
+      "step": 3812
+    },
+    {
+      "epoch": 1.8055045871559634,
+      "grad_norm": 2.6810529232025146,
+      "learning_rate": 7.216817747987208e-06,
+      "loss": 0.2098,
+      "step": 3813
+    },
+    {
+      "epoch": 1.8059781000295945,
+      "grad_norm": 1.186989426612854,
+      "learning_rate": 7.2119060741721435e-06,
+      "loss": 0.1888,
+      "step": 3814
+    },
+    {
+      "epoch": 1.8064516129032258,
+      "grad_norm": 0.9349979162216187,
+      "learning_rate": 7.206995129556151e-06,
+      "loss": 0.2345,
+      "step": 3815
+    },
+    {
+      "epoch": 1.806925125776857,
+      "grad_norm": 1.71229887008667,
+      "learning_rate": 7.202084915423636e-06,
+      "loss": 0.2431,
+      "step": 3816
+    },
+    {
+      "epoch": 1.8073986386504883,
+      "grad_norm": 1.5204129219055176,
+      "learning_rate": 7.197175433058818e-06,
+      "loss": 0.2119,
+      "step": 3817
+    },
+    {
+      "epoch": 1.8078721515241196,
+      "grad_norm": 1.7288326025009155,
+      "learning_rate": 7.192266683745728e-06,
+      "loss": 0.22,
+      "step": 3818
+    },
+    {
+      "epoch": 1.8083456643977507,
+      "grad_norm": 1.4279898405075073,
+      "learning_rate": 7.187358668768198e-06,
+      "loss": 0.2029,
+      "step": 3819
+    },
+    {
+      "epoch": 1.8088191772713822,
+      "grad_norm": 1.4333093166351318,
+      "learning_rate": 7.182451389409877e-06,
+      "loss": 0.2323,
+      "step": 3820
+    },
+    {
+      "epoch": 1.8092926901450133,
+      "grad_norm": 1.506713628768921,
+      "learning_rate": 7.177544846954212e-06,
+      "loss": 0.2408,
+      "step": 3821
+    },
+    {
+      "epoch": 1.8097662030186445,
+      "grad_norm": 1.236220121383667,
+      "learning_rate": 7.172639042684464e-06,
+      "loss": 0.2193,
+      "step": 3822
+    },
+    {
+      "epoch": 1.8102397158922758,
+      "grad_norm": 0.997626006603241,
+      "learning_rate": 7.1677339778836975e-06,
+      "loss": 0.2065,
+      "step": 3823
+    },
+    {
+      "epoch": 1.8107132287659071,
+      "grad_norm": 1.5409119129180908,
+      "learning_rate": 7.162829653834787e-06,
+      "loss": 0.205,
+      "step": 3824
+    },
+    {
+      "epoch": 1.8111867416395384,
+      "grad_norm": 1.3550500869750977,
+      "learning_rate": 7.157926071820411e-06,
+      "loss": 0.2062,
+      "step": 3825
+    },
+    {
+      "epoch": 1.8116602545131695,
+      "grad_norm": 1.6117483377456665,
+      "learning_rate": 7.153023233123047e-06,
+      "loss": 0.2185,
+      "step": 3826
+    },
+    {
+      "epoch": 1.812133767386801,
+      "grad_norm": 1.3827286958694458,
+      "learning_rate": 7.148121139024995e-06,
+      "loss": 0.1976,
+      "step": 3827
+    },
+    {
+      "epoch": 1.812607280260432,
+      "grad_norm": 1.7857599258422852,
+      "learning_rate": 7.143219790808347e-06,
+      "loss": 0.2123,
+      "step": 3828
+    },
+    {
+      "epoch": 1.8130807931340633,
+      "grad_norm": 1.2344205379486084,
+      "learning_rate": 7.138319189755002e-06,
+      "loss": 0.2103,
+      "step": 3829
+    },
+    {
+      "epoch": 1.8135543060076946,
+      "grad_norm": 2.3192245960235596,
+      "learning_rate": 7.1334193371466675e-06,
+      "loss": 0.231,
+      "step": 3830
+    },
+    {
+      "epoch": 1.8140278188813257,
+      "grad_norm": 1.136781096458435,
+      "learning_rate": 7.128520234264851e-06,
+      "loss": 0.2119,
+      "step": 3831
+    },
+    {
+      "epoch": 1.8145013317549572,
+      "grad_norm": 1.3078882694244385,
+      "learning_rate": 7.1236218823908645e-06,
+      "loss": 0.2428,
+      "step": 3832
+    },
+    {
+      "epoch": 1.8149748446285883,
+      "grad_norm": 1.2478915452957153,
+      "learning_rate": 7.118724282805825e-06,
+      "loss": 0.2439,
+      "step": 3833
+    },
+    {
+      "epoch": 1.8154483575022196,
+      "grad_norm": 1.50113844871521,
+      "learning_rate": 7.113827436790655e-06,
+      "loss": 0.2277,
+      "step": 3834
+    },
+    {
+      "epoch": 1.8159218703758508,
+      "grad_norm": 1.2147942781448364,
+      "learning_rate": 7.108931345626074e-06,
+      "loss": 0.2285,
+      "step": 3835
+    },
+    {
+      "epoch": 1.8163953832494821,
+      "grad_norm": 1.1065564155578613,
+      "learning_rate": 7.104036010592609e-06,
+      "loss": 0.1995,
+      "step": 3836
+    },
+    {
+      "epoch": 1.8168688961231134,
+      "grad_norm": 1.3819801807403564,
+      "learning_rate": 7.099141432970588e-06,
+      "loss": 0.1979,
+      "step": 3837
+    },
+    {
+      "epoch": 1.8173424089967445,
+      "grad_norm": 0.9944810271263123,
+      "learning_rate": 7.094247614040139e-06,
+      "loss": 0.2053,
+      "step": 3838
+    },
+    {
+      "epoch": 1.817815921870376,
+      "grad_norm": 1.5402950048446655,
+      "learning_rate": 7.0893545550811956e-06,
+      "loss": 0.2208,
+      "step": 3839
+    },
+    {
+      "epoch": 1.818289434744007,
+      "grad_norm": 1.463292121887207,
+      "learning_rate": 7.08446225737349e-06,
+      "loss": 0.2016,
+      "step": 3840
+    },
+    {
+      "epoch": 1.8187629476176383,
+      "grad_norm": 1.0113959312438965,
+      "learning_rate": 7.079570722196553e-06,
+      "loss": 0.2195,
+      "step": 3841
+    },
+    {
+      "epoch": 1.8192364604912696,
+      "grad_norm": 1.070245623588562,
+      "learning_rate": 7.074679950829719e-06,
+      "loss": 0.2084,
+      "step": 3842
+    },
+    {
+      "epoch": 1.8197099733649007,
+      "grad_norm": 1.2320399284362793,
+      "learning_rate": 7.069789944552124e-06,
+      "loss": 0.206,
+      "step": 3843
+    },
+    {
+      "epoch": 1.8201834862385322,
+      "grad_norm": 1.2585248947143555,
+      "learning_rate": 7.0649007046427006e-06,
+      "loss": 0.2279,
+      "step": 3844
+    },
+    {
+      "epoch": 1.8206569991121633,
+      "grad_norm": 0.9737508296966553,
+      "learning_rate": 7.060012232380182e-06,
+      "loss": 0.1996,
+      "step": 3845
+    },
+    {
+      "epoch": 1.8211305119857946,
+      "grad_norm": 1.2638822793960571,
+      "learning_rate": 7.0551245290431e-06,
+      "loss": 0.1924,
+      "step": 3846
+    },
+    {
+      "epoch": 1.8216040248594259,
+      "grad_norm": 1.16457200050354,
+      "learning_rate": 7.0502375959097904e-06,
+      "loss": 0.2235,
+      "step": 3847
+    },
+    {
+      "epoch": 1.8220775377330571,
+      "grad_norm": 1.078177809715271,
+      "learning_rate": 7.045351434258378e-06,
+      "loss": 0.2087,
+      "step": 3848
+    },
+    {
+      "epoch": 1.8225510506066884,
+      "grad_norm": 1.1603608131408691,
+      "learning_rate": 7.040466045366796e-06,
+      "loss": 0.2128,
+      "step": 3849
+    },
+    {
+      "epoch": 1.8230245634803195,
+      "grad_norm": 1.2323132753372192,
+      "learning_rate": 7.03558143051277e-06,
+      "loss": 0.2041,
+      "step": 3850
+    },
+    {
+      "epoch": 1.823498076353951,
+      "grad_norm": 1.5576890707015991,
+      "learning_rate": 7.0306975909738205e-06,
+      "loss": 0.2194,
+      "step": 3851
+    },
+    {
+      "epoch": 1.823971589227582,
+      "grad_norm": 0.993789553642273,
+      "learning_rate": 7.025814528027272e-06,
+      "loss": 0.1883,
+      "step": 3852
+    },
+    {
+      "epoch": 1.8244451021012134,
+      "grad_norm": 1.0642006397247314,
+      "learning_rate": 7.02093224295024e-06,
+      "loss": 0.1956,
+      "step": 3853
+    },
+    {
+      "epoch": 1.8249186149748446,
+      "grad_norm": 1.1008495092391968,
+      "learning_rate": 7.016050737019641e-06,
+      "loss": 0.2071,
+      "step": 3854
+    },
+    {
+      "epoch": 1.825392127848476,
+      "grad_norm": 1.0528384447097778,
+      "learning_rate": 7.0111700115121835e-06,
+      "loss": 0.2297,
+      "step": 3855
+    },
+    {
+      "epoch": 1.8258656407221072,
+      "grad_norm": 1.6767280101776123,
+      "learning_rate": 7.006290067704378e-06,
+      "loss": 0.2402,
+      "step": 3856
+    },
+    {
+      "epoch": 1.8263391535957383,
+      "grad_norm": 0.9541283249855042,
+      "learning_rate": 7.001410906872522e-06,
+      "loss": 0.1919,
+      "step": 3857
+    },
+    {
+      "epoch": 1.8268126664693698,
+      "grad_norm": 0.9566398859024048,
+      "learning_rate": 6.996532530292717e-06,
+      "loss": 0.1902,
+      "step": 3858
+    },
+    {
+      "epoch": 1.8272861793430009,
+      "grad_norm": 0.9043372869491577,
+      "learning_rate": 6.991654939240855e-06,
+      "loss": 0.2088,
+      "step": 3859
+    },
+    {
+      "epoch": 1.8277596922166321,
+      "grad_norm": 1.314922571182251,
+      "learning_rate": 6.98677813499262e-06,
+      "loss": 0.2167,
+      "step": 3860
+    },
+    {
+      "epoch": 1.8282332050902634,
+      "grad_norm": 1.1723600625991821,
+      "learning_rate": 6.981902118823495e-06,
+      "loss": 0.2107,
+      "step": 3861
+    },
+    {
+      "epoch": 1.8287067179638945,
+      "grad_norm": 1.5371065139770508,
+      "learning_rate": 6.977026892008753e-06,
+      "loss": 0.2167,
+      "step": 3862
+    },
+    {
+      "epoch": 1.829180230837526,
+      "grad_norm": 1.1447199583053589,
+      "learning_rate": 6.972152455823467e-06,
+      "loss": 0.2027,
+      "step": 3863
+    },
+    {
+      "epoch": 1.829653743711157,
+      "grad_norm": 1.6060473918914795,
+      "learning_rate": 6.967278811542495e-06,
+      "loss": 0.2164,
+      "step": 3864
+    },
+    {
+      "epoch": 1.8301272565847884,
+      "grad_norm": 1.6065421104431152,
+      "learning_rate": 6.96240596044049e-06,
+      "loss": 0.2087,
+      "step": 3865
+    },
+    {
+      "epoch": 1.8306007694584197,
+      "grad_norm": 1.0932152271270752,
+      "learning_rate": 6.957533903791904e-06,
+      "loss": 0.2294,
+      "step": 3866
+    },
+    {
+      "epoch": 1.831074282332051,
+      "grad_norm": 1.129197359085083,
+      "learning_rate": 6.9526626428709745e-06,
+      "loss": 0.1843,
+      "step": 3867
+    },
+    {
+      "epoch": 1.8315477952056822,
+      "grad_norm": 2.516554117202759,
+      "learning_rate": 6.947792178951733e-06,
+      "loss": 0.2076,
+      "step": 3868
+    },
+    {
+      "epoch": 1.8320213080793133,
+      "grad_norm": 1.335574984550476,
+      "learning_rate": 6.942922513308001e-06,
+      "loss": 0.2114,
+      "step": 3869
+    },
+    {
+      "epoch": 1.8324948209529448,
+      "grad_norm": 0.9613366723060608,
+      "learning_rate": 6.9380536472133945e-06,
+      "loss": 0.2074,
+      "step": 3870
+    },
+    {
+      "epoch": 1.8329683338265759,
+      "grad_norm": 1.1925537586212158,
+      "learning_rate": 6.933185581941316e-06,
+      "loss": 0.2034,
+      "step": 3871
+    },
+    {
+      "epoch": 1.8334418467002072,
+      "grad_norm": 1.1290981769561768,
+      "learning_rate": 6.928318318764964e-06,
+      "loss": 0.1972,
+      "step": 3872
+    },
+    {
+      "epoch": 1.8339153595738384,
+      "grad_norm": 1.4056495428085327,
+      "learning_rate": 6.923451858957322e-06,
+      "loss": 0.2269,
+      "step": 3873
+    },
+    {
+      "epoch": 1.8343888724474695,
+      "grad_norm": 1.2841758728027344,
+      "learning_rate": 6.918586203791169e-06,
+      "loss": 0.1974,
+      "step": 3874
+    },
+    {
+      "epoch": 1.834862385321101,
+      "grad_norm": 0.927555501461029,
+      "learning_rate": 6.913721354539065e-06,
+      "loss": 0.2272,
+      "step": 3875
+    },
+    {
+      "epoch": 1.835335898194732,
+      "grad_norm": 1.3648067712783813,
+      "learning_rate": 6.908857312473366e-06,
+      "loss": 0.2267,
+      "step": 3876
+    },
+    {
+      "epoch": 1.8358094110683634,
+      "grad_norm": 1.9164044857025146,
+      "learning_rate": 6.903994078866216e-06,
+      "loss": 0.226,
+      "step": 3877
+    },
+    {
+      "epoch": 1.8362829239419947,
+      "grad_norm": 1.0202974081039429,
+      "learning_rate": 6.899131654989548e-06,
+      "loss": 0.205,
+      "step": 3878
+    },
+    {
+      "epoch": 1.836756436815626,
+      "grad_norm": 1.1604101657867432,
+      "learning_rate": 6.894270042115081e-06,
+      "loss": 0.1836,
+      "step": 3879
+    },
+    {
+      "epoch": 1.8372299496892572,
+      "grad_norm": 1.1902738809585571,
+      "learning_rate": 6.889409241514323e-06,
+      "loss": 0.2165,
+      "step": 3880
+    },
+    {
+      "epoch": 1.8377034625628883,
+      "grad_norm": 1.178782343864441,
+      "learning_rate": 6.88454925445857e-06,
+      "loss": 0.1952,
+      "step": 3881
+    },
+    {
+      "epoch": 1.8381769754365198,
+      "grad_norm": 1.345608115196228,
+      "learning_rate": 6.879690082218903e-06,
+      "loss": 0.2005,
+      "step": 3882
+    },
+    {
+      "epoch": 1.8386504883101509,
+      "grad_norm": 1.6030305624008179,
+      "learning_rate": 6.874831726066194e-06,
+      "loss": 0.2087,
+      "step": 3883
+    },
+    {
+      "epoch": 1.8391240011837822,
+      "grad_norm": 1.1542789936065674,
+      "learning_rate": 6.869974187271098e-06,
+      "loss": 0.1984,
+      "step": 3884
+    },
+    {
+      "epoch": 1.8395975140574135,
+      "grad_norm": 1.1518371105194092,
+      "learning_rate": 6.865117467104058e-06,
+      "loss": 0.2063,
+      "step": 3885
+    },
+    {
+      "epoch": 1.8400710269310447,
+      "grad_norm": 1.9634912014007568,
+      "learning_rate": 6.8602615668353e-06,
+      "loss": 0.2074,
+      "step": 3886
+    },
+    {
+      "epoch": 1.840544539804676,
+      "grad_norm": 1.0860075950622559,
+      "learning_rate": 6.85540648773484e-06,
+      "loss": 0.202,
+      "step": 3887
+    },
+    {
+      "epoch": 1.841018052678307,
+      "grad_norm": 1.4793963432312012,
+      "learning_rate": 6.850552231072477e-06,
+      "loss": 0.229,
+      "step": 3888
+    },
+    {
+      "epoch": 1.8414915655519386,
+      "grad_norm": 1.0886497497558594,
+      "learning_rate": 6.845698798117795e-06,
+      "loss": 0.234,
+      "step": 3889
+    },
+    {
+      "epoch": 1.8419650784255697,
+      "grad_norm": 2.2416272163391113,
+      "learning_rate": 6.840846190140161e-06,
+      "loss": 0.2166,
+      "step": 3890
+    },
+    {
+      "epoch": 1.842438591299201,
+      "grad_norm": 1.1083406209945679,
+      "learning_rate": 6.83599440840873e-06,
+      "loss": 0.2016,
+      "step": 3891
+    },
+    {
+      "epoch": 1.8429121041728322,
+      "grad_norm": 1.5353022813796997,
+      "learning_rate": 6.831143454192437e-06,
+      "loss": 0.1789,
+      "step": 3892
+    },
+    {
+      "epoch": 1.8433856170464633,
+      "grad_norm": 1.475759744644165,
+      "learning_rate": 6.826293328760004e-06,
+      "loss": 0.2397,
+      "step": 3893
+    },
+    {
+      "epoch": 1.8438591299200948,
+      "grad_norm": 1.3669559955596924,
+      "learning_rate": 6.821444033379936e-06,
+      "loss": 0.2096,
+      "step": 3894
+    },
+    {
+      "epoch": 1.8443326427937259,
+      "grad_norm": 0.972876787185669,
+      "learning_rate": 6.816595569320514e-06,
+      "loss": 0.2265,
+      "step": 3895
+    },
+    {
+      "epoch": 1.8448061556673572,
+      "grad_norm": 1.7018649578094482,
+      "learning_rate": 6.8117479378498104e-06,
+      "loss": 0.2174,
+      "step": 3896
+    },
+    {
+      "epoch": 1.8452796685409885,
+      "grad_norm": 1.9252698421478271,
+      "learning_rate": 6.806901140235675e-06,
+      "loss": 0.2152,
+      "step": 3897
+    },
+    {
+      "epoch": 1.8457531814146197,
+      "grad_norm": 1.056630253791809,
+      "learning_rate": 6.802055177745743e-06,
+      "loss": 0.2357,
+      "step": 3898
+    },
+    {
+      "epoch": 1.846226694288251,
+      "grad_norm": 1.101554036140442,
+      "learning_rate": 6.79721005164743e-06,
+      "loss": 0.213,
+      "step": 3899
+    },
+    {
+      "epoch": 1.846700207161882,
+      "grad_norm": 1.172674536705017,
+      "learning_rate": 6.792365763207926e-06,
+      "loss": 0.1947,
+      "step": 3900
+    },
+    {
+      "epoch": 1.8471737200355136,
+      "grad_norm": 2.0155622959136963,
+      "learning_rate": 6.787522313694214e-06,
+      "loss": 0.2403,
+      "step": 3901
+    },
+    {
+      "epoch": 1.8476472329091447,
+      "grad_norm": 1.046518325805664,
+      "learning_rate": 6.782679704373051e-06,
+      "loss": 0.2045,
+      "step": 3902
+    },
+    {
+      "epoch": 1.848120745782776,
+      "grad_norm": 1.204442024230957,
+      "learning_rate": 6.777837936510971e-06,
+      "loss": 0.2257,
+      "step": 3903
+    },
+    {
+      "epoch": 1.8485942586564073,
+      "grad_norm": 1.0126386880874634,
+      "learning_rate": 6.772997011374294e-06,
+      "loss": 0.194,
+      "step": 3904
+    },
+    {
+      "epoch": 1.8490677715300383,
+      "grad_norm": 1.27089524269104,
+      "learning_rate": 6.768156930229118e-06,
+      "loss": 0.1956,
+      "step": 3905
+    },
+    {
+      "epoch": 1.8495412844036698,
+      "grad_norm": 1.372044563293457,
+      "learning_rate": 6.763317694341319e-06,
+      "loss": 0.1892,
+      "step": 3906
+    },
+    {
+      "epoch": 1.850014797277301,
+      "grad_norm": 1.0626308917999268,
+      "learning_rate": 6.758479304976553e-06,
+      "loss": 0.2172,
+      "step": 3907
+    },
+    {
+      "epoch": 1.8504883101509322,
+      "grad_norm": 0.8939222693443298,
+      "learning_rate": 6.753641763400252e-06,
+      "loss": 0.2101,
+      "step": 3908
+    },
+    {
+      "epoch": 1.8509618230245635,
+      "grad_norm": 1.0738952159881592,
+      "learning_rate": 6.748805070877632e-06,
+      "loss": 0.195,
+      "step": 3909
+    },
+    {
+      "epoch": 1.8514353358981948,
+      "grad_norm": 2.0406644344329834,
+      "learning_rate": 6.743969228673679e-06,
+      "loss": 0.2011,
+      "step": 3910
+    },
+    {
+      "epoch": 1.851908848771826,
+      "grad_norm": 1.1621339321136475,
+      "learning_rate": 6.739134238053162e-06,
+      "loss": 0.2406,
+      "step": 3911
+    },
+    {
+      "epoch": 1.852382361645457,
+      "grad_norm": 1.1607718467712402,
+      "learning_rate": 6.734300100280629e-06,
+      "loss": 0.2154,
+      "step": 3912
+    },
+    {
+      "epoch": 1.8528558745190886,
+      "grad_norm": 1.2867686748504639,
+      "learning_rate": 6.729466816620398e-06,
+      "loss": 0.2091,
+      "step": 3913
+    },
+    {
+      "epoch": 1.8533293873927197,
+      "grad_norm": 1.1185880899429321,
+      "learning_rate": 6.724634388336571e-06,
+      "loss": 0.2298,
+      "step": 3914
+    },
+    {
+      "epoch": 1.853802900266351,
+      "grad_norm": 1.1135728359222412,
+      "learning_rate": 6.71980281669302e-06,
+      "loss": 0.2092,
+      "step": 3915
+    },
+    {
+      "epoch": 1.8542764131399823,
+      "grad_norm": 0.8761497735977173,
+      "learning_rate": 6.714972102953399e-06,
+      "loss": 0.1895,
+      "step": 3916
+    },
+    {
+      "epoch": 1.8547499260136135,
+      "grad_norm": 1.1779834032058716,
+      "learning_rate": 6.710142248381133e-06,
+      "loss": 0.2315,
+      "step": 3917
+    },
+    {
+      "epoch": 1.8552234388872448,
+      "grad_norm": 0.9498329758644104,
+      "learning_rate": 6.705313254239424e-06,
+      "loss": 0.2117,
+      "step": 3918
+    },
+    {
+      "epoch": 1.855696951760876,
+      "grad_norm": 1.2298184633255005,
+      "learning_rate": 6.700485121791252e-06,
+      "loss": 0.2018,
+      "step": 3919
+    },
+    {
+      "epoch": 1.8561704646345074,
+      "grad_norm": 1.0038814544677734,
+      "learning_rate": 6.695657852299362e-06,
+      "loss": 0.1963,
+      "step": 3920
+    },
+    {
+      "epoch": 1.8566439775081385,
+      "grad_norm": 1.4736988544464111,
+      "learning_rate": 6.690831447026283e-06,
+      "loss": 0.2308,
+      "step": 3921
+    },
+    {
+      "epoch": 1.8571174903817698,
+      "grad_norm": 1.168468952178955,
+      "learning_rate": 6.686005907234317e-06,
+      "loss": 0.2056,
+      "step": 3922
+    },
+    {
+      "epoch": 1.857591003255401,
+      "grad_norm": 1.3906092643737793,
+      "learning_rate": 6.681181234185532e-06,
+      "loss": 0.2446,
+      "step": 3923
+    },
+    {
+      "epoch": 1.8580645161290321,
+      "grad_norm": 1.0936076641082764,
+      "learning_rate": 6.6763574291417795e-06,
+      "loss": 0.1935,
+      "step": 3924
+    },
+    {
+      "epoch": 1.8585380290026636,
+      "grad_norm": 1.0105923414230347,
+      "learning_rate": 6.67153449336468e-06,
+      "loss": 0.2192,
+      "step": 3925
+    },
+    {
+      "epoch": 1.8590115418762947,
+      "grad_norm": 1.5099976062774658,
+      "learning_rate": 6.666712428115621e-06,
+      "loss": 0.199,
+      "step": 3926
+    },
+    {
+      "epoch": 1.859485054749926,
+      "grad_norm": 1.0132044553756714,
+      "learning_rate": 6.661891234655769e-06,
+      "loss": 0.2101,
+      "step": 3927
+    },
+    {
+      "epoch": 1.8599585676235573,
+      "grad_norm": 0.927442193031311,
+      "learning_rate": 6.657070914246063e-06,
+      "loss": 0.1969,
+      "step": 3928
+    },
+    {
+      "epoch": 1.8604320804971886,
+      "grad_norm": 1.5215531587600708,
+      "learning_rate": 6.6522514681472105e-06,
+      "loss": 0.1945,
+      "step": 3929
+    },
+    {
+      "epoch": 1.8609055933708198,
+      "grad_norm": 1.445884108543396,
+      "learning_rate": 6.64743289761969e-06,
+      "loss": 0.2198,
+      "step": 3930
+    },
+    {
+      "epoch": 1.861379106244451,
+      "grad_norm": 1.817511796951294,
+      "learning_rate": 6.64261520392375e-06,
+      "loss": 0.2005,
+      "step": 3931
+    },
+    {
+      "epoch": 1.8618526191180824,
+      "grad_norm": 1.0783874988555908,
+      "learning_rate": 6.637798388319416e-06,
+      "loss": 0.1863,
+      "step": 3932
+    },
+    {
+      "epoch": 1.8623261319917135,
+      "grad_norm": 1.2979896068572998,
+      "learning_rate": 6.632982452066476e-06,
+      "loss": 0.2191,
+      "step": 3933
+    },
+    {
+      "epoch": 1.8627996448653448,
+      "grad_norm": 0.9577775001525879,
+      "learning_rate": 6.628167396424494e-06,
+      "loss": 0.1862,
+      "step": 3934
+    },
+    {
+      "epoch": 1.863273157738976,
+      "grad_norm": 1.0281299352645874,
+      "learning_rate": 6.623353222652802e-06,
+      "loss": 0.2308,
+      "step": 3935
+    },
+    {
+      "epoch": 1.8637466706126071,
+      "grad_norm": 1.4038194417953491,
+      "learning_rate": 6.6185399320105e-06,
+      "loss": 0.2115,
+      "step": 3936
+    },
+    {
+      "epoch": 1.8642201834862386,
+      "grad_norm": 2.125293254852295,
+      "learning_rate": 6.613727525756455e-06,
+      "loss": 0.2016,
+      "step": 3937
+    },
+    {
+      "epoch": 1.8646936963598697,
+      "grad_norm": 1.3290354013442993,
+      "learning_rate": 6.608916005149311e-06,
+      "loss": 0.1913,
+      "step": 3938
+    },
+    {
+      "epoch": 1.865167209233501,
+      "grad_norm": 1.1084948778152466,
+      "learning_rate": 6.604105371447469e-06,
+      "loss": 0.1794,
+      "step": 3939
+    },
+    {
+      "epoch": 1.8656407221071323,
+      "grad_norm": 1.267200231552124,
+      "learning_rate": 6.599295625909107e-06,
+      "loss": 0.1885,
+      "step": 3940
+    },
+    {
+      "epoch": 1.8661142349807636,
+      "grad_norm": 1.9131397008895874,
+      "learning_rate": 6.5944867697921654e-06,
+      "loss": 0.224,
+      "step": 3941
+    },
+    {
+      "epoch": 1.8665877478543949,
+      "grad_norm": 1.2804118394851685,
+      "learning_rate": 6.589678804354353e-06,
+      "loss": 0.1945,
+      "step": 3942
+    },
+    {
+      "epoch": 1.867061260728026,
+      "grad_norm": 0.9017341732978821,
+      "learning_rate": 6.584871730853153e-06,
+      "loss": 0.2059,
+      "step": 3943
+    },
+    {
+      "epoch": 1.8675347736016574,
+      "grad_norm": 1.365110158920288,
+      "learning_rate": 6.5800655505458065e-06,
+      "loss": 0.222,
+      "step": 3944
+    },
+    {
+      "epoch": 1.8680082864752885,
+      "grad_norm": 0.8447331190109253,
+      "learning_rate": 6.5752602646893224e-06,
+      "loss": 0.1945,
+      "step": 3945
+    },
+    {
+      "epoch": 1.8684817993489198,
+      "grad_norm": 1.1045807600021362,
+      "learning_rate": 6.5704558745404755e-06,
+      "loss": 0.1969,
+      "step": 3946
+    },
+    {
+      "epoch": 1.868955312222551,
+      "grad_norm": 1.3846262693405151,
+      "learning_rate": 6.56565238135581e-06,
+      "loss": 0.2017,
+      "step": 3947
+    },
+    {
+      "epoch": 1.8694288250961824,
+      "grad_norm": 1.107153058052063,
+      "learning_rate": 6.560849786391632e-06,
+      "loss": 0.2207,
+      "step": 3948
+    },
+    {
+      "epoch": 1.8699023379698136,
+      "grad_norm": 1.206408143043518,
+      "learning_rate": 6.556048090904015e-06,
+      "loss": 0.2088,
+      "step": 3949
+    },
+    {
+      "epoch": 1.8703758508434447,
+      "grad_norm": 1.063124418258667,
+      "learning_rate": 6.5512472961487946e-06,
+      "loss": 0.2066,
+      "step": 3950
+    },
+    {
+      "epoch": 1.8708493637170762,
+      "grad_norm": 1.2000980377197266,
+      "learning_rate": 6.5464474033815754e-06,
+      "loss": 0.1757,
+      "step": 3951
+    },
+    {
+      "epoch": 1.8713228765907073,
+      "grad_norm": 1.3470823764801025,
+      "learning_rate": 6.541648413857718e-06,
+      "loss": 0.2252,
+      "step": 3952
+    },
+    {
+      "epoch": 1.8717963894643386,
+      "grad_norm": 1.3533769845962524,
+      "learning_rate": 6.536850328832358e-06,
+      "loss": 0.2145,
+      "step": 3953
+    },
+    {
+      "epoch": 1.8722699023379699,
+      "grad_norm": 1.6405057907104492,
+      "learning_rate": 6.5320531495603825e-06,
+      "loss": 0.1892,
+      "step": 3954
+    },
+    {
+      "epoch": 1.872743415211601,
+      "grad_norm": 1.296522855758667,
+      "learning_rate": 6.527256877296449e-06,
+      "loss": 0.2205,
+      "step": 3955
+    },
+    {
+      "epoch": 1.8732169280852324,
+      "grad_norm": 1.0236841440200806,
+      "learning_rate": 6.522461513294979e-06,
+      "loss": 0.1865,
+      "step": 3956
+    },
+    {
+      "epoch": 1.8736904409588635,
+      "grad_norm": 1.0961860418319702,
+      "learning_rate": 6.517667058810151e-06,
+      "loss": 0.1896,
+      "step": 3957
+    },
+    {
+      "epoch": 1.8741639538324948,
+      "grad_norm": 1.3094490766525269,
+      "learning_rate": 6.5128735150959075e-06,
+      "loss": 0.2059,
+      "step": 3958
+    },
+    {
+      "epoch": 1.874637466706126,
+      "grad_norm": 1.5833834409713745,
+      "learning_rate": 6.508080883405957e-06,
+      "loss": 0.223,
+      "step": 3959
+    },
+    {
+      "epoch": 1.8751109795797574,
+      "grad_norm": 0.9895943999290466,
+      "learning_rate": 6.503289164993765e-06,
+      "loss": 0.2081,
+      "step": 3960
+    },
+    {
+      "epoch": 1.8755844924533887,
+      "grad_norm": 1.386003851890564,
+      "learning_rate": 6.498498361112557e-06,
+      "loss": 0.2138,
+      "step": 3961
+    },
+    {
+      "epoch": 1.8760580053270197,
+      "grad_norm": 1.1875851154327393,
+      "learning_rate": 6.4937084730153236e-06,
+      "loss": 0.2184,
+      "step": 3962
+    },
+    {
+      "epoch": 1.8765315182006512,
+      "grad_norm": 1.4439812898635864,
+      "learning_rate": 6.4889195019548155e-06,
+      "loss": 0.2191,
+      "step": 3963
+    },
+    {
+      "epoch": 1.8770050310742823,
+      "grad_norm": 1.2867056131362915,
+      "learning_rate": 6.484131449183537e-06,
+      "loss": 0.195,
+      "step": 3964
+    },
+    {
+      "epoch": 1.8774785439479136,
+      "grad_norm": 1.1023000478744507,
+      "learning_rate": 6.47934431595376e-06,
+      "loss": 0.2226,
+      "step": 3965
+    },
+    {
+      "epoch": 1.8779520568215449,
+      "grad_norm": 2.235792636871338,
+      "learning_rate": 6.474558103517513e-06,
+      "loss": 0.2143,
+      "step": 3966
+    },
+    {
+      "epoch": 1.878425569695176,
+      "grad_norm": 1.0556797981262207,
+      "learning_rate": 6.469772813126584e-06,
+      "loss": 0.1981,
+      "step": 3967
+    },
+    {
+      "epoch": 1.8788990825688074,
+      "grad_norm": 0.9840413928031921,
+      "learning_rate": 6.464988446032518e-06,
+      "loss": 0.2003,
+      "step": 3968
+    },
+    {
+      "epoch": 1.8793725954424385,
+      "grad_norm": 1.3949209451675415,
+      "learning_rate": 6.460205003486621e-06,
+      "loss": 0.2189,
+      "step": 3969
+    },
+    {
+      "epoch": 1.8798461083160698,
+      "grad_norm": 0.8188128471374512,
+      "learning_rate": 6.4554224867399575e-06,
+      "loss": 0.1956,
+      "step": 3970
+    },
+    {
+      "epoch": 1.880319621189701,
+      "grad_norm": 1.0726219415664673,
+      "learning_rate": 6.450640897043346e-06,
+      "loss": 0.2257,
+      "step": 3971
+    },
+    {
+      "epoch": 1.8807931340633324,
+      "grad_norm": 1.4881572723388672,
+      "learning_rate": 6.445860235647367e-06,
+      "loss": 0.1915,
+      "step": 3972
+    },
+    {
+      "epoch": 1.8812666469369637,
+      "grad_norm": 1.1142457723617554,
+      "learning_rate": 6.44108050380236e-06,
+      "loss": 0.1911,
+      "step": 3973
+    },
+    {
+      "epoch": 1.8817401598105947,
+      "grad_norm": 1.912439227104187,
+      "learning_rate": 6.43630170275841e-06,
+      "loss": 0.1943,
+      "step": 3974
+    },
+    {
+      "epoch": 1.8822136726842262,
+      "grad_norm": 1.1108665466308594,
+      "learning_rate": 6.431523833765369e-06,
+      "loss": 0.2103,
+      "step": 3975
+    },
+    {
+      "epoch": 1.8826871855578573,
+      "grad_norm": 1.0399905443191528,
+      "learning_rate": 6.426746898072845e-06,
+      "loss": 0.2138,
+      "step": 3976
+    },
+    {
+      "epoch": 1.8831606984314886,
+      "grad_norm": 0.9671503305435181,
+      "learning_rate": 6.421970896930199e-06,
+      "loss": 0.193,
+      "step": 3977
+    },
+    {
+      "epoch": 1.8836342113051199,
+      "grad_norm": 1.4614564180374146,
+      "learning_rate": 6.417195831586545e-06,
+      "loss": 0.2061,
+      "step": 3978
+    },
+    {
+      "epoch": 1.8841077241787512,
+      "grad_norm": 1.1578701734542847,
+      "learning_rate": 6.412421703290759e-06,
+      "loss": 0.2123,
+      "step": 3979
+    },
+    {
+      "epoch": 1.8845812370523825,
+      "grad_norm": 1.1793917417526245,
+      "learning_rate": 6.4076485132914644e-06,
+      "loss": 0.2134,
+      "step": 3980
+    },
+    {
+      "epoch": 1.8850547499260135,
+      "grad_norm": 1.1729707717895508,
+      "learning_rate": 6.402876262837045e-06,
+      "loss": 0.217,
+      "step": 3981
+    },
+    {
+      "epoch": 1.885528262799645,
+      "grad_norm": 1.058951735496521,
+      "learning_rate": 6.398104953175639e-06,
+      "loss": 0.2041,
+      "step": 3982
+    },
+    {
+      "epoch": 1.886001775673276,
+      "grad_norm": 1.7468572854995728,
+      "learning_rate": 6.393334585555133e-06,
+      "loss": 0.215,
+      "step": 3983
+    },
+    {
+      "epoch": 1.8864752885469074,
+      "grad_norm": 0.8419433832168579,
+      "learning_rate": 6.388565161223172e-06,
+      "loss": 0.1718,
+      "step": 3984
+    },
+    {
+      "epoch": 1.8869488014205387,
+      "grad_norm": 1.8909525871276855,
+      "learning_rate": 6.383796681427154e-06,
+      "loss": 0.1912,
+      "step": 3985
+    },
+    {
+      "epoch": 1.8874223142941697,
+      "grad_norm": 1.1229642629623413,
+      "learning_rate": 6.379029147414228e-06,
+      "loss": 0.2449,
+      "step": 3986
+    },
+    {
+      "epoch": 1.8878958271678012,
+      "grad_norm": 1.213046669960022,
+      "learning_rate": 6.374262560431297e-06,
+      "loss": 0.1952,
+      "step": 3987
+    },
+    {
+      "epoch": 1.8883693400414323,
+      "grad_norm": 1.6721361875534058,
+      "learning_rate": 6.369496921725016e-06,
+      "loss": 0.2018,
+      "step": 3988
+    },
+    {
+      "epoch": 1.8888428529150636,
+      "grad_norm": 1.6675399541854858,
+      "learning_rate": 6.364732232541788e-06,
+      "loss": 0.2019,
+      "step": 3989
+    },
+    {
+      "epoch": 1.8893163657886949,
+      "grad_norm": 1.043617606163025,
+      "learning_rate": 6.359968494127777e-06,
+      "loss": 0.1982,
+      "step": 3990
+    },
+    {
+      "epoch": 1.8897898786623262,
+      "grad_norm": 0.9541271924972534,
+      "learning_rate": 6.355205707728889e-06,
+      "loss": 0.2128,
+      "step": 3991
+    },
+    {
+      "epoch": 1.8902633915359575,
+      "grad_norm": 1.9646912813186646,
+      "learning_rate": 6.350443874590786e-06,
+      "loss": 0.1915,
+      "step": 3992
+    },
+    {
+      "epoch": 1.8907369044095885,
+      "grad_norm": 1.3205372095108032,
+      "learning_rate": 6.34568299595888e-06,
+      "loss": 0.1902,
+      "step": 3993
+    },
+    {
+      "epoch": 1.89121041728322,
+      "grad_norm": 2.0450875759124756,
+      "learning_rate": 6.340923073078333e-06,
+      "loss": 0.208,
+      "step": 3994
+    },
+    {
+      "epoch": 1.891683930156851,
+      "grad_norm": 1.3528573513031006,
+      "learning_rate": 6.336164107194056e-06,
+      "loss": 0.2167,
+      "step": 3995
+    },
+    {
+      "epoch": 1.8921574430304824,
+      "grad_norm": 1.0544376373291016,
+      "learning_rate": 6.331406099550711e-06,
+      "loss": 0.213,
+      "step": 3996
+    },
+    {
+      "epoch": 1.8926309559041137,
+      "grad_norm": 1.3501182794570923,
+      "learning_rate": 6.326649051392709e-06,
+      "loss": 0.1947,
+      "step": 3997
+    },
+    {
+      "epoch": 1.8931044687777447,
+      "grad_norm": 1.4328538179397583,
+      "learning_rate": 6.321892963964214e-06,
+      "loss": 0.1958,
+      "step": 3998
+    },
+    {
+      "epoch": 1.8935779816513763,
+      "grad_norm": 1.6328697204589844,
+      "learning_rate": 6.317137838509126e-06,
+      "loss": 0.2082,
+      "step": 3999
+    },
+    {
+      "epoch": 1.8940514945250073,
+      "grad_norm": 2.273808240890503,
+      "learning_rate": 6.3123836762711085e-06,
+      "loss": 0.2325,
+      "step": 4000
+    },
+    {
+      "epoch": 1.8945250073986386,
+      "grad_norm": 1.4385737180709839,
+      "learning_rate": 6.307630478493565e-06,
+      "loss": 0.2211,
+      "step": 4001
+    },
+    {
+      "epoch": 1.89499852027227,
+      "grad_norm": 1.7842854261398315,
+      "learning_rate": 6.30287824641965e-06,
+      "loss": 0.2317,
+      "step": 4002
+    },
+    {
+      "epoch": 1.8954720331459012,
+      "grad_norm": 1.4352833032608032,
+      "learning_rate": 6.298126981292263e-06,
+      "loss": 0.2032,
+      "step": 4003
+    },
+    {
+      "epoch": 1.8959455460195325,
+      "grad_norm": 1.0405994653701782,
+      "learning_rate": 6.29337668435405e-06,
+      "loss": 0.213,
+      "step": 4004
+    },
+    {
+      "epoch": 1.8964190588931635,
+      "grad_norm": 1.2680540084838867,
+      "learning_rate": 6.288627356847407e-06,
+      "loss": 0.2208,
+      "step": 4005
+    },
+    {
+      "epoch": 1.896892571766795,
+      "grad_norm": 0.9765031337738037,
+      "learning_rate": 6.283879000014476e-06,
+      "loss": 0.1896,
+      "step": 4006
+    },
+    {
+      "epoch": 1.897366084640426,
+      "grad_norm": 1.2945647239685059,
+      "learning_rate": 6.27913161509714e-06,
+      "loss": 0.2107,
+      "step": 4007
+    },
+    {
+      "epoch": 1.8978395975140574,
+      "grad_norm": 1.2577142715454102,
+      "learning_rate": 6.274385203337039e-06,
+      "loss": 0.2156,
+      "step": 4008
+    },
+    {
+      "epoch": 1.8983131103876887,
+      "grad_norm": 2.5358786582946777,
+      "learning_rate": 6.269639765975543e-06,
+      "loss": 0.2061,
+      "step": 4009
+    },
+    {
+      "epoch": 1.89878662326132,
+      "grad_norm": 0.9788986444473267,
+      "learning_rate": 6.264895304253779e-06,
+      "loss": 0.1898,
+      "step": 4010
+    },
+    {
+      "epoch": 1.8992601361349513,
+      "grad_norm": 1.0234096050262451,
+      "learning_rate": 6.260151819412616e-06,
+      "loss": 0.1886,
+      "step": 4011
+    },
+    {
+      "epoch": 1.8997336490085823,
+      "grad_norm": 1.4769940376281738,
+      "learning_rate": 6.255409312692664e-06,
+      "loss": 0.2173,
+      "step": 4012
+    },
+    {
+      "epoch": 1.9002071618822138,
+      "grad_norm": 1.2099794149398804,
+      "learning_rate": 6.250667785334282e-06,
+      "loss": 0.2085,
+      "step": 4013
+    },
+    {
+      "epoch": 1.900680674755845,
+      "grad_norm": 1.8601834774017334,
+      "learning_rate": 6.24592723857757e-06,
+      "loss": 0.2349,
+      "step": 4014
+    },
+    {
+      "epoch": 1.9011541876294762,
+      "grad_norm": 1.1096925735473633,
+      "learning_rate": 6.241187673662375e-06,
+      "loss": 0.2258,
+      "step": 4015
+    },
+    {
+      "epoch": 1.9016277005031075,
+      "grad_norm": 1.2290061712265015,
+      "learning_rate": 6.236449091828278e-06,
+      "loss": 0.2111,
+      "step": 4016
+    },
+    {
+      "epoch": 1.9021012133767385,
+      "grad_norm": 1.0335679054260254,
+      "learning_rate": 6.231711494314618e-06,
+      "loss": 0.1963,
+      "step": 4017
+    },
+    {
+      "epoch": 1.90257472625037,
+      "grad_norm": 1.2123810052871704,
+      "learning_rate": 6.226974882360462e-06,
+      "loss": 0.2317,
+      "step": 4018
+    },
+    {
+      "epoch": 1.9030482391240011,
+      "grad_norm": 1.043067216873169,
+      "learning_rate": 6.222239257204625e-06,
+      "loss": 0.2035,
+      "step": 4019
+    },
+    {
+      "epoch": 1.9035217519976324,
+      "grad_norm": 1.1024198532104492,
+      "learning_rate": 6.217504620085662e-06,
+      "loss": 0.2172,
+      "step": 4020
+    },
+    {
+      "epoch": 1.9039952648712637,
+      "grad_norm": 1.036494493484497,
+      "learning_rate": 6.21277097224188e-06,
+      "loss": 0.2044,
+      "step": 4021
+    },
+    {
+      "epoch": 1.904468777744895,
+      "grad_norm": 1.0444680452346802,
+      "learning_rate": 6.208038314911312e-06,
+      "loss": 0.1994,
+      "step": 4022
+    },
+    {
+      "epoch": 1.9049422906185263,
+      "grad_norm": 1.1238281726837158,
+      "learning_rate": 6.203306649331744e-06,
+      "loss": 0.1888,
+      "step": 4023
+    },
+    {
+      "epoch": 1.9054158034921573,
+      "grad_norm": 0.9902118444442749,
+      "learning_rate": 6.1985759767406925e-06,
+      "loss": 0.1989,
+      "step": 4024
+    },
+    {
+      "epoch": 1.9058893163657888,
+      "grad_norm": 1.583835244178772,
+      "learning_rate": 6.1938462983754235e-06,
+      "loss": 0.2093,
+      "step": 4025
+    },
+    {
+      "epoch": 1.90636282923942,
+      "grad_norm": 1.8781702518463135,
+      "learning_rate": 6.189117615472935e-06,
+      "loss": 0.2073,
+      "step": 4026
+    },
+    {
+      "epoch": 1.9068363421130512,
+      "grad_norm": 1.4825783967971802,
+      "learning_rate": 6.1843899292699695e-06,
+      "loss": 0.2244,
+      "step": 4027
+    },
+    {
+      "epoch": 1.9073098549866825,
+      "grad_norm": 1.3632996082305908,
+      "learning_rate": 6.179663241003008e-06,
+      "loss": 0.2145,
+      "step": 4028
+    },
+    {
+      "epoch": 1.9077833678603136,
+      "grad_norm": 1.502853274345398,
+      "learning_rate": 6.174937551908271e-06,
+      "loss": 0.2004,
+      "step": 4029
+    },
+    {
+      "epoch": 1.908256880733945,
+      "grad_norm": 1.4652563333511353,
+      "learning_rate": 6.170212863221715e-06,
+      "loss": 0.2041,
+      "step": 4030
+    },
+    {
+      "epoch": 1.9087303936075761,
+      "grad_norm": 1.1851582527160645,
+      "learning_rate": 6.165489176179039e-06,
+      "loss": 0.2071,
+      "step": 4031
+    },
+    {
+      "epoch": 1.9092039064812074,
+      "grad_norm": 1.044634222984314,
+      "learning_rate": 6.16076649201568e-06,
+      "loss": 0.203,
+      "step": 4032
+    },
+    {
+      "epoch": 1.9096774193548387,
+      "grad_norm": 1.1396992206573486,
+      "learning_rate": 6.1560448119668034e-06,
+      "loss": 0.2165,
+      "step": 4033
+    },
+    {
+      "epoch": 1.91015093222847,
+      "grad_norm": 1.1796884536743164,
+      "learning_rate": 6.151324137267322e-06,
+      "loss": 0.2182,
+      "step": 4034
+    },
+    {
+      "epoch": 1.9106244451021013,
+      "grad_norm": 1.1063097715377808,
+      "learning_rate": 6.146604469151886e-06,
+      "loss": 0.2015,
+      "step": 4035
+    },
+    {
+      "epoch": 1.9110979579757323,
+      "grad_norm": 2.4029173851013184,
+      "learning_rate": 6.141885808854877e-06,
+      "loss": 0.231,
+      "step": 4036
+    },
+    {
+      "epoch": 1.9115714708493639,
+      "grad_norm": 1.0543550252914429,
+      "learning_rate": 6.137168157610413e-06,
+      "loss": 0.1866,
+      "step": 4037
+    },
+    {
+      "epoch": 1.912044983722995,
+      "grad_norm": 1.8748868703842163,
+      "learning_rate": 6.1324515166523535e-06,
+      "loss": 0.1752,
+      "step": 4038
+    },
+    {
+      "epoch": 1.9125184965966262,
+      "grad_norm": 1.1984789371490479,
+      "learning_rate": 6.1277358872142875e-06,
+      "loss": 0.213,
+      "step": 4039
+    },
+    {
+      "epoch": 1.9129920094702575,
+      "grad_norm": 1.075285792350769,
+      "learning_rate": 6.1230212705295455e-06,
+      "loss": 0.2163,
+      "step": 4040
+    },
+    {
+      "epoch": 1.9134655223438886,
+      "grad_norm": 1.1514887809753418,
+      "learning_rate": 6.118307667831187e-06,
+      "loss": 0.1981,
+      "step": 4041
+    },
+    {
+      "epoch": 1.91393903521752,
+      "grad_norm": 1.0806608200073242,
+      "learning_rate": 6.113595080352013e-06,
+      "loss": 0.2059,
+      "step": 4042
+    },
+    {
+      "epoch": 1.9144125480911511,
+      "grad_norm": 0.9989210963249207,
+      "learning_rate": 6.10888350932455e-06,
+      "loss": 0.1973,
+      "step": 4043
+    },
+    {
+      "epoch": 1.9148860609647826,
+      "grad_norm": 1.4374021291732788,
+      "learning_rate": 6.104172955981069e-06,
+      "loss": 0.2153,
+      "step": 4044
+    },
+    {
+      "epoch": 1.9153595738384137,
+      "grad_norm": 1.3092314004898071,
+      "learning_rate": 6.0994634215535665e-06,
+      "loss": 0.1993,
+      "step": 4045
+    },
+    {
+      "epoch": 1.915833086712045,
+      "grad_norm": 1.3858625888824463,
+      "learning_rate": 6.094754907273777e-06,
+      "loss": 0.1935,
+      "step": 4046
+    },
+    {
+      "epoch": 1.9163065995856763,
+      "grad_norm": 1.9879931211471558,
+      "learning_rate": 6.090047414373166e-06,
+      "loss": 0.1879,
+      "step": 4047
+    },
+    {
+      "epoch": 1.9167801124593074,
+      "grad_norm": 1.0218373537063599,
+      "learning_rate": 6.085340944082935e-06,
+      "loss": 0.2103,
+      "step": 4048
+    },
+    {
+      "epoch": 1.9172536253329389,
+      "grad_norm": 1.021353840827942,
+      "learning_rate": 6.0806354976340145e-06,
+      "loss": 0.2111,
+      "step": 4049
+    },
+    {
+      "epoch": 1.91772713820657,
+      "grad_norm": 1.1734100580215454,
+      "learning_rate": 6.075931076257069e-06,
+      "loss": 0.2014,
+      "step": 4050
+    },
+    {
+      "epoch": 1.9182006510802012,
+      "grad_norm": 1.5947613716125488,
+      "learning_rate": 6.071227681182494e-06,
+      "loss": 0.1894,
+      "step": 4051
+    },
+    {
+      "epoch": 1.9186741639538325,
+      "grad_norm": 1.2673771381378174,
+      "learning_rate": 6.066525313640419e-06,
+      "loss": 0.1882,
+      "step": 4052
+    },
+    {
+      "epoch": 1.9191476768274638,
+      "grad_norm": 1.011744499206543,
+      "learning_rate": 6.061823974860699e-06,
+      "loss": 0.2165,
+      "step": 4053
+    },
+    {
+      "epoch": 1.919621189701095,
+      "grad_norm": 1.5951597690582275,
+      "learning_rate": 6.057123666072927e-06,
+      "loss": 0.232,
+      "step": 4054
+    },
+    {
+      "epoch": 1.9200947025747261,
+      "grad_norm": 1.3709065914154053,
+      "learning_rate": 6.052424388506421e-06,
+      "loss": 0.2051,
+      "step": 4055
+    },
+    {
+      "epoch": 1.9205682154483577,
+      "grad_norm": 1.1398015022277832,
+      "learning_rate": 6.047726143390236e-06,
+      "loss": 0.2136,
+      "step": 4056
+    },
+    {
+      "epoch": 1.9210417283219887,
+      "grad_norm": 1.450731873512268,
+      "learning_rate": 6.043028931953148e-06,
+      "loss": 0.2186,
+      "step": 4057
+    },
+    {
+      "epoch": 1.92151524119562,
+      "grad_norm": 1.0435858964920044,
+      "learning_rate": 6.03833275542367e-06,
+      "loss": 0.21,
+      "step": 4058
+    },
+    {
+      "epoch": 1.9219887540692513,
+      "grad_norm": 1.3922533988952637,
+      "learning_rate": 6.033637615030039e-06,
+      "loss": 0.2023,
+      "step": 4059
+    },
+    {
+      "epoch": 1.9224622669428824,
+      "grad_norm": 1.1114870309829712,
+      "learning_rate": 6.028943512000227e-06,
+      "loss": 0.2057,
+      "step": 4060
+    },
+    {
+      "epoch": 1.9229357798165139,
+      "grad_norm": 1.509615421295166,
+      "learning_rate": 6.02425044756193e-06,
+      "loss": 0.1999,
+      "step": 4061
+    },
+    {
+      "epoch": 1.923409292690145,
+      "grad_norm": 1.4083658456802368,
+      "learning_rate": 6.019558422942575e-06,
+      "loss": 0.2151,
+      "step": 4062
+    },
+    {
+      "epoch": 1.9238828055637762,
+      "grad_norm": 1.1245790719985962,
+      "learning_rate": 6.014867439369314e-06,
+      "loss": 0.2422,
+      "step": 4063
+    },
+    {
+      "epoch": 1.9243563184374075,
+      "grad_norm": 1.3989055156707764,
+      "learning_rate": 6.010177498069027e-06,
+      "loss": 0.1943,
+      "step": 4064
+    },
+    {
+      "epoch": 1.9248298313110388,
+      "grad_norm": 1.1599762439727783,
+      "learning_rate": 6.005488600268328e-06,
+      "loss": 0.2005,
+      "step": 4065
+    },
+    {
+      "epoch": 1.92530334418467,
+      "grad_norm": 1.421303391456604,
+      "learning_rate": 6.000800747193547e-06,
+      "loss": 0.2228,
+      "step": 4066
+    },
+    {
+      "epoch": 1.9257768570583012,
+      "grad_norm": 1.3245368003845215,
+      "learning_rate": 5.996113940070754e-06,
+      "loss": 0.2075,
+      "step": 4067
+    },
+    {
+      "epoch": 1.9262503699319327,
+      "grad_norm": 1.5470653772354126,
+      "learning_rate": 5.99142818012573e-06,
+      "loss": 0.2029,
+      "step": 4068
+    },
+    {
+      "epoch": 1.9267238828055637,
+      "grad_norm": 1.1906659603118896,
+      "learning_rate": 5.986743468583996e-06,
+      "loss": 0.2261,
+      "step": 4069
+    },
+    {
+      "epoch": 1.927197395679195,
+      "grad_norm": 1.2057496309280396,
+      "learning_rate": 5.982059806670788e-06,
+      "loss": 0.2086,
+      "step": 4070
+    },
+    {
+      "epoch": 1.9276709085528263,
+      "grad_norm": 1.0178190469741821,
+      "learning_rate": 5.9773771956110785e-06,
+      "loss": 0.1835,
+      "step": 4071
+    },
+    {
+      "epoch": 1.9281444214264574,
+      "grad_norm": 1.4230746030807495,
+      "learning_rate": 5.972695636629555e-06,
+      "loss": 0.2209,
+      "step": 4072
+    },
+    {
+      "epoch": 1.9286179343000889,
+      "grad_norm": 1.5597596168518066,
+      "learning_rate": 5.968015130950638e-06,
+      "loss": 0.1961,
+      "step": 4073
+    },
+    {
+      "epoch": 1.92909144717372,
+      "grad_norm": 0.9596446752548218,
+      "learning_rate": 5.963335679798465e-06,
+      "loss": 0.1941,
+      "step": 4074
+    },
+    {
+      "epoch": 1.9295649600473515,
+      "grad_norm": 1.1124833822250366,
+      "learning_rate": 5.958657284396902e-06,
+      "loss": 0.1979,
+      "step": 4075
+    },
+    {
+      "epoch": 1.9300384729209825,
+      "grad_norm": 1.0595910549163818,
+      "learning_rate": 5.953979945969539e-06,
+      "loss": 0.2157,
+      "step": 4076
+    },
+    {
+      "epoch": 1.9305119857946138,
+      "grad_norm": 0.9917038083076477,
+      "learning_rate": 5.949303665739689e-06,
+      "loss": 0.2023,
+      "step": 4077
+    },
+    {
+      "epoch": 1.930985498668245,
+      "grad_norm": 1.0265789031982422,
+      "learning_rate": 5.944628444930388e-06,
+      "loss": 0.2223,
+      "step": 4078
+    },
+    {
+      "epoch": 1.9314590115418762,
+      "grad_norm": 1.217456579208374,
+      "learning_rate": 5.9399542847643935e-06,
+      "loss": 0.2226,
+      "step": 4079
+    },
+    {
+      "epoch": 1.9319325244155077,
+      "grad_norm": 1.284399151802063,
+      "learning_rate": 5.935281186464188e-06,
+      "loss": 0.2251,
+      "step": 4080
+    },
+    {
+      "epoch": 1.9324060372891387,
+      "grad_norm": 1.6773645877838135,
+      "learning_rate": 5.930609151251975e-06,
+      "loss": 0.2266,
+      "step": 4081
+    },
+    {
+      "epoch": 1.93287955016277,
+      "grad_norm": 1.1189942359924316,
+      "learning_rate": 5.925938180349679e-06,
+      "loss": 0.1975,
+      "step": 4082
+    },
+    {
+      "epoch": 1.9333530630364013,
+      "grad_norm": 1.3536006212234497,
+      "learning_rate": 5.921268274978951e-06,
+      "loss": 0.2106,
+      "step": 4083
+    },
+    {
+      "epoch": 1.9338265759100326,
+      "grad_norm": 1.3153413534164429,
+      "learning_rate": 5.9165994363611565e-06,
+      "loss": 0.2272,
+      "step": 4084
+    },
+    {
+      "epoch": 1.9343000887836639,
+      "grad_norm": 1.37685227394104,
+      "learning_rate": 5.911931665717386e-06,
+      "loss": 0.2237,
+      "step": 4085
+    },
+    {
+      "epoch": 1.934773601657295,
+      "grad_norm": 1.1147427558898926,
+      "learning_rate": 5.907264964268451e-06,
+      "loss": 0.2185,
+      "step": 4086
+    },
+    {
+      "epoch": 1.9352471145309265,
+      "grad_norm": 1.5562469959259033,
+      "learning_rate": 5.902599333234882e-06,
+      "loss": 0.205,
+      "step": 4087
+    },
+    {
+      "epoch": 1.9357206274045575,
+      "grad_norm": 1.3844199180603027,
+      "learning_rate": 5.8979347738369276e-06,
+      "loss": 0.2183,
+      "step": 4088
+    },
+    {
+      "epoch": 1.9361941402781888,
+      "grad_norm": 1.3082056045532227,
+      "learning_rate": 5.89327128729456e-06,
+      "loss": 0.1965,
+      "step": 4089
+    },
+    {
+      "epoch": 1.93666765315182,
+      "grad_norm": 0.897994875907898,
+      "learning_rate": 5.888608874827469e-06,
+      "loss": 0.2081,
+      "step": 4090
+    },
+    {
+      "epoch": 1.9371411660254512,
+      "grad_norm": 0.9968763589859009,
+      "learning_rate": 5.883947537655061e-06,
+      "loss": 0.2208,
+      "step": 4091
+    },
+    {
+      "epoch": 1.9376146788990827,
+      "grad_norm": 1.1803154945373535,
+      "learning_rate": 5.8792872769964705e-06,
+      "loss": 0.1927,
+      "step": 4092
+    },
+    {
+      "epoch": 1.9380881917727137,
+      "grad_norm": 1.3356378078460693,
+      "learning_rate": 5.874628094070536e-06,
+      "loss": 0.1976,
+      "step": 4093
+    },
+    {
+      "epoch": 1.938561704646345,
+      "grad_norm": 1.4385789632797241,
+      "learning_rate": 5.869969990095828e-06,
+      "loss": 0.22,
+      "step": 4094
+    },
+    {
+      "epoch": 1.9390352175199763,
+      "grad_norm": 1.067402958869934,
+      "learning_rate": 5.865312966290624e-06,
+      "loss": 0.2028,
+      "step": 4095
+    },
+    {
+      "epoch": 1.9395087303936076,
+      "grad_norm": 1.020371675491333,
+      "learning_rate": 5.8606570238729286e-06,
+      "loss": 0.2087,
+      "step": 4096
+    },
+    {
+      "epoch": 1.939982243267239,
+      "grad_norm": 1.094397783279419,
+      "learning_rate": 5.856002164060453e-06,
+      "loss": 0.1967,
+      "step": 4097
+    },
+    {
+      "epoch": 1.94045575614087,
+      "grad_norm": 1.3213562965393066,
+      "learning_rate": 5.85134838807063e-06,
+      "loss": 0.2218,
+      "step": 4098
+    },
+    {
+      "epoch": 1.9409292690145015,
+      "grad_norm": 1.4701647758483887,
+      "learning_rate": 5.846695697120617e-06,
+      "loss": 0.2085,
+      "step": 4099
+    },
+    {
+      "epoch": 1.9414027818881325,
+      "grad_norm": 0.9358981847763062,
+      "learning_rate": 5.842044092427277e-06,
+      "loss": 0.2129,
+      "step": 4100
+    },
+    {
+      "epoch": 1.9418762947617638,
+      "grad_norm": 1.27782142162323,
+      "learning_rate": 5.837393575207194e-06,
+      "loss": 0.204,
+      "step": 4101
+    },
+    {
+      "epoch": 1.942349807635395,
+      "grad_norm": 1.1165828704833984,
+      "learning_rate": 5.832744146676661e-06,
+      "loss": 0.2169,
+      "step": 4102
+    },
+    {
+      "epoch": 1.9428233205090262,
+      "grad_norm": 1.0212310552597046,
+      "learning_rate": 5.828095808051697e-06,
+      "loss": 0.2175,
+      "step": 4103
+    },
+    {
+      "epoch": 1.9432968333826577,
+      "grad_norm": 0.9246084094047546,
+      "learning_rate": 5.823448560548024e-06,
+      "loss": 0.1908,
+      "step": 4104
+    },
+    {
+      "epoch": 1.9437703462562888,
+      "grad_norm": 1.1818164587020874,
+      "learning_rate": 5.818802405381091e-06,
+      "loss": 0.2047,
+      "step": 4105
+    },
+    {
+      "epoch": 1.9442438591299203,
+      "grad_norm": 1.5867642164230347,
+      "learning_rate": 5.814157343766049e-06,
+      "loss": 0.2179,
+      "step": 4106
+    },
+    {
+      "epoch": 1.9447173720035513,
+      "grad_norm": 1.1746879816055298,
+      "learning_rate": 5.8095133769177766e-06,
+      "loss": 0.1948,
+      "step": 4107
+    },
+    {
+      "epoch": 1.9451908848771826,
+      "grad_norm": 1.1779289245605469,
+      "learning_rate": 5.804870506050853e-06,
+      "loss": 0.2007,
+      "step": 4108
+    },
+    {
+      "epoch": 1.945664397750814,
+      "grad_norm": 0.9775161743164062,
+      "learning_rate": 5.800228732379574e-06,
+      "loss": 0.202,
+      "step": 4109
+    },
+    {
+      "epoch": 1.946137910624445,
+      "grad_norm": 1.0120432376861572,
+      "learning_rate": 5.795588057117958e-06,
+      "loss": 0.1857,
+      "step": 4110
+    },
+    {
+      "epoch": 1.9466114234980765,
+      "grad_norm": 1.3756533861160278,
+      "learning_rate": 5.790948481479721e-06,
+      "loss": 0.181,
+      "step": 4111
+    },
+    {
+      "epoch": 1.9470849363717075,
+      "grad_norm": 1.0722591876983643,
+      "learning_rate": 5.786310006678308e-06,
+      "loss": 0.2051,
+      "step": 4112
+    },
+    {
+      "epoch": 1.9475584492453388,
+      "grad_norm": 1.1463873386383057,
+      "learning_rate": 5.781672633926858e-06,
+      "loss": 0.1911,
+      "step": 4113
+    },
+    {
+      "epoch": 1.9480319621189701,
+      "grad_norm": 1.0058190822601318,
+      "learning_rate": 5.77703636443824e-06,
+      "loss": 0.2162,
+      "step": 4114
+    },
+    {
+      "epoch": 1.9485054749926014,
+      "grad_norm": 1.7192530632019043,
+      "learning_rate": 5.772401199425017e-06,
+      "loss": 0.2117,
+      "step": 4115
+    },
+    {
+      "epoch": 1.9489789878662327,
+      "grad_norm": 1.1650121212005615,
+      "learning_rate": 5.76776714009948e-06,
+      "loss": 0.2388,
+      "step": 4116
+    },
+    {
+      "epoch": 1.9494525007398638,
+      "grad_norm": 1.3511508703231812,
+      "learning_rate": 5.763134187673618e-06,
+      "loss": 0.2045,
+      "step": 4117
+    },
+    {
+      "epoch": 1.9499260136134953,
+      "grad_norm": 1.0509499311447144,
+      "learning_rate": 5.7585023433591315e-06,
+      "loss": 0.2094,
+      "step": 4118
+    },
+    {
+      "epoch": 1.9503995264871263,
+      "grad_norm": 1.0673551559448242,
+      "learning_rate": 5.7538716083674425e-06,
+      "loss": 0.1909,
+      "step": 4119
+    },
+    {
+      "epoch": 1.9508730393607576,
+      "grad_norm": 1.0487030744552612,
+      "learning_rate": 5.749241983909668e-06,
+      "loss": 0.2173,
+      "step": 4120
+    },
+    {
+      "epoch": 1.951346552234389,
+      "grad_norm": 1.2835848331451416,
+      "learning_rate": 5.744613471196648e-06,
+      "loss": 0.2258,
+      "step": 4121
+    },
+    {
+      "epoch": 1.95182006510802,
+      "grad_norm": 1.9468538761138916,
+      "learning_rate": 5.739986071438919e-06,
+      "loss": 0.2424,
+      "step": 4122
+    },
+    {
+      "epoch": 1.9522935779816515,
+      "grad_norm": 1.0411995649337769,
+      "learning_rate": 5.735359785846739e-06,
+      "loss": 0.1879,
+      "step": 4123
+    },
+    {
+      "epoch": 1.9527670908552826,
+      "grad_norm": 1.0039364099502563,
+      "learning_rate": 5.730734615630063e-06,
+      "loss": 0.1935,
+      "step": 4124
+    },
+    {
+      "epoch": 1.9532406037289138,
+      "grad_norm": 1.342782735824585,
+      "learning_rate": 5.7261105619985635e-06,
+      "loss": 0.2116,
+      "step": 4125
+    },
+    {
+      "epoch": 1.9537141166025451,
+      "grad_norm": 0.9727945923805237,
+      "learning_rate": 5.721487626161617e-06,
+      "loss": 0.2056,
+      "step": 4126
+    },
+    {
+      "epoch": 1.9541876294761764,
+      "grad_norm": 0.834372341632843,
+      "learning_rate": 5.7168658093283026e-06,
+      "loss": 0.1894,
+      "step": 4127
+    },
+    {
+      "epoch": 1.9546611423498077,
+      "grad_norm": 1.5934944152832031,
