diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,66551 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 9503,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 36.36130527170318,
+      "learning_rate": 3.4965034965034967e-08,
+      "loss": 1.8874,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 44.35509081795796,
+      "learning_rate": 6.993006993006993e-08,
+      "loss": 1.9623,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 113.15059681611625,
+      "learning_rate": 1.048951048951049e-07,
+      "loss": 1.8315,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 26.480686239957798,
+      "learning_rate": 1.3986013986013987e-07,
+      "loss": 1.8153,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 73.55538638754199,
+      "learning_rate": 1.7482517482517484e-07,
+      "loss": 1.8261,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 44.47995785747746,
+      "learning_rate": 2.097902097902098e-07,
+      "loss": 1.7684,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 41.19958226318581,
+      "learning_rate": 2.447552447552448e-07,
+      "loss": 1.7842,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 162.3686148569218,
+      "learning_rate": 2.7972027972027973e-07,
+      "loss": 1.8389,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 37.77584116684587,
+      "learning_rate": 3.1468531468531473e-07,
+      "loss": 1.739,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 33.567385322447045,
+      "learning_rate": 3.496503496503497e-07,
+      "loss": 1.823,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 79.25205422236616,
+      "learning_rate": 3.846153846153847e-07,
+      "loss": 1.8705,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 29.01030979777457,
+      "learning_rate": 4.195804195804196e-07,
+      "loss": 1.8595,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 32.29577345941253,
+      "learning_rate": 4.5454545454545457e-07,
+      "loss": 1.7776,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 75.68576026951285,
+      "learning_rate": 4.895104895104896e-07,
+      "loss": 1.7253,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 44.31708263717978,
+      "learning_rate": 5.244755244755246e-07,
+      "loss": 1.6779,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 219.5878152798679,
+      "learning_rate": 5.594405594405595e-07,
+      "loss": 1.702,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 33.33211308321619,
+      "learning_rate": 5.944055944055945e-07,
+      "loss": 1.5494,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 28.77943791137145,
+      "learning_rate": 6.293706293706295e-07,
+      "loss": 1.5089,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 60.96425095593503,
+      "learning_rate": 6.643356643356644e-07,
+      "loss": 1.4609,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 21.119045219567745,
+      "learning_rate": 6.993006993006994e-07,
+      "loss": 1.4632,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 30.087479668348955,
+      "learning_rate": 7.342657342657343e-07,
+      "loss": 1.5663,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 22.134522059806123,
+      "learning_rate": 7.692307692307694e-07,
+      "loss": 1.4837,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 29.13387995825167,
+      "learning_rate": 8.041958041958043e-07,
+      "loss": 1.4038,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.488083114385395,
+      "learning_rate": 8.391608391608393e-07,
+      "loss": 1.2151,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.410898426509407,
+      "learning_rate": 8.741258741258741e-07,
+      "loss": 1.2384,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.196734276367997,
+      "learning_rate": 9.090909090909091e-07,
+      "loss": 1.2946,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.855727873849648,
+      "learning_rate": 9.44055944055944e-07,
+      "loss": 1.3565,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.775814502806083,
+      "learning_rate": 9.790209790209791e-07,
+      "loss": 1.2246,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.850328552766668,
+      "learning_rate": 1.013986013986014e-06,
+      "loss": 0.8368,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.46226966648362,
+      "learning_rate": 1.0489510489510491e-06,
+      "loss": 0.8374,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.878572158964688,
+      "learning_rate": 1.083916083916084e-06,
+      "loss": 1.1998,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.823282919023532,
+      "learning_rate": 1.118881118881119e-06,
+      "loss": 1.3195,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.031172311822704,
+      "learning_rate": 1.153846153846154e-06,
+      "loss": 1.1547,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.192629192579879,
+      "learning_rate": 1.188811188811189e-06,
+      "loss": 1.1146,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.2298322921309905,
+      "learning_rate": 1.2237762237762238e-06,
+      "loss": 1.0787,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.136735379037229,
+      "learning_rate": 1.258741258741259e-06,
+      "loss": 1.1542,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.0826929997956345,
+      "learning_rate": 1.2937062937062938e-06,
+      "loss": 1.1515,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.642838223265395,
+      "learning_rate": 1.3286713286713287e-06,
+      "loss": 1.0195,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 21.361559216190493,
+      "learning_rate": 1.3636363636363636e-06,
+      "loss": 1.099,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.808653298973119,
+      "learning_rate": 1.3986013986013987e-06,
+      "loss": 1.0705,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.37717034792203,
+      "learning_rate": 1.4335664335664336e-06,
+      "loss": 1.0939,
+      "step": 41
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.615327284739696,
+      "learning_rate": 1.4685314685314685e-06,
+      "loss": 1.0464,
+      "step": 42
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.179391150292978,
+      "learning_rate": 1.5034965034965034e-06,
+      "loss": 1.0878,
+      "step": 43
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 20.006567323936093,
+      "learning_rate": 1.5384615384615387e-06,
+      "loss": 1.1121,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.213316956884923,
+      "learning_rate": 1.5734265734265736e-06,
+      "loss": 1.0629,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.22220233163102,
+      "learning_rate": 1.6083916083916085e-06,
+      "loss": 1.1175,
+      "step": 46
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.945556190990595,
+      "learning_rate": 1.6433566433566434e-06,
+      "loss": 1.0423,
+      "step": 47
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.6470904938849054,
+      "learning_rate": 1.6783216783216785e-06,
+      "loss": 1.0511,
+      "step": 48
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.446567502839809,
+      "learning_rate": 1.7132867132867134e-06,
+      "loss": 1.0117,
+      "step": 49
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.4133295800026815,
+      "learning_rate": 1.7482517482517483e-06,
+      "loss": 1.0825,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.263781826142708,
+      "learning_rate": 1.7832167832167834e-06,
+      "loss": 1.0928,
+      "step": 51
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3410522050072218,
+      "learning_rate": 1.8181818181818183e-06,
+      "loss": 0.9511,
+      "step": 52
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.078099442521805,
+      "learning_rate": 1.8531468531468532e-06,
+      "loss": 1.0446,
+      "step": 53
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7783601770563657,
+      "learning_rate": 1.888111888111888e-06,
+      "loss": 0.9961,
+      "step": 54
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.3245107534214275,
+      "learning_rate": 1.9230769230769234e-06,
+      "loss": 1.002,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.550873551909127,
+      "learning_rate": 1.9580419580419583e-06,
+      "loss": 1.0045,
+      "step": 56
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.298300519654384,
+      "learning_rate": 1.993006993006993e-06,
+      "loss": 0.9899,
+      "step": 57
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.636808401697184,
+      "learning_rate": 2.027972027972028e-06,
+      "loss": 1.0293,
+      "step": 58
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.989845861785016,
+      "learning_rate": 2.0629370629370634e-06,
+      "loss": 0.9782,
+      "step": 59
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.036151597648198,
+      "learning_rate": 2.0979020979020983e-06,
+      "loss": 0.9476,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.748491625435901,
+      "learning_rate": 2.132867132867133e-06,
+      "loss": 1.0607,
+      "step": 61
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.981488543652494,
+      "learning_rate": 2.167832167832168e-06,
+      "loss": 1.0103,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.747396934973974,
+      "learning_rate": 2.202797202797203e-06,
+      "loss": 0.9394,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.959028225064991,
+      "learning_rate": 2.237762237762238e-06,
+      "loss": 1.0223,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.1572731448374745,
+      "learning_rate": 2.2727272727272728e-06,
+      "loss": 0.8474,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.264763279942374,
+      "learning_rate": 2.307692307692308e-06,
+      "loss": 1.0108,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.095523926263317,
+      "learning_rate": 2.342657342657343e-06,
+      "loss": 0.9966,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0378141611911293,
+      "learning_rate": 2.377622377622378e-06,
+      "loss": 0.6418,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.350034886280204,
+      "learning_rate": 2.4125874125874128e-06,
+      "loss": 0.9833,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.400350695457104,
+      "learning_rate": 2.4475524475524477e-06,
+      "loss": 1.014,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.539091480211638,
+      "learning_rate": 2.4825174825174825e-06,
+      "loss": 0.9588,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.76197207482864,
+      "learning_rate": 2.517482517482518e-06,
+      "loss": 1.0505,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.105444561403831,
+      "learning_rate": 2.5524475524475528e-06,
+      "loss": 1.0268,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.808311827182721,
+      "learning_rate": 2.5874125874125877e-06,
+      "loss": 0.9266,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3062924637115896,
+      "learning_rate": 2.6223776223776225e-06,
+      "loss": 0.9497,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.359465673010584,
+      "learning_rate": 2.6573426573426574e-06,
+      "loss": 1.0061,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.5894653414802122,
+      "learning_rate": 2.6923076923076923e-06,
+      "loss": 0.9365,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.6964431945565215,
+      "learning_rate": 2.7272727272727272e-06,
+      "loss": 0.9945,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.554710415267673,
+      "learning_rate": 2.762237762237762e-06,
+      "loss": 0.909,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.6224966193912933,
+      "learning_rate": 2.7972027972027974e-06,
+      "loss": 0.9841,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.100484077566709,
+      "learning_rate": 2.8321678321678323e-06,
+      "loss": 0.9785,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7434549809880306,
+      "learning_rate": 2.8671328671328672e-06,
+      "loss": 0.9091,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.910473094338652,
+      "learning_rate": 2.902097902097902e-06,
+      "loss": 0.9128,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.681590491267232,
+      "learning_rate": 2.937062937062937e-06,
+      "loss": 0.9723,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8481519631092667,
+      "learning_rate": 2.972027972027972e-06,
+      "loss": 0.9126,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.061017336429524,
+      "learning_rate": 3.006993006993007e-06,
+      "loss": 0.949,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1054284861421526,
+      "learning_rate": 3.0419580419580425e-06,
+      "loss": 0.954,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.216222090801437,
+      "learning_rate": 3.0769230769230774e-06,
+      "loss": 0.999,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.344055127943243,
+      "learning_rate": 3.1118881118881123e-06,
+      "loss": 0.8865,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.916337109234207,
+      "learning_rate": 3.1468531468531472e-06,
+      "loss": 0.9985,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.646273915281262,
+      "learning_rate": 3.181818181818182e-06,
+      "loss": 0.9368,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2528928171773566,
+      "learning_rate": 3.216783216783217e-06,
+      "loss": 0.8946,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.670080997940339,
+      "learning_rate": 3.251748251748252e-06,
+      "loss": 0.883,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.3498888254458885,
+      "learning_rate": 3.286713286713287e-06,
+      "loss": 0.8334,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4044098902580067,
+      "learning_rate": 3.321678321678322e-06,
+      "loss": 0.8982,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.964267535512436,
+      "learning_rate": 3.356643356643357e-06,
+      "loss": 0.901,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.2547749034420175,
+      "learning_rate": 3.391608391608392e-06,
+      "loss": 0.888,
+      "step": 97
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.6445555952684594,
+      "learning_rate": 3.426573426573427e-06,
+      "loss": 0.9357,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.6183641170544676,
+      "learning_rate": 3.4615384615384617e-06,
+      "loss": 0.8898,
+      "step": 99
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.683704401531074,
+      "learning_rate": 3.4965034965034966e-06,
+      "loss": 0.9716,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.642503279279386,
+      "learning_rate": 3.5314685314685315e-06,
+      "loss": 0.8517,
+      "step": 101
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.5861147598949064,
+      "learning_rate": 3.566433566433567e-06,
+      "loss": 0.8891,
+      "step": 102
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.15820179982336,
+      "learning_rate": 3.6013986013986017e-06,
+      "loss": 0.9273,
+      "step": 103
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0966116013351526,
+      "learning_rate": 3.6363636363636366e-06,
+      "loss": 0.8954,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.590992779608164,
+      "learning_rate": 3.6713286713286715e-06,
+      "loss": 0.8935,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.257062049913251,
+      "learning_rate": 3.7062937062937064e-06,
+      "loss": 0.7903,
+      "step": 106
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1160354318045598,
+      "learning_rate": 3.7412587412587413e-06,
+      "loss": 0.9888,
+      "step": 107
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.557620314215135,
+      "learning_rate": 3.776223776223776e-06,
+      "loss": 0.8218,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.690586014565839,
+      "learning_rate": 3.811188811188811e-06,
+      "loss": 0.9652,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.649988211859253,
+      "learning_rate": 3.846153846153847e-06,
+      "loss": 0.874,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4358159594855593,
+      "learning_rate": 3.881118881118881e-06,
+      "loss": 0.8646,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.275525174285379,
+      "learning_rate": 3.916083916083917e-06,
+      "loss": 0.8884,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.163330070858521,
+      "learning_rate": 3.951048951048951e-06,
+      "loss": 0.8856,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0432246966048093,
+      "learning_rate": 3.986013986013986e-06,
+      "loss": 0.6153,
+      "step": 114
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.6243328469121807,
+      "learning_rate": 4.020979020979021e-06,
+      "loss": 0.9266,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7038828259874026,
+      "learning_rate": 4.055944055944056e-06,
+      "loss": 0.8613,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.587444425327127,
+      "learning_rate": 4.0909090909090915e-06,
+      "loss": 0.8677,
+      "step": 117
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8022237219414,
+      "learning_rate": 4.125874125874127e-06,
+      "loss": 0.9397,
+      "step": 118
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.591322074020894,
+      "learning_rate": 4.160839160839161e-06,
+      "loss": 0.8731,
+      "step": 119
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.273093982425321,
+      "learning_rate": 4.195804195804197e-06,
+      "loss": 0.8526,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.197128473907411,
+      "learning_rate": 4.230769230769231e-06,
+      "loss": 0.8944,
+      "step": 121
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.223590603807003,
+      "learning_rate": 4.265734265734266e-06,
+      "loss": 0.8939,
+      "step": 122
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7474696753810433,
+      "learning_rate": 4.300699300699301e-06,
+      "loss": 0.893,
+      "step": 123
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.509667764665236,
+      "learning_rate": 4.335664335664336e-06,
+      "loss": 0.8958,
+      "step": 124
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4746005923931715,
+      "learning_rate": 4.3706293706293715e-06,
+      "loss": 0.9511,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8752998189935024,
+      "learning_rate": 4.405594405594406e-06,
+      "loss": 0.9831,
+      "step": 126
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6750347835184956,
+      "learning_rate": 4.440559440559441e-06,
+      "loss": 0.8389,
+      "step": 127
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.059225334055095,
+      "learning_rate": 4.475524475524476e-06,
+      "loss": 0.9191,
+      "step": 128
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8492139079466123,
+      "learning_rate": 4.510489510489511e-06,
+      "loss": 0.9256,
+      "step": 129
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.466866349881368,
+      "learning_rate": 4.5454545454545455e-06,
+      "loss": 0.9199,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5173267002299946,
+      "learning_rate": 4.580419580419581e-06,
+      "loss": 0.8514,
+      "step": 131
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.6745126525258236,
+      "learning_rate": 4.615384615384616e-06,
+      "loss": 0.9873,
+      "step": 132
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9338742337892234,
+      "learning_rate": 4.650349650349651e-06,
+      "loss": 0.8569,
+      "step": 133
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.843844549233959,
+      "learning_rate": 4.685314685314686e-06,
+      "loss": 0.9136,
+      "step": 134
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.624661321804515,
+      "learning_rate": 4.72027972027972e-06,
+      "loss": 0.9501,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.175344406567936,
+      "learning_rate": 4.755244755244756e-06,
+      "loss": 0.7571,
+      "step": 136
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9193507761169832,
+      "learning_rate": 4.79020979020979e-06,
+      "loss": 0.8804,
+      "step": 137
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.136679694610541,
+      "learning_rate": 4.8251748251748255e-06,
+      "loss": 0.8397,
+      "step": 138
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.002722016275543,
+      "learning_rate": 4.86013986013986e-06,
+      "loss": 0.7947,
+      "step": 139
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.059401974397323,
+      "learning_rate": 4.895104895104895e-06,
+      "loss": 0.9127,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.278913674680803,
+      "learning_rate": 4.930069930069931e-06,
+      "loss": 0.88,
+      "step": 141
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.830877808380784,
+      "learning_rate": 4.965034965034965e-06,
+      "loss": 0.9,
+      "step": 142
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.983823394564358,
+      "learning_rate": 5e-06,
+      "loss": 0.9034,
+      "step": 143
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.804644998821451,
+      "learning_rate": 5.034965034965036e-06,
+      "loss": 0.7911,
+      "step": 144
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.991083677118832,
+      "learning_rate": 5.06993006993007e-06,
+      "loss": 0.9021,
+      "step": 145
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.151416376092996,
+      "learning_rate": 5.1048951048951055e-06,
+      "loss": 0.9136,
+      "step": 146
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.2803307360013227,
+      "learning_rate": 5.13986013986014e-06,
+      "loss": 0.862,
+      "step": 147
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.2255731360831343,
+      "learning_rate": 5.174825174825175e-06,
+      "loss": 0.8045,
+      "step": 148
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.688893991857855,
+      "learning_rate": 5.20979020979021e-06,
+      "loss": 0.8606,
+      "step": 149
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0166064964847963,
+      "learning_rate": 5.244755244755245e-06,
+      "loss": 0.8728,
+      "step": 150
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.697806716640856,
+      "learning_rate": 5.27972027972028e-06,
+      "loss": 0.8277,
+      "step": 151
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9797528895264196,
+      "learning_rate": 5.314685314685315e-06,
+      "loss": 0.8832,
+      "step": 152
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.5741820214637494,
+      "learning_rate": 5.34965034965035e-06,
+      "loss": 0.7884,
+      "step": 153
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.904019428236043,
+      "learning_rate": 5.384615384615385e-06,
+      "loss": 0.8818,
+      "step": 154
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.113543577877246,
+      "learning_rate": 5.41958041958042e-06,
+      "loss": 0.8515,
+      "step": 155
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.916916159534253,
+      "learning_rate": 5.4545454545454545e-06,
+      "loss": 0.883,
+      "step": 156
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.026332966896275,
+      "learning_rate": 5.48951048951049e-06,
+      "loss": 0.8202,
+      "step": 157
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.2747535576313895,
+      "learning_rate": 5.524475524475524e-06,
+      "loss": 0.8201,
+      "step": 158
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.915673342738748,
+      "learning_rate": 5.5594405594405596e-06,
+      "loss": 0.8054,
+      "step": 159
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8868567094503492,
+      "learning_rate": 5.594405594405595e-06,
+      "loss": 0.5897,
+      "step": 160
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.031567311252644,
+      "learning_rate": 5.629370629370629e-06,
+      "loss": 0.9305,
+      "step": 161
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.750704520517035,
+      "learning_rate": 5.664335664335665e-06,
+      "loss": 0.8506,
+      "step": 162
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9833189282740107,
+      "learning_rate": 5.699300699300699e-06,
+      "loss": 0.7708,
+      "step": 163
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.354822607813489,
+      "learning_rate": 5.7342657342657345e-06,
+      "loss": 0.8819,
+      "step": 164
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.130090251045954,
+      "learning_rate": 5.769230769230769e-06,
+      "loss": 0.769,
+      "step": 165
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.0268037848757725,
+      "learning_rate": 5.804195804195804e-06,
+      "loss": 0.8636,
+      "step": 166
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.590445411100409,
+      "learning_rate": 5.83916083916084e-06,
+      "loss": 0.9014,
+      "step": 167
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9573676244626235,
+      "learning_rate": 5.874125874125874e-06,
+      "loss": 0.9767,
+      "step": 168
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.7976970282517706,
+      "learning_rate": 5.90909090909091e-06,
+      "loss": 0.7677,
+      "step": 169
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.20200552669586,
+      "learning_rate": 5.944055944055944e-06,
+      "loss": 0.8744,
+      "step": 170
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.403103369719815,
+      "learning_rate": 5.97902097902098e-06,
+      "loss": 0.7684,
+      "step": 171
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.688672125161367,
+      "learning_rate": 6.013986013986014e-06,
+      "loss": 0.828,
+      "step": 172
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.300203165530322,
+      "learning_rate": 6.04895104895105e-06,
+      "loss": 0.8845,
+      "step": 173
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9549829165853074,
+      "learning_rate": 6.083916083916085e-06,
+      "loss": 0.8396,
+      "step": 174
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.137866556517222,
+      "learning_rate": 6.1188811188811196e-06,
+      "loss": 0.8187,
+      "step": 175
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.821237941276991,
+      "learning_rate": 6.153846153846155e-06,
+      "loss": 0.8752,
+      "step": 176
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.1343880638693324,
+      "learning_rate": 6.188811188811189e-06,
+      "loss": 0.8322,
+      "step": 177
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.187172327993686,
+      "learning_rate": 6.223776223776225e-06,
+      "loss": 0.7712,
+      "step": 178
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.822050694898768,
+      "learning_rate": 6.258741258741259e-06,
+      "loss": 0.8855,
+      "step": 179
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.567034848940679,
+      "learning_rate": 6.2937062937062944e-06,
+      "loss": 0.8808,
+      "step": 180
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.878919913487739,
+      "learning_rate": 6.32867132867133e-06,
+      "loss": 0.816,
+      "step": 181
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.235297169720441,
+      "learning_rate": 6.363636363636364e-06,
+      "loss": 0.7188,
+      "step": 182
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.930819892851912,
+      "learning_rate": 6.3986013986013996e-06,
+      "loss": 0.9284,
+      "step": 183
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.670121712591447,
+      "learning_rate": 6.433566433566434e-06,
+      "loss": 0.7796,
+      "step": 184
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6886770215888207,
+      "learning_rate": 6.468531468531469e-06,
+      "loss": 0.8803,
+      "step": 185
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.5148986203471853,
+      "learning_rate": 6.503496503496504e-06,
+      "loss": 0.8603,
+      "step": 186
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.671074065374444,
+      "learning_rate": 6.538461538461539e-06,
+      "loss": 0.8593,
+      "step": 187
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.011761895974061,
+      "learning_rate": 6.573426573426574e-06,
+      "loss": 0.8902,
+      "step": 188
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.48637995228731,
+      "learning_rate": 6.608391608391609e-06,
+      "loss": 0.7607,
+      "step": 189
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.3132599184181504,
+      "learning_rate": 6.643356643356644e-06,
+      "loss": 0.8014,
+      "step": 190
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.055806698354549,
+      "learning_rate": 6.678321678321679e-06,
+      "loss": 0.9812,
+      "step": 191
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.2543072248631977,
+      "learning_rate": 6.713286713286714e-06,
+      "loss": 0.8687,
+      "step": 192
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.040625426096151,
+      "learning_rate": 6.7482517482517485e-06,
+      "loss": 0.7797,
+      "step": 193
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.215620104060743,
+      "learning_rate": 6.783216783216784e-06,
+      "loss": 0.8274,
+      "step": 194
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.471180655628137,
+      "learning_rate": 6.818181818181818e-06,
+      "loss": 0.6187,
+      "step": 195
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.359167344334515,
+      "learning_rate": 6.853146853146854e-06,
+      "loss": 0.8114,
+      "step": 196
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.290199760313723,
+      "learning_rate": 6.888111888111889e-06,
+      "loss": 0.8188,
+      "step": 197
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.885960883267336,
+      "learning_rate": 6.923076923076923e-06,
+      "loss": 0.8622,
+      "step": 198
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.63144492234704,
+      "learning_rate": 6.958041958041959e-06,
+      "loss": 0.9063,
+      "step": 199
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.3612366109316167,
+      "learning_rate": 6.993006993006993e-06,
+      "loss": 0.862,
+      "step": 200
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.586306665535676,
+      "learning_rate": 7.0279720279720285e-06,
+      "loss": 0.8577,
+      "step": 201
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.2176272960473553,
+      "learning_rate": 7.062937062937063e-06,
+      "loss": 0.8225,
+      "step": 202
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.416044727353836,
+      "learning_rate": 7.097902097902098e-06,
+      "loss": 0.8681,
+      "step": 203
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.4811630787160945,
+      "learning_rate": 7.132867132867134e-06,
+      "loss": 0.8767,
+      "step": 204
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.567725865125878,
+      "learning_rate": 7.167832167832168e-06,
+      "loss": 0.8528,
+      "step": 205
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.2885446735714443,
+      "learning_rate": 7.202797202797203e-06,
+      "loss": 0.756,
+      "step": 206
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.884812749781261,
+      "learning_rate": 7.237762237762238e-06,
+      "loss": 0.8742,
+      "step": 207
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.2827496038865385,
+      "learning_rate": 7.272727272727273e-06,
+      "loss": 0.854,
+      "step": 208
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.8769620411554135,
+      "learning_rate": 7.307692307692308e-06,
+      "loss": 0.9437,
+      "step": 209
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.152946273455516,
+      "learning_rate": 7.342657342657343e-06,
+      "loss": 0.7662,
+      "step": 210
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7643870516857074,
+      "learning_rate": 7.377622377622379e-06,
+      "loss": 0.9363,
+      "step": 211
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.519504752467641,
+      "learning_rate": 7.412587412587413e-06,
+      "loss": 0.8056,
+      "step": 212
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7919893275590533,
+      "learning_rate": 7.447552447552449e-06,
+      "loss": 0.6353,
+      "step": 213
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9052671409624167,
+      "learning_rate": 7.4825174825174825e-06,
+      "loss": 0.8335,
+      "step": 214
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.479591408911271,
+      "learning_rate": 7.517482517482519e-06,
+      "loss": 0.8493,
+      "step": 215
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.377375722973853,
+      "learning_rate": 7.552447552447552e-06,
+      "loss": 0.7782,
+      "step": 216
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.2542755743953102,
+      "learning_rate": 7.5874125874125885e-06,
+      "loss": 0.8053,
+      "step": 217
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.2686387521722158,
+      "learning_rate": 7.622377622377622e-06,
+      "loss": 0.8326,
+      "step": 218
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.023160327593423,
+      "learning_rate": 7.657342657342658e-06,
+      "loss": 0.8383,
+      "step": 219
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0319106359639614,
+      "learning_rate": 7.692307692307694e-06,
+      "loss": 0.873,
+      "step": 220
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.857091847979987,
+      "learning_rate": 7.727272727272727e-06,
+      "loss": 0.8656,
+      "step": 221
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.137600063788741,
+      "learning_rate": 7.762237762237763e-06,
+      "loss": 0.7638,
+      "step": 222
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.911586562313934,
+      "learning_rate": 7.797202797202798e-06,
+      "loss": 0.7693,
+      "step": 223
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9977640714571807,
+      "learning_rate": 7.832167832167833e-06,
+      "loss": 0.8815,
+      "step": 224
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6947791098337848,
+      "learning_rate": 7.867132867132867e-06,
+      "loss": 0.8508,
+      "step": 225
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.70865229471086,
+      "learning_rate": 7.902097902097902e-06,
+      "loss": 0.8039,
+      "step": 226
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9607344558352198,
+      "learning_rate": 7.937062937062937e-06,
+      "loss": 0.8913,
+      "step": 227
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7336519000248947,
+      "learning_rate": 7.972027972027973e-06,
+      "loss": 0.855,
+      "step": 228
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.197027691020523,
+      "learning_rate": 8.006993006993008e-06,
+      "loss": 0.8883,
+      "step": 229
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8607529888322873,
+      "learning_rate": 8.041958041958042e-06,
+      "loss": 0.8003,
+      "step": 230
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.811064736453866,
+      "learning_rate": 8.076923076923077e-06,
+      "loss": 0.9102,
+      "step": 231
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9400768388006893,
+      "learning_rate": 8.111888111888112e-06,
+      "loss": 0.6545,
+      "step": 232
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.1601907658724624,
+      "learning_rate": 8.146853146853148e-06,
+      "loss": 0.8757,
+      "step": 233
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.646638408245371,
+      "learning_rate": 8.181818181818183e-06,
+      "loss": 0.8677,
+      "step": 234
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.665181271364294,
+      "learning_rate": 8.216783216783217e-06,
+      "loss": 0.7499,
+      "step": 235
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.519763391278135,
+      "learning_rate": 8.251748251748254e-06,
+      "loss": 0.8092,
+      "step": 236
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.799514568767844,
+      "learning_rate": 8.286713286713287e-06,
+      "loss": 0.8237,
+      "step": 237
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.3167393880149625,
+      "learning_rate": 8.321678321678323e-06,
+      "loss": 0.7958,
+      "step": 238
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.207750765393011,
+      "learning_rate": 8.356643356643356e-06,
+      "loss": 0.9301,
+      "step": 239
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0242485006730004,
+      "learning_rate": 8.391608391608393e-06,
+      "loss": 0.8689,
+      "step": 240
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.061479269990459,
+      "learning_rate": 8.426573426573428e-06,
+      "loss": 0.8528,
+      "step": 241
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.1543956915127973,
+      "learning_rate": 8.461538461538462e-06,
+      "loss": 0.8714,
+      "step": 242
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.237885935376305,
+      "learning_rate": 8.496503496503497e-06,
+      "loss": 0.7606,
+      "step": 243
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.950280179321886,
+      "learning_rate": 8.531468531468533e-06,
+      "loss": 0.6335,
+      "step": 244
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5896370335044245,
+      "learning_rate": 8.566433566433568e-06,
+      "loss": 0.6082,
+      "step": 245
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.404835463536644,
+      "learning_rate": 8.601398601398602e-06,
+      "loss": 0.8319,
+      "step": 246
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.238800843683949,
+      "learning_rate": 8.636363636363637e-06,
+      "loss": 0.7277,
+      "step": 247
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.816682891374414,
+      "learning_rate": 8.671328671328672e-06,
+      "loss": 0.8987,
+      "step": 248
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.937171146872307,
+      "learning_rate": 8.706293706293708e-06,
+      "loss": 0.8465,
+      "step": 249
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.9152180387496465,
+      "learning_rate": 8.741258741258743e-06,
+      "loss": 0.8588,
+      "step": 250
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.2233086259580013,
+      "learning_rate": 8.776223776223777e-06,
+      "loss": 0.7794,
+      "step": 251
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2240451951651212,
+      "learning_rate": 8.811188811188812e-06,
+      "loss": 0.6543,
+      "step": 252
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9198696085904783,
+      "learning_rate": 8.846153846153847e-06,
+      "loss": 0.8946,
+      "step": 253
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0548127678087447,
+      "learning_rate": 8.881118881118883e-06,
+      "loss": 0.8461,
+      "step": 254
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0034454684442564,
+      "learning_rate": 8.916083916083916e-06,
+      "loss": 0.836,
+      "step": 255
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.356298512009719,
+      "learning_rate": 8.951048951048951e-06,
+      "loss": 0.8044,
+      "step": 256
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.7719381130534395,
+      "learning_rate": 8.986013986013987e-06,
+      "loss": 0.8598,
+      "step": 257
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.961248303513742,
+      "learning_rate": 9.020979020979022e-06,
+      "loss": 0.8625,
+      "step": 258
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.494139526881577,
+      "learning_rate": 9.055944055944057e-06,
+      "loss": 0.8373,
+      "step": 259
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.62343900060472,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 0.8142,
+      "step": 260
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.976610214894441,
+      "learning_rate": 9.125874125874126e-06,
+      "loss": 0.8383,
+      "step": 261
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.152334899926377,
+      "learning_rate": 9.160839160839162e-06,
+      "loss": 0.831,
+      "step": 262
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.943305109872976,
+      "learning_rate": 9.195804195804197e-06,
+      "loss": 0.7919,
+      "step": 263
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6262854108938019,
+      "learning_rate": 9.230769230769232e-06,
+      "loss": 0.6016,
+      "step": 264
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.9820347523442305,
+      "learning_rate": 9.265734265734266e-06,
+      "loss": 0.842,
+      "step": 265
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.965150403672179,
+      "learning_rate": 9.300699300699301e-06,
+      "loss": 0.8197,
+      "step": 266
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.015217382657136,
+      "learning_rate": 9.335664335664337e-06,
+      "loss": 0.7963,
+      "step": 267
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.5798922752668783,
+      "learning_rate": 9.370629370629372e-06,
+      "loss": 0.8153,
+      "step": 268
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.150189471491218,
+      "learning_rate": 9.405594405594406e-06,
+      "loss": 0.7447,
+      "step": 269
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.8869071132238133,
+      "learning_rate": 9.44055944055944e-06,
+      "loss": 0.8291,
+      "step": 270
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.031692940937883,
+      "learning_rate": 9.475524475524476e-06,
+      "loss": 0.7182,
+      "step": 271
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.3293803718839365,
+      "learning_rate": 9.510489510489511e-06,
+      "loss": 0.7867,
+      "step": 272
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.411178547420612,
+      "learning_rate": 9.545454545454547e-06,
+      "loss": 0.849,
+      "step": 273
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8859064509058623,
+      "learning_rate": 9.58041958041958e-06,
+      "loss": 0.5875,
+      "step": 274
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9199928735928666,
+      "learning_rate": 9.615384615384616e-06,
+      "loss": 0.8625,
+      "step": 275
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.4204022429701837,
+      "learning_rate": 9.650349650349651e-06,
+      "loss": 0.787,
+      "step": 276
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.280881830881753,
+      "learning_rate": 9.685314685314686e-06,
+      "loss": 0.8484,
+      "step": 277
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9903254168619435,
+      "learning_rate": 9.72027972027972e-06,
+      "loss": 0.8444,
+      "step": 278
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.2907073267044566,
+      "learning_rate": 9.755244755244755e-06,
+      "loss": 0.7562,
+      "step": 279
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.454956915827184,
+      "learning_rate": 9.79020979020979e-06,
+      "loss": 0.7811,
+      "step": 280
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.564600921480734,
+      "learning_rate": 9.825174825174826e-06,
+      "loss": 0.7783,
+      "step": 281
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0241945476082597,
+      "learning_rate": 9.860139860139861e-06,
+      "loss": 0.8345,
+      "step": 282
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.548190169203221,
+      "learning_rate": 9.895104895104895e-06,
+      "loss": 0.7788,
+      "step": 283
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.732400299622023,
+      "learning_rate": 9.93006993006993e-06,
+      "loss": 0.7922,
+      "step": 284
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.878974528064686,
+      "learning_rate": 9.965034965034966e-06,
+      "loss": 0.8064,
+      "step": 285
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.264526689698259,
+      "learning_rate": 1e-05,
+      "loss": 0.8588,
+      "step": 286
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.20198908359177,
+      "learning_rate": 9.999999709557228e-06,
+      "loss": 0.8367,
+      "step": 287
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 20.715771128490395,
+      "learning_rate": 9.999998838228941e-06,
+      "loss": 0.7669,
+      "step": 288
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.3540036271091456,
+      "learning_rate": 9.999997386015244e-06,
+      "loss": 0.863,
+      "step": 289
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.1848307160272125,
+      "learning_rate": 9.999995352916303e-06,
+      "loss": 0.7439,
+      "step": 290
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.152572478387918,
+      "learning_rate": 9.999992738932357e-06,
+      "loss": 0.822,
+      "step": 291
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.5369549798522573,
+      "learning_rate": 9.999989544063708e-06,
+      "loss": 0.7687,
+      "step": 292
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5501867277358583,
+      "learning_rate": 9.999985768310726e-06,
+      "loss": 0.6133,
+      "step": 293
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.319103455407901,
+      "learning_rate": 9.99998141167385e-06,
+      "loss": 0.7729,
+      "step": 294
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.687148563168577,
+      "learning_rate": 9.999976474153589e-06,
+      "loss": 0.8048,
+      "step": 295
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.984437946192857,
+      "learning_rate": 9.999970955750516e-06,
+      "loss": 0.7737,
+      "step": 296
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.5653321300359915,
+      "learning_rate": 9.999964856465268e-06,
+      "loss": 0.7832,
+      "step": 297
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.6017277018504,
+      "learning_rate": 9.999958176298559e-06,
+      "loss": 0.8802,
+      "step": 298
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.4928074350110396,
+      "learning_rate": 9.99995091525116e-06,
+      "loss": 0.8067,
+      "step": 299
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7767296913858037,
+      "learning_rate": 9.999943073323919e-06,
+      "loss": 0.7946,
+      "step": 300
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.560126265511432,
+      "learning_rate": 9.999934650517743e-06,
+      "loss": 0.7549,
+      "step": 301
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.1779121129553207,
+      "learning_rate": 9.999925646833614e-06,
+      "loss": 0.8234,
+      "step": 302
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6269027998737364,
+      "learning_rate": 9.999916062272576e-06,
+      "loss": 0.747,
+      "step": 303
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0096315209518854,
+      "learning_rate": 9.999905896835745e-06,
+      "loss": 0.8391,
+      "step": 304
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.6570640261588303,
+      "learning_rate": 9.999895150524297e-06,
+      "loss": 0.8514,
+      "step": 305
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7168969428743295,
+      "learning_rate": 9.999883823339487e-06,
+      "loss": 0.7491,
+      "step": 306
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.929508346366845,
+      "learning_rate": 9.999871915282625e-06,
+      "loss": 0.8162,
+      "step": 307
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.431569543759362,
+      "learning_rate": 9.999859426355098e-06,
+      "loss": 0.8174,
+      "step": 308
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.6342519486169595,
+      "learning_rate": 9.999846356558356e-06,
+      "loss": 0.8412,
+      "step": 309
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.773720032260141,
+      "learning_rate": 9.999832705893919e-06,
+      "loss": 0.8367,
+      "step": 310
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.9812086033455607,
+      "learning_rate": 9.999818474363368e-06,
+      "loss": 0.8288,
+      "step": 311
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9790057425352723,
+      "learning_rate": 9.999803661968361e-06,
+      "loss": 0.8828,
+      "step": 312
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.06724404084709,
+      "learning_rate": 9.999788268710619e-06,
+      "loss": 0.8277,
+      "step": 313
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.020362580767787,
+      "learning_rate": 9.999772294591927e-06,
+      "loss": 0.7538,
+      "step": 314
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.4612261352729305,
+      "learning_rate": 9.999755739614144e-06,
+      "loss": 0.7895,
+      "step": 315
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.272614641342335,
+      "learning_rate": 9.999738603779192e-06,
+      "loss": 0.6866,
+      "step": 316
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7980881010106364,
+      "learning_rate": 9.999720887089062e-06,
+      "loss": 0.6429,
+      "step": 317
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.3263471950749786,
+      "learning_rate": 9.99970258954581e-06,
+      "loss": 0.8188,
+      "step": 318
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0088862503780187,
+      "learning_rate": 9.999683711151565e-06,
+      "loss": 0.7195,
+      "step": 319
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.87971268033177,
+      "learning_rate": 9.99966425190852e-06,
+      "loss": 0.8251,
+      "step": 320
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.8366240584831233,
+      "learning_rate": 9.999644211818934e-06,
+      "loss": 0.8274,
+      "step": 321
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.3752232102724986,
+      "learning_rate": 9.999623590885135e-06,
+      "loss": 0.7519,
+      "step": 322
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.248073936944215,
+      "learning_rate": 9.999602389109521e-06,
+      "loss": 0.8161,
+      "step": 323
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7887278560705093,
+      "learning_rate": 9.999580606494554e-06,
+      "loss": 0.8526,
+      "step": 324
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6703840967870525,
+      "learning_rate": 9.999558243042763e-06,
+      "loss": 0.763,
+      "step": 325
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.653691408190642,
+      "learning_rate": 9.999535298756749e-06,
+      "loss": 0.8728,
+      "step": 326
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6557511065861665,
+      "learning_rate": 9.999511773639177e-06,
+      "loss": 0.8372,
+      "step": 327
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.9851651458699573,
+      "learning_rate": 9.999487667692778e-06,
+      "loss": 0.8749,
+      "step": 328
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.556775744280704,
+      "learning_rate": 9.999462980920353e-06,
+      "loss": 0.9202,
+      "step": 329
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.1780645043985722,
+      "learning_rate": 9.99943771332477e-06,
+      "loss": 0.7793,
+      "step": 330
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.50054261428372,
+      "learning_rate": 9.999411864908967e-06,
+      "loss": 0.7445,
+      "step": 331
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.416951605036898,
+      "learning_rate": 9.999385435675947e-06,
+      "loss": 0.7777,
+      "step": 332
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7297205360660906,
+      "learning_rate": 9.999358425628777e-06,
+      "loss": 0.6569,
+      "step": 333
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.763919036592778,
+      "learning_rate": 9.999330834770598e-06,
+      "loss": 0.7999,
+      "step": 334
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.3468910031579515,
+      "learning_rate": 9.999302663104611e-06,
+      "loss": 0.8996,
+      "step": 335
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.52656655880427,
+      "learning_rate": 9.999273910634095e-06,
+      "loss": 0.8481,
+      "step": 336
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0210600507512853,
+      "learning_rate": 9.999244577362388e-06,
+      "loss": 0.7786,
+      "step": 337
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.449552196852401,
+      "learning_rate": 9.999214663292896e-06,
+      "loss": 0.7566,
+      "step": 338
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.5228157643635942,
+      "learning_rate": 9.999184168429095e-06,
+      "loss": 0.8273,
+      "step": 339
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8127747250358417,
+      "learning_rate": 9.99915309277453e-06,
+      "loss": 0.7628,
+      "step": 340
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.9561302000674052,
+      "learning_rate": 9.999121436332809e-06,
+      "loss": 0.7372,
+      "step": 341
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.635610731073348,
+      "learning_rate": 9.99908919910761e-06,
+      "loss": 0.8471,
+      "step": 342
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.974938379443744,
+      "learning_rate": 9.99905638110268e-06,
+      "loss": 0.8192,
+      "step": 343
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.7322849018695665,
+      "learning_rate": 9.99902298232183e-06,
+      "loss": 0.8051,
+      "step": 344
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.989115971364437,
+      "learning_rate": 9.998989002768939e-06,
+      "loss": 0.7731,
+      "step": 345
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7573253684038,
+      "learning_rate": 9.99895444244796e-06,
+      "loss": 0.7253,
+      "step": 346
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0060806455923066,
+      "learning_rate": 9.998919301362902e-06,
+      "loss": 0.8549,
+      "step": 347
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.813781196605346,
+      "learning_rate": 9.998883579517849e-06,
+      "loss": 0.8518,
+      "step": 348
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0126862479542558,
+      "learning_rate": 9.998847276916953e-06,
+      "loss": 0.7217,
+      "step": 349
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5842429131791644,
+      "learning_rate": 9.99881039356443e-06,
+      "loss": 0.8056,
+      "step": 350
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.153516334642533,
+      "learning_rate": 9.998772929464567e-06,
+      "loss": 0.827,
+      "step": 351
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.955277320770189,
+      "learning_rate": 9.998734884621714e-06,
+      "loss": 0.8264,
+      "step": 352
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3636272885829444,
+      "learning_rate": 9.998696259040292e-06,
+      "loss": 0.8436,
+      "step": 353
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6445985660522586,
+      "learning_rate": 9.99865705272479e-06,
+      "loss": 0.7755,
+      "step": 354
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.237081162467692,
+      "learning_rate": 9.99861726567976e-06,
+      "loss": 0.7811,
+      "step": 355
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6981850184698,
+      "learning_rate": 9.998576897909826e-06,
+      "loss": 0.7442,
+      "step": 356
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.1062173431704796,
+      "learning_rate": 9.998535949419676e-06,
+      "loss": 0.8203,
+      "step": 357
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.337004723227758,
+      "learning_rate": 9.99849442021407e-06,
+      "loss": 0.8208,
+      "step": 358
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.14701609390736,
+      "learning_rate": 9.99845231029783e-06,
+      "loss": 0.8006,
+      "step": 359
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.19324654121165,
+      "learning_rate": 9.998409619675852e-06,
+      "loss": 0.8317,
+      "step": 360
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.676166242806993,
+      "learning_rate": 9.998366348353092e-06,
+      "loss": 0.7146,
+      "step": 361
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7837495615392405,
+      "learning_rate": 9.998322496334579e-06,
+      "loss": 0.6307,
+      "step": 362
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6924764025176575,
+      "learning_rate": 9.998278063625407e-06,
+      "loss": 0.788,
+      "step": 363
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.771669821577446,
+      "learning_rate": 9.998233050230737e-06,
+      "loss": 0.7719,
+      "step": 364
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7022314056519003,
+      "learning_rate": 9.9981874561558e-06,
+      "loss": 0.7276,
+      "step": 365
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.4443745382241926,
+      "learning_rate": 9.998141281405892e-06,
+      "loss": 0.7159,
+      "step": 366
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.638255448321009,
+      "learning_rate": 9.99809452598638e-06,
+      "loss": 0.8132,
+      "step": 367
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.342150185068803,
+      "learning_rate": 9.998047189902693e-06,
+      "loss": 0.7607,
+      "step": 368
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.415218070936139,
+      "learning_rate": 9.997999273160333e-06,
+      "loss": 0.7152,
+      "step": 369
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.393978658986462,
+      "learning_rate": 9.997950775764862e-06,
+      "loss": 0.8176,
+      "step": 370
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.590618917457282,
+      "learning_rate": 9.99790169772192e-06,
+      "loss": 0.7372,
+      "step": 371
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.4930700657768323,
+      "learning_rate": 9.997852039037206e-06,
+      "loss": 0.823,
+      "step": 372
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.487379499776839,
+      "learning_rate": 9.99780179971649e-06,
+      "loss": 0.6554,
+      "step": 373
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.624919386906931,
+      "learning_rate": 9.997750979765606e-06,
+      "loss": 0.8252,
+      "step": 374
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6531319266118483,
+      "learning_rate": 9.997699579190462e-06,
+      "loss": 0.9001,
+      "step": 375
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.769920692517521,
+      "learning_rate": 9.997647597997025e-06,
+      "loss": 0.8726,
+      "step": 376
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8611168209874145,
+      "learning_rate": 9.997595036191338e-06,
+      "loss": 0.8799,
+      "step": 377
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.704800856083731,
+      "learning_rate": 9.997541893779507e-06,
+      "loss": 0.8497,
+      "step": 378
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4415710077044612,
+      "learning_rate": 9.997488170767706e-06,
+      "loss": 0.8527,
+      "step": 379
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.050991550010565,
+      "learning_rate": 9.997433867162174e-06,
+      "loss": 0.7829,
+      "step": 380
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.869557820847034,
+      "learning_rate": 9.997378982969223e-06,
+      "loss": 0.8413,
+      "step": 381
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5652433421128276,
+      "learning_rate": 9.997323518195227e-06,
+      "loss": 0.7852,
+      "step": 382
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0907376752843794,
+      "learning_rate": 9.99726747284663e-06,
+      "loss": 0.8476,
+      "step": 383
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.576052254739924,
+      "learning_rate": 9.997210846929945e-06,
+      "loss": 0.7156,
+      "step": 384
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.957239805878372,
+      "learning_rate": 9.997153640451748e-06,
+      "loss": 0.7981,
+      "step": 385
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.406716857617946,
+      "learning_rate": 9.997095853418685e-06,
+      "loss": 0.7176,
+      "step": 386
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5783679058938773,
+      "learning_rate": 9.997037485837474e-06,
+      "loss": 0.8001,
+      "step": 387
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.2799386939664985,
+      "learning_rate": 9.996978537714891e-06,
+      "loss": 0.7861,
+      "step": 388
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6320652922884773,
+      "learning_rate": 9.996919009057787e-06,
+      "loss": 0.844,
+      "step": 389
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.079074473308074,
+      "learning_rate": 9.996858899873076e-06,
+      "loss": 0.8376,
+      "step": 390
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.810909666983869,
+      "learning_rate": 9.996798210167745e-06,
+      "loss": 0.8102,
+      "step": 391
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.822002229623917,
+      "learning_rate": 9.996736939948838e-06,
+      "loss": 0.7358,
+      "step": 392
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.553150551330481,
+      "learning_rate": 9.996675089223481e-06,
+      "loss": 0.8613,
+      "step": 393
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6973816176709278,
+      "learning_rate": 9.996612657998856e-06,
+      "loss": 0.7688,
+      "step": 394
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.017782068668808,
+      "learning_rate": 9.996549646282214e-06,
+      "loss": 0.8728,
+      "step": 395
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4233265237128307,
+      "learning_rate": 9.99648605408088e-06,
+      "loss": 0.7584,
+      "step": 396
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5786786252897844,
+      "learning_rate": 9.996421881402238e-06,
+      "loss": 0.8619,
+      "step": 397
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.733216464863266,
+      "learning_rate": 9.996357128253747e-06,
+      "loss": 0.7994,
+      "step": 398
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1790269474955832,
+      "learning_rate": 9.996291794642924e-06,
+      "loss": 0.7871,
+      "step": 399
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.506896691185903,
+      "learning_rate": 9.996225880577366e-06,
+      "loss": 0.7843,
+      "step": 400
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3733740763366336,
+      "learning_rate": 9.996159386064728e-06,
+      "loss": 0.743,
+      "step": 401
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3332932132080813,
+      "learning_rate": 9.996092311112734e-06,
+      "loss": 0.7422,
+      "step": 402
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.521593140693123,
+      "learning_rate": 9.996024655729177e-06,
+      "loss": 0.8187,
+      "step": 403
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7935109625374253,
+      "learning_rate": 9.99595641992192e-06,
+      "loss": 0.8251,
+      "step": 404
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.26279670060322,
+      "learning_rate": 9.995887603698886e-06,
+      "loss": 0.7706,
+      "step": 405
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.41303802576548,
+      "learning_rate": 9.99581820706807e-06,
+      "loss": 0.7678,
+      "step": 406
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0303506168562255,
+      "learning_rate": 9.99574823003754e-06,
+      "loss": 0.7319,
+      "step": 407
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.486164975875137,
+      "learning_rate": 9.99567767261542e-06,
+      "loss": 0.8546,
+      "step": 408
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.9414654462050036,
+      "learning_rate": 9.995606534809909e-06,
+      "loss": 0.77,
+      "step": 409
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5654915500442366,
+      "learning_rate": 9.995534816629271e-06,
+      "loss": 0.7913,
+      "step": 410
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6295512034240995,
+      "learning_rate": 9.99546251808184e-06,
+      "loss": 0.8889,
+      "step": 411
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.789997301197883,
+      "learning_rate": 9.995389639176013e-06,
+      "loss": 0.848,
+      "step": 412
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.9172960774366037,
+      "learning_rate": 9.995316179920258e-06,
+      "loss": 0.729,
+      "step": 413
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5099680383958822,
+      "learning_rate": 9.99524214032311e-06,
+      "loss": 0.7368,
+      "step": 414
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.516008292563799,
+      "learning_rate": 9.99516752039317e-06,
+      "loss": 0.8693,
+      "step": 415
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2755006768965584,
+      "learning_rate": 9.995092320139106e-06,
+      "loss": 0.7947,
+      "step": 416
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.276675967191438,
+      "learning_rate": 9.995016539569656e-06,
+      "loss": 0.8244,
+      "step": 417
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4161022069453173,
+      "learning_rate": 9.994940178693624e-06,
+      "loss": 0.8016,
+      "step": 418
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8339702744181325,
+      "learning_rate": 9.99486323751988e-06,
+      "loss": 0.7548,
+      "step": 419
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.975612430718834,
+      "learning_rate": 9.994785716057364e-06,
+      "loss": 0.7163,
+      "step": 420
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5503239423291677,
+      "learning_rate": 9.994707614315084e-06,
+      "loss": 0.7693,
+      "step": 421
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8040417675794083,
+      "learning_rate": 9.99462893230211e-06,
+      "loss": 0.7943,
+      "step": 422
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6155920470952188,
+      "learning_rate": 9.994549670027584e-06,
+      "loss": 0.8104,
+      "step": 423
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6688952806199437,
+      "learning_rate": 9.994469827500716e-06,
+      "loss": 0.7914,
+      "step": 424
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.784372060150453,
+      "learning_rate": 9.99438940473078e-06,
+      "loss": 0.7975,
+      "step": 425
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5929689621810783,
+      "learning_rate": 9.994308401727122e-06,
+      "loss": 0.8029,
+      "step": 426
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.827200496684302,
+      "learning_rate": 9.99422681849915e-06,
+      "loss": 0.8063,
+      "step": 427
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.420731477742907,
+      "learning_rate": 9.994144655056343e-06,
+      "loss": 0.7089,
+      "step": 428
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.1411562499918078,
+      "learning_rate": 9.994061911408245e-06,
+      "loss": 0.8219,
+      "step": 429
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.296664953262252,
+      "learning_rate": 9.993978587564473e-06,
+      "loss": 0.748,
+      "step": 430
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.786778377368977,
+      "learning_rate": 9.993894683534704e-06,
+      "loss": 0.79,
+      "step": 431
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.9069867693590075,
+      "learning_rate": 9.993810199328687e-06,
+      "loss": 0.7826,
+      "step": 432
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7740885681631293,
+      "learning_rate": 9.993725134956235e-06,
+      "loss": 0.8472,
+      "step": 433
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0792423971075777,
+      "learning_rate": 9.993639490427235e-06,
+      "loss": 0.7272,
+      "step": 434
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.505248944163071,
+      "learning_rate": 9.993553265751632e-06,
+      "loss": 0.8125,
+      "step": 435
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.1705332082489175,
+      "learning_rate": 9.993466460939447e-06,
+      "loss": 0.7515,
+      "step": 436
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.068737573080172,
+      "learning_rate": 9.993379076000762e-06,
+      "loss": 0.6471,
+      "step": 437
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0565141356797696,
+      "learning_rate": 9.99329111094573e-06,
+      "loss": 0.7869,
+      "step": 438
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.418500158534013,
+      "learning_rate": 9.993202565784573e-06,
+      "loss": 0.8211,
+      "step": 439
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.423639948102629,
+      "learning_rate": 9.993113440527573e-06,
+      "loss": 0.815,
+      "step": 440
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2704412010849007,
+      "learning_rate": 9.993023735185088e-06,
+      "loss": 0.7843,
+      "step": 441
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.521200294230519,
+      "learning_rate": 9.992933449767538e-06,
+      "loss": 0.7483,
+      "step": 442
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.706970229721735,
+      "learning_rate": 9.992842584285416e-06,
+      "loss": 0.7187,
+      "step": 443
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.390640870602009,
+      "learning_rate": 9.992751138749273e-06,
+      "loss": 0.7721,
+      "step": 444
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5982385192863124,
+      "learning_rate": 9.992659113169736e-06,
+      "loss": 0.7153,
+      "step": 445
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4856303794193093,
+      "learning_rate": 9.992566507557495e-06,
+      "loss": 0.7573,
+      "step": 446
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5444768288105166,
+      "learning_rate": 9.99247332192331e-06,
+      "loss": 0.7668,
+      "step": 447
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.040260161584014,
+      "learning_rate": 9.992379556278006e-06,
+      "loss": 0.7644,
+      "step": 448
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3952337813826174,
+      "learning_rate": 9.992285210632476e-06,
+      "loss": 0.7813,
+      "step": 449
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7875393710927003,
+      "learning_rate": 9.992190284997683e-06,
+      "loss": 0.8625,
+      "step": 450
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3456211144367436,
+      "learning_rate": 9.992094779384651e-06,
+      "loss": 0.7421,
+      "step": 451
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2941981241460914,
+      "learning_rate": 9.991998693804482e-06,
+      "loss": 0.786,
+      "step": 452
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7710253391308113,
+      "learning_rate": 9.991902028268333e-06,
+      "loss": 0.8221,
+      "step": 453
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8127605742355093,
+      "learning_rate": 9.991804782787435e-06,
+      "loss": 0.8029,
+      "step": 454
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.21579763303864,
+      "learning_rate": 9.991706957373088e-06,
+      "loss": 0.8484,
+      "step": 455
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0567607175438702,
+      "learning_rate": 9.991608552036659e-06,
+      "loss": 0.7895,
+      "step": 456
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.337491818683799,
+      "learning_rate": 9.991509566789575e-06,
+      "loss": 0.8293,
+      "step": 457
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0862821534316573,
+      "learning_rate": 9.991410001643338e-06,
+      "loss": 0.7856,
+      "step": 458
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.278358994242391,
+      "learning_rate": 9.991309856609517e-06,
+      "loss": 0.7536,
+      "step": 459
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9648242121722526,
+      "learning_rate": 9.991209131699745e-06,
+      "loss": 0.7508,
+      "step": 460
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.454545521761428,
+      "learning_rate": 9.991107826925724e-06,
+      "loss": 0.7682,
+      "step": 461
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3135550329147647,
+      "learning_rate": 9.991005942299224e-06,
+      "loss": 0.794,
+      "step": 462
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.958497853786959,
+      "learning_rate": 9.990903477832081e-06,
+      "loss": 0.884,
+      "step": 463
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.559497699524832,
+      "learning_rate": 9.990800433536198e-06,
+      "loss": 0.7475,
+      "step": 464
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2979423649720405,
+      "learning_rate": 9.990696809423551e-06,
+      "loss": 0.8139,
+      "step": 465
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.477081620098315,
+      "learning_rate": 9.990592605506172e-06,
+      "loss": 0.7413,
+      "step": 466
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.145253092459398,
+      "learning_rate": 9.990487821796171e-06,
+      "loss": 0.816,
+      "step": 467
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6671842619211605,
+      "learning_rate": 9.99038245830572e-06,
+      "loss": 0.7634,
+      "step": 468
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4700914843679307,
+      "learning_rate": 9.990276515047063e-06,
+      "loss": 0.7116,
+      "step": 469
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2107804543409255,
+      "learning_rate": 9.990169992032506e-06,
+      "loss": 0.8021,
+      "step": 470
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1867548251004902,
+      "learning_rate": 9.990062889274423e-06,
+      "loss": 0.8221,
+      "step": 471
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.1546753580491793,
+      "learning_rate": 9.989955206785258e-06,
+      "loss": 0.8515,
+      "step": 472
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.845940673194977,
+      "learning_rate": 9.989846944577524e-06,
+      "loss": 0.7823,
+      "step": 473
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.782847174447964,
+      "learning_rate": 9.989738102663792e-06,
+      "loss": 0.8043,
+      "step": 474
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5304700280693413,
+      "learning_rate": 9.989628681056716e-06,
+      "loss": 0.6264,
+      "step": 475
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4549669242826684,
+      "learning_rate": 9.989518679769e-06,
+      "loss": 0.8646,
+      "step": 476
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.704340279572477,
+      "learning_rate": 9.989408098813429e-06,
+      "loss": 0.8055,
+      "step": 477
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5055698906785566,
+      "learning_rate": 9.989296938202846e-06,
+      "loss": 0.7233,
+      "step": 478
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.2218415084935477,
+      "learning_rate": 9.989185197950168e-06,
+      "loss": 0.8227,
+      "step": 479
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3617802926725933,
+      "learning_rate": 9.989072878068376e-06,
+      "loss": 0.8324,
+      "step": 480
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4468315450678713,
+      "learning_rate": 9.98895997857052e-06,
+      "loss": 0.7699,
+      "step": 481
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.667742814508917,
+      "learning_rate": 9.988846499469714e-06,
+      "loss": 0.9051,
+      "step": 482
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7213442322657078,
+      "learning_rate": 9.988732440779145e-06,
+      "loss": 0.7052,
+      "step": 483
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.326845379917123,
+      "learning_rate": 9.98861780251206e-06,
+      "loss": 0.7763,
+      "step": 484
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.9136376815635368,
+      "learning_rate": 9.98850258468178e-06,
+      "loss": 0.7103,
+      "step": 485
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2455870076795366,
+      "learning_rate": 9.988386787301689e-06,
+      "loss": 0.8498,
+      "step": 486
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.697658986196603,
+      "learning_rate": 9.988270410385242e-06,
+      "loss": 0.7247,
+      "step": 487
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6037250054010124,
+      "learning_rate": 9.98815345394596e-06,
+      "loss": 0.7998,
+      "step": 488
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2732857324114515,
+      "learning_rate": 9.988035917997426e-06,
+      "loss": 0.7466,
+      "step": 489
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.447532792950365,
+      "learning_rate": 9.9879178025533e-06,
+      "loss": 0.7684,
+      "step": 490
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4996822315622005,
+      "learning_rate": 9.987799107627301e-06,
+      "loss": 0.7892,
+      "step": 491
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4392244915064354,
+      "learning_rate": 9.987679833233219e-06,
+      "loss": 0.7154,
+      "step": 492
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.01944378814734,
+      "learning_rate": 9.987559979384913e-06,
+      "loss": 0.8231,
+      "step": 493
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3479953891689944,
+      "learning_rate": 9.987439546096309e-06,
+      "loss": 0.7696,
+      "step": 494
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.095707445377812,
+      "learning_rate": 9.987318533381391e-06,
+      "loss": 0.7402,
+      "step": 495
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.6125099373264056,
+      "learning_rate": 9.987196941254226e-06,
+      "loss": 0.8434,
+      "step": 496
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.407463112602799,
+      "learning_rate": 9.987074769728936e-06,
+      "loss": 0.7014,
+      "step": 497
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2803041227883676,
+      "learning_rate": 9.986952018819715e-06,
+      "loss": 0.7015,
+      "step": 498
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.850670147969541,
+      "learning_rate": 9.986828688540825e-06,
+      "loss": 0.8059,
+      "step": 499
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5894833396875727,
+      "learning_rate": 9.986704778906594e-06,
+      "loss": 0.8364,
+      "step": 500
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.310359350175941,
+      "learning_rate": 9.986580289931416e-06,
+      "loss": 0.854,
+      "step": 501
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3368374036518098,
+      "learning_rate": 9.986455221629754e-06,
+      "loss": 0.5978,
+      "step": 502
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4931722034378425,
+      "learning_rate": 9.98632957401614e-06,
+      "loss": 0.8525,
+      "step": 503
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.644787385474121,
+      "learning_rate": 9.986203347105168e-06,
+      "loss": 0.8037,
+      "step": 504
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.563442064696163,
+      "learning_rate": 9.986076540911507e-06,
+      "loss": 0.8657,
+      "step": 505
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1165485781851836,
+      "learning_rate": 9.985949155449885e-06,
+      "loss": 0.7919,
+      "step": 506
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.471404486523451,
+      "learning_rate": 9.985821190735104e-06,
+      "loss": 0.7542,
+      "step": 507
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3249602966837206,
+      "learning_rate": 9.98569264678203e-06,
+      "loss": 0.8402,
+      "step": 508
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.48495778605686,
+      "learning_rate": 9.985563523605597e-06,
+      "loss": 0.721,
+      "step": 509
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4276988470343577,
+      "learning_rate": 9.985433821220805e-06,
+      "loss": 0.8305,
+      "step": 510
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.266343385980247,
+      "learning_rate": 9.985303539642721e-06,
+      "loss": 0.8618,
+      "step": 511
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6368944571388737,
+      "learning_rate": 9.985172678886486e-06,
+      "loss": 0.7497,
+      "step": 512
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3362271184394703,
+      "learning_rate": 9.985041238967297e-06,
+      "loss": 0.7608,
+      "step": 513
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.323494738235158,
+      "learning_rate": 9.984909219900429e-06,
+      "loss": 0.8141,
+      "step": 514
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2944704925260733,
+      "learning_rate": 9.984776621701218e-06,
+      "loss": 0.7668,
+      "step": 515
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.539104026728401,
+      "learning_rate": 9.984643444385067e-06,
+      "loss": 0.6898,
+      "step": 516
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2795736059424563,
+      "learning_rate": 9.984509687967451e-06,
+      "loss": 0.8352,
+      "step": 517
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5063326762752127,
+      "learning_rate": 9.984375352463908e-06,
+      "loss": 0.7219,
+      "step": 518
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.875219661862659,
+      "learning_rate": 9.984240437890045e-06,
+      "loss": 0.7054,
+      "step": 519
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8779919468667536,
+      "learning_rate": 9.984104944261536e-06,
+      "loss": 0.8438,
+      "step": 520
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.222566644720778,
+      "learning_rate": 9.983968871594121e-06,
+      "loss": 0.7997,
+      "step": 521
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.959982902693521,
+      "learning_rate": 9.98383221990361e-06,
+      "loss": 0.7926,
+      "step": 522
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.1528886906910403,
+      "learning_rate": 9.983694989205882e-06,
+      "loss": 0.8242,
+      "step": 523
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.18977821015223,
+      "learning_rate": 9.983557179516872e-06,
+      "loss": 0.757,
+      "step": 524
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5734084664110641,
+      "learning_rate": 9.983418790852597e-06,
+      "loss": 0.6629,
+      "step": 525
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.324375311018105,
+      "learning_rate": 9.983279823229132e-06,
+      "loss": 0.8114,
+      "step": 526
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.124013190729634,
+      "learning_rate": 9.983140276662621e-06,
+      "loss": 0.779,
+      "step": 527
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.535001843346559,
+      "learning_rate": 9.98300015116928e-06,
+      "loss": 0.7235,
+      "step": 528
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1443674722870116,
+      "learning_rate": 9.982859446765385e-06,
+      "loss": 0.8205,
+      "step": 529
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.3022865226413534,
+      "learning_rate": 9.982718163467282e-06,
+      "loss": 0.7919,
+      "step": 530
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2210575024061376,
+      "learning_rate": 9.982576301291387e-06,
+      "loss": 0.8224,
+      "step": 531
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.266820994801646,
+      "learning_rate": 9.982433860254181e-06,
+      "loss": 0.8027,
+      "step": 532
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.8366928780887593,
+      "learning_rate": 9.982290840372212e-06,
+      "loss": 0.7632,
+      "step": 533
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7429115320708926,
+      "learning_rate": 9.982147241662097e-06,
+      "loss": 0.8406,
+      "step": 534
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.324430888530594,
+      "learning_rate": 9.982003064140515e-06,
+      "loss": 0.7881,
+      "step": 535
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.519378774537875,
+      "learning_rate": 9.98185830782422e-06,
+      "loss": 0.7492,
+      "step": 536
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2247325308571977,
+      "learning_rate": 9.981712972730027e-06,
+      "loss": 0.7321,
+      "step": 537
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.367419912672573,
+      "learning_rate": 9.981567058874822e-06,
+      "loss": 0.7559,
+      "step": 538
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.442752211560955,
+      "learning_rate": 9.981420566275554e-06,
+      "loss": 0.7823,
+      "step": 539
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.718090756388608,
+      "learning_rate": 9.981273494949247e-06,
+      "loss": 0.7689,
+      "step": 540
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3168535978993352,
+      "learning_rate": 9.981125844912985e-06,
+      "loss": 0.6097,
+      "step": 541
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.39423124719434,
+      "learning_rate": 9.98097761618392e-06,
+      "loss": 0.8062,
+      "step": 542
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6461362569919227,
+      "learning_rate": 9.980828808779275e-06,
+      "loss": 0.8499,
+      "step": 543
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6944205009407094,
+      "learning_rate": 9.980679422716336e-06,
+      "loss": 0.8563,
+      "step": 544
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.419123667806579,
+      "learning_rate": 9.98052945801246e-06,
+      "loss": 0.8623,
+      "step": 545
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7435396717539557,
+      "learning_rate": 9.980378914685069e-06,
+      "loss": 0.7444,
+      "step": 546
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.662949276618667,
+      "learning_rate": 9.980227792751653e-06,
+      "loss": 0.7362,
+      "step": 547
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.9210580658144902,
+      "learning_rate": 9.980076092229767e-06,
+      "loss": 0.7805,
+      "step": 548
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6271159104021105,
+      "learning_rate": 9.979923813137039e-06,
+      "loss": 0.7591,
+      "step": 549
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1688198714122695,
+      "learning_rate": 9.979770955491154e-06,
+      "loss": 0.7107,
+      "step": 550
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.3806001298954946,
+      "learning_rate": 9.979617519309878e-06,
+      "loss": 0.7611,
+      "step": 551
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6186345183581548,
+      "learning_rate": 9.97946350461103e-06,
+      "loss": 0.8052,
+      "step": 552
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.434402348311883,
+      "learning_rate": 9.979308911412508e-06,
+      "loss": 0.7546,
+      "step": 553
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1361701202500973,
+      "learning_rate": 9.979153739732273e-06,
+      "loss": 0.6987,
+      "step": 554
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.638578499216826,
+      "learning_rate": 9.978997989588346e-06,
+      "loss": 0.8184,
+      "step": 555
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.397169992403867,
+      "learning_rate": 9.978841660998827e-06,
+      "loss": 0.748,
+      "step": 556
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6871052832438886,
+      "learning_rate": 9.978684753981875e-06,
+      "loss": 0.8245,
+      "step": 557
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.127649979618681,
+      "learning_rate": 9.978527268555723e-06,
+      "loss": 0.7135,
+      "step": 558
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.500931002963333,
+      "learning_rate": 9.97836920473866e-06,
+      "loss": 0.7869,
+      "step": 559
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6065448044913357,
+      "learning_rate": 9.978210562549057e-06,
+      "loss": 0.7589,
+      "step": 560
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2848105655651274,
+      "learning_rate": 9.978051342005342e-06,
+      "loss": 0.7439,
+      "step": 561
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3300511317144,
+      "learning_rate": 9.97789154312601e-06,
+      "loss": 0.731,
+      "step": 562
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.53982707617505,
+      "learning_rate": 9.97773116592963e-06,
+      "loss": 0.7874,
+      "step": 563
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.0472971560396367,
+      "learning_rate": 9.977570210434831e-06,
+      "loss": 0.7261,
+      "step": 564
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.883803485999733,
+      "learning_rate": 9.977408676660314e-06,
+      "loss": 0.8079,
+      "step": 565
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 19.573290818215536,
+      "learning_rate": 9.977246564624845e-06,
+      "loss": 0.7604,
+      "step": 566
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6716091098325245,
+      "learning_rate": 9.977083874347258e-06,
+      "loss": 0.6414,
+      "step": 567
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6901090885051846,
+      "learning_rate": 9.976920605846452e-06,
+      "loss": 0.8105,
+      "step": 568
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7134792536923555,
+      "learning_rate": 9.976756759141399e-06,
+      "loss": 0.7635,
+      "step": 569
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7086458704405434,
+      "learning_rate": 9.976592334251132e-06,
+      "loss": 0.8302,
+      "step": 570
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1306874134793508,
+      "learning_rate": 9.976427331194753e-06,
+      "loss": 0.7362,
+      "step": 571
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.45895557799812,
+      "learning_rate": 9.976261749991433e-06,
+      "loss": 0.7854,
+      "step": 572
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2266882437527475,
+      "learning_rate": 9.976095590660407e-06,
+      "loss": 0.8152,
+      "step": 573
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.363404791350861,
+      "learning_rate": 9.975928853220979e-06,
+      "loss": 0.8157,
+      "step": 574
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4298149385499475,
+      "learning_rate": 9.975761537692522e-06,
+      "loss": 0.7796,
+      "step": 575
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2830133914169424,
+      "learning_rate": 9.975593644094472e-06,
+      "loss": 0.7881,
+      "step": 576
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.164741876077585,
+      "learning_rate": 9.975425172446336e-06,
+      "loss": 0.6478,
+      "step": 577
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1615284561481283,
+      "learning_rate": 9.975256122767687e-06,
+      "loss": 0.7038,
+      "step": 578
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.8554861521662316,
+      "learning_rate": 9.975086495078161e-06,
+      "loss": 0.8082,
+      "step": 579
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.8793525267467426,
+      "learning_rate": 9.974916289397469e-06,
+      "loss": 0.7612,
+      "step": 580
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2817512397632775,
+      "learning_rate": 9.974745505745385e-06,
+      "loss": 0.7507,
+      "step": 581
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3293064382183752,
+      "learning_rate": 9.974574144141746e-06,
+      "loss": 0.782,
+      "step": 582
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.420863828032633,
+      "learning_rate": 9.974402204606464e-06,
+      "loss": 0.7904,
+      "step": 583
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.467148177657104,
+      "learning_rate": 9.974229687159515e-06,
+      "loss": 0.7148,
+      "step": 584
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3692761510748506,
+      "learning_rate": 9.974056591820937e-06,
+      "loss": 0.7572,
+      "step": 585
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.410599414340472,
+      "learning_rate": 9.973882918610845e-06,
+      "loss": 0.7463,
+      "step": 586
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.407706605193204,
+      "learning_rate": 9.973708667549413e-06,
+      "loss": 0.7719,
+      "step": 587
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1487892620296885,
+      "learning_rate": 9.973533838656886e-06,
+      "loss": 0.737,
+      "step": 588
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.983088881371681,
+      "learning_rate": 9.973358431953574e-06,
+      "loss": 0.8082,
+      "step": 589
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.84274621933108,
+      "learning_rate": 9.973182447459856e-06,
+      "loss": 0.8005,
+      "step": 590
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.514268336749957,
+      "learning_rate": 9.973005885196177e-06,
+      "loss": 0.8384,
+      "step": 591
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.9786767920760076,
+      "learning_rate": 9.97282874518305e-06,
+      "loss": 0.8312,
+      "step": 592
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.5407053182524333,
+      "learning_rate": 9.972651027441053e-06,
+      "loss": 0.8417,
+      "step": 593
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2034669895851526,
+      "learning_rate": 9.972472731990836e-06,
+      "loss": 0.7066,
+      "step": 594
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3153560832834725,
+      "learning_rate": 9.972293858853111e-06,
+      "loss": 0.7032,
+      "step": 595
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1674873013648661,
+      "learning_rate": 9.972114408048658e-06,
+      "loss": 0.6294,
+      "step": 596
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4983825264182977,
+      "learning_rate": 9.971934379598327e-06,
+      "loss": 0.8138,
+      "step": 597
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.5640167266728517,
+      "learning_rate": 9.971753773523032e-06,
+      "loss": 0.7516,
+      "step": 598
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.819664317552869,
+      "learning_rate": 9.971572589843754e-06,
+      "loss": 0.7164,
+      "step": 599
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.8700762251626,
+      "learning_rate": 9.971390828581546e-06,
+      "loss": 0.7128,
+      "step": 600
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.2350947513381,
+      "learning_rate": 9.971208489757522e-06,
+      "loss": 0.7065,
+      "step": 601
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.78868545134538,
+      "learning_rate": 9.971025573392863e-06,
+      "loss": 0.7491,
+      "step": 602
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.620861295267675,
+      "learning_rate": 9.970842079508827e-06,
+      "loss": 0.8286,
+      "step": 603
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.387001074929076,
+      "learning_rate": 9.970658008126725e-06,
+      "loss": 0.7537,
+      "step": 604
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.356795460296981,
+      "learning_rate": 9.970473359267945e-06,
+      "loss": 0.7542,
+      "step": 605
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9886458699286347,
+      "learning_rate": 9.970288132953938e-06,
+      "loss": 0.6917,
+      "step": 606
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2751385664982706,
+      "learning_rate": 9.970102329206221e-06,
+      "loss": 0.7728,
+      "step": 607
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.495226626436587,
+      "learning_rate": 9.969915948046387e-06,
+      "loss": 0.6747,
+      "step": 608
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4123037303465609,
+      "learning_rate": 9.969728989496081e-06,
+      "loss": 0.6243,
+      "step": 609
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.510919227428744,
+      "learning_rate": 9.96954145357703e-06,
+      "loss": 0.7466,
+      "step": 610
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4441972726765404,
+      "learning_rate": 9.969353340311017e-06,
+      "loss": 0.7598,
+      "step": 611
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0740206980367337,
+      "learning_rate": 9.969164649719898e-06,
+      "loss": 0.7293,
+      "step": 612
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3874311549388647,
+      "learning_rate": 9.968975381825594e-06,
+      "loss": 0.6497,
+      "step": 613
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2328817638501284,
+      "learning_rate": 9.968785536650095e-06,
+      "loss": 0.7843,
+      "step": 614
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.039098314636674,
+      "learning_rate": 9.968595114215453e-06,
+      "loss": 0.7148,
+      "step": 615
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8974999023720256,
+      "learning_rate": 9.968404114543796e-06,
+      "loss": 0.7553,
+      "step": 616
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.363258301216911,
+      "learning_rate": 9.968212537657311e-06,
+      "loss": 0.7465,
+      "step": 617
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5506185481947714,
+      "learning_rate": 9.968020383578253e-06,
+      "loss": 0.8889,
+      "step": 618
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0478689904138068,
+      "learning_rate": 9.96782765232895e-06,
+      "loss": 0.7014,
+      "step": 619
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1272161891964703,
+      "learning_rate": 9.967634343931791e-06,
+      "loss": 0.8825,
+      "step": 620
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3598809615068794,
+      "learning_rate": 9.967440458409232e-06,
+      "loss": 0.7521,
+      "step": 621
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0759823166768627,
+      "learning_rate": 9.967245995783801e-06,
+      "loss": 0.7767,
+      "step": 622
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1835848414184875,
+      "learning_rate": 9.96705095607809e-06,
+      "loss": 0.8266,
+      "step": 623
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.6137036631937254,
+      "learning_rate": 9.966855339314756e-06,
+      "loss": 0.857,
+      "step": 624
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.516218552454981,
+      "learning_rate": 9.966659145516527e-06,
+      "loss": 0.8295,
+      "step": 625
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.179275872622473,
+      "learning_rate": 9.966462374706196e-06,
+      "loss": 0.8447,
+      "step": 626
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0487084736863883,
+      "learning_rate": 9.966265026906622e-06,
+      "loss": 0.7527,
+      "step": 627
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.394315996422968,
+      "learning_rate": 9.966067102140734e-06,
+      "loss": 0.7543,
+      "step": 628
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.0716112519627363,
+      "learning_rate": 9.965868600431525e-06,
+      "loss": 0.7141,
+      "step": 629
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.924340713473136,
+      "learning_rate": 9.965669521802057e-06,
+      "loss": 0.8291,
+      "step": 630
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.8596202937468727,
+      "learning_rate": 9.965469866275457e-06,
+      "loss": 0.7771,
+      "step": 631
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7307465501368644,
+      "learning_rate": 9.965269633874924e-06,
+      "loss": 0.7157,
+      "step": 632
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.407738691497479,
+      "learning_rate": 9.965068824623718e-06,
+      "loss": 0.6772,
+      "step": 633
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0971229549109176,
+      "learning_rate": 9.964867438545166e-06,
+      "loss": 0.7709,
+      "step": 634
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.90404822468166,
+      "learning_rate": 9.964665475662668e-06,
+      "loss": 0.7457,
+      "step": 635
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.213670720198414,
+      "learning_rate": 9.964462935999688e-06,
+      "loss": 0.8134,
+      "step": 636
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.8026473036892554,
+      "learning_rate": 9.964259819579754e-06,
+      "loss": 0.7484,
+      "step": 637
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4715032303759554,
+      "learning_rate": 9.964056126426464e-06,
+      "loss": 0.851,
+      "step": 638
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9713105204657861,
+      "learning_rate": 9.963851856563483e-06,
+      "loss": 0.7706,
+      "step": 639
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0696765042050407,
+      "learning_rate": 9.963647010014541e-06,
+      "loss": 0.7526,
+      "step": 640
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.61319517784048,
+      "learning_rate": 9.963441586803439e-06,
+      "loss": 0.7249,
+      "step": 641
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 35.735839142292704,
+      "learning_rate": 9.963235586954043e-06,
+      "loss": 0.807,
+      "step": 642
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.601165133281576,
+      "learning_rate": 9.963029010490281e-06,
+      "loss": 0.6441,
+      "step": 643
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.700590007062543,
+      "learning_rate": 9.962821857436156e-06,
+      "loss": 0.699,
+      "step": 644
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.83901661028331,
+      "learning_rate": 9.962614127815735e-06,
+      "loss": 0.715,
+      "step": 645
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.2052448381433063,
+      "learning_rate": 9.96240582165315e-06,
+      "loss": 0.7545,
+      "step": 646
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.50569734388418,
+      "learning_rate": 9.962196938972599e-06,
+      "loss": 0.7477,
+      "step": 647
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3913535336435077,
+      "learning_rate": 9.961987479798354e-06,
+      "loss": 0.6525,
+      "step": 648
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.97776453246874,
+      "learning_rate": 9.961777444154747e-06,
+      "loss": 0.7575,
+      "step": 649
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6326354312277207,
+      "learning_rate": 9.96156683206618e-06,
+      "loss": 0.7891,
+      "step": 650
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.590711433166387,
+      "learning_rate": 9.96135564355712e-06,
+      "loss": 0.8471,
+      "step": 651
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.694706096312992,
+      "learning_rate": 9.961143878652104e-06,
+      "loss": 0.7534,
+      "step": 652
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1493792498317736,
+      "learning_rate": 9.960931537375731e-06,
+      "loss": 0.7545,
+      "step": 653
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.547411269212289,
+      "learning_rate": 9.960718619752676e-06,
+      "loss": 0.7998,
+      "step": 654
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.380301424223435,
+      "learning_rate": 9.96050512580767e-06,
+      "loss": 0.6799,
+      "step": 655
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.317215606575152,
+      "learning_rate": 9.960291055565518e-06,
+      "loss": 0.8361,
+      "step": 656
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.76463251057736,
+      "learning_rate": 9.96007640905109e-06,
+      "loss": 0.7474,
+      "step": 657
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.337177345644392,
+      "learning_rate": 9.959861186289324e-06,
+      "loss": 0.7889,
+      "step": 658
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.549090640868925,
+      "learning_rate": 9.95964538730522e-06,
+      "loss": 0.8025,
+      "step": 659
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.530522238161703,
+      "learning_rate": 9.959429012123853e-06,
+      "loss": 0.715,
+      "step": 660
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.515390648368771,
+      "learning_rate": 9.95921206077036e-06,
+      "loss": 0.7851,
+      "step": 661
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.929755415997428,
+      "learning_rate": 9.958994533269947e-06,
+      "loss": 0.7982,
+      "step": 662
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.450266556660888,
+      "learning_rate": 9.958776429647882e-06,
+      "loss": 0.6835,
+      "step": 663
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5343938278654954,
+      "learning_rate": 9.958557749929507e-06,
+      "loss": 0.7832,
+      "step": 664
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2961561118476643,
+      "learning_rate": 9.958338494140226e-06,
+      "loss": 0.7488,
+      "step": 665
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0185829134658593,
+      "learning_rate": 9.958118662305512e-06,
+      "loss": 0.6639,
+      "step": 666
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6097834515400935,
+      "learning_rate": 9.957898254450904e-06,
+      "loss": 0.8354,
+      "step": 667
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.681329545781412,
+      "learning_rate": 9.957677270602009e-06,
+      "loss": 0.8243,
+      "step": 668
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.0534672454472056,
+      "learning_rate": 9.957455710784499e-06,
+      "loss": 0.7536,
+      "step": 669
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.306313918952685,
+      "learning_rate": 9.957233575024114e-06,
+      "loss": 0.831,
+      "step": 670
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.188405047316281,
+      "learning_rate": 9.957010863346665e-06,
+      "loss": 0.7803,
+      "step": 671
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.0129653737744073,
+      "learning_rate": 9.956787575778022e-06,
+      "loss": 0.721,
+      "step": 672
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3823769975743683,
+      "learning_rate": 9.956563712344127e-06,
+      "loss": 0.7449,
+      "step": 673
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5712155219456476,
+      "learning_rate": 9.956339273070988e-06,
+      "loss": 0.7693,
+      "step": 674
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.917447740270616,
+      "learning_rate": 9.95611425798468e-06,
+      "loss": 0.7564,
+      "step": 675
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.441012766615215,
+      "learning_rate": 9.955888667111341e-06,
+      "loss": 0.7311,
+      "step": 676
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3439122547341844,
+      "learning_rate": 9.955662500477185e-06,
+      "loss": 0.6111,
+      "step": 677
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3472763771803153,
+      "learning_rate": 9.955435758108488e-06,
+      "loss": 0.7722,
+      "step": 678
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.0118308796961055,
+      "learning_rate": 9.955208440031586e-06,
+      "loss": 0.7479,
+      "step": 679
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5164721040624083,
+      "learning_rate": 9.954980546272892e-06,
+      "loss": 0.7528,
+      "step": 680
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.349312862665128,
+      "learning_rate": 9.95475207685888e-06,
+      "loss": 0.7895,
+      "step": 681
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.538231437183247,
+      "learning_rate": 9.954523031816096e-06,
+      "loss": 0.7881,
+      "step": 682
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2748356407545756,
+      "learning_rate": 9.95429341117115e-06,
+      "loss": 0.7373,
+      "step": 683
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.690864131773263,
+      "learning_rate": 9.954063214950715e-06,
+      "loss": 0.7729,
+      "step": 684
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.441689564432529,
+      "learning_rate": 9.953832443181536e-06,
+      "loss": 0.6846,
+      "step": 685
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7063998770038253,
+      "learning_rate": 9.953601095890425e-06,
+      "loss": 0.7764,
+      "step": 686
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3897375113199755,
+      "learning_rate": 9.953369173104256e-06,
+      "loss": 0.7468,
+      "step": 687
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.623372216112194,
+      "learning_rate": 9.953136674849978e-06,
+      "loss": 0.7766,
+      "step": 688
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.33678410607246,
+      "learning_rate": 9.952903601154598e-06,
+      "loss": 0.7899,
+      "step": 689
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.7100874309999385,
+      "learning_rate": 9.952669952045196e-06,
+      "loss": 0.6976,
+      "step": 690
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4385704663154444,
+      "learning_rate": 9.952435727548915e-06,
+      "loss": 0.7834,
+      "step": 691
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.466989152583984,
+      "learning_rate": 9.952200927692965e-06,
+      "loss": 0.7762,
+      "step": 692
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4093135917267245,
+      "learning_rate": 9.95196555250463e-06,
+      "loss": 0.6769,
+      "step": 693
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3338351134830537,
+      "learning_rate": 9.95172960201125e-06,
+      "loss": 0.7501,
+      "step": 694
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3101279059282334,
+      "learning_rate": 9.95149307624024e-06,
+      "loss": 0.7034,
+      "step": 695
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2681781457320707,
+      "learning_rate": 9.951255975219076e-06,
+      "loss": 0.8163,
+      "step": 696
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.79211360891463,
+      "learning_rate": 9.951018298975306e-06,
+      "loss": 0.7878,
+      "step": 697
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9232352000611326,
+      "learning_rate": 9.950780047536543e-06,
+      "loss": 0.74,
+      "step": 698
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.750774856582151,
+      "learning_rate": 9.950541220930463e-06,
+      "loss": 0.7653,
+      "step": 699
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.046202452399702,
+      "learning_rate": 9.950301819184816e-06,
+      "loss": 0.6924,
+      "step": 700
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5862040137998457,
+      "learning_rate": 9.950061842327415e-06,
+      "loss": 0.7198,
+      "step": 701
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.9170844730186487,
+      "learning_rate": 9.949821290386137e-06,
+      "loss": 0.7769,
+      "step": 702
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.638513502534882,
+      "learning_rate": 9.94958016338893e-06,
+      "loss": 0.7465,
+      "step": 703
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.154543573950718,
+      "learning_rate": 9.949338461363807e-06,
+      "loss": 0.7234,
+      "step": 704
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4170777296764943,
+      "learning_rate": 9.949096184338849e-06,
+      "loss": 0.7579,
+      "step": 705
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.1649853433830772,
+      "learning_rate": 9.948853332342202e-06,
+      "loss": 0.6416,
+      "step": 706
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2732580001069485,
+      "learning_rate": 9.948609905402082e-06,
+      "loss": 0.7448,
+      "step": 707
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.0109624448260335,
+      "learning_rate": 9.948365903546766e-06,
+      "loss": 0.8066,
+      "step": 708
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5719103836155117,
+      "learning_rate": 9.948121326804604e-06,
+      "loss": 0.6297,
+      "step": 709
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6740846319075147,
+      "learning_rate": 9.947876175204013e-06,
+      "loss": 0.6353,
+      "step": 710
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.8270316407346163,
+      "learning_rate": 9.947630448773468e-06,
+      "loss": 0.7751,
+      "step": 711
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.8283132421827895,
+      "learning_rate": 9.94738414754152e-06,
+      "loss": 0.8027,
+      "step": 712
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.332076932637489,
+      "learning_rate": 9.947137271536784e-06,
+      "loss": 0.6781,
+      "step": 713
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3484004490599073,
+      "learning_rate": 9.94688982078794e-06,
+      "loss": 0.6904,
+      "step": 714
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1432128787457025,
+      "learning_rate": 9.946641795323737e-06,
+      "loss": 0.7361,
+      "step": 715
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6418533463624745,
+      "learning_rate": 9.946393195172987e-06,
+      "loss": 0.63,
+      "step": 716
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2838183539614234,
+      "learning_rate": 9.946144020364576e-06,
+      "loss": 0.7464,
+      "step": 717
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.955408112768042,
+      "learning_rate": 9.945894270927452e-06,
+      "loss": 0.7333,
+      "step": 718
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.565202439751648,
+      "learning_rate": 9.945643946890628e-06,
+      "loss": 0.6788,
+      "step": 719
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.987883523633849,
+      "learning_rate": 9.945393048283186e-06,
+      "loss": 0.7369,
+      "step": 720
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.305645743958634,
+      "learning_rate": 9.945141575134275e-06,
+      "loss": 0.8795,
+      "step": 721
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.342801724950221,
+      "learning_rate": 9.944889527473112e-06,
+      "loss": 0.7749,
+      "step": 722
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6867592587630953,
+      "learning_rate": 9.944636905328977e-06,
+      "loss": 0.7122,
+      "step": 723
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7963278085386087,
+      "learning_rate": 9.94438370873122e-06,
+      "loss": 0.7286,
+      "step": 724
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0500060514327645,
+      "learning_rate": 9.944129937709255e-06,
+      "loss": 0.8238,
+      "step": 725
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.518747853267898,
+      "learning_rate": 9.943875592292569e-06,
+      "loss": 0.7468,
+      "step": 726
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3307749408071023,
+      "learning_rate": 9.943620672510706e-06,
+      "loss": 0.784,
+      "step": 727
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2927298280547608,
+      "learning_rate": 9.943365178393283e-06,
+      "loss": 0.6643,
+      "step": 728
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.413030343240209,
+      "learning_rate": 9.943109109969985e-06,
+      "loss": 0.7718,
+      "step": 729
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.985911781478991,
+      "learning_rate": 9.94285246727056e-06,
+      "loss": 0.7437,
+      "step": 730
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8515953816737416,
+      "learning_rate": 9.942595250324823e-06,
+      "loss": 0.7,
+      "step": 731
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.567107917615768,
+      "learning_rate": 9.942337459162657e-06,
+      "loss": 0.6874,
+      "step": 732
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.389536459691384,
+      "learning_rate": 9.942079093814012e-06,
+      "loss": 0.7786,
+      "step": 733
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.456329041193247,
+      "learning_rate": 9.941820154308905e-06,
+      "loss": 0.6675,
+      "step": 734
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2902527313975454,
+      "learning_rate": 9.941560640677417e-06,
+      "loss": 0.7431,
+      "step": 735
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9299279456459286,
+      "learning_rate": 9.941300552949697e-06,
+      "loss": 0.7444,
+      "step": 736
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.894177677033512,
+      "learning_rate": 9.941039891155964e-06,
+      "loss": 0.6389,
+      "step": 737
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1178139348238716,
+      "learning_rate": 9.940778655326499e-06,
+      "loss": 0.7812,
+      "step": 738
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6341684151897957,
+      "learning_rate": 9.940516845491653e-06,
+      "loss": 0.7911,
+      "step": 739
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3572249944985764,
+      "learning_rate": 9.940254461681841e-06,
+      "loss": 0.7095,
+      "step": 740
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.392416707308911,
+      "learning_rate": 9.939991503927548e-06,
+      "loss": 0.7532,
+      "step": 741
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.892105469013813,
+      "learning_rate": 9.939727972259321e-06,
+      "loss": 0.8392,
+      "step": 742
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5792946557866454,
+      "learning_rate": 9.939463866707777e-06,
+      "loss": 0.7718,
+      "step": 743
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.382428104409083,
+      "learning_rate": 9.939199187303598e-06,
+      "loss": 0.6834,
+      "step": 744
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.963009035348591,
+      "learning_rate": 9.938933934077539e-06,
+      "loss": 0.749,
+      "step": 745
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3169431674281125,
+      "learning_rate": 9.93866810706041e-06,
+      "loss": 0.7598,
+      "step": 746
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.269068621083456,
+      "learning_rate": 9.938401706283096e-06,
+      "loss": 0.6919,
+      "step": 747
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.520110943561689,
+      "learning_rate": 9.93813473177655e-06,
+      "loss": 0.7574,
+      "step": 748
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6705782110520553,
+      "learning_rate": 9.937867183571784e-06,
+      "loss": 0.6852,
+      "step": 749
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.394466270287015,
+      "learning_rate": 9.93759906169988e-06,
+      "loss": 0.8214,
+      "step": 750
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4103257753815956,
+      "learning_rate": 9.937330366191994e-06,
+      "loss": 0.7784,
+      "step": 751
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2399480859042666,
+      "learning_rate": 9.937061097079337e-06,
+      "loss": 0.6601,
+      "step": 752
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0481316936580254,
+      "learning_rate": 9.936791254393193e-06,
+      "loss": 0.7356,
+      "step": 753
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.437583015333378,
+      "learning_rate": 9.936520838164912e-06,
+      "loss": 0.7158,
+      "step": 754
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.707441716431831,
+      "learning_rate": 9.93624984842591e-06,
+      "loss": 0.7215,
+      "step": 755
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.532030304474135,
+      "learning_rate": 9.93597828520767e-06,
+      "loss": 0.753,
+      "step": 756
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7461971545571786,
+      "learning_rate": 9.935706148541742e-06,
+      "loss": 0.7097,
+      "step": 757
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3627790545916216,
+      "learning_rate": 9.93543343845974e-06,
+      "loss": 0.8169,
+      "step": 758
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6693014549496814,
+      "learning_rate": 9.93516015499335e-06,
+      "loss": 0.8296,
+      "step": 759
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.7105561050330222,
+      "learning_rate": 9.934886298174317e-06,
+      "loss": 0.7099,
+      "step": 760
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.720683329507592,
+      "learning_rate": 9.93461186803446e-06,
+      "loss": 0.7632,
+      "step": 761
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.588291193800891,
+      "learning_rate": 9.934336864605663e-06,
+      "loss": 0.7789,
+      "step": 762
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5585071173949863,
+      "learning_rate": 9.934061287919869e-06,
+      "loss": 0.6833,
+      "step": 763
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2895773779403195,
+      "learning_rate": 9.9337851380091e-06,
+      "loss": 0.7367,
+      "step": 764
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0219975702999697,
+      "learning_rate": 9.933508414905434e-06,
+      "loss": 0.7717,
+      "step": 765
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6270513907040094,
+      "learning_rate": 9.933231118641025e-06,
+      "loss": 0.6998,
+      "step": 766
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.620673953552405,
+      "learning_rate": 9.932953249248082e-06,
+      "loss": 0.833,
+      "step": 767
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5153078631067154,
+      "learning_rate": 9.93267480675889e-06,
+      "loss": 0.7723,
+      "step": 768
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.951354250369123,
+      "learning_rate": 9.9323957912058e-06,
+      "loss": 0.7502,
+      "step": 769
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4394149168178134,
+      "learning_rate": 9.932116202621224e-06,
+      "loss": 0.6457,
+      "step": 770
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1095938889246026,
+      "learning_rate": 9.931836041037644e-06,
+      "loss": 0.7061,
+      "step": 771
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6819242824234957,
+      "learning_rate": 9.931555306487612e-06,
+      "loss": 0.7691,
+      "step": 772
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2043376051832424,
+      "learning_rate": 9.931273999003738e-06,
+      "loss": 0.6946,
+      "step": 773
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6902924884956425,
+      "learning_rate": 9.930992118618706e-06,
+      "loss": 0.6695,
+      "step": 774
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8989924651843326,
+      "learning_rate": 9.930709665365264e-06,
+      "loss": 0.7694,
+      "step": 775
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.224303330241575,
+      "learning_rate": 9.930426639276225e-06,
+      "loss": 0.7487,
+      "step": 776
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6983836303838826,
+      "learning_rate": 9.930143040384472e-06,
+      "loss": 0.7523,
+      "step": 777
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3780230591194065,
+      "learning_rate": 9.929858868722954e-06,
+      "loss": 0.7692,
+      "step": 778
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4634887523343942,
+      "learning_rate": 9.929574124324682e-06,
+      "loss": 0.8036,
+      "step": 779
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2694290012543052,
+      "learning_rate": 9.929288807222738e-06,
+      "loss": 0.7424,
+      "step": 780
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6592362586407394,
+      "learning_rate": 9.92900291745027e-06,
+      "loss": 0.6375,
+      "step": 781
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.495897539805403,
+      "learning_rate": 9.92871645504049e-06,
+      "loss": 0.7854,
+      "step": 782
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0961963685516727,
+      "learning_rate": 9.928429420026682e-06,
+      "loss": 0.7634,
+      "step": 783
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8509364984526253,
+      "learning_rate": 9.92814181244219e-06,
+      "loss": 0.6458,
+      "step": 784
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8241576523218956,
+      "learning_rate": 9.927853632320427e-06,
+      "loss": 0.7804,
+      "step": 785
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0001225664364055,
+      "learning_rate": 9.927564879694874e-06,
+      "loss": 0.7229,
+      "step": 786
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4475528816764944,
+      "learning_rate": 9.927275554599078e-06,
+      "loss": 0.8362,
+      "step": 787
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2308232356508526,
+      "learning_rate": 9.926985657066653e-06,
+      "loss": 0.7114,
+      "step": 788
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.341682432680137,
+      "learning_rate": 9.926695187131275e-06,
+      "loss": 0.7783,
+      "step": 789
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6052199283711874,
+      "learning_rate": 9.92640414482669e-06,
+      "loss": 0.8333,
+      "step": 790
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8370875665068858,
+      "learning_rate": 9.926112530186715e-06,
+      "loss": 0.7107,
+      "step": 791
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5275739897207865,
+      "learning_rate": 9.925820343245225e-06,
+      "loss": 0.7965,
+      "step": 792
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.241333301690011,
+      "learning_rate": 9.925527584036167e-06,
+      "loss": 0.8081,
+      "step": 793
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1980959634447164,
+      "learning_rate": 9.925234252593554e-06,
+      "loss": 0.7231,
+      "step": 794
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5031300722496965,
+      "learning_rate": 9.92494034895146e-06,
+      "loss": 0.6994,
+      "step": 795
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1430334991183604,
+      "learning_rate": 9.924645873144035e-06,
+      "loss": 0.6908,
+      "step": 796
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.350620970434163,
+      "learning_rate": 9.924350825205487e-06,
+      "loss": 0.8007,
+      "step": 797
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.579066826686899,
+      "learning_rate": 9.924055205170095e-06,
+      "loss": 0.8572,
+      "step": 798
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.9358520387744473,
+      "learning_rate": 9.923759013072205e-06,
+      "loss": 0.7289,
+      "step": 799
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4595413230957774,
+      "learning_rate": 9.923462248946224e-06,
+      "loss": 0.6404,
+      "step": 800
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.25406590521839,
+      "learning_rate": 9.923164912826631e-06,
+      "loss": 0.7915,
+      "step": 801
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5320545260007727,
+      "learning_rate": 9.922867004747971e-06,
+      "loss": 0.7355,
+      "step": 802
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0795033707901633,
+      "learning_rate": 9.922568524744854e-06,
+      "loss": 0.6239,
+      "step": 803
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.1305263819984,
+      "learning_rate": 9.922269472851953e-06,
+      "loss": 0.7992,
+      "step": 804
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4229395578571924,
+      "learning_rate": 9.921969849104015e-06,
+      "loss": 0.686,
+      "step": 805
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.925961640710328,
+      "learning_rate": 9.921669653535848e-06,
+      "loss": 0.7174,
+      "step": 806
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.77414781814395,
+      "learning_rate": 9.921368886182328e-06,
+      "loss": 0.7249,
+      "step": 807
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.732166312681211,
+      "learning_rate": 9.921067547078396e-06,
+      "loss": 0.7682,
+      "step": 808
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6581746298192357,
+      "learning_rate": 9.920765636259062e-06,
+      "loss": 0.7427,
+      "step": 809
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.420618373750652,
+      "learning_rate": 9.9204631537594e-06,
+      "loss": 0.8256,
+      "step": 810
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.452433300784335,
+      "learning_rate": 9.920160099614553e-06,
+      "loss": 0.7742,
+      "step": 811
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.347854862834246,
+      "learning_rate": 9.91985647385973e-06,
+      "loss": 0.6981,
+      "step": 812
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.335973187712152,
+      "learning_rate": 9.919552276530202e-06,
+      "loss": 0.6822,
+      "step": 813
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0899513351336645,
+      "learning_rate": 9.919247507661313e-06,
+      "loss": 0.6554,
+      "step": 814
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4579977944848657,
+      "learning_rate": 9.918942167288467e-06,
+      "loss": 0.7347,
+      "step": 815
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2321378141123316,
+      "learning_rate": 9.918636255447141e-06,
+      "loss": 0.7083,
+      "step": 816
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.337678644880633,
+      "learning_rate": 9.918329772172872e-06,
+      "loss": 0.7353,
+      "step": 817
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.484233611654691,
+      "learning_rate": 9.918022717501268e-06,
+      "loss": 0.7626,
+      "step": 818
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6233513206248698,
+      "learning_rate": 9.917715091467999e-06,
+      "loss": 0.7422,
+      "step": 819
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2460279378058137,
+      "learning_rate": 9.91740689410881e-06,
+      "loss": 0.8034,
+      "step": 820
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.014666827028448,
+      "learning_rate": 9.917098125459501e-06,
+      "loss": 0.7201,
+      "step": 821
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8852937872527904,
+      "learning_rate": 9.916788785555945e-06,
+      "loss": 0.702,
+      "step": 822
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.468109581522842,
+      "learning_rate": 9.91647887443408e-06,
+      "loss": 0.7292,
+      "step": 823
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0137798678101095,
+      "learning_rate": 9.916168392129914e-06,
+      "loss": 0.7081,
+      "step": 824
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1356848172886997,
+      "learning_rate": 9.915857338679515e-06,
+      "loss": 0.7477,
+      "step": 825
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.146941252431847,
+      "learning_rate": 9.91554571411902e-06,
+      "loss": 0.7231,
+      "step": 826
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.967160075670586,
+      "learning_rate": 9.915233518484633e-06,
+      "loss": 0.7984,
+      "step": 827
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0751156185412385,
+      "learning_rate": 9.914920751812626e-06,
+      "loss": 0.8102,
+      "step": 828
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3472721595318586,
+      "learning_rate": 9.914607414139332e-06,
+      "loss": 0.7313,
+      "step": 829
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.118986957021893,
+      "learning_rate": 9.914293505501155e-06,
+      "loss": 0.8668,
+      "step": 830
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.394587519870367,
+      "learning_rate": 9.913979025934566e-06,
+      "loss": 0.7498,
+      "step": 831
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0206794969562405,
+      "learning_rate": 9.913663975476099e-06,
+      "loss": 0.7633,
+      "step": 832
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.096598722449407,
+      "learning_rate": 9.913348354162353e-06,
+      "loss": 0.7422,
+      "step": 833
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4996308686533895,
+      "learning_rate": 9.913032162029999e-06,
+      "loss": 0.7345,
+      "step": 834
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.692833005695984,
+      "learning_rate": 9.91271539911577e-06,
+      "loss": 0.783,
+      "step": 835
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1228836691300037,
+      "learning_rate": 9.91239806545647e-06,
+      "loss": 0.7425,
+      "step": 836
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3986505144559427,
+      "learning_rate": 9.91208016108896e-06,
+      "loss": 0.7267,
+      "step": 837
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7101167395338273,
+      "learning_rate": 9.911761686050177e-06,
+      "loss": 0.7581,
+      "step": 838
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5806365307450667,
+      "learning_rate": 9.91144264037712e-06,
+      "loss": 0.7847,
+      "step": 839
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5964088279350312,
+      "learning_rate": 9.911123024106854e-06,
+      "loss": 0.6742,
+      "step": 840
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0458536491050143,
+      "learning_rate": 9.910802837276514e-06,
+      "loss": 0.6898,
+      "step": 841
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9738131081216486,
+      "learning_rate": 9.910482079923293e-06,
+      "loss": 0.7097,
+      "step": 842
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.033940638900864,
+      "learning_rate": 9.910160752084461e-06,
+      "loss": 0.6891,
+      "step": 843
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.493418453706031,
+      "learning_rate": 9.909838853797347e-06,
+      "loss": 0.8386,
+      "step": 844
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3609786228834397,
+      "learning_rate": 9.909516385099346e-06,
+      "loss": 0.6447,
+      "step": 845
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8826827313822074,
+      "learning_rate": 9.909193346027923e-06,
+      "loss": 0.7634,
+      "step": 846
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.192146232328817,
+      "learning_rate": 9.90886973662061e-06,
+      "loss": 0.6501,
+      "step": 847
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.924566828839797,
+      "learning_rate": 9.908545556915e-06,
+      "loss": 0.7382,
+      "step": 848
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.124722725213922,
+      "learning_rate": 9.908220806948755e-06,
+      "loss": 0.7342,
+      "step": 849
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.149040160809967,
+      "learning_rate": 9.90789548675961e-06,
+      "loss": 0.769,
+      "step": 850
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.016612347468723,
+      "learning_rate": 9.90756959638535e-06,
+      "loss": 0.7533,
+      "step": 851
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.21890626690869,
+      "learning_rate": 9.90724313586384e-06,
+      "loss": 0.7183,
+      "step": 852
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1739377588424182,
+      "learning_rate": 9.90691610523301e-06,
+      "loss": 0.725,
+      "step": 853
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4308506352472565,
+      "learning_rate": 9.906588504530852e-06,
+      "loss": 0.7721,
+      "step": 854
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.039428986286277,
+      "learning_rate": 9.906260333795423e-06,
+      "loss": 0.7862,
+      "step": 855
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.072889277401122,
+      "learning_rate": 9.905931593064852e-06,
+      "loss": 0.7057,
+      "step": 856
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5603746952912583,
+      "learning_rate": 9.905602282377331e-06,
+      "loss": 0.7845,
+      "step": 857
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2094754050609255,
+      "learning_rate": 9.905272401771115e-06,
+      "loss": 0.6726,
+      "step": 858
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.010211257078767,
+      "learning_rate": 9.904941951284535e-06,
+      "loss": 0.82,
+      "step": 859
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4294590301163907,
+      "learning_rate": 9.904610930955975e-06,
+      "loss": 0.7225,
+      "step": 860
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4545417076156157,
+      "learning_rate": 9.904279340823895e-06,
+      "loss": 0.7379,
+      "step": 861
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0489546331766704,
+      "learning_rate": 9.903947180926819e-06,
+      "loss": 0.7939,
+      "step": 862
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.328322416011042,
+      "learning_rate": 9.903614451303335e-06,
+      "loss": 0.7423,
+      "step": 863
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.673446796533317,
+      "learning_rate": 9.903281151992097e-06,
+      "loss": 0.684,
+      "step": 864
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6240021544906904,
+      "learning_rate": 9.902947283031833e-06,
+      "loss": 0.6573,
+      "step": 865
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4194535688154097,
+      "learning_rate": 9.902612844461322e-06,
+      "loss": 0.7671,
+      "step": 866
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.600458794788045,
+      "learning_rate": 9.902277836319424e-06,
+      "loss": 0.686,
+      "step": 867
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7708332752714226,
+      "learning_rate": 9.90194225864506e-06,
+      "loss": 0.7713,
+      "step": 868
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.345954440796902,
+      "learning_rate": 9.901606111477213e-06,
+      "loss": 0.7127,
+      "step": 869
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4380277637214958,
+      "learning_rate": 9.901269394854938e-06,
+      "loss": 0.7432,
+      "step": 870
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.535510929005561,
+      "learning_rate": 9.900932108817352e-06,
+      "loss": 0.735,
+      "step": 871
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0039289042478337,
+      "learning_rate": 9.900594253403642e-06,
+      "loss": 0.76,
+      "step": 872
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.350300158880335,
+      "learning_rate": 9.900255828653057e-06,
+      "loss": 0.6084,
+      "step": 873
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4958117500634938,
+      "learning_rate": 9.899916834604914e-06,
+      "loss": 0.7951,
+      "step": 874
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4483497475118843,
+      "learning_rate": 9.899577271298596e-06,
+      "loss": 0.7217,
+      "step": 875
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9481072329531595,
+      "learning_rate": 9.89923713877356e-06,
+      "loss": 0.6245,
+      "step": 876
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.765051932549656,
+      "learning_rate": 9.89889643706931e-06,
+      "loss": 0.6723,
+      "step": 877
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.739221117936455,
+      "learning_rate": 9.898555166225434e-06,
+      "loss": 0.7703,
+      "step": 878
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1418498868809768,
+      "learning_rate": 9.89821332628158e-06,
+      "loss": 0.7298,
+      "step": 879
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2036351653598367,
+      "learning_rate": 9.897870917277461e-06,
+      "loss": 0.8093,
+      "step": 880
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4323072705596505,
+      "learning_rate": 9.897527939252858e-06,
+      "loss": 0.7399,
+      "step": 881
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8657728288025477,
+      "learning_rate": 9.897184392247614e-06,