+      "learning_rate": 5.7122451127074185e-06,
+      "loss": 0.206,
+      "step": 4128
+    },
+    {
+      "epoch": 1.9551346552234388,
+      "grad_norm": 1.4059547185897827,
+      "learning_rate": 5.7076255375074574e-06,
+      "loss": 0.2246,
+      "step": 4129
+    },
+    {
+      "epoch": 1.9556081680970703,
+      "grad_norm": 1.7891411781311035,
+      "learning_rate": 5.703007084936631e-06,
+      "loss": 0.188,
+      "step": 4130
+    },
+    {
+      "epoch": 1.9560816809707013,
+      "grad_norm": 0.9586341381072998,
+      "learning_rate": 5.698389756202844e-06,
+      "loss": 0.2105,
+      "step": 4131
+    },
+    {
+      "epoch": 1.9565551938443326,
+      "grad_norm": 1.686652660369873,
+      "learning_rate": 5.693773552513723e-06,
+      "loss": 0.2437,
+      "step": 4132
+    },
+    {
+      "epoch": 1.957028706717964,
+      "grad_norm": 1.3186736106872559,
+      "learning_rate": 5.689158475076582e-06,
+      "loss": 0.2004,
+      "step": 4133
+    },
+    {
+      "epoch": 1.957502219591595,
+      "grad_norm": 1.3622384071350098,
+      "learning_rate": 5.6845445250984566e-06,
+      "loss": 0.2079,
+      "step": 4134
+    },
+    {
+      "epoch": 1.9579757324652265,
+      "grad_norm": 1.2416852712631226,
+      "learning_rate": 5.679931703786077e-06,
+      "loss": 0.2208,
+      "step": 4135
+    },
+    {
+      "epoch": 1.9584492453388576,
+      "grad_norm": 0.9193575978279114,
+      "learning_rate": 5.675320012345887e-06,
+      "loss": 0.2292,
+      "step": 4136
+    },
+    {
+      "epoch": 1.958922758212489,
+      "grad_norm": 1.2942508459091187,
+      "learning_rate": 5.670709451984022e-06,
+      "loss": 0.2146,
+      "step": 4137
+    },
+    {
+      "epoch": 1.9593962710861201,
+      "grad_norm": 1.3082016706466675,
+      "learning_rate": 5.666100023906336e-06,
+      "loss": 0.2173,
+      "step": 4138
+    },
+    {
+      "epoch": 1.9598697839597514,
+      "grad_norm": 1.0084069967269897,
+      "learning_rate": 5.66149172931838e-06,
+      "loss": 0.2172,
+      "step": 4139
+    },
+    {
+      "epoch": 1.9603432968333827,
+      "grad_norm": 1.1174718141555786,
+      "learning_rate": 5.656884569425407e-06,
+      "loss": 0.2017,
+      "step": 4140
+    },
+    {
+      "epoch": 1.9608168097070138,
+      "grad_norm": 1.1940710544586182,
+      "learning_rate": 5.6522785454323795e-06,
+      "loss": 0.2027,
+      "step": 4141
+    },
+    {
+      "epoch": 1.9612903225806453,
+      "grad_norm": 1.1007012128829956,
+      "learning_rate": 5.647673658543954e-06,
+      "loss": 0.2218,
+      "step": 4142
+    },
+    {
+      "epoch": 1.9617638354542764,
+      "grad_norm": 1.368607997894287,
+      "learning_rate": 5.643069909964491e-06,
+      "loss": 0.1888,
+      "step": 4143
+    },
+    {
+      "epoch": 1.9622373483279076,
+      "grad_norm": 1.2197048664093018,
+      "learning_rate": 5.638467300898067e-06,
+      "loss": 0.2261,
+      "step": 4144
+    },
+    {
+      "epoch": 1.962710861201539,
+      "grad_norm": 1.1047946214675903,
+      "learning_rate": 5.6338658325484395e-06,
+      "loss": 0.1864,
+      "step": 4145
+    },
+    {
+      "epoch": 1.9631843740751702,
+      "grad_norm": 1.3874999284744263,
+      "learning_rate": 5.629265506119086e-06,
+      "loss": 0.2419,
+      "step": 4146
+    },
+    {
+      "epoch": 1.9636578869488015,
+      "grad_norm": 1.1594977378845215,
+      "learning_rate": 5.62466632281317e-06,
+      "loss": 0.2087,
+      "step": 4147
+    },
+    {
+      "epoch": 1.9641313998224326,
+      "grad_norm": 1.0370484590530396,
+      "learning_rate": 5.620068283833573e-06,
+      "loss": 0.2138,
+      "step": 4148
+    },
+    {
+      "epoch": 1.964604912696064,
+      "grad_norm": 1.3975090980529785,
+      "learning_rate": 5.615471390382858e-06,
+      "loss": 0.1873,
+      "step": 4149
+    },
+    {
+      "epoch": 1.9650784255696951,
+      "grad_norm": 1.7761406898498535,
+      "learning_rate": 5.610875643663305e-06,
+      "loss": 0.215,
+      "step": 4150
+    },
+    {
+      "epoch": 1.9655519384433264,
+      "grad_norm": 1.3416228294372559,
+      "learning_rate": 5.606281044876887e-06,
+      "loss": 0.2152,
+      "step": 4151
+    },
+    {
+      "epoch": 1.9660254513169577,
+      "grad_norm": 1.7138333320617676,
+      "learning_rate": 5.601687595225269e-06,
+      "loss": 0.2265,
+      "step": 4152
+    },
+    {
+      "epoch": 1.9664989641905888,
+      "grad_norm": 1.7663260698318481,
+      "learning_rate": 5.597095295909833e-06,
+      "loss": 0.2109,
+      "step": 4153
+    },
+    {
+      "epoch": 1.9669724770642203,
+      "grad_norm": 1.0027509927749634,
+      "learning_rate": 5.592504148131645e-06,
+      "loss": 0.2036,
+      "step": 4154
+    },
+    {
+      "epoch": 1.9674459899378514,
+      "grad_norm": 0.9841294884681702,
+      "learning_rate": 5.587914153091479e-06,
+      "loss": 0.223,
+      "step": 4155
+    },
+    {
+      "epoch": 1.9679195028114826,
+      "grad_norm": 1.8754013776779175,
+      "learning_rate": 5.583325311989799e-06,
+      "loss": 0.1892,
+      "step": 4156
+    },
+    {
+      "epoch": 1.968393015685114,
+      "grad_norm": 1.138641357421875,
+      "learning_rate": 5.57873762602678e-06,
+      "loss": 0.2099,
+      "step": 4157
+    },
+    {
+      "epoch": 1.9688665285587452,
+      "grad_norm": 1.1637465953826904,
+      "learning_rate": 5.574151096402276e-06,
+      "loss": 0.1921,
+      "step": 4158
+    },
+    {
+      "epoch": 1.9693400414323765,
+      "grad_norm": 1.3505114316940308,
+      "learning_rate": 5.569565724315862e-06,
+      "loss": 0.2148,
+      "step": 4159
+    },
+    {
+      "epoch": 1.9698135543060076,
+      "grad_norm": 2.0613527297973633,
+      "learning_rate": 5.5649815109667874e-06,
+      "loss": 0.1949,
+      "step": 4160
+    },
+    {
+      "epoch": 1.970287067179639,
+      "grad_norm": 1.8924801349639893,
+      "learning_rate": 5.560398457554016e-06,
+      "loss": 0.2145,
+      "step": 4161
+    },
+    {
+      "epoch": 1.9707605800532702,
+      "grad_norm": 1.1464554071426392,
+      "learning_rate": 5.5558165652762e-06,
+      "loss": 0.2119,
+      "step": 4162
+    },
+    {
+      "epoch": 1.9712340929269014,
+      "grad_norm": 1.047352910041809,
+      "learning_rate": 5.551235835331682e-06,
+      "loss": 0.2148,
+      "step": 4163
+    },
+    {
+      "epoch": 1.9717076058005327,
+      "grad_norm": 1.259980320930481,
+      "learning_rate": 5.546656268918517e-06,
+      "loss": 0.2182,
+      "step": 4164
+    },
+    {
+      "epoch": 1.9721811186741638,
+      "grad_norm": 1.1720131635665894,
+      "learning_rate": 5.542077867234441e-06,
+      "loss": 0.1947,
+      "step": 4165
+    },
+    {
+      "epoch": 1.9726546315477953,
+      "grad_norm": 1.064708948135376,
+      "learning_rate": 5.537500631476895e-06,
+      "loss": 0.1935,
+      "step": 4166
+    },
+    {
+      "epoch": 1.9731281444214264,
+      "grad_norm": 1.1356375217437744,
+      "learning_rate": 5.5329245628430036e-06,
+      "loss": 0.1802,
+      "step": 4167
+    },
+    {
+      "epoch": 1.9736016572950577,
+      "grad_norm": 2.05705189704895,
+      "learning_rate": 5.528349662529604e-06,
+      "loss": 0.2095,
+      "step": 4168
+    },
+    {
+      "epoch": 1.974075170168689,
+      "grad_norm": 1.081849455833435,
+      "learning_rate": 5.5237759317332065e-06,
+      "loss": 0.2166,
+      "step": 4169
+    },
+    {
+      "epoch": 1.9745486830423202,
+      "grad_norm": 1.1514532566070557,
+      "learning_rate": 5.519203371650035e-06,
+      "loss": 0.2273,
+      "step": 4170
+    },
+    {
+      "epoch": 1.9750221959159515,
+      "grad_norm": 1.438201665878296,
+      "learning_rate": 5.514631983475995e-06,
+      "loss": 0.2014,
+      "step": 4171
+    },
+    {
+      "epoch": 1.9754957087895826,
+      "grad_norm": 1.2542989253997803,
+      "learning_rate": 5.510061768406683e-06,
+      "loss": 0.2045,
+      "step": 4172
+    },
+    {
+      "epoch": 1.975969221663214,
+      "grad_norm": 0.9703104496002197,
+      "learning_rate": 5.505492727637406e-06,
+      "loss": 0.1985,
+      "step": 4173
+    },
+    {
+      "epoch": 1.9764427345368452,
+      "grad_norm": 1.1615082025527954,
+      "learning_rate": 5.5009248623631416e-06,
+      "loss": 0.1902,
+      "step": 4174
+    },
+    {
+      "epoch": 1.9769162474104764,
+      "grad_norm": 1.1647366285324097,
+      "learning_rate": 5.496358173778582e-06,
+      "loss": 0.2114,
+      "step": 4175
+    },
+    {
+      "epoch": 1.9773897602841077,
+      "grad_norm": 1.780960202217102,
+      "learning_rate": 5.4917926630780895e-06,
+      "loss": 0.2196,
+      "step": 4176
+    },
+    {
+      "epoch": 1.977863273157739,
+      "grad_norm": 1.0951555967330933,
+      "learning_rate": 5.487228331455734e-06,
+      "loss": 0.1874,
+      "step": 4177
+    },
+    {
+      "epoch": 1.9783367860313703,
+      "grad_norm": 1.208298921585083,
+      "learning_rate": 5.482665180105278e-06,
+      "loss": 0.1986,
+      "step": 4178
+    },
+    {
+      "epoch": 1.9788102989050014,
+      "grad_norm": 1.1400575637817383,
+      "learning_rate": 5.4781032102201605e-06,
+      "loss": 0.2193,
+      "step": 4179
+    },
+    {
+      "epoch": 1.9792838117786329,
+      "grad_norm": 1.7641512155532837,
+      "learning_rate": 5.4735424229935274e-06,
+      "loss": 0.2038,
+      "step": 4180
+    },
+    {
+      "epoch": 1.979757324652264,
+      "grad_norm": 1.020308256149292,
+      "learning_rate": 5.468982819618204e-06,
+      "loss": 0.2143,
+      "step": 4181
+    },
+    {
+      "epoch": 1.9802308375258952,
+      "grad_norm": 0.9065782427787781,
+      "learning_rate": 5.464424401286715e-06,
+      "loss": 0.2096,
+      "step": 4182
+    },
+    {
+      "epoch": 1.9807043503995265,
+      "grad_norm": 1.3247904777526855,
+      "learning_rate": 5.459867169191267e-06,
+      "loss": 0.2037,
+      "step": 4183
+    },
+    {
+      "epoch": 1.9811778632731576,
+      "grad_norm": 1.2426649332046509,
+      "learning_rate": 5.455311124523762e-06,
+      "loss": 0.227,
+      "step": 4184
+    },
+    {
+      "epoch": 1.981651376146789,
+      "grad_norm": 0.9883340001106262,
+      "learning_rate": 5.4507562684757875e-06,
+      "loss": 0.1973,
+      "step": 4185
+    },
+    {
+      "epoch": 1.9821248890204202,
+      "grad_norm": 1.001715898513794,
+      "learning_rate": 5.446202602238626e-06,
+      "loss": 0.2071,
+      "step": 4186
+    },
+    {
+      "epoch": 1.9825984018940515,
+      "grad_norm": 0.9738106727600098,
+      "learning_rate": 5.441650127003244e-06,
+      "loss": 0.2014,
+      "step": 4187
+    },
+    {
+      "epoch": 1.9830719147676827,
+      "grad_norm": 1.1363893747329712,
+      "learning_rate": 5.4370988439602916e-06,
+      "loss": 0.1934,
+      "step": 4188
+    },
+    {
+      "epoch": 1.983545427641314,
+      "grad_norm": 1.2461233139038086,
+      "learning_rate": 5.4325487543001196e-06,
+      "loss": 0.2202,
+      "step": 4189
+    },
+    {
+      "epoch": 1.9840189405149453,
+      "grad_norm": 1.1095913648605347,
+      "learning_rate": 5.427999859212757e-06,
+      "loss": 0.2152,
+      "step": 4190
+    },
+    {
+      "epoch": 1.9844924533885764,
+      "grad_norm": 1.7337262630462646,
+      "learning_rate": 5.423452159887927e-06,
+      "loss": 0.2166,
+      "step": 4191
+    },
+    {
+      "epoch": 1.984965966262208,
+      "grad_norm": 0.9954274296760559,
+      "learning_rate": 5.41890565751503e-06,
+      "loss": 0.2174,
+      "step": 4192
+    },
+    {
+      "epoch": 1.985439479135839,
+      "grad_norm": 1.3146260976791382,
+      "learning_rate": 5.414360353283168e-06,
+      "loss": 0.1902,
+      "step": 4193
+    },
+    {
+      "epoch": 1.9859129920094702,
+      "grad_norm": 1.0898852348327637,
+      "learning_rate": 5.409816248381112e-06,
+      "loss": 0.2221,
+      "step": 4194
+    },
+    {
+      "epoch": 1.9863865048831015,
+      "grad_norm": 1.0851001739501953,
+      "learning_rate": 5.405273343997339e-06,
+      "loss": 0.1969,
+      "step": 4195
+    },
+    {
+      "epoch": 1.9868600177567326,
+      "grad_norm": 1.586696743965149,
+      "learning_rate": 5.400731641319996e-06,
+      "loss": 0.198,
+      "step": 4196
+    },
+    {
+      "epoch": 1.987333530630364,
+      "grad_norm": 1.0100387334823608,
+      "learning_rate": 5.39619114153692e-06,
+      "loss": 0.1984,
+      "step": 4197
+    },
+    {
+      "epoch": 1.9878070435039952,
+      "grad_norm": 1.0449753999710083,
+      "learning_rate": 5.39165184583564e-06,
+      "loss": 0.2017,
+      "step": 4198
+    },
+    {
+      "epoch": 1.9882805563776265,
+      "grad_norm": 1.3189187049865723,
+      "learning_rate": 5.387113755403357e-06,
+      "loss": 0.1889,
+      "step": 4199
+    },
+    {
+      "epoch": 1.9887540692512578,
+      "grad_norm": 1.8392877578735352,
+      "learning_rate": 5.3825768714269745e-06,
+      "loss": 0.2115,
+      "step": 4200
+    },
+    {
+      "epoch": 1.989227582124889,
+      "grad_norm": 0.9645484685897827,
+      "learning_rate": 5.378041195093063e-06,
+      "loss": 0.1919,
+      "step": 4201
+    },
+    {
+      "epoch": 1.9897010949985203,
+      "grad_norm": 1.2029494047164917,
+      "learning_rate": 5.3735067275878915e-06,
+      "loss": 0.2015,
+      "step": 4202
+    },
+    {
+      "epoch": 1.9901746078721514,
+      "grad_norm": 1.8107424974441528,
+      "learning_rate": 5.368973470097401e-06,
+      "loss": 0.209,
+      "step": 4203
+    },
+    {
+      "epoch": 1.990648120745783,
+      "grad_norm": 1.4543911218643188,
+      "learning_rate": 5.364441423807224e-06,
+      "loss": 0.2173,
+      "step": 4204
+    },
+    {
+      "epoch": 1.991121633619414,
+      "grad_norm": 1.1092660427093506,
+      "learning_rate": 5.359910589902674e-06,
+      "loss": 0.2003,
+      "step": 4205
+    },
+    {
+      "epoch": 1.9915951464930453,
+      "grad_norm": 1.1251716613769531,
+      "learning_rate": 5.355380969568742e-06,
+      "loss": 0.2065,
+      "step": 4206
+    },
+    {
+      "epoch": 1.9920686593666765,
+      "grad_norm": 1.116974949836731,
+      "learning_rate": 5.350852563990112e-06,
+      "loss": 0.2168,
+      "step": 4207
+    },
+    {
+      "epoch": 1.9925421722403078,
+      "grad_norm": 1.312705397605896,
+      "learning_rate": 5.34632537435114e-06,
+      "loss": 0.2283,
+      "step": 4208
+    },
+    {
+      "epoch": 1.9930156851139391,
+      "grad_norm": 1.2690129280090332,
+      "learning_rate": 5.341799401835877e-06,
+      "loss": 0.2151,
+      "step": 4209
+    },
+    {
+      "epoch": 1.9934891979875702,
+      "grad_norm": 0.9621937870979309,
+      "learning_rate": 5.3372746476280366e-06,
+      "loss": 0.2054,
+      "step": 4210
+    },
+    {
+      "epoch": 1.9939627108612017,
+      "grad_norm": 1.264472246170044,
+      "learning_rate": 5.3327511129110344e-06,
+      "loss": 0.2327,
+      "step": 4211
+    },
+    {
+      "epoch": 1.9944362237348328,
+      "grad_norm": 1.66592276096344,
+      "learning_rate": 5.328228798867947e-06,
+      "loss": 0.2133,
+      "step": 4212
+    },
+    {
+      "epoch": 1.994909736608464,
+      "grad_norm": 1.2937157154083252,
+      "learning_rate": 5.323707706681553e-06,
+      "loss": 0.2025,
+      "step": 4213
+    },
+    {
+      "epoch": 1.9953832494820953,
+      "grad_norm": 1.0050321817398071,
+      "learning_rate": 5.319187837534292e-06,
+      "loss": 0.2058,
+      "step": 4214
+    },
+    {
+      "epoch": 1.9958567623557264,
+      "grad_norm": 1.160569190979004,
+      "learning_rate": 5.314669192608296e-06,
+      "loss": 0.205,
+      "step": 4215
+    },
+    {
+      "epoch": 1.996330275229358,
+      "grad_norm": 0.9353100657463074,
+      "learning_rate": 5.310151773085376e-06,
+      "loss": 0.1983,
+      "step": 4216
+    },
+    {
+      "epoch": 1.996803788102989,
+      "grad_norm": 1.170721173286438,
+      "learning_rate": 5.3056355801470114e-06,
+      "loss": 0.2015,
+      "step": 4217
+    },
+    {
+      "epoch": 1.9972773009766203,
+      "grad_norm": 1.126014232635498,
+      "learning_rate": 5.301120614974378e-06,
+      "loss": 0.1832,
+      "step": 4218
+    },
+    {
+      "epoch": 1.9977508138502516,
+      "grad_norm": 0.9776692390441895,
+      "learning_rate": 5.296606878748313e-06,
+      "loss": 0.1971,
+      "step": 4219
+    },
+    {
+      "epoch": 1.9982243267238828,
+      "grad_norm": 1.0567349195480347,
+      "learning_rate": 5.292094372649348e-06,
+      "loss": 0.2223,
+      "step": 4220
+    },
+    {
+      "epoch": 1.9986978395975141,
+      "grad_norm": 1.1440393924713135,
+      "learning_rate": 5.287583097857682e-06,
+      "loss": 0.2532,
+      "step": 4221
+    },
+    {
+      "epoch": 1.9991713524711452,
+      "grad_norm": 0.9647483229637146,
+      "learning_rate": 5.283073055553191e-06,
+      "loss": 0.2087,
+      "step": 4222
+    },
+    {
+      "epoch": 1.9996448653447767,
+      "grad_norm": 1.0528312921524048,
+      "learning_rate": 5.278564246915441e-06,
+      "loss": 0.2172,
+      "step": 4223
+    },
+    {
+      "epoch": 2.0001183782184078,
+      "grad_norm": 1.2826507091522217,
+      "learning_rate": 5.27405667312366e-06,
+      "loss": 0.1835,
+      "step": 4224
+    },
+    {
+      "epoch": 2.0005918910920393,
+      "grad_norm": 2.0795340538024902,
+      "learning_rate": 5.269550335356769e-06,
+      "loss": 0.1905,
+      "step": 4225
+    },
+    {
+      "epoch": 2.0010654039656703,
+      "grad_norm": 1.348443627357483,
+      "learning_rate": 5.265045234793348e-06,
+      "loss": 0.1877,
+      "step": 4226
+    },
+    {
+      "epoch": 2.0015389168393014,
+      "grad_norm": 0.9791116118431091,
+      "learning_rate": 5.260541372611669e-06,
+      "loss": 0.2081,
+      "step": 4227
+    },
+    {
+      "epoch": 2.002012429712933,
+      "grad_norm": 0.9848561882972717,
+      "learning_rate": 5.256038749989671e-06,
+      "loss": 0.18,
+      "step": 4228
+    },
+    {
+      "epoch": 2.002485942586564,
+      "grad_norm": 1.2935410737991333,
+      "learning_rate": 5.251537368104974e-06,
+      "loss": 0.1972,
+      "step": 4229
+    },
+    {
+      "epoch": 2.0029594554601955,
+      "grad_norm": 1.8192050457000732,
+      "learning_rate": 5.2470372281348695e-06,
+      "loss": 0.2058,
+      "step": 4230
+    },
+    {
+      "epoch": 2.0034329683338266,
+      "grad_norm": 1.1813068389892578,
+      "learning_rate": 5.242538331256322e-06,
+      "loss": 0.2309,
+      "step": 4231
+    },
+    {
+      "epoch": 2.0039064812074576,
+      "grad_norm": 1.1924341917037964,
+      "learning_rate": 5.238040678645981e-06,
+      "loss": 0.1972,
+      "step": 4232
+    },
+    {
+      "epoch": 2.004379994081089,
+      "grad_norm": 1.2242668867111206,
+      "learning_rate": 5.233544271480158e-06,
+      "loss": 0.1799,
+      "step": 4233
+    },
+    {
+      "epoch": 2.00485350695472,
+      "grad_norm": 1.3374933004379272,
+      "learning_rate": 5.22904911093485e-06,
+      "loss": 0.2176,
+      "step": 4234
+    },
+    {
+      "epoch": 2.0053270198283517,
+      "grad_norm": 1.1583667993545532,
+      "learning_rate": 5.224555198185719e-06,
+      "loss": 0.1999,
+      "step": 4235
+    },
+    {
+      "epoch": 2.0058005327019828,
+      "grad_norm": 1.0789278745651245,
+      "learning_rate": 5.220062534408109e-06,
+      "loss": 0.2103,
+      "step": 4236
+    },
+    {
+      "epoch": 2.0062740455756143,
+      "grad_norm": 1.0617800951004028,
+      "learning_rate": 5.215571120777027e-06,
+      "loss": 0.1883,
+      "step": 4237
+    },
+    {
+      "epoch": 2.0067475584492453,
+      "grad_norm": 1.34521484375,
+      "learning_rate": 5.211080958467166e-06,
+      "loss": 0.2037,
+      "step": 4238
+    },
+    {
+      "epoch": 2.0072210713228764,
+      "grad_norm": 1.116936445236206,
+      "learning_rate": 5.206592048652876e-06,
+      "loss": 0.1971,
+      "step": 4239
+    },
+    {
+      "epoch": 2.007694584196508,
+      "grad_norm": 2.06673264503479,
+      "learning_rate": 5.202104392508198e-06,
+      "loss": 0.2102,
+      "step": 4240
+    },
+    {
+      "epoch": 2.008168097070139,
+      "grad_norm": 0.9581062197685242,
+      "learning_rate": 5.19761799120683e-06,
+      "loss": 0.214,
+      "step": 4241
+    },
+    {
+      "epoch": 2.0086416099437705,
+      "grad_norm": 1.1789650917053223,
+      "learning_rate": 5.193132845922143e-06,
+      "loss": 0.2078,
+      "step": 4242
+    },
+    {
+      "epoch": 2.0091151228174016,
+      "grad_norm": 1.1819120645523071,
+      "learning_rate": 5.188648957827191e-06,
+      "loss": 0.1923,
+      "step": 4243
+    },
+    {
+      "epoch": 2.0095886356910326,
+      "grad_norm": 2.0509867668151855,
+      "learning_rate": 5.184166328094684e-06,
+      "loss": 0.2147,
+      "step": 4244
+    },
+    {
+      "epoch": 2.010062148564664,
+      "grad_norm": 1.0577517747879028,
+      "learning_rate": 5.17968495789702e-06,
+      "loss": 0.2028,
+      "step": 4245
+    },
+    {
+      "epoch": 2.010535661438295,
+      "grad_norm": 1.2950037717819214,
+      "learning_rate": 5.175204848406248e-06,
+      "loss": 0.1884,
+      "step": 4246
+    },
+    {
+      "epoch": 2.0110091743119267,
+      "grad_norm": 1.7854464054107666,
+      "learning_rate": 5.170726000794105e-06,
+      "loss": 0.1976,
+      "step": 4247
+    },
+    {
+      "epoch": 2.011482687185558,
+      "grad_norm": 1.12384831905365,
+      "learning_rate": 5.166248416231985e-06,
+      "loss": 0.1866,
+      "step": 4248
+    },
+    {
+      "epoch": 2.0119562000591893,
+      "grad_norm": 1.3171573877334595,
+      "learning_rate": 5.161772095890963e-06,
+      "loss": 0.1952,
+      "step": 4249
+    },
+    {
+      "epoch": 2.0124297129328204,
+      "grad_norm": 1.4037394523620605,
+      "learning_rate": 5.157297040941775e-06,
+      "loss": 0.1977,
+      "step": 4250
+    },
+    {
+      "epoch": 2.0129032258064514,
+      "grad_norm": 1.2727725505828857,
+      "learning_rate": 5.152823252554824e-06,
+      "loss": 0.2269,
+      "step": 4251
+    },
+    {
+      "epoch": 2.013376738680083,
+      "grad_norm": 1.1373170614242554,
+      "learning_rate": 5.1483507319001925e-06,
+      "loss": 0.2105,
+      "step": 4252
+    },
+    {
+      "epoch": 2.013850251553714,
+      "grad_norm": 1.0914908647537231,
+      "learning_rate": 5.143879480147616e-06,
+      "loss": 0.225,
+      "step": 4253
+    },
+    {
+      "epoch": 2.0143237644273455,
+      "grad_norm": 1.7061387300491333,
+      "learning_rate": 5.1394094984665185e-06,
+      "loss": 0.1815,
+      "step": 4254
+    },
+    {
+      "epoch": 2.0147972773009766,
+      "grad_norm": 1.8330706357955933,
+      "learning_rate": 5.134940788025978e-06,
+      "loss": 0.1946,
+      "step": 4255
+    },
+    {
+      "epoch": 2.015270790174608,
+      "grad_norm": 1.9821844100952148,
+      "learning_rate": 5.130473349994737e-06,
+      "loss": 0.2131,
+      "step": 4256
+    },
+    {
+      "epoch": 2.015744303048239,
+      "grad_norm": 1.2151135206222534,
+      "learning_rate": 5.1260071855412175e-06,
+      "loss": 0.1673,
+      "step": 4257
+    },
+    {
+      "epoch": 2.01621781592187,
+      "grad_norm": 1.4756810665130615,
+      "learning_rate": 5.121542295833493e-06,
+      "loss": 0.2035,
+      "step": 4258
+    },
+    {
+      "epoch": 2.0166913287955017,
+      "grad_norm": 1.570881724357605,
+      "learning_rate": 5.117078682039323e-06,
+      "loss": 0.2112,
+      "step": 4259
+    },
+    {
+      "epoch": 2.017164841669133,
+      "grad_norm": 1.0294634103775024,
+      "learning_rate": 5.112616345326114e-06,
+      "loss": 0.2062,
+      "step": 4260
+    },
+    {
+      "epoch": 2.0176383545427643,
+      "grad_norm": 0.987274169921875,
+      "learning_rate": 5.108155286860953e-06,
+      "loss": 0.2028,
+      "step": 4261
+    },
+    {
+      "epoch": 2.0181118674163954,
+      "grad_norm": 1.121518611907959,
+      "learning_rate": 5.10369550781058e-06,
+      "loss": 0.2018,
+      "step": 4262
+    },
+    {
+      "epoch": 2.0185853802900264,
+      "grad_norm": 0.953002393245697,
+      "learning_rate": 5.099237009341417e-06,
+      "loss": 0.1732,
+      "step": 4263
+    },
+    {
+      "epoch": 2.019058893163658,
+      "grad_norm": 0.955437183380127,
+      "learning_rate": 5.094779792619531e-06,
+      "loss": 0.1911,
+      "step": 4264
+    },
+    {
+      "epoch": 2.019532406037289,
+      "grad_norm": 1.8718898296356201,
+      "learning_rate": 5.0903238588106725e-06,
+      "loss": 0.1966,
+      "step": 4265
+    },
+    {
+      "epoch": 2.0200059189109205,
+      "grad_norm": 1.4193190336227417,
+      "learning_rate": 5.085869209080246e-06,
+      "loss": 0.2138,
+      "step": 4266
+    },
+    {
+      "epoch": 2.0204794317845516,
+      "grad_norm": 1.1836130619049072,
+      "learning_rate": 5.081415844593314e-06,
+      "loss": 0.1817,
+      "step": 4267
+    },
+    {
+      "epoch": 2.020952944658183,
+      "grad_norm": 2.136206865310669,
+      "learning_rate": 5.076963766514622e-06,
+      "loss": 0.21,
+      "step": 4268
+    },
+    {
+      "epoch": 2.021426457531814,
+      "grad_norm": 1.0202592611312866,
+      "learning_rate": 5.072512976008559e-06,
+      "loss": 0.223,
+      "step": 4269
+    },
+    {
+      "epoch": 2.0218999704054452,
+      "grad_norm": 1.1156057119369507,
+      "learning_rate": 5.068063474239195e-06,
+      "loss": 0.2055,
+      "step": 4270
+    },
+    {
+      "epoch": 2.0223734832790767,
+      "grad_norm": 1.3342862129211426,
+      "learning_rate": 5.063615262370247e-06,
+      "loss": 0.1872,
+      "step": 4271
+    },
+    {
+      "epoch": 2.022846996152708,
+      "grad_norm": 1.031692624092102,
+      "learning_rate": 5.059168341565109e-06,
+      "loss": 0.2132,
+      "step": 4272
+    },
+    {
+      "epoch": 2.0233205090263393,
+      "grad_norm": 1.07035493850708,
+      "learning_rate": 5.0547227129868225e-06,
+      "loss": 0.2238,
+      "step": 4273
+    },
+    {
+      "epoch": 2.0237940218999704,
+      "grad_norm": 1.6974101066589355,
+      "learning_rate": 5.050278377798105e-06,
+      "loss": 0.194,
+      "step": 4274
+    },
+    {
+      "epoch": 2.0242675347736014,
+      "grad_norm": 1.17416512966156,
+      "learning_rate": 5.045835337161329e-06,
+      "loss": 0.2137,
+      "step": 4275
+    },
+    {
+      "epoch": 2.024741047647233,
+      "grad_norm": 1.0354957580566406,
+      "learning_rate": 5.041393592238521e-06,
+      "loss": 0.2022,
+      "step": 4276
+    },
+    {
+      "epoch": 2.025214560520864,
+      "grad_norm": 1.328599452972412,
+      "learning_rate": 5.036953144191388e-06,
+      "loss": 0.2116,
+      "step": 4277
+    },
+    {
+      "epoch": 2.0256880733944955,
+      "grad_norm": 0.9613640308380127,
+      "learning_rate": 5.032513994181276e-06,
+      "loss": 0.1968,
+      "step": 4278
+    },
+    {
+      "epoch": 2.0261615862681266,
+      "grad_norm": 1.4717971086502075,
+      "learning_rate": 5.0280761433692095e-06,
+      "loss": 0.2161,
+      "step": 4279
+    },
+    {
+      "epoch": 2.026635099141758,
+      "grad_norm": 1.2015409469604492,
+      "learning_rate": 5.023639592915861e-06,
+      "loss": 0.2002,
+      "step": 4280
+    },
+    {
+      "epoch": 2.027108612015389,
+      "grad_norm": 0.9941813945770264,
+      "learning_rate": 5.019204343981572e-06,
+      "loss": 0.1973,
+      "step": 4281
+    },
+    {
+      "epoch": 2.0275821248890202,
+      "grad_norm": 1.1398041248321533,
+      "learning_rate": 5.0147703977263315e-06,
+      "loss": 0.2201,
+      "step": 4282
+    },
+    {
+      "epoch": 2.0280556377626517,
+      "grad_norm": 1.4286599159240723,
+      "learning_rate": 5.010337755309804e-06,
+      "loss": 0.1974,
+      "step": 4283
+    },
+    {
+      "epoch": 2.028529150636283,
+      "grad_norm": 1.7072778940200806,
+      "learning_rate": 5.005906417891302e-06,
+      "loss": 0.204,
+      "step": 4284
+    },
+    {
+      "epoch": 2.0290026635099143,
+      "grad_norm": 1.2885264158248901,
+      "learning_rate": 5.0014763866297935e-06,
+      "loss": 0.2257,
+      "step": 4285
+    },
+    {
+      "epoch": 2.0294761763835454,
+      "grad_norm": 1.2946069240570068,
+      "learning_rate": 4.997047662683917e-06,
+      "loss": 0.2029,
+      "step": 4286
+    },
+    {
+      "epoch": 2.029949689257177,
+      "grad_norm": 1.0730814933776855,
+      "learning_rate": 4.992620247211957e-06,
+      "loss": 0.208,
+      "step": 4287
+    },
+    {
+      "epoch": 2.030423202130808,
+      "grad_norm": 1.4278727769851685,
+      "learning_rate": 4.988194141371868e-06,
+      "loss": 0.1989,
+      "step": 4288
+    },
+    {
+      "epoch": 2.030896715004439,
+      "grad_norm": 0.9635195136070251,
+      "learning_rate": 4.9837693463212474e-06,
+      "loss": 0.2119,
+      "step": 4289
+    },
+    {
+      "epoch": 2.0313702278780705,
+      "grad_norm": 1.49040949344635,
+      "learning_rate": 4.979345863217366e-06,
+      "loss": 0.1987,
+      "step": 4290
+    },
+    {
+      "epoch": 2.0318437407517016,
+      "grad_norm": 1.3482121229171753,
+      "learning_rate": 4.974923693217135e-06,
+      "loss": 0.2184,
+      "step": 4291
+    },
+    {
+      "epoch": 2.032317253625333,
+      "grad_norm": 1.1210942268371582,
+      "learning_rate": 4.970502837477135e-06,
+      "loss": 0.211,
+      "step": 4292
+    },
+    {
+      "epoch": 2.032790766498964,
+      "grad_norm": 1.4026209115982056,
+      "learning_rate": 4.9660832971536014e-06,
+      "loss": 0.1975,
+      "step": 4293
+    },
+    {
+      "epoch": 2.0332642793725952,
+      "grad_norm": 1.885514259338379,
+      "learning_rate": 4.961665073402413e-06,
+      "loss": 0.2031,
+      "step": 4294
+    },
+    {
+      "epoch": 2.0337377922462267,
+      "grad_norm": 1.712158441543579,
+      "learning_rate": 4.957248167379124e-06,
+      "loss": 0.202,
+      "step": 4295
+    },
+    {
+      "epoch": 2.034211305119858,
+      "grad_norm": 1.4192404747009277,
+      "learning_rate": 4.9528325802389246e-06,
+      "loss": 0.2093,
+      "step": 4296
+    },
+    {
+      "epoch": 2.0346848179934893,
+      "grad_norm": 1.8846938610076904,
+      "learning_rate": 4.948418313136676e-06,
+      "loss": 0.2055,
+      "step": 4297
+    },
+    {
+      "epoch": 2.0351583308671204,
+      "grad_norm": 1.1656869649887085,
+      "learning_rate": 4.94400536722688e-06,
+      "loss": 0.2003,
+      "step": 4298
+    },
+    {
+      "epoch": 2.035631843740752,
+      "grad_norm": 1.072436809539795,
+      "learning_rate": 4.9395937436637096e-06,
+      "loss": 0.1975,
+      "step": 4299
+    },
+    {
+      "epoch": 2.036105356614383,
+      "grad_norm": 1.109155297279358,
+      "learning_rate": 4.935183443600976e-06,
+      "loss": 0.191,
+      "step": 4300
+    },
+    {
+      "epoch": 2.036578869488014,
+      "grad_norm": 1.9992730617523193,
+      "learning_rate": 4.930774468192147e-06,
+      "loss": 0.2152,
+      "step": 4301
+    },
+    {
+      "epoch": 2.0370523823616455,
+      "grad_norm": 1.2080824375152588,
+      "learning_rate": 4.926366818590358e-06,
+      "loss": 0.2105,
+      "step": 4302
+    },
+    {
+      "epoch": 2.0375258952352766,
+      "grad_norm": 1.5259896516799927,
+      "learning_rate": 4.921960495948377e-06,
+      "loss": 0.2042,
+      "step": 4303
+    },
+    {
+      "epoch": 2.037999408108908,
+      "grad_norm": 1.8273221254348755,
+      "learning_rate": 4.917555501418643e-06,
+      "loss": 0.2145,
+      "step": 4304
+    },
+    {
+      "epoch": 2.038472920982539,
+      "grad_norm": 1.103759765625,
+      "learning_rate": 4.913151836153232e-06,
+      "loss": 0.1831,
+      "step": 4305
+    },
+    {
+      "epoch": 2.0389464338561702,
+      "grad_norm": 1.3072751760482788,
+      "learning_rate": 4.908749501303889e-06,
+      "loss": 0.1974,
+      "step": 4306
+    },
+    {
+      "epoch": 2.0394199467298018,
+      "grad_norm": 1.1292860507965088,
+      "learning_rate": 4.904348498021993e-06,
+      "loss": 0.1865,
+      "step": 4307
+    },
+    {
+      "epoch": 2.039893459603433,
+      "grad_norm": 1.2447689771652222,
+      "learning_rate": 4.8999488274585935e-06,
+      "loss": 0.2132,
+      "step": 4308
+    },
+    {
+      "epoch": 2.0403669724770643,
+      "grad_norm": 0.8923916816711426,
+      "learning_rate": 4.895550490764377e-06,
+      "loss": 0.1893,
+      "step": 4309
+    },
+    {
+      "epoch": 2.0408404853506954,
+      "grad_norm": 1.4532053470611572,
+      "learning_rate": 4.891153489089681e-06,
+      "loss": 0.2084,
+      "step": 4310
+    },
+    {
+      "epoch": 2.041313998224327,
+      "grad_norm": 1.329358696937561,
+      "learning_rate": 4.886757823584511e-06,
+      "loss": 0.1988,
+      "step": 4311
+    },
+    {
+      "epoch": 2.041787511097958,
+      "grad_norm": 1.5044658184051514,
+      "learning_rate": 4.882363495398498e-06,
+      "loss": 0.1831,
+      "step": 4312
+    },
+    {
+      "epoch": 2.042261023971589,
+      "grad_norm": 1.3755611181259155,
+      "learning_rate": 4.877970505680946e-06,
+      "loss": 0.2147,
+      "step": 4313
+    },
+    {
+      "epoch": 2.0427345368452205,
+      "grad_norm": 0.9415785074234009,
+      "learning_rate": 4.8735788555807905e-06,
+      "loss": 0.1909,
+      "step": 4314
+    },
+    {
+      "epoch": 2.0432080497188516,
+      "grad_norm": 1.1154485940933228,
+      "learning_rate": 4.8691885462466345e-06,
+      "loss": 0.2137,
+      "step": 4315
+    },
+    {
+      "epoch": 2.043681562592483,
+      "grad_norm": 1.1781600713729858,
+      "learning_rate": 4.864799578826713e-06,
+      "loss": 0.2118,
+      "step": 4316
+    },
+    {
+      "epoch": 2.044155075466114,
+      "grad_norm": 1.8692867755889893,
+      "learning_rate": 4.860411954468925e-06,
+      "loss": 0.2269,
+      "step": 4317
+    },
+    {
+      "epoch": 2.0446285883397457,
+      "grad_norm": 1.1391620635986328,
+      "learning_rate": 4.856025674320803e-06,
+      "loss": 0.199,
+      "step": 4318
+    },
+    {
+      "epoch": 2.0451021012133768,
+      "grad_norm": 1.0855448246002197,
+      "learning_rate": 4.851640739529547e-06,
+      "loss": 0.191,
+      "step": 4319
+    },
+    {
+      "epoch": 2.045575614087008,
+      "grad_norm": 1.8946772813796997,
+      "learning_rate": 4.847257151241987e-06,
+      "loss": 0.2126,
+      "step": 4320
+    },
+    {
+      "epoch": 2.0460491269606393,
+      "grad_norm": 1.7969681024551392,
+      "learning_rate": 4.842874910604606e-06,
+      "loss": 0.2058,
+      "step": 4321
+    },
+    {
+      "epoch": 2.0465226398342704,
+      "grad_norm": 1.028253436088562,
+      "learning_rate": 4.838494018763546e-06,
+      "loss": 0.2044,
+      "step": 4322
+    },
+    {
+      "epoch": 2.046996152707902,
+      "grad_norm": 1.3800286054611206,
+      "learning_rate": 4.8341144768645754e-06,
+      "loss": 0.2053,
+      "step": 4323
+    },
+    {
+      "epoch": 2.047469665581533,
+      "grad_norm": 0.9476808905601501,
+      "learning_rate": 4.829736286053131e-06,
+      "loss": 0.2135,
+      "step": 4324
+    },
+    {
+      "epoch": 2.047943178455164,
+      "grad_norm": 1.1911259889602661,
+      "learning_rate": 4.82535944747428e-06,
+      "loss": 0.2038,
+      "step": 4325
+    },
+    {
+      "epoch": 2.0484166913287956,
+      "grad_norm": 0.8657131791114807,
+      "learning_rate": 4.820983962272748e-06,
+      "loss": 0.1959,
+      "step": 4326
+    },
+    {
+      "epoch": 2.0488902042024266,
+      "grad_norm": 2.2093346118927,
+      "learning_rate": 4.816609831592895e-06,
+      "loss": 0.208,
+      "step": 4327
+    },
+    {
+      "epoch": 2.049363717076058,
+      "grad_norm": 1.8499171733856201,
+      "learning_rate": 4.812237056578738e-06,
+      "loss": 0.206,
+      "step": 4328
+    },
+    {
+      "epoch": 2.049837229949689,
+      "grad_norm": 1.404910683631897,
+      "learning_rate": 4.807865638373932e-06,
+      "loss": 0.1919,
+      "step": 4329
+    },
+    {
+      "epoch": 2.0503107428233207,
+      "grad_norm": 0.8892547488212585,
+      "learning_rate": 4.803495578121775e-06,
+      "loss": 0.1778,
+      "step": 4330
+    },
+    {
+      "epoch": 2.0507842556969518,
+      "grad_norm": 1.3549368381500244,
+      "learning_rate": 4.799126876965219e-06,
+      "loss": 0.202,
+      "step": 4331
+    },
+    {
+      "epoch": 2.051257768570583,
+      "grad_norm": 1.4522606134414673,
+      "learning_rate": 4.794759536046854e-06,
+      "loss": 0.1813,
+      "step": 4332
+    },
+    {
+      "epoch": 2.0517312814442143,
+      "grad_norm": 1.1356416940689087,
+      "learning_rate": 4.790393556508918e-06,
+      "loss": 0.1793,
+      "step": 4333
+    },
+    {
+      "epoch": 2.0522047943178454,
+      "grad_norm": 1.0403809547424316,
+      "learning_rate": 4.786028939493292e-06,
+      "loss": 0.2073,
+      "step": 4334
+    },
+    {
+      "epoch": 2.052678307191477,
+      "grad_norm": 1.0235661268234253,
+      "learning_rate": 4.781665686141493e-06,
+      "loss": 0.1865,
+      "step": 4335
+    },
+    {
+      "epoch": 2.053151820065108,
+      "grad_norm": 1.1404122114181519,
+      "learning_rate": 4.777303797594694e-06,
+      "loss": 0.1997,
+      "step": 4336
+    },
+    {
+      "epoch": 2.053625332938739,
+      "grad_norm": 1.2667036056518555,
+      "learning_rate": 4.772943274993701e-06,
+      "loss": 0.2054,
+      "step": 4337
+    },
+    {
+      "epoch": 2.0540988458123706,
+      "grad_norm": 1.177870273590088,
+      "learning_rate": 4.768584119478971e-06,
+      "loss": 0.2103,
+      "step": 4338
+    },
+    {
+      "epoch": 2.0545723586860016,
+      "grad_norm": 0.9611978530883789,
+      "learning_rate": 4.764226332190591e-06,
+      "loss": 0.203,
+      "step": 4339
+    },
+    {
+      "epoch": 2.055045871559633,
+      "grad_norm": 1.3527143001556396,
+      "learning_rate": 4.759869914268308e-06,
+      "loss": 0.2133,
+      "step": 4340
+    },
+    {
+      "epoch": 2.055519384433264,
+      "grad_norm": 0.833387017250061,
+      "learning_rate": 4.7555148668514925e-06,
+      "loss": 0.1858,
+      "step": 4341
+    },
+    {
+      "epoch": 2.0559928973068957,
+      "grad_norm": 1.0796325206756592,
+      "learning_rate": 4.751161191079173e-06,
+      "loss": 0.21,
+      "step": 4342
+    },
+    {
+      "epoch": 2.056466410180527,
+      "grad_norm": 0.8967043161392212,
+      "learning_rate": 4.746808888090004e-06,
+      "loss": 0.1963,
+      "step": 4343
+    },
+    {
+      "epoch": 2.056939923054158,
+      "grad_norm": 1.2309699058532715,
+      "learning_rate": 4.742457959022296e-06,
+      "loss": 0.2024,
+      "step": 4344
+    },
+    {
+      "epoch": 2.0574134359277894,
+      "grad_norm": 1.511290431022644,
+      "learning_rate": 4.738108405013988e-06,
+      "loss": 0.2285,
+      "step": 4345
+    },
+    {
+      "epoch": 2.0578869488014204,
+      "grad_norm": 2.3464651107788086,
+      "learning_rate": 4.73376022720266e-06,
+      "loss": 0.2228,
+      "step": 4346
+    },
+    {
+      "epoch": 2.058360461675052,
+      "grad_norm": 1.3974965810775757,
+      "learning_rate": 4.729413426725546e-06,
+      "loss": 0.2022,
+      "step": 4347
+    },
+    {
+      "epoch": 2.058833974548683,
+      "grad_norm": 1.8854213953018188,
+      "learning_rate": 4.725068004719499e-06,
+      "loss": 0.1891,
+      "step": 4348
+    },
+    {
+      "epoch": 2.0593074874223145,
+      "grad_norm": 1.3588758707046509,
+      "learning_rate": 4.72072396232103e-06,
+      "loss": 0.2084,
+      "step": 4349
+    },
+    {
+      "epoch": 2.0597810002959456,
+      "grad_norm": 1.5690404176712036,
+      "learning_rate": 4.716381300666275e-06,
+      "loss": 0.1923,
+      "step": 4350
+    },
+    {
+      "epoch": 2.0602545131695766,
+      "grad_norm": 1.721727967262268,
+      "learning_rate": 4.7120400208910235e-06,
+      "loss": 0.1893,
+      "step": 4351
+    },
+    {
+      "epoch": 2.060728026043208,
+      "grad_norm": 1.5722174644470215,
+      "learning_rate": 4.707700124130686e-06,
+      "loss": 0.1797,
+      "step": 4352
+    },
+    {
+      "epoch": 2.061201538916839,
+      "grad_norm": 1.4766589403152466,
+      "learning_rate": 4.703361611520331e-06,
+      "loss": 0.2187,
+      "step": 4353
+    },
+    {
+      "epoch": 2.0616750517904707,
+      "grad_norm": 1.067948341369629,
+      "learning_rate": 4.699024484194648e-06,
+      "loss": 0.2299,
+      "step": 4354
+    },
+    {
+      "epoch": 2.062148564664102,
+      "grad_norm": 1.0318337678909302,
+      "learning_rate": 4.694688743287966e-06,
+      "loss": 0.2086,
+      "step": 4355
+    },
+    {
+      "epoch": 2.062622077537733,
+      "grad_norm": 1.2142385244369507,
+      "learning_rate": 4.6903543899342685e-06,
+      "loss": 0.2005,
+      "step": 4356
+    },
+    {
+      "epoch": 2.0630955904113644,
+      "grad_norm": 1.0302040576934814,
+      "learning_rate": 4.686021425267152e-06,
+      "loss": 0.2019,
+      "step": 4357
+    },
+    {
+      "epoch": 2.0635691032849954,
+      "grad_norm": 0.8725736737251282,
+      "learning_rate": 4.681689850419871e-06,
+      "loss": 0.1938,
+      "step": 4358
+    },
+    {
+      "epoch": 2.064042616158627,
+      "grad_norm": 0.9284090995788574,
+      "learning_rate": 4.677359666525299e-06,
+      "loss": 0.1988,
+      "step": 4359
+    },
+    {
+      "epoch": 2.064516129032258,
+      "grad_norm": 1.0550601482391357,
+      "learning_rate": 4.673030874715961e-06,
+      "loss": 0.2268,
+      "step": 4360
+    },
+    {
+      "epoch": 2.0649896419058895,
+      "grad_norm": 2.1118485927581787,
+      "learning_rate": 4.668703476124005e-06,
+      "loss": 0.2087,
+      "step": 4361
+    },
+    {
+      "epoch": 2.0654631547795206,
+      "grad_norm": 1.2798197269439697,
+      "learning_rate": 4.664377471881226e-06,
+      "loss": 0.2028,
+      "step": 4362
+    },
+    {
+      "epoch": 2.0659366676531516,
+      "grad_norm": 1.4490396976470947,
+      "learning_rate": 4.660052863119046e-06,
+      "loss": 0.2108,
+      "step": 4363
+    },
+    {
+      "epoch": 2.066410180526783,
+      "grad_norm": 1.0242908000946045,
+      "learning_rate": 4.65572965096852e-06,
+      "loss": 0.1924,
+      "step": 4364
+    },
+    {
+      "epoch": 2.0668836934004142,
+      "grad_norm": 1.2563238143920898,
+      "learning_rate": 4.651407836560351e-06,
+      "loss": 0.1882,
+      "step": 4365
+    },
+    {
+      "epoch": 2.0673572062740457,
+      "grad_norm": 1.3273645639419556,
+      "learning_rate": 4.6470874210248595e-06,
+      "loss": 0.2313,
+      "step": 4366
+    },
+    {
+      "epoch": 2.067830719147677,
+      "grad_norm": 2.203810453414917,
+      "learning_rate": 4.642768405492016e-06,
+      "loss": 0.206,
+      "step": 4367
+    },
+    {
+      "epoch": 2.068304232021308,
+      "grad_norm": 2.2822744846343994,
+      "learning_rate": 4.63845079109141e-06,
+      "loss": 0.2121,
+      "step": 4368
+    },
+    {
+      "epoch": 2.0687777448949394,
+      "grad_norm": 1.70283043384552,
+      "learning_rate": 4.63413457895228e-06,
+      "loss": 0.1918,
+      "step": 4369
+    },
+    {
+      "epoch": 2.0692512577685704,
+      "grad_norm": 1.6546306610107422,
+      "learning_rate": 4.629819770203482e-06,
+      "loss": 0.2086,
+      "step": 4370
+    },
+    {
+      "epoch": 2.069724770642202,
+      "grad_norm": 1.611094355583191,
+      "learning_rate": 4.625506365973515e-06,
+      "loss": 0.2005,
+      "step": 4371
+    },
+    {
+      "epoch": 2.070198283515833,
+      "grad_norm": 1.6032445430755615,
+      "learning_rate": 4.621194367390515e-06,
+      "loss": 0.1965,
+      "step": 4372
+    },
+    {
+      "epoch": 2.0706717963894645,
+      "grad_norm": 1.065320372581482,
+      "learning_rate": 4.6168837755822326e-06,
+      "loss": 0.213,
+      "step": 4373
+    },
+    {
+      "epoch": 2.0711453092630956,
+      "grad_norm": 1.1174391508102417,
+      "learning_rate": 4.612574591676071e-06,
+      "loss": 0.1863,
+      "step": 4374
+    },
+    {
+      "epoch": 2.0716188221367267,
+      "grad_norm": 1.169047236442566,
+      "learning_rate": 4.608266816799049e-06,
+      "loss": 0.2003,
+      "step": 4375
+    },
+    {
+      "epoch": 2.072092335010358,
+      "grad_norm": 1.628919005393982,
+      "learning_rate": 4.603960452077828e-06,
+      "loss": 0.21,
+      "step": 4376
+    },
+    {
+      "epoch": 2.0725658478839892,
+      "grad_norm": 1.1893731355667114,
+      "learning_rate": 4.599655498638691e-06,
+      "loss": 0.2092,
+      "step": 4377
+    },
+    {
+      "epoch": 2.0730393607576207,
+      "grad_norm": 1.3188472986221313,
+      "learning_rate": 4.595351957607564e-06,
+      "loss": 0.2084,
+      "step": 4378
+    },
+    {
+      "epoch": 2.073512873631252,
+      "grad_norm": 1.0750354528427124,
+      "learning_rate": 4.5910498301099935e-06,
+      "loss": 0.2228,
+      "step": 4379
+    },
+    {
+      "epoch": 2.073986386504883,
+      "grad_norm": 1.0731077194213867,
+      "learning_rate": 4.5867491172711546e-06,
+      "loss": 0.2001,
+      "step": 4380
+    },
+    {
+      "epoch": 2.0744598993785144,
+      "grad_norm": 0.9855295419692993,
+      "learning_rate": 4.582449820215865e-06,
+      "loss": 0.2041,
+      "step": 4381
+    },
+    {
+      "epoch": 2.0749334122521454,
+      "grad_norm": 1.3802636861801147,
+      "learning_rate": 4.578151940068558e-06,
+      "loss": 0.2164,
+      "step": 4382
+    },
+    {
+      "epoch": 2.075406925125777,
+      "grad_norm": 1.4265203475952148,
+      "learning_rate": 4.57385547795331e-06,
+      "loss": 0.2146,
+      "step": 4383
+    },
+    {
+      "epoch": 2.075880437999408,
+      "grad_norm": 1.3063853979110718,
+      "learning_rate": 4.569560434993809e-06,
+      "loss": 0.1938,
+      "step": 4384
+    },
+    {
+      "epoch": 2.0763539508730395,
+      "grad_norm": 1.087546467781067,
+      "learning_rate": 4.5652668123133925e-06,
+      "loss": 0.1883,
+      "step": 4385
+    },
+    {
+      "epoch": 2.0768274637466706,
+      "grad_norm": 1.0859131813049316,
+      "learning_rate": 4.560974611035007e-06,
+      "loss": 0.2257,
+      "step": 4386
+    },
+    {
+      "epoch": 2.0773009766203017,
+      "grad_norm": 1.2528398036956787,
+      "learning_rate": 4.556683832281246e-06,
+      "loss": 0.1905,
+      "step": 4387
+    },
+    {
+      "epoch": 2.077774489493933,
+      "grad_norm": 1.1462734937667847,
+      "learning_rate": 4.552394477174316e-06,
+      "loss": 0.191,
+      "step": 4388
+    },
+    {
+      "epoch": 2.0782480023675642,
+      "grad_norm": 1.486925721168518,
+      "learning_rate": 4.5481065468360515e-06,
+      "loss": 0.1935,
+      "step": 4389
+    },
+    {
+      "epoch": 2.0787215152411957,
+      "grad_norm": 1.0987601280212402,
+      "learning_rate": 4.5438200423879285e-06,
+      "loss": 0.213,
+      "step": 4390
+    },
+    {
+      "epoch": 2.079195028114827,
+      "grad_norm": 0.965355396270752,
+      "learning_rate": 4.539534964951033e-06,
+      "loss": 0.2278,
+      "step": 4391
+    },
+    {
+      "epoch": 2.0796685409884583,
+      "grad_norm": 1.4318578243255615,
+      "learning_rate": 4.535251315646093e-06,
+      "loss": 0.2028,
+      "step": 4392
+    },
+    {
+      "epoch": 2.0801420538620894,
+      "grad_norm": 1.0449013710021973,
+      "learning_rate": 4.53096909559345e-06,
+      "loss": 0.2047,
+      "step": 4393
+    },
+    {
+      "epoch": 2.0806155667357205,
+      "grad_norm": 2.591721296310425,
+      "learning_rate": 4.526688305913081e-06,
+      "loss": 0.2128,
+      "step": 4394
+    },
+    {
+      "epoch": 2.081089079609352,
+      "grad_norm": 1.750844955444336,
+      "learning_rate": 4.52240894772458e-06,
+      "loss": 0.2035,
+      "step": 4395
+    },
+    {
+      "epoch": 2.081562592482983,
+      "grad_norm": 1.5162577629089355,
+      "learning_rate": 4.51813102214718e-06,
+      "loss": 0.2092,
+      "step": 4396
+    },
+    {
+      "epoch": 2.0820361053566145,
+      "grad_norm": 2.782047986984253,
+      "learning_rate": 4.513854530299723e-06,
+      "loss": 0.1874,
+      "step": 4397
+    },
+    {
+      "epoch": 2.0825096182302456,
+      "grad_norm": 2.013601779937744,
+      "learning_rate": 4.50957947330069e-06,
+      "loss": 0.2174,
+      "step": 4398
+    },
+    {
+      "epoch": 2.0829831311038767,
+      "grad_norm": 1.1781798601150513,
+      "learning_rate": 4.50530585226818e-06,
+      "loss": 0.1953,
+      "step": 4399
+    },
+    {
+      "epoch": 2.083456643977508,
+      "grad_norm": 1.7047263383865356,
+      "learning_rate": 4.501033668319913e-06,
+      "loss": 0.202,
+      "step": 4400
+    },
+    {
+      "epoch": 2.0839301568511392,
+      "grad_norm": 1.3971747159957886,
+      "learning_rate": 4.496762922573244e-06,
+      "loss": 0.2062,
+      "step": 4401
+    },
+    {
+      "epoch": 2.0844036697247708,
+      "grad_norm": 1.4985466003417969,
+      "learning_rate": 4.492493616145137e-06,
+      "loss": 0.1795,
+      "step": 4402
+    },
+    {
+      "epoch": 2.084877182598402,
+      "grad_norm": 1.2991846799850464,
+      "learning_rate": 4.4882257501521975e-06,
+      "loss": 0.1858,
+      "step": 4403
+    },
+    {
+      "epoch": 2.0853506954720333,
+      "grad_norm": 1.194708228111267,
+      "learning_rate": 4.483959325710636e-06,
+      "loss": 0.1944,
+      "step": 4404
+    },
+    {
+      "epoch": 2.0858242083456644,
+      "grad_norm": 1.1843652725219727,
+      "learning_rate": 4.479694343936303e-06,
+      "loss": 0.2054,
+      "step": 4405
+    },
+    {
+      "epoch": 2.0862977212192955,
+      "grad_norm": 1.1956802606582642,
+      "learning_rate": 4.4754308059446546e-06,
+      "loss": 0.2074,
+      "step": 4406
+    },
+    {
+      "epoch": 2.086771234092927,
+      "grad_norm": 1.1742326021194458,
+      "learning_rate": 4.471168712850787e-06,
+      "loss": 0.2047,
+      "step": 4407
+    },
+    {
+      "epoch": 2.087244746966558,
+      "grad_norm": 1.7787115573883057,
+      "learning_rate": 4.466908065769404e-06,
+      "loss": 0.2243,
+      "step": 4408
+    },
+    {
+      "epoch": 2.0877182598401895,
+      "grad_norm": 1.4370203018188477,
+      "learning_rate": 4.46264886581483e-06,
+      "loss": 0.25,
+      "step": 4409
+    },
+    {
+      "epoch": 2.0881917727138206,
+      "grad_norm": 1.0106841325759888,
+      "learning_rate": 4.458391114101034e-06,
+      "loss": 0.2019,
+      "step": 4410
+    },
+    {
+      "epoch": 2.088665285587452,
+      "grad_norm": 0.9871804714202881,
+      "learning_rate": 4.454134811741577e-06,
+      "loss": 0.1807,
+      "step": 4411
+    },
+    {
+      "epoch": 2.089138798461083,
+      "grad_norm": 1.2592934370040894,
+      "learning_rate": 4.449879959849662e-06,
+      "loss": 0.215,
+      "step": 4412
+    },
+    {
+      "epoch": 2.0896123113347143,
+      "grad_norm": 1.557584285736084,
+      "learning_rate": 4.445626559538101e-06,
+      "loss": 0.1972,
+      "step": 4413
+    },
+    {
+      "epoch": 2.0900858242083458,
+      "grad_norm": 1.67927885055542,
+      "learning_rate": 4.4413746119193245e-06,
+      "loss": 0.1969,
+      "step": 4414
+    },
+    {
+      "epoch": 2.090559337081977,
+      "grad_norm": 1.2148725986480713,
+      "learning_rate": 4.437124118105397e-06,
+      "loss": 0.2133,
+      "step": 4415
+    },
+    {
+      "epoch": 2.0910328499556083,
+      "grad_norm": 1.0764367580413818,
+      "learning_rate": 4.4328750792079875e-06,
+      "loss": 0.2014,
+      "step": 4416
+    },
+    {
+      "epoch": 2.0915063628292394,
+      "grad_norm": 0.9037055373191833,
+      "learning_rate": 4.428627496338398e-06,
+      "loss": 0.1938,
+      "step": 4417
+    },
+    {
+      "epoch": 2.0919798757028705,
+      "grad_norm": 1.0375968217849731,
+      "learning_rate": 4.424381370607535e-06,
+      "loss": 0.2043,
+      "step": 4418
+    },
+    {
+      "epoch": 2.092453388576502,
+      "grad_norm": 1.5051581859588623,
+      "learning_rate": 4.420136703125938e-06,
+      "loss": 0.2058,
+      "step": 4419
+    },
+    {
+      "epoch": 2.092926901450133,
+      "grad_norm": 1.5662999153137207,
+      "learning_rate": 4.415893495003753e-06,
+      "loss": 0.2044,
+      "step": 4420
+    },
+    {
+      "epoch": 2.0934004143237646,
+      "grad_norm": 1.0691457986831665,
+      "learning_rate": 4.411651747350758e-06,
+      "loss": 0.197,
+      "step": 4421
+    },
+    {
+      "epoch": 2.0938739271973956,
+      "grad_norm": 1.773221731185913,
+      "learning_rate": 4.407411461276333e-06,
+      "loss": 0.1969,
+      "step": 4422
+    },
+    {
+      "epoch": 2.094347440071027,
+      "grad_norm": 1.03744637966156,
+      "learning_rate": 4.4031726378894915e-06,
+      "loss": 0.2462,
+      "step": 4423
+    },
+    {
+      "epoch": 2.094820952944658,
+      "grad_norm": 1.8114067316055298,
+      "learning_rate": 4.3989352782988525e-06,
+      "loss": 0.1918,
+      "step": 4424
+    },
+    {
+      "epoch": 2.0952944658182893,
+      "grad_norm": 0.8494598269462585,
+      "learning_rate": 4.394699383612653e-06,
+      "loss": 0.1993,
+      "step": 4425
+    },
+    {
+      "epoch": 2.0957679786919208,
+      "grad_norm": 1.0043613910675049,
+      "learning_rate": 4.390464954938759e-06,
+      "loss": 0.202,
+      "step": 4426
+    },
+    {
+      "epoch": 2.096241491565552,
+      "grad_norm": 1.178818941116333,
+      "learning_rate": 4.386231993384635e-06,
+      "loss": 0.1917,
+      "step": 4427
+    },
+    {
+      "epoch": 2.0967150044391833,
+      "grad_norm": 1.1056585311889648,
+      "learning_rate": 4.382000500057381e-06,
+      "loss": 0.1876,
+      "step": 4428
+    },
+    {
+      "epoch": 2.0971885173128144,
+      "grad_norm": 1.0591120719909668,
+      "learning_rate": 4.377770476063694e-06,
+      "loss": 0.2058,
+      "step": 4429
+    },
+    {
+      "epoch": 2.0976620301864455,
+      "grad_norm": 1.0878396034240723,
+      "learning_rate": 4.373541922509905e-06,
+      "loss": 0.1971,
+      "step": 4430
+    },
+    {
+      "epoch": 2.098135543060077,
+      "grad_norm": 1.087065577507019,
+      "learning_rate": 4.369314840501943e-06,
+      "loss": 0.2182,
+      "step": 4431
+    },
+    {
+      "epoch": 2.098609055933708,
+      "grad_norm": 1.2058556079864502,
+      "learning_rate": 4.365089231145367e-06,
+      "loss": 0.203,
+      "step": 4432
+    },
+    {
+      "epoch": 2.0990825688073396,
+      "grad_norm": 1.8395373821258545,
+      "learning_rate": 4.360865095545343e-06,
+      "loss": 0.1915,
+      "step": 4433
+    },
+    {
+      "epoch": 2.0995560816809706,
+      "grad_norm": 1.0700697898864746,
+      "learning_rate": 4.356642434806646e-06,
+      "loss": 0.2025,
+      "step": 4434
+    },
+    {
+      "epoch": 2.100029594554602,
+      "grad_norm": 1.068076491355896,
+      "learning_rate": 4.352421250033683e-06,
+      "loss": 0.1822,
+      "step": 4435
+    },
+    {
+      "epoch": 2.100503107428233,
+      "grad_norm": 1.1136810779571533,
+      "learning_rate": 4.348201542330455e-06,
+      "loss": 0.1865,
+      "step": 4436
+    },
+    {
+      "epoch": 2.1009766203018643,
+      "grad_norm": 1.1286970376968384,
+      "learning_rate": 4.3439833128005925e-06,
+      "loss": 0.1867,
+      "step": 4437
+    },
+    {
+      "epoch": 2.101450133175496,
+      "grad_norm": 1.4740508794784546,
+      "learning_rate": 4.339766562547326e-06,
+      "loss": 0.2043,
+      "step": 4438
+    },
+    {
+      "epoch": 2.101923646049127,
+      "grad_norm": 0.8717254996299744,
+      "learning_rate": 4.335551292673515e-06,
+      "loss": 0.1854,
+      "step": 4439
+    },
+    {
+      "epoch": 2.1023971589227584,
+      "grad_norm": 1.1837518215179443,
+      "learning_rate": 4.331337504281613e-06,
+      "loss": 0.183,
+      "step": 4440
+    },
+    {
+      "epoch": 2.1028706717963894,
+      "grad_norm": 1.649964690208435,
+      "learning_rate": 4.327125198473704e-06,
+      "loss": 0.2153,
+      "step": 4441
+    },
+    {
+      "epoch": 2.1033441846700205,
+      "grad_norm": 1.4881905317306519,
+      "learning_rate": 4.322914376351472e-06,
+      "loss": 0.2027,
+      "step": 4442
+    },
+    {
+      "epoch": 2.103817697543652,
+      "grad_norm": 1.6832579374313354,
+      "learning_rate": 4.318705039016215e-06,
+      "loss": 0.2197,
+      "step": 4443
+    },
+    {
+      "epoch": 2.104291210417283,
+      "grad_norm": 1.136296033859253,
+      "learning_rate": 4.314497187568848e-06,
+      "loss": 0.2201,
+      "step": 4444
+    },
+    {
+      "epoch": 2.1047647232909146,
+      "grad_norm": 0.9659789800643921,
+      "learning_rate": 4.310290823109889e-06,
+      "loss": 0.1773,
+      "step": 4445
+    },
+    {
+      "epoch": 2.1052382361645456,
+      "grad_norm": 1.2847449779510498,
+      "learning_rate": 4.306085946739481e-06,
+      "loss": 0.1805,
+      "step": 4446
+    },
+    {
+      "epoch": 2.105711749038177,
+      "grad_norm": 1.6592435836791992,
+      "learning_rate": 4.301882559557359e-06,
+      "loss": 0.1944,
+      "step": 4447
+    },
+    {
+      "epoch": 2.106185261911808,
+      "grad_norm": 1.5656230449676514,
+      "learning_rate": 4.297680662662882e-06,
+      "loss": 0.2252,
+      "step": 4448
+    },
+    {
+      "epoch": 2.1066587747854393,
+      "grad_norm": 1.2571629285812378,
+      "learning_rate": 4.293480257155022e-06,
+      "loss": 0.224,
+      "step": 4449
+    },
+    {
+      "epoch": 2.107132287659071,
+      "grad_norm": 1.8102307319641113,
+      "learning_rate": 4.289281344132344e-06,
+      "loss": 0.1945,
+      "step": 4450
+    },
+    {
+      "epoch": 2.107605800532702,
+      "grad_norm": 1.113694190979004,
+      "learning_rate": 4.285083924693041e-06,
+      "loss": 0.1983,
+      "step": 4451
+    },
+    {
+      "epoch": 2.1080793134063334,
+      "grad_norm": 1.0855345726013184,
+      "learning_rate": 4.280887999934902e-06,
+      "loss": 0.2136,
+      "step": 4452
+    },
+    {
+      "epoch": 2.1085528262799644,
+      "grad_norm": 1.6805202960968018,
+      "learning_rate": 4.276693570955337e-06,
+      "loss": 0.1893,
+      "step": 4453
+    },
+    {
+      "epoch": 2.109026339153596,
+      "grad_norm": 1.4205180406570435,
+      "learning_rate": 4.272500638851351e-06,
+      "loss": 0.2042,
+      "step": 4454
+    },
+    {
+      "epoch": 2.109499852027227,
+      "grad_norm": 1.77024245262146,
+      "learning_rate": 4.2683092047195725e-06,
+      "loss": 0.2161,
+      "step": 4455
+    },
+    {
+      "epoch": 2.109973364900858,
+      "grad_norm": 1.2872737646102905,
+      "learning_rate": 4.264119269656224e-06,
+      "loss": 0.199,
+      "step": 4456
+    },
+    {
+      "epoch": 2.1104468777744896,
+      "grad_norm": 1.008962631225586,
+      "learning_rate": 4.259930834757149e-06,
+      "loss": 0.2022,
+      "step": 4457
+    },
+    {
+      "epoch": 2.1109203906481206,
+      "grad_norm": 1.3520052433013916,
+      "learning_rate": 4.255743901117788e-06,
+      "loss": 0.2068,
+      "step": 4458
+    },
+    {
+      "epoch": 2.111393903521752,
+      "grad_norm": 1.200093388557434,
+      "learning_rate": 4.25155846983319e-06,
+      "loss": 0.2165,
+      "step": 4459
+    },
+    {
+      "epoch": 2.1118674163953832,
+      "grad_norm": 0.9952120780944824,
+      "learning_rate": 4.247374541998022e-06,
+      "loss": 0.2021,
+      "step": 4460
+    },
+    {
+      "epoch": 2.1123409292690143,
+      "grad_norm": 2.106011390686035,
+      "learning_rate": 4.243192118706543e-06,
+      "loss": 0.1916,
+      "step": 4461
+    },
+    {
+      "epoch": 2.112814442142646,
+      "grad_norm": 1.3868449926376343,
+      "learning_rate": 4.239011201052631e-06,
+      "loss": 0.2154,
+      "step": 4462
+    },
+    {
+      "epoch": 2.113287955016277,
+      "grad_norm": 1.1115097999572754,
+      "learning_rate": 4.234831790129759e-06,
+      "loss": 0.2044,
+      "step": 4463
+    },
+    {
+      "epoch": 2.1137614678899084,
+      "grad_norm": 1.1005358695983887,
+      "learning_rate": 4.2306538870310185e-06,
+      "loss": 0.2041,
+      "step": 4464
+    },
+    {
+      "epoch": 2.1142349807635394,
+      "grad_norm": 1.5084381103515625,
+      "learning_rate": 4.226477492849092e-06,
+      "loss": 0.2113,
+      "step": 4465
+    },
+    {
+      "epoch": 2.114708493637171,
+      "grad_norm": 1.4259034395217896,
+      "learning_rate": 4.222302608676283e-06,
+      "loss": 0.2041,
+      "step": 4466
+    },
+    {
+      "epoch": 2.115182006510802,
+      "grad_norm": 1.2917234897613525,
+      "learning_rate": 4.218129235604488e-06,
+      "loss": 0.2141,
+      "step": 4467
+    },
+    {
+      "epoch": 2.115655519384433,
+      "grad_norm": 0.9340530633926392,
+      "learning_rate": 4.21395737472521e-06,
+      "loss": 0.1878,
+      "step": 4468
+    },
+    {
+      "epoch": 2.1161290322580646,
+      "grad_norm": 1.0747179985046387,
+      "learning_rate": 4.209787027129563e-06,
+      "loss": 0.2098,
+      "step": 4469
+    },
+    {
+      "epoch": 2.1166025451316957,
+      "grad_norm": 1.3655407428741455,
+      "learning_rate": 4.2056181939082584e-06,
+      "loss": 0.1985,
+      "step": 4470
+    },
+    {
+      "epoch": 2.117076058005327,
+      "grad_norm": 1.1653110980987549,
+      "learning_rate": 4.201450876151619e-06,
+      "loss": 0.2072,
+      "step": 4471
+    },
+    {
+      "epoch": 2.1175495708789582,
+      "grad_norm": 1.0803667306900024,
+      "learning_rate": 4.19728507494956e-06,
+      "loss": 0.2158,
+      "step": 4472
+    },
+    {
+      "epoch": 2.1180230837525897,
+      "grad_norm": 1.5151382684707642,
+      "learning_rate": 4.193120791391612e-06,
+      "loss": 0.1998,
+      "step": 4473
+    },
+    {
+      "epoch": 2.118496596626221,
+      "grad_norm": 1.5851473808288574,
+      "learning_rate": 4.1889580265669e-06,
+      "loss": 0.199,
+      "step": 4474
+    },
+    {
+      "epoch": 2.118970109499852,
+      "grad_norm": 1.5319643020629883,
+      "learning_rate": 4.184796781564158e-06,
+      "loss": 0.1875,
+      "step": 4475
+    },
+    {
+      "epoch": 2.1194436223734834,
+      "grad_norm": 1.3403736352920532,
+      "learning_rate": 4.180637057471714e-06,
+      "loss": 0.1847,
+      "step": 4476
+    },
+    {
+      "epoch": 2.1199171352471144,
+      "grad_norm": 1.2243744134902954,
+      "learning_rate": 4.1764788553775105e-06,
+      "loss": 0.1827,
+      "step": 4477
+    },
+    {
+      "epoch": 2.120390648120746,
+      "grad_norm": 0.9514744877815247,
+      "learning_rate": 4.1723221763690826e-06,
+      "loss": 0.1932,
+      "step": 4478
+    },
+    {
+      "epoch": 2.120864160994377,
+      "grad_norm": 0.9540281295776367,
+      "learning_rate": 4.1681670215335646e-06,
+      "loss": 0.1746,
+      "step": 4479
+    },
+    {
+      "epoch": 2.121337673868008,
+      "grad_norm": 1.6678895950317383,
+      "learning_rate": 4.1640133919577065e-06,
+      "loss": 0.2004,
+      "step": 4480
+    },
+    {
+      "epoch": 2.1218111867416396,
+      "grad_norm": 1.186750054359436,
+      "learning_rate": 4.15986128872784e-06,
+      "loss": 0.2015,
+      "step": 4481
+    },
+    {
+      "epoch": 2.1222846996152707,
+      "grad_norm": 1.694643259048462,
+      "learning_rate": 4.155710712929916e-06,
+      "loss": 0.1983,
+      "step": 4482
+    },
+    {
+      "epoch": 2.122758212488902,
+      "grad_norm": 1.0623250007629395,
+      "learning_rate": 4.151561665649471e-06,
+      "loss": 0.2111,
+      "step": 4483
+    },
+    {
+      "epoch": 2.1232317253625332,
+      "grad_norm": 1.233841061592102,
+      "learning_rate": 4.147414147971655e-06,
+      "loss": 0.2121,
+      "step": 4484
+    },
+    {
+      "epoch": 2.1237052382361647,
+      "grad_norm": 1.1146776676177979,
+      "learning_rate": 4.143268160981204e-06,
+      "loss": 0.1958,
+      "step": 4485
+    },
+    {
+      "epoch": 2.124178751109796,
+      "grad_norm": 1.02981436252594,
+      "learning_rate": 4.139123705762469e-06,
+      "loss": 0.1988,
+      "step": 4486
+    },
+    {
+      "epoch": 2.124652263983427,
+      "grad_norm": 1.6232107877731323,
+      "learning_rate": 4.134980783399384e-06,
+      "loss": 0.2104,
+      "step": 4487
+    },
+    {
+      "epoch": 2.1251257768570584,
+      "grad_norm": 1.1073497533798218,
+      "learning_rate": 4.130839394975493e-06,
+      "loss": 0.1978,
+      "step": 4488
+    },
+    {
+      "epoch": 2.1255992897306895,
+      "grad_norm": 1.430740475654602,
+      "learning_rate": 4.126699541573943e-06,
+      "loss": 0.1999,
+      "step": 4489
+    },
+    {
+      "epoch": 2.126072802604321,
+      "grad_norm": 1.1626049280166626,
+      "learning_rate": 4.122561224277463e-06,
+      "loss": 0.1954,
+      "step": 4490
+    },
+    {
+      "epoch": 2.126546315477952,
+      "grad_norm": 0.9593620300292969,
+      "learning_rate": 4.1184244441683965e-06,
+      "loss": 0.1728,
+      "step": 4491
+    },
+    {
+      "epoch": 2.1270198283515835,
+      "grad_norm": 1.0123560428619385,
+      "learning_rate": 4.114289202328678e-06,
+      "loss": 0.2072,
+      "step": 4492
+    },
+    {
+      "epoch": 2.1274933412252146,
+      "grad_norm": 1.1748878955841064,
+      "learning_rate": 4.110155499839833e-06,
+      "loss": 0.2059,
+      "step": 4493
+    },
+    {
+      "epoch": 2.1279668540988457,
+      "grad_norm": 1.172324538230896,
+      "learning_rate": 4.106023337783e-06,
+      "loss": 0.1993,
+      "step": 4494
+    },
+    {
+      "epoch": 2.128440366972477,
+      "grad_norm": 1.9978069067001343,
+      "learning_rate": 4.1018927172389e-06,
+      "loss": 0.2031,
+      "step": 4495
+    },
+    {
+      "epoch": 2.1289138798461082,
+      "grad_norm": 1.4836291074752808,
+      "learning_rate": 4.097763639287864e-06,
+      "loss": 0.2432,
+      "step": 4496
+    },
+    {
+      "epoch": 2.1293873927197398,
+      "grad_norm": 1.3724143505096436,
+      "learning_rate": 4.093636105009804e-06,
+      "loss": 0.2108,
+      "step": 4497
+    },
+    {
+      "epoch": 2.129860905593371,
+      "grad_norm": 1.8259536027908325,
+      "learning_rate": 4.0895101154842444e-06,
+      "loss": 0.1977,
+      "step": 4498
+    },
+    {
+      "epoch": 2.130334418467002,
+      "grad_norm": 1.0505609512329102,
+      "learning_rate": 4.08538567179029e-06,
+      "loss": 0.1911,
+      "step": 4499
+    },
+    {
+      "epoch": 2.1308079313406334,
+      "grad_norm": 1.6359678506851196,
+      "learning_rate": 4.081262775006659e-06,
+      "loss": 0.2034,
+      "step": 4500
+    },
+    {
+      "epoch": 2.1312814442142645,
+      "grad_norm": 1.0886156558990479,
+      "learning_rate": 4.077141426211647e-06,
+      "loss": 0.2053,
+      "step": 4501
+    },
+    {
+      "epoch": 2.131754957087896,
+      "grad_norm": 1.12444269657135,
+      "learning_rate": 4.073021626483159e-06,
+      "loss": 0.2192,
+      "step": 4502
+    },
+    {
+      "epoch": 2.132228469961527,
+      "grad_norm": 0.8436000347137451,
+      "learning_rate": 4.0689033768986855e-06,
+      "loss": 0.2068,
+      "step": 4503
+    },
+    {
+      "epoch": 2.132701982835158,
+      "grad_norm": 1.2649866342544556,
+      "learning_rate": 4.064786678535313e-06,
+      "loss": 0.2072,
+      "step": 4504
+    },
+    {
+      "epoch": 2.1331754957087896,
+      "grad_norm": 1.0428705215454102,
+      "learning_rate": 4.0606715324697285e-06,
+      "loss": 0.2224,
+      "step": 4505
+    },
+    {
+      "epoch": 2.1336490085824207,
+      "grad_norm": 1.1856803894042969,
+      "learning_rate": 4.056557939778205e-06,
+      "loss": 0.1947,
+      "step": 4506
+    },
+    {
+      "epoch": 2.134122521456052,
+      "grad_norm": 2.003635883331299,
+      "learning_rate": 4.052445901536618e-06,
+      "loss": 0.2234,
+      "step": 4507
+    },
+    {
+      "epoch": 2.1345960343296833,
+      "grad_norm": 1.0833152532577515,
+      "learning_rate": 4.048335418820425e-06,
+      "loss": 0.1878,
+      "step": 4508
+    },
+    {
+      "epoch": 2.1350695472033148,
+      "grad_norm": 1.1753066778182983,
+      "learning_rate": 4.04422649270469e-06,
+      "loss": 0.1935,
+      "step": 4509
+    },
+    {
+      "epoch": 2.135543060076946,
+      "grad_norm": 1.2887529134750366,
+      "learning_rate": 4.040119124264056e-06,
+      "loss": 0.21,
+      "step": 4510
+    },
+    {
+      "epoch": 2.136016572950577,
+      "grad_norm": 1.0373543500900269,
+      "learning_rate": 4.036013314572772e-06,
+      "loss": 0.2036,
+      "step": 4511
+    },
+    {
+      "epoch": 2.1364900858242084,
+      "grad_norm": 1.2322744131088257,
+      "learning_rate": 4.0319090647046714e-06,
+      "loss": 0.2086,
+      "step": 4512
+    },
+    {
+      "epoch": 2.1369635986978395,
+      "grad_norm": 1.1571698188781738,
+      "learning_rate": 4.0278063757331745e-06,
+      "loss": 0.1731,
+      "step": 4513
+    },
+    {
+      "epoch": 2.137437111571471,
+      "grad_norm": 1.0368672609329224,
+      "learning_rate": 4.0237052487313084e-06,
+      "loss": 0.204,
+      "step": 4514
+    },
+    {
+      "epoch": 2.137910624445102,
+      "grad_norm": 1.494619369506836,
+      "learning_rate": 4.0196056847716795e-06,
+      "loss": 0.2129,
+      "step": 4515
+    },
+    {
+      "epoch": 2.1383841373187336,
+      "grad_norm": 1.6594982147216797,
+      "learning_rate": 4.015507684926491e-06,
+      "loss": 0.1958,
+      "step": 4516
+    },
+    {
+      "epoch": 2.1388576501923646,
+      "grad_norm": 1.1065243482589722,
+      "learning_rate": 4.0114112502675305e-06,
+      "loss": 0.1961,
+      "step": 4517
+    },
+    {
+      "epoch": 2.1393311630659957,
+      "grad_norm": 1.0827975273132324,
+      "learning_rate": 4.007316381866188e-06,
+      "loss": 0.2111,
+      "step": 4518
+    },
+    {
+      "epoch": 2.139804675939627,
+      "grad_norm": 1.3087358474731445,
+      "learning_rate": 4.003223080793432e-06,
+      "loss": 0.2031,
+      "step": 4519
+    },
+    {
+      "epoch": 2.1402781888132583,
+      "grad_norm": 1.1883478164672852,
+      "learning_rate": 3.999131348119829e-06,
+      "loss": 0.198,
+      "step": 4520
+    },
+    {
+      "epoch": 2.1407517016868898,
+      "grad_norm": 1.2664716243743896,
+      "learning_rate": 3.995041184915531e-06,
+      "loss": 0.2151,
+      "step": 4521
+    },
+    {
+      "epoch": 2.141225214560521,
+      "grad_norm": 1.1769858598709106,
+      "learning_rate": 3.990952592250277e-06,
+      "loss": 0.2112,
+      "step": 4522
+    },
+    {
+      "epoch": 2.141698727434152,
+      "grad_norm": 1.0426677465438843,
+      "learning_rate": 3.986865571193404e-06,
+      "loss": 0.1892,
+      "step": 4523
+    },
+    {
+      "epoch": 2.1421722403077834,
+      "grad_norm": 1.4847642183303833,
+      "learning_rate": 3.98278012281383e-06,
+      "loss": 0.209,
+      "step": 4524
+    },
+    {
+      "epoch": 2.1426457531814145,
+      "grad_norm": 1.9673742055892944,
+      "learning_rate": 3.978696248180069e-06,
+      "loss": 0.2053,
+      "step": 4525
+    },
+    {
+      "epoch": 2.143119266055046,
+      "grad_norm": 1.1002095937728882,
+      "learning_rate": 3.9746139483602095e-06,
+      "loss": 0.1944,
+      "step": 4526
+    },
+    {
+      "epoch": 2.143592778928677,
+      "grad_norm": 1.0705955028533936,
+      "learning_rate": 3.970533224421947e-06,
+      "loss": 0.1889,
+      "step": 4527
+    },
+    {
+      "epoch": 2.1440662918023086,
+      "grad_norm": 1.1432850360870361,
+      "learning_rate": 3.9664540774325545e-06,
+      "loss": 0.1978,
+      "step": 4528
+    },
+    {
+      "epoch": 2.1445398046759396,
+      "grad_norm": 1.0200495719909668,
+      "learning_rate": 3.962376508458887e-06,
+      "loss": 0.2121,
+      "step": 4529
+    },
+    {
+      "epoch": 2.1450133175495707,
+      "grad_norm": 1.4120169878005981,
+      "learning_rate": 3.958300518567403e-06,
+      "loss": 0.2158,
+      "step": 4530
+    },
+    {
+      "epoch": 2.145486830423202,
+      "grad_norm": 1.242547631263733,
+      "learning_rate": 3.954226108824129e-06,
+      "loss": 0.2045,
+      "step": 4531
+    },
+    {
+      "epoch": 2.1459603432968333,
+      "grad_norm": 1.3954753875732422,
+      "learning_rate": 3.950153280294695e-06,
+      "loss": 0.1992,
+      "step": 4532
+    },
+    {
+      "epoch": 2.146433856170465,
+      "grad_norm": 1.1535682678222656,
+      "learning_rate": 3.946082034044303e-06,
+      "loss": 0.1887,
+      "step": 4533
+    },
+    {
+      "epoch": 2.146907369044096,
+      "grad_norm": 1.6597424745559692,
+      "learning_rate": 3.942012371137755e-06,
+      "loss": 0.206,
+      "step": 4534
+    },
+    {
+      "epoch": 2.1473808819177274,
+      "grad_norm": 1.2386748790740967,
+      "learning_rate": 3.937944292639426e-06,
+      "loss": 0.2018,
+      "step": 4535
+    },
+    {
+      "epoch": 2.1478543947913584,
+      "grad_norm": 1.133917212486267,
+      "learning_rate": 3.9338777996132885e-06,
+      "loss": 0.2068,
+      "step": 4536
+    },
+    {
+      "epoch": 2.1483279076649895,
+      "grad_norm": 1.2553430795669556,
+      "learning_rate": 3.929812893122892e-06,
+      "loss": 0.2355,
+      "step": 4537
+    },
+    {
+      "epoch": 2.148801420538621,
+      "grad_norm": 1.0734045505523682,
+      "learning_rate": 3.9257495742313704e-06,
+      "loss": 0.2102,
+      "step": 4538
+    },
+    {
+      "epoch": 2.149274933412252,
+      "grad_norm": 1.530061960220337,
+      "learning_rate": 3.9216878440014506e-06,
+      "loss": 0.2079,
+      "step": 4539
+    },
+    {
+      "epoch": 2.1497484462858836,
+      "grad_norm": 1.4186937808990479,
+      "learning_rate": 3.917627703495434e-06,
+      "loss": 0.1909,
+      "step": 4540
+    },
+    {
+      "epoch": 2.1502219591595146,
+      "grad_norm": 0.925179123878479,
+      "learning_rate": 3.913569153775216e-06,
+      "loss": 0.1862,
+      "step": 4541
+    },
+    {
+      "epoch": 2.1506954720331457,
+      "grad_norm": 1.8251923322677612,
+      "learning_rate": 3.909512195902266e-06,
+      "loss": 0.1955,
+      "step": 4542
+    },
+    {
+      "epoch": 2.151168984906777,
+      "grad_norm": 1.1590983867645264,
+      "learning_rate": 3.905456830937651e-06,
+      "loss": 0.1839,
+      "step": 4543
+    },
+    {
+      "epoch": 2.1516424977804083,
+      "grad_norm": 1.3032938241958618,
+      "learning_rate": 3.901403059942e-06,
+      "loss": 0.1992,
+      "step": 4544
+    },
+    {
+      "epoch": 2.15211601065404,
+      "grad_norm": 1.323080062866211,
+      "learning_rate": 3.897350883975551e-06,
+      "loss": 0.1924,
+      "step": 4545
+    },
+    {
+      "epoch": 2.152589523527671,
+      "grad_norm": 1.6115025281906128,
+      "learning_rate": 3.893300304098102e-06,
+      "loss": 0.1954,
+      "step": 4546
+    },
+    {
+      "epoch": 2.1530630364013024,
+      "grad_norm": 1.2296570539474487,
+      "learning_rate": 3.889251321369044e-06,
+      "loss": 0.2111,
+      "step": 4547
+    },
+    {
+      "epoch": 2.1535365492749334,
+      "grad_norm": 1.6636604070663452,
+      "learning_rate": 3.885203936847355e-06,
+      "loss": 0.2111,
+      "step": 4548
+    },
+    {
+      "epoch": 2.1540100621485645,
+      "grad_norm": 1.3433752059936523,
+      "learning_rate": 3.881158151591583e-06,
+      "loss": 0.227,
+      "step": 4549
+    },
+    {
+      "epoch": 2.154483575022196,
+      "grad_norm": 1.0224047899246216,
+      "learning_rate": 3.877113966659869e-06,
+      "loss": 0.1997,
+      "step": 4550
+    },
+    {
+      "epoch": 2.154957087895827,
+      "grad_norm": 1.1871498823165894,
+      "learning_rate": 3.8730713831099265e-06,
+      "loss": 0.2111,
+      "step": 4551
+    },
+    {
+      "epoch": 2.1554306007694586,
+      "grad_norm": 1.0613186359405518,
+      "learning_rate": 3.869030401999059e-06,
+      "loss": 0.2026,
+      "step": 4552
+    },
+    {
+      "epoch": 2.1559041136430896,
+      "grad_norm": 1.2302757501602173,
+      "learning_rate": 3.8649910243841395e-06,
+      "loss": 0.212,
+      "step": 4553
+    },
+    {
+      "epoch": 2.156377626516721,
+      "grad_norm": 0.9007354974746704,
+      "learning_rate": 3.860953251321635e-06,
+      "loss": 0.1887,
+      "step": 4554
+    },
+    {
+      "epoch": 2.1568511393903522,
+      "grad_norm": 1.7830326557159424,
+      "learning_rate": 3.856917083867581e-06,
+      "loss": 0.1895,
+      "step": 4555
+    },
+    {
+      "epoch": 2.1573246522639833,
+      "grad_norm": 1.7577792406082153,
+      "learning_rate": 3.852882523077604e-06,
+      "loss": 0.222,
+      "step": 4556
+    },
+    {
+      "epoch": 2.157798165137615,
+      "grad_norm": 1.0963820219039917,
+      "learning_rate": 3.8488495700068994e-06,
+      "loss": 0.2307,
+      "step": 4557
+    },
+    {
+      "epoch": 2.158271678011246,
+      "grad_norm": 1.826006531715393,
+      "learning_rate": 3.844818225710246e-06,
+      "loss": 0.1977,
+      "step": 4558
+    },
+    {
+      "epoch": 2.1587451908848774,
+      "grad_norm": 1.689234972000122,
+      "learning_rate": 3.840788491242009e-06,
+      "loss": 0.204,
+      "step": 4559
+    },
+    {
+      "epoch": 2.1592187037585084,
+      "grad_norm": 1.8260334730148315,
+      "learning_rate": 3.8367603676561195e-06,
+      "loss": 0.2045,
+      "step": 4560
+    },
+    {
+      "epoch": 2.1596922166321395,
+      "grad_norm": 1.4265202283859253,
+      "learning_rate": 3.832733856006103e-06,
+      "loss": 0.2122,
+      "step": 4561
+    },
+    {
+      "epoch": 2.160165729505771,
+      "grad_norm": 1.1697028875350952,
+      "learning_rate": 3.8287089573450444e-06,
+      "loss": 0.1932,
+      "step": 4562
+    },
+    {
+      "epoch": 2.160639242379402,
+      "grad_norm": 1.2971158027648926,
+      "learning_rate": 3.824685672725626e-06,
+      "loss": 0.1934,
+      "step": 4563
+    },
+    {
+      "epoch": 2.1611127552530336,
+      "grad_norm": 0.9162940382957458,
+      "learning_rate": 3.820664003200092e-06,
+      "loss": 0.2034,
+      "step": 4564
+    },
+    {
+      "epoch": 2.1615862681266647,
+      "grad_norm": 2.122776985168457,
+      "learning_rate": 3.816643949820275e-06,
+      "loss": 0.2026,
+      "step": 4565
+    },
+    {
+      "epoch": 2.1620597810002957,
+      "grad_norm": 1.0466620922088623,
+      "learning_rate": 3.8126255136375857e-06,
+      "loss": 0.2109,
+      "step": 4566
+    },
+    {
+      "epoch": 2.1625332938739272,
+      "grad_norm": 0.9423272013664246,
+      "learning_rate": 3.808608695702999e-06,
+      "loss": 0.1851,
+      "step": 4567
+    },
+    {
+      "epoch": 2.1630068067475583,
+      "grad_norm": 1.181909203529358,
+      "learning_rate": 3.8045934970670808e-06,
+      "loss": 0.2037,
+      "step": 4568
+    },
+    {
+      "epoch": 2.16348031962119,
+      "grad_norm": 1.0272080898284912,
+      "learning_rate": 3.800579918779963e-06,
+      "loss": 0.1763,
+      "step": 4569
+    },
+    {
+      "epoch": 2.163953832494821,
+      "grad_norm": 1.413796305656433,
+      "learning_rate": 3.796567961891363e-06,
+      "loss": 0.1911,
+      "step": 4570
+    },
+    {
+      "epoch": 2.1644273453684524,
+      "grad_norm": 1.3752124309539795,
+      "learning_rate": 3.792557627450568e-06,
+      "loss": 0.193,
+      "step": 4571
+    },
+    {
+      "epoch": 2.1649008582420834,
+      "grad_norm": 1.488099217414856,
+      "learning_rate": 3.788548916506437e-06,
+      "loss": 0.1938,
+      "step": 4572
+    },
+    {
+      "epoch": 2.1653743711157145,
+      "grad_norm": 1.0083427429199219,
+      "learning_rate": 3.7845418301074176e-06,
+      "loss": 0.2191,
+      "step": 4573
+    },
+    {
+      "epoch": 2.165847883989346,
+      "grad_norm": 1.5522511005401611,
+      "learning_rate": 3.7805363693015172e-06,
+      "loss": 0.1947,
+      "step": 4574
+    },
+    {
+      "epoch": 2.166321396862977,
+      "grad_norm": 1.5290082693099976,
+      "learning_rate": 3.7765325351363335e-06,
+      "loss": 0.1975,
+      "step": 4575
+    },
+    {
+      "epoch": 2.1667949097366086,
+      "grad_norm": 0.8700765371322632,
+      "learning_rate": 3.772530328659023e-06,
+      "loss": 0.2067,
+      "step": 4576
+    },
+    {
+      "epoch": 2.1672684226102397,
+      "grad_norm": 0.9798673987388611,
+      "learning_rate": 3.7685297509163297e-06,
+      "loss": 0.2168,
+      "step": 4577
+    },
+    {
+      "epoch": 2.167741935483871,
+      "grad_norm": 1.7563843727111816,
+      "learning_rate": 3.7645308029545623e-06,
+      "loss": 0.1976,
+      "step": 4578
+    },
+    {
+      "epoch": 2.1682154483575022,
+      "grad_norm": 1.0324368476867676,
+      "learning_rate": 3.7605334858196107e-06,
+      "loss": 0.204,
+      "step": 4579
+    },
+    {
+      "epoch": 2.1686889612311333,
+      "grad_norm": 1.488629937171936,
+      "learning_rate": 3.75653780055693e-06,
+      "loss": 0.2303,
+      "step": 4580
+    },
+    {
+      "epoch": 2.169162474104765,
+      "grad_norm": 1.2473211288452148,
+      "learning_rate": 3.752543748211559e-06,
+      "loss": 0.2015,
+      "step": 4581
+    },
+    {
+      "epoch": 2.169635986978396,
+      "grad_norm": 1.4931015968322754,
+      "learning_rate": 3.748551329828101e-06,
+      "loss": 0.2084,
+      "step": 4582
+    },
+    {
+      "epoch": 2.1701094998520274,
+      "grad_norm": 1.2696232795715332,
+      "learning_rate": 3.7445605464507295e-06,
+      "loss": 0.2173,
+      "step": 4583
+    },
+    {
+      "epoch": 2.1705830127256585,
+      "grad_norm": 0.9876007437705994,
+      "learning_rate": 3.740571399123204e-06,
+      "loss": 0.2009,
+      "step": 4584
+    },
+    {
+      "epoch": 2.1710565255992895,
+      "grad_norm": 1.6325695514678955,
+      "learning_rate": 3.7365838888888395e-06,
+      "loss": 0.2295,
+      "step": 4585
+    },
+    {
+      "epoch": 2.171530038472921,
+      "grad_norm": 1.6730865240097046,
+      "learning_rate": 3.732598016790537e-06,
+      "loss": 0.2208,
+      "step": 4586
+    },
+    {
+      "epoch": 2.172003551346552,
+      "grad_norm": 1.2315953969955444,
+      "learning_rate": 3.728613783870759e-06,
+      "loss": 0.1985,
+      "step": 4587
+    },
+    {
+      "epoch": 2.1724770642201836,
+      "grad_norm": 1.5150845050811768,
+      "learning_rate": 3.724631191171547e-06,
+      "loss": 0.194,
+      "step": 4588
+    },
+    {
+      "epoch": 2.1729505770938147,
+      "grad_norm": 1.2721232175827026,
+      "learning_rate": 3.7206502397345044e-06,
+      "loss": 0.1883,
+      "step": 4589
+    },
+    {
+      "epoch": 2.173424089967446,
+      "grad_norm": 1.0581114292144775,
+      "learning_rate": 3.7166709306008163e-06,
+      "loss": 0.2121,
+      "step": 4590
+    },
+    {
+      "epoch": 2.1738976028410772,
+      "grad_norm": 0.9920486211776733,
+      "learning_rate": 3.712693264811231e-06,
+      "loss": 0.192,
+      "step": 4591
+    },
+    {
+      "epoch": 2.1743711157147083,
+      "grad_norm": 1.3474196195602417,
+      "learning_rate": 3.7087172434060635e-06,
+      "loss": 0.2062,
+      "step": 4592
+    },
+    {
+      "epoch": 2.17484462858834,
+      "grad_norm": 1.098014235496521,
+      "learning_rate": 3.704742867425212e-06,
+      "loss": 0.2018,
+      "step": 4593
+    },
+    {
+      "epoch": 2.175318141461971,
+      "grad_norm": 1.0982664823532104,
+      "learning_rate": 3.7007701379081275e-06,
+      "loss": 0.204,
+      "step": 4594
+    },
+    {
+      "epoch": 2.1757916543356024,
+      "grad_norm": 1.2308850288391113,
+      "learning_rate": 3.6967990558938484e-06,
+      "loss": 0.2105,
+      "step": 4595
+    },
+    {
+      "epoch": 2.1762651672092335,
+      "grad_norm": 1.5977511405944824,
+      "learning_rate": 3.6928296224209636e-06,
+      "loss": 0.2215,
+      "step": 4596
+    },
+    {
+      "epoch": 2.176738680082865,
+      "grad_norm": 1.0719974040985107,
+      "learning_rate": 3.688861838527649e-06,
+      "loss": 0.196,
+      "step": 4597
+    },
+    {
+      "epoch": 2.177212192956496,
+      "grad_norm": 1.4752261638641357,
+      "learning_rate": 3.684895705251632e-06,
+      "loss": 0.1891,
+      "step": 4598
+    },
+    {
+      "epoch": 2.177685705830127,
+      "grad_norm": 1.9501125812530518,
+      "learning_rate": 3.6809312236302243e-06,
+      "loss": 0.208,
+      "step": 4599
+    },
+    {
+      "epoch": 2.1781592187037586,
+      "grad_norm": 1.51497483253479,
+      "learning_rate": 3.6769683947002934e-06,
+      "loss": 0.2276,
+      "step": 4600
+    },
+    {
+      "epoch": 2.1786327315773897,
+      "grad_norm": 1.680430293083191,
+      "learning_rate": 3.673007219498276e-06,
+      "loss": 0.1995,
+      "step": 4601
+    },
+    {
+      "epoch": 2.179106244451021,
+      "grad_norm": 1.073860764503479,
+      "learning_rate": 3.6690476990601866e-06,
+      "loss": 0.217,
+      "step": 4602
+    },
+    {
+      "epoch": 2.1795797573246523,
+      "grad_norm": 1.1040340662002563,
+      "learning_rate": 3.665089834421588e-06,
+      "loss": 0.2057,
+      "step": 4603
+    },
+    {
+      "epoch": 2.1800532701982833,
+      "grad_norm": 1.3993034362792969,
+      "learning_rate": 3.661133626617638e-06,
+      "loss": 0.2013,
+      "step": 4604
+    },
+    {
+      "epoch": 2.180526783071915,
+      "grad_norm": 1.149754524230957,
+      "learning_rate": 3.657179076683034e-06,
+      "loss": 0.1923,
+      "step": 4605
+    },
+    {
+      "epoch": 2.181000295945546,
+      "grad_norm": 1.6532198190689087,
+      "learning_rate": 3.653226185652049e-06,
+      "loss": 0.1997,
+      "step": 4606
+    },
+    {
+      "epoch": 2.1814738088191774,
+      "grad_norm": 1.3920501470565796,
+      "learning_rate": 3.6492749545585313e-06,
+      "loss": 0.2051,
+      "step": 4607
+    },
+    {
+      "epoch": 2.1819473216928085,
+      "grad_norm": 1.0357420444488525,
+      "learning_rate": 3.6453253844358783e-06,
+      "loss": 0.2199,
+      "step": 4608
+    },
+    {
+      "epoch": 2.18242083456644,
+      "grad_norm": 1.4226856231689453,
+      "learning_rate": 3.6413774763170707e-06,
+      "loss": 0.2089,
+      "step": 4609
+    },
+    {
+      "epoch": 2.182894347440071,
+      "grad_norm": 1.0830601453781128,
+      "learning_rate": 3.637431231234637e-06,
+      "loss": 0.2125,
+      "step": 4610
+    },
+    {
+      "epoch": 2.183367860313702,
+      "grad_norm": 1.3604761362075806,
+      "learning_rate": 3.6334866502206877e-06,
+      "loss": 0.203,
+      "step": 4611
+    },
+    {
+      "epoch": 2.1838413731873336,
+      "grad_norm": 1.3618247509002686,
+      "learning_rate": 3.6295437343068828e-06,
+      "loss": 0.1979,
+      "step": 4612
+    },
+    {
+      "epoch": 2.1843148860609647,
+      "grad_norm": 1.374878168106079,
+      "learning_rate": 3.625602484524461e-06,
+      "loss": 0.2205,
+      "step": 4613
+    },
+    {
+      "epoch": 2.184788398934596,
+      "grad_norm": 0.9489659667015076,
+      "learning_rate": 3.6216629019042106e-06,
+      "loss": 0.1847,
+      "step": 4614
+    },
+    {
+      "epoch": 2.1852619118082273,
+      "grad_norm": 1.074998378753662,
+      "learning_rate": 3.6177249874764986e-06,
+      "loss": 0.1974,
+      "step": 4615
+    },
+    {
+      "epoch": 2.1857354246818583,
+      "grad_norm": 1.1851049661636353,
+      "learning_rate": 3.613788742271246e-06,
+      "loss": 0.2032,
+      "step": 4616
+    },
+    {
+      "epoch": 2.18620893755549,
+      "grad_norm": 0.9041722416877747,
+      "learning_rate": 3.6098541673179353e-06,
+      "loss": 0.2215,
+      "step": 4617
+    },
+    {
+      "epoch": 2.186682450429121,
+      "grad_norm": 1.1576639413833618,
+      "learning_rate": 3.605921263645623e-06,
+      "loss": 0.21,
+      "step": 4618
+    },
+    {
+      "epoch": 2.1871559633027524,
+      "grad_norm": 1.2321959733963013,
+      "learning_rate": 3.6019900322829173e-06,
+      "loss": 0.2291,
+      "step": 4619
+    },
+    {
+      "epoch": 2.1876294761763835,
+      "grad_norm": 1.1647976636886597,
+      "learning_rate": 3.5980604742579985e-06,
+      "loss": 0.1897,
+      "step": 4620
+    },
+    {
+      "epoch": 2.188102989050015,
+      "grad_norm": 1.6139419078826904,
+      "learning_rate": 3.594132590598599e-06,
+      "loss": 0.2243,
+      "step": 4621
+    },
+    {
+      "epoch": 2.188576501923646,
+      "grad_norm": 1.1993911266326904,
+      "learning_rate": 3.5902063823320255e-06,
+      "loss": 0.2209,
+      "step": 4622
+    },
+    {
+      "epoch": 2.189050014797277,
+      "grad_norm": 1.2088009119033813,
+      "learning_rate": 3.5862818504851325e-06,
+      "loss": 0.1982,
+      "step": 4623
+    },
+    {
+      "epoch": 2.1895235276709086,
+      "grad_norm": 1.2603254318237305,
+      "learning_rate": 3.5823589960843506e-06,
+      "loss": 0.1903,
+      "step": 4624
+    },
+    {
+      "epoch": 2.1899970405445397,
+      "grad_norm": 1.01237952709198,
+      "learning_rate": 3.5784378201556612e-06,
+      "loss": 0.2093,
+      "step": 4625
+    },
+    {
+      "epoch": 2.190470553418171,
+      "grad_norm": 1.6355202198028564,
+      "learning_rate": 3.5745183237246074e-06,
+      "loss": 0.1958,
+      "step": 4626
+    },
+    {
+      "epoch": 2.1909440662918023,
+      "grad_norm": 1.0540543794631958,
+      "learning_rate": 3.570600507816301e-06,
+      "loss": 0.2108,
+      "step": 4627
+    },
+    {
+      "epoch": 2.1914175791654333,
+      "grad_norm": 2.3014087677001953,
+      "learning_rate": 3.5666843734554022e-06,
+      "loss": 0.1943,
+      "step": 4628
+    },
+    {
+      "epoch": 2.191891092039065,
+      "grad_norm": 1.5514744520187378,
+      "learning_rate": 3.562769921666147e-06,
+      "loss": 0.2027,
+      "step": 4629
+    },
+    {
+      "epoch": 2.192364604912696,
+      "grad_norm": 1.1093729734420776,
+      "learning_rate": 3.558857153472314e-06,
+      "loss": 0.1883,
+      "step": 4630
+    },
+    {
+      "epoch": 2.1928381177863274,
+      "grad_norm": 1.1554844379425049,
+      "learning_rate": 3.554946069897256e-06,
+      "loss": 0.1939,
+      "step": 4631
+    },
+    {
+      "epoch": 2.1933116306599585,
+      "grad_norm": 1.6279453039169312,
+      "learning_rate": 3.5510366719638745e-06,
+      "loss": 0.2058,
+      "step": 4632
+    },
+    {
+      "epoch": 2.19378514353359,
+      "grad_norm": 1.1952141523361206,
+      "learning_rate": 3.54712896069464e-06,
+      "loss": 0.1856,
+      "step": 4633
+    },
+    {
+      "epoch": 2.194258656407221,
+      "grad_norm": 1.0594127178192139,
+      "learning_rate": 3.543222937111571e-06,
+      "loss": 0.2095,
+      "step": 4634
+    },
+    {
+      "epoch": 2.194732169280852,
+      "grad_norm": 1.2939765453338623,
+      "learning_rate": 3.539318602236257e-06,
+      "loss": 0.1998,
+      "step": 4635
+    },
+    {
+      "epoch": 2.1952056821544836,
+      "grad_norm": 1.392008900642395,
+      "learning_rate": 3.535415957089835e-06,
+      "loss": 0.1864,
+      "step": 4636
+    },
+    {
+      "epoch": 2.1956791950281147,
+      "grad_norm": 1.3421459197998047,
+      "learning_rate": 3.531515002693e-06,
+      "loss": 0.2139,
+      "step": 4637
+    },
+    {
+      "epoch": 2.196152707901746,
+      "grad_norm": 1.5865654945373535,
+      "learning_rate": 3.5276157400660184e-06,
+      "loss": 0.2222,
+      "step": 4638
+    },
+    {
+      "epoch": 2.1966262207753773,
+      "grad_norm": 0.9927922487258911,
+      "learning_rate": 3.523718170228696e-06,
+      "loss": 0.1997,
+      "step": 4639
+    },
+    {
+      "epoch": 2.197099733649009,
+      "grad_norm": 1.2386114597320557,
+      "learning_rate": 3.5198222942004113e-06,
+      "loss": 0.1857,
+      "step": 4640
+    },
+    {
+      "epoch": 2.19757324652264,
+      "grad_norm": 0.8716875910758972,
+      "learning_rate": 3.5159281130000867e-06,
+      "loss": 0.1986,
+      "step": 4641
+    },
+    {
+      "epoch": 2.198046759396271,
+      "grad_norm": 1.1668813228607178,
+      "learning_rate": 3.512035627646211e-06,
+      "loss": 0.2301,
+      "step": 4642
+    },
+    {
+      "epoch": 2.1985202722699024,
+      "grad_norm": 1.0498262643814087,
+      "learning_rate": 3.5081448391568307e-06,
+      "loss": 0.2091,
+      "step": 4643
+    },
+    {
+      "epoch": 2.1989937851435335,
+      "grad_norm": 1.0276434421539307,
+      "learning_rate": 3.5042557485495355e-06,
+      "loss": 0.1967,
+      "step": 4644
+    },
+    {
+      "epoch": 2.199467298017165,
+      "grad_norm": 2.1473894119262695,
+      "learning_rate": 3.500368356841487e-06,
+      "loss": 0.2149,
+      "step": 4645
+    },
+    {
+      "epoch": 2.199940810890796,
+      "grad_norm": 1.263840913772583,
+      "learning_rate": 3.496482665049389e-06,
+      "loss": 0.2281,
+      "step": 4646
+    },
+    {
+      "epoch": 2.200414323764427,
+      "grad_norm": 1.1061254739761353,
+      "learning_rate": 3.4925986741895125e-06,
+      "loss": 0.1913,
+      "step": 4647
+    },
+    {
+      "epoch": 2.2008878366380586,
+      "grad_norm": 1.3180168867111206,
+      "learning_rate": 3.4887163852776716e-06,
+      "loss": 0.2162,
+      "step": 4648
+    },
+    {
+      "epoch": 2.2013613495116897,
+      "grad_norm": 1.3046200275421143,
+      "learning_rate": 3.484835799329248e-06,
+      "loss": 0.1811,
+      "step": 4649
+    },
+    {
+      "epoch": 2.2018348623853212,
+      "grad_norm": 1.533521056175232,
+      "learning_rate": 3.4809569173591683e-06,
+      "loss": 0.2237,
+      "step": 4650
+    },
+    {
+      "epoch": 2.2023083752589523,
+      "grad_norm": 1.1395829916000366,
+      "learning_rate": 3.4770797403819122e-06,
+      "loss": 0.1978,
+      "step": 4651
+    },
+    {
+      "epoch": 2.202781888132584,
+      "grad_norm": 1.0285067558288574,
+      "learning_rate": 3.4732042694115265e-06,
+      "loss": 0.1905,
+      "step": 4652
+    },
+    {
+      "epoch": 2.203255401006215,
+      "grad_norm": 1.173947811126709,
+      "learning_rate": 3.4693305054615957e-06,
+      "loss": 0.1928,
+      "step": 4653
+    },
+    {
+      "epoch": 2.203728913879846,
+      "grad_norm": 1.178647518157959,
+      "learning_rate": 3.4654584495452716e-06,
+      "loss": 0.1898,
+      "step": 4654
+    },
+    {
+      "epoch": 2.2042024267534774,
+      "grad_norm": 1.0134490728378296,
+      "learning_rate": 3.4615881026752473e-06,
+      "loss": 0.2047,
+      "step": 4655
+    },
+    {
+      "epoch": 2.2046759396271085,
+      "grad_norm": 2.0668044090270996,
+      "learning_rate": 3.4577194658637815e-06,
+      "loss": 0.2035,
+      "step": 4656
+    },
+    {
+      "epoch": 2.20514945250074,
+      "grad_norm": 1.4281859397888184,
+      "learning_rate": 3.4538525401226697e-06,
+      "loss": 0.1914,
+      "step": 4657
+    },
+    {
+      "epoch": 2.205622965374371,
+      "grad_norm": 1.2419672012329102,
+      "learning_rate": 3.4499873264632787e-06,
+      "loss": 0.2086,
+      "step": 4658
+    },
+    {
+      "epoch": 2.2060964782480026,
+      "grad_norm": 1.1350500583648682,
+      "learning_rate": 3.44612382589651e-06,
+      "loss": 0.1957,
+      "step": 4659
+    },
+    {
+      "epoch": 2.2065699911216337,
+      "grad_norm": 1.8303426504135132,
+      "learning_rate": 3.4422620394328322e-06,
+      "loss": 0.1979,
+      "step": 4660
+    },
+    {
+      "epoch": 2.2070435039952647,
+      "grad_norm": 1.1737298965454102,
+      "learning_rate": 3.438401968082253e-06,
+      "loss": 0.1961,
+      "step": 4661
+    },
+    {
+      "epoch": 2.2075170168688962,
+      "grad_norm": 1.198386311531067,
+      "learning_rate": 3.434543612854336e-06,
+      "loss": 0.2124,
+      "step": 4662
+    },
+    {
+      "epoch": 2.2079905297425273,
+      "grad_norm": 1.2552683353424072,
+      "learning_rate": 3.4306869747582016e-06,
+      "loss": 0.1846,
+      "step": 4663
+    },
+    {
+      "epoch": 2.208464042616159,
+      "grad_norm": 0.9571728110313416,
+      "learning_rate": 3.426832054802511e-06,
+      "loss": 0.1961,
+      "step": 4664
+    },
+    {
+      "epoch": 2.20893755548979,
+      "grad_norm": 1.5007789134979248,
+      "learning_rate": 3.422978853995487e-06,
+      "loss": 0.2061,
+      "step": 4665
+    },
+    {
+      "epoch": 2.209411068363421,
+      "grad_norm": 1.0542176961898804,
+      "learning_rate": 3.4191273733448916e-06,
+      "loss": 0.2199,
+      "step": 4666
+    },
+    {
+      "epoch": 2.2098845812370524,
+      "grad_norm": 1.1434391736984253,
+      "learning_rate": 3.4152776138580466e-06,
+      "loss": 0.2185,
+      "step": 4667
+    },
+    {
+      "epoch": 2.2103580941106835,
+      "grad_norm": 1.2754579782485962,
+      "learning_rate": 3.411429576541815e-06,
+      "loss": 0.2134,
+      "step": 4668
+    },
+    {
+      "epoch": 2.210831606984315,
+      "grad_norm": 0.9875036478042603,
+      "learning_rate": 3.4075832624026204e-06,
+      "loss": 0.1884,
+      "step": 4669
+    },
+    {
+      "epoch": 2.211305119857946,
+      "grad_norm": 1.2512601613998413,
+      "learning_rate": 3.403738672446425e-06,
+      "loss": 0.1965,
+      "step": 4670
+    },
+    {
+      "epoch": 2.211778632731577,
+      "grad_norm": 1.0288127660751343,
+      "learning_rate": 3.3998958076787415e-06,
+      "loss": 0.1922,
+      "step": 4671
+    },
+    {
+      "epoch": 2.2122521456052087,
+      "grad_norm": 1.011583924293518,
+      "learning_rate": 3.3960546691046405e-06,
+      "loss": 0.1961,
+      "step": 4672
+    },
+    {
+      "epoch": 2.2127256584788397,
+      "grad_norm": 1.268002986907959,
+      "learning_rate": 3.3922152577287284e-06,
+      "loss": 0.183,
+      "step": 4673
+    },
+    {
+      "epoch": 2.2131991713524712,
+      "grad_norm": 1.4242584705352783,
+      "learning_rate": 3.388377574555172e-06,
+      "loss": 0.2272,
+      "step": 4674
+    },
+    {
+      "epoch": 2.2136726842261023,
+      "grad_norm": 1.1874171495437622,
+      "learning_rate": 3.3845416205876737e-06,
+      "loss": 0.2049,
+      "step": 4675
+    },
+    {
+      "epoch": 2.214146197099734,
+      "grad_norm": 0.9578942060470581,
+      "learning_rate": 3.380707396829498e-06,
+      "loss": 0.2171,
+      "step": 4676
+    },
+    {
+      "epoch": 2.214619709973365,
+      "grad_norm": 1.165828824043274,
+      "learning_rate": 3.3768749042834416e-06,
+      "loss": 0.2164,
+      "step": 4677
+    },
+    {
+      "epoch": 2.215093222846996,
+      "grad_norm": 1.2874082326889038,
+      "learning_rate": 3.3730441439518637e-06,
+      "loss": 0.1935,
+      "step": 4678
+    },
+    {
+      "epoch": 2.2155667357206275,
+      "grad_norm": 1.277848720550537,
+      "learning_rate": 3.3692151168366573e-06,
+      "loss": 0.1781,
+      "step": 4679
+    },
+    {
+      "epoch": 2.2160402485942585,
+      "grad_norm": 2.349771499633789,
+      "learning_rate": 3.3653878239392668e-06,
+      "loss": 0.2103,
+      "step": 4680
+    },
+    {
+      "epoch": 2.21651376146789,
+      "grad_norm": 1.3279833793640137,
+      "learning_rate": 3.3615622662606852e-06,
+      "loss": 0.2103,
+      "step": 4681
+    },
+    {
+      "epoch": 2.216987274341521,
+      "grad_norm": 1.032465934753418,
+      "learning_rate": 3.357738444801449e-06,
+      "loss": 0.1824,
+      "step": 4682
+    },
+    {
+      "epoch": 2.2174607872151526,
+      "grad_norm": 1.4926496744155884,
+      "learning_rate": 3.35391636056165e-06,
+      "loss": 0.1904,
+      "step": 4683
+    },
+    {
+      "epoch": 2.2179343000887837,
+      "grad_norm": 0.948499858379364,
+      "learning_rate": 3.350096014540909e-06,
+      "loss": 0.1849,
+      "step": 4684
+    },
+    {
+      "epoch": 2.2184078129624147,
+      "grad_norm": 1.0163851976394653,
+      "learning_rate": 3.3462774077383996e-06,
+      "loss": 0.209,
+      "step": 4685
+    },
+    {
+      "epoch": 2.2188813258360462,
+      "grad_norm": 1.0877399444580078,
+      "learning_rate": 3.3424605411528476e-06,
+      "loss": 0.2033,
+      "step": 4686
+    },
+    {
+      "epoch": 2.2193548387096773,
+      "grad_norm": 1.5267970561981201,
+      "learning_rate": 3.338645415782512e-06,
+      "loss": 0.207,
+      "step": 4687
+    },
+    {
+      "epoch": 2.219828351583309,
+      "grad_norm": 1.1469260454177856,
+      "learning_rate": 3.334832032625208e-06,
+      "loss": 0.1791,
+      "step": 4688
+    },
+    {
+      "epoch": 2.22030186445694,
+      "grad_norm": 1.130159854888916,
+      "learning_rate": 3.3310203926782826e-06,
+      "loss": 0.1979,
+      "step": 4689
+    },
+    {
+      "epoch": 2.220775377330571,
+      "grad_norm": 0.9808071851730347,
+      "learning_rate": 3.3272104969386388e-06,
+      "loss": 0.178,
+      "step": 4690
+    },
+    {
+      "epoch": 2.2212488902042025,
+      "grad_norm": 1.1407606601715088,
+      "learning_rate": 3.3234023464027143e-06,
+      "loss": 0.1997,
+      "step": 4691
+    },
+    {
+      "epoch": 2.2217224030778335,
+      "grad_norm": 1.0201860666275024,
+      "learning_rate": 3.319595942066498e-06,
+      "loss": 0.2017,
+      "step": 4692
+    },
+    {
+      "epoch": 2.222195915951465,
+      "grad_norm": 1.1310131549835205,
+      "learning_rate": 3.3157912849255137e-06,
+      "loss": 0.2015,
+      "step": 4693
+    },
+    {
+      "epoch": 2.222669428825096,
+      "grad_norm": 1.1622865200042725,
+      "learning_rate": 3.311988375974837e-06,
+      "loss": 0.2173,
+      "step": 4694
+    },
+    {
+      "epoch": 2.2231429416987276,
+      "grad_norm": 1.5009393692016602,
+      "learning_rate": 3.308187216209082e-06,
+      "loss": 0.2147,
+      "step": 4695
+    },
+    {
+      "epoch": 2.2236164545723587,
+      "grad_norm": 1.1541341543197632,
+      "learning_rate": 3.304387806622399e-06,
+      "loss": 0.1975,
+      "step": 4696
+    },
+    {
+      "epoch": 2.2240899674459897,
+      "grad_norm": 1.204338788986206,
+      "learning_rate": 3.3005901482084947e-06,
+      "loss": 0.2123,
+      "step": 4697
+    },
+    {
+      "epoch": 2.2245634803196213,
+      "grad_norm": 1.4117945432662964,
+      "learning_rate": 3.2967942419606045e-06,
+      "loss": 0.2167,
+      "step": 4698
+    },
+    {
+      "epoch": 2.2250369931932523,
+      "grad_norm": 1.0476959943771362,
+      "learning_rate": 3.293000088871515e-06,
+      "loss": 0.2071,
+      "step": 4699
+    },
+    {
+      "epoch": 2.225510506066884,
+      "grad_norm": 1.0269898176193237,
+      "learning_rate": 3.289207689933547e-06,
+      "loss": 0.1831,
+      "step": 4700
+    },
+    {
+      "epoch": 2.225984018940515,
+      "grad_norm": 1.0802260637283325,
+      "learning_rate": 3.2854170461385705e-06,
+      "loss": 0.2284,
+      "step": 4701
+    },
+    {
+      "epoch": 2.2264575318141464,
+      "grad_norm": 1.3022472858428955,
+      "learning_rate": 3.2816281584779853e-06,
+      "loss": 0.2124,
+      "step": 4702
+    },
+    {
+      "epoch": 2.2269310446877775,
+      "grad_norm": 1.3747221231460571,
+      "learning_rate": 3.277841027942745e-06,
+      "loss": 0.203,
+      "step": 4703
+    },
+    {
+      "epoch": 2.2274045575614085,
+      "grad_norm": 1.0478261709213257,
+      "learning_rate": 3.274055655523335e-06,
+      "loss": 0.1948,
+      "step": 4704
+    },
+    {
+      "epoch": 2.22787807043504,
+      "grad_norm": 1.0842070579528809,
+      "learning_rate": 3.2702720422097777e-06,
+      "loss": 0.1837,
+      "step": 4705
+    },
+    {
+      "epoch": 2.228351583308671,
+      "grad_norm": 1.1079217195510864,
+      "learning_rate": 3.2664901889916477e-06,
+      "loss": 0.1974,
+      "step": 4706
+    },
+    {
+      "epoch": 2.2288250961823026,
+      "grad_norm": 1.3619805574417114,
+      "learning_rate": 3.2627100968580472e-06,
+      "loss": 0.1895,
+      "step": 4707
+    },
+    {
+      "epoch": 2.2292986090559337,
+      "grad_norm": 1.1514424085617065,
+      "learning_rate": 3.2589317667976286e-06,
+      "loss": 0.1885,
+      "step": 4708
+    },
+    {
+      "epoch": 2.2297721219295648,
+      "grad_norm": 1.3855830430984497,
+      "learning_rate": 3.25515519979857e-06,
+      "loss": 0.1978,
+      "step": 4709
+    },
+    {
+      "epoch": 2.2302456348031963,
+      "grad_norm": 1.361132025718689,
+      "learning_rate": 3.2513803968486037e-06,
+      "loss": 0.201,
+      "step": 4710
+    },
+    {
+      "epoch": 2.2307191476768273,
+      "grad_norm": 1.0529563426971436,
+      "learning_rate": 3.2476073589349866e-06,
+      "loss": 0.2019,
+      "step": 4711
+    },
+    {
+      "epoch": 2.231192660550459,
+      "grad_norm": 1.162382960319519,
+      "learning_rate": 3.2438360870445263e-06,
+      "loss": 0.2119,
+      "step": 4712
+    },
+    {
+      "epoch": 2.23166617342409,
+      "grad_norm": 0.9659311175346375,
+      "learning_rate": 3.2400665821635568e-06,
+      "loss": 0.2018,
+      "step": 4713
+    },
+    {
+      "epoch": 2.2321396862977214,
+      "grad_norm": 1.2184617519378662,
+      "learning_rate": 3.236298845277961e-06,
+      "loss": 0.2146,
+      "step": 4714
+    },
+    {
+      "epoch": 2.2326131991713525,
+      "grad_norm": 1.6080515384674072,
+      "learning_rate": 3.2325328773731524e-06,
+      "loss": 0.2,
+      "step": 4715
+    },
+    {
+      "epoch": 2.2330867120449835,
+      "grad_norm": 1.5760236978530884,
+      "learning_rate": 3.228768679434079e-06,
+      "loss": 0.2092,
+      "step": 4716
+    },
+    {
+      "epoch": 2.233560224918615,
+      "grad_norm": 1.4468612670898438,
+      "learning_rate": 3.2250062524452376e-06,
+      "loss": 0.2187,
+      "step": 4717
+    },
+    {
+      "epoch": 2.234033737792246,
+      "grad_norm": 1.2942707538604736,
+      "learning_rate": 3.2212455973906477e-06,
+      "loss": 0.1937,
+      "step": 4718
+    },
+    {
+      "epoch": 2.2345072506658776,
+      "grad_norm": 1.237175703048706,
+      "learning_rate": 3.2174867152538802e-06,
+      "loss": 0.213,
+      "step": 4719
+    },
+    {
+      "epoch": 2.2349807635395087,
+      "grad_norm": 1.146020770072937,
+      "learning_rate": 3.213729607018026e-06,
+      "loss": 0.2081,
+      "step": 4720
+    },
+    {
+      "epoch": 2.23545427641314,
+      "grad_norm": 1.1109809875488281,
+      "learning_rate": 3.209974273665726e-06,
+      "loss": 0.2109,
+      "step": 4721
+    },
+    {
+      "epoch": 2.2359277892867713,
+      "grad_norm": 1.0821495056152344,
+      "learning_rate": 3.2062207161791526e-06,
+      "loss": 0.2124,
+      "step": 4722
+    },
+    {
+      "epoch": 2.2364013021604023,
+      "grad_norm": 1.1807968616485596,
+      "learning_rate": 3.2024689355400063e-06,
+      "loss": 0.2362,
+      "step": 4723
+    },
+    {
+      "epoch": 2.236874815034034,
+      "grad_norm": 1.4384002685546875,
+      "learning_rate": 3.1987189327295377e-06,
+      "loss": 0.1825,
+      "step": 4724
+    },
+    {
+      "epoch": 2.237348327907665,
+      "grad_norm": 1.1353758573532104,
+      "learning_rate": 3.1949707087285144e-06,
+      "loss": 0.2115,
+      "step": 4725
+    },
+    {
+      "epoch": 2.2378218407812964,
+      "grad_norm": 2.6835150718688965,
+      "learning_rate": 3.1912242645172576e-06,
+      "loss": 0.2054,
+      "step": 4726
+    },
+    {
+      "epoch": 2.2382953536549275,
+      "grad_norm": 0.9744576811790466,
+      "learning_rate": 3.187479601075605e-06,
+      "loss": 0.1937,
+      "step": 4727
+    },
+    {
+      "epoch": 2.2387688665285586,
+      "grad_norm": 1.31288480758667,
+      "learning_rate": 3.183736719382944e-06,
+      "loss": 0.2199,
+      "step": 4728
+    },
+    {
+      "epoch": 2.23924237940219,
+      "grad_norm": 0.9159910678863525,
+      "learning_rate": 3.179995620418187e-06,
+      "loss": 0.2009,
+      "step": 4729
+    },
+    {
+      "epoch": 2.239715892275821,
+      "grad_norm": 1.6688588857650757,
+      "learning_rate": 3.176256305159778e-06,
+      "loss": 0.1832,
+      "step": 4730
+    },
+    {
+      "epoch": 2.2401894051494526,
+      "grad_norm": 0.9921554923057556,
+      "learning_rate": 3.1725187745857066e-06,
+      "loss": 0.2143,
+      "step": 4731
+    },
+    {
+      "epoch": 2.2406629180230837,
+      "grad_norm": 0.9964454770088196,
+      "learning_rate": 3.16878302967348e-06,
+      "loss": 0.1916,
+      "step": 4732
+    },
+    {
+      "epoch": 2.2411364308967148,
+      "grad_norm": 1.3738844394683838,
+      "learning_rate": 3.1650490714001536e-06,
+      "loss": 0.181,
+      "step": 4733
+    },
+    {
+      "epoch": 2.2416099437703463,
+      "grad_norm": 1.0650712251663208,
+      "learning_rate": 3.1613169007423016e-06,
+      "loss": 0.1985,
+      "step": 4734
+    },
+    {
+      "epoch": 2.2420834566439773,
+      "grad_norm": 0.9678623080253601,
+      "learning_rate": 3.1575865186760425e-06,
+      "loss": 0.1875,
+      "step": 4735
+    },
+    {
+      "epoch": 2.242556969517609,
+      "grad_norm": 1.2123451232910156,
+      "learning_rate": 3.1538579261770177e-06,
+      "loss": 0.2132,
+      "step": 4736
+    },
+    {
+      "epoch": 2.24303048239124,
+      "grad_norm": 1.060880422592163,
+      "learning_rate": 3.150131124220408e-06,
+      "loss": 0.1999,
+      "step": 4737
+    },
+    {
+      "epoch": 2.2435039952648714,
+      "grad_norm": 1.4341163635253906,
+      "learning_rate": 3.1464061137809187e-06,
+      "loss": 0.209,
+      "step": 4738
+    },
+    {
+      "epoch": 2.2439775081385025,
+      "grad_norm": 1.1810160875320435,
+      "learning_rate": 3.142682895832796e-06,
+      "loss": 0.2031,
+      "step": 4739
+    },
+    {
+      "epoch": 2.2444510210121336,
+      "grad_norm": 1.1386125087738037,
+      "learning_rate": 3.1389614713498073e-06,
+      "loss": 0.1958,
+      "step": 4740
+    },
+    {
+      "epoch": 2.244924533885765,
+      "grad_norm": 1.4858670234680176,
+      "learning_rate": 3.1352418413052543e-06,
+      "loss": 0.1942,
+      "step": 4741
+    },
+    {
+      "epoch": 2.245398046759396,
+      "grad_norm": 1.1537786722183228,
+      "learning_rate": 3.131524006671974e-06,
+      "loss": 0.1929,
+      "step": 4742
+    },
+    {
+      "epoch": 2.2458715596330276,
+      "grad_norm": 1.0491985082626343,
+      "learning_rate": 3.127807968422326e-06,
+      "loss": 0.2021,
+      "step": 4743
+    },
+    {
+      "epoch": 2.2463450725066587,
+      "grad_norm": 1.1256279945373535,
+      "learning_rate": 3.1240937275282103e-06,
+      "loss": 0.2091,
+      "step": 4744
+    },
+    {
+      "epoch": 2.2468185853802902,
+      "grad_norm": 1.0205121040344238,
+      "learning_rate": 3.120381284961043e-06,
+      "loss": 0.2042,
+      "step": 4745
+    },
+    {
+      "epoch": 2.2472920982539213,
+      "grad_norm": 1.3747820854187012,
+      "learning_rate": 3.116670641691785e-06,
+      "loss": 0.2037,
+      "step": 4746
+    },
+    {
+      "epoch": 2.2477656111275524,
+      "grad_norm": 1.127557396888733,
+      "learning_rate": 3.112961798690913e-06,
+      "loss": 0.2081,
+      "step": 4747
+    },
+    {
+      "epoch": 2.248239124001184,
+      "grad_norm": 1.0771230459213257,
+      "learning_rate": 3.109254756928445e-06,
+      "loss": 0.1914,
+      "step": 4748
+    },
+    {
+      "epoch": 2.248712636874815,
+      "grad_norm": 1.6226811408996582,
+      "learning_rate": 3.105549517373919e-06,
+      "loss": 0.2036,
+      "step": 4749
+    },
+    {
+      "epoch": 2.2491861497484464,
+      "grad_norm": 1.0237318277359009,
+      "learning_rate": 3.1018460809964025e-06,
+      "loss": 0.2105,
+      "step": 4750
+    },
+    {
+      "epoch": 2.2496596626220775,
+      "grad_norm": 1.0179623365402222,
+      "learning_rate": 3.0981444487644984e-06,
+      "loss": 0.2184,
+      "step": 4751
+    },
+    {
+      "epoch": 2.2501331754957086,
+      "grad_norm": 1.6776036024093628,
+      "learning_rate": 3.0944446216463276e-06,
+      "loss": 0.2038,
+      "step": 4752
+    },
+    {
+      "epoch": 2.25060668836934,
+      "grad_norm": 1.2936748266220093,
+      "learning_rate": 3.09074660060955e-06,
+      "loss": 0.2189,
+      "step": 4753
+    },
+    {
+      "epoch": 2.251080201242971,
+      "grad_norm": 0.9462767839431763,
+      "learning_rate": 3.087050386621341e-06,
+      "loss": 0.2057,
+      "step": 4754
+    },
+    {
+      "epoch": 2.2515537141166027,
+      "grad_norm": 1.1547857522964478,
+      "learning_rate": 3.083355980648416e-06,
+      "loss": 0.2095,
+      "step": 4755
+    },
+    {
+      "epoch": 2.2520272269902337,
+      "grad_norm": 2.062722682952881,
+      "learning_rate": 3.0796633836570055e-06,
+      "loss": 0.1905,
+      "step": 4756
+    },
+    {
+      "epoch": 2.2525007398638652,
+      "grad_norm": 1.355178952217102,
+      "learning_rate": 3.0759725966128774e-06,
+      "loss": 0.2056,
+      "step": 4757
+    },
+    {
+      "epoch": 2.2529742527374963,
+      "grad_norm": 1.0428045988082886,
+      "learning_rate": 3.072283620481321e-06,
+      "loss": 0.1885,
+      "step": 4758
+    },
+    {
+      "epoch": 2.2534477656111274,
+      "grad_norm": 1.1466599702835083,
+      "learning_rate": 3.068596456227143e-06,
+      "loss": 0.2027,