+      "loss": 0.7356,
+      "step": 882
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5414139234529642,
+      "learning_rate": 9.896840276301645e-06,
+      "loss": 0.6747,
+      "step": 883
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2994147096204327,
+      "learning_rate": 9.896495591454929e-06,
+      "loss": 0.7335,
+      "step": 884
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9851136351096752,
+      "learning_rate": 9.896150337747508e-06,
+      "loss": 0.7839,
+      "step": 885
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.475304204640165,
+      "learning_rate": 9.895804515219495e-06,
+      "loss": 0.7442,
+      "step": 886
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0051133893066178,
+      "learning_rate": 9.895458123911066e-06,
+      "loss": 0.819,
+      "step": 887
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.01882317139601,
+      "learning_rate": 9.895111163862464e-06,
+      "loss": 0.743,
+      "step": 888
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4869064278045154,
+      "learning_rate": 9.894763635113995e-06,
+      "loss": 0.673,
+      "step": 889
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0423193152498316,
+      "learning_rate": 9.894415537706036e-06,
+      "loss": 0.7938,
+      "step": 890
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3877929769561415,
+      "learning_rate": 9.89406687167903e-06,
+      "loss": 0.8399,
+      "step": 891
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3118433378041283,
+      "learning_rate": 9.893717637073483e-06,
+      "loss": 0.7365,
+      "step": 892
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1231159883183817,
+      "learning_rate": 9.893367833929965e-06,
+      "loss": 0.6975,
+      "step": 893
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.946732175793273,
+      "learning_rate": 9.893017462289119e-06,
+      "loss": 0.7355,
+      "step": 894
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.3717939300188275,
+      "learning_rate": 9.892666522191648e-06,
+      "loss": 0.6766,
+      "step": 895
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0279846422346446,
+      "learning_rate": 9.892315013678323e-06,
+      "loss": 0.7597,
+      "step": 896
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.55744056624668,
+      "learning_rate": 9.891962936789983e-06,
+      "loss": 0.7977,
+      "step": 897
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.569713950624268,
+      "learning_rate": 9.891610291567529e-06,
+      "loss": 0.7722,
+      "step": 898
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8251737732812003,
+      "learning_rate": 9.891257078051932e-06,
+      "loss": 0.8538,
+      "step": 899
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.940064605870108,
+      "learning_rate": 9.890903296284228e-06,
+      "loss": 0.6861,
+      "step": 900
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3962938305889727,
+      "learning_rate": 9.890548946305516e-06,
+      "loss": 0.7367,
+      "step": 901
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6270790503951877,
+      "learning_rate": 9.890194028156965e-06,
+      "loss": 0.6927,
+      "step": 902
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2464932261338486,
+      "learning_rate": 9.889838541879808e-06,
+      "loss": 0.6671,
+      "step": 903
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2068048828956184,
+      "learning_rate": 9.889482487515344e-06,
+      "loss": 0.6574,
+      "step": 904
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.131286705587521,
+      "learning_rate": 9.889125865104939e-06,
+      "loss": 0.6844,
+      "step": 905
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4854115277088384,
+      "learning_rate": 9.888768674690023e-06,
+      "loss": 0.7562,
+      "step": 906
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.430105603029439,
+      "learning_rate": 9.888410916312096e-06,
+      "loss": 0.8826,
+      "step": 907
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3307764886099447,
+      "learning_rate": 9.888052590012719e-06,
+      "loss": 0.6994,
+      "step": 908
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.399232411769506,
+      "learning_rate": 9.887693695833522e-06,
+      "loss": 0.7844,
+      "step": 909
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4841741299380127,
+      "learning_rate": 9.887334233816199e-06,
+      "loss": 0.7191,
+      "step": 910
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.231644133692005,
+      "learning_rate": 9.886974204002514e-06,
+      "loss": 0.7236,
+      "step": 911
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.589193939890633,
+      "learning_rate": 9.886613606434294e-06,
+      "loss": 0.8006,
+      "step": 912
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.8815065090619294,
+      "learning_rate": 9.886252441153428e-06,
+      "loss": 0.762,
+      "step": 913
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.458751138818673,
+      "learning_rate": 9.885890708201881e-06,
+      "loss": 0.748,
+      "step": 914
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.729352069908257,
+      "learning_rate": 9.885528407621674e-06,
+      "loss": 0.7319,
+      "step": 915
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7456542817248035,
+      "learning_rate": 9.885165539454898e-06,
+      "loss": 0.7691,
+      "step": 916
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.821954609815299,
+      "learning_rate": 9.884802103743712e-06,
+      "loss": 0.724,
+      "step": 917
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.507495458671125,
+      "learning_rate": 9.88443810053034e-06,
+      "loss": 0.7557,
+      "step": 918
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.813557690867528,
+      "learning_rate": 9.884073529857066e-06,
+      "loss": 0.7378,
+      "step": 919
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2109392193308994,
+      "learning_rate": 9.883708391766248e-06,
+      "loss": 0.655,
+      "step": 920
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7797266133676817,
+      "learning_rate": 9.88334268630031e-06,
+      "loss": 0.7568,
+      "step": 921
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.291991455347209,
+      "learning_rate": 9.882976413501733e-06,
+      "loss": 0.6939,
+      "step": 922
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2553429818506694,
+      "learning_rate": 9.88260957341307e-06,
+      "loss": 0.8572,
+      "step": 923
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.944662889076845,
+      "learning_rate": 9.882242166076942e-06,
+      "loss": 0.7171,
+      "step": 924
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6005068384270817,
+      "learning_rate": 9.881874191536032e-06,
+      "loss": 0.7304,
+      "step": 925
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.466317488664055,
+      "learning_rate": 9.881505649833091e-06,
+      "loss": 0.7116,
+      "step": 926
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4452143875236243,
+      "learning_rate": 9.881136541010934e-06,
+      "loss": 0.7629,
+      "step": 927
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5102243570459595,
+      "learning_rate": 9.880766865112444e-06,
+      "loss": 0.7286,
+      "step": 928
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7791143392672875,
+      "learning_rate": 9.880396622180567e-06,
+      "loss": 0.7923,
+      "step": 929
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.876441285426497,
+      "learning_rate": 9.880025812258322e-06,
+      "loss": 0.7381,
+      "step": 930
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2830140025930246,
+      "learning_rate": 9.879654435388781e-06,
+      "loss": 0.7657,
+      "step": 931
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1392167947242817,
+      "learning_rate": 9.879282491615096e-06,
+      "loss": 0.7584,
+      "step": 932
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5574554782173244,
+      "learning_rate": 9.878909980980475e-06,
+      "loss": 0.6736,
+      "step": 933
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.482015182428965,
+      "learning_rate": 9.878536903528195e-06,
+      "loss": 0.7855,
+      "step": 934
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.20969188023806,
+      "learning_rate": 9.8781632593016e-06,
+      "loss": 0.7142,
+      "step": 935
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6096755265839593,
+      "learning_rate": 9.8777890483441e-06,
+      "loss": 0.8059,
+      "step": 936
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6014399338081122,
+      "learning_rate": 9.877414270699168e-06,
+      "loss": 0.7378,
+      "step": 937
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1756243191451383,
+      "learning_rate": 9.877038926410346e-06,
+      "loss": 0.7268,
+      "step": 938
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.196316926201555,
+      "learning_rate": 9.876663015521237e-06,
+      "loss": 0.8158,
+      "step": 939
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.383983584002673,
+      "learning_rate": 9.876286538075519e-06,
+      "loss": 0.7215,
+      "step": 940
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7085992373351204,
+      "learning_rate": 9.875909494116925e-06,
+      "loss": 0.6575,
+      "step": 941
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5130773928374737,
+      "learning_rate": 9.875531883689262e-06,
+      "loss": 0.6911,
+      "step": 942
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.318178727992876,
+      "learning_rate": 9.875153706836397e-06,
+      "loss": 0.788,
+      "step": 943
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.286494331288682,
+      "learning_rate": 9.874774963602268e-06,
+      "loss": 0.7114,
+      "step": 944
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1591477947522186,
+      "learning_rate": 9.874395654030876e-06,
+      "loss": 0.7871,
+      "step": 945
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1547218552212177,
+      "learning_rate": 9.874015778166285e-06,
+      "loss": 0.8348,
+      "step": 946
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.762373245266379,
+      "learning_rate": 9.873635336052633e-06,
+      "loss": 0.7688,
+      "step": 947
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.309708677231106,
+      "learning_rate": 9.873254327734115e-06,
+      "loss": 0.747,
+      "step": 948
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.053888930700921,
+      "learning_rate": 9.872872753254996e-06,
+      "loss": 0.7223,
+      "step": 949
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5008788542760363,
+      "learning_rate": 9.872490612659607e-06,
+      "loss": 0.6766,
+      "step": 950
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3863679851555673,
+      "learning_rate": 9.872107905992343e-06,
+      "loss": 0.6972,
+      "step": 951
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.921615156601913,
+      "learning_rate": 9.871724633297666e-06,
+      "loss": 0.6999,
+      "step": 952
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2644813929698615,
+      "learning_rate": 9.871340794620103e-06,
+      "loss": 0.761,
+      "step": 953
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1191106722345814,
+      "learning_rate": 9.87095639000425e-06,
+      "loss": 0.7376,
+      "step": 954
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.026837083096435,
+      "learning_rate": 9.870571419494764e-06,
+      "loss": 0.7222,
+      "step": 955
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.336308324288583,
+      "learning_rate": 9.87018588313637e-06,
+      "loss": 0.7229,
+      "step": 956
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.456262394309276,
+      "learning_rate": 9.869799780973856e-06,
+      "loss": 0.7182,
+      "step": 957
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5276315796370676,
+      "learning_rate": 9.869413113052084e-06,
+      "loss": 0.6835,
+      "step": 958
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1051858480552497,
+      "learning_rate": 9.86902587941597e-06,
+      "loss": 0.7101,
+      "step": 959
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3731070604108844,
+      "learning_rate": 9.868638080110507e-06,
+      "loss": 0.7221,
+      "step": 960
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0867518495145223,
+      "learning_rate": 9.868249715180741e-06,
+      "loss": 0.7046,
+      "step": 961
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0764983780128166,
+      "learning_rate": 9.8678607846718e-06,
+      "loss": 0.7498,
+      "step": 962
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1440243110852037,
+      "learning_rate": 9.867471288628863e-06,
+      "loss": 0.6442,
+      "step": 963
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.559996501289641,
+      "learning_rate": 9.867081227097182e-06,
+      "loss": 0.7556,
+      "step": 964
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.546829215817774,
+      "learning_rate": 9.866690600122075e-06,
+      "loss": 0.7098,
+      "step": 965
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.692415588752108,
+      "learning_rate": 9.866299407748921e-06,
+      "loss": 0.694,
+      "step": 966
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.336011704071477,
+      "learning_rate": 9.865907650023167e-06,
+      "loss": 0.7751,
+      "step": 967
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3030234604337436,
+      "learning_rate": 9.865515326990332e-06,
+      "loss": 0.68,
+      "step": 968
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3788631634376527,
+      "learning_rate": 9.865122438695988e-06,
+      "loss": 0.7517,
+      "step": 969
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.594025015808721,
+      "learning_rate": 9.864728985185783e-06,
+      "loss": 0.7859,
+      "step": 970
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2839322712235326,
+      "learning_rate": 9.86433496650543e-06,
+      "loss": 0.7123,
+      "step": 971
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7189304806754397,
+      "learning_rate": 9.863940382700699e-06,
+      "loss": 0.7495,
+      "step": 972
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5542757749406033,
+      "learning_rate": 9.863545233817436e-06,
+      "loss": 0.7249,
+      "step": 973
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2818240056053565,
+      "learning_rate": 9.863149519901545e-06,
+      "loss": 0.7263,
+      "step": 974
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.980498735075583,
+      "learning_rate": 9.862753240999001e-06,
+      "loss": 0.796,
+      "step": 975
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.17696941282096,
+      "learning_rate": 9.862356397155843e-06,
+      "loss": 0.7528,
+      "step": 976
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7279007297671747,
+      "learning_rate": 9.861958988418174e-06,
+      "loss": 0.7417,
+      "step": 977
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7775418932594853,
+      "learning_rate": 9.861561014832166e-06,
+      "loss": 0.6685,
+      "step": 978
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0608876144824255,
+      "learning_rate": 9.86116247644405e-06,
+      "loss": 0.796,
+      "step": 979
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.230200096432408,
+      "learning_rate": 9.860763373300133e-06,
+      "loss": 0.682,
+      "step": 980
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.583892124389053,
+      "learning_rate": 9.860363705446776e-06,
+      "loss": 0.7154,
+      "step": 981
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.420436801225999,
+      "learning_rate": 9.859963472930413e-06,
+      "loss": 0.6849,
+      "step": 982
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5498244013352545,
+      "learning_rate": 9.859562675797543e-06,
+      "loss": 0.8011,
+      "step": 983
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1659421053037153,
+      "learning_rate": 9.85916131409473e-06,
+      "loss": 0.7394,
+      "step": 984
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.338196115649606,
+      "learning_rate": 9.858759387868601e-06,
+      "loss": 0.7245,
+      "step": 985
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7321452634255357,
+      "learning_rate": 9.858356897165853e-06,
+      "loss": 0.7919,
+      "step": 986
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6253699652801417,
+      "learning_rate": 9.857953842033243e-06,
+      "loss": 0.733,
+      "step": 987
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.59635003190081,
+      "learning_rate": 9.857550222517598e-06,
+      "loss": 0.7092,
+      "step": 988
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.621851871749562,
+      "learning_rate": 9.857146038665812e-06,
+      "loss": 0.7042,
+      "step": 989
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7774812468209973,
+      "learning_rate": 9.856741290524839e-06,
+      "loss": 0.7218,
+      "step": 990
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9702400566443012,
+      "learning_rate": 9.856335978141703e-06,
+      "loss": 0.7605,
+      "step": 991
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.229378466858724,
+      "learning_rate": 9.85593010156349e-06,
+      "loss": 0.7731,
+      "step": 992
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.985236389182805,
+      "learning_rate": 9.855523660837355e-06,
+      "loss": 0.7433,
+      "step": 993
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.418368994051385,
+      "learning_rate": 9.855116656010518e-06,
+      "loss": 0.6799,
+      "step": 994
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6770469216630866,
+      "learning_rate": 9.854709087130261e-06,
+      "loss": 0.6967,
+      "step": 995
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2596414821764004,
+      "learning_rate": 9.854300954243937e-06,
+      "loss": 0.6321,
+      "step": 996
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1913642125753563,
+      "learning_rate": 9.853892257398961e-06,
+      "loss": 0.6239,
+      "step": 997
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.038065448125383,
+      "learning_rate": 9.853482996642812e-06,
+      "loss": 0.8232,
+      "step": 998
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.80199548153629,
+      "learning_rate": 9.85307317202304e-06,
+      "loss": 0.7232,
+      "step": 999
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3246781936353464,
+      "learning_rate": 9.852662783587255e-06,
+      "loss": 0.6723,
+      "step": 1000
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.285393284696264,
+      "learning_rate": 9.852251831383136e-06,
+      "loss": 0.7249,
+      "step": 1001
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.6840301004398026,
+      "learning_rate": 9.851840315458424e-06,
+      "loss": 0.8205,
+      "step": 1002
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.635419321078597,
+      "learning_rate": 9.85142823586093e-06,
+      "loss": 0.7197,
+      "step": 1003
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.832069075664214,
+      "learning_rate": 9.851015592638528e-06,
+      "loss": 0.7241,
+      "step": 1004
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.533811572972949,
+      "learning_rate": 9.850602385839158e-06,
+      "loss": 0.7935,
+      "step": 1005
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.225027430000192,
+      "learning_rate": 9.850188615510824e-06,
+      "loss": 0.6579,
+      "step": 1006
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3382405131753483,
+      "learning_rate": 9.849774281701597e-06,
+      "loss": 0.7659,
+      "step": 1007
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.097379634081843,
+      "learning_rate": 9.849359384459614e-06,
+      "loss": 0.7244,
+      "step": 1008
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.882248967395376,
+      "learning_rate": 9.848943923833075e-06,
+      "loss": 0.699,
+      "step": 1009
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1835190905227715,
+      "learning_rate": 9.848527899870249e-06,
+      "loss": 0.729,
+      "step": 1010
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3344865062688966,
+      "learning_rate": 9.848111312619464e-06,
+      "loss": 0.7155,
+      "step": 1011
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5563191668434757,
+      "learning_rate": 9.847694162129124e-06,
+      "loss": 0.718,
+      "step": 1012
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.468563374343176,
+      "learning_rate": 9.84727644844769e-06,
+      "loss": 0.7124,
+      "step": 1013
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7246642917209405,
+      "learning_rate": 9.846858171623687e-06,
+      "loss": 0.6024,
+      "step": 1014
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0551445765951826,
+      "learning_rate": 9.846439331705715e-06,
+      "loss": 0.6882,
+      "step": 1015
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9450655022782932,
+      "learning_rate": 9.846019928742432e-06,
+      "loss": 0.7355,
+      "step": 1016
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1869980803278106,
+      "learning_rate": 9.84559996278256e-06,
+      "loss": 0.6998,
+      "step": 1017
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0957427212075115,
+      "learning_rate": 9.845179433874891e-06,
+      "loss": 0.7364,
+      "step": 1018
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2104276939981258,
+      "learning_rate": 9.844758342068284e-06,
+      "loss": 0.7528,
+      "step": 1019
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2088600503635325,
+      "learning_rate": 9.844336687411657e-06,
+      "loss": 0.637,
+      "step": 1020
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2733449333519444,
+      "learning_rate": 9.843914469953995e-06,
+      "loss": 0.7377,
+      "step": 1021
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.872012422821911,
+      "learning_rate": 9.843491689744354e-06,
+      "loss": 0.7699,
+      "step": 1022
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.471746142903541,
+      "learning_rate": 9.84306834683185e-06,
+      "loss": 0.7256,
+      "step": 1023
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5086959706339997,
+      "learning_rate": 9.842644441265664e-06,
+      "loss": 0.8149,
+      "step": 1024
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3150840456614925,
+      "learning_rate": 9.842219973095045e-06,
+      "loss": 0.7124,
+      "step": 1025
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3604359593997164,
+      "learning_rate": 9.841794942369309e-06,
+      "loss": 0.6731,
+      "step": 1026
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.439817785760237,
+      "learning_rate": 9.841369349137832e-06,
+      "loss": 0.7622,
+      "step": 1027
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9509635034152497,
+      "learning_rate": 9.840943193450059e-06,
+      "loss": 0.7069,
+      "step": 1028
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9934027081797234,
+      "learning_rate": 9.840516475355499e-06,
+      "loss": 0.7784,
+      "step": 1029
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.533628591878186,
+      "learning_rate": 9.840089194903729e-06,
+      "loss": 0.7079,
+      "step": 1030
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4056446800859272,
+      "learning_rate": 9.839661352144386e-06,
+      "loss": 0.761,
+      "step": 1031
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.515210654037056,
+      "learning_rate": 9.839232947127178e-06,
+      "loss": 0.7748,
+      "step": 1032
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.108290692947581,
+      "learning_rate": 9.838803979901874e-06,
+      "loss": 0.7569,
+      "step": 1033
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.079443665645577,
+      "learning_rate": 9.838374450518311e-06,
+      "loss": 0.7073,
+      "step": 1034
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3720082565519367,
+      "learning_rate": 9.837944359026392e-06,
+      "loss": 0.735,
+      "step": 1035
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.6597037634218994,
+      "learning_rate": 9.837513705476082e-06,
+      "loss": 0.745,
+      "step": 1036
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6631008749810534,
+      "learning_rate": 9.837082489917413e-06,
+      "loss": 0.7522,
+      "step": 1037
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7564725036319344,
+      "learning_rate": 9.836650712400484e-06,
+      "loss": 0.6971,
+      "step": 1038
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7150369300373387,
+      "learning_rate": 9.836218372975456e-06,
+      "loss": 0.7348,
+      "step": 1039
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.456649521111631,
+      "learning_rate": 9.835785471692559e-06,
+      "loss": 0.7276,
+      "step": 1040
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4789195883853457,
+      "learning_rate": 9.835352008602081e-06,
+      "loss": 0.7406,
+      "step": 1041
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.240016595350792,
+      "learning_rate": 9.834917983754388e-06,
+      "loss": 0.7213,
+      "step": 1042
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4292109109445925,
+      "learning_rate": 9.834483397199897e-06,
+      "loss": 0.7573,
+      "step": 1043
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7009207551734242,
+      "learning_rate": 9.834048248989101e-06,
+      "loss": 0.7654,
+      "step": 1044
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.359205277225582,
+      "learning_rate": 9.833612539172554e-06,
+      "loss": 0.6987,
+      "step": 1045
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.052083146121145,
+      "learning_rate": 9.833176267800874e-06,
+      "loss": 0.8056,
+      "step": 1046
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7708114222605413,
+      "learning_rate": 9.832739434924747e-06,
+      "loss": 0.7814,
+      "step": 1047
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4023504113769785,
+      "learning_rate": 9.832302040594923e-06,
+      "loss": 0.7072,
+      "step": 1048
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.926390430878239,
+      "learning_rate": 9.831864084862216e-06,
+      "loss": 0.7338,
+      "step": 1049
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.18607735411363,
+      "learning_rate": 9.831425567777506e-06,
+      "loss": 0.7444,
+      "step": 1050
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.494103967751944,
+      "learning_rate": 9.830986489391743e-06,
+      "loss": 0.7925,
+      "step": 1051
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.65375185592133,
+      "learning_rate": 9.830546849755932e-06,
+      "loss": 0.7176,
+      "step": 1052
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.350034885411002,
+      "learning_rate": 9.830106648921152e-06,
+      "loss": 0.6827,
+      "step": 1053
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2408010200349135,
+      "learning_rate": 9.829665886938544e-06,
+      "loss": 0.7066,
+      "step": 1054
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0868922059592543,
+      "learning_rate": 9.829224563859314e-06,
+      "loss": 0.6116,
+      "step": 1055
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.941803666500437,
+      "learning_rate": 9.828782679734737e-06,
+      "loss": 0.8022,
+      "step": 1056
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.823063887767071,
+      "learning_rate": 9.828340234616142e-06,
+      "loss": 0.7427,
+      "step": 1057
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.8844101888229514,
+      "learning_rate": 9.827897228554939e-06,
+      "loss": 0.7707,
+      "step": 1058
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0160506838501684,
+      "learning_rate": 9.827453661602592e-06,
+      "loss": 0.7326,
+      "step": 1059
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.300778641620372,
+      "learning_rate": 9.827009533810632e-06,
+      "loss": 0.7175,
+      "step": 1060
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.6516868846525616,
+      "learning_rate": 9.82656484523066e-06,
+      "loss": 0.7898,
+      "step": 1061
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4715363916836623,
+      "learning_rate": 9.826119595914334e-06,
+      "loss": 0.6921,
+      "step": 1062
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.327340583098541,
+      "learning_rate": 9.825673785913385e-06,
+      "loss": 0.6928,
+      "step": 1063
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.759469881660751,
+      "learning_rate": 9.825227415279606e-06,
+      "loss": 0.7855,
+      "step": 1064
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.534088445808419,
+      "learning_rate": 9.824780484064853e-06,
+      "loss": 0.7307,
+      "step": 1065
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1112574816361924,
+      "learning_rate": 9.824332992321052e-06,
+      "loss": 0.6811,
+      "step": 1066
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.373627167824405,
+      "learning_rate": 9.823884940100188e-06,
+      "loss": 0.7312,
+      "step": 1067
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7883580813619697,
+      "learning_rate": 9.823436327454318e-06,
+      "loss": 0.7676,
+      "step": 1068
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1899506229168537,
+      "learning_rate": 9.822987154435557e-06,
+      "loss": 0.7408,
+      "step": 1069
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0479852936546252,
+      "learning_rate": 9.82253742109609e-06,
+      "loss": 0.6622,
+      "step": 1070
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7416313050535015,
+      "learning_rate": 9.822087127488167e-06,
+      "loss": 0.7698,
+      "step": 1071
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.105517261367347,
+      "learning_rate": 9.821636273664102e-06,
+      "loss": 0.6482,
+      "step": 1072
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.280144562837707,
+      "learning_rate": 9.821184859676269e-06,
+      "loss": 0.6423,
+      "step": 1073
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.134409942350642,
+      "learning_rate": 9.820732885577117e-06,
+      "loss": 0.6664,
+      "step": 1074
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.912802423096457,
+      "learning_rate": 9.820280351419155e-06,
+      "loss": 0.826,
+      "step": 1075
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1928081836845386,
+      "learning_rate": 9.819827257254957e-06,
+      "loss": 0.7791,
+      "step": 1076
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.779451497550286,
+      "learning_rate": 9.81937360313716e-06,
+      "loss": 0.8068,
+      "step": 1077
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.32410870620609,
+      "learning_rate": 9.818919389118466e-06,
+      "loss": 0.7509,
+      "step": 1078
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5323788910344027,
+      "learning_rate": 9.81846461525165e-06,
+      "loss": 0.7576,
+      "step": 1079
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.439458544255558,
+      "learning_rate": 9.818009281589545e-06,
+      "loss": 0.7202,
+      "step": 1080
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0195258140646195,
+      "learning_rate": 9.817553388185046e-06,
+      "loss": 0.6668,
+      "step": 1081
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5034092301568605,
+      "learning_rate": 9.817096935091123e-06,
+      "loss": 0.6359,
+      "step": 1082
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.4584143921665347,
+      "learning_rate": 9.816639922360802e-06,
+      "loss": 0.8676,
+      "step": 1083
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1995319474627197,
+      "learning_rate": 9.816182350047179e-06,
+      "loss": 0.6817,
+      "step": 1084
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.416376258425686,
+      "learning_rate": 9.815724218203411e-06,
+      "loss": 0.7575,
+      "step": 1085
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.4149722424830817,
+      "learning_rate": 9.815265526882726e-06,
+      "loss": 0.7805,
+      "step": 1086
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3282208941621216,
+      "learning_rate": 9.814806276138412e-06,
+      "loss": 0.8052,
+      "step": 1087
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4514470562750907,
+      "learning_rate": 9.81434646602382e-06,
+      "loss": 0.7267,
+      "step": 1088
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.976660270631451,
+      "learning_rate": 9.813886096592376e-06,
+      "loss": 0.6789,
+      "step": 1089
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0977217217797643,
+      "learning_rate": 9.81342516789756e-06,
+      "loss": 0.7351,
+      "step": 1090
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 16.753908355958515,
+      "learning_rate": 9.81296367999292e-06,
+      "loss": 0.6394,
+      "step": 1091
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3815484353523138,
+      "learning_rate": 9.812501632932074e-06,
+      "loss": 0.7619,
+      "step": 1092
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.14830234781379,
+      "learning_rate": 9.8120390267687e-06,
+      "loss": 0.7141,
+      "step": 1093
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4341062988650624,
+      "learning_rate": 9.811575861556541e-06,
+      "loss": 0.776,
+      "step": 1094
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.290585984126302,
+      "learning_rate": 9.811112137349407e-06,
+      "loss": 0.8117,
+      "step": 1095
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.555561103046378,
+      "learning_rate": 9.810647854201174e-06,
+      "loss": 0.748,
+      "step": 1096
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.710466022236052,
+      "learning_rate": 9.81018301216578e-06,
+      "loss": 0.7782,
+      "step": 1097
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3963252605264445,
+      "learning_rate": 9.809717611297227e-06,
+      "loss": 0.6976,
+      "step": 1098
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6354731547094254,
+      "learning_rate": 9.809251651649586e-06,
+      "loss": 0.8057,
+      "step": 1099
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4944044637050682,
+      "learning_rate": 9.80878513327699e-06,
+      "loss": 0.6846,
+      "step": 1100
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2779646813088203,
+      "learning_rate": 9.80831805623364e-06,
+      "loss": 0.6475,
+      "step": 1101
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3609991882243317,
+      "learning_rate": 9.807850420573794e-06,
+      "loss": 0.6642,
+      "step": 1102
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2887645711932185,
+      "learning_rate": 9.807382226351786e-06,
+      "loss": 0.7091,
+      "step": 1103
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1955085410313404,
+      "learning_rate": 9.806913473622008e-06,
+      "loss": 0.6662,
+      "step": 1104
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2036596843300829,
+      "learning_rate": 9.806444162438917e-06,
+      "loss": 0.6465,
+      "step": 1105
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1349422385810124,
+      "learning_rate": 9.805974292857038e-06,
+      "loss": 0.6237,
+      "step": 1106
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5361645121575336,
+      "learning_rate": 9.805503864930958e-06,
+      "loss": 0.9142,
+      "step": 1107
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.0395622006277123,
+      "learning_rate": 9.80503287871533e-06,
+      "loss": 0.6835,
+      "step": 1108
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.422353959280125,
+      "learning_rate": 9.804561334264872e-06,
+      "loss": 0.7633,
+      "step": 1109
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5480040358296767,
+      "learning_rate": 9.804089231634368e-06,
+      "loss": 0.8203,
+      "step": 1110
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.8729382273172983,
+      "learning_rate": 9.803616570878664e-06,
+      "loss": 0.6498,
+      "step": 1111
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7811166027244028,
+      "learning_rate": 9.803143352052674e-06,
+      "loss": 0.5825,
+      "step": 1112
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.12202620735347,
+      "learning_rate": 9.802669575211369e-06,
+      "loss": 0.6428,
+      "step": 1113
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.0006468325620315,
+      "learning_rate": 9.8021952404098e-06,
+      "loss": 0.7803,
+      "step": 1114
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1558917309253327,
+      "learning_rate": 9.80172034770307e-06,
+      "loss": 0.7181,
+      "step": 1115
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.333799697036943,
+      "learning_rate": 9.801244897146348e-06,
+      "loss": 0.8098,
+      "step": 1116
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3764518557927974,
+      "learning_rate": 9.800768888794874e-06,
+      "loss": 0.6982,
+      "step": 1117
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.553267042004277,
+      "learning_rate": 9.800292322703949e-06,
+      "loss": 0.7525,
+      "step": 1118
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.418379261387043,
+      "learning_rate": 9.799815198928937e-06,
+      "loss": 0.7675,
+      "step": 1119
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.8872564473960765,
+      "learning_rate": 9.79933751752527e-06,
+      "loss": 0.6478,
+      "step": 1120
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.052357385896143,
+      "learning_rate": 9.798859278548443e-06,
+      "loss": 0.7681,
+      "step": 1121
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2984285503416286,
+      "learning_rate": 9.798380482054019e-06,
+      "loss": 0.6921,
+      "step": 1122
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.422351898914631,
+      "learning_rate": 9.79790112809762e-06,
+      "loss": 0.7374,
+      "step": 1123
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2410791102035756,
+      "learning_rate": 9.797421216734938e-06,
+      "loss": 0.664,
+      "step": 1124
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.470129472972923,
+      "learning_rate": 9.796940748021727e-06,
+      "loss": 0.7462,
+      "step": 1125
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.414314965943791,
+      "learning_rate": 9.796459722013804e-06,
+      "loss": 0.7995,
+      "step": 1126
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.270258241210417,
+      "learning_rate": 9.795978138767059e-06,
+      "loss": 0.7509,
+      "step": 1127
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.353498649269498,
+      "learning_rate": 9.795495998337436e-06,
+      "loss": 0.8035,
+      "step": 1128
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.8698790645980807,
+      "learning_rate": 9.795013300780951e-06,
+      "loss": 0.7457,
+      "step": 1129
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.6109946770346673,
+      "learning_rate": 9.794530046153681e-06,
+      "loss": 0.7238,
+      "step": 1130
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.488339655347247,
+      "learning_rate": 9.79404623451177e-06,
+      "loss": 0.7118,
+      "step": 1131
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9913223737937966,
+      "learning_rate": 9.793561865911425e-06,
+      "loss": 0.76,
+      "step": 1132
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.4767066337380035,
+      "learning_rate": 9.793076940408921e-06,
+      "loss": 0.6867,
+      "step": 1133
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.657626022399911,
+      "learning_rate": 9.792591458060592e-06,
+      "loss": 0.7424,
+      "step": 1134
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.0031975051776887,
+      "learning_rate": 9.792105418922842e-06,
+      "loss": 0.6451,
+      "step": 1135
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.515412387819786,
+      "learning_rate": 9.791618823052137e-06,
+      "loss": 0.654,
+      "step": 1136
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2988508182868506,
+      "learning_rate": 9.791131670505008e-06,
+      "loss": 0.7053,
+      "step": 1137
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.392212351593364,
+      "learning_rate": 9.790643961338051e-06,
+      "loss": 0.6784,
+      "step": 1138
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3321628230545013,
+      "learning_rate": 9.790155695607927e-06,
+      "loss": 0.747,
+      "step": 1139
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5520902735537194,
+      "learning_rate": 9.789666873371361e-06,
+      "loss": 0.7747,
+      "step": 1140
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.0561347446951537,
+      "learning_rate": 9.789177494685146e-06,
+      "loss": 0.7497,
+      "step": 1141
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.42722295869959,
+      "learning_rate": 9.788687559606131e-06,
+      "loss": 0.7141,
+      "step": 1142
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.087792011146286,
+      "learning_rate": 9.788197068191237e-06,
+      "loss": 0.78,
+      "step": 1143
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.742833380122242,
+      "learning_rate": 9.787706020497451e-06,
+      "loss": 0.702,
+      "step": 1144
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2623934679268323,
+      "learning_rate": 9.787214416581818e-06,
+      "loss": 0.6702,
+      "step": 1145
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.105793673298671,
+      "learning_rate": 9.786722256501454e-06,
+      "loss": 0.742,
+      "step": 1146
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.771703607077318,
+      "learning_rate": 9.786229540313534e-06,
+      "loss": 0.6825,
+      "step": 1147
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.833220908887274,
+      "learning_rate": 9.785736268075303e-06,
+      "loss": 0.6402,
+      "step": 1148
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.252609192328139,
+      "learning_rate": 9.785242439844064e-06,
+      "loss": 0.7542,
+      "step": 1149
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6298483732855993,
+      "learning_rate": 9.784748055677193e-06,
+      "loss": 0.758,
+      "step": 1150
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1632996471894916,
+      "learning_rate": 9.784253115632125e-06,
+      "loss": 0.7049,
+      "step": 1151
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.34611962892255,
+      "learning_rate": 9.783757619766359e-06,
+      "loss": 0.6605,
+      "step": 1152
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3000195548875086,
+      "learning_rate": 9.783261568137461e-06,
+      "loss": 0.6681,
+      "step": 1153
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.644856326527659,
+      "learning_rate": 9.78276496080306e-06,
+      "loss": 0.7254,
+      "step": 1154
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.363577339292242,
+      "learning_rate": 9.782267797820852e-06,
+      "loss": 0.6255,
+      "step": 1155
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.178294432104343,
+      "learning_rate": 9.781770079248597e-06,
+      "loss": 0.6964,
+      "step": 1156
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.351307497143435,
+      "learning_rate": 9.781271805144115e-06,
+      "loss": 0.672,
+      "step": 1157
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.317111010522036,
+      "learning_rate": 9.780772975565297e-06,
+      "loss": 0.716,
+      "step": 1158
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.74680266959224,
+      "learning_rate": 9.780273590570095e-06,
+      "loss": 0.7632,
+      "step": 1159
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.8612186188840654,
+      "learning_rate": 9.779773650216524e-06,
+      "loss": 0.688,
+      "step": 1160
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.73767312113061,
+      "learning_rate": 9.779273154562668e-06,
+      "loss": 0.732,
+      "step": 1161
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.195963737502883,
+      "learning_rate": 9.778772103666672e-06,
+      "loss": 0.7151,
+      "step": 1162
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.842525432869639,
+      "learning_rate": 9.778270497586747e-06,
+      "loss": 0.6845,
+      "step": 1163
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.134809149072051,
+      "learning_rate": 9.77776833638117e-06,
+      "loss": 0.7597,
+      "step": 1164
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7009490835040846,
+      "learning_rate": 9.777265620108277e-06,
+      "loss": 0.7779,
+      "step": 1165
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.317959306339774,
+      "learning_rate": 9.776762348826474e-06,
+      "loss": 0.7794,
+      "step": 1166
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.109072732996193,
+      "learning_rate": 9.776258522594231e-06,
+      "loss": 0.7258,
+      "step": 1167
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1691103374743497,
+      "learning_rate": 9.775754141470077e-06,
+      "loss": 0.7235,
+      "step": 1168
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2658839932977033,
+      "learning_rate": 9.775249205512614e-06,
+      "loss": 0.7238,
+      "step": 1169
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4497606850003515,
+      "learning_rate": 9.774743714780502e-06,
+      "loss": 0.7825,
+      "step": 1170
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7201508522085223,
+      "learning_rate": 9.774237669332467e-06,
+      "loss": 0.7096,
+      "step": 1171
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.194344873698593,
+      "learning_rate": 9.7737310692273e-06,
+      "loss": 0.7806,
+      "step": 1172
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4546802907254675,
+      "learning_rate": 9.77322391452386e-06,
+      "loss": 0.7075,
+      "step": 1173
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.8217799648099753,
+      "learning_rate": 9.772716205281061e-06,
+      "loss": 0.6641,
+      "step": 1174
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3940574360026314,
+      "learning_rate": 9.772207941557889e-06,
+      "loss": 0.7329,
+      "step": 1175
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.8617824561808356,
+      "learning_rate": 9.771699123413396e-06,
+      "loss": 0.6231,
+      "step": 1176
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4154376282640087,
+      "learning_rate": 9.77118975090669e-06,
+      "loss": 0.6116,
+      "step": 1177
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.378924743383222,
+      "learning_rate": 9.770679824096952e-06,
+      "loss": 0.6989,
+      "step": 1178
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7001805333163587,
+      "learning_rate": 9.770169343043423e-06,
+      "loss": 0.7121,
+      "step": 1179
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5402821969530742,
+      "learning_rate": 9.769658307805408e-06,
+      "loss": 0.7407,
+      "step": 1180
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4849409843596,
+      "learning_rate": 9.769146718442279e-06,
+      "loss": 0.7695,
+      "step": 1181
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.36462157041008,
+      "learning_rate": 9.76863457501347e-06,
+      "loss": 0.7366,
+      "step": 1182
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5569947068679664,
+      "learning_rate": 9.76812187757848e-06,
+      "loss": 0.7354,
+      "step": 1183
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.230897787375336,
+      "learning_rate": 9.767608626196877e-06,
+      "loss": 0.6896,
+      "step": 1184
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.322170991929863,
+      "learning_rate": 9.767094820928282e-06,
+      "loss": 0.7377,
+      "step": 1185
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.766972909981728,
+      "learning_rate": 9.766580461832391e-06,
+      "loss": 0.7322,
+      "step": 1186
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9936899282121772,
+      "learning_rate": 9.766065548968962e-06,
+      "loss": 0.7133,
+      "step": 1187
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9245738454667074,
+      "learning_rate": 9.765550082397815e-06,
+      "loss": 0.7225,
+      "step": 1188
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9105025784741505,
+      "learning_rate": 9.765034062178836e-06,
+      "loss": 0.773,
+      "step": 1189
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3491105179808205,
+      "learning_rate": 9.764517488371971e-06,
+      "loss": 0.7345,
+      "step": 1190
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.042045767945915,
+      "learning_rate": 9.76400036103724e-06,
+      "loss": 0.6984,
+      "step": 1191
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.452386829371169,
+      "learning_rate": 9.763482680234718e-06,
+      "loss": 0.7039,
+      "step": 1192
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3004751526217735,
+      "learning_rate": 9.762964446024547e-06,
+      "loss": 0.7878,
+      "step": 1193
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3235197535321637,
+      "learning_rate": 9.762445658466935e-06,
+      "loss": 0.6694,
+      "step": 1194
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.539843716603838,
+      "learning_rate": 9.761926317622154e-06,
+      "loss": 0.741,
+      "step": 1195
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.849640472644429,
+      "learning_rate": 9.761406423550539e-06,
+      "loss": 0.7408,
+      "step": 1196
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5166258673564514,
+      "learning_rate": 9.760885976312488e-06,
+      "loss": 0.7223,
+      "step": 1197
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.725239057841064,
+      "learning_rate": 9.760364975968469e-06,
+      "loss": 0.7007,
+      "step": 1198
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.7294366804059464,
+      "learning_rate": 9.759843422579005e-06,
+      "loss": 0.6748,
+      "step": 1199
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.7738891360490694,
+      "learning_rate": 9.759321316204693e-06,
+      "loss": 0.6751,
+      "step": 1200
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 16.17668799017805,
+      "learning_rate": 9.758798656906187e-06,
+      "loss": 0.7039,
+      "step": 1201
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8538609592861137,
+      "learning_rate": 9.758275444744211e-06,
+      "loss": 0.7457,
+      "step": 1202
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8412177031201455,
+      "learning_rate": 9.757751679779549e-06,
+      "loss": 0.6996,
+      "step": 1203
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4794823846058103,
+      "learning_rate": 9.757227362073048e-06,
+      "loss": 0.7336,
+      "step": 1204
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9919272034069202,
+      "learning_rate": 9.756702491685626e-06,
+      "loss": 0.7235,
+      "step": 1205
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.682936702377847,
+      "learning_rate": 9.756177068678258e-06,
+      "loss": 0.6468,
+      "step": 1206
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2979583262392227,
+      "learning_rate": 9.755651093111987e-06,
+      "loss": 0.7306,
+      "step": 1207
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.631349641332311,
+      "learning_rate": 9.755124565047918e-06,
+      "loss": 0.7321,
+      "step": 1208
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.09817214920078,
+      "learning_rate": 9.754597484547223e-06,
+      "loss": 0.6638,
+      "step": 1209
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.909553694559456,
+      "learning_rate": 9.754069851671138e-06,
+      "loss": 0.7271,
+      "step": 1210
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0625344333106828,
+      "learning_rate": 9.753541666480959e-06,
+      "loss": 0.6552,
+      "step": 1211
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2516687148488987,
+      "learning_rate": 9.75301292903805e-06,
+      "loss": 0.7763,
+      "step": 1212
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0525856091150914,
+      "learning_rate": 9.752483639403839e-06,
+      "loss": 0.7237,
+      "step": 1213
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3517703119170414,
+      "learning_rate": 9.751953797639817e-06,
+      "loss": 0.6346,
+      "step": 1214
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5336050253735287,
+      "learning_rate": 9.751423403807539e-06,
+      "loss": 0.7032,
+      "step": 1215
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.274623839914132,
+      "learning_rate": 9.750892457968626e-06,
+      "loss": 0.7113,
+      "step": 1216
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.502789805441234,
+      "learning_rate": 9.75036096018476e-06,
+      "loss": 0.6948,
+      "step": 1217
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.362716780752687,
+      "learning_rate": 9.749828910517688e-06,
+      "loss": 0.6797,
+      "step": 1218
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9013855117687954,
+      "learning_rate": 9.749296309029224e-06,
+      "loss": 0.7267,
+      "step": 1219
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5275074129127404,
+      "learning_rate": 9.748763155781244e-06,
+      "loss": 0.719,
+      "step": 1220
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.7639603698409827,
+      "learning_rate": 9.748229450835689e-06,
+      "loss": 0.6755,
+      "step": 1221
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8030238737546624,
+      "learning_rate": 9.747695194254561e-06,
+      "loss": 0.7427,
+      "step": 1222
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.820939766808103,
+      "learning_rate": 9.74716038609993e-06,
+      "loss": 0.7745,
+      "step": 1223
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.48939064756324,
+      "learning_rate": 9.746625026433929e-06,
+      "loss": 0.6752,
+      "step": 1224
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.7729898492722955,
+      "learning_rate": 9.746089115318751e-06,
+      "loss": 0.6633,
+      "step": 1225
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5642579652047615,
+      "learning_rate": 9.745552652816662e-06,
+      "loss": 0.6911,
+      "step": 1226
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.479378462117706,
+      "learning_rate": 9.74501563898998e-06,
+      "loss": 0.6961,
+      "step": 1227
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.517261027107456,
+      "learning_rate": 9.744478073901102e-06,
+      "loss": 0.6875,
+      "step": 1228
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.143063814038163,
+      "learning_rate": 9.743939957612473e-06,
+      "loss": 0.7862,
+      "step": 1229
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.186066134453419,
+      "learning_rate": 9.743401290186615e-06,
+      "loss": 0.6905,
+      "step": 1230
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4988157638795383,
+      "learning_rate": 9.742862071686105e-06,
+      "loss": 0.6065,
+      "step": 1231
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.944980331009129,
+      "learning_rate": 9.742322302173591e-06,
+      "loss": 0.8424,
+      "step": 1232
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4840746884722997,
+      "learning_rate": 9.74178198171178e-06,
+      "loss": 0.7242,
+      "step": 1233
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3564467576176615,
+      "learning_rate": 9.741241110363446e-06,
+      "loss": 0.5671,
+      "step": 1234
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5428398872084914,
+      "learning_rate": 9.740699688191426e-06,
+      "loss": 0.7558,
+      "step": 1235
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.987401358325348,
+      "learning_rate": 9.74015771525862e-06,
+      "loss": 0.7976,
+      "step": 1236
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.288527677436737,
+      "learning_rate": 9.73961519162799e-06,
+      "loss": 0.7495,
+      "step": 1237
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.203656023458977,
+      "learning_rate": 9.739072117362572e-06,
+      "loss": 0.6546,
+      "step": 1238
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.893448106608211,
+      "learning_rate": 9.738528492525454e-06,
+      "loss": 0.7536,
+      "step": 1239
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.562798821263755,
+      "learning_rate": 9.737984317179792e-06,
+      "loss": 0.6901,
+      "step": 1240
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.525831897954803,
+      "learning_rate": 9.737439591388808e-06,
+      "loss": 0.7072,
+      "step": 1241
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.390376616064533,
+      "learning_rate": 9.736894315215788e-06,
+      "loss": 0.6709,
+      "step": 1242
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.290595026707291,
+      "learning_rate": 9.736348488724078e-06,
+      "loss": 0.7022,
+      "step": 1243
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0135726421842928,
+      "learning_rate": 9.735802111977093e-06,
+      "loss": 0.7398,
+      "step": 1244
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.4970745285376745,
+      "learning_rate": 9.735255185038308e-06,
+      "loss": 0.7238,
+      "step": 1245
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5294324112992945,
+      "learning_rate": 9.734707707971265e-06,
+      "loss": 0.6654,
+      "step": 1246
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.769086981727464,
+      "learning_rate": 9.734159680839566e-06,
+      "loss": 0.7625,
+      "step": 1247
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.425192205003877,
+      "learning_rate": 9.733611103706882e-06,
+      "loss": 0.7288,
+      "step": 1248
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4432588819312016,
+      "learning_rate": 9.73306197663694e-06,
+      "loss": 0.7453,
+      "step": 1249
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4916017795323224,
+      "learning_rate": 9.732512299693542e-06,
+      "loss": 0.6687,
+      "step": 1250
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.171083266500421,
+      "learning_rate": 9.731962072940545e-06,
+      "loss": 0.7657,
+      "step": 1251
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2828106585662606,
+      "learning_rate": 9.731411296441873e-06,
+      "loss": 0.7871,
+      "step": 1252
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.2796867996521595,
+      "learning_rate": 9.730859970261514e-06,
+      "loss": 0.7616,
+      "step": 1253
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9365596197667454,
+      "learning_rate": 9.730308094463519e-06,
+      "loss": 0.7857,
+      "step": 1254
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.571904631777549,
+      "learning_rate": 9.729755669112003e-06,
+      "loss": 0.7116,
+      "step": 1255
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5467525821131054,
+      "learning_rate": 9.729202694271145e-06,
+      "loss": 0.6199,
+      "step": 1256
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2962244446782196,
+      "learning_rate": 9.72864917000519e-06,
+      "loss": 0.758,
+      "step": 1257
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.140094567957693,
+      "learning_rate": 9.728095096378443e-06,
+      "loss": 0.7457,
+      "step": 1258
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4307560819993066,
+      "learning_rate": 9.727540473455277e-06,
+      "loss": 0.7374,
+      "step": 1259
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4574501390436145,
+      "learning_rate": 9.726985301300122e-06,
+      "loss": 0.7201,
+      "step": 1260
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5770663455788885,
+      "learning_rate": 9.72642957997748e-06,
+      "loss": 0.6912,
+      "step": 1261
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6051783139637252,
+      "learning_rate": 9.725873309551915e-06,
+      "loss": 0.7906,
+      "step": 1262
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6247042237572757,
+      "learning_rate": 9.725316490088046e-06,
+      "loss": 0.7596,
+      "step": 1263
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.208978552737289,
+      "learning_rate": 9.724759121650569e-06,
+      "loss": 0.6893,
+      "step": 1264
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.818237537685943,
+      "learning_rate": 9.724201204304234e-06,
+      "loss": 0.6509,
+      "step": 1265
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.516551806674273,
+      "learning_rate": 9.72364273811386e-06,
+      "loss": 0.8091,
+      "step": 1266
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8775423326241003,
+      "learning_rate": 9.723083723144326e-06,
+      "loss": 0.6664,
+      "step": 1267
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1104989369101292,
+      "learning_rate": 9.722524159460579e-06,
+      "loss": 0.7484,
+      "step": 1268
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.284433621967399,
+      "learning_rate": 9.721964047127627e-06,
+      "loss": 0.7694,
+      "step": 1269
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.528762803348098,
+      "learning_rate": 9.721403386210542e-06,
+      "loss": 0.7047,
+      "step": 1270
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3698967383567053,
+      "learning_rate": 9.720842176774458e-06,
+      "loss": 0.7303,
+      "step": 1271
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2862180267688195,
+      "learning_rate": 9.720280418884578e-06,
+      "loss": 0.6719,
+      "step": 1272
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.585341957705507,
+      "learning_rate": 9.719718112606163e-06,
+      "loss": 0.6893,
+      "step": 1273
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.931292145704906,
+      "learning_rate": 9.719155258004542e-06,
+      "loss": 0.7129,
+      "step": 1274
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.930050695251463,
+      "learning_rate": 9.718591855145105e-06,
+      "loss": 0.7024,
+      "step": 1275
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3725890130265377,
+      "learning_rate": 9.718027904093306e-06,
+      "loss": 0.6779,
+      "step": 1276
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.667713730201913,
+      "learning_rate": 9.717463404914661e-06,
+      "loss": 0.7128,
+      "step": 1277
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4476996596913376,
+      "learning_rate": 9.716898357674757e-06,
+      "loss": 0.725,
+      "step": 1278
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.286403952573892,
+      "learning_rate": 9.716332762439238e-06,
+      "loss": 0.6882,
+      "step": 1279
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.187379321806065,
+      "learning_rate": 9.71576661927381e-06,
+      "loss": 0.6514,
+      "step": 1280
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.048461731248665,
+      "learning_rate": 9.71519992824425e-06,
+      "loss": 0.69,
+      "step": 1281
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4471121642879634,
+      "learning_rate": 9.714632689416392e-06,
+      "loss": 0.6853,
+      "step": 1282
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3522287088330436,
+      "learning_rate": 9.714064902856136e-06,
+      "loss": 0.6857,
+      "step": 1283
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8039331155505693,
+      "learning_rate": 9.713496568629447e-06,
+      "loss": 0.7953,
+      "step": 1284
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5431568191768257,
+      "learning_rate": 9.71292768680235e-06,
+      "loss": 0.6906,
+      "step": 1285
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.734010132334214,
+      "learning_rate": 9.712358257440942e-06,
+      "loss": 0.7469,
+      "step": 1286
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.766837316455564,
+      "learning_rate": 9.711788280611371e-06,
+      "loss": 0.7425,
+      "step": 1287
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2607094221443815,
+      "learning_rate": 9.711217756379859e-06,
+      "loss": 0.6745,
+      "step": 1288
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.718072945572998,
+      "learning_rate": 9.710646684812686e-06,
+      "loss": 0.6632,
+      "step": 1289
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.220340692493552,
+      "learning_rate": 9.7100750659762e-06,
+      "loss": 0.7289,
+      "step": 1290
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.630360305784288,
+      "learning_rate": 9.709502899936805e-06,
+      "loss": 0.6569,
+      "step": 1291
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7311986179292194,
+      "learning_rate": 9.70893018676098e-06,
+      "loss": 0.7763,
+      "step": 1292
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.292026620226449,
+      "learning_rate": 9.708356926515256e-06,
+      "loss": 0.7046,
+      "step": 1293
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.099523665954287,
+      "learning_rate": 9.707783119266236e-06,
+      "loss": 0.7467,
+      "step": 1294
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9843246706602047,
+      "learning_rate": 9.707208765080583e-06,
+      "loss": 0.7871,
+      "step": 1295
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.603495110130337,
+      "learning_rate": 9.706633864025021e-06,
+      "loss": 0.6149,
+      "step": 1296
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9957523727453244,
+      "learning_rate": 9.706058416166342e-06,
+      "loss": 0.6866,
+      "step": 1297
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4581527501195923,
+      "learning_rate": 9.705482421571401e-06,
+      "loss": 0.7306,
+      "step": 1298
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.75978405172835,
+      "learning_rate": 9.704905880307113e-06,
+      "loss": 0.704,
+      "step": 1299
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6219298892401675,
+      "learning_rate": 9.704328792440462e-06,
+      "loss": 0.7072,
+      "step": 1300
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1583848744671967,
+      "learning_rate": 9.70375115803849e-06,
+      "loss": 0.6726,
+      "step": 1301
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7991267040996943,
+      "learning_rate": 9.703172977168307e-06,
+      "loss": 0.6575,
+      "step": 1302
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.725583520285742,
+      "learning_rate": 9.702594249897082e-06,
+      "loss": 0.735,
+      "step": 1303
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2733331053327124,
+      "learning_rate": 9.70201497629205e-06,
+      "loss": 0.7678,
+      "step": 1304
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3225063544881346,
+      "learning_rate": 9.701435156420511e-06,
+      "loss": 0.6528,
+      "step": 1305
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5695618934438342,
+      "learning_rate": 9.700854790349826e-06,
+      "loss": 0.7385,
+      "step": 1306
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.165748397571203,
+      "learning_rate": 9.700273878147419e-06,
+      "loss": 0.6923,
+      "step": 1307
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.946583806133681,
+      "learning_rate": 9.699692419880782e-06,
+      "loss": 0.6818,
+      "step": 1308
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7377235660563417,
+      "learning_rate": 9.699110415617464e-06,
+      "loss": 0.6344,
+      "step": 1309
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.216540668421752,
+      "learning_rate": 9.698527865425083e-06,
+      "loss": 0.7325,
+      "step": 1310
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5772415638304755,
+      "learning_rate": 9.697944769371315e-06,
+      "loss": 0.6738,
+      "step": 1311
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4296412096784854,
+      "learning_rate": 9.697361127523905e-06,
+      "loss": 0.7011,
+      "step": 1312
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.0442601673343894,
+      "learning_rate": 9.696776939950657e-06,
+      "loss": 0.7084,
+      "step": 1313
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.075407781881538,
+      "learning_rate": 9.696192206719441e-06,
+      "loss": 0.7175,
+      "step": 1314
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.190682088347276,
+      "learning_rate": 9.69560692789819e-06,
+      "loss": 0.718,
+      "step": 1315
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.203702468047881,
+      "learning_rate": 9.695021103554901e-06,
+      "loss": 0.78,
+      "step": 1316
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.334461505397333,
+      "learning_rate": 9.694434733757632e-06,
+      "loss": 0.7062,
+      "step": 1317
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3787385280317968,
+      "learning_rate": 9.693847818574504e-06,
+      "loss": 0.6883,
+      "step": 1318
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7061551304241709,
+      "learning_rate": 9.693260358073707e-06,
+      "loss": 0.7412,
+      "step": 1319
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.5189339975333427,
+      "learning_rate": 9.692672352323486e-06,
+      "loss": 0.7493,
+      "step": 1320
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.718227304089354,
+      "learning_rate": 9.692083801392158e-06,
+      "loss": 0.7298,
+      "step": 1321
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.863146583225161,
+      "learning_rate": 9.691494705348097e-06,
+      "loss": 0.685,
+      "step": 1322
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.224062335592531,
+      "learning_rate": 9.690905064259744e-06,
+      "loss": 0.7131,
+      "step": 1323
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3482124963584043,
+      "learning_rate": 9.690314878195599e-06,
+      "loss": 0.7642,
+      "step": 1324
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.802943387618265,
+      "learning_rate": 9.68972414722423e-06,
+      "loss": 0.7402,
+      "step": 1325
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.439325155682455,
+      "learning_rate": 9.689132871414266e-06,
+      "loss": 0.6827,
+      "step": 1326
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.283544545333155,
+      "learning_rate": 9.688541050834402e-06,
+      "loss": 0.7649,
+      "step": 1327
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8881726166539816,
+      "learning_rate": 9.68794868555339e-06,
+      "loss": 0.7581,
+      "step": 1328
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5386184576990725,
+      "learning_rate": 9.687355775640052e-06,
+      "loss": 0.7252,
+      "step": 1329
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6298695027488703,
+      "learning_rate": 9.68676232116327e-06,
+      "loss": 0.7742,
+      "step": 1330
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.084061296828642,
+      "learning_rate": 9.686168322191988e-06,
+      "loss": 0.6,
+      "step": 1331
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3716239247862054,
+      "learning_rate": 9.685573778795218e-06,
+      "loss": 0.5694,
+      "step": 1332
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2053716555139404,
+      "learning_rate": 9.684978691042031e-06,
+      "loss": 0.7231,
+      "step": 1333
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.844182707802076,
+      "learning_rate": 9.684383059001562e-06,
+      "loss": 0.6835,
+      "step": 1334
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.390207972128154,
+      "learning_rate": 9.68378688274301e-06,
+      "loss": 0.7385,
+      "step": 1335
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4980529568022267,
+      "learning_rate": 9.683190162335638e-06,
+      "loss": 0.8148,
+      "step": 1336
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.308242765317922,
+      "learning_rate": 9.68259289784877e-06,
+      "loss": 0.604,
+      "step": 1337
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5111104713428425,
+      "learning_rate": 9.681995089351797e-06,
+      "loss": 0.6423,
+      "step": 1338
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.184046389884775,
+      "learning_rate": 9.681396736914169e-06,
+      "loss": 0.8212,
+      "step": 1339
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3349747833725316,
+      "learning_rate": 9.680797840605398e-06,
+      "loss": 0.724,
+      "step": 1340
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.357395284413713,
+      "learning_rate": 9.680198400495067e-06,
+      "loss": 0.7505,
+      "step": 1341
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.237004929584163,
+      "learning_rate": 9.679598416652814e-06,
+      "loss": 0.7294,
+      "step": 1342
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3140384711893978,
+      "learning_rate": 9.678997889148342e-06,
+      "loss": 0.7354,
+      "step": 1343
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.396103673854701,
+      "learning_rate": 9.678396818051423e-06,
+      "loss": 0.7875,
+      "step": 1344
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4741313274912935,
+      "learning_rate": 9.677795203431886e-06,
+      "loss": 0.7377,
+      "step": 1345
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4119548772690003,
+      "learning_rate": 9.677193045359626e-06,
+      "loss": 0.6744,
+      "step": 1346
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.0195566082707535,
+      "learning_rate": 9.676590343904595e-06,
+      "loss": 0.5983,
+      "step": 1347
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.429934472959647,
+      "learning_rate": 9.675987099136817e-06,
+      "loss": 0.6282,
+      "step": 1348
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.399876917459758,
+      "learning_rate": 9.675383311126376e-06,
+      "loss": 0.7758,
+      "step": 1349
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3597263333081977,
+      "learning_rate": 9.674778979943417e-06,
+      "loss": 0.7706,
+      "step": 1350
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.582735032637409,
+      "learning_rate": 9.67417410565815e-06,
+      "loss": 0.6888,
+      "step": 1351
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2457596734399106,
+      "learning_rate": 9.673568688340846e-06,
+      "loss": 0.6083,
+      "step": 1352
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2475959881242735,
+      "learning_rate": 9.672962728061842e-06,
+      "loss": 0.7384,
+      "step": 1353
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7160143206592346,
+      "learning_rate": 9.672356224891536e-06,
+      "loss": 0.6937,
+      "step": 1354
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5502591319299843,
+      "learning_rate": 9.671749178900392e-06,
+      "loss": 0.7563,
+      "step": 1355
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6327155581660455,
+      "learning_rate": 9.67114159015893e-06,
+      "loss": 0.6819,
+      "step": 1356
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.131038039139984,
+      "learning_rate": 9.670533458737744e-06,
+      "loss": 0.7373,
+      "step": 1357
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.385584377854225,
+      "learning_rate": 9.66992478470748e-06,
+      "loss": 0.7657,
+      "step": 1358
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.320376159534197,
+      "learning_rate": 9.669315568138854e-06,
+      "loss": 0.6898,
+      "step": 1359
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8074892746810742,
+      "learning_rate": 9.668705809102644e-06,
+      "loss": 0.6807,
+      "step": 1360
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8098437749259717,
+      "learning_rate": 9.668095507669688e-06,
+      "loss": 0.6959,
+      "step": 1361
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7943446077345855,
+      "learning_rate": 9.66748466391089e-06,
+      "loss": 0.7087,
+      "step": 1362
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.059820137036976,
+      "learning_rate": 9.666873277897216e-06,
+      "loss": 0.6634,
+      "step": 1363
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.575688586030959,
+      "learning_rate": 9.666261349699696e-06,
+      "loss": 0.6982,
+      "step": 1364
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.262608946469177,
+      "learning_rate": 9.66564887938942e-06,
+      "loss": 0.8071,
+      "step": 1365
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.311694063851022,
+      "learning_rate": 9.665035867037545e-06,
+      "loss": 0.7185,
+      "step": 1366
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1721646707712954,
+      "learning_rate": 9.66442231271529e-06,
+      "loss": 0.6839,
+      "step": 1367
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.056959685399615,
+      "learning_rate": 9.663808216493931e-06,
+      "loss": 0.6246,
+      "step": 1368
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.211417541737079,
+      "learning_rate": 9.663193578444815e-06,
+      "loss": 0.7382,
+      "step": 1369
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.482753174472199,
+      "learning_rate": 9.662578398639353e-06,
+      "loss": 0.7261,
+      "step": 1370
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8419175901293308,
+      "learning_rate": 9.661962677149007e-06,
+      "loss": 0.7479,
+      "step": 1371
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.562383209374972,
+      "learning_rate": 9.661346414045315e-06,
+      "loss": 0.6338,
+      "step": 1372
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1133024703227266,
+      "learning_rate": 9.66072960939987e-06,
+      "loss": 0.8398,
+      "step": 1373
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.209697515295642,
+      "learning_rate": 9.660112263284334e-06,
+      "loss": 0.6979,
+      "step": 1374
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.316723331737145,
+      "learning_rate": 9.659494375770424e-06,
+      "loss": 0.7773,
+      "step": 1375
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3531621544495867,
+      "learning_rate": 9.65887594692993e-06,
+      "loss": 0.7506,
+      "step": 1376
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.6434780012213634,
+      "learning_rate": 9.658256976834692e-06,
+      "loss": 0.7223,
+      "step": 1377
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.899949137691732,
+      "learning_rate": 9.657637465556626e-06,
+      "loss": 0.7059,
+      "step": 1378
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.154298474199794,
+      "learning_rate": 9.657017413167702e-06,
+      "loss": 0.6951,
+      "step": 1379
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.762468831498145,
+      "learning_rate": 9.656396819739959e-06,
+      "loss": 0.7888,
+      "step": 1380
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3312043862215792,
+      "learning_rate": 9.655775685345493e-06,
+      "loss": 0.6494,
+      "step": 1381
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5501629949975944,
+      "learning_rate": 9.655154010056464e-06,
+      "loss": 0.8004,
+      "step": 1382
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.64326272620766,
+      "learning_rate": 9.654531793945102e-06,
+      "loss": 0.7567,
+      "step": 1383
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.575524144576267,
+      "learning_rate": 9.653909037083689e-06,
+      "loss": 0.7945,
+      "step": 1384
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9295057714053,
+      "learning_rate": 9.653285739544578e-06,
+      "loss": 0.7318,
+      "step": 1385
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0997333286555233,
+      "learning_rate": 9.65266190140018e-06,
+      "loss": 0.6631,
+      "step": 1386
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2304269290377787,
+      "learning_rate": 9.652037522722974e-06,
+      "loss": 0.6277,
+      "step": 1387
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.44783772978294,
+      "learning_rate": 9.651412603585495e-06,
+      "loss": 0.7528,
+      "step": 1388
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.178234675495518,
+      "learning_rate": 9.650787144060345e-06,
+      "loss": 0.753,
+      "step": 1389
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2283048639192446,
+      "learning_rate": 9.65016114422019e-06,
+      "loss": 0.7198,
+      "step": 1390
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4993099361247664,
+      "learning_rate": 9.649534604137755e-06,
+      "loss": 0.7052,
+      "step": 1391
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1995602061556725,
+      "learning_rate": 9.64890752388583e-06,
+      "loss": 0.7257,
+      "step": 1392
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2307125057623227,
+      "learning_rate": 9.648279903537268e-06,
+      "loss": 0.7356,
+      "step": 1393
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1622062738792347,
+      "learning_rate": 9.647651743164983e-06,
+      "loss": 0.6852,
+      "step": 1394
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3927400290041,
+      "learning_rate": 9.647023042841953e-06,
+      "loss": 0.6911,
+      "step": 1395
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.504113313434371,
+      "learning_rate": 9.64639380264122e-06,
+      "loss": 0.7242,
+      "step": 1396
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.837839765380396,
+      "learning_rate": 9.645764022635886e-06,
+      "loss": 0.6788,
+      "step": 1397
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2209452355398236,
+      "learning_rate": 9.645133702899116e-06,
+      "loss": 0.6795,
+      "step": 1398
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3171450711888,
+      "learning_rate": 9.644502843504141e-06,
+      "loss": 0.7098,
+      "step": 1399
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.433039777766387,
+      "learning_rate": 9.643871444524253e-06,
+      "loss": 0.7227,
+      "step": 1400
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.572768002891731,
+      "learning_rate": 9.643239506032805e-06,
+      "loss": 0.7593,
+      "step": 1401
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.125858662151184,
+      "learning_rate": 9.642607028103212e-06,
+      "loss": 0.7119,
+      "step": 1402
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.685061632087422,
+      "learning_rate": 9.641974010808954e-06,
+      "loss": 0.6543,
+      "step": 1403
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.22659599285853,
+      "learning_rate": 9.641340454223576e-06,
+      "loss": 0.7721,
+      "step": 1404
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4662300157631596,
+      "learning_rate": 9.64070635842068e-06,
+      "loss": 0.7117,
+      "step": 1405
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.643822549271141,
+      "learning_rate": 9.640071723473934e-06,
+      "loss": 0.7299,
+      "step": 1406
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.595323686997289,
+      "learning_rate": 9.639436549457069e-06,
+      "loss": 0.7492,
+      "step": 1407
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4646272228107784,
+      "learning_rate": 9.638800836443875e-06,
+      "loss": 0.7249,
+      "step": 1408
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.889739658027975,
+      "learning_rate": 9.638164584508211e-06,
+      "loss": 0.7073,
+      "step": 1409
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.731955437980805,
+      "learning_rate": 9.637527793723993e-06,
+      "loss": 0.6637,
+      "step": 1410
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.330236462820327,
+      "learning_rate": 9.6368904641652e-06,
+      "loss": 0.6849,
+      "step": 1411
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.09841411709079,
+      "learning_rate": 9.63625259590588e-06,
+      "loss": 0.7158,
+      "step": 1412
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.629199349370734,
+      "learning_rate": 9.635614189020133e-06,
+      "loss": 0.7321,
+      "step": 1413
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.344946491188394,
+      "learning_rate": 9.63497524358213e-06,
+      "loss": 0.746,
+      "step": 1414
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6910188223643856,
+      "learning_rate": 9.6343357596661e-06,
+      "loss": 0.6378,
+      "step": 1415
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2643901325386224,
+      "learning_rate": 9.633695737346341e-06,
+      "loss": 0.7012,
+      "step": 1416
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.741008316052005,
+      "learning_rate": 9.633055176697205e-06,
+      "loss": 0.7338,
+      "step": 1417
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.204256419036011,
+      "learning_rate": 9.632414077793111e-06,
+      "loss": 0.7241,
+      "step": 1418
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.440995779151848,
+      "learning_rate": 9.63177244070854e-06,
+      "loss": 0.6625,
+      "step": 1419
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5374105809523626,
+      "learning_rate": 9.631130265518036e-06,
+      "loss": 0.7329,
+      "step": 1420
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.46110371173565,
+      "learning_rate": 9.630487552296207e-06,
+      "loss": 0.7371,
+      "step": 1421
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3468353330936393,
+      "learning_rate": 9.629844301117717e-06,
+      "loss": 0.6405,
+      "step": 1422
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.397830492051804,
+      "learning_rate": 9.6292005120573e-06,
+      "loss": 0.6774,
+      "step": 1423
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.7532545714870977,
+      "learning_rate": 9.62855618518975e-06,
+      "loss": 0.7351,
+      "step": 1424
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1816654437320833,
+      "learning_rate": 9.627911320589922e-06,
+      "loss": 0.7497,
+      "step": 1425
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4023123862610016,
+      "learning_rate": 9.627265918332734e-06,
+      "loss": 0.7702,
+      "step": 1426
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6310701737673368,
+      "learning_rate": 9.626619978493168e-06,
+      "loss": 0.7071,
+      "step": 1427
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.44463817108983,
+      "learning_rate": 9.62597350114627e-06,
+      "loss": 0.6781,
+      "step": 1428
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4602674966430977,
+      "learning_rate": 9.625326486367139e-06,
+      "loss": 0.7439,
+      "step": 1429
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4828542109646614,
+      "learning_rate": 9.624678934230948e-06,
+      "loss": 0.7377,
+      "step": 1430
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.846123388535948,
+      "learning_rate": 9.624030844812926e-06,
+      "loss": 0.6388,
+      "step": 1431
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1757720444416084,
+      "learning_rate": 9.623382218188371e-06,
+      "loss": 0.6985,
+      "step": 1432
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.259786537034406,
+      "learning_rate": 9.622733054432631e-06,
+      "loss": 0.7484,
+      "step": 1433
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.212671951693864,
+      "learning_rate": 9.62208335362113e-06,
+      "loss": 0.6293,
+      "step": 1434
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.818588711614401,
+      "learning_rate": 9.621433115829344e-06,
+      "loss": 0.7193,
+      "step": 1435
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.029835287597727,
+      "learning_rate": 9.62078234113282e-06,
+      "loss": 0.6895,
+      "step": 1436
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1696773604444783,
+      "learning_rate": 9.62013102960716e-06,
+      "loss": 0.7032,
+      "step": 1437
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5749029914926895,
+      "learning_rate": 9.619479181328034e-06,
+      "loss": 0.6764,
+      "step": 1438
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8794399028585693,
+      "learning_rate": 9.618826796371168e-06,
+      "loss": 0.7734,
+      "step": 1439
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.516706221446295,
+      "learning_rate": 9.618173874812357e-06,
+      "loss": 0.7291,
+      "step": 1440
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1628068758073637,
+      "learning_rate": 9.617520416727456e-06,
+      "loss": 0.7082,
+      "step": 1441
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8419799887337898,
+      "learning_rate": 9.61686642219238e-06,
+      "loss": 0.7383,
+      "step": 1442
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6343772160057055,
+      "learning_rate": 9.616211891283108e-06,
+      "loss": 0.6451,
+      "step": 1443
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1705848350056636,
+      "learning_rate": 9.615556824075684e-06,
+      "loss": 0.7876,
+      "step": 1444
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.483339076402114,
+      "learning_rate": 9.61490122064621e-06,
+      "loss": 0.7049,
+      "step": 1445
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0799562658988946,
+      "learning_rate": 9.614245081070851e-06,
+      "loss": 0.7073,
+      "step": 1446
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6861424428613367,
+      "learning_rate": 9.61358840542584e-06,
+      "loss": 0.6411,
+      "step": 1447
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7191955155420042,
+      "learning_rate": 9.612931193787464e-06,
+      "loss": 0.6745,
+      "step": 1448
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.472474388518654,
+      "learning_rate": 9.612273446232075e-06,
+      "loss": 0.7655,
+      "step": 1449
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9357438196561434,
+      "learning_rate": 9.61161516283609e-06,
+      "loss": 0.6609,
+      "step": 1450
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.590924049130375,
+      "learning_rate": 9.610956343675988e-06,
+      "loss": 0.7238,
+      "step": 1451
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.879538354937373,
+      "learning_rate": 9.610296988828305e-06,
+      "loss": 0.7834,
+      "step": 1452
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1041931126773212,
+      "learning_rate": 9.609637098369646e-06,
+      "loss": 0.6807,
+      "step": 1453
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.261974070639065,
+      "learning_rate": 9.608976672376673e-06,
+      "loss": 0.7105,
+      "step": 1454
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1210663189380314,
+      "learning_rate": 9.608315710926113e-06,
+      "loss": 0.6593,
+      "step": 1455
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.050620721243887,
+      "learning_rate": 9.607654214094757e-06,
+      "loss": 0.6594,
+      "step": 1456
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7905140014141727,
+      "learning_rate": 9.606992181959451e-06,
+      "loss": 0.7085,
+      "step": 1457
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7770890627009153,
+      "learning_rate": 9.606329614597114e-06,
+      "loss": 0.7159,
+      "step": 1458
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9828864391785557,
+      "learning_rate": 9.605666512084716e-06,
+      "loss": 0.6708,
+      "step": 1459
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.266774311682029,
+      "learning_rate": 9.605002874499296e-06,
+      "loss": 0.6211,
+      "step": 1460
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 11.988026431087897,
+      "learning_rate": 9.604338701917954e-06,
+      "loss": 0.7906,
+      "step": 1461
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.5367891929196,
+      "learning_rate": 9.603673994417853e-06,
+      "loss": 0.6816,
+      "step": 1462
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3227720167759194,
+      "learning_rate": 9.603008752076213e-06,
+      "loss": 0.7428,
+      "step": 1463
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.309441828787185,
+      "learning_rate": 9.602342974970323e-06,
+      "loss": 0.675,
+      "step": 1464
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.523141442933751,
+      "learning_rate": 9.60167666317753e-06,
+      "loss": 0.6574,
+      "step": 1465
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9705719635611074,
+      "learning_rate": 9.601009816775244e-06,
+      "loss": 0.6318,
+      "step": 1466
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4004597591752566,
+      "learning_rate": 9.600342435840939e-06,
+      "loss": 0.7256,
+      "step": 1467
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9444494206781227,
+      "learning_rate": 9.599674520452148e-06,
+      "loss": 0.7762,
+      "step": 1468
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.7455953837505005,
+      "learning_rate": 9.599006070686467e-06,
+      "loss": 0.7091,
+      "step": 1469
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6074677438299947,
+      "learning_rate": 9.598337086621555e-06,
+      "loss": 0.6716,
+      "step": 1470
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9274341045897745,
+      "learning_rate": 9.597667568335133e-06,
+      "loss": 0.6385,
+      "step": 1471
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.7346926939077543,
+      "learning_rate": 9.596997515904983e-06,
+      "loss": 0.7569,
+      "step": 1472
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.5131800878247748,
+      "learning_rate": 9.59632692940895e-06,
+      "loss": 0.8693,
+      "step": 1473
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.097568621283518,
+      "learning_rate": 9.595655808924942e-06,
+      "loss": 0.7473,
+      "step": 1474
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.067535905696128,
+      "learning_rate": 9.594984154530926e-06,
+      "loss": 0.7308,
+      "step": 1475
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.508065359892658,
+      "learning_rate": 9.594311966304933e-06,
+      "loss": 0.7814,
+      "step": 1476
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2827015265359845,
+      "learning_rate": 9.593639244325057e-06,
+      "loss": 0.7721,
+      "step": 1477
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.738236724185965,
+      "learning_rate": 9.592965988669454e-06,
+      "loss": 0.7184,
+      "step": 1478
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6198830644856206,
+      "learning_rate": 9.592292199416338e-06,
+      "loss": 0.9262,
+      "step": 1479
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.552236548583453,
+      "learning_rate": 9.591617876643989e-06,
+      "loss": 0.7082,
+      "step": 1480
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.826249720921075,
+      "learning_rate": 9.590943020430747e-06,
+      "loss": 0.6825,
+      "step": 1481
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.860146961164607,
+      "learning_rate": 9.59026763085502e-06,
+      "loss": 0.6227,
+      "step": 1482
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2203871767796366,
+      "learning_rate": 9.589591707995265e-06,
+      "loss": 0.6181,
+      "step": 1483
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.401274298711083,
+      "learning_rate": 9.588915251930013e-06,
+      "loss": 0.6813,
+      "step": 1484
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.331113197722243,
+      "learning_rate": 9.588238262737853e-06,
+      "loss": 0.6847,
+      "step": 1485
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7893674560880104,
+      "learning_rate": 9.587560740497434e-06,
+      "loss": 0.6684,
+      "step": 1486
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.812079227043104,
+      "learning_rate": 9.586882685287471e-06,
+      "loss": 0.6859,
+      "step": 1487
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.690594004910145,
+      "learning_rate": 9.586204097186738e-06,
+      "loss": 0.6719,
+      "step": 1488
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.67029996471702,
+      "learning_rate": 9.585524976274068e-06,
+      "loss": 0.6752,
+      "step": 1489
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.783949918181014,
+      "learning_rate": 9.584845322628363e-06,
+      "loss": 0.6744,
+      "step": 1490
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.768161593958554,
+      "learning_rate": 9.58416513632858e-06,
+      "loss": 0.6731,
+      "step": 1491
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7128750709002785,
+      "learning_rate": 9.583484417453744e-06,
+      "loss": 0.6595,
+      "step": 1492
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1967481849379173,
+      "learning_rate": 9.582803166082938e-06,
+      "loss": 0.7063,
+      "step": 1493
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5603570119876258,
+      "learning_rate": 9.582121382295309e-06,
+      "loss": 0.7163,
+      "step": 1494
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.275503794973825,
+      "learning_rate": 9.58143906617006e-06,
+      "loss": 0.7148,
+      "step": 1495
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4819881710650984,
+      "learning_rate": 9.580756217786466e-06,
+      "loss": 0.6752,
+      "step": 1496
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.274480650090562,
+      "learning_rate": 9.580072837223857e-06,
+      "loss": 0.6004,
+      "step": 1497
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4619991807752157,
+      "learning_rate": 9.579388924561625e-06,
+      "loss": 0.7528,
+      "step": 1498
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.69869982453989,
+      "learning_rate": 9.578704479879225e-06,
+      "loss": 0.6724,
+      "step": 1499
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.587551021620836,
+      "learning_rate": 9.578019503256175e-06,
+      "loss": 0.661,
+      "step": 1500
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5984643485005297,
+      "learning_rate": 9.577333994772052e-06,
+      "loss": 0.7543,
+      "step": 1501
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.106286735099979,
+      "learning_rate": 9.576647954506498e-06,
+      "loss": 0.7503,
+      "step": 1502
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5619268778049284,
+      "learning_rate": 9.575961382539211e-06,
+      "loss": 0.7057,
+      "step": 1503
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6307425334372945,
+      "learning_rate": 9.575274278949962e-06,
+      "loss": 0.7321,
+      "step": 1504
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8345146127858487,
+      "learning_rate": 9.574586643818572e-06,
+      "loss": 0.7287,
+      "step": 1505
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.26642048267166,
+      "learning_rate": 9.57389847722493e-06,
+      "loss": 0.7251,
+      "step": 1506
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5229552614331263,
+      "learning_rate": 9.573209779248985e-06,
+      "loss": 0.7519,
+      "step": 1507
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.80455601855804,
+      "learning_rate": 9.572520549970746e-06,
+      "loss": 0.7422,
+      "step": 1508
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5636231475624394,
+      "learning_rate": 9.571830789470288e-06,
+      "loss": 0.6824,
+      "step": 1509
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.455344120878093,
+      "learning_rate": 9.571140497827746e-06,
+      "loss": 0.7862,
+      "step": 1510
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9947982231220225,
+      "learning_rate": 9.570449675123313e-06,
+      "loss": 0.8056,
+      "step": 1511
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.199158359526649,
+      "learning_rate": 9.56975832143725e-06,
+      "loss": 0.7388,
+      "step": 1512
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.481563398604672,
+      "learning_rate": 9.569066436849875e-06,
+      "loss": 0.7722,
+      "step": 1513
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7913702575178583,
+      "learning_rate": 9.568374021441567e-06,
+      "loss": 0.6772,
+      "step": 1514
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3892593790051593,
+      "learning_rate": 9.567681075292774e-06,
+      "loss": 0.7162,
+      "step": 1515
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3903154524770183,
+      "learning_rate": 9.566987598483995e-06,
+      "loss": 0.6662,
+      "step": 1516
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6872816236265624,
+      "learning_rate": 9.5662935910958e-06,
+      "loss": 0.6746,
+      "step": 1517
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1845449131190704,
+      "learning_rate": 9.565599053208815e-06,
+      "loss": 0.713,
+      "step": 1518
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.520639847963599,
+      "learning_rate": 9.564903984903731e-06,
+      "loss": 0.6745,
+      "step": 1519
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3349303925717697,
+      "learning_rate": 9.564208386261296e-06,
+      "loss": 0.6643,
+      "step": 1520
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.883067515708215,
+      "learning_rate": 9.563512257362325e-06,
+      "loss": 0.7143,
+      "step": 1521
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.403562641572306,
+      "learning_rate": 9.56281559828769e-06,
+      "loss": 0.5652,
+      "step": 1522
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.692764536743657,
+      "learning_rate": 9.562118409118334e-06,
+      "loss": 0.7098,
+      "step": 1523
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.24853739667993,
+      "learning_rate": 9.561420689935243e-06,
+      "loss": 0.7346,
+      "step": 1524
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.847945930120676,
+      "learning_rate": 9.560722440819485e-06,
+      "loss": 0.6834,
+      "step": 1525
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9793899903106524,
+      "learning_rate": 9.560023661852178e-06,
+      "loss": 0.6984,
+      "step": 1526
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.696356145427722,
+      "learning_rate": 9.559324353114503e-06,
+      "loss": 0.7272,
+      "step": 1527
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2812497097713216,
+      "learning_rate": 9.558624514687703e-06,
+      "loss": 0.761,
+      "step": 1528
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.271576386159374,
+      "learning_rate": 9.557924146653087e-06,
+      "loss": 0.7657,
+      "step": 1529
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7171270620464365,
+      "learning_rate": 9.557223249092017e-06,
+      "loss": 0.7263,
+      "step": 1530
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8535792247554834,
+      "learning_rate": 9.556521822085924e-06,
+      "loss": 0.7297,
+      "step": 1531
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4140439514306706,
+      "learning_rate": 9.555819865716298e-06,
+      "loss": 0.6768,
+      "step": 1532
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9490982350238957,
+      "learning_rate": 9.555117380064689e-06,
+      "loss": 0.698,
+      "step": 1533
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.49780482074063,
+      "learning_rate": 9.55441436521271e-06,
+      "loss": 0.6453,
+      "step": 1534
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.075811387978172,
+      "learning_rate": 9.553710821242036e-06,
+      "loss": 0.6666,
+      "step": 1535
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0553855565019585,
+      "learning_rate": 9.553006748234402e-06,
+      "loss": 0.7491,
+      "step": 1536
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.1401242046590845,
+      "learning_rate": 9.552302146271606e-06,
+      "loss": 0.7782,
+      "step": 1537
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.42939226201379,
+      "learning_rate": 9.551597015435505e-06,
+      "loss": 0.7143,
+      "step": 1538
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.82555463707085,
+      "learning_rate": 9.550891355808022e-06,
+      "loss": 0.7285,
+      "step": 1539
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6022856991586036,
+      "learning_rate": 9.550185167471134e-06,
+      "loss": 0.6323,
+      "step": 1540
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.716216112330096,
+      "learning_rate": 9.549478450506888e-06,
+      "loss": 0.6955,
+      "step": 1541
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9288414453383163,
+      "learning_rate": 9.548771204997387e-06,
+      "loss": 0.6652,
+      "step": 1542
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9260455534051664,
+      "learning_rate": 9.548063431024797e-06,
+      "loss": 0.6677,
+      "step": 1543
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6013304107381066,
+      "learning_rate": 9.547355128671343e-06,
+      "loss": 0.6632,
+      "step": 1544
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7326280366062523,
+      "learning_rate": 9.546646298019315e-06,
+      "loss": 0.7204,
+      "step": 1545
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1529382962472567,
+      "learning_rate": 9.545936939151066e-06,
+      "loss": 0.7363,
+      "step": 1546
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.48827212379924,
+      "learning_rate": 9.545227052149002e-06,
+      "loss": 0.761,
+      "step": 1547
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6924419534835637,
+      "learning_rate": 9.544516637095598e-06,
+      "loss": 0.7262,
+      "step": 1548
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4210226974371554,
+      "learning_rate": 9.54380569407339e-06,
+      "loss": 0.6652,
+      "step": 1549
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.91586157827423,
+      "learning_rate": 9.543094223164967e-06,
+      "loss": 0.6746,
+      "step": 1550
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8878136097168228,
+      "learning_rate": 9.542382224452995e-06,
+      "loss": 0.6712,
+      "step": 1551
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.485973778059242,
+      "learning_rate": 9.541669698020184e-06,
+      "loss": 0.7157,
+      "step": 1552
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6385514951897298,
+      "learning_rate": 9.540956643949317e-06,
+      "loss": 0.7383,
+      "step": 1553
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2286222175015378,
+      "learning_rate": 9.540243062323234e-06,
+      "loss": 0.6431,
+      "step": 1554
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6255728686862043,
+      "learning_rate": 9.539528953224835e-06,
+      "loss": 0.6488,
+      "step": 1555
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.294558982143112,
+      "learning_rate": 9.538814316737085e-06,
+      "loss": 0.7291,
+      "step": 1556
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7649291179617075,
+      "learning_rate": 9.53809915294301e-06,
+      "loss": 0.7938,
+      "step": 1557
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6080959910423047,
+      "learning_rate": 9.53738346192569e-06,
+      "loss": 0.6932,
+      "step": 1558
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0121425622421936,
+      "learning_rate": 9.536667243768279e-06,
+      "loss": 0.689,
+      "step": 1559
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.332705210852358,
+      "learning_rate": 9.53595049855398e-06,
+      "loss": 0.7032,
+      "step": 1560
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5824358159901704,
+      "learning_rate": 9.535233226366067e-06,
+      "loss": 0.6877,
+      "step": 1561
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7223244944225486,
+      "learning_rate": 9.534515427287865e-06,
+      "loss": 0.6517,
+      "step": 1562
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0686395234828074,
+      "learning_rate": 9.53379710140277e-06,
+      "loss": 0.7079,
+      "step": 1563
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0167637777660223,
+      "learning_rate": 9.533078248794232e-06,
+      "loss": 0.6687,
+      "step": 1564
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.432973784171849,
+      "learning_rate": 9.532358869545767e-06,
+      "loss": 0.7802,
+      "step": 1565
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2926310921200885,
+      "learning_rate": 9.531638963740953e-06,
+      "loss": 0.7231,
+      "step": 1566
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6393048434940276,
+      "learning_rate": 9.530918531463423e-06,
+      "loss": 0.6623,
+      "step": 1567
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.481069201927267,
+      "learning_rate": 9.530197572796873e-06,
+      "loss": 0.735,
+      "step": 1568
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6013622318011276,
+      "learning_rate": 9.529476087825067e-06,
+      "loss": 0.6855,
+      "step": 1569
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4249010091697434,
+      "learning_rate": 9.528754076631821e-06,
+      "loss": 0.7398,
+      "step": 1570
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2070633925039465,
+      "learning_rate": 9.528031539301016e-06,
+      "loss": 0.7574,
+      "step": 1571
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.555841750294239,
+      "learning_rate": 9.5273084759166e-06,
+      "loss": 0.6432,
+      "step": 1572
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.457477490220107,
+      "learning_rate": 9.526584886562571e-06,
+      "loss": 0.6151,
+      "step": 1573
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.048705570425487,
+      "learning_rate": 9.525860771322995e-06,
+      "loss": 0.7126,
+      "step": 1574
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.438910164173861,
+      "learning_rate": 9.525136130281995e-06,
+      "loss": 0.79,
+      "step": 1575
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.28221697375792,
+      "learning_rate": 9.524410963523763e-06,
+      "loss": 0.7654,
+      "step": 1576
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0467966030740645,
+      "learning_rate": 9.523685271132543e-06,
+      "loss": 0.7272,
+      "step": 1577
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3181479200807398,
+      "learning_rate": 9.522959053192646e-06,
+      "loss": 0.7779,
+      "step": 1578
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.571771113992117,
+      "learning_rate": 9.522232309788439e-06,
+      "loss": 0.7544,
+      "step": 1579
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1183269342095508,
+      "learning_rate": 9.521505041004356e-06,
+      "loss": 0.7422,
+      "step": 1580
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.416853737352066,
+      "learning_rate": 9.520777246924887e-06,
+      "loss": 0.7424,
+      "step": 1581
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.885932690234875,
+      "learning_rate": 9.520048927634587e-06,
+      "loss": 0.6748,
+      "step": 1582
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.615546636686311,
+      "learning_rate": 9.519320083218067e-06,
+      "loss": 0.7059,
+      "step": 1583
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.755154151212538,
+      "learning_rate": 9.518590713760004e-06,
+      "loss": 0.7826,
+      "step": 1584
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.132180252839951,
+      "learning_rate": 9.517860819345136e-06,
+      "loss": 0.6665,
+      "step": 1585
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.255957083552413,
+      "learning_rate": 9.517130400058255e-06,
+      "loss": 0.7323,
+      "step": 1586
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2483764789100205,
+      "learning_rate": 9.516399455984222e-06,
+      "loss": 0.7601,
+      "step": 1587
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.352631648887012,
+      "learning_rate": 9.515667987207958e-06,
+      "loss": 0.7785,
+      "step": 1588
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0869685444798676,
+      "learning_rate": 9.514935993814438e-06,
+      "loss": 0.6793,
+      "step": 1589
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3425785545443816,
+      "learning_rate": 9.514203475888706e-06,
+      "loss": 0.781,
+      "step": 1590
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.86315297570834,
+      "learning_rate": 9.513470433515866e-06,
+      "loss": 0.6503,
+      "step": 1591
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1652116829667896,
+      "learning_rate": 9.512736866781076e-06,
+      "loss": 0.681,
+      "step": 1592
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.766273085621574,
+      "learning_rate": 9.512002775769562e-06,
+      "loss": 0.7644,
+      "step": 1593
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2909986490602074,
+      "learning_rate": 9.511268160566609e-06,
+      "loss": 0.6739,
+      "step": 1594
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.075660637359557,
+      "learning_rate": 9.510533021257562e-06,
+      "loss": 0.6721,
+      "step": 1595
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.21371811437727,
+      "learning_rate": 9.509797357927826e-06,
+      "loss": 0.6669,
+      "step": 1596
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2650689770495087,
+      "learning_rate": 9.50906117066287e-06,
+      "loss": 0.6818,
+      "step": 1597
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.894316490655317,
+      "learning_rate": 9.508324459548221e-06,
+      "loss": 0.6128,
+      "step": 1598
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.363686348557055,
+      "learning_rate": 9.50758722466947e-06,
+      "loss": 0.6766,
+      "step": 1599
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1204545096681504,
+      "learning_rate": 9.506849466112264e-06,
+      "loss": 0.7487,
+      "step": 1600
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.445721487013043,
+      "learning_rate": 9.506111183962316e-06,
+      "loss": 0.66,
+      "step": 1601
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9704281101668144,
+      "learning_rate": 9.505372378305398e-06,
+      "loss": 0.6566,
+      "step": 1602
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.032858460518241,
+      "learning_rate": 9.504633049227338e-06,
+      "loss": 0.7025,
+      "step": 1603
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3197467695242198,
+      "learning_rate": 9.503893196814034e-06,
+      "loss": 0.6593,
+      "step": 1604
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.003637291105467,
+      "learning_rate": 9.503152821151435e-06,
+      "loss": 0.7282,
+      "step": 1605
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.974823409792684,
+      "learning_rate": 9.502411922325561e-06,
+      "loss": 0.644,
+      "step": 1606
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3999354547850578,
+      "learning_rate": 9.501670500422483e-06,
+      "loss": 0.7695,
+      "step": 1607
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.142048720747408,
+      "learning_rate": 9.500928555528341e-06,
+      "loss": 0.697,
+      "step": 1608
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1918483417994215,
+      "learning_rate": 9.500186087729331e-06,
+      "loss": 0.6865,
+      "step": 1609
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3130950683881673,
+      "learning_rate": 9.49944309711171e-06,
+      "loss": 0.6882,
+      "step": 1610
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2722152426733166,
+      "learning_rate": 9.498699583761795e-06,
+      "loss": 0.7057,
+      "step": 1611
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3067065149092674,
+      "learning_rate": 9.497955547765966e-06,
+      "loss": 0.6129,
+      "step": 1612
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.000454383678098,
+      "learning_rate": 9.497210989210665e-06,
+      "loss": 0.709,
+      "step": 1613
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6049845885606016,
+      "learning_rate": 9.49646590818239e-06,
+      "loss": 0.7035,
+      "step": 1614
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.5829176336530795,
+      "learning_rate": 9.495720304767705e-06,
+      "loss": 0.6982,
+      "step": 1615
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8362441150996003,
+      "learning_rate": 9.494974179053233e-06,
+      "loss": 0.7086,
+      "step": 1616
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.6175372207495045,
+      "learning_rate": 9.494227531125652e-06,
+      "loss": 0.7028,
+      "step": 1617
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.522650359801413,
+      "learning_rate": 9.493480361071707e-06,
+      "loss": 0.7281,
+      "step": 1618
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4815378042784713,
+      "learning_rate": 9.492732668978205e-06,
+      "loss": 0.6899,
+      "step": 1619
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.439653241146134,
+      "learning_rate": 9.491984454932009e-06,
+      "loss": 0.6359,
+      "step": 1620
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.322479399522989,
+      "learning_rate": 9.491235719020042e-06,
+      "loss": 0.6365,
+      "step": 1621
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8722680257735367,
+      "learning_rate": 9.490486461329293e-06,
+      "loss": 0.7157,
+      "step": 1622
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8325212931006036,
+      "learning_rate": 9.489736681946809e-06,
+      "loss": 0.7014,
+      "step": 1623
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8149300073276424,
+      "learning_rate": 9.488986380959694e-06,
+      "loss": 0.7507,
+      "step": 1624
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.399930194308783,
+      "learning_rate": 9.488235558455118e-06,
+      "loss": 0.7731,
+      "step": 1625
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4240902602318832,
+      "learning_rate": 9.487484214520308e-06,
+      "loss": 0.6616,
+      "step": 1626
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3880823758905465,
+      "learning_rate": 9.486732349242556e-06,
+      "loss": 0.6865,
+      "step": 1627
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.434462113134042,
+      "learning_rate": 9.485979962709209e-06,
+      "loss": 0.7546,
+      "step": 1628
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.140235982386299,
+      "learning_rate": 9.485227055007676e-06,
+      "loss": 0.6478,
+      "step": 1629
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6979154962012895,
+      "learning_rate": 9.48447362622543e-06,
+      "loss": 0.7336,
+      "step": 1630
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.564891032405616,
+      "learning_rate": 9.483719676450003e-06,
+      "loss": 0.6995,
+      "step": 1631
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.186995520230965,
+      "learning_rate": 9.482965205768983e-06,
+      "loss": 0.7284,
+      "step": 1632
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.93274944592198,
+      "learning_rate": 9.482210214270026e-06,
+      "loss": 0.7065,
+      "step": 1633
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3964166552732427,
+      "learning_rate": 9.481454702040842e-06,
+      "loss": 0.7386,
+      "step": 1634
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.290346577796857,
+      "learning_rate": 9.480698669169207e-06,
+      "loss": 0.6791,
+      "step": 1635
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.335965960819836,
+      "learning_rate": 9.479942115742951e-06,
+      "loss": 0.6365,
+      "step": 1636
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9605346112705178,
+      "learning_rate": 9.47918504184997e-06,
+      "loss": 0.6899,
+      "step": 1637
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.7418810164216705,
+      "learning_rate": 9.47842744757822e-06,
+      "loss": 0.6344,
+      "step": 1638
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2563707246896096,
+      "learning_rate": 9.477669333015714e-06,
+      "loss": 0.5757,
+      "step": 1639
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.013900658553373,
+      "learning_rate": 9.476910698250529e-06,
+      "loss": 0.7077,
+      "step": 1640
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.396575665358366,
+      "learning_rate": 9.4761515433708e-06,
+      "loss": 0.7228,
+      "step": 1641
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.40289964027477,
+      "learning_rate": 9.475391868464725e-06,
+      "loss": 0.7134,
+      "step": 1642
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.387400572157838,
+      "learning_rate": 9.474631673620558e-06,
+      "loss": 0.7249,
+      "step": 1643
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.470393309910611,
+      "learning_rate": 9.47387095892662e-06,
+      "loss": 0.7756,
+      "step": 1644
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.318559436180576,
+      "learning_rate": 9.473109724471287e-06,
+      "loss": 0.6502,
+      "step": 1645
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5401989407597005,
+      "learning_rate": 9.472347970342995e-06,
+      "loss": 0.6817,
+      "step": 1646
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2594402322006175,
+      "learning_rate": 9.471585696630245e-06,
+      "loss": 0.7104,
+      "step": 1647
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1769351591530914,
+      "learning_rate": 9.470822903421595e-06,
+      "loss": 0.6472,
+      "step": 1648
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.540121387003377,
+      "learning_rate": 9.470059590805663e-06,
+      "loss": 0.715,
+      "step": 1649
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3993005705838404,
+      "learning_rate": 9.46929575887113e-06,
+      "loss": 0.7426,
+      "step": 1650
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4076747230692574,
+      "learning_rate": 9.468531407706733e-06,
+      "loss": 0.6518,
+      "step": 1651
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3345495527645443,
+      "learning_rate": 9.467766537401278e-06,
+      "loss": 0.6538,
+      "step": 1652
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2658417056945064,
+      "learning_rate": 9.46700114804362e-06,
+      "loss": 0.6715,
+      "step": 1653
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2719118199932598,
+      "learning_rate": 9.46623523972268e-06,
+      "loss": 0.6762,
+      "step": 1654
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0582788277872037,
+      "learning_rate": 9.465468812527443e-06,
+      "loss": 0.629,
+      "step": 1655
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.247235978920937,
+      "learning_rate": 9.464701866546945e-06,
+      "loss": 0.7408,
+      "step": 1656
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.155468792797194,
+      "learning_rate": 9.463934401870292e-06,
+      "loss": 0.613,
+      "step": 1657
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.544248700229081,
+      "learning_rate": 9.463166418586645e-06,
+      "loss": 0.7269,
+      "step": 1658
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3313482073212315,
+      "learning_rate": 9.462397916785222e-06,
+      "loss": 0.7243,
+      "step": 1659
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.82456167117737,
+      "learning_rate": 9.461628896555312e-06,
+      "loss": 0.7121,
+      "step": 1660
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.726685554503556,
+      "learning_rate": 9.460859357986251e-06,
+      "loss": 0.6316,
+      "step": 1661
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.542875855778497,
+      "learning_rate": 9.460089301167448e-06,
+      "loss": 0.6625,
+      "step": 1662
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.86229240642419,
+      "learning_rate": 9.45931872618836e-06,
+      "loss": 0.6818,
+      "step": 1663
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.656197150207348,
+      "learning_rate": 9.458547633138515e-06,
+      "loss": 0.6801,
+      "step": 1664
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.243631283233658,
+      "learning_rate": 9.457776022107494e-06,
+      "loss": 0.6679,
+      "step": 1665
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5293787058884636,
+      "learning_rate": 9.45700389318494e-06,
+      "loss": 0.7386,
+      "step": 1666
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.310070540436612,
+      "learning_rate": 9.456231246460557e-06,
+      "loss": 0.774,
+      "step": 1667
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.508606440342495,
+      "learning_rate": 9.455458082024112e-06,
+      "loss": 0.7392,
+      "step": 1668
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2920775313114405,
+      "learning_rate": 9.454684399965423e-06,
+      "loss": 0.758,
+      "step": 1669
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.041893234843545,
+      "learning_rate": 9.453910200374382e-06,
+      "loss": 0.6369,
+      "step": 1670
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2931690055491833,
+      "learning_rate": 9.453135483340925e-06,
+      "loss": 0.6547,
+      "step": 1671
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3115301263653227,
+      "learning_rate": 9.452360248955062e-06,
+      "loss": 0.6449,
+      "step": 1672
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9637645722219323,
+      "learning_rate": 9.451584497306856e-06,
+      "loss": 0.6999,
+      "step": 1673
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3068845068680326,
+      "learning_rate": 9.45080822848643e-06,
+      "loss": 0.6948,
+      "step": 1674
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2231756130525673,
+      "learning_rate": 9.45003144258397e-06,
+      "loss": 0.6791,
+      "step": 1675
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.320104844855623,
+      "learning_rate": 9.449254139689721e-06,
+      "loss": 0.6537,
+      "step": 1676
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.471206074285933,
+      "learning_rate": 9.448476319893989e-06,
+      "loss": 0.7204,
+      "step": 1677
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4573874653738934,
+      "learning_rate": 9.447697983287136e-06,
+      "loss": 0.6883,
+      "step": 1678
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2768018435285424,
+      "learning_rate": 9.446919129959589e-06,
+      "loss": 0.6646,
+      "step": 1679
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5304009939179433,
+      "learning_rate": 9.44613976000183e-06,
+      "loss": 0.6923,
+      "step": 1680
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3183304739700534,
+      "learning_rate": 9.44535987350441e-06,
+      "loss": 0.82,
+      "step": 1681
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.293573173274393,
+      "learning_rate": 9.44457947055793e-06,
+      "loss": 0.7597,
+      "step": 1682
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.820715534237888,
+      "learning_rate": 9.443798551253052e-06,
+      "loss": 0.7216,
+      "step": 1683
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.122577759545682,
+      "learning_rate": 9.443017115680503e-06,
+      "loss": 0.7303,
+      "step": 1684
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.653511448510485,
+      "learning_rate": 9.442235163931072e-06,
+      "loss": 0.6314,
+      "step": 1685
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.186415024401008,
+      "learning_rate": 9.441452696095601e-06,
+      "loss": 0.755,
+      "step": 1686
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3274053893757563,
+      "learning_rate": 9.440669712264994e-06,
+      "loss": 0.7157,
+      "step": 1687
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9206038169379611,
+      "learning_rate": 9.439886212530217e-06,
+      "loss": 0.7278,
+      "step": 1688
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.974321572145424,
+      "learning_rate": 9.439102196982292e-06,
+      "loss": 0.6442,
+      "step": 1689
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9034939575257075,
+      "learning_rate": 9.438317665712308e-06,
+      "loss": 0.7244,
+      "step": 1690
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5630336170505945,
+      "learning_rate": 9.437532618811407e-06,
+      "loss": 0.7238,
+      "step": 1691
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.7600588498607768,
+      "learning_rate": 9.436747056370794e-06,
+      "loss": 0.652,
+      "step": 1692
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4504076582596195,
+      "learning_rate": 9.435960978481734e-06,
+      "loss": 0.7314,
+      "step": 1693
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.51073812241736,
+      "learning_rate": 9.435174385235548e-06,
+      "loss": 0.7227,
+      "step": 1694
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.846356911208405,
+      "learning_rate": 9.434387276723624e-06,
+      "loss": 0.6857,
+      "step": 1695
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.703615406650167,
+      "learning_rate": 9.433599653037406e-06,
+      "loss": 0.5352,
+      "step": 1696
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3457312870011955,
+      "learning_rate": 9.432811514268396e-06,
+      "loss": 0.7758,
+      "step": 1697
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.271626563468226,
+      "learning_rate": 9.432022860508158e-06,
+      "loss": 0.7422,
+      "step": 1698
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3311293386557175,
+      "learning_rate": 9.431233691848316e-06,
+      "loss": 0.6922,
+      "step": 1699
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.13275822233704,
+      "learning_rate": 9.430444008380553e-06,
+      "loss": 0.6746,
+      "step": 1700
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3910164531032594,
+      "learning_rate": 9.429653810196611e-06,
+      "loss": 0.7156,
+      "step": 1701
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.25390169947036,
+      "learning_rate": 9.428863097388295e-06,
+      "loss": 0.7435,
+      "step": 1702
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9560405228811595,
+      "learning_rate": 9.428071870047469e-06,
+      "loss": 0.7,
+      "step": 1703
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8824075331234948,
+      "learning_rate": 9.427280128266049e-06,
+      "loss": 0.7596,
+      "step": 1704
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.286053461352808,
+      "learning_rate": 9.426487872136025e-06,
+      "loss": 0.7239,
+      "step": 1705
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.008603418544589,
+      "learning_rate": 9.425695101749435e-06,
+      "loss": 0.7413,
+      "step": 1706
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.10028582452023,
+      "learning_rate": 9.424901817198381e-06,
+      "loss": 0.6829,
+      "step": 1707
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1535649860037007,
+      "learning_rate": 9.424108018575026e-06,
+      "loss": 0.7089,
+      "step": 1708
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8660284485190903,
+      "learning_rate": 9.42331370597159e-06,
+      "loss": 0.7544,
+      "step": 1709
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5095078891354508,
+      "learning_rate": 9.422518879480353e-06,
+      "loss": 0.6405,
+      "step": 1710
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1188014421663874,
+      "learning_rate": 9.421723539193657e-06,
+      "loss": 0.6818,
+      "step": 1711
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.163280526338815,
+      "learning_rate": 9.420927685203901e-06,
+      "loss": 0.6727,
+      "step": 1712
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7994284590304366,
+      "learning_rate": 9.42013131760355e-06,
+      "loss": 0.6459,
+      "step": 1713
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.098234987273422,
+      "learning_rate": 9.419334436485117e-06,
+      "loss": 0.7406,
+      "step": 1714
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.204421750962229,
+      "learning_rate": 9.418537041941185e-06,
+      "loss": 0.75,
+      "step": 1715
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.551866812050452,
+      "learning_rate": 9.417739134064392e-06,
+      "loss": 0.7352,
+      "step": 1716
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1052536649230147,
+      "learning_rate": 9.416940712947436e-06,
+      "loss": 0.7457,
+      "step": 1717
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.552123073690243,
+      "learning_rate": 9.416141778683077e-06,
+      "loss": 0.7008,
+      "step": 1718
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.386726881220311,
+      "learning_rate": 9.415342331364132e-06,
+      "loss": 0.6845,
+      "step": 1719
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0231356594407064,
+      "learning_rate": 9.414542371083477e-06,
+      "loss": 0.6269,
+      "step": 1720
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3944077369294376,
+      "learning_rate": 9.413741897934052e-06,
+      "loss": 0.7178,
+      "step": 1721
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4163679275832997,
+      "learning_rate": 9.412940912008852e-06,
+      "loss": 0.6554,
+      "step": 1722
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1677504987549554,
+      "learning_rate": 9.412139413400933e-06,
+      "loss": 0.7565,
+      "step": 1723
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.628059217990637,
+      "learning_rate": 9.41133740220341e-06,
+      "loss": 0.6513,
+      "step": 1724
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2085858844991972,
+      "learning_rate": 9.410534878509461e-06,
+      "loss": 0.6636,
+      "step": 1725
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9079694945157286,
+      "learning_rate": 9.40973184241232e-06,
+      "loss": 0.6428,
+      "step": 1726
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2280940372738156,
+      "learning_rate": 9.408928294005279e-06,
+      "loss": 0.6991,
+      "step": 1727
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2464560200093553,
+      "learning_rate": 9.408124233381695e-06,
+      "loss": 0.7057,
+      "step": 1728
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.41547778019205,
+      "learning_rate": 9.40731966063498e-06,
+      "loss": 0.7175,
+      "step": 1729
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9493683126801489,
+      "learning_rate": 9.406514575858606e-06,
+      "loss": 0.6848,
+      "step": 1730
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.653450440842846,
+      "learning_rate": 9.405708979146106e-06,
+      "loss": 0.7155,
+      "step": 1731
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2614446312235077,
+      "learning_rate": 9.404902870591076e-06,
+      "loss": 0.7237,
+      "step": 1732
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.236025418767225,
+      "learning_rate": 9.40409625028716e-06,
+      "loss": 0.7383,
+      "step": 1733
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0068155617046903,
+      "learning_rate": 9.403289118328074e-06,
+      "loss": 0.624,
+      "step": 1734
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.586872590910335,
+      "learning_rate": 9.402481474807588e-06,
+      "loss": 0.6978,
+      "step": 1735
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.05535830543558,
+      "learning_rate": 9.401673319819529e-06,
+      "loss": 0.6669,
+      "step": 1736
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8884455652664238,
+      "learning_rate": 9.400864653457789e-06,
+      "loss": 0.697,
+      "step": 1737
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1851392337369266,
+      "learning_rate": 9.400055475816313e-06,
+      "loss": 0.6524,
+      "step": 1738
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.576173165092392,
+      "learning_rate": 9.399245786989112e-06,
+      "loss": 0.7119,
+      "step": 1739
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.620809902632731,
+      "learning_rate": 9.398435587070254e-06,
+      "loss": 0.7611,
+      "step": 1740
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8033777586962114,
+      "learning_rate": 9.397624876153862e-06,
+      "loss": 0.6888,
+      "step": 1741
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1952968707428666,
+      "learning_rate": 9.396813654334124e-06,
+      "loss": 0.729,
+      "step": 1742
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.184788154360516,
+      "learning_rate": 9.396001921705287e-06,
+      "loss": 0.7522,
+      "step": 1743
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5561173456941186,
+      "learning_rate": 9.395189678361655e-06,
+      "loss": 0.6819,
+      "step": 1744
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9560468612635016,
+      "learning_rate": 9.39437692439759e-06,
+      "loss": 0.7499,
+      "step": 1745
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.170377346969926,
+      "learning_rate": 9.393563659907516e-06,
+      "loss": 0.7307,
+      "step": 1746
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0529367396092386,
+      "learning_rate": 9.392749884985918e-06,
+      "loss": 0.6618,
+      "step": 1747
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5770411994159867,
+      "learning_rate": 9.391935599727336e-06,
+      "loss": 0.702,
+      "step": 1748
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1159695596217305,