+      "step": 4759
+    },
+    {
+      "epoch": 2.253921278484759,
+      "grad_norm": 1.3236037492752075,
+      "learning_rate": 3.0649111048147006e-06,
+      "loss": 0.2144,
+      "step": 4760
+    },
+    {
+      "epoch": 2.25439479135839,
+      "grad_norm": 1.5951242446899414,
+      "learning_rate": 3.061227567207852e-06,
+      "loss": 0.1973,
+      "step": 4761
+    },
+    {
+      "epoch": 2.2548683042320214,
+      "grad_norm": 1.3796095848083496,
+      "learning_rate": 3.0575458443699957e-06,
+      "loss": 0.1869,
+      "step": 4762
+    },
+    {
+      "epoch": 2.2553418171056525,
+      "grad_norm": 1.0382217168807983,
+      "learning_rate": 3.053865937264049e-06,
+      "loss": 0.1942,
+      "step": 4763
+    },
+    {
+      "epoch": 2.255815329979284,
+      "grad_norm": 0.9059621095657349,
+      "learning_rate": 3.0501878468524525e-06,
+      "loss": 0.1961,
+      "step": 4764
+    },
+    {
+      "epoch": 2.256288842852915,
+      "grad_norm": 1.007012963294983,
+      "learning_rate": 3.046511574097183e-06,
+      "loss": 0.2033,
+      "step": 4765
+    },
+    {
+      "epoch": 2.256762355726546,
+      "grad_norm": 0.9761565327644348,
+      "learning_rate": 3.042837119959726e-06,
+      "loss": 0.2006,
+      "step": 4766
+    },
+    {
+      "epoch": 2.2572358686001777,
+      "grad_norm": 1.1975845098495483,
+      "learning_rate": 3.039164485401106e-06,
+      "loss": 0.2131,
+      "step": 4767
+    },
+    {
+      "epoch": 2.2577093814738087,
+      "grad_norm": 1.2189717292785645,
+      "learning_rate": 3.0354936713818594e-06,
+      "loss": 0.1828,
+      "step": 4768
+    },
+    {
+      "epoch": 2.2581828943474402,
+      "grad_norm": 1.35395085811615,
+      "learning_rate": 3.0318246788620588e-06,
+      "loss": 0.1928,
+      "step": 4769
+    },
+    {
+      "epoch": 2.2586564072210713,
+      "grad_norm": 0.9459072947502136,
+      "learning_rate": 3.028157508801287e-06,
+      "loss": 0.2134,
+      "step": 4770
+    },
+    {
+      "epoch": 2.2591299200947024,
+      "grad_norm": 1.2507688999176025,
+      "learning_rate": 3.0244921621586643e-06,
+      "loss": 0.207,
+      "step": 4771
+    },
+    {
+      "epoch": 2.259603432968334,
+      "grad_norm": 1.5289151668548584,
+      "learning_rate": 3.020828639892818e-06,
+      "loss": 0.2006,
+      "step": 4772
+    },
+    {
+      "epoch": 2.260076945841965,
+      "grad_norm": 1.1962858438491821,
+      "learning_rate": 3.0171669429619154e-06,
+      "loss": 0.2226,
+      "step": 4773
+    },
+    {
+      "epoch": 2.2605504587155965,
+      "grad_norm": 1.2389626502990723,
+      "learning_rate": 3.0135070723236346e-06,
+      "loss": 0.2018,
+      "step": 4774
+    },
+    {
+      "epoch": 2.2610239715892275,
+      "grad_norm": 1.1596938371658325,
+      "learning_rate": 3.0098490289351756e-06,
+      "loss": 0.1978,
+      "step": 4775
+    },
+    {
+      "epoch": 2.2614974844628586,
+      "grad_norm": 1.206876516342163,
+      "learning_rate": 3.0061928137532713e-06,
+      "loss": 0.1913,
+      "step": 4776
+    },
+    {
+      "epoch": 2.26197099733649,
+      "grad_norm": 1.2051137685775757,
+      "learning_rate": 3.002538427734163e-06,
+      "loss": 0.2041,
+      "step": 4777
+    },
+    {
+      "epoch": 2.262444510210121,
+      "grad_norm": 0.8974770903587341,
+      "learning_rate": 2.9988858718336256e-06,
+      "loss": 0.2054,
+      "step": 4778
+    },
+    {
+      "epoch": 2.2629180230837527,
+      "grad_norm": 1.372613787651062,
+      "learning_rate": 2.995235147006945e-06,
+      "loss": 0.1953,
+      "step": 4779
+    },
+    {
+      "epoch": 2.2633915359573837,
+      "grad_norm": 1.115798830986023,
+      "learning_rate": 2.991586254208939e-06,
+      "loss": 0.213,
+      "step": 4780
+    },
+    {
+      "epoch": 2.2638650488310152,
+      "grad_norm": 1.4372373819351196,
+      "learning_rate": 2.987939194393933e-06,
+      "loss": 0.2302,
+      "step": 4781
+    },
+    {
+      "epoch": 2.2643385617046463,
+      "grad_norm": 0.9313597679138184,
+      "learning_rate": 2.984293968515788e-06,
+      "loss": 0.2021,
+      "step": 4782
+    },
+    {
+      "epoch": 2.264812074578278,
+      "grad_norm": 1.0332454442977905,
+      "learning_rate": 2.9806505775278738e-06,
+      "loss": 0.1874,
+      "step": 4783
+    },
+    {
+      "epoch": 2.265285587451909,
+      "grad_norm": 1.0937074422836304,
+      "learning_rate": 2.9770090223830803e-06,
+      "loss": 0.199,
+      "step": 4784
+    },
+    {
+      "epoch": 2.26575910032554,
+      "grad_norm": 1.022531509399414,
+      "learning_rate": 2.9733693040338286e-06,
+      "loss": 0.1963,
+      "step": 4785
+    },
+    {
+      "epoch": 2.2662326131991715,
+      "grad_norm": 1.1225355863571167,
+      "learning_rate": 2.969731423432045e-06,
+      "loss": 0.2178,
+      "step": 4786
+    },
+    {
+      "epoch": 2.2667061260728025,
+      "grad_norm": 1.8823810815811157,
+      "learning_rate": 2.9660953815291893e-06,
+      "loss": 0.2103,
+      "step": 4787
+    },
+    {
+      "epoch": 2.267179638946434,
+      "grad_norm": 1.7231241464614868,
+      "learning_rate": 2.962461179276225e-06,
+      "loss": 0.1777,
+      "step": 4788
+    },
+    {
+      "epoch": 2.267653151820065,
+      "grad_norm": 1.254069447517395,
+      "learning_rate": 2.9588288176236502e-06,
+      "loss": 0.2163,
+      "step": 4789
+    },
+    {
+      "epoch": 2.268126664693696,
+      "grad_norm": 1.0961583852767944,
+      "learning_rate": 2.955198297521469e-06,
+      "loss": 0.2097,
+      "step": 4790
+    },
+    {
+      "epoch": 2.2686001775673277,
+      "grad_norm": 1.0620346069335938,
+      "learning_rate": 2.9515696199192123e-06,
+      "loss": 0.1878,
+      "step": 4791
+    },
+    {
+      "epoch": 2.2690736904409587,
+      "grad_norm": 1.0803302526474,
+      "learning_rate": 2.9479427857659213e-06,
+      "loss": 0.2128,
+      "step": 4792
+    },
+    {
+      "epoch": 2.2695472033145903,
+      "grad_norm": 1.3696365356445312,
+      "learning_rate": 2.9443177960101653e-06,
+      "loss": 0.2047,
+      "step": 4793
+    },
+    {
+      "epoch": 2.2700207161882213,
+      "grad_norm": 1.0663868188858032,
+      "learning_rate": 2.9406946516000236e-06,
+      "loss": 0.1987,
+      "step": 4794
+    },
+    {
+      "epoch": 2.2704942290618524,
+      "grad_norm": 1.149569034576416,
+      "learning_rate": 2.9370733534830887e-06,
+      "loss": 0.1871,
+      "step": 4795
+    },
+    {
+      "epoch": 2.270967741935484,
+      "grad_norm": 1.892318606376648,
+      "learning_rate": 2.9334539026064847e-06,
+      "loss": 0.1883,
+      "step": 4796
+    },
+    {
+      "epoch": 2.271441254809115,
+      "grad_norm": 1.3495607376098633,
+      "learning_rate": 2.9298362999168373e-06,
+      "loss": 0.2193,
+      "step": 4797
+    },
+    {
+      "epoch": 2.2719147676827465,
+      "grad_norm": 1.4502182006835938,
+      "learning_rate": 2.926220546360299e-06,
+      "loss": 0.1985,
+      "step": 4798
+    },
+    {
+      "epoch": 2.2723882805563775,
+      "grad_norm": 1.812647819519043,
+      "learning_rate": 2.922606642882537e-06,
+      "loss": 0.2017,
+      "step": 4799
+    },
+    {
+      "epoch": 2.272861793430009,
+      "grad_norm": 1.912377953529358,
+      "learning_rate": 2.9189945904287287e-06,
+      "loss": 0.2218,
+      "step": 4800
+    },
+    {
+      "epoch": 2.27333530630364,
+      "grad_norm": 1.0828890800476074,
+      "learning_rate": 2.915384389943576e-06,
+      "loss": 0.1887,
+      "step": 4801
+    },
+    {
+      "epoch": 2.2738088191772716,
+      "grad_norm": 1.1113115549087524,
+      "learning_rate": 2.911776042371286e-06,
+      "loss": 0.1963,
+      "step": 4802
+    },
+    {
+      "epoch": 2.2742823320509027,
+      "grad_norm": 1.5018105506896973,
+      "learning_rate": 2.908169548655595e-06,
+      "loss": 0.1994,
+      "step": 4803
+    },
+    {
+      "epoch": 2.2747558449245338,
+      "grad_norm": 1.3521610498428345,
+      "learning_rate": 2.9045649097397386e-06,
+      "loss": 0.2026,
+      "step": 4804
+    },
+    {
+      "epoch": 2.2752293577981653,
+      "grad_norm": 1.7627965211868286,
+      "learning_rate": 2.9009621265664832e-06,
+      "loss": 0.1797,
+      "step": 4805
+    },
+    {
+      "epoch": 2.2757028706717963,
+      "grad_norm": 1.138641357421875,
+      "learning_rate": 2.8973612000780937e-06,
+      "loss": 0.2053,
+      "step": 4806
+    },
+    {
+      "epoch": 2.276176383545428,
+      "grad_norm": 1.8081318140029907,
+      "learning_rate": 2.8937621312163653e-06,
+      "loss": 0.1976,
+      "step": 4807
+    },
+    {
+      "epoch": 2.276649896419059,
+      "grad_norm": 1.7190033197402954,
+      "learning_rate": 2.890164920922597e-06,
+      "loss": 0.2069,
+      "step": 4808
+    },
+    {
+      "epoch": 2.27712340929269,
+      "grad_norm": 1.0892248153686523,
+      "learning_rate": 2.8865695701376005e-06,
+      "loss": 0.2048,
+      "step": 4809
+    },
+    {
+      "epoch": 2.2775969221663215,
+      "grad_norm": 1.098090648651123,
+      "learning_rate": 2.8829760798017115e-06,
+      "loss": 0.193,
+      "step": 4810
+    },
+    {
+      "epoch": 2.2780704350399525,
+      "grad_norm": 1.2340128421783447,
+      "learning_rate": 2.8793844508547664e-06,
+      "loss": 0.1822,
+      "step": 4811
+    },
+    {
+      "epoch": 2.278543947913584,
+      "grad_norm": 0.9539898037910461,
+      "learning_rate": 2.875794684236127e-06,
+      "loss": 0.1829,
+      "step": 4812
+    },
+    {
+      "epoch": 2.279017460787215,
+      "grad_norm": 1.3292118310928345,
+      "learning_rate": 2.8722067808846575e-06,
+      "loss": 0.2284,
+      "step": 4813
+    },
+    {
+      "epoch": 2.279490973660846,
+      "grad_norm": 1.208977460861206,
+      "learning_rate": 2.8686207417387446e-06,
+      "loss": 0.1792,
+      "step": 4814
+    },
+    {
+      "epoch": 2.2799644865344777,
+      "grad_norm": 1.084511637687683,
+      "learning_rate": 2.8650365677362756e-06,
+      "loss": 0.1716,
+      "step": 4815
+    },
+    {
+      "epoch": 2.2804379994081088,
+      "grad_norm": 1.649825930595398,
+      "learning_rate": 2.861454259814662e-06,
+      "loss": 0.2078,
+      "step": 4816
+    },
+    {
+      "epoch": 2.2809115122817403,
+      "grad_norm": 1.3843883275985718,
+      "learning_rate": 2.857873818910821e-06,
+      "loss": 0.219,
+      "step": 4817
+    },
+    {
+      "epoch": 2.2813850251553713,
+      "grad_norm": 1.0740282535552979,
+      "learning_rate": 2.854295245961178e-06,
+      "loss": 0.1984,
+      "step": 4818
+    },
+    {
+      "epoch": 2.281858538029003,
+      "grad_norm": 1.2983367443084717,
+      "learning_rate": 2.8507185419016813e-06,
+      "loss": 0.213,
+      "step": 4819
+    },
+    {
+      "epoch": 2.282332050902634,
+      "grad_norm": 1.1610127687454224,
+      "learning_rate": 2.8471437076677767e-06,
+      "loss": 0.2434,
+      "step": 4820
+    },
+    {
+      "epoch": 2.282805563776265,
+      "grad_norm": 1.033072829246521,
+      "learning_rate": 2.8435707441944325e-06,
+      "loss": 0.195,
+      "step": 4821
+    },
+    {
+      "epoch": 2.2832790766498965,
+      "grad_norm": 1.0944322347640991,
+      "learning_rate": 2.839999652416119e-06,
+      "loss": 0.2027,
+      "step": 4822
+    },
+    {
+      "epoch": 2.2837525895235276,
+      "grad_norm": 1.1411755084991455,
+      "learning_rate": 2.8364304332668245e-06,
+      "loss": 0.2242,
+      "step": 4823
+    },
+    {
+      "epoch": 2.284226102397159,
+      "grad_norm": 1.6203806400299072,
+      "learning_rate": 2.83286308768004e-06,
+      "loss": 0.2057,
+      "step": 4824
+    },
+    {
+      "epoch": 2.28469961527079,
+      "grad_norm": 1.048741102218628,
+      "learning_rate": 2.829297616588775e-06,
+      "loss": 0.2284,
+      "step": 4825
+    },
+    {
+      "epoch": 2.2851731281444216,
+      "grad_norm": 1.2415732145309448,
+      "learning_rate": 2.825734020925538e-06,
+      "loss": 0.1923,
+      "step": 4826
+    },
+    {
+      "epoch": 2.2856466410180527,
+      "grad_norm": 1.5212392807006836,
+      "learning_rate": 2.82217230162236e-06,
+      "loss": 0.2163,
+      "step": 4827
+    },
+    {
+      "epoch": 2.2861201538916838,
+      "grad_norm": 1.1868995428085327,
+      "learning_rate": 2.8186124596107713e-06,
+      "loss": 0.201,
+      "step": 4828
+    },
+    {
+      "epoch": 2.2865936667653153,
+      "grad_norm": 1.1392196416854858,
+      "learning_rate": 2.8150544958218097e-06,
+      "loss": 0.2098,
+      "step": 4829
+    },
+    {
+      "epoch": 2.2870671796389463,
+      "grad_norm": 1.0334511995315552,
+      "learning_rate": 2.8114984111860334e-06,
+      "loss": 0.2125,
+      "step": 4830
+    },
+    {
+      "epoch": 2.287540692512578,
+      "grad_norm": 1.3680204153060913,
+      "learning_rate": 2.8079442066334963e-06,
+      "loss": 0.1946,
+      "step": 4831
+    },
+    {
+      "epoch": 2.288014205386209,
+      "grad_norm": 1.040935754776001,
+      "learning_rate": 2.8043918830937713e-06,
+      "loss": 0.1835,
+      "step": 4832
+    },
+    {
+      "epoch": 2.28848771825984,
+      "grad_norm": 0.9355210661888123,
+      "learning_rate": 2.8008414414959295e-06,
+      "loss": 0.1898,
+      "step": 4833
+    },
+    {
+      "epoch": 2.2889612311334715,
+      "grad_norm": 1.3512656688690186,
+      "learning_rate": 2.7972928827685597e-06,
+      "loss": 0.1793,
+      "step": 4834
+    },
+    {
+      "epoch": 2.2894347440071026,
+      "grad_norm": 1.103294849395752,
+      "learning_rate": 2.793746207839748e-06,
+      "loss": 0.2012,
+      "step": 4835
+    },
+    {
+      "epoch": 2.289908256880734,
+      "grad_norm": 1.275721549987793,
+      "learning_rate": 2.7902014176370996e-06,
+      "loss": 0.2176,
+      "step": 4836
+    },
+    {
+      "epoch": 2.290381769754365,
+      "grad_norm": 1.1361020803451538,
+      "learning_rate": 2.786658513087712e-06,
+      "loss": 0.2,
+      "step": 4837
+    },
+    {
+      "epoch": 2.290855282627996,
+      "grad_norm": 1.2212600708007812,
+      "learning_rate": 2.7831174951182027e-06,
+      "loss": 0.2056,
+      "step": 4838
+    },
+    {
+      "epoch": 2.2913287955016277,
+      "grad_norm": 1.6602025032043457,
+      "learning_rate": 2.7795783646546924e-06,
+      "loss": 0.2165,
+      "step": 4839
+    },
+    {
+      "epoch": 2.291802308375259,
+      "grad_norm": 1.3278297185897827,
+      "learning_rate": 2.7760411226228022e-06,
+      "loss": 0.193,
+      "step": 4840
+    },
+    {
+      "epoch": 2.2922758212488903,
+      "grad_norm": 1.0122475624084473,
+      "learning_rate": 2.7725057699476674e-06,
+      "loss": 0.1895,
+      "step": 4841
+    },
+    {
+      "epoch": 2.2927493341225214,
+      "grad_norm": 1.3454642295837402,
+      "learning_rate": 2.7689723075539245e-06,
+      "loss": 0.1943,
+      "step": 4842
+    },
+    {
+      "epoch": 2.293222846996153,
+      "grad_norm": 1.102555274963379,
+      "learning_rate": 2.765440736365712e-06,
+      "loss": 0.2111,
+      "step": 4843
+    },
+    {
+      "epoch": 2.293696359869784,
+      "grad_norm": 1.9383556842803955,
+      "learning_rate": 2.7619110573066856e-06,
+      "loss": 0.1992,
+      "step": 4844
+    },
+    {
+      "epoch": 2.2941698727434154,
+      "grad_norm": 1.0931776762008667,
+      "learning_rate": 2.7583832712999912e-06,
+      "loss": 0.2254,
+      "step": 4845
+    },
+    {
+      "epoch": 2.2946433856170465,
+      "grad_norm": 1.0424646139144897,
+      "learning_rate": 2.754857379268294e-06,
+      "loss": 0.2012,
+      "step": 4846
+    },
+    {
+      "epoch": 2.2951168984906776,
+      "grad_norm": 1.0645688772201538,
+      "learning_rate": 2.7513333821337516e-06,
+      "loss": 0.1864,
+      "step": 4847
+    },
+    {
+      "epoch": 2.295590411364309,
+      "grad_norm": 1.6237589120864868,
+      "learning_rate": 2.7478112808180378e-06,
+      "loss": 0.175,
+      "step": 4848
+    },
+    {
+      "epoch": 2.29606392423794,
+      "grad_norm": 1.3338507413864136,
+      "learning_rate": 2.744291076242317e-06,
+      "loss": 0.1883,
+      "step": 4849
+    },
+    {
+      "epoch": 2.2965374371115717,
+      "grad_norm": 2.062040328979492,
+      "learning_rate": 2.74077276932727e-06,
+      "loss": 0.1894,
+      "step": 4850
+    },
+    {
+      "epoch": 2.2970109499852027,
+      "grad_norm": 1.056857705116272,
+      "learning_rate": 2.7372563609930726e-06,
+      "loss": 0.1875,
+      "step": 4851
+    },
+    {
+      "epoch": 2.297484462858834,
+      "grad_norm": 1.0261149406433105,
+      "learning_rate": 2.7337418521594107e-06,
+      "loss": 0.19,
+      "step": 4852
+    },
+    {
+      "epoch": 2.2979579757324653,
+      "grad_norm": 1.3251570463180542,
+      "learning_rate": 2.730229243745469e-06,
+      "loss": 0.215,
+      "step": 4853
+    },
+    {
+      "epoch": 2.2984314886060964,
+      "grad_norm": 0.9587031602859497,
+      "learning_rate": 2.726718536669933e-06,
+      "loss": 0.2052,
+      "step": 4854
+    },
+    {
+      "epoch": 2.298905001479728,
+      "grad_norm": 0.987009584903717,
+      "learning_rate": 2.7232097318510007e-06,
+      "loss": 0.1793,
+      "step": 4855
+    },
+    {
+      "epoch": 2.299378514353359,
+      "grad_norm": 1.4788868427276611,
+      "learning_rate": 2.7197028302063587e-06,
+      "loss": 0.2226,
+      "step": 4856
+    },
+    {
+      "epoch": 2.29985202722699,
+      "grad_norm": 1.2045164108276367,
+      "learning_rate": 2.716197832653211e-06,
+      "loss": 0.1932,
+      "step": 4857
+    },
+    {
+      "epoch": 2.3003255401006215,
+      "grad_norm": 1.0435082912445068,
+      "learning_rate": 2.7126947401082494e-06,
+      "loss": 0.2083,
+      "step": 4858
+    },
+    {
+      "epoch": 2.3007990529742526,
+      "grad_norm": 1.5519726276397705,
+      "learning_rate": 2.709193553487679e-06,
+      "loss": 0.2008,
+      "step": 4859
+    },
+    {
+      "epoch": 2.301272565847884,
+      "grad_norm": 1.805208444595337,
+      "learning_rate": 2.7056942737071955e-06,
+      "loss": 0.2021,
+      "step": 4860
+    },
+    {
+      "epoch": 2.301746078721515,
+      "grad_norm": 1.2867292165756226,
+      "learning_rate": 2.702196901682009e-06,
+      "loss": 0.1852,
+      "step": 4861
+    },
+    {
+      "epoch": 2.3022195915951467,
+      "grad_norm": 1.190337061882019,
+      "learning_rate": 2.6987014383268196e-06,
+      "loss": 0.1956,
+      "step": 4862
+    },
+    {
+      "epoch": 2.3026931044687777,
+      "grad_norm": 1.159690499305725,
+      "learning_rate": 2.6952078845558292e-06,
+      "loss": 0.1994,
+      "step": 4863
+    },
+    {
+      "epoch": 2.3031666173424092,
+      "grad_norm": 1.3060331344604492,
+      "learning_rate": 2.691716241282748e-06,
+      "loss": 0.2027,
+      "step": 4864
+    },
+    {
+      "epoch": 2.3036401302160403,
+      "grad_norm": 1.173386812210083,
+      "learning_rate": 2.6882265094207783e-06,
+      "loss": 0.1967,
+      "step": 4865
+    },
+    {
+      "epoch": 2.3041136430896714,
+      "grad_norm": 1.255037784576416,
+      "learning_rate": 2.684738689882629e-06,
+      "loss": 0.1947,
+      "step": 4866
+    },
+    {
+      "epoch": 2.304587155963303,
+      "grad_norm": 1.0719294548034668,
+      "learning_rate": 2.6812527835805013e-06,
+      "loss": 0.2067,
+      "step": 4867
+    },
+    {
+      "epoch": 2.305060668836934,
+      "grad_norm": 1.0298558473587036,
+      "learning_rate": 2.6777687914261054e-06,
+      "loss": 0.1918,
+      "step": 4868
+    },
+    {
+      "epoch": 2.3055341817105655,
+      "grad_norm": 1.2399965524673462,
+      "learning_rate": 2.6742867143306404e-06,
+      "loss": 0.2119,
+      "step": 4869
+    },
+    {
+      "epoch": 2.3060076945841965,
+      "grad_norm": 1.3060729503631592,
+      "learning_rate": 2.6708065532048167e-06,
+      "loss": 0.2244,
+      "step": 4870
+    },
+    {
+      "epoch": 2.3064812074578276,
+      "grad_norm": 1.0601017475128174,
+      "learning_rate": 2.6673283089588286e-06,
+      "loss": 0.1678,
+      "step": 4871
+    },
+    {
+      "epoch": 2.306954720331459,
+      "grad_norm": 1.461519718170166,
+      "learning_rate": 2.6638519825023855e-06,
+      "loss": 0.2043,
+      "step": 4872
+    },
+    {
+      "epoch": 2.30742823320509,
+      "grad_norm": 1.202025055885315,
+      "learning_rate": 2.6603775747446836e-06,
+      "loss": 0.219,
+      "step": 4873
+    },
+    {
+      "epoch": 2.3079017460787217,
+      "grad_norm": 1.5408061742782593,
+      "learning_rate": 2.6569050865944168e-06,
+      "loss": 0.2058,
+      "step": 4874
+    },
+    {
+      "epoch": 2.3083752589523527,
+      "grad_norm": 1.1081178188323975,
+      "learning_rate": 2.653434518959788e-06,
+      "loss": 0.2056,
+      "step": 4875
+    },
+    {
+      "epoch": 2.308848771825984,
+      "grad_norm": 1.2208150625228882,
+      "learning_rate": 2.649965872748481e-06,
+      "loss": 0.2119,
+      "step": 4876
+    },
+    {
+      "epoch": 2.3093222846996153,
+      "grad_norm": 0.9531561732292175,
+      "learning_rate": 2.6464991488676996e-06,
+      "loss": 0.175,
+      "step": 4877
+    },
+    {
+      "epoch": 2.3097957975732464,
+      "grad_norm": 1.0431057214736938,
+      "learning_rate": 2.6430343482241237e-06,
+      "loss": 0.2357,
+      "step": 4878
+    },
+    {
+      "epoch": 2.310269310446878,
+      "grad_norm": 1.079884648323059,
+      "learning_rate": 2.6395714717239384e-06,
+      "loss": 0.2059,
+      "step": 4879
+    },
+    {
+      "epoch": 2.310742823320509,
+      "grad_norm": 1.2502458095550537,
+      "learning_rate": 2.63611052027283e-06,
+      "loss": 0.2051,
+      "step": 4880
+    },
+    {
+      "epoch": 2.3112163361941405,
+      "grad_norm": 1.1873927116394043,
+      "learning_rate": 2.6326514947759718e-06,
+      "loss": 0.2142,
+      "step": 4881
+    },
+    {
+      "epoch": 2.3116898490677715,
+      "grad_norm": 1.1415022611618042,
+      "learning_rate": 2.629194396138044e-06,
+      "loss": 0.1781,
+      "step": 4882
+    },
+    {
+      "epoch": 2.3121633619414026,
+      "grad_norm": 1.0206491947174072,
+      "learning_rate": 2.625739225263211e-06,
+      "loss": 0.1994,
+      "step": 4883
+    },
+    {
+      "epoch": 2.312636874815034,
+      "grad_norm": 1.136160969734192,
+      "learning_rate": 2.622285983055146e-06,
+      "loss": 0.1869,
+      "step": 4884
+    },
+    {
+      "epoch": 2.313110387688665,
+      "grad_norm": 1.2701917886734009,
+      "learning_rate": 2.6188346704170053e-06,
+      "loss": 0.2147,
+      "step": 4885
+    },
+    {
+      "epoch": 2.3135839005622967,
+      "grad_norm": 1.050111174583435,
+      "learning_rate": 2.615385288251452e-06,
+      "loss": 0.2092,
+      "step": 4886
+    },
+    {
+      "epoch": 2.3140574134359277,
+      "grad_norm": 1.1938689947128296,
+      "learning_rate": 2.6119378374606354e-06,
+      "loss": 0.2094,
+      "step": 4887
+    },
+    {
+      "epoch": 2.3145309263095593,
+      "grad_norm": 1.2285950183868408,
+      "learning_rate": 2.608492318946201e-06,
+      "loss": 0.1953,
+      "step": 4888
+    },
+    {
+      "epoch": 2.3150044391831903,
+      "grad_norm": 2.1110057830810547,
+      "learning_rate": 2.6050487336092967e-06,
+      "loss": 0.2099,
+      "step": 4889
+    },
+    {
+      "epoch": 2.3154779520568214,
+      "grad_norm": 1.1653103828430176,
+      "learning_rate": 2.6016070823505525e-06,
+      "loss": 0.1689,
+      "step": 4890
+    },
+    {
+      "epoch": 2.315951464930453,
+      "grad_norm": 1.4252010583877563,
+      "learning_rate": 2.5981673660701055e-06,
+      "loss": 0.1917,
+      "step": 4891
+    },
+    {
+      "epoch": 2.316424977804084,
+      "grad_norm": 1.0870301723480225,
+      "learning_rate": 2.5947295856675737e-06,
+      "loss": 0.1839,
+      "step": 4892
+    },
+    {
+      "epoch": 2.3168984906777155,
+      "grad_norm": 1.1563901901245117,
+      "learning_rate": 2.5912937420420823e-06,
+      "loss": 0.1985,
+      "step": 4893
+    },
+    {
+      "epoch": 2.3173720035513465,
+      "grad_norm": 0.9440892338752747,
+      "learning_rate": 2.587859836092237e-06,
+      "loss": 0.1894,
+      "step": 4894
+    },
+    {
+      "epoch": 2.3178455164249776,
+      "grad_norm": 1.289351463317871,
+      "learning_rate": 2.5844278687161474e-06,
+      "loss": 0.2051,
+      "step": 4895
+    },
+    {
+      "epoch": 2.318319029298609,
+      "grad_norm": 1.7012381553649902,
+      "learning_rate": 2.58099784081141e-06,
+      "loss": 0.1951,
+      "step": 4896
+    },
+    {
+      "epoch": 2.31879254217224,
+      "grad_norm": 0.9597797393798828,
+      "learning_rate": 2.577569753275112e-06,
+      "loss": 0.2101,
+      "step": 4897
+    },
+    {
+      "epoch": 2.3192660550458717,
+      "grad_norm": 1.141579270362854,
+      "learning_rate": 2.574143607003843e-06,
+      "loss": 0.2081,
+      "step": 4898
+    },
+    {
+      "epoch": 2.3197395679195028,
+      "grad_norm": 1.3628860712051392,
+      "learning_rate": 2.570719402893671e-06,
+      "loss": 0.2154,
+      "step": 4899
+    },
+    {
+      "epoch": 2.320213080793134,
+      "grad_norm": 1.7435754537582397,
+      "learning_rate": 2.5672971418401716e-06,
+      "loss": 0.1822,
+      "step": 4900
+    },
+    {
+      "epoch": 2.3206865936667653,
+      "grad_norm": 1.5572841167449951,
+      "learning_rate": 2.5638768247383962e-06,
+      "loss": 0.1956,
+      "step": 4901
+    },
+    {
+      "epoch": 2.3211601065403964,
+      "grad_norm": 1.8090298175811768,
+      "learning_rate": 2.5604584524829036e-06,
+      "loss": 0.1927,
+      "step": 4902
+    },
+    {
+      "epoch": 2.321633619414028,
+      "grad_norm": 1.089689016342163,
+      "learning_rate": 2.5570420259677285e-06,
+      "loss": 0.2198,
+      "step": 4903
+    },
+    {
+      "epoch": 2.322107132287659,
+      "grad_norm": 0.9589508175849915,
+      "learning_rate": 2.553627546086411e-06,
+      "loss": 0.2007,
+      "step": 4904
+    },
+    {
+      "epoch": 2.3225806451612905,
+      "grad_norm": 0.8565977811813354,
+      "learning_rate": 2.55021501373197e-06,
+      "loss": 0.1745,
+      "step": 4905
+    },
+    {
+      "epoch": 2.3230541580349215,
+      "grad_norm": 1.0628879070281982,
+      "learning_rate": 2.5468044297969265e-06,
+      "loss": 0.1977,
+      "step": 4906
+    },
+    {
+      "epoch": 2.323527670908553,
+      "grad_norm": 1.0389453172683716,
+      "learning_rate": 2.543395795173281e-06,
+      "loss": 0.1992,
+      "step": 4907
+    },
+    {
+      "epoch": 2.324001183782184,
+      "grad_norm": 1.0597020387649536,
+      "learning_rate": 2.5399891107525277e-06,
+      "loss": 0.2128,
+      "step": 4908
+    },
+    {
+      "epoch": 2.324474696655815,
+      "grad_norm": 1.1213783025741577,
+      "learning_rate": 2.5365843774256573e-06,
+      "loss": 0.1916,
+      "step": 4909
+    },
+    {
+      "epoch": 2.3249482095294467,
+      "grad_norm": 1.2535953521728516,
+      "learning_rate": 2.5331815960831387e-06,
+      "loss": 0.1868,
+      "step": 4910
+    },
+    {
+      "epoch": 2.3254217224030778,
+      "grad_norm": 1.0827935934066772,
+      "learning_rate": 2.5297807676149435e-06,
+      "loss": 0.2103,
+      "step": 4911
+    },
+    {
+      "epoch": 2.3258952352767093,
+      "grad_norm": 1.4444578886032104,
+      "learning_rate": 2.526381892910519e-06,
+      "loss": 0.2075,
+      "step": 4912
+    },
+    {
+      "epoch": 2.3263687481503403,
+      "grad_norm": 1.0366746187210083,
+      "learning_rate": 2.5229849728588143e-06,
+      "loss": 0.2126,
+      "step": 4913
+    },
+    {
+      "epoch": 2.3268422610239714,
+      "grad_norm": 1.0630645751953125,
+      "learning_rate": 2.519590008348255e-06,
+      "loss": 0.2182,
+      "step": 4914
+    },
+    {
+      "epoch": 2.327315773897603,
+      "grad_norm": 1.4311585426330566,
+      "learning_rate": 2.5161970002667636e-06,
+      "loss": 0.2224,
+      "step": 4915
+    },
+    {
+      "epoch": 2.327789286771234,
+      "grad_norm": 1.9269888401031494,
+      "learning_rate": 2.512805949501752e-06,
+      "loss": 0.1924,
+      "step": 4916
+    },
+    {
+      "epoch": 2.3282627996448655,
+      "grad_norm": 1.697356939315796,
+      "learning_rate": 2.5094168569401123e-06,
+      "loss": 0.2131,
+      "step": 4917
+    },
+    {
+      "epoch": 2.3287363125184966,
+      "grad_norm": 1.6500767469406128,
+      "learning_rate": 2.5060297234682328e-06,
+      "loss": 0.2155,
+      "step": 4918
+    },
+    {
+      "epoch": 2.3292098253921276,
+      "grad_norm": 1.0592091083526611,
+      "learning_rate": 2.502644549971981e-06,
+      "loss": 0.209,
+      "step": 4919
+    },
+    {
+      "epoch": 2.329683338265759,
+      "grad_norm": 1.3273735046386719,
+      "learning_rate": 2.499261337336721e-06,
+      "loss": 0.2025,
+      "step": 4920
+    },
+    {
+      "epoch": 2.33015685113939,
+      "grad_norm": 2.2133803367614746,
+      "learning_rate": 2.4958800864472974e-06,
+      "loss": 0.1929,
+      "step": 4921
+    },
+    {
+      "epoch": 2.3306303640130217,
+      "grad_norm": 1.3577215671539307,
+      "learning_rate": 2.492500798188042e-06,
+      "loss": 0.2091,
+      "step": 4922
+    },
+    {
+      "epoch": 2.3311038768866528,
+      "grad_norm": 1.0294976234436035,
+      "learning_rate": 2.4891234734427784e-06,
+      "loss": 0.226,
+      "step": 4923
+    },
+    {
+      "epoch": 2.3315773897602843,
+      "grad_norm": 1.083409070968628,
+      "learning_rate": 2.485748113094809e-06,
+      "loss": 0.208,
+      "step": 4924
+    },
+    {
+      "epoch": 2.3320509026339153,
+      "grad_norm": 1.1658552885055542,
+      "learning_rate": 2.4823747180269332e-06,
+      "loss": 0.2045,
+      "step": 4925
+    },
+    {
+      "epoch": 2.332524415507547,
+      "grad_norm": 1.3609403371810913,
+      "learning_rate": 2.479003289121422e-06,
+      "loss": 0.2003,
+      "step": 4926
+    },
+    {
+      "epoch": 2.332997928381178,
+      "grad_norm": 0.9218862652778625,
+      "learning_rate": 2.4756338272600476e-06,
+      "loss": 0.1993,
+      "step": 4927
+    },
+    {
+      "epoch": 2.333471441254809,
+      "grad_norm": 1.0133236646652222,
+      "learning_rate": 2.472266333324054e-06,
+      "loss": 0.2027,
+      "step": 4928
+    },
+    {
+      "epoch": 2.3339449541284405,
+      "grad_norm": 0.9589722752571106,
+      "learning_rate": 2.4689008081941825e-06,
+      "loss": 0.1872,
+      "step": 4929
+    },
+    {
+      "epoch": 2.3344184670020716,
+      "grad_norm": 1.2544256448745728,
+      "learning_rate": 2.4655372527506473e-06,
+      "loss": 0.2071,
+      "step": 4930
+    },
+    {
+      "epoch": 2.334891979875703,
+      "grad_norm": 1.001051902770996,
+      "learning_rate": 2.462175667873161e-06,
+      "loss": 0.2062,
+      "step": 4931
+    },
+    {
+      "epoch": 2.335365492749334,
+      "grad_norm": 1.142121434211731,
+      "learning_rate": 2.45881605444091e-06,
+      "loss": 0.1953,
+      "step": 4932
+    },
+    {
+      "epoch": 2.335839005622965,
+      "grad_norm": 1.1744766235351562,
+      "learning_rate": 2.4554584133325653e-06,
+      "loss": 0.1917,
+      "step": 4933
+    },
+    {
+      "epoch": 2.3363125184965967,
+      "grad_norm": 1.3197706937789917,
+      "learning_rate": 2.4521027454262925e-06,
+      "loss": 0.1952,
+      "step": 4934
+    },
+    {
+      "epoch": 2.336786031370228,
+      "grad_norm": 1.6744234561920166,
+      "learning_rate": 2.4487490515997282e-06,
+      "loss": 0.2136,
+      "step": 4935
+    },
+    {
+      "epoch": 2.3372595442438593,
+      "grad_norm": 1.0961428880691528,
+      "learning_rate": 2.445397332730004e-06,
+      "loss": 0.2046,
+      "step": 4936
+    },
+    {
+      "epoch": 2.3377330571174904,
+      "grad_norm": 0.8409469127655029,
+      "learning_rate": 2.4420475896937236e-06,
+      "loss": 0.1953,
+      "step": 4937
+    },
+    {
+      "epoch": 2.3382065699911214,
+      "grad_norm": 1.0160478353500366,
+      "learning_rate": 2.4386998233669867e-06,
+      "loss": 0.1908,
+      "step": 4938
+    },
+    {
+      "epoch": 2.338680082864753,
+      "grad_norm": 1.2315906286239624,
+      "learning_rate": 2.4353540346253622e-06,
+      "loss": 0.212,
+      "step": 4939
+    },
+    {
+      "epoch": 2.339153595738384,
+      "grad_norm": 1.112743616104126,
+      "learning_rate": 2.4320102243439157e-06,
+      "loss": 0.2025,
+      "step": 4940
+    },
+    {
+      "epoch": 2.3396271086120155,
+      "grad_norm": 1.1238675117492676,
+      "learning_rate": 2.4286683933971853e-06,
+      "loss": 0.2032,
+      "step": 4941
+    },
+    {
+      "epoch": 2.3401006214856466,
+      "grad_norm": 1.2250267267227173,
+      "learning_rate": 2.4253285426591923e-06,
+      "loss": 0.2298,
+      "step": 4942
+    },
+    {
+      "epoch": 2.340574134359278,
+      "grad_norm": 1.391561508178711,
+      "learning_rate": 2.4219906730034472e-06,
+      "loss": 0.2103,
+      "step": 4943
+    },
+    {
+      "epoch": 2.341047647232909,
+      "grad_norm": 1.2246688604354858,
+      "learning_rate": 2.418654785302933e-06,
+      "loss": 0.202,
+      "step": 4944
+    },
+    {
+      "epoch": 2.34152116010654,
+      "grad_norm": 1.2264790534973145,
+      "learning_rate": 2.4153208804301232e-06,
+      "loss": 0.2202,
+      "step": 4945
+    },
+    {
+      "epoch": 2.3419946729801717,
+      "grad_norm": 0.9517511129379272,
+      "learning_rate": 2.4119889592569633e-06,
+      "loss": 0.1852,
+      "step": 4946
+    },
+    {
+      "epoch": 2.342468185853803,
+      "grad_norm": 0.9769893288612366,
+      "learning_rate": 2.408659022654892e-06,
+      "loss": 0.1959,
+      "step": 4947
+    },
+    {
+      "epoch": 2.3429416987274343,
+      "grad_norm": 1.0385048389434814,
+      "learning_rate": 2.4053310714948153e-06,
+      "loss": 0.1984,
+      "step": 4948
+    },
+    {
+      "epoch": 2.3434152116010654,
+      "grad_norm": 0.9590575098991394,
+      "learning_rate": 2.402005106647133e-06,
+      "loss": 0.1941,
+      "step": 4949
+    },
+    {
+      "epoch": 2.343888724474697,
+      "grad_norm": 1.5553905963897705,
+      "learning_rate": 2.3986811289817126e-06,
+      "loss": 0.2009,
+      "step": 4950
+    },
+    {
+      "epoch": 2.344362237348328,
+      "grad_norm": 1.0020135641098022,
+      "learning_rate": 2.3953591393679143e-06,
+      "loss": 0.1865,
+      "step": 4951
+    },
+    {
+      "epoch": 2.344835750221959,
+      "grad_norm": 1.1729081869125366,
+      "learning_rate": 2.39203913867457e-06,
+      "loss": 0.2014,
+      "step": 4952
+    },
+    {
+      "epoch": 2.3453092630955905,
+      "grad_norm": 1.1632317304611206,
+      "learning_rate": 2.3887211277699883e-06,
+      "loss": 0.2208,
+      "step": 4953
+    },
+    {
+      "epoch": 2.3457827759692216,
+      "grad_norm": 1.0477259159088135,
+      "learning_rate": 2.3854051075219744e-06,
+      "loss": 0.2035,
+      "step": 4954
+    },
+    {
+      "epoch": 2.346256288842853,
+      "grad_norm": 1.2778170108795166,
+      "learning_rate": 2.382091078797791e-06,
+      "loss": 0.2128,
+      "step": 4955
+    },
+    {
+      "epoch": 2.346729801716484,
+      "grad_norm": 1.322662115097046,
+      "learning_rate": 2.3787790424641986e-06,
+      "loss": 0.2053,
+      "step": 4956
+    },
+    {
+      "epoch": 2.347203314590115,
+      "grad_norm": 1.149414300918579,
+      "learning_rate": 2.3754689993874247e-06,
+      "loss": 0.2045,
+      "step": 4957
+    },
+    {
+      "epoch": 2.3476768274637467,
+      "grad_norm": 1.00741446018219,
+      "learning_rate": 2.3721609504331755e-06,
+      "loss": 0.1899,
+      "step": 4958
+    },
+    {
+      "epoch": 2.348150340337378,
+      "grad_norm": 1.745548963546753,
+      "learning_rate": 2.3688548964666446e-06,
+      "loss": 0.1886,
+      "step": 4959
+    },
+    {
+      "epoch": 2.3486238532110093,
+      "grad_norm": 1.0258804559707642,
+      "learning_rate": 2.365550838352495e-06,
+      "loss": 0.1878,
+      "step": 4960
+    },
+    {
+      "epoch": 2.3490973660846404,
+      "grad_norm": 1.286632776260376,
+      "learning_rate": 2.3622487769548754e-06,
+      "loss": 0.2032,
+      "step": 4961
+    },
+    {
+      "epoch": 2.3495708789582714,
+      "grad_norm": 1.0202562808990479,
+      "learning_rate": 2.3589487131374023e-06,
+      "loss": 0.2097,
+      "step": 4962
+    },
+    {
+      "epoch": 2.350044391831903,
+      "grad_norm": 1.0657576322555542,
+      "learning_rate": 2.3556506477631826e-06,
+      "loss": 0.1977,
+      "step": 4963
+    },
+    {
+      "epoch": 2.350517904705534,
+      "grad_norm": 1.0342962741851807,
+      "learning_rate": 2.3523545816947856e-06,
+      "loss": 0.1915,
+      "step": 4964
+    },
+    {
+      "epoch": 2.3509914175791655,
+      "grad_norm": 1.115942120552063,
+      "learning_rate": 2.3490605157942726e-06,
+      "loss": 0.2007,
+      "step": 4965
+    },
+    {
+      "epoch": 2.3514649304527966,
+      "grad_norm": 2.0979726314544678,
+      "learning_rate": 2.3457684509231725e-06,
+      "loss": 0.1993,
+      "step": 4966
+    },
+    {
+      "epoch": 2.351938443326428,
+      "grad_norm": 1.075711965560913,
+      "learning_rate": 2.3424783879424894e-06,
+      "loss": 0.1756,
+      "step": 4967
+    },
+    {
+      "epoch": 2.352411956200059,
+      "grad_norm": 1.1330883502960205,
+      "learning_rate": 2.3391903277127127e-06,
+      "loss": 0.2115,
+      "step": 4968
+    },
+    {
+      "epoch": 2.3528854690736907,
+      "grad_norm": 1.1201201677322388,
+      "learning_rate": 2.3359042710937986e-06,
+      "loss": 0.1995,
+      "step": 4969
+    },
+    {
+      "epoch": 2.3533589819473217,
+      "grad_norm": 1.2727965116500854,
+      "learning_rate": 2.3326202189451873e-06,
+      "loss": 0.2007,
+      "step": 4970
+    },
+    {
+      "epoch": 2.353832494820953,
+      "grad_norm": 1.3274377584457397,
+      "learning_rate": 2.3293381721257868e-06,
+      "loss": 0.1829,
+      "step": 4971
+    },
+    {
+      "epoch": 2.3543060076945843,
+      "grad_norm": 0.9202346205711365,
+      "learning_rate": 2.326058131493991e-06,
+      "loss": 0.1861,
+      "step": 4972
+    },
+    {
+      "epoch": 2.3547795205682154,
+      "grad_norm": 1.0923426151275635,
+      "learning_rate": 2.3227800979076552e-06,
+      "loss": 0.2093,
+      "step": 4973
+    },
+    {
+      "epoch": 2.355253033441847,
+      "grad_norm": 1.1253713369369507,
+      "learning_rate": 2.319504072224125e-06,
+      "loss": 0.2014,
+      "step": 4974
+    },
+    {
+      "epoch": 2.355726546315478,
+      "grad_norm": 1.5804685354232788,
+      "learning_rate": 2.316230055300208e-06,
+      "loss": 0.1789,
+      "step": 4975
+    },
+    {
+      "epoch": 2.356200059189109,
+      "grad_norm": 2.09173321723938,
+      "learning_rate": 2.312958047992192e-06,
+      "loss": 0.1968,
+      "step": 4976
+    },
+    {
+      "epoch": 2.3566735720627405,
+      "grad_norm": 1.405809760093689,
+      "learning_rate": 2.3096880511558427e-06,
+      "loss": 0.2299,
+      "step": 4977
+    },
+    {
+      "epoch": 2.3571470849363716,
+      "grad_norm": 1.0536633729934692,
+      "learning_rate": 2.306420065646392e-06,
+      "loss": 0.2048,
+      "step": 4978
+    },
+    {
+      "epoch": 2.357620597810003,
+      "grad_norm": 1.2741084098815918,
+      "learning_rate": 2.3031540923185536e-06,
+      "loss": 0.2172,
+      "step": 4979
+    },
+    {
+      "epoch": 2.358094110683634,
+      "grad_norm": 0.9234102964401245,
+      "learning_rate": 2.299890132026508e-06,
+      "loss": 0.2006,
+      "step": 4980
+    },
+    {
+      "epoch": 2.3585676235572652,
+      "grad_norm": 1.0131012201309204,
+      "learning_rate": 2.296628185623915e-06,
+      "loss": 0.1732,
+      "step": 4981
+    },
+    {
+      "epoch": 2.3590411364308967,
+      "grad_norm": 1.1941182613372803,
+      "learning_rate": 2.2933682539639026e-06,
+      "loss": 0.2364,
+      "step": 4982
+    },
+    {
+      "epoch": 2.359514649304528,
+      "grad_norm": 1.0108896493911743,
+      "learning_rate": 2.2901103378990785e-06,
+      "loss": 0.1966,
+      "step": 4983
+    },
+    {
+      "epoch": 2.3599881621781593,
+      "grad_norm": 0.8393535017967224,
+      "learning_rate": 2.286854438281515e-06,
+      "loss": 0.194,
+      "step": 4984
+    },
+    {
+      "epoch": 2.3604616750517904,
+      "grad_norm": 1.4110928773880005,
+      "learning_rate": 2.283600555962765e-06,
+      "loss": 0.1968,
+      "step": 4985
+    },
+    {
+      "epoch": 2.360935187925422,
+      "grad_norm": 1.11850905418396,
+      "learning_rate": 2.2803486917938487e-06,
+      "loss": 0.2094,
+      "step": 4986
+    },
+    {
+      "epoch": 2.361408700799053,
+      "grad_norm": 1.1148712635040283,
+      "learning_rate": 2.2770988466252565e-06,
+      "loss": 0.2258,
+      "step": 4987
+    },
+    {
+      "epoch": 2.3618822136726845,
+      "grad_norm": 1.2502351999282837,
+      "learning_rate": 2.273851021306959e-06,
+      "loss": 0.1848,
+      "step": 4988
+    },
+    {
+      "epoch": 2.3623557265463155,
+      "grad_norm": 1.252803087234497,
+      "learning_rate": 2.27060521668839e-06,
+      "loss": 0.2035,
+      "step": 4989
+    },
+    {
+      "epoch": 2.3628292394199466,
+      "grad_norm": 1.0564323663711548,
+      "learning_rate": 2.26736143361846e-06,
+      "loss": 0.1949,
+      "step": 4990
+    },
+    {
+      "epoch": 2.363302752293578,
+      "grad_norm": 1.3723267316818237,
+      "learning_rate": 2.2641196729455482e-06,
+      "loss": 0.211,
+      "step": 4991
+    },
+    {
+      "epoch": 2.363776265167209,
+      "grad_norm": 1.104577660560608,
+      "learning_rate": 2.2608799355175058e-06,
+      "loss": 0.1932,
+      "step": 4992
+    },
+    {
+      "epoch": 2.3642497780408407,
+      "grad_norm": 1.5187643766403198,
+      "learning_rate": 2.2576422221816596e-06,
+      "loss": 0.1772,
+      "step": 4993
+    },
+    {
+      "epoch": 2.3647232909144718,
+      "grad_norm": 1.233651041984558,
+      "learning_rate": 2.254406533784794e-06,
+      "loss": 0.2034,
+      "step": 4994
+    },
+    {
+      "epoch": 2.365196803788103,
+      "grad_norm": 1.5177334547042847,
+      "learning_rate": 2.2511728711731806e-06,
+      "loss": 0.2064,
+      "step": 4995
+    },
+    {
+      "epoch": 2.3656703166617343,
+      "grad_norm": 1.8442257642745972,
+      "learning_rate": 2.2479412351925444e-06,
+      "loss": 0.2119,
+      "step": 4996
+    },
+    {
+      "epoch": 2.3661438295353654,
+      "grad_norm": 1.1477844715118408,
+      "learning_rate": 2.2447116266880964e-06,
+      "loss": 0.1973,
+      "step": 4997
+    },
+    {
+      "epoch": 2.366617342408997,
+      "grad_norm": 1.1657443046569824,
+      "learning_rate": 2.241484046504503e-06,
+      "loss": 0.2081,
+      "step": 4998
+    },
+    {
+      "epoch": 2.367090855282628,
+      "grad_norm": 1.501852035522461,
+      "learning_rate": 2.238258495485912e-06,
+      "loss": 0.2391,
+      "step": 4999
+    },
+    {
+      "epoch": 2.367564368156259,
+      "grad_norm": 1.438940167427063,
+      "learning_rate": 2.2350349744759324e-06,
+      "loss": 0.2117,
+      "step": 5000
+    },
+    {
+      "epoch": 2.3680378810298905,
+      "grad_norm": 2.076133966445923,
+      "learning_rate": 2.231813484317643e-06,
+      "loss": 0.1915,
+      "step": 5001
+    },
+    {
+      "epoch": 2.3685113939035216,
+      "grad_norm": 0.9920544624328613,
+      "learning_rate": 2.2285940258535987e-06,
+      "loss": 0.2254,
+      "step": 5002
+    },
+    {
+      "epoch": 2.368984906777153,
+      "grad_norm": 0.9885861873626709,
+      "learning_rate": 2.2253765999258115e-06,
+      "loss": 0.2179,
+      "step": 5003
+    },
+    {
+      "epoch": 2.369458419650784,
+      "grad_norm": 1.199271559715271,
+      "learning_rate": 2.222161207375775e-06,
+      "loss": 0.1974,
+      "step": 5004
+    },
+    {
+      "epoch": 2.3699319325244157,
+      "grad_norm": 0.865847110748291,
+      "learning_rate": 2.2189478490444373e-06,
+      "loss": 0.1935,
+      "step": 5005
+    },
+    {
+      "epoch": 2.3704054453980468,
+      "grad_norm": 1.360724925994873,
+      "learning_rate": 2.2157365257722273e-06,
+      "loss": 0.1836,
+      "step": 5006
+    },
+    {
+      "epoch": 2.370878958271678,
+      "grad_norm": 1.0131431818008423,
+      "learning_rate": 2.2125272383990304e-06,
+      "loss": 0.1864,
+      "step": 5007
+    },
+    {
+      "epoch": 2.3713524711453093,
+      "grad_norm": 0.978959858417511,
+      "learning_rate": 2.209319987764209e-06,
+      "loss": 0.2075,
+      "step": 5008
+    },
+    {
+      "epoch": 2.3718259840189404,
+      "grad_norm": 1.2233644723892212,
+      "learning_rate": 2.2061147747065847e-06,
+      "loss": 0.2237,
+      "step": 5009
+    },
+    {
+      "epoch": 2.372299496892572,
+      "grad_norm": 1.8459240198135376,
+      "learning_rate": 2.2029116000644544e-06,
+      "loss": 0.1897,
+      "step": 5010
+    },
+    {
+      "epoch": 2.372773009766203,
+      "grad_norm": 1.2090134620666504,
+      "learning_rate": 2.1997104646755763e-06,
+      "loss": 0.2038,
+      "step": 5011
+    },
+    {
+      "epoch": 2.3732465226398345,
+      "grad_norm": 0.9758379459381104,
+      "learning_rate": 2.1965113693771725e-06,
+      "loss": 0.1924,
+      "step": 5012
+    },
+    {
+      "epoch": 2.3737200355134656,
+      "grad_norm": 1.6864513158798218,
+      "learning_rate": 2.193314315005941e-06,
+      "loss": 0.2004,
+      "step": 5013
+    },
+    {
+      "epoch": 2.3741935483870966,
+      "grad_norm": 1.3247301578521729,
+      "learning_rate": 2.190119302398035e-06,
+      "loss": 0.1929,
+      "step": 5014
+    },
+    {
+      "epoch": 2.374667061260728,
+      "grad_norm": 1.4092978239059448,
+      "learning_rate": 2.186926332389084e-06,
+      "loss": 0.2078,
+      "step": 5015
+    },
+    {
+      "epoch": 2.375140574134359,
+      "grad_norm": 0.9686800241470337,
+      "learning_rate": 2.1837354058141756e-06,
+      "loss": 0.1993,
+      "step": 5016
+    },
+    {
+      "epoch": 2.3756140870079907,
+      "grad_norm": 1.527159571647644,
+      "learning_rate": 2.1805465235078695e-06,
+      "loss": 0.191,
+      "step": 5017
+    },
+    {
+      "epoch": 2.3760875998816218,
+      "grad_norm": 1.4575700759887695,
+      "learning_rate": 2.177359686304181e-06,
+      "loss": 0.203,
+      "step": 5018
+    },
+    {
+      "epoch": 2.376561112755253,
+      "grad_norm": 1.0823571681976318,
+      "learning_rate": 2.1741748950366036e-06,
+      "loss": 0.2032,
+      "step": 5019
+    },
+    {
+      "epoch": 2.3770346256288843,
+      "grad_norm": 1.1299647092819214,
+      "learning_rate": 2.170992150538085e-06,
+      "loss": 0.1946,
+      "step": 5020
+    },
+    {
+      "epoch": 2.3775081385025154,
+      "grad_norm": 1.379338264465332,
+      "learning_rate": 2.167811453641039e-06,
+      "loss": 0.2056,
+      "step": 5021
+    },
+    {
+      "epoch": 2.377981651376147,
+      "grad_norm": 1.223885416984558,
+      "learning_rate": 2.1646328051773512e-06,
+      "loss": 0.2183,
+      "step": 5022
+    },
+    {
+      "epoch": 2.378455164249778,
+      "grad_norm": 1.0248559713363647,
+      "learning_rate": 2.1614562059783627e-06,
+      "loss": 0.1936,
+      "step": 5023
+    },
+    {
+      "epoch": 2.378928677123409,
+      "grad_norm": 1.0929266214370728,
+      "learning_rate": 2.1582816568748856e-06,
+      "loss": 0.1954,
+      "step": 5024
+    },
+    {
+      "epoch": 2.3794021899970406,
+      "grad_norm": 1.3425700664520264,
+      "learning_rate": 2.155109158697187e-06,
+      "loss": 0.1968,
+      "step": 5025
+    },
+    {
+      "epoch": 2.3798757028706716,
+      "grad_norm": 1.4846118688583374,
+      "learning_rate": 2.151938712275011e-06,
+      "loss": 0.1819,
+      "step": 5026
+    },
+    {
+      "epoch": 2.380349215744303,
+      "grad_norm": 1.0532817840576172,
+      "learning_rate": 2.14877031843755e-06,
+      "loss": 0.1961,
+      "step": 5027
+    },
+    {
+      "epoch": 2.380822728617934,
+      "grad_norm": 1.0059374570846558,
+      "learning_rate": 2.145603978013473e-06,
+      "loss": 0.2051,
+      "step": 5028
+    },
+    {
+      "epoch": 2.3812962414915657,
+      "grad_norm": 1.0236760377883911,
+      "learning_rate": 2.1424396918309e-06,
+      "loss": 0.2084,
+      "step": 5029
+    },
+    {
+      "epoch": 2.381769754365197,
+      "grad_norm": 1.142202377319336,
+      "learning_rate": 2.139277460717425e-06,
+      "loss": 0.1947,
+      "step": 5030
+    },
+    {
+      "epoch": 2.3822432672388283,
+      "grad_norm": 1.1363552808761597,
+      "learning_rate": 2.1361172855000943e-06,
+      "loss": 0.2102,
+      "step": 5031
+    },
+    {
+      "epoch": 2.3827167801124594,
+      "grad_norm": 1.0805346965789795,
+      "learning_rate": 2.1329591670054227e-06,
+      "loss": 0.1846,
+      "step": 5032
+    },
+    {
+      "epoch": 2.3831902929860904,
+      "grad_norm": 1.1634663343429565,
+      "learning_rate": 2.1298031060593893e-06,
+      "loss": 0.2285,
+      "step": 5033
+    },
+    {
+      "epoch": 2.383663805859722,
+      "grad_norm": 0.9185308814048767,
+      "learning_rate": 2.1266491034874247e-06,
+      "loss": 0.1965,
+      "step": 5034
+    },
+    {
+      "epoch": 2.384137318733353,
+      "grad_norm": 0.9834035038948059,
+      "learning_rate": 2.1234971601144362e-06,
+      "loss": 0.2108,
+      "step": 5035
+    },
+    {
+      "epoch": 2.3846108316069845,
+      "grad_norm": 1.1302801370620728,
+      "learning_rate": 2.1203472767647782e-06,
+      "loss": 0.1893,
+      "step": 5036
+    },
+    {
+      "epoch": 2.3850843444806156,
+      "grad_norm": 1.2327231168746948,
+      "learning_rate": 2.117199454262271e-06,
+      "loss": 0.1993,
+      "step": 5037
+    },
+    {
+      "epoch": 2.3855578573542466,
+      "grad_norm": 0.890938401222229,
+      "learning_rate": 2.1140536934302024e-06,
+      "loss": 0.1921,
+      "step": 5038
+    },
+    {
+      "epoch": 2.386031370227878,
+      "grad_norm": 0.9080513715744019,
+      "learning_rate": 2.1109099950913105e-06,
+      "loss": 0.1937,
+      "step": 5039
+    },
+    {
+      "epoch": 2.386504883101509,
+      "grad_norm": 0.9706762433052063,
+      "learning_rate": 2.107768360067806e-06,
+      "loss": 0.2035,
+      "step": 5040
+    },
+    {
+      "epoch": 2.3869783959751407,
+      "grad_norm": 1.112209439277649,
+      "learning_rate": 2.1046287891813445e-06,
+      "loss": 0.2072,
+      "step": 5041
+    },
+    {
+      "epoch": 2.387451908848772,
+      "grad_norm": 1.2673587799072266,
+      "learning_rate": 2.1014912832530587e-06,
+      "loss": 0.1955,
+      "step": 5042
+    },
+    {
+      "epoch": 2.387925421722403,
+      "grad_norm": 1.109052300453186,
+      "learning_rate": 2.0983558431035266e-06,
+      "loss": 0.1937,
+      "step": 5043
+    },
+    {
+      "epoch": 2.3883989345960344,
+      "grad_norm": 0.9777413606643677,
+      "learning_rate": 2.095222469552799e-06,
+      "loss": 0.2019,
+      "step": 5044
+    },
+    {
+      "epoch": 2.3888724474696654,
+      "grad_norm": 1.3833292722702026,
+      "learning_rate": 2.0920911634203745e-06,
+      "loss": 0.1948,
+      "step": 5045
+    },
+    {
+      "epoch": 2.389345960343297,
+      "grad_norm": 1.2488656044006348,
+      "learning_rate": 2.0889619255252147e-06,
+      "loss": 0.1998,
+      "step": 5046
+    },
+    {
+      "epoch": 2.389819473216928,
+      "grad_norm": 1.1769556999206543,
+      "learning_rate": 2.085834756685747e-06,
+      "loss": 0.2172,
+      "step": 5047
+    },
+    {
+      "epoch": 2.3902929860905595,
+      "grad_norm": 1.6330255270004272,
+      "learning_rate": 2.082709657719848e-06,
+      "loss": 0.1885,
+      "step": 5048
+    },
+    {
+      "epoch": 2.3907664989641906,
+      "grad_norm": 1.0924394130706787,
+      "learning_rate": 2.0795866294448596e-06,
+      "loss": 0.2175,
+      "step": 5049
+    },
+    {
+      "epoch": 2.391240011837822,
+      "grad_norm": 1.4790942668914795,
+      "learning_rate": 2.0764656726775767e-06,
+      "loss": 0.1987,
+      "step": 5050
+    },
+    {
+      "epoch": 2.391713524711453,
+      "grad_norm": 0.9769918918609619,
+      "learning_rate": 2.0733467882342597e-06,
+      "loss": 0.2065,
+      "step": 5051
+    },
+    {
+      "epoch": 2.392187037585084,
+      "grad_norm": 1.1905534267425537,
+      "learning_rate": 2.0702299769306177e-06,
+      "loss": 0.2126,
+      "step": 5052
+    },
+    {
+      "epoch": 2.3926605504587157,
+      "grad_norm": 1.0014641284942627,
+      "learning_rate": 2.067115239581828e-06,
+      "loss": 0.1905,
+      "step": 5053
+    },
+    {
+      "epoch": 2.393134063332347,
+      "grad_norm": 1.234556794166565,
+      "learning_rate": 2.0640025770025186e-06,
+      "loss": 0.2117,
+      "step": 5054
+    },
+    {
+      "epoch": 2.3936075762059783,
+      "grad_norm": 1.1088292598724365,
+      "learning_rate": 2.0608919900067716e-06,
+      "loss": 0.2065,
+      "step": 5055
+    },
+    {
+      "epoch": 2.3940810890796094,
+      "grad_norm": 0.9730679988861084,
+      "learning_rate": 2.0577834794081377e-06,
+      "loss": 0.1976,
+      "step": 5056
+    },
+    {
+      "epoch": 2.3945546019532404,
+      "grad_norm": 1.591340184211731,
+      "learning_rate": 2.0546770460196117e-06,
+      "loss": 0.1804,
+      "step": 5057
+    },
+    {
+      "epoch": 2.395028114826872,
+      "grad_norm": 1.2157362699508667,
+      "learning_rate": 2.051572690653658e-06,
+      "loss": 0.1855,
+      "step": 5058
+    },
+    {
+      "epoch": 2.395501627700503,
+      "grad_norm": 1.2474086284637451,
+      "learning_rate": 2.0484704141221845e-06,
+      "loss": 0.1692,
+      "step": 5059
+    },
+    {
+      "epoch": 2.3959751405741345,
+      "grad_norm": 0.999910831451416,
+      "learning_rate": 2.0453702172365663e-06,
+      "loss": 0.1992,
+      "step": 5060
+    },
+    {
+      "epoch": 2.3964486534477656,
+      "grad_norm": 1.1098980903625488,
+      "learning_rate": 2.0422721008076264e-06,
+      "loss": 0.2495,
+      "step": 5061
+    },
+    {
+      "epoch": 2.3969221663213967,
+      "grad_norm": 1.1341627836227417,
+      "learning_rate": 2.0391760656456506e-06,
+      "loss": 0.1914,
+      "step": 5062
+    },
+    {
+      "epoch": 2.397395679195028,
+      "grad_norm": 0.9773929119110107,
+      "learning_rate": 2.0360821125603726e-06,
+      "loss": 0.204,
+      "step": 5063
+    },
+    {
+      "epoch": 2.3978691920686592,
+      "grad_norm": 1.09663724899292,
+      "learning_rate": 2.0329902423609926e-06,
+      "loss": 0.2104,
+      "step": 5064
+    },
+    {
+      "epoch": 2.3983427049422907,
+      "grad_norm": 1.1711968183517456,
+      "learning_rate": 2.0299004558561554e-06,
+      "loss": 0.2043,
+      "step": 5065
+    },
+    {
+      "epoch": 2.398816217815922,
+      "grad_norm": 1.7636314630508423,
+      "learning_rate": 2.026812753853962e-06,
+      "loss": 0.1984,
+      "step": 5066
+    },
+    {
+      "epoch": 2.3992897306895533,
+      "grad_norm": 0.9198000431060791,
+      "learning_rate": 2.023727137161976e-06,
+      "loss": 0.1887,
+      "step": 5067
+    },
+    {
+      "epoch": 2.3997632435631844,
+      "grad_norm": 1.0493696928024292,
+      "learning_rate": 2.020643606587207e-06,
+      "loss": 0.1972,
+      "step": 5068
+    },
+    {
+      "epoch": 2.4002367564368154,
+      "grad_norm": 1.075859785079956,
+      "learning_rate": 2.0175621629361274e-06,
+      "loss": 0.1963,
+      "step": 5069
+    },
+    {
+      "epoch": 2.400710269310447,
+      "grad_norm": 1.5328881740570068,
+      "learning_rate": 2.0144828070146528e-06,
+      "loss": 0.1893,
+      "step": 5070
+    },
+    {
+      "epoch": 2.401183782184078,
+      "grad_norm": 1.1692156791687012,
+      "learning_rate": 2.011405539628163e-06,
+      "loss": 0.181,
+      "step": 5071
+    },
+    {
+      "epoch": 2.4016572950577095,
+      "grad_norm": 2.3541417121887207,
+      "learning_rate": 2.0083303615814896e-06,
+      "loss": 0.2215,
+      "step": 5072
+    },
+    {
+      "epoch": 2.4021308079313406,
+      "grad_norm": 1.5428338050842285,
+      "learning_rate": 2.00525727367891e-06,
+      "loss": 0.1991,
+      "step": 5073
+    },
+    {
+      "epoch": 2.402604320804972,
+      "grad_norm": 0.920405924320221,
+      "learning_rate": 2.002186276724166e-06,
+      "loss": 0.1992,