+      "learning_rate": 9.391120804226372e-06,
+      "loss": 0.7339,
+      "step": 1749
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9557470831906363,
+      "learning_rate": 9.390305498577685e-06,
+      "loss": 0.829,
+      "step": 1750
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.066429118959667,
+      "learning_rate": 9.389489682875999e-06,
+      "loss": 0.7344,
+      "step": 1751
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3728866613105315,
+      "learning_rate": 9.388673357216088e-06,
+      "loss": 0.6821,
+      "step": 1752
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5738421703420835,
+      "learning_rate": 9.387856521692795e-06,
+      "loss": 0.6881,
+      "step": 1753
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2173777484723933,
+      "learning_rate": 9.387039176401013e-06,
+      "loss": 0.6497,
+      "step": 1754
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8326929223772916,
+      "learning_rate": 9.386221321435702e-06,
+      "loss": 0.7752,
+      "step": 1755
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.342472475816738,
+      "learning_rate": 9.385402956891878e-06,
+      "loss": 0.6642,
+      "step": 1756
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8421908380492378,
+      "learning_rate": 9.384584082864614e-06,
+      "loss": 0.628,
+      "step": 1757
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.293083670289089,
+      "learning_rate": 9.383764699449047e-06,
+      "loss": 0.7642,
+      "step": 1758
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1881257410006234,
+      "learning_rate": 9.382944806740369e-06,
+      "loss": 0.697,
+      "step": 1759
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0454383098040125,
+      "learning_rate": 9.382124404833832e-06,
+      "loss": 0.6683,
+      "step": 1760
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.39746694551876,
+      "learning_rate": 9.38130349382475e-06,
+      "loss": 0.7014,
+      "step": 1761
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.646324899559431,
+      "learning_rate": 9.380482073808493e-06,
+      "loss": 0.6275,
+      "step": 1762
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.31148966058067,
+      "learning_rate": 9.379660144880491e-06,
+      "loss": 0.6855,
+      "step": 1763
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6610941993286072,
+      "learning_rate": 9.378837707136235e-06,
+      "loss": 0.721,
+      "step": 1764
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.181626035362435,
+      "learning_rate": 9.37801476067127e-06,
+      "loss": 0.7382,
+      "step": 1765
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5121904009240774,
+      "learning_rate": 9.377191305581208e-06,
+      "loss": 0.6953,
+      "step": 1766
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.324736532829736,
+      "learning_rate": 9.376367341961712e-06,
+      "loss": 0.6536,
+      "step": 1767
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1731590595672206,
+      "learning_rate": 9.375542869908509e-06,
+      "loss": 0.6795,
+      "step": 1768
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0441280442692396,
+      "learning_rate": 9.374717889517384e-06,
+      "loss": 0.6707,
+      "step": 1769
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.188945519966826,
+      "learning_rate": 9.373892400884182e-06,
+      "loss": 0.6427,
+      "step": 1770
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8292074188975915,
+      "learning_rate": 9.373066404104803e-06,
+      "loss": 0.6416,
+      "step": 1771
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2353048574201884,
+      "learning_rate": 9.37223989927521e-06,
+      "loss": 0.7069,
+      "step": 1772
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.313904910526145,
+      "learning_rate": 9.371412886491424e-06,
+      "loss": 0.7505,
+      "step": 1773
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.177329261346893,
+      "learning_rate": 9.370585365849527e-06,
+      "loss": 0.6287,
+      "step": 1774
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.509687395524509,
+      "learning_rate": 9.369757337445655e-06,
+      "loss": 0.6938,
+      "step": 1775
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.50959271620824,
+      "learning_rate": 9.368928801376009e-06,
+      "loss": 0.7594,
+      "step": 1776
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3425929238257943,
+      "learning_rate": 9.368099757736843e-06,
+      "loss": 0.6348,
+      "step": 1777
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.571462951149562,
+      "learning_rate": 9.367270206624474e-06,
+      "loss": 0.7839,
+      "step": 1778
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.288592181156132,
+      "learning_rate": 9.366440148135276e-06,
+      "loss": 0.7086,
+      "step": 1779
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.7452409081529328,
+      "learning_rate": 9.365609582365685e-06,
+      "loss": 0.7787,
+      "step": 1780
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.188224299368682,
+      "learning_rate": 9.364778509412191e-06,
+      "loss": 0.7622,
+      "step": 1781
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.7450685765521223,
+      "learning_rate": 9.363946929371349e-06,
+      "loss": 0.6863,
+      "step": 1782
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5045275357571284,
+      "learning_rate": 9.363114842339767e-06,
+      "loss": 0.7075,
+      "step": 1783
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0164414465737837,
+      "learning_rate": 9.362282248414114e-06,
+      "loss": 0.6598,
+      "step": 1784
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.35492294691701,
+      "learning_rate": 9.361449147691122e-06,
+      "loss": 0.7235,
+      "step": 1785
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2619503151970335,
+      "learning_rate": 9.360615540267572e-06,
+      "loss": 0.6997,
+      "step": 1786
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5899599164267766,
+      "learning_rate": 9.359781426240316e-06,
+      "loss": 0.7257,
+      "step": 1787
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5634208857345366,
+      "learning_rate": 9.358946805706257e-06,
+      "loss": 0.664,
+      "step": 1788
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.51798972826991,
+      "learning_rate": 9.358111678762359e-06,
+      "loss": 0.6728,
+      "step": 1789
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3326773791425466,
+      "learning_rate": 9.357276045505643e-06,
+      "loss": 0.7358,
+      "step": 1790
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2350420427602833,
+      "learning_rate": 9.35643990603319e-06,
+      "loss": 0.6856,
+      "step": 1791
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.608868685684469,
+      "learning_rate": 9.355603260442145e-06,
+      "loss": 0.6177,
+      "step": 1792
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9713835504144617,
+      "learning_rate": 9.354766108829703e-06,
+      "loss": 0.7943,
+      "step": 1793
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8062689555008857,
+      "learning_rate": 9.353928451293122e-06,
+      "loss": 0.6725,
+      "step": 1794
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.671272851573433,
+      "learning_rate": 9.35309028792972e-06,
+      "loss": 0.6848,
+      "step": 1795
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.066156049708186,
+      "learning_rate": 9.352251618836872e-06,
+      "loss": 0.7521,
+      "step": 1796
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0657191943549273,
+      "learning_rate": 9.351412444112013e-06,
+      "loss": 0.6063,
+      "step": 1797
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.647445442676486,
+      "learning_rate": 9.350572763852633e-06,
+      "loss": 0.6984,
+      "step": 1798
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2824264713651994,
+      "learning_rate": 9.349732578156286e-06,
+      "loss": 0.6747,
+      "step": 1799
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.232278265988478,
+      "learning_rate": 9.348891887120582e-06,
+      "loss": 0.7481,
+      "step": 1800
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2260356050576937,
+      "learning_rate": 9.348050690843192e-06,
+      "loss": 0.7101,
+      "step": 1801
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6245692650813734,
+      "learning_rate": 9.347208989421838e-06,
+      "loss": 0.6776,
+      "step": 1802
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8932084963461766,
+      "learning_rate": 9.346366782954313e-06,
+      "loss": 0.7145,
+      "step": 1803
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6293013765675806,
+      "learning_rate": 9.345524071538457e-06,
+      "loss": 0.762,
+      "step": 1804
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4909478746794926,
+      "learning_rate": 9.344680855272178e-06,
+      "loss": 0.7797,
+      "step": 1805
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.678911462141494,
+      "learning_rate": 9.343837134253434e-06,
+      "loss": 0.7017,
+      "step": 1806
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.236697329924541,
+      "learning_rate": 9.342992908580252e-06,
+      "loss": 0.6629,
+      "step": 1807
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6302188588519484,
+      "learning_rate": 9.342148178350705e-06,
+      "loss": 0.7272,
+      "step": 1808
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3111476074297292,
+      "learning_rate": 9.341302943662937e-06,
+      "loss": 0.6723,
+      "step": 1809
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9282424193071446,
+      "learning_rate": 9.34045720461514e-06,
+      "loss": 0.8171,
+      "step": 1810
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9521986711914656,
+      "learning_rate": 9.339610961305575e-06,
+      "loss": 0.7266,
+      "step": 1811
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6008867763760595,
+      "learning_rate": 9.33876421383255e-06,
+      "loss": 0.6991,
+      "step": 1812
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4219540920578213,
+      "learning_rate": 9.337916962294443e-06,
+      "loss": 0.7751,
+      "step": 1813
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3994526596884787,
+      "learning_rate": 9.337069206789681e-06,
+      "loss": 0.7132,
+      "step": 1814
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.141697670450548,
+      "learning_rate": 9.336220947416757e-06,
+      "loss": 0.6879,
+      "step": 1815
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.442700420921565,
+      "learning_rate": 9.335372184274219e-06,
+      "loss": 0.725,
+      "step": 1816
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9922843786433275,
+      "learning_rate": 9.334522917460671e-06,
+      "loss": 0.7077,
+      "step": 1817
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9847132878411518,
+      "learning_rate": 9.33367314707478e-06,
+      "loss": 0.6625,
+      "step": 1818
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2682583137130137,
+      "learning_rate": 9.332822873215273e-06,
+      "loss": 0.6533,
+      "step": 1819
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1574351053442125,
+      "learning_rate": 9.331972095980927e-06,
+      "loss": 0.6908,
+      "step": 1820
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.495175865873836,
+      "learning_rate": 9.331120815470586e-06,
+      "loss": 0.69,
+      "step": 1821
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5556685920907243,
+      "learning_rate": 9.330269031783147e-06,
+      "loss": 0.6904,
+      "step": 1822
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.50380795052326,
+      "learning_rate": 9.329416745017573e-06,
+      "loss": 0.6958,
+      "step": 1823
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.195310629239766,
+      "learning_rate": 9.328563955272873e-06,
+      "loss": 0.718,
+      "step": 1824
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.596171146233937,
+      "learning_rate": 9.327710662648128e-06,
+      "loss": 0.6523,
+      "step": 1825
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1903089534866957,
+      "learning_rate": 9.326856867242467e-06,
+      "loss": 0.6914,
+      "step": 1826
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1949315431307785,
+      "learning_rate": 9.326002569155084e-06,
+      "loss": 0.6121,
+      "step": 1827
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1196053273242206,
+      "learning_rate": 9.325147768485226e-06,
+      "loss": 0.7247,
+      "step": 1828
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.927602122432581,
+      "learning_rate": 9.324292465332205e-06,
+      "loss": 0.7023,
+      "step": 1829
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.59333950514103,
+      "learning_rate": 9.323436659795384e-06,
+      "loss": 0.774,
+      "step": 1830
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1860597709229794,
+      "learning_rate": 9.32258035197419e-06,
+      "loss": 0.6294,
+      "step": 1831
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.237668840439711,
+      "learning_rate": 9.321723541968106e-06,
+      "loss": 0.6915,
+      "step": 1832
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.013676832988803,
+      "learning_rate": 9.320866229876674e-06,
+      "loss": 0.7642,
+      "step": 1833
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.357450534390374,
+      "learning_rate": 9.320008415799496e-06,
+      "loss": 0.7363,
+      "step": 1834
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2990217042269863,
+      "learning_rate": 9.319150099836225e-06,
+      "loss": 0.6567,
+      "step": 1835
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4112889638082033,
+      "learning_rate": 9.318291282086582e-06,
+      "loss": 0.6427,
+      "step": 1836
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5245440646141706,
+      "learning_rate": 9.317431962650339e-06,
+      "loss": 0.6699,
+      "step": 1837
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1773949952495566,
+      "learning_rate": 9.316572141627334e-06,
+      "loss": 0.6629,
+      "step": 1838
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0104341618703967,
+      "learning_rate": 9.315711819117452e-06,
+      "loss": 0.6719,
+      "step": 1839
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5108935311841467,
+      "learning_rate": 9.31485099522065e-06,
+      "loss": 0.7134,
+      "step": 1840
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.695150335447739,
+      "learning_rate": 9.31398967003693e-06,
+      "loss": 0.8007,
+      "step": 1841
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.7187367089479895,
+      "learning_rate": 9.31312784366636e-06,
+      "loss": 0.6938,
+      "step": 1842
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.266932399300048,
+      "learning_rate": 9.312265516209068e-06,
+      "loss": 0.7279,
+      "step": 1843
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.064288579022452,
+      "learning_rate": 9.311402687765231e-06,
+      "loss": 0.6903,
+      "step": 1844
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3054354931895165,
+      "learning_rate": 9.310539358435095e-06,
+      "loss": 0.6779,
+      "step": 1845
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3704699001159244,
+      "learning_rate": 9.309675528318955e-06,
+      "loss": 0.7828,
+      "step": 1846
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5359614229510274,
+      "learning_rate": 9.308811197517172e-06,
+      "loss": 0.6946,
+      "step": 1847
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.433416978342713,
+      "learning_rate": 9.307946366130158e-06,
+      "loss": 0.8163,
+      "step": 1848
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6463567241215493,
+      "learning_rate": 9.307081034258389e-06,
+      "loss": 0.7162,
+      "step": 1849
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2116023992699523,
+      "learning_rate": 9.306215202002396e-06,
+      "loss": 0.6796,
+      "step": 1850
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0445330680873828,
+      "learning_rate": 9.305348869462768e-06,
+      "loss": 0.7149,
+      "step": 1851
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.881081284323706,
+      "learning_rate": 9.304482036740154e-06,
+      "loss": 0.6436,
+      "step": 1852
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4175242650695687,
+      "learning_rate": 9.30361470393526e-06,
+      "loss": 0.6238,
+      "step": 1853
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.430329856197137,
+      "learning_rate": 9.302746871148852e-06,
+      "loss": 0.7318,
+      "step": 1854
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1962368489134803,
+      "learning_rate": 9.301878538481748e-06,
+      "loss": 0.7108,
+      "step": 1855
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.169152259585836,
+      "learning_rate": 9.30100970603483e-06,
+      "loss": 0.7173,
+      "step": 1856
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.463790448876148,
+      "learning_rate": 9.30014037390904e-06,
+      "loss": 0.689,
+      "step": 1857
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.368852330455152,
+      "learning_rate": 9.299270542205372e-06,
+      "loss": 0.6933,
+      "step": 1858
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.444593231468713,
+      "learning_rate": 9.298400211024878e-06,
+      "loss": 0.6799,
+      "step": 1859
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8680880820924584,
+      "learning_rate": 9.297529380468675e-06,
+      "loss": 0.6574,
+      "step": 1860
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.948905283236975,
+      "learning_rate": 9.29665805063793e-06,
+      "loss": 0.6756,
+      "step": 1861
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1963367833574643,
+      "learning_rate": 9.295786221633874e-06,
+      "loss": 0.634,
+      "step": 1862
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.139790541038441,
+      "learning_rate": 9.294913893557792e-06,
+      "loss": 0.6734,
+      "step": 1863
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.619706209228136,
+      "learning_rate": 9.294041066511031e-06,
+      "loss": 0.6999,
+      "step": 1864
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.409480108291939,
+      "learning_rate": 9.29316774059499e-06,
+      "loss": 0.6769,
+      "step": 1865
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.410329086374503,
+      "learning_rate": 9.29229391591113e-06,
+      "loss": 0.7089,
+      "step": 1866
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.690751097339351,
+      "learning_rate": 9.291419592560973e-06,
+      "loss": 0.7696,
+      "step": 1867
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.69073549819373,
+      "learning_rate": 9.290544770646092e-06,
+      "loss": 0.6698,
+      "step": 1868
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.9808783369984093,
+      "learning_rate": 9.289669450268122e-06,
+      "loss": 0.6039,
+      "step": 1869
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.15890459589978,
+      "learning_rate": 9.288793631528757e-06,
+      "loss": 0.6616,
+      "step": 1870
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.081034392032343,
+      "learning_rate": 9.287917314529743e-06,
+      "loss": 0.682,
+      "step": 1871
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1849941363664382,
+      "learning_rate": 9.287040499372893e-06,
+      "loss": 0.7173,
+      "step": 1872
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4741373441095202,
+      "learning_rate": 9.286163186160067e-06,
+      "loss": 0.6348,
+      "step": 1873
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4013819058393944,
+      "learning_rate": 9.285285374993195e-06,
+      "loss": 0.641,
+      "step": 1874
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3332689108954336,
+      "learning_rate": 9.284407065974254e-06,
+      "loss": 0.7153,
+      "step": 1875
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3415860761989378,
+      "learning_rate": 9.283528259205287e-06,
+      "loss": 0.7111,
+      "step": 1876
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5710052029402166,
+      "learning_rate": 9.282648954788387e-06,
+      "loss": 0.6825,
+      "step": 1877
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.172242362761334,
+      "learning_rate": 9.281769152825713e-06,
+      "loss": 0.6001,
+      "step": 1878
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.202637528829396,
+      "learning_rate": 9.280888853419476e-06,
+      "loss": 0.6635,
+      "step": 1879
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.4006855761361345,
+      "learning_rate": 9.280008056671947e-06,
+      "loss": 0.694,
+      "step": 1880
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.831327947150695,
+      "learning_rate": 9.279126762685454e-06,
+      "loss": 0.636,
+      "step": 1881
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6062353877854436,
+      "learning_rate": 9.278244971562382e-06,
+      "loss": 0.8014,
+      "step": 1882
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.870705908895546,
+      "learning_rate": 9.277362683405177e-06,
+      "loss": 0.6577,
+      "step": 1883
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3116536973353057,
+      "learning_rate": 9.276479898316341e-06,
+      "loss": 0.6627,
+      "step": 1884
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.325166740616653,
+      "learning_rate": 9.275596616398431e-06,
+      "loss": 0.7138,
+      "step": 1885
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.764579252240071,
+      "learning_rate": 9.274712837754068e-06,
+      "loss": 0.6182,
+      "step": 1886
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.610119518405415,
+      "learning_rate": 9.273828562485923e-06,
+      "loss": 0.6369,
+      "step": 1887
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3604350623510846,
+      "learning_rate": 9.272943790696728e-06,
+      "loss": 0.6644,
+      "step": 1888
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2652286769748495,
+      "learning_rate": 9.272058522489277e-06,
+      "loss": 0.7804,
+      "step": 1889
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9511210502310745,
+      "learning_rate": 9.271172757966418e-06,
+      "loss": 0.6344,
+      "step": 1890
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8169929100268867,
+      "learning_rate": 9.270286497231052e-06,
+      "loss": 0.697,
+      "step": 1891
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.7252767248143597,
+      "learning_rate": 9.269399740386146e-06,
+      "loss": 0.7239,
+      "step": 1892
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.507005417930868,
+      "learning_rate": 9.26851248753472e-06,
+      "loss": 0.6427,
+      "step": 1893
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8482802966189373,
+      "learning_rate": 9.267624738779853e-06,
+      "loss": 0.6975,
+      "step": 1894
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8213003925408153,
+      "learning_rate": 9.266736494224677e-06,
+      "loss": 0.6433,
+      "step": 1895
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.436716166307937,
+      "learning_rate": 9.265847753972392e-06,
+      "loss": 0.6206,
+      "step": 1896
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0067406772274063,
+      "learning_rate": 9.264958518126246e-06,
+      "loss": 0.7178,
+      "step": 1897
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.652398231372856,
+      "learning_rate": 9.264068786789546e-06,
+      "loss": 0.7469,
+      "step": 1898
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.903751071631814,
+      "learning_rate": 9.263178560065664e-06,
+      "loss": 0.6284,
+      "step": 1899
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.233324204587242,
+      "learning_rate": 9.262287838058017e-06,
+      "loss": 0.654,
+      "step": 1900
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.426621816553852,
+      "learning_rate": 9.261396620870092e-06,
+      "loss": 0.6874,
+      "step": 1901
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9434733409299283,
+      "learning_rate": 9.260504908605425e-06,
+      "loss": 0.6296,
+      "step": 1902
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.370239117918409,
+      "learning_rate": 9.259612701367615e-06,
+      "loss": 0.6566,
+      "step": 1903
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.306513335576572,
+      "learning_rate": 9.258719999260315e-06,
+      "loss": 0.6967,
+      "step": 1904
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.385612710261463,
+      "learning_rate": 9.257826802387234e-06,
+      "loss": 0.6995,
+      "step": 1905
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8909613530557707,
+      "learning_rate": 9.256933110852145e-06,
+      "loss": 0.6234,
+      "step": 1906
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.956138783299861,
+      "learning_rate": 9.25603892475887e-06,
+      "loss": 0.7436,
+      "step": 1907
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.9130492645144934,
+      "learning_rate": 9.255144244211299e-06,
+      "loss": 0.7194,
+      "step": 1908
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8433841114866922,
+      "learning_rate": 9.254249069313368e-06,
+      "loss": 0.6685,
+      "step": 1909
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0403922052487427,
+      "learning_rate": 9.253353400169078e-06,
+      "loss": 0.6988,
+      "step": 1910
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3083987036017835,
+      "learning_rate": 9.252457236882487e-06,
+      "loss": 0.659,
+      "step": 1911
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.636186973720048,
+      "learning_rate": 9.251560579557705e-06,
+      "loss": 0.6196,
+      "step": 1912
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3354347387696337,
+      "learning_rate": 9.250663428298906e-06,
+      "loss": 0.7811,
+      "step": 1913
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.3940926366504227,
+      "learning_rate": 9.249765783210316e-06,
+      "loss": 0.6968,
+      "step": 1914
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9781144300478661,
+      "learning_rate": 9.248867644396224e-06,
+      "loss": 0.7138,
+      "step": 1915
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.221082262034515,
+      "learning_rate": 9.24796901196097e-06,
+      "loss": 0.7758,
+      "step": 1916
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.281625973109882,
+      "learning_rate": 9.247069886008957e-06,
+      "loss": 0.6453,
+      "step": 1917
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0495291535897127,
+      "learning_rate": 9.24617026664464e-06,
+      "loss": 0.6724,
+      "step": 1918
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.294555365241956,
+      "learning_rate": 9.245270153972537e-06,
+      "loss": 0.6348,
+      "step": 1919
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6583206062968143,
+      "learning_rate": 9.244369548097218e-06,
+      "loss": 0.7125,
+      "step": 1920
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2992496345079894,
+      "learning_rate": 9.243468449123316e-06,
+      "loss": 0.6501,
+      "step": 1921
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.294260337838386,
+      "learning_rate": 9.242566857155515e-06,
+      "loss": 0.6783,
+      "step": 1922
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2677177768740107,
+      "learning_rate": 9.241664772298561e-06,
+      "loss": 0.7314,
+      "step": 1923
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5676248737182172,
+      "learning_rate": 9.240762194657254e-06,
+      "loss": 0.7354,
+      "step": 1924
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.456584308459011,
+      "learning_rate": 9.239859124336457e-06,
+      "loss": 0.7148,
+      "step": 1925
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5953088609259134,
+      "learning_rate": 9.23895556144108e-06,
+      "loss": 0.7198,
+      "step": 1926
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.144466625240076,
+      "learning_rate": 9.2380515060761e-06,
+      "loss": 0.6582,
+      "step": 1927
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.821280176023632,
+      "learning_rate": 9.237146958346549e-06,
+      "loss": 0.6836,
+      "step": 1928
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.012627315683036,
+      "learning_rate": 9.236241918357511e-06,
+      "loss": 0.7583,
+      "step": 1929
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3886139997345825,
+      "learning_rate": 9.235336386214133e-06,
+      "loss": 0.6987,
+      "step": 1930
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.5518418067148287,
+      "learning_rate": 9.234430362021615e-06,
+      "loss": 0.7171,
+      "step": 1931
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0640715116734265,
+      "learning_rate": 9.233523845885221e-06,
+      "loss": 0.5934,
+      "step": 1932
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5462408850965303,
+      "learning_rate": 9.232616837910263e-06,
+      "loss": 0.6674,
+      "step": 1933
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1027514803586316,
+      "learning_rate": 9.231709338202117e-06,
+      "loss": 0.7143,
+      "step": 1934
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.981837777190229,
+      "learning_rate": 9.230801346866212e-06,
+      "loss": 0.6824,
+      "step": 1935
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0072257637585995,
+      "learning_rate": 9.229892864008037e-06,
+      "loss": 0.6737,
+      "step": 1936
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.404301234252323,
+      "learning_rate": 9.228983889733135e-06,
+      "loss": 0.72,
+      "step": 1937
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1099503574763405,
+      "learning_rate": 9.228074424147111e-06,
+      "loss": 0.7708,
+      "step": 1938
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1564706992929987,
+      "learning_rate": 9.227164467355621e-06,
+      "loss": 0.7047,
+      "step": 1939
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.717145887514134,
+      "learning_rate": 9.226254019464384e-06,
+      "loss": 0.7386,
+      "step": 1940
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.434512344610932,
+      "learning_rate": 9.225343080579171e-06,
+      "loss": 0.7586,
+      "step": 1941
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.516975033266961,
+      "learning_rate": 9.224431650805814e-06,
+      "loss": 0.721,
+      "step": 1942
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0749482225336697,
+      "learning_rate": 9.223519730250198e-06,
+      "loss": 0.6007,
+      "step": 1943
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5096562394219775,
+      "learning_rate": 9.222607319018271e-06,
+      "loss": 0.7729,
+      "step": 1944
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9924376463945759,
+      "learning_rate": 9.221694417216031e-06,
+      "loss": 0.7065,
+      "step": 1945
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4082767933565594,
+      "learning_rate": 9.220781024949536e-06,
+      "loss": 0.7151,
+      "step": 1946
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.157525030220561,
+      "learning_rate": 9.219867142324904e-06,
+      "loss": 0.6323,
+      "step": 1947
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.809345903838342,
+      "learning_rate": 9.218952769448307e-06,
+      "loss": 0.6643,
+      "step": 1948
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1126315310721426,
+      "learning_rate": 9.218037906425971e-06,
+      "loss": 0.66,
+      "step": 1949
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.106123096050007,
+      "learning_rate": 9.217122553364184e-06,
+      "loss": 0.779,
+      "step": 1950
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4264702932506927,
+      "learning_rate": 9.21620671036929e-06,
+      "loss": 0.7279,
+      "step": 1951
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.387568989843238,
+      "learning_rate": 9.215290377547688e-06,
+      "loss": 0.576,
+      "step": 1952
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.721569909570348,
+      "learning_rate": 9.214373555005834e-06,
+      "loss": 0.6169,
+      "step": 1953
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5751826747005553,
+      "learning_rate": 9.213456242850245e-06,
+      "loss": 0.7445,
+      "step": 1954
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.238179252301112,
+      "learning_rate": 9.21253844118749e-06,
+      "loss": 0.6893,
+      "step": 1955
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.230727560495232,
+      "learning_rate": 9.211620150124192e-06,
+      "loss": 0.7169,
+      "step": 1956
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5021330652535645,
+      "learning_rate": 9.210701369767043e-06,
+      "loss": 0.7869,
+      "step": 1957
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7468363192392427,
+      "learning_rate": 9.20978210022278e-06,
+      "loss": 0.7359,
+      "step": 1958
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0826919441428458,
+      "learning_rate": 9.208862341598201e-06,
+      "loss": 0.7324,
+      "step": 1959
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7555569094425993,
+      "learning_rate": 9.207942094000163e-06,
+      "loss": 0.7457,
+      "step": 1960
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1278899651708088,
+      "learning_rate": 9.207021357535576e-06,
+      "loss": 0.7093,
+      "step": 1961
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.304923508024911,
+      "learning_rate": 9.206100132311408e-06,
+      "loss": 0.6779,
+      "step": 1962
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3962706909638007,
+      "learning_rate": 9.205178418434687e-06,
+      "loss": 0.7315,
+      "step": 1963
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3044315673666205,
+      "learning_rate": 9.204256216012493e-06,
+      "loss": 0.7382,
+      "step": 1964
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4732648879760863,
+      "learning_rate": 9.203333525151964e-06,
+      "loss": 0.7135,
+      "step": 1965
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.206789161392687,
+      "learning_rate": 9.202410345960298e-06,
+      "loss": 0.7343,
+      "step": 1966
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4651985133702827,
+      "learning_rate": 9.201486678544745e-06,
+      "loss": 0.718,
+      "step": 1967
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2568110947549407,
+      "learning_rate": 9.200562523012615e-06,
+      "loss": 0.6678,
+      "step": 1968
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4466689065325182,
+      "learning_rate": 9.199637879471272e-06,
+      "loss": 0.7278,
+      "step": 1969
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7412446183420593,
+      "learning_rate": 9.198712748028142e-06,
+      "loss": 0.6746,
+      "step": 1970
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.090127601569364,
+      "learning_rate": 9.197787128790702e-06,
+      "loss": 0.7454,
+      "step": 1971
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2493962206295457,
+      "learning_rate": 9.19686102186649e-06,
+      "loss": 0.6949,
+      "step": 1972
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4780855364247603,
+      "learning_rate": 9.195934427363093e-06,
+      "loss": 0.7695,
+      "step": 1973
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.720134076492023,
+      "learning_rate": 9.195007345388165e-06,
+      "loss": 0.7939,
+      "step": 1974
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6619081847464705,
+      "learning_rate": 9.19407977604941e-06,
+      "loss": 0.7645,
+      "step": 1975
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1669764589008986,
+      "learning_rate": 9.193151719454591e-06,
+      "loss": 0.658,
+      "step": 1976
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7606276779843375,
+      "learning_rate": 9.192223175711526e-06,
+      "loss": 0.641,
+      "step": 1977
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5503781978393776,
+      "learning_rate": 9.191294144928091e-06,
+      "loss": 0.7459,
+      "step": 1978
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2113194458441345,
+      "learning_rate": 9.190364627212216e-06,
+      "loss": 0.6788,
+      "step": 1979
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.178914025765916,
+      "learning_rate": 9.189434622671894e-06,
+      "loss": 0.6719,
+      "step": 1980
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.684741530038339,
+      "learning_rate": 9.188504131415167e-06,
+      "loss": 0.6779,
+      "step": 1981
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.245253383160478,
+      "learning_rate": 9.187573153550139e-06,
+      "loss": 0.6422,
+      "step": 1982
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5644319947882055,
+      "learning_rate": 9.186641689184966e-06,
+      "loss": 0.7558,
+      "step": 1983
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6626949713612134,
+      "learning_rate": 9.185709738427864e-06,
+      "loss": 0.731,
+      "step": 1984
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.227738894627522,
+      "learning_rate": 9.184777301387104e-06,
+      "loss": 0.7663,
+      "step": 1985
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3446022774136024,
+      "learning_rate": 9.183844378171016e-06,
+      "loss": 0.6564,
+      "step": 1986
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.432195543728687,
+      "learning_rate": 9.182910968887982e-06,
+      "loss": 0.7728,
+      "step": 1987
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4823577607004146,
+      "learning_rate": 9.181977073646442e-06,
+      "loss": 0.6328,
+      "step": 1988
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.177220583837391,
+      "learning_rate": 9.181042692554894e-06,
+      "loss": 0.6911,
+      "step": 1989
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.483575743653078,
+      "learning_rate": 9.180107825721891e-06,
+      "loss": 0.7007,
+      "step": 1990
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.109985363397772,
+      "learning_rate": 9.179172473256046e-06,
+      "loss": 0.688,
+      "step": 1991
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0644562546243423,
+      "learning_rate": 9.178236635266025e-06,
+      "loss": 0.6994,
+      "step": 1992
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1056210922690526,
+      "learning_rate": 9.17730031186055e-06,
+      "loss": 0.7325,
+      "step": 1993
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.249585098272686,
+      "learning_rate": 9.176363503148397e-06,
+      "loss": 0.698,
+      "step": 1994
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.57223165623908,
+      "learning_rate": 9.175426209238407e-06,
+      "loss": 0.6476,
+      "step": 1995
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.6183458052615665,
+      "learning_rate": 9.17448843023947e-06,
+      "loss": 0.8366,
+      "step": 1996
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.072274750411721,
+      "learning_rate": 9.173550166260533e-06,
+      "loss": 0.7383,
+      "step": 1997
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3127023094424963,
+      "learning_rate": 9.172611417410604e-06,
+      "loss": 0.6212,
+      "step": 1998
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3387267206899556,
+      "learning_rate": 9.17167218379874e-06,
+      "loss": 0.7201,
+      "step": 1999
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.311361411354398,
+      "learning_rate": 9.170732465534062e-06,
+      "loss": 0.7375,
+      "step": 2000
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2786653146755618,
+      "learning_rate": 9.169792262725744e-06,
+      "loss": 0.7097,
+      "step": 2001
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.107809653312181,
+      "learning_rate": 9.168851575483013e-06,
+      "loss": 0.6939,
+      "step": 2002
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.057099189874853,
+      "learning_rate": 9.167910403915157e-06,
+      "loss": 0.7254,
+      "step": 2003
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.445940571938724,
+      "learning_rate": 9.16696874813152e-06,
+      "loss": 0.6684,
+      "step": 2004
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4846164924547747,
+      "learning_rate": 9.166026608241496e-06,
+      "loss": 0.6799,
+      "step": 2005
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2002524931768814,
+      "learning_rate": 9.165083984354545e-06,
+      "loss": 0.7163,
+      "step": 2006
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.2128685263981924,
+      "learning_rate": 9.164140876580179e-06,
+      "loss": 0.6353,
+      "step": 2007
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.487800260585636,
+      "learning_rate": 9.16319728502796e-06,
+      "loss": 0.7001,
+      "step": 2008
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3851629284461553,
+      "learning_rate": 9.162253209807517e-06,
+      "loss": 0.6288,
+      "step": 2009
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4232158597086744,
+      "learning_rate": 9.161308651028527e-06,
+      "loss": 0.7754,
+      "step": 2010
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.154013851517835,
+      "learning_rate": 9.160363608800728e-06,
+      "loss": 0.7117,
+      "step": 2011
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.896276107997591,
+      "learning_rate": 9.159418083233911e-06,
+      "loss": 0.6859,
+      "step": 2012
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.44486977130411,
+      "learning_rate": 9.158472074437923e-06,
+      "loss": 0.6467,
+      "step": 2013
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0090988720613,
+      "learning_rate": 9.157525582522673e-06,
+      "loss": 0.6484,
+      "step": 2014
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.34089707383189,
+      "learning_rate": 9.156578607598118e-06,
+      "loss": 0.7049,
+      "step": 2015
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2880641379260889,
+      "learning_rate": 9.155631149774276e-06,
+      "loss": 0.6175,
+      "step": 2016
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.229526829371313,
+      "learning_rate": 9.15468320916122e-06,
+      "loss": 0.6662,
+      "step": 2017
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.156990061837107,
+      "learning_rate": 9.153734785869077e-06,
+      "loss": 0.7708,
+      "step": 2018
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0523799529588602,
+      "learning_rate": 9.152785880008035e-06,
+      "loss": 0.6244,
+      "step": 2019
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.993605252322678,
+      "learning_rate": 9.151836491688334e-06,
+      "loss": 0.6563,
+      "step": 2020
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3078838508848922,
+      "learning_rate": 9.15088662102027e-06,
+      "loss": 0.6904,
+      "step": 2021
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.063801331956025,
+      "learning_rate": 9.149936268114199e-06,
+      "loss": 0.676,
+      "step": 2022
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.295192468843143,
+      "learning_rate": 9.148985433080528e-06,
+      "loss": 0.6834,
+      "step": 2023
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0564302867231383,
+      "learning_rate": 9.148034116029723e-06,
+      "loss": 0.704,
+      "step": 2024
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5228740207896085,
+      "learning_rate": 9.147082317072305e-06,
+      "loss": 0.6683,
+      "step": 2025
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7052297386153015,
+      "learning_rate": 9.146130036318853e-06,
+      "loss": 0.7911,
+      "step": 2026
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.601417655514352,
+      "learning_rate": 9.145177273879995e-06,
+      "loss": 0.6382,
+      "step": 2027
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9751425113181098,
+      "learning_rate": 9.144224029866426e-06,
+      "loss": 0.6693,
+      "step": 2028
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5895232014896266,
+      "learning_rate": 9.14327030438889e-06,
+      "loss": 0.6712,
+      "step": 2029
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.747457510645605,
+      "learning_rate": 9.142316097558185e-06,
+      "loss": 0.6682,
+      "step": 2030
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.220857892024144,
+      "learning_rate": 9.14136140948517e-06,
+      "loss": 0.7372,
+      "step": 2031
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3266528015138386,
+      "learning_rate": 9.14040624028076e-06,
+      "loss": 0.7786,
+      "step": 2032
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.348078882997178,
+      "learning_rate": 9.13945059005592e-06,
+      "loss": 0.7961,
+      "step": 2033
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.81669382916974,
+      "learning_rate": 9.138494458921676e-06,
+      "loss": 0.6599,
+      "step": 2034
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.170306767141564,
+      "learning_rate": 9.137537846989111e-06,
+      "loss": 0.6299,
+      "step": 2035
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5696842899169385,
+      "learning_rate": 9.136580754369357e-06,
+      "loss": 0.6784,
+      "step": 2036
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3739145799623045,
+      "learning_rate": 9.135623181173609e-06,
+      "loss": 0.7959,
+      "step": 2037
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3501976834313174,
+      "learning_rate": 9.134665127513116e-06,
+      "loss": 0.6405,
+      "step": 2038
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7821580766766676,
+      "learning_rate": 9.133706593499181e-06,
+      "loss": 0.7582,
+      "step": 2039
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0217636722769323,
+      "learning_rate": 9.132747579243163e-06,
+      "loss": 0.7176,
+      "step": 2040
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.149948219126222,
+      "learning_rate": 9.131788084856477e-06,
+      "loss": 0.7303,
+      "step": 2041
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.279544855221473,
+      "learning_rate": 9.130828110450593e-06,
+      "loss": 0.6588,
+      "step": 2042
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.49303969504543,
+      "learning_rate": 9.129867656137044e-06,
+      "loss": 0.7645,
+      "step": 2043
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.143115712697355,
+      "learning_rate": 9.128906722027406e-06,
+      "loss": 0.6674,
+      "step": 2044
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0564042062634225,
+      "learning_rate": 9.127945308233322e-06,
+      "loss": 0.7074,
+      "step": 2045
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.147223116012419,
+      "learning_rate": 9.126983414866486e-06,
+      "loss": 0.7134,
+      "step": 2046
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.272635221500622,
+      "learning_rate": 9.126021042038644e-06,
+      "loss": 0.7515,
+      "step": 2047
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7298434180359954,
+      "learning_rate": 9.125058189861607e-06,
+      "loss": 0.7381,
+      "step": 2048
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0486731591766913,
+      "learning_rate": 9.124094858447233e-06,
+      "loss": 0.6821,
+      "step": 2049
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.6588169832809205,
+      "learning_rate": 9.123131047907439e-06,
+      "loss": 0.8302,
+      "step": 2050
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.07023150767486,
+      "learning_rate": 9.122166758354199e-06,
+      "loss": 0.6876,
+      "step": 2051
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7807238495464133,
+      "learning_rate": 9.12120198989954e-06,
+      "loss": 0.5919,
+      "step": 2052
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1579164615666837,
+      "learning_rate": 9.120236742655548e-06,
+      "loss": 0.7351,
+      "step": 2053
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.5527053709438214,
+      "learning_rate": 9.11927101673436e-06,
+      "loss": 0.6993,
+      "step": 2054
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.261502991172867,
+      "learning_rate": 9.118304812248177e-06,
+      "loss": 0.6502,
+      "step": 2055
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.242461395330852,
+      "learning_rate": 9.117338129309243e-06,
+      "loss": 0.6605,
+      "step": 2056
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.914414634765642,
+      "learning_rate": 9.116370968029867e-06,
+      "loss": 0.6981,
+      "step": 2057
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.344979314174602,
+      "learning_rate": 9.115403328522412e-06,
+      "loss": 0.7227,
+      "step": 2058
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.191238107674176,
+      "learning_rate": 9.114435210899296e-06,
+      "loss": 0.6631,
+      "step": 2059
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4602672291177528,
+      "learning_rate": 9.113466615272988e-06,
+      "loss": 0.7243,
+      "step": 2060
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.44867772558428,
+      "learning_rate": 9.11249754175602e-06,
+      "loss": 0.6982,
+      "step": 2061
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.651963613477152,
+      "learning_rate": 9.111527990460977e-06,
+      "loss": 0.7013,
+      "step": 2062
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.5214633687617325,
+      "learning_rate": 9.110557961500496e-06,
+      "loss": 0.6273,
+      "step": 2063
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1745678771179233,
+      "learning_rate": 9.109587454987274e-06,
+      "loss": 0.6482,
+      "step": 2064
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.489922591200943,
+      "learning_rate": 9.108616471034061e-06,
+      "loss": 0.6609,
+      "step": 2065
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0973955866357814,
+      "learning_rate": 9.107645009753663e-06,
+      "loss": 0.6722,
+      "step": 2066
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9996934192215,
+      "learning_rate": 9.106673071258942e-06,
+      "loss": 0.7301,
+      "step": 2067
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.494129615110545,
+      "learning_rate": 9.105700655662815e-06,
+      "loss": 0.6926,
+      "step": 2068
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.264357578268509,
+      "learning_rate": 9.104727763078253e-06,
+      "loss": 0.7061,
+      "step": 2069
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.8039561431694544,
+      "learning_rate": 9.103754393618287e-06,
+      "loss": 0.6591,
+      "step": 2070
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2260708554024737,
+      "learning_rate": 9.102780547395997e-06,
+      "loss": 0.7733,
+      "step": 2071
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6838150314170997,
+      "learning_rate": 9.101806224524524e-06,
+      "loss": 0.7586,
+      "step": 2072
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.5259424740449896,
+      "learning_rate": 9.10083142511706e-06,
+      "loss": 0.6863,
+      "step": 2073
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.894372573153219,
+      "learning_rate": 9.099856149286857e-06,
+      "loss": 0.6572,
+      "step": 2074
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.246841628686618,
+      "learning_rate": 9.098880397147215e-06,
+      "loss": 0.6908,
+      "step": 2075
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.039572983541654,
+      "learning_rate": 9.0979041688115e-06,
+      "loss": 0.6255,
+      "step": 2076
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5644692341003028,
+      "learning_rate": 9.096927464393123e-06,
+      "loss": 0.6359,
+      "step": 2077
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.305502204546247,
+      "learning_rate": 9.095950284005557e-06,
+      "loss": 0.7748,
+      "step": 2078
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2388140399919334,
+      "learning_rate": 9.094972627762326e-06,
+      "loss": 0.7147,
+      "step": 2079
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.137643134632365,
+      "learning_rate": 9.093994495777014e-06,
+      "loss": 0.7239,
+      "step": 2080
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.166788507712685,
+      "learning_rate": 9.093015888163255e-06,
+      "loss": 0.7105,
+      "step": 2081
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.811849358016705,
+      "learning_rate": 9.09203680503474e-06,
+      "loss": 0.6821,
+      "step": 2082
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.196127603607177,
+      "learning_rate": 9.091057246505221e-06,
+      "loss": 0.7968,
+      "step": 2083
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.922784029473444,
+      "learning_rate": 9.090077212688496e-06,
+      "loss": 0.7587,
+      "step": 2084
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.039930805628341,
+      "learning_rate": 9.089096703698423e-06,
+      "loss": 0.6852,
+      "step": 2085
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 30.95824912540301,
+      "learning_rate": 9.088115719648917e-06,
+      "loss": 0.7022,
+      "step": 2086
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.181623054651788,
+      "learning_rate": 9.087134260653943e-06,
+      "loss": 0.6246,
+      "step": 2087
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9997909852846925,
+      "learning_rate": 9.086152326827527e-06,
+      "loss": 0.6697,
+      "step": 2088
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.395290205085553,
+      "learning_rate": 9.085169918283744e-06,
+      "loss": 0.6567,
+      "step": 2089
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.836594783866522,
+      "learning_rate": 9.084187035136727e-06,
+      "loss": 0.7324,
+      "step": 2090
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3072720294313496,
+      "learning_rate": 9.08320367750067e-06,
+      "loss": 0.6529,
+      "step": 2091
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8931936091262302,
+      "learning_rate": 9.08221984548981e-06,
+      "loss": 0.7267,
+      "step": 2092
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4682881700848234,
+      "learning_rate": 9.081235539218451e-06,
+      "loss": 0.6981,
+      "step": 2093
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1040997554296026,
+      "learning_rate": 9.080250758800944e-06,
+      "loss": 0.7395,
+      "step": 2094
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.148229839051349,
+      "learning_rate": 9.0792655043517e-06,
+      "loss": 0.6395,
+      "step": 2095
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.008482369488449,
+      "learning_rate": 9.078279775985179e-06,
+      "loss": 0.665,
+      "step": 2096
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.414670215539318,
+      "learning_rate": 9.077293573815905e-06,
+      "loss": 0.7037,
+      "step": 2097
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6045420628577474,
+      "learning_rate": 9.07630689795845e-06,
+      "loss": 0.6876,
+      "step": 2098
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.089544444738824,
+      "learning_rate": 9.075319748527442e-06,
+      "loss": 0.6634,
+      "step": 2099
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.068471813524876,
+      "learning_rate": 9.074332125637564e-06,
+      "loss": 0.7743,
+      "step": 2100
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.390556215843803,
+      "learning_rate": 9.073344029403562e-06,
+      "loss": 0.6952,
+      "step": 2101
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.5241030806585005,
+      "learning_rate": 9.072355459940222e-06,
+      "loss": 0.7437,
+      "step": 2102
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.752003680927352,
+      "learning_rate": 9.071366417362398e-06,
+      "loss": 0.7585,
+      "step": 2103
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.5306889032364492,
+      "learning_rate": 9.070376901784992e-06,
+      "loss": 0.7722,
+      "step": 2104
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.5152078389519015,
+      "learning_rate": 9.069386913322964e-06,
+      "loss": 0.7526,
+      "step": 2105
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.300437974231803,
+      "learning_rate": 9.068396452091328e-06,
+      "loss": 0.7412,
+      "step": 2106
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.966239911106796,
+      "learning_rate": 9.067405518205153e-06,
+      "loss": 0.6955,
+      "step": 2107
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.374383053660284,
+      "learning_rate": 9.066414111779562e-06,
+      "loss": 0.7065,
+      "step": 2108
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.95341575255716,
+      "learning_rate": 9.065422232929735e-06,
+      "loss": 0.6911,
+      "step": 2109
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.38959249885153,
+      "learning_rate": 9.064429881770905e-06,
+      "loss": 0.5927,
+      "step": 2110
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.354874954642544,
+      "learning_rate": 9.063437058418361e-06,
+      "loss": 0.6709,
+      "step": 2111
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.1182596136285308,
+      "learning_rate": 9.062443762987442e-06,
+      "loss": 0.7256,
+      "step": 2112
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3451511992294085,
+      "learning_rate": 9.061449995593554e-06,
+      "loss": 0.6163,
+      "step": 2113
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6148189214617816,
+      "learning_rate": 9.060455756352144e-06,
+      "loss": 0.7164,
+      "step": 2114
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.984293021281382,
+      "learning_rate": 9.059461045378723e-06,
+      "loss": 0.5894,
+      "step": 2115
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.254747474890875,
+      "learning_rate": 9.058465862788852e-06,
+      "loss": 0.6708,
+      "step": 2116
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.363609404179215,
+      "learning_rate": 9.05747020869815e-06,
+      "loss": 0.6888,
+      "step": 2117
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.37996717795513,
+      "learning_rate": 9.056474083222286e-06,
+      "loss": 0.5809,
+      "step": 2118
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7306453255932848,
+      "learning_rate": 9.055477486476992e-06,
+      "loss": 0.7184,
+      "step": 2119
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3108089867660713,
+      "learning_rate": 9.054480418578044e-06,
+      "loss": 0.6179,
+      "step": 2120
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.5666873554391665,
+      "learning_rate": 9.053482879641283e-06,
+      "loss": 0.701,
+      "step": 2121
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.492602226878482,
+      "learning_rate": 9.052484869782597e-06,
+      "loss": 0.6191,
+      "step": 2122
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.660745248304112,
+      "learning_rate": 9.051486389117933e-06,
+      "loss": 0.6498,
+      "step": 2123
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.593579048719051,
+      "learning_rate": 9.050487437763294e-06,
+      "loss": 0.7411,
+      "step": 2124
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.193130270383422,
+      "learning_rate": 9.049488015834731e-06,
+      "loss": 0.6509,
+      "step": 2125
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.286545999733999,
+      "learning_rate": 9.048488123448357e-06,
+      "loss": 0.7333,
+      "step": 2126
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7285901550002145,
+      "learning_rate": 9.047487760720338e-06,
+      "loss": 0.755,
+      "step": 2127
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.5323414148927963,
+      "learning_rate": 9.046486927766889e-06,
+      "loss": 0.6686,
+      "step": 2128
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2528013077555502,
+      "learning_rate": 9.045485624704287e-06,
+      "loss": 0.6362,
+      "step": 2129
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.170043281237849,
+      "learning_rate": 9.044483851648858e-06,
+      "loss": 0.6531,
+      "step": 2130
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.537814352100077,
+      "learning_rate": 9.043481608716987e-06,
+      "loss": 0.7327,
+      "step": 2131
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1536774455658683,
+      "learning_rate": 9.042478896025113e-06,
+      "loss": 0.737,
+      "step": 2132
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4395488467277504,
+      "learning_rate": 9.041475713689725e-06,
+      "loss": 0.7198,
+      "step": 2133
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.612160402864214,
+      "learning_rate": 9.04047206182737e-06,
+      "loss": 0.7047,
+      "step": 2134
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.361860087108175,
+      "learning_rate": 9.039467940554651e-06,
+      "loss": 0.8051,
+      "step": 2135
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1118384611691083,
+      "learning_rate": 9.038463349988226e-06,
+      "loss": 0.6651,
+      "step": 2136
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1279878773488656,
+      "learning_rate": 9.0374582902448e-06,
+      "loss": 0.7212,
+      "step": 2137
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.984425840534222,
+      "learning_rate": 9.036452761441143e-06,
+      "loss": 0.6806,
+      "step": 2138
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2564393133187943,
+      "learning_rate": 9.035446763694073e-06,
+      "loss": 0.6898,
+      "step": 2139
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.2675730698127854,
+      "learning_rate": 9.034440297120461e-06,
+      "loss": 0.6907,
+      "step": 2140
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1457055260967333,
+      "learning_rate": 9.03343336183724e-06,
+      "loss": 0.6275,
+      "step": 2141
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.699650466296311,
+      "learning_rate": 9.032425957961388e-06,
+      "loss": 0.7142,
+      "step": 2142
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.361799060850231,
+      "learning_rate": 9.031418085609946e-06,
+      "loss": 0.7878,
+      "step": 2143
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0889387826729044,
+      "learning_rate": 9.030409744900005e-06,
+      "loss": 0.7566,
+      "step": 2144
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6149850709834492,
+      "learning_rate": 9.029400935948712e-06,
+      "loss": 0.6909,
+      "step": 2145
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.102880043260286,
+      "learning_rate": 9.028391658873264e-06,
+      "loss": 0.6655,
+      "step": 2146
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6572624912551666,
+      "learning_rate": 9.027381913790916e-06,
+      "loss": 0.627,
+      "step": 2147
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6045372231550585,
+      "learning_rate": 9.026371700818982e-06,
+      "loss": 0.779,
+      "step": 2148
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.662458429721062,
+      "learning_rate": 9.025361020074823e-06,
+      "loss": 0.7057,
+      "step": 2149
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8868287840968536,
+      "learning_rate": 9.024349871675855e-06,
+      "loss": 0.7235,
+      "step": 2150
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2117676794317105,
+      "learning_rate": 9.023338255739553e-06,
+      "loss": 0.7088,
+      "step": 2151
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2014960756353057,
+      "learning_rate": 9.022326172383444e-06,
+      "loss": 0.7346,
+      "step": 2152
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.14684126625403,
+      "learning_rate": 9.021313621725106e-06,
+      "loss": 0.734,
+      "step": 2153
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.558372722059645,
+      "learning_rate": 9.020300603882178e-06,
+      "loss": 0.6929,
+      "step": 2154
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.442785125856375,
+      "learning_rate": 9.019287118972343e-06,
+      "loss": 0.7285,
+      "step": 2155
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5672631424976617,
+      "learning_rate": 9.018273167113354e-06,
+      "loss": 0.7115,
+      "step": 2156
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.9219116603164372,
+      "learning_rate": 9.017258748423e-06,
+      "loss": 0.7292,
+      "step": 2157
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1655630933315155,
+      "learning_rate": 9.01624386301914e-06,
+      "loss": 0.6373,
+      "step": 2158
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.9959938750864805,
+      "learning_rate": 9.015228511019678e-06,
+      "loss": 0.6972,
+      "step": 2159
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.451365705730853,
+      "learning_rate": 9.014212692542573e-06,
+      "loss": 0.7195,
+      "step": 2160
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.277660307422319,
+      "learning_rate": 9.013196407705842e-06,
+      "loss": 0.704,
+      "step": 2161
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.015544240787718,
+      "learning_rate": 9.012179656627553e-06,
+      "loss": 0.6184,
+      "step": 2162
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.046162331434754,
+      "learning_rate": 9.011162439425831e-06,
+      "loss": 0.7308,
+      "step": 2163
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.62859526794707,
+      "learning_rate": 9.010144756218851e-06,
+      "loss": 0.6979,
+      "step": 2164
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.026397922330441,
+      "learning_rate": 9.009126607124844e-06,
+      "loss": 0.6918,
+      "step": 2165
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3621280368628996,
+      "learning_rate": 9.008107992262098e-06,
+      "loss": 0.7819,
+      "step": 2166
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.176552442201047,
+      "learning_rate": 9.00708891174895e-06,
+      "loss": 0.7668,
+      "step": 2167
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.2070738864736765,
+      "learning_rate": 9.006069365703799e-06,
+      "loss": 0.6924,
+      "step": 2168
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.168031678031707,
+      "learning_rate": 9.005049354245088e-06,
+      "loss": 0.6743,
+      "step": 2169
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8808305867742656,
+      "learning_rate": 9.004028877491319e-06,
+      "loss": 0.631,
+      "step": 2170
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.295861751174668,
+      "learning_rate": 9.003007935561052e-06,
+      "loss": 0.6949,
+      "step": 2171
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.49760284826448,
+      "learning_rate": 9.001986528572892e-06,
+      "loss": 0.743,
+      "step": 2172
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.296535119642908,
+      "learning_rate": 9.000964656645508e-06,
+      "loss": 0.7973,
+      "step": 2173
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2214283799588825,
+      "learning_rate": 8.999942319897615e-06,
+      "loss": 0.6888,
+      "step": 2174
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1319580111833254,
+      "learning_rate": 8.998919518447986e-06,
+      "loss": 0.6926,
+      "step": 2175
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4770875145538556,
+      "learning_rate": 8.997896252415445e-06,
+      "loss": 0.7683,
+      "step": 2176
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.646693120678769,
+      "learning_rate": 8.996872521918877e-06,
+      "loss": 0.8287,
+      "step": 2177
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.343228561686917,
+      "learning_rate": 8.995848327077211e-06,
+      "loss": 0.7529,
+      "step": 2178
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.810342504369392,
+      "learning_rate": 8.994823668009437e-06,
+      "loss": 0.7357,
+      "step": 2179
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1482922949969145,
+      "learning_rate": 8.9937985448346e-06,
+      "loss": 0.7304,
+      "step": 2180
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8551204580841676,
+      "learning_rate": 8.992772957671791e-06,
+      "loss": 0.7101,
+      "step": 2181
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2466577666192475,
+      "learning_rate": 8.991746906640162e-06,
+      "loss": 0.6895,
+      "step": 2182
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.168650877854395,
+      "learning_rate": 8.990720391858915e-06,
+      "loss": 0.6479,
+      "step": 2183
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.300343690084461,
+      "learning_rate": 8.98969341344731e-06,
+      "loss": 0.6546,
+      "step": 2184
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3637093422065734,
+      "learning_rate": 8.98866597152466e-06,
+      "loss": 0.6307,
+      "step": 2185
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.754311126650472,
+      "learning_rate": 8.987638066210325e-06,
+      "loss": 0.6912,
+      "step": 2186
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1562502830328474,
+      "learning_rate": 8.986609697623724e-06,
+      "loss": 0.6785,
+      "step": 2187
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.327649052593783,
+      "learning_rate": 8.985580865884336e-06,
+      "loss": 0.7294,
+      "step": 2188
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.408598312141122,
+      "learning_rate": 8.984551571111683e-06,
+      "loss": 0.6296,
+      "step": 2189
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.142678163175262,
+      "learning_rate": 8.983521813425348e-06,
+      "loss": 0.6944,
+      "step": 2190
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.175847871801838,
+      "learning_rate": 8.982491592944962e-06,
+      "loss": 0.6382,
+      "step": 2191
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.228095121751495,
+      "learning_rate": 8.981460909790216e-06,
+      "loss": 0.5974,
+      "step": 2192
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.052783456234635,
+      "learning_rate": 8.98042976408085e-06,
+      "loss": 0.6731,
+      "step": 2193
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0986872293091468,
+      "learning_rate": 8.97939815593666e-06,
+      "loss": 0.6891,
+      "step": 2194
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.154720097555765,
+      "learning_rate": 8.978366085477497e-06,
+      "loss": 0.6185,
+      "step": 2195
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1403969118588067,
+      "learning_rate": 8.977333552823261e-06,
+      "loss": 0.6918,
+      "step": 2196
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.33214529211307,
+      "learning_rate": 8.976300558093911e-06,
+      "loss": 0.6466,
+      "step": 2197
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.260333763778915,
+      "learning_rate": 8.975267101409458e-06,
+      "loss": 0.7482,
+      "step": 2198
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0914836565162496,
+      "learning_rate": 8.974233182889961e-06,
+      "loss": 0.6254,
+      "step": 2199
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3762255505538676,
+      "learning_rate": 8.973198802655543e-06,
+      "loss": 0.7553,
+      "step": 2200
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1282785961850506,
+      "learning_rate": 8.972163960826375e-06,
+      "loss": 0.7332,
+      "step": 2201
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9694503720386745,
+      "learning_rate": 8.971128657522677e-06,
+      "loss": 0.7163,
+      "step": 2202
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1801548709290546,
+      "learning_rate": 8.970092892864732e-06,
+      "loss": 0.7532,
+      "step": 2203
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2491318122078092,
+      "learning_rate": 8.969056666972874e-06,
+      "loss": 0.716,
+      "step": 2204
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0893548904362413,
+      "learning_rate": 8.968019979967482e-06,
+      "loss": 0.6952,
+      "step": 2205
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2759911727787907,
+      "learning_rate": 8.966982831969001e-06,
+      "loss": 0.7345,
+      "step": 2206
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.33173403077009,
+      "learning_rate": 8.965945223097922e-06,
+      "loss": 0.7495,
+      "step": 2207
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9341550603493136,
+      "learning_rate": 8.964907153474791e-06,
+      "loss": 0.7013,
+      "step": 2208
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1147262266169173,
+      "learning_rate": 8.963868623220208e-06,
+      "loss": 0.6709,
+      "step": 2209
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8936564339718123,
+      "learning_rate": 8.962829632454829e-06,
+      "loss": 0.7317,
+      "step": 2210
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4903063833013555,
+      "learning_rate": 8.961790181299354e-06,
+      "loss": 0.8207,
+      "step": 2211
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2804678844589246,
+      "learning_rate": 8.960750269874552e-06,
+      "loss": 0.6467,
+      "step": 2212
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4641706147225713,
+      "learning_rate": 8.959709898301232e-06,
+      "loss": 0.7512,
+      "step": 2213
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6089124316964005,
+      "learning_rate": 8.958669066700261e-06,
+      "loss": 0.6522,
+      "step": 2214
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.19069049516705,
+      "learning_rate": 8.957627775192564e-06,
+      "loss": 0.7337,
+      "step": 2215
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.500190008138319,
+      "learning_rate": 8.956586023899109e-06,
+      "loss": 0.681,
+      "step": 2216
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.159720426862354,
+      "learning_rate": 8.95554381294093e-06,
+      "loss": 0.7571,
+      "step": 2217
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0686321610309095,
+      "learning_rate": 8.954501142439105e-06,
+      "loss": 0.6577,
+      "step": 2218
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2475898480773995,
+      "learning_rate": 8.953458012514766e-06,
+      "loss": 0.7263,
+      "step": 2219
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.4884995247883603,
+      "learning_rate": 8.952414423289107e-06,
+      "loss": 0.7057,
+      "step": 2220
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.4448700774798064,
+      "learning_rate": 8.951370374883362e-06,
+      "loss": 0.7204,
+      "step": 2221
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.441270491294434,
+      "learning_rate": 8.950325867418831e-06,
+      "loss": 0.7727,
+      "step": 2222
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.310391536836991,
+      "learning_rate": 8.949280901016859e-06,
+      "loss": 0.7408,
+      "step": 2223
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.894271235429747,
+      "learning_rate": 8.94823547579885e-06,
+      "loss": 0.6713,
+      "step": 2224
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.731116389165696,
+      "learning_rate": 8.947189591886255e-06,
+      "loss": 0.7204,
+      "step": 2225
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3805815349526687,
+      "learning_rate": 8.946143249400582e-06,
+      "loss": 0.6696,
+      "step": 2226
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.354183378868526,
+      "learning_rate": 8.945096448463397e-06,
+      "loss": 0.7481,
+      "step": 2227
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.189115002118964,
+      "learning_rate": 8.944049189196308e-06,
+      "loss": 0.7049,
+      "step": 2228
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 15.927462117356988,
+      "learning_rate": 8.943001471720987e-06,
+      "loss": 0.63,
+      "step": 2229
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.588326102390013,
+      "learning_rate": 8.941953296159153e-06,
+      "loss": 0.7003,
+      "step": 2230
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0744521000216753,
+      "learning_rate": 8.940904662632579e-06,
+      "loss": 0.7521,
+      "step": 2231
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.4628646582770877,
+      "learning_rate": 8.939855571263095e-06,
+      "loss": 0.6593,
+      "step": 2232
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.273569477502985,
+      "learning_rate": 8.938806022172578e-06,
+      "loss": 0.6958,
+      "step": 2233
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3321350896927435,
+      "learning_rate": 8.937756015482962e-06,
+      "loss": 0.706,
+      "step": 2234
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.249108717782656,
+      "learning_rate": 8.936705551316238e-06,
+      "loss": 0.5939,
+      "step": 2235
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4287842458282207,
+      "learning_rate": 8.935654629794442e-06,
+      "loss": 0.7158,
+      "step": 2236
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7772840916896127,
+      "learning_rate": 8.934603251039667e-06,
+      "loss": 0.6861,
+      "step": 2237
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.350481813294236,
+      "learning_rate": 8.93355141517406e-06,
+      "loss": 0.7134,
+      "step": 2238
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9394185187015496,
+      "learning_rate": 8.932499122319821e-06,
+      "loss": 0.6629,
+      "step": 2239
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.664123857241033,
+      "learning_rate": 8.931446372599202e-06,
+      "loss": 0.6152,
+      "step": 2240
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.349915878479327,
+      "learning_rate": 8.930393166134507e-06,
+      "loss": 0.6989,
+      "step": 2241
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.206930105187941,
+      "learning_rate": 8.929339503048096e-06,
+      "loss": 0.6799,
+      "step": 2242
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.647423217564447,
+      "learning_rate": 8.92828538346238e-06,
+      "loss": 0.7513,
+      "step": 2243
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4210465633816582,
+      "learning_rate": 8.927230807499824e-06,
+      "loss": 0.6743,
+      "step": 2244
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5877919600311934,
+      "learning_rate": 8.926175775282946e-06,
+      "loss": 0.6994,
+      "step": 2245
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.158209740707754,
+      "learning_rate": 8.925120286934315e-06,
+      "loss": 0.6448,
+      "step": 2246
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3832699071525774,
+      "learning_rate": 8.924064342576554e-06,
+      "loss": 0.6684,
+      "step": 2247
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.400928597949794,
+      "learning_rate": 8.923007942332345e-06,
+      "loss": 0.73,
+      "step": 2248
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9437835667880945,
+      "learning_rate": 8.92195108632441e-06,
+      "loss": 0.7273,
+      "step": 2249
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.870436670961615,
+      "learning_rate": 8.920893774675536e-06,
+      "loss": 0.705,
+      "step": 2250
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.3747630286954555,
+      "learning_rate": 8.919836007508558e-06,
+      "loss": 0.7267,
+      "step": 2251
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.341185883155674,
+      "learning_rate": 8.918777784946364e-06,
+      "loss": 0.712,
+      "step": 2252
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0954214888627747,
+      "learning_rate": 8.917719107111893e-06,
+      "loss": 0.6333,
+      "step": 2253
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2833413585564724,
+      "learning_rate": 8.916659974128144e-06,
+      "loss": 0.6779,
+      "step": 2254
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1835249116285995,
+      "learning_rate": 8.91560038611816e-06,
+      "loss": 0.705,
+      "step": 2255
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.568232696610119,
+      "learning_rate": 8.91454034320504e-06,
+      "loss": 0.6866,
+      "step": 2256
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7059228367154904,
+      "learning_rate": 8.913479845511942e-06,
+      "loss": 0.6874,
+      "step": 2257
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1413067070108274,
+      "learning_rate": 8.912418893162066e-06,
+      "loss": 0.6511,
+      "step": 2258
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.553360298084785,
+      "learning_rate": 8.91135748627867e-06,
+      "loss": 0.6896,
+      "step": 2259
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5267231876409033,
+      "learning_rate": 8.910295624985072e-06,
+      "loss": 0.7684,
+      "step": 2260
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.267588533534899,
+      "learning_rate": 8.909233309404632e-06,
+      "loss": 0.7048,
+      "step": 2261
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.118523741223378,
+      "learning_rate": 8.908170539660766e-06,
+      "loss": 0.7692,
+      "step": 2262
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.7350137185652335,
+      "learning_rate": 8.907107315876942e-06,
+      "loss": 0.5702,
+      "step": 2263
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2334438560399243,
+      "learning_rate": 8.906043638176686e-06,
+      "loss": 0.6064,
+      "step": 2264
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.408881941929093,
+      "learning_rate": 8.90497950668357e-06,
+      "loss": 0.7069,
+      "step": 2265
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.985892002965994,
+      "learning_rate": 8.903914921521226e-06,
+      "loss": 0.6155,
+      "step": 2266
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0808888218270587,
+      "learning_rate": 8.90284988281333e-06,
+      "loss": 0.6486,
+      "step": 2267
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.235696390043599,
+      "learning_rate": 8.901784390683616e-06,
+      "loss": 0.7458,
+      "step": 2268
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.545711880315011,
+      "learning_rate": 8.90071844525587e-06,
+      "loss": 0.6872,
+      "step": 2269
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1494684836406015,
+      "learning_rate": 8.89965204665393e-06,
+      "loss": 0.6339,
+      "step": 2270
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2914981848806717,
+      "learning_rate": 8.898585195001691e-06,
+      "loss": 0.6513,
+      "step": 2271
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6252182724430613,
+      "learning_rate": 8.897517890423092e-06,
+      "loss": 0.6416,
+      "step": 2272
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3145266443168397,
+      "learning_rate": 8.896450133042132e-06,
+      "loss": 0.5625,
+      "step": 2273
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9425490217808346,
+      "learning_rate": 8.895381922982857e-06,
+      "loss": 0.7486,
+      "step": 2274
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4409922366780656,
+      "learning_rate": 8.894313260369372e-06,
+      "loss": 0.7289,
+      "step": 2275
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7296566582591337,
+      "learning_rate": 8.89324414532583e-06,
+      "loss": 0.652,
+      "step": 2276
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7073730009269488,
+      "learning_rate": 8.892174577976438e-06,
+      "loss": 0.6029,
+      "step": 2277
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.984224980813416,
+      "learning_rate": 8.891104558445454e-06,
+      "loss": 0.6809,
+      "step": 2278
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5742027243121153,
+      "learning_rate": 8.890034086857189e-06,
+      "loss": 0.6906,
+      "step": 2279
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0483532696980427,
+      "learning_rate": 8.88896316333601e-06,
+      "loss": 0.7207,
+      "step": 2280
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7333128521952657,
+      "learning_rate": 8.887891788006334e-06,
+      "loss": 0.7079,
+      "step": 2281
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.860179858149371,
+      "learning_rate": 8.886819960992626e-06,
+      "loss": 0.6179,
+      "step": 2282
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.398502797457853,
+      "learning_rate": 8.885747682419413e-06,
+      "loss": 0.6413,
+      "step": 2283
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3814307011526807,
+      "learning_rate": 8.884674952411265e-06,
+      "loss": 0.815,
+      "step": 2284
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.91263925062964,
+      "learning_rate": 8.883601771092812e-06,
+      "loss": 0.6907,
+      "step": 2285
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.470495388049324,
+      "learning_rate": 8.882528138588729e-06,
+      "loss": 0.7293,
+      "step": 2286
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.206151479410179,
+      "learning_rate": 8.881454055023752e-06,
+      "loss": 0.7459,
+      "step": 2287
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.432016434550301,
+      "learning_rate": 8.880379520522664e-06,
+      "loss": 0.762,
+      "step": 2288
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.8140952416127694,
+      "learning_rate": 8.879304535210298e-06,
+      "loss": 0.6734,
+      "step": 2289
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.473443750742672,
+      "learning_rate": 8.878229099211548e-06,
+      "loss": 0.75,
+      "step": 2290
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.536037475518659,
+      "learning_rate": 8.87715321265135e-06,
+      "loss": 0.6536,
+      "step": 2291
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.268832339589105,
+      "learning_rate": 8.8760768756547e-06,
+      "loss": 0.7971,
+      "step": 2292
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.784421921853551,
+      "learning_rate": 8.875000088346642e-06,
+      "loss": 0.7665,
+      "step": 2293
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.494207647604574,
+      "learning_rate": 8.873922850852276e-06,
+      "loss": 0.5876,
+      "step": 2294
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.963714518286977,
+      "learning_rate": 8.872845163296752e-06,
+      "loss": 0.7046,
+      "step": 2295
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1676492228416104,
+      "learning_rate": 8.87176702580527e-06,
+      "loss": 0.6656,
+      "step": 2296
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5119192830739734,
+      "learning_rate": 8.87068843850309e-06,
+      "loss": 0.7083,
+      "step": 2297
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.943714911949743,
+      "learning_rate": 8.869609401515516e-06,
+      "loss": 0.6959,
+      "step": 2298
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6887677900783884,
+      "learning_rate": 8.868529914967908e-06,
+      "loss": 0.6394,
+      "step": 2299
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2402749666716884,
+      "learning_rate": 8.867449978985676e-06,
+      "loss": 0.7139,
+      "step": 2300
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.8829424956949716,
+      "learning_rate": 8.866369593694285e-06,
+      "loss": 0.6697,
+      "step": 2301
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.746440356252218,
+      "learning_rate": 8.865288759219251e-06,
+      "loss": 0.7202,
+      "step": 2302
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1527046705438075,
+      "learning_rate": 8.864207475686142e-06,
+      "loss": 0.724,
+      "step": 2303
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.95970505604412,
+      "learning_rate": 8.86312574322058e-06,
+      "loss": 0.7037,
+      "step": 2304
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.271026223729123,
+      "learning_rate": 8.862043561948237e-06,
+      "loss": 0.6285,
+      "step": 2305
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8402186327283752,
+      "learning_rate": 8.860960931994835e-06,
+      "loss": 0.6876,
+      "step": 2306
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.098504551725227,
+      "learning_rate": 8.859877853486154e-06,
+      "loss": 0.7273,
+      "step": 2307
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2831168072415884,
+      "learning_rate": 8.85879432654802e-06,
+      "loss": 0.7357,
+      "step": 2308
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.372416894210338,
+      "learning_rate": 8.85771035130632e-06,
+      "loss": 0.6096,
+      "step": 2309
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.296320517752463,
+      "learning_rate": 8.85662592788698e-06,
+      "loss": 0.6875,
+      "step": 2310
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.8593869238201894,
+      "learning_rate": 8.855541056415988e-06,
+      "loss": 0.7534,
+      "step": 2311
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.688210282987243,
+      "learning_rate": 8.854455737019381e-06,
+      "loss": 0.7085,
+      "step": 2312
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2398910749922645,
+      "learning_rate": 8.853369969823249e-06,
+      "loss": 0.7067,
+      "step": 2313
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5141437787544123,
+      "learning_rate": 8.852283754953734e-06,
+      "loss": 0.6465,
+      "step": 2314
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.3203384815403836,
+      "learning_rate": 8.851197092537027e-06,
+      "loss": 0.6464,
+      "step": 2315
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.071240596546394,
+      "learning_rate": 8.850109982699375e-06,
+      "loss": 0.7057,
+      "step": 2316
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.27586745399632,
+      "learning_rate": 8.849022425567074e-06,
+      "loss": 0.5833,
+      "step": 2317
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.242964035717761,
+      "learning_rate": 8.847934421266475e-06,
+      "loss": 0.6137,
+      "step": 2318
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.005217367036941,
+      "learning_rate": 8.846845969923977e-06,
+      "loss": 0.6083,
+      "step": 2319
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9145747544167695,
+      "learning_rate": 8.845757071666035e-06,
+      "loss": 0.6655,
+      "step": 2320
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0295015098864075,
+      "learning_rate": 8.844667726619153e-06,
+      "loss": 0.8086,
+      "step": 2321
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2677233953465716,
+      "learning_rate": 8.843577934909888e-06,
+      "loss": 0.7134,
+      "step": 2322
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0387979236232696,
+      "learning_rate": 8.84248769666485e-06,
+      "loss": 0.6452,
+      "step": 2323
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3092678024403184,
+      "learning_rate": 8.8413970120107e-06,
+      "loss": 0.6905,
+      "step": 2324
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.441938437783016,
+      "learning_rate": 8.840305881074147e-06,
+      "loss": 0.7386,
+      "step": 2325
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.299488064549423,
+      "learning_rate": 8.83921430398196e-06,
+      "loss": 0.5852,
+      "step": 2326
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.346680432521479,
+      "learning_rate": 8.838122280860953e-06,
+      "loss": 0.6816,
+      "step": 2327
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1971116175516738,
+      "learning_rate": 8.837029811837991e-06,
+      "loss": 0.6475,
+      "step": 2328
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.033820645161564,
+      "learning_rate": 8.83593689704e-06,
+      "loss": 0.6569,
+      "step": 2329
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2373130942751507,
+      "learning_rate": 8.834843536593949e-06,
+      "loss": 0.7147,
+      "step": 2330
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.462343548899127,
+      "learning_rate": 8.833749730626862e-06,
+      "loss": 0.6969,
+      "step": 2331
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.597236651843794,
+      "learning_rate": 8.832655479265812e-06,
+      "loss": 0.6882,
+      "step": 2332
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2374268550953933,
+      "learning_rate": 8.831560782637929e-06,
+      "loss": 0.562,
+      "step": 2333
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.383491282270233,
+      "learning_rate": 8.830465640870388e-06,
+      "loss": 0.7145,
+      "step": 2334
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1925762438333,
+      "learning_rate": 8.829370054090423e-06,
+      "loss": 0.6832,
+      "step": 2335
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.41126511819354,
+      "learning_rate": 8.828274022425316e-06,
+      "loss": 0.7105,
+      "step": 2336
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.863081103637933,
+      "learning_rate": 8.827177546002398e-06,
+      "loss": 0.7167,
+      "step": 2337
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3220167881162324,
+      "learning_rate": 8.826080624949056e-06,
+      "loss": 0.6032,
+      "step": 2338
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4710434596501796,
+      "learning_rate": 8.824983259392727e-06,
+      "loss": 0.6848,
+      "step": 2339
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0728765590634004,
+      "learning_rate": 8.823885449460899e-06,
+      "loss": 0.6752,
+      "step": 2340
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.173323985472954,
+      "learning_rate": 8.822787195281114e-06,
+      "loss": 0.6507,
+      "step": 2341
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.022192983168258,
+      "learning_rate": 8.821688496980964e-06,
+      "loss": 0.5687,
+      "step": 2342
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.321047814509788,
+      "learning_rate": 8.82058935468809e-06,
+      "loss": 0.7194,
+      "step": 2343
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4035818387743038,
+      "learning_rate": 8.819489768530192e-06,
+      "loss": 0.6671,
+      "step": 2344
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2287384454438013,
+      "learning_rate": 8.818389738635012e-06,
+      "loss": 0.6707,
+      "step": 2345
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1051049030090847,
+      "learning_rate": 8.817289265130348e-06,
+      "loss": 0.7092,
+      "step": 2346
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6891874499459885,
+      "learning_rate": 8.816188348144054e-06,
+      "loss": 0.7263,
+      "step": 2347
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8695048876833085,
+      "learning_rate": 8.815086987804029e-06,
+      "loss": 0.661,
+      "step": 2348
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.307276114427358,
+      "learning_rate": 8.813985184238226e-06,
+      "loss": 0.7653,
+      "step": 2349
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.273797735104588,
+      "learning_rate": 8.81288293757465e-06,
+      "loss": 0.6886,
+      "step": 2350
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8942145715992846,
+      "learning_rate": 8.811780247941354e-06,
+      "loss": 0.7078,
+      "step": 2351
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3063601691994826,
+      "learning_rate": 8.810677115466451e-06,
+      "loss": 0.677,
+      "step": 2352
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1245376854410516,
+      "learning_rate": 8.809573540278094e-06,
+      "loss": 0.6966,
+      "step": 2353
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8737504795386495,
+      "learning_rate": 8.808469522504495e-06,
+      "loss": 0.6428,
+      "step": 2354
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.578969771967044,
+      "learning_rate": 8.807365062273917e-06,
+      "loss": 0.7379,
+      "step": 2355
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.771616985464036,
+      "learning_rate": 8.806260159714672e-06,
+      "loss": 0.5943,
+      "step": 2356
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0923888169384046,
+      "learning_rate": 8.805154814955124e-06,
+      "loss": 0.6965,
+      "step": 2357
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4647946308004243,
+      "learning_rate": 8.80404902812369e-06,
+      "loss": 0.6628,
+      "step": 2358
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8740732083977183,
+      "learning_rate": 8.802942799348836e-06,
+      "loss": 0.7136,
+      "step": 2359
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.023161383444029,
+      "learning_rate": 8.80183612875908e-06,
+      "loss": 0.6532,
+      "step": 2360
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2951972055338783,
+      "learning_rate": 8.800729016482993e-06,
+      "loss": 0.7056,
+      "step": 2361
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.005412561447531,
+      "learning_rate": 8.799621462649198e-06,
+      "loss": 0.6511,
+      "step": 2362
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.347273517470351,
+      "learning_rate": 8.798513467386361e-06,
+      "loss": 0.6972,
+      "step": 2363
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2061420190350576,
+      "learning_rate": 8.797405030823212e-06,
+      "loss": 0.6577,
+      "step": 2364
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.209573303684799,
+      "learning_rate": 8.796296153088523e-06,
+      "loss": 0.6624,
+      "step": 2365
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0157084559249987,
+      "learning_rate": 8.79518683431112e-06,
+      "loss": 0.6218,
+      "step": 2366
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6605469423214507,
+      "learning_rate": 8.794077074619884e-06,
+      "loss": 0.6776,
+      "step": 2367
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.329846735401453,
+      "learning_rate": 8.79296687414374e-06,
+      "loss": 0.6263,
+      "step": 2368
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.16211910196471,
+      "learning_rate": 8.791856233011668e-06,
+      "loss": 0.7662,
+      "step": 2369
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5132087358973325,
+      "learning_rate": 8.7907451513527e-06,
+      "loss": 0.5662,
+      "step": 2370
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4691914644939135,
+      "learning_rate": 8.78963362929592e-06,
+      "loss": 0.6809,
+      "step": 2371
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.517049664768459,
+      "learning_rate": 8.788521666970458e-06,
+      "loss": 0.6717,
+      "step": 2372
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.083726661084472,
+      "learning_rate": 8.7874092645055e-06,
+      "loss": 0.6727,
+      "step": 2373
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.259534102965764,
+      "learning_rate": 8.786296422030283e-06,
+      "loss": 0.7514,
+      "step": 2374
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.0201738332244945,
+      "learning_rate": 8.785183139674093e-06,
+      "loss": 0.7253,
+      "step": 2375
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7540666834409913,
+      "learning_rate": 8.784069417566268e-06,
+      "loss": 0.6923,
+      "step": 2376
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4727794350012213,
+      "learning_rate": 8.782955255836194e-06,
+      "loss": 0.6761,
+      "step": 2377
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.642090445706794,
+      "learning_rate": 8.781840654613317e-06,
+      "loss": 0.6928,
+      "step": 2378
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3378650663235514,
+      "learning_rate": 8.780725614027123e-06,
+      "loss": 0.6407,
+      "step": 2379
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.5463507569750945,
+      "learning_rate": 8.779610134207157e-06,
+      "loss": 0.731,
+      "step": 2380
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.476687518226198,
+      "learning_rate": 8.778494215283011e-06,
+      "loss": 0.6693,
+      "step": 2381
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6190587133475143,
+      "learning_rate": 8.777377857384329e-06,
+      "loss": 0.7043,
+      "step": 2382
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.276304071392182,
+      "learning_rate": 8.776261060640807e-06,
+      "loss": 0.6189,
+      "step": 2383
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5067148512484305,
+      "learning_rate": 8.775143825182192e-06,
+      "loss": 0.6429,
+      "step": 2384
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.533233559475545,
+      "learning_rate": 8.77402615113828e-06,
+      "loss": 0.7218,
+      "step": 2385
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3752543251935956,
+      "learning_rate": 8.77290803863892e-06,
+      "loss": 0.6615,
+      "step": 2386
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6509057573466417,
+      "learning_rate": 8.771789487814009e-06,
+      "loss": 0.6687,
+      "step": 2387
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.322946052629726,
+      "learning_rate": 8.770670498793498e-06,
+      "loss": 0.71,
+      "step": 2388
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.555317772836953,
+      "learning_rate": 8.76955107170739e-06,
+      "loss": 0.6629,
+      "step": 2389
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0462771901928227,
+      "learning_rate": 8.768431206685735e-06,
+      "loss": 0.6768,
+      "step": 2390
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4959674203980957,
+      "learning_rate": 8.767310903858635e-06,
+      "loss": 0.7051,
+      "step": 2391
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.400493851416112,
+      "learning_rate": 8.766190163356243e-06,
+      "loss": 0.7141,
+      "step": 2392
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.447160327467119,
+      "learning_rate": 8.765068985308768e-06,
+      "loss": 0.6782,
+      "step": 2393
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.205789371123814,
+      "learning_rate": 8.76394736984646e-06,
+      "loss": 0.6606,
+      "step": 2394
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2524271530733375,
+      "learning_rate": 8.762825317099628e-06,
+      "loss": 0.623,
+      "step": 2395
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1224476313578493,
+      "learning_rate": 8.761702827198626e-06,
+      "loss": 0.6808,
+      "step": 2396
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2653362778423383,
+      "learning_rate": 8.760579900273865e-06,
+      "loss": 0.636,
+      "step": 2397
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2781604321793023,
+      "learning_rate": 8.759456536455802e-06,
+      "loss": 0.6624,
+      "step": 2398
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4477894582450905,
+      "learning_rate": 8.758332735874946e-06,
+      "loss": 0.6002,
+      "step": 2399
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.197585400074303,
+      "learning_rate": 8.757208498661857e-06,
+      "loss": 0.6718,
+      "step": 2400
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.278739347444417,
+      "learning_rate": 8.756083824947145e-06,
+      "loss": 0.7983,
+      "step": 2401
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.585442200635953,
+      "learning_rate": 8.754958714861474e-06,
+      "loss": 0.7644,
+      "step": 2402
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0314353743944635,
+      "learning_rate": 8.753833168535551e-06,
+      "loss": 0.7351,
+      "step": 2403
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7580999310075356,
+      "learning_rate": 8.752707186100144e-06,
+      "loss": 0.6929,
+      "step": 2404
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.720900436087719,
+      "learning_rate": 8.751580767686063e-06,
+      "loss": 0.6816,
+      "step": 2405
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7870764847387455,
+      "learning_rate": 8.750453913424172e-06,
+      "loss": 0.6466,
+      "step": 2406
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5240294889336234,
+      "learning_rate": 8.74932662344539e-06,
+      "loss": 0.6982,
+      "step": 2407
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 12.532901910391917,
+      "learning_rate": 8.748198897880677e-06,
+      "loss": 0.68,
+      "step": 2408
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2585279019476845,
+      "learning_rate": 8.747070736861052e-06,
+      "loss": 0.6859,
+      "step": 2409
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.950693720936932,
+      "learning_rate": 8.745942140517579e-06,
+      "loss": 0.6846,
+      "step": 2410
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8818540494898084,
+      "learning_rate": 8.744813108981377e-06,
+      "loss": 0.7454,
+      "step": 2411
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.03077990529205,
+      "learning_rate": 8.743683642383613e-06,
+      "loss": 0.7515,
+      "step": 2412
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3199732724761954,
+      "learning_rate": 8.742553740855507e-06,
+      "loss": 0.6335,
+      "step": 2413
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5030676274429786,
+      "learning_rate": 8.741423404528325e-06,
+      "loss": 0.6373,
+      "step": 2414
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.576788831727748,
+      "learning_rate": 8.740292633533387e-06,
+      "loss": 0.6511,
+      "step": 2415
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.292929750954021,
+      "learning_rate": 8.739161428002061e-06,
+      "loss": 0.6665,
+      "step": 2416
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.495581313970782,
+      "learning_rate": 8.738029788065772e-06,
+      "loss": 0.6986,
+      "step": 2417
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7267216304430106,
+      "learning_rate": 8.736897713855988e-06,
+      "loss": 0.6805,
+      "step": 2418
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5310007726717654,
+      "learning_rate": 8.735765205504228e-06,
+      "loss": 0.6876,
+      "step": 2419
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2473961744466824,
+      "learning_rate": 8.734632263142066e-06,
+      "loss": 0.626,
+      "step": 2420
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6025083204104194,
+      "learning_rate": 8.733498886901123e-06,
+      "loss": 0.6504,
+      "step": 2421
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.885692680206437,
+      "learning_rate": 8.732365076913072e-06,
+      "loss": 0.6555,
+      "step": 2422
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.032356607171435,
+      "learning_rate": 8.731230833309637e-06,
+      "loss": 0.6401,
+      "step": 2423
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1972012220479074,
+      "learning_rate": 8.730096156222586e-06,
+      "loss": 0.6757,
+      "step": 2424
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.291089360356186,
+      "learning_rate": 8.728961045783751e-06,
+      "loss": 0.7047,
+      "step": 2425
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2487186243095816,
+      "learning_rate": 8.727825502124998e-06,
+      "loss": 0.7178,
+      "step": 2426
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.417453807033552,
+      "learning_rate": 8.726689525378254e-06,
+      "loss": 0.6669,
+      "step": 2427
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3540762968676656,
+      "learning_rate": 8.725553115675496e-06,
+      "loss": 0.5727,
+      "step": 2428
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.006417010869224,
+      "learning_rate": 8.724416273148745e-06,
+      "loss": 0.6692,
+      "step": 2429
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.24813269883343,
+      "learning_rate": 8.723278997930078e-06,
+      "loss": 0.6947,
+      "step": 2430
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.568072157207673,
+      "learning_rate": 8.722141290151618e-06,
+      "loss": 0.6721,
+      "step": 2431
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.681084669318892,
+      "learning_rate": 8.721003149945545e-06,
+      "loss": 0.7217,
+      "step": 2432
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5296915450142516,
+      "learning_rate": 8.719864577444082e-06,
+      "loss": 0.6596,
+      "step": 2433
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.750107096539297,
+      "learning_rate": 8.718725572779505e-06,
+      "loss": 0.835,
+      "step": 2434
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.444233296643607,
+      "learning_rate": 8.71758613608414e-06,
+      "loss": 0.7186,
+      "step": 2435
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.270574624183063,
+      "learning_rate": 8.716446267490365e-06,
+      "loss": 0.7578,
+      "step": 2436
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.8389444303037235,
+      "learning_rate": 8.715305967130604e-06,
+      "loss": 0.734,
+      "step": 2437
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1979511502659714,
+      "learning_rate": 8.71416523513734e-06,
+      "loss": 0.6708,
+      "step": 2438
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8417444339853852,
+      "learning_rate": 8.713024071643092e-06,
+      "loss": 0.6287,
+      "step": 2439
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0558749007403,
+      "learning_rate": 8.71188247678044e-06,
+      "loss": 0.7621,
+      "step": 2440
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.291175152328059,
+      "learning_rate": 8.710740450682013e-06,
+      "loss": 0.6667,
+      "step": 2441
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.141308772601036,
+      "learning_rate": 8.709597993480489e-06,
+      "loss": 0.6674,
+      "step": 2442
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0735451596302137,
+      "learning_rate": 8.708455105308591e-06,
+      "loss": 0.6599,
+      "step": 2443
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.481850558377454,
+      "learning_rate": 8.7073117862991e-06,
+      "loss": 0.783,
+      "step": 2444
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9268542381826397,
+      "learning_rate": 8.706168036584843e-06,
+      "loss": 0.6356,
+      "step": 2445
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.349321958616555,
+      "learning_rate": 8.705023856298695e-06,
+      "loss": 0.611,
+      "step": 2446
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1941989903979624,
+      "learning_rate": 8.703879245573588e-06,
+      "loss": 0.6716,
+      "step": 2447
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1349012651457664,
+      "learning_rate": 8.702734204542494e-06,
+      "loss": 0.7377,
+      "step": 2448
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.648744197677024,
+      "learning_rate": 8.701588733338446e-06,
+      "loss": 0.7413,
+      "step": 2449
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.3362813946257623,
+      "learning_rate": 8.700442832094517e-06,
+      "loss": 0.5947,
+      "step": 2450
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0804046111905614,
+      "learning_rate": 8.699296500943839e-06,
+      "loss": 0.6786,
+      "step": 2451
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.661217136166758,
+      "learning_rate": 8.698149740019587e-06,
+      "loss": 0.7303,
+      "step": 2452
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5190865956506547,
+      "learning_rate": 8.697002549454988e-06,
+      "loss": 0.7269,
+      "step": 2453
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.070053475596138,
+      "learning_rate": 8.695854929383318e-06,
+      "loss": 0.7228,
+      "step": 2454
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.441251249535768,
+      "learning_rate": 8.694706879937909e-06,
+      "loss": 0.6955,
+      "step": 2455
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2612646874942746,
+      "learning_rate": 8.693558401252132e-06,
+      "loss": 0.656,
+      "step": 2456
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4863160385294063,
+      "learning_rate": 8.69240949345942e-06,
+      "loss": 0.7101,
+      "step": 2457
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.43116626808894,
+      "learning_rate": 8.691260156693245e-06,
+      "loss": 0.6744,
+      "step": 2458
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5200768749768963,
+      "learning_rate": 8.690110391087134e-06,
+      "loss": 0.6949,
+      "step": 2459
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.525389539161813,
+      "learning_rate": 8.688960196774668e-06,
+      "loss": 0.6323,
+      "step": 2460
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.102109783358811,
+      "learning_rate": 8.687809573889467e-06,
+      "loss": 0.6833,
+      "step": 2461
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.222074030086465,
+      "learning_rate": 8.686658522565211e-06,
+      "loss": 0.7645,
+      "step": 2462
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6743137482807753,
+      "learning_rate": 8.685507042935627e-06,
+      "loss": 0.7335,
+      "step": 2463
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8113533406525595,
+      "learning_rate": 8.684355135134486e-06,
+      "loss": 0.6922,
+      "step": 2464
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9458115809108403,
+      "learning_rate": 8.683202799295616e-06,
+      "loss": 0.6556,
+      "step": 2465
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1842127186245333,
+      "learning_rate": 8.682050035552891e-06,
+      "loss": 0.6102,
+      "step": 2466
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1705972141925733,
+      "learning_rate": 8.680896844040238e-06,
+      "loss": 0.706,
+      "step": 2467
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2360348544184308,
+      "learning_rate": 8.67974322489163e-06,
+      "loss": 0.7269,
+      "step": 2468
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.569034689616941,
+      "learning_rate": 8.678589178241092e-06,
+      "loss": 0.737,
+      "step": 2469
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.729945172388795,
+      "learning_rate": 8.677434704222697e-06,
+      "loss": 0.7642,
+      "step": 2470
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.577038874216306,
+      "learning_rate": 8.676279802970566e-06,
+      "loss": 0.6741,
+      "step": 2471
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.477474813381419,
+      "learning_rate": 8.675124474618876e-06,
+      "loss": 0.6524,
+      "step": 2472
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.281055787854727,
+      "learning_rate": 8.673968719301849e-06,
+      "loss": 0.6976,
+      "step": 2473
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.506648427439492,
+      "learning_rate": 8.672812537153757e-06,
+      "loss": 0.6998,
+      "step": 2474
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.568840294148294,
+      "learning_rate": 8.67165592830892e-06,
+      "loss": 0.692,
+      "step": 2475
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.257378434912131,
+      "learning_rate": 8.670498892901712e-06,
+      "loss": 0.7227,
+      "step": 2476
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9597763138644833,
+      "learning_rate": 8.669341431066552e-06,
+      "loss": 0.6658,
+      "step": 2477
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4823038278036846,
+      "learning_rate": 8.668183542937912e-06,
+      "loss": 0.6355,
+      "step": 2478
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2040856463646663,
+      "learning_rate": 8.66702522865031e-06,
+      "loss": 0.6359,
+      "step": 2479
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7952172523154166,
+      "learning_rate": 8.66586648833832e-06,
+      "loss": 0.684,
+      "step": 2480
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.400485734318114,
+      "learning_rate": 8.664707322136556e-06,
+      "loss": 0.6433,
+      "step": 2481
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.306795903415676,
+      "learning_rate": 8.663547730179692e-06,
+      "loss": 0.651,
+      "step": 2482
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.985126814934233,
+      "learning_rate": 8.662387712602438e-06,
+      "loss": 0.6954,
+      "step": 2483
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2505071915852075,
+      "learning_rate": 8.661227269539572e-06,
+      "loss": 0.6472,
+      "step": 2484
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8826833418351787,
+      "learning_rate": 8.660066401125902e-06,
+      "loss": 0.6311,
+      "step": 2485
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.818214783238789,
+      "learning_rate": 8.658905107496299e-06,
+      "loss": 0.6617,
+      "step": 2486
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6246974001427716,
+      "learning_rate": 8.657743388785676e-06,
+      "loss": 0.7234,
+      "step": 2487
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5053395649894235,
+      "learning_rate": 8.656581245129e-06,
+      "loss": 0.7851,
+      "step": 2488
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5788879956613666,
+      "learning_rate": 8.655418676661285e-06,
+      "loss": 0.6737,
+      "step": 2489
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5695918451833752,
+      "learning_rate": 8.654255683517595e-06,
+      "loss": 0.66,
+      "step": 2490
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2259903673736456,
+      "learning_rate": 8.653092265833044e-06,
+      "loss": 0.7476,
+      "step": 2491
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0361177334211376,
+      "learning_rate": 8.651928423742793e-06,
+      "loss": 0.6448,
+      "step": 2492
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2871956578155253,
+      "learning_rate": 8.650764157382054e-06,
+      "loss": 0.6647,
+      "step": 2493
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.278582756268501,
+      "learning_rate": 8.64959946688609e-06,
+      "loss": 0.663,
+      "step": 2494
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5143339131874147,
+      "learning_rate": 8.648434352390209e-06,
+      "loss": 0.6837,
+      "step": 2495
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4594476956792586,
+      "learning_rate": 8.64726881402977e-06,
+      "loss": 0.6644,
+      "step": 2496
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1170695576353,
+      "learning_rate": 8.646102851940184e-06,
+      "loss": 0.766,
+      "step": 2497
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2169468874210065,
+      "learning_rate": 8.64493646625691e-06,
+      "loss": 0.664,
+      "step": 2498
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0516103481287034,
+      "learning_rate": 8.643769657115452e-06,
+      "loss": 0.6429,
+      "step": 2499
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.027841909715622,
+      "learning_rate": 8.642602424651369e-06,
+      "loss": 0.7947,
+      "step": 2500
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.292377856177173,
+      "learning_rate": 8.641434769000267e-06,
+      "loss": 0.6661,
+      "step": 2501
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.591736810360195,
+      "learning_rate": 8.640266690297797e-06,
+      "loss": 0.5915,
+      "step": 2502
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.751465380246021,
+      "learning_rate": 8.639098188679668e-06,
+      "loss": 0.6897,
+      "step": 2503
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0621395224891077,
+      "learning_rate": 8.637929264281632e-06,
+      "loss": 0.6687,
+      "step": 2504
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.16826930507742,
+      "learning_rate": 8.63675991723949e-06,
+      "loss": 0.6106,
+      "step": 2505
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.7332955065147577,
+      "learning_rate": 8.635590147689092e-06,
+      "loss": 0.7138,
+      "step": 2506
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7397582619681597,
+      "learning_rate": 8.634419955766342e-06,
+      "loss": 0.6154,
+      "step": 2507
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6765754635699213,
+      "learning_rate": 8.633249341607186e-06,
+      "loss": 0.6402,
+      "step": 2508
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.147939650717752,
+      "learning_rate": 8.632078305347623e-06,
+      "loss": 0.6187,
+      "step": 2509
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5410600160091876,
+      "learning_rate": 8.630906847123704e-06,
+      "loss": 0.7342,
+      "step": 2510
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2309952636873485,
+      "learning_rate": 8.629734967071522e-06,
+      "loss": 0.6337,
+      "step": 2511
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.221441936274778,
+      "learning_rate": 8.628562665327224e-06,
+      "loss": 0.6374,
+      "step": 2512
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4825649248711654,
+      "learning_rate": 8.627389942027008e-06,
+      "loss": 0.7285,
+      "step": 2513
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.3071050264366133,
+      "learning_rate": 8.62621679730711e-06,
+      "loss": 0.5995,
+      "step": 2514
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9177350717101802,
+      "learning_rate": 8.62504323130383e-06,
+      "loss": 0.7131,
+      "step": 2515
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.510316012626707,
+      "learning_rate": 8.623869244153504e-06,
+      "loss": 0.7768,
+      "step": 2516
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.497041444497354,
+      "learning_rate": 8.622694835992525e-06,
+      "loss": 0.6689,
+      "step": 2517
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.564165380926844,
+      "learning_rate": 8.621520006957334e-06,
+      "loss": 0.7526,
+      "step": 2518
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6172118910010704,
+      "learning_rate": 8.620344757184415e-06,
+      "loss": 0.6911,
+      "step": 2519
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2241203927304367,
+      "learning_rate": 8.619169086810308e-06,
+      "loss": 0.7166,
+      "step": 2520
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.3008760240604422,
+      "learning_rate": 8.6179929959716e-06,
+      "loss": 0.7909,
+      "step": 2521
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.3505911458947626,
+      "learning_rate": 8.616816484804922e-06,
+      "loss": 0.7658,
+      "step": 2522
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.268553301879235,
+      "learning_rate": 8.615639553446961e-06,
+      "loss": 0.642,
+      "step": 2523
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9320366440133188,
+      "learning_rate": 8.614462202034449e-06,
+      "loss": 0.6443,
+      "step": 2524
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5459832781821428,
+      "learning_rate": 8.613284430704165e-06,
+      "loss": 0.662,
+      "step": 2525
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4781573079223422,
+      "learning_rate": 8.612106239592944e-06,
+      "loss": 0.6541,
+      "step": 2526
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.971517376818193,
+      "learning_rate": 8.610927628837658e-06,
+      "loss": 0.6708,
+      "step": 2527
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.465844796112776,
+      "learning_rate": 8.60974859857524e-06,
+      "loss": 0.6641,
+      "step": 2528
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3666940157342347,
+      "learning_rate": 8.608569148942664e-06,
+      "loss": 0.6648,
+      "step": 2529
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.3906416500624106,
+      "learning_rate": 8.607389280076956e-06,
+      "loss": 0.7074,
+      "step": 2530
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.55985900963281,
+      "learning_rate": 8.606208992115191e-06,
+      "loss": 0.6741,
+      "step": 2531
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.088312059952641,
+      "learning_rate": 8.605028285194487e-06,
+      "loss": 0.6705,
+      "step": 2532
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.4291577802630435,
+      "learning_rate": 8.60384715945202e-06,
+      "loss": 0.7239,
+      "step": 2533
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.431296481740759,
+      "learning_rate": 8.602665615025006e-06,
+      "loss": 0.6131,
+      "step": 2534
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9635644510309531,
+      "learning_rate": 8.601483652050717e-06,
+      "loss": 0.5776,
+      "step": 2535
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.460305744619151,
+      "learning_rate": 8.600301270666467e-06,
+      "loss": 0.7602,
+      "step": 2536
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2925216721758674,
+      "learning_rate": 8.599118471009622e-06,
+      "loss": 0.6207,
+      "step": 2537
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4578397320289453,
+      "learning_rate": 8.597935253217598e-06,
+      "loss": 0.7284,
+      "step": 2538
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.148544072195046,
+      "learning_rate": 8.596751617427856e-06,
+      "loss": 0.609,
+      "step": 2539
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.5641003536407068,
+      "learning_rate": 8.595567563777909e-06,
+      "loss": 0.6304,
+      "step": 2540
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5993299217825125,
+      "learning_rate": 8.594383092405317e-06,
+      "loss": 0.773,
+      "step": 2541
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.261057636960014,
+      "learning_rate": 8.593198203447685e-06,
+      "loss": 0.6816,
+      "step": 2542
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7270686627916243,
+      "learning_rate": 8.592012897042677e-06,
+      "loss": 0.6008,
+      "step": 2543
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.593307670860837,
+      "learning_rate": 8.59082717332799e-06,
+      "loss": 0.7299,
+      "step": 2544
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.701277864197115,
+      "learning_rate": 8.589641032441384e-06,
+      "loss": 0.6848,
+      "step": 2545
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.085418911458651,
+      "learning_rate": 8.588454474520657e-06,
+      "loss": 0.6909,
+      "step": 2546
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.8232006350273355,
+      "learning_rate": 8.587267499703667e-06,
+      "loss": 0.6507,
+      "step": 2547
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4387883280629845,
+      "learning_rate": 8.586080108128304e-06,
+      "loss": 0.7263,
+      "step": 2548
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2908278171042737,
+      "learning_rate": 8.584892299932523e-06,
+      "loss": 0.6689,
+      "step": 2549
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.573968919899555,
+      "learning_rate": 8.583704075254315e-06,
+      "loss": 0.7485,
+      "step": 2550
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9765316518612366,
+      "learning_rate": 8.582515434231729e-06,
+      "loss": 0.6095,
+      "step": 2551
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1051643768949975,
+      "learning_rate": 8.581326377002857e-06,
+      "loss": 0.6412,
+      "step": 2552
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0858357891432138,
+      "learning_rate": 8.580136903705838e-06,
+      "loss": 0.7277,
+      "step": 2553
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3915626172098765,
+      "learning_rate": 8.578947014478861e-06,
+      "loss": 0.6705,
+      "step": 2554
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.234544105188988,
+      "learning_rate": 8.577756709460167e-06,
+      "loss": 0.6751,
+      "step": 2555
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.99895822558418,
+      "learning_rate": 8.576565988788042e-06,
+      "loss": 0.6858,
+      "step": 2556
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.176379192047237,
+      "learning_rate": 8.575374852600816e-06,
+      "loss": 0.7022,
+      "step": 2557
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.882909865354107,
+      "learning_rate": 8.574183301036877e-06,
+      "loss": 0.7089,
+      "step": 2558
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6446090141728487,
+      "learning_rate": 8.572991334234654e-06,
+      "loss": 0.7491,
+      "step": 2559
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.631308560093212,
+      "learning_rate": 8.571798952332625e-06,
+      "loss": 0.7464,
+      "step": 2560
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3075799632747174,
+      "learning_rate": 8.57060615546932e-06,
+      "loss": 0.6408,
+      "step": 2561
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.280083800016366,
+      "learning_rate": 8.569412943783313e-06,
+      "loss": 0.6251,
+      "step": 2562
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.37287864006488,
+      "learning_rate": 8.56821931741323e-06,
+      "loss": 0.609,
+      "step": 2563
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5322472513374734,
+      "learning_rate": 8.567025276497739e-06,
+      "loss": 0.7268,
+      "step": 2564
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4974087371265776,
+      "learning_rate": 8.565830821175563e-06,
+      "loss": 0.7007,
+      "step": 2565
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.042712640061327,
+      "learning_rate": 8.56463595158547e-06,
+      "loss": 0.6754,
+      "step": 2566
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.414866039042617,
+      "learning_rate": 8.563440667866278e-06,
+      "loss": 0.5793,
+      "step": 2567
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.186808534700377,
+      "learning_rate": 8.56224497015685e-06,
+      "loss": 0.6083,
+      "step": 2568
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7550982917448628,
+      "learning_rate": 8.561048858596097e-06,
+      "loss": 0.6927,
+      "step": 2569
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.307539577711527,
+      "learning_rate": 8.559852333322982e-06,
+      "loss": 0.7618,
+      "step": 2570
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6597721789148605,
+      "learning_rate": 8.558655394476513e-06,
+      "loss": 0.7079,
+      "step": 2571
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.345267249284243,
+      "learning_rate": 8.557458042195748e-06,
+      "loss": 0.723,
+      "step": 2572
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.194152292832171,
+      "learning_rate": 8.556260276619792e-06,
+      "loss": 0.7222,
+      "step": 2573
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.8628566104318725,
+      "learning_rate": 8.555062097887796e-06,
+      "loss": 0.6131,
+      "step": 2574
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2175177567930087,
+      "learning_rate": 8.553863506138962e-06,
+      "loss": 0.6847,
+      "step": 2575
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.19010534016988,
+      "learning_rate": 8.55266450151254e-06,
+      "loss": 0.6466,
+      "step": 2576
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.030066766241923,
+      "learning_rate": 8.551465084147826e-06,
+      "loss": 0.5981,
+      "step": 2577
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5819351901307446,
+      "learning_rate": 8.550265254184163e-06,
+      "loss": 0.6906,
+      "step": 2578
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.837791982234303,
+      "learning_rate": 8.549065011760948e-06,
+      "loss": 0.6177,
+      "step": 2579
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.767739222282302,
+      "learning_rate": 8.547864357017618e-06,
+      "loss": 0.6715,
+      "step": 2580
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6955870932338923,
+      "learning_rate": 8.546663290093663e-06,
+      "loss": 0.6986,
+      "step": 2581
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.542337333218461,
+      "learning_rate": 8.545461811128618e-06,
+      "loss": 0.7228,
+      "step": 2582
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.345404508684496,
+      "learning_rate": 8.54425992026207e-06,
+      "loss": 0.6988,
+      "step": 2583
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.046812475477053,
+      "learning_rate": 8.54305761763365e-06,
+      "loss": 0.6544,
+      "step": 2584
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0768028881923986,
+      "learning_rate": 8.541854903383038e-06,
+      "loss": 0.6944,
+      "step": 2585
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4296065875213158,
+      "learning_rate": 8.54065177764996e-06,
+      "loss": 0.6207,
+      "step": 2586
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.333406207555947,
+      "learning_rate": 8.539448240574196e-06,
+      "loss": 0.7403,
+      "step": 2587
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2395561665415036,
+      "learning_rate": 8.538244292295565e-06,
+      "loss": 0.6861,
+      "step": 2588
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1745677467987092,
+      "learning_rate": 8.537039932953941e-06,
+      "loss": 0.6158,
+      "step": 2589
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7633198958097256,
+      "learning_rate": 8.535835162689243e-06,
+      "loss": 0.6364,
+      "step": 2590
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5596406892643273,
+      "learning_rate": 8.534629981641435e-06,
+      "loss": 0.6835,
+      "step": 2591
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.182600890353231,
+      "learning_rate": 8.533424389950534e-06,
+      "loss": 0.574,
+      "step": 2592
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6915694598896676,
+      "learning_rate": 8.532218387756603e-06,
+      "loss": 0.6927,
+      "step": 2593
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.6283644209612125,
+      "learning_rate": 8.531011975199747e-06,
+      "loss": 0.6385,
+      "step": 2594
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9155113489734292,
+      "learning_rate": 8.52980515242013e-06,
+      "loss": 0.6983,
+      "step": 2595
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.980349217997206,
+      "learning_rate": 8.528597919557953e-06,
+      "loss": 0.7198,
+      "step": 2596
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.810695167253675,
+      "learning_rate": 8.52739027675347e-06,
+      "loss": 0.6109,
+      "step": 2597
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9757821780269134,
+      "learning_rate": 8.526182224146982e-06,
+      "loss": 0.6589,
+      "step": 2598
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.5030227005870316,
+      "learning_rate": 8.524973761878834e-06,
+      "loss": 0.7574,
+      "step": 2599
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9534355234534457,
+      "learning_rate": 8.523764890089425e-06,
+      "loss": 0.6311,
+      "step": 2600
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2968695529272085,
+      "learning_rate": 8.522555608919198e-06,
+      "loss": 0.7159,
+      "step": 2601
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5862520567000935,
+      "learning_rate": 8.521345918508644e-06,
+      "loss": 0.6494,
+      "step": 2602
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2346246576533058,
+      "learning_rate": 8.520135818998299e-06,
+      "loss": 0.6798,
+      "step": 2603
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3638719754838653,
+      "learning_rate": 8.518925310528749e-06,
+      "loss": 0.728,
+      "step": 2604
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.03918196825084,
+      "learning_rate": 8.51771439324063e-06,
+      "loss": 0.7459,
+      "step": 2605
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3225257145217433,
+      "learning_rate": 8.516503067274622e-06,
+      "loss": 0.7274,
+      "step": 2606
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.186434199840289,
+      "learning_rate": 8.515291332771452e-06,
+      "loss": 0.6888,
+      "step": 2607
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1134392839374618,
+      "learning_rate": 8.514079189871898e-06,
+      "loss": 0.6312,
+      "step": 2608
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.563735931453355,
+      "learning_rate": 8.51286663871678e-06,
+      "loss": 0.7038,
+      "step": 2609
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.940504117749469,
+      "learning_rate": 8.511653679446972e-06,
+      "loss": 0.7079,
+      "step": 2610
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2533130654227143,
+      "learning_rate": 8.51044031220339e-06,
+      "loss": 0.6543,
+      "step": 2611
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.8589646098695116,
+      "learning_rate": 8.509226537127e-06,
+      "loss": 0.6759,
+      "step": 2612
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.5253019336589153,
+      "learning_rate": 8.508012354358815e-06,
+      "loss": 0.5805,
+      "step": 2613
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1367218095988973,
+      "learning_rate": 8.506797764039895e-06,
+      "loss": 0.6941,
+      "step": 2614
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0713095113351727,
+      "learning_rate": 8.505582766311349e-06,
+      "loss": 0.6673,
+      "step": 2615
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.312239589857587,
+      "learning_rate": 8.504367361314329e-06,
+      "loss": 0.6638,
+      "step": 2616
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8300865651331475,
+      "learning_rate": 8.50315154919004e-06,
+      "loss": 0.7389,
+      "step": 2617
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.5470160475934462,
+      "learning_rate": 8.501935330079732e-06,
+      "loss": 0.7534,
+      "step": 2618
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.316557953808444,
+      "learning_rate": 8.5007187041247e-06,
+      "loss": 0.7406,
+      "step": 2619
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.146381681148758,
+      "learning_rate": 8.499501671466287e-06,
+      "loss": 0.663,
+      "step": 2620
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7424691680667523,
+      "learning_rate": 8.498284232245888e-06,
+      "loss": 0.6684,
+      "step": 2621
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.812793415389665,
+      "learning_rate": 8.497066386604937e-06,
+      "loss": 0.7148,
+      "step": 2622
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4179027330742247,
+      "learning_rate": 8.495848134684924e-06,
+      "loss": 0.5672,
+      "step": 2623
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9317748456252173,
+      "learning_rate": 8.494629476627378e-06,
+      "loss": 0.6196,
+      "step": 2624
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.8552043750504303,
+      "learning_rate": 8.493410412573883e-06,
+      "loss": 0.76,
+      "step": 2625
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1671342068280426,
+      "learning_rate": 8.492190942666065e-06,
+      "loss": 0.7228,
+      "step": 2626
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.4361650708105707,
+      "learning_rate": 8.490971067045596e-06,
+      "loss": 0.7221,
+      "step": 2627
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6358964795479687,
+      "learning_rate": 8.489750785854203e-06,
+      "loss": 0.6363,
+      "step": 2628
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1358464929580174,
+      "learning_rate": 8.48853009923365e-06,
+      "loss": 0.7216,
+      "step": 2629
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3541129682126383,
+      "learning_rate": 8.487309007325755e-06,
+      "loss": 0.5821,
+      "step": 2630
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.420390390909207,
+      "learning_rate": 8.48608751027238e-06,
+      "loss": 0.6743,
+      "step": 2631
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4451760999988363,
+      "learning_rate": 8.484865608215435e-06,
+      "loss": 0.6701,
+      "step": 2632
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.4307024526167944,
+      "learning_rate": 8.483643301296877e-06,
+      "loss": 0.7317,
+      "step": 2633
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2798188313827032,
+      "learning_rate": 8.482420589658712e-06,
+      "loss": 0.6603,
+      "step": 2634
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7920937123015785,
+      "learning_rate": 8.481197473442989e-06,
+      "loss": 0.7351,
+      "step": 2635
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1650536164648404,
+      "learning_rate": 8.479973952791805e-06,
+      "loss": 0.747,
+      "step": 2636
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4408149875447256,
+      "learning_rate": 8.478750027847308e-06,
+      "loss": 0.7117,
+      "step": 2637
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.5845290741492817,
+      "learning_rate": 8.477525698751688e-06,
+      "loss": 0.6305,
+      "step": 2638
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.311318523859558,
+      "learning_rate": 8.476300965647186e-06,
+      "loss": 0.6609,
+      "step": 2639
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0266589746186376,
+      "learning_rate": 8.475075828676086e-06,
+      "loss": 0.663,
+      "step": 2640
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8468358467655803,
+      "learning_rate": 8.473850287980721e-06,
+      "loss": 0.6316,
+      "step": 2641
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1150563470392996,
+      "learning_rate": 8.472624343703473e-06,
+      "loss": 0.713,
+      "step": 2642
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.3018689988839918,
+      "learning_rate": 8.471397995986766e-06,
+      "loss": 0.5945,
+      "step": 2643
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.299828029802301,
+      "learning_rate": 8.470171244973075e-06,
+      "loss": 0.6362,
+      "step": 2644
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.400265553940191,
+      "learning_rate": 8.46894409080492e-06,
+      "loss": 0.7142,
+      "step": 2645
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3965587472793275,
+      "learning_rate": 8.467716533624869e-06,
+      "loss": 0.6693,
+      "step": 2646
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1424468064421527,
+      "learning_rate": 8.466488573575536e-06,
+      "loss": 0.6621,
+      "step": 2647
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.058208255714223,
+      "learning_rate": 8.465260210799579e-06,
+      "loss": 0.5876,
+      "step": 2648
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.051342205579871,
+      "learning_rate": 8.464031445439708e-06,
+      "loss": 0.6876,
+      "step": 2649
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.435428283210832,
+      "learning_rate": 8.462802277638677e-06,
+      "loss": 0.7141,
+      "step": 2650
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2536304400179357,
+      "learning_rate": 8.461572707539288e-06,
+      "loss": 0.7239,
+      "step": 2651
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2033188343885404,
+      "learning_rate": 8.460342735284388e-06,
+      "loss": 0.5988,
+      "step": 2652
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6488901006220136,
+      "learning_rate": 8.459112361016873e-06,
+      "loss": 0.7123,
+      "step": 2653
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.454166766453256,
+      "learning_rate": 8.457881584879681e-06,
+      "loss": 0.7274,
+      "step": 2654
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.338529320134492,
+      "learning_rate": 8.456650407015804e-06,
+      "loss": 0.7723,
+      "step": 2655
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7241780151773933,
+      "learning_rate": 8.455418827568275e-06,
+      "loss": 0.619,
+      "step": 2656
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2430549835972053,
+      "learning_rate": 8.454186846680174e-06,
+      "loss": 0.6374,
+      "step": 2657
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.090373805102784,
+      "learning_rate": 8.452954464494631e-06,
+      "loss": 0.7013,
+      "step": 2658
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.50997857984063,
+      "learning_rate": 8.451721681154819e-06,
+      "loss": 0.6987,
+      "step": 2659
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.318114958320992,
+      "learning_rate": 8.45048849680396e-06,
+      "loss": 0.7233,
+      "step": 2660
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2346055238002462,
+      "learning_rate": 8.449254911585323e-06,
+      "loss": 0.6527,
+      "step": 2661
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.293606280499452,
+      "learning_rate": 8.44802092564222e-06,
+      "loss": 0.6759,
+      "step": 2662
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.206662259361233,
+      "learning_rate": 8.446786539118014e-06,
+      "loss": 0.6139,
+      "step": 2663
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 12.24962776644313,
+      "learning_rate": 8.445551752156111e-06,
+      "loss": 0.6434,
+      "step": 2664
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2212865543394447,
+      "learning_rate": 8.444316564899966e-06,
+      "loss": 0.6465,
+      "step": 2665
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.714117520697914,
+      "learning_rate": 8.443080977493078e-06,
+      "loss": 0.5998,
+      "step": 2666
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.336735248734921,
+      "learning_rate": 8.441844990078995e-06,
+      "loss": 0.7573,
+      "step": 2667
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.403703415234515,
+      "learning_rate": 8.44060860280131e-06,
+      "loss": 0.6043,
+      "step": 2668
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.382422434010672,
+      "learning_rate": 8.439371815803666e-06,
+      "loss": 0.7587,
+      "step": 2669
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0968077866837533,
+      "learning_rate": 8.438134629229746e-06,
+      "loss": 0.6986,
+      "step": 2670
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.43034993732329,
+      "learning_rate": 8.436897043223282e-06,
+      "loss": 0.5956,
+      "step": 2671
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.396150179364192,
+      "learning_rate": 8.435659057928054e-06,
+      "loss": 0.6747,
+      "step": 2672
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7277764541975973,
+      "learning_rate": 8.434420673487888e-06,
+      "loss": 0.6954,
+      "step": 2673
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.217561657414955,
+      "learning_rate": 8.433181890046658e-06,
+      "loss": 0.6458,
+      "step": 2674
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.236257143877165,
+      "learning_rate": 8.431942707748279e-06,
+      "loss": 0.6915,
+      "step": 2675
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0699718693696636,
+      "learning_rate": 8.430703126736717e-06,
+      "loss": 0.6652,
+      "step": 2676
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.175699865000509,
+      "learning_rate": 8.429463147155984e-06,
+      "loss": 0.772,
+      "step": 2677
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7545634656991043,
+      "learning_rate": 8.428222769150137e-06,
+      "loss": 0.6886,
+      "step": 2678
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.448219091164623,
+      "learning_rate": 8.426981992863276e-06,
+      "loss": 0.6719,
+      "step": 2679
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2382890240649553,
+      "learning_rate": 8.425740818439553e-06,
+      "loss": 0.7212,
+      "step": 2680
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.483770641222447,
+      "learning_rate": 8.424499246023168e-06,
+      "loss": 0.7427,
+      "step": 2681
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0361221580953,
+      "learning_rate": 8.42325727575836e-06,
+      "loss": 0.7023,
+      "step": 2682
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1519597344947137,
+      "learning_rate": 8.422014907789413e-06,
+      "loss": 0.7392,
+      "step": 2683
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1345686803799038,
+      "learning_rate": 8.420772142260667e-06,
+      "loss": 0.6333,
+      "step": 2684
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.287284430544546,
+      "learning_rate": 8.419528979316505e-06,
+      "loss": 0.7573,
+      "step": 2685
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4381162142757145,
+      "learning_rate": 8.41828541910135e-06,
+      "loss": 0.7724,
+      "step": 2686
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8332535794197633,
+      "learning_rate": 8.417041461759674e-06,
+      "loss": 0.6979,
+      "step": 2687
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1728298056179722,
+      "learning_rate": 8.415797107436e-06,
+      "loss": 0.627,
+      "step": 2688
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1601804909844005,
+      "learning_rate": 8.414552356274891e-06,
+      "loss": 0.6739,
+      "step": 2689
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.679136900723749,
+      "learning_rate": 8.413307208420963e-06,
+      "loss": 0.7404,
+      "step": 2690
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.5824680380348557,
+      "learning_rate": 8.412061664018869e-06,
+      "loss": 0.745,
+      "step": 2691
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.209895645078567,
+      "learning_rate": 8.410815723213312e-06,
+      "loss": 0.6478,
+      "step": 2692
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1856854173402795,
+      "learning_rate": 8.409569386149046e-06,
+      "loss": 0.6822,
+      "step": 2693
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3084967091893653,
+      "learning_rate": 8.408322652970866e-06,
+      "loss": 0.7768,
+      "step": 2694
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.9063292951795425,
+      "learning_rate": 8.40707552382361e-06,
+      "loss": 0.6608,
+      "step": 2695
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.051324396490213,
+      "learning_rate": 8.40582799885217e-06,
+      "loss": 0.5923,
+      "step": 2696
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.765515717742255,
+      "learning_rate": 8.404580078201476e-06,
+      "loss": 0.7872,
+      "step": 2697
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1789517487510066,
+      "learning_rate": 8.403331762016514e-06,
+      "loss": 0.6043,
+      "step": 2698
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2655735243519515,
+      "learning_rate": 8.402083050442302e-06,
+      "loss": 0.5652,
+      "step": 2699
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2602508493232687,
+      "learning_rate": 8.400833943623919e-06,
+      "loss": 0.7123,
+      "step": 2700
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.5454383180770535,
+      "learning_rate": 8.399584441706477e-06,
+      "loss": 0.699,
+      "step": 2701
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2564015333342606,
+      "learning_rate": 8.398334544835143e-06,
+      "loss": 0.6204,
+      "step": 2702
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.418647080107335,
+      "learning_rate": 8.397084253155125e-06,
+      "loss": 0.6295,
+      "step": 2703
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.25654659367241,
+      "learning_rate": 8.395833566811677e-06,
+      "loss": 0.692,
+      "step": 2704
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.666744974325964,
+      "learning_rate": 8.394582485950103e-06,
+      "loss": 0.6038,
+      "step": 2705
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4348825850755493,
+      "learning_rate": 8.393331010715749e-06,
+      "loss": 0.6997,
+      "step": 2706
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0592895390345376,
+      "learning_rate": 8.392079141254006e-06,
+      "loss": 0.6366,
+      "step": 2707
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0843302668304746,
+      "learning_rate": 8.390826877710314e-06,
+      "loss": 0.595,
+      "step": 2708
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.693813971402732,
+      "learning_rate": 8.38957422023016e-06,
+      "loss": 0.6939,
+      "step": 2709
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4215728777192447,
+      "learning_rate": 8.388321168959068e-06,
+      "loss": 0.7119,
+      "step": 2710
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4275301307354837,
+      "learning_rate": 8.387067724042618e-06,
+      "loss": 0.6875,
+      "step": 2711
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.524758976644784,
+      "learning_rate": 8.38581388562643e-06,
+      "loss": 0.7328,
+      "step": 2712
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.529001789661384,
+      "learning_rate": 8.384559653856174e-06,
+      "loss": 0.6467,
+      "step": 2713
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.339231251838686,
+      "learning_rate": 8.383305028877559e-06,
+      "loss": 0.6954,
+      "step": 2714
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.422685898720981,
+      "learning_rate": 8.382050010836349e-06,
+      "loss": 0.7583,
+      "step": 2715
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.610823185514183,
+      "learning_rate": 8.380794599878343e-06,
+      "loss": 0.67,
+      "step": 2716
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.642245922603603,
+      "learning_rate": 8.379538796149395e-06,
+      "loss": 0.6711,
+      "step": 2717
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3135555055352635,
+      "learning_rate": 8.378282599795397e-06,
+      "loss": 0.6705,
+      "step": 2718
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.447841830239155,
+      "learning_rate": 8.377026010962293e-06,
+      "loss": 0.6548,
+      "step": 2719
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.349506243780416,
+      "learning_rate": 8.375769029796068e-06,
+      "loss": 0.6988,
+      "step": 2720
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.4916280933140373,
+      "learning_rate": 8.374511656442756e-06,
+      "loss": 0.6912,
+      "step": 2721
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.008729698463561,
+      "learning_rate": 8.373253891048436e-06,
+      "loss": 0.6498,
+      "step": 2722
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.504995655860519,
+      "learning_rate": 8.371995733759228e-06,
+      "loss": 0.6303,
+      "step": 2723
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.465351579775575,
+      "learning_rate": 8.370737184721305e-06,
+      "loss": 0.6931,
+      "step": 2724
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.786999536883786,
+      "learning_rate": 8.369478244080878e-06,
+      "loss": 0.6444,
+      "step": 2725
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.089894612238134,
+      "learning_rate": 8.368218911984211e-06,
+      "loss": 0.7627,
+      "step": 2726
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.929159775372233,
+      "learning_rate": 8.366959188577606e-06,
+      "loss": 0.6889,
+      "step": 2727
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.440489695426411,
+      "learning_rate": 8.365699074007416e-06,
+      "loss": 0.6793,
+      "step": 2728
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.058496978919441,
+      "learning_rate": 8.364438568420034e-06,
+      "loss": 0.6912,
+      "step": 2729
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4524683201064064,
+      "learning_rate": 8.363177671961908e-06,
+      "loss": 0.6447,
+      "step": 2730
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0795075403169156,
+      "learning_rate": 8.36191638477952e-06,
+      "loss": 0.6589,
+      "step": 2731
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9902950430871291,
+      "learning_rate": 8.360654707019406e-06,
+      "loss": 0.6756,
+      "step": 2732
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.568567286867176,
+      "learning_rate": 8.359392638828142e-06,
+      "loss": 0.7202,
+      "step": 2733
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7851953645089007,
+      "learning_rate": 8.358130180352353e-06,
+      "loss": 0.6707,
+      "step": 2734
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7057531159089776,
+      "learning_rate": 8.356867331738706e-06,
+      "loss": 0.6586,
+      "step": 2735
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.2347571171882237,
+      "learning_rate": 8.355604093133916e-06,
+      "loss": 0.6705,
+      "step": 2736
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.625192631188438,
+      "learning_rate": 8.354340464684745e-06,
+      "loss": 0.7096,
+      "step": 2737
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7137203886388295,
+      "learning_rate": 8.353076446537993e-06,
+      "loss": 0.6789,
+      "step": 2738
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1425928647546533,
+      "learning_rate": 8.351812038840513e-06,
+      "loss": 0.6174,
+      "step": 2739
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1199663116285286,
+      "learning_rate": 8.3505472417392e-06,
+      "loss": 0.6875,
+      "step": 2740
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3558771465349433,
+      "learning_rate": 8.349282055380992e-06,
+      "loss": 0.6542,
+      "step": 2741
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9753417802732893,
+      "learning_rate": 8.348016479912877e-06,
+      "loss": 0.6566,
+      "step": 2742
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5685130835209935,
+      "learning_rate": 8.346750515481888e-06,
+      "loss": 0.7154,
+      "step": 2743
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1983652825774427,
+      "learning_rate": 8.345484162235096e-06,
+      "loss": 0.6878,
+      "step": 2744
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.279656071212713,
+      "learning_rate": 8.344217420319624e-06,
+      "loss": 0.6388,
+      "step": 2745
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.6025787077693563,
+      "learning_rate": 8.342950289882641e-06,
+      "loss": 0.6941,
+      "step": 2746
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.614210374161948,
+      "learning_rate": 8.341682771071357e-06,
+      "loss": 0.6266,
+      "step": 2747
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1048003157133306,
+      "learning_rate": 8.340414864033028e-06,
+      "loss": 0.6507,
+      "step": 2748
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1566816578883072,
+      "learning_rate": 8.339146568914958e-06,
+      "loss": 0.7176,
+      "step": 2749
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.6545077340607643,
+      "learning_rate": 8.337877885864489e-06,
+      "loss": 0.7292,
+      "step": 2750
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.9130923410553775,
+      "learning_rate": 8.336608815029018e-06,
+      "loss": 0.7339,
+      "step": 2751
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.184098187059293,
+      "learning_rate": 8.335339356555981e-06,
+      "loss": 0.6965,
+      "step": 2752
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.14723331071444,
+      "learning_rate": 8.334069510592857e-06,
+      "loss": 0.5925,
+      "step": 2753
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9405681252036464,
+      "learning_rate": 8.332799277287175e-06,
+      "loss": 0.6977,
+      "step": 2754
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3880455307462882,
+      "learning_rate": 8.331528656786508e-06,
+      "loss": 0.6346,
+      "step": 2755
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.322406078061077,
+      "learning_rate": 8.330257649238472e-06,
+      "loss": 0.7288,
+      "step": 2756
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1014011587142933,
+      "learning_rate": 8.328986254790729e-06,
+      "loss": 0.6754,
+      "step": 2757
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3514936631244956,
+      "learning_rate": 8.327714473590986e-06,
+      "loss": 0.646,
+      "step": 2758
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.299770815204609,
+      "learning_rate": 8.326442305786995e-06,
+      "loss": 0.6682,
+      "step": 2759
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4432174028048284,
+      "learning_rate": 8.325169751526552e-06,
+      "loss": 0.7419,
+      "step": 2760
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7511522431665765,
+      "learning_rate": 8.323896810957501e-06,
+      "loss": 0.5777,
+      "step": 2761
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7500144099538226,
+      "learning_rate": 8.322623484227725e-06,
+      "loss": 0.6688,
+      "step": 2762
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.442208106975574,
+      "learning_rate": 8.321349771485159e-06,
+      "loss": 0.7288,
+      "step": 2763
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4073668296629323,
+      "learning_rate": 8.320075672877776e-06,
+      "loss": 0.6964,
+      "step": 2764
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.973204687169632,
+      "learning_rate": 8.3188011885536e-06,
+      "loss": 0.665,
+      "step": 2765
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9341022002413333,
+      "learning_rate": 8.317526318660695e-06,
+      "loss": 0.6552,
+      "step": 2766
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.41457288085297,
+      "learning_rate": 8.316251063347175e-06,
+      "loss": 0.7097,
+      "step": 2767
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5800526852961263,
+      "learning_rate": 8.314975422761187e-06,
+      "loss": 0.5778,
+      "step": 2768
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.752826125090857,
+      "learning_rate": 8.313699397050941e-06,
+      "loss": 0.6198,
+      "step": 2769
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3181883296463717,
+      "learning_rate": 8.312422986364677e-06,
+      "loss": 0.546,
+      "step": 2770
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3923531457438534,
+      "learning_rate": 8.311146190850687e-06,
+      "loss": 0.7019,
+      "step": 2771
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.57435657748528,
+      "learning_rate": 8.309869010657303e-06,
+      "loss": 0.6215,
+      "step": 2772
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9661259285016583,
+      "learning_rate": 8.308591445932905e-06,
+      "loss": 0.6052,
+      "step": 2773
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1300824586691043,
+      "learning_rate": 8.307313496825918e-06,
+      "loss": 0.6245,
+      "step": 2774
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7492441683809816,
+      "learning_rate": 8.306035163484806e-06,
+      "loss": 0.6852,
+      "step": 2775
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4951222437462484,
+      "learning_rate": 8.30475644605809e-06,
+      "loss": 0.6103,
+      "step": 2776
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2323509496392493,
+      "learning_rate": 8.30347734469432e-06,
+      "loss": 0.6149,
+      "step": 2777
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.702733593456729,
+      "learning_rate": 8.302197859542104e-06,
+      "loss": 0.6772,
+      "step": 2778
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.029812775065599,
+      "learning_rate": 8.300917990750085e-06,
+      "loss": 0.669,
+      "step": 2779
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.436939042536463,
+      "learning_rate": 8.299637738466956e-06,
+      "loss": 0.6703,
+      "step": 2780
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.064859457627149,
+      "learning_rate": 8.298357102841452e-06,
+      "loss": 0.6941,
+      "step": 2781
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.290933554699684,
+      "learning_rate": 8.297076084022355e-06,
+      "loss": 0.585,
+      "step": 2782
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7292159300881327,
+      "learning_rate": 8.29579468215849e-06,
+      "loss": 0.6463,
+      "step": 2783
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4076744166382755,
+      "learning_rate": 8.294512897398725e-06,
+      "loss": 0.6891,
+      "step": 2784
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.400183386823079,
+      "learning_rate": 8.293230729891976e-06,
+      "loss": 0.6797,
+      "step": 2785
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5300566156875215,
+      "learning_rate": 8.2919481797872e-06,
+      "loss": 0.6729,
+      "step": 2786
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.504139135974389,
+      "learning_rate": 8.2906652472334e-06,
+      "loss": 0.7771,
+      "step": 2787
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4033755762522127,
+      "learning_rate": 8.289381932379625e-06,
+      "loss": 0.7255,
+      "step": 2788
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.130391692034721,
+      "learning_rate": 8.288098235374966e-06,
+      "loss": 0.6504,
+      "step": 2789
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3119004788902524,
+      "learning_rate": 8.286814156368559e-06,
+      "loss": 0.7532,
+      "step": 2790
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 12.473032165918204,
+      "learning_rate": 8.285529695509585e-06,
+      "loss": 0.7055,
+      "step": 2791
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.874206287893155,
+      "learning_rate": 8.284244852947265e-06,
+      "loss": 0.7682,
+      "step": 2792
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3354999888656627,
+      "learning_rate": 8.282959628830875e-06,
+      "loss": 0.7107,
+      "step": 2793
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.6099032292361617,
+      "learning_rate": 8.281674023309725e-06,
+      "loss": 0.6281,
+      "step": 2794
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.232202918075071,
+      "learning_rate": 8.280388036533171e-06,
+      "loss": 0.6399,
+      "step": 2795
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.491569330119017,
+      "learning_rate": 8.27910166865062e-06,
+      "loss": 0.626,
+      "step": 2796
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2273954138081073,
+      "learning_rate": 8.277814919811516e-06,
+      "loss": 0.6008,
+      "step": 2797
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2945544456627385,
+      "learning_rate": 8.276527790165349e-06,
+      "loss": 0.6461,
+      "step": 2798
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.277058416467495,
+      "learning_rate": 8.275240279861655e-06,
+      "loss": 0.69,
+      "step": 2799
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2731253809981653,
+      "learning_rate": 8.273952389050015e-06,
+      "loss": 0.6764,
+      "step": 2800
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1917315114226636,
+      "learning_rate": 8.272664117880047e-06,
+      "loss": 0.6476,
+      "step": 2801
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.25326375571558,
+      "learning_rate": 8.271375466501424e-06,
+      "loss": 0.7102,
+      "step": 2802
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.302993394481045,
+      "learning_rate": 8.270086435063856e-06,
+      "loss": 0.6434,
+      "step": 2803
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7166993208874195,
+      "learning_rate": 8.268797023717098e-06,
+      "loss": 0.6369,
+      "step": 2804
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2576333540166673,
+      "learning_rate": 8.267507232610952e-06,
+      "loss": 0.7091,
+      "step": 2805
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.834052961933635,
+      "learning_rate": 8.26621706189526e-06,
+      "loss": 0.6522,
+      "step": 2806
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4814364257476966,
+      "learning_rate": 8.264926511719912e-06,
+      "loss": 0.6399,
+      "step": 2807
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8842459517526953,
+      "learning_rate": 8.26363558223484e-06,
+      "loss": 0.7116,
+      "step": 2808
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2082080106969904,
+      "learning_rate": 8.26234427359002e-06,
+      "loss": 0.6688,
+      "step": 2809
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.234644633018572,
+      "learning_rate": 8.261052585935471e-06,
+      "loss": 0.6346,
+      "step": 2810
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4322865695650178,
+      "learning_rate": 8.259760519421263e-06,
+      "loss": 0.7054,
+      "step": 2811
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.6564893401792675,
+      "learning_rate": 8.258468074197499e-06,
+      "loss": 0.6955,
+      "step": 2812
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.266036275854144,
+      "learning_rate": 8.257175250414333e-06,
+      "loss": 0.7159,
+      "step": 2813
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.762553186968921,
+      "learning_rate": 8.255882048221961e-06,
+      "loss": 0.7056,
+      "step": 2814
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.139967136595774,
+      "learning_rate": 8.254588467770628e-06,
+      "loss": 0.707,
+      "step": 2815
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.9568494330619317,
+      "learning_rate": 8.253294509210612e-06,
+      "loss": 0.6702,
+      "step": 2816
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.031085393262171,
+      "learning_rate": 8.252000172692244e-06,
+      "loss": 0.6967,
+      "step": 2817
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.3355616466171094,
+      "learning_rate": 8.250705458365897e-06,
+      "loss": 0.7047,
+      "step": 2818
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3126709857674936,
+      "learning_rate": 8.249410366381987e-06,
+      "loss": 0.6727,
+      "step": 2819
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.857054433557827,
+      "learning_rate": 8.248114896890975e-06,
+      "loss": 0.6336,
+      "step": 2820
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.0926046214596106,
+      "learning_rate": 8.246819050043363e-06,
+      "loss": 0.6691,
+      "step": 2821
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.618242351375955,
+      "learning_rate": 8.245522825989697e-06,
+      "loss": 0.6733,
+      "step": 2822
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7640385714230673,
+      "learning_rate": 8.244226224880574e-06,
+      "loss": 0.6313,
+      "step": 2823
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5354134106997948,
+      "learning_rate": 8.242929246866624e-06,
+      "loss": 0.7069,
+      "step": 2824
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.933933284256052,
+      "learning_rate": 8.24163189209853e-06,
+      "loss": 0.749,
+      "step": 2825
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8633014033698627,
+      "learning_rate": 8.240334160727013e-06,
+      "loss": 0.6782,
+      "step": 2826
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7941442636390543,
+      "learning_rate": 8.23903605290284e-06,
+      "loss": 0.6433,
+      "step": 2827
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3734190039511978,
+      "learning_rate": 8.23773756877682e-06,
+      "loss": 0.6341,
+      "step": 2828
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.306290887870174,
+      "learning_rate": 8.236438708499811e-06,
+      "loss": 0.7399,
+      "step": 2829
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.327313393460941,
+      "learning_rate": 8.235139472222708e-06,
+      "loss": 0.6106,
+      "step": 2830
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.842091527633625,
+      "learning_rate": 8.233839860096453e-06,
+      "loss": 0.6594,
+      "step": 2831
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.427751709970699,
+      "learning_rate": 8.23253987227203e-06,
+      "loss": 0.757,
+      "step": 2832
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.827636901802302,
+      "learning_rate": 8.23123950890047e-06,
+      "loss": 0.7158,
+      "step": 2833
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.310647756260461,
+      "learning_rate": 8.229938770132843e-06,
+      "loss": 0.6142,
+      "step": 2834
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.439737457921449,
+      "learning_rate": 8.228637656120268e-06,
+      "loss": 0.6859,
+      "step": 2835
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.465299362478447,
+      "learning_rate": 8.227336167013901e-06,
+      "loss": 0.619,
+      "step": 2836
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.161811987054101,
+      "learning_rate": 8.22603430296495e-06,
+      "loss": 0.6885,
+      "step": 2837
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 13.016909893025721,
+      "learning_rate": 8.224732064124658e-06,
+      "loss": 0.7578,
+      "step": 2838
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.29770830365415,
+      "learning_rate": 8.223429450644317e-06,
+      "loss": 0.5983,
+      "step": 2839
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.74582961488932,
+      "learning_rate": 8.222126462675259e-06,
+      "loss": 0.703,
+      "step": 2840
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.839825341685874,
+      "learning_rate": 8.220823100368865e-06,
+      "loss": 0.7577,
+      "step": 2841
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.151612849557347,
+      "learning_rate": 8.219519363876552e-06,
+      "loss": 0.7437,
+      "step": 2842
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2813563053260157,
+      "learning_rate": 8.218215253349785e-06,
+      "loss": 0.7145,
+      "step": 2843
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8740576992118227,
+      "learning_rate": 8.216910768940075e-06,
+      "loss": 0.6663,
+      "step": 2844
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3832922463971293,
+      "learning_rate": 8.215605910798972e-06,
+      "loss": 0.6398,
+      "step": 2845
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.087398052989275,
+      "learning_rate": 8.21430067907807e-06,
+      "loss": 0.6233,
+      "step": 2846
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1121220094877007,
+      "learning_rate": 8.212995073929002e-06,
+      "loss": 0.7069,
+      "step": 2847
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1733089591500396,
+      "learning_rate": 8.211689095503457e-06,
+      "loss": 0.6373,
+      "step": 2848
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.53436764490174,
+      "learning_rate": 8.210382743953159e-06,
+      "loss": 0.6675,
+      "step": 2849
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9962280045699328,
+      "learning_rate": 8.20907601942987e-06,
+      "loss": 0.6339,
+      "step": 2850
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.6687863772007625,
+      "learning_rate": 8.207768922085408e-06,
+      "loss": 0.6427,
+      "step": 2851
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4367057789328523,
+      "learning_rate": 8.206461452071625e-06,
+      "loss": 0.6382,
+      "step": 2852
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.494084327088906,
+      "learning_rate": 8.20515360954042e-06,
+      "loss": 0.7028,
+      "step": 2853
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.0800899409941738,
+      "learning_rate": 8.203845394643732e-06,
+      "loss": 0.6296,
+      "step": 2854
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.158844023386168,
+      "learning_rate": 8.202536807533548e-06,
+      "loss": 0.7244,
+      "step": 2855
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.667508619561635,
+      "learning_rate": 8.201227848361895e-06,
+      "loss": 0.7001,
+      "step": 2856
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.53043885098298,
+      "learning_rate": 8.199918517280848e-06,
+      "loss": 0.6844,
+      "step": 2857
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.114762138835378,
+      "learning_rate": 8.198608814442513e-06,
+      "loss": 0.6376,
+      "step": 2858
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9348916221619814,
+      "learning_rate": 8.197298739999055e-06,
+      "loss": 0.6631,
+      "step": 2859
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2713830464167466,
+      "learning_rate": 8.19598829410267e-06,
+      "loss": 0.5696,
+      "step": 2860
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.640428459107109,
+      "learning_rate": 8.194677476905604e-06,
+      "loss": 0.6494,
+      "step": 2861
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.458537500123125,
+      "learning_rate": 8.193366288560144e-06,
+      "loss": 0.7073,
+      "step": 2862
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.789431886976531,
+      "learning_rate": 8.192054729218621e-06,
+      "loss": 0.6223,
+      "step": 2863
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0346142539870478,
+      "learning_rate": 8.190742799033404e-06,
+      "loss": 0.6502,
+      "step": 2864
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.523541617507126,
+      "learning_rate": 8.189430498156914e-06,
+      "loss": 0.6346,
+      "step": 2865
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2087194523547145,
+      "learning_rate": 8.18811782674161e-06,
+      "loss": 0.6108,
+      "step": 2866
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.491402481550859,
+      "learning_rate": 8.18680478493999e-06,
+      "loss": 0.6853,
+      "step": 2867
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.026321341321937,
+      "learning_rate": 8.185491372904604e-06,
+      "loss": 0.6313,
+      "step": 2868
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.6138591078805526,
+      "learning_rate": 8.184177590788038e-06,
+      "loss": 0.6652,
+      "step": 2869
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.519595862218996,
+      "learning_rate": 8.182863438742922e-06,
+      "loss": 0.7254,
+      "step": 2870
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4026200308540617,
+      "learning_rate": 8.181548916921935e-06,
+      "loss": 0.5704,
+      "step": 2871
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.580908432962159,
+      "learning_rate": 8.180234025477792e-06,
+      "loss": 0.6507,
+      "step": 2872
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.268967709119169,
+      "learning_rate": 8.178918764563251e-06,
+      "loss": 0.718,
+      "step": 2873
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2396586103792795,
+      "learning_rate": 8.177603134331119e-06,
+      "loss": 0.59,
+      "step": 2874
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7093881022757165,
+      "learning_rate": 8.17628713493424e-06,
+      "loss": 0.6465,
+      "step": 2875
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2292433893864105,
+      "learning_rate": 8.174970766525503e-06,
+      "loss": 0.6516,
+      "step": 2876
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.0744331784154943,
+      "learning_rate": 8.17365402925784e-06,
+      "loss": 0.6089,
+      "step": 2877
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.563571554597488,
+      "learning_rate": 8.172336923284225e-06,
+      "loss": 0.6704,
+      "step": 2878
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8782161919357705,
+      "learning_rate": 8.17101944875768e-06,
+      "loss": 0.7115,
+      "step": 2879
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.662681630857156,
+      "learning_rate": 8.16970160583126e-06,
+      "loss": 0.6443,
+      "step": 2880
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.35955388779492,
+      "learning_rate": 8.16838339465807e-06,
+      "loss": 0.7116,
+      "step": 2881
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1676445753173974,
+      "learning_rate": 8.167064815391254e-06,
+      "loss": 0.6053,
+      "step": 2882
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.466081183525901,
+      "learning_rate": 8.165745868184006e-06,
+      "loss": 0.6041,
+      "step": 2883
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0720974841936783,
+      "learning_rate": 8.164426553189553e-06,
+      "loss": 0.6562,
+      "step": 2884
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.740189082503985,
+      "learning_rate": 8.16310687056117e-06,
+      "loss": 0.7227,
+      "step": 2885
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.978902500808749,
+      "learning_rate": 8.161786820452176e-06,
+      "loss": 0.6807,
+      "step": 2886
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.301320445626182,
+      "learning_rate": 8.160466403015928e-06,
+      "loss": 0.726,
+      "step": 2887
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.267037663828458,
+      "learning_rate": 8.159145618405828e-06,
+      "loss": 0.6489,
+      "step": 2888
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.015189747260077,
+      "learning_rate": 8.157824466775324e-06,
+      "loss": 0.6653,
+      "step": 2889
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.521922180245439,
+      "learning_rate": 8.156502948277902e-06,
+      "loss": 0.6805,
+      "step": 2890
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.12799521453685,
+      "learning_rate": 8.15518106306709e-06,
+      "loss": 0.6851,
+      "step": 2891
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0483331306231944,
+      "learning_rate": 8.153858811296465e-06,
+      "loss": 0.6434,
+      "step": 2892
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.707423136219375,
+      "learning_rate": 8.152536193119638e-06,
+      "loss": 0.577,
+      "step": 2893
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.092278245285496,
+      "learning_rate": 8.151213208690271e-06,
+      "loss": 0.6721,
+      "step": 2894
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.561997031833164,
+      "learning_rate": 8.149889858162062e-06,
+      "loss": 0.6855,
+      "step": 2895
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0803284506968662,
+      "learning_rate": 8.148566141688755e-06,
+      "loss": 0.6528,
+      "step": 2896
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9962095481419624,
+      "learning_rate": 8.147242059424134e-06,
+      "loss": 0.7353,
+      "step": 2897
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5237602734129516,
+      "learning_rate": 8.145917611522029e-06,
+      "loss": 0.6057,
+      "step": 2898
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8369538414675164,
+      "learning_rate": 8.14459279813631e-06,
+      "loss": 0.7189,
+      "step": 2899
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2530174295755743,
+      "learning_rate": 8.143267619420892e-06,
+      "loss": 0.605,
+      "step": 2900
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6147582967437226,
+      "learning_rate": 8.141942075529725e-06,
+      "loss": 0.7003,
+      "step": 2901
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8824381178183907,
+      "learning_rate": 8.14061616661681e-06,
+      "loss": 0.7087,
+      "step": 2902
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.993274788781763,
+      "learning_rate": 8.13928989283619e-06,
+      "loss": 0.5397,
+      "step": 2903
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5531708615561,
+      "learning_rate": 8.137963254341944e-06,
+      "loss": 0.6569,
+      "step": 2904
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.666082677328026,
+      "learning_rate": 8.136636251288197e-06,
+      "loss": 0.7032,
+      "step": 2905
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.483321235755457,
+      "learning_rate": 8.135308883829119e-06,
+      "loss": 0.6559,
+      "step": 2906
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2907791966791415,
+      "learning_rate": 8.133981152118916e-06,
+      "loss": 0.6794,
+      "step": 2907
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2809901014129657,
+      "learning_rate": 8.132653056311844e-06,
+      "loss": 0.6586,
+      "step": 2908
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.5398481577715,
+      "learning_rate": 8.131324596562195e-06,
+      "loss": 0.6637,
+      "step": 2909
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7594732513859443,
+      "learning_rate": 8.129995773024306e-06,
+      "loss": 0.7316,
+      "step": 2910
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.30105743088516,
+      "learning_rate": 8.128666585852556e-06,
+      "loss": 0.6668,
+      "step": 2911
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.731674755400316,
+      "learning_rate": 8.127337035201365e-06,
+      "loss": 0.6782,
+      "step": 2912
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.186947648727673,
+      "learning_rate": 8.1260071212252e-06,
+      "loss": 0.6693,
+      "step": 2913
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.198681447100411,
+      "learning_rate": 8.12467684407856e-06,
+      "loss": 0.6396,
+      "step": 2914
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5485456110616918,
+      "learning_rate": 8.123346203916e-06,
+      "loss": 0.6218,
+      "step": 2915
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.914548621339626,
+      "learning_rate": 8.122015200892106e-06,
+      "loss": 0.6717,
+      "step": 2916
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.593755867982301,
+      "learning_rate": 8.120683835161511e-06,
+      "loss": 0.6373,
+      "step": 2917
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.062797790828946,
+      "learning_rate": 8.11935210687889e-06,
+      "loss": 0.6702,
+      "step": 2918
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9875147078800683,
+      "learning_rate": 8.118020016198957e-06,
+      "loss": 0.6398,
+      "step": 2919
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.022004557471706,
+      "learning_rate": 8.11668756327647e-06,
+      "loss": 0.7186,
+      "step": 2920
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2197328167155737,
+      "learning_rate": 8.115354748266233e-06,
+      "loss": 0.652,
+      "step": 2921
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0617705515592957,
+      "learning_rate": 8.114021571323089e-06,
+      "loss": 0.5757,
+      "step": 2922
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3145783721939996,
+      "learning_rate": 8.112688032601919e-06,
+      "loss": 0.6625,
+      "step": 2923
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.531303219988851,
+      "learning_rate": 8.111354132257651e-06,
+      "loss": 0.6679,
+      "step": 2924
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6045952874222555,
+      "learning_rate": 8.110019870445254e-06,
+      "loss": 0.7008,
+      "step": 2925
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9426229731805704,
+      "learning_rate": 8.10868524731974e-06,
+      "loss": 0.6898,
+      "step": 2926
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.261653088656059,
+      "learning_rate": 8.107350263036157e-06,
+      "loss": 0.6312,
+      "step": 2927
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.259944992592717,
+      "learning_rate": 8.106014917749605e-06,
+      "loss": 0.6676,
+      "step": 2928
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6421624462915765,
+      "learning_rate": 8.104679211615218e-06,
+      "loss": 0.7101,
+      "step": 2929
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3871886404739504,
+      "learning_rate": 8.103343144788177e-06,
+      "loss": 0.6416,
+      "step": 2930
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1379036146977164,
+      "learning_rate": 8.102006717423695e-06,
+      "loss": 0.6405,
+      "step": 2931
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0829424463195303,
+      "learning_rate": 8.100669929677044e-06,
+      "loss": 0.6991,
+      "step": 2932
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.45544537675686,
+      "learning_rate": 8.099332781703523e-06,
+      "loss": 0.6083,
+      "step": 2933
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5320119199003224,
+      "learning_rate": 8.097995273658479e-06,
+      "loss": 0.6114,
+      "step": 2934
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2301392753727916,
+      "learning_rate": 8.0966574056973e-06,
+      "loss": 0.7644,
+      "step": 2935
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.397360362520422,
+      "learning_rate": 8.095319177975412e-06,
+      "loss": 0.5961,
+      "step": 2936
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.5622955344160414,
+      "learning_rate": 8.093980590648291e-06,
+      "loss": 0.6874,
+      "step": 2937
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.481733875250971,
+      "learning_rate": 8.092641643871451e-06,
+      "loss": 0.5639,
+      "step": 2938
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.237192875379186,
+      "learning_rate": 8.091302337800441e-06,
+      "loss": 0.6794,
+      "step": 2939
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.364977168624842,
+      "learning_rate": 8.089962672590865e-06,
+      "loss": 0.6851,
+      "step": 2940
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.155447396944477,
+      "learning_rate": 8.088622648398357e-06,
+      "loss": 0.6662,
+      "step": 2941
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8646626316625636,
+      "learning_rate": 8.087282265378596e-06,
+      "loss": 0.6547,
+      "step": 2942
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.999962898982827,
+      "learning_rate": 8.085941523687309e-06,
+      "loss": 0.6913,
+      "step": 2943
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.793000783068167,
+      "learning_rate": 8.084600423480253e-06,
+      "loss": 0.7647,
+      "step": 2944
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.294318321894251,
+      "learning_rate": 8.083258964913238e-06,
+      "loss": 0.6202,
+      "step": 2945
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3681448665076648,
+      "learning_rate": 8.08191714814211e-06,
+      "loss": 0.6897,
+      "step": 2946
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.4035241766844258,
+      "learning_rate": 8.080574973322755e-06,
+      "loss": 0.6347,
+      "step": 2947
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.697359357745482,
+      "learning_rate": 8.079232440611106e-06,
+      "loss": 0.6607,
+      "step": 2948
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 11.08038730758078,
+      "learning_rate": 8.077889550163133e-06,
+      "loss": 0.6913,
+      "step": 2949
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.289498594594031,
+      "learning_rate": 8.076546302134849e-06,
+      "loss": 0.5815,
+      "step": 2950
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.8411656849036024,
+      "learning_rate": 8.07520269668231e-06,
+      "loss": 0.6598,
+      "step": 2951
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.9514846702421127,
+      "learning_rate": 8.073858733961609e-06,
+      "loss": 0.7163,
+      "step": 2952
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.287925053791738,
+      "learning_rate": 8.072514414128886e-06,
+      "loss": 0.6914,
+      "step": 2953
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.850131167037012,
+      "learning_rate": 8.071169737340322e-06,
+      "loss": 0.6972,
+      "step": 2954
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.3564428711936714,
+      "learning_rate": 8.069824703752136e-06,
+      "loss": 0.6299,
+      "step": 2955
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5337256748284793,
+      "learning_rate": 8.068479313520589e-06,
+      "loss": 0.7192,
+      "step": 2956
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7829671234993767,
+      "learning_rate": 8.067133566801986e-06,
+      "loss": 0.6942,
+      "step": 2957
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0780064728452814,
+      "learning_rate": 8.06578746375267e-06,
+      "loss": 0.699,
+      "step": 2958
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.066654395430205,
+      "learning_rate": 8.06444100452903e-06,
+      "loss": 0.6052,
+      "step": 2959
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3241867147767636,
+      "learning_rate": 8.063094189287492e-06,
+      "loss": 0.6425,
+      "step": 2960
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.936750856792946,
+      "learning_rate": 8.061747018184525e-06,
+      "loss": 0.6816,
+      "step": 2961
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.564282733037237,
+      "learning_rate": 8.06039949137664e-06,
+      "loss": 0.7117,
+      "step": 2962
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 35.37441915823146,
+      "learning_rate": 8.05905160902039e-06,
+      "loss": 0.7604,
+      "step": 2963
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.470994439733572,
+      "learning_rate": 8.057703371272368e-06,
+      "loss": 0.6648,
+      "step": 2964
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.653406363973704,
+      "learning_rate": 8.056354778289204e-06,
+      "loss": 0.6818,
+      "step": 2965
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7162242932530374,
+      "learning_rate": 8.055005830227578e-06,
+      "loss": 0.6176,
+      "step": 2966
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0336246489108563,
+      "learning_rate": 8.053656527244206e-06,
+      "loss": 0.6957,
+      "step": 2967
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.404566189632472,
+      "learning_rate": 8.052306869495847e-06,
+      "loss": 0.6862,
+      "step": 2968
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.662561717072185,
+      "learning_rate": 8.050956857139298e-06,
+      "loss": 0.7162,
+      "step": 2969
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5875784364886254,
+      "learning_rate": 8.049606490331403e-06,
+      "loss": 0.6401,
+      "step": 2970
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.978839811791059,
+      "learning_rate": 8.048255769229038e-06,
+      "loss": 0.6423,
+      "step": 2971
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.159696497187548,
+      "learning_rate": 8.046904693989132e-06,
+      "loss": 0.6541,
+      "step": 2972
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.108106082999746,
+      "learning_rate": 8.045553264768645e-06,
+      "loss": 0.7498,
+      "step": 2973
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4365012913470747,
+      "learning_rate": 8.044201481724582e-06,
+      "loss": 0.6099,
+      "step": 2974
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9007518834879593,
+      "learning_rate": 8.042849345013995e-06,
+      "loss": 0.6726,
+      "step": 2975
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2970067282582476,
+      "learning_rate": 8.041496854793964e-06,
+      "loss": 0.5851,
+      "step": 2976
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.006328818924341,
+      "learning_rate": 8.040144011221621e-06,
+      "loss": 0.6791,
+      "step": 2977
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6939058418287805,
+      "learning_rate": 8.038790814454137e-06,
+      "loss": 0.726,
+      "step": 2978
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2054039052694705,
+      "learning_rate": 8.037437264648717e-06,
+      "loss": 0.6719,
+      "step": 2979
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.41222955568751,
+      "learning_rate": 8.036083361962616e-06,
+      "loss": 0.6879,
+      "step": 2980
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4774726350606158,
+      "learning_rate": 8.03472910655313e-06,
+      "loss": 0.7345,
+      "step": 2981
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3042323688271327,
+      "learning_rate": 8.033374498577586e-06,
+      "loss": 0.7063,
+      "step": 2982
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7730077023368938,
+      "learning_rate": 8.032019538193363e-06,
+      "loss": 0.7008,
+      "step": 2983
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7189215222404144,
+      "learning_rate": 8.030664225557873e-06,
+      "loss": 0.6169,
+      "step": 2984
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.845427019172591,
+      "learning_rate": 8.029308560828574e-06,
+      "loss": 0.7042,
+      "step": 2985
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.311773529022882,
+      "learning_rate": 8.027952544162965e-06,
+      "loss": 0.7276,
+      "step": 2986
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.498629343600977,
+      "learning_rate": 8.026596175718582e-06,
+      "loss": 0.65,
+      "step": 2987
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.527545636273197,
+      "learning_rate": 8.025239455653003e-06,
+      "loss": 0.6596,
+      "step": 2988
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3431814650204354,
+      "learning_rate": 8.023882384123851e-06,
+      "loss": 0.5784,
+      "step": 2989
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.266244506572552,
+      "learning_rate": 8.022524961288783e-06,
+      "loss": 0.6536,
+      "step": 2990
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4413399629821395,
+      "learning_rate": 8.021167187305504e-06,
+      "loss": 0.648,
+      "step": 2991
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.8044222712960485,
+      "learning_rate": 8.019809062331754e-06,
+      "loss": 0.6785,
+      "step": 2992
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8009795918380527,
+      "learning_rate": 8.018450586525314e-06,
+      "loss": 0.6825,
+      "step": 2993
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.044733004957816,
+      "learning_rate": 8.017091760044014e-06,
+      "loss": 0.7218,
+      "step": 2994
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5348478649648625,
+      "learning_rate": 8.015732583045713e-06,
+      "loss": 0.6306,
+      "step": 2995
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.820292578056801,
+      "learning_rate": 8.014373055688319e-06,
+      "loss": 0.6103,
+      "step": 2996
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.83941151917699,
+      "learning_rate": 8.013013178129775e-06,
+      "loss": 0.739,
+      "step": 2997
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.562461642222233,
+      "learning_rate": 8.01165295052807e-06,
+      "loss": 0.6528,
+      "step": 2998
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2991584011557946,
+      "learning_rate": 8.010292373041233e-06,
+      "loss": 0.6633,
+      "step": 2999
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0164089815235755,
+      "learning_rate": 8.008931445827329e-06,
+      "loss": 0.6508,
+      "step": 3000
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2488209086571036,
+      "learning_rate": 8.007570169044467e-06,
+      "loss": 0.6924,
+      "step": 3001
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.067344347818427,
+      "learning_rate": 8.006208542850797e-06,
+      "loss": 0.716,
+      "step": 3002
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.433534837689038,
+      "learning_rate": 8.004846567404509e-06,
+      "loss": 0.6682,
+      "step": 3003
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1675254720528176,
+      "learning_rate": 8.003484242863833e-06,
+      "loss": 0.744,
+      "step": 3004
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3134441373453916,
+      "learning_rate": 8.00212156938704e-06,
+      "loss": 0.5428,
+      "step": 3005
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.8236083439191355,
+      "learning_rate": 8.000758547132441e-06,
+      "loss": 0.66,
+      "step": 3006
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.158397208676714,
+      "learning_rate": 7.99939517625839e-06,
+      "loss": 0.7367,
+      "step": 3007
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2654650330105257,
+      "learning_rate": 7.998031456923274e-06,
+      "loss": 0.5894,
+      "step": 3008
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5662295257161833,
+      "learning_rate": 7.996667389285532e-06,
+      "loss": 0.6596,
+      "step": 3009
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.677019268255404,
+      "learning_rate": 7.995302973503636e-06,
+      "loss": 0.6978,
+      "step": 3010
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.371436079489015,
+      "learning_rate": 7.993938209736097e-06,
+      "loss": 0.69,
+      "step": 3011
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1346580337024186,
+      "learning_rate": 7.992573098141472e-06,
+      "loss": 0.644,
+      "step": 3012
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9295251025511633,
+      "learning_rate": 7.991207638878356e-06,
+      "loss": 0.6617,
+      "step": 3013
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.558479337294168,
+      "learning_rate": 7.989841832105382e-06,
+      "loss": 0.7159,
+      "step": 3014
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.147625038233415,
+      "learning_rate": 7.988475677981229e-06,
+      "loss": 0.6179,
+      "step": 3015
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1693505791045053,
+      "learning_rate": 7.98710917666461e-06,
+      "loss": 0.6942,
+      "step": 3016
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1568770622983475,
+      "learning_rate": 7.985742328314279e-06,
+      "loss": 0.7053,
+      "step": 3017
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1959613312211084,
+      "learning_rate": 7.984375133089038e-06,
+      "loss": 0.637,
+      "step": 3018
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.895723607078637,
+      "learning_rate": 7.98300759114772e-06,
+      "loss": 0.7098,
+      "step": 3019
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.234596285844836,
+      "learning_rate": 7.981639702649204e-06,
+      "loss": 0.6282,
+      "step": 3020
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2374506416442133,
+      "learning_rate": 7.980271467752405e-06,
+      "loss": 0.6773,
+      "step": 3021
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6097297477113326,
+      "learning_rate": 7.97890288661628e-06,
+      "loss": 0.7456,
+      "step": 3022
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.8225474093932426,
+      "learning_rate": 7.977533959399833e-06,
+      "loss": 0.68,
+      "step": 3023
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6648211533685795,
+      "learning_rate": 7.976164686262096e-06,
+      "loss": 0.686,
+      "step": 3024
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3486556822537206,
+      "learning_rate": 7.974795067362148e-06,
+      "loss": 0.7185,
+      "step": 3025
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.529824254487311,
+      "learning_rate": 7.97342510285911e-06,
+      "loss": 0.6447,
+      "step": 3026
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7391077394652763,
+      "learning_rate": 7.972054792912138e-06,
+      "loss": 0.6169,
+      "step": 3027
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.237331937538726,
+      "learning_rate": 7.970684137680431e-06,
+      "loss": 0.6488,
+      "step": 3028
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2823597001812614,
+      "learning_rate": 7.969313137323228e-06,
+      "loss": 0.6577,
+      "step": 3029
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6046928602923143,
+      "learning_rate": 7.96794179199981e-06,
+      "loss": 0.682,
+      "step": 3030
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.018838373919387,
+      "learning_rate": 7.966570101869494e-06,
+      "loss": 0.6314,
+      "step": 3031
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.190243611941645,
+      "learning_rate": 7.965198067091637e-06,
+      "loss": 0.6224,
+      "step": 3032
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4033899287674183,
+      "learning_rate": 7.96382568782564e-06,
+      "loss": 0.6538,
+      "step": 3033
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.57052475740044,
+      "learning_rate": 7.962452964230944e-06,
+      "loss": 0.6519,
+      "step": 3034
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0579037048889752,
+      "learning_rate": 7.961079896467025e-06,
+      "loss": 0.5823,
+      "step": 3035
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.8012393029432645,
+      "learning_rate": 7.959706484693405e-06,
+      "loss": 0.6989,
+      "step": 3036
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7558779760997525,
+      "learning_rate": 7.95833272906964e-06,
+      "loss": 0.6581,
+      "step": 3037
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1874011393784665,
+      "learning_rate": 7.95695862975533e-06,
+      "loss": 0.6699,
+      "step": 3038
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6164914345320107,
+      "learning_rate": 7.955584186910115e-06,
+      "loss": 0.7054,
+      "step": 3039
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0292864485504594,
+      "learning_rate": 7.954209400693673e-06,
+      "loss": 0.6537,
+      "step": 3040
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.905511291216553,
+      "learning_rate": 7.95283427126572e-06,
+      "loss": 0.6966,
+      "step": 3041
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.208663400211376,
+      "learning_rate": 7.95145879878602e-06,
+      "loss": 0.5702,
+      "step": 3042
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.559598093856362,
+      "learning_rate": 7.950082983414367e-06,
+      "loss": 0.6517,
+      "step": 3043
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.389487342933906,
+      "learning_rate": 7.948706825310601e-06,
+      "loss": 0.6289,
+      "step": 3044
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.675926284212583,
+      "learning_rate": 7.947330324634601e-06,
+      "loss": 0.6088,
+      "step": 3045
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.648358995913935,
+      "learning_rate": 7.945953481546282e-06,
+      "loss": 0.6291,
+      "step": 3046
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2023556963692583,
+      "learning_rate": 7.944576296205603e-06,
+      "loss": 0.5934,
+      "step": 3047
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.627906298885444,
+      "learning_rate": 7.943198768772565e-06,
+      "loss": 0.743,
+      "step": 3048
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.399467853821857,
+      "learning_rate": 7.9418208994072e-06,
+      "loss": 0.6587,
+      "step": 3049
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6749330583941258,
+      "learning_rate": 7.940442688269587e-06,
+      "loss": 0.6825,
+      "step": 3050
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9142489850897517,
+      "learning_rate": 7.939064135519844e-06,
+      "loss": 0.6535,
+      "step": 3051
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3087341250267763,
+      "learning_rate": 7.937685241318122e-06,
+      "loss": 0.6461,
+      "step": 3052
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.292432896926418,
+      "learning_rate": 7.936306005824624e-06,
+      "loss": 0.6497,
+      "step": 3053
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.3219986590091435,
+      "learning_rate": 7.93492642919958e-06,
+      "loss": 0.6326,
+      "step": 3054
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.119470589054828,
+      "learning_rate": 7.933546511603269e-06,
+      "loss": 0.6417,
+      "step": 3055
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.774960844710111,
+      "learning_rate": 7.932166253196004e-06,
+      "loss": 0.5901,
+      "step": 3056
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4456975659415847,
+      "learning_rate": 7.93078565413814e-06,
+      "loss": 0.6138,
+      "step": 3057
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.8996471024029513,
+      "learning_rate": 7.92940471459007e-06,
+      "loss": 0.6084,
+      "step": 3058
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1348588208140797,
+      "learning_rate": 7.928023434712227e-06,
+      "loss": 0.6235,
+      "step": 3059
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1903589219833037,
+      "learning_rate": 7.926641814665088e-06,
+      "loss": 0.6413,
+      "step": 3060
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6063098506456126,
+      "learning_rate": 7.925259854609162e-06,
+      "loss": 0.6501,
+      "step": 3061
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.68937491835249,
+      "learning_rate": 7.923877554705002e-06,
+      "loss": 0.632,
+      "step": 3062
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2423734100409196,
+      "learning_rate": 7.9224949151132e-06,
+      "loss": 0.7238,
+      "step": 3063
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.607658237678769,
+      "learning_rate": 7.921111935994388e-06,
+      "loss": 0.6373,
+      "step": 3064
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.727385481962859,
+      "learning_rate": 7.919728617509233e-06,
+      "loss": 0.6826,
+      "step": 3065
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6366816990101145,
+      "learning_rate": 7.91834495981845e-06,
+      "loss": 0.6223,
+      "step": 3066
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.388243265491952,
+      "learning_rate": 7.916960963082783e-06,
+      "loss": 0.6517,
+      "step": 3067
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.384651407591471,
+      "learning_rate": 7.915576627463024e-06,
+      "loss": 0.6422,
+      "step": 3068
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.315078446495304,
+      "learning_rate": 7.91419195312e-06,
+      "loss": 0.7492,
+      "step": 3069
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5016948568104214,
+      "learning_rate": 7.91280694021458e-06,
+      "loss": 0.5617,
+      "step": 3070
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.267739297104988,
+      "learning_rate": 7.91142158890767e-06,
+      "loss": 0.6419,
+      "step": 3071
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9294081180768607,
+      "learning_rate": 7.910035899360215e-06,
+      "loss": 0.6419,
+      "step": 3072
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5533958607494083,
+      "learning_rate": 7.908649871733202e-06,
+      "loss": 0.6337,
+      "step": 3073
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.016887449204389,
+      "learning_rate": 7.907263506187655e-06,
+      "loss": 0.577,
+      "step": 3074
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.229569058131735,
+      "learning_rate": 7.905876802884639e-06,
+      "loss": 0.6829,
+      "step": 3075
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.650463843334712,
+      "learning_rate": 7.904489761985254e-06,
+      "loss": 0.6764,
+      "step": 3076
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5592076303986047,
+      "learning_rate": 7.903102383650645e-06,
+      "loss": 0.623,
+      "step": 3077
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.032748852875668,
+      "learning_rate": 7.901714668041993e-06,
+      "loss": 0.6891,
+      "step": 3078
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.317465920800203,
+      "learning_rate": 7.90032661532052e-06,
+      "loss": 0.6487,
+      "step": 3079
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.305967746812263,
+      "learning_rate": 7.898938225647484e-06,
+      "loss": 0.6823,
+      "step": 3080
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.8884399615654206,
+      "learning_rate": 7.897549499184184e-06,
+      "loss": 0.6296,
+      "step": 3081
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0770385988445907,
+      "learning_rate": 7.896160436091961e-06,
+      "loss": 0.6943,
+      "step": 3082
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4056547215114055,
+      "learning_rate": 7.894771036532189e-06,
+      "loss": 0.5929,
+      "step": 3083
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4323552445732304,
+      "learning_rate": 7.893381300666287e-06,
+      "loss": 0.6887,
+      "step": 3084
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.343998812827096,
+      "learning_rate": 7.89199122865571e-06,
+      "loss": 0.6709,
+      "step": 3085
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0146219044287164,
+      "learning_rate": 7.89060082066195e-06,
+      "loss": 0.6724,
+      "step": 3086
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6033290148752357,
+      "learning_rate": 7.889210076846544e-06,
+      "loss": 0.6815,
+      "step": 3087
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.381411272217176,
+      "learning_rate": 7.887818997371062e-06,
+      "loss": 0.6404,
+      "step": 3088
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.904456744874637,
+      "learning_rate": 7.886427582397117e-06,
+      "loss": 0.7417,
+      "step": 3089
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2786773692260547,
+      "learning_rate": 7.88503583208636e-06,
+      "loss": 0.5864,
+      "step": 3090
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.81932535707139,
+      "learning_rate": 7.88364374660048e-06,
+      "loss": 0.7001,
+      "step": 3091
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6849729938128997,
+      "learning_rate": 7.882251326101205e-06,
+      "loss": 0.6875,
+      "step": 3092
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.3310488217574488,
+      "learning_rate": 7.8808585707503e-06,
+      "loss": 0.698,
+      "step": 3093
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6002798711965975,
+      "learning_rate": 7.879465480709577e-06,
+      "loss": 0.6669,
+      "step": 3094
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8876056113874142,
+      "learning_rate": 7.878072056140878e-06,
+      "loss": 0.6834,
+      "step": 3095
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.537815529224319,
+      "learning_rate": 7.876678297206086e-06,
+      "loss": 0.7348,
+      "step": 3096
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8620170280847104,
+      "learning_rate": 7.875284204067127e-06,
+      "loss": 0.6779,
+      "step": 3097
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.283985994119008,
+      "learning_rate": 7.873889776885959e-06,
+      "loss": 0.6443,
+      "step": 3098
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.3105599250764297,
+      "learning_rate": 7.872495015824586e-06,
+      "loss": 0.6744,
+      "step": 3099
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9406649266002853,
+      "learning_rate": 7.871099921045042e-06,
+      "loss": 0.6953,
+      "step": 3100
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.574644681179438,
+      "learning_rate": 7.869704492709412e-06,
+      "loss": 0.6711,
+      "step": 3101
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.449965941126261,
+      "learning_rate": 7.868308730979809e-06,
+      "loss": 0.628,
+      "step": 3102
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.772397515441643,
+      "learning_rate": 7.866912636018389e-06,
+      "loss": 0.7371,
+      "step": 3103
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.39442828453295,
+      "learning_rate": 7.865516207987344e-06,
+      "loss": 0.63,
+      "step": 3104
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.3246795963005713,
+      "learning_rate": 7.864119447048912e-06,
+      "loss": 0.7073,
+      "step": 3105
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0064966422057773,
+      "learning_rate": 7.862722353365361e-06,
+      "loss": 0.7265,
+      "step": 3106
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0807358878776006,
+      "learning_rate": 7.861324927099004e-06,
+      "loss": 0.7145,
+      "step": 3107
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.069135073025866,
+      "learning_rate": 7.859927168412186e-06,
+      "loss": 0.6135,
+      "step": 3108
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.286312915224411,
+      "learning_rate": 7.858529077467298e-06,
+      "loss": 0.6363,
+      "step": 3109
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2389602662700385,
+      "learning_rate": 7.857130654426764e-06,
+      "loss": 0.6131,
+      "step": 3110
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.36015735420768,
+      "learning_rate": 7.85573189945305e-06,
+      "loss": 0.5841,
+      "step": 3111
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7462815401691096,
+      "learning_rate": 7.854332812708661e-06,
+      "loss": 0.6193,
+      "step": 3112
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0603590845258815,
+      "learning_rate": 7.852933394356134e-06,
+      "loss": 0.7154,
+      "step": 3113
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6664238931594584,
+      "learning_rate": 7.851533644558054e-06,
+      "loss": 0.747,
+      "step": 3114
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.859101742503454,
+      "learning_rate": 7.850133563477037e-06,
+      "loss": 0.6958,
+      "step": 3115
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.669270847924662,
+      "learning_rate": 7.848733151275741e-06,
+      "loss": 0.5718,
+      "step": 3116
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.7219654697356863,
+      "learning_rate": 7.847332408116863e-06,
+      "loss": 0.6531,
+      "step": 3117
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.7912563805814945,
+      "learning_rate": 7.845931334163138e-06,
+      "loss": 0.7386,
+      "step": 3118
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8886193603297348,
+      "learning_rate": 7.844529929577336e-06,
+      "loss": 0.6827,
+      "step": 3119
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.586772606669079,
+      "learning_rate": 7.84312819452227e-06,
+      "loss": 0.6084,
+      "step": 3120
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6184892406372193,
+      "learning_rate": 7.841726129160789e-06,
+      "loss": 0.6775,
+      "step": 3121
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3983415224825637,
+      "learning_rate": 7.84032373365578e-06,
+      "loss": 0.6357,
+      "step": 3122
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1697316223767333,
+      "learning_rate": 7.838921008170171e-06,
+      "loss": 0.7038,
+      "step": 3123
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.538074643605018,
+      "learning_rate": 7.837517952866924e-06,
+      "loss": 0.6816,
+      "step": 3124
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.723960615531358,
+      "learning_rate": 7.836114567909046e-06,
+      "loss": 0.6351,
+      "step": 3125
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.3141201817543715,
+      "learning_rate": 7.834710853459575e-06,
+      "loss": 0.6886,
+      "step": 3126
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.935659578134951,
+      "learning_rate": 7.833306809681593e-06,
+      "loss": 0.6418,
+      "step": 3127
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8856037758667292,
+      "learning_rate": 7.831902436738215e-06,
+      "loss": 0.6849,
+      "step": 3128
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.563709509210556,
+      "learning_rate": 7.830497734792597e-06,
+      "loss": 0.6117,
+      "step": 3129
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.927718266986384,
+      "learning_rate": 7.829092704007935e-06,
+      "loss": 0.6543,
+      "step": 3130
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.675602881652366,
+      "learning_rate": 7.827687344547459e-06,
+      "loss": 0.6443,
+      "step": 3131
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.732964806627827,
+      "learning_rate": 7.826281656574444e-06,
+      "loss": 0.6704,
+      "step": 3132
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9065789120991763,
+      "learning_rate": 7.824875640252195e-06,
+      "loss": 0.7202,
+      "step": 3133
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5054335560181444,
+      "learning_rate": 7.82346929574406e-06,
+      "loss": 0.7212,
+      "step": 3134
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5085507480711065,
+      "learning_rate": 7.822062623213424e-06,
+      "loss": 0.7014,
+      "step": 3135
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.434812127737116,
+      "learning_rate": 7.820655622823712e-06,
+      "loss": 0.6936,
+      "step": 3136
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.281720661318549,
+      "learning_rate": 7.819248294738381e-06,
+      "loss": 0.6299,
+      "step": 3137
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.336679866157924,
+      "learning_rate": 7.817840639120932e-06,
+      "loss": 0.6982,
+      "step": 3138
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7369577275240657,
+      "learning_rate": 7.816432656134907e-06,
+      "loss": 0.6836,
+      "step": 3139
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.408523586518461,
+      "learning_rate": 7.815024345943874e-06,
+      "loss": 0.6304,
+      "step": 3140
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7339532046846244,
+      "learning_rate": 7.81361570871145e-06,
+      "loss": 0.6901,
+      "step": 3141
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.55643397634351,
+      "learning_rate": 7.812206744601288e-06,
+      "loss": 0.7034,
+      "step": 3142
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7516095557400058,
+      "learning_rate": 7.810797453777076e-06,
+      "loss": 0.5308,
+      "step": 3143
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5846496687334355,
+      "learning_rate": 7.80938783640254e-06,
+      "loss": 0.7117,
+      "step": 3144
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8518803010333,
+      "learning_rate": 7.807977892641446e-06,
+      "loss": 0.7384,
+      "step": 3145
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.978801136142066,
+      "learning_rate": 7.806567622657598e-06,
+      "loss": 0.6169,
+      "step": 3146
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.8381379717790853,
+      "learning_rate": 7.805157026614836e-06,
+      "loss": 0.6696,
+      "step": 3147
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6451501046738706,
+      "learning_rate": 7.80374610467704e-06,
+      "loss": 0.695,
+      "step": 3148
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6556926713393936,
+      "learning_rate": 7.802334857008127e-06,
+      "loss": 0.5896,
+      "step": 3149
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3511219203796054,
+      "learning_rate": 7.800923283772051e-06,
+      "loss": 0.6078,
+      "step": 3150
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3055445730449935,
+      "learning_rate": 7.799511385132803e-06,
+      "loss": 0.7246,
+      "step": 3151
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.807571982019713,
+      "learning_rate": 7.798099161254415e-06,
+      "loss": 0.5812,
+      "step": 3152
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9073765971064183,
+      "learning_rate": 7.796686612300957e-06,
+      "loss": 0.5974,
+      "step": 3153
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.580983486608408,
+      "learning_rate": 7.795273738436531e-06,
+      "loss": 0.6505,
+      "step": 3154
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.27033419980578,
+      "learning_rate": 7.793860539825282e-06,
+      "loss": 0.6077,
+      "step": 3155
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.736113101860133,
+      "learning_rate": 7.792447016631392e-06,
+      "loss": 0.5951,
+      "step": 3156
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1064869720748733,
+      "learning_rate": 7.79103316901908e-06,
+      "loss": 0.6632,
+      "step": 3157
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0989709229973887,
+      "learning_rate": 7.789618997152603e-06,
+      "loss": 0.7188,
+      "step": 3158
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.06812401144027,
+      "learning_rate": 7.788204501196255e-06,
+      "loss": 0.6171,
+      "step": 3159
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.691041503384547,
+      "learning_rate": 7.786789681314368e-06,
+      "loss": 0.604,
+      "step": 3160
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.614511086385431,
+      "learning_rate": 7.785374537671311e-06,
+      "loss": 0.6663,
+      "step": 3161
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6590381539539054,
+      "learning_rate": 7.783959070431492e-06,
+      "loss": 0.7339,
+      "step": 3162
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1374038834717708,
+      "learning_rate": 7.782543279759356e-06,
+      "loss": 0.5921,
+      "step": 3163
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7699969206806414,
+      "learning_rate": 7.781127165819386e-06,
+      "loss": 0.767,
+      "step": 3164
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0646485786942144,
+      "learning_rate": 7.7797107287761e-06,
+      "loss": 0.657,
+      "step": 3165
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.383226812212046,
+      "learning_rate": 7.778293968794056e-06,
+      "loss": 0.6619,
+      "step": 3166
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3946861066654246,
+      "learning_rate": 7.776876886037852e-06,
+      "loss": 0.7148,
+      "step": 3167
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.2920273098631965,
+      "learning_rate": 7.775459480672117e-06,
+      "loss": 0.6964,
+      "step": 3168
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.74757852495417,
+      "learning_rate": 7.774041752861524e-06,
+      "loss": 0.6485,
+      "step": 3169
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.8429087492176057,
+      "learning_rate": 7.772623702770779e-06,
+      "loss": 0.7287,
+      "step": 3170
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6463677367731218,
+      "learning_rate": 7.771205330564626e-06,
+      "loss": 0.6197,
+      "step": 3171
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3146582536567784,
+      "learning_rate": 7.769786636407849e-06,
+      "loss": 0.673,
+      "step": 3172
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.097668002633075,
+      "learning_rate": 7.768367620465267e-06,
+      "loss": 0.6628,
+      "step": 3173
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6251746945849934,
+      "learning_rate": 7.766948282901738e-06,
+      "loss": 0.6698,
+      "step": 3174
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7114682635233325,
+      "learning_rate": 7.765528623882155e-06,
+      "loss": 0.6594,
+      "step": 3175
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.4612588856168838,
+      "learning_rate": 7.76410864357145e-06,
+      "loss": 0.6582,
+      "step": 3176
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.874426831691072,
+      "learning_rate": 7.762688342134597e-06,
+      "loss": 0.6661,
+      "step": 3177
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9148574564549485,
+      "learning_rate": 7.761267719736593e-06,
+      "loss": 0.6976,
+      "step": 3178
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0616632712894885,
+      "learning_rate": 7.759846776542492e-06,
+      "loss": 0.6677,
+      "step": 3179
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.561508539134428,
+      "learning_rate": 7.75842551271737e-06,
+      "loss": 0.6977,
+      "step": 3180
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.298240212255653,
+      "learning_rate": 7.757003928426342e-06,
+      "loss": 0.6149,
+      "step": 3181
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.4188530769511982,
+      "learning_rate": 7.755582023834572e-06,
+      "loss": 0.6496,
+      "step": 3182
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8398810101886345,
+      "learning_rate": 7.754159799107244e-06,
+      "loss": 0.6351,
+      "step": 3183
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2499424827555927,
+      "learning_rate": 7.752737254409594e-06,
+      "loss": 0.7088,
+      "step": 3184
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3814957039535374,
+      "learning_rate": 7.751314389906887e-06,
+      "loss": 0.6245,
+      "step": 3185
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.666046394806816,
+      "learning_rate": 7.749891205764427e-06,
+      "loss": 0.6133,
+      "step": 3186
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.7799564388267295,
+      "learning_rate": 7.748467702147555e-06,
+      "loss": 0.605,
+      "step": 3187
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.281029160468379,
+      "learning_rate": 7.747043879221653e-06,
+      "loss": 0.6321,
+      "step": 3188
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.934308558200218,
+      "learning_rate": 7.745619737152133e-06,
+      "loss": 0.7053,
+      "step": 3189
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1545251666263026,
+      "learning_rate": 7.744195276104447e-06,
+      "loss": 0.6998,
+      "step": 3190
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.642275030044787,
+      "learning_rate": 7.742770496244087e-06,
+      "loss": 0.7263,
+      "step": 3191
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9956469346760177,
+      "learning_rate": 7.74134539773658e-06,
+      "loss": 0.6707,
+      "step": 3192
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.099408512549327,
+      "learning_rate": 7.73991998074749e-06,
+      "loss": 0.6832,
+      "step": 3193
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.473717647736508,
+      "learning_rate": 7.738494245442415e-06,
+      "loss": 0.5969,
+      "step": 3194
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.67068434985646,
+      "learning_rate": 7.737068191986995e-06,
+      "loss": 0.7505,
+      "step": 3195
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.413731979539875,
+      "learning_rate": 7.735641820546906e-06,
+      "loss": 0.7376,
+      "step": 3196
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.619766771399215,
+      "learning_rate": 7.73421513128786e-06,
+      "loss": 0.712,
+      "step": 3197
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0705943862250664,
+      "learning_rate": 7.7327881243756e-06,
+      "loss": 0.701,
+      "step": 3198
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3139029563356512,
+      "learning_rate": 7.731360799975916e-06,
+      "loss": 0.6423,
+      "step": 3199
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4451859469193082,
+      "learning_rate": 7.72993315825463e-06,
+      "loss": 0.6318,
+      "step": 3200
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2475814172050104,
+      "learning_rate": 7.728505199377603e-06,
+      "loss": 0.6631,
+      "step": 3201
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.3486986779317154,
+      "learning_rate": 7.727076923510727e-06,
+      "loss": 0.7424,
+      "step": 3202
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.121154876974287,
+      "learning_rate": 7.72564833081994e-06,
+      "loss": 0.6947,
+      "step": 3203
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.4236551024113093,
+      "learning_rate": 7.724219421471206e-06,
+      "loss": 0.727,
+      "step": 3204
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.346428595824038,
+      "learning_rate": 7.722790195630536e-06,
+      "loss": 0.6948,
+      "step": 3205
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.092820968756066,
+      "learning_rate": 7.721360653463971e-06,
+      "loss": 0.6849,
+      "step": 3206
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5447700966964466,
+      "learning_rate": 7.719930795137592e-06,
+      "loss": 0.7097,
+      "step": 3207
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2830897925687825,
+      "learning_rate": 7.718500620817517e-06,
+      "loss": 0.6177,
+      "step": 3208
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8208598121038104,
+      "learning_rate": 7.717070130669896e-06,
+      "loss": 0.6102,
+      "step": 3209
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9927032819623642,
+      "learning_rate": 7.715639324860925e-06,
+      "loss": 0.6379,
+      "step": 3210
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6289931425920803,
+      "learning_rate": 7.714208203556825e-06,
+      "loss": 0.7293,
+      "step": 3211
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.899374378134415,
+      "learning_rate": 7.712776766923862e-06,
+      "loss": 0.6116,
+      "step": 3212
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2414892114848803,
+      "learning_rate": 7.711345015128335e-06,
+      "loss": 0.7262,
+      "step": 3213
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.45926603241094,
+      "learning_rate": 7.709912948336583e-06,
+      "loss": 0.7132,
+      "step": 3214
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.182829948766781,
+      "learning_rate": 7.70848056671498e-06,
+      "loss": 0.6255,
+      "step": 3215
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.7748630647111376,
+      "learning_rate": 7.707047870429931e-06,
+      "loss": 0.6655,
+      "step": 3216
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.9189572306348666,
+      "learning_rate": 7.705614859647888e-06,
+      "loss": 0.6888,
+      "step": 3217
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1326017663302763,
+      "learning_rate": 7.704181534535332e-06,
+      "loss": 0.6407,
+      "step": 3218
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.352263696228539,
+      "learning_rate": 7.70274789525878e-06,
+      "loss": 0.6605,
+      "step": 3219
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.035499997279176,
+      "learning_rate": 7.701313941984791e-06,
+      "loss": 0.6789,
+      "step": 3220
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.629717191952034,
+      "learning_rate": 7.699879674879958e-06,
+      "loss": 0.5778,
+      "step": 3221
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.130587806018568,
+      "learning_rate": 7.698445094110909e-06,
+      "loss": 0.7025,
+      "step": 3222
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8760063671669753,
+      "learning_rate": 7.697010199844308e-06,
+      "loss": 0.7081,
+      "step": 3223
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.7331704381503226,
+      "learning_rate": 7.69557499224686e-06,
+      "loss": 0.7452,
+      "step": 3224
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.850116734761613,
+      "learning_rate": 7.694139471485301e-06,
+      "loss": 0.6846,
+      "step": 3225
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.489662975427095,
+      "learning_rate": 7.692703637726407e-06,
+      "loss": 0.7061,
+      "step": 3226
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.903189368179508,
+      "learning_rate": 7.691267491136986e-06,
+      "loss": 0.5947,
+      "step": 3227
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.316176181783629,
+      "learning_rate": 7.689831031883887e-06,
+      "loss": 0.6344,
+      "step": 3228
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4581575403347564,
+      "learning_rate": 7.688394260133997e-06,
+      "loss": 0.6042,
+      "step": 3229
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.424492880509628,
+      "learning_rate": 7.686957176054231e-06,
+      "loss": 0.6886,
+      "step": 3230
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4711470411348553,
+      "learning_rate": 7.68551977981155e-06,
+      "loss": 0.5719,
+      "step": 3231
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4487165637771358,
+      "learning_rate": 7.684082071572943e-06,
+      "loss": 0.7407,
+      "step": 3232
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8173730592149235,
+      "learning_rate": 7.68264405150544e-06,
+      "loss": 0.6361,
+      "step": 3233
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.721164285756093,
+      "learning_rate": 7.681205719776104e-06,
+      "loss": 0.7631,
+      "step": 3234
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.9251106459657077,
+      "learning_rate": 7.679767076552038e-06,
+      "loss": 0.6352,
+      "step": 3235
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0737715770377105,
+      "learning_rate": 7.678328122000382e-06,
+      "loss": 0.6233,
+      "step": 3236
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5541768150933235,
+      "learning_rate": 7.676888856288307e-06,
+      "loss": 0.6348,
+      "step": 3237
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.580142219433405,
+      "learning_rate": 7.67544927958302e-06,
+      "loss": 0.6439,
+      "step": 3238
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.654991640432859,
+      "learning_rate": 7.67400939205177e-06,
+      "loss": 0.6801,
+      "step": 3239
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.6174398276239623,
+      "learning_rate": 7.67256919386184e-06,
+      "loss": 0.7203,
+      "step": 3240
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.5610803553484955,
+      "learning_rate": 7.671128685180547e-06,
+      "loss": 0.6511,
+      "step": 3241
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.938858056226461,
+      "learning_rate": 7.669687866175245e-06,
+      "loss": 0.6039,
+      "step": 3242
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8043292999636313,
+      "learning_rate": 7.668246737013323e-06,
+      "loss": 0.6662,
+      "step": 3243
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3949517216338907,
+      "learning_rate": 7.666805297862208e-06,
+      "loss": 0.6496,
+      "step": 3244
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1462781898491157,
+      "learning_rate": 7.665363548889362e-06,
+      "loss": 0.6178,
+      "step": 3245
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5506311534896433,
+      "learning_rate": 7.663921490262286e-06,
+      "loss": 0.6798,
+      "step": 3246
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.48469185368361,
+      "learning_rate": 7.66247912214851e-06,
+      "loss": 0.623,
+      "step": 3247
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.841001179637787,
+      "learning_rate": 7.661036444715608e-06,
+      "loss": 0.6621,
+      "step": 3248
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1276649407769432,
+      "learning_rate": 7.659593458131181e-06,
+      "loss": 0.6079,
+      "step": 3249
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.967235234356677,
+      "learning_rate": 7.658150162562875e-06,
+      "loss": 0.6655,
+      "step": 3250
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.935135934795403,
+      "learning_rate": 7.656706558178368e-06,
+      "loss": 0.6983,
+      "step": 3251
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.7507617496596493,
+      "learning_rate": 7.655262645145374e-06,
+      "loss": 0.6997,
+      "step": 3252
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.6113941756706422,
+      "learning_rate": 7.65381842363164e-06,
+      "loss": 0.6512,
+      "step": 3253
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9939074711670517,
+      "learning_rate": 7.652373893804952e-06,
+      "loss": 0.6817,
+      "step": 3254
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.5665404033863157,
+      "learning_rate": 7.650929055833135e-06,
+      "loss": 0.6852,
+      "step": 3255
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.7571299007941965,
+      "learning_rate": 7.64948390988404e-06,
+      "loss": 0.6688,
+      "step": 3256
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.93253267201708,
+      "learning_rate": 7.648038456125566e-06,
+      "loss": 0.6498,
+      "step": 3257
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2564745859826627,
+      "learning_rate": 7.646592694725638e-06,
+      "loss": 0.6059,
+      "step": 3258
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.460398219897918,
+      "learning_rate": 7.64514662585222e-06,
+      "loss": 0.7199,
+      "step": 3259
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.095756389781973,
+      "learning_rate": 7.643700249673315e-06,
+      "loss": 0.6769,
+      "step": 3260
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6122927168443058,
+      "learning_rate": 7.642253566356957e-06,
+      "loss": 0.6627,
+      "step": 3261
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.825071935901371,
+      "learning_rate": 7.640806576071215e-06,
+      "loss": 0.6343,
+      "step": 3262
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.968048064848337,
+      "learning_rate": 7.639359278984202e-06,
+      "loss": 0.7232,
+      "step": 3263
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.636260757493031,
+      "learning_rate": 7.637911675264056e-06,
+      "loss": 0.6242,
+      "step": 3264
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.452731556132531,
+      "learning_rate": 7.636463765078958e-06,
+      "loss": 0.6242,
+      "step": 3265
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0833107459100242,
+      "learning_rate": 7.63501554859712e-06,
+      "loss": 0.5713,
+      "step": 3266
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2859379838193097,
+      "learning_rate": 7.633567025986795e-06,
+      "loss": 0.7321,
+      "step": 3267
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9027700596440162,
+      "learning_rate": 7.632118197416263e-06,
+      "loss": 0.574,
+      "step": 3268
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.7949610995170224,
+      "learning_rate": 7.630669063053849e-06,
+      "loss": 0.6283,
+      "step": 3269
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.724700144475404,
+      "learning_rate": 7.629219623067907e-06,
+      "loss": 0.6921,
+      "step": 3270
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.384530330225657,
+      "learning_rate": 7.62776987762683e-06,
+      "loss": 0.6618,
+      "step": 3271
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.591079385525758,
+      "learning_rate": 7.626319826899045e-06,
+      "loss": 0.6482,
+      "step": 3272
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.314517492095959,
+      "learning_rate": 7.624869471053014e-06,
+      "loss": 0.6968,
+      "step": 3273
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.71453402590448,
+      "learning_rate": 7.623418810257234e-06,
+      "loss": 0.6831,
+      "step": 3274
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.710436771293437,
+      "learning_rate": 7.621967844680241e-06,
+      "loss": 0.6333,
+      "step": 3275
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.8259084062940265,
+      "learning_rate": 7.620516574490604e-06,
+      "loss": 0.6768,
+      "step": 3276
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.291949490786709,
+      "learning_rate": 7.6190649998569265e-06,
+      "loss": 0.6547,
+      "step": 3277
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.7125389974746223,
+      "learning_rate": 7.617613120947848e-06,
+      "loss": 0.5768,
+      "step": 3278
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.167970886019283,
+      "learning_rate": 7.616160937932045e-06,
+      "loss": 0.7309,
+      "step": 3279
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.549733398070985,
+      "learning_rate": 7.614708450978226e-06,
+      "loss": 0.7135,
+      "step": 3280
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9374280731445555,
+      "learning_rate": 7.613255660255137e-06,
+      "loss": 0.7211,
+      "step": 3281
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.870278169530705,
+      "learning_rate": 7.611802565931559e-06,
+      "loss": 0.6897,
+      "step": 3282
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2543736203150375,
+      "learning_rate": 7.610349168176309e-06,
+      "loss": 0.6159,
+      "step": 3283
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.738519475571267,
+      "learning_rate": 7.608895467158241e-06,
+      "loss": 0.7084,
+      "step": 3284
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.479720977262225,
+      "learning_rate": 7.607441463046236e-06,
+      "loss": 0.7224,
+      "step": 3285
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.4942767891910593,
+      "learning_rate": 7.60598715600922e-06,
+      "loss": 0.6773,
+      "step": 3286
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.350906281284041,
+      "learning_rate": 7.60453254621615e-06,
+      "loss": 0.6071,
+      "step": 3287
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.8683834683641827,
+      "learning_rate": 7.603077633836018e-06,
+      "loss": 0.6792,
+      "step": 3288
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7330042305762703,
+      "learning_rate": 7.601622419037851e-06,
+      "loss": 0.5959,
+      "step": 3289
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.4753065005990313,
+      "learning_rate": 7.600166901990711e-06,
+      "loss": 0.6422,
+      "step": 3290
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6481317364356585,
+      "learning_rate": 7.5987110828636966e-06,
+      "loss": 0.7076,
+      "step": 3291
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.978965610289502,
+      "learning_rate": 7.5972549618259415e-06,
+      "loss": 0.6281,
+      "step": 3292
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.1836343422551043,
+      "learning_rate": 7.595798539046612e-06,
+      "loss": 0.6259,
+      "step": 3293
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6636282179606776,
+      "learning_rate": 7.594341814694914e-06,
+      "loss": 0.5633,
+      "step": 3294
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6492054525812354,
+      "learning_rate": 7.592884788940082e-06,
+      "loss": 0.7039,
+      "step": 3295
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6199101782391185,
+      "learning_rate": 7.59142746195139e-06,
+      "loss": 0.6398,
+      "step": 3296
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4865546985556777,
+      "learning_rate": 7.5899698338981475e-06,
+      "loss": 0.625,
+      "step": 3297
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3823056427702305,
+      "learning_rate": 7.588511904949696e-06,
+      "loss": 0.7082,
+      "step": 3298
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.653522038656345,
+      "learning_rate": 7.587053675275413e-06,
+      "loss": 0.7159,
+      "step": 3299
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.007493250962484,
+      "learning_rate": 7.585595145044714e-06,
+      "loss": 0.6847,
+      "step": 3300
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.9762142271332412,
+      "learning_rate": 7.5841363144270445e-06,
+      "loss": 0.6299,
+      "step": 3301
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.081453490480924,
+      "learning_rate": 7.582677183591889e-06,
+      "loss": 0.6112,
+      "step": 3302
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0011529464421827,
+      "learning_rate": 7.581217752708763e-06,
+      "loss": 0.6528,
+      "step": 3303
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.8502588551289163,
+      "learning_rate": 7.579758021947221e-06,
+      "loss": 0.7399,
+      "step": 3304
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.6979556515844454,
+      "learning_rate": 7.578297991476848e-06,
+      "loss": 0.7052,
+      "step": 3305
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.409793154362905,
+      "learning_rate": 7.576837661467269e-06,
+      "loss": 0.5654,
+      "step": 3306
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0297576136517264,
+      "learning_rate": 7.575377032088138e-06,
+      "loss": 0.5874,
+      "step": 3307
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.335566510834448,
+      "learning_rate": 7.573916103509149e-06,
+      "loss": 0.6567,
+      "step": 3308
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8306846989581675,
+      "learning_rate": 7.572454875900026e-06,
+      "loss": 0.6771,
+      "step": 3309
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.914166159192387,
+      "learning_rate": 7.570993349430533e-06,
+      "loss": 0.6979,
+      "step": 3310
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.172792917556958,
+      "learning_rate": 7.569531524270465e-06,
+      "loss": 0.6625,
+      "step": 3311
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.322440067377819,
+      "learning_rate": 7.568069400589651e-06,
+      "loss": 0.6143,
+      "step": 3312
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7366214062810466,
+      "learning_rate": 7.566606978557959e-06,
+      "loss": 0.6719,
+      "step": 3313
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.6661060876658205,
+      "learning_rate": 7.565144258345287e-06,
+      "loss": 0.7597,
+      "step": 3314
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.841666028729839,
+      "learning_rate": 7.563681240121569e-06,
+      "loss": 0.6452,
+      "step": 3315
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9148308706374637,
+      "learning_rate": 7.562217924056777e-06,
+      "loss": 0.7037,
+      "step": 3316
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.493732792998696,
+      "learning_rate": 7.560754310320912e-06,
+      "loss": 0.665,
+      "step": 3317
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3318417413774633,
+      "learning_rate": 7.559290399084016e-06,
+      "loss": 0.6403,
+      "step": 3318
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.400731362221395,
+      "learning_rate": 7.5578261905161575e-06,
+      "loss": 0.6265,
+      "step": 3319
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.040790993560575,
+      "learning_rate": 7.556361684787446e-06,
+      "loss": 0.6665,
+      "step": 3320
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2499211076771033,
+      "learning_rate": 7.554896882068025e-06,
+      "loss": 0.594,
+      "step": 3321
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.357387837819163,
+      "learning_rate": 7.5534317825280664e-06,
+      "loss": 0.6814,
+      "step": 3322
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.712351178001371,
+      "learning_rate": 7.551966386337788e-06,
+      "loss": 0.7282,
+      "step": 3323
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4161878209245478,
+      "learning_rate": 7.5505006936674304e-06,
+      "loss": 0.6134,
+      "step": 3324
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.640259293020909,
+      "learning_rate": 7.5490347046872755e-06,
+      "loss": 0.7243,
+      "step": 3325
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.240206873144137,
+      "learning_rate": 7.547568419567637e-06,
+      "loss": 0.6741,
+      "step": 3326
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.394746942635679,
+      "learning_rate": 7.546101838478864e-06,
+      "loss": 0.6578,
+      "step": 3327
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.144423846324425,
+      "learning_rate": 7.54463496159134e-06,
+      "loss": 0.6511,
+      "step": 3328
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6453748407725106,
+      "learning_rate": 7.543167789075481e-06,
+      "loss": 0.6275,
+      "step": 3329
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3080573199067187,
+      "learning_rate": 7.54170032110174e-06,
+      "loss": 0.6676,
+      "step": 3330
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.714660296548131,
+      "learning_rate": 7.540232557840604e-06,
+      "loss": 0.611,
+      "step": 3331
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.047279583079742,
+      "learning_rate": 7.53876449946259e-06,
+      "loss": 0.6867,
+      "step": 3332
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.8795995067294924,
+      "learning_rate": 7.537296146138255e-06,
+      "loss": 0.5692,
+      "step": 3333
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8734422567926003,
+      "learning_rate": 7.535827498038192e-06,
+      "loss": 0.7021,
+      "step": 3334
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.854617594622321,
+      "learning_rate": 7.534358555333018e-06,
+      "loss": 0.6198,
+      "step": 3335
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.60018430557542,
+      "learning_rate": 7.532889318193393e-06,
+      "loss": 0.6089,
+      "step": 3336
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.2343822540898515,
+      "learning_rate": 7.531419786790011e-06,
+      "loss": 0.7023,
+      "step": 3337
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.095566888586303,
+      "learning_rate": 7.5299499612935934e-06,
+      "loss": 0.6577,
+      "step": 3338
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.136010955006197,
+      "learning_rate": 7.528479841874904e-06,
+      "loss": 0.6715,
+      "step": 3339
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.077841229423998,
+      "learning_rate": 7.527009428704735e-06,
+      "loss": 0.683,
+      "step": 3340
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.840010075056747,
+      "learning_rate": 7.525538721953915e-06,
+      "loss": 0.7238,
+      "step": 3341
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1576704441730468,
+      "learning_rate": 7.524067721793309e-06,
+      "loss": 0.6221,
+      "step": 3342
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6260852715338414,
+      "learning_rate": 7.522596428393809e-06,
+      "loss": 0.6732,
+      "step": 3343
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.35119357031428,
+      "learning_rate": 7.521124841926348e-06,
+      "loss": 0.6466,
+      "step": 3344
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.410888681828178,
+      "learning_rate": 7.519652962561894e-06,
+      "loss": 0.7642,
+      "step": 3345
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.7276855469070074,
+      "learning_rate": 7.5181807904714385e-06,
+      "loss": 0.6249,
+      "step": 3346
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.415586524281074,
+      "learning_rate": 7.516708325826021e-06,
+      "loss": 0.5827,
+      "step": 3347
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8263139482195503,
+      "learning_rate": 7.515235568796704e-06,
+      "loss": 0.662,
+      "step": 3348
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.664326847472684,
+      "learning_rate": 7.513762519554588e-06,
+      "loss": 0.5015,
+      "step": 3349
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.618966692292694,
+      "learning_rate": 7.51228917827081e-06,
+      "loss": 0.6738,
+      "step": 3350
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.633158245759139,
+      "learning_rate": 7.510815545116539e-06,
+      "loss": 0.6204,
+      "step": 3351
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4185537158238395,
+      "learning_rate": 7.509341620262976e-06,
+      "loss": 0.5918,
+      "step": 3352
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4156711101816497,
+      "learning_rate": 7.507867403881356e-06,
+      "loss": 0.7105,
+      "step": 3353
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.148153998559598,
+      "learning_rate": 7.506392896142951e-06,
+      "loss": 0.6219,
+      "step": 3354
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.6450108914306396,
+      "learning_rate": 7.5049180972190646e-06,
+      "loss": 0.6179,
+      "step": 3355
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.794686070537919,
+      "learning_rate": 7.503443007281035e-06,
+      "loss": 0.662,
+      "step": 3356
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.1106881667918755,
+      "learning_rate": 7.501967626500231e-06,
+      "loss": 0.695,
+      "step": 3357
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.700458795600681,
+      "learning_rate": 7.500491955048063e-06,
+      "loss": 0.6794,
+      "step": 3358
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.419751161900197,
+      "learning_rate": 7.499015993095968e-06,
+      "loss": 0.6671,
+      "step": 3359
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8449869838184343,
+      "learning_rate": 7.497539740815419e-06,
+      "loss": 0.7413,
+      "step": 3360
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.4184394602863226,
+      "learning_rate": 7.4960631983779205e-06,
+      "loss": 0.6781,
+      "step": 3361
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.796963292810003,
+      "learning_rate": 7.494586365955017e-06,
+      "loss": 0.667,
+      "step": 3362
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0483662745104665,
+      "learning_rate": 7.49310924371828e-06,
+      "loss": 0.715,
+      "step": 3363
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.5907131005662767,
+      "learning_rate": 7.491631831839318e-06,
+      "loss": 0.7097,
+      "step": 3364
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.175466104453125,
+      "learning_rate": 7.490154130489773e-06,
+      "loss": 0.6217,
+      "step": 3365
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.423899220213644,
+      "learning_rate": 7.488676139841318e-06,
+      "loss": 0.6247,
+      "step": 3366
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8572229700965646,
+      "learning_rate": 7.487197860065664e-06,
+      "loss": 0.613,
+      "step": 3367
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8587163319916042,
+      "learning_rate": 7.485719291334551e-06,
+      "loss": 0.6576,
+      "step": 3368
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3991473483168626,
+      "learning_rate": 7.484240433819758e-06,
+      "loss": 0.66,
+      "step": 3369
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.965531907514714,
+      "learning_rate": 7.482761287693092e-06,
+      "loss": 0.6466,
+      "step": 3370
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.048642664328739,
+      "learning_rate": 7.481281853126397e-06,
+      "loss": 0.715,
+      "step": 3371
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6309407059230296,
+      "learning_rate": 7.479802130291548e-06,
+      "loss": 0.6181,
+      "step": 3372
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.222323555003956,
+      "learning_rate": 7.478322119360457e-06,
+      "loss": 0.7331,
+      "step": 3373
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2573259131196384,
+      "learning_rate": 7.476841820505065e-06,
+      "loss": 0.6684,
+      "step": 3374
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.072317633131639,
+      "learning_rate": 7.475361233897352e-06,
+      "loss": 0.6969,
+      "step": 3375
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.046601351124791,
+      "learning_rate": 7.473880359709324e-06,
+      "loss": 0.6585,
+      "step": 3376
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4729492389743917,
+      "learning_rate": 7.472399198113029e-06,
+      "loss": 0.677,
+      "step": 3377
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.2903365979880492,
+      "learning_rate": 7.4709177492805405e-06,
+      "loss": 0.6252,
+      "step": 3378
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9983558599680125,
+      "learning_rate": 7.46943601338397e-06,
+      "loss": 0.7068,
+      "step": 3379
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0004226327693235,
+      "learning_rate": 7.4679539905954655e-06,
+      "loss": 0.6208,
+      "step": 3380
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.898934649269225,
+      "learning_rate": 7.4664716810871975e-06,
+      "loss": 0.6644,
+      "step": 3381
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.194372324636005,
+      "learning_rate": 7.464989085031381e-06,
+      "loss": 0.6673,
+      "step": 3382
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1182459004852183,
+      "learning_rate": 7.463506202600257e-06,
+      "loss": 0.5741,
+      "step": 3383
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.36761957421957,
+      "learning_rate": 7.462023033966104e-06,
+      "loss": 0.6101,
+      "step": 3384
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2861432544796156,
+      "learning_rate": 7.4605395793012325e-06,
+      "loss": 0.6921,
+      "step": 3385
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7603235128046846,
+      "learning_rate": 7.459055838777984e-06,
+      "loss": 0.6734,
+      "step": 3386
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.9793695446983532,
+      "learning_rate": 7.457571812568738e-06,
+      "loss": 0.6595,
+      "step": 3387
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.311706185746525,
+      "learning_rate": 7.4560875008459035e-06,
+      "loss": 0.7186,
+      "step": 3388
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.597103883565191,
+      "learning_rate": 7.454602903781921e-06,
+      "loss": 0.7305,
+      "step": 3389
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.501333529220921,
+      "learning_rate": 7.45311802154927e-06,
+      "loss": 0.6342,
+      "step": 3390
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5800190663717655,
+      "learning_rate": 7.451632854320459e-06,
+      "loss": 0.568,
+      "step": 3391
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3701799197834,
+      "learning_rate": 7.4501474022680265e-06,
+      "loss": 0.6836,
+      "step": 3392
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3772231724274557,
+      "learning_rate": 7.4486616655645565e-06,
+      "loss": 0.619,
+      "step": 3393
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8160484269096595,
+      "learning_rate": 7.447175644382648e-06,
+      "loss": 0.7025,
+      "step": 3394
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.7832742183500967,
+      "learning_rate": 7.445689338894949e-06,
+      "loss": 0.7469,
+      "step": 3395
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.0059621903562,
+      "learning_rate": 7.444202749274133e-06,
+      "loss": 0.7397,
+      "step": 3396
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.892558568857166,
+      "learning_rate": 7.442715875692908e-06,
+      "loss": 0.6436,
+      "step": 3397
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8280194358791824,
+      "learning_rate": 7.4412287183240115e-06,
+      "loss": 0.6798,
+      "step": 3398
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.837303017856792,
+      "learning_rate": 7.43974127734022e-06,
+      "loss": 0.6466,
+      "step": 3399
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.734694065110121,
+      "learning_rate": 7.4382535529143395e-06,
+      "loss": 0.6031,
+      "step": 3400
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.959288671850776,
+      "learning_rate": 7.43676554521921e-06,
+      "loss": 0.6906,
+      "step": 3401
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.198201079133412,
+      "learning_rate": 7.435277254427704e-06,
+      "loss": 0.6888,
+      "step": 3402
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.386419840965464,
+      "learning_rate": 7.4337886807127235e-06,
+      "loss": 0.6252,
+      "step": 3403
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.1534089316760054,
+      "learning_rate": 7.4322998242472135e-06,
+      "loss": 0.6707,
+      "step": 3404
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7988368901968124,
+      "learning_rate": 7.430810685204137e-06,
+      "loss": 0.6584,
+      "step": 3405
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.919583307019518,
+      "learning_rate": 7.4293212637565045e-06,
+      "loss": 0.6966,
+      "step": 3406
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.508307648411482,
+      "learning_rate": 7.427831560077349e-06,
+      "loss": 0.6444,
+      "step": 3407
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3042696713402557,
+      "learning_rate": 7.426341574339741e-06,
+      "loss": 0.5891,
+      "step": 3408
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.582107451003599,
+      "learning_rate": 7.424851306716783e-06,
+      "loss": 0.6359,
+      "step": 3409
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.904925684194061,
+      "learning_rate": 7.42336075738161e-06,
+      "loss": 0.6671,
+      "step": 3410
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.5167258077031267,
+      "learning_rate": 7.421869926507389e-06,
+      "loss": 0.7021,
+      "step": 3411
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.358687817951523,
+      "learning_rate": 7.420378814267322e-06,
+      "loss": 0.7419,
+      "step": 3412
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.152148471276924,
+      "learning_rate": 7.41888742083464e-06,
+      "loss": 0.6744,
+      "step": 3413
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3915603282771976,
+      "learning_rate": 7.417395746382608e-06,
+      "loss": 0.6771,
+      "step": 3414
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.7148070409783536,
+      "learning_rate": 7.415903791084529e-06,
+      "loss": 0.7244,
+      "step": 3415
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.752027193158189,
+      "learning_rate": 7.41441155511373e-06,
+      "loss": 0.6177,
+      "step": 3416
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.62863853424837,
+      "learning_rate": 7.412919038643577e-06,
+      "loss": 0.6457,
+      "step": 3417
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5794611356979993,
+      "learning_rate": 7.411426241847463e-06,
+      "loss": 0.6303,
+      "step": 3418
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.9676318300037527,
+      "learning_rate": 7.409933164898819e-06,
+      "loss": 0.6817,
+      "step": 3419
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8550149054743,
+      "learning_rate": 7.408439807971108e-06,
+      "loss": 0.654,
+      "step": 3420
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.195120384669288,
+      "learning_rate": 7.406946171237822e-06,
+      "loss": 0.579,
+      "step": 3421
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0078016488240107,
+      "learning_rate": 7.4054522548724874e-06,
+      "loss": 0.6311,
+      "step": 3422
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4860091683873784,
+      "learning_rate": 7.403958059048662e-06,
+      "loss": 0.6168,
+      "step": 3423
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.570809434838443,
+      "learning_rate": 7.40246358393994e-06,
+      "loss": 0.591,
+      "step": 3424
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.7959371110166256,
+      "learning_rate": 7.4009688297199436e-06,
+      "loss": 0.7263,
+      "step": 3425
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.774483241184963,
+      "learning_rate": 7.3994737965623285e-06,
+      "loss": 0.6566,
+      "step": 3426
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.455385395465524,
+      "learning_rate": 7.397978484640783e-06,
+      "loss": 0.5688,
+      "step": 3427
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.422389158395626,
+      "learning_rate": 7.396482894129031e-06,
+      "loss": 0.6501,
+      "step": 3428
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9579176252075983,
+      "learning_rate": 7.3949870252008215e-06,
+      "loss": 0.5639,
+      "step": 3429
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.2151993534590684,
+      "learning_rate": 7.393490878029945e-06,
+      "loss": 0.6368,
+      "step": 3430
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4041060387118276,
+      "learning_rate": 7.391994452790217e-06,
+      "loss": 0.7303,
+      "step": 3431
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7465187882412563,
+      "learning_rate": 7.390497749655487e-06,
+      "loss": 0.6067,
+      "step": 3432
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.714071334159705,
+      "learning_rate": 7.389000768799638e-06,
+      "loss": 0.6604,
+      "step": 3433
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9662799463771505,
+      "learning_rate": 7.387503510396586e-06,
+      "loss": 0.6275,
+      "step": 3434
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3746212916638507,
+      "learning_rate": 7.386005974620278e-06,
+      "loss": 0.7533,
+      "step": 3435
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.6759513228513154,
+      "learning_rate": 7.384508161644694e-06,
+      "loss": 0.6726,
+      "step": 3436
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.2759734265835565,
+      "learning_rate": 7.383010071643844e-06,
+      "loss": 0.6219,
+      "step": 3437
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0676737541014298,
+      "learning_rate": 7.381511704791771e-06,
+      "loss": 0.5884,
+      "step": 3438
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.292189795768927,
+      "learning_rate": 7.380013061262557e-06,
+      "loss": 0.5909,
+      "step": 3439
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.097298037082653,
+      "learning_rate": 7.3785141412303e-06,
+      "loss": 0.6163,
+      "step": 3440
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.816854958679348,
+      "learning_rate": 7.37701494486915e-06,
+      "loss": 0.6845,
+      "step": 3441
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.788505661813447,
+      "learning_rate": 7.375515472353272e-06,
+      "loss": 0.6201,
+      "step": 3442
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.129127828825236,
+      "learning_rate": 7.374015723856873e-06,
+      "loss": 0.6793,
+      "step": 3443
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3574849469565544,
+      "learning_rate": 7.372515699554191e-06,
+      "loss": 0.6221,
+      "step": 3444
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.3812374504332805,
+      "learning_rate": 7.371015399619494e-06,
+      "loss": 0.6603,
+      "step": 3445
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0367925120459303,
+      "learning_rate": 7.369514824227082e-06,
+      "loss": 0.6616,
+      "step": 3446
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6543688097654665,
+      "learning_rate": 7.368013973551286e-06,
+      "loss": 0.7013,
+      "step": 3447
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1635818274199434,
+      "learning_rate": 7.366512847766472e-06,
+      "loss": 0.6885,
+      "step": 3448
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.085467390645828,
+      "learning_rate": 7.365011447047036e-06,
+      "loss": 0.7165,
+      "step": 3449
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.283937568641037,
+      "learning_rate": 7.363509771567408e-06,
+      "loss": 0.6338,
+      "step": 3450
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.1768009624928166,
+      "learning_rate": 7.362007821502045e-06,
+      "loss": 0.747,
+      "step": 3451
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0324905424749597,
+      "learning_rate": 7.360505597025442e-06,
+      "loss": 0.685,
+      "step": 3452
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.983148422243415,
+      "learning_rate": 7.359003098312123e-06,
+      "loss": 0.6152,
+      "step": 3453
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9793497750002658,
+      "learning_rate": 7.357500325536644e-06,
+      "loss": 0.7456,
+      "step": 3454
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.941874609510829,
+      "learning_rate": 7.355997278873589e-06,
+      "loss": 0.6748,
+      "step": 3455
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1486776506884153,
+      "learning_rate": 7.354493958497583e-06,
+      "loss": 0.661,
+      "step": 3456
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0170095119117828,
+      "learning_rate": 7.3529903645832744e-06,
+      "loss": 0.5884,
+      "step": 3457
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4036589735821483,
+      "learning_rate": 7.351486497305347e-06,
+      "loss": 0.5853,
+      "step": 3458
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.310188596068514,
+      "learning_rate": 7.349982356838515e-06,
+      "loss": 0.7171,
+      "step": 3459
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.942433337193868,
+      "learning_rate": 7.348477943357527e-06,
+      "loss": 0.6472,
+      "step": 3460
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.969354980228407,
+      "learning_rate": 7.34697325703716e-06,
+      "loss": 0.6456,
+      "step": 3461
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.8519505648735737,
+      "learning_rate": 7.345468298052224e-06,
+      "loss": 0.6262,
+      "step": 3462
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.685187241926782,
+      "learning_rate": 7.343963066577563e-06,
+      "loss": 0.6571,
+      "step": 3463
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.437533087132148,
+      "learning_rate": 7.342457562788046e-06,
+      "loss": 0.6799,
+      "step": 3464
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.281370556976629,
+      "learning_rate": 7.340951786858583e-06,
+      "loss": 0.688,
+      "step": 3465
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.1063629908728014,
+      "learning_rate": 7.339445738964106e-06,
+      "loss": 0.7737,
+      "step": 3466
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0889025130238092,
+      "learning_rate": 7.337939419279588e-06,
+      "loss": 0.6248,
+      "step": 3467
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.815104069216375,
+      "learning_rate": 7.336432827980026e-06,
+      "loss": 0.6681,
+      "step": 3468
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.1118170144076265,
+      "learning_rate": 7.334925965240451e-06,
+      "loss": 0.6273,
+      "step": 3469
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.2170988071407147,
+      "learning_rate": 7.333418831235928e-06,
+      "loss": 0.6412,
+      "step": 3470
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.4121748554575984,
+      "learning_rate": 7.33191142614155e-06,
+      "loss": 0.698,
+      "step": 3471
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0702451497078247,
+      "learning_rate": 7.330403750132443e-06,
+      "loss": 0.5974,
+      "step": 3472
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.896556923363,
+      "learning_rate": 7.328895803383764e-06,
+      "loss": 0.6431,
+      "step": 3473
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.849283645322886,
+      "learning_rate": 7.327387586070705e-06,
+      "loss": 0.6288,
+      "step": 3474
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.6979247535807063,
+      "learning_rate": 7.325879098368483e-06,
+      "loss": 0.6332,
+      "step": 3475
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7475025526366377,
+      "learning_rate": 7.324370340452351e-06,
+      "loss": 0.6657,
+      "step": 3476
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9659631728372777,
+      "learning_rate": 7.322861312497591e-06,
+      "loss": 0.6951,
+      "step": 3477
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1844754621892837,
+      "learning_rate": 7.321352014679522e-06,
+      "loss": 0.5929,
+      "step": 3478
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.974837282969043,
+      "learning_rate": 7.319842447173482e-06,
+      "loss": 0.5501,
+      "step": 3479
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.892368554345106,