+      "step": 5074
+    },
+    {
+      "epoch": 2.403077833678603,
+      "grad_norm": 1.1372102499008179,
+      "learning_rate": 1.999117371520444e-06,
+      "loss": 0.2079,
+      "step": 5075
+    },
+    {
+      "epoch": 2.4035513465522342,
+      "grad_norm": 1.1284953355789185,
+      "learning_rate": 1.9960505588703893e-06,
+      "loss": 0.2023,
+      "step": 5076
+    },
+    {
+      "epoch": 2.4040248594258657,
+      "grad_norm": 1.1499290466308594,
+      "learning_rate": 1.9929858395760927e-06,
+      "loss": 0.1872,
+      "step": 5077
+    },
+    {
+      "epoch": 2.404498372299497,
+      "grad_norm": 1.525108814239502,
+      "learning_rate": 1.9899232144391077e-06,
+      "loss": 0.2028,
+      "step": 5078
+    },
+    {
+      "epoch": 2.4049718851731283,
+      "grad_norm": 1.109409213066101,
+      "learning_rate": 1.986862684260431e-06,
+      "loss": 0.2014,
+      "step": 5079
+    },
+    {
+      "epoch": 2.4054453980467594,
+      "grad_norm": 1.8481568098068237,
+      "learning_rate": 1.983804249840513e-06,
+      "loss": 0.1663,
+      "step": 5080
+    },
+    {
+      "epoch": 2.4059189109203905,
+      "grad_norm": 0.8443739414215088,
+      "learning_rate": 1.9807479119792618e-06,
+      "loss": 0.1905,
+      "step": 5081
+    },
+    {
+      "epoch": 2.406392423794022,
+      "grad_norm": 1.2149391174316406,
+      "learning_rate": 1.9776936714760297e-06,
+      "loss": 0.1994,
+      "step": 5082
+    },
+    {
+      "epoch": 2.406865936667653,
+      "grad_norm": 1.261169672012329,
+      "learning_rate": 1.974641529129626e-06,
+      "loss": 0.2054,
+      "step": 5083
+    },
+    {
+      "epoch": 2.4073394495412845,
+      "grad_norm": 1.2343029975891113,
+      "learning_rate": 1.971591485738308e-06,
+      "loss": 0.1861,
+      "step": 5084
+    },
+    {
+      "epoch": 2.4078129624149156,
+      "grad_norm": 1.0303642749786377,
+      "learning_rate": 1.968543542099787e-06,
+      "loss": 0.2034,
+      "step": 5085
+    },
+    {
+      "epoch": 2.4082864752885467,
+      "grad_norm": 1.204028606414795,
+      "learning_rate": 1.9654976990112184e-06,
+      "loss": 0.1955,
+      "step": 5086
+    },
+    {
+      "epoch": 2.408759988162178,
+      "grad_norm": 1.6389491558074951,
+      "learning_rate": 1.9624539572692193e-06,
+      "loss": 0.2175,
+      "step": 5087
+    },
+    {
+      "epoch": 2.4092335010358092,
+      "grad_norm": 1.201258897781372,
+      "learning_rate": 1.9594123176698467e-06,
+      "loss": 0.1972,
+      "step": 5088
+    },
+    {
+      "epoch": 2.4097070139094408,
+      "grad_norm": 1.1844000816345215,
+      "learning_rate": 1.9563727810086155e-06,
+      "loss": 0.193,
+      "step": 5089
+    },
+    {
+      "epoch": 2.410180526783072,
+      "grad_norm": 1.0625587701797485,
+      "learning_rate": 1.953335348080484e-06,
+      "loss": 0.2036,
+      "step": 5090
+    },
+    {
+      "epoch": 2.4106540396567033,
+      "grad_norm": 1.2302863597869873,
+      "learning_rate": 1.950300019679863e-06,
+      "loss": 0.2189,
+      "step": 5091
+    },
+    {
+      "epoch": 2.4111275525303344,
+      "grad_norm": 0.8708613514900208,
+      "learning_rate": 1.9472667966006177e-06,
+      "loss": 0.1779,
+      "step": 5092
+    },
+    {
+      "epoch": 2.411601065403966,
+      "grad_norm": 1.382256031036377,
+      "learning_rate": 1.944235679636053e-06,
+      "loss": 0.1904,
+      "step": 5093
+    },
+    {
+      "epoch": 2.412074578277597,
+      "grad_norm": 1.4713544845581055,
+      "learning_rate": 1.941206669578933e-06,
+      "loss": 0.218,
+      "step": 5094
+    },
+    {
+      "epoch": 2.412548091151228,
+      "grad_norm": 1.513837456703186,
+      "learning_rate": 1.9381797672214618e-06,
+      "loss": 0.201,
+      "step": 5095
+    },
+    {
+      "epoch": 2.4130216040248595,
+      "grad_norm": 1.2420634031295776,
+      "learning_rate": 1.935154973355302e-06,
+      "loss": 0.225,
+      "step": 5096
+    },
+    {
+      "epoch": 2.4134951168984906,
+      "grad_norm": 1.2236741781234741,
+      "learning_rate": 1.9321322887715533e-06,
+      "loss": 0.1928,
+      "step": 5097
+    },
+    {
+      "epoch": 2.413968629772122,
+      "grad_norm": 1.1855653524398804,
+      "learning_rate": 1.929111714260774e-06,
+      "loss": 0.2099,
+      "step": 5098
+    },
+    {
+      "epoch": 2.414442142645753,
+      "grad_norm": 1.738868236541748,
+      "learning_rate": 1.9260932506129647e-06,
+      "loss": 0.1922,
+      "step": 5099
+    },
+    {
+      "epoch": 2.4149156555193843,
+      "grad_norm": 1.1198941469192505,
+      "learning_rate": 1.923076898617574e-06,
+      "loss": 0.2124,
+      "step": 5100
+    },
+    {
+      "epoch": 2.4153891683930158,
+      "grad_norm": 0.9823029041290283,
+      "learning_rate": 1.920062659063503e-06,
+      "loss": 0.2041,
+      "step": 5101
+    },
+    {
+      "epoch": 2.415862681266647,
+      "grad_norm": 1.4813225269317627,
+      "learning_rate": 1.917050532739092e-06,
+      "loss": 0.2001,
+      "step": 5102
+    },
+    {
+      "epoch": 2.4163361941402783,
+      "grad_norm": 1.1506394147872925,
+      "learning_rate": 1.9140405204321387e-06,
+      "loss": 0.1975,
+      "step": 5103
+    },
+    {
+      "epoch": 2.4168097070139094,
+      "grad_norm": 1.1538875102996826,
+      "learning_rate": 1.911032622929879e-06,
+      "loss": 0.1927,
+      "step": 5104
+    },
+    {
+      "epoch": 2.4172832198875405,
+      "grad_norm": 1.162630558013916,
+      "learning_rate": 1.908026841019003e-06,
+      "loss": 0.2028,
+      "step": 5105
+    },
+    {
+      "epoch": 2.417756732761172,
+      "grad_norm": 1.90775465965271,
+      "learning_rate": 1.9050231754856407e-06,
+      "loss": 0.2086,
+      "step": 5106
+    },
+    {
+      "epoch": 2.418230245634803,
+      "grad_norm": 1.8542652130126953,
+      "learning_rate": 1.9020216271153747e-06,
+      "loss": 0.2132,
+      "step": 5107
+    },
+    {
+      "epoch": 2.4187037585084346,
+      "grad_norm": 1.7880303859710693,
+      "learning_rate": 1.8990221966932266e-06,
+      "loss": 0.207,
+      "step": 5108
+    },
+    {
+      "epoch": 2.4191772713820656,
+      "grad_norm": 1.4888118505477905,
+      "learning_rate": 1.8960248850036722e-06,
+      "loss": 0.234,
+      "step": 5109
+    },
+    {
+      "epoch": 2.419650784255697,
+      "grad_norm": 1.0973647832870483,
+      "learning_rate": 1.8930296928306313e-06,
+      "loss": 0.1932,
+      "step": 5110
+    },
+    {
+      "epoch": 2.420124297129328,
+      "grad_norm": 1.4137020111083984,
+      "learning_rate": 1.8900366209574627e-06,
+      "loss": 0.199,
+      "step": 5111
+    },
+    {
+      "epoch": 2.4205978100029597,
+      "grad_norm": 1.0636194944381714,
+      "learning_rate": 1.8870456701669792e-06,
+      "loss": 0.1843,
+      "step": 5112
+    },
+    {
+      "epoch": 2.4210713228765908,
+      "grad_norm": 0.9531354308128357,
+      "learning_rate": 1.8840568412414318e-06,
+      "loss": 0.2034,
+      "step": 5113
+    },
+    {
+      "epoch": 2.421544835750222,
+      "grad_norm": 1.3613364696502686,
+      "learning_rate": 1.8810701349625237e-06,
+      "loss": 0.1957,
+      "step": 5114
+    },
+    {
+      "epoch": 2.4220183486238533,
+      "grad_norm": 1.619372844696045,
+      "learning_rate": 1.8780855521113983e-06,
+      "loss": 0.1774,
+      "step": 5115
+    },
+    {
+      "epoch": 2.4224918614974844,
+      "grad_norm": 1.429699182510376,
+      "learning_rate": 1.8751030934686398e-06,
+      "loss": 0.2106,
+      "step": 5116
+    },
+    {
+      "epoch": 2.422965374371116,
+      "grad_norm": 1.0113288164138794,
+      "learning_rate": 1.8721227598142876e-06,
+      "loss": 0.2062,
+      "step": 5117
+    },
+    {
+      "epoch": 2.423438887244747,
+      "grad_norm": 1.9973174333572388,
+      "learning_rate": 1.869144551927814e-06,
+      "loss": 0.2193,
+      "step": 5118
+    },
+    {
+      "epoch": 2.423912400118378,
+      "grad_norm": 1.051254153251648,
+      "learning_rate": 1.8661684705881456e-06,
+      "loss": 0.198,
+      "step": 5119
+    },
+    {
+      "epoch": 2.4243859129920096,
+      "grad_norm": 1.627144455909729,
+      "learning_rate": 1.863194516573642e-06,
+      "loss": 0.1883,
+      "step": 5120
+    },
+    {
+      "epoch": 2.4248594258656406,
+      "grad_norm": 1.1323554515838623,
+      "learning_rate": 1.860222690662119e-06,
+      "loss": 0.2051,
+      "step": 5121
+    },
+    {
+      "epoch": 2.425332938739272,
+      "grad_norm": 1.0321543216705322,
+      "learning_rate": 1.8572529936308225e-06,
+      "loss": 0.2016,
+      "step": 5122
+    },
+    {
+      "epoch": 2.425806451612903,
+      "grad_norm": 1.293243646621704,
+      "learning_rate": 1.8542854262564537e-06,
+      "loss": 0.2199,
+      "step": 5123
+    },
+    {
+      "epoch": 2.4262799644865343,
+      "grad_norm": 1.043289065361023,
+      "learning_rate": 1.851319989315149e-06,
+      "loss": 0.1967,
+      "step": 5124
+    },
+    {
+      "epoch": 2.426753477360166,
+      "grad_norm": 1.376531720161438,
+      "learning_rate": 1.8483566835824862e-06,
+      "loss": 0.2124,
+      "step": 5125
+    },
+    {
+      "epoch": 2.427226990233797,
+      "grad_norm": 1.3981205224990845,
+      "learning_rate": 1.8453955098334953e-06,
+      "loss": 0.2134,
+      "step": 5126
+    },
+    {
+      "epoch": 2.4277005031074284,
+      "grad_norm": 1.1886367797851562,
+      "learning_rate": 1.8424364688426365e-06,
+      "loss": 0.1731,
+      "step": 5127
+    },
+    {
+      "epoch": 2.4281740159810594,
+      "grad_norm": 0.9211165308952332,
+      "learning_rate": 1.8394795613838256e-06,
+      "loss": 0.1854,
+      "step": 5128
+    },
+    {
+      "epoch": 2.428647528854691,
+      "grad_norm": 1.1072946786880493,
+      "learning_rate": 1.8365247882304061e-06,
+      "loss": 0.1827,
+      "step": 5129
+    },
+    {
+      "epoch": 2.429121041728322,
+      "grad_norm": 0.8516340851783752,
+      "learning_rate": 1.8335721501551774e-06,
+      "loss": 0.2105,
+      "step": 5130
+    },
+    {
+      "epoch": 2.429594554601953,
+      "grad_norm": 1.1309757232666016,
+      "learning_rate": 1.8306216479303663e-06,
+      "loss": 0.195,
+      "step": 5131
+    },
+    {
+      "epoch": 2.4300680674755846,
+      "grad_norm": 1.1593215465545654,
+      "learning_rate": 1.8276732823276556e-06,
+      "loss": 0.2171,
+      "step": 5132
+    },
+    {
+      "epoch": 2.4305415803492156,
+      "grad_norm": 0.949464738368988,
+      "learning_rate": 1.8247270541181572e-06,
+      "loss": 0.2146,
+      "step": 5133
+    },
+    {
+      "epoch": 2.431015093222847,
+      "grad_norm": 1.4666881561279297,
+      "learning_rate": 1.8217829640724271e-06,
+      "loss": 0.2085,
+      "step": 5134
+    },
+    {
+      "epoch": 2.431488606096478,
+      "grad_norm": 1.130859136581421,
+      "learning_rate": 1.8188410129604684e-06,
+      "loss": 0.2297,
+      "step": 5135
+    },
+    {
+      "epoch": 2.4319621189701097,
+      "grad_norm": 1.016351342201233,
+      "learning_rate": 1.8159012015517152e-06,
+      "loss": 0.1865,
+      "step": 5136
+    },
+    {
+      "epoch": 2.432435631843741,
+      "grad_norm": 1.0443476438522339,
+      "learning_rate": 1.8129635306150517e-06,
+      "loss": 0.2097,
+      "step": 5137
+    },
+    {
+      "epoch": 2.432909144717372,
+      "grad_norm": 1.0505341291427612,
+      "learning_rate": 1.8100280009187931e-06,
+      "loss": 0.1951,
+      "step": 5138
+    },
+    {
+      "epoch": 2.4333826575910034,
+      "grad_norm": 1.0258119106292725,
+      "learning_rate": 1.8070946132307033e-06,
+      "loss": 0.2119,
+      "step": 5139
+    },
+    {
+      "epoch": 2.4338561704646344,
+      "grad_norm": 1.3897591829299927,
+      "learning_rate": 1.804163368317976e-06,
+      "loss": 0.2081,
+      "step": 5140
+    },
+    {
+      "epoch": 2.434329683338266,
+      "grad_norm": 1.1299906969070435,
+      "learning_rate": 1.801234266947256e-06,
+      "loss": 0.187,
+      "step": 5141
+    },
+    {
+      "epoch": 2.434803196211897,
+      "grad_norm": 1.4262118339538574,
+      "learning_rate": 1.798307309884616e-06,
+      "loss": 0.2022,
+      "step": 5142
+    },
+    {
+      "epoch": 2.435276709085528,
+      "grad_norm": 1.6670175790786743,
+      "learning_rate": 1.795382497895578e-06,
+      "loss": 0.2016,
+      "step": 5143
+    },
+    {
+      "epoch": 2.4357502219591596,
+      "grad_norm": 1.0870014429092407,
+      "learning_rate": 1.792459831745097e-06,
+      "loss": 0.205,
+      "step": 5144
+    },
+    {
+      "epoch": 2.4362237348327906,
+      "grad_norm": 1.5366194248199463,
+      "learning_rate": 1.7895393121975646e-06,
+      "loss": 0.1978,
+      "step": 5145
+    },
+    {
+      "epoch": 2.436697247706422,
+      "grad_norm": 1.1896754503250122,
+      "learning_rate": 1.7866209400168211e-06,
+      "loss": 0.2016,
+      "step": 5146
+    },
+    {
+      "epoch": 2.437170760580053,
+      "grad_norm": 1.7006498575210571,
+      "learning_rate": 1.7837047159661302e-06,
+      "loss": 0.1941,
+      "step": 5147
+    },
+    {
+      "epoch": 2.4376442734536843,
+      "grad_norm": 1.0235669612884521,
+      "learning_rate": 1.7807906408082087e-06,
+      "loss": 0.2014,
+      "step": 5148
+    },
+    {
+      "epoch": 2.438117786327316,
+      "grad_norm": 1.008035659790039,
+      "learning_rate": 1.7778787153052045e-06,
+      "loss": 0.2089,
+      "step": 5149
+    },
+    {
+      "epoch": 2.438591299200947,
+      "grad_norm": 1.0255095958709717,
+      "learning_rate": 1.7749689402186998e-06,
+      "loss": 0.1944,
+      "step": 5150
+    },
+    {
+      "epoch": 2.4390648120745784,
+      "grad_norm": 1.1309661865234375,
+      "learning_rate": 1.7720613163097233e-06,
+      "loss": 0.1999,
+      "step": 5151
+    },
+    {
+      "epoch": 2.4395383249482094,
+      "grad_norm": 1.1396714448928833,
+      "learning_rate": 1.7691558443387302e-06,
+      "loss": 0.1979,
+      "step": 5152
+    },
+    {
+      "epoch": 2.440011837821841,
+      "grad_norm": 1.2275151014328003,
+      "learning_rate": 1.766252525065625e-06,
+      "loss": 0.1792,
+      "step": 5153
+    },
+    {
+      "epoch": 2.440485350695472,
+      "grad_norm": 1.1925066709518433,
+      "learning_rate": 1.7633513592497354e-06,
+      "loss": 0.2024,
+      "step": 5154
+    },
+    {
+      "epoch": 2.4409588635691035,
+      "grad_norm": 0.9196386933326721,
+      "learning_rate": 1.7604523476498413e-06,
+      "loss": 0.1739,
+      "step": 5155
+    },
+    {
+      "epoch": 2.4414323764427346,
+      "grad_norm": 1.4410473108291626,
+      "learning_rate": 1.7575554910241444e-06,
+      "loss": 0.1839,
+      "step": 5156
+    },
+    {
+      "epoch": 2.4419058893163657,
+      "grad_norm": 1.2430617809295654,
+      "learning_rate": 1.7546607901302948e-06,
+      "loss": 0.2111,
+      "step": 5157
+    },
+    {
+      "epoch": 2.442379402189997,
+      "grad_norm": 1.0924980640411377,
+      "learning_rate": 1.7517682457253715e-06,
+      "loss": 0.218,
+      "step": 5158
+    },
+    {
+      "epoch": 2.4428529150636282,
+      "grad_norm": 1.1421464681625366,
+      "learning_rate": 1.7488778585658894e-06,
+      "loss": 0.2173,
+      "step": 5159
+    },
+    {
+      "epoch": 2.4433264279372597,
+      "grad_norm": 1.0318294763565063,
+      "learning_rate": 1.745989629407806e-06,
+      "loss": 0.2207,
+      "step": 5160
+    },
+    {
+      "epoch": 2.443799940810891,
+      "grad_norm": 1.0395057201385498,
+      "learning_rate": 1.7431035590065037e-06,
+      "loss": 0.174,
+      "step": 5161
+    },
+    {
+      "epoch": 2.444273453684522,
+      "grad_norm": 1.2233282327651978,
+      "learning_rate": 1.7402196481168132e-06,
+      "loss": 0.2084,
+      "step": 5162
+    },
+    {
+      "epoch": 2.4447469665581534,
+      "grad_norm": 1.3540483713150024,
+      "learning_rate": 1.7373378974929878e-06,
+      "loss": 0.1778,
+      "step": 5163
+    },
+    {
+      "epoch": 2.4452204794317844,
+      "grad_norm": 1.648417353630066,
+      "learning_rate": 1.7344583078887255e-06,
+      "loss": 0.2016,
+      "step": 5164
+    },
+    {
+      "epoch": 2.445693992305416,
+      "grad_norm": 1.0285402536392212,
+      "learning_rate": 1.731580880057152e-06,
+      "loss": 0.2081,
+      "step": 5165
+    },
+    {
+      "epoch": 2.446167505179047,
+      "grad_norm": 1.1315886974334717,
+      "learning_rate": 1.7287056147508353e-06,
+      "loss": 0.2134,
+      "step": 5166
+    },
+    {
+      "epoch": 2.446641018052678,
+      "grad_norm": 1.0791397094726562,
+      "learning_rate": 1.7258325127217668e-06,
+      "loss": 0.1921,
+      "step": 5167
+    },
+    {
+      "epoch": 2.4471145309263096,
+      "grad_norm": 1.0540249347686768,
+      "learning_rate": 1.7229615747213858e-06,
+      "loss": 0.2018,
+      "step": 5168
+    },
+    {
+      "epoch": 2.4475880437999407,
+      "grad_norm": 0.8960132002830505,
+      "learning_rate": 1.7200928015005546e-06,
+      "loss": 0.1999,
+      "step": 5169
+    },
+    {
+      "epoch": 2.448061556673572,
+      "grad_norm": 1.0253756046295166,
+      "learning_rate": 1.7172261938095713e-06,
+      "loss": 0.202,
+      "step": 5170
+    },
+    {
+      "epoch": 2.4485350695472032,
+      "grad_norm": 1.6087161302566528,
+      "learning_rate": 1.7143617523981737e-06,
+      "loss": 0.208,
+      "step": 5171
+    },
+    {
+      "epoch": 2.4490085824208347,
+      "grad_norm": 1.3046700954437256,
+      "learning_rate": 1.7114994780155236e-06,
+      "loss": 0.2251,
+      "step": 5172
+    },
+    {
+      "epoch": 2.449482095294466,
+      "grad_norm": 0.9480929374694824,
+      "learning_rate": 1.7086393714102278e-06,
+      "loss": 0.1973,
+      "step": 5173
+    },
+    {
+      "epoch": 2.449955608168097,
+      "grad_norm": 1.570953369140625,
+      "learning_rate": 1.7057814333303146e-06,
+      "loss": 0.1994,
+      "step": 5174
+    },
+    {
+      "epoch": 2.4504291210417284,
+      "grad_norm": 0.91864013671875,
+      "learning_rate": 1.7029256645232529e-06,
+      "loss": 0.1762,
+      "step": 5175
+    },
+    {
+      "epoch": 2.4509026339153595,
+      "grad_norm": 1.4908578395843506,
+      "learning_rate": 1.7000720657359383e-06,
+      "loss": 0.2058,
+      "step": 5176
+    },
+    {
+      "epoch": 2.451376146788991,
+      "grad_norm": 1.3648159503936768,
+      "learning_rate": 1.6972206377147072e-06,
+      "loss": 0.2089,
+      "step": 5177
+    },
+    {
+      "epoch": 2.451849659662622,
+      "grad_norm": 1.1475952863693237,
+      "learning_rate": 1.6943713812053185e-06,
+      "loss": 0.2,
+      "step": 5178
+    },
+    {
+      "epoch": 2.4523231725362535,
+      "grad_norm": 1.2770181894302368,
+      "learning_rate": 1.6915242969529676e-06,
+      "loss": 0.2011,
+      "step": 5179
+    },
+    {
+      "epoch": 2.4527966854098846,
+      "grad_norm": 0.8871325850486755,
+      "learning_rate": 1.6886793857022866e-06,
+      "loss": 0.1983,
+      "step": 5180
+    },
+    {
+      "epoch": 2.4532701982835157,
+      "grad_norm": 1.2607122659683228,
+      "learning_rate": 1.6858366481973288e-06,
+      "loss": 0.1899,
+      "step": 5181
+    },
+    {
+      "epoch": 2.453743711157147,
+      "grad_norm": 0.8917462825775146,
+      "learning_rate": 1.6829960851815896e-06,
+      "loss": 0.1821,
+      "step": 5182
+    },
+    {
+      "epoch": 2.4542172240307782,
+      "grad_norm": 0.9355318546295166,
+      "learning_rate": 1.680157697397986e-06,
+      "loss": 0.1975,
+      "step": 5183
+    },
+    {
+      "epoch": 2.4546907369044098,
+      "grad_norm": 2.099505662918091,
+      "learning_rate": 1.6773214855888765e-06,
+      "loss": 0.1737,
+      "step": 5184
+    },
+    {
+      "epoch": 2.455164249778041,
+      "grad_norm": 1.3527806997299194,
+      "learning_rate": 1.6744874504960395e-06,
+      "loss": 0.1821,
+      "step": 5185
+    },
+    {
+      "epoch": 2.455637762651672,
+      "grad_norm": 1.080901026725769,
+      "learning_rate": 1.6716555928606959e-06,
+      "loss": 0.1998,
+      "step": 5186
+    },
+    {
+      "epoch": 2.4561112755253034,
+      "grad_norm": 1.0348796844482422,
+      "learning_rate": 1.668825913423483e-06,
+      "loss": 0.204,
+      "step": 5187
+    },
+    {
+      "epoch": 2.4565847883989345,
+      "grad_norm": 1.0990900993347168,
+      "learning_rate": 1.665998412924481e-06,
+      "loss": 0.1958,
+      "step": 5188
+    },
+    {
+      "epoch": 2.457058301272566,
+      "grad_norm": 1.2333770990371704,
+      "learning_rate": 1.6631730921031964e-06,
+      "loss": 0.2101,
+      "step": 5189
+    },
+    {
+      "epoch": 2.457531814146197,
+      "grad_norm": 1.7082300186157227,
+      "learning_rate": 1.660349951698561e-06,
+      "loss": 0.1913,
+      "step": 5190
+    },
+    {
+      "epoch": 2.4580053270198285,
+      "grad_norm": 1.0094937086105347,
+      "learning_rate": 1.6575289924489435e-06,
+      "loss": 0.1842,
+      "step": 5191
+    },
+    {
+      "epoch": 2.4584788398934596,
+      "grad_norm": 1.1216864585876465,
+      "learning_rate": 1.6547102150921346e-06,
+      "loss": 0.2108,
+      "step": 5192
+    },
+    {
+      "epoch": 2.4589523527670907,
+      "grad_norm": 2.20843505859375,
+      "learning_rate": 1.6518936203653636e-06,
+      "loss": 0.184,
+      "step": 5193
+    },
+    {
+      "epoch": 2.459425865640722,
+      "grad_norm": 0.8962819576263428,
+      "learning_rate": 1.6490792090052799e-06,
+      "loss": 0.2049,
+      "step": 5194
+    },
+    {
+      "epoch": 2.4598993785143533,
+      "grad_norm": 1.1314330101013184,
+      "learning_rate": 1.6462669817479638e-06,
+      "loss": 0.21,
+      "step": 5195
+    },
+    {
+      "epoch": 2.4603728913879848,
+      "grad_norm": 1.110121726989746,
+      "learning_rate": 1.6434569393289313e-06,
+      "loss": 0.1899,
+      "step": 5196
+    },
+    {
+      "epoch": 2.460846404261616,
+      "grad_norm": 1.1242990493774414,
+      "learning_rate": 1.6406490824831166e-06,
+      "loss": 0.2234,
+      "step": 5197
+    },
+    {
+      "epoch": 2.4613199171352473,
+      "grad_norm": 0.946533739566803,
+      "learning_rate": 1.6378434119448939e-06,
+      "loss": 0.1964,
+      "step": 5198
+    },
+    {
+      "epoch": 2.4617934300088784,
+      "grad_norm": 1.0624065399169922,
+      "learning_rate": 1.6350399284480523e-06,
+      "loss": 0.1787,
+      "step": 5199
+    },
+    {
+      "epoch": 2.4622669428825095,
+      "grad_norm": 0.9491204619407654,
+      "learning_rate": 1.632238632725821e-06,
+      "loss": 0.1927,
+      "step": 5200
+    },
+    {
+      "epoch": 2.462740455756141,
+      "grad_norm": 1.2071702480316162,
+      "learning_rate": 1.6294395255108487e-06,
+      "loss": 0.1941,
+      "step": 5201
+    },
+    {
+      "epoch": 2.463213968629772,
+      "grad_norm": 1.1479326486587524,
+      "learning_rate": 1.6266426075352182e-06,
+      "loss": 0.2133,
+      "step": 5202
+    },
+    {
+      "epoch": 2.4636874815034036,
+      "grad_norm": 0.9430690407752991,
+      "learning_rate": 1.6238478795304346e-06,
+      "loss": 0.2007,
+      "step": 5203
+    },
+    {
+      "epoch": 2.4641609943770346,
+      "grad_norm": 1.0291746854782104,
+      "learning_rate": 1.62105534222743e-06,
+      "loss": 0.1892,
+      "step": 5204
+    },
+    {
+      "epoch": 2.4646345072506657,
+      "grad_norm": 1.4124772548675537,
+      "learning_rate": 1.61826499635657e-06,
+      "loss": 0.1997,
+      "step": 5205
+    },
+    {
+      "epoch": 2.465108020124297,
+      "grad_norm": 1.0831562280654907,
+      "learning_rate": 1.6154768426476375e-06,
+      "loss": 0.2015,
+      "step": 5206
+    },
+    {
+      "epoch": 2.4655815329979283,
+      "grad_norm": 1.5308729410171509,
+      "learning_rate": 1.6126908818298514e-06,
+      "loss": 0.1886,
+      "step": 5207
+    },
+    {
+      "epoch": 2.4660550458715598,
+      "grad_norm": 1.0939233303070068,
+      "learning_rate": 1.6099071146318502e-06,
+      "loss": 0.2037,
+      "step": 5208
+    },
+    {
+      "epoch": 2.466528558745191,
+      "grad_norm": 1.4337128400802612,
+      "learning_rate": 1.6071255417817045e-06,
+      "loss": 0.1942,
+      "step": 5209
+    },
+    {
+      "epoch": 2.467002071618822,
+      "grad_norm": 1.2987715005874634,
+      "learning_rate": 1.6043461640069025e-06,
+      "loss": 0.1953,
+      "step": 5210
+    },
+    {
+      "epoch": 2.4674755844924534,
+      "grad_norm": 1.0721392631530762,
+      "learning_rate": 1.6015689820343705e-06,
+      "loss": 0.2063,
+      "step": 5211
+    },
+    {
+      "epoch": 2.4679490973660845,
+      "grad_norm": 1.0339511632919312,
+      "learning_rate": 1.5987939965904498e-06,
+      "loss": 0.1766,
+      "step": 5212
+    },
+    {
+      "epoch": 2.468422610239716,
+      "grad_norm": 1.3787801265716553,
+      "learning_rate": 1.5960212084009097e-06,
+      "loss": 0.2037,
+      "step": 5213
+    },
+    {
+      "epoch": 2.468896123113347,
+      "grad_norm": 1.1688666343688965,
+      "learning_rate": 1.593250618190949e-06,
+      "loss": 0.1851,
+      "step": 5214
+    },
+    {
+      "epoch": 2.4693696359869786,
+      "grad_norm": 1.2239266633987427,
+      "learning_rate": 1.590482226685186e-06,
+      "loss": 0.2122,
+      "step": 5215
+    },
+    {
+      "epoch": 2.4698431488606096,
+      "grad_norm": 1.1540725231170654,
+      "learning_rate": 1.5877160346076714e-06,
+      "loss": 0.213,
+      "step": 5216
+    },
+    {
+      "epoch": 2.470316661734241,
+      "grad_norm": 0.8362421989440918,
+      "learning_rate": 1.584952042681871e-06,
+      "loss": 0.1734,
+      "step": 5217
+    },
+    {
+      "epoch": 2.470790174607872,
+      "grad_norm": 1.1025753021240234,
+      "learning_rate": 1.5821902516306842e-06,
+      "loss": 0.1877,
+      "step": 5218
+    },
+    {
+      "epoch": 2.4712636874815033,
+      "grad_norm": 1.23174250125885,
+      "learning_rate": 1.5794306621764265e-06,
+      "loss": 0.1998,
+      "step": 5219
+    },
+    {
+      "epoch": 2.471737200355135,
+      "grad_norm": 1.2430462837219238,
+      "learning_rate": 1.5766732750408465e-06,
+      "loss": 0.2024,
+      "step": 5220
+    },
+    {
+      "epoch": 2.472210713228766,
+      "grad_norm": 1.0162687301635742,
+      "learning_rate": 1.573918090945109e-06,
+      "loss": 0.2102,
+      "step": 5221
+    },
+    {
+      "epoch": 2.4726842261023974,
+      "grad_norm": 1.1504173278808594,
+      "learning_rate": 1.571165110609808e-06,
+      "loss": 0.2077,
+      "step": 5222
+    },
+    {
+      "epoch": 2.4731577389760284,
+      "grad_norm": 0.9471516609191895,
+      "learning_rate": 1.5684143347549586e-06,
+      "loss": 0.1864,
+      "step": 5223
+    },
+    {
+      "epoch": 2.4736312518496595,
+      "grad_norm": 1.431763768196106,
+      "learning_rate": 1.5656657640999973e-06,
+      "loss": 0.1995,
+      "step": 5224
+    },
+    {
+      "epoch": 2.474104764723291,
+      "grad_norm": 1.2276103496551514,
+      "learning_rate": 1.562919399363787e-06,
+      "loss": 0.2024,
+      "step": 5225
+    },
+    {
+      "epoch": 2.474578277596922,
+      "grad_norm": 1.1695070266723633,
+      "learning_rate": 1.5601752412646143e-06,
+      "loss": 0.1863,
+      "step": 5226
+    },
+    {
+      "epoch": 2.4750517904705536,
+      "grad_norm": 1.6259397268295288,
+      "learning_rate": 1.5574332905201883e-06,
+      "loss": 0.1961,
+      "step": 5227
+    },
+    {
+      "epoch": 2.4755253033441846,
+      "grad_norm": 1.0487055778503418,
+      "learning_rate": 1.554693547847639e-06,
+      "loss": 0.1864,
+      "step": 5228
+    },
+    {
+      "epoch": 2.4759988162178157,
+      "grad_norm": 1.0676391124725342,
+      "learning_rate": 1.551956013963517e-06,
+      "loss": 0.2179,
+      "step": 5229
+    },
+    {
+      "epoch": 2.476472329091447,
+      "grad_norm": 1.1580907106399536,
+      "learning_rate": 1.5492206895838013e-06,
+      "loss": 0.198,
+      "step": 5230
+    },
+    {
+      "epoch": 2.4769458419650783,
+      "grad_norm": 1.0551162958145142,
+      "learning_rate": 1.546487575423886e-06,
+      "loss": 0.1941,
+      "step": 5231
+    },
+    {
+      "epoch": 2.47741935483871,
+      "grad_norm": 1.4372506141662598,
+      "learning_rate": 1.5437566721985952e-06,
+      "loss": 0.1876,
+      "step": 5232
+    },
+    {
+      "epoch": 2.477892867712341,
+      "grad_norm": 1.0641188621520996,
+      "learning_rate": 1.5410279806221662e-06,
+      "loss": 0.2156,
+      "step": 5233
+    },
+    {
+      "epoch": 2.4783663805859724,
+      "grad_norm": 1.8951334953308105,
+      "learning_rate": 1.5383015014082659e-06,
+      "loss": 0.1862,
+      "step": 5234
+    },
+    {
+      "epoch": 2.4788398934596034,
+      "grad_norm": 1.3256562948226929,
+      "learning_rate": 1.5355772352699738e-06,
+      "loss": 0.1894,
+      "step": 5235
+    },
+    {
+      "epoch": 2.4793134063332345,
+      "grad_norm": 0.990827739238739,
+      "learning_rate": 1.5328551829198e-06,
+      "loss": 0.1896,
+      "step": 5236
+    },
+    {
+      "epoch": 2.479786919206866,
+      "grad_norm": 1.4543442726135254,
+      "learning_rate": 1.53013534506967e-06,
+      "loss": 0.2072,
+      "step": 5237
+    },
+    {
+      "epoch": 2.480260432080497,
+      "grad_norm": 1.17643141746521,
+      "learning_rate": 1.5274177224309273e-06,
+      "loss": 0.2118,
+      "step": 5238
+    },
+    {
+      "epoch": 2.4807339449541286,
+      "grad_norm": 1.1279579401016235,
+      "learning_rate": 1.5247023157143459e-06,
+      "loss": 0.2043,
+      "step": 5239
+    },
+    {
+      "epoch": 2.4812074578277596,
+      "grad_norm": 1.2052768468856812,
+      "learning_rate": 1.5219891256301079e-06,
+      "loss": 0.2013,
+      "step": 5240
+    },
+    {
+      "epoch": 2.481680970701391,
+      "grad_norm": 1.2631618976593018,
+      "learning_rate": 1.5192781528878285e-06,
+      "loss": 0.2011,
+      "step": 5241
+    },
+    {
+      "epoch": 2.482154483575022,
+      "grad_norm": 1.0374186038970947,
+      "learning_rate": 1.5165693981965302e-06,
+      "loss": 0.1748,
+      "step": 5242
+    },
+    {
+      "epoch": 2.4826279964486533,
+      "grad_norm": 1.1341975927352905,
+      "learning_rate": 1.513862862264668e-06,
+      "loss": 0.214,
+      "step": 5243
+    },
+    {
+      "epoch": 2.483101509322285,
+      "grad_norm": 0.9370356202125549,
+      "learning_rate": 1.5111585458001032e-06,
+      "loss": 0.1778,
+      "step": 5244
+    },
+    {
+      "epoch": 2.483575022195916,
+      "grad_norm": 1.78024423122406,
+      "learning_rate": 1.5084564495101306e-06,
+      "loss": 0.1897,
+      "step": 5245
+    },
+    {
+      "epoch": 2.4840485350695474,
+      "grad_norm": 1.1708461046218872,
+      "learning_rate": 1.5057565741014513e-06,
+      "loss": 0.2309,
+      "step": 5246
+    },
+    {
+      "epoch": 2.4845220479431784,
+      "grad_norm": 1.8832025527954102,
+      "learning_rate": 1.5030589202801982e-06,
+      "loss": 0.2073,
+      "step": 5247
+    },
+    {
+      "epoch": 2.4849955608168095,
+      "grad_norm": 1.0782551765441895,
+      "learning_rate": 1.5003634887519126e-06,
+      "loss": 0.2115,
+      "step": 5248
+    },
+    {
+      "epoch": 2.485469073690441,
+      "grad_norm": 2.393939733505249,
+      "learning_rate": 1.497670280221556e-06,
+      "loss": 0.202,
+      "step": 5249
+    },
+    {
+      "epoch": 2.485942586564072,
+      "grad_norm": 1.293050765991211,
+      "learning_rate": 1.4949792953935172e-06,
+      "loss": 0.1859,
+      "step": 5250
+    },
+    {
+      "epoch": 2.4864160994377036,
+      "grad_norm": 1.0516685247421265,
+      "learning_rate": 1.4922905349715922e-06,
+      "loss": 0.2269,
+      "step": 5251
+    },
+    {
+      "epoch": 2.4868896123113347,
+      "grad_norm": 0.9482864141464233,
+      "learning_rate": 1.489603999659004e-06,
+      "loss": 0.1947,
+      "step": 5252
+    },
+    {
+      "epoch": 2.4873631251849657,
+      "grad_norm": 1.2541604042053223,
+      "learning_rate": 1.486919690158386e-06,
+      "loss": 0.2006,
+      "step": 5253
+    },
+    {
+      "epoch": 2.4878366380585972,
+      "grad_norm": 1.3161418437957764,
+      "learning_rate": 1.4842376071717989e-06,
+      "loss": 0.2067,
+      "step": 5254
+    },
+    {
+      "epoch": 2.4883101509322283,
+      "grad_norm": 1.2702239751815796,
+      "learning_rate": 1.4815577514007106e-06,
+      "loss": 0.1866,
+      "step": 5255
+    },
+    {
+      "epoch": 2.48878366380586,
+      "grad_norm": 1.2881168127059937,
+      "learning_rate": 1.478880123546015e-06,
+      "loss": 0.1928,
+      "step": 5256
+    },
+    {
+      "epoch": 2.489257176679491,
+      "grad_norm": 1.0193349123001099,
+      "learning_rate": 1.476204724308019e-06,
+      "loss": 0.2026,
+      "step": 5257
+    },
+    {
+      "epoch": 2.4897306895531224,
+      "grad_norm": 1.3195552825927734,
+      "learning_rate": 1.4735315543864436e-06,
+      "loss": 0.1803,
+      "step": 5258
+    },
+    {
+      "epoch": 2.4902042024267534,
+      "grad_norm": 1.2184456586837769,
+      "learning_rate": 1.4708606144804371e-06,
+      "loss": 0.2057,
+      "step": 5259
+    },
+    {
+      "epoch": 2.490677715300385,
+      "grad_norm": 1.221314549446106,
+      "learning_rate": 1.468191905288553e-06,
+      "loss": 0.1912,
+      "step": 5260
+    },
+    {
+      "epoch": 2.491151228174016,
+      "grad_norm": 1.2005162239074707,
+      "learning_rate": 1.4655254275087693e-06,
+      "loss": 0.2101,
+      "step": 5261
+    },
+    {
+      "epoch": 2.491624741047647,
+      "grad_norm": 2.1400723457336426,
+      "learning_rate": 1.4628611818384753e-06,
+      "loss": 0.1901,
+      "step": 5262
+    },
+    {
+      "epoch": 2.4920982539212786,
+      "grad_norm": 1.6290489435195923,
+      "learning_rate": 1.460199168974481e-06,
+      "loss": 0.1929,
+      "step": 5263
+    },
+    {
+      "epoch": 2.4925717667949097,
+      "grad_norm": 1.020664095878601,
+      "learning_rate": 1.4575393896130073e-06,
+      "loss": 0.1879,
+      "step": 5264
+    },
+    {
+      "epoch": 2.493045279668541,
+      "grad_norm": 0.9219028353691101,
+      "learning_rate": 1.454881844449697e-06,
+      "loss": 0.2014,
+      "step": 5265
+    },
+    {
+      "epoch": 2.4935187925421722,
+      "grad_norm": 1.7500580549240112,
+      "learning_rate": 1.4522265341796048e-06,
+      "loss": 0.2044,
+      "step": 5266
+    },
+    {
+      "epoch": 2.4939923054158033,
+      "grad_norm": 1.1425151824951172,
+      "learning_rate": 1.4495734594971988e-06,
+      "loss": 0.2053,
+      "step": 5267
+    },
+    {
+      "epoch": 2.494465818289435,
+      "grad_norm": 1.0946452617645264,
+      "learning_rate": 1.4469226210963693e-06,
+      "loss": 0.2082,
+      "step": 5268
+    },
+    {
+      "epoch": 2.494939331163066,
+      "grad_norm": 1.198346495628357,
+      "learning_rate": 1.444274019670413e-06,
+      "loss": 0.1991,
+      "step": 5269
+    },
+    {
+      "epoch": 2.4954128440366974,
+      "grad_norm": 1.0651274919509888,
+      "learning_rate": 1.4416276559120511e-06,
+      "loss": 0.2051,
+      "step": 5270
+    },
+    {
+      "epoch": 2.4958863569103285,
+      "grad_norm": 1.0308165550231934,
+      "learning_rate": 1.4389835305134092e-06,
+      "loss": 0.1973,
+      "step": 5271
+    },
+    {
+      "epoch": 2.4963598697839595,
+      "grad_norm": 1.7552647590637207,
+      "learning_rate": 1.436341644166037e-06,
+      "loss": 0.2061,
+      "step": 5272
+    },
+    {
+      "epoch": 2.496833382657591,
+      "grad_norm": 1.7813093662261963,
+      "learning_rate": 1.4337019975608934e-06,
+      "loss": 0.1912,
+      "step": 5273
+    },
+    {
+      "epoch": 2.497306895531222,
+      "grad_norm": 1.5826293230056763,
+      "learning_rate": 1.4310645913883493e-06,
+      "loss": 0.2167,
+      "step": 5274
+    },
+    {
+      "epoch": 2.4977804084048536,
+      "grad_norm": 0.9368717670440674,
+      "learning_rate": 1.4284294263381982e-06,
+      "loss": 0.2067,
+      "step": 5275
+    },
+    {
+      "epoch": 2.4982539212784847,
+      "grad_norm": 1.000964879989624,
+      "learning_rate": 1.4257965030996357e-06,
+      "loss": 0.1868,
+      "step": 5276
+    },
+    {
+      "epoch": 2.498727434152116,
+      "grad_norm": 1.1660606861114502,
+      "learning_rate": 1.4231658223612842e-06,
+      "loss": 0.2016,
+      "step": 5277
+    },
+    {
+      "epoch": 2.4992009470257472,
+      "grad_norm": 1.216518521308899,
+      "learning_rate": 1.420537384811167e-06,
+      "loss": 0.2274,
+      "step": 5278
+    },
+    {
+      "epoch": 2.4996744598993788,
+      "grad_norm": 1.3850187063217163,
+      "learning_rate": 1.4179111911367315e-06,
+      "loss": 0.2104,
+      "step": 5279
+    },
+    {
+      "epoch": 2.50014797277301,
+      "grad_norm": 1.306488275527954,
+      "learning_rate": 1.4152872420248288e-06,
+      "loss": 0.2175,
+      "step": 5280
+    },
+    {
+      "epoch": 2.500621485646641,
+      "grad_norm": 1.0250792503356934,
+      "learning_rate": 1.4126655381617327e-06,
+      "loss": 0.181,
+      "step": 5281
+    },
+    {
+      "epoch": 2.5010949985202724,
+      "grad_norm": 1.5189791917800903,
+      "learning_rate": 1.4100460802331205e-06,
+      "loss": 0.2075,
+      "step": 5282
+    },
+    {
+      "epoch": 2.5015685113939035,
+      "grad_norm": 1.1778138875961304,
+      "learning_rate": 1.4074288689240856e-06,
+      "loss": 0.1985,
+      "step": 5283
+    },
+    {
+      "epoch": 2.502042024267535,
+      "grad_norm": 1.0283559560775757,
+      "learning_rate": 1.4048139049191389e-06,
+      "loss": 0.2098,
+      "step": 5284
+    },
+    {
+      "epoch": 2.502515537141166,
+      "grad_norm": 0.9769028425216675,
+      "learning_rate": 1.4022011889021936e-06,
+      "loss": 0.2068,
+      "step": 5285
+    },
+    {
+      "epoch": 2.502989050014797,
+      "grad_norm": 1.1810842752456665,
+      "learning_rate": 1.399590721556584e-06,
+      "loss": 0.2083,
+      "step": 5286
+    },
+    {
+      "epoch": 2.5034625628884286,
+      "grad_norm": 1.018278956413269,
+      "learning_rate": 1.396982503565051e-06,
+      "loss": 0.2003,
+      "step": 5287
+    },
+    {
+      "epoch": 2.5039360757620597,
+      "grad_norm": 1.1597832441329956,
+      "learning_rate": 1.3943765356097505e-06,
+      "loss": 0.1902,
+      "step": 5288
+    },
+    {
+      "epoch": 2.504409588635691,
+      "grad_norm": 1.001120686531067,
+      "learning_rate": 1.3917728183722456e-06,
+      "loss": 0.1733,
+      "step": 5289
+    },
+    {
+      "epoch": 2.5048831015093223,
+      "grad_norm": 1.1091008186340332,
+      "learning_rate": 1.389171352533517e-06,
+      "loss": 0.2251,
+      "step": 5290
+    },
+    {
+      "epoch": 2.5053566143829533,
+      "grad_norm": 0.9414849877357483,
+      "learning_rate": 1.3865721387739507e-06,
+      "loss": 0.1797,
+      "step": 5291
+    },
+    {
+      "epoch": 2.505830127256585,
+      "grad_norm": 1.9510407447814941,
+      "learning_rate": 1.3839751777733445e-06,
+      "loss": 0.1881,
+      "step": 5292
+    },
+    {
+      "epoch": 2.506303640130216,
+      "grad_norm": 1.3246495723724365,
+      "learning_rate": 1.3813804702109124e-06,
+      "loss": 0.2084,
+      "step": 5293
+    },
+    {
+      "epoch": 2.5067771530038474,
+      "grad_norm": 1.0592890977859497,
+      "learning_rate": 1.37878801676527e-06,
+      "loss": 0.2072,
+      "step": 5294
+    },
+    {
+      "epoch": 2.5072506658774785,
+      "grad_norm": 1.3438637256622314,
+      "learning_rate": 1.3761978181144542e-06,
+      "loss": 0.2262,
+      "step": 5295
+    },
+    {
+      "epoch": 2.5077241787511095,
+      "grad_norm": 1.0453280210494995,
+      "learning_rate": 1.373609874935903e-06,
+      "loss": 0.1857,
+      "step": 5296
+    },
+    {
+      "epoch": 2.508197691624741,
+      "grad_norm": 1.3020319938659668,
+      "learning_rate": 1.3710241879064689e-06,
+      "loss": 0.1875,
+      "step": 5297
+    },
+    {
+      "epoch": 2.5086712044983726,
+      "grad_norm": 1.245324969291687,
+      "learning_rate": 1.3684407577024116e-06,
+      "loss": 0.1978,
+      "step": 5298
+    },
+    {
+      "epoch": 2.5091447173720036,
+      "grad_norm": 1.0261917114257812,
+      "learning_rate": 1.3658595849994072e-06,
+      "loss": 0.199,
+      "step": 5299
+    },
+    {
+      "epoch": 2.5096182302456347,
+      "grad_norm": 1.699816107749939,
+      "learning_rate": 1.36328067047253e-06,
+      "loss": 0.2069,
+      "step": 5300
+    },
+    {
+      "epoch": 2.510091743119266,
+      "grad_norm": 2.2417402267456055,
+      "learning_rate": 1.360704014796277e-06,
+      "loss": 0.1876,
+      "step": 5301
+    },
+    {
+      "epoch": 2.5105652559928973,
+      "grad_norm": 0.9877926707267761,
+      "learning_rate": 1.3581296186445426e-06,
+      "loss": 0.1888,
+      "step": 5302
+    },
+    {
+      "epoch": 2.5110387688665288,
+      "grad_norm": 1.183518648147583,
+      "learning_rate": 1.3555574826906337e-06,
+      "loss": 0.1995,
+      "step": 5303
+    },
+    {
+      "epoch": 2.51151228174016,
+      "grad_norm": 1.0759626626968384,
+      "learning_rate": 1.3529876076072746e-06,
+      "loss": 0.1998,
+      "step": 5304
+    },
+    {
+      "epoch": 2.511985794613791,
+      "grad_norm": 1.130675196647644,
+      "learning_rate": 1.3504199940665852e-06,
+      "loss": 0.1963,
+      "step": 5305
+    },
+    {
+      "epoch": 2.5124593074874224,
+      "grad_norm": 1.4707733392715454,
+      "learning_rate": 1.347854642740104e-06,
+      "loss": 0.216,
+      "step": 5306
+    },
+    {
+      "epoch": 2.5129328203610535,
+      "grad_norm": 1.2437856197357178,
+      "learning_rate": 1.3452915542987732e-06,
+      "loss": 0.2187,
+      "step": 5307
+    },
+    {
+      "epoch": 2.513406333234685,
+      "grad_norm": 1.0514782667160034,
+      "learning_rate": 1.3427307294129411e-06,
+      "loss": 0.2178,
+      "step": 5308
+    },
+    {
+      "epoch": 2.513879846108316,
+      "grad_norm": 1.6245321035385132,
+      "learning_rate": 1.3401721687523706e-06,
+      "loss": 0.2095,
+      "step": 5309
+    },
+    {
+      "epoch": 2.514353358981947,
+      "grad_norm": 1.0891262292861938,
+      "learning_rate": 1.3376158729862232e-06,
+      "loss": 0.1807,
+      "step": 5310
+    },
+    {
+      "epoch": 2.5148268718555786,
+      "grad_norm": 1.054901361465454,
+      "learning_rate": 1.3350618427830796e-06,
+      "loss": 0.2074,
+      "step": 5311
+    },
+    {
+      "epoch": 2.5153003847292097,
+      "grad_norm": 1.5967611074447632,
+      "learning_rate": 1.3325100788109168e-06,
+      "loss": 0.1997,
+      "step": 5312
+    },
+    {
+      "epoch": 2.515773897602841,
+      "grad_norm": 1.1578254699707031,
+      "learning_rate": 1.3299605817371285e-06,
+      "loss": 0.2056,
+      "step": 5313
+    },
+    {
+      "epoch": 2.5162474104764723,
+      "grad_norm": 0.9452877044677734,
+      "learning_rate": 1.327413352228506e-06,
+      "loss": 0.2101,
+      "step": 5314
+    },
+    {
+      "epoch": 2.5167209233501033,
+      "grad_norm": 1.0103105306625366,
+      "learning_rate": 1.3248683909512584e-06,
+      "loss": 0.1875,
+      "step": 5315
+    },
+    {
+      "epoch": 2.517194436223735,
+      "grad_norm": 1.4317337274551392,
+      "learning_rate": 1.322325698570992e-06,
+      "loss": 0.1962,
+      "step": 5316
+    },
+    {
+      "epoch": 2.5176679490973664,
+      "grad_norm": 1.23985755443573,
+      "learning_rate": 1.3197852757527219e-06,
+      "loss": 0.2054,
+      "step": 5317
+    },
+    {
+      "epoch": 2.5181414619709974,
+      "grad_norm": 1.0402408838272095,
+      "learning_rate": 1.3172471231608753e-06,
+      "loss": 0.1856,
+      "step": 5318
+    },
+    {
+      "epoch": 2.5186149748446285,
+      "grad_norm": 1.3654496669769287,
+      "learning_rate": 1.3147112414592777e-06,
+      "loss": 0.2111,
+      "step": 5319
+    },
+    {
+      "epoch": 2.51908848771826,
+      "grad_norm": 1.1006624698638916,
+      "learning_rate": 1.312177631311169e-06,
+      "loss": 0.2167,
+      "step": 5320
+    },
+    {
+      "epoch": 2.519562000591891,
+      "grad_norm": 0.9989845752716064,
+      "learning_rate": 1.3096462933791853e-06,
+      "loss": 0.1995,
+      "step": 5321
+    },
+    {
+      "epoch": 2.5200355134655226,
+      "grad_norm": 1.1869349479675293,
+      "learning_rate": 1.3071172283253786e-06,
+      "loss": 0.2215,
+      "step": 5322
+    },
+    {
+      "epoch": 2.5205090263391536,
+      "grad_norm": 0.9278461933135986,
+      "learning_rate": 1.3045904368111973e-06,
+      "loss": 0.194,
+      "step": 5323
+    },
+    {
+      "epoch": 2.5209825392127847,
+      "grad_norm": 1.345420241355896,
+      "learning_rate": 1.3020659194975028e-06,
+      "loss": 0.2158,
+      "step": 5324
+    },
+    {
+      "epoch": 2.521456052086416,
+      "grad_norm": 1.0108526945114136,
+      "learning_rate": 1.2995436770445547e-06,
+      "loss": 0.1979,
+      "step": 5325
+    },
+    {
+      "epoch": 2.5219295649600473,
+      "grad_norm": 1.0445326566696167,
+      "learning_rate": 1.2970237101120253e-06,
+      "loss": 0.1861,
+      "step": 5326
+    },
+    {
+      "epoch": 2.522403077833679,
+      "grad_norm": 1.2074165344238281,
+      "learning_rate": 1.2945060193589852e-06,
+      "loss": 0.2149,
+      "step": 5327
+    },
+    {
+      "epoch": 2.52287659070731,
+      "grad_norm": 0.880549430847168,
+      "learning_rate": 1.2919906054439103e-06,
+      "loss": 0.1945,
+      "step": 5328
+    },
+    {
+      "epoch": 2.523350103580941,
+      "grad_norm": 1.272971749305725,
+      "learning_rate": 1.289477469024687e-06,
+      "loss": 0.1983,
+      "step": 5329
+    },
+    {
+      "epoch": 2.5238236164545724,
+      "grad_norm": 1.3522940874099731,
+      "learning_rate": 1.2869666107585975e-06,
+      "loss": 0.2175,
+      "step": 5330
+    },
+    {
+      "epoch": 2.5242971293282035,
+      "grad_norm": 1.0059926509857178,
+      "learning_rate": 1.2844580313023368e-06,
+      "loss": 0.1978,
+      "step": 5331
+    },
+    {
+      "epoch": 2.524770642201835,
+      "grad_norm": 0.781419038772583,
+      "learning_rate": 1.2819517313119956e-06,
+      "loss": 0.1766,
+      "step": 5332
+    },
+    {
+      "epoch": 2.525244155075466,
+      "grad_norm": 1.2181997299194336,
+      "learning_rate": 1.279447711443077e-06,
+      "loss": 0.1921,
+      "step": 5333
+    },
+    {
+      "epoch": 2.525717667949097,
+      "grad_norm": 1.028064489364624,
+      "learning_rate": 1.2769459723504795e-06,
+      "loss": 0.1777,
+      "step": 5334
+    },
+    {
+      "epoch": 2.5261911808227286,
+      "grad_norm": 1.0090110301971436,
+      "learning_rate": 1.274446514688511e-06,
+      "loss": 0.2018,
+      "step": 5335
+    },
+    {
+      "epoch": 2.5266646936963597,
+      "grad_norm": 1.4929478168487549,
+      "learning_rate": 1.2719493391108806e-06,
+      "loss": 0.1938,
+      "step": 5336
+    },
+    {
+      "epoch": 2.527138206569991,
+      "grad_norm": 1.0513087511062622,
+      "learning_rate": 1.2694544462706959e-06,
+      "loss": 0.1879,
+      "step": 5337
+    },
+    {
+      "epoch": 2.5276117194436223,
+      "grad_norm": 1.1161446571350098,
+      "learning_rate": 1.2669618368204795e-06,
+      "loss": 0.1997,
+      "step": 5338
+    },
+    {
+      "epoch": 2.5280852323172534,
+      "grad_norm": 1.209606647491455,
+      "learning_rate": 1.2644715114121432e-06,
+      "loss": 0.1928,
+      "step": 5339
+    },
+    {
+      "epoch": 2.528558745190885,
+      "grad_norm": 1.125483512878418,
+      "learning_rate": 1.2619834706970113e-06,
+      "loss": 0.1892,
+      "step": 5340
+    },
+    {
+      "epoch": 2.5290322580645164,
+      "grad_norm": 1.3125486373901367,
+      "learning_rate": 1.2594977153258036e-06,
+      "loss": 0.2068,
+      "step": 5341
+    },
+    {
+      "epoch": 2.5295057709381474,
+      "grad_norm": 1.0961823463439941,
+      "learning_rate": 1.2570142459486478e-06,
+      "loss": 0.1934,
+      "step": 5342
+    },
+    {
+      "epoch": 2.5299792838117785,
+      "grad_norm": 1.161319375038147,
+      "learning_rate": 1.254533063215072e-06,
+      "loss": 0.1815,
+      "step": 5343
+    },
+    {
+      "epoch": 2.53045279668541,
+      "grad_norm": 1.1228846311569214,
+      "learning_rate": 1.2520541677740038e-06,
+      "loss": 0.1877,
+      "step": 5344
+    },
+    {
+      "epoch": 2.530926309559041,
+      "grad_norm": 1.2616682052612305,
+      "learning_rate": 1.2495775602737759e-06,
+      "loss": 0.2004,
+      "step": 5345
+    },
+    {
+      "epoch": 2.5313998224326726,
+      "grad_norm": 1.3314753770828247,
+      "learning_rate": 1.2471032413621188e-06,
+      "loss": 0.223,
+      "step": 5346
+    },
+    {
+      "epoch": 2.5318733353063037,
+      "grad_norm": 1.303731918334961,
+      "learning_rate": 1.2446312116861703e-06,
+      "loss": 0.1988,
+      "step": 5347
+    },
+    {
+      "epoch": 2.5323468481799347,
+      "grad_norm": 1.0389883518218994,
+      "learning_rate": 1.2421614718924623e-06,
+      "loss": 0.194,
+      "step": 5348
+    },
+    {
+      "epoch": 2.5328203610535662,
+      "grad_norm": 1.4113802909851074,
+      "learning_rate": 1.239694022626935e-06,
+      "loss": 0.1868,
+      "step": 5349
+    },
+    {
+      "epoch": 2.5332938739271973,
+      "grad_norm": 1.1160602569580078,
+      "learning_rate": 1.2372288645349207e-06,
+      "loss": 0.1914,
+      "step": 5350
+    },
+    {
+      "epoch": 2.533767386800829,
+      "grad_norm": 1.1180630922317505,
+      "learning_rate": 1.2347659982611637e-06,
+      "loss": 0.187,
+      "step": 5351
+    },
+    {
+      "epoch": 2.53424089967446,
+      "grad_norm": 2.2258567810058594,
+      "learning_rate": 1.2323054244498001e-06,
+      "loss": 0.1949,
+      "step": 5352
+    },
+    {
+      "epoch": 2.534714412548091,
+      "grad_norm": 1.0925301313400269,
+      "learning_rate": 1.2298471437443671e-06,
+      "loss": 0.1731,
+      "step": 5353
+    },
+    {
+      "epoch": 2.5351879254217224,
+      "grad_norm": 1.3025660514831543,
+      "learning_rate": 1.2273911567878095e-06,
+      "loss": 0.2104,
+      "step": 5354
+    },
+    {
+      "epoch": 2.5356614382953535,
+      "grad_norm": 1.7932027578353882,
+      "learning_rate": 1.224937464222461e-06,
+      "loss": 0.202,
+      "step": 5355
+    },
+    {
+      "epoch": 2.536134951168985,
+      "grad_norm": 1.089361310005188,
+      "learning_rate": 1.222486066690066e-06,
+      "loss": 0.2003,
+      "step": 5356
+    },
+    {
+      "epoch": 2.536608464042616,
+      "grad_norm": 0.9862686395645142,
+      "learning_rate": 1.22003696483176e-06,
+      "loss": 0.218,
+      "step": 5357
+    },
+    {
+      "epoch": 2.537081976916247,
+      "grad_norm": 1.0599945783615112,
+      "learning_rate": 1.2175901592880867e-06,
+      "loss": 0.2115,
+      "step": 5358
+    },
+    {
+      "epoch": 2.5375554897898787,
+      "grad_norm": 1.217585563659668,
+      "learning_rate": 1.21514565069898e-06,
+      "loss": 0.1972,
+      "step": 5359
+    },
+    {
+      "epoch": 2.53802900266351,
+      "grad_norm": 1.0531421899795532,
+      "learning_rate": 1.2127034397037808e-06,
+      "loss": 0.1863,
+      "step": 5360
+    },
+    {
+      "epoch": 2.5385025155371412,
+      "grad_norm": 1.6633493900299072,
+      "learning_rate": 1.2102635269412244e-06,
+      "loss": 0.184,
+      "step": 5361
+    },
+    {
+      "epoch": 2.5389760284107723,
+      "grad_norm": 1.1086888313293457,
+      "learning_rate": 1.207825913049445e-06,
+      "loss": 0.2196,
+      "step": 5362
+    },
+    {
+      "epoch": 2.539449541284404,
+      "grad_norm": 1.2374646663665771,
+      "learning_rate": 1.2053905986659798e-06,
+      "loss": 0.2041,
+      "step": 5363
+    },
+    {
+      "epoch": 2.539923054158035,
+      "grad_norm": 0.9073995351791382,
+      "learning_rate": 1.2029575844277585e-06,
+      "loss": 0.1934,
+      "step": 5364
+    },
+    {
+      "epoch": 2.5403965670316664,
+      "grad_norm": 1.31955885887146,
+      "learning_rate": 1.2005268709711172e-06,
+      "loss": 0.1862,
+      "step": 5365
+    },
+    {
+      "epoch": 2.5408700799052975,
+      "grad_norm": 0.9501339793205261,
+      "learning_rate": 1.1980984589317802e-06,
+      "loss": 0.1849,
+      "step": 5366
+    },
+    {
+      "epoch": 2.5413435927789285,
+      "grad_norm": 1.0616928339004517,
+      "learning_rate": 1.1956723489448796e-06,
+      "loss": 0.2111,
+      "step": 5367
+    },
+    {
+      "epoch": 2.54181710565256,
+      "grad_norm": 1.4748432636260986,
+      "learning_rate": 1.1932485416449369e-06,
+      "loss": 0.1916,
+      "step": 5368
+    },
+    {
+      "epoch": 2.542290618526191,
+      "grad_norm": 1.1778125762939453,
+      "learning_rate": 1.1908270376658804e-06,
+      "loss": 0.2055,
+      "step": 5369
+    },
+    {
+      "epoch": 2.5427641313998226,
+      "grad_norm": 1.297905445098877,
+      "learning_rate": 1.1884078376410291e-06,
+      "loss": 0.2062,
+      "step": 5370
+    },
+    {
+      "epoch": 2.5432376442734537,
+      "grad_norm": 0.9151560664176941,
+      "learning_rate": 1.1859909422030991e-06,
+      "loss": 0.1974,
+      "step": 5371
+    },
+    {
+      "epoch": 2.5437111571470847,
+      "grad_norm": 1.6805721521377563,
+      "learning_rate": 1.1835763519842092e-06,
+      "loss": 0.2103,
+      "step": 5372
+    },
+    {
+      "epoch": 2.5441846700207162,
+      "grad_norm": 1.0770697593688965,
+      "learning_rate": 1.1811640676158686e-06,
+      "loss": 0.1965,
+      "step": 5373
+    },
+    {
+      "epoch": 2.5446581828943473,
+      "grad_norm": 1.3352454900741577,
+      "learning_rate": 1.1787540897289918e-06,
+      "loss": 0.2031,
+      "step": 5374
+    },
+    {
+      "epoch": 2.545131695767979,
+      "grad_norm": 1.8438594341278076,
+      "learning_rate": 1.176346418953881e-06,
+      "loss": 0.1951,
+      "step": 5375
+    },
+    {
+      "epoch": 2.54560520864161,
+      "grad_norm": 1.525089144706726,
+      "learning_rate": 1.1739410559202425e-06,
+      "loss": 0.2543,
+      "step": 5376
+    },
+    {
+      "epoch": 2.546078721515241,
+      "grad_norm": 1.3220195770263672,
+      "learning_rate": 1.171538001257172e-06,
+      "loss": 0.2211,
+      "step": 5377
+    },
+    {
+      "epoch": 2.5465522343888725,
+      "grad_norm": 1.0386881828308105,
+      "learning_rate": 1.16913725559317e-06,
+      "loss": 0.2002,
+      "step": 5378
+    },
+    {
+      "epoch": 2.547025747262504,
+      "grad_norm": 1.009458303451538,
+      "learning_rate": 1.1667388195561247e-06,
+      "loss": 0.1873,
+      "step": 5379
+    },
+    {
+      "epoch": 2.547499260136135,
+      "grad_norm": 1.3089097738265991,
+      "learning_rate": 1.164342693773326e-06,
+      "loss": 0.1998,
+      "step": 5380
+    },
+    {
+      "epoch": 2.547972773009766,
+      "grad_norm": 1.061797022819519,
+      "learning_rate": 1.161948878871455e-06,
+      "loss": 0.2146,
+      "step": 5381
+    },
+    {
+      "epoch": 2.5484462858833976,
+      "grad_norm": 0.9451477527618408,
+      "learning_rate": 1.1595573754765932e-06,
+      "loss": 0.2215,
+      "step": 5382
+    },
+    {
+      "epoch": 2.5489197987570287,
+      "grad_norm": 1.7291032075881958,
+      "learning_rate": 1.1571681842142158e-06,
+      "loss": 0.1936,
+      "step": 5383
+    },
+    {
+      "epoch": 2.54939331163066,
+      "grad_norm": 1.9363996982574463,
+      "learning_rate": 1.1547813057091906e-06,
+      "loss": 0.2136,
+      "step": 5384
+    },
+    {
+      "epoch": 2.5498668245042913,
+      "grad_norm": 0.9909132122993469,
+      "learning_rate": 1.1523967405857838e-06,
+      "loss": 0.1925,
+      "step": 5385
+    },
+    {
+      "epoch": 2.5503403373779223,
+      "grad_norm": 1.205962896347046,
+      "learning_rate": 1.1500144894676568e-06,
+      "loss": 0.2009,
+      "step": 5386
+    },
+    {
+      "epoch": 2.550813850251554,
+      "grad_norm": 0.9973340034484863,
+      "learning_rate": 1.14763455297786e-06,
+      "loss": 0.1876,
+      "step": 5387
+    },
+    {
+      "epoch": 2.551287363125185,
+      "grad_norm": 1.2037270069122314,
+      "learning_rate": 1.1452569317388474e-06,
+      "loss": 0.2058,
+      "step": 5388
+    },
+    {
+      "epoch": 2.5517608759988164,
+      "grad_norm": 0.9326767325401306,
+      "learning_rate": 1.1428816263724596e-06,