+      "learning_rate": 7.318332610154854e-06,
+      "loss": 0.7127,
+      "step": 3480
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.715197258419502,
+      "learning_rate": 7.3168225037990434e-06,
+      "loss": 0.6408,
+      "step": 3481
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5710173525042554,
+      "learning_rate": 7.315312128281493e-06,
+      "loss": 0.6488,
+      "step": 3482
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0270385193056204,
+      "learning_rate": 7.313801483777674e-06,
+      "loss": 0.6552,
+      "step": 3483
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.205749249065878,
+      "learning_rate": 7.3122905704630845e-06,
+      "loss": 0.6876,
+      "step": 3484
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.688817976879616,
+      "learning_rate": 7.310779388513263e-06,
+      "loss": 0.6637,
+      "step": 3485
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.401643485510619,
+      "learning_rate": 7.309267938103769e-06,
+      "loss": 0.6614,
+      "step": 3486
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4560485153621237,
+      "learning_rate": 7.307756219410205e-06,
+      "loss": 0.6329,
+      "step": 3487
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5820481308855046,
+      "learning_rate": 7.306244232608191e-06,
+      "loss": 0.7006,
+      "step": 3488
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1020552524421927,
+      "learning_rate": 7.304731977873392e-06,
+      "loss": 0.5822,
+      "step": 3489
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.584430521631613,
+      "learning_rate": 7.303219455381491e-06,
+      "loss": 0.6619,
+      "step": 3490
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.298444021464788,
+      "learning_rate": 7.301706665308212e-06,
+      "loss": 0.7223,
+      "step": 3491
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.985437156669851,
+      "learning_rate": 7.300193607829308e-06,
+      "loss": 0.6624,
+      "step": 3492
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.590700474555624,
+      "learning_rate": 7.298680283120558e-06,
+      "loss": 0.6532,
+      "step": 3493
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3977866755703614,
+      "learning_rate": 7.29716669135778e-06,
+      "loss": 0.7138,
+      "step": 3494
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7237583049566525,
+      "learning_rate": 7.295652832716814e-06,
+      "loss": 0.6091,
+      "step": 3495
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.789323251837038,
+      "learning_rate": 7.294138707373539e-06,
+      "loss": 0.6027,
+      "step": 3496
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1445610603977494,
+      "learning_rate": 7.29262431550386e-06,
+      "loss": 0.6358,
+      "step": 3497
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7828156452167434,
+      "learning_rate": 7.2911096572837155e-06,
+      "loss": 0.7049,
+      "step": 3498
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.062889194963662,
+      "learning_rate": 7.289594732889073e-06,
+      "loss": 0.6906,
+      "step": 3499
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.187017863959395,
+      "learning_rate": 7.288079542495936e-06,
+      "loss": 0.6046,
+      "step": 3500
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.80365868383857,
+      "learning_rate": 7.286564086280329e-06,
+      "loss": 0.6339,
+      "step": 3501
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.978700832429271,
+      "learning_rate": 7.285048364418319e-06,
+      "loss": 0.6488,
+      "step": 3502
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9173805232519245,
+      "learning_rate": 7.283532377085992e-06,
+      "loss": 0.6956,
+      "step": 3503
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.020587219160722,
+      "learning_rate": 7.282016124459477e-06,
+      "loss": 0.6508,
+      "step": 3504
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3101783919909056,
+      "learning_rate": 7.280499606714923e-06,
+      "loss": 0.6795,
+      "step": 3505
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.9706549307214556,
+      "learning_rate": 7.27898282402852e-06,
+      "loss": 0.7841,
+      "step": 3506
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.2123361694164423,
+      "learning_rate": 7.277465776576478e-06,
+      "loss": 0.6485,
+      "step": 3507
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.833849508292802,
+      "learning_rate": 7.275948464535045e-06,
+      "loss": 0.6073,
+      "step": 3508
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.190248056297931,
+      "learning_rate": 7.274430888080502e-06,
+      "loss": 0.6136,
+      "step": 3509
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.801011293418992,
+      "learning_rate": 7.27291304738915e-06,
+      "loss": 0.6847,
+      "step": 3510
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.725920278243489,
+      "learning_rate": 7.271394942637332e-06,
+      "loss": 0.6099,
+      "step": 3511
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3603811567927466,
+      "learning_rate": 7.269876574001414e-06,
+      "loss": 0.6546,
+      "step": 3512
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.775937881236346,
+      "learning_rate": 7.2683579416578e-06,
+      "loss": 0.6915,
+      "step": 3513
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.542726679655711,
+      "learning_rate": 7.266839045782914e-06,
+      "loss": 0.6684,
+      "step": 3514
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7964307299094453,
+      "learning_rate": 7.265319886553223e-06,
+      "loss": 0.594,
+      "step": 3515
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.37740356376484,
+      "learning_rate": 7.263800464145214e-06,
+      "loss": 0.7419,
+      "step": 3516
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.485584641652323,
+      "learning_rate": 7.262280778735412e-06,
+      "loss": 0.6703,
+      "step": 3517
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.4839449059905907,
+      "learning_rate": 7.26076083050037e-06,
+      "loss": 0.675,
+      "step": 3518
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.12300936003989,
+      "learning_rate": 7.259240619616668e-06,
+      "loss": 0.6224,
+      "step": 3519
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6719730804814863,
+      "learning_rate": 7.257720146260923e-06,
+      "loss": 0.659,
+      "step": 3520
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.73292327896022,
+      "learning_rate": 7.256199410609776e-06,
+      "loss": 0.6476,
+      "step": 3521
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.636014633612924,
+      "learning_rate": 7.254678412839905e-06,
+      "loss": 0.6879,
+      "step": 3522
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5306894259549075,
+      "learning_rate": 7.253157153128012e-06,
+      "loss": 0.6785,
+      "step": 3523
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.976569034282962,
+      "learning_rate": 7.251635631650838e-06,
+      "loss": 0.7064,
+      "step": 3524
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.45250759237423,
+      "learning_rate": 7.250113848585141e-06,
+      "loss": 0.6499,
+      "step": 3525
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.060166639945123,
+      "learning_rate": 7.248591804107724e-06,
+      "loss": 0.6757,
+      "step": 3526
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.716487897796609,
+      "learning_rate": 7.247069498395409e-06,
+      "loss": 0.6512,
+      "step": 3527
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.4915603318297346,
+      "learning_rate": 7.245546931625057e-06,
+      "loss": 0.6518,
+      "step": 3528
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.657417413138783,
+      "learning_rate": 7.244024103973553e-06,
+      "loss": 0.6662,
+      "step": 3529
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.6985543615408805,
+      "learning_rate": 7.242501015617815e-06,
+      "loss": 0.5643,
+      "step": 3530
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7512005261618566,
+      "learning_rate": 7.240977666734793e-06,
+      "loss": 0.6004,
+      "step": 3531
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3062901989306015,
+      "learning_rate": 7.239454057501462e-06,
+      "loss": 0.6528,
+      "step": 3532
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1690879827237803,
+      "learning_rate": 7.237930188094834e-06,
+      "loss": 0.6433,
+      "step": 3533
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.189006985940273,
+      "learning_rate": 7.236406058691944e-06,
+      "loss": 0.6005,
+      "step": 3534
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5436760311571067,
+      "learning_rate": 7.234881669469864e-06,
+      "loss": 0.6815,
+      "step": 3535
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5440441226370254,
+      "learning_rate": 7.233357020605692e-06,
+      "loss": 0.5973,
+      "step": 3536
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8717435554951423,
+      "learning_rate": 7.2318321122765575e-06,
+      "loss": 0.7263,
+      "step": 3537
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5676017996288674,
+      "learning_rate": 7.230306944659618e-06,
+      "loss": 0.6429,
+      "step": 3538
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.320166184610786,
+      "learning_rate": 7.2287815179320665e-06,
+      "loss": 0.7412,
+      "step": 3539
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.887680699145069,
+      "learning_rate": 7.227255832271122e-06,
+      "loss": 0.7,
+      "step": 3540
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2913081303644938,
+      "learning_rate": 7.225729887854032e-06,
+      "loss": 0.6098,
+      "step": 3541
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.5364276369948415,
+      "learning_rate": 7.224203684858078e-06,
+      "loss": 0.639,
+      "step": 3542
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9611456139801065,
+      "learning_rate": 7.222677223460567e-06,
+      "loss": 0.626,
+      "step": 3543
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.607908981424635,
+      "learning_rate": 7.221150503838844e-06,
+      "loss": 0.6371,
+      "step": 3544
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6691678732909114,
+      "learning_rate": 7.219623526170275e-06,
+      "loss": 0.6519,
+      "step": 3545
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.914049840880877,
+      "learning_rate": 7.218096290632263e-06,
+      "loss": 0.6737,
+      "step": 3546
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.551071513521762,
+      "learning_rate": 7.216568797402232e-06,
+      "loss": 0.616,
+      "step": 3547
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5813083708314255,
+      "learning_rate": 7.2150410466576495e-06,
+      "loss": 0.7268,
+      "step": 3548
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0795987810833583,
+      "learning_rate": 7.213513038575999e-06,
+      "loss": 0.6526,
+      "step": 3549
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2721784087350216,
+      "learning_rate": 7.211984773334803e-06,
+      "loss": 0.6302,
+      "step": 3550
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.928510250413681,
+      "learning_rate": 7.210456251111611e-06,
+      "loss": 0.6411,
+      "step": 3551
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.445810874079125,
+      "learning_rate": 7.208927472084e-06,
+      "loss": 0.6141,
+      "step": 3552
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3202477177734657,
+      "learning_rate": 7.207398436429581e-06,
+      "loss": 0.6578,
+      "step": 3553
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.305688026429538,
+      "learning_rate": 7.205869144325992e-06,
+      "loss": 0.6092,
+      "step": 3554
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.346775916923627,
+      "learning_rate": 7.204339595950904e-06,
+      "loss": 0.5894,
+      "step": 3555
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.8477133741441674,
+      "learning_rate": 7.202809791482013e-06,
+      "loss": 0.6427,
+      "step": 3556
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.235974425228947,
+      "learning_rate": 7.201279731097048e-06,
+      "loss": 0.6674,
+      "step": 3557
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2493046168393445,
+      "learning_rate": 7.199749414973767e-06,
+      "loss": 0.6528,
+      "step": 3558
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.107170893614456,
+      "learning_rate": 7.1982188432899595e-06,
+      "loss": 0.5877,
+      "step": 3559
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.3599103332329485,
+      "learning_rate": 7.196688016223439e-06,
+      "loss": 0.7369,
+      "step": 3560
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4363610701892657,
+      "learning_rate": 7.195156933952055e-06,
+      "loss": 0.6762,
+      "step": 3561
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7535120357249556,
+      "learning_rate": 7.193625596653684e-06,
+      "loss": 0.6812,
+      "step": 3562
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.099885019771948,
+      "learning_rate": 7.1920940045062335e-06,
+      "loss": 0.6133,
+      "step": 3563
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.93938976849278,
+      "learning_rate": 7.1905621576876375e-06,
+      "loss": 0.6401,
+      "step": 3564
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4121806622384656,
+      "learning_rate": 7.189030056375862e-06,
+      "loss": 0.629,
+      "step": 3565
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.832137503000308,
+      "learning_rate": 7.187497700748903e-06,
+      "loss": 0.7839,
+      "step": 3566
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.892305433301476,
+      "learning_rate": 7.185965090984783e-06,
+      "loss": 0.6009,
+      "step": 3567
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.701267405207593,
+      "learning_rate": 7.184432227261561e-06,
+      "loss": 0.7288,
+      "step": 3568
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.065332041666608,
+      "learning_rate": 7.182899109757314e-06,
+      "loss": 0.6107,
+      "step": 3569
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3541441748774177,
+      "learning_rate": 7.181365738650161e-06,
+      "loss": 0.6027,
+      "step": 3570
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.271716533928777,
+      "learning_rate": 7.17983211411824e-06,
+      "loss": 0.6606,
+      "step": 3571
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.88759903004812,
+      "learning_rate": 7.178298236339727e-06,
+      "loss": 0.5967,
+      "step": 3572
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.01835950623007,
+      "learning_rate": 7.176764105492821e-06,
+      "loss": 0.6285,
+      "step": 3573
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7434215149338397,
+      "learning_rate": 7.175229721755753e-06,
+      "loss": 0.6626,
+      "step": 3574
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.346997092680918,
+      "learning_rate": 7.173695085306785e-06,
+      "loss": 0.7167,
+      "step": 3575
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.140529020266596,
+      "learning_rate": 7.172160196324205e-06,
+      "loss": 0.658,
+      "step": 3576
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.120165751814182,
+      "learning_rate": 7.1706250549863335e-06,
+      "loss": 0.69,
+      "step": 3577
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2247914008766796,
+      "learning_rate": 7.1690896614715155e-06,
+      "loss": 0.6288,
+      "step": 3578
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.605629864739022,
+      "learning_rate": 7.167554015958133e-06,
+      "loss": 0.6357,
+      "step": 3579
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3904906015835947,
+      "learning_rate": 7.166018118624588e-06,
+      "loss": 0.5556,
+      "step": 3580
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6192217487542586,
+      "learning_rate": 7.164481969649323e-06,
+      "loss": 0.6576,
+      "step": 3581
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.076216402779878,
+      "learning_rate": 7.162945569210796e-06,
+      "loss": 0.6749,
+      "step": 3582
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.889072644554949,
+      "learning_rate": 7.161408917487509e-06,
+      "loss": 0.6673,
+      "step": 3583
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.537856847059156,
+      "learning_rate": 7.159872014657978e-06,
+      "loss": 0.6752,
+      "step": 3584
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.221005642218842,
+      "learning_rate": 7.158334860900762e-06,
+      "loss": 0.5711,
+      "step": 3585
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4533316517178387,
+      "learning_rate": 7.156797456394441e-06,
+      "loss": 0.7351,
+      "step": 3586
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.886040834314495,
+      "learning_rate": 7.1552598013176264e-06,
+      "loss": 0.6767,
+      "step": 3587
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.874147523804201,
+      "learning_rate": 7.1537218958489575e-06,
+      "loss": 0.6642,
+      "step": 3588
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.722531139220571,
+      "learning_rate": 7.152183740167105e-06,
+      "loss": 0.6678,
+      "step": 3589
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.187200392218844,
+      "learning_rate": 7.150645334450767e-06,
+      "loss": 0.7528,
+      "step": 3590
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3536496010454147,
+      "learning_rate": 7.14910667887867e-06,
+      "loss": 0.6902,
+      "step": 3591
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0650748967886057,
+      "learning_rate": 7.147567773629573e-06,
+      "loss": 0.6505,
+      "step": 3592
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.86825048250956,
+      "learning_rate": 7.146028618882258e-06,
+      "loss": 0.7501,
+      "step": 3593
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5434125470032756,
+      "learning_rate": 7.1444892148155445e-06,
+      "loss": 0.7357,
+      "step": 3594
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4145813864328374,
+      "learning_rate": 7.14294956160827e-06,
+      "loss": 0.613,
+      "step": 3595
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.141187484264333,
+      "learning_rate": 7.141409659439313e-06,
+      "loss": 0.6496,
+      "step": 3596
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9374788957358136,
+      "learning_rate": 7.139869508487569e-06,
+      "loss": 0.6382,
+      "step": 3597
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4608281779178722,
+      "learning_rate": 7.138329108931974e-06,
+      "loss": 0.6155,
+      "step": 3598
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9810869246330727,
+      "learning_rate": 7.136788460951482e-06,
+      "loss": 0.6844,
+      "step": 3599
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.261119179477242,
+      "learning_rate": 7.135247564725085e-06,
+      "loss": 0.6355,
+      "step": 3600
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.655641062191612,
+      "learning_rate": 7.133706420431799e-06,
+      "loss": 0.578,
+      "step": 3601
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0452192913458778,
+      "learning_rate": 7.132165028250666e-06,
+      "loss": 0.6081,
+      "step": 3602
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5869892334643882,
+      "learning_rate": 7.130623388360767e-06,
+      "loss": 0.7272,
+      "step": 3603
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7640836198719785,
+      "learning_rate": 7.129081500941199e-06,
+      "loss": 0.6978,
+      "step": 3604
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.768985249973699,
+      "learning_rate": 7.127539366171099e-06,
+      "loss": 0.5961,
+      "step": 3605
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.9567725278063506,
+      "learning_rate": 7.125996984229623e-06,
+      "loss": 0.685,
+      "step": 3606
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.879031446261374,
+      "learning_rate": 7.124454355295966e-06,
+      "loss": 0.6856,
+      "step": 3607
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4194323937650752,
+      "learning_rate": 7.1229114795493405e-06,
+      "loss": 0.6894,
+      "step": 3608
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9796362738749906,
+      "learning_rate": 7.121368357168997e-06,
+      "loss": 0.5798,
+      "step": 3609
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.51120233981379,
+      "learning_rate": 7.11982498833421e-06,
+      "loss": 0.7307,
+      "step": 3610
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9959995491929295,
+      "learning_rate": 7.1182813732242835e-06,
+      "loss": 0.5706,
+      "step": 3611
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.693687958001663,
+      "learning_rate": 7.116737512018551e-06,
+      "loss": 0.598,
+      "step": 3612
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.298264703609383,
+      "learning_rate": 7.115193404896372e-06,
+      "loss": 0.6928,
+      "step": 3613
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.7510712526228716,
+      "learning_rate": 7.1136490520371394e-06,
+      "loss": 0.6345,
+      "step": 3614
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.545812637717511,
+      "learning_rate": 7.112104453620269e-06,
+      "loss": 0.7136,
+      "step": 3615
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4425094561499003,
+      "learning_rate": 7.11055960982521e-06,
+      "loss": 0.6611,
+      "step": 3616
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2838270941484837,
+      "learning_rate": 7.109014520831433e-06,
+      "loss": 0.6004,
+      "step": 3617
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.6464684533156175,
+      "learning_rate": 7.10746918681845e-06,
+      "loss": 0.7205,
+      "step": 3618
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.494247980704131,
+      "learning_rate": 7.105923607965786e-06,
+      "loss": 0.6274,
+      "step": 3619
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4050078671251107,
+      "learning_rate": 7.104377784453005e-06,
+      "loss": 0.6664,
+      "step": 3620
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.404246434896617,
+      "learning_rate": 7.102831716459696e-06,
+      "loss": 0.5954,
+      "step": 3621
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.623044062785884,
+      "learning_rate": 7.101285404165478e-06,
+      "loss": 0.6359,
+      "step": 3622
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5546602923700013,
+      "learning_rate": 7.099738847749995e-06,
+      "loss": 0.6142,
+      "step": 3623
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5511546021541984,
+      "learning_rate": 7.098192047392923e-06,
+      "loss": 0.6326,
+      "step": 3624
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.863354029549815,
+      "learning_rate": 7.096645003273964e-06,
+      "loss": 0.7335,
+      "step": 3625
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6001421531585835,
+      "learning_rate": 7.095097715572849e-06,
+      "loss": 0.6356,
+      "step": 3626
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6189887318128875,
+      "learning_rate": 7.093550184469339e-06,
+      "loss": 0.6734,
+      "step": 3627
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9686532488381214,
+      "learning_rate": 7.092002410143218e-06,
+      "loss": 0.6598,
+      "step": 3628
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.885198141114095,
+      "learning_rate": 7.0904543927743066e-06,
+      "loss": 0.7185,
+      "step": 3629
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.655688880025789,
+      "learning_rate": 7.088906132542446e-06,
+      "loss": 0.6541,
+      "step": 3630
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7287819033410154,
+      "learning_rate": 7.0873576296275096e-06,
+      "loss": 0.6549,
+      "step": 3631
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6674782827937946,
+      "learning_rate": 7.085808884209396e-06,
+      "loss": 0.5593,
+      "step": 3632
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3031490250344473,
+      "learning_rate": 7.084259896468038e-06,
+      "loss": 0.6337,
+      "step": 3633
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.243801593410752,
+      "learning_rate": 7.082710666583389e-06,
+      "loss": 0.5897,
+      "step": 3634
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.3382949550635908,
+      "learning_rate": 7.081161194735435e-06,
+      "loss": 0.7044,
+      "step": 3635
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2631769374260293,
+      "learning_rate": 7.0796114811041905e-06,
+      "loss": 0.6655,
+      "step": 3636
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1036281150291933,
+      "learning_rate": 7.078061525869695e-06,
+      "loss": 0.6465,
+      "step": 3637
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7686792433647867,
+      "learning_rate": 7.076511329212019e-06,
+      "loss": 0.674,
+      "step": 3638
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.7550739067072625,
+      "learning_rate": 7.074960891311258e-06,
+      "loss": 0.6734,
+      "step": 3639
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.177433134571054,
+      "learning_rate": 7.073410212347541e-06,
+      "loss": 0.6277,
+      "step": 3640
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7882879054667487,
+      "learning_rate": 7.071859292501018e-06,
+      "loss": 0.6493,
+      "step": 3641
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.434366748004624,
+      "learning_rate": 7.070308131951872e-06,
+      "loss": 0.6723,
+      "step": 3642
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7276152477544935,
+      "learning_rate": 7.068756730880311e-06,
+      "loss": 0.622,
+      "step": 3643
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.0417680257370083,
+      "learning_rate": 7.067205089466574e-06,
+      "loss": 0.6209,
+      "step": 3644
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.321254791228394,
+      "learning_rate": 7.065653207890924e-06,
+      "loss": 0.6661,
+      "step": 3645
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.8612826916827196,
+      "learning_rate": 7.064101086333657e-06,
+      "loss": 0.6949,
+      "step": 3646
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.1520069324380366,
+      "learning_rate": 7.0625487249750915e-06,
+      "loss": 0.6501,
+      "step": 3647
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.964198431485837,
+      "learning_rate": 7.060996123995576e-06,
+      "loss": 0.6407,
+      "step": 3648
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.93236095260122,
+      "learning_rate": 7.059443283575492e-06,
+      "loss": 0.5939,
+      "step": 3649
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6552920819656887,
+      "learning_rate": 7.0578902038952375e-06,
+      "loss": 0.6932,
+      "step": 3650
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9297734442881906,
+      "learning_rate": 7.056336885135251e-06,
+      "loss": 0.6419,
+      "step": 3651
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6224581309833943,
+      "learning_rate": 7.054783327475987e-06,
+      "loss": 0.675,
+      "step": 3652
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7160780202997263,
+      "learning_rate": 7.053229531097937e-06,
+      "loss": 0.656,
+      "step": 3653
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.3774431250824373,
+      "learning_rate": 7.051675496181614e-06,
+      "loss": 0.5999,
+      "step": 3654
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2783376063724854,
+      "learning_rate": 7.050121222907564e-06,
+      "loss": 0.5744,
+      "step": 3655
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3589367318753514,
+      "learning_rate": 7.048566711456355e-06,
+      "loss": 0.6371,
+      "step": 3656
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6774468417048696,
+      "learning_rate": 7.047011962008589e-06,
+      "loss": 0.6447,
+      "step": 3657
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4342842413375734,
+      "learning_rate": 7.04545697474489e-06,
+      "loss": 0.6903,
+      "step": 3658
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.252872706576466,
+      "learning_rate": 7.043901749845913e-06,
+      "loss": 0.6952,
+      "step": 3659
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3175374076123303,
+      "learning_rate": 7.042346287492339e-06,
+      "loss": 0.708,
+      "step": 3660
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.634929578368015,
+      "learning_rate": 7.040790587864875e-06,
+      "loss": 0.6424,
+      "step": 3661
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.618636727946655,
+      "learning_rate": 7.039234651144262e-06,
+      "loss": 0.6591,
+      "step": 3662
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.904861681615746,
+      "learning_rate": 7.037678477511261e-06,
+      "loss": 0.7491,
+      "step": 3663
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5649713303543047,
+      "learning_rate": 7.036122067146667e-06,
+      "loss": 0.703,
+      "step": 3664
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.220337768555499,
+      "learning_rate": 7.034565420231294e-06,
+      "loss": 0.7034,
+      "step": 3665
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.390572089955541,
+      "learning_rate": 7.033008536945994e-06,
+      "loss": 0.6504,
+      "step": 3666
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.7954821263514296,
+      "learning_rate": 7.031451417471638e-06,
+      "loss": 0.6958,
+      "step": 3667
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8545469996532997,
+      "learning_rate": 7.029894061989128e-06,
+      "loss": 0.7058,
+      "step": 3668
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.741678996189431,
+      "learning_rate": 7.0283364706793954e-06,
+      "loss": 0.6367,
+      "step": 3669
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.677654834887371,
+      "learning_rate": 7.026778643723393e-06,
+      "loss": 0.6911,
+      "step": 3670
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.061593434569085,
+      "learning_rate": 7.025220581302108e-06,
+      "loss": 0.6083,
+      "step": 3671
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.4271014763245535,
+      "learning_rate": 7.02366228359655e-06,
+      "loss": 0.632,
+      "step": 3672
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.875367995763348,
+      "learning_rate": 7.022103750787759e-06,
+      "loss": 0.7067,
+      "step": 3673
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7873572210988407,
+      "learning_rate": 7.020544983056796e-06,
+      "loss": 0.7194,
+      "step": 3674
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.662020767735063,
+      "learning_rate": 7.0189859805847615e-06,
+      "loss": 0.6952,
+      "step": 3675
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.8783628852915077,
+      "learning_rate": 7.017426743552769e-06,
+      "loss": 0.6343,
+      "step": 3676
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.5280735743445337,
+      "learning_rate": 7.015867272141972e-06,
+      "loss": 0.6424,
+      "step": 3677
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.457030504295436,
+      "learning_rate": 7.014307566533541e-06,
+      "loss": 0.6985,
+      "step": 3678
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.102622932285392,
+      "learning_rate": 7.0127476269086796e-06,
+      "loss": 0.6618,
+      "step": 3679
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.953298825950946,
+      "learning_rate": 7.011187453448617e-06,
+      "loss": 0.6986,
+      "step": 3680
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2057102088439478,
+      "learning_rate": 7.009627046334611e-06,
+      "loss": 0.7022,
+      "step": 3681
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.414332975549559,
+      "learning_rate": 7.008066405747943e-06,
+      "loss": 0.6429,
+      "step": 3682
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.484038423521051,
+      "learning_rate": 7.006505531869925e-06,
+      "loss": 0.7395,
+      "step": 3683
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.441293974717065,
+      "learning_rate": 7.004944424881894e-06,
+      "loss": 0.6179,
+      "step": 3684
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.12476887533077,
+      "learning_rate": 7.003383084965215e-06,
+      "loss": 0.5975,
+      "step": 3685
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4875957020716335,
+      "learning_rate": 7.001821512301283e-06,
+      "loss": 0.701,
+      "step": 3686
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2622990175164963,
+      "learning_rate": 7.000259707071512e-06,
+      "loss": 0.7091,
+      "step": 3687
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.4195893855491506,
+      "learning_rate": 6.9986976694573515e-06,
+      "loss": 0.6841,
+      "step": 3688
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.6453885611711816,
+      "learning_rate": 6.997135399640273e-06,
+      "loss": 0.6114,
+      "step": 3689
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.129611856788084,
+      "learning_rate": 6.9955728978017775e-06,
+      "loss": 0.6429,
+      "step": 3690
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2384526562197875,
+      "learning_rate": 6.99401016412339e-06,
+      "loss": 0.7059,
+      "step": 3691
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.206543127801006,
+      "learning_rate": 6.992447198786666e-06,
+      "loss": 0.6889,
+      "step": 3692
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5820156224344046,
+      "learning_rate": 6.990884001973187e-06,
+      "loss": 0.5918,
+      "step": 3693
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.705800904753045,
+      "learning_rate": 6.98932057386456e-06,
+      "loss": 0.6739,
+      "step": 3694
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.2287782223395998,
+      "learning_rate": 6.987756914642418e-06,
+      "loss": 0.5849,
+      "step": 3695
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.4820384692081654,
+      "learning_rate": 6.986193024488423e-06,
+      "loss": 0.6841,
+      "step": 3696
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.90798808346997,
+      "learning_rate": 6.984628903584266e-06,
+      "loss": 0.6208,
+      "step": 3697
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.045138928005131,
+      "learning_rate": 6.983064552111658e-06,
+      "loss": 0.6761,
+      "step": 3698
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.763765878384353,
+      "learning_rate": 6.981499970252345e-06,
+      "loss": 0.6429,
+      "step": 3699
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.8969396881884393,
+      "learning_rate": 6.979935158188091e-06,
+      "loss": 0.6543,
+      "step": 3700
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.144538256745873,
+      "learning_rate": 6.9783701161006965e-06,
+      "loss": 0.6848,
+      "step": 3701
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2691334424929632,
+      "learning_rate": 6.976804844171978e-06,
+      "loss": 0.682,
+      "step": 3702
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.853203277677684,
+      "learning_rate": 6.975239342583789e-06,
+      "loss": 0.5798,
+      "step": 3703
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.984156618342594,
+      "learning_rate": 6.973673611518003e-06,
+      "loss": 0.6253,
+      "step": 3704
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.6403380840996964,
+      "learning_rate": 6.972107651156521e-06,
+      "loss": 0.7099,
+      "step": 3705
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.0398511893391715,
+      "learning_rate": 6.970541461681274e-06,
+      "loss": 0.6499,
+      "step": 3706
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5614405274937453,
+      "learning_rate": 6.968975043274215e-06,
+      "loss": 0.5853,
+      "step": 3707
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.34052750735692,
+      "learning_rate": 6.9674083961173276e-06,
+      "loss": 0.7104,
+      "step": 3708
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5089454252540015,
+      "learning_rate": 6.96584152039262e-06,
+      "loss": 0.6719,
+      "step": 3709
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9541879261553783,
+      "learning_rate": 6.964274416282129e-06,
+      "loss": 0.6482,
+      "step": 3710
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3885655104600265,
+      "learning_rate": 6.962707083967911e-06,
+      "loss": 0.5981,
+      "step": 3711
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4321001454354527,
+      "learning_rate": 6.9611395236320615e-06,
+      "loss": 0.6967,
+      "step": 3712
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1448344326564757,
+      "learning_rate": 6.959571735456687e-06,
+      "loss": 0.6432,
+      "step": 3713
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.652298799399858,
+      "learning_rate": 6.958003719623936e-06,
+      "loss": 0.7021,
+      "step": 3714
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2558034126837194,
+      "learning_rate": 6.956435476315972e-06,
+      "loss": 0.7334,
+      "step": 3715
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.089987806720731,
+      "learning_rate": 6.9548670057149896e-06,
+      "loss": 0.6904,
+      "step": 3716
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.0275655323101884,
+      "learning_rate": 6.953298308003209e-06,
+      "loss": 0.6388,
+      "step": 3717
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6883335061425595,
+      "learning_rate": 6.9517293833628785e-06,
+      "loss": 0.666,
+      "step": 3718
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.102080658442691,
+      "learning_rate": 6.950160231976269e-06,
+      "loss": 0.6471,
+      "step": 3719
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.547523830477747,
+      "learning_rate": 6.948590854025681e-06,
+      "loss": 0.6384,
+      "step": 3720
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9940684741992807,
+      "learning_rate": 6.947021249693442e-06,
+      "loss": 0.6313,
+      "step": 3721
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.388166051557115,
+      "learning_rate": 6.9454514191619e-06,
+      "loss": 0.6452,
+      "step": 3722
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5275961581874453,
+      "learning_rate": 6.9438813626134395e-06,
+      "loss": 0.5957,
+      "step": 3723
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9456477007058033,
+      "learning_rate": 6.942311080230458e-06,
+      "loss": 0.6231,
+      "step": 3724
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.085258818767328,
+      "learning_rate": 6.940740572195392e-06,
+      "loss": 0.6739,
+      "step": 3725
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.189417657638878,
+      "learning_rate": 6.939169838690695e-06,
+      "loss": 0.6297,
+      "step": 3726
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.502191462085905,
+      "learning_rate": 6.937598879898853e-06,
+      "loss": 0.6238,
+      "step": 3727
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4771620597343773,
+      "learning_rate": 6.936027696002373e-06,
+      "loss": 0.7406,
+      "step": 3728
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3821967150760526,
+      "learning_rate": 6.934456287183793e-06,
+      "loss": 0.7131,
+      "step": 3729
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.537080782846546,
+      "learning_rate": 6.932884653625672e-06,
+      "loss": 0.587,
+      "step": 3730
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.044830093724147,
+      "learning_rate": 6.931312795510601e-06,
+      "loss": 0.6819,
+      "step": 3731
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.060390451396102,
+      "learning_rate": 6.929740713021192e-06,
+      "loss": 0.6625,
+      "step": 3732
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1547981802320297,
+      "learning_rate": 6.928168406340082e-06,
+      "loss": 0.7245,
+      "step": 3733
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.526368800735357,
+      "learning_rate": 6.926595875649944e-06,
+      "loss": 0.6918,
+      "step": 3734
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.751346484300469,
+      "learning_rate": 6.925023121133465e-06,
+      "loss": 0.5705,
+      "step": 3735
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.5756950930937323,
+      "learning_rate": 6.923450142973366e-06,
+      "loss": 0.7081,
+      "step": 3736
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3081997699324925,
+      "learning_rate": 6.921876941352388e-06,
+      "loss": 0.6519,
+      "step": 3737
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.6782336966893183,
+      "learning_rate": 6.920303516453302e-06,
+      "loss": 0.6644,
+      "step": 3738
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1786958735173174,
+      "learning_rate": 6.918729868458905e-06,
+      "loss": 0.6086,
+      "step": 3739
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1080710822626574,
+      "learning_rate": 6.91715599755202e-06,
+      "loss": 0.5619,
+      "step": 3740
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.3882194917466837,
+      "learning_rate": 6.9155819039154914e-06,
+      "loss": 0.6884,
+      "step": 3741
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2474420615290263,
+      "learning_rate": 6.9140075877321955e-06,
+      "loss": 0.7152,
+      "step": 3742
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3178490077460467,
+      "learning_rate": 6.91243304918503e-06,
+      "loss": 0.7279,
+      "step": 3743
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.915704362648865,
+      "learning_rate": 6.9108582884569206e-06,
+      "loss": 0.6947,
+      "step": 3744
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4243198642597314,
+      "learning_rate": 6.909283305730822e-06,
+      "loss": 0.7129,
+      "step": 3745
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.961387262313486,
+      "learning_rate": 6.907708101189705e-06,
+      "loss": 0.7032,
+      "step": 3746
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7889776855961554,
+      "learning_rate": 6.906132675016577e-06,
+      "loss": 0.5773,
+      "step": 3747
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.106825781647384,
+      "learning_rate": 6.904557027394464e-06,
+      "loss": 0.5643,
+      "step": 3748
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.128759223475058,
+      "learning_rate": 6.902981158506421e-06,
+      "loss": 0.6652,
+      "step": 3749
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0212995877258428,
+      "learning_rate": 6.90140506853553e-06,
+      "loss": 0.5981,
+      "step": 3750
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.410233497856306,
+      "learning_rate": 6.899828757664892e-06,
+      "loss": 0.6367,
+      "step": 3751
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5548357809398037,
+      "learning_rate": 6.898252226077642e-06,
+      "loss": 0.6518,
+      "step": 3752
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8928951894823056,
+      "learning_rate": 6.896675473956935e-06,
+      "loss": 0.6251,
+      "step": 3753
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8686199304183004,
+      "learning_rate": 6.895098501485955e-06,
+      "loss": 0.6949,
+      "step": 3754
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4005163798318754,
+      "learning_rate": 6.8935213088479096e-06,
+      "loss": 0.6532,
+      "step": 3755
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.083088075895648,
+      "learning_rate": 6.891943896226031e-06,
+      "loss": 0.6361,
+      "step": 3756
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3207947203497694,
+      "learning_rate": 6.890366263803579e-06,
+      "loss": 0.6775,
+      "step": 3757
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8029690208205515,
+      "learning_rate": 6.88878841176384e-06,
+      "loss": 0.614,
+      "step": 3758
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.913955180413663,
+      "learning_rate": 6.887210340290124e-06,
+      "loss": 0.6718,
+      "step": 3759
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4336755807961237,
+      "learning_rate": 6.885632049565766e-06,
+      "loss": 0.6733,
+      "step": 3760
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0818816161132827,
+      "learning_rate": 6.884053539774125e-06,
+      "loss": 0.6907,
+      "step": 3761
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.6742613246704785,
+      "learning_rate": 6.882474811098592e-06,
+      "loss": 0.6829,
+      "step": 3762
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9849851340007725,
+      "learning_rate": 6.880895863722576e-06,
+      "loss": 0.7172,
+      "step": 3763
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0447314238097247,
+      "learning_rate": 6.8793166978295166e-06,
+      "loss": 0.631,
+      "step": 3764
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.91470175770067,
+      "learning_rate": 6.877737313602876e-06,
+      "loss": 0.6654,
+      "step": 3765
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1785559232413867,
+      "learning_rate": 6.8761577112261425e-06,
+      "loss": 0.6318,
+      "step": 3766
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.669250198673815,
+      "learning_rate": 6.874577890882829e-06,
+      "loss": 0.6375,
+      "step": 3767
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9415389760168593,
+      "learning_rate": 6.872997852756474e-06,
+      "loss": 0.6013,
+      "step": 3768
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4526688080708596,
+      "learning_rate": 6.871417597030644e-06,
+      "loss": 0.6684,
+      "step": 3769
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.788823894332462,
+      "learning_rate": 6.869837123888926e-06,
+      "loss": 0.6253,
+      "step": 3770
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.477851029463746,
+      "learning_rate": 6.868256433514938e-06,
+      "loss": 0.7383,
+      "step": 3771
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.269796038141395,
+      "learning_rate": 6.866675526092317e-06,
+      "loss": 0.7487,
+      "step": 3772
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.648425239153188,
+      "learning_rate": 6.86509440180473e-06,
+      "loss": 0.7038,
+      "step": 3773
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.375374473672006,
+      "learning_rate": 6.863513060835866e-06,
+      "loss": 0.6327,
+      "step": 3774
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1344879105982528,
+      "learning_rate": 6.861931503369441e-06,
+      "loss": 0.6503,
+      "step": 3775
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.840442440534222,
+      "learning_rate": 6.8603497295891975e-06,
+      "loss": 0.5909,
+      "step": 3776
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.9504848651490403,
+      "learning_rate": 6.8587677396789e-06,
+      "loss": 0.6947,
+      "step": 3777
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.764177497995866,
+      "learning_rate": 6.8571855338223395e-06,
+      "loss": 0.6062,
+      "step": 3778
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.440160369843259,
+      "learning_rate": 6.855603112203333e-06,
+      "loss": 0.6853,
+      "step": 3779
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2362406157836525,
+      "learning_rate": 6.854020475005719e-06,
+      "loss": 0.6329,
+      "step": 3780
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1361521836950608,
+      "learning_rate": 6.852437622413366e-06,
+      "loss": 0.5591,
+      "step": 3781
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2958928068872213,
+      "learning_rate": 6.850854554610167e-06,
+      "loss": 0.7093,
+      "step": 3782
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.311428807633486,
+      "learning_rate": 6.849271271780034e-06,
+      "loss": 0.6284,
+      "step": 3783
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8763588330667194,
+      "learning_rate": 6.847687774106911e-06,
+      "loss": 0.6576,
+      "step": 3784
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.382119861964838,
+      "learning_rate": 6.846104061774763e-06,
+      "loss": 0.6577,
+      "step": 3785
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0655614717842052,
+      "learning_rate": 6.8445201349675825e-06,
+      "loss": 0.6832,
+      "step": 3786
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8485737316689894,
+      "learning_rate": 6.842935993869385e-06,
+      "loss": 0.6554,
+      "step": 3787
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6641033052353102,
+      "learning_rate": 6.841351638664211e-06,
+      "loss": 0.7066,
+      "step": 3788
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.636994092546543,
+      "learning_rate": 6.839767069536126e-06,
+      "loss": 0.5704,
+      "step": 3789
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1059693268960524,
+      "learning_rate": 6.838182286669222e-06,
+      "loss": 0.6862,
+      "step": 3790
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.389714565065946,
+      "learning_rate": 6.8365972902476115e-06,
+      "loss": 0.6674,
+      "step": 3791
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3867920333404187,
+      "learning_rate": 6.835012080455439e-06,
+      "loss": 0.6521,
+      "step": 3792
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6159918503535615,
+      "learning_rate": 6.833426657476866e-06,
+      "loss": 0.6519,
+      "step": 3793
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.226258850434491,
+      "learning_rate": 6.831841021496084e-06,
+      "loss": 0.7133,
+      "step": 3794
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.899461642291162,
+      "learning_rate": 6.830255172697309e-06,
+      "loss": 0.6671,
+      "step": 3795
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3305187723569447,
+      "learning_rate": 6.828669111264776e-06,
+      "loss": 0.6724,
+      "step": 3796
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4055301005112946,
+      "learning_rate": 6.8270828373827536e-06,
+      "loss": 0.8054,
+      "step": 3797
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2070086468105874,
+      "learning_rate": 6.825496351235528e-06,
+      "loss": 0.6663,
+      "step": 3798
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.465861609693202,
+      "learning_rate": 6.823909653007414e-06,
+      "loss": 0.6668,
+      "step": 3799
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7446295867469264,
+      "learning_rate": 6.8223227428827485e-06,
+      "loss": 0.6663,
+      "step": 3800
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2768346355361744,
+      "learning_rate": 6.820735621045895e-06,
+      "loss": 0.6675,
+      "step": 3801
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3820614356193843,
+      "learning_rate": 6.81914828768124e-06,
+      "loss": 0.689,
+      "step": 3802
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.993505071374713,
+      "learning_rate": 6.817560742973196e-06,
+      "loss": 0.7702,
+      "step": 3803
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.25574301607733,
+      "learning_rate": 6.8159729871061984e-06,
+      "loss": 0.6731,
+      "step": 3804
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.880274384584933,
+      "learning_rate": 6.814385020264708e-06,
+      "loss": 0.7064,
+      "step": 3805
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6969745237833136,
+      "learning_rate": 6.812796842633213e-06,
+      "loss": 0.7117,
+      "step": 3806
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.10614106669541,
+      "learning_rate": 6.811208454396218e-06,
+      "loss": 0.7491,
+      "step": 3807
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.469718689532213,
+      "learning_rate": 6.809619855738262e-06,
+      "loss": 0.6633,
+      "step": 3808
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6641844592932244,
+      "learning_rate": 6.8080310468439015e-06,
+      "loss": 0.6784,
+      "step": 3809
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.960823525262656,
+      "learning_rate": 6.806442027897722e-06,
+      "loss": 0.6114,
+      "step": 3810
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.963680773700924,
+      "learning_rate": 6.804852799084329e-06,
+      "loss": 0.6952,
+      "step": 3811
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.327360252561889,
+      "learning_rate": 6.803263360588355e-06,
+      "loss": 0.6871,
+      "step": 3812
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6946911855419358,
+      "learning_rate": 6.801673712594456e-06,
+      "loss": 0.7192,
+      "step": 3813
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7849793322013197,
+      "learning_rate": 6.8000838552873135e-06,
+      "loss": 0.6403,
+      "step": 3814
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3535336364449804,
+      "learning_rate": 6.7984937888516325e-06,
+      "loss": 0.708,
+      "step": 3815
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.483684306418722,
+      "learning_rate": 6.796903513472142e-06,
+      "loss": 0.6605,
+      "step": 3816
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.157785406165776,
+      "learning_rate": 6.795313029333596e-06,
+      "loss": 0.576,
+      "step": 3817
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.243032193761467,
+      "learning_rate": 6.79372233662077e-06,
+      "loss": 0.5717,
+      "step": 3818
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0101289906232367,
+      "learning_rate": 6.79213143551847e-06,
+      "loss": 0.5542,
+      "step": 3819
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.8401966485891643,
+      "learning_rate": 6.790540326211519e-06,
+      "loss": 0.6712,
+      "step": 3820
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.757074235173517,
+      "learning_rate": 6.78894900888477e-06,
+      "loss": 0.6846,
+      "step": 3821
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.674263473701696,
+      "learning_rate": 6.787357483723096e-06,
+      "loss": 0.7419,
+      "step": 3822
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.606110641324874,
+      "learning_rate": 6.785765750911396e-06,
+      "loss": 0.6317,
+      "step": 3823
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1675322487659723,
+      "learning_rate": 6.7841738106345935e-06,
+      "loss": 0.6373,
+      "step": 3824
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2312923273279153,
+      "learning_rate": 6.782581663077637e-06,
+      "loss": 0.6132,
+      "step": 3825
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6581842644039164,
+      "learning_rate": 6.780989308425493e-06,
+      "loss": 0.6928,
+      "step": 3826
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.246504818054555,
+      "learning_rate": 6.77939674686316e-06,
+      "loss": 0.6127,
+      "step": 3827
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7698626969798905,
+      "learning_rate": 6.7778039785756575e-06,
+      "loss": 0.6525,
+      "step": 3828
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4450311084573038,
+      "learning_rate": 6.7762110037480265e-06,
+      "loss": 0.6343,
+      "step": 3829
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.123523847157009,
+      "learning_rate": 6.774617822565338e-06,
+      "loss": 0.6966,
+      "step": 3830
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7410973711990247,
+      "learning_rate": 6.773024435212678e-06,
+      "loss": 0.6834,
+      "step": 3831
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.729971927019776,
+      "learning_rate": 6.771430841875166e-06,
+      "loss": 0.7335,
+      "step": 3832
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.461919317112508,
+      "learning_rate": 6.7698370427379405e-06,
+      "loss": 0.7462,
+      "step": 3833
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.764361831143325,
+      "learning_rate": 6.7682430379861615e-06,
+      "loss": 0.6316,
+      "step": 3834
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.09261926247929,
+      "learning_rate": 6.766648827805019e-06,
+      "loss": 0.6006,
+      "step": 3835
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.804234407107111,
+      "learning_rate": 6.765054412379722e-06,
+      "loss": 0.6852,
+      "step": 3836
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.798217238649048,
+      "learning_rate": 6.763459791895506e-06,
+      "loss": 0.6397,
+      "step": 3837
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6552060440599377,
+      "learning_rate": 6.761864966537629e-06,
+      "loss": 0.5773,
+      "step": 3838
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0551205343106513,
+      "learning_rate": 6.760269936491373e-06,
+      "loss": 0.7127,
+      "step": 3839
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2555093947538323,
+      "learning_rate": 6.7586747019420444e-06,
+      "loss": 0.6463,
+      "step": 3840
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.508518018543992,
+      "learning_rate": 6.7570792630749725e-06,
+      "loss": 0.6151,
+      "step": 3841
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2115480509362024,
+      "learning_rate": 6.755483620075509e-06,
+      "loss": 0.6153,
+      "step": 3842
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5571778252409416,
+      "learning_rate": 6.753887773129036e-06,
+      "loss": 0.6822,
+      "step": 3843
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.3827224492687704,
+      "learning_rate": 6.752291722420951e-06,
+      "loss": 0.6651,
+      "step": 3844
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.068263890882636,
+      "learning_rate": 6.750695468136679e-06,
+      "loss": 0.7441,
+      "step": 3845
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.599488472042048,
+      "learning_rate": 6.7490990104616684e-06,
+      "loss": 0.6053,
+      "step": 3846
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.068319078405129,
+      "learning_rate": 6.74750234958139e-06,
+      "loss": 0.6563,
+      "step": 3847
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.4661100791938293,
+      "learning_rate": 6.745905485681341e-06,
+      "loss": 0.605,
+      "step": 3848
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8783604735605857,
+      "learning_rate": 6.7443084189470385e-06,
+      "loss": 0.6415,
+      "step": 3849
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.159785400047253,
+      "learning_rate": 6.742711149564028e-06,
+      "loss": 0.7075,
+      "step": 3850
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5896401663997524,
+      "learning_rate": 6.741113677717872e-06,
+      "loss": 0.7611,
+      "step": 3851
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.423614667213638,
+      "learning_rate": 6.7395160035941624e-06,
+      "loss": 0.7156,
+      "step": 3852
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.874352439060714,
+      "learning_rate": 6.737918127378511e-06,
+      "loss": 0.5727,
+      "step": 3853
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2310646192841106,
+      "learning_rate": 6.736320049256557e-06,
+      "loss": 0.6234,
+      "step": 3854
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.078116761718466,
+      "learning_rate": 6.734721769413959e-06,
+      "loss": 0.6838,
+      "step": 3855
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.047167191192513,
+      "learning_rate": 6.733123288036399e-06,
+      "loss": 0.739,
+      "step": 3856
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.98595396084378,
+      "learning_rate": 6.731524605309587e-06,
+      "loss": 0.6049,
+      "step": 3857
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.152132778242366,
+      "learning_rate": 6.72992572141925e-06,
+      "loss": 0.6779,
+      "step": 3858
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.411218764669821,
+      "learning_rate": 6.728326636551145e-06,
+      "loss": 0.6093,
+      "step": 3859
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1157066121453587,
+      "learning_rate": 6.726727350891047e-06,
+      "loss": 0.6165,
+      "step": 3860
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.068433705431225,
+      "learning_rate": 6.725127864624757e-06,
+      "loss": 0.5947,
+      "step": 3861
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8708815594354373,
+      "learning_rate": 6.723528177938097e-06,
+      "loss": 0.6354,
+      "step": 3862
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.33467858012365,
+      "learning_rate": 6.721928291016917e-06,
+      "loss": 0.7318,
+      "step": 3863
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.803975371742299,
+      "learning_rate": 6.720328204047085e-06,
+      "loss": 0.6814,
+      "step": 3864
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1642857039266927,
+      "learning_rate": 6.718727917214496e-06,
+      "loss": 0.6067,
+      "step": 3865
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.80771219731655,
+      "learning_rate": 6.7171274307050645e-06,
+      "loss": 0.6535,
+      "step": 3866
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0379961191421874,
+      "learning_rate": 6.715526744704732e-06,
+      "loss": 0.6385,
+      "step": 3867
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2538626104452626,
+      "learning_rate": 6.7139258593994625e-06,
+      "loss": 0.6347,
+      "step": 3868
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5560569234327355,
+      "learning_rate": 6.712324774975241e-06,
+      "loss": 0.6909,
+      "step": 3869
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4724982923209953,
+      "learning_rate": 6.710723491618077e-06,
+      "loss": 0.676,
+      "step": 3870
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.692509799497406,
+      "learning_rate": 6.709122009514003e-06,
+      "loss": 0.673,
+      "step": 3871
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2919297459345502,
+      "learning_rate": 6.707520328849074e-06,
+      "loss": 0.6929,
+      "step": 3872
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.882493860740183,
+      "learning_rate": 6.7059184498093696e-06,
+      "loss": 0.6791,
+      "step": 3873
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.6357067311902735,
+      "learning_rate": 6.70431637258099e-06,
+      "loss": 0.7293,
+      "step": 3874
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.322268264406318,
+      "learning_rate": 6.702714097350063e-06,
+      "loss": 0.6628,
+      "step": 3875
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4783117896301388,
+      "learning_rate": 6.701111624302732e-06,
+      "loss": 0.6576,
+      "step": 3876
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5768699541037674,
+      "learning_rate": 6.69950895362517e-06,
+      "loss": 0.7493,
+      "step": 3877
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.682147247687123,
+      "learning_rate": 6.697906085503572e-06,
+      "loss": 0.6796,
+      "step": 3878
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2802199978055913,
+      "learning_rate": 6.696303020124152e-06,
+      "loss": 0.6168,
+      "step": 3879
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.264182484401708,
+      "learning_rate": 6.694699757673151e-06,
+      "loss": 0.6551,
+      "step": 3880
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5897290859559083,
+      "learning_rate": 6.693096298336832e-06,
+      "loss": 0.6541,
+      "step": 3881
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.05265338675788,
+      "learning_rate": 6.691492642301478e-06,
+      "loss": 0.6983,
+      "step": 3882
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.55619801462537,
+      "learning_rate": 6.6898887897533985e-06,
+      "loss": 0.7378,
+      "step": 3883
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.7122207089430885,
+      "learning_rate": 6.6882847408789255e-06,
+      "loss": 0.6772,
+      "step": 3884
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8904289019733236,
+      "learning_rate": 6.686680495864411e-06,
+      "loss": 0.6001,
+      "step": 3885
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1642154485237306,
+      "learning_rate": 6.6850760548962315e-06,
+      "loss": 0.7032,
+      "step": 3886
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.085180115876389,
+      "learning_rate": 6.683471418160787e-06,
+      "loss": 0.6206,
+      "step": 3887
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.33182419264645,
+      "learning_rate": 6.681866585844501e-06,
+      "loss": 0.5902,
+      "step": 3888
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.158864307849004,
+      "learning_rate": 6.680261558133816e-06,
+      "loss": 0.5937,
+      "step": 3889
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7965197556825463,
+      "learning_rate": 6.678656335215199e-06,
+      "loss": 0.6601,
+      "step": 3890
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8204118744480993,
+      "learning_rate": 6.677050917275143e-06,
+      "loss": 0.6472,
+      "step": 3891
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.4806015074983594,
+      "learning_rate": 6.675445304500159e-06,
+      "loss": 0.6548,
+      "step": 3892
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1229091161769098,
+      "learning_rate": 6.673839497076783e-06,
+      "loss": 0.5833,
+      "step": 3893
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1338450581533546,
+      "learning_rate": 6.672233495191572e-06,
+      "loss": 0.6849,
+      "step": 3894
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3209497422936622,
+      "learning_rate": 6.670627299031109e-06,
+      "loss": 0.7225,
+      "step": 3895
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.76307749364276,
+      "learning_rate": 6.669020908781994e-06,
+      "loss": 0.6026,
+      "step": 3896
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4047612077616933,
+      "learning_rate": 6.667414324630856e-06,
+      "loss": 0.6428,
+      "step": 3897
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.962203572754492,
+      "learning_rate": 6.665807546764341e-06,
+      "loss": 0.6385,
+      "step": 3898
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3629283595989983,
+      "learning_rate": 6.664200575369121e-06,
+      "loss": 0.6119,
+      "step": 3899
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1049318573672307,
+      "learning_rate": 6.662593410631888e-06,
+      "loss": 0.6005,
+      "step": 3900
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1219922103429094,
+      "learning_rate": 6.660986052739357e-06,
+      "loss": 0.6908,
+      "step": 3901
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.5048628027957625,
+      "learning_rate": 6.659378501878271e-06,
+      "loss": 0.5462,
+      "step": 3902
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.081524730634284,
+      "learning_rate": 6.657770758235387e-06,
+      "loss": 0.6065,
+      "step": 3903
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.286231378737835,
+      "learning_rate": 6.656162821997487e-06,
+      "loss": 0.6641,
+      "step": 3904
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.018088065090707,
+      "learning_rate": 6.654554693351379e-06,
+      "loss": 0.5698,
+      "step": 3905
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0133508154340487,
+      "learning_rate": 6.652946372483889e-06,
+      "loss": 0.6064,
+      "step": 3906
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.681912809689596,
+      "learning_rate": 6.651337859581868e-06,
+      "loss": 0.6913,
+      "step": 3907
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0310073690024075,
+      "learning_rate": 6.6497291548321876e-06,
+      "loss": 0.6523,
+      "step": 3908
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3565442897512345,
+      "learning_rate": 6.648120258421744e-06,
+      "loss": 0.6395,
+      "step": 3909
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.52692979561079,
+      "learning_rate": 6.646511170537452e-06,
+      "loss": 0.6792,
+      "step": 3910
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.21140083501598,
+      "learning_rate": 6.644901891366251e-06,
+      "loss": 0.5643,
+      "step": 3911
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0497136827512572,
+      "learning_rate": 6.643292421095105e-06,
+      "loss": 0.5934,
+      "step": 3912
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.4574095439368064,
+      "learning_rate": 6.641682759910993e-06,
+      "loss": 0.6165,
+      "step": 3913
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1072803681357684,
+      "learning_rate": 6.640072908000926e-06,
+      "loss": 0.7101,
+      "step": 3914
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.449280579698595,
+      "learning_rate": 6.638462865551929e-06,
+      "loss": 0.631,
+      "step": 3915
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.300143552458361,
+      "learning_rate": 6.6368526327510515e-06,
+      "loss": 0.6856,
+      "step": 3916
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1842280323695746,
+      "learning_rate": 6.635242209785369e-06,
+      "loss": 0.7335,
+      "step": 3917
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.49088278011825,
+      "learning_rate": 6.633631596841972e-06,
+      "loss": 0.7089,
+      "step": 3918
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9551946728374507,
+      "learning_rate": 6.63202079410798e-06,
+      "loss": 0.7198,
+      "step": 3919
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5169889269326773,
+      "learning_rate": 6.630409801770528e-06,
+      "loss": 0.6483,
+      "step": 3920
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5244649430685646,
+      "learning_rate": 6.6287986200167785e-06,
+      "loss": 0.6174,
+      "step": 3921
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6269573809565423,
+      "learning_rate": 6.627187249033915e-06,
+      "loss": 0.6274,
+      "step": 3922
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.855773301785785,
+      "learning_rate": 6.6255756890091394e-06,
+      "loss": 0.6258,
+      "step": 3923
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4211612959153817,
+      "learning_rate": 6.6239639401296796e-06,
+      "loss": 0.6101,
+      "step": 3924
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.259452031681648,
+      "learning_rate": 6.622352002582783e-06,
+      "loss": 0.6934,
+      "step": 3925
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0056254388720265,
+      "learning_rate": 6.620739876555721e-06,
+      "loss": 0.7024,
+      "step": 3926
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.950640775072797,
+      "learning_rate": 6.619127562235786e-06,
+      "loss": 0.7432,
+      "step": 3927
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.111457506555049,
+      "learning_rate": 6.61751505981029e-06,
+      "loss": 0.6785,
+      "step": 3928
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.205473900545925,
+      "learning_rate": 6.615902369466571e-06,
+      "loss": 0.6155,
+      "step": 3929
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1442082817667982,
+      "learning_rate": 6.614289491391985e-06,
+      "loss": 0.6085,
+      "step": 3930
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.097758491032023,
+      "learning_rate": 6.612676425773914e-06,
+      "loss": 0.6214,
+      "step": 3931
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.102972042758122,
+      "learning_rate": 6.6110631727997566e-06,
+      "loss": 0.6056,
+      "step": 3932
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9937462472618483,
+      "learning_rate": 6.609449732656936e-06,
+      "loss": 0.6597,
+      "step": 3933
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5897301973961366,
+      "learning_rate": 6.6078361055328986e-06,
+      "loss": 0.6214,
+      "step": 3934
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0149618917681156,
+      "learning_rate": 6.606222291615112e-06,
+      "loss": 0.6023,
+      "step": 3935
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.253201954300004,
+      "learning_rate": 6.60460829109106e-06,
+      "loss": 0.6611,
+      "step": 3936
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9592197675312657,
+      "learning_rate": 6.602994104148256e-06,
+      "loss": 0.7145,
+      "step": 3937
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2284657591030883,
+      "learning_rate": 6.601379730974231e-06,
+      "loss": 0.618,
+      "step": 3938
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.557356204052841,
+      "learning_rate": 6.599765171756538e-06,
+      "loss": 0.6578,
+      "step": 3939
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.091674295085208,
+      "learning_rate": 6.598150426682752e-06,
+      "loss": 0.6179,
+      "step": 3940
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0618356021886943,
+      "learning_rate": 6.596535495940468e-06,
+      "loss": 0.6168,
+      "step": 3941
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4424041643505827,
+      "learning_rate": 6.594920379717307e-06,
+      "loss": 0.6277,
+      "step": 3942
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6813309064292485,
+      "learning_rate": 6.593305078200907e-06,
+      "loss": 0.6051,
+      "step": 3943
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.067586995607296,
+      "learning_rate": 6.591689591578927e-06,
+      "loss": 0.631,
+      "step": 3944
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1568079707086967,
+      "learning_rate": 6.590073920039052e-06,
+      "loss": 0.6446,
+      "step": 3945
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7031629573317324,
+      "learning_rate": 6.588458063768985e-06,
+      "loss": 0.6328,
+      "step": 3946
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.406212144524127,
+      "learning_rate": 6.586842022956453e-06,
+      "loss": 0.6518,
+      "step": 3947
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.66916693541006,
+      "learning_rate": 6.585225797789201e-06,
+      "loss": 0.6624,
+      "step": 3948
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2582234615243375,
+      "learning_rate": 6.583609388454998e-06,
+      "loss": 0.555,
+      "step": 3949
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7199146063706134,
+      "learning_rate": 6.581992795141634e-06,
+      "loss": 0.6526,
+      "step": 3950
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5390407294947397,
+      "learning_rate": 6.580376018036921e-06,
+      "loss": 0.6671,
+      "step": 3951
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.7476904706174,
+      "learning_rate": 6.578759057328691e-06,
+      "loss": 0.6829,
+      "step": 3952
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5382661667053528,
+      "learning_rate": 6.5771419132047965e-06,
+      "loss": 0.6645,
+      "step": 3953
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4043076498085805,
+      "learning_rate": 6.5755245858531135e-06,
+      "loss": 0.5974,
+      "step": 3954
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0677814480436205,
+      "learning_rate": 6.573907075461538e-06,
+      "loss": 0.6752,
+      "step": 3955
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3299465309321636,
+      "learning_rate": 6.57228938221799e-06,
+      "loss": 0.7099,
+      "step": 3956
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.5992623283864615,
+      "learning_rate": 6.5706715063104065e-06,
+      "loss": 0.7091,
+      "step": 3957
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1185187454894776,
+      "learning_rate": 6.569053447926746e-06,
+      "loss": 0.5476,
+      "step": 3958
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.4670327552541123,
+      "learning_rate": 6.567435207254993e-06,
+      "loss": 0.679,
+      "step": 3959
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.602615705445102,
+      "learning_rate": 6.565816784483147e-06,
+      "loss": 0.6774,
+      "step": 3960
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.735033391286663,
+      "learning_rate": 6.564198179799234e-06,
+      "loss": 0.6466,
+      "step": 3961
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.8877061902133954,
+      "learning_rate": 6.5625793933912985e-06,
+      "loss": 0.594,
+      "step": 3962
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.814956124692221,
+      "learning_rate": 6.5609604254474065e-06,
+      "loss": 0.6185,
+      "step": 3963
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1484699815057058,
+      "learning_rate": 6.559341276155644e-06,
+      "loss": 0.6048,
+      "step": 3964
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.872534078697945,
+      "learning_rate": 6.55772194570412e-06,
+      "loss": 0.6446,
+      "step": 3965
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.229090156228293,
+      "learning_rate": 6.5561024342809625e-06,
+      "loss": 0.6325,
+      "step": 3966
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5077624702277195,
+      "learning_rate": 6.554482742074323e-06,
+      "loss": 0.633,
+      "step": 3967
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6547530593774695,
+      "learning_rate": 6.552862869272371e-06,
+      "loss": 0.6035,
+      "step": 3968
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.367826287135605,
+      "learning_rate": 6.551242816063302e-06,
+      "loss": 0.6763,
+      "step": 3969
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4066120293512174,
+      "learning_rate": 6.5496225826353254e-06,
+      "loss": 0.6515,
+      "step": 3970
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5260789922901496,
+      "learning_rate": 6.548002169176677e-06,
+      "loss": 0.651,
+      "step": 3971
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3066437486350457,
+      "learning_rate": 6.54638157587561e-06,
+      "loss": 0.5797,
+      "step": 3972
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.181813637695597,
+      "learning_rate": 6.544760802920402e-06,
+      "loss": 0.6741,
+      "step": 3973
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.3250608385498466,
+      "learning_rate": 6.543139850499352e-06,
+      "loss": 0.6409,
+      "step": 3974
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3644447611901835,
+      "learning_rate": 6.5415187188007726e-06,
+      "loss": 0.6754,
+      "step": 3975
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.405053664018577,
+      "learning_rate": 6.539897408013005e-06,
+      "loss": 0.6084,
+      "step": 3976
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1188551440111305,
+      "learning_rate": 6.538275918324408e-06,
+      "loss": 0.7079,
+      "step": 3977
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.439917462899966,
+      "learning_rate": 6.536654249923361e-06,
+      "loss": 0.6137,
+      "step": 3978
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6239270579211404,
+      "learning_rate": 6.535032402998266e-06,
+      "loss": 0.5237,
+      "step": 3979
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.4664442974636303,
+      "learning_rate": 6.533410377737544e-06,
+      "loss": 0.6955,
+      "step": 3980
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3044985882207807,
+      "learning_rate": 6.531788174329636e-06,
+      "loss": 0.5246,
+      "step": 3981
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.007003515117586,
+      "learning_rate": 6.5301657929630055e-06,
+      "loss": 0.6039,
+      "step": 3982
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2164380112159745,
+      "learning_rate": 6.5285432338261365e-06,
+      "loss": 0.616,
+      "step": 3983
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.257824172908572,
+      "learning_rate": 6.526920497107535e-06,
+      "loss": 0.7186,
+      "step": 3984
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.093528093998501,
+      "learning_rate": 6.525297582995722e-06,
+      "loss": 0.5979,
+      "step": 3985
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3370494508343467,
+      "learning_rate": 6.523674491679246e-06,
+      "loss": 0.639,
+      "step": 3986
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1729179222230477,
+      "learning_rate": 6.522051223346672e-06,
+      "loss": 0.6788,
+      "step": 3987
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.462532611308276,
+      "learning_rate": 6.520427778186586e-06,
+      "loss": 0.7091,
+      "step": 3988
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.673478312971706,
+      "learning_rate": 6.518804156387597e-06,
+      "loss": 0.6831,
+      "step": 3989
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.9158479320514985,
+      "learning_rate": 6.517180358138332e-06,
+      "loss": 0.6926,
+      "step": 3990
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.635411987857093,
+      "learning_rate": 6.515556383627437e-06,
+      "loss": 0.666,
+      "step": 3991
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2626676614651906,
+      "learning_rate": 6.513932233043584e-06,
+      "loss": 0.7135,
+      "step": 3992
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.83263520238375,
+      "learning_rate": 6.512307906575459e-06,
+      "loss": 0.5915,
+      "step": 3993
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.204046953504845,
+      "learning_rate": 6.510683404411774e-06,
+      "loss": 0.6271,
+      "step": 3994
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.593299117358978,
+      "learning_rate": 6.509058726741258e-06,
+      "loss": 0.6616,
+      "step": 3995
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.996690826929998,
+      "learning_rate": 6.50743387375266e-06,
+      "loss": 0.6885,
+      "step": 3996
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 12.933011472212376,
+      "learning_rate": 6.505808845634753e-06,
+      "loss": 0.6603,
+      "step": 3997
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.453187947445511,
+      "learning_rate": 6.504183642576327e-06,
+      "loss": 0.7046,
+      "step": 3998
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4465722034806157,
+      "learning_rate": 6.502558264766194e-06,
+      "loss": 0.5633,
+      "step": 3999
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.341833452210207,
+      "learning_rate": 6.500932712393185e-06,
+      "loss": 0.6048,
+      "step": 4000
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5314039211160955,
+      "learning_rate": 6.499306985646152e-06,
+      "loss": 0.6565,
+      "step": 4001
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.307652014877147,
+      "learning_rate": 6.497681084713969e-06,
+      "loss": 0.6175,
+      "step": 4002
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.7936798318747624,
+      "learning_rate": 6.496055009785526e-06,
+      "loss": 0.6992,
+      "step": 4003
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6272219714248983,
+      "learning_rate": 6.494428761049736e-06,
+      "loss": 0.6342,
+      "step": 4004
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1079646961470906,
+      "learning_rate": 6.492802338695533e-06,
+      "loss": 0.6606,
+      "step": 4005
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.297324004495899,
+      "learning_rate": 6.491175742911869e-06,
+      "loss": 0.6246,
+      "step": 4006
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6530141636498663,
+      "learning_rate": 6.489548973887717e-06,
+      "loss": 0.6609,
+      "step": 4007
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1063440835651397,
+      "learning_rate": 6.4879220318120735e-06,
+      "loss": 0.6974,
+      "step": 4008
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3068250586790517,
+      "learning_rate": 6.486294916873947e-06,
+      "loss": 0.6624,
+      "step": 4009
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.459942839783369,
+      "learning_rate": 6.484667629262375e-06,
+      "loss": 0.6707,
+      "step": 4010
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1027394322760666,
+      "learning_rate": 6.4830401691664106e-06,
+      "loss": 0.6924,
+      "step": 4011
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1296970727448212,
+      "learning_rate": 6.481412536775125e-06,
+      "loss": 0.5682,
+      "step": 4012
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.62301953143833,
+      "learning_rate": 6.479784732277612e-06,
+      "loss": 0.6738,
+      "step": 4013
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8179450664566996,
+      "learning_rate": 6.478156755862988e-06,
+      "loss": 0.6378,
+      "step": 4014
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7307924595408544,
+      "learning_rate": 6.4765286077203844e-06,
+      "loss": 0.5877,
+      "step": 4015
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.571571986675738,
+      "learning_rate": 6.4749002880389566e-06,
+      "loss": 0.6638,
+      "step": 4016
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.603346420805311,
+      "learning_rate": 6.473271797007876e-06,
+      "loss": 0.6619,
+      "step": 4017
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.174453465113902,
+      "learning_rate": 6.471643134816336e-06,
+      "loss": 0.7298,
+      "step": 4018
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5043155473625514,
+      "learning_rate": 6.470014301653552e-06,
+      "loss": 0.6076,
+      "step": 4019
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8733996645916116,
+      "learning_rate": 6.468385297708755e-06,
+      "loss": 0.6422,
+      "step": 4020
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0053641792809447,
+      "learning_rate": 6.466756123171199e-06,
+      "loss": 0.6017,
+      "step": 4021
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.170505948062594,
+      "learning_rate": 6.465126778230156e-06,
+      "loss": 0.6564,
+      "step": 4022
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2632578787097453,
+      "learning_rate": 6.46349726307492e-06,
+      "loss": 0.6804,
+      "step": 4023
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.936556398536724,
+      "learning_rate": 6.4618675778948026e-06,
+      "loss": 0.6739,
+      "step": 4024
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1638172540039973,
+      "learning_rate": 6.4602377228791345e-06,
+      "loss": 0.6198,
+      "step": 4025
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.599114630574275,
+      "learning_rate": 6.458607698217271e-06,
+      "loss": 0.5974,
+      "step": 4026
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.359851858598546,
+      "learning_rate": 6.45697750409858e-06,
+      "loss": 0.6207,
+      "step": 4027
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0414819211011148,
+      "learning_rate": 6.455347140712455e-06,
+      "loss": 0.5482,
+      "step": 4028
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9073262379332099,
+      "learning_rate": 6.453716608248306e-06,
+      "loss": 0.7179,
+      "step": 4029
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.894734987817326,
+      "learning_rate": 6.452085906895564e-06,
+      "loss": 0.6275,
+      "step": 4030
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.409131230764491,
+      "learning_rate": 6.450455036843677e-06,
+      "loss": 0.5708,
+      "step": 4031
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.626357432074768,
+      "learning_rate": 6.448823998282119e-06,
+      "loss": 0.6892,
+      "step": 4032
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3318787973963424,
+      "learning_rate": 6.447192791400375e-06,
+      "loss": 0.6684,
+      "step": 4033
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.341295423290045,
+      "learning_rate": 6.4455614163879576e-06,
+      "loss": 0.7042,
+      "step": 4034
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.51877291354241,
+      "learning_rate": 6.443929873434392e-06,
+      "loss": 0.6783,
+      "step": 4035
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.530723558434043,
+      "learning_rate": 6.442298162729229e-06,
+      "loss": 0.652,
+      "step": 4036
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.544615833177302,
+      "learning_rate": 6.4406662844620346e-06,
+      "loss": 0.6204,
+      "step": 4037
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6992866068672643,
+      "learning_rate": 6.439034238822396e-06,
+      "loss": 0.6114,
+      "step": 4038
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.361537811446634,
+      "learning_rate": 6.4374020259999194e-06,
+      "loss": 0.6084,
+      "step": 4039
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.3932419299570737,
+      "learning_rate": 6.43576964618423e-06,
+      "loss": 0.5435,
+      "step": 4040
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2247947497001896,
+      "learning_rate": 6.4341370995649735e-06,
+      "loss": 0.6274,
+      "step": 4041
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.203698930496953,
+      "learning_rate": 6.4325043863318136e-06,
+      "loss": 0.6563,
+      "step": 4042
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6533644667178184,
+      "learning_rate": 6.430871506674437e-06,
+      "loss": 0.5898,
+      "step": 4043
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.265730806748852,
+      "learning_rate": 6.429238460782543e-06,
+      "loss": 0.5698,
+      "step": 4044
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1642899955947548,
+      "learning_rate": 6.427605248845859e-06,
+      "loss": 0.6199,
+      "step": 4045
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6003134035460604,
+      "learning_rate": 6.42597187105412e-06,
+      "loss": 0.6055,
+      "step": 4046
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2193001480951926,
+      "learning_rate": 6.4243383275970924e-06,
+      "loss": 0.671,
+      "step": 4047
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.5572778898263016,
+      "learning_rate": 6.422704618664557e-06,
+      "loss": 0.6435,
+      "step": 4048
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.222157891641835,
+      "learning_rate": 6.42107074444631e-06,
+      "loss": 0.6501,
+      "step": 4049
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.725601175114304,
+      "learning_rate": 6.419436705132172e-06,
+      "loss": 0.6936,
+      "step": 4050
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4112413353819484,
+      "learning_rate": 6.4178025009119825e-06,
+      "loss": 0.6389,
+      "step": 4051
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.589327647378624,
+      "learning_rate": 6.416168131975595e-06,
+      "loss": 0.6985,
+      "step": 4052
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4983806609239703,
+      "learning_rate": 6.414533598512887e-06,
+      "loss": 0.6664,
+      "step": 4053
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.240075636586892,
+      "learning_rate": 6.412898900713757e-06,
+      "loss": 0.6559,
+      "step": 4054
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8847812663112897,
+      "learning_rate": 6.411264038768115e-06,
+      "loss": 0.6703,
+      "step": 4055
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.367558885670087,
+      "learning_rate": 6.409629012865899e-06,
+      "loss": 0.5792,
+      "step": 4056
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.3855942924416382,
+      "learning_rate": 6.407993823197056e-06,
+      "loss": 0.7401,
+      "step": 4057
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.876331931348521,
+      "learning_rate": 6.406358469951562e-06,
+      "loss": 0.6485,
+      "step": 4058
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.206685527616314,
+      "learning_rate": 6.404722953319406e-06,
+      "loss": 0.6424,
+      "step": 4059
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.075266887505169,
+      "learning_rate": 6.403087273490599e-06,
+      "loss": 0.6713,
+      "step": 4060
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.856229721355921,
+      "learning_rate": 6.401451430655168e-06,
+      "loss": 0.628,
+      "step": 4061
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0646056376603608,
+      "learning_rate": 6.399815425003161e-06,
+      "loss": 0.6839,
+      "step": 4062
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0561582164002683,
+      "learning_rate": 6.398179256724644e-06,
+      "loss": 0.6327,
+      "step": 4063
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1686914862568094,
+      "learning_rate": 6.396542926009703e-06,
+      "loss": 0.5924,
+      "step": 4064
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.452702288596918,
+      "learning_rate": 6.394906433048442e-06,
+      "loss": 0.6387,
+      "step": 4065
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.400913216633786,
+      "learning_rate": 6.3932697780309825e-06,
+      "loss": 0.6913,
+      "step": 4066
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.108636102788443,
+      "learning_rate": 6.3916329611474705e-06,
+      "loss": 0.6406,
+      "step": 4067
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.411671185444273,
+      "learning_rate": 6.389995982588061e-06,
+      "loss": 0.6552,
+      "step": 4068
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.186150326143783,
+      "learning_rate": 6.388358842542939e-06,
+      "loss": 0.6234,
+      "step": 4069
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.859739939616308,
+      "learning_rate": 6.386721541202296e-06,
+      "loss": 0.6838,
+      "step": 4070
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.460365538292539,
+      "learning_rate": 6.3850840787563565e-06,
+      "loss": 0.5673,
+      "step": 4071
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.402069347480076,
+      "learning_rate": 6.383446455395352e-06,
+      "loss": 0.6837,
+      "step": 4072
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.2533740088151664,
+      "learning_rate": 6.3818086713095374e-06,
+      "loss": 0.6788,
+      "step": 4073
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2858335624830324,
+      "learning_rate": 6.380170726689185e-06,
+      "loss": 0.7238,
+      "step": 4074
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.484411238766058,
+      "learning_rate": 6.378532621724588e-06,
+      "loss": 0.6292,
+      "step": 4075
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.867064268787102,
+      "learning_rate": 6.376894356606056e-06,
+      "loss": 0.5916,
+      "step": 4076
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.291422789683661,
+      "learning_rate": 6.375255931523916e-06,
+      "loss": 0.6287,
+      "step": 4077
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1675402432820405,
+      "learning_rate": 6.373617346668519e-06,
+      "loss": 0.5687,
+      "step": 4078
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2952408596599088,
+      "learning_rate": 6.371978602230229e-06,
+      "loss": 0.6288,
+      "step": 4079
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.141696812562484,
+      "learning_rate": 6.370339698399432e-06,
+      "loss": 0.6161,
+      "step": 4080
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.8668906189282164,
+      "learning_rate": 6.3687006353665285e-06,
+      "loss": 0.6,
+      "step": 4081
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.230532760665731,
+      "learning_rate": 6.367061413321942e-06,
+      "loss": 0.6102,
+      "step": 4082
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.68067571548342,
+      "learning_rate": 6.365422032456113e-06,
+      "loss": 0.64,
+      "step": 4083
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.574913062898836,
+      "learning_rate": 6.363782492959499e-06,
+      "loss": 0.6528,
+      "step": 4084
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.168916407705001,
+      "learning_rate": 6.362142795022578e-06,
+      "loss": 0.6213,
+      "step": 4085
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7508588377575767,
+      "learning_rate": 6.360502938835844e-06,
+      "loss": 0.6605,
+      "step": 4086
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.4135918342912577,
+      "learning_rate": 6.35886292458981e-06,
+      "loss": 0.6797,
+      "step": 4087
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.555329036173374,
+      "learning_rate": 6.35722275247501e-06,
+      "loss": 0.691,
+      "step": 4088
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.3576452705524025,
+      "learning_rate": 6.355582422681996e-06,
+      "loss": 0.6184,
+      "step": 4089
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.01931185233685,
+      "learning_rate": 6.353941935401333e-06,
+      "loss": 0.6206,
+      "step": 4090
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5007928234352677,
+      "learning_rate": 6.352301290823611e-06,
+      "loss": 0.6169,
+      "step": 4091
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.556883601705964,
+      "learning_rate": 6.350660489139433e-06,
+      "loss": 0.6329,
+      "step": 4092
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5252161752585427,
+      "learning_rate": 6.349019530539425e-06,
+      "loss": 0.6555,
+      "step": 4093
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5247540508658526,
+      "learning_rate": 6.347378415214226e-06,
+      "loss": 0.6158,
+      "step": 4094
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.182461495472191,
+      "learning_rate": 6.3457371433544975e-06,
+      "loss": 0.7378,
+      "step": 4095
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6257667849308226,
+      "learning_rate": 6.34409571515092e-06,
+      "loss": 0.6825,
+      "step": 4096
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4163772839213733,
+      "learning_rate": 6.342454130794186e-06,
+      "loss": 0.5942,
+      "step": 4097
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.548472388600027,
+      "learning_rate": 6.340812390475012e-06,
+      "loss": 0.6612,
+      "step": 4098
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6494112784435204,
+      "learning_rate": 6.33917049438413e-06,
+      "loss": 0.7117,
+      "step": 4099
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.975703696322628,
+      "learning_rate": 6.3375284427122915e-06,
+      "loss": 0.726,
+      "step": 4100
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.124929550258763,
+      "learning_rate": 6.335886235650264e-06,
+      "loss": 0.599,
+      "step": 4101
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5941633006024762,
+      "learning_rate": 6.334243873388838e-06,
+      "loss": 0.7624,
+      "step": 4102
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.256377886619029,
+      "learning_rate": 6.332601356118813e-06,
+      "loss": 0.5449,
+      "step": 4103
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.9556721482690977,
+      "learning_rate": 6.330958684031016e-06,
+      "loss": 0.6911,
+      "step": 4104
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.733540693074302,
+      "learning_rate": 6.329315857316285e-06,
+      "loss": 0.6602,
+      "step": 4105
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.101209748842294,
+      "learning_rate": 6.327672876165481e-06,
+      "loss": 0.6334,
+      "step": 4106
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.230682903167034,
+      "learning_rate": 6.326029740769481e-06,
+      "loss": 0.6723,
+      "step": 4107
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.8797719596760216,
+      "learning_rate": 6.324386451319179e-06,
+      "loss": 0.6211,
+      "step": 4108
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1406370239153274,
+      "learning_rate": 6.322743008005488e-06,
+      "loss": 0.6444,
+      "step": 4109
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.712312589801154,
+      "learning_rate": 6.321099411019336e-06,
+      "loss": 0.6464,
+      "step": 4110
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.233736620239088,
+      "learning_rate": 6.319455660551674e-06,
+      "loss": 0.6325,
+      "step": 4111
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.038358257077351,
+      "learning_rate": 6.317811756793467e-06,
+      "loss": 0.6742,
+      "step": 4112
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.143450702921909,
+      "learning_rate": 6.316167699935702e-06,
+      "loss": 0.6708,
+      "step": 4113
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.8705182545419534,
+      "learning_rate": 6.314523490169375e-06,
+      "loss": 0.5526,
+      "step": 4114
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.073511807722659,
+      "learning_rate": 6.312879127685512e-06,
+      "loss": 0.6384,
+      "step": 4115
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.653135511720908,
+      "learning_rate": 6.311234612675143e-06,
+      "loss": 0.709,
+      "step": 4116
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0677937210863715,
+      "learning_rate": 6.309589945329332e-06,
+      "loss": 0.6508,
+      "step": 4117
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.338910592281395,
+      "learning_rate": 6.307945125839143e-06,
+      "loss": 0.678,
+      "step": 4118
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.715799303343155,
+      "learning_rate": 6.3063001543956715e-06,
+      "loss": 0.6564,
+      "step": 4119
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4933350769017357,
+      "learning_rate": 6.304655031190024e-06,
+      "loss": 0.5819,
+      "step": 4120
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.044652665239333,
+      "learning_rate": 6.303009756413327e-06,
+      "loss": 0.6325,
+      "step": 4121
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.095867866117808,
+      "learning_rate": 6.3013643302567225e-06,
+      "loss": 0.6967,
+      "step": 4122
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.849872100475443,
+      "learning_rate": 6.299718752911371e-06,
+      "loss": 0.7464,
+      "step": 4123
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9748453249208904,
+      "learning_rate": 6.298073024568454e-06,
+      "loss": 0.6542,
+      "step": 4124
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2600324826958365,
+      "learning_rate": 6.296427145419164e-06,
+      "loss": 0.5836,
+      "step": 4125
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.802818402196517,
+      "learning_rate": 6.294781115654718e-06,
+      "loss": 0.6668,
+      "step": 4126
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.676232174418196,
+      "learning_rate": 6.293134935466342e-06,
+      "loss": 0.6596,
+      "step": 4127
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.228739627526291,
+      "learning_rate": 6.291488605045288e-06,
+      "loss": 0.5731,
+      "step": 4128
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6049506326721703,
+      "learning_rate": 6.289842124582822e-06,
+      "loss": 0.5658,
+      "step": 4129
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6541990876497694,
+      "learning_rate": 6.2881954942702265e-06,
+      "loss": 0.6145,
+      "step": 4130
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0595704649971625,
+      "learning_rate": 6.286548714298801e-06,
+      "loss": 0.6131,
+      "step": 4131
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.420647946200874,
+      "learning_rate": 6.284901784859866e-06,
+      "loss": 0.6737,
+      "step": 4132
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.041851672678566,
+      "learning_rate": 6.283254706144756e-06,
+      "loss": 0.6566,
+      "step": 4133
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.712373901252542,
+      "learning_rate": 6.281607478344823e-06,
+      "loss": 0.6394,
+      "step": 4134
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4741338459449853,
+      "learning_rate": 6.279960101651439e-06,
+      "loss": 0.7035,
+      "step": 4135
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9857951374629823,
+      "learning_rate": 6.278312576255988e-06,
+      "loss": 0.6226,
+      "step": 4136
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.3494664192954384,
+      "learning_rate": 6.276664902349881e-06,
+      "loss": 0.6609,
+      "step": 4137
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2270633620793228,
+      "learning_rate": 6.275017080124533e-06,
+      "loss": 0.6487,
+      "step": 4138
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.2352633292562363,
+      "learning_rate": 6.273369109771387e-06,
+      "loss": 0.7118,
+      "step": 4139
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.268087699809955,
+      "learning_rate": 6.271720991481897e-06,
+      "loss": 0.6689,
+      "step": 4140
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.099756078374374,
+      "learning_rate": 6.270072725447542e-06,
+      "loss": 0.7536,
+      "step": 4141
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4160875739712644,
+      "learning_rate": 6.268424311859808e-06,
+      "loss": 0.6797,
+      "step": 4142
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.1608916128751066,
+      "learning_rate": 6.266775750910203e-06,
+      "loss": 0.6239,
+      "step": 4143
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.642655322409538,
+      "learning_rate": 6.265127042790253e-06,
+      "loss": 0.7106,
+      "step": 4144
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.48267619018257,
+      "learning_rate": 6.263478187691502e-06,
+      "loss": 0.6842,
+      "step": 4145
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2073464735347583,
+      "learning_rate": 6.2618291858055065e-06,
+      "loss": 0.716,
+      "step": 4146
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.2192677534009446,
+      "learning_rate": 6.260180037323843e-06,
+      "loss": 0.6008,
+      "step": 4147
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8880065557661028,
+      "learning_rate": 6.258530742438107e-06,
+      "loss": 0.6526,
+      "step": 4148
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.923108042812208,
+      "learning_rate": 6.256881301339907e-06,
+      "loss": 0.5658,
+      "step": 4149
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3215056242914405,
+      "learning_rate": 6.255231714220871e-06,
+      "loss": 0.664,
+      "step": 4150
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.598137114476907,
+      "learning_rate": 6.253581981272641e-06,
+      "loss": 0.6033,
+      "step": 4151
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.292031250462775,
+      "learning_rate": 6.251932102686883e-06,
+      "loss": 0.6393,
+      "step": 4152
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9711599193058196,
+      "learning_rate": 6.2502820786552695e-06,
+      "loss": 0.7229,
+      "step": 4153
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.4776790489556118,
+      "learning_rate": 6.2486319093695006e-06,
+      "loss": 0.6444,
+      "step": 4154
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.574952252391362,
+      "learning_rate": 6.246981595021284e-06,
+      "loss": 0.6376,
+      "step": 4155
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.485791306054375,
+      "learning_rate": 6.245331135802351e-06,
+      "loss": 0.6843,
+      "step": 4156
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.581681311574103,
+      "learning_rate": 6.243680531904448e-06,
+      "loss": 0.6037,
+      "step": 4157
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.6015305354626954,
+      "learning_rate": 6.242029783519334e-06,
+      "loss": 0.6377,
+      "step": 4158
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.514525749517879,
+      "learning_rate": 6.240378890838792e-06,
+      "loss": 0.688,
+      "step": 4159
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1131329820828673,
+      "learning_rate": 6.238727854054614e-06,
+      "loss": 0.6627,
+      "step": 4160
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3327635807486757,
+      "learning_rate": 6.237076673358616e-06,
+      "loss": 0.6625,
+      "step": 4161
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.564116461115227,
+      "learning_rate": 6.235425348942625e-06,
+      "loss": 0.6025,
+      "step": 4162
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 25.51131291171202,
+      "learning_rate": 6.2337738809984905e-06,
+      "loss": 0.6809,
+      "step": 4163
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.404343930899044,
+      "learning_rate": 6.23212226971807e-06,
+      "loss": 0.6425,
+      "step": 4164
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.986695766080571,
+      "learning_rate": 6.230470515293248e-06,
+      "loss": 0.674,
+      "step": 4165
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5976631521397846,
+      "learning_rate": 6.2288186179159175e-06,
+      "loss": 0.6709,
+      "step": 4166
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 12.729513809408251,
+      "learning_rate": 6.227166577777992e-06,
+      "loss": 0.6686,
+      "step": 4167
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.51466132483747,
+      "learning_rate": 6.225514395071401e-06,
+      "loss": 0.7087,
+      "step": 4168
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7437128880343717,
+      "learning_rate": 6.223862069988091e-06,
+      "loss": 0.6417,
+      "step": 4169
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.7709270732185365,
+      "learning_rate": 6.222209602720023e-06,
+      "loss": 0.6492,
+      "step": 4170
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6653325226917635,
+      "learning_rate": 6.220556993459174e-06,
+      "loss": 0.6882,
+      "step": 4171
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.290790865423674,
+      "learning_rate": 6.218904242397546e-06,
+      "loss": 0.6277,
+      "step": 4172
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.0699405954785104,
+      "learning_rate": 6.217251349727145e-06,
+      "loss": 0.7121,
+      "step": 4173
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.734217356286444,
+      "learning_rate": 6.215598315640001e-06,
+      "loss": 0.7427,
+      "step": 4174
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.713834982597247,
+      "learning_rate": 6.213945140328157e-06,
+      "loss": 0.6955,
+      "step": 4175
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.311432316870064,
+      "learning_rate": 6.212291823983678e-06,
+      "loss": 0.6851,
+      "step": 4176
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.399335810386324,
+      "learning_rate": 6.2106383667986385e-06,
+      "loss": 0.6589,
+      "step": 4177
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1179610916249185,
+      "learning_rate": 6.208984768965133e-06,
+      "loss": 0.585,
+      "step": 4178
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.492230402979365,
+      "learning_rate": 6.207331030675272e-06,
+      "loss": 0.6906,
+      "step": 4179
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.05084961096072,
+      "learning_rate": 6.2056771521211815e-06,
+      "loss": 0.6719,
+      "step": 4180
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.525761336265004,
+      "learning_rate": 6.204023133495005e-06,
+      "loss": 0.6352,
+      "step": 4181
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2251077581550414,
+      "learning_rate": 6.2023689749889e-06,
+      "loss": 0.6294,
+      "step": 4182
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3917017204363615,
+      "learning_rate": 6.2007146767950455e-06,
+      "loss": 0.6562,
+      "step": 4183
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.7924703663410675,
+      "learning_rate": 6.199060239105628e-06,
+      "loss": 0.6758,
+      "step": 4184
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3713543077213854,
+      "learning_rate": 6.197405662112862e-06,
+      "loss": 0.6653,
+      "step": 4185
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.439148199597049,
+      "learning_rate": 6.195750946008965e-06,
+      "loss": 0.6896,
+      "step": 4186
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.7616942745004667,
+      "learning_rate": 6.19409609098618e-06,
+      "loss": 0.6445,
+      "step": 4187
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.610494475761507,
+      "learning_rate": 6.192441097236762e-06,
+      "loss": 0.7334,
+      "step": 4188
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.098962128255649,
+      "learning_rate": 6.190785964952985e-06,
+      "loss": 0.634,
+      "step": 4189
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.11012265793801,
+      "learning_rate": 6.189130694327138e-06,
+      "loss": 0.6304,
+      "step": 4190
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6761648104534377,
+      "learning_rate": 6.187475285551523e-06,
+      "loss": 0.6746,
+      "step": 4191
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6521807011331933,
+      "learning_rate": 6.185819738818463e-06,
+      "loss": 0.6235,
+      "step": 4192
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.263291536626014,
+      "learning_rate": 6.184164054320293e-06,
+      "loss": 0.6194,
+      "step": 4193
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0093262515523898,
+      "learning_rate": 6.1825082322493655e-06,
+      "loss": 0.6021,
+      "step": 4194
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.4984369533755295,
+      "learning_rate": 6.18085227279805e-06,
+      "loss": 0.6466,
+      "step": 4195
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.873242485567296,
+      "learning_rate": 6.179196176158733e-06,
+      "loss": 0.6112,
+      "step": 4196
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2153603654893232,
+      "learning_rate": 6.17753994252381e-06,
+      "loss": 0.677,
+      "step": 4197
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.885706072220139,
+      "learning_rate": 6.175883572085703e-06,
+      "loss": 0.6734,
+      "step": 4198
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.306773750236914,
+      "learning_rate": 6.1742270650368395e-06,
+      "loss": 0.6614,
+      "step": 4199
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.0757107827698014,
+      "learning_rate": 6.172570421569672e-06,
+      "loss": 0.6607,
+      "step": 4200
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6497916669129955,
+      "learning_rate": 6.170913641876662e-06,
+      "loss": 0.6356,
+      "step": 4201
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4729495304399536,
+      "learning_rate": 6.1692567261502885e-06,
+      "loss": 0.5511,
+      "step": 4202
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4085363284193635,
+      "learning_rate": 6.167599674583049e-06,
+      "loss": 0.6771,
+      "step": 4203
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5904214468463556,
+      "learning_rate": 6.165942487367456e-06,
+      "loss": 0.6186,
+      "step": 4204
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 11.74561826819688,
+      "learning_rate": 6.164285164696034e-06,
+      "loss": 0.698,
+      "step": 4205
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4895776594825447,
+      "learning_rate": 6.162627706761326e-06,
+      "loss": 0.6161,
+      "step": 4206
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.9883010766926974,
+      "learning_rate": 6.160970113755894e-06,
+      "loss": 0.6723,
+      "step": 4207
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.412743390251808,
+      "learning_rate": 6.159312385872309e-06,
+      "loss": 0.6407,
+      "step": 4208
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.988545368414137,
+      "learning_rate": 6.157654523303164e-06,
+      "loss": 0.6851,
+      "step": 4209
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4530378165050104,
+      "learning_rate": 6.15599652624106e-06,
+      "loss": 0.6844,
+      "step": 4210
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.747802711725356,
+      "learning_rate": 6.154338394878624e-06,
+      "loss": 0.6251,
+      "step": 4211
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.6279563635896923,
+      "learning_rate": 6.152680129408488e-06,
+      "loss": 0.6315,
+      "step": 4212
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6101031201569262,
+      "learning_rate": 6.151021730023308e-06,
+      "loss": 0.6899,
+      "step": 4213
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8097748740269637,
+      "learning_rate": 6.14936319691575e-06,
+      "loss": 0.6828,
+      "step": 4214
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5163859758968785,
+      "learning_rate": 6.147704530278497e-06,
+      "loss": 0.7089,
+      "step": 4215
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8122326932337884,
+      "learning_rate": 6.146045730304252e-06,
+      "loss": 0.7728,
+      "step": 4216
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2991108660602873,
+      "learning_rate": 6.144386797185724e-06,
+      "loss": 0.6771,
+      "step": 4217
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.82067248565608,
+      "learning_rate": 6.14272773111565e-06,
+      "loss": 0.6825,
+      "step": 4218
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.292155901876321,
+      "learning_rate": 6.141068532286768e-06,
+      "loss": 0.7232,
+      "step": 4219
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.2138852203020436,
+      "learning_rate": 6.139409200891845e-06,
+      "loss": 0.6331,
+      "step": 4220
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.733361759557919,
+      "learning_rate": 6.137749737123652e-06,
+      "loss": 0.5726,
+      "step": 4221
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.045897121812701,
+      "learning_rate": 6.136090141174986e-06,
+      "loss": 0.6303,
+      "step": 4222
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.215595577195097,
+      "learning_rate": 6.134430413238649e-06,
+      "loss": 0.7165,
+      "step": 4223
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4720795018859287,
+      "learning_rate": 6.132770553507468e-06,
+      "loss": 0.5848,
+      "step": 4224
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.469345000674683,
+      "learning_rate": 6.1311105621742775e-06,
+      "loss": 0.6669,
+      "step": 4225
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.625205929898641,
+      "learning_rate": 6.129450439431932e-06,
+      "loss": 0.6472,
+      "step": 4226
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.5548855414175695,
+      "learning_rate": 6.1277901854732994e-06,
+      "loss": 0.6817,
+      "step": 4227
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1027568650104005,
+      "learning_rate": 6.126129800491263e-06,
+      "loss": 0.6514,
+      "step": 4228
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.045046875891458,
+      "learning_rate": 6.124469284678721e-06,
+      "loss": 0.5825,
+      "step": 4229
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1138142631995698,
+      "learning_rate": 6.122808638228588e-06,
+      "loss": 0.6002,
+      "step": 4230
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.3093105459556447,
+      "learning_rate": 6.121147861333795e-06,
+      "loss": 0.7347,
+      "step": 4231
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.637992867101944,
+      "learning_rate": 6.119486954187283e-06,
+      "loss": 0.6293,
+      "step": 4232
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.2443163383582645,
+      "learning_rate": 6.117825916982013e-06,
+      "loss": 0.6475,
+      "step": 4233
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3645005998517554,
+      "learning_rate": 6.116164749910959e-06,
+      "loss": 0.7004,
+      "step": 4234
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 12.488986762043982,
+      "learning_rate": 6.114503453167112e-06,
+      "loss": 0.6375,
+      "step": 4235
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.456714677650742,
+      "learning_rate": 6.112842026943473e-06,
+      "loss": 0.7201,
+      "step": 4236
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.792896979064036,
+      "learning_rate": 6.111180471433067e-06,
+      "loss": 0.7046,
+      "step": 4237
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.311085039466233,
+      "learning_rate": 6.109518786828924e-06,
+      "loss": 0.6333,
+      "step": 4238
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.082195260565265,
+      "learning_rate": 6.107856973324097e-06,
+      "loss": 0.7398,
+      "step": 4239
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.422658221464742,
+      "learning_rate": 6.106195031111648e-06,
+      "loss": 0.626,
+      "step": 4240
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.357834278269825,
+      "learning_rate": 6.104532960384658e-06,
+      "loss": 0.6259,
+      "step": 4241
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0886646213733924,
+      "learning_rate": 6.1028707613362236e-06,
+      "loss": 0.7135,
+      "step": 4242
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.8932851903398698,
+      "learning_rate": 6.101208434159451e-06,
+      "loss": 0.6996,
+      "step": 4243
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3623621800314547,
+      "learning_rate": 6.099545979047465e-06,
+      "loss": 0.6463,
+      "step": 4244
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3794475175322924,
+      "learning_rate": 6.097883396193406e-06,
+      "loss": 0.6134,
+      "step": 4245
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2159715788742234,
+      "learning_rate": 6.09622068579043e-06,
+      "loss": 0.6379,
+      "step": 4246
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.65089220209315,
+      "learning_rate": 6.094557848031699e-06,
+      "loss": 0.6913,
+      "step": 4247
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.238068508379566,
+      "learning_rate": 6.092894883110405e-06,
+      "loss": 0.6267,
+      "step": 4248
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.62840874216548,
+      "learning_rate": 6.0912317912197416e-06,
+      "loss": 0.6864,
+      "step": 4249
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7152502107110603,
+      "learning_rate": 6.089568572552923e-06,
+      "loss": 0.6414,
+      "step": 4250
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0634647134438302,
+      "learning_rate": 6.087905227303177e-06,
+      "loss": 0.5915,
+      "step": 4251
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6867407708114026,
+      "learning_rate": 6.086241755663746e-06,
+      "loss": 0.668,
+      "step": 4252
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0245261285927338,
+      "learning_rate": 6.08457815782789e-06,
+      "loss": 0.5945,
+      "step": 4253
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.184617873138533,
+      "learning_rate": 6.082914433988875e-06,
+      "loss": 0.566,
+      "step": 4254
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2773216580867945,
+      "learning_rate": 6.081250584339996e-06,
+      "loss": 0.5985,
+      "step": 4255
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2623346319662763,
+      "learning_rate": 6.079586609074547e-06,
+      "loss": 0.6648,
+      "step": 4256
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.594546200291953,
+      "learning_rate": 6.077922508385849e-06,
+      "loss": 0.6895,
+      "step": 4257
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2014493515781806,
+      "learning_rate": 6.076258282467227e-06,
+      "loss": 0.6097,
+      "step": 4258
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6694207927575233,
+      "learning_rate": 6.074593931512031e-06,
+      "loss": 0.6801,
+      "step": 4259
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.451259021434374,
+      "learning_rate": 6.072929455713616e-06,
+      "loss": 0.6692,
+      "step": 4260
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9337354819470554,
+      "learning_rate": 6.07126485526536e-06,
+      "loss": 0.5854,
+      "step": 4261
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.767008015068956,
+      "learning_rate": 6.0696001303606486e-06,
+      "loss": 0.6037,
+      "step": 4262
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0086485367597264,
+      "learning_rate": 6.067935281192887e-06,
+      "loss": 0.6432,
+      "step": 4263
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.153528393185996,
+      "learning_rate": 6.066270307955492e-06,
+      "loss": 0.643,
+      "step": 4264
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.1645611909710882,
+      "learning_rate": 6.064605210841893e-06,
+      "loss": 0.6053,
+      "step": 4265
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.309866791740079,
+      "learning_rate": 6.062939990045541e-06,
+      "loss": 0.6321,
+      "step": 4266
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.156210421450635,
+      "learning_rate": 6.06127464575989e-06,
+      "loss": 0.5757,
+      "step": 4267
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.244885997343262,
+      "learning_rate": 6.059609178178423e-06,
+      "loss": 0.6942,
+      "step": 4268
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.152053051101937,
+      "learning_rate": 6.0579435874946205e-06,
+      "loss": 0.6633,
+      "step": 4269
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.478964275396135,
+      "learning_rate": 6.056277873901993e-06,
+      "loss": 0.6697,
+      "step": 4270
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.9729373074740866,
+      "learning_rate": 6.054612037594053e-06,
+      "loss": 0.7223,
+      "step": 4271
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0783619133176625,
+      "learning_rate": 6.052946078764337e-06,
+      "loss": 0.6165,
+      "step": 4272
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2263809793723803,
+      "learning_rate": 6.0512799976063885e-06,
+      "loss": 0.586,
+      "step": 4273
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.84734542991169,
+      "learning_rate": 6.049613794313769e-06,
+      "loss": 0.651,
+      "step": 4274
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3793929939042915,
+      "learning_rate": 6.047947469080053e-06,
+      "loss": 0.7264,
+      "step": 4275
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.7894502963984307,
+      "learning_rate": 6.0462810220988284e-06,
+      "loss": 0.5848,
+      "step": 4276
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.385912284914175,
+      "learning_rate": 6.044614453563702e-06,
+      "loss": 0.6738,
+      "step": 4277
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0450272266233553,
+      "learning_rate": 6.042947763668285e-06,
+      "loss": 0.746,
+      "step": 4278
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0469407502606782,
+      "learning_rate": 6.041280952606214e-06,
+      "loss": 0.5847,
+      "step": 4279
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.531872581619871,
+      "learning_rate": 6.03961402057113e-06,
+      "loss": 0.5961,
+      "step": 4280
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9848807465163134,
+      "learning_rate": 6.037946967756696e-06,
+      "loss": 0.6215,
+      "step": 4281
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.090185818907428,
+      "learning_rate": 6.036279794356582e-06,
+      "loss": 0.6028,
+      "step": 4282
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.995346619697124,
+      "learning_rate": 6.034612500564479e-06,
+      "loss": 0.5208,
+      "step": 4283
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7529881198497277,
+      "learning_rate": 6.032945086574085e-06,
+      "loss": 0.6228,
+      "step": 4284
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.3379083789144857,
+      "learning_rate": 6.0312775525791165e-06,
+      "loss": 0.6813,
+      "step": 4285
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.119809893657813,
+      "learning_rate": 6.029609898773305e-06,
+      "loss": 0.6076,
+      "step": 4286
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5052511733476788,
+      "learning_rate": 6.027942125350389e-06,
+      "loss": 0.6299,
+      "step": 4287
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.655210831114237,
+      "learning_rate": 6.02627423250413e-06,
+      "loss": 0.6856,
+      "step": 4288
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.625661240338276,
+      "learning_rate": 6.024606220428297e-06,
+      "loss": 0.647,
+      "step": 4289
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.2243027452718698,
+      "learning_rate": 6.022938089316677e-06,
+      "loss": 0.7011,
+      "step": 4290
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3340700591649557,
+      "learning_rate": 6.021269839363063e-06,
+      "loss": 0.6817,
+      "step": 4291
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0637410458294068,
+      "learning_rate": 6.019601470761275e-06,
+      "loss": 0.668,
+      "step": 4292
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.865208891942265,
+      "learning_rate": 6.017932983705132e-06,
+      "loss": 0.6142,
+      "step": 4293
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.9027349130767828,
+      "learning_rate": 6.016264378388481e-06,
+      "loss": 0.6485,
+      "step": 4294
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.9168453504696266,
+      "learning_rate": 6.0145956550051694e-06,
+      "loss": 0.6187,
+      "step": 4295
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7032692104534073,
+      "learning_rate": 6.01292681374907e-06,
+      "loss": 0.6664,
+      "step": 4296
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.223566307096264,
+      "learning_rate": 6.01125785481406e-06,
+      "loss": 0.5642,
+      "step": 4297
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.504081111094006,
+      "learning_rate": 6.009588778394035e-06,
+      "loss": 0.6222,
+      "step": 4298
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8721849635784116,
+      "learning_rate": 6.0079195846829055e-06,
+      "loss": 0.629,
+      "step": 4299
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9713414450562043,
+      "learning_rate": 6.006250273874591e-06,
+      "loss": 0.6071,
+      "step": 4300
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0636910457431736,
+      "learning_rate": 6.0045808461630295e-06,
+      "loss": 0.606,
+      "step": 4301
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4935316363007756,
+      "learning_rate": 6.002911301742168e-06,
+      "loss": 0.6487,
+      "step": 4302
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0612655080904037,
+      "learning_rate": 6.001241640805973e-06,
+      "loss": 0.6267,
+      "step": 4303
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.5302669242733815,
+      "learning_rate": 5.999571863548416e-06,
+      "loss": 0.6001,
+      "step": 4304
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4210531577615226,
+      "learning_rate": 5.997901970163491e-06,
+      "loss": 0.6732,
+      "step": 4305
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4958888183227415,
+      "learning_rate": 5.996231960845198e-06,
+      "loss": 0.6682,
+      "step": 4306
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5503369133292892,
+      "learning_rate": 5.994561835787558e-06,
+      "loss": 0.5959,
+      "step": 4307
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.176063899739418,
+      "learning_rate": 5.992891595184596e-06,
+      "loss": 0.6718,
+      "step": 4308
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4856814525301485,
+      "learning_rate": 5.991221239230362e-06,
+      "loss": 0.6848,
+      "step": 4309
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.966837942768863,
+      "learning_rate": 5.989550768118908e-06,
+      "loss": 0.5977,
+      "step": 4310
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4789769096013554,
+      "learning_rate": 5.987880182044304e-06,
+      "loss": 0.5662,
+      "step": 4311
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0922816492045966,
+      "learning_rate": 5.98620948120064e-06,
+      "loss": 0.7099,
+      "step": 4312
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.144043338021814,
+      "learning_rate": 5.984538665782007e-06,
+      "loss": 0.6219,
+      "step": 4313
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9684756423653305,
+      "learning_rate": 5.9828677359825196e-06,
+      "loss": 0.6493,
+      "step": 4314
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3925762779344515,
+      "learning_rate": 5.981196691996298e-06,
+      "loss": 0.5853,
+      "step": 4315
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.9740384297980214,
+      "learning_rate": 5.9795255340174825e-06,
+      "loss": 0.656,
+      "step": 4316
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1922968760567727,
+      "learning_rate": 5.9778542622402205e-06,
+      "loss": 0.5823,
+      "step": 4317
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.08742766373127,
+      "learning_rate": 5.976182876858679e-06,
+      "loss": 0.6744,
+      "step": 4318
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4973148138070638,
+      "learning_rate": 5.9745113780670305e-06,
+      "loss": 0.6447,
+      "step": 4319
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6221490595037755,
+      "learning_rate": 5.972839766059469e-06,
+      "loss": 0.6632,
+      "step": 4320
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.182690583898645,
+      "learning_rate": 5.971168041030194e-06,
+      "loss": 0.7078,
+      "step": 4321
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.984007580074134,
+      "learning_rate": 5.969496203173424e-06,
+      "loss": 0.5901,
+      "step": 4322
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.2119980194558324,
+      "learning_rate": 5.967824252683389e-06,
+      "loss": 0.6961,
+      "step": 4323
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9424736990101,
+      "learning_rate": 5.9661521897543276e-06,
+      "loss": 0.5905,
+      "step": 4324
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3950357257174666,
+      "learning_rate": 5.9644800145805e-06,
+      "loss": 0.6498,
+      "step": 4325
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9016254383942606,
+      "learning_rate": 5.962807727356169e-06,
+      "loss": 0.7386,
+      "step": 4326
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.187079791651539,
+      "learning_rate": 5.9611353282756235e-06,
+      "loss": 0.6256,
+      "step": 4327
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.203908956516435,
+      "learning_rate": 5.95946281753315e-06,
+      "loss": 0.5533,
+      "step": 4328
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3206237982379285,
+      "learning_rate": 5.957790195323064e-06,
+      "loss": 0.5879,
+      "step": 4329
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0758013751273796,
+      "learning_rate": 5.956117461839679e-06,
+      "loss": 0.7067,
+      "step": 4330
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9307153953571254,
+      "learning_rate": 5.954444617277332e-06,
+      "loss": 0.6043,
+      "step": 4331
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.1653578493636227,
+      "learning_rate": 5.952771661830368e-06,
+      "loss": 0.6076,
+      "step": 4332
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.835128332613066,
+      "learning_rate": 5.951098595693146e-06,
+      "loss": 0.5411,
+      "step": 4333
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.314934556444207,
+      "learning_rate": 5.9494254190600395e-06,
+      "loss": 0.7049,
+      "step": 4334
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.742652891296002,
+      "learning_rate": 5.947752132125432e-06,
+      "loss": 0.6817,
+      "step": 4335
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3350065294888402,
+      "learning_rate": 5.946078735083723e-06,
+      "loss": 0.6664,
+      "step": 4336
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.878910527625717,
+      "learning_rate": 5.944405228129318e-06,
+      "loss": 0.6108,
+      "step": 4337
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8697645174477455,
+      "learning_rate": 5.942731611456647e-06,
+      "loss": 0.6308,
+      "step": 4338
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.8193425849051312,
+      "learning_rate": 5.941057885260141e-06,
+      "loss": 0.6536,
+      "step": 4339
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3962102257159876,
+      "learning_rate": 5.939384049734252e-06,
+      "loss": 0.6428,
+      "step": 4340
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.897199349105444,
+      "learning_rate": 5.937710105073436e-06,
+      "loss": 0.6204,
+      "step": 4341
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.592219177725684,
+      "learning_rate": 5.936036051472173e-06,
+      "loss": 0.6533,
+      "step": 4342
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.856840556341079,
+      "learning_rate": 5.934361889124946e-06,
+      "loss": 0.6482,
+      "step": 4343
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.8920368478735567,