+      "loss": 0.1998,
+      "step": 5389
+    },
+    {
+      "epoch": 2.5522343888724475,
+      "grad_norm": 1.0340579748153687,
+      "learning_rate": 1.1405086374999386e-06,
+      "loss": 0.1951,
+      "step": 5390
+    },
+    {
+      "epoch": 2.5527079017460785,
+      "grad_norm": 1.1469632387161255,
+      "learning_rate": 1.1381379657419112e-06,
+      "loss": 0.2013,
+      "step": 5391
+    },
+    {
+      "epoch": 2.55318141461971,
+      "grad_norm": 1.0741297006607056,
+      "learning_rate": 1.1357696117184103e-06,
+      "loss": 0.183,
+      "step": 5392
+    },
+    {
+      "epoch": 2.553654927493341,
+      "grad_norm": 1.3129887580871582,
+      "learning_rate": 1.1334035760488493e-06,
+      "loss": 0.2213,
+      "step": 5393
+    },
+    {
+      "epoch": 2.5541284403669726,
+      "grad_norm": 1.0129112005233765,
+      "learning_rate": 1.1310398593520488e-06,
+      "loss": 0.1879,
+      "step": 5394
+    },
+    {
+      "epoch": 2.5546019532406037,
+      "grad_norm": 0.8904241323471069,
+      "learning_rate": 1.128678462246212e-06,
+      "loss": 0.2156,
+      "step": 5395
+    },
+    {
+      "epoch": 2.5550754661142348,
+      "grad_norm": 1.214227557182312,
+      "learning_rate": 1.1263193853489384e-06,
+      "loss": 0.2161,
+      "step": 5396
+    },
+    {
+      "epoch": 2.5555489789878663,
+      "grad_norm": 1.472237229347229,
+      "learning_rate": 1.1239626292772254e-06,
+      "loss": 0.2028,
+      "step": 5397
+    },
+    {
+      "epoch": 2.5560224918614973,
+      "grad_norm": 1.0698444843292236,
+      "learning_rate": 1.1216081946474566e-06,
+      "loss": 0.2018,
+      "step": 5398
+    },
+    {
+      "epoch": 2.556496004735129,
+      "grad_norm": 1.198815107345581,
+      "learning_rate": 1.1192560820754151e-06,
+      "loss": 0.2144,
+      "step": 5399
+    },
+    {
+      "epoch": 2.55696951760876,
+      "grad_norm": 2.1452925205230713,
+      "learning_rate": 1.1169062921762686e-06,
+      "loss": 0.2089,
+      "step": 5400
+    },
+    {
+      "epoch": 2.557443030482391,
+      "grad_norm": 1.0623699426651,
+      "learning_rate": 1.1145588255645868e-06,
+      "loss": 0.2058,
+      "step": 5401
+    },
+    {
+      "epoch": 2.5579165433560225,
+      "grad_norm": 1.0359307527542114,
+      "learning_rate": 1.112213682854323e-06,
+      "loss": 0.2041,
+      "step": 5402
+    },
+    {
+      "epoch": 2.558390056229654,
+      "grad_norm": 1.2944672107696533,
+      "learning_rate": 1.1098708646588308e-06,
+      "loss": 0.211,
+      "step": 5403
+    },
+    {
+      "epoch": 2.558863569103285,
+      "grad_norm": 1.033818006515503,
+      "learning_rate": 1.107530371590847e-06,
+      "loss": 0.1941,
+      "step": 5404
+    },
+    {
+      "epoch": 2.559337081976916,
+      "grad_norm": 1.210862398147583,
+      "learning_rate": 1.1051922042625096e-06,
+      "loss": 0.2018,
+      "step": 5405
+    },
+    {
+      "epoch": 2.5598105948505476,
+      "grad_norm": 0.9782442450523376,
+      "learning_rate": 1.1028563632853407e-06,
+      "loss": 0.197,
+      "step": 5406
+    },
+    {
+      "epoch": 2.5602841077241787,
+      "grad_norm": 1.218672752380371,
+      "learning_rate": 1.1005228492702557e-06,
+      "loss": 0.2224,
+      "step": 5407
+    },
+    {
+      "epoch": 2.56075762059781,
+      "grad_norm": 1.1376310586929321,
+      "learning_rate": 1.0981916628275679e-06,
+      "loss": 0.2009,
+      "step": 5408
+    },
+    {
+      "epoch": 2.5612311334714413,
+      "grad_norm": 1.071939468383789,
+      "learning_rate": 1.0958628045669705e-06,
+      "loss": 0.194,
+      "step": 5409
+    },
+    {
+      "epoch": 2.5617046463450723,
+      "grad_norm": 1.157542109489441,
+      "learning_rate": 1.0935362750975597e-06,
+      "loss": 0.2161,
+      "step": 5410
+    },
+    {
+      "epoch": 2.562178159218704,
+      "grad_norm": 1.7687501907348633,
+      "learning_rate": 1.091212075027811e-06,
+      "loss": 0.1928,
+      "step": 5411
+    },
+    {
+      "epoch": 2.562651672092335,
+      "grad_norm": 1.131037950515747,
+      "learning_rate": 1.0888902049656014e-06,
+      "loss": 0.2192,
+      "step": 5412
+    },
+    {
+      "epoch": 2.5631251849659664,
+      "grad_norm": 1.0506967306137085,
+      "learning_rate": 1.0865706655181907e-06,
+      "loss": 0.1816,
+      "step": 5413
+    },
+    {
+      "epoch": 2.5635986978395975,
+      "grad_norm": 1.2665033340454102,
+      "learning_rate": 1.0842534572922348e-06,
+      "loss": 0.1969,
+      "step": 5414
+    },
+    {
+      "epoch": 2.5640722107132286,
+      "grad_norm": 1.2768964767456055,
+      "learning_rate": 1.0819385808937743e-06,
+      "loss": 0.2108,
+      "step": 5415
+    },
+    {
+      "epoch": 2.56454572358686,
+      "grad_norm": 1.3844908475875854,
+      "learning_rate": 1.0796260369282429e-06,
+      "loss": 0.2119,
+      "step": 5416
+    },
+    {
+      "epoch": 2.565019236460491,
+      "grad_norm": 1.3508644104003906,
+      "learning_rate": 1.0773158260004668e-06,
+      "loss": 0.1835,
+      "step": 5417
+    },
+    {
+      "epoch": 2.5654927493341226,
+      "grad_norm": 1.0464918613433838,
+      "learning_rate": 1.0750079487146558e-06,
+      "loss": 0.2183,
+      "step": 5418
+    },
+    {
+      "epoch": 2.5659662622077537,
+      "grad_norm": 1.4549388885498047,
+      "learning_rate": 1.0727024056744172e-06,
+      "loss": 0.2108,
+      "step": 5419
+    },
+    {
+      "epoch": 2.5664397750813848,
+      "grad_norm": 1.0538500547409058,
+      "learning_rate": 1.0703991974827399e-06,
+      "loss": 0.2007,
+      "step": 5420
+    },
+    {
+      "epoch": 2.5669132879550163,
+      "grad_norm": 1.0848989486694336,
+      "learning_rate": 1.0680983247420062e-06,
+      "loss": 0.2185,
+      "step": 5421
+    },
+    {
+      "epoch": 2.567386800828648,
+      "grad_norm": 0.9970195889472961,
+      "learning_rate": 1.0657997880539894e-06,
+      "loss": 0.2014,
+      "step": 5422
+    },
+    {
+      "epoch": 2.567860313702279,
+      "grad_norm": 1.1892081499099731,
+      "learning_rate": 1.0635035880198474e-06,
+      "loss": 0.1993,
+      "step": 5423
+    },
+    {
+      "epoch": 2.56833382657591,
+      "grad_norm": 1.3775814771652222,
+      "learning_rate": 1.061209725240132e-06,
+      "loss": 0.1976,
+      "step": 5424
+    },
+    {
+      "epoch": 2.5688073394495414,
+      "grad_norm": 1.3159923553466797,
+      "learning_rate": 1.0589182003147758e-06,
+      "loss": 0.2057,
+      "step": 5425
+    },
+    {
+      "epoch": 2.5692808523231725,
+      "grad_norm": 1.09856116771698,
+      "learning_rate": 1.0566290138431113e-06,
+      "loss": 0.1962,
+      "step": 5426
+    },
+    {
+      "epoch": 2.569754365196804,
+      "grad_norm": 1.0620033740997314,
+      "learning_rate": 1.0543421664238473e-06,
+      "loss": 0.2097,
+      "step": 5427
+    },
+    {
+      "epoch": 2.570227878070435,
+      "grad_norm": 1.6413558721542358,
+      "learning_rate": 1.0520576586550923e-06,
+      "loss": 0.2114,
+      "step": 5428
+    },
+    {
+      "epoch": 2.570701390944066,
+      "grad_norm": 1.1933810710906982,
+      "learning_rate": 1.0497754911343316e-06,
+      "loss": 0.2145,
+      "step": 5429
+    },
+    {
+      "epoch": 2.5711749038176976,
+      "grad_norm": 1.7124356031417847,
+      "learning_rate": 1.0474956644584488e-06,
+      "loss": 0.1995,
+      "step": 5430
+    },
+    {
+      "epoch": 2.5716484166913287,
+      "grad_norm": 0.9946605563163757,
+      "learning_rate": 1.0452181792237092e-06,
+      "loss": 0.2177,
+      "step": 5431
+    },
+    {
+      "epoch": 2.57212192956496,
+      "grad_norm": 1.332301139831543,
+      "learning_rate": 1.0429430360257642e-06,
+      "loss": 0.1792,
+      "step": 5432
+    },
+    {
+      "epoch": 2.5725954424385913,
+      "grad_norm": 1.497750163078308,
+      "learning_rate": 1.0406702354596598e-06,
+      "loss": 0.196,
+      "step": 5433
+    },
+    {
+      "epoch": 2.5730689553122224,
+      "grad_norm": 1.034303069114685,
+      "learning_rate": 1.0383997781198218e-06,
+      "loss": 0.1817,
+      "step": 5434
+    },
+    {
+      "epoch": 2.573542468185854,
+      "grad_norm": 0.7837559580802917,
+      "learning_rate": 1.0361316646000686e-06,
+      "loss": 0.1871,
+      "step": 5435
+    },
+    {
+      "epoch": 2.574015981059485,
+      "grad_norm": 1.18108069896698,
+      "learning_rate": 1.0338658954936008e-06,
+      "loss": 0.1877,
+      "step": 5436
+    },
+    {
+      "epoch": 2.5744894939331164,
+      "grad_norm": 0.9967142939567566,
+      "learning_rate": 1.0316024713930129e-06,
+      "loss": 0.1927,
+      "step": 5437
+    },
+    {
+      "epoch": 2.5749630068067475,
+      "grad_norm": 1.073191523551941,
+      "learning_rate": 1.0293413928902761e-06,
+      "loss": 0.2007,
+      "step": 5438
+    },
+    {
+      "epoch": 2.5754365196803786,
+      "grad_norm": 1.0731115341186523,
+      "learning_rate": 1.0270826605767592e-06,
+      "loss": 0.2134,
+      "step": 5439
+    },
+    {
+      "epoch": 2.57591003255401,
+      "grad_norm": 1.0234590768814087,
+      "learning_rate": 1.024826275043209e-06,
+      "loss": 0.1902,
+      "step": 5440
+    },
+    {
+      "epoch": 2.5763835454276416,
+      "grad_norm": 1.0381513833999634,
+      "learning_rate": 1.0225722368797598e-06,
+      "loss": 0.1957,
+      "step": 5441
+    },
+    {
+      "epoch": 2.5768570583012727,
+      "grad_norm": 1.2362014055252075,
+      "learning_rate": 1.020320546675937e-06,
+      "loss": 0.186,
+      "step": 5442
+    },
+    {
+      "epoch": 2.5773305711749037,
+      "grad_norm": 0.9183345437049866,
+      "learning_rate": 1.0180712050206442e-06,
+      "loss": 0.1847,
+      "step": 5443
+    },
+    {
+      "epoch": 2.5778040840485352,
+      "grad_norm": 1.0637094974517822,
+      "learning_rate": 1.01582421250218e-06,
+      "loss": 0.1914,
+      "step": 5444
+    },
+    {
+      "epoch": 2.5782775969221663,
+      "grad_norm": 1.023342490196228,
+      "learning_rate": 1.0135795697082195e-06,
+      "loss": 0.2049,
+      "step": 5445
+    },
+    {
+      "epoch": 2.578751109795798,
+      "grad_norm": 0.9113678336143494,
+      "learning_rate": 1.0113372772258302e-06,
+      "loss": 0.185,
+      "step": 5446
+    },
+    {
+      "epoch": 2.579224622669429,
+      "grad_norm": 1.1002506017684937,
+      "learning_rate": 1.009097335641459e-06,
+      "loss": 0.2028,
+      "step": 5447
+    },
+    {
+      "epoch": 2.57969813554306,
+      "grad_norm": 1.069926142692566,
+      "learning_rate": 1.0068597455409458e-06,
+      "loss": 0.1777,
+      "step": 5448
+    },
+    {
+      "epoch": 2.5801716484166914,
+      "grad_norm": 1.127890944480896,
+      "learning_rate": 1.0046245075095074e-06,
+      "loss": 0.1816,
+      "step": 5449
+    },
+    {
+      "epoch": 2.5806451612903225,
+      "grad_norm": 0.9867877960205078,
+      "learning_rate": 1.0023916221317465e-06,
+      "loss": 0.199,
+      "step": 5450
+    },
+    {
+      "epoch": 2.581118674163954,
+      "grad_norm": 1.3227022886276245,
+      "learning_rate": 1.000161089991658e-06,
+      "loss": 0.196,
+      "step": 5451
+    },
+    {
+      "epoch": 2.581592187037585,
+      "grad_norm": 1.1155279874801636,
+      "learning_rate": 9.979329116726111e-07,
+      "loss": 0.1934,
+      "step": 5452
+    },
+    {
+      "epoch": 2.582065699911216,
+      "grad_norm": 1.0326663255691528,
+      "learning_rate": 9.957070877573682e-07,
+      "loss": 0.1942,
+      "step": 5453
+    },
+    {
+      "epoch": 2.5825392127848477,
+      "grad_norm": 1.2042585611343384,
+      "learning_rate": 9.934836188280693e-07,
+      "loss": 0.189,
+      "step": 5454
+    },
+    {
+      "epoch": 2.5830127256584787,
+      "grad_norm": 1.3072971105575562,
+      "learning_rate": 9.91262505466245e-07,
+      "loss": 0.1837,
+      "step": 5455
+    },
+    {
+      "epoch": 2.5834862385321102,
+      "grad_norm": 1.351815938949585,
+      "learning_rate": 9.890437482528004e-07,
+      "loss": 0.1906,
+      "step": 5456
+    },
+    {
+      "epoch": 2.5839597514057413,
+      "grad_norm": 1.0295909643173218,
+      "learning_rate": 9.868273477680357e-07,
+      "loss": 0.2168,
+      "step": 5457
+    },
+    {
+      "epoch": 2.5844332642793724,
+      "grad_norm": 0.970486044883728,
+      "learning_rate": 9.84613304591625e-07,
+      "loss": 0.1817,
+      "step": 5458
+    },
+    {
+      "epoch": 2.584906777153004,
+      "grad_norm": 1.0597081184387207,
+      "learning_rate": 9.824016193026308e-07,
+      "loss": 0.1955,
+      "step": 5459
+    },
+    {
+      "epoch": 2.585380290026635,
+      "grad_norm": 1.491175889968872,
+      "learning_rate": 9.801922924795004e-07,
+      "loss": 0.2162,
+      "step": 5460
+    },
+    {
+      "epoch": 2.5858538029002665,
+      "grad_norm": 1.3303380012512207,
+      "learning_rate": 9.779853247000593e-07,
+      "loss": 0.1997,
+      "step": 5461
+    },
+    {
+      "epoch": 2.5863273157738975,
+      "grad_norm": 1.9732062816619873,
+      "learning_rate": 9.757807165415213e-07,
+      "loss": 0.2225,
+      "step": 5462
+    },
+    {
+      "epoch": 2.5868008286475286,
+      "grad_norm": 1.0555258989334106,
+      "learning_rate": 9.735784685804773e-07,
+      "loss": 0.1806,
+      "step": 5463
+    },
+    {
+      "epoch": 2.58727434152116,
+      "grad_norm": 1.026889681816101,
+      "learning_rate": 9.713785813929056e-07,
+      "loss": 0.2107,
+      "step": 5464
+    },
+    {
+      "epoch": 2.5877478543947916,
+      "grad_norm": 1.0289431810379028,
+      "learning_rate": 9.69181055554167e-07,
+      "loss": 0.1958,
+      "step": 5465
+    },
+    {
+      "epoch": 2.5882213672684227,
+      "grad_norm": 0.9716818332672119,
+      "learning_rate": 9.669858916389985e-07,
+      "loss": 0.1865,
+      "step": 5466
+    },
+    {
+      "epoch": 2.5886948801420537,
+      "grad_norm": 1.6020885705947876,
+      "learning_rate": 9.647930902215296e-07,
+      "loss": 0.1944,
+      "step": 5467
+    },
+    {
+      "epoch": 2.5891683930156852,
+      "grad_norm": 1.1589189767837524,
+      "learning_rate": 9.626026518752619e-07,
+      "loss": 0.1926,
+      "step": 5468
+    },
+    {
+      "epoch": 2.5896419058893163,
+      "grad_norm": 1.059776782989502,
+      "learning_rate": 9.604145771730865e-07,
+      "loss": 0.1716,
+      "step": 5469
+    },
+    {
+      "epoch": 2.590115418762948,
+      "grad_norm": 1.3090096712112427,
+      "learning_rate": 9.582288666872708e-07,
+      "loss": 0.2115,
+      "step": 5470
+    },
+    {
+      "epoch": 2.590588931636579,
+      "grad_norm": 1.174440860748291,
+      "learning_rate": 9.560455209894691e-07,
+      "loss": 0.1893,
+      "step": 5471
+    },
+    {
+      "epoch": 2.59106244451021,
+      "grad_norm": 1.4484925270080566,
+      "learning_rate": 9.538645406507108e-07,
+      "loss": 0.2198,
+      "step": 5472
+    },
+    {
+      "epoch": 2.5915359573838415,
+      "grad_norm": 1.0521667003631592,
+      "learning_rate": 9.516859262414147e-07,
+      "loss": 0.1962,
+      "step": 5473
+    },
+    {
+      "epoch": 2.5920094702574725,
+      "grad_norm": 1.0996438264846802,
+      "learning_rate": 9.495096783313729e-07,
+      "loss": 0.2061,
+      "step": 5474
+    },
+    {
+      "epoch": 2.592482983131104,
+      "grad_norm": 1.0292750597000122,
+      "learning_rate": 9.473357974897623e-07,
+      "loss": 0.1941,
+      "step": 5475
+    },
+    {
+      "epoch": 2.592956496004735,
+      "grad_norm": 1.131027340888977,
+      "learning_rate": 9.451642842851427e-07,
+      "loss": 0.1894,
+      "step": 5476
+    },
+    {
+      "epoch": 2.593430008878366,
+      "grad_norm": 1.0126436948776245,
+      "learning_rate": 9.429951392854486e-07,
+      "loss": 0.1936,
+      "step": 5477
+    },
+    {
+      "epoch": 2.5939035217519977,
+      "grad_norm": 1.0096503496170044,
+      "learning_rate": 9.40828363058004e-07,
+      "loss": 0.1973,
+      "step": 5478
+    },
+    {
+      "epoch": 2.5943770346256287,
+      "grad_norm": 1.16969895362854,
+      "learning_rate": 9.386639561695043e-07,
+      "loss": 0.2078,
+      "step": 5479
+    },
+    {
+      "epoch": 2.5948505474992603,
+      "grad_norm": 1.0769869089126587,
+      "learning_rate": 9.365019191860314e-07,
+      "loss": 0.2057,
+      "step": 5480
+    },
+    {
+      "epoch": 2.5953240603728913,
+      "grad_norm": 1.0034693479537964,
+      "learning_rate": 9.343422526730428e-07,
+      "loss": 0.199,
+      "step": 5481
+    },
+    {
+      "epoch": 2.5957975732465224,
+      "grad_norm": 1.0947496891021729,
+      "learning_rate": 9.321849571953822e-07,
+      "loss": 0.1851,
+      "step": 5482
+    },
+    {
+      "epoch": 2.596271086120154,
+      "grad_norm": 1.386936068534851,
+      "learning_rate": 9.300300333172652e-07,
+      "loss": 0.2034,
+      "step": 5483
+    },
+    {
+      "epoch": 2.5967445989937854,
+      "grad_norm": 1.0438120365142822,
+      "learning_rate": 9.27877481602295e-07,
+      "loss": 0.2129,
+      "step": 5484
+    },
+    {
+      "epoch": 2.5972181118674165,
+      "grad_norm": 1.0546690225601196,
+      "learning_rate": 9.257273026134494e-07,
+      "loss": 0.2108,
+      "step": 5485
+    },
+    {
+      "epoch": 2.5976916247410475,
+      "grad_norm": 1.3479506969451904,
+      "learning_rate": 9.235794969130851e-07,
+      "loss": 0.184,
+      "step": 5486
+    },
+    {
+      "epoch": 2.598165137614679,
+      "grad_norm": 1.1528756618499756,
+      "learning_rate": 9.214340650629439e-07,
+      "loss": 0.1933,
+      "step": 5487
+    },
+    {
+      "epoch": 2.59863865048831,
+      "grad_norm": 1.3162215948104858,
+      "learning_rate": 9.192910076241379e-07,
+      "loss": 0.1848,
+      "step": 5488
+    },
+    {
+      "epoch": 2.5991121633619416,
+      "grad_norm": 1.3679094314575195,
+      "learning_rate": 9.171503251571678e-07,
+      "loss": 0.2207,
+      "step": 5489
+    },
+    {
+      "epoch": 2.5995856762355727,
+      "grad_norm": 1.234590768814087,
+      "learning_rate": 9.150120182219046e-07,
+      "loss": 0.1839,
+      "step": 5490
+    },
+    {
+      "epoch": 2.6000591891092037,
+      "grad_norm": 1.105852484703064,
+      "learning_rate": 9.128760873776054e-07,
+      "loss": 0.2014,
+      "step": 5491
+    },
+    {
+      "epoch": 2.6005327019828353,
+      "grad_norm": 1.1004148721694946,
+      "learning_rate": 9.107425331828989e-07,
+      "loss": 0.1932,
+      "step": 5492
+    },
+    {
+      "epoch": 2.6010062148564663,
+      "grad_norm": 0.9159629940986633,
+      "learning_rate": 9.086113561957987e-07,
+      "loss": 0.2114,
+      "step": 5493
+    },
+    {
+      "epoch": 2.601479727730098,
+      "grad_norm": 1.061231255531311,
+      "learning_rate": 9.064825569736924e-07,
+      "loss": 0.1762,
+      "step": 5494
+    },
+    {
+      "epoch": 2.601953240603729,
+      "grad_norm": 1.741815209388733,
+      "learning_rate": 9.043561360733444e-07,
+      "loss": 0.187,
+      "step": 5495
+    },
+    {
+      "epoch": 2.60242675347736,
+      "grad_norm": 1.0610990524291992,
+      "learning_rate": 9.022320940509033e-07,
+      "loss": 0.1988,
+      "step": 5496
+    },
+    {
+      "epoch": 2.6029002663509915,
+      "grad_norm": 1.0070830583572388,
+      "learning_rate": 9.001104314618892e-07,
+      "loss": 0.2234,
+      "step": 5497
+    },
+    {
+      "epoch": 2.6033737792246225,
+      "grad_norm": 0.9660130143165588,
+      "learning_rate": 8.979911488612037e-07,
+      "loss": 0.2007,
+      "step": 5498
+    },
+    {
+      "epoch": 2.603847292098254,
+      "grad_norm": 1.0970488786697388,
+      "learning_rate": 8.958742468031257e-07,
+      "loss": 0.2136,
+      "step": 5499
+    },
+    {
+      "epoch": 2.604320804971885,
+      "grad_norm": 0.9795224666595459,
+      "learning_rate": 8.937597258413078e-07,
+      "loss": 0.2162,
+      "step": 5500
+    },
+    {
+      "epoch": 2.604794317845516,
+      "grad_norm": 1.2171189785003662,
+      "learning_rate": 8.916475865287855e-07,
+      "loss": 0.2051,
+      "step": 5501
+    },
+    {
+      "epoch": 2.6052678307191477,
+      "grad_norm": 1.032201886177063,
+      "learning_rate": 8.895378294179658e-07,
+      "loss": 0.1856,
+      "step": 5502
+    },
+    {
+      "epoch": 2.605741343592779,
+      "grad_norm": 1.1588233709335327,
+      "learning_rate": 8.874304550606383e-07,
+      "loss": 0.2028,
+      "step": 5503
+    },
+    {
+      "epoch": 2.6062148564664103,
+      "grad_norm": 0.9060316681861877,
+      "learning_rate": 8.853254640079633e-07,
+      "loss": 0.1907,
+      "step": 5504
+    },
+    {
+      "epoch": 2.6066883693400413,
+      "grad_norm": 1.5966558456420898,
+      "learning_rate": 8.832228568104839e-07,
+      "loss": 0.2036,
+      "step": 5505
+    },
+    {
+      "epoch": 2.607161882213673,
+      "grad_norm": 1.2059063911437988,
+      "learning_rate": 8.811226340181133e-07,
+      "loss": 0.2239,
+      "step": 5506
+    },
+    {
+      "epoch": 2.607635395087304,
+      "grad_norm": 1.165710210800171,
+      "learning_rate": 8.79024796180149e-07,
+      "loss": 0.2087,
+      "step": 5507
+    },
+    {
+      "epoch": 2.6081089079609354,
+      "grad_norm": 1.2334532737731934,
+      "learning_rate": 8.769293438452553e-07,
+      "loss": 0.203,
+      "step": 5508
+    },
+    {
+      "epoch": 2.6085824208345665,
+      "grad_norm": 1.1886234283447266,
+      "learning_rate": 8.748362775614816e-07,
+      "loss": 0.2152,
+      "step": 5509
+    },
+    {
+      "epoch": 2.6090559337081975,
+      "grad_norm": 1.4193661212921143,
+      "learning_rate": 8.727455978762478e-07,
+      "loss": 0.1982,
+      "step": 5510
+    },
+    {
+      "epoch": 2.609529446581829,
+      "grad_norm": 1.0052508115768433,
+      "learning_rate": 8.706573053363487e-07,
+      "loss": 0.2157,
+      "step": 5511
+    },
+    {
+      "epoch": 2.61000295945546,
+      "grad_norm": 1.1085777282714844,
+      "learning_rate": 8.685714004879608e-07,
+      "loss": 0.1973,
+      "step": 5512
+    },
+    {
+      "epoch": 2.6104764723290916,
+      "grad_norm": 1.2764101028442383,
+      "learning_rate": 8.664878838766289e-07,
+      "loss": 0.1935,
+      "step": 5513
+    },
+    {
+      "epoch": 2.6109499852027227,
+      "grad_norm": 0.94977867603302,
+      "learning_rate": 8.644067560472802e-07,
+      "loss": 0.1907,
+      "step": 5514
+    },
+    {
+      "epoch": 2.6114234980763538,
+      "grad_norm": 0.928570032119751,
+      "learning_rate": 8.623280175442094e-07,
+      "loss": 0.1929,
+      "step": 5515
+    },
+    {
+      "epoch": 2.6118970109499853,
+      "grad_norm": 1.3341639041900635,
+      "learning_rate": 8.602516689110952e-07,
+      "loss": 0.1987,
+      "step": 5516
+    },
+    {
+      "epoch": 2.6123705238236163,
+      "grad_norm": 1.2463881969451904,
+      "learning_rate": 8.581777106909827e-07,
+      "loss": 0.1788,
+      "step": 5517
+    },
+    {
+      "epoch": 2.612844036697248,
+      "grad_norm": 1.0751234292984009,
+      "learning_rate": 8.561061434262996e-07,
+      "loss": 0.2133,
+      "step": 5518
+    },
+    {
+      "epoch": 2.613317549570879,
+      "grad_norm": 0.9469509124755859,
+      "learning_rate": 8.540369676588411e-07,
+      "loss": 0.1925,
+      "step": 5519
+    },
+    {
+      "epoch": 2.61379106244451,
+      "grad_norm": 1.0061715841293335,
+      "learning_rate": 8.51970183929779e-07,
+      "loss": 0.2128,
+      "step": 5520
+    },
+    {
+      "epoch": 2.6142645753181415,
+      "grad_norm": 1.019024133682251,
+      "learning_rate": 8.49905792779665e-07,
+      "loss": 0.2089,
+      "step": 5521
+    },
+    {
+      "epoch": 2.6147380881917726,
+      "grad_norm": 2.2128031253814697,
+      "learning_rate": 8.478437947484164e-07,
+      "loss": 0.2122,
+      "step": 5522
+    },
+    {
+      "epoch": 2.615211601065404,
+      "grad_norm": 1.6629959344863892,
+      "learning_rate": 8.457841903753327e-07,
+      "loss": 0.2341,
+      "step": 5523
+    },
+    {
+      "epoch": 2.615685113939035,
+      "grad_norm": 1.15640127658844,
+      "learning_rate": 8.4372698019908e-07,
+      "loss": 0.2024,
+      "step": 5524
+    },
+    {
+      "epoch": 2.616158626812666,
+      "grad_norm": 0.9604775905609131,
+      "learning_rate": 8.416721647577053e-07,
+      "loss": 0.1944,
+      "step": 5525
+    },
+    {
+      "epoch": 2.6166321396862977,
+      "grad_norm": 0.9418638348579407,
+      "learning_rate": 8.396197445886223e-07,
+      "loss": 0.1688,
+      "step": 5526
+    },
+    {
+      "epoch": 2.617105652559929,
+      "grad_norm": 1.1123701333999634,
+      "learning_rate": 8.375697202286248e-07,
+      "loss": 0.2019,
+      "step": 5527
+    },
+    {
+      "epoch": 2.6175791654335603,
+      "grad_norm": 1.4473880529403687,
+      "learning_rate": 8.355220922138762e-07,
+      "loss": 0.2034,
+      "step": 5528
+    },
+    {
+      "epoch": 2.6180526783071913,
+      "grad_norm": 1.3960907459259033,
+      "learning_rate": 8.334768610799104e-07,
+      "loss": 0.1939,
+      "step": 5529
+    },
+    {
+      "epoch": 2.618526191180823,
+      "grad_norm": 0.9807230830192566,
+      "learning_rate": 8.314340273616428e-07,
+      "loss": 0.2027,
+      "step": 5530
+    },
+    {
+      "epoch": 2.618999704054454,
+      "grad_norm": 1.4325486421585083,
+      "learning_rate": 8.293935915933526e-07,
+      "loss": 0.2103,
+      "step": 5531
+    },
+    {
+      "epoch": 2.6194732169280854,
+      "grad_norm": 1.228115200996399,
+      "learning_rate": 8.273555543087009e-07,
+      "loss": 0.1993,
+      "step": 5532
+    },
+    {
+      "epoch": 2.6199467298017165,
+      "grad_norm": 1.0208629369735718,
+      "learning_rate": 8.25319916040711e-07,
+      "loss": 0.1961,
+      "step": 5533
+    },
+    {
+      "epoch": 2.6204202426753476,
+      "grad_norm": 1.7419202327728271,
+      "learning_rate": 8.232866773217896e-07,
+      "loss": 0.2002,
+      "step": 5534
+    },
+    {
+      "epoch": 2.620893755548979,
+      "grad_norm": 1.0411862134933472,
+      "learning_rate": 8.212558386837067e-07,
+      "loss": 0.1958,
+      "step": 5535
+    },
+    {
+      "epoch": 2.62136726842261,
+      "grad_norm": 1.4005458354949951,
+      "learning_rate": 8.192274006576095e-07,
+      "loss": 0.2136,
+      "step": 5536
+    },
+    {
+      "epoch": 2.6218407812962417,
+      "grad_norm": 1.0770565271377563,
+      "learning_rate": 8.172013637740195e-07,
+      "loss": 0.2002,
+      "step": 5537
+    },
+    {
+      "epoch": 2.6223142941698727,
+      "grad_norm": 1.2719935178756714,
+      "learning_rate": 8.151777285628226e-07,
+      "loss": 0.2034,
+      "step": 5538
+    },
+    {
+      "epoch": 2.622787807043504,
+      "grad_norm": 1.5015056133270264,
+      "learning_rate": 8.131564955532856e-07,
+      "loss": 0.1889,
+      "step": 5539
+    },
+    {
+      "epoch": 2.6232613199171353,
+      "grad_norm": 1.192428469657898,
+      "learning_rate": 8.111376652740388e-07,
+      "loss": 0.1872,
+      "step": 5540
+    },
+    {
+      "epoch": 2.6237348327907664,
+      "grad_norm": 1.306041955947876,
+      "learning_rate": 8.091212382530899e-07,
+      "loss": 0.2135,
+      "step": 5541
+    },
+    {
+      "epoch": 2.624208345664398,
+      "grad_norm": 1.4788362979888916,
+      "learning_rate": 8.071072150178138e-07,
+      "loss": 0.2017,
+      "step": 5542
+    },
+    {
+      "epoch": 2.624681858538029,
+      "grad_norm": 1.2858426570892334,
+      "learning_rate": 8.050955960949625e-07,
+      "loss": 0.2031,
+      "step": 5543
+    },
+    {
+      "epoch": 2.62515537141166,
+      "grad_norm": 1.2146573066711426,
+      "learning_rate": 8.030863820106527e-07,
+      "loss": 0.1757,
+      "step": 5544
+    },
+    {
+      "epoch": 2.6256288842852915,
+      "grad_norm": 1.650189995765686,
+      "learning_rate": 8.010795732903731e-07,
+      "loss": 0.2143,
+      "step": 5545
+    },
+    {
+      "epoch": 2.626102397158923,
+      "grad_norm": 1.076585292816162,
+      "learning_rate": 7.990751704589906e-07,
+      "loss": 0.1987,
+      "step": 5546
+    },
+    {
+      "epoch": 2.626575910032554,
+      "grad_norm": 1.0550071001052856,
+      "learning_rate": 7.970731740407311e-07,
+      "loss": 0.1886,
+      "step": 5547
+    },
+    {
+      "epoch": 2.627049422906185,
+      "grad_norm": 1.0130494832992554,
+      "learning_rate": 7.950735845592039e-07,
+      "loss": 0.2068,
+      "step": 5548
+    },
+    {
+      "epoch": 2.6275229357798167,
+      "grad_norm": 1.2154542207717896,
+      "learning_rate": 7.93076402537376e-07,
+      "loss": 0.2018,
+      "step": 5549
+    },
+    {
+      "epoch": 2.6279964486534477,
+      "grad_norm": 0.8742989897727966,
+      "learning_rate": 7.910816284975975e-07,
+      "loss": 0.1784,
+      "step": 5550
+    },
+    {
+      "epoch": 2.6284699615270792,
+      "grad_norm": 1.0601682662963867,
+      "learning_rate": 7.890892629615765e-07,
+      "loss": 0.1972,
+      "step": 5551
+    },
+    {
+      "epoch": 2.6289434744007103,
+      "grad_norm": 1.270751953125,
+      "learning_rate": 7.870993064504018e-07,
+      "loss": 0.2142,
+      "step": 5552
+    },
+    {
+      "epoch": 2.6294169872743414,
+      "grad_norm": 1.1302870512008667,
+      "learning_rate": 7.851117594845237e-07,
+      "loss": 0.2138,
+      "step": 5553
+    },
+    {
+      "epoch": 2.629890500147973,
+      "grad_norm": 1.3007394075393677,
+      "learning_rate": 7.831266225837675e-07,
+      "loss": 0.1978,
+      "step": 5554
+    },
+    {
+      "epoch": 2.630364013021604,
+      "grad_norm": 0.9886184930801392,
+      "learning_rate": 7.811438962673268e-07,
+      "loss": 0.1897,
+      "step": 5555
+    },
+    {
+      "epoch": 2.6308375258952355,
+      "grad_norm": 1.188515305519104,
+      "learning_rate": 7.791635810537624e-07,
+      "loss": 0.2089,
+      "step": 5556
+    },
+    {
+      "epoch": 2.6313110387688665,
+      "grad_norm": 1.18173086643219,
+      "learning_rate": 7.771856774610109e-07,
+      "loss": 0.2251,
+      "step": 5557
+    },
+    {
+      "epoch": 2.6317845516424976,
+      "grad_norm": 1.0726786851882935,
+      "learning_rate": 7.752101860063687e-07,
+      "loss": 0.2004,
+      "step": 5558
+    },
+    {
+      "epoch": 2.632258064516129,
+      "grad_norm": 1.0728563070297241,
+      "learning_rate": 7.732371072065126e-07,
+      "loss": 0.1939,
+      "step": 5559
+    },
+    {
+      "epoch": 2.63273157738976,
+      "grad_norm": 1.0991253852844238,
+      "learning_rate": 7.712664415774762e-07,
+      "loss": 0.1699,
+      "step": 5560
+    },
+    {
+      "epoch": 2.6332050902633917,
+      "grad_norm": 0.9829975962638855,
+      "learning_rate": 7.692981896346718e-07,
+      "loss": 0.199,
+      "step": 5561
+    },
+    {
+      "epoch": 2.6336786031370227,
+      "grad_norm": 0.8773795366287231,
+      "learning_rate": 7.673323518928755e-07,
+      "loss": 0.1894,
+      "step": 5562
+    },
+    {
+      "epoch": 2.634152116010654,
+      "grad_norm": 1.0769257545471191,
+      "learning_rate": 7.653689288662335e-07,
+      "loss": 0.2144,
+      "step": 5563
+    },
+    {
+      "epoch": 2.6346256288842853,
+      "grad_norm": 1.1146119832992554,
+      "learning_rate": 7.63407921068261e-07,
+      "loss": 0.2237,
+      "step": 5564
+    },
+    {
+      "epoch": 2.635099141757917,
+      "grad_norm": 0.9974523782730103,
+      "learning_rate": 7.614493290118386e-07,
+      "loss": 0.2062,
+      "step": 5565
+    },
+    {
+      "epoch": 2.635572654631548,
+      "grad_norm": 1.053133249282837,
+      "learning_rate": 7.594931532092198e-07,
+      "loss": 0.2005,
+      "step": 5566
+    },
+    {
+      "epoch": 2.636046167505179,
+      "grad_norm": 1.3148210048675537,
+      "learning_rate": 7.575393941720199e-07,
+      "loss": 0.2097,
+      "step": 5567
+    },
+    {
+      "epoch": 2.6365196803788105,
+      "grad_norm": 1.292978286743164,
+      "learning_rate": 7.555880524112291e-07,
+      "loss": 0.1951,
+      "step": 5568
+    },
+    {
+      "epoch": 2.6369931932524415,
+      "grad_norm": 1.3344711065292358,
+      "learning_rate": 7.536391284372002e-07,
+      "loss": 0.1978,
+      "step": 5569
+    },
+    {
+      "epoch": 2.637466706126073,
+      "grad_norm": 1.8182041645050049,
+      "learning_rate": 7.516926227596566e-07,
+      "loss": 0.2056,
+      "step": 5570
+    },
+    {
+      "epoch": 2.637940218999704,
+      "grad_norm": 1.400888204574585,
+      "learning_rate": 7.497485358876866e-07,
+      "loss": 0.2021,
+      "step": 5571
+    },
+    {
+      "epoch": 2.638413731873335,
+      "grad_norm": 0.8689830303192139,
+      "learning_rate": 7.478068683297501e-07,
+      "loss": 0.1753,
+      "step": 5572
+    },
+    {
+      "epoch": 2.6388872447469667,
+      "grad_norm": 1.108304738998413,
+      "learning_rate": 7.458676205936688e-07,
+      "loss": 0.1851,
+      "step": 5573
+    },
+    {
+      "epoch": 2.6393607576205977,
+      "grad_norm": 1.0863293409347534,
+      "learning_rate": 7.439307931866346e-07,
+      "loss": 0.2266,
+      "step": 5574
+    },
+    {
+      "epoch": 2.6398342704942293,
+      "grad_norm": 1.085361361503601,
+      "learning_rate": 7.419963866152058e-07,
+      "loss": 0.2166,
+      "step": 5575
+    },
+    {
+      "epoch": 2.6403077833678603,
+      "grad_norm": 1.4498753547668457,
+      "learning_rate": 7.400644013853087e-07,
+      "loss": 0.1799,
+      "step": 5576
+    },
+    {
+      "epoch": 2.6407812962414914,
+      "grad_norm": 1.1030598878860474,
+      "learning_rate": 7.381348380022368e-07,
+      "loss": 0.1678,
+      "step": 5577
+    },
+    {
+      "epoch": 2.641254809115123,
+      "grad_norm": 1.5014503002166748,
+      "learning_rate": 7.362076969706478e-07,
+      "loss": 0.2074,
+      "step": 5578
+    },
+    {
+      "epoch": 2.641728321988754,
+      "grad_norm": 1.086707353591919,
+      "learning_rate": 7.342829787945638e-07,
+      "loss": 0.2029,
+      "step": 5579
+    },
+    {
+      "epoch": 2.6422018348623855,
+      "grad_norm": 1.4167083501815796,
+      "learning_rate": 7.323606839773811e-07,
+      "loss": 0.1902,
+      "step": 5580
+    },
+    {
+      "epoch": 2.6426753477360165,
+      "grad_norm": 0.9145298600196838,
+      "learning_rate": 7.304408130218532e-07,
+      "loss": 0.2039,
+      "step": 5581
+    },
+    {
+      "epoch": 2.6431488606096476,
+      "grad_norm": 2.068638563156128,
+      "learning_rate": 7.285233664301073e-07,
+      "loss": 0.2075,
+      "step": 5582
+    },
+    {
+      "epoch": 2.643622373483279,
+      "grad_norm": 1.0407607555389404,
+      "learning_rate": 7.266083447036287e-07,
+      "loss": 0.1838,
+      "step": 5583
+    },
+    {
+      "epoch": 2.64409588635691,
+      "grad_norm": 1.2109540700912476,
+      "learning_rate": 7.246957483432782e-07,
+      "loss": 0.2107,
+      "step": 5584
+    },
+    {
+      "epoch": 2.6445693992305417,
+      "grad_norm": 1.51192045211792,
+      "learning_rate": 7.227855778492732e-07,
+      "loss": 0.1798,
+      "step": 5585
+    },
+    {
+      "epoch": 2.6450429121041727,
+      "grad_norm": 1.015613317489624,
+      "learning_rate": 7.208778337212019e-07,
+      "loss": 0.2107,
+      "step": 5586
+    },
+    {
+      "epoch": 2.645516424977804,
+      "grad_norm": 1.1152918338775635,
+      "learning_rate": 7.189725164580152e-07,
+      "loss": 0.2313,
+      "step": 5587
+    },
+    {
+      "epoch": 2.6459899378514353,
+      "grad_norm": 1.4424123764038086,
+      "learning_rate": 7.170696265580323e-07,
+      "loss": 0.1907,
+      "step": 5588
+    },
+    {
+      "epoch": 2.646463450725067,
+      "grad_norm": 1.2000633478164673,
+      "learning_rate": 7.15169164518934e-07,
+      "loss": 0.1903,
+      "step": 5589
+    },
+    {
+      "epoch": 2.646936963598698,
+      "grad_norm": 0.9677756428718567,
+      "learning_rate": 7.132711308377682e-07,
+      "loss": 0.2042,
+      "step": 5590
+    },
+    {
+      "epoch": 2.647410476472329,
+      "grad_norm": 1.519376516342163,
+      "learning_rate": 7.113755260109478e-07,
+      "loss": 0.1914,
+      "step": 5591
+    },
+    {
+      "epoch": 2.6478839893459605,
+      "grad_norm": 0.9879403114318848,
+      "learning_rate": 7.094823505342485e-07,
+      "loss": 0.2147,
+      "step": 5592
+    },
+    {
+      "epoch": 2.6483575022195915,
+      "grad_norm": 1.1486883163452148,
+      "learning_rate": 7.075916049028142e-07,
+      "loss": 0.2034,
+      "step": 5593
+    },
+    {
+      "epoch": 2.648831015093223,
+      "grad_norm": 0.992047905921936,
+      "learning_rate": 7.057032896111494e-07,
+      "loss": 0.1918,
+      "step": 5594
+    },
+    {
+      "epoch": 2.649304527966854,
+      "grad_norm": 1.395175576210022,
+      "learning_rate": 7.038174051531266e-07,
+      "loss": 0.1973,
+      "step": 5595
+    },
+    {
+      "epoch": 2.649778040840485,
+      "grad_norm": 0.9354866743087769,
+      "learning_rate": 7.019339520219793e-07,
+      "loss": 0.1899,
+      "step": 5596
+    },
+    {
+      "epoch": 2.6502515537141167,
+      "grad_norm": 1.0331776142120361,
+      "learning_rate": 7.000529307103066e-07,
+      "loss": 0.1847,
+      "step": 5597
+    },
+    {
+      "epoch": 2.6507250665877478,
+      "grad_norm": 1.2936712503433228,
+      "learning_rate": 6.981743417100728e-07,
+      "loss": 0.1875,
+      "step": 5598
+    },
+    {
+      "epoch": 2.6511985794613793,
+      "grad_norm": 1.1697713136672974,
+      "learning_rate": 6.962981855126017e-07,
+      "loss": 0.1885,
+      "step": 5599
+    },
+    {
+      "epoch": 2.6516720923350103,
+      "grad_norm": 1.2552714347839355,
+      "learning_rate": 6.944244626085872e-07,
+      "loss": 0.2045,
+      "step": 5600
+    },
+    {
+      "epoch": 2.6521456052086414,
+      "grad_norm": 1.1409389972686768,
+      "learning_rate": 6.925531734880808e-07,
+      "loss": 0.2088,
+      "step": 5601
+    },
+    {
+      "epoch": 2.652619118082273,
+      "grad_norm": 1.1405903100967407,
+      "learning_rate": 6.906843186405032e-07,
+      "loss": 0.2132,
+      "step": 5602
+    },
+    {
+      "epoch": 2.653092630955904,
+      "grad_norm": 1.2994462251663208,
+      "learning_rate": 6.888178985546312e-07,
+      "loss": 0.201,
+      "step": 5603
+    },
+    {
+      "epoch": 2.6535661438295355,
+      "grad_norm": 1.3358666896820068,
+      "learning_rate": 6.869539137186132e-07,
+      "loss": 0.208,
+      "step": 5604
+    },
+    {
+      "epoch": 2.6540396567031665,
+      "grad_norm": 1.0983041524887085,
+      "learning_rate": 6.850923646199526e-07,
+      "loss": 0.2056,
+      "step": 5605
+    },
+    {
+      "epoch": 2.6545131695767976,
+      "grad_norm": 1.1728966236114502,
+      "learning_rate": 6.832332517455242e-07,
+      "loss": 0.2267,
+      "step": 5606
+    },
+    {
+      "epoch": 2.654986682450429,
+      "grad_norm": 1.0750831365585327,
+      "learning_rate": 6.813765755815571e-07,
+      "loss": 0.204,
+      "step": 5607
+    },
+    {
+      "epoch": 2.6554601953240606,
+      "grad_norm": 1.2680811882019043,
+      "learning_rate": 6.795223366136471e-07,
+      "loss": 0.1889,
+      "step": 5608
+    },
+    {
+      "epoch": 2.6559337081976917,
+      "grad_norm": 1.0505774021148682,
+      "learning_rate": 6.776705353267554e-07,
+      "loss": 0.1937,
+      "step": 5609
+    },
+    {
+      "epoch": 2.6564072210713228,
+      "grad_norm": 1.042385220527649,
+      "learning_rate": 6.758211722052e-07,
+      "loss": 0.2031,
+      "step": 5610
+    },
+    {
+      "epoch": 2.6568807339449543,
+      "grad_norm": 1.3553048372268677,
+      "learning_rate": 6.73974247732666e-07,
+      "loss": 0.2098,
+      "step": 5611
+    },
+    {
+      "epoch": 2.6573542468185853,
+      "grad_norm": 1.1963160037994385,
+      "learning_rate": 6.721297623921963e-07,
+      "loss": 0.1899,
+      "step": 5612
+    },
+    {
+      "epoch": 2.657827759692217,
+      "grad_norm": 1.010690450668335,
+      "learning_rate": 6.702877166662014e-07,
+      "loss": 0.2022,
+      "step": 5613
+    },
+    {
+      "epoch": 2.658301272565848,
+      "grad_norm": 1.4362280368804932,
+      "learning_rate": 6.684481110364471e-07,
+      "loss": 0.1826,
+      "step": 5614
+    },
+    {
+      "epoch": 2.658774785439479,
+      "grad_norm": 1.1894739866256714,
+      "learning_rate": 6.666109459840664e-07,
+      "loss": 0.1979,
+      "step": 5615
+    },
+    {
+      "epoch": 2.6592482983131105,
+      "grad_norm": 1.0000640153884888,
+      "learning_rate": 6.647762219895526e-07,
+      "loss": 0.1917,
+      "step": 5616
+    },
+    {
+      "epoch": 2.6597218111867416,
+      "grad_norm": 1.6517083644866943,
+      "learning_rate": 6.629439395327597e-07,
+      "loss": 0.1963,
+      "step": 5617
+    },
+    {
+      "epoch": 2.660195324060373,
+      "grad_norm": 1.1130201816558838,
+      "learning_rate": 6.611140990929032e-07,
+      "loss": 0.1901,
+      "step": 5618
+    },
+    {
+      "epoch": 2.660668836934004,
+      "grad_norm": 1.1720373630523682,
+      "learning_rate": 6.592867011485593e-07,
+      "loss": 0.1891,
+      "step": 5619
+    },
+    {
+      "epoch": 2.661142349807635,
+      "grad_norm": 1.1503329277038574,
+      "learning_rate": 6.574617461776689e-07,
+      "loss": 0.211,
+      "step": 5620
+    },
+    {
+      "epoch": 2.6616158626812667,
+      "grad_norm": 1.2841466665267944,
+      "learning_rate": 6.55639234657528e-07,
+      "loss": 0.192,
+      "step": 5621
+    },
+    {
+      "epoch": 2.6620893755548978,
+      "grad_norm": 1.003335952758789,
+      "learning_rate": 6.538191670648008e-07,
+      "loss": 0.1809,
+      "step": 5622
+    },
+    {
+      "epoch": 2.6625628884285293,
+      "grad_norm": 1.1069647073745728,
+      "learning_rate": 6.520015438755056e-07,
+      "loss": 0.2017,
+      "step": 5623
+    },
+    {
+      "epoch": 2.6630364013021603,
+      "grad_norm": 1.0702106952667236,
+      "learning_rate": 6.501863655650243e-07,
+      "loss": 0.2038,
+      "step": 5624
+    },
+    {
+      "epoch": 2.6635099141757914,
+      "grad_norm": 0.9489043951034546,
+      "learning_rate": 6.483736326081003e-07,
+      "loss": 0.1976,
+      "step": 5625
+    },
+    {
+      "epoch": 2.663983427049423,
+      "grad_norm": 1.254457712173462,
+      "learning_rate": 6.465633454788345e-07,
+      "loss": 0.2198,
+      "step": 5626
+    },
+    {
+      "epoch": 2.6644569399230544,
+      "grad_norm": 1.0318970680236816,
+      "learning_rate": 6.447555046506937e-07,
+      "loss": 0.1741,
+      "step": 5627
+    },
+    {
+      "epoch": 2.6649304527966855,
+      "grad_norm": 1.3176510334014893,
+      "learning_rate": 6.429501105964964e-07,
+      "loss": 0.2041,
+      "step": 5628
+    },
+    {
+      "epoch": 2.6654039656703166,
+      "grad_norm": 1.1123440265655518,
+      "learning_rate": 6.411471637884315e-07,
+      "loss": 0.2024,
+      "step": 5629
+    },
+    {
+      "epoch": 2.665877478543948,
+      "grad_norm": 0.9631196856498718,
+      "learning_rate": 6.393466646980362e-07,
+      "loss": 0.1872,
+      "step": 5630
+    },
+    {
+      "epoch": 2.666350991417579,
+      "grad_norm": 1.3509668111801147,
+      "learning_rate": 6.375486137962194e-07,
+      "loss": 0.1934,
+      "step": 5631
+    },
+    {
+      "epoch": 2.6668245042912107,
+      "grad_norm": 1.3398867845535278,
+      "learning_rate": 6.357530115532417e-07,
+      "loss": 0.1883,
+      "step": 5632
+    },
+    {
+      "epoch": 2.6672980171648417,
+      "grad_norm": 1.4973032474517822,
+      "learning_rate": 6.339598584387241e-07,
+      "loss": 0.2354,
+      "step": 5633
+    },
+    {
+      "epoch": 2.667771530038473,
+      "grad_norm": 1.0023930072784424,
+      "learning_rate": 6.321691549216502e-07,
+      "loss": 0.2074,
+      "step": 5634
+    },
+    {
+      "epoch": 2.6682450429121043,
+      "grad_norm": 1.0620249509811401,
+      "learning_rate": 6.303809014703599e-07,
+      "loss": 0.1999,
+      "step": 5635
+    },
+    {
+      "epoch": 2.6687185557857354,
+      "grad_norm": 1.0061266422271729,
+      "learning_rate": 6.285950985525569e-07,
+      "loss": 0.2038,
+      "step": 5636
+    },
+    {
+      "epoch": 2.669192068659367,
+      "grad_norm": 0.9382733106613159,
+      "learning_rate": 6.268117466352952e-07,
+      "loss": 0.1932,
+      "step": 5637
+    },
+    {
+      "epoch": 2.669665581532998,
+      "grad_norm": 1.1017144918441772,
+      "learning_rate": 6.250308461849986e-07,
+      "loss": 0.2087,
+      "step": 5638
+    },
+    {
+      "epoch": 2.670139094406629,
+      "grad_norm": 1.4511491060256958,
+      "learning_rate": 6.232523976674409e-07,
+      "loss": 0.2105,
+      "step": 5639
+    },
+    {
+      "epoch": 2.6706126072802605,
+      "grad_norm": 1.016521692276001,
+      "learning_rate": 6.214764015477614e-07,
+      "loss": 0.1869,
+      "step": 5640
+    },
+    {
+      "epoch": 2.6710861201538916,
+      "grad_norm": 1.1885292530059814,
+      "learning_rate": 6.197028582904507e-07,
+      "loss": 0.2107,
+      "step": 5641
+    },
+    {
+      "epoch": 2.671559633027523,
+      "grad_norm": 1.1228227615356445,
+      "learning_rate": 6.179317683593656e-07,
+      "loss": 0.2178,
+      "step": 5642
+    },
+    {
+      "epoch": 2.672033145901154,
+      "grad_norm": 1.4119621515274048,
+      "learning_rate": 6.161631322177164e-07,
+      "loss": 0.2027,
+      "step": 5643
+    },
+    {
+      "epoch": 2.672506658774785,
+      "grad_norm": 1.3242872953414917,
+      "learning_rate": 6.1439695032807e-07,
+      "loss": 0.2273,
+      "step": 5644
+    },
+    {
+      "epoch": 2.6729801716484167,
+      "grad_norm": 1.2599126100540161,
+      "learning_rate": 6.126332231523591e-07,
+      "loss": 0.2258,
+      "step": 5645
+    },
+    {
+      "epoch": 2.673453684522048,
+      "grad_norm": 1.113526701927185,
+      "learning_rate": 6.108719511518658e-07,
+      "loss": 0.1929,
+      "step": 5646
+    },
+    {
+      "epoch": 2.6739271973956793,
+      "grad_norm": 1.0311851501464844,
+      "learning_rate": 6.09113134787237e-07,
+      "loss": 0.2111,
+      "step": 5647
+    },
+    {
+      "epoch": 2.6744007102693104,
+      "grad_norm": 1.3474483489990234,
+      "learning_rate": 6.073567745184694e-07,
+      "loss": 0.1973,
+      "step": 5648
+    },
+    {
+      "epoch": 2.6748742231429414,
+      "grad_norm": 1.5055440664291382,
+      "learning_rate": 6.056028708049278e-07,
+      "loss": 0.2036,
+      "step": 5649
+    },
+    {
+      "epoch": 2.675347736016573,
+      "grad_norm": 1.0800116062164307,
+      "learning_rate": 6.038514241053239e-07,
+      "loss": 0.1864,
+      "step": 5650
+    },
+    {
+      "epoch": 2.6758212488902045,
+      "grad_norm": 1.0215858221054077,
+      "learning_rate": 6.021024348777349e-07,
+      "loss": 0.1721,
+      "step": 5651
+    },
+    {
+      "epoch": 2.6762947617638355,
+      "grad_norm": 1.2525755167007446,
+      "learning_rate": 6.003559035795914e-07,
+      "loss": 0.1724,
+      "step": 5652
+    },
+    {
+      "epoch": 2.6767682746374666,
+      "grad_norm": 1.2781281471252441,
+      "learning_rate": 5.986118306676791e-07,
+      "loss": 0.2081,
+      "step": 5653
+    },
+    {
+      "epoch": 2.677241787511098,
+      "grad_norm": 1.0916087627410889,
+      "learning_rate": 5.968702165981477e-07,
+      "loss": 0.1801,
+      "step": 5654
+    },
+    {
+      "epoch": 2.677715300384729,
+      "grad_norm": 0.9678347706794739,
+      "learning_rate": 5.95131061826496e-07,
+      "loss": 0.1771,
+      "step": 5655
+    },
+    {
+      "epoch": 2.6781888132583607,
+      "grad_norm": 1.1609114408493042,
+      "learning_rate": 5.933943668075869e-07,
+      "loss": 0.1954,
+      "step": 5656
+    },
+    {
+      "epoch": 2.6786623261319917,
+      "grad_norm": 1.1677464246749878,
+      "learning_rate": 5.916601319956339e-07,
+      "loss": 0.1781,
+      "step": 5657
+    },
+    {
+      "epoch": 2.679135839005623,
+      "grad_norm": 1.1138585805892944,
+      "learning_rate": 5.899283578442073e-07,
+      "loss": 0.1877,
+      "step": 5658
+    },
+    {
+      "epoch": 2.6796093518792543,
+      "grad_norm": 1.0214651823043823,
+      "learning_rate": 5.881990448062402e-07,
+      "loss": 0.2018,
+      "step": 5659
+    },
+    {
+      "epoch": 2.6800828647528854,
+      "grad_norm": 1.117795467376709,
+      "learning_rate": 5.864721933340145e-07,
+      "loss": 0.2012,
+      "step": 5660
+    },
+    {
+      "epoch": 2.680556377626517,
+      "grad_norm": 1.546120524406433,
+      "learning_rate": 5.847478038791732e-07,
+      "loss": 0.2145,
+      "step": 5661
+    },
+    {
+      "epoch": 2.681029890500148,
+      "grad_norm": 1.232239842414856,
+      "learning_rate": 5.830258768927122e-07,
+      "loss": 0.181,
+      "step": 5662
+    },
+    {
+      "epoch": 2.681503403373779,
+      "grad_norm": 1.0880526304244995,
+      "learning_rate": 5.813064128249879e-07,
+      "loss": 0.1968,
+      "step": 5663
+    },
+    {
+      "epoch": 2.6819769162474105,
+      "grad_norm": 1.078931450843811,
+      "learning_rate": 5.795894121257062e-07,
+      "loss": 0.2053,
+      "step": 5664
+    },
+    {
+      "epoch": 2.6824504291210416,
+      "grad_norm": 2.163848400115967,
+      "learning_rate": 5.778748752439345e-07,
+      "loss": 0.1973,
+      "step": 5665
+    },
+    {
+      "epoch": 2.682923941994673,
+      "grad_norm": 1.1884369850158691,
+      "learning_rate": 5.761628026280908e-07,
+      "loss": 0.2006,
+      "step": 5666
+    },
+    {
+      "epoch": 2.683397454868304,
+      "grad_norm": 1.1050399541854858,
+      "learning_rate": 5.744531947259535e-07,
+      "loss": 0.1786,
+      "step": 5667
+    },
+    {
+      "epoch": 2.6838709677419352,
+      "grad_norm": 1.53378427028656,
+      "learning_rate": 5.727460519846539e-07,
+      "loss": 0.2083,
+      "step": 5668
+    },
+    {
+      "epoch": 2.6843444806155667,
+      "grad_norm": 1.0245704650878906,
+      "learning_rate": 5.710413748506772e-07,
+      "loss": 0.2315,
+      "step": 5669
+    },
+    {
+      "epoch": 2.6848179934891983,
+      "grad_norm": 1.2128366231918335,
+      "learning_rate": 5.693391637698664e-07,
+      "loss": 0.2033,
+      "step": 5670
+    },
+    {
+      "epoch": 2.6852915063628293,
+      "grad_norm": 1.1133742332458496,
+      "learning_rate": 5.676394191874179e-07,
+      "loss": 0.1781,
+      "step": 5671
+    },
+    {
+      "epoch": 2.6857650192364604,
+      "grad_norm": 1.095633864402771,
+      "learning_rate": 5.65942141547885e-07,
+      "loss": 0.2086,
+      "step": 5672
+    },
+    {
+      "epoch": 2.686238532110092,
+      "grad_norm": 1.5314477682113647,
+      "learning_rate": 5.642473312951713e-07,
+      "loss": 0.2064,
+      "step": 5673
+    },
+    {
+      "epoch": 2.686712044983723,
+      "grad_norm": 1.4954949617385864,
+      "learning_rate": 5.625549888725401e-07,
+      "loss": 0.1821,
+      "step": 5674
+    },
+    {
+      "epoch": 2.6871855578573545,
+      "grad_norm": 1.1045236587524414,
+      "learning_rate": 5.608651147226074e-07,
+      "loss": 0.1925,
+      "step": 5675
+    },
+    {
+      "epoch": 2.6876590707309855,
+      "grad_norm": 1.7516658306121826,
+      "learning_rate": 5.591777092873429e-07,
+      "loss": 0.2061,
+      "step": 5676
+    },
+    {
+      "epoch": 2.6881325836046166,
+      "grad_norm": 1.2861231565475464,
+      "learning_rate": 5.574927730080725e-07,
+      "loss": 0.2016,
+      "step": 5677
+    },
+    {
+      "epoch": 2.688606096478248,
+      "grad_norm": 1.7489197254180908,
+      "learning_rate": 5.558103063254716e-07,
+      "loss": 0.2002,
+      "step": 5678
+    },
+    {
+      "epoch": 2.689079609351879,
+      "grad_norm": 1.1831798553466797,
+      "learning_rate": 5.541303096795769e-07,
+      "loss": 0.2127,
+      "step": 5679
+    },
+    {
+      "epoch": 2.6895531222255107,
+      "grad_norm": 1.17523992061615,
+      "learning_rate": 5.524527835097726e-07,
+      "loss": 0.1985,
+      "step": 5680
+    },
+    {
+      "epoch": 2.6900266350991417,
+      "grad_norm": 1.547934651374817,
+      "learning_rate": 5.507777282548021e-07,
+      "loss": 0.2032,
+      "step": 5681
+    },
+    {
+      "epoch": 2.690500147972773,
+      "grad_norm": 1.073725700378418,
+      "learning_rate": 5.491051443527573e-07,
+      "loss": 0.2342,
+      "step": 5682
+    },
+    {
+      "epoch": 2.6909736608464043,
+      "grad_norm": 0.9526923298835754,
+      "learning_rate": 5.474350322410882e-07,
+      "loss": 0.1952,
+      "step": 5683
+    },
+    {
+      "epoch": 2.6914471737200354,
+      "grad_norm": 1.3431147336959839,
+      "learning_rate": 5.457673923565954e-07,
+      "loss": 0.1967,
+      "step": 5684
+    },
+    {
+      "epoch": 2.691920686593667,
+      "grad_norm": 1.127583622932434,
+      "learning_rate": 5.441022251354355e-07,
+      "loss": 0.2074,
+      "step": 5685
+    },
+    {
+      "epoch": 2.692394199467298,
+      "grad_norm": 0.9880527257919312,
+      "learning_rate": 5.424395310131159e-07,
+      "loss": 0.2143,
+      "step": 5686
+    },
+    {
+      "epoch": 2.692867712340929,
+      "grad_norm": 1.1635633707046509,
+      "learning_rate": 5.407793104244963e-07,
+      "loss": 0.2018,
+      "step": 5687
+    },
+    {
+      "epoch": 2.6933412252145605,
+      "grad_norm": 1.3094615936279297,
+      "learning_rate": 5.391215638037961e-07,
+      "loss": 0.229,
+      "step": 5688
+    },
+    {
+      "epoch": 2.6938147380881916,
+      "grad_norm": 1.3654303550720215,
+      "learning_rate": 5.374662915845774e-07,
+      "loss": 0.2072,
+      "step": 5689
+    },
+    {
+      "epoch": 2.694288250961823,
+      "grad_norm": 1.0598773956298828,
+      "learning_rate": 5.358134941997661e-07,
+      "loss": 0.219,
+      "step": 5690
+    },
+    {
+      "epoch": 2.694761763835454,
+      "grad_norm": 1.110864281654358,
+      "learning_rate": 5.341631720816309e-07,
+      "loss": 0.192,
+      "step": 5691
+    },
+    {
+      "epoch": 2.6952352767090857,
+      "grad_norm": 1.1458567380905151,
+      "learning_rate": 5.325153256617988e-07,
+      "loss": 0.1928,
+      "step": 5692
+    },
+    {
+      "epoch": 2.6957087895827168,
+      "grad_norm": 1.2874677181243896,
+      "learning_rate": 5.308699553712515e-07,
+      "loss": 0.2226,
+      "step": 5693
+    },
+    {
+      "epoch": 2.6961823024563483,
+      "grad_norm": 1.3333238363265991,
+      "learning_rate": 5.29227061640315e-07,
+      "loss": 0.1834,
+      "step": 5694
+    },
+    {
+      "epoch": 2.6966558153299793,
+      "grad_norm": 1.119017481803894,
+      "learning_rate": 5.275866448986755e-07,
+      "loss": 0.1743,
+      "step": 5695
+    },
+    {
+      "epoch": 2.6971293282036104,
+      "grad_norm": 1.0952125787734985,
+      "learning_rate": 5.259487055753653e-07,
+      "loss": 0.213,
+      "step": 5696
+    },
+    {
+      "epoch": 2.697602841077242,
+      "grad_norm": 1.2849918603897095,
+      "learning_rate": 5.243132440987752e-07,
+      "loss": 0.1902,
+      "step": 5697
+    },
+    {
+      "epoch": 2.698076353950873,
+      "grad_norm": 1.2609485387802124,
+      "learning_rate": 5.226802608966419e-07,
+      "loss": 0.2232,
+      "step": 5698
+    },
+    {
+      "epoch": 2.6985498668245045,
+      "grad_norm": 1.236099123954773,
+      "learning_rate": 5.210497563960581e-07,
+      "loss": 0.2236,
+      "step": 5699
+    },
+    {
+      "epoch": 2.6990233796981355,
+      "grad_norm": 1.6993776559829712,
+      "learning_rate": 5.19421731023464e-07,
+      "loss": 0.1883,
+      "step": 5700
+    },
+    {
+      "epoch": 2.6994968925717666,
+      "grad_norm": 2.0070505142211914,
+      "learning_rate": 5.177961852046565e-07,
+      "loss": 0.202,
+      "step": 5701
+    },
+    {
+      "epoch": 2.699970405445398,
+      "grad_norm": 1.0293316841125488,
+      "learning_rate": 5.161731193647801e-07,
+      "loss": 0.1863,
+      "step": 5702
+    },
+    {
+      "epoch": 2.700443918319029,
+      "grad_norm": 0.9846745133399963,
+      "learning_rate": 5.145525339283308e-07,
+      "loss": 0.1895,
+      "step": 5703
+    },
+    {
+      "epoch": 2.7009174311926607,
+      "grad_norm": 1.1934609413146973,
+      "learning_rate": 5.129344293191607e-07,
+      "loss": 0.1918,
+      "step": 5704
+    },
+    {
+      "epoch": 2.7013909440662918,
+      "grad_norm": 1.1649199724197388,
+      "learning_rate": 5.113188059604657e-07,