+      "learning_rate": 5.9326876182262575e-06,
+      "loss": 0.6338,
+      "step": 4344
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3189541995763316,
+      "learning_rate": 5.931013238970616e-06,
+      "loss": 0.6674,
+      "step": 4345
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.1842357570851374,
+      "learning_rate": 5.929338751552549e-06,
+      "loss": 0.6539,
+      "step": 4346
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.412864761506081,
+      "learning_rate": 5.927664156166592e-06,
+      "loss": 0.6452,
+      "step": 4347
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.450085408288252,
+      "learning_rate": 5.925989453007294e-06,
+      "loss": 0.5489,
+      "step": 4348
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.413183425668527,
+      "learning_rate": 5.924314642269219e-06,
+      "loss": 0.5971,
+      "step": 4349
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.077588646518163,
+      "learning_rate": 5.922639724146939e-06,
+      "loss": 0.6276,
+      "step": 4350
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.440785017604606,
+      "learning_rate": 5.920964698835047e-06,
+      "loss": 0.6674,
+      "step": 4351
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.242550745644832,
+      "learning_rate": 5.919289566528135e-06,
+      "loss": 0.5893,
+      "step": 4352
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.571180994234153,
+      "learning_rate": 5.9176143274208185e-06,
+      "loss": 0.7091,
+      "step": 4353
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4118573325477386,
+      "learning_rate": 5.915938981707724e-06,
+      "loss": 0.6622,
+      "step": 4354
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5558862540573424,
+      "learning_rate": 5.914263529583485e-06,
+      "loss": 0.7812,
+      "step": 4355
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2737963351658887,
+      "learning_rate": 5.9125879712427525e-06,
+      "loss": 0.5881,
+      "step": 4356
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4431483934766702,
+      "learning_rate": 5.9109123068801875e-06,
+      "loss": 0.7217,
+      "step": 4357
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3370688862721347,
+      "learning_rate": 5.909236536690464e-06,
+      "loss": 0.6663,
+      "step": 4358
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3073085136490574,
+      "learning_rate": 5.907560660868266e-06,
+      "loss": 0.6542,
+      "step": 4359
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.2027134863893343,
+      "learning_rate": 5.905884679608297e-06,
+      "loss": 0.6156,
+      "step": 4360
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5426901682583063,
+      "learning_rate": 5.904208593105263e-06,
+      "loss": 0.7358,
+      "step": 4361
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.4513894931625066,
+      "learning_rate": 5.902532401553888e-06,
+      "loss": 0.6191,
+      "step": 4362
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4589931992350498,
+      "learning_rate": 5.900856105148908e-06,
+      "loss": 0.6734,
+      "step": 4363
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.33735601801206,
+      "learning_rate": 5.899179704085072e-06,
+      "loss": 0.6383,
+      "step": 4364
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.153407332381854,
+      "learning_rate": 5.897503198557134e-06,
+      "loss": 0.6444,
+      "step": 4365
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8332048431773433,
+      "learning_rate": 5.89582658875987e-06,
+      "loss": 0.5908,
+      "step": 4366
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8707757219917687,
+      "learning_rate": 5.8941498748880635e-06,
+      "loss": 0.6854,
+      "step": 4367
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.80342904637755,
+      "learning_rate": 5.892473057136508e-06,
+      "loss": 0.5482,
+      "step": 4368
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0613535298316306,
+      "learning_rate": 5.890796135700013e-06,
+      "loss": 0.607,
+      "step": 4369
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.544050244229915,
+      "learning_rate": 5.889119110773398e-06,
+      "loss": 0.6613,
+      "step": 4370
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.225190268668775,
+      "learning_rate": 5.887441982551495e-06,
+      "loss": 0.684,
+      "step": 4371
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.406353835293022,
+      "learning_rate": 5.885764751229146e-06,
+      "loss": 0.5922,
+      "step": 4372
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.660963513951452,
+      "learning_rate": 5.884087417001212e-06,
+      "loss": 0.6724,
+      "step": 4373
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.151920515191103,
+      "learning_rate": 5.882409980062554e-06,
+      "loss": 0.6619,
+      "step": 4374
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.056944032442784,
+      "learning_rate": 5.880732440608059e-06,
+      "loss": 0.689,
+      "step": 4375
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8201326385968266,
+      "learning_rate": 5.879054798832612e-06,
+      "loss": 0.6328,
+      "step": 4376
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9966584556095535,
+      "learning_rate": 5.877377054931122e-06,
+      "loss": 0.6494,
+      "step": 4377
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3102019347667095,
+      "learning_rate": 5.8756992090985e-06,
+      "loss": 0.6443,
+      "step": 4378
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7187452579507503,
+      "learning_rate": 5.874021261529675e-06,
+      "loss": 0.778,
+      "step": 4379
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.9391200420774966,
+      "learning_rate": 5.872343212419589e-06,
+      "loss": 0.7402,
+      "step": 4380
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.342650913183981,
+      "learning_rate": 5.870665061963188e-06,
+      "loss": 0.6401,
+      "step": 4381
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.10923752534272,
+      "learning_rate": 5.868986810355437e-06,
+      "loss": 0.6497,
+      "step": 4382
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9439328630070154,
+      "learning_rate": 5.867308457791311e-06,
+      "loss": 0.6287,
+      "step": 4383
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1724951171650644,
+      "learning_rate": 5.865630004465796e-06,
+      "loss": 0.6773,
+      "step": 4384
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2545948900225095,
+      "learning_rate": 5.8639514505738885e-06,
+      "loss": 0.7132,
+      "step": 4385
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.339764871054603,
+      "learning_rate": 5.8622727963106e-06,
+      "loss": 0.6558,
+      "step": 4386
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.264396932081793,
+      "learning_rate": 5.860594041870948e-06,
+      "loss": 0.564,
+      "step": 4387
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.0726383532297126,
+      "learning_rate": 5.85891518744997e-06,
+      "loss": 0.6943,
+      "step": 4388
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.819599638111385,
+      "learning_rate": 5.857236233242709e-06,
+      "loss": 0.6919,
+      "step": 4389
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.55942614466353,
+      "learning_rate": 5.855557179444219e-06,
+      "loss": 0.6243,
+      "step": 4390
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1630742426007186,
+      "learning_rate": 5.8538780262495695e-06,
+      "loss": 0.59,
+      "step": 4391
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6158551871983122,
+      "learning_rate": 5.85219877385384e-06,
+      "loss": 0.7309,
+      "step": 4392
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.212409520666976,
+      "learning_rate": 5.8505194224521204e-06,
+      "loss": 0.5875,
+      "step": 4393
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3813643477613318,
+      "learning_rate": 5.848839972239512e-06,
+      "loss": 0.6635,
+      "step": 4394
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4275155873597285,
+      "learning_rate": 5.847160423411129e-06,
+      "loss": 0.6709,
+      "step": 4395
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.826967263615847,
+      "learning_rate": 5.845480776162097e-06,
+      "loss": 0.6407,
+      "step": 4396
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.306596082817607,
+      "learning_rate": 5.843801030687555e-06,
+      "loss": 0.6385,
+      "step": 4397
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9985283159296727,
+      "learning_rate": 5.842121187182644e-06,
+      "loss": 0.6064,
+      "step": 4398
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.2091727302195725,
+      "learning_rate": 5.840441245842532e-06,
+      "loss": 0.74,
+      "step": 4399
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8692803940553713,
+      "learning_rate": 5.838761206862382e-06,
+      "loss": 0.6428,
+      "step": 4400
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.125632738179534,
+      "learning_rate": 5.837081070437383e-06,
+      "loss": 0.7399,
+      "step": 4401
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.397361281528623,
+      "learning_rate": 5.835400836762723e-06,
+      "loss": 0.6768,
+      "step": 4402
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.962196122812087,
+      "learning_rate": 5.833720506033609e-06,
+      "loss": 0.5604,
+      "step": 4403
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4524501870949416,
+      "learning_rate": 5.8320400784452566e-06,
+      "loss": 0.6212,
+      "step": 4404
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.242962121068071,
+      "learning_rate": 5.830359554192894e-06,
+      "loss": 0.6391,
+      "step": 4405
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4880079804921844,
+      "learning_rate": 5.828678933471758e-06,
+      "loss": 0.6894,
+      "step": 4406
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9982078926059621,
+      "learning_rate": 5.826998216477097e-06,
+      "loss": 0.5715,
+      "step": 4407
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.39200422891883,
+      "learning_rate": 5.825317403404177e-06,
+      "loss": 0.7124,
+      "step": 4408
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5758571917643183,
+      "learning_rate": 5.823636494448265e-06,
+      "loss": 0.6916,
+      "step": 4409
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0155720379239783,
+      "learning_rate": 5.821955489804647e-06,
+      "loss": 0.5974,
+      "step": 4410
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4690764237144096,
+      "learning_rate": 5.820274389668614e-06,
+      "loss": 0.6291,
+      "step": 4411
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4569169908883968,
+      "learning_rate": 5.818593194235475e-06,
+      "loss": 0.675,
+      "step": 4412
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0672128706646213,
+      "learning_rate": 5.816911903700546e-06,
+      "loss": 0.6804,
+      "step": 4413
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9976719347095646,
+      "learning_rate": 5.815230518259153e-06,
+      "loss": 0.6844,
+      "step": 4414
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.564923614303906,
+      "learning_rate": 5.813549038106635e-06,
+      "loss": 0.6927,
+      "step": 4415
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3811759370101084,
+      "learning_rate": 5.811867463438341e-06,
+      "loss": 0.628,
+      "step": 4416
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0628079283250282,
+      "learning_rate": 5.810185794449633e-06,
+      "loss": 0.5905,
+      "step": 4417
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.582538159728672,
+      "learning_rate": 5.80850403133588e-06,
+      "loss": 0.7048,
+      "step": 4418
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4157594950729466,
+      "learning_rate": 5.806822174292467e-06,
+      "loss": 0.5557,
+      "step": 4419
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.990310949015315,
+      "learning_rate": 5.805140223514785e-06,
+      "loss": 0.5572,
+      "step": 4420
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0489038974046543,
+      "learning_rate": 5.80345817919824e-06,
+      "loss": 0.6249,
+      "step": 4421
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.060292665650145,
+      "learning_rate": 5.801776041538245e-06,
+      "loss": 0.6035,
+      "step": 4422
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.091882121485412,
+      "learning_rate": 5.800093810730229e-06,
+      "loss": 0.6245,
+      "step": 4423
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.179974949588236,
+      "learning_rate": 5.798411486969626e-06,
+      "loss": 0.6589,
+      "step": 4424
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.138657402068139,
+      "learning_rate": 5.796729070451884e-06,
+      "loss": 0.6545,
+      "step": 4425
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.233621588999477,
+      "learning_rate": 5.795046561372463e-06,
+      "loss": 0.681,
+      "step": 4426
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.004996496095365,
+      "learning_rate": 5.79336395992683e-06,
+      "loss": 0.6399,
+      "step": 4427
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.069729876929192,
+      "learning_rate": 5.791681266310465e-06,
+      "loss": 0.6049,
+      "step": 4428
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4465891459726623,
+      "learning_rate": 5.78999848071886e-06,
+      "loss": 0.6035,
+      "step": 4429
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2899552504847716,
+      "learning_rate": 5.788315603347515e-06,
+      "loss": 0.6646,
+      "step": 4430
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.750752115170631,
+      "learning_rate": 5.78663263439194e-06,
+      "loss": 0.6953,
+      "step": 4431
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2813221898068594,
+      "learning_rate": 5.7849495740476625e-06,
+      "loss": 0.7107,
+      "step": 4432
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.412245921640548,
+      "learning_rate": 5.783266422510211e-06,
+      "loss": 0.6862,
+      "step": 4433
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1713157933842573,
+      "learning_rate": 5.781583179975132e-06,
+      "loss": 0.7024,
+      "step": 4434
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5292807136122497,
+      "learning_rate": 5.779899846637976e-06,
+      "loss": 0.6344,
+      "step": 4435
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.638992479723519,
+      "learning_rate": 5.778216422694312e-06,
+      "loss": 0.6147,
+      "step": 4436
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.253511934867257,
+      "learning_rate": 5.776532908339713e-06,
+      "loss": 0.6775,
+      "step": 4437
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5189103997047404,
+      "learning_rate": 5.774849303769767e-06,
+      "loss": 0.6079,
+      "step": 4438
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9551608372390805,
+      "learning_rate": 5.773165609180067e-06,
+      "loss": 0.6591,
+      "step": 4439
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0567111687412636,
+      "learning_rate": 5.771481824766222e-06,
+      "loss": 0.6149,
+      "step": 4440
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.063349581374693,
+      "learning_rate": 5.769797950723848e-06,
+      "loss": 0.6967,
+      "step": 4441
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1228306731836386,
+      "learning_rate": 5.7681139872485744e-06,
+      "loss": 0.6664,
+      "step": 4442
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.793644978792627,
+      "learning_rate": 5.766429934536037e-06,
+      "loss": 0.5862,
+      "step": 4443
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0595037162948158,
+      "learning_rate": 5.764745792781886e-06,
+      "loss": 0.5796,
+      "step": 4444
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6360229263832173,
+      "learning_rate": 5.763061562181781e-06,
+      "loss": 0.5961,
+      "step": 4445
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.317300668949924,
+      "learning_rate": 5.761377242931386e-06,
+      "loss": 0.6888,
+      "step": 4446
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3873429958437904,
+      "learning_rate": 5.759692835226387e-06,
+      "loss": 0.6889,
+      "step": 4447
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2136534312552785,
+      "learning_rate": 5.75800833926247e-06,
+      "loss": 0.6915,
+      "step": 4448
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0974828764619353,
+      "learning_rate": 5.756323755235334e-06,
+      "loss": 0.5588,
+      "step": 4449
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4333697468641398,
+      "learning_rate": 5.754639083340691e-06,
+      "loss": 0.7895,
+      "step": 4450
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.480667649075754,
+      "learning_rate": 5.752954323774261e-06,
+      "loss": 0.6759,
+      "step": 4451
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.26250369737515,
+      "learning_rate": 5.751269476731775e-06,
+      "loss": 0.5934,
+      "step": 4452
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1003678544450188,
+      "learning_rate": 5.749584542408971e-06,
+      "loss": 0.5815,
+      "step": 4453
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3078642309343578,
+      "learning_rate": 5.747899521001603e-06,
+      "loss": 0.6309,
+      "step": 4454
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9538174909326671,
+      "learning_rate": 5.74621441270543e-06,
+      "loss": 0.5846,
+      "step": 4455
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3481363612900386,
+      "learning_rate": 5.744529217716225e-06,
+      "loss": 0.6627,
+      "step": 4456
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4288667378870494,
+      "learning_rate": 5.742843936229765e-06,
+      "loss": 0.6473,
+      "step": 4457
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 12.117953706010812,
+      "learning_rate": 5.741158568441846e-06,
+      "loss": 0.6602,
+      "step": 4458
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9795673957075657,
+      "learning_rate": 5.739473114548266e-06,
+      "loss": 0.6127,
+      "step": 4459
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.184917637480841,
+      "learning_rate": 5.737787574744837e-06,
+      "loss": 0.7399,
+      "step": 4460
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0453232292377943,
+      "learning_rate": 5.736101949227382e-06,
+      "loss": 0.6086,
+      "step": 4461
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.146930741168219,
+      "learning_rate": 5.734416238191729e-06,
+      "loss": 0.6511,
+      "step": 4462
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0161096829790184,
+      "learning_rate": 5.73273044183372e-06,
+      "loss": 0.5465,
+      "step": 4463
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.4064385189492246,
+      "learning_rate": 5.7310445603492085e-06,
+      "loss": 0.6235,
+      "step": 4464
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1516478687948544,
+      "learning_rate": 5.729358593934051e-06,
+      "loss": 0.6553,
+      "step": 4465
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.013275127090678,
+      "learning_rate": 5.727672542784122e-06,
+      "loss": 0.631,
+      "step": 4466
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.21296220940741,
+      "learning_rate": 5.7259864070953e-06,
+      "loss": 0.6722,
+      "step": 4467
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0788929579185234,
+      "learning_rate": 5.724300187063474e-06,
+      "loss": 0.6972,
+      "step": 4468
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2057902914842415,
+      "learning_rate": 5.722613882884549e-06,
+      "loss": 0.6598,
+      "step": 4469
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3957705076731886,
+      "learning_rate": 5.720927494754429e-06,
+      "loss": 0.6991,
+      "step": 4470
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.12211879589646,
+      "learning_rate": 5.719241022869039e-06,
+      "loss": 0.6594,
+      "step": 4471
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.490904530002642,
+      "learning_rate": 5.7175544674243044e-06,
+      "loss": 0.5995,
+      "step": 4472
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.934704124433464,
+      "learning_rate": 5.715867828616167e-06,
+      "loss": 0.6718,
+      "step": 4473
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.518726432040352,
+      "learning_rate": 5.714181106640575e-06,
+      "loss": 0.7023,
+      "step": 4474
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2024550220706205,
+      "learning_rate": 5.712494301693486e-06,
+      "loss": 0.6435,
+      "step": 4475
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.867057001543077,
+      "learning_rate": 5.710807413970868e-06,
+      "loss": 0.6585,
+      "step": 4476
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3892252500285447,
+      "learning_rate": 5.709120443668701e-06,
+      "loss": 0.6105,
+      "step": 4477
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.751702391769238,
+      "learning_rate": 5.707433390982969e-06,
+      "loss": 0.6801,
+      "step": 4478
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2281246093897633,
+      "learning_rate": 5.705746256109671e-06,
+      "loss": 0.6254,
+      "step": 4479
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2065787831492654,
+      "learning_rate": 5.704059039244814e-06,
+      "loss": 0.6247,
+      "step": 4480
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.4211374534691665,
+      "learning_rate": 5.7023717405844114e-06,
+      "loss": 0.6601,
+      "step": 4481
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.97147154570612,
+      "learning_rate": 5.700684360324492e-06,
+      "loss": 0.6468,
+      "step": 4482
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.36936656327254,
+      "learning_rate": 5.6989968986610876e-06,
+      "loss": 0.6256,
+      "step": 4483
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6783669889814394,
+      "learning_rate": 5.697309355790246e-06,
+      "loss": 0.7214,
+      "step": 4484
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3024950794881716,
+      "learning_rate": 5.695621731908018e-06,
+      "loss": 0.6098,
+      "step": 4485
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2375276123528454,
+      "learning_rate": 5.693934027210468e-06,
+      "loss": 0.6077,
+      "step": 4486
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3754390849560383,
+      "learning_rate": 5.692246241893669e-06,
+      "loss": 0.6369,
+      "step": 4487
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1211748838319964,
+      "learning_rate": 5.6905583761537034e-06,
+      "loss": 0.6388,
+      "step": 4488
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.868019959818687,
+      "learning_rate": 5.68887043018666e-06,
+      "loss": 0.6607,
+      "step": 4489
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.33570658417535,
+      "learning_rate": 5.687182404188642e-06,
+      "loss": 0.6159,
+      "step": 4490
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8823599155269486,
+      "learning_rate": 5.68549429835576e-06,
+      "loss": 0.535,
+      "step": 4491
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.743721547670784,
+      "learning_rate": 5.6838061128841294e-06,
+      "loss": 0.6538,
+      "step": 4492
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.8002300637615645,
+      "learning_rate": 5.682117847969884e-06,
+      "loss": 0.6659,
+      "step": 4493
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0584871886426965,
+      "learning_rate": 5.680429503809157e-06,
+      "loss": 0.5985,
+      "step": 4494
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.525503447996165,
+      "learning_rate": 5.678741080598098e-06,
+      "loss": 0.5908,
+      "step": 4495
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.687907042607034,
+      "learning_rate": 5.6770525785328625e-06,
+      "loss": 0.6845,
+      "step": 4496
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.304902889828247,
+      "learning_rate": 5.675363997809616e-06,
+      "loss": 0.6827,
+      "step": 4497
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4406252620899282,
+      "learning_rate": 5.6736753386245315e-06,
+      "loss": 0.5224,
+      "step": 4498
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5325199094461035,
+      "learning_rate": 5.6719866011737934e-06,
+      "loss": 0.6689,
+      "step": 4499
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.375682344749398,
+      "learning_rate": 5.670297785653596e-06,
+      "loss": 0.6651,
+      "step": 4500
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.425658142737432,
+      "learning_rate": 5.668608892260138e-06,
+      "loss": 0.7422,
+      "step": 4501
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0617808401615054,
+      "learning_rate": 5.666919921189632e-06,
+      "loss": 0.6425,
+      "step": 4502
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.272673910971418,
+      "learning_rate": 5.665230872638297e-06,
+      "loss": 0.648,
+      "step": 4503
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.572982484112035,
+      "learning_rate": 5.6635417468023635e-06,
+      "loss": 0.6951,
+      "step": 4504
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0084581361854075,
+      "learning_rate": 5.661852543878067e-06,
+      "loss": 0.6481,
+      "step": 4505
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0157127811635567,
+      "learning_rate": 5.660163264061656e-06,
+      "loss": 0.585,
+      "step": 4506
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5432077638654644,
+      "learning_rate": 5.6584739075493835e-06,
+      "loss": 0.6619,
+      "step": 4507
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5133640279691036,
+      "learning_rate": 5.656784474537518e-06,
+      "loss": 0.612,
+      "step": 4508
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9315990055007775,
+      "learning_rate": 5.65509496522233e-06,
+      "loss": 0.6136,
+      "step": 4509
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2810358608137027,
+      "learning_rate": 5.653405379800102e-06,
+      "loss": 0.6993,
+      "step": 4510
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.493229989865747,
+      "learning_rate": 5.651715718467127e-06,
+      "loss": 0.657,
+      "step": 4511
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6802549573204453,
+      "learning_rate": 5.6500259814197025e-06,
+      "loss": 0.6624,
+      "step": 4512
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.625253414667082,
+      "learning_rate": 5.648336168854139e-06,
+      "loss": 0.6646,
+      "step": 4513
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3868204390514567,
+      "learning_rate": 5.646646280966755e-06,
+      "loss": 0.6926,
+      "step": 4514
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4290428676146343,
+      "learning_rate": 5.6449563179538734e-06,
+      "loss": 0.5732,
+      "step": 4515
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6813004334794983,
+      "learning_rate": 5.64326628001183e-06,
+      "loss": 0.6179,
+      "step": 4516
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2908656332176767,
+      "learning_rate": 5.641576167336972e-06,
+      "loss": 0.7049,
+      "step": 4517
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5593906258069583,
+      "learning_rate": 5.639885980125649e-06,
+      "loss": 0.5926,
+      "step": 4518
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7153221598085593,
+      "learning_rate": 5.638195718574222e-06,
+      "loss": 0.7009,
+      "step": 4519
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.570454238663235,
+      "learning_rate": 5.636505382879061e-06,
+      "loss": 0.6327,
+      "step": 4520
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8531174704597646,
+      "learning_rate": 5.6348149732365465e-06,
+      "loss": 0.6287,
+      "step": 4521
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.518867643695685,
+      "learning_rate": 5.633124489843063e-06,
+      "loss": 0.6584,
+      "step": 4522
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.940669081934954,
+      "learning_rate": 5.631433932895005e-06,
+      "loss": 0.7226,
+      "step": 4523
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.262500330185508,
+      "learning_rate": 5.62974330258878e-06,
+      "loss": 0.5943,
+      "step": 4524
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.668293506438283,
+      "learning_rate": 5.6280525991207954e-06,
+      "loss": 0.6074,
+      "step": 4525
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5241175283242465,
+      "learning_rate": 5.626361822687478e-06,
+      "loss": 0.6424,
+      "step": 4526
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3985700650396047,
+      "learning_rate": 5.6246709734852535e-06,
+      "loss": 0.6714,
+      "step": 4527
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.762272787782166,
+      "learning_rate": 5.6229800517105615e-06,
+      "loss": 0.7199,
+      "step": 4528
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.74146694769342,
+      "learning_rate": 5.621289057559847e-06,
+      "loss": 0.6983,
+      "step": 4529
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5748292058521938,
+      "learning_rate": 5.619597991229566e-06,
+      "loss": 0.6199,
+      "step": 4530
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9437147149445138,
+      "learning_rate": 5.617906852916183e-06,
+      "loss": 0.6011,
+      "step": 4531
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 23.92235581251259,
+      "learning_rate": 5.6162156428161665e-06,
+      "loss": 0.6263,
+      "step": 4532
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.087366386179771,
+      "learning_rate": 5.614524361125998e-06,
+      "loss": 0.5655,
+      "step": 4533
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.503816305177784,
+      "learning_rate": 5.612833008042166e-06,
+      "loss": 0.6989,
+      "step": 4534
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0399171517394388,
+      "learning_rate": 5.611141583761167e-06,
+      "loss": 0.5931,
+      "step": 4535
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4370559310970306,
+      "learning_rate": 5.609450088479506e-06,
+      "loss": 0.6178,
+      "step": 4536
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.033431572385822,
+      "learning_rate": 5.607758522393693e-06,
+      "loss": 0.5607,
+      "step": 4537
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.41604755439263,
+      "learning_rate": 5.6060668857002545e-06,
+      "loss": 0.6314,
+      "step": 4538
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3686567490270187,
+      "learning_rate": 5.604375178595715e-06,
+      "loss": 0.6557,
+      "step": 4539
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1562680126957847,
+      "learning_rate": 5.6026834012766155e-06,
+      "loss": 0.6312,
+      "step": 4540
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.498322304260171,
+      "learning_rate": 5.600991553939501e-06,
+      "loss": 0.7011,
+      "step": 4541
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7681759119595353,
+      "learning_rate": 5.5992996367809236e-06,
+      "loss": 0.6749,
+      "step": 4542
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.1869566753671816,
+      "learning_rate": 5.597607649997449e-06,
+      "loss": 0.6855,
+      "step": 4543
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3926645915251172,
+      "learning_rate": 5.595915593785644e-06,
+      "loss": 0.676,
+      "step": 4544
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7435380969095537,
+      "learning_rate": 5.594223468342087e-06,
+      "loss": 0.5655,
+      "step": 4545
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0698113787263352,
+      "learning_rate": 5.592531273863367e-06,
+      "loss": 0.5551,
+      "step": 4546
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.66821684730331,
+      "learning_rate": 5.590839010546074e-06,
+      "loss": 0.6269,
+      "step": 4547
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6735786158961927,
+      "learning_rate": 5.589146678586814e-06,
+      "loss": 0.5896,
+      "step": 4548
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.556035989864172,
+      "learning_rate": 5.587454278182196e-06,
+      "loss": 0.7271,
+      "step": 4549
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5030779108925056,
+      "learning_rate": 5.585761809528839e-06,
+      "loss": 0.5887,
+      "step": 4550
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6285299123478363,
+      "learning_rate": 5.584069272823367e-06,
+      "loss": 0.6228,
+      "step": 4551
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.742738953366397,
+      "learning_rate": 5.582376668262415e-06,
+      "loss": 0.5916,
+      "step": 4552
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.051227649046525,
+      "learning_rate": 5.580683996042625e-06,
+      "loss": 0.5793,
+      "step": 4553
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.683953068169987,
+      "learning_rate": 5.578991256360649e-06,
+      "loss": 0.6996,
+      "step": 4554
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4168596403111824,
+      "learning_rate": 5.577298449413141e-06,
+      "loss": 0.59,
+      "step": 4555
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5891727673576144,
+      "learning_rate": 5.575605575396767e-06,
+      "loss": 0.635,
+      "step": 4556
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.164368406974795,
+      "learning_rate": 5.573912634508203e-06,
+      "loss": 0.6719,
+      "step": 4557
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1467160524829803,
+      "learning_rate": 5.572219626944128e-06,
+      "loss": 0.6302,
+      "step": 4558
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9545578011130686,
+      "learning_rate": 5.5705265529012295e-06,
+      "loss": 0.6129,
+      "step": 4559
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.397337716611551,
+      "learning_rate": 5.5688334125762065e-06,
+      "loss": 0.577,
+      "step": 4560
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5141297243805707,
+      "learning_rate": 5.567140206165762e-06,
+      "loss": 0.6222,
+      "step": 4561
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5018217036974577,
+      "learning_rate": 5.565446933866607e-06,
+      "loss": 0.6108,
+      "step": 4562
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.280655074209046,
+      "learning_rate": 5.563753595875463e-06,
+      "loss": 0.6282,
+      "step": 4563
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.359831147248012,
+      "learning_rate": 5.562060192389054e-06,
+      "loss": 0.6622,
+      "step": 4564
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2107889624685333,
+      "learning_rate": 5.560366723604117e-06,
+      "loss": 0.6035,
+      "step": 4565
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5994180305445815,
+      "learning_rate": 5.558673189717395e-06,
+      "loss": 0.6777,
+      "step": 4566
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7545314951405593,
+      "learning_rate": 5.556979590925636e-06,
+      "loss": 0.5888,
+      "step": 4567
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.364332000424429,
+      "learning_rate": 5.555285927425599e-06,
+      "loss": 0.6485,
+      "step": 4568
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5857119479401773,
+      "learning_rate": 5.553592199414047e-06,
+      "loss": 0.6616,
+      "step": 4569
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.15710865319317,
+      "learning_rate": 5.551898407087754e-06,
+      "loss": 0.6485,
+      "step": 4570
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8154040551745236,
+      "learning_rate": 5.550204550643501e-06,
+      "loss": 0.6448,
+      "step": 4571
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8762614479514133,
+      "learning_rate": 5.548510630278073e-06,
+      "loss": 0.5749,
+      "step": 4572
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.430020443992809,
+      "learning_rate": 5.5468166461882645e-06,
+      "loss": 0.694,
+      "step": 4573
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.010935891713146,
+      "learning_rate": 5.545122598570879e-06,
+      "loss": 0.5862,
+      "step": 4574
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.755230997400163,
+      "learning_rate": 5.543428487622727e-06,
+      "loss": 0.6561,
+      "step": 4575
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.059923185258385,
+      "learning_rate": 5.5417343135406206e-06,
+      "loss": 0.6963,
+      "step": 4576
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.790230956470769,
+      "learning_rate": 5.54004007652139e-06,
+      "loss": 0.5903,
+      "step": 4577
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.829573606090716,
+      "learning_rate": 5.5383457767618655e-06,
+      "loss": 0.6433,
+      "step": 4578
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.3802992751931424,
+      "learning_rate": 5.5366514144588835e-06,
+      "loss": 0.6504,
+      "step": 4579
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.87227842810873,
+      "learning_rate": 5.534956989809293e-06,
+      "loss": 0.6567,
+      "step": 4580
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.680652735847989,
+      "learning_rate": 5.533262503009944e-06,
+      "loss": 0.6889,
+      "step": 4581
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.381512231517992,
+      "learning_rate": 5.5315679542577e-06,
+      "loss": 0.6544,
+      "step": 4582
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.546877424221789,
+      "learning_rate": 5.529873343749428e-06,
+      "loss": 0.6369,
+      "step": 4583
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.319093075019239,
+      "learning_rate": 5.528178671682002e-06,
+      "loss": 0.5969,
+      "step": 4584
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.453302660315127,
+      "learning_rate": 5.5264839382523035e-06,
+      "loss": 0.5925,
+      "step": 4585
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3352208256933125,
+      "learning_rate": 5.524789143657226e-06,
+      "loss": 0.6313,
+      "step": 4586
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.6803200585589213,
+      "learning_rate": 5.523094288093659e-06,
+      "loss": 0.6757,
+      "step": 4587
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.4253926993814092,
+      "learning_rate": 5.521399371758511e-06,
+      "loss": 0.6969,
+      "step": 4588
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.223973921567959,
+      "learning_rate": 5.519704394848693e-06,
+      "loss": 0.6356,
+      "step": 4589
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.6909496271121074,
+      "learning_rate": 5.518009357561119e-06,
+      "loss": 0.6763,
+      "step": 4590
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.315901845822254,
+      "learning_rate": 5.516314260092717e-06,
+      "loss": 0.6356,
+      "step": 4591
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.885139970332389,
+      "learning_rate": 5.514619102640415e-06,
+      "loss": 0.6069,
+      "step": 4592
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.566257463657484,
+      "learning_rate": 5.512923885401154e-06,
+      "loss": 0.6844,
+      "step": 4593
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1239709742116073,
+      "learning_rate": 5.511228608571879e-06,
+      "loss": 0.5754,
+      "step": 4594
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.1944023152638357,
+      "learning_rate": 5.5095332723495425e-06,
+      "loss": 0.6047,
+      "step": 4595
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.0417150262269077,
+      "learning_rate": 5.507837876931102e-06,
+      "loss": 0.6709,
+      "step": 4596
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.1835449500447512,
+      "learning_rate": 5.506142422513525e-06,
+      "loss": 0.6766,
+      "step": 4597
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4531359904321675,
+      "learning_rate": 5.504446909293786e-06,
+      "loss": 0.6872,
+      "step": 4598
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9575039883577126,
+      "learning_rate": 5.502751337468862e-06,
+      "loss": 0.6198,
+      "step": 4599
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.254352953269388,
+      "learning_rate": 5.5010557072357395e-06,
+      "loss": 0.7175,
+      "step": 4600
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5317638836045275,
+      "learning_rate": 5.499360018791416e-06,
+      "loss": 0.626,
+      "step": 4601
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.080250323596298,
+      "learning_rate": 5.497664272332888e-06,
+      "loss": 0.6564,
+      "step": 4602
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.791033529732766,
+      "learning_rate": 5.495968468057164e-06,
+      "loss": 0.6691,
+      "step": 4603
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.576002951182326,
+      "learning_rate": 5.4942726061612564e-06,
+      "loss": 0.7081,
+      "step": 4604
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.002512038489308,
+      "learning_rate": 5.492576686842186e-06,
+      "loss": 0.5352,
+      "step": 4605
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.3897993264643347,
+      "learning_rate": 5.49088071029698e-06,
+      "loss": 0.6308,
+      "step": 4606
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.409849318201643,
+      "learning_rate": 5.489184676722673e-06,
+      "loss": 0.6008,
+      "step": 4607
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5209078919290677,
+      "learning_rate": 5.487488586316304e-06,
+      "loss": 0.609,
+      "step": 4608
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3412970675636466,
+      "learning_rate": 5.485792439274919e-06,
+      "loss": 0.644,
+      "step": 4609
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9448249147025708,
+      "learning_rate": 5.484096235795574e-06,
+      "loss": 0.5841,
+      "step": 4610
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.4569880167790625,
+      "learning_rate": 5.482399976075327e-06,
+      "loss": 0.6169,
+      "step": 4611
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.038821439830503,
+      "learning_rate": 5.4807036603112465e-06,
+      "loss": 0.6924,
+      "step": 4612
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1488134768098157,
+      "learning_rate": 5.479007288700403e-06,
+      "loss": 0.6386,
+      "step": 4613
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0696890955921132,
+      "learning_rate": 5.477310861439877e-06,
+      "loss": 0.6493,
+      "step": 4614
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5185747574560793,
+      "learning_rate": 5.475614378726757e-06,
+      "loss": 0.6438,
+      "step": 4615
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0897327507354406,
+      "learning_rate": 5.4739178407581315e-06,
+      "loss": 0.5929,
+      "step": 4616
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6311884639260477,
+      "learning_rate": 5.4722212477311025e-06,
+      "loss": 0.6394,
+      "step": 4617
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.269733417286331,
+      "learning_rate": 5.470524599842773e-06,
+      "loss": 0.5718,
+      "step": 4618
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.021071958636412,
+      "learning_rate": 5.468827897290256e-06,
+      "loss": 0.5784,
+      "step": 4619
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.5006735943403235,
+      "learning_rate": 5.46713114027067e-06,
+      "loss": 0.6056,
+      "step": 4620
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0266506883030095,
+      "learning_rate": 5.465434328981136e-06,
+      "loss": 0.5808,
+      "step": 4621
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.8304477124966176,
+      "learning_rate": 5.463737463618788e-06,
+      "loss": 0.7006,
+      "step": 4622
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.515291735666162,
+      "learning_rate": 5.462040544380764e-06,
+      "loss": 0.7003,
+      "step": 4623
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.387353672942185,
+      "learning_rate": 5.460343571464203e-06,
+      "loss": 0.6504,
+      "step": 4624
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.141288708078109,
+      "learning_rate": 5.458646545066258e-06,
+      "loss": 0.5682,
+      "step": 4625
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5542956306423674,
+      "learning_rate": 5.456949465384082e-06,
+      "loss": 0.6874,
+      "step": 4626
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.9922167702054705,
+      "learning_rate": 5.455252332614839e-06,
+      "loss": 0.6838,
+      "step": 4627
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.914946760609991,
+      "learning_rate": 5.453555146955696e-06,
+      "loss": 0.6331,
+      "step": 4628
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.352573285600877,
+      "learning_rate": 5.451857908603826e-06,
+      "loss": 0.635,
+      "step": 4629
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6733703812603755,
+      "learning_rate": 5.450160617756411e-06,
+      "loss": 0.6508,
+      "step": 4630
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.349826396279672,
+      "learning_rate": 5.448463274610637e-06,
+      "loss": 0.6848,
+      "step": 4631
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.81980906330448,
+      "learning_rate": 5.446765879363697e-06,
+      "loss": 0.6457,
+      "step": 4632
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.651707782689453,
+      "learning_rate": 5.445068432212787e-06,
+      "loss": 0.6972,
+      "step": 4633
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3655792841646512,
+      "learning_rate": 5.443370933355114e-06,
+      "loss": 0.5924,
+      "step": 4634
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7313353670254203,
+      "learning_rate": 5.441673382987886e-06,
+      "loss": 0.616,
+      "step": 4635
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4130469073654335,
+      "learning_rate": 5.439975781308322e-06,
+      "loss": 0.6315,
+      "step": 4636
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.597140459042695,
+      "learning_rate": 5.4382781285136445e-06,
+      "loss": 0.6355,
+      "step": 4637
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2052663360438993,
+      "learning_rate": 5.436580424801081e-06,
+      "loss": 0.6403,
+      "step": 4638
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0955992547403572,
+      "learning_rate": 5.434882670367865e-06,
+      "loss": 0.5755,
+      "step": 4639
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7478098709325316,
+      "learning_rate": 5.4331848654112374e-06,
+      "loss": 0.713,
+      "step": 4640
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.099989483770273,
+      "learning_rate": 5.431487010128445e-06,
+      "loss": 0.6117,
+      "step": 4641
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7349941290390514,
+      "learning_rate": 5.4297891047167385e-06,
+      "loss": 0.6223,
+      "step": 4642
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0659832601686883,
+      "learning_rate": 5.428091149373377e-06,
+      "loss": 0.5199,
+      "step": 4643
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.61686418980415,
+      "learning_rate": 5.426393144295623e-06,
+      "loss": 0.6323,
+      "step": 4644
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.013180159122992,
+      "learning_rate": 5.4246950896807445e-06,
+      "loss": 0.6634,
+      "step": 4645
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.7524905717817365,
+      "learning_rate": 5.422996985726019e-06,
+      "loss": 0.5842,
+      "step": 4646
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2515177137365296,
+      "learning_rate": 5.421298832628729e-06,
+      "loss": 0.6197,
+      "step": 4647
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.427073007294642,
+      "learning_rate": 5.419600630586155e-06,
+      "loss": 0.6538,
+      "step": 4648
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.743461241178892,
+      "learning_rate": 5.417902379795593e-06,
+      "loss": 0.6647,
+      "step": 4649
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3931710581662897,
+      "learning_rate": 5.416204080454343e-06,
+      "loss": 0.6355,
+      "step": 4650
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.8194753476800356,
+      "learning_rate": 5.414505732759704e-06,
+      "loss": 0.6645,
+      "step": 4651
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6326764418481194,
+      "learning_rate": 5.412807336908987e-06,
+      "loss": 0.6086,
+      "step": 4652
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.284399679999572,
+      "learning_rate": 5.411108893099508e-06,
+      "loss": 0.5637,
+      "step": 4653
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.7384502492836775,
+      "learning_rate": 5.409410401528586e-06,
+      "loss": 0.622,
+      "step": 4654
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.61822403374306,
+      "learning_rate": 5.4077118623935476e-06,
+      "loss": 0.6361,
+      "step": 4655
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.2720764272007425,
+      "learning_rate": 5.406013275891723e-06,
+      "loss": 0.7198,
+      "step": 4656
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.533871073074416,
+      "learning_rate": 5.404314642220448e-06,
+      "loss": 0.6521,
+      "step": 4657
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5014645838243834,
+      "learning_rate": 5.40261596157707e-06,
+      "loss": 0.623,
+      "step": 4658
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.187282413928454,
+      "learning_rate": 5.40091723415893e-06,
+      "loss": 0.6687,
+      "step": 4659
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.210081592011959,
+      "learning_rate": 5.399218460163387e-06,
+      "loss": 0.5783,
+      "step": 4660
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.9686826745928863,
+      "learning_rate": 5.397519639787796e-06,
+      "loss": 0.6148,
+      "step": 4661
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8313621442070436,
+      "learning_rate": 5.395820773229523e-06,
+      "loss": 0.6224,
+      "step": 4662
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.837090983146463,
+      "learning_rate": 5.394121860685937e-06,
+      "loss": 0.5873,
+      "step": 4663
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7825956148863606,
+      "learning_rate": 5.392422902354413e-06,
+      "loss": 0.6691,
+      "step": 4664
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.3329282867648136,
+      "learning_rate": 5.39072389843233e-06,
+      "loss": 0.6057,
+      "step": 4665
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.5608187127868387,
+      "learning_rate": 5.389024849117074e-06,
+      "loss": 0.705,
+      "step": 4666
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.706689822341293,
+      "learning_rate": 5.387325754606035e-06,
+      "loss": 0.7137,
+      "step": 4667
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5224351370813674,
+      "learning_rate": 5.3856266150966094e-06,
+      "loss": 0.6446,
+      "step": 4668
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.120747037478334,
+      "learning_rate": 5.3839274307862e-06,
+      "loss": 0.6706,
+      "step": 4669
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.6521412088636325,
+      "learning_rate": 5.3822282018722085e-06,
+      "loss": 0.7024,
+      "step": 4670
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.364628947418871,
+      "learning_rate": 5.380528928552052e-06,
+      "loss": 0.7261,
+      "step": 4671
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7339482051464534,
+      "learning_rate": 5.378829611023144e-06,
+      "loss": 0.6781,
+      "step": 4672
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.478305759343481,
+      "learning_rate": 5.377130249482907e-06,
+      "loss": 0.6225,
+      "step": 4673
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.411215799916555,
+      "learning_rate": 5.3754308441287675e-06,
+      "loss": 0.6103,
+      "step": 4674
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.42557822909133,
+      "learning_rate": 5.3737313951581575e-06,
+      "loss": 0.5748,
+      "step": 4675
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6013955730359046,
+      "learning_rate": 5.372031902768514e-06,
+      "loss": 0.5715,
+      "step": 4676
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1819025201627946,
+      "learning_rate": 5.370332367157281e-06,
+      "loss": 0.7339,
+      "step": 4677
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.588545731682473,
+      "learning_rate": 5.368632788521903e-06,
+      "loss": 0.6315,
+      "step": 4678
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.008669398229263,
+      "learning_rate": 5.3669331670598335e-06,
+      "loss": 0.7626,
+      "step": 4679
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.43609159650542,
+      "learning_rate": 5.36523350296853e-06,
+      "loss": 0.7089,
+      "step": 4680
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.121151492979661,
+      "learning_rate": 5.363533796445452e-06,
+      "loss": 0.6616,
+      "step": 4681
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7034182260790494,
+      "learning_rate": 5.361834047688071e-06,
+      "loss": 0.6528,
+      "step": 4682
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.478076643546118,
+      "learning_rate": 5.360134256893854e-06,
+      "loss": 0.6114,
+      "step": 4683
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6079287424630264,
+      "learning_rate": 5.35843442426028e-06,
+      "loss": 0.6421,
+      "step": 4684
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.685462278835273,
+      "learning_rate": 5.356734549984832e-06,
+      "loss": 0.5111,
+      "step": 4685
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.979817349950144,
+      "learning_rate": 5.355034634264996e-06,
+      "loss": 0.6508,
+      "step": 4686
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0200409785087854,
+      "learning_rate": 5.353334677298261e-06,
+      "loss": 0.6656,
+      "step": 4687
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.646554726463927,
+      "learning_rate": 5.351634679282125e-06,
+      "loss": 0.5935,
+      "step": 4688
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6235037363889915,
+      "learning_rate": 5.349934640414089e-06,
+      "loss": 0.6812,
+      "step": 4689
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.07421728360287,
+      "learning_rate": 5.348234560891657e-06,
+      "loss": 0.6569,
+      "step": 4690
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 36.97843704858725,
+      "learning_rate": 5.346534440912341e-06,
+      "loss": 0.6354,
+      "step": 4691
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4594375427225232,
+      "learning_rate": 5.3448342806736545e-06,
+      "loss": 0.6174,
+      "step": 4692
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8018238508168425,
+      "learning_rate": 5.34313408037312e-06,
+      "loss": 0.5667,
+      "step": 4693
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.009916868253691,
+      "learning_rate": 5.341433840208258e-06,
+      "loss": 0.7038,
+      "step": 4694
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.258548590941279,
+      "learning_rate": 5.339733560376601e-06,
+      "loss": 0.719,
+      "step": 4695
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.61623774709776,
+      "learning_rate": 5.33803324107568e-06,
+      "loss": 0.7362,
+      "step": 4696
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7165699079262935,
+      "learning_rate": 5.336332882503034e-06,
+      "loss": 0.6397,
+      "step": 4697
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.479128754751253,
+      "learning_rate": 5.334632484856206e-06,
+      "loss": 0.6062,
+      "step": 4698
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0822937031597597,
+      "learning_rate": 5.332932048332744e-06,
+      "loss": 0.6856,
+      "step": 4699
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.670085477402618,
+      "learning_rate": 5.331231573130199e-06,
+      "loss": 0.6451,
+      "step": 4700
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.710855683571757,
+      "learning_rate": 5.329531059446127e-06,
+      "loss": 0.5651,
+      "step": 4701
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.256008895457131,
+      "learning_rate": 5.327830507478089e-06,
+      "loss": 0.6252,
+      "step": 4702
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1096958021617995,
+      "learning_rate": 5.32612991742365e-06,
+      "loss": 0.5849,
+      "step": 4703
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.5793139887216263,
+      "learning_rate": 5.32442928948038e-06,
+      "loss": 0.6785,
+      "step": 4704
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7963244411549644,
+      "learning_rate": 5.322728623845853e-06,
+      "loss": 0.6408,
+      "step": 4705
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.182317971739211,
+      "learning_rate": 5.321027920717649e-06,
+      "loss": 0.6363,
+      "step": 4706
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.7507644271214393,
+      "learning_rate": 5.319327180293347e-06,
+      "loss": 0.635,
+      "step": 4707
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.075170533407163,
+      "learning_rate": 5.317626402770537e-06,
+      "loss": 0.6228,
+      "step": 4708
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.841933692244469,
+      "learning_rate": 5.3159255883468095e-06,
+      "loss": 0.6796,
+      "step": 4709
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.6806157165268347,
+      "learning_rate": 5.314224737219761e-06,
+      "loss": 0.6646,
+      "step": 4710
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.533486664828576,
+      "learning_rate": 5.31252384958699e-06,
+      "loss": 0.711,
+      "step": 4711
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7604461850517406,
+      "learning_rate": 5.310822925646103e-06,
+      "loss": 0.604,
+      "step": 4712
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.578358994492397,
+      "learning_rate": 5.309121965594706e-06,
+      "loss": 0.6351,
+      "step": 4713
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.579513564368866,
+      "learning_rate": 5.307420969630412e-06,
+      "loss": 0.7123,
+      "step": 4714
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.582532356341639,
+      "learning_rate": 5.30571993795084e-06,
+      "loss": 0.6549,
+      "step": 4715
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5183375541671986,
+      "learning_rate": 5.304018870753608e-06,
+      "loss": 0.6463,
+      "step": 4716
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0420447788319127,
+      "learning_rate": 5.3023177682363435e-06,
+      "loss": 0.5367,
+      "step": 4717
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.294424245337193,
+      "learning_rate": 5.300616630596673e-06,
+      "loss": 0.6515,
+      "step": 4718
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0335405308223342,
+      "learning_rate": 5.298915458032233e-06,
+      "loss": 0.5989,
+      "step": 4719
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0427016248786705,
+      "learning_rate": 5.297214250740658e-06,
+      "loss": 0.5918,
+      "step": 4720
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.004325778129872,
+      "learning_rate": 5.295513008919592e-06,
+      "loss": 0.6773,
+      "step": 4721
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3701521965706323,
+      "learning_rate": 5.293811732766677e-06,
+      "loss": 0.6755,
+      "step": 4722
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.111246958207434,
+      "learning_rate": 5.292110422479565e-06,
+      "loss": 0.6284,
+      "step": 4723
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.018425742396122,
+      "learning_rate": 5.290409078255909e-06,
+      "loss": 0.6031,
+      "step": 4724
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9869690779073568,
+      "learning_rate": 5.288707700293365e-06,
+      "loss": 0.5695,
+      "step": 4725
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1072157586028053,
+      "learning_rate": 5.287006288789596e-06,
+      "loss": 0.5593,
+      "step": 4726
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3663187157291676,
+      "learning_rate": 5.285304843942265e-06,
+      "loss": 0.702,
+      "step": 4727
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.046850596094636,
+      "learning_rate": 5.283603365949043e-06,
+      "loss": 0.7287,
+      "step": 4728
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9462619367998221,
+      "learning_rate": 5.2819018550076e-06,
+      "loss": 0.5409,
+      "step": 4729
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1605371729694633,
+      "learning_rate": 5.280200311315616e-06,
+      "loss": 0.6896,
+      "step": 4730
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9185781197431124,
+      "learning_rate": 5.278498735070769e-06,
+      "loss": 0.6377,
+      "step": 4731
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.5719498173680098,
+      "learning_rate": 5.2767971264707445e-06,
+      "loss": 0.6695,
+      "step": 4732
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8324365081609333,
+      "learning_rate": 5.27509548571323e-06,
+      "loss": 0.5886,
+      "step": 4733
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.705346452985015,
+      "learning_rate": 5.273393812995917e-06,
+      "loss": 0.6509,
+      "step": 4734
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.9814173861586815,
+      "learning_rate": 5.271692108516501e-06,
+      "loss": 0.7016,
+      "step": 4735
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.424011098877338,
+      "learning_rate": 5.269990372472682e-06,
+      "loss": 0.6616,
+      "step": 4736
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1493045527463837,
+      "learning_rate": 5.2682886050621604e-06,
+      "loss": 0.6617,
+      "step": 4737
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.716615071908531,
+      "learning_rate": 5.266586806482646e-06,
+      "loss": 0.6362,
+      "step": 4738
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.118626444685795,
+      "learning_rate": 5.264884976931845e-06,
+      "loss": 0.6424,
+      "step": 4739
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.541880011041433,
+      "learning_rate": 5.263183116607474e-06,
+      "loss": 0.5951,
+      "step": 4740
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1285558503464246,
+      "learning_rate": 5.261481225707251e-06,
+      "loss": 0.588,
+      "step": 4741
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.916993130135385,
+      "learning_rate": 5.259779304428893e-06,
+      "loss": 0.6042,
+      "step": 4742
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7254833052810423,
+      "learning_rate": 5.258077352970128e-06,
+      "loss": 0.6621,
+      "step": 4743
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.462602034967814,
+      "learning_rate": 5.256375371528681e-06,
+      "loss": 0.5998,
+      "step": 4744
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.520686792805083,
+      "learning_rate": 5.254673360302284e-06,
+      "loss": 0.6472,
+      "step": 4745
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.459934646842888,
+      "learning_rate": 5.252971319488672e-06,
+      "loss": 0.6504,
+      "step": 4746
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0506864779125666,
+      "learning_rate": 5.2512692492855845e-06,
+      "loss": 0.5606,
+      "step": 4747
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.191202095280542,
+      "learning_rate": 5.249567149890762e-06,
+      "loss": 0.6614,
+      "step": 4748
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.032018653356111,
+      "learning_rate": 5.247865021501949e-06,
+      "loss": 0.608,
+      "step": 4749
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.5143563476752124,
+      "learning_rate": 5.2461628643168935e-06,
+      "loss": 0.5829,
+      "step": 4750
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.066214123433812,
+      "learning_rate": 5.244460678533349e-06,
+      "loss": 0.6338,
+      "step": 4751
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.7234112282653116,
+      "learning_rate": 5.24275846434907e-06,
+      "loss": 0.6476,
+      "step": 4752
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5886795596777463,
+      "learning_rate": 5.2410562219618135e-06,
+      "loss": 0.652,
+      "step": 4753
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5480441407074443,
+      "learning_rate": 5.239353951569342e-06,
+      "loss": 0.5687,
+      "step": 4754
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4463031639209403,
+      "learning_rate": 5.2376516533694196e-06,
+      "loss": 0.5338,
+      "step": 4755
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.257125530689949,
+      "learning_rate": 5.235949327559817e-06,
+      "loss": 0.6191,
+      "step": 4756
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1011431880055143,
+      "learning_rate": 5.2342469743383026e-06,
+      "loss": 0.6559,
+      "step": 4757
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.487907737324848,
+      "learning_rate": 5.232544593902652e-06,
+      "loss": 0.6932,
+      "step": 4758
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.402578163767159,
+      "learning_rate": 5.230842186450642e-06,
+      "loss": 0.6343,
+      "step": 4759
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.077478548938603,
+      "learning_rate": 5.2291397521800545e-06,
+      "loss": 0.6685,
+      "step": 4760
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2841356416265044,
+      "learning_rate": 5.227437291288674e-06,
+      "loss": 0.6646,
+      "step": 4761
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7800487949493458,
+      "learning_rate": 5.225734803974285e-06,
+      "loss": 0.6371,
+      "step": 4762
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2976515988461337,
+      "learning_rate": 5.22403229043468e-06,
+      "loss": 0.6387,
+      "step": 4763
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6430943743263358,
+      "learning_rate": 5.222329750867649e-06,
+      "loss": 0.6613,
+      "step": 4764
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.659848728533461,
+      "learning_rate": 5.220627185470993e-06,
+      "loss": 0.614,
+      "step": 4765
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.5083748901897884,
+      "learning_rate": 5.218924594442507e-06,
+      "loss": 0.5716,
+      "step": 4766
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8298517158269325,
+      "learning_rate": 5.217221977979996e-06,
+      "loss": 0.6882,
+      "step": 4767
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.7161705013278534,
+      "learning_rate": 5.215519336281261e-06,
+      "loss": 0.7084,
+      "step": 4768
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4592185775923303,
+      "learning_rate": 5.213816669544114e-06,
+      "loss": 0.5649,
+      "step": 4769
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7485436454433247,
+      "learning_rate": 5.2121139779663645e-06,
+      "loss": 0.6664,
+      "step": 4770
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0981595313863677,
+      "learning_rate": 5.2104112617458254e-06,
+      "loss": 0.6766,
+      "step": 4771
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.0052501193362255,
+      "learning_rate": 5.2087085210803145e-06,
+      "loss": 0.7007,
+      "step": 4772
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.1047565880215737,
+      "learning_rate": 5.207005756167651e-06,
+      "loss": 0.6448,
+      "step": 4773
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.792375318373496,
+      "learning_rate": 5.205302967205657e-06,
+      "loss": 0.6724,
+      "step": 4774
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.51567986610389,
+      "learning_rate": 5.203600154392158e-06,
+      "loss": 0.606,
+      "step": 4775
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5125662323316287,
+      "learning_rate": 5.2018973179249824e-06,
+      "loss": 0.6156,
+      "step": 4776
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2159857718006224,
+      "learning_rate": 5.200194458001958e-06,
+      "loss": 0.5654,
+      "step": 4777
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9452532347483427,
+      "learning_rate": 5.198491574820923e-06,
+      "loss": 0.603,
+      "step": 4778
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.142355113115631,
+      "learning_rate": 5.196788668579708e-06,
+      "loss": 0.6565,
+      "step": 4779
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2887002234975764,
+      "learning_rate": 5.195085739476156e-06,
+      "loss": 0.6851,
+      "step": 4780
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.085541461688833,
+      "learning_rate": 5.193382787708106e-06,
+      "loss": 0.5886,
+      "step": 4781
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.500928456155343,
+      "learning_rate": 5.191679813473402e-06,
+      "loss": 0.5474,
+      "step": 4782
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.985875727924238,
+      "learning_rate": 5.189976816969892e-06,
+      "loss": 0.6326,
+      "step": 4783
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.1970563840062383,
+      "learning_rate": 5.188273798395425e-06,
+      "loss": 0.7014,
+      "step": 4784
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.241455526944972,
+      "learning_rate": 5.186570757947852e-06,
+      "loss": 0.731,
+      "step": 4785
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.576374543674953,
+      "learning_rate": 5.1848676958250265e-06,
+      "loss": 0.6363,
+      "step": 4786
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.335548183754556,
+      "learning_rate": 5.183164612224809e-06,
+      "loss": 0.583,
+      "step": 4787
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4154405135354136,
+      "learning_rate": 5.181461507345054e-06,
+      "loss": 0.5799,
+      "step": 4788
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.5423104288104916,
+      "learning_rate": 5.1797583813836285e-06,
+      "loss": 0.6326,
+      "step": 4789
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5045963993199747,
+      "learning_rate": 5.178055234538391e-06,
+      "loss": 0.6283,
+      "step": 4790
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0244708876329987,
+      "learning_rate": 5.176352067007213e-06,
+      "loss": 0.6263,
+      "step": 4791
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.255397053263638,
+      "learning_rate": 5.174648878987959e-06,
+      "loss": 0.6359,
+      "step": 4792
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.722706053854962,
+      "learning_rate": 5.1729456706785055e-06,
+      "loss": 0.6435,
+      "step": 4793
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.075448749546548,
+      "learning_rate": 5.1712424422767224e-06,
+      "loss": 0.6811,
+      "step": 4794
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.688883878063154,
+      "learning_rate": 5.169539193980489e-06,
+      "loss": 0.6266,
+      "step": 4795
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9204150200499615,
+      "learning_rate": 5.1678359259876824e-06,
+      "loss": 0.6795,
+      "step": 4796
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1559522338055792,
+      "learning_rate": 5.1661326384961805e-06,
+      "loss": 0.5607,
+      "step": 4797
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9541032407056105,
+      "learning_rate": 5.164429331703871e-06,
+      "loss": 0.6785,
+      "step": 4798
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8942781057490377,
+      "learning_rate": 5.162726005808636e-06,
+      "loss": 0.5714,
+      "step": 4799
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.079312812042359,
+      "learning_rate": 5.1610226610083655e-06,
+      "loss": 0.5694,
+      "step": 4800
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.277923548372197,
+      "learning_rate": 5.159319297500945e-06,
+      "loss": 0.6173,
+      "step": 4801
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.986097736979872,
+      "learning_rate": 5.157615915484273e-06,
+      "loss": 0.6071,
+      "step": 4802
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.2600607771581642,
+      "learning_rate": 5.155912515156236e-06,
+      "loss": 0.6338,
+      "step": 4803
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.444448387829765,
+      "learning_rate": 5.154209096714736e-06,
+      "loss": 0.635,
+      "step": 4804
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2143014150947318,
+      "learning_rate": 5.152505660357667e-06,
+      "loss": 0.6542,
+      "step": 4805
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.685218090583335,
+      "learning_rate": 5.150802206282932e-06,
+      "loss": 0.5947,
+      "step": 4806
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.593209127978214,
+      "learning_rate": 5.149098734688434e-06,
+      "loss": 0.6186,
+      "step": 4807
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.2086494702555886,
+      "learning_rate": 5.147395245772074e-06,
+      "loss": 0.737,
+      "step": 4808
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.188890598572003,
+      "learning_rate": 5.145691739731761e-06,
+      "loss": 0.6247,
+      "step": 4809
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.44137252928556,
+      "learning_rate": 5.143988216765402e-06,
+      "loss": 0.6595,
+      "step": 4810
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.172419643746679,
+      "learning_rate": 5.142284677070911e-06,
+      "loss": 0.6992,
+      "step": 4811
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.952102588392533,
+      "learning_rate": 5.140581120846194e-06,
+      "loss": 0.6163,
+      "step": 4812
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6735180292043665,
+      "learning_rate": 5.138877548289173e-06,
+      "loss": 0.6745,
+      "step": 4813
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.3271844953089964,
+      "learning_rate": 5.137173959597755e-06,
+      "loss": 0.7083,
+      "step": 4814
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.779885373526377,
+      "learning_rate": 5.135470354969867e-06,
+      "loss": 0.5758,
+      "step": 4815
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.880291332502046,
+      "learning_rate": 5.1337667346034226e-06,
+      "loss": 0.6426,
+      "step": 4816
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.789682189366147,
+      "learning_rate": 5.132063098696346e-06,
+      "loss": 0.6936,
+      "step": 4817
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5838279711205714,
+      "learning_rate": 5.130359447446561e-06,
+      "loss": 0.6659,
+      "step": 4818
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.4258511041007518,
+      "learning_rate": 5.128655781051991e-06,
+      "loss": 0.6169,
+      "step": 4819
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.6136439580060657,
+      "learning_rate": 5.126952099710566e-06,
+      "loss": 0.6806,
+      "step": 4820
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.209126945224101,
+      "learning_rate": 5.125248403620211e-06,
+      "loss": 0.6683,
+      "step": 4821
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.809317990932967,
+      "learning_rate": 5.12354469297886e-06,
+      "loss": 0.6826,
+      "step": 4822
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.2384938127477425,
+      "learning_rate": 5.121840967984443e-06,
+      "loss": 0.6434,
+      "step": 4823
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2368138450617656,
+      "learning_rate": 5.120137228834896e-06,
+      "loss": 0.6217,
+      "step": 4824
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.452259843593572,
+      "learning_rate": 5.1184334757281506e-06,
+      "loss": 0.6101,
+      "step": 4825
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.585727076856822,
+      "learning_rate": 5.1167297088621485e-06,
+      "loss": 0.6537,
+      "step": 4826
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.454616726505247,
+      "learning_rate": 5.1150259284348246e-06,
+      "loss": 0.571,
+      "step": 4827
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.132538433226866,
+      "learning_rate": 5.113322134644122e-06,
+      "loss": 0.6571,
+      "step": 4828
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.143250401873857,
+      "learning_rate": 5.111618327687981e-06,
+      "loss": 0.5883,
+      "step": 4829
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.356716637447877,
+      "learning_rate": 5.109914507764345e-06,
+      "loss": 0.6736,
+      "step": 4830
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7253830188411547,
+      "learning_rate": 5.108210675071159e-06,
+      "loss": 0.6881,
+      "step": 4831
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.4516235016563788,
+      "learning_rate": 5.1065068298063705e-06,
+      "loss": 0.6512,
+      "step": 4832
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7073002106917237,
+      "learning_rate": 5.104802972167926e-06,
+      "loss": 0.6957,
+      "step": 4833
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7841704207983984,
+      "learning_rate": 5.103099102353775e-06,
+      "loss": 0.5775,
+      "step": 4834
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.9292838099679965,
+      "learning_rate": 5.101395220561869e-06,
+      "loss": 0.7005,
+      "step": 4835
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7773812390741077,
+      "learning_rate": 5.099691326990158e-06,
+      "loss": 0.5243,
+      "step": 4836
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.26055135706108,
+      "learning_rate": 5.0979874218365985e-06,
+      "loss": 0.6554,
+      "step": 4837
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.93701844263028,
+      "learning_rate": 5.096283505299142e-06,
+      "loss": 0.5827,
+      "step": 4838
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.661195675756214,
+      "learning_rate": 5.094579577575748e-06,
+      "loss": 0.7464,
+      "step": 4839
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.4434201960292805,
+      "learning_rate": 5.09287563886437e-06,
+      "loss": 0.6926,
+      "step": 4840
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.8619375312083424,
+      "learning_rate": 5.0911716893629695e-06,
+      "loss": 0.7326,
+      "step": 4841
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.179918987044076,
+      "learning_rate": 5.089467729269506e-06,
+      "loss": 0.6202,
+      "step": 4842
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3674261810438875,
+      "learning_rate": 5.087763758781941e-06,
+      "loss": 0.5401,
+      "step": 4843
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.954478033527312,
+      "learning_rate": 5.0860597780982345e-06,
+      "loss": 0.6457,
+      "step": 4844
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.1055773678841705,
+      "learning_rate": 5.084355787416352e-06,
+      "loss": 0.7077,
+      "step": 4845
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.104870089140017,
+      "learning_rate": 5.08265178693426e-06,
+      "loss": 0.5571,
+      "step": 4846
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.282087698903521,
+      "learning_rate": 5.08094777684992e-06,
+      "loss": 0.7171,
+      "step": 4847
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.212353373051497,
+      "learning_rate": 5.079243757361304e-06,
+      "loss": 0.6634,
+      "step": 4848
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.151447480509918,
+      "learning_rate": 5.077539728666374e-06,
+      "loss": 0.6715,
+      "step": 4849
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7192044541711002,
+      "learning_rate": 5.0758356909631055e-06,
+      "loss": 0.6798,
+      "step": 4850
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7229617539344644,
+      "learning_rate": 5.074131644449462e-06,
+      "loss": 0.6132,
+      "step": 4851
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.5465212384001665,
+      "learning_rate": 5.072427589323422e-06,
+      "loss": 0.6457,
+      "step": 4852
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.539686113044392,
+      "learning_rate": 5.0707235257829525e-06,
+      "loss": 0.6064,
+      "step": 4853
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.8359956675113813,
+      "learning_rate": 5.069019454026028e-06,
+      "loss": 0.6889,
+      "step": 4854
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.075569786140454,
+      "learning_rate": 5.067315374250623e-06,
+      "loss": 0.6786,
+      "step": 4855
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.936345810031271,
+      "learning_rate": 5.065611286654712e-06,
+      "loss": 0.6724,
+      "step": 4856
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7223753328820366,
+      "learning_rate": 5.063907191436274e-06,
+      "loss": 0.6038,
+      "step": 4857
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.747137280849216,
+      "learning_rate": 5.062203088793279e-06,
+      "loss": 0.6451,
+      "step": 4858
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.534090883953568,
+      "learning_rate": 5.060498978923713e-06,
+      "loss": 0.7221,
+      "step": 4859
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.8222282010733877,
+      "learning_rate": 5.058794862025548e-06,
+      "loss": 0.6084,
+      "step": 4860
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.0443913355409364,
+      "learning_rate": 5.057090738296767e-06,
+      "loss": 0.6554,
+      "step": 4861
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.2515144608074653,
+      "learning_rate": 5.055386607935347e-06,
+      "loss": 0.5725,
+      "step": 4862
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1037680116930089,
+      "learning_rate": 5.053682471139275e-06,
+      "loss": 0.5484,
+      "step": 4863
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.5575549338006436,
+      "learning_rate": 5.051978328106525e-06,
+      "loss": 0.6632,
+      "step": 4864
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.962207242548026,
+      "learning_rate": 5.050274179035084e-06,
+      "loss": 0.6792,
+      "step": 4865
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.5877705132080306,
+      "learning_rate": 5.048570024122935e-06,
+      "loss": 0.6326,
+      "step": 4866
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.322301722453874,
+      "learning_rate": 5.046865863568061e-06,
+      "loss": 0.7552,
+      "step": 4867
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.682490269010231,
+      "learning_rate": 5.045161697568446e-06,
+      "loss": 0.6046,
+      "step": 4868
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.7731626743807523,
+      "learning_rate": 5.0434575263220745e-06,
+      "loss": 0.6787,
+      "step": 4869
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.898042881785744,
+      "learning_rate": 5.041753350026936e-06,
+      "loss": 0.5913,
+      "step": 4870
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.927842102069006,
+      "learning_rate": 5.0400491688810105e-06,
+      "loss": 0.6178,
+      "step": 4871
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1051954463354348,
+      "learning_rate": 5.038344983082292e-06,
+      "loss": 0.5853,
+      "step": 4872
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.71794682870089,
+      "learning_rate": 5.036640792828761e-06,
+      "loss": 0.6183,
+      "step": 4873
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.746461374115521,
+      "learning_rate": 5.0349365983184105e-06,
+      "loss": 0.6803,
+      "step": 4874
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2340308167373597,
+      "learning_rate": 5.033232399749226e-06,
+      "loss": 0.6801,
+      "step": 4875
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.9189306232219194,
+      "learning_rate": 5.031528197319197e-06,
+      "loss": 0.6212,
+      "step": 4876
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3120947610822853,
+      "learning_rate": 5.0298239912263145e-06,
+      "loss": 0.6743,
+      "step": 4877
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.838555185528894,
+      "learning_rate": 5.028119781668566e-06,
+      "loss": 0.6603,
+      "step": 4878
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8865667092436296,
+      "learning_rate": 5.026415568843943e-06,
+      "loss": 0.4771,
+      "step": 4879
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7027338628136923,
+      "learning_rate": 5.024711352950435e-06,
+      "loss": 0.6176,
+      "step": 4880
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.3089641577867184,
+      "learning_rate": 5.023007134186035e-06,
+      "loss": 0.5705,
+      "step": 4881
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1851775810718976,
+      "learning_rate": 5.0213029127487315e-06,
+      "loss": 0.5628,
+      "step": 4882
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.5359776818740016,
+      "learning_rate": 5.0195986888365175e-06,
+      "loss": 0.6628,
+      "step": 4883
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6067456420321014,
+      "learning_rate": 5.017894462647383e-06,
+      "loss": 0.7029,
+      "step": 4884
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.532052200950741,
+      "learning_rate": 5.0161902343793245e-06,
+      "loss": 0.6756,
+      "step": 4885
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4514973771189186,
+      "learning_rate": 5.014486004230329e-06,
+      "loss": 0.6228,
+      "step": 4886
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.9144203054617583,
+      "learning_rate": 5.012781772398392e-06,
+      "loss": 0.5954,
+      "step": 4887
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.5304442748367006,
+      "learning_rate": 5.011077539081506e-06,
+      "loss": 0.6309,
+      "step": 4888
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5115992891771706,
+      "learning_rate": 5.009373304477663e-06,
+      "loss": 0.65,
+      "step": 4889
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.842298359852735,
+      "learning_rate": 5.007669068784857e-06,
+      "loss": 0.5684,
+      "step": 4890
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.293124571221643,
+      "learning_rate": 5.005964832201079e-06,
+      "loss": 0.6347,
+      "step": 4891
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 12.289205214511725,
+      "learning_rate": 5.004260594924327e-06,
+      "loss": 0.649,
+      "step": 4892
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6036414189539125,
+      "learning_rate": 5.002556357152589e-06,
+      "loss": 0.6041,
+      "step": 4893
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7202200557498992,
+      "learning_rate": 5.000852119083863e-06,
+      "loss": 0.5758,
+      "step": 4894
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.769388945590125,
+      "learning_rate": 4.999147880916139e-06,
+      "loss": 0.6113,
+      "step": 4895
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.9093269736434353,
+      "learning_rate": 4.997443642847412e-06,
+      "loss": 0.737,
+      "step": 4896
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6169699031806175,
+      "learning_rate": 4.995739405075674e-06,
+      "loss": 0.6174,
+      "step": 4897
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.102025420126629,
+      "learning_rate": 4.994035167798921e-06,
+      "loss": 0.601,
+      "step": 4898
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5096200600239045,
+      "learning_rate": 4.992330931215146e-06,
+      "loss": 0.7385,
+      "step": 4899
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5045722346398853,
+      "learning_rate": 4.990626695522339e-06,
+      "loss": 0.7197,
+      "step": 4900
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.487306202546889,
+      "learning_rate": 4.988922460918496e-06,
+      "loss": 0.5743,
+      "step": 4901
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.319440294752822,
+      "learning_rate": 4.98721822760161e-06,
+      "loss": 0.7375,
+      "step": 4902
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1289416212364505,
+      "learning_rate": 4.985513995769672e-06,
+      "loss": 0.575,
+      "step": 4903
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.504486312284673,
+      "learning_rate": 4.983809765620678e-06,
+      "loss": 0.689,
+      "step": 4904
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.312174671775115,
+      "learning_rate": 4.9821055373526175e-06,
+      "loss": 0.5768,
+      "step": 4905
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0889029546172004,
+      "learning_rate": 4.980401311163483e-06,
+      "loss": 0.6912,
+      "step": 4906
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.355632254622633,
+      "learning_rate": 4.978697087251269e-06,
+      "loss": 0.6775,
+      "step": 4907
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2880860706153845,
+      "learning_rate": 4.976992865813968e-06,
+      "loss": 0.6585,
+      "step": 4908
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9846108300900807,
+      "learning_rate": 4.975288647049566e-06,
+      "loss": 0.5791,
+      "step": 4909
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2342087894997813,
+      "learning_rate": 4.9735844311560574e-06,
+      "loss": 0.6124,
+      "step": 4910
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9947124356007078,
+      "learning_rate": 4.9718802183314345e-06,
+      "loss": 0.6461,
+      "step": 4911
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.197647296374431,
+      "learning_rate": 4.970176008773688e-06,
+      "loss": 0.6122,
+      "step": 4912
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9932317458658896,
+      "learning_rate": 4.9684718026808035e-06,
+      "loss": 0.5758,
+      "step": 4913
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.709681927841805,
+      "learning_rate": 4.966767600250776e-06,
+      "loss": 0.6107,
+      "step": 4914
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1508024576423437,
+      "learning_rate": 4.965063401681591e-06,
+      "loss": 0.598,
+      "step": 4915
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.146135541337258,
+      "learning_rate": 4.963359207171239e-06,
+      "loss": 0.5641,
+      "step": 4916
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.3285165686253713,
+      "learning_rate": 4.961655016917712e-06,
+      "loss": 0.7023,
+      "step": 4917
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.943768718695757,
+      "learning_rate": 4.959950831118991e-06,
+      "loss": 0.6485,
+      "step": 4918
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.023540049447175,
+      "learning_rate": 4.958246649973066e-06,
+      "loss": 0.6233,
+      "step": 4919
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.000434315098372,
+      "learning_rate": 4.956542473677926e-06,
+      "loss": 0.709,
+      "step": 4920
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.067727072792758,
+      "learning_rate": 4.954838302431556e-06,
+      "loss": 0.5751,
+      "step": 4921
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3191980917060753,
+      "learning_rate": 4.9531341364319404e-06,
+      "loss": 0.6596,
+      "step": 4922
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3586105417389245,
+      "learning_rate": 4.951429975877066e-06,
+      "loss": 0.6919,
+      "step": 4923
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1526033976465793,
+      "learning_rate": 4.9497258209649165e-06,
+      "loss": 0.6568,
+      "step": 4924
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.133729675007697,
+      "learning_rate": 4.948021671893475e-06,
+      "loss": 0.6752,
+      "step": 4925
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.159426842721098,
+      "learning_rate": 4.946317528860728e-06,
+      "loss": 0.7144,
+      "step": 4926
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.115435574837891,
+      "learning_rate": 4.9446133920646535e-06,
+      "loss": 0.5658,
+      "step": 4927
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7390237750782194,
+      "learning_rate": 4.942909261703234e-06,
+      "loss": 0.606,
+      "step": 4928
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8886081183240573,
+      "learning_rate": 4.941205137974453e-06,
+      "loss": 0.6459,
+      "step": 4929
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1116737366663343,
+      "learning_rate": 4.93950102107629e-06,
+      "loss": 0.5802,
+      "step": 4930
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.26047601506244,
+      "learning_rate": 4.937796911206722e-06,
+      "loss": 0.6255,
+      "step": 4931
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1539115192013214,
+      "learning_rate": 4.936092808563729e-06,
+      "loss": 0.6115,
+      "step": 4932
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0145846571869437,
+      "learning_rate": 4.9343887133452885e-06,
+      "loss": 0.6485,
+      "step": 4933
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.099479126502059,
+      "learning_rate": 4.932684625749379e-06,
+      "loss": 0.6624,
+      "step": 4934
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5769731100724083,
+      "learning_rate": 4.930980545973973e-06,
+      "loss": 0.6308,
+      "step": 4935
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6783179427641115,
+      "learning_rate": 4.929276474217049e-06,
+      "loss": 0.6243,
+      "step": 4936
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.943547559066303,
+      "learning_rate": 4.92757241067658e-06,
+      "loss": 0.6204,
+      "step": 4937
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3266756319552204,
+      "learning_rate": 4.925868355550537e-06,
+      "loss": 0.5994,
+      "step": 4938
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.310739073972056,
+      "learning_rate": 4.924164309036897e-06,
+      "loss": 0.6684,
+      "step": 4939
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3395607545275916,
+      "learning_rate": 4.922460271333627e-06,
+      "loss": 0.5414,
+      "step": 4940
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4676331611691746,
+      "learning_rate": 4.920756242638698e-06,
+      "loss": 0.6746,
+      "step": 4941
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2941161951403686,
+      "learning_rate": 4.919052223150081e-06,
+      "loss": 0.7604,
+      "step": 4942
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.929369540693781,
+      "learning_rate": 4.917348213065742e-06,
+      "loss": 0.6284,
+      "step": 4943
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.196953658021791,
+      "learning_rate": 4.915644212583649e-06,
+      "loss": 0.5309,
+      "step": 4944
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.277862980217032,
+      "learning_rate": 4.913940221901766e-06,
+      "loss": 0.6333,
+      "step": 4945
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1353131160650287,
+      "learning_rate": 4.912236241218061e-06,
+      "loss": 0.6933,
+      "step": 4946
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1309552775913865,
+      "learning_rate": 4.910532270730497e-06,
+      "loss": 0.6451,
+      "step": 4947
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.226377878125629,
+      "learning_rate": 4.908828310637031e-06,
+      "loss": 0.6428,
+      "step": 4948
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.218293921242868,
+      "learning_rate": 4.907124361135632e-06,
+      "loss": 0.6537,
+      "step": 4949
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4059507731751615,
+      "learning_rate": 4.905420422424254e-06,
+      "loss": 0.5964,
+      "step": 4950
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4481730880265014,
+      "learning_rate": 4.903716494700859e-06,
+      "loss": 0.6094,
+      "step": 4951
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.684768787067311,
+      "learning_rate": 4.902012578163404e-06,
+      "loss": 0.6083,
+      "step": 4952
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9937131948780513,
+      "learning_rate": 4.900308673009843e-06,
+      "loss": 0.5961,
+      "step": 4953
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.274862105439873,
+      "learning_rate": 4.8986047794381325e-06,
+      "loss": 0.6355,
+      "step": 4954
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.081137560086168,
+      "learning_rate": 4.896900897646226e-06,
+      "loss": 0.6181,
+      "step": 4955
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1349551752747886,
+      "learning_rate": 4.8951970278320765e-06,
+      "loss": 0.6046,
+      "step": 4956
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.541248755489936,
+      "learning_rate": 4.89349317019363e-06,
+      "loss": 0.6012,
+      "step": 4957
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.565627078399468,
+      "learning_rate": 4.891789324928842e-06,
+      "loss": 0.6558,
+      "step": 4958
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.109453742078581,
+      "learning_rate": 4.890085492235657e-06,
+      "loss": 0.5778,
+      "step": 4959
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0050613585910755,
+      "learning_rate": 4.888381672312022e-06,
+      "loss": 0.6251,
+      "step": 4960
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.339550540816934,
+      "learning_rate": 4.88667786535588e-06,
+      "loss": 0.6687,
+      "step": 4961
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.290758913793059,
+      "learning_rate": 4.884974071565177e-06,
+      "loss": 0.6846,
+      "step": 4962
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.077911583636364,
+      "learning_rate": 4.883270291137852e-06,
+      "loss": 0.6473,
+      "step": 4963
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.985960627648404,
+      "learning_rate": 4.88156652427185e-06,
+      "loss": 0.6044,
+      "step": 4964
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.305367885235591,
+      "learning_rate": 4.879862771165107e-06,
+      "loss": 0.5774,
+      "step": 4965
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4768244502117036,
+      "learning_rate": 4.878159032015559e-06,
+      "loss": 0.6634,
+      "step": 4966
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4276873504006815,
+      "learning_rate": 4.8764553070211415e-06,
+      "loss": 0.6828,
+      "step": 4967
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4060474857332013,
+      "learning_rate": 4.87475159637979e-06,
+      "loss": 0.6065,
+      "step": 4968
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7772890860805246,
+      "learning_rate": 4.873047900289437e-06,
+      "loss": 0.6808,
+      "step": 4969
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0625237060008588,
+      "learning_rate": 4.87134421894801e-06,
+      "loss": 0.6523,
+      "step": 4970
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2524142418639466,
+      "learning_rate": 4.869640552553441e-06,
+      "loss": 0.6733,
+      "step": 4971
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2414743848382006,
+      "learning_rate": 4.867936901303656e-06,
+      "loss": 0.6021,
+      "step": 4972
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.338392581877551,
+      "learning_rate": 4.866233265396577e-06,
+      "loss": 0.7099,
+      "step": 4973
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0935049485955495,
+      "learning_rate": 4.8645296450301345e-06,
+      "loss": 0.6684,
+      "step": 4974
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1007971339772427,
+      "learning_rate": 4.862826040402246e-06,
+      "loss": 0.6992,
+      "step": 4975
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9696344105471323,
+      "learning_rate": 4.861122451710829e-06,
+      "loss": 0.5749,
+      "step": 4976
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.553225906237905,
+      "learning_rate": 4.859418879153805e-06,
+      "loss": 0.7279,
+      "step": 4977
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.067442914319545,
+      "learning_rate": 4.857715322929091e-06,
+      "loss": 0.6576,
+      "step": 4978
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.149325406618078,
+      "learning_rate": 4.856011783234599e-06,
+      "loss": 0.7101,
+      "step": 4979
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9706548407978104,
+      "learning_rate": 4.85430826026824e-06,
+      "loss": 0.6659,
+      "step": 4980
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6696818093641395,
+      "learning_rate": 4.852604754227927e-06,
+      "loss": 0.6047,
+      "step": 4981
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.3802463897178785,
+      "learning_rate": 4.8509012653115695e-06,
+      "loss": 0.7413,
+      "step": 4982
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1818945622881833,
+      "learning_rate": 4.849197793717069e-06,
+      "loss": 0.5983,
+      "step": 4983
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.14901122734723,
+      "learning_rate": 4.847494339642334e-06,
+      "loss": 0.7177,
+      "step": 4984
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.918569152656593,
+      "learning_rate": 4.8457909032852654e-06,
+      "loss": 0.6908,
+      "step": 4985
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9406184537970192,
+      "learning_rate": 4.844087484843764e-06,
+      "loss": 0.6133,
+      "step": 4986
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0469495720368256,
+      "learning_rate": 4.84238408451573e-06,
+      "loss": 0.6769,
+      "step": 4987
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7726823392613116,
+      "learning_rate": 4.840680702499056e-06,
+      "loss": 0.6546,
+      "step": 4988
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.272990604413833,
+      "learning_rate": 4.838977338991636e-06,
+      "loss": 0.647,
+      "step": 4989
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9330631724535494,
+      "learning_rate": 4.837273994191364e-06,
+      "loss": 0.6251,
+      "step": 4990
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2631226425982933,
+      "learning_rate": 4.835570668296131e-06,
+      "loss": 0.676,
+      "step": 4991
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.13674093049108,
+      "learning_rate": 4.83386736150382e-06,
+      "loss": 0.6693,
+      "step": 4992
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.133284920671136,
+      "learning_rate": 4.83216407401232e-06,
+      "loss": 0.5922,
+      "step": 4993
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.520394477013768,
+      "learning_rate": 4.830460806019512e-06,
+      "loss": 0.6195,
+      "step": 4994
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.693745272656559,
+      "learning_rate": 4.828757557723279e-06,
+      "loss": 0.6932,
+      "step": 4995
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4450291798500845,
+      "learning_rate": 4.827054329321496e-06,
+      "loss": 0.537,
+      "step": 4996
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0458948423518097,
+      "learning_rate": 4.825351121012042e-06,
+      "loss": 0.6444,
+      "step": 4997
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6712241240805357,
+      "learning_rate": 4.823647932992788e-06,
+      "loss": 0.6205,
+      "step": 4998
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.207121252476882,
+      "learning_rate": 4.82194476546161e-06,
+      "loss": 0.5541,
+      "step": 4999
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.289069632699829,
+      "learning_rate": 4.820241618616375e-06,
+      "loss": 0.544,
+      "step": 5000
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2675898245567687,
+      "learning_rate": 4.818538492654947e-06,
+      "loss": 0.6852,
+      "step": 5001
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.343735840653348,
+      "learning_rate": 4.816835387775193e-06,
+      "loss": 0.6876,
+      "step": 5002
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.246807379369579,
+      "learning_rate": 4.8151323041749734e-06,
+      "loss": 0.6657,
+      "step": 5003
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4237118515960447,
+      "learning_rate": 4.8134292420521505e-06,
+      "loss": 0.6224,
+      "step": 5004
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.289453411888955,
+      "learning_rate": 4.811726201604576e-06,
+      "loss": 0.655,
+      "step": 5005
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0026507977297503,
+      "learning_rate": 4.810023183030109e-06,
+      "loss": 0.6169,
+      "step": 5006
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.0079499431943195,
+      "learning_rate": 4.808320186526599e-06,
+      "loss": 0.6452,
+      "step": 5007
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2628812402524896,
+      "learning_rate": 4.806617212291898e-06,
+      "loss": 0.6522,
+      "step": 5008
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2505910057014997,
+      "learning_rate": 4.804914260523847e-06,
+      "loss": 0.5728,
+      "step": 5009
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2225142986972495,
+      "learning_rate": 4.803211331420294e-06,
+      "loss": 0.636,
+      "step": 5010
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5157421199340506,
+      "learning_rate": 4.801508425179079e-06,
+      "loss": 0.6838,
+      "step": 5011
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.362812838563486,
+      "learning_rate": 4.799805541998042e-06,
+      "loss": 0.6514,
+      "step": 5012
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.460937036790075,
+      "learning_rate": 4.79810268207502e-06,
+      "loss": 0.63,
+      "step": 5013
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.3836688933373873,
+      "learning_rate": 4.796399845607844e-06,
+      "loss": 0.6927,
+      "step": 5014
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1163838752069926,
+      "learning_rate": 4.7946970327943435e-06,
+      "loss": 0.6223,
+      "step": 5015
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4368026381336314,
+      "learning_rate": 4.79299424383235e-06,
+      "loss": 0.5904,
+      "step": 5016
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0227052397175225,
+      "learning_rate": 4.791291478919688e-06,
+      "loss": 0.6301,
+      "step": 5017
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.4149521326579464,
+      "learning_rate": 4.789588738254176e-06,
+      "loss": 0.6325,
+      "step": 5018
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.03492009399757,
+      "learning_rate": 4.787886022033637e-06,
+      "loss": 0.6345,
+      "step": 5019
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.045894564408748,
+      "learning_rate": 4.786183330455886e-06,
+      "loss": 0.6108,
+      "step": 5020
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0734976435360384,
+      "learning_rate": 4.784480663718742e-06,
+      "loss": 0.5606,
+      "step": 5021
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.08406849984304,
+      "learning_rate": 4.782778022020006e-06,
+      "loss": 0.6312,
+      "step": 5022
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8678914982078667,
+      "learning_rate": 4.7810754055574945e-06,
+      "loss": 0.6431,
+      "step": 5023
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.010459684826099,
+      "learning_rate": 4.779372814529008e-06,
+      "loss": 0.6737,
+      "step": 5024
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.274385411512584,
+      "learning_rate": 4.7776702491323506e-06,
+      "loss": 0.6473,
+      "step": 5025
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4893768635552216,
+      "learning_rate": 4.775967709565323e-06,
+      "loss": 0.5868,
+      "step": 5026
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8594743184322855,
+      "learning_rate": 4.774265196025716e-06,
+      "loss": 0.6205,
+      "step": 5027
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.0896156585418804,
+      "learning_rate": 4.772562708711328e-06,
+      "loss": 0.5436,
+      "step": 5028
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1024636760857023,
+      "learning_rate": 4.770860247819946e-06,
+      "loss": 0.6145,
+      "step": 5029
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7483433705268467,
+      "learning_rate": 4.7691578135493595e-06,
+      "loss": 0.6789,
+      "step": 5030
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.7267937561387456,
+      "learning_rate": 4.76745540609735e-06,
+      "loss": 0.651,
+      "step": 5031
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.930921651098154,
+      "learning_rate": 4.765753025661699e-06,
+      "loss": 0.5867,
+      "step": 5032
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0301984146112435,
+      "learning_rate": 4.764050672440184e-06,
+      "loss": 0.633,
+      "step": 5033
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.3826775767799564,
+      "learning_rate": 4.76234834663058e-06,
+      "loss": 0.6399,
+      "step": 5034
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.648287363125826,
+      "learning_rate": 4.76064604843066e-06,
+      "loss": 0.6734,
+      "step": 5035
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6257884700520076,
+      "learning_rate": 4.758943778038189e-06,
+      "loss": 0.5742,
+      "step": 5036
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.790385480712019,
+      "learning_rate": 4.757241535650931e-06,
+      "loss": 0.5854,
+      "step": 5037
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3321216810115755,
+      "learning_rate": 4.755539321466652e-06,
+      "loss": 0.7305,
+      "step": 5038
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.470106783507069,
+      "learning_rate": 4.753837135683108e-06,
+      "loss": 0.7199,
+      "step": 5039
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.519684963476052,
+      "learning_rate": 4.752134978498052e-06,
+      "loss": 0.6208,
+      "step": 5040
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9933322283576163,
+      "learning_rate": 4.750432850109239e-06,
+      "loss": 0.5654,
+      "step": 5041
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1628910176613396,
+      "learning_rate": 4.748730750714417e-06,
+      "loss": 0.5597,
+      "step": 5042
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.2601189066847938,
+      "learning_rate": 4.74702868051133e-06,
+      "loss": 0.5868,
+      "step": 5043
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4445761523300917,
+      "learning_rate": 4.745326639697718e-06,
+      "loss": 0.6006,
+      "step": 5044
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.157421144865131,
+      "learning_rate": 4.743624628471322e-06,
+      "loss": 0.5829,
+      "step": 5045
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.608165987972704,
+      "learning_rate": 4.741922647029873e-06,
+      "loss": 0.5994,
+      "step": 5046
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6959202608616892,
+      "learning_rate": 4.740220695571108e-06,
+      "loss": 0.7195,
+      "step": 5047
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.374505085575942,
+      "learning_rate": 4.738518774292752e-06,
+      "loss": 0.5836,
+      "step": 5048
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4217957922423077,
+      "learning_rate": 4.736816883392527e-06,
+      "loss": 0.5942,
+      "step": 5049
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.6160178947490085,
+      "learning_rate": 4.735115023068155e-06,
+      "loss": 0.6114,
+      "step": 5050
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9430364282301278,
+      "learning_rate": 4.733413193517355e-06,
+      "loss": 0.6125,
+      "step": 5051
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.886628277780551,
+      "learning_rate": 4.731711394937842e-06,
+      "loss": 0.6733,
+      "step": 5052
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7819669075459643,
+      "learning_rate": 4.73000962752732e-06,
+      "loss": 0.607,
+      "step": 5053
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.3962133589136707,
+      "learning_rate": 4.7283078914835e-06,
+      "loss": 0.6421,
+      "step": 5054
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8324268366863623,
+      "learning_rate": 4.726606187004084e-06,
+      "loss": 0.6786,
+      "step": 5055
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.009696361121063,
+      "learning_rate": 4.724904514286773e-06,
+      "loss": 0.6692,
+      "step": 5056
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6870058183715755,
+      "learning_rate": 4.723202873529256e-06,
+      "loss": 0.6067,
+      "step": 5057
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.358795360202302,
+      "learning_rate": 4.721501264929232e-06,
+      "loss": 0.6416,
+      "step": 5058
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2459402545172984,
+      "learning_rate": 4.719799688684385e-06,
+      "loss": 0.7759,
+      "step": 5059
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1705987722229483,
+      "learning_rate": 4.7180981449924006e-06,
+      "loss": 0.5818,
+      "step": 5060
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.764535131155558,
+      "learning_rate": 4.716396634050959e-06,
+      "loss": 0.6444,
+      "step": 5061
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7065511765293717,
+      "learning_rate": 4.714695156057737e-06,
+      "loss": 0.7239,
+      "step": 5062
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.614545604700766,
+      "learning_rate": 4.712993711210405e-06,
+      "loss": 0.5962,
+      "step": 5063
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.2589815632805568,
+      "learning_rate": 4.711292299706636e-06,
+      "loss": 0.6851,
+      "step": 5064
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.695743902716769,
+      "learning_rate": 4.709590921744093e-06,
+      "loss": 0.5757,
+      "step": 5065
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.0665293362792982,
+      "learning_rate": 4.707889577520436e-06,
+      "loss": 0.6153,
+      "step": 5066
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6924889301708195,
+      "learning_rate": 4.706188267233324e-06,
+      "loss": 0.6596,
+      "step": 5067
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3595683992153256,
+      "learning_rate": 4.704486991080409e-06,
+      "loss": 0.6294,
+      "step": 5068
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.669021302249065,
+      "learning_rate": 4.7027857492593445e-06,
+      "loss": 0.6792,
+      "step": 5069
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0161808520148305,
+      "learning_rate": 4.701084541967769e-06,
+      "loss": 0.5579,
+      "step": 5070
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2988515242704346,
+      "learning_rate": 4.699383369403329e-06,
+      "loss": 0.6003,
+      "step": 5071
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.252022330411591,
+      "learning_rate": 4.697682231763658e-06,
+      "loss": 0.6515,
+      "step": 5072
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0233341986752094,
+      "learning_rate": 4.695981129246393e-06,
+      "loss": 0.5974,
+      "step": 5073
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2983932146999533,
+      "learning_rate": 4.694280062049163e-06,
+      "loss": 0.5946,
+      "step": 5074
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.894694889398997,
+      "learning_rate": 4.6925790303695886e-06,
+      "loss": 0.5984,
+      "step": 5075
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.374417964315601,
+      "learning_rate": 4.690878034405296e-06,
+      "loss": 0.5768,
+      "step": 5076
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.310052580071245,
+      "learning_rate": 4.689177074353899e-06,
+      "loss": 0.559,
+      "step": 5077
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4835325819520286,
+      "learning_rate": 4.687476150413012e-06,
+      "loss": 0.6051,
+      "step": 5078
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.57984204212142,
+      "learning_rate": 4.6857752627802405e-06,
+      "loss": 0.5899,
+      "step": 5079
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6143636376389408,
+      "learning_rate": 4.684074411653192e-06,
+      "loss": 0.6276,
+      "step": 5080
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.065161043806273,
+      "learning_rate": 4.682373597229464e-06,
+      "loss": 0.6279,
+      "step": 5081
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.501763492850309,
+      "learning_rate": 4.680672819706654e-06,
+      "loss": 0.681,
+      "step": 5082
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.563624459964035,
+      "learning_rate": 4.678972079282354e-06,
+      "loss": 0.64,
+      "step": 5083
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0259420193680757,
+      "learning_rate": 4.677271376154149e-06,
+      "loss": 0.7082,
+      "step": 5084
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.20008658827479,
+      "learning_rate": 4.6755707105196204e-06,
+      "loss": 0.7601,
+      "step": 5085
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5221950538299738,
+      "learning_rate": 4.673870082576351e-06,
+      "loss": 0.5972,
+      "step": 5086
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.312808929084742,
+      "learning_rate": 4.672169492521914e-06,
+      "loss": 0.6986,
+      "step": 5087
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.5730283833471397,
+      "learning_rate": 4.670468940553875e-06,
+      "loss": 0.6108,
+      "step": 5088
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.6696729783798485,
+      "learning_rate": 4.6687684268698034e-06,
+      "loss": 0.6344,
+      "step": 5089
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.465514843046939,
+      "learning_rate": 4.667067951667256e-06,
+      "loss": 0.622,
+      "step": 5090
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.13181319546967,
+      "learning_rate": 4.665367515143797e-06,
+      "loss": 0.5712,
+      "step": 5091
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.425783344218305,
+      "learning_rate": 4.663667117496968e-06,
+      "loss": 0.6551,
+      "step": 5092
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5727074812927193,
+      "learning_rate": 4.6619667589243225e-06,
+      "loss": 0.6365,
+      "step": 5093
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3970092435076134,
+      "learning_rate": 4.6602664396234e-06,
+      "loss": 0.5804,
+      "step": 5094
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.186207872425851,
+      "learning_rate": 4.658566159791742e-06,
+      "loss": 0.6546,
+      "step": 5095
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.655332565894822,
+      "learning_rate": 4.656865919626883e-06,
+      "loss": 0.5582,
+      "step": 5096
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.041070063381227,
+      "learning_rate": 4.655165719326347e-06,
+      "loss": 0.7031,
+      "step": 5097
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9879284762088225,
+      "learning_rate": 4.653465559087661e-06,
+      "loss": 0.6188,
+      "step": 5098
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.362708201617922,
+      "learning_rate": 4.651765439108344e-06,
+      "loss": 0.5938,
+      "step": 5099
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3036918192067968,
+      "learning_rate": 4.650065359585914e-06,
+      "loss": 0.6185,
+      "step": 5100
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.862595099667224,
+      "learning_rate": 4.648365320717876e-06,
+      "loss": 0.6754,
+      "step": 5101
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3230616471795256,
+      "learning_rate": 4.64666532270174e-06,
+      "loss": 0.6647,
+      "step": 5102
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0507051961729577,
+      "learning_rate": 4.644965365735004e-06,
+      "loss": 0.5953,
+      "step": 5103
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.422141046773203,
+      "learning_rate": 4.643265450015169e-06,
+      "loss": 0.7227,
+      "step": 5104
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2097746173914743,
+      "learning_rate": 4.6415655757397206e-06,
+      "loss": 0.6461,
+      "step": 5105
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0454571953490075,
+      "learning_rate": 4.639865743106148e-06,
+      "loss": 0.5875,
+      "step": 5106
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.145981246348931,
+      "learning_rate": 4.63816595231193e-06,
+      "loss": 0.6611,
+      "step": 5107
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9222457727504096,
+      "learning_rate": 4.636466203554548e-06,
+      "loss": 0.5957,
+      "step": 5108
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.509150332110932,
+      "learning_rate": 4.634766497031472e-06,
+      "loss": 0.6091,
+      "step": 5109
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8618331161141985,
+      "learning_rate": 4.633066832940167e-06,
+      "loss": 0.6478,
+      "step": 5110
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3088562094725518,
+      "learning_rate": 4.631367211478098e-06,
+      "loss": 0.5967,
+      "step": 5111
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2252841546630666,
+      "learning_rate": 4.62966763284272e-06,
+      "loss": 0.703,
+      "step": 5112
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9848922788689293,
+      "learning_rate": 4.6279680972314875e-06,
+      "loss": 0.5806,
+      "step": 5113
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.421707800583379,
+      "learning_rate": 4.626268604841844e-06,
+      "loss": 0.655,
+      "step": 5114
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.060075796638142,
+      "learning_rate": 4.624569155871235e-06,
+      "loss": 0.6389,
+      "step": 5115
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.708021895892409,
+      "learning_rate": 4.622869750517094e-06,
+      "loss": 0.7085,
+      "step": 5116
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.198154906455453,
+      "learning_rate": 4.621170388976858e-06,
+      "loss": 0.5702,
+      "step": 5117
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.074916489895744,
+      "learning_rate": 4.619471071447949e-06,
+      "loss": 0.514,
+      "step": 5118
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.89007197368037,
+      "learning_rate": 4.617771798127792e-06,
+      "loss": 0.6663,
+      "step": 5119
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.186100478989604,
+      "learning_rate": 4.616072569213802e-06,
+      "loss": 0.5598,
+      "step": 5120
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.737181403903594,
+      "learning_rate": 4.614373384903391e-06,
+      "loss": 0.6478,
+      "step": 5121
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.744234393991175,
+      "learning_rate": 4.612674245393967e-06,
+      "loss": 0.6454,
+      "step": 5122
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4651854631448478,
+      "learning_rate": 4.610975150882928e-06,
+      "loss": 0.6289,
+      "step": 5123
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.106591381662164,
+      "learning_rate": 4.609276101567672e-06,
+      "loss": 0.6305,
+      "step": 5124
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8645757518422106,
+      "learning_rate": 4.607577097645587e-06,
+      "loss": 0.5743,
+      "step": 5125
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8329814893534113,
+      "learning_rate": 4.605878139314065e-06,
+      "loss": 0.6463,
+      "step": 5126
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3889289397117075,
+      "learning_rate": 4.604179226770478e-06,
+      "loss": 0.7184,
+      "step": 5127
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.124562346527691,
+      "learning_rate": 4.602480360212205e-06,
+      "loss": 0.6488,
+      "step": 5128
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7833987833164624,
+      "learning_rate": 4.600781539836614e-06,
+      "loss": 0.6184,
+      "step": 5129
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.554887684552683,
+      "learning_rate": 4.5990827658410705e-06,
+      "loss": 0.6443,
+      "step": 5130
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.83397128521315,
+      "learning_rate": 4.597384038422933e-06,
+      "loss": 0.6566,
+      "step": 5131
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.322027615656651,
+      "learning_rate": 4.595685357779553e-06,
+      "loss": 0.6787,
+      "step": 5132
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.634011797281909,
+      "learning_rate": 4.593986724108279e-06,
+      "loss": 0.5787,
+      "step": 5133
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0666278509393818,
+      "learning_rate": 4.592288137606454e-06,
+      "loss": 0.5666,
+      "step": 5134
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.207516428553423,
+      "learning_rate": 4.590589598471416e-06,
+      "loss": 0.6368,
+      "step": 5135
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.49123455675996,
+      "learning_rate": 4.588891106900493e-06,
+      "loss": 0.7018,
+      "step": 5136
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9847912096139533,
+      "learning_rate": 4.587192663091014e-06,
+      "loss": 0.5966,
+      "step": 5137
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.22624018717781,
+      "learning_rate": 4.5854942672402965e-06,
+      "loss": 0.582,
+      "step": 5138
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9077695725913995,
+      "learning_rate": 4.5837959195456605e-06,
+      "loss": 0.6262,
+      "step": 5139
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5077328392848472,
+      "learning_rate": 4.5820976202044085e-06,
+      "loss": 0.678,
+      "step": 5140
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.107242753594195,
+      "learning_rate": 4.580399369413847e-06,
+      "loss": 0.5716,
+      "step": 5141
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2886530499565723,
+      "learning_rate": 4.578701167371274e-06,
+      "loss": 0.6729,
+      "step": 5142
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.806852048189923,
+      "learning_rate": 4.577003014273981e-06,
+      "loss": 0.6226,
+      "step": 5143
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0468082920924573,
+      "learning_rate": 4.575304910319257e-06,
+      "loss": 0.7215,
+      "step": 5144
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3451100058055863,
+      "learning_rate": 4.573606855704379e-06,
+      "loss": 0.5165,
+      "step": 5145
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.122583359383461,
+      "learning_rate": 4.571908850626625e-06,
+      "loss": 0.558,
+      "step": 5146
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1021680512947545,
+      "learning_rate": 4.570210895283262e-06,
+      "loss": 0.5882,
+      "step": 5147
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.613254013725175,
+      "learning_rate": 4.568512989871557e-06,
+      "loss": 0.627,
+      "step": 5148
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.063740812597954,
+      "learning_rate": 4.566815134588763e-06,
+      "loss": 0.5857,
+      "step": 5149
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3292145678287235,
+      "learning_rate": 4.565117329632137e-06,
+      "loss": 0.5244,
+      "step": 5150
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2621172262271285,
+      "learning_rate": 4.5634195751989195e-06,
+      "loss": 0.6286,
+      "step": 5151
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.305527091957668,
+      "learning_rate": 4.561721871486357e-06,
+      "loss": 0.6073,
+      "step": 5152
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.664097973486397,
+      "learning_rate": 4.5600242186916786e-06,
+      "loss": 0.7036,
+      "step": 5153
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.907247212421276,
+      "learning_rate": 4.5583266170121155e-06,
+      "loss": 0.6086,
+      "step": 5154
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.356425324401827,
+      "learning_rate": 4.556629066644888e-06,
+      "loss": 0.5769,
+      "step": 5155
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.328669215805127,
+      "learning_rate": 4.554931567787214e-06,
+      "loss": 0.6026,
+      "step": 5156
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2570183445717817,
+      "learning_rate": 4.553234120636306e-06,
+      "loss": 0.6124,
+      "step": 5157
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.081027926632191,
+      "learning_rate": 4.551536725389364e-06,
+      "loss": 0.6321,
+      "step": 5158
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.273359671518665,
+      "learning_rate": 4.54983938224359e-06,
+      "loss": 0.6413,
+      "step": 5159
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9426125776938763,
+      "learning_rate": 4.5481420913961734e-06,
+      "loss": 0.6623,
+      "step": 5160
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.272805905164049,
+      "learning_rate": 4.546444853044308e-06,
+      "loss": 0.6284,
+      "step": 5161
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.032007123280218,
+      "learning_rate": 4.544747667385163e-06,
+      "loss": 0.671,
+      "step": 5162
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.705544234308645,
+      "learning_rate": 4.543050534615919e-06,
+      "loss": 0.6616,
+      "step": 5163
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3270348164118886,
+      "learning_rate": 4.541353454933743e-06,
+      "loss": 0.6169,
+      "step": 5164
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.1781706730214743,
+      "learning_rate": 4.539656428535799e-06,
+      "loss": 0.6087,
+      "step": 5165
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.157315390560079,
+      "learning_rate": 4.537959455619238e-06,
+      "loss": 0.5943,
+      "step": 5166
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.909391157704517,
+      "learning_rate": 4.536262536381213e-06,
+      "loss": 0.6192,
+      "step": 5167
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3877838658782027,
+      "learning_rate": 4.5345656710188645e-06,
+      "loss": 0.6268,
+      "step": 5168
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8840280487338967,
+      "learning_rate": 4.532868859729333e-06,
+      "loss": 0.6004,
+      "step": 5169
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.113177902066499,
+      "learning_rate": 4.531172102709746e-06,
+      "loss": 0.6332,
+      "step": 5170
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0427183759210084,
+      "learning_rate": 4.529475400157228e-06,
+      "loss": 0.6404,
+      "step": 5171
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.552191146094726,
+      "learning_rate": 4.527778752268899e-06,
+      "loss": 0.6292,
+      "step": 5172
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3830504963095254,
+      "learning_rate": 4.5260821592418685e-06,
+      "loss": 0.6213,
+      "step": 5173
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.240778096163022,
+      "learning_rate": 4.5243856212732466e-06,
+      "loss": 0.5673,
+      "step": 5174
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0053178425031761,
+      "learning_rate": 4.5226891385601235e-06,
+      "loss": 0.5822,
+      "step": 5175
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4335630650323252,
+      "learning_rate": 4.520992711299599e-06,
+      "loss": 0.6692,
+      "step": 5176
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.058047240249313,
+      "learning_rate": 4.519296339688754e-06,
+      "loss": 0.5852,
+      "step": 5177
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.063974457211975,
+      "learning_rate": 4.517600023924673e-06,
+      "loss": 0.6141,
+      "step": 5178
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1386266960462503,
+      "learning_rate": 4.515903764204428e-06,
+      "loss": 0.6138,
+      "step": 5179
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5711607873760527,
+      "learning_rate": 4.514207560725082e-06,
+      "loss": 0.6987,
+      "step": 5180
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.979892923073177,
+      "learning_rate": 4.512511413683698e-06,
+      "loss": 0.6426,
+      "step": 5181
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1211499569858487,
+      "learning_rate": 4.510815323277329e-06,
+      "loss": 0.6801,
+      "step": 5182
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.415322447097737,
+      "learning_rate": 4.509119289703023e-06,
+      "loss": 0.5961,
+      "step": 5183
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4841757246588174,
+      "learning_rate": 4.507423313157815e-06,
+      "loss": 0.6813,
+      "step": 5184
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.409140496949704,
+      "learning_rate": 4.505727393838746e-06,
+      "loss": 0.6362,
+      "step": 5185
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3420213528636924,
+      "learning_rate": 4.504031531942837e-06,
+      "loss": 0.6293,
+      "step": 5186
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.454227338659623,
+      "learning_rate": 4.502335727667114e-06,
+      "loss": 0.6149,
+      "step": 5187
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7418078304273394,
+      "learning_rate": 4.500639981208586e-06,
+      "loss": 0.6847,
+      "step": 5188
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9687799092010179,
+      "learning_rate": 4.498944292764261e-06,
+      "loss": 0.6455,
+      "step": 5189
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3321151600204844,
+      "learning_rate": 4.497248662531139e-06,
+      "loss": 0.605,
+      "step": 5190
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.4606611035859682,
+      "learning_rate": 4.495553090706216e-06,
+      "loss": 0.6776,
+      "step": 5191
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3731291130732086,
+      "learning_rate": 4.493857577486477e-06,
+      "loss": 0.6971,
+      "step": 5192
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.041114475320568,
+      "learning_rate": 4.492162123068899e-06,
+      "loss": 0.5753,
+      "step": 5193
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5283267101737827,
+      "learning_rate": 4.49046672765046e-06,
+      "loss": 0.6632,
+      "step": 5194
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.535327625572618,
+      "learning_rate": 4.488771391428122e-06,
+      "loss": 0.7297,
+      "step": 5195
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3383531443300956,
+      "learning_rate": 4.487076114598848e-06,
+      "loss": 0.5403,
+      "step": 5196
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1712529025180665,
+      "learning_rate": 4.485380897359587e-06,
+      "loss": 0.658,
+      "step": 5197
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2880236706230876,
+      "learning_rate": 4.483685739907285e-06,
+      "loss": 0.6671,
+      "step": 5198
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.923715631742324,
+      "learning_rate": 4.481990642438881e-06,
+      "loss": 0.6322,
+      "step": 5199
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5156565772915966,
+      "learning_rate": 4.480295605151308e-06,
+      "loss": 0.779,
+      "step": 5200
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.279437155096648,
+      "learning_rate": 4.47860062824149e-06,
+      "loss": 0.5941,
+      "step": 5201
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8210671148826783,
+      "learning_rate": 4.4769057119063425e-06,
+      "loss": 0.6129,
+      "step": 5202
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1034516568430606,
+      "learning_rate": 4.475210856342777e-06,
+      "loss": 0.655,
+      "step": 5203
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1311538275721205,
+      "learning_rate": 4.473516061747697e-06,
+      "loss": 0.537,
+      "step": 5204
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9788330045930997,
+      "learning_rate": 4.471821328318001e-06,
+      "loss": 0.5896,
+      "step": 5205
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.8737593347024135,
+      "learning_rate": 4.470126656250574e-06,
+      "loss": 0.5861,
+      "step": 5206
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5511717033127996,
+      "learning_rate": 4.468432045742301e-06,
+      "loss": 0.5941,
+      "step": 5207
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.9074504611235072,