+      "loss": 0.2288,
+      "step": 5705
+    },
+    {
+      "epoch": 2.701864456939923,
+      "grad_norm": 1.2630609273910522,
+      "learning_rate": 5.097056642747988e-07,
+      "loss": 0.1989,
+      "step": 5706
+    },
+    {
+      "epoch": 2.7023379698135543,
+      "grad_norm": 1.5064605474472046,
+      "learning_rate": 5.080950046840594e-07,
+      "loss": 0.2033,
+      "step": 5707
+    },
+    {
+      "epoch": 2.7028114826871854,
+      "grad_norm": 1.2008765935897827,
+      "learning_rate": 5.064868276095036e-07,
+      "loss": 0.2044,
+      "step": 5708
+    },
+    {
+      "epoch": 2.703284995560817,
+      "grad_norm": 1.0486986637115479,
+      "learning_rate": 5.048811334717307e-07,
+      "loss": 0.1992,
+      "step": 5709
+    },
+    {
+      "epoch": 2.703758508434448,
+      "grad_norm": 1.0270819664001465,
+      "learning_rate": 5.032779226906981e-07,
+      "loss": 0.2147,
+      "step": 5710
+    },
+    {
+      "epoch": 2.704232021308079,
+      "grad_norm": 0.9577698707580566,
+      "learning_rate": 5.016771956857081e-07,
+      "loss": 0.1697,
+      "step": 5711
+    },
+    {
+      "epoch": 2.7047055341817106,
+      "grad_norm": 1.1315481662750244,
+      "learning_rate": 5.000789528754147e-07,
+      "loss": 0.201,
+      "step": 5712
+    },
+    {
+      "epoch": 2.705179047055342,
+      "grad_norm": 1.1655607223510742,
+      "learning_rate": 4.984831946778246e-07,
+      "loss": 0.2057,
+      "step": 5713
+    },
+    {
+      "epoch": 2.705652559928973,
+      "grad_norm": 1.2996375560760498,
+      "learning_rate": 4.968899215102907e-07,
+      "loss": 0.2162,
+      "step": 5714
+    },
+    {
+      "epoch": 2.706126072802604,
+      "grad_norm": 1.351178765296936,
+      "learning_rate": 4.952991337895219e-07,
+      "loss": 0.1751,
+      "step": 5715
+    },
+    {
+      "epoch": 2.7065995856762357,
+      "grad_norm": 0.9919388890266418,
+      "learning_rate": 4.937108319315687e-07,
+      "loss": 0.2041,
+      "step": 5716
+    },
+    {
+      "epoch": 2.7070730985498668,
+      "grad_norm": 1.0616121292114258,
+      "learning_rate": 4.92125016351841e-07,
+      "loss": 0.2077,
+      "step": 5717
+    },
+    {
+      "epoch": 2.7075466114234983,
+      "grad_norm": 1.1046313047409058,
+      "learning_rate": 4.905416874650892e-07,
+      "loss": 0.2069,
+      "step": 5718
+    },
+    {
+      "epoch": 2.7080201242971293,
+      "grad_norm": 1.44788658618927,
+      "learning_rate": 4.889608456854211e-07,
+      "loss": 0.1919,
+      "step": 5719
+    },
+    {
+      "epoch": 2.7084936371707604,
+      "grad_norm": 0.9813434481620789,
+      "learning_rate": 4.873824914262882e-07,
+      "loss": 0.1997,
+      "step": 5720
+    },
+    {
+      "epoch": 2.708967150044392,
+      "grad_norm": 1.2912452220916748,
+      "learning_rate": 4.858066251004956e-07,
+      "loss": 0.2067,
+      "step": 5721
+    },
+    {
+      "epoch": 2.709440662918023,
+      "grad_norm": 0.9664787650108337,
+      "learning_rate": 4.842332471201961e-07,
+      "loss": 0.1958,
+      "step": 5722
+    },
+    {
+      "epoch": 2.7099141757916545,
+      "grad_norm": 1.488752841949463,
+      "learning_rate": 4.826623578968881e-07,
+      "loss": 0.2134,
+      "step": 5723
+    },
+    {
+      "epoch": 2.7103876886652856,
+      "grad_norm": 1.1605441570281982,
+      "learning_rate": 4.810939578414265e-07,
+      "loss": 0.2028,
+      "step": 5724
+    },
+    {
+      "epoch": 2.7108612015389166,
+      "grad_norm": 1.2621954679489136,
+      "learning_rate": 4.795280473640085e-07,
+      "loss": 0.2198,
+      "step": 5725
+    },
+    {
+      "epoch": 2.711334714412548,
+      "grad_norm": 1.275181531906128,
+      "learning_rate": 4.779646268741866e-07,
+      "loss": 0.1786,
+      "step": 5726
+    },
+    {
+      "epoch": 2.711808227286179,
+      "grad_norm": 0.8626628518104553,
+      "learning_rate": 4.7640369678085275e-07,
+      "loss": 0.1957,
+      "step": 5727
+    },
+    {
+      "epoch": 2.7122817401598107,
+      "grad_norm": 1.1933006048202515,
+      "learning_rate": 4.7484525749225907e-07,
+      "loss": 0.1729,
+      "step": 5728
+    },
+    {
+      "epoch": 2.712755253033442,
+      "grad_norm": 0.9497954249382019,
+      "learning_rate": 4.7328930941599514e-07,
+      "loss": 0.1848,
+      "step": 5729
+    },
+    {
+      "epoch": 2.713228765907073,
+      "grad_norm": 1.1757985353469849,
+      "learning_rate": 4.717358529590077e-07,
+      "loss": 0.2026,
+      "step": 5730
+    },
+    {
+      "epoch": 2.7137022787807044,
+      "grad_norm": 1.2278982400894165,
+      "learning_rate": 4.7018488852758727e-07,
+      "loss": 0.2011,
+      "step": 5731
+    },
+    {
+      "epoch": 2.714175791654336,
+      "grad_norm": 1.4735335111618042,
+      "learning_rate": 4.6863641652737157e-07,
+      "loss": 0.2165,
+      "step": 5732
+    },
+    {
+      "epoch": 2.714649304527967,
+      "grad_norm": 1.0341819524765015,
+      "learning_rate": 4.6709043736335334e-07,
+      "loss": 0.2221,
+      "step": 5733
+    },
+    {
+      "epoch": 2.715122817401598,
+      "grad_norm": 1.042711853981018,
+      "learning_rate": 4.655469514398636e-07,
+      "loss": 0.208,
+      "step": 5734
+    },
+    {
+      "epoch": 2.7155963302752295,
+      "grad_norm": 1.383939504623413,
+      "learning_rate": 4.6400595916058944e-07,
+      "loss": 0.2037,
+      "step": 5735
+    },
+    {
+      "epoch": 2.7160698431488606,
+      "grad_norm": 0.9150691032409668,
+      "learning_rate": 4.6246746092856176e-07,
+      "loss": 0.1982,
+      "step": 5736
+    },
+    {
+      "epoch": 2.716543356022492,
+      "grad_norm": 1.098061203956604,
+      "learning_rate": 4.6093145714615763e-07,
+      "loss": 0.2108,
+      "step": 5737
+    },
+    {
+      "epoch": 2.717016868896123,
+      "grad_norm": 0.9107375144958496,
+      "learning_rate": 4.5939794821510785e-07,
+      "loss": 0.1822,
+      "step": 5738
+    },
+    {
+      "epoch": 2.717490381769754,
+      "grad_norm": 1.2272454500198364,
+      "learning_rate": 4.578669345364828e-07,
+      "loss": 0.1883,
+      "step": 5739
+    },
+    {
+      "epoch": 2.7179638946433857,
+      "grad_norm": 1.202950119972229,
+      "learning_rate": 4.5633841651070766e-07,
+      "loss": 0.1952,
+      "step": 5740
+    },
+    {
+      "epoch": 2.718437407517017,
+      "grad_norm": 1.1974061727523804,
+      "learning_rate": 4.548123945375493e-07,
+      "loss": 0.2013,
+      "step": 5741
+    },
+    {
+      "epoch": 2.7189109203906483,
+      "grad_norm": 1.0299830436706543,
+      "learning_rate": 4.5328886901612743e-07,
+      "loss": 0.1803,
+      "step": 5742
+    },
+    {
+      "epoch": 2.7193844332642794,
+      "grad_norm": 0.9163728356361389,
+      "learning_rate": 4.5176784034489993e-07,
+      "loss": 0.1797,
+      "step": 5743
+    },
+    {
+      "epoch": 2.7198579461379104,
+      "grad_norm": 1.6618443727493286,
+      "learning_rate": 4.5024930892168305e-07,
+      "loss": 0.2081,
+      "step": 5744
+    },
+    {
+      "epoch": 2.720331459011542,
+      "grad_norm": 1.2780410051345825,
+      "learning_rate": 4.487332751436302e-07,
+      "loss": 0.1836,
+      "step": 5745
+    },
+    {
+      "epoch": 2.720804971885173,
+      "grad_norm": 0.9704563617706299,
+      "learning_rate": 4.472197394072464e-07,
+      "loss": 0.1998,
+      "step": 5746
+    },
+    {
+      "epoch": 2.7212784847588045,
+      "grad_norm": 1.1087617874145508,
+      "learning_rate": 4.457087021083839e-07,
+      "loss": 0.2173,
+      "step": 5747
+    },
+    {
+      "epoch": 2.7217519976324356,
+      "grad_norm": 1.3192325830459595,
+      "learning_rate": 4.442001636422366e-07,
+      "loss": 0.2108,
+      "step": 5748
+    },
+    {
+      "epoch": 2.7222255105060666,
+      "grad_norm": 0.8854762315750122,
+      "learning_rate": 4.4269412440335114e-07,
+      "loss": 0.2035,
+      "step": 5749
+    },
+    {
+      "epoch": 2.722699023379698,
+      "grad_norm": 1.296974778175354,
+      "learning_rate": 4.411905847856157e-07,
+      "loss": 0.1926,
+      "step": 5750
+    },
+    {
+      "epoch": 2.7231725362533292,
+      "grad_norm": 1.4468486309051514,
+      "learning_rate": 4.39689545182268e-07,
+      "loss": 0.198,
+      "step": 5751
+    },
+    {
+      "epoch": 2.7236460491269607,
+      "grad_norm": 1.2393492460250854,
+      "learning_rate": 4.381910059858896e-07,
+      "loss": 0.1975,
+      "step": 5752
+    },
+    {
+      "epoch": 2.724119562000592,
+      "grad_norm": 1.3274880647659302,
+      "learning_rate": 4.366949675884091e-07,
+      "loss": 0.201,
+      "step": 5753
+    },
+    {
+      "epoch": 2.7245930748742233,
+      "grad_norm": 1.3509947061538696,
+      "learning_rate": 4.352014303811003e-07,
+      "loss": 0.1904,
+      "step": 5754
+    },
+    {
+      "epoch": 2.7250665877478544,
+      "grad_norm": 0.9841063022613525,
+      "learning_rate": 4.337103947545862e-07,
+      "loss": 0.1994,
+      "step": 5755
+    },
+    {
+      "epoch": 2.725540100621486,
+      "grad_norm": 1.1662771701812744,
+      "learning_rate": 4.3222186109882933e-07,
+      "loss": 0.2036,
+      "step": 5756
+    },
+    {
+      "epoch": 2.726013613495117,
+      "grad_norm": 0.9167471528053284,
+      "learning_rate": 4.307358298031428e-07,
+      "loss": 0.1971,
+      "step": 5757
+    },
+    {
+      "epoch": 2.726487126368748,
+      "grad_norm": 1.2513035535812378,
+      "learning_rate": 4.2925230125618336e-07,
+      "loss": 0.2269,
+      "step": 5758
+    },
+    {
+      "epoch": 2.7269606392423795,
+      "grad_norm": 1.0162153244018555,
+      "learning_rate": 4.2777127584595403e-07,
+      "loss": 0.1965,
+      "step": 5759
+    },
+    {
+      "epoch": 2.7274341521160106,
+      "grad_norm": 1.4992437362670898,
+      "learning_rate": 4.2629275395980275e-07,
+      "loss": 0.1795,
+      "step": 5760
+    },
+    {
+      "epoch": 2.727907664989642,
+      "grad_norm": 1.2534233331680298,
+      "learning_rate": 4.248167359844224e-07,
+      "loss": 0.2045,
+      "step": 5761
+    },
+    {
+      "epoch": 2.728381177863273,
+      "grad_norm": 1.2920854091644287,
+      "learning_rate": 4.23343222305852e-07,
+      "loss": 0.1998,
+      "step": 5762
+    },
+    {
+      "epoch": 2.7288546907369042,
+      "grad_norm": 0.9584261775016785,
+      "learning_rate": 4.2187221330947216e-07,
+      "loss": 0.1905,
+      "step": 5763
+    },
+    {
+      "epoch": 2.7293282036105357,
+      "grad_norm": 1.5315953493118286,
+      "learning_rate": 4.2040370938001507e-07,
+      "loss": 0.21,
+      "step": 5764
+    },
+    {
+      "epoch": 2.729801716484167,
+      "grad_norm": 1.48392915725708,
+      "learning_rate": 4.1893771090155246e-07,
+      "loss": 0.1877,
+      "step": 5765
+    },
+    {
+      "epoch": 2.7302752293577983,
+      "grad_norm": 0.9685987830162048,
+      "learning_rate": 4.174742182574998e-07,
+      "loss": 0.2019,
+      "step": 5766
+    },
+    {
+      "epoch": 2.7307487422314294,
+      "grad_norm": 1.2201647758483887,
+      "learning_rate": 4.1601323183062205e-07,
+      "loss": 0.1967,
+      "step": 5767
+    },
+    {
+      "epoch": 2.7312222551050604,
+      "grad_norm": 0.9604336023330688,
+      "learning_rate": 4.1455475200302353e-07,
+      "loss": 0.1898,
+      "step": 5768
+    },
+    {
+      "epoch": 2.731695767978692,
+      "grad_norm": 0.9989910125732422,
+      "learning_rate": 4.1309877915615913e-07,
+      "loss": 0.2053,
+      "step": 5769
+    },
+    {
+      "epoch": 2.732169280852323,
+      "grad_norm": 1.192683219909668,
+      "learning_rate": 4.116453136708187e-07,
+      "loss": 0.2246,
+      "step": 5770
+    },
+    {
+      "epoch": 2.7326427937259545,
+      "grad_norm": 1.0321522951126099,
+      "learning_rate": 4.101943559271504e-07,
+      "loss": 0.2037,
+      "step": 5771
+    },
+    {
+      "epoch": 2.7331163065995856,
+      "grad_norm": 1.1789295673370361,
+      "learning_rate": 4.0874590630463283e-07,
+      "loss": 0.2316,
+      "step": 5772
+    },
+    {
+      "epoch": 2.7335898194732167,
+      "grad_norm": 1.1018298864364624,
+      "learning_rate": 4.072999651820941e-07,
+      "loss": 0.2042,
+      "step": 5773
+    },
+    {
+      "epoch": 2.734063332346848,
+      "grad_norm": 0.9121476411819458,
+      "learning_rate": 4.058565329377073e-07,
+      "loss": 0.1769,
+      "step": 5774
+    },
+    {
+      "epoch": 2.7345368452204797,
+      "grad_norm": 1.112981915473938,
+      "learning_rate": 4.044156099489882e-07,
+      "loss": 0.189,
+      "step": 5775
+    },
+    {
+      "epoch": 2.7350103580941107,
+      "grad_norm": 1.084700345993042,
+      "learning_rate": 4.0297719659279645e-07,
+      "loss": 0.2051,
+      "step": 5776
+    },
+    {
+      "epoch": 2.735483870967742,
+      "grad_norm": 1.0018030405044556,
+      "learning_rate": 4.015412932453333e-07,
+      "loss": 0.2046,
+      "step": 5777
+    },
+    {
+      "epoch": 2.7359573838413733,
+      "grad_norm": 1.5083683729171753,
+      "learning_rate": 4.0010790028214843e-07,
+      "loss": 0.1895,
+      "step": 5778
+    },
+    {
+      "epoch": 2.7364308967150044,
+      "grad_norm": 1.0385921001434326,
+      "learning_rate": 3.9867701807812963e-07,
+      "loss": 0.2093,
+      "step": 5779
+    },
+    {
+      "epoch": 2.736904409588636,
+      "grad_norm": 1.0134193897247314,
+      "learning_rate": 3.9724864700751207e-07,
+      "loss": 0.2117,
+      "step": 5780
+    },
+    {
+      "epoch": 2.737377922462267,
+      "grad_norm": 1.1653934717178345,
+      "learning_rate": 3.9582278744387137e-07,
+      "loss": 0.2079,
+      "step": 5781
+    },
+    {
+      "epoch": 2.737851435335898,
+      "grad_norm": 1.1373720169067383,
+      "learning_rate": 3.9439943976012696e-07,
+      "loss": 0.2118,
+      "step": 5782
+    },
+    {
+      "epoch": 2.7383249482095295,
+      "grad_norm": 1.2817039489746094,
+      "learning_rate": 3.929786043285433e-07,
+      "loss": 0.2094,
+      "step": 5783
+    },
+    {
+      "epoch": 2.7387984610831606,
+      "grad_norm": 0.9910814166069031,
+      "learning_rate": 3.915602815207231e-07,
+      "loss": 0.2151,
+      "step": 5784
+    },
+    {
+      "epoch": 2.739271973956792,
+      "grad_norm": 1.1655786037445068,
+      "learning_rate": 3.901444717076186e-07,
+      "loss": 0.2014,
+      "step": 5785
+    },
+    {
+      "epoch": 2.739745486830423,
+      "grad_norm": 1.3065234422683716,
+      "learning_rate": 3.8873117525951797e-07,
+      "loss": 0.1806,
+      "step": 5786
+    },
+    {
+      "epoch": 2.7402189997040542,
+      "grad_norm": 1.1026694774627686,
+      "learning_rate": 3.873203925460589e-07,
+      "loss": 0.183,
+      "step": 5787
+    },
+    {
+      "epoch": 2.7406925125776858,
+      "grad_norm": 1.210478663444519,
+      "learning_rate": 3.8591212393621405e-07,
+      "loss": 0.216,
+      "step": 5788
+    },
+    {
+      "epoch": 2.741166025451317,
+      "grad_norm": 1.6913577318191528,
+      "learning_rate": 3.845063697983065e-07,
+      "loss": 0.2123,
+      "step": 5789
+    },
+    {
+      "epoch": 2.7416395383249483,
+      "grad_norm": 1.2592370510101318,
+      "learning_rate": 3.8310313049999546e-07,
+      "loss": 0.1888,
+      "step": 5790
+    },
+    {
+      "epoch": 2.7421130511985794,
+      "grad_norm": 0.9264201521873474,
+      "learning_rate": 3.8170240640828304e-07,
+      "loss": 0.1949,
+      "step": 5791
+    },
+    {
+      "epoch": 2.7425865640722105,
+      "grad_norm": 1.0950411558151245,
+      "learning_rate": 3.8030419788951834e-07,
+      "loss": 0.215,
+      "step": 5792
+    },
+    {
+      "epoch": 2.743060076945842,
+      "grad_norm": 1.2804242372512817,
+      "learning_rate": 3.789085053093866e-07,
+      "loss": 0.2031,
+      "step": 5793
+    },
+    {
+      "epoch": 2.7435335898194735,
+      "grad_norm": 1.092449426651001,
+      "learning_rate": 3.775153290329203e-07,
+      "loss": 0.1747,
+      "step": 5794
+    },
+    {
+      "epoch": 2.7440071026931045,
+      "grad_norm": 0.9500091075897217,
+      "learning_rate": 3.7612466942448797e-07,
+      "loss": 0.2102,
+      "step": 5795
+    },
+    {
+      "epoch": 2.7444806155667356,
+      "grad_norm": 1.2651680707931519,
+      "learning_rate": 3.747365268478076e-07,
+      "loss": 0.1895,
+      "step": 5796
+    },
+    {
+      "epoch": 2.744954128440367,
+      "grad_norm": 1.0188058614730835,
+      "learning_rate": 3.733509016659298e-07,
+      "loss": 0.2146,
+      "step": 5797
+    },
+    {
+      "epoch": 2.745427641313998,
+      "grad_norm": 1.008729338645935,
+      "learning_rate": 3.7196779424125585e-07,
+      "loss": 0.2023,
+      "step": 5798
+    },
+    {
+      "epoch": 2.7459011541876297,
+      "grad_norm": 1.0331710577011108,
+      "learning_rate": 3.705872049355208e-07,
+      "loss": 0.1947,
+      "step": 5799
+    },
+    {
+      "epoch": 2.7463746670612608,
+      "grad_norm": 0.8746491074562073,
+      "learning_rate": 3.6920913410980585e-07,
+      "loss": 0.1955,
+      "step": 5800
+    },
+    {
+      "epoch": 2.746848179934892,
+      "grad_norm": 0.9344222545623779,
+      "learning_rate": 3.678335821245327e-07,
+      "loss": 0.1811,
+      "step": 5801
+    },
+    {
+      "epoch": 2.7473216928085233,
+      "grad_norm": 0.9031004309654236,
+      "learning_rate": 3.664605493394624e-07,
+      "loss": 0.1813,
+      "step": 5802
+    },
+    {
+      "epoch": 2.7477952056821544,
+      "grad_norm": 1.2377285957336426,
+      "learning_rate": 3.6509003611369884e-07,
+      "loss": 0.2265,
+      "step": 5803
+    },
+    {
+      "epoch": 2.748268718555786,
+      "grad_norm": 1.0772106647491455,
+      "learning_rate": 3.6372204280568644e-07,
+      "loss": 0.2122,
+      "step": 5804
+    },
+    {
+      "epoch": 2.748742231429417,
+      "grad_norm": 0.8862806558609009,
+      "learning_rate": 3.623565697732123e-07,
+      "loss": 0.1984,
+      "step": 5805
+    },
+    {
+      "epoch": 2.749215744303048,
+      "grad_norm": 1.0365098714828491,
+      "learning_rate": 3.6099361737339965e-07,
+      "loss": 0.1737,
+      "step": 5806
+    },
+    {
+      "epoch": 2.7496892571766796,
+      "grad_norm": 1.0085512399673462,
+      "learning_rate": 3.596331859627189e-07,
+      "loss": 0.1949,
+      "step": 5807
+    },
+    {
+      "epoch": 2.7501627700503106,
+      "grad_norm": 1.1064403057098389,
+      "learning_rate": 3.582752758969743e-07,
+      "loss": 0.1957,
+      "step": 5808
+    },
+    {
+      "epoch": 2.750636282923942,
+      "grad_norm": 0.9437001347541809,
+      "learning_rate": 3.5691988753131625e-07,
+      "loss": 0.1861,
+      "step": 5809
+    },
+    {
+      "epoch": 2.751109795797573,
+      "grad_norm": 0.9680258631706238,
+      "learning_rate": 3.5556702122023444e-07,
+      "loss": 0.2232,
+      "step": 5810
+    },
+    {
+      "epoch": 2.7515833086712043,
+      "grad_norm": 0.9630969762802124,
+      "learning_rate": 3.5421667731755484e-07,
+      "loss": 0.196,
+      "step": 5811
+    },
+    {
+      "epoch": 2.7520568215448358,
+      "grad_norm": 1.0958917140960693,
+      "learning_rate": 3.528688561764515e-07,
+      "loss": 0.2042,
+      "step": 5812
+    },
+    {
+      "epoch": 2.752530334418467,
+      "grad_norm": 1.1072863340377808,
+      "learning_rate": 3.5152355814942916e-07,
+      "loss": 0.2031,
+      "step": 5813
+    },
+    {
+      "epoch": 2.7530038472920983,
+      "grad_norm": 1.231067180633545,
+      "learning_rate": 3.5018078358834084e-07,
+      "loss": 0.1989,
+      "step": 5814
+    },
+    {
+      "epoch": 2.7534773601657294,
+      "grad_norm": 1.1398643255233765,
+      "learning_rate": 3.4884053284437444e-07,
+      "loss": 0.2065,
+      "step": 5815
+    },
+    {
+      "epoch": 2.7539508730393605,
+      "grad_norm": 1.0876073837280273,
+      "learning_rate": 3.4750280626805964e-07,
+      "loss": 0.1859,
+      "step": 5816
+    },
+    {
+      "epoch": 2.754424385912992,
+      "grad_norm": 1.428344964981079,
+      "learning_rate": 3.461676042092688e-07,
+      "loss": 0.2038,
+      "step": 5817
+    },
+    {
+      "epoch": 2.7548978987866235,
+      "grad_norm": 0.8772413730621338,
+      "learning_rate": 3.4483492701720687e-07,
+      "loss": 0.1925,
+      "step": 5818
+    },
+    {
+      "epoch": 2.7553714116602546,
+      "grad_norm": 1.0749726295471191,
+      "learning_rate": 3.435047750404252e-07,
+      "loss": 0.2179,
+      "step": 5819
+    },
+    {
+      "epoch": 2.7558449245338856,
+      "grad_norm": 1.8996601104736328,
+      "learning_rate": 3.4217714862681215e-07,
+      "loss": 0.1664,
+      "step": 5820
+    },
+    {
+      "epoch": 2.756318437407517,
+      "grad_norm": 1.382574200630188,
+      "learning_rate": 3.408520481235955e-07,
+      "loss": 0.2156,
+      "step": 5821
+    },
+    {
+      "epoch": 2.756791950281148,
+      "grad_norm": 1.1029624938964844,
+      "learning_rate": 3.395294738773403e-07,
+      "loss": 0.2294,
+      "step": 5822
+    },
+    {
+      "epoch": 2.7572654631547797,
+      "grad_norm": 1.0061315298080444,
+      "learning_rate": 3.382094262339575e-07,
+      "loss": 0.1968,
+      "step": 5823
+    },
+    {
+      "epoch": 2.757738976028411,
+      "grad_norm": 1.1826341152191162,
+      "learning_rate": 3.368919055386888e-07,
+      "loss": 0.1868,
+      "step": 5824
+    },
+    {
+      "epoch": 2.758212488902042,
+      "grad_norm": 1.690619707107544,
+      "learning_rate": 3.3557691213612075e-07,
+      "loss": 0.2114,
+      "step": 5825
+    },
+    {
+      "epoch": 2.7586860017756734,
+      "grad_norm": 1.034387469291687,
+      "learning_rate": 3.342644463701783e-07,
+      "loss": 0.2105,
+      "step": 5826
+    },
+    {
+      "epoch": 2.7591595146493044,
+      "grad_norm": 0.9920729398727417,
+      "learning_rate": 3.3295450858412125e-07,
+      "loss": 0.1833,
+      "step": 5827
+    },
+    {
+      "epoch": 2.759633027522936,
+      "grad_norm": 1.3329393863677979,
+      "learning_rate": 3.316470991205534e-07,
+      "loss": 0.2245,
+      "step": 5828
+    },
+    {
+      "epoch": 2.760106540396567,
+      "grad_norm": 1.0018433332443237,
+      "learning_rate": 3.3034221832141446e-07,
+      "loss": 0.1915,
+      "step": 5829
+    },
+    {
+      "epoch": 2.760580053270198,
+      "grad_norm": 1.5985726118087769,
+      "learning_rate": 3.2903986652798367e-07,
+      "loss": 0.1633,
+      "step": 5830
+    },
+    {
+      "epoch": 2.7610535661438296,
+      "grad_norm": 2.1785120964050293,
+      "learning_rate": 3.2774004408087756e-07,
+      "loss": 0.1907,
+      "step": 5831
+    },
+    {
+      "epoch": 2.7615270790174606,
+      "grad_norm": 0.9680808782577515,
+      "learning_rate": 3.264427513200552e-07,
+      "loss": 0.1905,
+      "step": 5832
+    },
+    {
+      "epoch": 2.762000591891092,
+      "grad_norm": 1.224849820137024,
+      "learning_rate": 3.2514798858480635e-07,
+      "loss": 0.192,
+      "step": 5833
+    },
+    {
+      "epoch": 2.762474104764723,
+      "grad_norm": 1.091199517250061,
+      "learning_rate": 3.238557562137679e-07,
+      "loss": 0.212,
+      "step": 5834
+    },
+    {
+      "epoch": 2.7629476176383543,
+      "grad_norm": 1.2063572406768799,
+      "learning_rate": 3.225660545449083e-07,
+      "loss": 0.1971,
+      "step": 5835
+    },
+    {
+      "epoch": 2.763421130511986,
+      "grad_norm": 1.4341157674789429,
+      "learning_rate": 3.2127888391553674e-07,
+      "loss": 0.1935,
+      "step": 5836
+    },
+    {
+      "epoch": 2.7638946433856173,
+      "grad_norm": 1.0425024032592773,
+      "learning_rate": 3.1999424466230166e-07,
+      "loss": 0.2098,
+      "step": 5837
+    },
+    {
+      "epoch": 2.7643681562592484,
+      "grad_norm": 1.3151661157608032,
+      "learning_rate": 3.187121371211854e-07,
+      "loss": 0.1904,
+      "step": 5838
+    },
+    {
+      "epoch": 2.7648416691328794,
+      "grad_norm": 1.4269354343414307,
+      "learning_rate": 3.1743256162751534e-07,
+      "loss": 0.2162,
+      "step": 5839
+    },
+    {
+      "epoch": 2.765315182006511,
+      "grad_norm": 0.9791456460952759,
+      "learning_rate": 3.161555185159471e-07,
+      "loss": 0.177,
+      "step": 5840
+    },
+    {
+      "epoch": 2.765788694880142,
+      "grad_norm": 1.1203821897506714,
+      "learning_rate": 3.1488100812048337e-07,
+      "loss": 0.2033,
+      "step": 5841
+    },
+    {
+      "epoch": 2.7662622077537735,
+      "grad_norm": 1.1284160614013672,
+      "learning_rate": 3.136090307744555e-07,
+      "loss": 0.1727,
+      "step": 5842
+    },
+    {
+      "epoch": 2.7667357206274046,
+      "grad_norm": 1.2004402875900269,
+      "learning_rate": 3.1233958681054164e-07,
+      "loss": 0.183,
+      "step": 5843
+    },
+    {
+      "epoch": 2.7672092335010356,
+      "grad_norm": 2.777414083480835,
+      "learning_rate": 3.110726765607497e-07,
+      "loss": 0.2068,
+      "step": 5844
+    },
+    {
+      "epoch": 2.767682746374667,
+      "grad_norm": 1.0925432443618774,
+      "learning_rate": 3.0980830035642783e-07,
+      "loss": 0.2027,
+      "step": 5845
+    },
+    {
+      "epoch": 2.7681562592482982,
+      "grad_norm": 0.955273449420929,
+      "learning_rate": 3.085464585282627e-07,
+      "loss": 0.2024,
+      "step": 5846
+    },
+    {
+      "epoch": 2.7686297721219297,
+      "grad_norm": 1.7570418119430542,
+      "learning_rate": 3.0728715140627584e-07,
+      "loss": 0.199,
+      "step": 5847
+    },
+    {
+      "epoch": 2.769103284995561,
+      "grad_norm": 1.181026577949524,
+      "learning_rate": 3.0603037931982603e-07,
+      "loss": 0.2072,
+      "step": 5848
+    },
+    {
+      "epoch": 2.769576797869192,
+      "grad_norm": 1.4937407970428467,
+      "learning_rate": 3.0477614259761254e-07,
+      "loss": 0.2191,
+      "step": 5849
+    },
+    {
+      "epoch": 2.7700503107428234,
+      "grad_norm": 0.9318932890892029,
+      "learning_rate": 3.035244415676675e-07,
+      "loss": 0.1817,
+      "step": 5850
+    },
+    {
+      "epoch": 2.7705238236164544,
+      "grad_norm": 1.0921509265899658,
+      "learning_rate": 3.0227527655736223e-07,
+      "loss": 0.2072,
+      "step": 5851
+    },
+    {
+      "epoch": 2.770997336490086,
+      "grad_norm": 1.1498230695724487,
+      "learning_rate": 3.01028647893401e-07,
+      "loss": 0.2132,
+      "step": 5852
+    },
+    {
+      "epoch": 2.771470849363717,
+      "grad_norm": 1.4725837707519531,
+      "learning_rate": 2.9978455590182974e-07,
+      "loss": 0.2007,
+      "step": 5853
+    },
+    {
+      "epoch": 2.771944362237348,
+      "grad_norm": 1.1440794467926025,
+      "learning_rate": 2.985430009080281e-07,
+      "loss": 0.2044,
+      "step": 5854
+    },
+    {
+      "epoch": 2.7724178751109796,
+      "grad_norm": 1.0810843706130981,
+      "learning_rate": 2.9730398323671415e-07,
+      "loss": 0.175,
+      "step": 5855
+    },
+    {
+      "epoch": 2.772891387984611,
+      "grad_norm": 1.1769734621047974,
+      "learning_rate": 2.960675032119387e-07,
+      "loss": 0.2085,
+      "step": 5856
+    },
+    {
+      "epoch": 2.773364900858242,
+      "grad_norm": 0.9242314100265503,
+      "learning_rate": 2.9483356115709295e-07,
+      "loss": 0.2023,
+      "step": 5857
+    },
+    {
+      "epoch": 2.7738384137318732,
+      "grad_norm": 0.8873565793037415,
+      "learning_rate": 2.936021573949011e-07,
+      "loss": 0.2078,
+      "step": 5858
+    },
+    {
+      "epoch": 2.7743119266055047,
+      "grad_norm": 1.0390340089797974,
+      "learning_rate": 2.923732922474265e-07,
+      "loss": 0.2173,
+      "step": 5859
+    },
+    {
+      "epoch": 2.774785439479136,
+      "grad_norm": 1.056854248046875,
+      "learning_rate": 2.911469660360655e-07,
+      "loss": 0.2255,
+      "step": 5860
+    },
+    {
+      "epoch": 2.7752589523527673,
+      "grad_norm": 0.9877737164497375,
+      "learning_rate": 2.8992317908155263e-07,
+      "loss": 0.1992,
+      "step": 5861
+    },
+    {
+      "epoch": 2.7757324652263984,
+      "grad_norm": 1.1483983993530273,
+      "learning_rate": 2.8870193170395855e-07,
+      "loss": 0.2159,
+      "step": 5862
+    },
+    {
+      "epoch": 2.7762059781000294,
+      "grad_norm": 1.4070454835891724,
+      "learning_rate": 2.874832242226866e-07,
+      "loss": 0.2035,
+      "step": 5863
+    },
+    {
+      "epoch": 2.776679490973661,
+      "grad_norm": 0.9695731401443481,
+      "learning_rate": 2.862670569564796e-07,
+      "loss": 0.2004,
+      "step": 5864
+    },
+    {
+      "epoch": 2.777153003847292,
+      "grad_norm": 1.5455920696258545,
+      "learning_rate": 2.850534302234131e-07,
+      "loss": 0.1964,
+      "step": 5865
+    },
+    {
+      "epoch": 2.7776265167209235,
+      "grad_norm": 1.2403937578201294,
+      "learning_rate": 2.838423443409011e-07,
+      "loss": 0.22,
+      "step": 5866
+    },
+    {
+      "epoch": 2.7781000295945546,
+      "grad_norm": 1.3792948722839355,
+      "learning_rate": 2.8263379962568894e-07,
+      "loss": 0.2188,
+      "step": 5867
+    },
+    {
+      "epoch": 2.7785735424681857,
+      "grad_norm": 0.9905843138694763,
+      "learning_rate": 2.8142779639386275e-07,
+      "loss": 0.1821,
+      "step": 5868
+    },
+    {
+      "epoch": 2.779047055341817,
+      "grad_norm": 0.9501548409461975,
+      "learning_rate": 2.8022433496084024e-07,
+      "loss": 0.2016,
+      "step": 5869
+    },
+    {
+      "epoch": 2.7795205682154482,
+      "grad_norm": 0.990218460559845,
+      "learning_rate": 2.7902341564137294e-07,
+      "loss": 0.1896,
+      "step": 5870
+    },
+    {
+      "epoch": 2.7799940810890797,
+      "grad_norm": 0.9677179455757141,
+      "learning_rate": 2.778250387495529e-07,
+      "loss": 0.21,
+      "step": 5871
+    },
+    {
+      "epoch": 2.780467593962711,
+      "grad_norm": 1.4591187238693237,
+      "learning_rate": 2.766292045988006e-07,
+      "loss": 0.2014,
+      "step": 5872
+    },
+    {
+      "epoch": 2.780941106836342,
+      "grad_norm": 1.055142879486084,
+      "learning_rate": 2.754359135018791e-07,
+      "loss": 0.1694,
+      "step": 5873
+    },
+    {
+      "epoch": 2.7814146197099734,
+      "grad_norm": 1.0192638635635376,
+      "learning_rate": 2.742451657708778e-07,
+      "loss": 0.1877,
+      "step": 5874
+    },
+    {
+      "epoch": 2.7818881325836045,
+      "grad_norm": 1.215868353843689,
+      "learning_rate": 2.730569617172296e-07,
+      "loss": 0.1723,
+      "step": 5875
+    },
+    {
+      "epoch": 2.782361645457236,
+      "grad_norm": 1.0110760927200317,
+      "learning_rate": 2.7187130165169383e-07,
+      "loss": 0.2039,
+      "step": 5876
+    },
+    {
+      "epoch": 2.782835158330867,
+      "grad_norm": 1.146170735359192,
+      "learning_rate": 2.706881858843702e-07,
+      "loss": 0.1955,
+      "step": 5877
+    },
+    {
+      "epoch": 2.783308671204498,
+      "grad_norm": 0.9560944437980652,
+      "learning_rate": 2.695076147246911e-07,
+      "loss": 0.1919,
+      "step": 5878
+    },
+    {
+      "epoch": 2.7837821840781296,
+      "grad_norm": 1.2928173542022705,
+      "learning_rate": 2.683295884814252e-07,
+      "loss": 0.1987,
+      "step": 5879
+    },
+    {
+      "epoch": 2.784255696951761,
+      "grad_norm": 1.1882565021514893,
+      "learning_rate": 2.6715410746267155e-07,
+      "loss": 0.2041,
+      "step": 5880
+    },
+    {
+      "epoch": 2.784729209825392,
+      "grad_norm": 1.3022156953811646,
+      "learning_rate": 2.6598117197586646e-07,
+      "loss": 0.1777,
+      "step": 5881
+    },
+    {
+      "epoch": 2.7852027226990232,
+      "grad_norm": 1.2867157459259033,
+      "learning_rate": 2.6481078232778013e-07,
+      "loss": 0.1863,
+      "step": 5882
+    },
+    {
+      "epoch": 2.7856762355726548,
+      "grad_norm": 1.3041218519210815,
+      "learning_rate": 2.6364293882451655e-07,
+      "loss": 0.1834,
+      "step": 5883
+    },
+    {
+      "epoch": 2.786149748446286,
+      "grad_norm": 1.1009517908096313,
+      "learning_rate": 2.624776417715147e-07,
+      "loss": 0.1964,
+      "step": 5884
+    },
+    {
+      "epoch": 2.7866232613199173,
+      "grad_norm": 1.0136332511901855,
+      "learning_rate": 2.6131489147354527e-07,
+      "loss": 0.1895,
+      "step": 5885
+    },
+    {
+      "epoch": 2.7870967741935484,
+      "grad_norm": 1.4136160612106323,
+      "learning_rate": 2.60154688234715e-07,
+      "loss": 0.2031,
+      "step": 5886
+    },
+    {
+      "epoch": 2.7875702870671795,
+      "grad_norm": 1.1578394174575806,
+      "learning_rate": 2.589970323584645e-07,
+      "loss": 0.2127,
+      "step": 5887
+    },
+    {
+      "epoch": 2.788043799940811,
+      "grad_norm": 1.2432137727737427,
+      "learning_rate": 2.5784192414756714e-07,
+      "loss": 0.1898,
+      "step": 5888
+    },
+    {
+      "epoch": 2.788517312814442,
+      "grad_norm": 1.029590368270874,
+      "learning_rate": 2.5668936390413014e-07,
+      "loss": 0.1919,
+      "step": 5889
+    },
+    {
+      "epoch": 2.7889908256880735,
+      "grad_norm": 1.444937825202942,
+      "learning_rate": 2.5553935192959457e-07,
+      "loss": 0.2137,
+      "step": 5890
+    },
+    {
+      "epoch": 2.7894643385617046,
+      "grad_norm": 1.5030622482299805,
+      "learning_rate": 2.543918885247354e-07,
+      "loss": 0.1897,
+      "step": 5891
+    },
+    {
+      "epoch": 2.7899378514353357,
+      "grad_norm": 0.9852031469345093,
+      "learning_rate": 2.5324697398965927e-07,
+      "loss": 0.2015,
+      "step": 5892
+    },
+    {
+      "epoch": 2.790411364308967,
+      "grad_norm": 1.4873343706130981,
+      "learning_rate": 2.5210460862380993e-07,
+      "loss": 0.1801,
+      "step": 5893
+    },
+    {
+      "epoch": 2.7908848771825983,
+      "grad_norm": 1.5172432661056519,
+      "learning_rate": 2.5096479272596064e-07,
+      "loss": 0.2038,
+      "step": 5894
+    },
+    {
+      "epoch": 2.7913583900562298,
+      "grad_norm": 1.1348522901535034,
+      "learning_rate": 2.4982752659421736e-07,
+      "loss": 0.1754,
+      "step": 5895
+    },
+    {
+      "epoch": 2.791831902929861,
+      "grad_norm": 0.959349513053894,
+      "learning_rate": 2.4869281052602447e-07,
+      "loss": 0.1908,
+      "step": 5896
+    },
+    {
+      "epoch": 2.792305415803492,
+      "grad_norm": 1.4271739721298218,
+      "learning_rate": 2.4756064481815445e-07,
+      "loss": 0.1959,
+      "step": 5897
+    },
+    {
+      "epoch": 2.7927789286771234,
+      "grad_norm": 1.5378594398498535,
+      "learning_rate": 2.4643102976671383e-07,
+      "loss": 0.2231,
+      "step": 5898
+    },
+    {
+      "epoch": 2.793252441550755,
+      "grad_norm": 1.1620539426803589,
+      "learning_rate": 2.45303965667143e-07,
+      "loss": 0.2145,
+      "step": 5899
+    },
+    {
+      "epoch": 2.793725954424386,
+      "grad_norm": 1.0995843410491943,
+      "learning_rate": 2.4417945281421495e-07,
+      "loss": 0.204,
+      "step": 5900
+    },
+    {
+      "epoch": 2.794199467298017,
+      "grad_norm": 0.9468101859092712,
+      "learning_rate": 2.430574915020345e-07,
+      "loss": 0.1756,
+      "step": 5901
+    },
+    {
+      "epoch": 2.7946729801716486,
+      "grad_norm": 1.0497629642486572,
+      "learning_rate": 2.419380820240413e-07,
+      "loss": 0.2112,
+      "step": 5902
+    },
+    {
+      "epoch": 2.7951464930452796,
+      "grad_norm": 1.391728401184082,
+      "learning_rate": 2.408212246730035e-07,
+      "loss": 0.2024,
+      "step": 5903
+    },
+    {
+      "epoch": 2.795620005918911,
+      "grad_norm": 0.9138317108154297,
+      "learning_rate": 2.397069197410273e-07,
+      "loss": 0.2109,
+      "step": 5904
+    },
+    {
+      "epoch": 2.796093518792542,
+      "grad_norm": 0.9192736148834229,
+      "learning_rate": 2.3859516751954637e-07,
+      "loss": 0.2016,
+      "step": 5905
+    },
+    {
+      "epoch": 2.7965670316661733,
+      "grad_norm": 1.567175030708313,
+      "learning_rate": 2.3748596829932914e-07,
+      "loss": 0.2126,
+      "step": 5906
+    },
+    {
+      "epoch": 2.7970405445398048,
+      "grad_norm": 1.9273465871810913,
+      "learning_rate": 2.36379322370478e-07,
+      "loss": 0.2062,
+      "step": 5907
+    },
+    {
+      "epoch": 2.797514057413436,
+      "grad_norm": 1.1850330829620361,
+      "learning_rate": 2.3527523002242147e-07,
+      "loss": 0.2089,
+      "step": 5908
+    },
+    {
+      "epoch": 2.7979875702870673,
+      "grad_norm": 1.0225036144256592,
+      "learning_rate": 2.3417369154392854e-07,
+      "loss": 0.2057,
+      "step": 5909
+    },
+    {
+      "epoch": 2.7984610831606984,
+      "grad_norm": 0.9364516735076904,
+      "learning_rate": 2.330747072230921e-07,
+      "loss": 0.1966,
+      "step": 5910
+    },
+    {
+      "epoch": 2.7989345960343295,
+      "grad_norm": 1.2841675281524658,
+      "learning_rate": 2.3197827734734446e-07,
+      "loss": 0.1784,
+      "step": 5911
+    },
+    {
+      "epoch": 2.799408108907961,
+      "grad_norm": 1.0790499448776245,
+      "learning_rate": 2.3088440220344288e-07,
+      "loss": 0.1821,
+      "step": 5912
+    },
+    {
+      "epoch": 2.799881621781592,
+      "grad_norm": 1.2959651947021484,
+      "learning_rate": 2.2979308207748295e-07,
+      "loss": 0.1877,
+      "step": 5913
+    },
+    {
+      "epoch": 2.8003551346552236,
+      "grad_norm": 1.3200701475143433,
+      "learning_rate": 2.287043172548875e-07,
+      "loss": 0.2125,
+      "step": 5914
+    },
+    {
+      "epoch": 2.8008286475288546,
+      "grad_norm": 1.6399303674697876,
+      "learning_rate": 2.2761810802041205e-07,
+      "loss": 0.2024,
+      "step": 5915
+    },
+    {
+      "epoch": 2.8013021604024857,
+      "grad_norm": 1.029131293296814,
+      "learning_rate": 2.2653445465814493e-07,
+      "loss": 0.2218,
+      "step": 5916
+    },
+    {
+      "epoch": 2.801775673276117,
+      "grad_norm": 1.1296693086624146,
+      "learning_rate": 2.2545335745150387e-07,
+      "loss": 0.1993,
+      "step": 5917
+    },
+    {
+      "epoch": 2.8022491861497487,
+      "grad_norm": 1.231637954711914,
+      "learning_rate": 2.2437481668324268e-07,
+      "loss": 0.1831,
+      "step": 5918
+    },
+    {
+      "epoch": 2.80272269902338,
+      "grad_norm": 1.1062796115875244,
+      "learning_rate": 2.2329883263543905e-07,
+      "loss": 0.19,
+      "step": 5919
+    },
+    {
+      "epoch": 2.803196211897011,
+      "grad_norm": 0.9966011643409729,
+      "learning_rate": 2.22225405589509e-07,
+      "loss": 0.1985,
+      "step": 5920
+    },
+    {
+      "epoch": 2.8036697247706424,
+      "grad_norm": 0.9549378752708435,
+      "learning_rate": 2.2115453582619573e-07,
+      "loss": 0.1719,
+      "step": 5921
+    },
+    {
+      "epoch": 2.8041432376442734,
+      "grad_norm": 1.1001967191696167,
+      "learning_rate": 2.200862236255763e-07,
+      "loss": 0.2117,
+      "step": 5922
+    },
+    {
+      "epoch": 2.804616750517905,
+      "grad_norm": 1.617583990097046,
+      "learning_rate": 2.1902046926705611e-07,
+      "loss": 0.1837,
+      "step": 5923
+    },
+    {
+      "epoch": 2.805090263391536,
+      "grad_norm": 1.1403623819351196,
+      "learning_rate": 2.179572730293733e-07,
+      "loss": 0.2129,
+      "step": 5924
+    },
+    {
+      "epoch": 2.805563776265167,
+      "grad_norm": 1.1201438903808594,
+      "learning_rate": 2.1689663519059545e-07,
+      "loss": 0.1948,
+      "step": 5925
+    },
+    {
+      "epoch": 2.8060372891387986,
+      "grad_norm": 1.2484549283981323,
+      "learning_rate": 2.158385560281251e-07,
+      "loss": 0.197,
+      "step": 5926
+    },
+    {
+      "epoch": 2.8065108020124296,
+      "grad_norm": 1.149869441986084,
+      "learning_rate": 2.1478303581869087e-07,
+      "loss": 0.1816,
+      "step": 5927
+    },
+    {
+      "epoch": 2.806984314886061,
+      "grad_norm": 1.0050691366195679,
+      "learning_rate": 2.1373007483835306e-07,
+      "loss": 0.1902,
+      "step": 5928
+    },
+    {
+      "epoch": 2.807457827759692,
+      "grad_norm": 1.3803904056549072,
+      "learning_rate": 2.1267967336250472e-07,
+      "loss": 0.2059,
+      "step": 5929
+    },
+    {
+      "epoch": 2.8079313406333233,
+      "grad_norm": 1.0127631425857544,
+      "learning_rate": 2.1163183166586943e-07,
+      "loss": 0.1983,
+      "step": 5930
+    },
+    {
+      "epoch": 2.808404853506955,
+      "grad_norm": 1.1045966148376465,
+      "learning_rate": 2.1058655002249573e-07,
+      "loss": 0.2302,
+      "step": 5931
+    },
+    {
+      "epoch": 2.808878366380586,
+      "grad_norm": 1.2346396446228027,
+      "learning_rate": 2.0954382870577162e-07,
+      "loss": 0.193,
+      "step": 5932
+    },
+    {
+      "epoch": 2.8093518792542174,
+      "grad_norm": 1.0287126302719116,
+      "learning_rate": 2.0850366798840892e-07,
+      "loss": 0.2041,
+      "step": 5933
+    },
+    {
+      "epoch": 2.8098253921278484,
+      "grad_norm": 1.2126495838165283,
+      "learning_rate": 2.074660681424512e-07,
+      "loss": 0.198,
+      "step": 5934
+    },
+    {
+      "epoch": 2.8102989050014795,
+      "grad_norm": 0.9380123615264893,
+      "learning_rate": 2.064310294392724e-07,
+      "loss": 0.1903,
+      "step": 5935
+    },
+    {
+      "epoch": 2.810772417875111,
+      "grad_norm": 1.0035574436187744,
+      "learning_rate": 2.0539855214957828e-07,
+      "loss": 0.2078,
+      "step": 5936
+    },
+    {
+      "epoch": 2.811245930748742,
+      "grad_norm": 1.22806715965271,
+      "learning_rate": 2.043686365434028e-07,
+      "loss": 0.2211,
+      "step": 5937
+    },
+    {
+      "epoch": 2.8117194436223736,
+      "grad_norm": 1.3754026889801025,
+      "learning_rate": 2.0334128289011046e-07,
+      "loss": 0.1877,
+      "step": 5938
+    },
+    {
+      "epoch": 2.8121929564960046,
+      "grad_norm": 1.7814515829086304,
+      "learning_rate": 2.0231649145839528e-07,
+      "loss": 0.2241,
+      "step": 5939
+    },
+    {
+      "epoch": 2.8126664693696357,
+      "grad_norm": 1.2736109495162964,
+      "learning_rate": 2.012942625162817e-07,
+      "loss": 0.186,
+      "step": 5940
+    },
+    {
+      "epoch": 2.8131399822432672,
+      "grad_norm": 1.2346388101577759,
+      "learning_rate": 2.002745963311248e-07,
+      "loss": 0.1886,
+      "step": 5941
+    },
+    {
+      "epoch": 2.8136134951168987,
+      "grad_norm": 0.9936687350273132,
+      "learning_rate": 1.9925749316960563e-07,
+      "loss": 0.2186,
+      "step": 5942
+    },
+    {
+      "epoch": 2.81408700799053,
+      "grad_norm": 0.9427998065948486,
+      "learning_rate": 1.9824295329774145e-07,
+      "loss": 0.1811,
+      "step": 5943
+    },
+    {
+      "epoch": 2.814560520864161,
+      "grad_norm": 1.290932059288025,
+      "learning_rate": 1.9723097698087336e-07,
+      "loss": 0.1865,
+      "step": 5944
+    },
+    {
+      "epoch": 2.8150340337377924,
+      "grad_norm": 1.3386746644973755,
+      "learning_rate": 1.9622156448367403e-07,
+      "loss": 0.1974,
+      "step": 5945
+    },
+    {
+      "epoch": 2.8155075466114234,
+      "grad_norm": 0.9645776748657227,
+      "learning_rate": 1.9521471607014565e-07,
+      "loss": 0.1782,
+      "step": 5946
+    },
+    {
+      "epoch": 2.815981059485055,
+      "grad_norm": 1.2050148248672485,
+      "learning_rate": 1.9421043200361976e-07,
+      "loss": 0.1887,
+      "step": 5947
+    },
+    {
+      "epoch": 2.816454572358686,
+      "grad_norm": 1.0068618059158325,
+      "learning_rate": 1.9320871254675745e-07,
+      "loss": 0.2107,
+      "step": 5948
+    },
+    {
+      "epoch": 2.816928085232317,
+      "grad_norm": 2.5484302043914795,
+      "learning_rate": 1.9220955796154794e-07,
+      "loss": 0.1889,
+      "step": 5949
+    },
+    {
+      "epoch": 2.8174015981059486,
+      "grad_norm": 1.0873286724090576,
+      "learning_rate": 1.9121296850931225e-07,
+      "loss": 0.2036,
+      "step": 5950
+    },
+    {
+      "epoch": 2.8178751109795797,
+      "grad_norm": 1.162097454071045,
+      "learning_rate": 1.9021894445069744e-07,
+      "loss": 0.2023,
+      "step": 5951
+    },
+    {
+      "epoch": 2.818348623853211,
+      "grad_norm": 1.1545616388320923,
+      "learning_rate": 1.892274860456811e-07,
+      "loss": 0.2182,
+      "step": 5952
+    },
+    {
+      "epoch": 2.8188221367268422,
+      "grad_norm": 1.2992243766784668,
+      "learning_rate": 1.8823859355356798e-07,
+      "loss": 0.1967,
+      "step": 5953
+    },
+    {
+      "epoch": 2.8192956496004733,
+      "grad_norm": 0.9885454177856445,
+      "learning_rate": 1.8725226723299682e-07,
+      "loss": 0.1857,
+      "step": 5954
+    },
+    {
+      "epoch": 2.819769162474105,
+      "grad_norm": 1.3030709028244019,
+      "learning_rate": 1.8626850734192904e-07,
+      "loss": 0.1846,
+      "step": 5955
+    },
+    {
+      "epoch": 2.820242675347736,
+      "grad_norm": 1.0009846687316895,
+      "learning_rate": 1.8528731413765877e-07,
+      "loss": 0.2014,
+      "step": 5956
+    },
+    {
+      "epoch": 2.8207161882213674,
+      "grad_norm": 1.367098331451416,
+      "learning_rate": 1.8430868787680633e-07,
+      "loss": 0.2086,
+      "step": 5957
+    },
+    {
+      "epoch": 2.8211897010949984,
+      "grad_norm": 1.151480793952942,
+      "learning_rate": 1.8333262881532476e-07,
+      "loss": 0.2091,
+      "step": 5958
+    },
+    {
+      "epoch": 2.8216632139686295,
+      "grad_norm": 1.2363460063934326,
+      "learning_rate": 1.8235913720848985e-07,
+      "loss": 0.2001,
+      "step": 5959
+    },
+    {
+      "epoch": 2.822136726842261,
+      "grad_norm": 1.2438163757324219,
+      "learning_rate": 1.8138821331091015e-07,
+      "loss": 0.2125,
+      "step": 5960
+    },
+    {
+      "epoch": 2.8226102397158925,
+      "grad_norm": 1.2525054216384888,
+      "learning_rate": 1.8041985737652256e-07,
+      "loss": 0.2103,
+      "step": 5961
+    },
+    {
+      "epoch": 2.8230837525895236,
+      "grad_norm": 1.3402167558670044,
+      "learning_rate": 1.7945406965858892e-07,
+      "loss": 0.1944,
+      "step": 5962
+    },
+    {
+      "epoch": 2.8235572654631547,
+      "grad_norm": 1.0227385759353638,
+      "learning_rate": 1.784908504097027e-07,
+      "loss": 0.2063,
+      "step": 5963
+    },
+    {
+      "epoch": 2.824030778336786,
+      "grad_norm": 1.5239243507385254,
+      "learning_rate": 1.7753019988178577e-07,
+      "loss": 0.1862,
+      "step": 5964
+    },
+    {
+      "epoch": 2.8245042912104172,
+      "grad_norm": 1.001605749130249,
+      "learning_rate": 1.765721183260849e-07,
+      "loss": 0.1981,
+      "step": 5965
+    },
+    {
+      "epoch": 2.8249778040840487,
+      "grad_norm": 1.0575411319732666,
+      "learning_rate": 1.7561660599317853e-07,
+      "loss": 0.2199,
+      "step": 5966
+    },
+    {
+      "epoch": 2.82545131695768,
+      "grad_norm": 1.230980396270752,
+      "learning_rate": 1.7466366313297123e-07,
+      "loss": 0.2134,
+      "step": 5967
+    },
+    {
+      "epoch": 2.825924829831311,
+      "grad_norm": 1.0976228713989258,
+      "learning_rate": 1.7371328999469695e-07,
+      "loss": 0.2227,
+      "step": 5968
+    },
+    {
+      "epoch": 2.8263983427049424,
+      "grad_norm": 1.2216429710388184,
+      "learning_rate": 1.7276548682691463e-07,
+      "loss": 0.1866,
+      "step": 5969
+    },
+    {
+      "epoch": 2.8268718555785735,
+      "grad_norm": 1.946909785270691,
+      "learning_rate": 1.71820253877516e-07,
+      "loss": 0.2028,
+      "step": 5970
+    },
+    {
+      "epoch": 2.827345368452205,
+      "grad_norm": 1.2762550115585327,
+      "learning_rate": 1.7087759139371442e-07,
+      "loss": 0.1854,
+      "step": 5971
+    },
+    {
+      "epoch": 2.827818881325836,
+      "grad_norm": 1.4644784927368164,
+      "learning_rate": 1.6993749962205597e-07,
+      "loss": 0.1905,
+      "step": 5972
+    },
+    {
+      "epoch": 2.828292394199467,
+      "grad_norm": 1.5430243015289307,
+      "learning_rate": 1.689999788084129e-07,
+      "loss": 0.2004,
+      "step": 5973
+    },
+    {
+      "epoch": 2.8287659070730986,
+      "grad_norm": 1.2102375030517578,
+      "learning_rate": 1.680650291979824e-07,
+      "loss": 0.1786,
+      "step": 5974
+    },
+    {
+      "epoch": 2.8292394199467297,
+      "grad_norm": 1.029317855834961,
+      "learning_rate": 1.671326510352944e-07,
+      "loss": 0.1895,
+      "step": 5975
+    },
+    {
+      "epoch": 2.829712932820361,
+      "grad_norm": 1.2381408214569092,
+      "learning_rate": 1.6620284456420167e-07,
+      "loss": 0.2229,
+      "step": 5976
+    },
+    {
+      "epoch": 2.8301864456939922,
+      "grad_norm": 0.9947028160095215,
+      "learning_rate": 1.652756100278874e-07,
+      "loss": 0.204,
+      "step": 5977
+    },
+    {
+      "epoch": 2.8306599585676233,
+      "grad_norm": 1.2137975692749023,
+      "learning_rate": 1.6435094766885873e-07,
+      "loss": 0.2174,
+      "step": 5978
+    },
+    {
+      "epoch": 2.831133471441255,
+      "grad_norm": 1.7809674739837646,
+      "learning_rate": 1.6342885772895445e-07,
+      "loss": 0.1848,
+      "step": 5979
+    },
+    {
+      "epoch": 2.8316069843148863,
+      "grad_norm": 1.0932118892669678,
+      "learning_rate": 1.625093404493372e-07,
+      "loss": 0.182,
+      "step": 5980
+    },
+    {
+      "epoch": 2.8320804971885174,
+      "grad_norm": 1.009130835533142,
+      "learning_rate": 1.6159239607049793e-07,
+      "loss": 0.187,
+      "step": 5981
+    },
+    {
+      "epoch": 2.8325540100621485,
+      "grad_norm": 1.0555768013000488,
+      "learning_rate": 1.606780248322548e-07,
+      "loss": 0.218,
+      "step": 5982
+    },
+    {
+      "epoch": 2.83302752293578,
+      "grad_norm": 1.0417250394821167,
+      "learning_rate": 1.597662269737521e-07,
+      "loss": 0.2098,
+      "step": 5983
+    },
+    {
+      "epoch": 2.833501035809411,
+      "grad_norm": 1.1635463237762451,
+      "learning_rate": 1.588570027334635e-07,
+      "loss": 0.2166,
+      "step": 5984
+    },
+    {
+      "epoch": 2.8339745486830425,
+      "grad_norm": 0.8822041749954224,
+      "learning_rate": 1.5795035234918543e-07,
+      "loss": 0.1992,
+      "step": 5985
+    },
+    {
+      "epoch": 2.8344480615566736,
+      "grad_norm": 1.387373685836792,
+      "learning_rate": 1.5704627605804601e-07,
+      "loss": 0.2067,
+      "step": 5986
+    },
+    {
+      "epoch": 2.8349215744303047,
+      "grad_norm": 1.6232830286026,
+      "learning_rate": 1.5614477409649497e-07,
+      "loss": 0.1954,
+      "step": 5987
+    },
+    {
+      "epoch": 2.835395087303936,
+      "grad_norm": 1.3384531736373901,
+      "learning_rate": 1.5524584670031372e-07,
+      "loss": 0.2028,
+      "step": 5988
+    },
+    {
+      "epoch": 2.8358686001775673,
+      "grad_norm": 1.060403823852539,
+      "learning_rate": 1.5434949410460753e-07,
+      "loss": 0.1862,
+      "step": 5989
+    },
+    {
+      "epoch": 2.8363421130511988,
+      "grad_norm": 1.383388876914978,
+      "learning_rate": 1.5345571654380775e-07,
+      "loss": 0.2104,
+      "step": 5990
+    },
+    {
+      "epoch": 2.83681562592483,
+      "grad_norm": 1.2764064073562622,
+      "learning_rate": 1.5256451425167406e-07,
+      "loss": 0.2042,
+      "step": 5991
+    },
+    {
+      "epoch": 2.837289138798461,
+      "grad_norm": 1.2864024639129639,
+      "learning_rate": 1.5167588746129224e-07,
+      "loss": 0.209,
+      "step": 5992
+    },
+    {
+      "epoch": 2.8377626516720924,
+      "grad_norm": 0.9310190081596375,
+      "learning_rate": 1.5078983640507416e-07,
+      "loss": 0.1899,
+      "step": 5993
+    },
+    {
+      "epoch": 2.8382361645457235,
+      "grad_norm": 1.218672752380371,
+      "learning_rate": 1.4990636131475554e-07,
+      "loss": 0.1884,
+      "step": 5994
+    },
+    {
+      "epoch": 2.838709677419355,
+      "grad_norm": 1.2877182960510254,
+      "learning_rate": 1.490254624214049e-07,
+      "loss": 0.1916,
+      "step": 5995
+    },
+    {
+      "epoch": 2.839183190292986,
+      "grad_norm": 1.661100149154663,
+      "learning_rate": 1.4814713995540797e-07,
+      "loss": 0.1996,
+      "step": 5996
+    },
+    {
+      "epoch": 2.839656703166617,
+      "grad_norm": 1.07488214969635,
+      "learning_rate": 1.472713941464865e-07,
+      "loss": 0.2045,
+      "step": 5997
+    },
+    {
+      "epoch": 2.8401302160402486,
+      "grad_norm": 1.1992322206497192,
+      "learning_rate": 1.4639822522367952e-07,
+      "loss": 0.1849,
+      "step": 5998
+    },
+    {
+      "epoch": 2.8406037289138797,
+      "grad_norm": 1.5754666328430176,
+      "learning_rate": 1.455276334153577e-07,
+      "loss": 0.21,
+      "step": 5999
+    },
+    {
+      "epoch": 2.841077241787511,
+      "grad_norm": 1.114166021347046,
+      "learning_rate": 1.4465961894921555e-07,
+      "loss": 0.2,
+      "step": 6000
+    },
+    {
+      "epoch": 2.8415507546611423,
+      "grad_norm": 1.2283138036727905,
+      "learning_rate": 1.4379418205227368e-07,
+      "loss": 0.1895,
+      "step": 6001
+    },
+    {
+      "epoch": 2.8420242675347733,
+      "grad_norm": 1.11212158203125,
+      "learning_rate": 1.4293132295087998e-07,
+      "loss": 0.205,
+      "step": 6002
+    },
+    {
+      "epoch": 2.842497780408405,
+      "grad_norm": 1.7816588878631592,
+      "learning_rate": 1.4207104187070386e-07,
+      "loss": 0.235,
+      "step": 6003
+    },
+    {
+      "epoch": 2.8429712932820363,
+      "grad_norm": 1.1824668645858765,
+      "learning_rate": 1.4121333903674762e-07,
+      "loss": 0.2125,
+      "step": 6004
+    },
+    {
+      "epoch": 2.8434448061556674,
+      "grad_norm": 1.4136723279953003,
+      "learning_rate": 1.4035821467333177e-07,
+      "loss": 0.2042,
+      "step": 6005
+    },
+    {
+      "epoch": 2.8439183190292985,
+      "grad_norm": 1.4217491149902344,
+      "learning_rate": 1.3950566900410856e-07,
+      "loss": 0.1737,
+      "step": 6006
+    },
+    {
+      "epoch": 2.84439183190293,
+      "grad_norm": 1.3726348876953125,
+      "learning_rate": 1.3865570225205072e-07,
+      "loss": 0.2252,
+      "step": 6007
+    },
+    {
+      "epoch": 2.844865344776561,
+      "grad_norm": 0.8112472891807556,
+      "learning_rate": 1.3780831463946042e-07,
+      "loss": 0.1636,
+      "step": 6008
+    },
+    {
+      "epoch": 2.8453388576501926,
+      "grad_norm": 1.479446530342102,
+      "learning_rate": 1.3696350638796263e-07,
+      "loss": 0.2228,
+      "step": 6009
+    },
+    {
+      "epoch": 2.8458123705238236,
+      "grad_norm": 1.054337501525879,
+      "learning_rate": 1.3612127771850947e-07,
+      "loss": 0.2159,
+      "step": 6010
+    },
+    {
+      "epoch": 2.8462858833974547,
+      "grad_norm": 1.3831230401992798,
+      "learning_rate": 1.3528162885137919e-07,
+      "loss": 0.1956,
+      "step": 6011
+    },
+    {
+      "epoch": 2.846759396271086,
+      "grad_norm": 0.9735446572303772,
+      "learning_rate": 1.3444456000617056e-07,
+      "loss": 0.1992,
+      "step": 6012
+    },
+    {
+      "epoch": 2.8472329091447173,
+      "grad_norm": 1.2813206911087036,
+      "learning_rate": 1.3361007140181293e-07,
+      "loss": 0.1998,
+      "step": 6013
+    },
+    {
+      "epoch": 2.847706422018349,
+      "grad_norm": 1.0976970195770264,
+      "learning_rate": 1.3277816325655835e-07,
+      "loss": 0.1954,
+      "step": 6014
+    },
+    {
+      "epoch": 2.84817993489198,
+      "grad_norm": 1.1628068685531616,
+      "learning_rate": 1.319488357879839e-07,
+      "loss": 0.2002,
+      "step": 6015
+    },
+    {
+      "epoch": 2.848653447765611,
+      "grad_norm": 1.114010214805603,
+      "learning_rate": 1.3112208921299274e-07,
+      "loss": 0.1749,
+      "step": 6016
+    },
+    {
+      "epoch": 2.8491269606392424,
+      "grad_norm": 1.5268833637237549,
+      "learning_rate": 1.3029792374781413e-07,
+      "loss": 0.2106,
+      "step": 6017
+    },
+    {
+      "epoch": 2.8496004735128735,
+      "grad_norm": 0.991341769695282,
+      "learning_rate": 1.294763396079979e-07,
+      "loss": 0.1836,
+      "step": 6018
+    },
+    {
+      "epoch": 2.850073986386505,
+      "grad_norm": 1.0029536485671997,
+      "learning_rate": 1.2865733700842098e-07,
+      "loss": 0.2071,
+      "step": 6019
+    },
+    {
+      "epoch": 2.850547499260136,
+      "grad_norm": 2.5262866020202637,
+      "learning_rate": 1.2784091616328876e-07,
+      "loss": 0.1866,