+      "learning_rate": 4.466737496990057e-06,
+      "loss": 0.6367,
+      "step": 5208
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.069907121072369,
+      "learning_rate": 4.46504301019071e-06,
+      "loss": 0.6646,
+      "step": 5209
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6033731154873,
+      "learning_rate": 4.463348585541117e-06,
+      "loss": 0.6088,
+      "step": 5210
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.188979272302636,
+      "learning_rate": 4.461654223238136e-06,
+      "loss": 0.5968,
+      "step": 5211
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.743951094727751,
+      "learning_rate": 4.459959923478609e-06,
+      "loss": 0.6687,
+      "step": 5212
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.379573949144014,
+      "learning_rate": 4.45826568645938e-06,
+      "loss": 0.6304,
+      "step": 5213
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.454626913642241,
+      "learning_rate": 4.456571512377277e-06,
+      "loss": 0.601,
+      "step": 5214
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.822548767237185,
+      "learning_rate": 4.454877401429123e-06,
+      "loss": 0.5249,
+      "step": 5215
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6657172122549992,
+      "learning_rate": 4.453183353811737e-06,
+      "loss": 0.5839,
+      "step": 5216
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3013033806129486,
+      "learning_rate": 4.45148936972193e-06,
+      "loss": 0.5733,
+      "step": 5217
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2066422534371846,
+      "learning_rate": 4.449795449356502e-06,
+      "loss": 0.6463,
+      "step": 5218
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.360497161476003,
+      "learning_rate": 4.4481015929122465e-06,
+      "loss": 0.6128,
+      "step": 5219
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5466902571725396,
+      "learning_rate": 4.446407800585954e-06,
+      "loss": 0.6788,
+      "step": 5220
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.202579197824117,
+      "learning_rate": 4.444714072574401e-06,
+      "loss": 0.6787,
+      "step": 5221
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.010123101753715,
+      "learning_rate": 4.443020409074365e-06,
+      "loss": 0.7269,
+      "step": 5222
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8939895157613513,
+      "learning_rate": 4.441326810282606e-06,
+      "loss": 0.7057,
+      "step": 5223
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7291258805819063,
+      "learning_rate": 4.4396332763958835e-06,
+      "loss": 0.6379,
+      "step": 5224
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8152570916653485,
+      "learning_rate": 4.437939807610947e-06,
+      "loss": 0.6389,
+      "step": 5225
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.863829933630771,
+      "learning_rate": 4.436246404124539e-06,
+      "loss": 0.592,
+      "step": 5226
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9228368391092252,
+      "learning_rate": 4.4345530661333955e-06,
+      "loss": 0.6296,
+      "step": 5227
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4158924275798084,
+      "learning_rate": 4.432859793834239e-06,
+      "loss": 0.582,
+      "step": 5228
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.572611497694934,
+      "learning_rate": 4.431166587423794e-06,
+      "loss": 0.6169,
+      "step": 5229
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.268729295515817,
+      "learning_rate": 4.42947344709877e-06,
+      "loss": 0.6666,
+      "step": 5230
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0952795027263966,
+      "learning_rate": 4.4277803730558746e-06,
+      "loss": 0.6486,
+      "step": 5231
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0968523158920367,
+      "learning_rate": 4.426087365491798e-06,
+      "loss": 0.5293,
+      "step": 5232
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.088438078890414,
+      "learning_rate": 4.424394424603234e-06,
+      "loss": 0.5566,
+      "step": 5233
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1632948290431497,
+      "learning_rate": 4.42270155058686e-06,
+      "loss": 0.6262,
+      "step": 5234
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9284998520052347,
+      "learning_rate": 4.421008743639353e-06,
+      "loss": 0.5875,
+      "step": 5235
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0523984063768803,
+      "learning_rate": 4.419316003957376e-06,
+      "loss": 0.7087,
+      "step": 5236
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.443819742443756,
+      "learning_rate": 4.417623331737587e-06,
+      "loss": 0.6464,
+      "step": 5237
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0896384959029604,
+      "learning_rate": 4.415930727176634e-06,
+      "loss": 0.5857,
+      "step": 5238
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.230344099608812,
+      "learning_rate": 4.414238190471163e-06,
+      "loss": 0.6313,
+      "step": 5239
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.3364275587007395,
+      "learning_rate": 4.412545721817806e-06,
+      "loss": 0.7127,
+      "step": 5240
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9806871576287524,
+      "learning_rate": 4.410853321413187e-06,
+      "loss": 0.6476,
+      "step": 5241
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5890707975835805,
+      "learning_rate": 4.409160989453927e-06,
+      "loss": 0.6181,
+      "step": 5242
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.0789313597402534,
+      "learning_rate": 4.407468726136634e-06,
+      "loss": 0.637,
+      "step": 5243
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1476761639567976,
+      "learning_rate": 4.405776531657916e-06,
+      "loss": 0.6924,
+      "step": 5244
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.4858241564055628,
+      "learning_rate": 4.404084406214358e-06,
+      "loss": 0.7234,
+      "step": 5245
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3970949139274835,
+      "learning_rate": 4.402392350002554e-06,
+      "loss": 0.5235,
+      "step": 5246
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.235687197461085,
+      "learning_rate": 4.400700363219076e-06,
+      "loss": 0.6731,
+      "step": 5247
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.656344736133104,
+      "learning_rate": 4.399008446060501e-06,
+      "loss": 0.6499,
+      "step": 5248
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.0639313280738216,
+      "learning_rate": 4.397316598723385e-06,
+      "loss": 0.5468,
+      "step": 5249
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0867057122263857,
+      "learning_rate": 4.3956248214042855e-06,
+      "loss": 0.6578,
+      "step": 5250
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.599040114763554,
+      "learning_rate": 4.393933114299746e-06,
+      "loss": 0.6618,
+      "step": 5251
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6001557142240945,
+      "learning_rate": 4.3922414776063075e-06,
+      "loss": 0.6799,
+      "step": 5252
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1507305611894223,
+      "learning_rate": 4.390549911520497e-06,
+      "loss": 0.663,
+      "step": 5253
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7080296933337946,
+      "learning_rate": 4.388858416238834e-06,
+      "loss": 0.5737,
+      "step": 5254
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.185440108618935,
+      "learning_rate": 4.3871669919578345e-06,
+      "loss": 0.6253,
+      "step": 5255
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.41165829284016,
+      "learning_rate": 4.385475638874001e-06,
+      "loss": 0.6524,
+      "step": 5256
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0980213739464113,
+      "learning_rate": 4.383784357183835e-06,
+      "loss": 0.5656,
+      "step": 5257
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.656438895282767,
+      "learning_rate": 4.382093147083819e-06,
+      "loss": 0.6449,
+      "step": 5258
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4496335492831403,
+      "learning_rate": 4.380402008770435e-06,
+      "loss": 0.6254,
+      "step": 5259
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.959663295531009,
+      "learning_rate": 4.378710942440153e-06,
+      "loss": 0.6069,
+      "step": 5260
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7471051867065492,
+      "learning_rate": 4.377019948289441e-06,
+      "loss": 0.6661,
+      "step": 5261
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.8039135221418805,
+      "learning_rate": 4.375329026514749e-06,
+      "loss": 0.6398,
+      "step": 5262
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.677966696492936,
+      "learning_rate": 4.373638177312524e-06,
+      "loss": 0.5647,
+      "step": 5263
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6177835119924455,
+      "learning_rate": 4.371947400879205e-06,
+      "loss": 0.6016,
+      "step": 5264
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.233055028774205,
+      "learning_rate": 4.370256697411221e-06,
+      "loss": 0.5827,
+      "step": 5265
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2362734819525723,
+      "learning_rate": 4.368566067104998e-06,
+      "loss": 0.6496,
+      "step": 5266
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.211364135754513,
+      "learning_rate": 4.366875510156939e-06,
+      "loss": 0.557,
+      "step": 5267
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8665096255402696,
+      "learning_rate": 4.365185026763455e-06,
+      "loss": 0.5965,
+      "step": 5268
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.0564195544942065,
+      "learning_rate": 4.363494617120938e-06,
+      "loss": 0.7145,
+      "step": 5269
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.64369421900083,
+      "learning_rate": 4.361804281425779e-06,
+      "loss": 0.725,
+      "step": 5270
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.12214434238347,
+      "learning_rate": 4.360114019874353e-06,
+      "loss": 0.6916,
+      "step": 5271
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.609092742818454,
+      "learning_rate": 4.35842383266303e-06,
+      "loss": 0.6325,
+      "step": 5272
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.300634353062687,
+      "learning_rate": 4.35673371998817e-06,
+      "loss": 0.6284,
+      "step": 5273
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.7838481612576356,
+      "learning_rate": 4.355043682046129e-06,
+      "loss": 0.6038,
+      "step": 5274
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7986771832900788,
+      "learning_rate": 4.353353719033249e-06,
+      "loss": 0.654,
+      "step": 5275
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.87051694324319,
+      "learning_rate": 4.3516638311458624e-06,
+      "loss": 0.5146,
+      "step": 5276
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.349730181771263,
+      "learning_rate": 4.349974018580298e-06,
+      "loss": 0.6656,
+      "step": 5277
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.335395790852798,
+      "learning_rate": 4.348284281532874e-06,
+      "loss": 0.5557,
+      "step": 5278
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7325757799709605,
+      "learning_rate": 4.3465946201999e-06,
+      "loss": 0.7216,
+      "step": 5279
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.104588371726466,
+      "learning_rate": 4.344905034777672e-06,
+      "loss": 0.6013,
+      "step": 5280
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3109494580443304,
+      "learning_rate": 4.343215525462484e-06,
+      "loss": 0.6356,
+      "step": 5281
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.017653085191442,
+      "learning_rate": 4.3415260924506165e-06,
+      "loss": 0.6454,
+      "step": 5282
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.571020375717761,
+      "learning_rate": 4.339836735938347e-06,
+      "loss": 0.5784,
+      "step": 5283
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4571187576232276,
+      "learning_rate": 4.338147456121935e-06,
+      "loss": 0.6483,
+      "step": 5284
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5644236337830937,
+      "learning_rate": 4.336458253197637e-06,
+      "loss": 0.5715,
+      "step": 5285
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.511149918383041,
+      "learning_rate": 4.334769127361703e-06,
+      "loss": 0.6012,
+      "step": 5286
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.883950372753124,
+      "learning_rate": 4.333080078810369e-06,
+      "loss": 0.6156,
+      "step": 5287
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.4555634315635984,
+      "learning_rate": 4.331391107739864e-06,
+      "loss": 0.6276,
+      "step": 5288
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8118042237460923,
+      "learning_rate": 4.329702214346406e-06,
+      "loss": 0.6439,
+      "step": 5289
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.141667067861921,
+      "learning_rate": 4.328013398826207e-06,
+      "loss": 0.6292,
+      "step": 5290
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1279882409294806,
+      "learning_rate": 4.3263246613754685e-06,
+      "loss": 0.6465,
+      "step": 5291
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.1797440641409,
+      "learning_rate": 4.324636002190386e-06,
+      "loss": 0.6019,
+      "step": 5292
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5518905433521217,
+      "learning_rate": 4.322947421467138e-06,
+      "loss": 0.6755,
+      "step": 5293
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.707501918054458,
+      "learning_rate": 4.321258919401903e-06,
+      "loss": 0.7213,
+      "step": 5294
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.090219508336725,
+      "learning_rate": 4.319570496190843e-06,
+      "loss": 0.5684,
+      "step": 5295
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5586775104390838,
+      "learning_rate": 4.317882152030118e-06,
+      "loss": 0.6409,
+      "step": 5296
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7425290806450615,
+      "learning_rate": 4.316193887115871e-06,
+      "loss": 0.6583,
+      "step": 5297
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.929685634618751,
+      "learning_rate": 4.314505701644242e-06,
+      "loss": 0.6087,
+      "step": 5298
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.037591260391563,
+      "learning_rate": 4.3128175958113585e-06,
+      "loss": 0.5755,
+      "step": 5299
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.459322238903088,
+      "learning_rate": 4.311129569813341e-06,
+      "loss": 0.6523,
+      "step": 5300
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.3425180981358253,
+      "learning_rate": 4.3094416238463e-06,
+      "loss": 0.6665,
+      "step": 5301
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2917138959022267,
+      "learning_rate": 4.307753758106332e-06,
+      "loss": 0.6679,
+      "step": 5302
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9682044085217711,
+      "learning_rate": 4.306065972789533e-06,
+      "loss": 0.6181,
+      "step": 5303
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.339833756953796,
+      "learning_rate": 4.304378268091982e-06,
+      "loss": 0.6613,
+      "step": 5304
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.445630959329979,
+      "learning_rate": 4.302690644209756e-06,
+      "loss": 0.6045,
+      "step": 5305
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5016790643491187,
+      "learning_rate": 4.301003101338913e-06,
+      "loss": 0.5814,
+      "step": 5306
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.6755679527300735,
+      "learning_rate": 4.29931563967551e-06,
+      "loss": 0.6186,
+      "step": 5307
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4930015287099927,
+      "learning_rate": 4.2976282594155885e-06,
+      "loss": 0.6656,
+      "step": 5308
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3871048273612114,
+      "learning_rate": 4.2959409607551885e-06,
+      "loss": 0.6458,
+      "step": 5309
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6870418365333726,
+      "learning_rate": 4.294253743890331e-06,
+      "loss": 0.5847,
+      "step": 5310
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.453388094507715,
+      "learning_rate": 4.292566609017032e-06,
+      "loss": 0.5881,
+      "step": 5311
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.1874846967139456,
+      "learning_rate": 4.290879556331301e-06,
+      "loss": 0.5902,
+      "step": 5312
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.5444861113536374,
+      "learning_rate": 4.2891925860291315e-06,
+      "loss": 0.4916,
+      "step": 5313
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2307192839807906,
+      "learning_rate": 4.287505698306517e-06,
+      "loss": 0.6353,
+      "step": 5314
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.316081716986372,
+      "learning_rate": 4.285818893359427e-06,
+      "loss": 0.6589,
+      "step": 5315
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.455941628341801,
+      "learning_rate": 4.284132171383834e-06,
+      "loss": 0.6596,
+      "step": 5316
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.220993513664449,
+      "learning_rate": 4.2824455325756955e-06,
+      "loss": 0.6346,
+      "step": 5317
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.19089056244659,
+      "learning_rate": 4.2807589771309635e-06,
+      "loss": 0.5772,
+      "step": 5318
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.731529565019871,
+      "learning_rate": 4.2790725052455726e-06,
+      "loss": 0.6435,
+      "step": 5319
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.5785634546615195,
+      "learning_rate": 4.2773861171154525e-06,
+      "loss": 0.7559,
+      "step": 5320
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.480519667349258,
+      "learning_rate": 4.275699812936526e-06,
+      "loss": 0.6401,
+      "step": 5321
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4318321347360388,
+      "learning_rate": 4.2740135929047034e-06,
+      "loss": 0.6463,
+      "step": 5322
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5507047618208354,
+      "learning_rate": 4.2723274572158805e-06,
+      "loss": 0.6721,
+      "step": 5323
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4756784180683753,
+      "learning_rate": 4.27064140606595e-06,
+      "loss": 0.6507,
+      "step": 5324
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5608707563373527,
+      "learning_rate": 4.268955439650793e-06,
+      "loss": 0.5792,
+      "step": 5325
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.551538680904513,
+      "learning_rate": 4.267269558166279e-06,
+      "loss": 0.6255,
+      "step": 5326
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2584831084156374,
+      "learning_rate": 4.265583761808272e-06,
+      "loss": 0.7075,
+      "step": 5327
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4024189459356524,
+      "learning_rate": 4.26389805077262e-06,
+      "loss": 0.6714,
+      "step": 5328
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5974293268410182,
+      "learning_rate": 4.262212425255164e-06,
+      "loss": 0.6049,
+      "step": 5329
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.222020544696383,
+      "learning_rate": 4.260526885451734e-06,
+      "loss": 0.6188,
+      "step": 5330
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.637638372678464,
+      "learning_rate": 4.258841431558156e-06,
+      "loss": 0.6554,
+      "step": 5331
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0803098388886165,
+      "learning_rate": 4.257156063770237e-06,
+      "loss": 0.6647,
+      "step": 5332
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.038548102890104,
+      "learning_rate": 4.2554707822837775e-06,
+      "loss": 0.576,
+      "step": 5333
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4381809992111414,
+      "learning_rate": 4.253785587294571e-06,
+      "loss": 0.6267,
+      "step": 5334
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9095370655137764,
+      "learning_rate": 4.252100478998398e-06,
+      "loss": 0.5394,
+      "step": 5335
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.208800982681307,
+      "learning_rate": 4.250415457591031e-06,
+      "loss": 0.5668,
+      "step": 5336
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.237278462901377,
+      "learning_rate": 4.248730523268227e-06,
+      "loss": 0.554,
+      "step": 5337
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.159594855549469,
+      "learning_rate": 4.24704567622574e-06,
+      "loss": 0.5775,
+      "step": 5338
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0495262092053586,
+      "learning_rate": 4.245360916659309e-06,
+      "loss": 0.6269,
+      "step": 5339
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7328601805196513,
+      "learning_rate": 4.243676244764667e-06,
+      "loss": 0.5666,
+      "step": 5340
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.599232977470765,
+      "learning_rate": 4.241991660737532e-06,
+      "loss": 0.5742,
+      "step": 5341
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6689500818787217,
+      "learning_rate": 4.240307164773615e-06,
+      "loss": 0.6964,
+      "step": 5342
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0051800373166295,
+      "learning_rate": 4.238622757068614e-06,
+      "loss": 0.6334,
+      "step": 5343
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.296191095173538,
+      "learning_rate": 4.2369384378182216e-06,
+      "loss": 0.5819,
+      "step": 5344
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.7526757822737973,
+      "learning_rate": 4.2352542072181156e-06,
+      "loss": 0.5592,
+      "step": 5345
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.769568113242851,
+      "learning_rate": 4.233570065463964e-06,
+      "loss": 0.6135,
+      "step": 5346
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2793506199943447,
+      "learning_rate": 4.231886012751427e-06,
+      "loss": 0.5955,
+      "step": 5347
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4668991593558562,
+      "learning_rate": 4.230202049276152e-06,
+      "loss": 0.6746,
+      "step": 5348
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.273428983253158,
+      "learning_rate": 4.228518175233781e-06,
+      "loss": 0.6367,
+      "step": 5349
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.012893755691541,
+      "learning_rate": 4.226834390819935e-06,
+      "loss": 0.6371,
+      "step": 5350
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.892193534980026,
+      "learning_rate": 4.225150696230236e-06,
+      "loss": 0.6217,
+      "step": 5351
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4872260622117532,
+      "learning_rate": 4.223467091660287e-06,
+      "loss": 0.6131,
+      "step": 5352
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.493780184626296,
+      "learning_rate": 4.22178357730569e-06,
+      "loss": 0.6648,
+      "step": 5353
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1706205845561226,
+      "learning_rate": 4.220100153362026e-06,
+      "loss": 0.6315,
+      "step": 5354
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2135535393045607,
+      "learning_rate": 4.2184168200248695e-06,
+      "loss": 0.6127,
+      "step": 5355
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.6077079918211687,
+      "learning_rate": 4.21673357748979e-06,
+      "loss": 0.5827,
+      "step": 5356
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2105823742172683,
+      "learning_rate": 4.215050425952339e-06,
+      "loss": 0.6933,
+      "step": 5357
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2055870618699362,
+      "learning_rate": 4.213367365608061e-06,
+      "loss": 0.638,
+      "step": 5358
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.588655796581996,
+      "learning_rate": 4.211684396652487e-06,
+      "loss": 0.6038,
+      "step": 5359
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0448283559826437,
+      "learning_rate": 4.210001519281142e-06,
+      "loss": 0.7124,
+      "step": 5360
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.062696584888601,
+      "learning_rate": 4.208318733689535e-06,
+      "loss": 0.626,
+      "step": 5361
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.604045649910468,
+      "learning_rate": 4.206636040073172e-06,
+      "loss": 0.6666,
+      "step": 5362
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.508672230990463,
+      "learning_rate": 4.204953438627539e-06,
+      "loss": 0.6892,
+      "step": 5363
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.199729614432598,
+      "learning_rate": 4.203270929548117e-06,
+      "loss": 0.6088,
+      "step": 5364
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.315338636083697,
+      "learning_rate": 4.2015885130303745e-06,
+      "loss": 0.5451,
+      "step": 5365
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8563637400295274,
+      "learning_rate": 4.199906189269773e-06,
+      "loss": 0.6707,
+      "step": 5366
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.3092435696741203,
+      "learning_rate": 4.198223958461756e-06,
+      "loss": 0.5999,
+      "step": 5367
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.284714130272184,
+      "learning_rate": 4.196541820801761e-06,
+      "loss": 0.5714,
+      "step": 5368
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.208365754435486,
+      "learning_rate": 4.194859776485216e-06,
+      "loss": 0.6477,
+      "step": 5369
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.9161689800766633,
+      "learning_rate": 4.193177825707535e-06,
+      "loss": 0.605,
+      "step": 5370
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5071516041978534,
+      "learning_rate": 4.191495968664122e-06,
+      "loss": 0.6405,
+      "step": 5371
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.374183355215098,
+      "learning_rate": 4.189814205550369e-06,
+      "loss": 0.6325,
+      "step": 5372
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.9716578098089905,
+      "learning_rate": 4.18813253656166e-06,
+      "loss": 0.5657,
+      "step": 5373
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5849556565957026,
+      "learning_rate": 4.186450961893366e-06,
+      "loss": 0.5766,
+      "step": 5374
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6139215621325977,
+      "learning_rate": 4.184769481740848e-06,
+      "loss": 0.5995,
+      "step": 5375
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.035112537909609,
+      "learning_rate": 4.183088096299455e-06,
+      "loss": 0.6741,
+      "step": 5376
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.3310970193321285,
+      "learning_rate": 4.1814068057645255e-06,
+      "loss": 0.6492,
+      "step": 5377
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.581325181469474,
+      "learning_rate": 4.179725610331385e-06,
+      "loss": 0.5451,
+      "step": 5378
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.498160977405595,
+      "learning_rate": 4.178044510195356e-06,
+      "loss": 0.657,
+      "step": 5379
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.631101141420637,
+      "learning_rate": 4.176363505551737e-06,
+      "loss": 0.7063,
+      "step": 5380
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.528098787527815,
+      "learning_rate": 4.174682596595824e-06,
+      "loss": 0.6972,
+      "step": 5381
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3803147400781186,
+      "learning_rate": 4.173001783522903e-06,
+      "loss": 0.6797,
+      "step": 5382
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.483697902714156,
+      "learning_rate": 4.171321066528243e-06,
+      "loss": 0.659,
+      "step": 5383
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1064216692361284,
+      "learning_rate": 4.16964044580711e-06,
+      "loss": 0.6183,
+      "step": 5384
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.750461435944443,
+      "learning_rate": 4.167959921554745e-06,
+      "loss": 0.6222,
+      "step": 5385
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0857966955698677,
+      "learning_rate": 4.166279493966393e-06,
+      "loss": 0.6282,
+      "step": 5386
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.2748873110006333,
+      "learning_rate": 4.164599163237277e-06,
+      "loss": 0.627,
+      "step": 5387
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7282550442214744,
+      "learning_rate": 4.1629189295626195e-06,
+      "loss": 0.6795,
+      "step": 5388
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.174622024247829,
+      "learning_rate": 4.161238793137619e-06,
+      "loss": 0.5899,
+      "step": 5389
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.068250197507869,
+      "learning_rate": 4.159558754157469e-06,
+      "loss": 0.5851,
+      "step": 5390
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1242958834633563,
+      "learning_rate": 4.157878812817356e-06,
+      "loss": 0.6659,
+      "step": 5391
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7032218490370274,
+      "learning_rate": 4.1561989693124485e-06,
+      "loss": 0.6705,
+      "step": 5392
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.753703066203034,
+      "learning_rate": 4.154519223837904e-06,
+      "loss": 0.6477,
+      "step": 5393
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5189720424400934,
+      "learning_rate": 4.1528395765888716e-06,
+      "loss": 0.6326,
+      "step": 5394
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1872662432303587,
+      "learning_rate": 4.15116002776049e-06,
+      "loss": 0.6172,
+      "step": 5395
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8914596193035806,
+      "learning_rate": 4.14948057754788e-06,
+      "loss": 0.6174,
+      "step": 5396
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.0372536675287805,
+      "learning_rate": 4.147801226146163e-06,
+      "loss": 0.5389,
+      "step": 5397
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.7867058905780673,
+      "learning_rate": 4.146121973750431e-06,
+      "loss": 0.6527,
+      "step": 5398
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3654720115387655,
+      "learning_rate": 4.144442820555782e-06,
+      "loss": 0.6418,
+      "step": 5399
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.179581313117845,
+      "learning_rate": 4.142763766757292e-06,
+      "loss": 0.635,
+      "step": 5400
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2479561580681238,
+      "learning_rate": 4.141084812550031e-06,
+      "loss": 0.6461,
+      "step": 5401
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.841191478849272,
+      "learning_rate": 4.139405958129053e-06,
+      "loss": 0.6007,
+      "step": 5402
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0859232520906836,
+      "learning_rate": 4.137727203689402e-06,
+      "loss": 0.5766,
+      "step": 5403
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.3485825067338117,
+      "learning_rate": 4.136048549426112e-06,
+      "loss": 0.6512,
+      "step": 5404
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.49496623222408,
+      "learning_rate": 4.134369995534206e-06,
+      "loss": 0.6146,
+      "step": 5405
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.324007129830475,
+      "learning_rate": 4.132691542208691e-06,
+      "loss": 0.6011,
+      "step": 5406
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0793244375392315,
+      "learning_rate": 4.1310131896445635e-06,
+      "loss": 0.5288,
+      "step": 5407
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.2583878078923725,
+      "learning_rate": 4.129334938036813e-06,
+      "loss": 0.6705,
+      "step": 5408
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.307137719361831,
+      "learning_rate": 4.127656787580412e-06,
+      "loss": 0.6923,
+      "step": 5409
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.578677072913258,
+      "learning_rate": 4.125978738470326e-06,
+      "loss": 0.6375,
+      "step": 5410
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4643009262796918,
+      "learning_rate": 4.1243007909015016e-06,
+      "loss": 0.5542,
+      "step": 5411
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2146931372972523,
+      "learning_rate": 4.1226229450688805e-06,
+      "loss": 0.6535,
+      "step": 5412
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.119950795155613,
+      "learning_rate": 4.120945201167388e-06,
+      "loss": 0.6085,
+      "step": 5413
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.5062101551407485,
+      "learning_rate": 4.119267559391944e-06,
+      "loss": 0.6207,
+      "step": 5414
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8128176772340931,
+      "learning_rate": 4.117590019937447e-06,
+      "loss": 0.5534,
+      "step": 5415
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4544616665994945,
+      "learning_rate": 4.11591258299879e-06,
+      "loss": 0.664,
+      "step": 5416
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.5069164956287646,
+      "learning_rate": 4.114235248770854e-06,
+      "loss": 0.6393,
+      "step": 5417
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7344010000318297,
+      "learning_rate": 4.112558017448508e-06,
+      "loss": 0.6276,
+      "step": 5418
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.5983760316820343,
+      "learning_rate": 4.1108808892266045e-06,
+      "loss": 0.6458,
+      "step": 5419
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8082114352695724,
+      "learning_rate": 4.109203864299989e-06,
+      "loss": 0.5609,
+      "step": 5420
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.364880898558155,
+      "learning_rate": 4.107526942863493e-06,
+      "loss": 0.7066,
+      "step": 5421
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9084674442912224,
+      "learning_rate": 4.105850125111937e-06,
+      "loss": 0.613,
+      "step": 5422
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.039149108932299,
+      "learning_rate": 4.104173411240131e-06,
+      "loss": 0.5501,
+      "step": 5423
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.760756741136062,
+      "learning_rate": 4.102496801442868e-06,
+      "loss": 0.7026,
+      "step": 5424
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2973399753902295,
+      "learning_rate": 4.10082029591493e-06,
+      "loss": 0.643,
+      "step": 5425
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5509810297461826,
+      "learning_rate": 4.099143894851092e-06,
+      "loss": 0.5958,
+      "step": 5426
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0570994896608783,
+      "learning_rate": 4.097467598446113e-06,
+      "loss": 0.5708,
+      "step": 5427
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0781495093655433,
+      "learning_rate": 4.095791406894739e-06,
+      "loss": 0.5697,
+      "step": 5428
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.46415369130286,
+      "learning_rate": 4.094115320391704e-06,
+      "loss": 0.6917,
+      "step": 5429
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1679299121769535,
+      "learning_rate": 4.0924393391317344e-06,
+      "loss": 0.6069,
+      "step": 5430
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0173189356350105,
+      "learning_rate": 4.090763463309536e-06,
+      "loss": 0.5557,
+      "step": 5431
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.570745562650635,
+      "learning_rate": 4.089087693119815e-06,
+      "loss": 0.6819,
+      "step": 5432
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7474195051489185,
+      "learning_rate": 4.087412028757249e-06,
+      "loss": 0.621,
+      "step": 5433
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9523833561555928,
+      "learning_rate": 4.085736470416517e-06,
+      "loss": 0.6772,
+      "step": 5434
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4968268241742986,
+      "learning_rate": 4.084061018292277e-06,
+      "loss": 0.6174,
+      "step": 5435
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.53792252658528,
+      "learning_rate": 4.082385672579182e-06,
+      "loss": 0.637,
+      "step": 5436
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.964810455872569,
+      "learning_rate": 4.0807104334718674e-06,
+      "loss": 0.6321,
+      "step": 5437
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2794226187051545,
+      "learning_rate": 4.079035301164955e-06,
+      "loss": 0.5753,
+      "step": 5438
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.488542771422994,
+      "learning_rate": 4.0773602758530606e-06,
+      "loss": 0.6145,
+      "step": 5439
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.149185123490166,
+      "learning_rate": 4.0756853577307835e-06,
+      "loss": 0.7298,
+      "step": 5440
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 19.446457701330363,
+      "learning_rate": 4.0740105469927084e-06,
+      "loss": 0.6299,
+      "step": 5441
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.921362940815699,
+      "learning_rate": 4.07233584383341e-06,
+      "loss": 0.5932,
+      "step": 5442
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.436933837657758,
+      "learning_rate": 4.070661248447453e-06,
+      "loss": 0.6107,
+      "step": 5443
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4936425865614624,
+      "learning_rate": 4.0689867610293845e-06,
+      "loss": 0.6755,
+      "step": 5444
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3563232675472374,
+      "learning_rate": 4.067312381773744e-06,
+      "loss": 0.5669,
+      "step": 5445
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.811216819647008,
+      "learning_rate": 4.065638110875055e-06,
+      "loss": 0.6642,
+      "step": 5446
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.9029408699267543,
+      "learning_rate": 4.063963948527829e-06,
+      "loss": 0.6854,
+      "step": 5447
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3885007406079173,
+      "learning_rate": 4.062289894926564e-06,
+      "loss": 0.6045,
+      "step": 5448
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5249267659429764,
+      "learning_rate": 4.060615950265752e-06,
+      "loss": 0.7814,
+      "step": 5449
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5576000799315537,
+      "learning_rate": 4.058942114739861e-06,
+      "loss": 0.6392,
+      "step": 5450
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.340626814388993,
+      "learning_rate": 4.057268388543354e-06,
+      "loss": 0.6515,
+      "step": 5451
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.366975049960535,
+      "learning_rate": 4.055594771870682e-06,
+      "loss": 0.5461,
+      "step": 5452
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.5250602391277406,
+      "learning_rate": 4.05392126491628e-06,
+      "loss": 0.6166,
+      "step": 5453
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7710266206375826,
+      "learning_rate": 4.052247867874569e-06,
+      "loss": 0.5996,
+      "step": 5454
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5532961974178727,
+      "learning_rate": 4.050574580939961e-06,
+      "loss": 0.6165,
+      "step": 5455
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.2749550989913345,
+      "learning_rate": 4.0489014043068545e-06,
+      "loss": 0.6004,
+      "step": 5456
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.131553043318907,
+      "learning_rate": 4.047228338169632e-06,
+      "loss": 0.6498,
+      "step": 5457
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.574736501197967,
+      "learning_rate": 4.04555538272267e-06,
+      "loss": 0.5856,
+      "step": 5458
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1788021789713112,
+      "learning_rate": 4.0438825381603225e-06,
+      "loss": 0.6355,
+      "step": 5459
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.4984416537675296,
+      "learning_rate": 4.042209804676937e-06,
+      "loss": 0.6518,
+      "step": 5460
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5902715960904383,
+      "learning_rate": 4.040537182466849e-06,
+      "loss": 0.538,
+      "step": 5461
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1931566795393427,
+      "learning_rate": 4.038864671724379e-06,
+      "loss": 0.6478,
+      "step": 5462
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.817321693238749,
+      "learning_rate": 4.0371922726438314e-06,
+      "loss": 0.6962,
+      "step": 5463
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1447873012073355,
+      "learning_rate": 4.035519985419502e-06,
+      "loss": 0.5852,
+      "step": 5464
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.884660635376069,
+      "learning_rate": 4.033847810245673e-06,
+      "loss": 0.6715,
+      "step": 5465
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.4056554597314035,
+      "learning_rate": 4.0321757473166145e-06,
+      "loss": 0.6417,
+      "step": 5466
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4034378137851213,
+      "learning_rate": 4.030503796826578e-06,
+      "loss": 0.6369,
+      "step": 5467
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.279509860886487,
+      "learning_rate": 4.028831958969807e-06,
+      "loss": 0.6686,
+      "step": 5468
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6768593400635177,
+      "learning_rate": 4.027160233940534e-06,
+      "loss": 0.6936,
+      "step": 5469
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.696418739353531,
+      "learning_rate": 4.02548862193297e-06,
+      "loss": 0.5815,
+      "step": 5470
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8470170762266642,
+      "learning_rate": 4.023817123141324e-06,
+      "loss": 0.6268,
+      "step": 5471
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3205700797742703,
+      "learning_rate": 4.022145737759781e-06,
+      "loss": 0.7149,
+      "step": 5472
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4124617955340772,
+      "learning_rate": 4.020474465982519e-06,
+      "loss": 0.5885,
+      "step": 5473
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4297597298078606,
+      "learning_rate": 4.0188033080037025e-06,
+      "loss": 0.632,
+      "step": 5474
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7294892840221427,
+      "learning_rate": 4.017132264017483e-06,
+      "loss": 0.6413,
+      "step": 5475
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0863421441237322,
+      "learning_rate": 4.015461334217995e-06,
+      "loss": 0.5971,
+      "step": 5476
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5430832765858735,
+      "learning_rate": 4.013790518799361e-06,
+      "loss": 0.6309,
+      "step": 5477
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.900637212310984,
+      "learning_rate": 4.012119817955696e-06,
+      "loss": 0.5844,
+      "step": 5478
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1161956263445076,
+      "learning_rate": 4.010449231881093e-06,
+      "loss": 0.6315,
+      "step": 5479
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.31955950639972,
+      "learning_rate": 4.00877876076964e-06,
+      "loss": 0.64,
+      "step": 5480
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1236471054278256,
+      "learning_rate": 4.0071084048154044e-06,
+      "loss": 0.6076,
+      "step": 5481
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.246468971263171,
+      "learning_rate": 4.005438164212444e-06,
+      "loss": 0.6235,
+      "step": 5482
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0097201604995623,
+      "learning_rate": 4.0037680391548015e-06,
+      "loss": 0.5754,
+      "step": 5483
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2344373050878206,
+      "learning_rate": 4.002098029836511e-06,
+      "loss": 0.6543,
+      "step": 5484
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.9506649394699247,
+      "learning_rate": 4.000428136451585e-06,
+      "loss": 0.5946,
+      "step": 5485
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.350295409726678,
+      "learning_rate": 3.998758359194028e-06,
+      "loss": 0.5822,
+      "step": 5486
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4139769523797048,
+      "learning_rate": 3.9970886982578314e-06,
+      "loss": 0.6689,
+      "step": 5487
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.322280956259487,
+      "learning_rate": 3.995419153836972e-06,
+      "loss": 0.5373,
+      "step": 5488
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.334612127720499,
+      "learning_rate": 3.9937497261254114e-06,
+      "loss": 0.741,
+      "step": 5489
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6565331932796057,
+      "learning_rate": 3.992080415317096e-06,
+      "loss": 0.6195,
+      "step": 5490
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.485017788078939,
+      "learning_rate": 3.9904112216059656e-06,
+      "loss": 0.5669,
+      "step": 5491
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.189719358830894,
+      "learning_rate": 3.988742145185941e-06,
+      "loss": 0.6253,
+      "step": 5492
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8776383714391742,
+      "learning_rate": 3.987073186250932e-06,
+      "loss": 0.6488,
+      "step": 5493
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.0757770946653458,
+      "learning_rate": 3.985404344994831e-06,
+      "loss": 0.7339,
+      "step": 5494
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.9151134884480556,
+      "learning_rate": 3.9837356216115205e-06,
+      "loss": 0.6524,
+      "step": 5495
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5361472839977455,
+      "learning_rate": 3.982067016294868e-06,
+      "loss": 0.6586,
+      "step": 5496
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1819398494243964,
+      "learning_rate": 3.980398529238728e-06,
+      "loss": 0.6144,
+      "step": 5497
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.277473751575629,
+      "learning_rate": 3.978730160636938e-06,
+      "loss": 0.5899,
+      "step": 5498
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.865134875134504,
+      "learning_rate": 3.977061910683325e-06,
+      "loss": 0.6446,
+      "step": 5499
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0322930621616004,
+      "learning_rate": 3.975393779571704e-06,
+      "loss": 0.6317,
+      "step": 5500
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.853800658322435,
+      "learning_rate": 3.9737257674958714e-06,
+      "loss": 0.5481,
+      "step": 5501
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9879336943874001,
+      "learning_rate": 3.972057874649613e-06,
+      "loss": 0.5534,
+      "step": 5502
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.222678097797317,
+      "learning_rate": 3.970390101226697e-06,
+      "loss": 0.5955,
+      "step": 5503
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.269757412226817,
+      "learning_rate": 3.968722447420884e-06,
+      "loss": 0.6627,
+      "step": 5504
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.009770633477226,
+      "learning_rate": 3.967054913425916e-06,
+      "loss": 0.6221,
+      "step": 5505
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8457569192736396,
+      "learning_rate": 3.965387499435524e-06,
+      "loss": 0.6292,
+      "step": 5506
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.254292979448844,
+      "learning_rate": 3.963720205643419e-06,
+      "loss": 0.6363,
+      "step": 5507
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.300083482731035,
+      "learning_rate": 3.962053032243305e-06,
+      "loss": 0.6737,
+      "step": 5508
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.126068404526892,
+      "learning_rate": 3.96038597942887e-06,
+      "loss": 0.689,
+      "step": 5509
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.441809460166265,
+      "learning_rate": 3.958719047393789e-06,
+      "loss": 0.6798,
+      "step": 5510
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8015689631584326,
+      "learning_rate": 3.9570522363317165e-06,
+      "loss": 0.6766,
+      "step": 5511
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8277968259678123,
+      "learning_rate": 3.955385546436299e-06,
+      "loss": 0.5986,
+      "step": 5512
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.335643392259737,
+      "learning_rate": 3.9537189779011715e-06,
+      "loss": 0.5452,
+      "step": 5513
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3708189781396936,
+      "learning_rate": 3.952052530919948e-06,
+      "loss": 0.6386,
+      "step": 5514
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.838879644566167,
+      "learning_rate": 3.9503862056862315e-06,
+      "loss": 0.7274,
+      "step": 5515
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4970962843979057,
+      "learning_rate": 3.948720002393613e-06,
+      "loss": 0.6018,
+      "step": 5516
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1528209154994675,
+      "learning_rate": 3.947053921235665e-06,
+      "loss": 0.6035,
+      "step": 5517
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.223370620526591,
+      "learning_rate": 3.945387962405946e-06,
+      "loss": 0.6433,
+      "step": 5518
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.583378068602465,
+      "learning_rate": 3.943722126098009e-06,
+      "loss": 0.5599,
+      "step": 5519
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.934321798163941,
+      "learning_rate": 3.94205641250538e-06,
+      "loss": 0.7043,
+      "step": 5520
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7272967348881823,
+      "learning_rate": 3.940390821821579e-06,
+      "loss": 0.6763,
+      "step": 5521
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4185559104349137,
+      "learning_rate": 3.93872535424011e-06,
+      "loss": 0.6508,
+      "step": 5522
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4863515025722562,
+      "learning_rate": 3.937060009954462e-06,
+      "loss": 0.6582,
+      "step": 5523
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.89940109019646,
+      "learning_rate": 3.935394789158108e-06,
+      "loss": 0.6821,
+      "step": 5524
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.850326326013693,
+      "learning_rate": 3.93372969204451e-06,
+      "loss": 0.5794,
+      "step": 5525
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.1372923087151254,
+      "learning_rate": 3.932064718807114e-06,
+      "loss": 0.6803,
+      "step": 5526
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0327932638867807,
+      "learning_rate": 3.930399869639353e-06,
+      "loss": 0.6513,
+      "step": 5527
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.519770785227865,
+      "learning_rate": 3.9287351447346424e-06,
+      "loss": 0.6086,
+      "step": 5528
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0115479234131517,
+      "learning_rate": 3.927070544286385e-06,
+      "loss": 0.5529,
+      "step": 5529
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0358399237096076,
+      "learning_rate": 3.925406068487972e-06,
+      "loss": 0.5233,
+      "step": 5530
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7457465246262487,
+      "learning_rate": 3.923741717532774e-06,
+      "loss": 0.6148,
+      "step": 5531
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7191030649642522,
+      "learning_rate": 3.922077491614155e-06,
+      "loss": 0.653,
+      "step": 5532
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.914553835743117,
+      "learning_rate": 3.920413390925454e-06,
+      "loss": 0.6825,
+      "step": 5533
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2209209916064294,
+      "learning_rate": 3.918749415660005e-06,
+      "loss": 0.6485,
+      "step": 5534
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6768274712210633,
+      "learning_rate": 3.917085566011124e-06,
+      "loss": 0.6383,
+      "step": 5535
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.07522859990918,
+      "learning_rate": 3.915421842172113e-06,
+      "loss": 0.6638,
+      "step": 5536
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4434649901286885,
+      "learning_rate": 3.913758244336255e-06,
+      "loss": 0.6297,
+      "step": 5537
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.82081715284726,
+      "learning_rate": 3.912094772696825e-06,
+      "loss": 0.6508,
+      "step": 5538
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2583424490630546,
+      "learning_rate": 3.910431427447079e-06,
+      "loss": 0.5839,
+      "step": 5539
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.663703287673878,
+      "learning_rate": 3.908768208780259e-06,
+      "loss": 0.5589,
+      "step": 5540
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.274839976109476,
+      "learning_rate": 3.907105116889597e-06,
+      "loss": 0.6395,
+      "step": 5541
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4342153540549925,
+      "learning_rate": 3.905442151968302e-06,
+      "loss": 0.6024,
+      "step": 5542
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.4254973540309996,
+      "learning_rate": 3.903779314209573e-06,
+      "loss": 0.612,
+      "step": 5543
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.9209988262515627,
+      "learning_rate": 3.902116603806594e-06,
+      "loss": 0.635,
+      "step": 5544
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5257899785531785,
+      "learning_rate": 3.900454020952537e-06,
+      "loss": 0.6775,
+      "step": 5545
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.231536697694151,
+      "learning_rate": 3.898791565840552e-06,
+      "loss": 0.6288,
+      "step": 5546
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7797880961916452,
+      "learning_rate": 3.897129238663777e-06,
+      "loss": 0.5785,
+      "step": 5547
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1833271590239662,
+      "learning_rate": 3.895467039615342e-06,
+      "loss": 0.6688,
+      "step": 5548
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.5023552603785757,
+      "learning_rate": 3.893804968888354e-06,
+      "loss": 0.5951,
+      "step": 5549
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1779919766332645,
+      "learning_rate": 3.892143026675905e-06,
+      "loss": 0.6629,
+      "step": 5550
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.937739628950341,
+      "learning_rate": 3.8904812131710776e-06,
+      "loss": 0.6831,
+      "step": 5551
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7094582537792893,
+      "learning_rate": 3.888819528566935e-06,
+      "loss": 0.6519,
+      "step": 5552
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.507846503527637,
+      "learning_rate": 3.8871579730565265e-06,
+      "loss": 0.6874,
+      "step": 5553
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3128367049716334,
+      "learning_rate": 3.885496546832891e-06,
+      "loss": 0.6468,
+      "step": 5554
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.18788284410742,
+      "learning_rate": 3.883835250089043e-06,
+      "loss": 0.608,
+      "step": 5555
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.0367412519509323,
+      "learning_rate": 3.8821740830179876e-06,
+      "loss": 0.6377,
+      "step": 5556
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.587808272946125,
+      "learning_rate": 3.880513045812718e-06,
+      "loss": 0.6709,
+      "step": 5557
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8752893486354014,
+      "learning_rate": 3.8788521386662076e-06,
+      "loss": 0.6109,
+      "step": 5558
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.853722055103752,
+      "learning_rate": 3.8771913617714135e-06,
+      "loss": 0.5964,
+      "step": 5559
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0139273382638472,
+      "learning_rate": 3.87553071532128e-06,
+      "loss": 0.57,
+      "step": 5560
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.753533947661753,
+      "learning_rate": 3.873870199508739e-06,
+      "loss": 0.6884,
+      "step": 5561
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.59431286674271,
+      "learning_rate": 3.872209814526703e-06,
+      "loss": 0.6489,
+      "step": 5562
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.118810061041824,
+      "learning_rate": 3.87054956056807e-06,
+      "loss": 0.6611,
+      "step": 5563
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.2188855308440183,
+      "learning_rate": 3.868889437825724e-06,
+      "loss": 0.6555,
+      "step": 5564
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4614632914433274,
+      "learning_rate": 3.867229446492533e-06,
+      "loss": 0.6215,
+      "step": 5565
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6768898443750375,
+      "learning_rate": 3.865569586761352e-06,
+      "loss": 0.6264,
+      "step": 5566
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4319918129394216,
+      "learning_rate": 3.863909858825016e-06,
+      "loss": 0.6583,
+      "step": 5567
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.70131547101183,
+      "learning_rate": 3.86225026287635e-06,
+      "loss": 0.5622,
+      "step": 5568
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5863671241991466,
+      "learning_rate": 3.860590799108157e-06,
+      "loss": 0.6492,
+      "step": 5569
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.603038839665001,
+      "learning_rate": 3.858931467713233e-06,
+      "loss": 0.7282,
+      "step": 5570
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.2162140609274386,
+      "learning_rate": 3.857272268884353e-06,
+      "loss": 0.5941,
+      "step": 5571
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2617877144746505,
+      "learning_rate": 3.855613202814277e-06,
+      "loss": 0.6509,
+      "step": 5572
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.560892583789322,
+      "learning_rate": 3.853954269695749e-06,
+      "loss": 0.6604,
+      "step": 5573
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.487420247946007,
+      "learning_rate": 3.8522954697215034e-06,
+      "loss": 0.6302,
+      "step": 5574
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.821258946431581,
+      "learning_rate": 3.8506368030842525e-06,
+      "loss": 0.6311,
+      "step": 5575
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3313444270659462,
+      "learning_rate": 3.848978269976694e-06,
+      "loss": 0.5945,
+      "step": 5576
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.469073388544526,
+      "learning_rate": 3.8473198705915135e-06,
+      "loss": 0.7422,
+      "step": 5577
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.193626538455637,
+      "learning_rate": 3.845661605121377e-06,
+      "loss": 0.6453,
+      "step": 5578
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5193329581868897,
+      "learning_rate": 3.844003473758941e-06,
+      "loss": 0.6308,
+      "step": 5579
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.157883246431803,
+      "learning_rate": 3.8423454766968394e-06,
+      "loss": 0.6685,
+      "step": 5580
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.718500163349186,
+      "learning_rate": 3.8406876141276924e-06,
+      "loss": 0.5817,
+      "step": 5581
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3243517458977667,
+      "learning_rate": 3.8390298862441075e-06,
+      "loss": 0.7218,
+      "step": 5582
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.612911018329449,
+      "learning_rate": 3.8373722932386745e-06,
+      "loss": 0.6655,
+      "step": 5583
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.2278773536072167,
+      "learning_rate": 3.835714835303969e-06,
+      "loss": 0.6834,
+      "step": 5584
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6365356341700603,
+      "learning_rate": 3.834057512632546e-06,
+      "loss": 0.6425,
+      "step": 5585
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4664103802063475,
+      "learning_rate": 3.832400325416952e-06,
+      "loss": 0.5572,
+      "step": 5586
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0773321625767425,
+      "learning_rate": 3.830743273849713e-06,
+      "loss": 0.6554,
+      "step": 5587
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.630335903459665,
+      "learning_rate": 3.829086358123339e-06,
+      "loss": 0.5597,
+      "step": 5588
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.038054933842016,
+      "learning_rate": 3.82742957843033e-06,
+      "loss": 0.5566,
+      "step": 5589
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.81111225996943,
+      "learning_rate": 3.825772934963161e-06,
+      "loss": 0.7147,
+      "step": 5590
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.130623296542242,
+      "learning_rate": 3.824116427914298e-06,
+      "loss": 0.5917,
+      "step": 5591
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.22553668110589,
+      "learning_rate": 3.82246005747619e-06,
+      "loss": 0.5317,
+      "step": 5592
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.470012398893969,
+      "learning_rate": 3.82080382384127e-06,
+      "loss": 0.5783,
+      "step": 5593
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.112233161530673,
+      "learning_rate": 3.819147727201951e-06,
+      "loss": 0.625,
+      "step": 5594
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.4003860684934244,
+      "learning_rate": 3.817491767750635e-06,
+      "loss": 0.656,
+      "step": 5595
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6648181797893433,
+      "learning_rate": 3.815835945679709e-06,
+      "loss": 0.6308,
+      "step": 5596
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.488479873290116,
+      "learning_rate": 3.81418026118154e-06,
+      "loss": 0.5743,
+      "step": 5597
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.520362514666106,
+      "learning_rate": 3.8125247144484777e-06,
+      "loss": 0.5858,
+      "step": 5598
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.938146052846746,
+      "learning_rate": 3.8108693056728636e-06,
+      "loss": 0.6713,
+      "step": 5599
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.143207176889715,
+      "learning_rate": 3.809214035047016e-06,
+      "loss": 0.6833,
+      "step": 5600
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.205971327946951,
+      "learning_rate": 3.8075589027632376e-06,
+      "loss": 0.6633,
+      "step": 5601
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2680778570668756,
+      "learning_rate": 3.805903909013822e-06,
+      "loss": 0.5748,
+      "step": 5602
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.501171349333083,
+      "learning_rate": 3.804249053991037e-06,
+      "loss": 0.687,
+      "step": 5603
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.102343060157223,
+      "learning_rate": 3.8025943378871394e-06,
+      "loss": 0.6342,
+      "step": 5604
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.894149017249174,
+      "learning_rate": 3.800939760894371e-06,
+      "loss": 0.5931,
+      "step": 5605
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1805773899796685,
+      "learning_rate": 3.7992853232049566e-06,
+      "loss": 0.5769,
+      "step": 5606
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4362136260737426,
+      "learning_rate": 3.7976310250111013e-06,
+      "loss": 0.6543,
+      "step": 5607
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9225984516059653,
+      "learning_rate": 3.7959768665049967e-06,
+      "loss": 0.6982,
+      "step": 5608
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.86310538278994,
+      "learning_rate": 3.7943228478788198e-06,
+      "loss": 0.5438,
+      "step": 5609
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2140763934345613,
+      "learning_rate": 3.792668969324731e-06,
+      "loss": 0.5698,
+      "step": 5610
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5016813794002126,
+      "learning_rate": 3.7910152310348686e-06,
+      "loss": 0.7066,
+      "step": 5611
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1455037865154742,
+      "learning_rate": 3.789361633201363e-06,
+      "loss": 0.6227,
+      "step": 5612
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0200967442200817,
+      "learning_rate": 3.7877081760163225e-06,
+      "loss": 0.6153,
+      "step": 5613
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.386311762553055,
+      "learning_rate": 3.7860548596718427e-06,
+      "loss": 0.6632,
+      "step": 5614
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.264330176961888,
+      "learning_rate": 3.784401684360001e-06,
+      "loss": 0.6568,
+      "step": 5615
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.761814617242924,
+      "learning_rate": 3.782748650272857e-06,
+      "loss": 0.6565,
+      "step": 5616
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.076122600652747,
+      "learning_rate": 3.781095757602455e-06,
+      "loss": 0.6303,
+      "step": 5617
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3150873597883947,
+      "learning_rate": 3.779443006540825e-06,
+      "loss": 0.6462,
+      "step": 5618
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.940963699254998,
+      "learning_rate": 3.7777903972799794e-06,
+      "loss": 0.564,
+      "step": 5619
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4027103997644392,
+      "learning_rate": 3.7761379300119104e-06,
+      "loss": 0.61,
+      "step": 5620
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2952639954909526,
+      "learning_rate": 3.7744856049286e-06,
+      "loss": 0.5822,
+      "step": 5621
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.0792638125829637,
+      "learning_rate": 3.77283342222201e-06,
+      "loss": 0.6502,
+      "step": 5622
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9784622364316484,
+      "learning_rate": 3.7711813820840854e-06,
+      "loss": 0.61,
+      "step": 5623
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.0398501089056658,
+      "learning_rate": 3.7695294847067544e-06,
+      "loss": 0.641,
+      "step": 5624
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.801664578402942,
+      "learning_rate": 3.7678777302819314e-06,
+      "loss": 0.5783,
+      "step": 5625
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4911044008283074,
+      "learning_rate": 3.7662261190015116e-06,
+      "loss": 0.7241,
+      "step": 5626
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3853448216389594,
+      "learning_rate": 3.7645746510573754e-06,
+      "loss": 0.6626,
+      "step": 5627
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.131535326971549,
+      "learning_rate": 3.7629233266413866e-06,
+      "loss": 0.64,
+      "step": 5628
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9389020030909176,
+      "learning_rate": 3.7612721459453883e-06,
+      "loss": 0.5731,
+      "step": 5629
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7087185491547947,
+      "learning_rate": 3.75962110916121e-06,
+      "loss": 0.582,
+      "step": 5630
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.475217811511566,
+      "learning_rate": 3.757970216480667e-06,
+      "loss": 0.6372,
+      "step": 5631
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4037718481514765,
+      "learning_rate": 3.756319468095555e-06,
+      "loss": 0.6331,
+      "step": 5632
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.020609700875131,
+      "learning_rate": 3.7546688641976496e-06,
+      "loss": 0.6931,
+      "step": 5633
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.592025998510074,
+      "learning_rate": 3.753018404978717e-06,
+      "loss": 0.5209,
+      "step": 5634
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.501978181917583,
+      "learning_rate": 3.7513680906305015e-06,
+      "loss": 0.5572,
+      "step": 5635
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.965457135634429,
+      "learning_rate": 3.7497179213447305e-06,
+      "loss": 0.6254,
+      "step": 5636
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.294791076548485,
+      "learning_rate": 3.7480678973131198e-06,
+      "loss": 0.6553,
+      "step": 5637
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4361429134308956,
+      "learning_rate": 3.74641801872736e-06,
+      "loss": 0.645,
+      "step": 5638
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.229172173470131,
+      "learning_rate": 3.7447682857791307e-06,
+      "loss": 0.6655,
+      "step": 5639
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5927694332695186,
+      "learning_rate": 3.743118698660094e-06,
+      "loss": 0.5953,
+      "step": 5640
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.419854503906456,
+      "learning_rate": 3.741469257561895e-06,
+      "loss": 0.6415,
+      "step": 5641
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.454076430219235,
+      "learning_rate": 3.739819962676159e-06,
+      "loss": 0.677,
+      "step": 5642
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.333031240171135,
+      "learning_rate": 3.738170814194495e-06,
+      "loss": 0.5559,
+      "step": 5643
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.427282486291471,
+      "learning_rate": 3.7365218123084996e-06,
+      "loss": 0.6336,
+      "step": 5644
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.692927524150578,
+      "learning_rate": 3.7348729572097487e-06,
+      "loss": 0.5909,
+      "step": 5645
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.406995986734407,
+      "learning_rate": 3.7332242490897985e-06,
+      "loss": 0.6221,
+      "step": 5646
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.735312603120759,
+      "learning_rate": 3.7315756881401944e-06,
+      "loss": 0.6724,
+      "step": 5647
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.4788360617314944,
+      "learning_rate": 3.7299272745524583e-06,
+      "loss": 0.634,
+      "step": 5648
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.280769795287397,
+      "learning_rate": 3.728279008518102e-06,
+      "loss": 0.5864,
+      "step": 5649
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7064070384851253,
+      "learning_rate": 3.726630890228615e-06,
+      "loss": 0.6402,
+      "step": 5650
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.801131241287733,
+      "learning_rate": 3.7249829198754694e-06,
+      "loss": 0.6281,
+      "step": 5651
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.172639276357326,
+      "learning_rate": 3.7233350976501217e-06,
+      "loss": 0.6297,
+      "step": 5652
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9375269933216765,
+      "learning_rate": 3.7216874237440127e-06,
+      "loss": 0.5696,
+      "step": 5653
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3229567583419217,
+      "learning_rate": 3.7200398983485643e-06,
+      "loss": 0.6104,
+      "step": 5654
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.523904866991665,
+      "learning_rate": 3.7183925216551784e-06,
+      "loss": 0.6879,
+      "step": 5655
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1913564898938622,
+      "learning_rate": 3.716745293855246e-06,
+      "loss": 0.5874,
+      "step": 5656
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2827810373182724,
+      "learning_rate": 3.715098215140136e-06,
+      "loss": 0.6078,
+      "step": 5657
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.128545296474594,
+      "learning_rate": 3.7134512857012017e-06,
+      "loss": 0.6683,
+      "step": 5658
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.9857064043435133,
+      "learning_rate": 3.711804505729776e-06,
+      "loss": 0.5813,
+      "step": 5659
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.632396638894393,
+      "learning_rate": 3.7101578754171797e-06,
+      "loss": 0.6517,
+      "step": 5660
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2170898683792815,
+      "learning_rate": 3.7085113949547126e-06,
+      "loss": 0.6747,
+      "step": 5661
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1974382543005357,
+      "learning_rate": 3.706865064533659e-06,
+      "loss": 0.6875,
+      "step": 5662
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.3819380364992115,
+      "learning_rate": 3.7052188843452854e-06,
+      "loss": 0.622,
+      "step": 5663
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.682752126582236,
+      "learning_rate": 3.7035728545808367e-06,
+      "loss": 0.6522,
+      "step": 5664
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.042183074281712,
+      "learning_rate": 3.701926975431547e-06,
+      "loss": 0.6193,
+      "step": 5665
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0167480891393266,
+      "learning_rate": 3.700281247088629e-06,
+      "loss": 0.5478,
+      "step": 5666
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.190095939887631,
+      "learning_rate": 3.6986356697432796e-06,
+      "loss": 0.6033,
+      "step": 5667
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3881960936207682,
+      "learning_rate": 3.6969902435866743e-06,
+      "loss": 0.5799,
+      "step": 5668
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4289590644832786,
+      "learning_rate": 3.6953449688099774e-06,
+      "loss": 0.7183,
+      "step": 5669
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.15648426463186,
+      "learning_rate": 3.69369984560433e-06,
+      "loss": 0.6219,
+      "step": 5670
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.679139642427304,
+      "learning_rate": 3.69205487416086e-06,
+      "loss": 0.6284,
+      "step": 5671
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.199487911029984,
+      "learning_rate": 3.690410054670671e-06,
+      "loss": 0.6225,
+      "step": 5672
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.927509209876508,
+      "learning_rate": 3.6887653873248575e-06,
+      "loss": 0.6316,
+      "step": 5673
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.9002460059852595,
+      "learning_rate": 3.6871208723144903e-06,
+      "loss": 0.6826,
+      "step": 5674
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2930525319126507,
+      "learning_rate": 3.6854765098306254e-06,
+      "loss": 0.5643,
+      "step": 5675
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.809686719399776,
+      "learning_rate": 3.6838323000643013e-06,
+      "loss": 0.6805,
+      "step": 5676
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6784212261701166,
+      "learning_rate": 3.682188243206535e-06,
+      "loss": 0.6072,
+      "step": 5677
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1335859555703096,
+      "learning_rate": 3.6805443394483275e-06,
+      "loss": 0.5365,
+      "step": 5678
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.477277761969307,
+      "learning_rate": 3.6789005889806656e-06,
+      "loss": 0.7198,
+      "step": 5679
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1807624837158626,
+      "learning_rate": 3.6772569919945157e-06,
+      "loss": 0.6629,
+      "step": 5680
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.278161138807236,
+      "learning_rate": 3.6756135486808227e-06,
+      "loss": 0.5702,
+      "step": 5681
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.433937775862514,
+      "learning_rate": 3.6739702592305205e-06,
+      "loss": 0.6464,
+      "step": 5682
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.0575854854067104,
+      "learning_rate": 3.6723271238345187e-06,
+      "loss": 0.674,
+      "step": 5683
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.607667973961037,
+      "learning_rate": 3.6706841426837145e-06,
+      "loss": 0.6982,
+      "step": 5684
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3855563143332543,
+      "learning_rate": 3.669041315968986e-06,
+      "loss": 0.6425,
+      "step": 5685
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.551518686262234,
+      "learning_rate": 3.667398643881189e-06,
+      "loss": 0.6379,
+      "step": 5686
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2344906686707855,
+      "learning_rate": 3.665756126611164e-06,
+      "loss": 0.6542,
+      "step": 5687
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.892933147579963,
+      "learning_rate": 3.664113764349736e-06,
+      "loss": 0.6753,
+      "step": 5688
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.899716923671387,
+      "learning_rate": 3.6624715572877106e-06,
+      "loss": 0.5615,
+      "step": 5689
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.52564711595407,
+      "learning_rate": 3.6608295056158717e-06,
+      "loss": 0.6426,
+      "step": 5690
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.2313531578252412,
+      "learning_rate": 3.65918760952499e-06,
+      "loss": 0.5662,
+      "step": 5691
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5379514454376904,
+      "learning_rate": 3.657545869205816e-06,
+      "loss": 0.5762,
+      "step": 5692
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.9538894091444807,
+      "learning_rate": 3.6559042848490835e-06,
+      "loss": 0.5906,
+      "step": 5693
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6382651155593604,
+      "learning_rate": 3.654262856645503e-06,
+      "loss": 0.6264,
+      "step": 5694
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.482444707821686,
+      "learning_rate": 3.652621584785776e-06,
+      "loss": 0.6596,
+      "step": 5695
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2566669898698146,
+      "learning_rate": 3.6509804694605768e-06,
+      "loss": 0.7118,
+      "step": 5696
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.5669855228140213,
+      "learning_rate": 3.649339510860568e-06,
+      "loss": 0.5655,
+      "step": 5697
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3661369567618866,
+      "learning_rate": 3.647698709176391e-06,
+      "loss": 0.6338,
+      "step": 5698
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3484815033901807,
+      "learning_rate": 3.6460580645986685e-06,
+      "loss": 0.6255,
+      "step": 5699
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.53755595996656,
+      "learning_rate": 3.6444175773180045e-06,
+      "loss": 0.5636,
+      "step": 5700
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5222433345809763,
+      "learning_rate": 3.6427772475249896e-06,
+      "loss": 0.6518,
+      "step": 5701
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.586007819868345,
+      "learning_rate": 3.6411370754101915e-06,
+      "loss": 0.6029,
+      "step": 5702
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.428940429081987,
+      "learning_rate": 3.639497061164158e-06,
+      "loss": 0.5528,
+      "step": 5703
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.60801073186904,
+      "learning_rate": 3.637857204977424e-06,
+      "loss": 0.5935,
+      "step": 5704
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.8916382563951037,
+      "learning_rate": 3.636217507040502e-06,
+      "loss": 0.6435,
+      "step": 5705
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9863942243849351,
+      "learning_rate": 3.6345779675438897e-06,
+      "loss": 0.5717,
+      "step": 5706
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.47039659299787,
+      "learning_rate": 3.6329385866780587e-06,
+      "loss": 0.5755,
+      "step": 5707
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7898417007114333,
+      "learning_rate": 3.6312993646334727e-06,
+      "loss": 0.5874,
+      "step": 5708
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.273917598473412,
+      "learning_rate": 3.6296603016005693e-06,
+      "loss": 0.6866,
+      "step": 5709
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6107303675720432,
+      "learning_rate": 3.6280213977697715e-06,
+      "loss": 0.6771,
+      "step": 5710
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2956736561042357,
+      "learning_rate": 3.6263826533314827e-06,
+      "loss": 0.6167,
+      "step": 5711
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.384801633397614,
+      "learning_rate": 3.624744068476086e-06,
+      "loss": 0.6664,
+      "step": 5712
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0974504782665337,
+      "learning_rate": 3.623105643393946e-06,
+      "loss": 0.6318,
+      "step": 5713
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.3889022639258015,
+      "learning_rate": 3.621467378275414e-06,
+      "loss": 0.6735,
+      "step": 5714
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.18791054241475,
+      "learning_rate": 3.6198292733108177e-06,
+      "loss": 0.6781,
+      "step": 5715
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7589589062878197,
+      "learning_rate": 3.6181913286904647e-06,
+      "loss": 0.5676,
+      "step": 5716
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9737675399727652,
+      "learning_rate": 3.6165535446046497e-06,
+      "loss": 0.6197,
+      "step": 5717
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.520077841376911,
+      "learning_rate": 3.6149159212436435e-06,
+      "loss": 0.5949,
+      "step": 5718
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2678271084992514,
+      "learning_rate": 3.6132784587977053e-06,
+      "loss": 0.5782,
+      "step": 5719
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.736740775923461,
+      "learning_rate": 3.611641157457064e-06,
+      "loss": 0.5897,
+      "step": 5720
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.503689546593751,
+      "learning_rate": 3.6100040174119403e-06,
+      "loss": 0.6358,
+      "step": 5721
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9542516668761488,
+      "learning_rate": 3.6083670388525316e-06,
+      "loss": 0.6144,
+      "step": 5722
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.151965071313513,
+      "learning_rate": 3.6067302219690175e-06,
+      "loss": 0.6571,
+      "step": 5723
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.083294772175323,
+      "learning_rate": 3.6050935669515604e-06,
+      "loss": 0.6405,
+      "step": 5724
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.308319919656163,
+      "learning_rate": 3.603457073990298e-06,
+      "loss": 0.5927,
+      "step": 5725
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5585087791382075,
+      "learning_rate": 3.6018207432753572e-06,
+      "loss": 0.5503,
+      "step": 5726
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 14.380449615576731,
+      "learning_rate": 3.60018457499684e-06,
+      "loss": 0.6287,
+      "step": 5727
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.450787029376905,
+      "learning_rate": 3.598548569344834e-06,
+      "loss": 0.6453,
+      "step": 5728
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.1421747068388366,
+      "learning_rate": 3.596912726509402e-06,
+      "loss": 0.7465,
+      "step": 5729
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.304462145687152,
+      "learning_rate": 3.595277046680594e-06,
+      "loss": 0.6277,
+      "step": 5730
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.179794005325396,
+      "learning_rate": 3.5936415300484383e-06,
+      "loss": 0.6019,
+      "step": 5731
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0233482875110236,
+      "learning_rate": 3.592006176802944e-06,
+      "loss": 0.5521,
+      "step": 5732
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2805038648763887,
+      "learning_rate": 3.5903709871341034e-06,
+      "loss": 0.6422,
+      "step": 5733
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7182569922936426,
+      "learning_rate": 3.5887359612318862e-06,
+      "loss": 0.5947,
+      "step": 5734
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.992203373180557,
+      "learning_rate": 3.5871010992862436e-06,
+      "loss": 0.6566,
+      "step": 5735
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.1660591900577493,
+      "learning_rate": 3.5854664014871128e-06,
+      "loss": 0.6317,
+      "step": 5736
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.645200532572237,
+      "learning_rate": 3.5838318680244067e-06,
+      "loss": 0.6629,
+      "step": 5737
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2006735106420225,
+      "learning_rate": 3.582197499088019e-06,
+      "loss": 0.7078,
+      "step": 5738
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7870416290233466,
+      "learning_rate": 3.580563294867828e-06,
+      "loss": 0.6616,
+      "step": 5739
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2284086143558057,
+      "learning_rate": 3.5789292555536907e-06,
+      "loss": 0.644,
+      "step": 5740
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0680596768407415,
+      "learning_rate": 3.5772953813354455e-06,
+      "loss": 0.5917,
+      "step": 5741
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6380941942532816,
+      "learning_rate": 3.575661672402908e-06,
+      "loss": 0.5598,
+      "step": 5742
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1088238905647647,
+      "learning_rate": 3.5740281289458812e-06,
+      "loss": 0.6548,
+      "step": 5743
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.364337376015603,
+      "learning_rate": 3.5723947511541435e-06,
+      "loss": 0.6301,
+      "step": 5744
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.155417858599172,
+      "learning_rate": 3.5707615392174576e-06,
+      "loss": 0.6011,
+      "step": 5745
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.789662978157353,
+      "learning_rate": 3.5691284933255653e-06,
+      "loss": 0.6223,
+      "step": 5746
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.436792685729838,
+      "learning_rate": 3.567495613668188e-06,
+      "loss": 0.6495,
+      "step": 5747
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2959050511929493,
+      "learning_rate": 3.565862900435028e-06,
+      "loss": 0.6366,
+      "step": 5748
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.0678303289148188,
+      "learning_rate": 3.564230353815772e-06,
+      "loss": 0.5961,
+      "step": 5749
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2062343340152553,
+      "learning_rate": 3.562597974000084e-06,
+      "loss": 0.5445,
+      "step": 5750
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1459202020440926,
+      "learning_rate": 3.5609657611776055e-06,
+      "loss": 0.5367,
+      "step": 5751
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.892508323233743,
+      "learning_rate": 3.5593337155379663e-06,
+      "loss": 0.5296,
+      "step": 5752
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 13.502367015161873,
+      "learning_rate": 3.5577018372707706e-06,
+      "loss": 0.5599,
+      "step": 5753
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.60449934352309,
+      "learning_rate": 3.5560701265656096e-06,
+      "loss": 0.5771,
+      "step": 5754
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0330026690570517,
+      "learning_rate": 3.5544385836120445e-06,
+      "loss": 0.6543,
+      "step": 5755
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.246716338488602,
+      "learning_rate": 3.552807208599626e-06,
+      "loss": 0.6199,
+      "step": 5756
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.880693301049812,
+      "learning_rate": 3.551176001717882e-06,
+      "loss": 0.5812,
+      "step": 5757
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.374133819647502,
+      "learning_rate": 3.549544963156324e-06,
+      "loss": 0.6039,
+      "step": 5758
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.651695196080499,
+      "learning_rate": 3.5479140931044393e-06,
+      "loss": 0.6026,
+      "step": 5759
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.256349484082301,
+      "learning_rate": 3.546283391751696e-06,
+      "loss": 0.5949,
+      "step": 5760
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2426448275819504,
+      "learning_rate": 3.5446528592875464e-06,
+      "loss": 0.6444,
+      "step": 5761
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0973705329694723,
+      "learning_rate": 3.5430224959014215e-06,
+      "loss": 0.7379,
+      "step": 5762
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2641582085188,
+      "learning_rate": 3.5413923017827317e-06,
+      "loss": 0.5749,
+      "step": 5763
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.058982375661702,
+      "learning_rate": 3.5397622771208663e-06,
+      "loss": 0.5927,
+      "step": 5764
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.998975467492036,
+      "learning_rate": 3.5381324221051995e-06,
+      "loss": 0.611,
+      "step": 5765
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.436460113249485,
+      "learning_rate": 3.5365027369250804e-06,
+      "loss": 0.5845,
+      "step": 5766
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.336543180347027,
+      "learning_rate": 3.5348732217698466e-06,
+      "loss": 0.6229,
+      "step": 5767
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.005942928681562,
+      "learning_rate": 3.533243876828803e-06,
+      "loss": 0.6442,
+      "step": 5768
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8740853995553226,
+      "learning_rate": 3.531614702291247e-06,
+      "loss": 0.5655,
+      "step": 5769
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 25.902005172145653,
+      "learning_rate": 3.5299856983464497e-06,
+      "loss": 0.6208,
+      "step": 5770
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.499670510652518,
+      "learning_rate": 3.528356865183665e-06,
+      "loss": 0.5878,
+      "step": 5771
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5307387633144414,
+      "learning_rate": 3.526728202992127e-06,
+      "loss": 0.7085,
+      "step": 5772
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.510774660981912,
+      "learning_rate": 3.525099711961045e-06,
+      "loss": 0.6365,
+      "step": 5773
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6261622750991034,
+      "learning_rate": 3.523471392279616e-06,
+      "loss": 0.6566,
+      "step": 5774
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.445814884829749,
+      "learning_rate": 3.521843244137013e-06,
+      "loss": 0.7112,
+      "step": 5775
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.355458478806883,
+      "learning_rate": 3.52021526772239e-06,
+      "loss": 0.6402,
+      "step": 5776
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3733230146293263,
+      "learning_rate": 3.5185874632248775e-06,
+      "loss": 0.6198,
+      "step": 5777
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8989292888055886,
+      "learning_rate": 3.5169598308335915e-06,
+      "loss": 0.592,
+      "step": 5778
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1665097816063295,
+      "learning_rate": 3.515332370737625e-06,
+      "loss": 0.669,
+      "step": 5779
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.005926002720759,
+      "learning_rate": 3.513705083126054e-06,
+      "loss": 0.6319,
+      "step": 5780
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.336459256200082,
+      "learning_rate": 3.5120779681879286e-06,
+      "loss": 0.6206,
+      "step": 5781
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.298134036866341,
+      "learning_rate": 3.5104510261122836e-06,
+      "loss": 0.6818,
+      "step": 5782
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8034284235986724,
+      "learning_rate": 3.508824257088132e-06,
+      "loss": 0.653,
+      "step": 5783
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.123317385052922,
+      "learning_rate": 3.507197661304469e-06,
+      "loss": 0.6499,
+      "step": 5784
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.655862388540164,
+      "learning_rate": 3.505571238950267e-06,
+      "loss": 0.5429,
+      "step": 5785
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.490391865079292,
+      "learning_rate": 3.5039449902144763e-06,
+      "loss": 0.6385,
+      "step": 5786
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.384431381049101,
+      "learning_rate": 3.5023189152860325e-06,
+      "loss": 0.7115,
+      "step": 5787
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6398628266384607,
+      "learning_rate": 3.5006930143538477e-06,
+      "loss": 0.6195,
+      "step": 5788
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.043162425725876,
+      "learning_rate": 3.499067287606817e-06,
+      "loss": 0.5859,
+      "step": 5789
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.925176048693361,
+      "learning_rate": 3.4974417352338074e-06,
+      "loss": 0.6185,
+      "step": 5790
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3132796544844303,
+      "learning_rate": 3.495816357423674e-06,
+      "loss": 0.6343,
+      "step": 5791
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2778784069626004,
+      "learning_rate": 3.494191154365247e-06,
+      "loss": 0.6054,
+      "step": 5792
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7460899113803503,
+      "learning_rate": 3.492566126247341e-06,
+      "loss": 0.6063,
+      "step": 5793
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.201639636764578,
+      "learning_rate": 3.4909412732587444e-06,
+      "loss": 0.6561,
+      "step": 5794
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.547089477361622,
+      "learning_rate": 3.4893165955882275e-06,
+      "loss": 0.5957,
+      "step": 5795
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.598764658378845,
+      "learning_rate": 3.4876920934245423e-06,
+      "loss": 0.6941,
+      "step": 5796
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3890690610014795,
+      "learning_rate": 3.486067766956418e-06,
+      "loss": 0.6066,
+      "step": 5797
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.9128726002118785,
+      "learning_rate": 3.4844436163725648e-06,
+      "loss": 0.5642,
+      "step": 5798
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.8378397404411326,
+      "learning_rate": 3.48281964186167e-06,
+      "loss": 0.6248,
+      "step": 5799
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3946114293745913,
+      "learning_rate": 3.4811958436124036e-06,
+      "loss": 0.6376,
+      "step": 5800
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.903630763714226,
+      "learning_rate": 3.479572221813413e-06,
+      "loss": 0.5643,
+      "step": 5801
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.522387352098266,
+      "learning_rate": 3.4779487766533306e-06,
+      "loss": 0.6095,
+      "step": 5802
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.539618765581498,
+      "learning_rate": 3.4763255083207547e-06,
+      "loss": 0.5834,
+      "step": 5803
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.498863654137628,
+      "learning_rate": 3.4747024170042785e-06,
+      "loss": 0.6581,
+      "step": 5804
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0699886181482836,
+      "learning_rate": 3.473079502892466e-06,
+      "loss": 0.686,
+      "step": 5805
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.56411416397428,
+      "learning_rate": 3.4714567661738635e-06,
+      "loss": 0.6205,
+      "step": 5806
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.6804375443854056,
+      "learning_rate": 3.469834207036996e-06,
+      "loss": 0.5698,
+      "step": 5807
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3140301754849704,
+      "learning_rate": 3.4682118256703657e-06,
+      "loss": 0.6665,
+      "step": 5808
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.132346843122038,
+      "learning_rate": 3.4665896222624585e-06,
+      "loss": 0.6473,
+      "step": 5809
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1626805670201685,
+      "learning_rate": 3.4649675970017355e-06,
+      "loss": 0.5748,
+      "step": 5810
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6486782602012684,
+      "learning_rate": 3.4633457500766413e-06,
+      "loss": 0.6031,
+      "step": 5811
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5370141596453624,
+      "learning_rate": 3.4617240816755937e-06,
+      "loss": 0.5877,
+      "step": 5812
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5534598070476,
+      "learning_rate": 3.460102591986997e-06,
+      "loss": 0.6301,
+      "step": 5813
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.3454701642307962,
+      "learning_rate": 3.4584812811992287e-06,
+      "loss": 0.6339,
+      "step": 5814
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0095813678630763,
+      "learning_rate": 3.4568601495006503e-06,
+      "loss": 0.6619,
+      "step": 5815
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4805455518296835,
+      "learning_rate": 3.4552391970795984e-06,
+      "loss": 0.6543,
+      "step": 5816
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.491050633881681,
+      "learning_rate": 3.453618424124392e-06,
+      "loss": 0.6431,
+      "step": 5817
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.733199748643421,
+      "learning_rate": 3.451997830823325e-06,
+      "loss": 0.5999,
+      "step": 5818
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0026297623529605,
+      "learning_rate": 3.4503774173646767e-06,
+      "loss": 0.5972,
+      "step": 5819
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.1106086481537787,
+      "learning_rate": 3.448757183936701e-06,
+      "loss": 0.6202,
+      "step": 5820
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.983654597238341,
+      "learning_rate": 3.447137130727629e-06,
+      "loss": 0.6625,
+      "step": 5821
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.81371149233971,
+      "learning_rate": 3.4455172579256784e-06,
+      "loss": 0.6177,
+      "step": 5822
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.458016439088697,
+      "learning_rate": 3.4438975657190375e-06,
+      "loss": 0.7018,
+      "step": 5823
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9033971357544983,
+      "learning_rate": 3.442278054295883e-06,
+      "loss": 0.5718,
+      "step": 5824
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.9518489778855823,
+      "learning_rate": 3.440658723844358e-06,
+      "loss": 0.6359,
+      "step": 5825
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6141808916194367,
+      "learning_rate": 3.439039574552595e-06,
+      "loss": 0.5865,
+      "step": 5826
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2429938266365808,
+      "learning_rate": 3.437420606608701e-06,
+      "loss": 0.6993,
+      "step": 5827
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1988445058564614,
+      "learning_rate": 3.435801820200767e-06,
+      "loss": 0.6538,
+      "step": 5828
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.093978386990422,
+      "learning_rate": 3.4341832155168547e-06,
+      "loss": 0.6003,
+      "step": 5829
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.311995229831535,
+      "learning_rate": 3.432564792745009e-06,
+      "loss": 0.6298,
+      "step": 5830
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.324588374264805,
+      "learning_rate": 3.4309465520732556e-06,
+      "loss": 0.5572,
+      "step": 5831
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0044411358111915,
+      "learning_rate": 3.4293284936895956e-06,
+      "loss": 0.548,
+      "step": 5832
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.285667049229145,
+      "learning_rate": 3.4277106177820123e-06,
+      "loss": 0.6496,
+      "step": 5833
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.389684858763753,
+      "learning_rate": 3.426092924538462e-06,
+      "loss": 0.654,
+      "step": 5834
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2261660392871327,
+      "learning_rate": 3.4244754141468878e-06,
+      "loss": 0.7062,
+      "step": 5835
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.237433451917938,
+      "learning_rate": 3.4228580867952044e-06,
+      "loss": 0.6249,
+      "step": 5836
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0863620341722386,
+      "learning_rate": 3.421240942671312e-06,
+      "loss": 0.6019,
+      "step": 5837
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5102635909640636,
+      "learning_rate": 3.4196239819630806e-06,
+      "loss": 0.5542,
+      "step": 5838
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.377771418582857,
+      "learning_rate": 3.4180072048583667e-06,
+      "loss": 0.7179,
+      "step": 5839
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8495071063526494,
+      "learning_rate": 3.4163906115450025e-06,
+      "loss": 0.6351,
+      "step": 5840
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.504564584687121,
+      "learning_rate": 3.4147742022108e-06,
+      "loss": 0.5398,
+      "step": 5841
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7848784093251355,
+      "learning_rate": 3.4131579770435495e-06,
+      "loss": 0.67,
+      "step": 5842
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7193060597973844,
+      "learning_rate": 3.411541936231016e-06,
+      "loss": 0.6779,
+      "step": 5843
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.9893759030534155,
+      "learning_rate": 3.409926079960949e-06,
+      "loss": 0.5734,
+      "step": 5844
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.402119046084388,
+      "learning_rate": 3.4083104084210746e-06,
+      "loss": 0.7261,
+      "step": 5845
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.2792238487813976,
+      "learning_rate": 3.4066949217990964e-06,
+      "loss": 0.6428,
+      "step": 5846
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.754234915442587,
+      "learning_rate": 3.4050796202826943e-06,
+      "loss": 0.6201,
+      "step": 5847
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5663658021698423,
+      "learning_rate": 3.4034645040595325e-06,
+      "loss": 0.6867,
+      "step": 5848
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.808416018507515,
+      "learning_rate": 3.4018495733172485e-06,
+      "loss": 0.7043,
+      "step": 5849
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.945233847925162,
+      "learning_rate": 3.4002348282434637e-06,
+      "loss": 0.5994,
+      "step": 5850
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.242388743778958,
+      "learning_rate": 3.3986202690257707e-06,
+      "loss": 0.689,
+      "step": 5851
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.706078895924368,
+      "learning_rate": 3.397005895851746e-06,
+      "loss": 0.6305,
+      "step": 5852
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4900554279730427,
+      "learning_rate": 3.395391708908941e-06,
+      "loss": 0.6242,
+      "step": 5853
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.2235848293555365,
+      "learning_rate": 3.39377770838489e-06,
+      "loss": 0.6205,
+      "step": 5854
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1780532875830025,
+      "learning_rate": 3.392163894467103e-06,
+      "loss": 0.6479,
+      "step": 5855
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.180928205646365,
+      "learning_rate": 3.3905502673430648e-06,
+      "loss": 0.6474,
+      "step": 5856
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0354934950603507,
+      "learning_rate": 3.3889368272002455e-06,
+      "loss": 0.6783,
+      "step": 5857
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 13.984789819926613,
+      "learning_rate": 3.387323574226087e-06,
+      "loss": 0.6263,
+      "step": 5858
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.841089761628964,
+      "learning_rate": 3.385710508608017e-06,
+      "loss": 0.6473,
+      "step": 5859
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.189754673822482,
+      "learning_rate": 3.3840976305334304e-06,
+      "loss": 0.5959,
+      "step": 5860
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5889741809528988,
+      "learning_rate": 3.382484940189711e-06,
+      "loss": 0.6292,
+      "step": 5861
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.289269598614149,
+      "learning_rate": 3.380872437764215e-06,
+      "loss": 0.6444,
+      "step": 5862
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.422036993894899,
+      "learning_rate": 3.37926012344428e-06,
+      "loss": 0.5488,
+      "step": 5863
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.456124041276871,
+      "learning_rate": 3.3776479974172184e-06,
+      "loss": 0.6358,
+      "step": 5864
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4906368084019817,
+      "learning_rate": 3.3760360598703217e-06,
+      "loss": 0.5991,
+      "step": 5865
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6266749863791805,
+      "learning_rate": 3.374424310990862e-06,
+      "loss": 0.6148,
+      "step": 5866
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.513373641942084,
+      "learning_rate": 3.372812750966087e-06,
+      "loss": 0.6433,
+      "step": 5867
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.884160237208583,
+      "learning_rate": 3.371201379983223e-06,
+      "loss": 0.5771,
+      "step": 5868
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.2671886681135485,
+      "learning_rate": 3.369590198229473e-06,
+      "loss": 0.5949,
+      "step": 5869
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2010932339105223,
+      "learning_rate": 3.3679792058920223e-06,
+      "loss": 0.677,
+      "step": 5870
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8453784039703374,
+      "learning_rate": 3.366368403158028e-06,
+      "loss": 0.604,
+      "step": 5871
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.641273584480205,
+      "learning_rate": 3.3647577902146334e-06,
+      "loss": 0.6553,
+      "step": 5872
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3540207447949997,
+      "learning_rate": 3.363147367248949e-06,
+      "loss": 0.6874,
+      "step": 5873
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1061808611226644,
+      "learning_rate": 3.3615371344480725e-06,
+      "loss": 0.5198,
+      "step": 5874
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.610784500535765,
+      "learning_rate": 3.3599270919990744e-06,
+      "loss": 0.6982,
+      "step": 5875
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 17.35217476906217,
+      "learning_rate": 3.358317240089008e-06,
+      "loss": 0.6382,
+      "step": 5876
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3499225456097874,
+      "learning_rate": 3.3567075789048973e-06,
+      "loss": 0.6571,
+      "step": 5877
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.846781393555933,
+      "learning_rate": 3.35509810863375e-06,
+      "loss": 0.7104,
+      "step": 5878
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0399172317674497,
+      "learning_rate": 3.35348882946255e-06,
+      "loss": 0.6536,
+      "step": 5879
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.2181809417118203,
+      "learning_rate": 3.3518797415782577e-06,
+      "loss": 0.5946,
+      "step": 5880
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.7651214632668175,
+      "learning_rate": 3.3502708451678145e-06,
+      "loss": 0.6062,
+      "step": 5881
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5978846672711,
+      "learning_rate": 3.348662140418133e-06,
+      "loss": 0.6363,
+      "step": 5882
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.165618488483955,
+      "learning_rate": 3.3470536275161126e-06,
+      "loss": 0.5837,
+      "step": 5883
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.6883277185410073,
+      "learning_rate": 3.3454453066486214e-06,
+      "loss": 0.6308,
+      "step": 5884
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1630804016878504,
+      "learning_rate": 3.3438371780025136e-06,
+      "loss": 0.6886,
+      "step": 5885
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.269499258419974,
+      "learning_rate": 3.342229241764615e-06,
+      "loss": 0.6706,
+      "step": 5886
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3398677700347497,
+      "learning_rate": 3.34062149812173e-06,
+      "loss": 0.6235,
+      "step": 5887
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.507757222319039,
+      "learning_rate": 3.339013947260642e-06,
+      "loss": 0.5739,
+      "step": 5888
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.9578807377909313,
+      "learning_rate": 3.337406589368113e-06,
+      "loss": 0.6063,
+      "step": 5889
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2486291156021263,
+      "learning_rate": 3.3357994246308815e-06,
+      "loss": 0.6258,
+      "step": 5890
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4560317757961596,
+      "learning_rate": 3.3341924532356605e-06,
+      "loss": 0.6704,
+      "step": 5891
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4000882896893856,
+      "learning_rate": 3.3325856753691453e-06,
+      "loss": 0.6653,
+      "step": 5892
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.309148094791573,
+      "learning_rate": 3.3309790912180056e-06,
+      "loss": 0.5464,
+      "step": 5893
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.9793468104494654,
+      "learning_rate": 3.329372700968894e-06,
+      "loss": 0.6522,
+      "step": 5894
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.234552857190179,
+      "learning_rate": 3.3277665048084283e-06,
+      "loss": 0.6763,
+      "step": 5895
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.305618418961012,
+      "learning_rate": 3.326160502923218e-06,
+      "loss": 0.5466,
+      "step": 5896
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4393905813677113,
+      "learning_rate": 3.324554695499841e-06,
+      "loss": 0.6869,
+      "step": 5897
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.1678663773433686,
+      "learning_rate": 3.3229490827248585e-06,
+      "loss": 0.6572,
+      "step": 5898
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.810314446746764,
+      "learning_rate": 3.3213436647848017e-06,
+      "loss": 0.6039,
+      "step": 5899
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.616553362514333,
+      "learning_rate": 3.3197384418661854e-06,
+      "loss": 0.6404,
+      "step": 5900
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.606456810157447,
+      "learning_rate": 3.3181334141555003e-06,
+      "loss": 0.6813,
+      "step": 5901
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.352922919680765,
+      "learning_rate": 3.3165285818392133e-06,
+      "loss": 0.5888,
+      "step": 5902
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.765095323296954,
+      "learning_rate": 3.3149239451037706e-06,
+      "loss": 0.6115,
+      "step": 5903
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0882597233379716,
+      "learning_rate": 3.313319504135591e-06,
+      "loss": 0.6357,
+      "step": 5904
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3840323224262576,
+      "learning_rate": 3.3117152591210765e-06,
+      "loss": 0.6076,
+      "step": 5905
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 12.879550786415358,
+      "learning_rate": 3.3101112102466014e-06,
+      "loss": 0.6847,
+      "step": 5906
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9258667628138425,
+      "learning_rate": 3.308507357698525e-06,
+      "loss": 0.5716,
+      "step": 5907
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.468162429286832,
+      "learning_rate": 3.30690370166317e-06,
+      "loss": 0.6944,
+      "step": 5908
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.4587947376198205,
+      "learning_rate": 3.30530024232685e-06,
+      "loss": 0.6611,
+      "step": 5909
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7167118186662447,
+      "learning_rate": 3.3036969798758486e-06,
+      "loss": 0.6919,
+      "step": 5910
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8540713326378624,
+      "learning_rate": 3.3020939144964298e-06,
+      "loss": 0.5524,
+      "step": 5911
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.633945852372088,
+      "learning_rate": 3.3004910463748315e-06,
+      "loss": 0.6379,
+      "step": 5912
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6471695256154977,
+      "learning_rate": 3.298888375697269e-06,
+      "loss": 0.6573,
+      "step": 5913
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7889126114993785,
+      "learning_rate": 3.2972859026499395e-06,
+      "loss": 0.5756,
+      "step": 5914
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4597559748985436,
+      "learning_rate": 3.2956836274190107e-06,
+      "loss": 0.6519,
+      "step": 5915
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.22590279694477,
+      "learning_rate": 3.294081550190633e-06,
+      "loss": 0.5535,
+      "step": 5916
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.1209221601644255,
+      "learning_rate": 3.292479671150927e-06,
+      "loss": 0.6076,
+      "step": 5917
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6511690459462853,
+      "learning_rate": 3.290877990485999e-06,
+      "loss": 0.5224,
+      "step": 5918
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.017349317050653,
+      "learning_rate": 3.2892765083819233e-06,
+      "loss": 0.515,
+      "step": 5919
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7093325400474417,
+      "learning_rate": 3.28767522502476e-06,
+      "loss": 0.6475,
+      "step": 5920
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.471371183047004,
+      "learning_rate": 3.2860741406005383e-06,
+      "loss": 0.5238,
+      "step": 5921
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6312257555514824,
+      "learning_rate": 3.2844732552952686e-06,
+      "loss": 0.6347,
+      "step": 5922
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.411758053136448,
+      "learning_rate": 3.282872569294936e-06,
+      "loss": 0.6182,
+      "step": 5923
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5478750822742935,
+      "learning_rate": 3.281272082785506e-06,
+      "loss": 0.481,
+      "step": 5924
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3428946425529644,
+      "learning_rate": 3.2796717959529167e-06,
+      "loss": 0.6574,
+      "step": 5925
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5726580568876902,
+      "learning_rate": 3.2780717089830845e-06,
+      "loss": 0.6016,
+      "step": 5926
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.284631597557582,
+      "learning_rate": 3.276471822061904e-06,
+      "loss": 0.5906,
+      "step": 5927
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.003461067354797,
+      "learning_rate": 3.2748721353752445e-06,
+      "loss": 0.683,
+      "step": 5928
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.570345614518788,
+      "learning_rate": 3.2732726491089562e-06,
+      "loss": 0.6233,
+      "step": 5929
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7306210262884796,
+      "learning_rate": 3.2716733634488563e-06,
+      "loss": 0.6855,
+      "step": 5930
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4961398274059046,
+      "learning_rate": 3.2700742785807503e-06,
+      "loss": 0.6488,
+      "step": 5931
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.674949765424381,
+      "learning_rate": 3.2684753946904136e-06,
+      "loss": 0.5726,
+      "step": 5932
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.1722015237208283,
+      "learning_rate": 3.266876711963602e-06,
+      "loss": 0.5482,
+      "step": 5933
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4532631137041836,
+      "learning_rate": 3.265278230586043e-06,
+      "loss": 0.6033,
+      "step": 5934
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.84912133996752,
+      "learning_rate": 3.2636799507434447e-06,
+      "loss": 0.6797,
+      "step": 5935
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.687307510260449,
+      "learning_rate": 3.2620818726214888e-06,
+      "loss": 0.6185,
+      "step": 5936
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3580307891356673,
+      "learning_rate": 3.260483996405839e-06,
+      "loss": 0.6161,
+      "step": 5937
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.074986511593617,
+      "learning_rate": 3.2588863222821306e-06,
+      "loss": 0.6061,
+      "step": 5938
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6795167053076145,
+      "learning_rate": 3.2572888504359743e-06,
+      "loss": 0.648,
+      "step": 5939
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9485630736671,
+      "learning_rate": 3.2556915810529627e-06,
+      "loss": 0.5871,
+      "step": 5940
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.589804098570901,
+      "learning_rate": 3.25409451431866e-06,
+      "loss": 0.5836,
+      "step": 5941
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.029028547694551,
+      "learning_rate": 3.2524976504186125e-06,
+      "loss": 0.6208,
+      "step": 5942
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.457015002957502,
+      "learning_rate": 3.2509009895383337e-06,
+      "loss": 0.5133,
+      "step": 5943
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7782124039908958,
+      "learning_rate": 3.249304531863323e-06,
+      "loss": 0.6363,
+      "step": 5944
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7887314599627198,
+      "learning_rate": 3.247708277579049e-06,
+      "loss": 0.693,
+      "step": 5945
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.059722447279897,
+      "learning_rate": 3.2461122268709657e-06,
+      "loss": 0.6958,
+      "step": 5946
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.609162372409086,
+      "learning_rate": 3.244516379924492e-06,
+      "loss": 0.5227,
+      "step": 5947
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5711727147391903,
+      "learning_rate": 3.242920736925029e-06,
+      "loss": 0.5873,
+      "step": 5948
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.965512354865726,
+      "learning_rate": 3.2413252980579572e-06,
+      "loss": 0.5924,
+      "step": 5949
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.480684034499855,
+      "learning_rate": 3.239730063508629e-06,
+      "loss": 0.6539,
+      "step": 5950
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.0554230163758693,
+      "learning_rate": 3.238135033462374e-06,
+      "loss": 0.6008,
+      "step": 5951
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.264871265257945,
+      "learning_rate": 3.2365402081044955e-06,
+      "loss": 0.7099,
+      "step": 5952
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4974903193187594,
+      "learning_rate": 3.2349455876202797e-06,
+      "loss": 0.6184,
+      "step": 5953
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.18292964645188,
+      "learning_rate": 3.2333511721949816e-06,
+      "loss": 0.576,
+      "step": 5954
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.187323524231644,
+      "learning_rate": 3.23175696201384e-06,
+      "loss": 0.6029,
+      "step": 5955
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.42051305833246,
+      "learning_rate": 3.230162957262062e-06,
+      "loss": 0.5949,
+      "step": 5956
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.437281602803996,
+      "learning_rate": 3.2285691581248345e-06,
+      "loss": 0.6308,
+      "step": 5957
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.188705793501174,
+      "learning_rate": 3.226975564787322e-06,
+      "loss": 0.6577,
+      "step": 5958
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.095497000196284,
+      "learning_rate": 3.2253821774346644e-06,
+      "loss": 0.613,
+      "step": 5959
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.217153208860546,
+      "learning_rate": 3.2237889962519748e-06,
+      "loss": 0.5494,
+      "step": 5960
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.0997362892532534,
+      "learning_rate": 3.2221960214243437e-06,
+      "loss": 0.6854,
+      "step": 5961
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.176771504376497,
+      "learning_rate": 3.2206032531368407e-06,
+      "loss": 0.625,
+      "step": 5962
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.3368855193919957,
+      "learning_rate": 3.2190106915745077e-06,
+      "loss": 0.6591,
+      "step": 5963
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.221750102788297,
+      "learning_rate": 3.2174183369223667e-06,
+      "loss": 0.5971,
+      "step": 5964
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3884878579108473,
+      "learning_rate": 3.2158261893654073e-06,
+      "loss": 0.7223,
+      "step": 5965
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.8667030280760954,
+      "learning_rate": 3.214234249088605e-06,
+      "loss": 0.6087,
+      "step": 5966
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.172551788871772,
+      "learning_rate": 3.212642516276905e-06,
+      "loss": 0.5243,
+      "step": 5967
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7291377150296534,
+      "learning_rate": 3.2110509911152315e-06,
+      "loss": 0.5997,
+      "step": 5968
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9384746293091961,
+      "learning_rate": 3.209459673788482e-06,
+      "loss": 0.5464,
+      "step": 5969
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.058472029705662,
+      "learning_rate": 3.207868564481532e-06,
+      "loss": 0.5822,
+      "step": 5970
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.567086993673489,
+      "learning_rate": 3.2062776633792303e-06,
+      "loss": 0.6797,
+      "step": 5971
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.1555293622681844,
+      "learning_rate": 3.2046869706664074e-06,
+      "loss": 0.5843,
+      "step": 5972
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3596960035354035,
+      "learning_rate": 3.2030964865278604e-06,
+      "loss": 0.5639,
+      "step": 5973
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0640908869935912,
+      "learning_rate": 3.2015062111483688e-06,
+      "loss": 0.6841,
+      "step": 5974
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.689198480543773,
+      "learning_rate": 3.199916144712688e-06,
+      "loss": 0.6605,
+      "step": 5975
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6877494467316985,
+      "learning_rate": 3.1983262874055442e-06,
+      "loss": 0.5771,
+      "step": 5976
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.772844595362914,
+      "learning_rate": 3.1967366394116477e-06,
+      "loss": 0.6137,
+      "step": 5977
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0203362646742848,
+      "learning_rate": 3.1951472009156725e-06,
+      "loss": 0.5498,
+      "step": 5978
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.242251234684735,
+      "learning_rate": 3.193557972102279e-06,
+      "loss": 0.6582,
+      "step": 5979
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.8382421017907684,
+      "learning_rate": 3.191968953156098e-06,
+      "loss": 0.6473,
+      "step": 5980
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1209413221002222,
+      "learning_rate": 3.19038014426174e-06,
+      "loss": 0.6004,
+      "step": 5981
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2255883161813963,
+      "learning_rate": 3.1887915456037833e-06,
+      "loss": 0.6218,
+      "step": 5982
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.8783014772228532,
+      "learning_rate": 3.1872031573667895e-06,
+      "loss": 0.7142,
+      "step": 5983
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5531410132078363,
+      "learning_rate": 3.185614979735293e-06,
+      "loss": 0.5936,
+      "step": 5984
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.565229169837344,
+      "learning_rate": 3.1840270128938032e-06,
+      "loss": 0.6143,
+      "step": 5985
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.045126577655692,
+      "learning_rate": 3.182439257026807e-06,
+      "loss": 0.5457,
+      "step": 5986
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0857258123928033,
+      "learning_rate": 3.180851712318761e-06,
+      "loss": 0.6753,
+      "step": 5987
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.631692022599449,
+      "learning_rate": 3.179264378954106e-06,
+      "loss": 0.5913,
+      "step": 5988
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.8525652346655925,
+      "learning_rate": 3.1776772571172514e-06,
+      "loss": 0.7053,
+      "step": 5989
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2544931285363856,
+      "learning_rate": 3.1760903469925874e-06,
+      "loss": 0.6392,
+      "step": 5990
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2143334987948213,
+      "learning_rate": 3.174503648764473e-06,
+      "loss": 0.6196,
+      "step": 5991
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7607820463483015,
+      "learning_rate": 3.1729171626172485e-06,
+      "loss": 0.6129,
+      "step": 5992
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.431241441047194,
+      "learning_rate": 3.1713308887352244e-06,
+      "loss": 0.5663,
+      "step": 5993
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9957317360641763,
+      "learning_rate": 3.1697448273026944e-06,
+      "loss": 0.5867,
+      "step": 5994
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.718815023192582,
+      "learning_rate": 3.1681589785039178e-06,
+      "loss": 0.5582,
+      "step": 5995
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3173917102419392,
+      "learning_rate": 3.1665733425231356e-06,
+      "loss": 0.5831,
+      "step": 5996
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.9747101713599116,
+      "learning_rate": 3.164987919544563e-06,
+      "loss": 0.6145,
+      "step": 5997
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.0576887431965263,
+      "learning_rate": 3.163402709752389e-06,
+      "loss": 0.6303,
+      "step": 5998
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2892471503762697,
+      "learning_rate": 3.1618177133307813e-06,
+      "loss": 0.5384,
+      "step": 5999
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0697756117981876,
+      "learning_rate": 3.1602329304638755e-06,
+      "loss": 0.6807,
+      "step": 6000
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1548228132567897,
+      "learning_rate": 3.158648361335791e-06,
+      "loss": 0.6769,
+      "step": 6001
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.171892545453622,
+      "learning_rate": 3.1570640061306153e-06,
+      "loss": 0.5813,
+      "step": 6002
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.5558082679769543,
+      "learning_rate": 3.1554798650324183e-06,
+      "loss": 0.5725,
+      "step": 6003
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1783862920998778,
+      "learning_rate": 3.1538959382252378e-06,
+      "loss": 0.6095,
+      "step": 6004
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.439670545080028,
+      "learning_rate": 3.1523122258930904e-06,
+      "loss": 0.6402,
+      "step": 6005
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1135984445220846,
+      "learning_rate": 3.150728728219966e-06,
+      "loss": 0.5736,
+      "step": 6006
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2613893512798544,
+      "learning_rate": 3.149145445389835e-06,
+      "loss": 0.6038,
+      "step": 6007
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7430614334073287,
+      "learning_rate": 3.147562377586635e-06,
+      "loss": 0.6266,
+      "step": 6008
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.8377286942815885,
+      "learning_rate": 3.1459795249942815e-06,
+      "loss": 0.6455,
+      "step": 6009
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9591836036301431,
+      "learning_rate": 3.144396887796669e-06,
+      "loss": 0.5318,
+      "step": 6010
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0086517588320363,
+      "learning_rate": 3.1428144661776605e-06,
+      "loss": 0.5734,
+      "step": 6011
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.218164590981601,
+      "learning_rate": 3.141232260321102e-06,
+      "loss": 0.683,
+      "step": 6012
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.503329798201055,
+      "learning_rate": 3.1396502704108034e-06,
+      "loss": 0.6242,
+      "step": 6013
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.221827577676699,
+      "learning_rate": 3.13806849663056e-06,
+      "loss": 0.5933,
+      "step": 6014
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.6189027947412518,
+      "learning_rate": 3.1364869391641343e-06,
+      "loss": 0.5451,
+      "step": 6015
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2116203062161435,
+      "learning_rate": 3.1349055981952725e-06,
+      "loss": 0.5863,
+      "step": 6016
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4603820875117735,
+      "learning_rate": 3.133324473907685e-06,
+      "loss": 0.6867,
+      "step": 6017
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.688134030426931,
+      "learning_rate": 3.1317435664850626e-06,
+      "loss": 0.6107,
+      "step": 6018
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2474502089494233,
+      "learning_rate": 3.130162876111074e-06,
+      "loss": 0.665,
+      "step": 6019
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.235530948111977,
+      "learning_rate": 3.128582402969358e-06,
+      "loss": 0.6928,
+      "step": 6020
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5821548582113008,
+      "learning_rate": 3.1270021472435276e-06,
+      "loss": 0.6203,
+      "step": 6021
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.637774136880771,
+      "learning_rate": 3.125422109117173e-06,
+      "loss": 0.66,
+      "step": 6022
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1103955615847716,
+      "learning_rate": 3.1238422887738596e-06,
+      "loss": 0.5965,
+      "step": 6023
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2341447165344492,
+      "learning_rate": 3.122262686397124e-06,
+      "loss": 0.741,
+      "step": 6024
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2231015638638967,
+      "learning_rate": 3.1206833021704843e-06,
+      "loss": 0.6181,
+      "step": 6025
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.8157201183565475,
+      "learning_rate": 3.1191041362774246e-06,
+      "loss": 0.688,
+      "step": 6026
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6112171825573403,
+      "learning_rate": 3.117525188901409e-06,
+      "loss": 0.6337,
+      "step": 6027
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.591086899022109,
+      "learning_rate": 3.115946460225875e-06,
+      "loss": 0.6666,
+      "step": 6028
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.990144411097681,
+      "learning_rate": 3.1143679504342367e-06,
+      "loss": 0.607,
+      "step": 6029
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.097695113019066,
+      "learning_rate": 3.1127896597098784e-06,
+      "loss": 0.5598,
+      "step": 6030
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.043305343302304,
+      "learning_rate": 3.1112115882361605e-06,
+      "loss": 0.6536,
+      "step": 6031
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1224861128520143,
+      "learning_rate": 3.1096337361964213e-06,
+      "loss": 0.6197,
+      "step": 6032
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2716747161287523,
+      "learning_rate": 3.108056103773972e-06,
+      "loss": 0.6368,
+      "step": 6033
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3153801155163998,
+      "learning_rate": 3.106478691152094e-06,
+      "loss": 0.6325,
+      "step": 6034
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5772248457210116,
+      "learning_rate": 3.1049014985140468e-06,
+      "loss": 0.6982,
+      "step": 6035
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.92355644648877,
+      "learning_rate": 3.103324526043066e-06,
+      "loss": 0.6235,
+      "step": 6036
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.394493973433464,
+      "learning_rate": 3.101747773922359e-06,
+      "loss": 0.5862,
+      "step": 6037
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.840825066417566,
+      "learning_rate": 3.100171242335109e-06,
+      "loss": 0.6773,
+      "step": 6038
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3142902400992083,
+      "learning_rate": 3.0985949314644724e-06,
+      "loss": 0.6347,
+      "step": 6039
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.1693698090325952,
+      "learning_rate": 3.09701884149358e-06,
+      "loss": 0.632,
+      "step": 6040
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.174190785161442,
+      "learning_rate": 3.0954429726055367e-06,
+      "loss": 0.5664,
+      "step": 6041
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.427368074809071,
+      "learning_rate": 3.093867324983425e-06,
+      "loss": 0.6002,
+      "step": 6042
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.98991358532084,
+      "learning_rate": 3.0922918988102968e-06,
+      "loss": 0.5889,
+      "step": 6043
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.16305476049039,
+      "learning_rate": 3.0907166942691804e-06,
+      "loss": 0.5656,
+      "step": 6044
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3426174204082426,
+      "learning_rate": 3.0891417115430794e-06,
+      "loss": 0.5778,
+      "step": 6045
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.695431267278061,
+      "learning_rate": 3.08756695081497e-06,
+      "loss": 0.662,
+      "step": 6046
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4120245938951017,
+      "learning_rate": 3.085992412267807e-06,
+      "loss": 0.5995,
+      "step": 6047
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0993135251738906,
+      "learning_rate": 3.08441809608451e-06,
+      "loss": 0.6046,
+      "step": 6048
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.120694022813388,
+      "learning_rate": 3.0828440024479823e-06,
+      "loss": 0.597,
+      "step": 6049
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.164550468698215,
+      "learning_rate": 3.081270131541094e-06,
+      "loss": 0.6243,
+      "step": 6050
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.499720888315672,
+      "learning_rate": 3.079696483546699e-06,
+      "loss": 0.5662,
+      "step": 6051
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.19241543023071,
+      "learning_rate": 3.078123058647614e-06,
+      "loss": 0.645,
+      "step": 6052
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1759008939791316,
+      "learning_rate": 3.0765498570266354e-06,
+      "loss": 0.5636,
+      "step": 6053
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.481766219102531,
+      "learning_rate": 3.074976878866536e-06,
+      "loss": 0.5773,
+      "step": 6054
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6119108357768677,
+      "learning_rate": 3.0734041243500578e-06,
+      "loss": 0.6492,
+      "step": 6055
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1005618832159336,
+      "learning_rate": 3.0718315936599184e-06,
+      "loss": 0.6326,
+      "step": 6056
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1970627941200704,
+      "learning_rate": 3.0702592869788105e-06,
+      "loss": 0.5606,
+      "step": 6057
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0059803541239813,
+      "learning_rate": 3.0686872044894014e-06,
+      "loss": 0.5722,
+      "step": 6058
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2015112152252043,
+      "learning_rate": 3.0671153463743282e-06,
+      "loss": 0.641,
+      "step": 6059
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0191280750132719,
+      "learning_rate": 3.0655437128162093e-06,
+      "loss": 0.5936,
+      "step": 6060
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.045650809744414,
+      "learning_rate": 3.0639723039976284e-06,
+      "loss": 0.6325,
+      "step": 6061
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2246240058605222,
+      "learning_rate": 3.062401120101149e-06,
+      "loss": 0.6262,
+      "step": 6062
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.146655954063514,
+      "learning_rate": 3.060830161309305e-06,
+      "loss": 0.699,
+      "step": 6063
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0936615072845663,
+      "learning_rate": 3.05925942780461e-06,
+      "loss": 0.5818,
+      "step": 6064
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.660109768200281,
+      "learning_rate": 3.0576889197695435e-06,
+      "loss": 0.5617,
+      "step": 6065
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2782490312138446,
+      "learning_rate": 3.0561186373865625e-06,
+      "loss": 0.5962,
+      "step": 6066
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5339115449409535,
+      "learning_rate": 3.054548580838099e-06,
+      "loss": 0.6213,
+      "step": 6067
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.145966200390791,
+      "learning_rate": 3.05297875030656e-06,
+      "loss": 0.5802,
+      "step": 6068
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.023290228663878,
+      "learning_rate": 3.05140914597432e-06,
+      "loss": 0.6221,
+      "step": 6069
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7573240624259157,
+      "learning_rate": 3.049839768023732e-06,
+      "loss": 0.5638,
+      "step": 6070