+      "step": 6020
+    },
+    {
+      "epoch": 2.851021012133767,
+      "grad_norm": 1.2454925775527954,
+      "learning_rate": 1.2702707728612596e-07,
+      "loss": 0.175,
+      "step": 6021
+    },
+    {
+      "epoch": 2.8514945250073986,
+      "grad_norm": 1.1130245923995972,
+      "learning_rate": 1.2621582058978455e-07,
+      "loss": 0.196,
+      "step": 6022
+    },
+    {
+      "epoch": 2.85196803788103,
+      "grad_norm": 1.1060099601745605,
+      "learning_rate": 1.2540714628644146e-07,
+      "loss": 0.191,
+      "step": 6023
+    },
+    {
+      "epoch": 2.852441550754661,
+      "grad_norm": 0.9595826268196106,
+      "learning_rate": 1.2460105458759753e-07,
+      "loss": 0.207,
+      "step": 6024
+    },
+    {
+      "epoch": 2.8529150636282923,
+      "grad_norm": 1.54879891872406,
+      "learning_rate": 1.2379754570407742e-07,
+      "loss": 0.2039,
+      "step": 6025
+    },
+    {
+      "epoch": 2.853388576501924,
+      "grad_norm": 1.1590107679367065,
+      "learning_rate": 1.2299661984603307e-07,
+      "loss": 0.2101,
+      "step": 6026
+    },
+    {
+      "epoch": 2.853862089375555,
+      "grad_norm": 1.0808899402618408,
+      "learning_rate": 1.2219827722293687e-07,
+      "loss": 0.1841,
+      "step": 6027
+    },
+    {
+      "epoch": 2.8543356022491864,
+      "grad_norm": 1.6629842519760132,
+      "learning_rate": 1.214025180435885e-07,
+      "loss": 0.2059,
+      "step": 6028
+    },
+    {
+      "epoch": 2.8548091151228174,
+      "grad_norm": 0.9478086829185486,
+      "learning_rate": 1.2060934251611146e-07,
+      "loss": 0.1861,
+      "step": 6029
+    },
+    {
+      "epoch": 2.8552826279964485,
+      "grad_norm": 1.5418405532836914,
+      "learning_rate": 1.1981875084795202e-07,
+      "loss": 0.1955,
+      "step": 6030
+    },
+    {
+      "epoch": 2.85575614087008,
+      "grad_norm": 1.2840735912322998,
+      "learning_rate": 1.190307432458826e-07,
+      "loss": 0.2217,
+      "step": 6031
+    },
+    {
+      "epoch": 2.856229653743711,
+      "grad_norm": 0.9639061689376831,
+      "learning_rate": 1.1824531991599831e-07,
+      "loss": 0.1806,
+      "step": 6032
+    },
+    {
+      "epoch": 2.8567031666173426,
+      "grad_norm": 1.2272212505340576,
+      "learning_rate": 1.1746248106372149e-07,
+      "loss": 0.1941,
+      "step": 6033
+    },
+    {
+      "epoch": 2.8571766794909736,
+      "grad_norm": 1.424233078956604,
+      "learning_rate": 1.1668222689379172e-07,
+      "loss": 0.1933,
+      "step": 6034
+    },
+    {
+      "epoch": 2.8576501923646047,
+      "grad_norm": 1.2118334770202637,
+      "learning_rate": 1.159045576102813e-07,
+      "loss": 0.2095,
+      "step": 6035
+    },
+    {
+      "epoch": 2.8581237052382362,
+      "grad_norm": 1.1151028871536255,
+      "learning_rate": 1.1512947341657976e-07,
+      "loss": 0.1909,
+      "step": 6036
+    },
+    {
+      "epoch": 2.8585972181118673,
+      "grad_norm": 1.1803522109985352,
+      "learning_rate": 1.1435697451540385e-07,
+      "loss": 0.1931,
+      "step": 6037
+    },
+    {
+      "epoch": 2.859070730985499,
+      "grad_norm": 1.8160289525985718,
+      "learning_rate": 1.1358706110879302e-07,
+      "loss": 0.2031,
+      "step": 6038
+    },
+    {
+      "epoch": 2.85954424385913,
+      "grad_norm": 1.0799731016159058,
+      "learning_rate": 1.1281973339810847e-07,
+      "loss": 0.1992,
+      "step": 6039
+    },
+    {
+      "epoch": 2.860017756732761,
+      "grad_norm": 1.1362674236297607,
+      "learning_rate": 1.1205499158404187e-07,
+      "loss": 0.1917,
+      "step": 6040
+    },
+    {
+      "epoch": 2.8604912696063924,
+      "grad_norm": 2.0242366790771484,
+      "learning_rate": 1.1129283586659988e-07,
+      "loss": 0.2059,
+      "step": 6041
+    },
+    {
+      "epoch": 2.860964782480024,
+      "grad_norm": 0.9954387545585632,
+      "learning_rate": 1.1053326644511863e-07,
+      "loss": 0.1961,
+      "step": 6042
+    },
+    {
+      "epoch": 2.861438295353655,
+      "grad_norm": 1.4265193939208984,
+      "learning_rate": 1.0977628351825697e-07,
+      "loss": 0.2002,
+      "step": 6043
+    },
+    {
+      "epoch": 2.861911808227286,
+      "grad_norm": 1.1525962352752686,
+      "learning_rate": 1.0902188728399433e-07,
+      "loss": 0.2124,
+      "step": 6044
+    },
+    {
+      "epoch": 2.8623853211009176,
+      "grad_norm": 0.9790692925453186,
+      "learning_rate": 1.082700779396384e-07,
+      "loss": 0.1971,
+      "step": 6045
+    },
+    {
+      "epoch": 2.8628588339745487,
+      "grad_norm": 1.0439430475234985,
+      "learning_rate": 1.0752085568181524e-07,
+      "loss": 0.191,
+      "step": 6046
+    },
+    {
+      "epoch": 2.86333234684818,
+      "grad_norm": 0.9470370411872864,
+      "learning_rate": 1.0677422070647924e-07,
+      "loss": 0.1869,
+      "step": 6047
+    },
+    {
+      "epoch": 2.8638058597218112,
+      "grad_norm": 0.9897739291191101,
+      "learning_rate": 1.0603017320890307e-07,
+      "loss": 0.1836,
+      "step": 6048
+    },
+    {
+      "epoch": 2.8642793725954423,
+      "grad_norm": 0.8592694997787476,
+      "learning_rate": 1.0528871338368773e-07,
+      "loss": 0.1808,
+      "step": 6049
+    },
+    {
+      "epoch": 2.864752885469074,
+      "grad_norm": 1.6525866985321045,
+      "learning_rate": 1.0454984142475145e-07,
+      "loss": 0.2002,
+      "step": 6050
+    },
+    {
+      "epoch": 2.865226398342705,
+      "grad_norm": 1.4348740577697754,
+      "learning_rate": 1.0381355752534295e-07,
+      "loss": 0.179,
+      "step": 6051
+    },
+    {
+      "epoch": 2.8656999112163364,
+      "grad_norm": 0.9945125579833984,
+      "learning_rate": 1.0307986187802709e-07,
+      "loss": 0.1853,
+      "step": 6052
+    },
+    {
+      "epoch": 2.8661734240899674,
+      "grad_norm": 1.0377004146575928,
+      "learning_rate": 1.0234875467469707e-07,
+      "loss": 0.1957,
+      "step": 6053
+    },
+    {
+      "epoch": 2.8666469369635985,
+      "grad_norm": 1.3555017709732056,
+      "learning_rate": 1.0162023610656547e-07,
+      "loss": 0.177,
+      "step": 6054
+    },
+    {
+      "epoch": 2.86712044983723,
+      "grad_norm": 1.3089735507965088,
+      "learning_rate": 1.0089430636416875e-07,
+      "loss": 0.1857,
+      "step": 6055
+    },
+    {
+      "epoch": 2.867593962710861,
+      "grad_norm": 1.1990454196929932,
+      "learning_rate": 1.001709656373695e-07,
+      "loss": 0.1987,
+      "step": 6056
+    },
+    {
+      "epoch": 2.8680674755844926,
+      "grad_norm": 0.9925453066825867,
+      "learning_rate": 9.94502141153475e-08,
+      "loss": 0.1858,
+      "step": 6057
+    },
+    {
+      "epoch": 2.8685409884581237,
+      "grad_norm": 1.205861210823059,
+      "learning_rate": 9.873205198660974e-08,
+      "loss": 0.182,
+      "step": 6058
+    },
+    {
+      "epoch": 2.8690145013317547,
+      "grad_norm": 1.1431076526641846,
+      "learning_rate": 9.801647943898484e-08,
+      "loss": 0.1988,
+      "step": 6059
+    },
+    {
+      "epoch": 2.8694880142053862,
+      "grad_norm": 1.1238492727279663,
+      "learning_rate": 9.730349665962424e-08,
+      "loss": 0.2117,
+      "step": 6060
+    },
+    {
+      "epoch": 2.8699615270790173,
+      "grad_norm": 1.089830994606018,
+      "learning_rate": 9.659310383499986e-08,
+      "loss": 0.2003,
+      "step": 6061
+    },
+    {
+      "epoch": 2.870435039952649,
+      "grad_norm": 1.0016067028045654,
+      "learning_rate": 9.588530115091088e-08,
+      "loss": 0.1881,
+      "step": 6062
+    },
+    {
+      "epoch": 2.87090855282628,
+      "grad_norm": 1.2619209289550781,
+      "learning_rate": 9.518008879247365e-08,
+      "loss": 0.2057,
+      "step": 6063
+    },
+    {
+      "epoch": 2.871382065699911,
+      "grad_norm": 1.600156307220459,
+      "learning_rate": 9.447746694413063e-08,
+      "loss": 0.1807,
+      "step": 6064
+    },
+    {
+      "epoch": 2.8718555785735425,
+      "grad_norm": 1.0998058319091797,
+      "learning_rate": 9.377743578964704e-08,
+      "loss": 0.2085,
+      "step": 6065
+    },
+    {
+      "epoch": 2.872329091447174,
+      "grad_norm": 1.203202486038208,
+      "learning_rate": 9.307999551210645e-08,
+      "loss": 0.1866,
+      "step": 6066
+    },
+    {
+      "epoch": 2.872802604320805,
+      "grad_norm": 1.0956839323043823,
+      "learning_rate": 9.23851462939207e-08,
+      "loss": 0.2027,
+      "step": 6067
+    },
+    {
+      "epoch": 2.873276117194436,
+      "grad_norm": 1.0432785749435425,
+      "learning_rate": 9.169288831681889e-08,
+      "loss": 0.1942,
+      "step": 6068
+    },
+    {
+      "epoch": 2.8737496300680676,
+      "grad_norm": 1.4329001903533936,
+      "learning_rate": 9.100322176185505e-08,
+      "loss": 0.1827,
+      "step": 6069
+    },
+    {
+      "epoch": 2.8742231429416987,
+      "grad_norm": 0.974470317363739,
+      "learning_rate": 9.031614680940381e-08,
+      "loss": 0.2002,
+      "step": 6070
+    },
+    {
+      "epoch": 2.87469665581533,
+      "grad_norm": 1.5563040971755981,
+      "learning_rate": 8.963166363916586e-08,
+      "loss": 0.188,
+      "step": 6071
+    },
+    {
+      "epoch": 2.8751701686889612,
+      "grad_norm": 1.2367172241210938,
+      "learning_rate": 8.894977243015801e-08,
+      "loss": 0.2312,
+      "step": 6072
+    },
+    {
+      "epoch": 2.8756436815625923,
+      "grad_norm": 1.5719220638275146,
+      "learning_rate": 8.827047336072426e-08,
+      "loss": 0.1985,
+      "step": 6073
+    },
+    {
+      "epoch": 2.876117194436224,
+      "grad_norm": 1.091109037399292,
+      "learning_rate": 8.759376660852803e-08,
+      "loss": 0.1942,
+      "step": 6074
+    },
+    {
+      "epoch": 2.876590707309855,
+      "grad_norm": 0.9197636842727661,
+      "learning_rate": 8.691965235055444e-08,
+      "loss": 0.1919,
+      "step": 6075
+    },
+    {
+      "epoch": 2.8770642201834864,
+      "grad_norm": 1.4586553573608398,
+      "learning_rate": 8.624813076311356e-08,
+      "loss": 0.2025,
+      "step": 6076
+    },
+    {
+      "epoch": 2.8775377330571175,
+      "grad_norm": 1.1022119522094727,
+      "learning_rate": 8.557920202183379e-08,
+      "loss": 0.2004,
+      "step": 6077
+    },
+    {
+      "epoch": 2.8780112459307485,
+      "grad_norm": 1.0540403127670288,
+      "learning_rate": 8.491286630166851e-08,
+      "loss": 0.2178,
+      "step": 6078
+    },
+    {
+      "epoch": 2.87848475880438,
+      "grad_norm": 1.0437299013137817,
+      "learning_rate": 8.424912377688943e-08,
+      "loss": 0.1937,
+      "step": 6079
+    },
+    {
+      "epoch": 2.878958271678011,
+      "grad_norm": 1.272997498512268,
+      "learning_rate": 8.358797462109325e-08,
+      "loss": 0.186,
+      "step": 6080
+    },
+    {
+      "epoch": 2.8794317845516426,
+      "grad_norm": 1.2749963998794556,
+      "learning_rate": 8.29294190071972e-08,
+      "loss": 0.2155,
+      "step": 6081
+    },
+    {
+      "epoch": 2.8799052974252737,
+      "grad_norm": 0.9653557538986206,
+      "learning_rate": 8.227345710744018e-08,
+      "loss": 0.1891,
+      "step": 6082
+    },
+    {
+      "epoch": 2.8803788102989047,
+      "grad_norm": 1.1792298555374146,
+      "learning_rate": 8.16200890933827e-08,
+      "loss": 0.1945,
+      "step": 6083
+    },
+    {
+      "epoch": 2.8808523231725363,
+      "grad_norm": 1.3401811122894287,
+      "learning_rate": 8.096931513590589e-08,
+      "loss": 0.1901,
+      "step": 6084
+    },
+    {
+      "epoch": 2.8813258360461678,
+      "grad_norm": 0.9627504944801331,
+      "learning_rate": 8.03211354052147e-08,
+      "loss": 0.2,
+      "step": 6085
+    },
+    {
+      "epoch": 2.881799348919799,
+      "grad_norm": 1.163562536239624,
+      "learning_rate": 7.967555007083239e-08,
+      "loss": 0.1979,
+      "step": 6086
+    },
+    {
+      "epoch": 2.88227286179343,
+      "grad_norm": 1.035914421081543,
+      "learning_rate": 7.903255930160836e-08,
+      "loss": 0.1956,
+      "step": 6087
+    },
+    {
+      "epoch": 2.8827463746670614,
+      "grad_norm": 1.300705075263977,
+      "learning_rate": 7.83921632657092e-08,
+      "loss": 0.2012,
+      "step": 6088
+    },
+    {
+      "epoch": 2.8832198875406925,
+      "grad_norm": 1.1752310991287231,
+      "learning_rate": 7.775436213062426e-08,
+      "loss": 0.1685,
+      "step": 6089
+    },
+    {
+      "epoch": 2.883693400414324,
+      "grad_norm": 0.986862301826477,
+      "learning_rate": 7.711915606316345e-08,
+      "loss": 0.1726,
+      "step": 6090
+    },
+    {
+      "epoch": 2.884166913287955,
+      "grad_norm": 1.0843511819839478,
+      "learning_rate": 7.648654522946053e-08,
+      "loss": 0.1928,
+      "step": 6091
+    },
+    {
+      "epoch": 2.884640426161586,
+      "grad_norm": 1.3774464130401611,
+      "learning_rate": 7.58565297949676e-08,
+      "loss": 0.1992,
+      "step": 6092
+    },
+    {
+      "epoch": 2.8851139390352176,
+      "grad_norm": 0.9635332822799683,
+      "learning_rate": 7.522910992445842e-08,
+      "loss": 0.2011,
+      "step": 6093
+    },
+    {
+      "epoch": 2.8855874519088487,
+      "grad_norm": 0.9744061827659607,
+      "learning_rate": 7.46042857820306e-08,
+      "loss": 0.1843,
+      "step": 6094
+    },
+    {
+      "epoch": 2.88606096478248,
+      "grad_norm": 1.0627609491348267,
+      "learning_rate": 7.39820575311001e-08,
+      "loss": 0.2044,
+      "step": 6095
+    },
+    {
+      "epoch": 2.8865344776561113,
+      "grad_norm": 1.7792444229125977,
+      "learning_rate": 7.33624253344034e-08,
+      "loss": 0.1901,
+      "step": 6096
+    },
+    {
+      "epoch": 2.8870079905297423,
+      "grad_norm": 1.1715210676193237,
+      "learning_rate": 7.274538935400199e-08,
+      "loss": 0.2155,
+      "step": 6097
+    },
+    {
+      "epoch": 2.887481503403374,
+      "grad_norm": 1.223122000694275,
+      "learning_rate": 7.213094975127233e-08,
+      "loss": 0.1888,
+      "step": 6098
+    },
+    {
+      "epoch": 2.887955016277005,
+      "grad_norm": 1.16334068775177,
+      "learning_rate": 7.151910668691808e-08,
+      "loss": 0.209,
+      "step": 6099
+    },
+    {
+      "epoch": 2.8884285291506364,
+      "grad_norm": 1.090516209602356,
+      "learning_rate": 7.090986032095903e-08,
+      "loss": 0.1873,
+      "step": 6100
+    },
+    {
+      "epoch": 2.8889020420242675,
+      "grad_norm": 0.9286671876907349,
+      "learning_rate": 7.030321081273883e-08,
+      "loss": 0.1888,
+      "step": 6101
+    },
+    {
+      "epoch": 2.8893755548978985,
+      "grad_norm": 0.9373717904090881,
+      "learning_rate": 6.969915832092056e-08,
+      "loss": 0.1875,
+      "step": 6102
+    },
+    {
+      "epoch": 2.88984906777153,
+      "grad_norm": 1.0975648164749146,
+      "learning_rate": 6.909770300348784e-08,
+      "loss": 0.1918,
+      "step": 6103
+    },
+    {
+      "epoch": 2.8903225806451616,
+      "grad_norm": 1.237648844718933,
+      "learning_rate": 6.849884501774484e-08,
+      "loss": 0.2088,
+      "step": 6104
+    },
+    {
+      "epoch": 2.8907960935187926,
+      "grad_norm": 1.114311933517456,
+      "learning_rate": 6.790258452031962e-08,
+      "loss": 0.1929,
+      "step": 6105
+    },
+    {
+      "epoch": 2.8912696063924237,
+      "grad_norm": 1.3131335973739624,
+      "learning_rate": 6.73089216671563e-08,
+      "loss": 0.1966,
+      "step": 6106
+    },
+    {
+      "epoch": 2.891743119266055,
+      "grad_norm": 1.0691511631011963,
+      "learning_rate": 6.671785661352182e-08,
+      "loss": 0.195,
+      "step": 6107
+    },
+    {
+      "epoch": 2.8922166321396863,
+      "grad_norm": 1.2618228197097778,
+      "learning_rate": 6.612938951400472e-08,
+      "loss": 0.1809,
+      "step": 6108
+    },
+    {
+      "epoch": 2.892690145013318,
+      "grad_norm": 1.2614898681640625,
+      "learning_rate": 6.554352052251079e-08,
+      "loss": 0.2123,
+      "step": 6109
+    },
+    {
+      "epoch": 2.893163657886949,
+      "grad_norm": 1.0897921323776245,
+      "learning_rate": 6.496024979226967e-08,
+      "loss": 0.2032,
+      "step": 6110
+    },
+    {
+      "epoch": 2.89363717076058,
+      "grad_norm": 1.0847673416137695,
+      "learning_rate": 6.437957747583046e-08,
+      "loss": 0.1998,
+      "step": 6111
+    },
+    {
+      "epoch": 2.8941106836342114,
+      "grad_norm": 1.2519115209579468,
+      "learning_rate": 6.380150372506277e-08,
+      "loss": 0.1941,
+      "step": 6112
+    },
+    {
+      "epoch": 2.8945841965078425,
+      "grad_norm": 1.4796113967895508,
+      "learning_rate": 6.322602869115568e-08,
+      "loss": 0.2203,
+      "step": 6113
+    },
+    {
+      "epoch": 2.895057709381474,
+      "grad_norm": 1.2136837244033813,
+      "learning_rate": 6.265315252461878e-08,
+      "loss": 0.1934,
+      "step": 6114
+    },
+    {
+      "epoch": 2.895531222255105,
+      "grad_norm": 1.1030209064483643,
+      "learning_rate": 6.208287537528223e-08,
+      "loss": 0.2029,
+      "step": 6115
+    },
+    {
+      "epoch": 2.896004735128736,
+      "grad_norm": 0.975267231464386,
+      "learning_rate": 6.151519739229672e-08,
+      "loss": 0.1909,
+      "step": 6116
+    },
+    {
+      "epoch": 2.8964782480023676,
+      "grad_norm": 1.1102718114852905,
+      "learning_rate": 6.095011872413347e-08,
+      "loss": 0.1771,
+      "step": 6117
+    },
+    {
+      "epoch": 2.8969517608759987,
+      "grad_norm": 1.001387596130371,
+      "learning_rate": 6.038763951858206e-08,
+      "loss": 0.1903,
+      "step": 6118
+    },
+    {
+      "epoch": 2.89742527374963,
+      "grad_norm": 0.9320979118347168,
+      "learning_rate": 5.982775992275592e-08,
+      "loss": 0.2005,
+      "step": 6119
+    },
+    {
+      "epoch": 2.8978987866232613,
+      "grad_norm": 1.0721495151519775,
+      "learning_rate": 5.9270480083083445e-08,
+      "loss": 0.2199,
+      "step": 6120
+    },
+    {
+      "epoch": 2.8983722994968923,
+      "grad_norm": 1.3128505945205688,
+      "learning_rate": 5.871580014531697e-08,
+      "loss": 0.2027,
+      "step": 6121
+    },
+    {
+      "epoch": 2.898845812370524,
+      "grad_norm": 0.9453513622283936,
+      "learning_rate": 5.816372025452821e-08,
+      "loss": 0.1975,
+      "step": 6122
+    },
+    {
+      "epoch": 2.899319325244155,
+      "grad_norm": 1.4196256399154663,
+      "learning_rate": 5.7614240555107224e-08,
+      "loss": 0.2019,
+      "step": 6123
+    },
+    {
+      "epoch": 2.8997928381177864,
+      "grad_norm": 1.2253470420837402,
+      "learning_rate": 5.706736119076683e-08,
+      "loss": 0.2197,
+      "step": 6124
+    },
+    {
+      "epoch": 2.9002663509914175,
+      "grad_norm": 1.5370304584503174,
+      "learning_rate": 5.652308230453596e-08,
+      "loss": 0.198,
+      "step": 6125
+    },
+    {
+      "epoch": 2.9007398638650486,
+      "grad_norm": 1.08665931224823,
+      "learning_rate": 5.5981404038767394e-08,
+      "loss": 0.1892,
+      "step": 6126
+    },
+    {
+      "epoch": 2.90121337673868,
+      "grad_norm": 1.2362626791000366,
+      "learning_rate": 5.5442326535130044e-08,
+      "loss": 0.19,
+      "step": 6127
+    },
+    {
+      "epoch": 2.9016868896123116,
+      "grad_norm": 1.4645034074783325,
+      "learning_rate": 5.490584993461556e-08,
+      "loss": 0.1764,
+      "step": 6128
+    },
+    {
+      "epoch": 2.9021604024859426,
+      "grad_norm": 1.3126651048660278,
+      "learning_rate": 5.4371974377533944e-08,
+      "loss": 0.18,
+      "step": 6129
+    },
+    {
+      "epoch": 2.9026339153595737,
+      "grad_norm": 0.9817326068878174,
+      "learning_rate": 5.384070000351571e-08,
+      "loss": 0.1861,
+      "step": 6130
+    },
+    {
+      "epoch": 2.9031074282332052,
+      "grad_norm": 1.1825339794158936,
+      "learning_rate": 5.331202695151083e-08,
+      "loss": 0.1834,
+      "step": 6131
+    },
+    {
+      "epoch": 2.9035809411068363,
+      "grad_norm": 1.0101470947265625,
+      "learning_rate": 5.278595535978648e-08,
+      "loss": 0.2065,
+      "step": 6132
+    },
+    {
+      "epoch": 2.904054453980468,
+      "grad_norm": 0.9827751517295837,
+      "learning_rate": 5.22624853659337e-08,
+      "loss": 0.2021,
+      "step": 6133
+    },
+    {
+      "epoch": 2.904527966854099,
+      "grad_norm": 1.8256947994232178,
+      "learning_rate": 5.174161710685965e-08,
+      "loss": 0.2104,
+      "step": 6134
+    },
+    {
+      "epoch": 2.90500147972773,
+      "grad_norm": 1.0786854028701782,
+      "learning_rate": 5.122335071879425e-08,
+      "loss": 0.2024,
+      "step": 6135
+    },
+    {
+      "epoch": 2.9054749926013614,
+      "grad_norm": 1.2459017038345337,
+      "learning_rate": 5.0707686337282404e-08,
+      "loss": 0.1871,
+      "step": 6136
+    },
+    {
+      "epoch": 2.9059485054749925,
+      "grad_norm": 1.1489919424057007,
+      "learning_rate": 5.0194624097194e-08,
+      "loss": 0.1885,
+      "step": 6137
+    },
+    {
+      "epoch": 2.906422018348624,
+      "grad_norm": 1.2657074928283691,
+      "learning_rate": 4.968416413271393e-08,
+      "loss": 0.1907,
+      "step": 6138
+    },
+    {
+      "epoch": 2.906895531222255,
+      "grad_norm": 1.3893004655838013,
+      "learning_rate": 4.9176306577347624e-08,
+      "loss": 0.2081,
+      "step": 6139
+    },
+    {
+      "epoch": 2.907369044095886,
+      "grad_norm": 1.1752687692642212,
+      "learning_rate": 4.8671051563922156e-08,
+      "loss": 0.2268,
+      "step": 6140
+    },
+    {
+      "epoch": 2.9078425569695177,
+      "grad_norm": 0.9716362953186035,
+      "learning_rate": 4.816839922457961e-08,
+      "loss": 0.188,
+      "step": 6141
+    },
+    {
+      "epoch": 2.9083160698431487,
+      "grad_norm": 1.7129836082458496,
+      "learning_rate": 4.766834969078704e-08,
+      "loss": 0.2392,
+      "step": 6142
+    },
+    {
+      "epoch": 2.9087895827167802,
+      "grad_norm": 1.019442081451416,
+      "learning_rate": 4.717090309332428e-08,
+      "loss": 0.1874,
+      "step": 6143
+    },
+    {
+      "epoch": 2.9092630955904113,
+      "grad_norm": 1.5065118074417114,
+      "learning_rate": 4.667605956229615e-08,
+      "loss": 0.1921,
+      "step": 6144
+    },
+    {
+      "epoch": 2.9097366084640424,
+      "grad_norm": 0.9485598206520081,
+      "learning_rate": 4.618381922712245e-08,
+      "loss": 0.1998,
+      "step": 6145
+    },
+    {
+      "epoch": 2.910210121337674,
+      "grad_norm": 0.9879134297370911,
+      "learning_rate": 4.5694182216544645e-08,
+      "loss": 0.2106,
+      "step": 6146
+    },
+    {
+      "epoch": 2.9106836342113054,
+      "grad_norm": 1.3310444355010986,
+      "learning_rate": 4.520714865862252e-08,
+      "loss": 0.2027,
+      "step": 6147
+    },
+    {
+      "epoch": 2.9111571470849364,
+      "grad_norm": 1.6184104681015015,
+      "learning_rate": 4.47227186807353e-08,
+      "loss": 0.1938,
+      "step": 6148
+    },
+    {
+      "epoch": 2.9116306599585675,
+      "grad_norm": 0.9067788124084473,
+      "learning_rate": 4.4240892409580516e-08,
+      "loss": 0.1856,
+      "step": 6149
+    },
+    {
+      "epoch": 2.912104172832199,
+      "grad_norm": 1.6381840705871582,
+      "learning_rate": 4.3761669971176255e-08,
+      "loss": 0.2039,
+      "step": 6150
+    },
+    {
+      "epoch": 2.91257768570583,
+      "grad_norm": 1.0911681652069092,
+      "learning_rate": 4.328505149085782e-08,
+      "loss": 0.1867,
+      "step": 6151
+    },
+    {
+      "epoch": 2.9130511985794616,
+      "grad_norm": 0.9727244973182678,
+      "learning_rate": 4.281103709327883e-08,
+      "loss": 0.2022,
+      "step": 6152
+    },
+    {
+      "epoch": 2.9135247114530927,
+      "grad_norm": 0.9916009306907654,
+      "learning_rate": 4.233962690241567e-08,
+      "loss": 0.1753,
+      "step": 6153
+    },
+    {
+      "epoch": 2.9139982243267237,
+      "grad_norm": 1.301115870475769,
+      "learning_rate": 4.18708210415586e-08,
+      "loss": 0.1829,
+      "step": 6154
+    },
+    {
+      "epoch": 2.9144717372003552,
+      "grad_norm": 1.1094136238098145,
+      "learning_rate": 4.140461963332065e-08,
+      "loss": 0.2055,
+      "step": 6155
+    },
+    {
+      "epoch": 2.9149452500739863,
+      "grad_norm": 0.9840229749679565,
+      "learning_rate": 4.094102279963319e-08,
+      "loss": 0.2242,
+      "step": 6156
+    },
+    {
+      "epoch": 2.915418762947618,
+      "grad_norm": 1.3108941316604614,
+      "learning_rate": 4.048003066174366e-08,
+      "loss": 0.1853,
+      "step": 6157
+    },
+    {
+      "epoch": 2.915892275821249,
+      "grad_norm": 1.0700182914733887,
+      "learning_rate": 4.002164334022118e-08,
+      "loss": 0.2149,
+      "step": 6158
+    },
+    {
+      "epoch": 2.91636578869488,
+      "grad_norm": 1.2526887655258179,
+      "learning_rate": 3.956586095495207e-08,
+      "loss": 0.1841,
+      "step": 6159
+    },
+    {
+      "epoch": 2.9168393015685115,
+      "grad_norm": 1.097798466682434,
+      "learning_rate": 3.911268362514209e-08,
+      "loss": 0.1977,
+      "step": 6160
+    },
+    {
+      "epoch": 2.9173128144421425,
+      "grad_norm": 1.293502688407898,
+      "learning_rate": 3.866211146931531e-08,
+      "loss": 0.2091,
+      "step": 6161
+    },
+    {
+      "epoch": 2.917786327315774,
+      "grad_norm": 1.0824062824249268,
+      "learning_rate": 3.821414460531414e-08,
+      "loss": 0.2008,
+      "step": 6162
+    },
+    {
+      "epoch": 2.918259840189405,
+      "grad_norm": 1.2927944660186768,
+      "learning_rate": 3.776878315030042e-08,
+      "loss": 0.2031,
+      "step": 6163
+    },
+    {
+      "epoch": 2.918733353063036,
+      "grad_norm": 1.0434043407440186,
+      "learning_rate": 3.73260272207554e-08,
+      "loss": 0.204,
+      "step": 6164
+    },
+    {
+      "epoch": 2.9192068659366677,
+      "grad_norm": 0.9378446936607361,
+      "learning_rate": 3.688587693247536e-08,
+      "loss": 0.1847,
+      "step": 6165
+    },
+    {
+      "epoch": 2.919680378810299,
+      "grad_norm": 1.1560747623443604,
+      "learning_rate": 3.644833240057821e-08,
+      "loss": 0.2163,
+      "step": 6166
+    },
+    {
+      "epoch": 2.9201538916839302,
+      "grad_norm": 0.9858490228652954,
+      "learning_rate": 3.601339373950019e-08,
+      "loss": 0.1798,
+      "step": 6167
+    },
+    {
+      "epoch": 2.9206274045575613,
+      "grad_norm": 1.1007888317108154,
+      "learning_rate": 3.558106106299475e-08,
+      "loss": 0.1843,
+      "step": 6168
+    },
+    {
+      "epoch": 2.921100917431193,
+      "grad_norm": 1.1641054153442383,
+      "learning_rate": 3.515133448413366e-08,
+      "loss": 0.187,
+      "step": 6169
+    },
+    {
+      "epoch": 2.921574430304824,
+      "grad_norm": 1.4323322772979736,
+      "learning_rate": 3.472421411530924e-08,
+      "loss": 0.1945,
+      "step": 6170
+    },
+    {
+      "epoch": 2.9220479431784554,
+      "grad_norm": 1.1009232997894287,
+      "learning_rate": 3.429970006822991e-08,
+      "loss": 0.1778,
+      "step": 6171
+    },
+    {
+      "epoch": 2.9225214560520865,
+      "grad_norm": 1.0793761014938354,
+      "learning_rate": 3.387779245392242e-08,
+      "loss": 0.1852,
+      "step": 6172
+    },
+    {
+      "epoch": 2.9229949689257175,
+      "grad_norm": 1.315026879310608,
+      "learning_rate": 3.345849138273405e-08,
+      "loss": 0.2095,
+      "step": 6173
+    },
+    {
+      "epoch": 2.923468481799349,
+      "grad_norm": 1.0471131801605225,
+      "learning_rate": 3.304179696432708e-08,
+      "loss": 0.2058,
+      "step": 6174
+    },
+    {
+      "epoch": 2.92394199467298,
+      "grad_norm": 1.2734088897705078,
+      "learning_rate": 3.262770930768655e-08,
+      "loss": 0.1976,
+      "step": 6175
+    },
+    {
+      "epoch": 2.9244155075466116,
+      "grad_norm": 1.2894588708877563,
+      "learning_rate": 3.2216228521111393e-08,
+      "loss": 0.2161,
+      "step": 6176
+    },
+    {
+      "epoch": 2.9248890204202427,
+      "grad_norm": 0.9765024781227112,
+      "learning_rate": 3.180735471222107e-08,
+      "loss": 0.2171,
+      "step": 6177
+    },
+    {
+      "epoch": 2.9253625332938737,
+      "grad_norm": 0.9731594324111938,
+      "learning_rate": 3.140108798795227e-08,
+      "loss": 0.1879,
+      "step": 6178
+    },
+    {
+      "epoch": 2.9258360461675053,
+      "grad_norm": 1.8532750606536865,
+      "learning_rate": 3.099742845455889e-08,
+      "loss": 0.1931,
+      "step": 6179
+    },
+    {
+      "epoch": 2.9263095590411363,
+      "grad_norm": 1.2193810939788818,
+      "learning_rate": 3.059637621761646e-08,
+      "loss": 0.1973,
+      "step": 6180
+    },
+    {
+      "epoch": 2.926783071914768,
+      "grad_norm": 1.1351593732833862,
+      "learning_rate": 3.019793138201554e-08,
+      "loss": 0.2097,
+      "step": 6181
+    },
+    {
+      "epoch": 2.927256584788399,
+      "grad_norm": 1.9396110773086548,
+      "learning_rate": 2.9802094051964993e-08,
+      "loss": 0.2438,
+      "step": 6182
+    },
+    {
+      "epoch": 2.92773009766203,
+      "grad_norm": 0.9938279986381531,
+      "learning_rate": 2.9408864330991993e-08,
+      "loss": 0.198,
+      "step": 6183
+    },
+    {
+      "epoch": 2.9282036105356615,
+      "grad_norm": 1.4137035608291626,
+      "learning_rate": 2.9018242321943168e-08,
+      "loss": 0.2003,
+      "step": 6184
+    },
+    {
+      "epoch": 2.9286771234092925,
+      "grad_norm": 1.076438546180725,
+      "learning_rate": 2.8630228126981242e-08,
+      "loss": 0.2066,
+      "step": 6185
+    },
+    {
+      "epoch": 2.929150636282924,
+      "grad_norm": 1.398909330368042,
+      "learning_rate": 2.8244821847587256e-08,
+      "loss": 0.1752,
+      "step": 6186
+    },
+    {
+      "epoch": 2.929624149156555,
+      "grad_norm": 0.9109712243080139,
+      "learning_rate": 2.7862023584561692e-08,
+      "loss": 0.196,
+      "step": 6187
+    },
+    {
+      "epoch": 2.930097662030186,
+      "grad_norm": 1.3671661615371704,
+      "learning_rate": 2.748183343802002e-08,
+      "loss": 0.2127,
+      "step": 6188
+    },
+    {
+      "epoch": 2.9305711749038177,
+      "grad_norm": 1.2796677350997925,
+      "learning_rate": 2.7104251507398262e-08,
+      "loss": 0.2206,
+      "step": 6189
+    },
+    {
+      "epoch": 2.931044687777449,
+      "grad_norm": 0.9807784557342529,
+      "learning_rate": 2.6729277891449634e-08,
+      "loss": 0.1877,
+      "step": 6190
+    },
+    {
+      "epoch": 2.9315182006510803,
+      "grad_norm": 0.9959766864776611,
+      "learning_rate": 2.6356912688244585e-08,
+      "loss": 0.187,
+      "step": 6191
+    },
+    {
+      "epoch": 2.9319917135247113,
+      "grad_norm": 1.6035785675048828,
+      "learning_rate": 2.5987155995171876e-08,
+      "loss": 0.1968,
+      "step": 6192
+    },
+    {
+      "epoch": 2.932465226398343,
+      "grad_norm": 1.377586841583252,
+      "learning_rate": 2.5620007908937483e-08,
+      "loss": 0.2178,
+      "step": 6193
+    },
+    {
+      "epoch": 2.932938739271974,
+      "grad_norm": 1.7171190977096558,
+      "learning_rate": 2.5255468525564598e-08,
+      "loss": 0.2138,
+      "step": 6194
+    },
+    {
+      "epoch": 2.9334122521456054,
+      "grad_norm": 1.3508435487747192,
+      "learning_rate": 2.489353794039695e-08,
+      "loss": 0.2017,
+      "step": 6195
+    },
+    {
+      "epoch": 2.9338857650192365,
+      "grad_norm": 1.0717988014221191,
+      "learning_rate": 2.4534216248092158e-08,
+      "loss": 0.1961,
+      "step": 6196
+    },
+    {
+      "epoch": 2.9343592778928675,
+      "grad_norm": 1.060861349105835,
+      "learning_rate": 2.4177503542627266e-08,
+      "loss": 0.2179,
+      "step": 6197
+    },
+    {
+      "epoch": 2.934832790766499,
+      "grad_norm": 1.0644173622131348,
+      "learning_rate": 2.382339991729987e-08,
+      "loss": 0.2089,
+      "step": 6198
+    },
+    {
+      "epoch": 2.93530630364013,
+      "grad_norm": 1.3779876232147217,
+      "learning_rate": 2.3471905464719226e-08,
+      "loss": 0.2026,
+      "step": 6199
+    },
+    {
+      "epoch": 2.9357798165137616,
+      "grad_norm": 1.585129737854004,
+      "learning_rate": 2.312302027681623e-08,
+      "loss": 0.2053,
+      "step": 6200
+    },
+    {
+      "epoch": 2.9362533293873927,
+      "grad_norm": 1.6859186887741089,
+      "learning_rate": 2.2776744444839017e-08,
+      "loss": 0.2124,
+      "step": 6201
+    },
+    {
+      "epoch": 2.9367268422610238,
+      "grad_norm": 1.0570391416549683,
+      "learning_rate": 2.24330780593518e-08,
+      "loss": 0.1986,
+      "step": 6202
+    },
+    {
+      "epoch": 2.9372003551346553,
+      "grad_norm": 1.2987060546875,
+      "learning_rate": 2.2092021210238233e-08,
+      "loss": 0.2184,
+      "step": 6203
+    },
+    {
+      "epoch": 2.9376738680082863,
+      "grad_norm": 1.2632862329483032,
+      "learning_rate": 2.1753573986698086e-08,
+      "loss": 0.2145,
+      "step": 6204
+    },
+    {
+      "epoch": 2.938147380881918,
+      "grad_norm": 0.9195615649223328,
+      "learning_rate": 2.141773647724832e-08,
+      "loss": 0.1825,
+      "step": 6205
+    },
+    {
+      "epoch": 2.938620893755549,
+      "grad_norm": 1.0756436586380005,
+      "learning_rate": 2.1084508769725344e-08,
+      "loss": 0.2099,
+      "step": 6206
+    },
+    {
+      "epoch": 2.93909440662918,
+      "grad_norm": 1.0231488943099976,
+      "learning_rate": 2.0753890951280554e-08,
+      "loss": 0.2167,
+      "step": 6207
+    },
+    {
+      "epoch": 2.9395679195028115,
+      "grad_norm": 1.5014095306396484,
+      "learning_rate": 2.0425883108383672e-08,
+      "loss": 0.1996,
+      "step": 6208
+    },
+    {
+      "epoch": 2.940041432376443,
+      "grad_norm": 1.0937631130218506,
+      "learning_rate": 2.010048532682274e-08,
+      "loss": 0.223,
+      "step": 6209
+    },
+    {
+      "epoch": 2.940514945250074,
+      "grad_norm": 1.103412389755249,
+      "learning_rate": 1.9777697691701904e-08,
+      "loss": 0.1888,
+      "step": 6210
+    },
+    {
+      "epoch": 2.940988458123705,
+      "grad_norm": 1.1209019422531128,
+      "learning_rate": 1.945752028744252e-08,
+      "loss": 0.1824,
+      "step": 6211
+    },
+    {
+      "epoch": 2.9414619709973366,
+      "grad_norm": 0.9927660226821899,
+      "learning_rate": 1.913995319778539e-08,
+      "loss": 0.186,
+      "step": 6212
+    },
+    {
+      "epoch": 2.9419354838709677,
+      "grad_norm": 1.021451473236084,
+      "learning_rate": 1.8824996505787398e-08,
+      "loss": 0.1983,
+      "step": 6213
+    },
+    {
+      "epoch": 2.942408996744599,
+      "grad_norm": 1.1407397985458374,
+      "learning_rate": 1.8512650293820433e-08,
+      "loss": 0.1997,
+      "step": 6214
+    },
+    {
+      "epoch": 2.9428825096182303,
+      "grad_norm": 0.8883314728736877,
+      "learning_rate": 1.820291464357693e-08,
+      "loss": 0.1719,
+      "step": 6215
+    },
+    {
+      "epoch": 2.9433560224918613,
+      "grad_norm": 1.1217479705810547,
+      "learning_rate": 1.789578963606431e-08,
+      "loss": 0.1944,
+      "step": 6216
+    },
+    {
+      "epoch": 2.943829535365493,
+      "grad_norm": 1.5166805982589722,
+      "learning_rate": 1.7591275351609428e-08,
+      "loss": 0.215,
+      "step": 6217
+    },
+    {
+      "epoch": 2.944303048239124,
+      "grad_norm": 1.0720276832580566,
+      "learning_rate": 1.7289371869854132e-08,
+      "loss": 0.1917,
+      "step": 6218
+    },
+    {
+      "epoch": 2.9447765611127554,
+      "grad_norm": 1.4796210527420044,
+      "learning_rate": 1.699007926975971e-08,
+      "loss": 0.2059,
+      "step": 6219
+    },
+    {
+      "epoch": 2.9452500739863865,
+      "grad_norm": 1.3808337450027466,
+      "learning_rate": 1.6693397629601317e-08,
+      "loss": 0.1986,
+      "step": 6220
+    },
+    {
+      "epoch": 2.9457235868600176,
+      "grad_norm": 1.275921106338501,
+      "learning_rate": 1.6399327026974666e-08,
+      "loss": 0.1888,
+      "step": 6221
+    },
+    {
+      "epoch": 2.946197099733649,
+      "grad_norm": 1.051120638847351,
+      "learning_rate": 1.6107867538790456e-08,
+      "loss": 0.2187,
+      "step": 6222
+    },
+    {
+      "epoch": 2.94667061260728,
+      "grad_norm": 1.0114779472351074,
+      "learning_rate": 1.58190192412766e-08,
+      "loss": 0.2084,
+      "step": 6223
+    },
+    {
+      "epoch": 2.9471441254809116,
+      "grad_norm": 1.3596932888031006,
+      "learning_rate": 1.5532782209979336e-08,
+      "loss": 0.1849,
+      "step": 6224
+    },
+    {
+      "epoch": 2.9476176383545427,
+      "grad_norm": 2.326784372329712,
+      "learning_rate": 1.524915651976211e-08,
+      "loss": 0.184,
+      "step": 6225
+    },
+    {
+      "epoch": 2.948091151228174,
+      "grad_norm": 1.3456201553344727,
+      "learning_rate": 1.4968142244802254e-08,
+      "loss": 0.2014,
+      "step": 6226
+    },
+    {
+      "epoch": 2.9485646641018053,
+      "grad_norm": 1.592707872390747,
+      "learning_rate": 1.4689739458597641e-08,
+      "loss": 0.193,
+      "step": 6227
+    },
+    {
+      "epoch": 2.9490381769754364,
+      "grad_norm": 1.3673793077468872,
+      "learning_rate": 1.4413948233961138e-08,
+      "loss": 0.2131,
+      "step": 6228
+    },
+    {
+      "epoch": 2.949511689849068,
+      "grad_norm": 1.1322108507156372,
+      "learning_rate": 1.414076864302505e-08,
+      "loss": 0.2056,
+      "step": 6229
+    },
+    {
+      "epoch": 2.949985202722699,
+      "grad_norm": 0.8551841974258423,
+      "learning_rate": 1.3870200757235552e-08,
+      "loss": 0.1825,
+      "step": 6230
+    },
+    {
+      "epoch": 2.9504587155963304,
+      "grad_norm": 0.9732584357261658,
+      "learning_rate": 1.3602244647356044e-08,
+      "loss": 0.208,
+      "step": 6231
+    },
+    {
+      "epoch": 2.9509322284699615,
+      "grad_norm": 1.1727527379989624,
+      "learning_rate": 1.3336900383469353e-08,
+      "loss": 0.1963,
+      "step": 6232
+    },
+    {
+      "epoch": 2.951405741343593,
+      "grad_norm": 0.9997616410255432,
+      "learning_rate": 1.30741680349733e-08,
+      "loss": 0.1923,
+      "step": 6233
+    },
+    {
+      "epoch": 2.951879254217224,
+      "grad_norm": 1.2932562828063965,
+      "learning_rate": 1.2814047670584028e-08,
+      "loss": 0.1968,
+      "step": 6234
+    },
+    {
+      "epoch": 2.952352767090855,
+      "grad_norm": 1.0885539054870605,
+      "learning_rate": 1.2556539358331566e-08,
+      "loss": 0.1931,
+      "step": 6235
+    },
+    {
+      "epoch": 2.9528262799644867,
+      "grad_norm": 1.0896440744400024,
+      "learning_rate": 1.230164316556537e-08,
+      "loss": 0.1941,
+      "step": 6236
+    },
+    {
+      "epoch": 2.9532997928381177,
+      "grad_norm": 1.0477862358093262,
+      "learning_rate": 1.2049359158952111e-08,
+      "loss": 0.192,
+      "step": 6237
+    },
+    {
+      "epoch": 2.9537733057117492,
+      "grad_norm": 0.844715416431427,
+      "learning_rate": 1.1799687404473458e-08,
+      "loss": 0.1807,
+      "step": 6238
+    },
+    {
+      "epoch": 2.9542468185853803,
+      "grad_norm": 1.3414781093597412,
+      "learning_rate": 1.1552627967428288e-08,
+      "loss": 0.2202,
+      "step": 6239
+    },
+    {
+      "epoch": 2.9547203314590114,
+      "grad_norm": 0.9424867630004883,
+      "learning_rate": 1.1308180912432688e-08,
+      "loss": 0.1943,
+      "step": 6240
+    },
+    {
+      "epoch": 2.955193844332643,
+      "grad_norm": 1.0376644134521484,
+      "learning_rate": 1.1066346303421071e-08,
+      "loss": 0.1946,
+      "step": 6241
+    },
+    {
+      "epoch": 2.955667357206274,
+      "grad_norm": 1.2721221446990967,
+      "learning_rate": 1.0827124203640627e-08,
+      "loss": 0.1955,
+      "step": 6242
+    },
+    {
+      "epoch": 2.9561408700799054,
+      "grad_norm": 1.0730029344558716,
+      "learning_rate": 1.059051467565797e-08,
+      "loss": 0.1921,
+      "step": 6243
+    },
+    {
+      "epoch": 2.9566143829535365,
+      "grad_norm": 1.2354124784469604,
+      "learning_rate": 1.0356517781358044e-08,
+      "loss": 0.2074,
+      "step": 6244
+    },
+    {
+      "epoch": 2.9570878958271676,
+      "grad_norm": 0.9237876534461975,
+      "learning_rate": 1.0125133581938562e-08,
+      "loss": 0.2075,
+      "step": 6245
+    },
+    {
+      "epoch": 2.957561408700799,
+      "grad_norm": 1.4493162631988525,
+      "learning_rate": 9.896362137916672e-09,
+      "loss": 0.2207,
+      "step": 6246
+    },
+    {
+      "epoch": 2.95803492157443,
+      "grad_norm": 1.3988970518112183,
+      "learning_rate": 9.670203509124509e-09,
+      "loss": 0.197,
+      "step": 6247
+    },
+    {
+      "epoch": 2.9585084344480617,
+      "grad_norm": 1.1738380193710327,
+      "learning_rate": 9.44665775471254e-09,
+      "loss": 0.2102,
+      "step": 6248
+    },
+    {
+      "epoch": 2.9589819473216927,
+      "grad_norm": 1.2801897525787354,
+      "learning_rate": 9.225724933146218e-09,
+      "loss": 0.2061,
+      "step": 6249
+    },
+    {
+      "epoch": 2.959455460195324,
+      "grad_norm": 1.5334160327911377,
+      "learning_rate": 9.007405102209321e-09,
+      "loss": 0.1896,
+      "step": 6250
+    },
+    {
+      "epoch": 2.9599289730689553,
+      "grad_norm": 1.0199025869369507,
+      "learning_rate": 8.791698318999508e-09,
+      "loss": 0.2034,
+      "step": 6251
+    },
+    {
+      "epoch": 2.960402485942587,
+      "grad_norm": 0.8134233951568604,
+      "learning_rate": 8.578604639936095e-09,
+      "loss": 0.1722,
+      "step": 6252
+    },
+    {
+      "epoch": 2.960875998816218,
+      "grad_norm": 1.0595088005065918,
+      "learning_rate": 8.368124120747833e-09,
+      "loss": 0.1856,
+      "step": 6253
+    },
+    {
+      "epoch": 2.961349511689849,
+      "grad_norm": 0.9498534202575684,
+      "learning_rate": 8.160256816487355e-09,
+      "loss": 0.1602,
+      "step": 6254
+    },
+    {
+      "epoch": 2.9618230245634805,
+      "grad_norm": 0.8854544162750244,
+      "learning_rate": 7.95500278151784e-09,
+      "loss": 0.1959,
+      "step": 6255
+    },
+    {
+      "epoch": 2.9622965374371115,
+      "grad_norm": 1.1654853820800781,
+      "learning_rate": 7.752362069523012e-09,
+      "loss": 0.2078,
+      "step": 6256
+    },
+    {
+      "epoch": 2.962770050310743,
+      "grad_norm": 0.9548454880714417,
+      "learning_rate": 7.552334733500477e-09,
+      "loss": 0.1876,
+      "step": 6257
+    },
+    {
+      "epoch": 2.963243563184374,
+      "grad_norm": 1.966020107269287,
+      "learning_rate": 7.354920825766166e-09,
+      "loss": 0.1918,
+      "step": 6258
+    },
+    {
+      "epoch": 2.963717076058005,
+      "grad_norm": 1.6142258644104004,
+      "learning_rate": 7.160120397950998e-09,
+      "loss": 0.2173,
+      "step": 6259
+    },
+    {
+      "epoch": 2.9641905889316367,
+      "grad_norm": 0.9371702075004578,
+      "learning_rate": 6.967933501004221e-09,
+      "loss": 0.1948,
+      "step": 6260
+    },
+    {
+      "epoch": 2.9646641018052677,
+      "grad_norm": 1.0012017488479614,
+      "learning_rate": 6.778360185190069e-09,
+      "loss": 0.196,
+      "step": 6261
+    },
+    {
+      "epoch": 2.9651376146788992,
+      "grad_norm": 1.5582655668258667,
+      "learning_rate": 6.591400500088885e-09,
+      "loss": 0.197,
+      "step": 6262
+    },
+    {
+      "epoch": 2.9656111275525303,
+      "grad_norm": 1.4498084783554077,
+      "learning_rate": 6.407054494599329e-09,
+      "loss": 0.2011,
+      "step": 6263
+    },
+    {
+      "epoch": 2.9660846404261614,
+      "grad_norm": 1.6158671379089355,
+      "learning_rate": 6.2253222169339485e-09,
+      "loss": 0.1903,
+      "step": 6264
+    },
+    {
+      "epoch": 2.966558153299793,
+      "grad_norm": 1.1202692985534668,
+      "learning_rate": 6.046203714624721e-09,
+      "loss": 0.2186,
+      "step": 6265
+    },
+    {
+      "epoch": 2.967031666173424,
+      "grad_norm": 1.2885193824768066,
+      "learning_rate": 5.8696990345175064e-09,
+      "loss": 0.2111,
+      "step": 6266
+    },
+    {
+      "epoch": 2.9675051790470555,
+      "grad_norm": 0.9069783687591553,
+      "learning_rate": 5.695808222775379e-09,
+      "loss": 0.2006,
+      "step": 6267
+    },
+    {
+      "epoch": 2.9679786919206865,
+      "grad_norm": 1.5070737600326538,
+      "learning_rate": 5.524531324877513e-09,
+      "loss": 0.1914,
+      "step": 6268
+    },
+    {
+      "epoch": 2.9684522047943176,
+      "grad_norm": 1.019160509109497,
+      "learning_rate": 5.3558683856203e-09,
+      "loss": 0.21,
+      "step": 6269
+    },
+    {
+      "epoch": 2.968925717667949,
+      "grad_norm": 1.3049955368041992,
+      "learning_rate": 5.189819449116229e-09,
+      "loss": 0.1916,
+      "step": 6270
+    },
+    {
+      "epoch": 2.9693992305415806,
+      "grad_norm": 1.0356775522232056,
+      "learning_rate": 5.026384558792785e-09,
+      "loss": 0.2048,
+      "step": 6271
+    },
+    {
+      "epoch": 2.9698727434152117,
+      "grad_norm": 1.2920750379562378,
+      "learning_rate": 4.865563757394665e-09,
+      "loss": 0.2184,
+      "step": 6272
+    },
+    {
+      "epoch": 2.9703462562888427,
+      "grad_norm": 1.0191415548324585,
+      "learning_rate": 4.707357086983777e-09,
+      "loss": 0.1999,
+      "step": 6273
+    },
+    {
+      "epoch": 2.9708197691624743,
+      "grad_norm": 1.0479013919830322,
+      "learning_rate": 4.5517645889381346e-09,
+      "loss": 0.2006,
+      "step": 6274
+    },
+    {
+      "epoch": 2.9712932820361053,
+      "grad_norm": 1.0281206369400024,
+      "learning_rate": 4.398786303949632e-09,
+      "loss": 0.1796,
+      "step": 6275
+    },
+    {
+      "epoch": 2.971766794909737,
+      "grad_norm": 1.0202572345733643,
+      "learning_rate": 4.248422272029596e-09,
+      "loss": 0.2112,
+      "step": 6276
+    },
+    {
+      "epoch": 2.972240307783368,
+      "grad_norm": 1.0091017484664917,
+      "learning_rate": 4.100672532504346e-09,
+      "loss": 0.1999,
+      "step": 6277
+    },
+    {
+      "epoch": 2.972713820656999,
+      "grad_norm": 0.9518880844116211,
+      "learning_rate": 3.955537124016306e-09,
+      "loss": 0.1934,
+      "step": 6278
+    },
+    {
+      "epoch": 2.9731873335306305,
+      "grad_norm": 1.1970912218093872,
+      "learning_rate": 3.813016084522892e-09,
+      "loss": 0.1956,
+      "step": 6279
+    },
+    {
+      "epoch": 2.9736608464042615,
+      "grad_norm": 1.150801181793213,
+      "learning_rate": 3.673109451300949e-09,
+      "loss": 0.1675,
+      "step": 6280
+    },
+    {
+      "epoch": 2.974134359277893,
+      "grad_norm": 1.025273323059082,
+      "learning_rate": 3.53581726094121e-09,
+      "loss": 0.1821,
+      "step": 6281
+    },
+    {
+      "epoch": 2.974607872151524,
+      "grad_norm": 1.1569122076034546,
+      "learning_rate": 3.4011395493505073e-09,
+      "loss": 0.1819,
+      "step": 6282
+    },
+    {
+      "epoch": 2.975081385025155,
+      "grad_norm": 2.0140018463134766,
+      "learning_rate": 3.269076351752887e-09,
+      "loss": 0.1852,
+      "step": 6283
+    },
+    {
+      "epoch": 2.9755548978987867,
+      "grad_norm": 1.07260000705719,
+      "learning_rate": 3.139627702688497e-09,
+      "loss": 0.2121,
+      "step": 6284
+    },
+    {
+      "epoch": 2.9760284107724178,
+      "grad_norm": 1.5674619674682617,
+      "learning_rate": 3.0127936360124786e-09,
+      "loss": 0.1916,
+      "step": 6285
+    },
+    {
+      "epoch": 2.9765019236460493,
+      "grad_norm": 1.289867639541626,
+      "learning_rate": 2.888574184898296e-09,
+      "loss": 0.1823,
+      "step": 6286
+    },
+    {
+      "epoch": 2.9769754365196803,
+      "grad_norm": 1.5313868522644043,
+      "learning_rate": 2.7669693818332954e-09,
+      "loss": 0.1996,
+      "step": 6287
+    },
+    {
+      "epoch": 2.9774489493933114,
+      "grad_norm": 1.160865306854248,
+      "learning_rate": 2.6479792586220356e-09,
+      "loss": 0.211,
+      "step": 6288
+    },
+    {
+      "epoch": 2.977922462266943,
+      "grad_norm": 1.0990283489227295,
+      "learning_rate": 2.531603846386288e-09,
+      "loss": 0.1845,
+      "step": 6289
+    },
+    {
+      "epoch": 2.978395975140574,
+      "grad_norm": 1.126089096069336,
+      "learning_rate": 2.417843175561707e-09,
+      "loss": 0.189,
+      "step": 6290
+    },
+    {
+      "epoch": 2.9788694880142055,
+      "grad_norm": 1.5801364183425903,
+      "learning_rate": 2.30669727590227e-09,
+      "loss": 0.214,
+      "step": 6291
+    },
+    {
+      "epoch": 2.9793430008878365,
+      "grad_norm": 0.9013964533805847,
+      "learning_rate": 2.1981661764769456e-09,
+      "loss": 0.2052,
+      "step": 6292
+    },
+    {
+      "epoch": 2.979816513761468,
+      "grad_norm": 1.815775752067566,
+      "learning_rate": 2.0922499056708066e-09,
+      "loss": 0.1886,
+      "step": 6293
+    },
+    {
+      "epoch": 2.980290026635099,
+      "grad_norm": 1.0103025436401367,
+      "learning_rate": 1.9889484911850276e-09,
+      "loss": 0.1864,
+      "step": 6294
+    },
+    {
+      "epoch": 2.9807635395087306,
+      "grad_norm": 1.127058744430542,
+      "learning_rate": 1.8882619600368855e-09,
+      "loss": 0.2151,
+      "step": 6295
+    },
+    {
+      "epoch": 2.9812370523823617,
+      "grad_norm": 1.0886855125427246,
+      "learning_rate": 1.7901903385597607e-09,
+      "loss": 0.1955,
+      "step": 6296
+    },
+    {
+      "epoch": 2.9817105652559928,
+      "grad_norm": 1.1773751974105835,
+      "learning_rate": 1.694733652405356e-09,
+      "loss": 0.172,
+      "step": 6297
+    },
+    {
+      "epoch": 2.9821840781296243,
+      "grad_norm": 0.9003830552101135,
+      "learning_rate": 1.601891926537036e-09,
+      "loss": 0.2013,
+      "step": 6298
+    },
+    {
+      "epoch": 2.9826575910032553,
+      "grad_norm": 1.0472919940948486,
+      "learning_rate": 1.5116651852375985e-09,
+      "loss": 0.1943,
+      "step": 6299
+    },
+    {
+      "epoch": 2.983131103876887,
+      "grad_norm": 1.1670929193496704,
+      "learning_rate": 1.4240534521059447e-09,
+      "loss": 0.1955,
+      "step": 6300
+    },
+    {
+      "epoch": 2.983604616750518,
+      "grad_norm": 1.7558685541152954,
+      "learning_rate": 1.3390567500537466e-09,
+      "loss": 0.2085,
+      "step": 6301
+    },
+    {
+      "epoch": 2.984078129624149,
+      "grad_norm": 1.4146206378936768,
+      "learning_rate": 1.2566751013132205e-09,
+      "loss": 0.2048,
+      "step": 6302
+    },
+    {
+      "epoch": 2.9845516424977805,
+      "grad_norm": 1.2582781314849854,
+      "learning_rate": 1.1769085274304648e-09,
+      "loss": 0.2005,
+      "step": 6303
+    },
+    {
+      "epoch": 2.9850251553714116,
+      "grad_norm": 1.0818257331848145,
+      "learning_rate": 1.0997570492654597e-09,
+      "loss": 0.2016,
+      "step": 6304
+    },
+    {
+      "epoch": 2.985498668245043,
+      "grad_norm": 1.0899266004562378,
+      "learning_rate": 1.025220686998729e-09,
+      "loss": 0.2178,
+      "step": 6305
+    },
+    {
+      "epoch": 2.985972181118674,
+      "grad_norm": 0.9373279809951782,
+      "learning_rate": 9.53299460123569e-10,
+      "loss": 0.1966,
+      "step": 6306
+    },
+    {
+      "epoch": 2.986445693992305,
+      "grad_norm": 1.2788419723510742,
+      "learning_rate": 8.83993387450488e-10,
+      "loss": 0.2131,
+      "step": 6307
+    },
+    {
+      "epoch": 2.9869192068659367,
+      "grad_norm": 1.247537612915039,
+      "learning_rate": 8.173024871060974e-10,
+      "loss": 0.1796,
+      "step": 6308
+    },
+    {
+      "epoch": 2.9873927197395678,
+      "grad_norm": 1.1506285667419434,
+      "learning_rate": 7.532267765320012e-10,
+      "loss": 0.1883,
+      "step": 6309
+    },
+    {
+      "epoch": 2.9878662326131993,
+      "grad_norm": 1.517842173576355,
+      "learning_rate": 6.917662724870155e-10,
+      "loss": 0.2005,
+      "step": 6310
+    },
+    {
+      "epoch": 2.9883397454868303,
+      "grad_norm": 1.6192688941955566,
+      "learning_rate": 6.329209910460598e-10,
+      "loss": 0.1934,
+      "step": 6311
+    },
+    {
+      "epoch": 2.9888132583604614,
+      "grad_norm": 1.194815754890442,
+      "learning_rate": 5.766909475979354e-10,
+      "loss": 0.2053,
+      "step": 6312
+    },
+    {
+      "epoch": 2.989286771234093,
+      "grad_norm": 0.9704394340515137,
+      "learning_rate": 5.230761568508769e-10,
+      "loss": 0.1886,
+      "step": 6313
+    },
+    {
+      "epoch": 2.9897602841077244,
+      "grad_norm": 1.4745657444000244,
+      "learning_rate": 4.72076632827001e-10,
+      "loss": 0.2025,
+      "step": 6314
+    },
+    {
+      "epoch": 2.9902337969813555,
+      "grad_norm": 1.8984787464141846,
+      "learning_rate": 4.2369238886341704e-10,
+      "loss": 0.2104,
+      "step": 6315
+    },
+    {
+      "epoch": 2.9907073098549866,
+      "grad_norm": 1.1849009990692139,
+      "learning_rate": 3.7792343761555717e-10,
+      "loss": 0.1879,
+      "step": 6316
+    },
+    {
+      "epoch": 2.991180822728618,
+      "grad_norm": 1.43865168094635,
+      "learning_rate": 3.347697910538461e-10,
+      "loss": 0.2133,
+      "step": 6317
+    },
+    {
+      "epoch": 2.991654335602249,
+      "grad_norm": 1.0743141174316406,
+      "learning_rate": 2.942314604648111e-10,
+      "loss": 0.1879,
+      "step": 6318
+    },
+    {
+      "epoch": 2.9921278484758806,
+      "grad_norm": 0.9571679830551147,
+      "learning_rate": 2.5630845645108207e-10,
+      "loss": 0.1998,
+      "step": 6319
+    },
+    {
+      "epoch": 2.9926013613495117,
+      "grad_norm": 0.9296308755874634,
+      "learning_rate": 2.210007889302812e-10,
+      "loss": 0.1982,
+      "step": 6320
+    },
+    {
+      "epoch": 2.993074874223143,
+      "grad_norm": 1.0719934701919556,
+      "learning_rate": 1.883084671372437e-10,
+      "loss": 0.1833,
+      "step": 6321
+    },
+    {
+      "epoch": 2.9935483870967743,
+      "grad_norm": 1.165964961051941,
+      "learning_rate": 1.5823149962179707e-10,
+      "loss": 0.198,
+      "step": 6322
+    },
+    {
+      "epoch": 2.9940218999704054,
+      "grad_norm": 1.052899956703186,
+      "learning_rate": 1.3076989425098162e-10,
+      "loss": 0.1998,
+      "step": 6323
+    },
+    {
+      "epoch": 2.994495412844037,
+      "grad_norm": 1.0535532236099243,
+      "learning_rate": 1.0592365820683015e-10,
+      "loss": 0.2141,
+      "step": 6324
+    },
+    {
+      "epoch": 2.994968925717668,
+      "grad_norm": 0.9778185486793518,
+      "learning_rate": 8.369279798747798e-11,
+      "loss": 0.1918,
+      "step": 6325
+    },
+    {
+      "epoch": 2.995442438591299,
+      "grad_norm": 1.1342120170593262,
+      "learning_rate": 6.407731940827333e-11,
+      "loss": 0.1983,
+      "step": 6326
+    },
+    {
+      "epoch": 2.9959159514649305,
+      "grad_norm": 1.020653247833252,
+      "learning_rate": 4.707722759733635e-11,
+      "loss": 0.1911,
+      "step": 6327
+    },
+    {
+      "epoch": 2.9963894643385616,
+      "grad_norm": 0.9396451115608215,
+      "learning_rate": 3.26925270033307e-11,
+      "loss": 0.1879,
+      "step": 6328
+    },
+    {
+      "epoch": 2.996862977212193,
+      "grad_norm": 0.9087719321250916,
+      "learning_rate": 2.0923221385471538e-11,
+      "loss": 0.1948,
+      "step": 6329
+    },
+    {
+      "epoch": 2.997336490085824,
+      "grad_norm": 1.0705500841140747,
+      "learning_rate": 1.1769313825737983e-11,
+      "loss": 0.1777,
+      "step": 6330
+    },
+    {
+      "epoch": 2.997810002959455,
+      "grad_norm": 1.0305906534194946,
+      "learning_rate": 5.230806714440206e-12,
+      "loss": 0.1944,
+      "step": 6331
+    },
+    {
+      "epoch": 2.9982835158330867,
+      "grad_norm": 1.1944894790649414,
+      "learning_rate": 1.307701764652336e-12,
+      "loss": 0.1901,
+      "step": 6332
+    },
+    {
+      "epoch": 2.9987570287067182,
+      "grad_norm": 1.1254169940948486,
+      "learning_rate": 0.0,
+      "loss": 0.1845,
+      "step": 6333
+    },
+    {
+      "epoch": 2.9987570287067182,
+      "step": 6333,
+      "total_flos": 6.209261276653158e+16,
+      "train_loss": 0.25655047147971993,
+      "train_runtime": 12932.3665,
+      "train_samples_per_second": 62.704,
+      "train_steps_per_second": 0.49
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 6333,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 6.209261276653158e+16,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}