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.360362698449237,
+      "learning_rate": 3.0482706166371236e-06,
+      "loss": 0.6423,
+      "step": 6071
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7058889588734343,
+      "learning_rate": 3.0467016919967908e-06,
+      "loss": 0.6319,
+      "step": 6072
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.38161674800116,
+      "learning_rate": 3.0451329942850117e-06,
+      "loss": 0.6777,
+      "step": 6073
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4754236044435,
+      "learning_rate": 3.0435645236840296e-06,
+      "loss": 0.6553,
+      "step": 6074
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.505835959851159,
+      "learning_rate": 3.041996280376066e-06,
+      "loss": 0.5921,
+      "step": 6075
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2325223472969085,
+      "learning_rate": 3.0404282645433125e-06,
+      "loss": 0.6708,
+      "step": 6076
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4229637130342065,
+      "learning_rate": 3.038860476367942e-06,
+      "loss": 0.5817,
+      "step": 6077
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2424616876949837,
+      "learning_rate": 3.03729291603209e-06,
+      "loss": 0.6665,
+      "step": 6078
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.196951713754301,
+      "learning_rate": 3.0357255837178733e-06,
+      "loss": 0.6736,
+      "step": 6079
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.412747433968496,
+      "learning_rate": 3.034158479607381e-06,
+      "loss": 0.6189,
+      "step": 6080
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3144405798445638,
+      "learning_rate": 3.032591603882674e-06,
+      "loss": 0.6844,
+      "step": 6081
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7027103937151957,
+      "learning_rate": 3.031024956725787e-06,
+      "loss": 0.6343,
+      "step": 6082
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.612524722737757,
+      "learning_rate": 3.029458538318728e-06,
+      "loss": 0.6779,
+      "step": 6083
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5457464342733767,
+      "learning_rate": 3.02789234884348e-06,
+      "loss": 0.6377,
+      "step": 6084
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9720571167689611,
+      "learning_rate": 3.0263263884819975e-06,
+      "loss": 0.6417,
+      "step": 6085
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.847776105778556,
+      "learning_rate": 3.0247606574162127e-06,
+      "loss": 0.6391,
+      "step": 6086
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.33929101232662,
+      "learning_rate": 3.0231951558280226e-06,
+      "loss": 0.6156,
+      "step": 6087
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5013360523106787,
+      "learning_rate": 3.0216298838993043e-06,
+      "loss": 0.7201,
+      "step": 6088
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.089321590479491,
+      "learning_rate": 3.020064841811908e-06,
+      "loss": 0.6186,
+      "step": 6089
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7046583661780823,
+      "learning_rate": 3.018500029747657e-06,
+      "loss": 0.6985,
+      "step": 6090
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5221299383717657,
+      "learning_rate": 3.016935447888343e-06,
+      "loss": 0.6104,
+      "step": 6091
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1978338732670597,
+      "learning_rate": 3.015371096415735e-06,
+      "loss": 0.5294,
+      "step": 6092
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.685524749270053,
+      "learning_rate": 3.0138069755115772e-06,
+      "loss": 0.672,
+      "step": 6093
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1195488615414546,
+      "learning_rate": 3.012243085357582e-06,
+      "loss": 0.6829,
+      "step": 6094
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.577287209734468,
+      "learning_rate": 3.010679426135442e-06,
+      "loss": 0.7072,
+      "step": 6095
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.13324035955566,
+      "learning_rate": 3.009115998026815e-06,
+      "loss": 0.5589,
+      "step": 6096
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.274593312395244,
+      "learning_rate": 3.007552801213335e-06,
+      "loss": 0.6797,
+      "step": 6097
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8495761189029405,
+      "learning_rate": 3.0059898358766102e-06,
+      "loss": 0.5548,
+      "step": 6098
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.263736591389369,
+      "learning_rate": 3.004427102198225e-06,
+      "loss": 0.6734,
+      "step": 6099
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.428586454310511,
+      "learning_rate": 3.002864600359729e-06,
+      "loss": 0.6665,
+      "step": 6100
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6142648776527855,
+      "learning_rate": 3.0013023305426493e-06,
+      "loss": 0.665,
+      "step": 6101
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.093099723594316,
+      "learning_rate": 2.9997402929284886e-06,
+      "loss": 0.5937,
+      "step": 6102
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.981769329689394,
+      "learning_rate": 2.9981784876987195e-06,
+      "loss": 0.5496,
+      "step": 6103
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.089324365056355,
+      "learning_rate": 2.996616915034786e-06,
+      "loss": 0.582,
+      "step": 6104
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.283444978516847,
+      "learning_rate": 2.9950555751181067e-06,
+      "loss": 0.6011,
+      "step": 6105
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1322926764233148,
+      "learning_rate": 2.9934944681300764e-06,
+      "loss": 0.595,
+      "step": 6106
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6785042803171994,
+      "learning_rate": 2.9919335942520577e-06,
+      "loss": 0.604,
+      "step": 6107
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.525026434174256,
+      "learning_rate": 2.9903729536653908e-06,
+      "loss": 0.5419,
+      "step": 6108
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.332421592393996,
+      "learning_rate": 2.9888125465513838e-06,
+      "loss": 0.6079,
+      "step": 6109
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7390363662038246,
+      "learning_rate": 2.987252373091322e-06,
+      "loss": 0.5825,
+      "step": 6110
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6929057930774722,
+      "learning_rate": 2.9856924334664607e-06,
+      "loss": 0.5882,
+      "step": 6111
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8780256095081334,
+      "learning_rate": 2.9841327278580306e-06,
+      "loss": 0.5991,
+      "step": 6112
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.211030034668102,
+      "learning_rate": 2.982573256447232e-06,
+      "loss": 0.6217,
+      "step": 6113
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.076849408004536,
+      "learning_rate": 2.98101401941524e-06,
+      "loss": 0.6127,
+      "step": 6114
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8399499017577328,
+      "learning_rate": 2.979455016943204e-06,
+      "loss": 0.6985,
+      "step": 6115
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5380298634074285,
+      "learning_rate": 2.977896249212244e-06,
+      "loss": 0.6568,
+      "step": 6116
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1139041232082096,
+      "learning_rate": 2.976337716403452e-06,
+      "loss": 0.6043,
+      "step": 6117
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.386270978789105,
+      "learning_rate": 2.974779418697893e-06,
+      "loss": 0.58,
+      "step": 6118
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.249981645922609,
+      "learning_rate": 2.9732213562766076e-06,
+      "loss": 0.7059,
+      "step": 6119
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7953475276968995,
+      "learning_rate": 2.9716635293206054e-06,
+      "loss": 0.5731,
+      "step": 6120
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1595949678722386,
+      "learning_rate": 2.9701059380108732e-06,
+      "loss": 0.6399,
+      "step": 6121
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.250031089313065,
+      "learning_rate": 2.9685485825283646e-06,
+      "loss": 0.6693,
+      "step": 6122
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.716682335751081,
+      "learning_rate": 2.9669914630540074e-06,
+      "loss": 0.5684,
+      "step": 6123
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.11893087796511,
+      "learning_rate": 2.9654345797687067e-06,
+      "loss": 0.6003,
+      "step": 6124
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.1494558741601413,
+      "learning_rate": 2.9638779328533363e-06,
+      "loss": 0.5924,
+      "step": 6125
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6969375293675673,
+      "learning_rate": 2.9623215224887405e-06,
+      "loss": 0.6277,
+      "step": 6126
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.129836907789918,
+      "learning_rate": 2.9607653488557385e-06,
+      "loss": 0.6265,
+      "step": 6127
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.24499842881353,
+      "learning_rate": 2.9592094121351257e-06,
+      "loss": 0.5141,
+      "step": 6128
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1727226381138738,
+      "learning_rate": 2.9576537125076644e-06,
+      "loss": 0.5722,
+      "step": 6129
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4525235655286766,
+      "learning_rate": 2.956098250154089e-06,
+      "loss": 0.628,
+      "step": 6130
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.0032073192917212,
+      "learning_rate": 2.954543025255111e-06,
+      "loss": 0.6489,
+      "step": 6131
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2967514557097366,
+      "learning_rate": 2.9529880379914123e-06,
+      "loss": 0.6079,
+      "step": 6132
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.9813396803758367,
+      "learning_rate": 2.9514332885436447e-06,
+      "loss": 0.6043,
+      "step": 6133
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.374827795762819,
+      "learning_rate": 2.9498787770924375e-06,
+      "loss": 0.6733,
+      "step": 6134
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8305018905675694,
+      "learning_rate": 2.9483245038183874e-06,
+      "loss": 0.6773,
+      "step": 6135
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.729421853905016,
+      "learning_rate": 2.946770468902064e-06,
+      "loss": 0.6898,
+      "step": 6136
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9984646956788515,
+      "learning_rate": 2.945216672524014e-06,
+      "loss": 0.6088,
+      "step": 6137
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.667763450610834,
+      "learning_rate": 2.943663114864752e-06,
+      "loss": 0.577,
+      "step": 6138
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.802310695654871,
+      "learning_rate": 2.9421097961047633e-06,
+      "loss": 0.5924,
+      "step": 6139
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2866904253496174,
+      "learning_rate": 2.9405567164245096e-06,
+      "loss": 0.6054,
+      "step": 6140
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.2953941016991246,
+      "learning_rate": 2.939003876004424e-06,
+      "loss": 0.6667,
+      "step": 6141
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.230139296054894,
+      "learning_rate": 2.9374512750249098e-06,
+      "loss": 0.594,
+      "step": 6142
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2675620169968975,
+      "learning_rate": 2.935898913666345e-06,
+      "loss": 0.6655,
+      "step": 6143
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.459430858178251,
+      "learning_rate": 2.9343467921090774e-06,
+      "loss": 0.655,
+      "step": 6144
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.267669278148576,
+      "learning_rate": 2.9327949105334284e-06,
+      "loss": 0.633,
+      "step": 6145
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4328862273124567,
+      "learning_rate": 2.93124326911969e-06,
+      "loss": 0.6138,
+      "step": 6146
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8997361338805473,
+      "learning_rate": 2.9296918680481308e-06,
+      "loss": 0.5905,
+      "step": 6147
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.7491447988759568,
+      "learning_rate": 2.928140707498984e-06,
+      "loss": 0.6503,
+      "step": 6148
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.6968575644202506,
+      "learning_rate": 2.92658978765246e-06,
+      "loss": 0.5916,
+      "step": 6149
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.5697760256627373,
+      "learning_rate": 2.925039108688742e-06,
+      "loss": 0.638,
+      "step": 6150
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.049482671759394,
+      "learning_rate": 2.9234886707879827e-06,
+      "loss": 0.6333,
+      "step": 6151
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.446291237259445,
+      "learning_rate": 2.921938474130307e-06,
+      "loss": 0.652,
+      "step": 6152
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.200019946516808,
+      "learning_rate": 2.9203885188958103e-06,
+      "loss": 0.5058,
+      "step": 6153
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.0764287200010685,
+      "learning_rate": 2.9188388052645656e-06,
+      "loss": 0.6562,
+      "step": 6154
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.3263889642902593,
+      "learning_rate": 2.9172893334166108e-06,
+      "loss": 0.5888,
+      "step": 6155
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.164471451455957,
+      "learning_rate": 2.915740103531963e-06,
+      "loss": 0.5748,
+      "step": 6156
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.0166454764249417,
+      "learning_rate": 2.9141911157906032e-06,
+      "loss": 0.6495,
+      "step": 6157
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0796493091839743,
+      "learning_rate": 2.9126423703724925e-06,
+      "loss": 0.5941,
+      "step": 6158
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.472853403516785,
+      "learning_rate": 2.911093867457555e-06,
+      "loss": 0.6391,
+      "step": 6159
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2590642055205663,
+      "learning_rate": 2.9095456072256955e-06,
+      "loss": 0.6039,
+      "step": 6160
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1770319394327617,
+      "learning_rate": 2.9079975898567823e-06,
+      "loss": 0.6178,
+      "step": 6161
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4623318761495,
+      "learning_rate": 2.906449815530664e-06,
+      "loss": 0.6449,
+      "step": 6162
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4317346877522916,
+      "learning_rate": 2.9049022844271517e-06,
+      "loss": 0.688,
+      "step": 6163
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.427285093977721,
+      "learning_rate": 2.9033549967260383e-06,
+      "loss": 0.6544,
+      "step": 6164
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4834448464193644,
+      "learning_rate": 2.9018079526070786e-06,
+      "loss": 0.5716,
+      "step": 6165
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.7395209019829614,
+      "learning_rate": 2.900261152250007e-06,
+      "loss": 0.542,
+      "step": 6166
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.0529654367037993,
+      "learning_rate": 2.8987145958345235e-06,
+      "loss": 0.5837,
+      "step": 6167
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.1062690132576405,
+      "learning_rate": 2.8971682835403043e-06,
+      "loss": 0.6021,
+      "step": 6168
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8007780098866983,
+      "learning_rate": 2.895622215546997e-06,
+      "loss": 0.5921,
+      "step": 6169
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.200411944525784,
+      "learning_rate": 2.8940763920342153e-06,
+      "loss": 0.6037,
+      "step": 6170
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.858054420142283,
+      "learning_rate": 2.892530813181553e-06,
+      "loss": 0.6641,
+      "step": 6171
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5625816954495666,
+      "learning_rate": 2.8909854791685666e-06,
+      "loss": 0.5688,
+      "step": 6172
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.715416045079819,
+      "learning_rate": 2.889440390174793e-06,
+      "loss": 0.5802,
+      "step": 6173
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6641433870808893,
+      "learning_rate": 2.887895546379732e-06,
+      "loss": 0.5821,
+      "step": 6174
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.416515122562737,
+      "learning_rate": 2.8863509479628626e-06,
+      "loss": 0.6671,
+      "step": 6175
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.836952564545351,
+      "learning_rate": 2.884806595103628e-06,
+      "loss": 0.6201,
+      "step": 6176
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.588917907113869,
+      "learning_rate": 2.8832624879814507e-06,
+      "loss": 0.5886,
+      "step": 6177
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.571817028335935,
+      "learning_rate": 2.8817186267757173e-06,
+      "loss": 0.5718,
+      "step": 6178
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.812640707900028,
+      "learning_rate": 2.88017501166579e-06,
+      "loss": 0.6379,
+      "step": 6179
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.626920559291416,
+      "learning_rate": 2.8786316428310046e-06,
+      "loss": 0.5918,
+      "step": 6180
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.798186376430685,
+      "learning_rate": 2.8770885204506603e-06,
+      "loss": 0.6388,
+      "step": 6181
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.751589330763615,
+      "learning_rate": 2.8755456447040362e-06,
+      "loss": 0.573,
+      "step": 6182
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.28473507954531,
+      "learning_rate": 2.874003015770377e-06,
+      "loss": 0.6162,
+      "step": 6183
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.557119980309637,
+      "learning_rate": 2.872460633828904e-06,
+      "loss": 0.5865,
+      "step": 6184
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.1804128048003784,
+      "learning_rate": 2.8709184990588012e-06,
+      "loss": 0.6432,
+      "step": 6185
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6384781468932905,
+      "learning_rate": 2.869376611639236e-06,
+      "loss": 0.6571,
+      "step": 6186
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.595212904243066,
+      "learning_rate": 2.8678349717493343e-06,
+      "loss": 0.6985,
+      "step": 6187
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.0950307332286857,
+      "learning_rate": 2.8662935795682046e-06,
+      "loss": 0.6029,
+      "step": 6188
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.345799472219859,
+      "learning_rate": 2.864752435274916e-06,
+      "loss": 0.6059,
+      "step": 6189
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.25918445844699,
+      "learning_rate": 2.8632115390485176e-06,
+      "loss": 0.676,
+      "step": 6190
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.322802206550409,
+      "learning_rate": 2.8616708910680278e-06,
+      "loss": 0.6211,
+      "step": 6191
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.523140689849376,
+      "learning_rate": 2.8601304915124305e-06,
+      "loss": 0.6485,
+      "step": 6192
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.013530763468245,
+      "learning_rate": 2.85859034056069e-06,
+      "loss": 0.6619,
+      "step": 6193
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.291612700420486,
+      "learning_rate": 2.8570504383917296e-06,
+      "loss": 0.5981,
+      "step": 6194
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4140328386952463,
+      "learning_rate": 2.8555107851844576e-06,
+      "loss": 0.5461,
+      "step": 6195
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.687609928584908,
+      "learning_rate": 2.8539713811177418e-06,
+      "loss": 0.6283,
+      "step": 6196
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3244100962522163,
+      "learning_rate": 2.8524322263704297e-06,
+      "loss": 0.5716,
+      "step": 6197
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1973658455412997,
+      "learning_rate": 2.8508933211213306e-06,
+      "loss": 0.531,
+      "step": 6198
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0935858472905258,
+      "learning_rate": 2.8493546655492356e-06,
+      "loss": 0.598,
+      "step": 6199
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.77092033346428,
+      "learning_rate": 2.8478162598328963e-06,
+      "loss": 0.552,
+      "step": 6200
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.624275390466366,
+      "learning_rate": 2.8462781041510446e-06,
+      "loss": 0.666,
+      "step": 6201
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1540146591782405,
+      "learning_rate": 2.8447401986823752e-06,
+      "loss": 0.6687,
+      "step": 6202
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.263646788490095,
+      "learning_rate": 2.8432025436055593e-06,
+      "loss": 0.6402,
+      "step": 6203
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.537185698965246,
+      "learning_rate": 2.841665139099239e-06,
+      "loss": 0.6249,
+      "step": 6204
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4385362216870554,
+      "learning_rate": 2.8401279853420216e-06,
+      "loss": 0.6824,
+      "step": 6205
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1466991106289137,
+      "learning_rate": 2.838591082512494e-06,
+      "loss": 0.5709,
+      "step": 6206
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.076550784919779,
+      "learning_rate": 2.837054430789204e-06,
+      "loss": 0.6901,
+      "step": 6207
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.7605170118341413,
+      "learning_rate": 2.8355180303506803e-06,
+      "loss": 0.6498,
+      "step": 6208
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.224223567454739,
+      "learning_rate": 2.8339818813754115e-06,
+      "loss": 0.6497,
+      "step": 6209
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.348664612330819,
+      "learning_rate": 2.8324459840418694e-06,
+      "loss": 0.6267,
+      "step": 6210
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.9347639572736144,
+      "learning_rate": 2.8309103385284853e-06,
+      "loss": 0.6423,
+      "step": 6211
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.408795609146555,
+      "learning_rate": 2.82937494501367e-06,
+      "loss": 0.6295,
+      "step": 6212
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.8368693625535277,
+      "learning_rate": 2.8278398036757963e-06,
+      "loss": 0.5594,
+      "step": 6213
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.51892710402293,
+      "learning_rate": 2.8263049146932153e-06,
+      "loss": 0.5737,
+      "step": 6214
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.302046825233197,
+      "learning_rate": 2.8247702782442483e-06,
+      "loss": 0.6769,
+      "step": 6215
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.85067453672738,
+      "learning_rate": 2.8232358945071804e-06,
+      "loss": 0.5899,
+      "step": 6216
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.9563736592529963,
+      "learning_rate": 2.821701763660275e-06,
+      "loss": 0.6199,
+      "step": 6217
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8979485285850846,
+      "learning_rate": 2.8201678858817604e-06,
+      "loss": 0.5747,
+      "step": 6218
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.242630295137544,
+      "learning_rate": 2.818634261349842e-06,
+      "loss": 0.5828,
+      "step": 6219
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8275662714905065,
+      "learning_rate": 2.8171008902426865e-06,
+      "loss": 0.702,
+      "step": 6220
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2586239782876385,
+      "learning_rate": 2.8155677727384422e-06,
+      "loss": 0.5648,
+      "step": 6221
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.5846393384463773,
+      "learning_rate": 2.814034909015217e-06,
+      "loss": 0.6001,
+      "step": 6222
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.736699486370083,
+      "learning_rate": 2.8125022992510997e-06,
+      "loss": 0.5504,
+      "step": 6223
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.982453703063961,
+      "learning_rate": 2.810969943624139e-06,
+      "loss": 0.6271,
+      "step": 6224
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0378158485658642,
+      "learning_rate": 2.8094378423123646e-06,
+      "loss": 0.5252,
+      "step": 6225
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.368853983880845,
+      "learning_rate": 2.807905995493768e-06,
+      "loss": 0.5657,
+      "step": 6226
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6453989305193906,
+      "learning_rate": 2.8063744033463157e-06,
+      "loss": 0.6209,
+      "step": 6227
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.5318565272857363,
+      "learning_rate": 2.8048430660479463e-06,
+      "loss": 0.5866,
+      "step": 6228
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.8249719217174523,
+      "learning_rate": 2.803311983776562e-06,
+      "loss": 0.6002,
+      "step": 6229
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4969861662636537,
+      "learning_rate": 2.8017811567100434e-06,
+      "loss": 0.6462,
+      "step": 6230
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3785984897135353,
+      "learning_rate": 2.8002505850262334e-06,
+      "loss": 0.5927,
+      "step": 6231
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.141017713761464,
+      "learning_rate": 2.7987202689029535e-06,
+      "loss": 0.6416,
+      "step": 6232
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.9090874878260973,
+      "learning_rate": 2.797190208517988e-06,
+      "loss": 0.6156,
+      "step": 6233
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0212371235951685,
+      "learning_rate": 2.795660404049098e-06,
+      "loss": 0.6247,
+      "step": 6234
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2096556030865004,
+      "learning_rate": 2.794130855674009e-06,
+      "loss": 0.6012,
+      "step": 6235
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2361232833661924,
+      "learning_rate": 2.7926015635704216e-06,
+      "loss": 0.6237,
+      "step": 6236
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.605730880983019,
+      "learning_rate": 2.7910725279160016e-06,
+      "loss": 0.6243,
+      "step": 6237
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.33390242289277,
+      "learning_rate": 2.78954374888839e-06,
+      "loss": 0.6674,
+      "step": 6238
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.329689514740396,
+      "learning_rate": 2.7880152266651985e-06,
+      "loss": 0.6576,
+      "step": 6239
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.400052397965454,
+      "learning_rate": 2.7864869614240013e-06,
+      "loss": 0.5916,
+      "step": 6240
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.437094841587654,
+      "learning_rate": 2.7849589533423526e-06,
+      "loss": 0.6769,
+      "step": 6241
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.158441047716093,
+      "learning_rate": 2.783431202597767e-06,
+      "loss": 0.5817,
+      "step": 6242
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.202778092080676,
+      "learning_rate": 2.78190370936774e-06,
+      "loss": 0.6905,
+      "step": 6243
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.181636718933403,
+      "learning_rate": 2.7803764738297257e-06,
+      "loss": 0.6753,
+      "step": 6244
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.169908589296973,
+      "learning_rate": 2.7788494961611577e-06,
+      "loss": 0.6062,
+      "step": 6245
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1889402639310336,
+      "learning_rate": 2.7773227765394335e-06,
+      "loss": 0.6787,
+      "step": 6246
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6886102911065244,
+      "learning_rate": 2.7757963151419255e-06,
+      "loss": 0.6211,
+      "step": 6247
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.620084649426985,
+      "learning_rate": 2.7742701121459703e-06,
+      "loss": 0.5493,
+      "step": 6248
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0808691207938845,
+      "learning_rate": 2.772744167728879e-06,
+      "loss": 0.6245,
+      "step": 6249
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.94937587598107,
+      "learning_rate": 2.7712184820679343e-06,
+      "loss": 0.5034,
+      "step": 6250
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.750670337900543,
+      "learning_rate": 2.7696930553403817e-06,
+      "loss": 0.6075,
+      "step": 6251
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8167968242924053,
+      "learning_rate": 2.7681678877234446e-06,
+      "loss": 0.567,
+      "step": 6252
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3048720790451163,
+      "learning_rate": 2.7666429793943087e-06,
+      "loss": 0.6424,
+      "step": 6253
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.978504924873048,
+      "learning_rate": 2.765118330530138e-06,
+      "loss": 0.6175,
+      "step": 6254
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.9075301291638174,
+      "learning_rate": 2.763593941308057e-06,
+      "loss": 0.6043,
+      "step": 6255
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.3429392341032482,
+      "learning_rate": 2.7620698119051687e-06,
+      "loss": 0.6713,
+      "step": 6256
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.077361667669363,
+      "learning_rate": 2.7605459424985387e-06,
+      "loss": 0.7026,
+      "step": 6257
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5927187596499923,
+      "learning_rate": 2.7590223332652096e-06,
+      "loss": 0.6464,
+      "step": 6258
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.588598586482582,
+      "learning_rate": 2.7574989843821855e-06,
+      "loss": 0.7161,
+      "step": 6259
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.9831247521107374,
+      "learning_rate": 2.7559758960264492e-06,
+      "loss": 0.6043,
+      "step": 6260
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 13.839156626103529,
+      "learning_rate": 2.7544530683749447e-06,
+      "loss": 0.6524,
+      "step": 6261
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.302955982405863,
+      "learning_rate": 2.7529305016045917e-06,
+      "loss": 0.6315,
+      "step": 6262
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.745223080871031,
+      "learning_rate": 2.751408195892279e-06,
+      "loss": 0.6543,
+      "step": 6263
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.2451920481026892,
+      "learning_rate": 2.74988615141486e-06,
+      "loss": 0.5986,
+      "step": 6264
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.1406872121844263,
+      "learning_rate": 2.7483643683491658e-06,
+      "loss": 0.5726,
+      "step": 6265
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.484418230228393,
+      "learning_rate": 2.7468428468719877e-06,
+      "loss": 0.7162,
+      "step": 6266
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.329591736950578,
+      "learning_rate": 2.7453215871600967e-06,
+      "loss": 0.5947,
+      "step": 6267
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4045111597868307,
+      "learning_rate": 2.743800589390225e-06,
+      "loss": 0.6726,
+      "step": 6268
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3627978273372885,
+      "learning_rate": 2.74227985373908e-06,
+      "loss": 0.5535,
+      "step": 6269
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.6457501096909266,
+      "learning_rate": 2.7407593803833333e-06,
+      "loss": 0.6602,
+      "step": 6270
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.926138821288341,
+      "learning_rate": 2.7392391694996335e-06,
+      "loss": 0.7381,
+      "step": 6271
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1259023466885845,
+      "learning_rate": 2.7377192212645888e-06,
+      "loss": 0.5446,
+      "step": 6272
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.231315217630402,
+      "learning_rate": 2.736199535854788e-06,
+      "loss": 0.566,
+      "step": 6273
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7856102084315983,
+      "learning_rate": 2.7346801134467794e-06,
+      "loss": 0.606,
+      "step": 6274
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6026849229772218,
+      "learning_rate": 2.733160954217086e-06,
+      "loss": 0.6111,
+      "step": 6275
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5134836440676724,
+      "learning_rate": 2.731642058342203e-06,
+      "loss": 0.5827,
+      "step": 6276
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2996578580915323,
+      "learning_rate": 2.7301234259985863e-06,
+      "loss": 0.5993,
+      "step": 6277
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.937914077737516,
+      "learning_rate": 2.72860505736267e-06,
+      "loss": 0.6798,
+      "step": 6278
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.9983367460524004,
+      "learning_rate": 2.7270869526108507e-06,
+      "loss": 0.6152,
+      "step": 6279
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6496495044995605,
+      "learning_rate": 2.7255691119195005e-06,
+      "loss": 0.6753,
+      "step": 6280
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.138925305786583,
+      "learning_rate": 2.7240515354649545e-06,
+      "loss": 0.6051,
+      "step": 6281
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.490293141221245,
+      "learning_rate": 2.722534223423524e-06,
+      "loss": 0.6335,
+      "step": 6282
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.322910264522336,
+      "learning_rate": 2.721017175971482e-06,
+      "loss": 0.6405,
+      "step": 6283
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.643925276634918,
+      "learning_rate": 2.719500393285076e-06,
+      "loss": 0.6404,
+      "step": 6284
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0080720283021827,
+      "learning_rate": 2.7179838755405253e-06,
+      "loss": 0.5701,
+      "step": 6285
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5525745426515476,
+      "learning_rate": 2.7164676229140098e-06,
+      "loss": 0.6459,
+      "step": 6286
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.512640406594001,
+      "learning_rate": 2.714951635581684e-06,
+      "loss": 0.6733,
+      "step": 6287
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9599284170567002,
+      "learning_rate": 2.713435913719671e-06,
+      "loss": 0.5056,
+      "step": 6288
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2612027984032435,
+      "learning_rate": 2.7119204575040666e-06,
+      "loss": 0.6296,
+      "step": 6289
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.856455315447502,
+      "learning_rate": 2.7104052671109267e-06,
+      "loss": 0.6801,
+      "step": 6290
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.527783580967457,
+      "learning_rate": 2.708890342716286e-06,
+      "loss": 0.5643,
+      "step": 6291
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.314451080373654,
+      "learning_rate": 2.7073756844961407e-06,
+      "loss": 0.6189,
+      "step": 6292
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.232485359666232,
+      "learning_rate": 2.7058612926264634e-06,
+      "loss": 0.6106,
+      "step": 6293
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9749775537277983,
+      "learning_rate": 2.7043471672831866e-06,
+      "loss": 0.597,
+      "step": 6294
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.570926833062,
+      "learning_rate": 2.7028333086422232e-06,
+      "loss": 0.5325,
+      "step": 6295
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.58096398136124,
+      "learning_rate": 2.7013197168794424e-06,
+      "loss": 0.6335,
+      "step": 6296
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1566887676933293,
+      "learning_rate": 2.699806392170693e-06,
+      "loss": 0.6744,
+      "step": 6297
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2365763878000284,
+      "learning_rate": 2.698293334691789e-06,
+      "loss": 0.6178,
+      "step": 6298
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.206253358431679,
+      "learning_rate": 2.69678054461851e-06,
+      "loss": 0.6438,
+      "step": 6299
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4308841962891514,
+      "learning_rate": 2.6952680221266116e-06,
+      "loss": 0.6907,
+      "step": 6300
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1155846785270302,
+      "learning_rate": 2.6937557673918096e-06,
+      "loss": 0.5513,
+      "step": 6301
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.102074075556992,
+      "learning_rate": 2.692243780589798e-06,
+      "loss": 0.6146,
+      "step": 6302
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.119358235355663,
+      "learning_rate": 2.6907320618962312e-06,
+      "loss": 0.6069,
+      "step": 6303
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.52908697678268,
+      "learning_rate": 2.6892206114867402e-06,
+      "loss": 0.5663,
+      "step": 6304
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6044059425378716,
+      "learning_rate": 2.6877094295369167e-06,
+      "loss": 0.5864,
+      "step": 6305
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3129237260574933,
+      "learning_rate": 2.686198516222329e-06,
+      "loss": 0.6366,
+      "step": 6306
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.699401934868816,
+      "learning_rate": 2.6846878717185076e-06,
+      "loss": 0.5805,
+      "step": 6307
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5503693982830806,
+      "learning_rate": 2.6831774962009582e-06,
+      "loss": 0.5873,
+      "step": 6308
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3528557272139152,
+      "learning_rate": 2.6816673898451486e-06,
+      "loss": 0.6644,
+      "step": 6309
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.178316075395149,
+      "learning_rate": 2.680157552826519e-06,
+      "loss": 0.5794,
+      "step": 6310
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.640485918878129,
+      "learning_rate": 2.6786479853204817e-06,
+      "loss": 0.7067,
+      "step": 6311
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.296387374049739,
+      "learning_rate": 2.6771386875024087e-06,
+      "loss": 0.6266,
+      "step": 6312
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1896654863553193,
+      "learning_rate": 2.6756296595476504e-06,
+      "loss": 0.5763,
+      "step": 6313
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.110961387612748,
+      "learning_rate": 2.674120901631517e-06,
+      "loss": 0.5546,
+      "step": 6314
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.028470459580835,
+      "learning_rate": 2.6726124139292964e-06,
+      "loss": 0.59,
+      "step": 6315
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3573162744903935,
+      "learning_rate": 2.6711041966162356e-06,
+      "loss": 0.638,
+      "step": 6316
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.738834086837961,
+      "learning_rate": 2.6695962498675588e-06,
+      "loss": 0.6098,
+      "step": 6317
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3705691571110727,
+      "learning_rate": 2.6680885738584512e-06,
+      "loss": 0.6687,
+      "step": 6318
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.548261604744699,
+      "learning_rate": 2.6665811687640723e-06,
+      "loss": 0.6152,
+      "step": 6319
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3601953378527485,
+      "learning_rate": 2.66507403475955e-06,
+      "loss": 0.5781,
+      "step": 6320
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.7282332847963287,
+      "learning_rate": 2.663567172019977e-06,
+      "loss": 0.6588,
+      "step": 6321
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.647581964093641,
+      "learning_rate": 2.6620605807204134e-06,
+      "loss": 0.5947,
+      "step": 6322
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.55204790802501,
+      "learning_rate": 2.660554261035894e-06,
+      "loss": 0.6317,
+      "step": 6323
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.794476567030957,
+      "learning_rate": 2.659048213141419e-06,
+      "loss": 0.5403,
+      "step": 6324
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5186567163194873,
+      "learning_rate": 2.657542437211954e-06,
+      "loss": 0.6475,
+      "step": 6325
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.715952090453819,
+      "learning_rate": 2.6560369334224396e-06,
+      "loss": 0.565,
+      "step": 6326
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.700955615524671,
+      "learning_rate": 2.6545317019477764e-06,
+      "loss": 0.6937,
+      "step": 6327
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.7950586784557565,
+      "learning_rate": 2.653026742962842e-06,
+      "loss": 0.6282,
+      "step": 6328
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.932271397052837,
+      "learning_rate": 2.6515220566424735e-06,
+      "loss": 0.5145,
+      "step": 6329
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.426433667015017,
+      "learning_rate": 2.6500176431614866e-06,
+      "loss": 0.6701,
+      "step": 6330
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.267463500168662,
+      "learning_rate": 2.6485135026946545e-06,
+      "loss": 0.62,
+      "step": 6331
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4952915986161748,
+      "learning_rate": 2.6470096354167264e-06,
+      "loss": 0.6529,
+      "step": 6332
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.28610014397335,
+      "learning_rate": 2.645506041502419e-06,
+      "loss": 0.5839,
+      "step": 6333
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8188123888897585,
+      "learning_rate": 2.644002721126413e-06,
+      "loss": 0.6255,
+      "step": 6334
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.831571559662532,
+      "learning_rate": 2.642499674463359e-06,
+      "loss": 0.536,
+      "step": 6335
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.58837692606335,
+      "learning_rate": 2.640996901687878e-06,
+      "loss": 0.6061,
+      "step": 6336
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3253447670569956,
+      "learning_rate": 2.6394944029745594e-06,
+      "loss": 0.6083,
+      "step": 6337
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.1161500571551355,
+      "learning_rate": 2.6379921784979556e-06,
+      "loss": 0.6626,
+      "step": 6338
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3957780596036167,
+      "learning_rate": 2.6364902284325943e-06,
+      "loss": 0.5869,
+      "step": 6339
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.14685635706374,
+      "learning_rate": 2.634988552952965e-06,
+      "loss": 0.6182,
+      "step": 6340
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.175501966326075,
+      "learning_rate": 2.63348715223353e-06,
+      "loss": 0.5976,
+      "step": 6341
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.2865872103188165,
+      "learning_rate": 2.6319860264487156e-06,
+      "loss": 0.6499,
+      "step": 6342
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.282251344333547,
+      "learning_rate": 2.630485175772921e-06,
+      "loss": 0.6299,
+      "step": 6343
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.959066483609487,
+      "learning_rate": 2.6289846003805073e-06,
+      "loss": 0.7163,
+      "step": 6344
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6533855270391533,
+      "learning_rate": 2.6274843004458083e-06,
+      "loss": 0.6091,
+      "step": 6345
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.092500503413888,
+      "learning_rate": 2.6259842761431275e-06,
+      "loss": 0.6512,
+      "step": 6346
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 11.63033721500452,
+      "learning_rate": 2.624484527646729e-06,
+      "loss": 0.5766,
+      "step": 6347
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.344316205856234,
+      "learning_rate": 2.6229850551308533e-06,
+      "loss": 0.6893,
+      "step": 6348
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.7670480341648984,
+      "learning_rate": 2.6214858587697e-06,
+      "loss": 0.6076,
+      "step": 6349
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.7552634595866894,
+      "learning_rate": 2.6199869387374465e-06,
+      "loss": 0.626,
+      "step": 6350
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.087916955904863,
+      "learning_rate": 2.6184882952082286e-06,
+      "loss": 0.6331,
+      "step": 6351
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2553742098034797,
+      "learning_rate": 2.616989928356158e-06,
+      "loss": 0.6029,
+      "step": 6352
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4877624793792266,
+      "learning_rate": 2.6154918383553075e-06,
+      "loss": 0.5829,
+      "step": 6353
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.164663373664098,
+      "learning_rate": 2.6139940253797237e-06,
+      "loss": 0.5857,
+      "step": 6354
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2967240401034346,
+      "learning_rate": 2.6124964896034143e-06,
+      "loss": 0.5886,
+      "step": 6355
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.099408237564359,
+      "learning_rate": 2.610999231200364e-06,
+      "loss": 0.6851,
+      "step": 6356
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3689665901694243,
+      "learning_rate": 2.6095022503445155e-06,
+      "loss": 0.6386,
+      "step": 6357
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2222788180650475,
+      "learning_rate": 2.6080055472097844e-06,
+      "loss": 0.638,
+      "step": 6358
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4258550761659854,
+      "learning_rate": 2.6065091219700568e-06,
+      "loss": 0.5755,
+      "step": 6359
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5822158507581823,
+      "learning_rate": 2.605012974799178e-06,
+      "loss": 0.6109,
+      "step": 6360
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.344258642918491,
+      "learning_rate": 2.603517105870971e-06,
+      "loss": 0.5318,
+      "step": 6361
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.661742078883693,
+      "learning_rate": 2.602021515359218e-06,
+      "loss": 0.5598,
+      "step": 6362
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2545924179160295,
+      "learning_rate": 2.600526203437674e-06,
+      "loss": 0.5684,
+      "step": 6363
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.7470556591551687,
+      "learning_rate": 2.5990311702800573e-06,
+      "loss": 0.6562,
+      "step": 6364
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.130529569635912,
+      "learning_rate": 2.597536416060062e-06,
+      "loss": 0.6661,
+      "step": 6365
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.848914985108963,
+      "learning_rate": 2.5960419409513386e-06,
+      "loss": 0.4984,
+      "step": 6366
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.47331013855302,
+      "learning_rate": 2.594547745127514e-06,
+      "loss": 0.6535,
+      "step": 6367
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1487059851875383,
+      "learning_rate": 2.5930538287621797e-06,
+      "loss": 0.6188,
+      "step": 6368
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2329154737791517,
+      "learning_rate": 2.591560192028894e-06,
+      "loss": 0.5626,
+      "step": 6369
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2654061918388013,
+      "learning_rate": 2.5900668351011815e-06,
+      "loss": 0.6215,
+      "step": 6370
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.256650550756588,
+      "learning_rate": 2.588573758152538e-06,
+      "loss": 0.631,
+      "step": 6371
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.001940811368443,
+      "learning_rate": 2.5870809613564264e-06,
+      "loss": 0.615,
+      "step": 6372
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.097482873027538,
+      "learning_rate": 2.585588444886271e-06,
+      "loss": 0.6168,
+      "step": 6373
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.401718053393286,
+      "learning_rate": 2.584096208915473e-06,
+      "loss": 0.6036,
+      "step": 6374
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.7170902456231816,
+      "learning_rate": 2.5826042536173923e-06,
+      "loss": 0.6435,
+      "step": 6375
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0656142552031445,
+      "learning_rate": 2.581112579165363e-06,
+      "loss": 0.6798,
+      "step": 6376
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.885892388490017,
+      "learning_rate": 2.5796211857326805e-06,
+      "loss": 0.5807,
+      "step": 6377
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.357443586015753,
+      "learning_rate": 2.578130073492613e-06,
+      "loss": 0.5938,
+      "step": 6378
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3457746446640377,
+      "learning_rate": 2.576639242618391e-06,
+      "loss": 0.6046,
+      "step": 6379
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0130372996055184,
+      "learning_rate": 2.575148693283217e-06,
+      "loss": 0.5469,
+      "step": 6380
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2471983613829685,
+      "learning_rate": 2.5736584256602604e-06,
+      "loss": 0.6914,
+      "step": 6381
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8049842929766076,
+      "learning_rate": 2.572168439922653e-06,
+      "loss": 0.6207,
+      "step": 6382
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5242050580331536,
+      "learning_rate": 2.570678736243497e-06,
+      "loss": 0.5769,
+      "step": 6383
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.7371668921754995,
+      "learning_rate": 2.569189314795863e-06,
+      "loss": 0.6394,
+      "step": 6384
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.541518894090096,
+      "learning_rate": 2.56770017575279e-06,
+      "loss": 0.5863,
+      "step": 6385
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.104105510648952,
+      "learning_rate": 2.566211319287276e-06,
+      "loss": 0.5289,
+      "step": 6386
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5996518622471667,
+      "learning_rate": 2.564722745572299e-06,
+      "loss": 0.6135,
+      "step": 6387
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2459414237155215,
+      "learning_rate": 2.563234454780791e-06,
+      "loss": 0.5344,
+      "step": 6388
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6467790790692405,
+      "learning_rate": 2.561746447085662e-06,
+      "loss": 0.5519,
+      "step": 6389
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.9589463310028505,
+      "learning_rate": 2.5602587226597813e-06,
+      "loss": 0.61,
+      "step": 6390
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.81740329183419,
+      "learning_rate": 2.5587712816759914e-06,
+      "loss": 0.6797,
+      "step": 6391
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4235942299105417,
+      "learning_rate": 2.5572841243070944e-06,
+      "loss": 0.6292,
+      "step": 6392
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.7257768033845178,
+      "learning_rate": 2.5557972507258676e-06,
+      "loss": 0.5951,
+      "step": 6393
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.9980654800137065,
+      "learning_rate": 2.554310661105052e-06,
+      "loss": 0.5879,
+      "step": 6394
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.45714129335474,
+      "learning_rate": 2.5528243556173526e-06,
+      "loss": 0.5924,
+      "step": 6395
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.524381334524273,
+      "learning_rate": 2.551338334435447e-06,
+      "loss": 0.5851,
+      "step": 6396
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2895599260578168,
+      "learning_rate": 2.549852597731973e-06,
+      "loss": 0.6139,
+      "step": 6397
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2734502037433173,
+      "learning_rate": 2.5483671456795446e-06,
+      "loss": 0.5641,
+      "step": 6398
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.10732024204948,
+      "learning_rate": 2.5468819784507314e-06,
+      "loss": 0.6338,
+      "step": 6399
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.831440417203488,
+      "learning_rate": 2.545397096218081e-06,
+      "loss": 0.632,
+      "step": 6400
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2175477294639503,
+      "learning_rate": 2.5439124991540986e-06,
+      "loss": 0.5228,
+      "step": 6401
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 11.26293371493835,
+      "learning_rate": 2.5424281874312616e-06,
+      "loss": 0.638,
+      "step": 6402
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.070512413502359,
+      "learning_rate": 2.5409441612220163e-06,
+      "loss": 0.5331,
+      "step": 6403
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.43982362945709,
+      "learning_rate": 2.53946042069877e-06,
+      "loss": 0.682,
+      "step": 6404
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5481585409694536,
+      "learning_rate": 2.537976966033897e-06,
+      "loss": 0.6913,
+      "step": 6405
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.780707259096246,
+      "learning_rate": 2.5364937973997433e-06,
+      "loss": 0.637,
+      "step": 6406
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.059638272022161,
+      "learning_rate": 2.535010914968621e-06,
+      "loss": 0.6197,
+      "step": 6407
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.198201640329154,
+      "learning_rate": 2.533528318912803e-06,
+      "loss": 0.544,
+      "step": 6408
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3114583069479973,
+      "learning_rate": 2.532046009404537e-06,
+      "loss": 0.5643,
+      "step": 6409
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.0785166014115655,
+      "learning_rate": 2.5305639866160293e-06,
+      "loss": 0.6458,
+      "step": 6410
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4149803287414717,
+      "learning_rate": 2.529082250719461e-06,
+      "loss": 0.7269,
+      "step": 6411
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.160446913236817,
+      "learning_rate": 2.5276008018869722e-06,
+      "loss": 0.5832,
+      "step": 6412
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6820344561724,
+      "learning_rate": 2.526119640290678e-06,
+      "loss": 0.5942,
+      "step": 6413
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.559166061869424,
+      "learning_rate": 2.5246387661026504e-06,
+      "loss": 0.6239,
+      "step": 6414
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.35834972171533,
+      "learning_rate": 2.5231581794949356e-06,
+      "loss": 0.535,
+      "step": 6415
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.964214903325439,
+      "learning_rate": 2.5216778806395448e-06,
+      "loss": 0.6452,
+      "step": 6416
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.200465378179132,
+      "learning_rate": 2.520197869708454e-06,
+      "loss": 0.6412,
+      "step": 6417
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5628414197927607,
+      "learning_rate": 2.518718146873605e-06,
+      "loss": 0.6189,
+      "step": 6418
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3739952466153516,
+      "learning_rate": 2.5172387123069085e-06,
+      "loss": 0.5864,
+      "step": 6419
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2088475058648496,
+      "learning_rate": 2.5157595661802437e-06,
+      "loss": 0.6126,
+      "step": 6420
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.8096193612985947,
+      "learning_rate": 2.514280708665449e-06,
+      "loss": 0.48,
+      "step": 6421
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8758991836794494,
+      "learning_rate": 2.5128021399343385e-06,
+      "loss": 0.601,
+      "step": 6422
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.354943435805703,
+      "learning_rate": 2.5113238601586833e-06,
+      "loss": 0.5763,
+      "step": 6423
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8991492339114561,
+      "learning_rate": 2.50984586951023e-06,
+      "loss": 0.571,
+      "step": 6424
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.988868169199592,
+      "learning_rate": 2.508368168160683e-06,
+      "loss": 0.6109,
+      "step": 6425
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0181480966879133,
+      "learning_rate": 2.5068907562817223e-06,
+      "loss": 0.6475,
+      "step": 6426
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.607017873285922,
+      "learning_rate": 2.505413634044984e-06,
+      "loss": 0.6466,
+      "step": 6427
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5325358752326745,
+      "learning_rate": 2.5039368016220795e-06,
+      "loss": 0.707,
+      "step": 6428
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.495898484741088,
+      "learning_rate": 2.502460259184584e-06,
+      "loss": 0.6143,
+      "step": 6429
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.27497515485443,
+      "learning_rate": 2.500984006904035e-06,
+      "loss": 0.6036,
+      "step": 6430
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.384554479372134,
+      "learning_rate": 2.4995080449519383e-06,
+      "loss": 0.632,
+      "step": 6431
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.136269004392376,
+      "learning_rate": 2.498032373499769e-06,
+      "loss": 0.5979,
+      "step": 6432
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9591439151746047,
+      "learning_rate": 2.496556992718968e-06,
+      "loss": 0.5752,
+      "step": 6433
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.625755986560923,
+      "learning_rate": 2.495081902780937e-06,
+      "loss": 0.6251,
+      "step": 6434
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.6993322682560046,
+      "learning_rate": 2.4936071038570514e-06,
+      "loss": 0.6101,
+      "step": 6435
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.606064994507977,
+      "learning_rate": 2.4921325961186455e-06,
+      "loss": 0.6113,
+      "step": 6436
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.8817794595876305,
+      "learning_rate": 2.490658379737025e-06,
+      "loss": 0.6152,
+      "step": 6437
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.592817222457228,
+      "learning_rate": 2.489184454883462e-06,
+      "loss": 0.6595,
+      "step": 6438
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.112454316014022,
+      "learning_rate": 2.4877108217291913e-06,
+      "loss": 0.6833,
+      "step": 6439
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.089441060610151,
+      "learning_rate": 2.4862374804454127e-06,
+      "loss": 0.6656,
+      "step": 6440
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1810446892787994,
+      "learning_rate": 2.484764431203297e-06,
+      "loss": 0.655,
+      "step": 6441
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2182929541942453,
+      "learning_rate": 2.483291674173981e-06,
+      "loss": 0.6376,
+      "step": 6442
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.620256297362581,
+      "learning_rate": 2.4818192095285615e-06,
+      "loss": 0.6077,
+      "step": 6443
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.959409963615247,
+      "learning_rate": 2.4803470374381084e-06,
+      "loss": 0.5402,
+      "step": 6444
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.690295499658671,
+      "learning_rate": 2.4788751580736516e-06,
+      "loss": 0.637,
+      "step": 6445
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2246955758634246,
+      "learning_rate": 2.4774035716061924e-06,
+      "loss": 0.6423,
+      "step": 6446
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4753333122148784,
+      "learning_rate": 2.4759322782066924e-06,
+      "loss": 0.6151,
+      "step": 6447
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3261687913389704,
+      "learning_rate": 2.4744612780460863e-06,
+      "loss": 0.6234,
+      "step": 6448
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.050091521356472,
+      "learning_rate": 2.472990571295266e-06,
+      "loss": 0.528,
+      "step": 6449
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.266129287841651,
+      "learning_rate": 2.4715201581250962e-06,
+      "loss": 0.6456,
+      "step": 6450
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5040000923789902,
+      "learning_rate": 2.4700500387064074e-06,
+      "loss": 0.6231,
+      "step": 6451
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.405010632193548,
+      "learning_rate": 2.4685802132099923e-06,
+      "loss": 0.5844,
+      "step": 6452
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.6378566394399967,
+      "learning_rate": 2.4671106818066076e-06,
+      "loss": 0.6313,
+      "step": 6453
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.429331136222701,
+      "learning_rate": 2.465641444666983e-06,
+      "loss": 0.6776,
+      "step": 6454
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.3515232887180457,
+      "learning_rate": 2.4641725019618107e-06,
+      "loss": 0.5311,
+      "step": 6455
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.516974554330432,
+      "learning_rate": 2.4627038538617447e-06,
+      "loss": 0.5724,
+      "step": 6456
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1215303565599735,
+      "learning_rate": 2.461235500537412e-06,
+      "loss": 0.5501,
+      "step": 6457
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0665664683899,
+      "learning_rate": 2.4597674421593985e-06,
+      "loss": 0.6118,
+      "step": 6458
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.723504300822161,
+      "learning_rate": 2.458299678898263e-06,
+      "loss": 0.6764,
+      "step": 6459
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4054502421315638,
+      "learning_rate": 2.456832210924521e-06,
+      "loss": 0.6773,
+      "step": 6460
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.408884797009261,
+      "learning_rate": 2.455365038408663e-06,
+      "loss": 0.6166,
+      "step": 6461
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.446288286832739,
+      "learning_rate": 2.453898161521137e-06,
+      "loss": 0.703,
+      "step": 6462
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2957817402286795,
+      "learning_rate": 2.4524315804323627e-06,
+      "loss": 0.6984,
+      "step": 6463
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.401524156010693,
+      "learning_rate": 2.4509652953127257e-06,
+      "loss": 0.6117,
+      "step": 6464
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.328086050605523,
+      "learning_rate": 2.4494993063325716e-06,
+      "loss": 0.5302,
+      "step": 6465
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.358380771233323,
+      "learning_rate": 2.4480336136622133e-06,
+      "loss": 0.6143,
+      "step": 6466
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.6392348923503626,
+      "learning_rate": 2.446568217471933e-06,
+      "loss": 0.5916,
+      "step": 6467
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7595861356010456,
+      "learning_rate": 2.445103117931978e-06,
+      "loss": 0.5833,
+      "step": 6468
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1230177292364067,
+      "learning_rate": 2.443638315212555e-06,
+      "loss": 0.5654,
+      "step": 6469
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.8062776463140517,
+      "learning_rate": 2.442173809483845e-06,
+      "loss": 0.7762,
+      "step": 6470
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.716571799479343,
+      "learning_rate": 2.440709600915986e-06,
+      "loss": 0.6991,
+      "step": 6471
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7930103373086728,
+      "learning_rate": 2.4392456896790874e-06,
+      "loss": 0.5541,
+      "step": 6472
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.224499151376592,
+      "learning_rate": 2.437782075943224e-06,
+      "loss": 0.6459,
+      "step": 6473
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.395183867623381,
+      "learning_rate": 2.4363187598784323e-06,
+      "loss": 0.6556,
+      "step": 6474
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1451834322017946,
+      "learning_rate": 2.4348557416547146e-06,
+      "loss": 0.6845,
+      "step": 6475
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.6454970175141983,
+      "learning_rate": 2.4333930214420414e-06,
+      "loss": 0.6793,
+      "step": 6476
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1614550552826097,
+      "learning_rate": 2.43193059941035e-06,
+      "loss": 0.6409,
+      "step": 6477
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9560127452631026,
+      "learning_rate": 2.4304684757295376e-06,
+      "loss": 0.6685,
+      "step": 6478
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.385228890681212,
+      "learning_rate": 2.429006650569468e-06,
+      "loss": 0.6054,
+      "step": 6479
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7023174068498594,
+      "learning_rate": 2.4275451240999743e-06,
+      "loss": 0.6734,
+      "step": 6480
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.8800257059035994,
+      "learning_rate": 2.4260838964908534e-06,
+      "loss": 0.6168,
+      "step": 6481
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3173592174188915,
+      "learning_rate": 2.424622967911863e-06,
+      "loss": 0.6082,
+      "step": 6482
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9950821121430008,
+      "learning_rate": 2.4231623385327337e-06,
+      "loss": 0.5374,
+      "step": 6483
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9226988485243646,
+      "learning_rate": 2.421702008523153e-06,
+      "loss": 0.5921,
+      "step": 6484
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2019190672941673,
+      "learning_rate": 2.4202419780527796e-06,
+      "loss": 0.6282,
+      "step": 6485
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 18.87914111722867,
+      "learning_rate": 2.418782247291238e-06,
+      "loss": 0.5984,
+      "step": 6486
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9756191834087302,
+      "learning_rate": 2.4173228164081135e-06,
+      "loss": 0.5355,
+      "step": 6487
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.229047102096132,
+      "learning_rate": 2.4158636855729563e-06,
+      "loss": 0.6097,
+      "step": 6488
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.160615604668489,
+      "learning_rate": 2.414404854955286e-06,
+      "loss": 0.5794,
+      "step": 6489
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5744631310901793,
+      "learning_rate": 2.4129463247245877e-06,
+      "loss": 0.6365,
+      "step": 6490
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4664632502489976,
+      "learning_rate": 2.411488095050305e-06,
+      "loss": 0.6087,
+      "step": 6491
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2462459640267336,
+      "learning_rate": 2.410030166101855e-06,
+      "loss": 0.6314,
+      "step": 6492
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.7690797181208953,
+      "learning_rate": 2.4085725380486106e-06,
+      "loss": 0.6211,
+      "step": 6493
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3187693057872476,
+      "learning_rate": 2.4071152110599204e-06,
+      "loss": 0.5885,
+      "step": 6494
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4839016053608605,
+      "learning_rate": 2.4056581853050877e-06,
+      "loss": 0.6435,
+      "step": 6495
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.5230696548262035,
+      "learning_rate": 2.4042014609533894e-06,
+      "loss": 0.5636,
+      "step": 6496
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.54975560910205,
+      "learning_rate": 2.4027450381740598e-06,
+      "loss": 0.6318,
+      "step": 6497
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1179176495059173,
+      "learning_rate": 2.4012889171363034e-06,
+      "loss": 0.6436,
+      "step": 6498
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.219085902942063,
+      "learning_rate": 2.3998330980092906e-06,
+      "loss": 0.5653,
+      "step": 6499
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4057964800651495,
+      "learning_rate": 2.3983775809621525e-06,
+      "loss": 0.6256,
+      "step": 6500
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9344149719041113,
+      "learning_rate": 2.3969223661639838e-06,
+      "loss": 0.6745,
+      "step": 6501
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.7761690986099326,
+      "learning_rate": 2.395467453783851e-06,
+      "loss": 0.5358,
+      "step": 6502
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3942022138130055,
+      "learning_rate": 2.394012843990781e-06,
+      "loss": 0.664,
+      "step": 6503
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3279945476576263,
+      "learning_rate": 2.3925585369537647e-06,
+      "loss": 0.6307,
+      "step": 6504
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.007225583568133,
+      "learning_rate": 2.391104532841762e-06,
+      "loss": 0.6179,
+      "step": 6505
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.6218819183172104,
+      "learning_rate": 2.389650831823691e-06,
+      "loss": 0.6501,
+      "step": 6506
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3080723378936723,
+      "learning_rate": 2.388197434068441e-06,
+      "loss": 0.6166,
+      "step": 6507
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2881275417992613,
+      "learning_rate": 2.3867443397448646e-06,
+      "loss": 0.622,
+      "step": 6508
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.241683983187079,
+      "learning_rate": 2.3852915490217772e-06,
+      "loss": 0.6207,
+      "step": 6509
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.556095245479891,
+      "learning_rate": 2.383839062067957e-06,
+      "loss": 0.5998,
+      "step": 6510
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9799784118007415,
+      "learning_rate": 2.382386879052152e-06,
+      "loss": 0.6074,
+      "step": 6511
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3182312355329513,
+      "learning_rate": 2.3809350001430743e-06,
+      "loss": 0.5951,
+      "step": 6512
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3694748756782027,
+      "learning_rate": 2.3794834255093977e-06,
+      "loss": 0.6304,
+      "step": 6513
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2001457523740147,
+      "learning_rate": 2.37803215531976e-06,
+      "loss": 0.6331,
+      "step": 6514
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3212863403370796,
+      "learning_rate": 2.3765811897427667e-06,
+      "loss": 0.53,
+      "step": 6515
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.160468095861783,
+      "learning_rate": 2.375130528946989e-06,
+      "loss": 0.7405,
+      "step": 6516
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6855144317046697,
+      "learning_rate": 2.373680173100957e-06,
+      "loss": 0.5849,
+      "step": 6517
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.341892568924303,
+      "learning_rate": 2.3722301223731724e-06,
+      "loss": 0.5603,
+      "step": 6518
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.283085350966203,
+      "learning_rate": 2.3707803769320943e-06,
+      "loss": 0.6285,
+      "step": 6519
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.545450875140134,
+      "learning_rate": 2.3693309369461514e-06,
+      "loss": 0.643,
+      "step": 6520
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.34738226126757,
+      "learning_rate": 2.367881802583738e-06,
+      "loss": 0.6825,
+      "step": 6521
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6575846961053333,
+      "learning_rate": 2.366432974013208e-06,
+      "loss": 0.5756,
+      "step": 6522
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6040764810860146,
+      "learning_rate": 2.36498445140288e-06,
+      "loss": 0.6637,
+      "step": 6523
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.575034774458675,
+      "learning_rate": 2.3635362349210423e-06,
+      "loss": 0.6506,
+      "step": 6524
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.561154528891339,
+      "learning_rate": 2.362088324735945e-06,
+      "loss": 0.585,
+      "step": 6525
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.557511457700825,
+      "learning_rate": 2.3606407210158007e-06,
+      "loss": 0.7341,
+      "step": 6526
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.151485466378373,
+      "learning_rate": 2.3591934239287858e-06,
+      "loss": 0.5364,
+      "step": 6527
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.2369051385886,
+      "learning_rate": 2.3577464336430446e-06,
+      "loss": 0.5904,
+      "step": 6528
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4237630485347683,
+      "learning_rate": 2.356299750326687e-06,
+      "loss": 0.6413,
+      "step": 6529
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.791157720967069,
+      "learning_rate": 2.3548533741477807e-06,
+      "loss": 0.6735,
+      "step": 6530
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3572226180240285,
+      "learning_rate": 2.353407305274365e-06,
+      "loss": 0.6745,
+      "step": 6531
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.756415728746884,
+      "learning_rate": 2.3519615438744358e-06,
+      "loss": 0.6098,
+      "step": 6532
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.9211036631117326,
+      "learning_rate": 2.3505160901159596e-06,
+      "loss": 0.6391,
+      "step": 6533
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.804798236915875,
+      "learning_rate": 2.3490709441668673e-06,
+      "loss": 0.6013,
+      "step": 6534
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3231398228966333,
+      "learning_rate": 2.34762610619505e-06,
+      "loss": 0.6146,
+      "step": 6535
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.496092282256762,
+      "learning_rate": 2.346181576368362e-06,
+      "loss": 0.7055,
+      "step": 6536
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.487885892384019,
+      "learning_rate": 2.344737354854627e-06,
+      "loss": 0.7101,
+      "step": 6537
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.2049421861180685,
+      "learning_rate": 2.343293441821633e-06,
+      "loss": 0.672,
+      "step": 6538
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9953645112584768,
+      "learning_rate": 2.3418498374371266e-06,
+      "loss": 0.6358,
+      "step": 6539
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.7737257283166983,
+      "learning_rate": 2.3404065418688203e-06,
+      "loss": 0.6623,
+      "step": 6540
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2828290607769466,
+      "learning_rate": 2.3389635552843943e-06,
+      "loss": 0.6073,
+      "step": 6541
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.7921110029438876,
+      "learning_rate": 2.3375208778514903e-06,
+      "loss": 0.6668,
+      "step": 6542
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.963063546111307,
+      "learning_rate": 2.336078509737715e-06,
+      "loss": 0.7043,
+      "step": 6543
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9709336196500105,
+      "learning_rate": 2.334636451110639e-06,
+      "loss": 0.5716,
+      "step": 6544
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.052297383518725,
+      "learning_rate": 2.333194702137793e-06,
+      "loss": 0.6293,
+      "step": 6545
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2880945024427444,
+      "learning_rate": 2.331753262986678e-06,
+      "loss": 0.5453,
+      "step": 6546
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.578393672808325,
+      "learning_rate": 2.330312133824757e-06,
+      "loss": 0.7206,
+      "step": 6547
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.010515317246663,
+      "learning_rate": 2.3288713148194554e-06,
+      "loss": 0.6199,
+      "step": 6548
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9927985217423352,
+      "learning_rate": 2.3274308061381605e-06,
+      "loss": 0.6139,
+      "step": 6549
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9806575010532403,
+      "learning_rate": 2.32599060794823e-06,
+      "loss": 0.6179,
+      "step": 6550
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.541089321186657,
+      "learning_rate": 2.324550720416982e-06,
+      "loss": 0.5371,
+      "step": 6551
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.060488905830918,
+      "learning_rate": 2.3231111437116954e-06,
+      "loss": 0.6226,
+      "step": 6552
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.944439269515085,
+      "learning_rate": 2.3216718779996205e-06,
+      "loss": 0.6055,
+      "step": 6553
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8098524227724737,
+      "learning_rate": 2.320232923447962e-06,
+      "loss": 0.6007,
+      "step": 6554
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.017792625313086,
+      "learning_rate": 2.318794280223897e-06,
+      "loss": 0.6693,
+      "step": 6555
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2085939033581856,
+      "learning_rate": 2.317355948494563e-06,
+      "loss": 0.5985,
+      "step": 6556
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4413921495254867,
+      "learning_rate": 2.31591792842706e-06,
+      "loss": 0.6141,
+      "step": 6557
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.270526009720749,
+      "learning_rate": 2.314480220188452e-06,
+      "loss": 0.6078,
+      "step": 6558
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3574965924340434,
+      "learning_rate": 2.3130428239457688e-06,
+      "loss": 0.6009,
+      "step": 6559
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.673316161467211,
+      "learning_rate": 2.3116057398660046e-06,
+      "loss": 0.5492,
+      "step": 6560
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.7207174172029065,
+      "learning_rate": 2.3101689681161142e-06,
+      "loss": 0.6563,
+      "step": 6561
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.307943954138673,
+      "learning_rate": 2.308732508863016e-06,
+      "loss": 0.5498,
+      "step": 6562
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.299414059442358,
+      "learning_rate": 2.307296362273595e-06,
+      "loss": 0.6527,
+      "step": 6563
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.0267194850070918,
+      "learning_rate": 2.305860528514701e-06,
+      "loss": 0.6284,
+      "step": 6564
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6543376447272022,
+      "learning_rate": 2.304425007753141e-06,
+      "loss": 0.6456,
+      "step": 6565
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2467729720442384,
+      "learning_rate": 2.3029898001556928e-06,
+      "loss": 0.5627,
+      "step": 6566
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.086285543773585,
+      "learning_rate": 2.301554905889092e-06,
+      "loss": 0.5767,
+      "step": 6567
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.212741108195205,
+      "learning_rate": 2.3001203251200417e-06,
+      "loss": 0.5448,
+      "step": 6568
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.394587869638318,
+      "learning_rate": 2.2986860580152095e-06,
+      "loss": 0.6423,
+      "step": 6569
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.4031910872820763,
+      "learning_rate": 2.2972521047412223e-06,
+      "loss": 0.6362,
+      "step": 6570
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.483025314036182,
+      "learning_rate": 2.2958184654646705e-06,
+      "loss": 0.4954,
+      "step": 6571
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.548631805057781,
+      "learning_rate": 2.2943851403521123e-06,
+      "loss": 0.7172,
+      "step": 6572
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.61360591907676,
+      "learning_rate": 2.2929521295700695e-06,
+      "loss": 0.6463,
+      "step": 6573
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8748275398323564,
+      "learning_rate": 2.2915194332850233e-06,
+      "loss": 0.6434,
+      "step": 6574
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9897551268506732,
+      "learning_rate": 2.290087051663418e-06,
+      "loss": 0.5195,
+      "step": 6575
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3753623260621426,
+      "learning_rate": 2.288654984871665e-06,
+      "loss": 0.485,
+      "step": 6576
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.635343007038432,
+      "learning_rate": 2.2872232330761383e-06,
+      "loss": 0.6737,
+      "step": 6577
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5493524141073918,
+      "learning_rate": 2.285791796443176e-06,
+      "loss": 0.6043,
+      "step": 6578
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9622930982127853,
+      "learning_rate": 2.284360675139078e-06,
+      "loss": 0.6278,
+      "step": 6579
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.097813348559686,
+      "learning_rate": 2.282929869330104e-06,
+      "loss": 0.6418,
+      "step": 6580
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6376842210758853,
+      "learning_rate": 2.2814993791824836e-06,
+      "loss": 0.6048,
+      "step": 6581
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.058425958808465,
+      "learning_rate": 2.2800692048624092e-06,
+      "loss": 0.6329,
+      "step": 6582
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9682730563914436,
+      "learning_rate": 2.278639346536031e-06,
+      "loss": 0.5591,
+      "step": 6583
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.6023672656105106,
+      "learning_rate": 2.2772098043694656e-06,
+      "loss": 0.6068,
+      "step": 6584
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5349956031101852,
+      "learning_rate": 2.2757805785287946e-06,
+      "loss": 0.5998,
+      "step": 6585
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9259914387049755,
+      "learning_rate": 2.274351669180063e-06,
+      "loss": 0.5596,
+      "step": 6586
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1838133733921383,
+      "learning_rate": 2.272923076489275e-06,
+      "loss": 0.6713,
+      "step": 6587
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.354737589610576,
+      "learning_rate": 2.271494800622399e-06,
+      "loss": 0.5937,
+      "step": 6588
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.3360105669346956,
+      "learning_rate": 2.2700668417453703e-06,
+      "loss": 0.6004,
+      "step": 6589
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.521181364398551,
+      "learning_rate": 2.2686392000240838e-06,
+      "loss": 0.6411,
+      "step": 6590
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3567654473778523,
+      "learning_rate": 2.2672118756244014e-06,
+      "loss": 0.6683,
+      "step": 6591
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2629509475772647,
+      "learning_rate": 2.2657848687121444e-06,
+      "loss": 0.5745,
+      "step": 6592
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1762686821367487,
+      "learning_rate": 2.2643581794530943e-06,
+      "loss": 0.6356,
+      "step": 6593
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4123508723145535,
+      "learning_rate": 2.2629318080130042e-06,
+      "loss": 0.638,
+      "step": 6594
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.803556073787599,
+      "learning_rate": 2.261505754557586e-06,
+      "loss": 0.6717,
+      "step": 6595
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8959725718418525,
+      "learning_rate": 2.260080019252513e-06,
+      "loss": 0.5497,
+      "step": 6596
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3343782073258676,
+      "learning_rate": 2.258654602263421e-06,
+      "loss": 0.6782,
+      "step": 6597
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0752300588139856,
+      "learning_rate": 2.2572295037559135e-06,
+      "loss": 0.6159,
+      "step": 6598
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4484059409198364,
+      "learning_rate": 2.2558047238955547e-06,
+      "loss": 0.6265,
+      "step": 6599
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.0167746887631335,
+      "learning_rate": 2.2543802628478695e-06,
+      "loss": 0.6153,
+      "step": 6600
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.7309284321227167,
+      "learning_rate": 2.2529561207783495e-06,
+      "loss": 0.5535,
+      "step": 6601
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5843645635460657,
+      "learning_rate": 2.251532297852445e-06,
+      "loss": 0.5995,
+      "step": 6602
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1738162447033127,
+      "learning_rate": 2.2501087942355736e-06,
+      "loss": 0.7412,
+      "step": 6603
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4229284055465707,
+      "learning_rate": 2.2486856100931146e-06,
+      "loss": 0.6312,
+      "step": 6604
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8696482572875674,
+      "learning_rate": 2.2472627455904086e-06,
+      "loss": 0.5519,
+      "step": 6605
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.844232232179864,
+      "learning_rate": 2.2458402008927578e-06,
+      "loss": 0.6271,
+      "step": 6606
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.404083137621484,
+      "learning_rate": 2.24441797616543e-06,
+      "loss": 0.5466,
+      "step": 6607
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.567530863025546,
+      "learning_rate": 2.2429960715736588e-06,
+      "loss": 0.5551,
+      "step": 6608
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2882668446276555,
+      "learning_rate": 2.241574487282634e-06,
+      "loss": 0.6007,
+      "step": 6609
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.110230304278789,
+      "learning_rate": 2.24015322345751e-06,
+      "loss": 0.6316,
+      "step": 6610
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 56.41508535977514,
+      "learning_rate": 2.2387322802634065e-06,
+      "loss": 0.6579,
+      "step": 6611
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2521890735234713,
+      "learning_rate": 2.2373116578654042e-06,
+      "loss": 0.6235,
+      "step": 6612
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.234967262280706,
+      "learning_rate": 2.2358913564285496e-06,
+      "loss": 0.5794,
+      "step": 6613
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.976223938280554,
+      "learning_rate": 2.234471376117847e-06,
+      "loss": 0.6562,
+      "step": 6614
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5700256793799388,
+      "learning_rate": 2.2330517170982634e-06,
+      "loss": 0.5574,
+      "step": 6615
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.3477002796845468,
+      "learning_rate": 2.2316323795347334e-06,
+      "loss": 0.6504,
+      "step": 6616
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.155545624420643,
+      "learning_rate": 2.2302133635921524e-06,
+      "loss": 0.6096,
+      "step": 6617
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.271504833154178,
+      "learning_rate": 2.2287946694353764e-06,
+      "loss": 0.6292,
+      "step": 6618
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9615477986067251,
+      "learning_rate": 2.2273762972292227e-06,
+      "loss": 0.5254,
+      "step": 6619
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.347528126787625,
+      "learning_rate": 2.2259582471384765e-06,
+      "loss": 0.5771,
+      "step": 6620
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0790518781036624,
+      "learning_rate": 2.224540519327884e-06,
+      "loss": 0.5152,
+      "step": 6621
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4932893443972515,
+      "learning_rate": 2.2231231139621505e-06,
+      "loss": 0.6342,
+      "step": 6622
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.40013914291326,
+      "learning_rate": 2.2217060312059453e-06,
+      "loss": 0.6919,
+      "step": 6623
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.7257495545768244,
+      "learning_rate": 2.2202892712239016e-06,
+      "loss": 0.6643,
+      "step": 6624
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.029296071523241,
+      "learning_rate": 2.2188728341806153e-06,
+      "loss": 0.5941,
+      "step": 6625
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.587522142486169,
+      "learning_rate": 2.2174567202406455e-06,
+      "loss": 0.689,
+      "step": 6626
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.102337698108396,
+      "learning_rate": 2.2160409295685105e-06,
+      "loss": 0.6263,
+      "step": 6627
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.442746996525038,
+      "learning_rate": 2.2146254623286905e-06,
+      "loss": 0.6833,
+      "step": 6628
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.678411820711548,
+      "learning_rate": 2.213210318685633e-06,
+      "loss": 0.6105,
+      "step": 6629
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.314390976374499,
+      "learning_rate": 2.2117954988037467e-06,
+      "loss": 0.6826,
+      "step": 6630
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3034289358297007,
+      "learning_rate": 2.210381002847399e-06,
+      "loss": 0.6343,
+      "step": 6631
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.9230039460865647,
+      "learning_rate": 2.208966830980921e-06,
+      "loss": 0.629,
+      "step": 6632
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.215059832781754,
+      "learning_rate": 2.207552983368608e-06,
+      "loss": 0.5855,
+      "step": 6633
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.733877578923876,
+      "learning_rate": 2.20613946017472e-06,
+      "loss": 0.561,
+      "step": 6634
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.184821592844948,
+      "learning_rate": 2.2047262615634723e-06,
+      "loss": 0.5693,
+      "step": 6635
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3808532135782983,
+      "learning_rate": 2.203313387699046e-06,
+      "loss": 0.61,
+      "step": 6636
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.185810849417573,
+      "learning_rate": 2.201900838745586e-06,
+      "loss": 0.6877,
+      "step": 6637
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.901166357224412,
+      "learning_rate": 2.2004886148671978e-06,
+      "loss": 0.6226,
+      "step": 6638
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8487030064529053,
+      "learning_rate": 2.1990767162279515e-06,
+      "loss": 0.5769,
+      "step": 6639
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9023289850683952,
+      "learning_rate": 2.197665142991876e-06,
+      "loss": 0.4856,
+      "step": 6640
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.343960901159718,
+      "learning_rate": 2.196253895322961e-06,
+      "loss": 0.5961,
+      "step": 6641
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.467203125378956,
+      "learning_rate": 2.1948429733851646e-06,
+      "loss": 0.6625,
+      "step": 6642
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.702936295244413,
+      "learning_rate": 2.193432377342404e-06,
+      "loss": 0.6413,
+      "step": 6643
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8056618447753277,
+      "learning_rate": 2.1920221073585564e-06,
+      "loss": 0.5931,
+      "step": 6644
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.79010389004188,
+      "learning_rate": 2.190612163597462e-06,
+      "loss": 0.6772,
+      "step": 6645
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.5825220168109295,
+      "learning_rate": 2.189202546222925e-06,
+      "loss": 0.6462,
+      "step": 6646
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.6022305800555974,
+      "learning_rate": 2.1877932553987114e-06,
+      "loss": 0.6701,
+      "step": 6647
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1361674026822683,
+      "learning_rate": 2.1863842912885496e-06,
+      "loss": 0.569,
+      "step": 6648
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.1167050044456364,
+      "learning_rate": 2.184975654056128e-06,
+      "loss": 0.745,
+      "step": 6649
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.735863377596502,
+      "learning_rate": 2.183567343865095e-06,
+      "loss": 0.607,
+      "step": 6650
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4462949583121554,
+      "learning_rate": 2.182159360879067e-06,
+      "loss": 0.587,
+      "step": 6651
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9840779320355102,
+      "learning_rate": 2.1807517052616205e-06,
+      "loss": 0.5347,
+      "step": 6652
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 16.6142045279403,
+      "learning_rate": 2.1793443771762912e-06,
+      "loss": 0.5878,
+      "step": 6653
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1547249179603583,
+      "learning_rate": 2.177937376786577e-06,
+      "loss": 0.5679,
+      "step": 6654
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.7649378439066905,
+      "learning_rate": 2.17653070425594e-06,
+      "loss": 0.6205,
+      "step": 6655
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.1627346207064804,
+      "learning_rate": 2.175124359747806e-06,
+      "loss": 0.653,
+      "step": 6656
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.3916267839635585,
+      "learning_rate": 2.173718343425558e-06,
+      "loss": 0.5896,
+      "step": 6657
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.269804794003583,
+      "learning_rate": 2.1723126554525415e-06,
+      "loss": 0.6527,
+      "step": 6658
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0320795601481907,
+      "learning_rate": 2.1709072959920667e-06,
+      "loss": 0.525,
+      "step": 6659
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3228826509557523,
+      "learning_rate": 2.169502265207404e-06,
+      "loss": 0.5739,
+      "step": 6660
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.952927663361331,
+      "learning_rate": 2.168097563261787e-06,
+      "loss": 0.6047,
+      "step": 6661
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.120080095383413,
+      "learning_rate": 2.1666931903184103e-06,
+      "loss": 0.5871,
+      "step": 6662
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.141689567051116,
+      "learning_rate": 2.1652891465404257e-06,
+      "loss": 0.6272,
+      "step": 6663
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7913975017146035,
+      "learning_rate": 2.1638854320909542e-06,
+      "loss": 0.7013,
+      "step": 6664
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2400371485167927,
+      "learning_rate": 2.162482047133076e-06,
+      "loss": 0.56,
+      "step": 6665
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.3342661126733666,
+      "learning_rate": 2.161078991829832e-06,
+      "loss": 0.635,
+      "step": 6666
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.536358295411925,
+      "learning_rate": 2.159676266344222e-06,
+      "loss": 0.6228,
+      "step": 6667
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.503851980349559,
+      "learning_rate": 2.1582738708392127e-06,
+      "loss": 0.5741,
+      "step": 6668
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.0980965236347804,
+      "learning_rate": 2.1568718054777322e-06,
+      "loss": 0.6151,
+      "step": 6669
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.792431692180165,
+      "learning_rate": 2.1554700704226673e-06,
+      "loss": 0.6017,
+      "step": 6670
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0346739479145444,
+      "learning_rate": 2.1540686658368643e-06,
+      "loss": 0.5778,
+      "step": 6671
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.666257894353589,
+      "learning_rate": 2.1526675918831373e-06,
+      "loss": 0.6578,
+      "step": 6672
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.726228194370597,
+      "learning_rate": 2.151266848724259e-06,
+      "loss": 0.6432,
+      "step": 6673
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5487776785030176,
+      "learning_rate": 2.149866436522965e-06,
+      "loss": 0.6072,
+      "step": 6674
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8073196930271074,
+      "learning_rate": 2.1484663554419495e-06,
+      "loss": 0.6205,
+      "step": 6675
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2783150428762,
+      "learning_rate": 2.147066605643868e-06,
+      "loss": 0.6129,
+      "step": 6676
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.203415894181792,
+      "learning_rate": 2.145667187291341e-06,
+      "loss": 0.6235,
+      "step": 6677
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0439147089340683,
+      "learning_rate": 2.144268100546951e-06,
+      "loss": 0.5993,
+      "step": 6678
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.351141372515001,
+      "learning_rate": 2.1428693455732384e-06,
+      "loss": 0.5968,
+      "step": 6679
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2933943463025726,
+      "learning_rate": 2.141470922532704e-06,
+      "loss": 0.6489,
+      "step": 6680
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3273415621641216,
+      "learning_rate": 2.140072831587815e-06,
+      "loss": 0.635,
+      "step": 6681
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2821768908749736,
+      "learning_rate": 2.138675072900997e-06,
+      "loss": 0.5997,
+      "step": 6682
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.089940278859609,
+      "learning_rate": 2.1372776466346414e-06,
+      "loss": 0.5724,
+      "step": 6683
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.251126225010119,
+      "learning_rate": 2.1358805529510896e-06,
+      "loss": 0.617,
+      "step": 6684
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.760541069813522,
+      "learning_rate": 2.134483792012656e-06,
+      "loss": 0.6546,
+      "step": 6685
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2370760853157,
+      "learning_rate": 2.1330873639816125e-06,
+      "loss": 0.5493,
+      "step": 6686
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.9935072593068037,
+      "learning_rate": 2.131691269020193e-06,
+      "loss": 0.556,
+      "step": 6687
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0100434248241155,
+      "learning_rate": 2.130295507290591e-06,
+      "loss": 0.6344,
+      "step": 6688
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.244649889387862,
+      "learning_rate": 2.1289000789549586e-06,
+      "loss": 0.5579,
+      "step": 6689
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.704981131283807,
+      "learning_rate": 2.1275049841754165e-06,
+      "loss": 0.5746,
+      "step": 6690
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.29539621168737,
+      "learning_rate": 2.126110223114043e-06,
+      "loss": 0.6104,
+      "step": 6691
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.205856297808691,
+      "learning_rate": 2.1247157959328763e-06,
+      "loss": 0.5662,
+      "step": 6692
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8836298348060035,
+      "learning_rate": 2.1233217027939153e-06,
+      "loss": 0.562,
+      "step": 6693
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3976000029984372,
+      "learning_rate": 2.121927943859123e-06,
+      "loss": 0.6278,
+      "step": 6694
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.331546227933603,
+      "learning_rate": 2.1205345192904224e-06,
+      "loss": 0.6523,
+      "step": 6695
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3605376267410083,
+      "learning_rate": 2.1191414292497e-06,
+      "loss": 0.6496,
+      "step": 6696
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.102974401678194,
+      "learning_rate": 2.1177486738987984e-06,
+      "loss": 0.6491,
+      "step": 6697
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.26075427680184,
+      "learning_rate": 2.116356253399522e-06,
+      "loss": 0.5821,
+      "step": 6698
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8908330043379147,
+      "learning_rate": 2.114964167913641e-06,
+      "loss": 0.537,
+      "step": 6699
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.579961013538668,
+      "learning_rate": 2.1135724176028844e-06,
+      "loss": 0.5184,
+      "step": 6700
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.08377361893251,
+      "learning_rate": 2.1121810026289404e-06,
+      "loss": 0.544,
+      "step": 6701
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.241375532358382,
+      "learning_rate": 2.110789923153458e-06,
+      "loss": 0.6789,
+      "step": 6702
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.9160714063650266,
+      "learning_rate": 2.109399179338051e-06,
+      "loss": 0.5792,
+      "step": 6703
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1537715668446755,
+      "learning_rate": 2.1080087713442928e-06,
+      "loss": 0.5489,
+      "step": 6704
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.053631859586391,
+      "learning_rate": 2.1066186993337158e-06,
+      "loss": 0.5573,
+      "step": 6705
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.425494797474714,
+      "learning_rate": 2.105228963467812e-06,
+      "loss": 0.5586,
+      "step": 6706
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.748837747672611,
+      "learning_rate": 2.10383956390804e-06,
+      "loss": 0.5483,
+      "step": 6707
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2002541684184727,
+      "learning_rate": 2.1024505008158153e-06,
+      "loss": 0.6831,
+      "step": 6708
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.309736548355432,
+      "learning_rate": 2.101061774352517e-06,
+      "loss": 0.5418,
+      "step": 6709
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4178982369090267,
+      "learning_rate": 2.099673384679482e-06,
+      "loss": 0.5158,
+      "step": 6710
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.18817854947126,
+      "learning_rate": 2.0982853319580075e-06,
+      "loss": 0.5532,
+      "step": 6711
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 15.098923040125959,
+      "learning_rate": 2.096897616349355e-06,
+      "loss": 0.6052,
+      "step": 6712
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.248992003023729,
+      "learning_rate": 2.0955102380147474e-06,
+      "loss": 0.6149,
+      "step": 6713
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.785961960691239,
+      "learning_rate": 2.0941231971153644e-06,
+      "loss": 0.6334,
+      "step": 6714
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.878088474717053,
+      "learning_rate": 2.0927364938123457e-06,
+      "loss": 0.6535,
+      "step": 6715
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4454378764029507,
+      "learning_rate": 2.0913501282667975e-06,
+      "loss": 0.6169,
+      "step": 6716
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.1028098404336815,
+      "learning_rate": 2.0899641006397836e-06,
+      "loss": 0.6849,
+      "step": 6717
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5675573992295457,
+      "learning_rate": 2.0885784110923325e-06,
+      "loss": 0.6658,
+      "step": 6718
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.0159340806610255,
+      "learning_rate": 2.087193059785421e-06,
+      "loss": 0.6471,
+      "step": 6719
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.879084843294698,
+      "learning_rate": 2.08580804688e-06,
+      "loss": 0.5826,
+      "step": 6720
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.7524142612645415,
+      "learning_rate": 2.084423372536976e-06,
+      "loss": 0.6057,
+      "step": 6721
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.2595597192294785,
+      "learning_rate": 2.083039036917219e-06,
+      "loss": 0.6349,
+      "step": 6722
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.6944542500831257,
+      "learning_rate": 2.0816550401815538e-06,
+      "loss": 0.5963,
+      "step": 6723
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1565697221035833,
+      "learning_rate": 2.0802713824907683e-06,
+      "loss": 0.5843,
+      "step": 6724
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4041247947594564,
+      "learning_rate": 2.0788880640056137e-06,
+      "loss": 0.5929,
+      "step": 6725
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6733679379720794,
+      "learning_rate": 2.077505084886802e-06,
+      "loss": 0.6643,
+      "step": 6726
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2387112348073126,
+      "learning_rate": 2.0761224452950003e-06,
+      "loss": 0.6349,
+      "step": 6727
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8590997656724713,
+      "learning_rate": 2.07474014539084e-06,
+      "loss": 0.6274,
+      "step": 6728
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6204018448125233,
+      "learning_rate": 2.0733581853349128e-06,
+      "loss": 0.699,
+      "step": 6729
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.7073322849788988,
+      "learning_rate": 2.071976565287772e-06,
+      "loss": 0.553,
+      "step": 6730
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0723138136947097,
+      "learning_rate": 2.0705952854099337e-06,
+      "loss": 0.5749,
+      "step": 6731
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.9024005003511815,
+      "learning_rate": 2.069214345861863e-06,
+      "loss": 0.5859,
+      "step": 6732
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.007051839789599,
+      "learning_rate": 2.067833746803998e-06,
+      "loss": 0.5448,
+      "step": 6733
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.16193794853113,
+      "learning_rate": 2.0664534883967315e-06,
+      "loss": 0.6046,
+      "step": 6734
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0022011347301887,
+      "learning_rate": 2.065073570800421e-06,
+      "loss": 0.6004,
+      "step": 6735
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5866230804925356,
+      "learning_rate": 2.0636939941753793e-06,
+      "loss": 0.6142,
+      "step": 6736
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9455346482011558,
+      "learning_rate": 2.0623147586818786e-06,
+      "loss": 0.5795,
+      "step": 6737
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8544304813633348,
+      "learning_rate": 2.060935864480158e-06,
+      "loss": 0.6272,
+      "step": 6738
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.139954944371839,
+      "learning_rate": 2.0595573117304147e-06,
+      "loss": 0.5931,
+      "step": 6739
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.141121600735202,
+      "learning_rate": 2.0581791005928024e-06,
+      "loss": 0.5734,
+      "step": 6740
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1663356934725844,
+      "learning_rate": 2.0568012312274367e-06,
+      "loss": 0.4842,
+      "step": 6741
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1593137328818948,
+      "learning_rate": 2.0554237037943966e-06,
+      "loss": 0.5317,
+      "step": 6742
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.225299678738348,
+      "learning_rate": 2.054046518453718e-06,
+      "loss": 0.6648,
+      "step": 6743
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.46890585145027,
+      "learning_rate": 2.0526696753654008e-06,
+      "loss": 0.648,
+      "step": 6744
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.384393534696619,
+      "learning_rate": 2.051293174689401e-06,
+      "loss": 0.6517,
+      "step": 6745
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3553993298810063,
+      "learning_rate": 2.0499170165856343e-06,
+      "loss": 0.5796,
+      "step": 6746
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.2778199669618373,
+      "learning_rate": 2.048541201213981e-06,
+      "loss": 0.6517,
+      "step": 6747
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1601593601154647,
+      "learning_rate": 2.0471657287342813e-06,
+      "loss": 0.598,
+      "step": 6748
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.7950345405844166,
+      "learning_rate": 2.0457905993063306e-06,
+      "loss": 0.6188,
+      "step": 6749
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3229400348434717,
+      "learning_rate": 2.044415813089887e-06,
+      "loss": 0.5894,
+      "step": 6750
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5323121960930086,
+      "learning_rate": 2.0430413702446707e-06,
+      "loss": 0.6715,
+      "step": 6751
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1152568643865597,
+      "learning_rate": 2.0416672709303597e-06,
+      "loss": 0.591,
+      "step": 6752
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.241190616242863,
+      "learning_rate": 2.0402935153065976e-06,
+      "loss": 0.5456,
+      "step": 6753
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8952230145025495,
+      "learning_rate": 2.0389201035329754e-06,
+      "loss": 0.6587,
+      "step": 6754
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.913524657839511,
+      "learning_rate": 2.0375470357690564e-06,
+      "loss": 0.6103,
+      "step": 6755
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.357411676416585,
+      "learning_rate": 2.036174312174359e-06,
+      "loss": 0.6013,
+      "step": 6756
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.164348116661981,
+      "learning_rate": 2.034801932908364e-06,
+      "loss": 0.641,
+      "step": 6757
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.285714519634651,
+      "learning_rate": 2.033429898130509e-06,
+      "loss": 0.5707,
+      "step": 6758
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.314588495421655,
+      "learning_rate": 2.032058208000191e-06,
+      "loss": 0.6353,
+      "step": 6759
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.1963968095024904,
+      "learning_rate": 2.030686862676771e-06,
+      "loss": 0.6579,
+      "step": 6760
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.234190092875219,
+      "learning_rate": 2.0293158623195702e-06,
+      "loss": 0.6601,
+      "step": 6761
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.350692178674305,
+      "learning_rate": 2.0279452070878647e-06,
+      "loss": 0.6188,
+      "step": 6762
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.564102840538574,
+      "learning_rate": 2.026574897140892e-06,
+      "loss": 0.6149,
+      "step": 6763
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.174466257356336,
+      "learning_rate": 2.0252049326378524e-06,
+      "loss": 0.605,
+      "step": 6764
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.149315755348158,
+      "learning_rate": 2.0238353137379047e-06,
+      "loss": 0.6217,
+      "step": 6765
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.22159691006164,
+      "learning_rate": 2.02246604060017e-06,
+      "loss": 0.4916,
+      "step": 6766
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0623059295189003,
+      "learning_rate": 2.0210971133837208e-06,
+      "loss": 0.6159,
+      "step": 6767
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.244551913598113,
+      "learning_rate": 2.0197285322475975e-06,
+      "loss": 0.5866,
+      "step": 6768
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.429308309812128,
+      "learning_rate": 2.0183602973507977e-06,
+      "loss": 0.6469,
+      "step": 6769
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2426310766438093,
+      "learning_rate": 2.016992408852282e-06,
+      "loss": 0.6122,
+      "step": 6770
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.147899821152736,
+      "learning_rate": 2.0156248669109645e-06,
+      "loss": 0.6084,
+      "step": 6771
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5207952773245057,
+      "learning_rate": 2.014257671685722e-06,
+      "loss": 0.6181,
+      "step": 6772
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.290492088782644,
+      "learning_rate": 2.012890823335392e-06,
+      "loss": 0.6861,
+      "step": 6773
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9788946060137566,
+      "learning_rate": 2.011524322018773e-06,
+      "loss": 0.5617,
+      "step": 6774
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.02202729427753,
+      "learning_rate": 2.01015816789462e-06,
+      "loss": 0.5491,
+      "step": 6775
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.644055938189567,
+      "learning_rate": 2.0087923611216452e-06,
+      "loss": 0.6326,
+      "step": 6776
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6229423897634088,
+      "learning_rate": 2.0074269018585286e-06,
+      "loss": 0.6593,
+      "step": 6777
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.657797931574321,
+      "learning_rate": 2.006061790263903e-06,
+      "loss": 0.6879,
+      "step": 6778
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.731539046733333,
+      "learning_rate": 2.004697026496366e-06,
+      "loss": 0.6025,
+      "step": 6779
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3300238922256162,
+      "learning_rate": 2.00333261071447e-06,
+      "loss": 0.5737,
+      "step": 6780
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5105007437944,
+      "learning_rate": 2.001968543076727e-06,
+      "loss": 0.5984,
+      "step": 6781
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.9355723240439238,
+      "learning_rate": 2.0006048237416127e-06,
+      "loss": 0.6853,
+      "step": 6782
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.7838842692244414,
+      "learning_rate": 1.9992414528675607e-06,
+      "loss": 0.5972,
+      "step": 6783
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.409380507657977,
+      "learning_rate": 1.997878430612963e-06,
+      "loss": 0.6305,
+      "step": 6784
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.610091326027608,
+      "learning_rate": 1.9965157571361688e-06,
+      "loss": 0.6465,
+      "step": 6785
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0465580485685355,
+      "learning_rate": 1.9951534325954913e-06,
+      "loss": 0.5599,
+      "step": 6786
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.7696890221497825,
+      "learning_rate": 1.9937914571492024e-06,
+      "loss": 0.5447,
+      "step": 6787
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.854812905102868,
+      "learning_rate": 1.9924298309555355e-06,
+      "loss": 0.5373,
+      "step": 6788
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.144295935862878,
+      "learning_rate": 1.991068554172673e-06,
+      "loss": 0.5129,
+      "step": 6789
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.235492207734396,
+      "learning_rate": 1.9897076269587686e-06,
+      "loss": 0.5211,
+      "step": 6790
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0026420194820682,
+      "learning_rate": 1.98834704947193e-06,
+      "loss": 0.5368,
+      "step": 6791
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.2741376437674825,
+      "learning_rate": 1.9869868218702266e-06,
+      "loss": 0.6746,
+      "step": 6792
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4315638104365025,
+      "learning_rate": 1.985626944311685e-06,
+      "loss": 0.4932,
+      "step": 6793
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.005978231078567,
+      "learning_rate": 1.984267416954289e-06,
+      "loss": 0.694,
+      "step": 6794
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9172999980041083,
+      "learning_rate": 1.9829082399559872e-06,
+      "loss": 0.5798,
+      "step": 6795
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2051339131239667,
+      "learning_rate": 1.9815494134746866e-06,
+      "loss": 0.6441,
+      "step": 6796
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4405798502942733,
+      "learning_rate": 1.98019093766825e-06,
+      "loss": 0.6649,
+      "step": 6797
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.505836213580318,
+      "learning_rate": 1.9788328126944984e-06,
+      "loss": 0.5017,
+      "step": 6798
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5383152998963965,
+      "learning_rate": 1.9774750387112176e-06,
+      "loss": 0.6945,
+      "step": 6799
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.295043766361787,
+      "learning_rate": 1.976117615876149e-06,
+      "loss": 0.5951,
+      "step": 6800
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.0818322840049546,
+      "learning_rate": 1.974760544346999e-06,
+      "loss": 0.6048,
+      "step": 6801
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1718689507776063,
+      "learning_rate": 1.9734038242814203e-06,
+      "loss": 0.5392,
+      "step": 6802
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.5743079582613,
+      "learning_rate": 1.9720474558370356e-06,
+      "loss": 0.6438,
+      "step": 6803
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4649750346849424,
+      "learning_rate": 1.970691439171425e-06,
+      "loss": 0.5826,
+      "step": 6804
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.251453377955877,
+      "learning_rate": 1.9693357744421282e-06,
+      "loss": 0.6129,
+      "step": 6805
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7067846238307616,
+      "learning_rate": 1.96798046180664e-06,
+      "loss": 0.5928,
+      "step": 6806
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9657749101823887,
+      "learning_rate": 1.966625501422415e-06,
+      "loss": 0.6502,
+      "step": 6807
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.885763329620718,
+      "learning_rate": 1.965270893446871e-06,
+      "loss": 0.6607,
+      "step": 6808
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.516314689530151,
+      "learning_rate": 1.963916638037384e-06,
+      "loss": 0.6097,
+      "step": 6809
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.819130855149682,
+      "learning_rate": 1.9625627353512854e-06,
+      "loss": 0.647,
+      "step": 6810
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.1510683550167125,
+      "learning_rate": 1.9612091855458663e-06,
+      "loss": 0.5885,
+      "step": 6811
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.8987176112644457,
+      "learning_rate": 1.9598559887783797e-06,
+      "loss": 0.5884,
+      "step": 6812
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9904600312949907,
+      "learning_rate": 1.958503145206036e-06,
+      "loss": 0.5806,
+      "step": 6813
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5080351376008374,
+      "learning_rate": 1.9571506549860065e-06,
+      "loss": 0.6038,
+      "step": 6814
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.939203442577235,
+      "learning_rate": 1.955798518275418e-06,
+      "loss": 0.5762,
+      "step": 6815
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.228977059412211,
+      "learning_rate": 1.954446735231356e-06,
+      "loss": 0.5609,
+      "step": 6816
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.63612086695577,
+      "learning_rate": 1.953095306010869e-06,
+      "loss": 0.5973,
+      "step": 6817
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.463307921855391,
+      "learning_rate": 1.9517442307709626e-06,
+      "loss": 0.5971,
+      "step": 6818
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9678254750031574,
+      "learning_rate": 1.9503935096686004e-06,
+      "loss": 0.5483,
+      "step": 6819
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4727076802540133,
+      "learning_rate": 1.9490431428607027e-06,
+      "loss": 0.626,
+      "step": 6820
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0978475149269795,
+      "learning_rate": 1.947693130504153e-06,
+      "loss": 0.5705,
+      "step": 6821
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.10936758831189,
+      "learning_rate": 1.9463434727557927e-06,
+      "loss": 0.6207,
+      "step": 6822
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4647814076539816,
+      "learning_rate": 1.9449941697724233e-06,
+      "loss": 0.5815,
+      "step": 6823
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2186953732210632,
+      "learning_rate": 1.943645221710797e-06,
+      "loss": 0.5623,
+      "step": 6824
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3686123006295414,
+      "learning_rate": 1.942296628727634e-06,
+      "loss": 0.6202,
+      "step": 6825
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4056025600068973,
+      "learning_rate": 1.9409483909796096e-06,
+      "loss": 0.5397,
+      "step": 6826
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.843753987586021,
+      "learning_rate": 1.93960050862336e-06,
+      "loss": 0.6794,
+      "step": 6827
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.0253380845711915,
+      "learning_rate": 1.9382529818154765e-06,
+      "loss": 0.6079,
+      "step": 6828
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.226590014372156,
+      "learning_rate": 1.9369058107125094e-06,
+      "loss": 0.5139,
+      "step": 6829
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6079029240171265,
+      "learning_rate": 1.935558995470971e-06,
+      "loss": 0.5842,
+      "step": 6830
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5961478591051663,
+      "learning_rate": 1.9342125362473313e-06,
+      "loss": 0.7159,
+      "step": 6831
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.508735121085522,
+      "learning_rate": 1.9328664331980175e-06,
+      "loss": 0.6256,
+      "step": 6832
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.0787485825721332,
+      "learning_rate": 1.931520686479413e-06,
+      "loss": 0.7014,
+      "step": 6833
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2749815114493286,
+      "learning_rate": 1.9301752962478646e-06,
+      "loss": 0.6009,
+      "step": 6834
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9794025506603408,
+      "learning_rate": 1.9288302626596772e-06,
+      "loss": 0.6428,
+      "step": 6835
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.039751673052599,
+      "learning_rate": 1.9274855858711157e-06,
+      "loss": 0.6417,
+      "step": 6836
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.8877557291619853,
+      "learning_rate": 1.9261412660383927e-06,
+      "loss": 0.6086,
+      "step": 6837
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.692848677613801,
+      "learning_rate": 1.924797303317692e-06,
+      "loss": 0.6517,
+      "step": 6838
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4817682883772334,
+      "learning_rate": 1.9234536978651514e-06,
+      "loss": 0.6697,
+      "step": 6839
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.126381340774734,
+      "learning_rate": 1.922110449836869e-06,
+      "loss": 0.6342,
+      "step": 6840
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.565024160486094,
+      "learning_rate": 1.920767559388896e-06,
+      "loss": 0.629,
+      "step": 6841
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.208434714049697,
+      "learning_rate": 1.919425026677246e-06,
+      "loss": 0.6768,
+      "step": 6842
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7771674907930843,
+      "learning_rate": 1.9180828518578907e-06,
+      "loss": 0.6414,
+      "step": 6843
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.972741578017247,
+      "learning_rate": 1.9167410350867634e-06,
+      "loss": 0.5638,
+      "step": 6844
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.408080059628293,
+      "learning_rate": 1.9153995765197492e-06,
+      "loss": 0.5538,
+      "step": 6845
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.463378183778841,
+      "learning_rate": 1.9140584763126942e-06,
+      "loss": 0.6613,
+      "step": 6846
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.6438629469274177,
+      "learning_rate": 1.912717734621404e-06,
+      "loss": 0.6311,
+      "step": 6847
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4515690885324446,
+      "learning_rate": 1.911377351601644e-06,
+      "loss": 0.587,
+      "step": 6848
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9571991246007077,
+      "learning_rate": 1.910037327409136e-06,
+      "loss": 0.5853,
+      "step": 6849
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0579456833316514,
+      "learning_rate": 1.9086976621995595e-06,
+      "loss": 0.5346,
+      "step": 6850
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.360162048127551,
+      "learning_rate": 1.9073583561285507e-06,
+      "loss": 0.5493,
+      "step": 6851
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9782767148226186,
+      "learning_rate": 1.9060194093517082e-06,
+      "loss": 0.6485,
+      "step": 6852
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.172949041709697,
+      "learning_rate": 1.9046808220245888e-06,
+      "loss": 0.6456,
+      "step": 6853
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4126025762580485,
+      "learning_rate": 1.903342594302704e-06,
+      "loss": 0.6233,
+      "step": 6854
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.111218794053248,
+      "learning_rate": 1.9020047263415226e-06,
+      "loss": 0.564,
+      "step": 6855
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.112267233952676,
+      "learning_rate": 1.9006672182964776e-06,
+      "loss": 0.5294,
+      "step": 6856
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1191102063033305,
+      "learning_rate": 1.899330070322955e-06,
+      "loss": 0.5833,
+      "step": 6857
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.8084622563861643,
+      "learning_rate": 1.8979932825763058e-06,
+      "loss": 0.6078,
+      "step": 6858
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.731001054966648,
+      "learning_rate": 1.8966568552118265e-06,
+      "loss": 0.6116,
+      "step": 6859
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4118252942652885,
+      "learning_rate": 1.895320788384783e-06,
+      "loss": 0.6477,
+      "step": 6860
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4880736901897076,
+      "learning_rate": 1.8939850822503953e-06,
+      "loss": 0.6904,
+      "step": 6861
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2508815431544282,
+      "learning_rate": 1.8926497369638435e-06,
+      "loss": 0.5529,
+      "step": 6862
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.475826447334696,
+      "learning_rate": 1.8913147526802633e-06,
+      "loss": 0.5821,
+      "step": 6863
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.106833308302653,
+      "learning_rate": 1.8899801295547476e-06,
+      "loss": 0.6153,
+      "step": 6864
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.202917393863976,
+      "learning_rate": 1.8886458677423497e-06,
+      "loss": 0.5526,
+      "step": 6865
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4715070549141216,
+      "learning_rate": 1.8873119673980828e-06,
+      "loss": 0.6657,
+      "step": 6866
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.967796086555414,
+      "learning_rate": 1.8859784286769133e-06,
+      "loss": 0.5738,
+      "step": 6867
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.265639351961703,
+      "learning_rate": 1.8846452517337665e-06,
+      "loss": 0.676,
+      "step": 6868
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2231883474524,
+      "learning_rate": 1.8833124367235294e-06,
+      "loss": 0.6419,
+      "step": 6869
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.0688124326659016,
+      "learning_rate": 1.8819799838010434e-06,
+      "loss": 0.617,
+      "step": 6870
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5790235530152508,
+      "learning_rate": 1.8806478931211137e-06,
+      "loss": 0.6266,
+      "step": 6871
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.248003195138114,
+      "learning_rate": 1.8793161648384905e-06,
+      "loss": 0.6454,
+      "step": 6872
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4974510096976377,
+      "learning_rate": 1.8779847991078943e-06,
+      "loss": 0.6192,
+      "step": 6873
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7581329611825183,
+      "learning_rate": 1.8766537960839997e-06,
+      "loss": 0.6176,
+      "step": 6874
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.5694350624696978,
+      "learning_rate": 1.8753231559214402e-06,
+      "loss": 0.4594,
+      "step": 6875
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.608950453790209,
+      "learning_rate": 1.8739928787748035e-06,
+      "loss": 0.5928,
+      "step": 6876
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.277145432538584,
+      "learning_rate": 1.872662964798636e-06,
+      "loss": 0.6286,
+      "step": 6877
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.794994081731649,
+      "learning_rate": 1.8713334141474454e-06,
+      "loss": 0.6851,
+      "step": 6878
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.249886037254742,
+      "learning_rate": 1.8700042269756964e-06,
+      "loss": 0.633,
+      "step": 6879
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2734533778197794,
+      "learning_rate": 1.8686754034378085e-06,
+      "loss": 0.6386,
+      "step": 6880
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.128369477606803,
+      "learning_rate": 1.867346943688158e-06,
+      "loss": 0.6784,
+      "step": 6881
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.304439961783062,
+      "learning_rate": 1.8660188478810848e-06,
+      "loss": 0.5845,
+      "step": 6882
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.995582630490625,
+      "learning_rate": 1.8646911161708824e-06,
+      "loss": 0.6216,
+      "step": 6883
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3675530017194064,
+      "learning_rate": 1.8633637487118046e-06,
+      "loss": 0.5987,
+      "step": 6884
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9813523750802173,
+      "learning_rate": 1.862036745658059e-06,
+      "loss": 0.6377,
+      "step": 6885
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6903692135653428,
+      "learning_rate": 1.8607101071638117e-06,
+      "loss": 0.6237,
+      "step": 6886
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.329812249626914,
+      "learning_rate": 1.8593838333831893e-06,
+      "loss": 0.6296,
+      "step": 6887
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.501685681461769,
+      "learning_rate": 1.8580579244702762e-06,
+      "loss": 0.6258,
+      "step": 6888
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.055180231542555,
+      "learning_rate": 1.8567323805791116e-06,
+      "loss": 0.539,
+      "step": 6889
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6544723103678916,
+      "learning_rate": 1.8554072018636903e-06,
+      "loss": 0.5934,
+      "step": 6890
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4845123889284735,
+      "learning_rate": 1.8540823884779708e-06,
+      "loss": 0.5886,
+      "step": 6891
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.1419531063459485,
+      "learning_rate": 1.8527579405758672e-06,
+      "loss": 0.5611,
+      "step": 6892
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4840964920780957,
+      "learning_rate": 1.851433858311248e-06,
+      "loss": 0.5448,
+      "step": 6893
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7386535448522418,
+      "learning_rate": 1.8501101418379398e-06,
+      "loss": 0.6005,
+      "step": 6894
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.879316470041651,
+      "learning_rate": 1.8487867913097301e-06,
+      "loss": 0.5817,
+      "step": 6895
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0545439469322904,
+      "learning_rate": 1.8474638068803612e-06,
+      "loss": 0.5266,
+      "step": 6896
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.215830096903211,
+      "learning_rate": 1.8461411887035368e-06,
+      "loss": 0.5681,
+      "step": 6897
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3679216916484243,
+      "learning_rate": 1.8448189369329117e-06,
+      "loss": 0.6838,
+      "step": 6898
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.423382212663329,
+      "learning_rate": 1.8434970517221e-06,
+      "loss": 0.6508,
+      "step": 6899
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0003477907669218,
+      "learning_rate": 1.8421755332246765e-06,
+      "loss": 0.5632,
+      "step": 6900
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3039328325659962,
+      "learning_rate": 1.840854381594173e-06,
+      "loss": 0.5946,
+      "step": 6901
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.9075536999442937,
+      "learning_rate": 1.8395335969840749e-06,
+      "loss": 0.6977,
+      "step": 6902
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.653021560915067,
+      "learning_rate": 1.8382131795478265e-06,
+      "loss": 0.6104,
+      "step": 6903
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.459736159668085,
+      "learning_rate": 1.8368931294388303e-06,
+      "loss": 0.6558,
+      "step": 6904
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.145122811581977,
+      "learning_rate": 1.8355734468104476e-06,
+      "loss": 0.6376,
+      "step": 6905
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3194721091248622,
+      "learning_rate": 1.8342541318159967e-06,
+      "loss": 0.6556,
+      "step": 6906
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.380316773328887,
+      "learning_rate": 1.8329351846087467e-06,
+      "loss": 0.4877,
+      "step": 6907
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.059952338938169,
+      "learning_rate": 1.8316166053419321e-06,
+      "loss": 0.648,
+      "step": 6908
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3302793061982854,
+      "learning_rate": 1.8302983941687414e-06,
+      "loss": 0.665,
+      "step": 6909
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.196962381204859,
+      "learning_rate": 1.828980551242322e-06,
+      "loss": 0.6074,
+      "step": 6910
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3653436980498133,
+      "learning_rate": 1.827663076715776e-06,
+      "loss": 0.5736,
+      "step": 6911
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3272350514303515,
+      "learning_rate": 1.8263459707421617e-06,
+      "loss": 0.5789,
+      "step": 6912
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.843158913797961,
+      "learning_rate": 1.8250292334744979e-06,
+      "loss": 0.6096,
+      "step": 6913
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3445312077529357,
+      "learning_rate": 1.8237128650657621e-06,
+      "loss": 0.6748,
+      "step": 6914
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7392105169159904,
+      "learning_rate": 1.8223968656688834e-06,
+      "loss": 0.5316,
+      "step": 6915
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.381012515570709,
+      "learning_rate": 1.8210812354367501e-06,
+      "loss": 0.6245,
+      "step": 6916
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0098706319158968,
+      "learning_rate": 1.8197659745222095e-06,
+      "loss": 0.5341,
+      "step": 6917
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3320907079732485,
+      "learning_rate": 1.818451083078065e-06,
+      "loss": 0.6104,
+      "step": 6918
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.243267962515471,
+      "learning_rate": 1.817136561257078e-06,
+      "loss": 0.5681,
+      "step": 6919
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.244715320509986,
+      "learning_rate": 1.8158224092119648e-06,
+      "loss": 0.5684,
+      "step": 6920
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.9394817301430924,
+      "learning_rate": 1.8145086270953977e-06,
+      "loss": 0.576,
+      "step": 6921
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.421731912227469,
+      "learning_rate": 1.8131952150600101e-06,
+      "loss": 0.573,
+      "step": 6922
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7932028278457115,
+      "learning_rate": 1.8118821732583918e-06,
+      "loss": 0.5806,
+      "step": 6923
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.9079436194870394,
+      "learning_rate": 1.8105695018430873e-06,
+      "loss": 0.6072,
+      "step": 6924
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.583809450277425,
+      "learning_rate": 1.8092572009665965e-06,
+      "loss": 0.628,
+      "step": 6925
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1674044267667516,
+      "learning_rate": 1.80794527078138e-06,
+      "loss": 0.5968,
+      "step": 6926
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6185013262869354,
+      "learning_rate": 1.8066337114398568e-06,
+      "loss": 0.5478,
+      "step": 6927
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.158397579484625,
+      "learning_rate": 1.8053225230943982e-06,
+      "loss": 0.6001,
+      "step": 6928
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.264667477188976,
+      "learning_rate": 1.8040117058973317e-06,
+      "loss": 0.6074,
+      "step": 6929
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7602307379116064,
+      "learning_rate": 1.802701260000947e-06,
+      "loss": 0.6428,
+      "step": 6930
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.986327201725227,
+      "learning_rate": 1.8013911855574874e-06,
+      "loss": 0.5278,
+      "step": 6931
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0385916995999906,
+      "learning_rate": 1.8000814827191548e-06,
+      "loss": 0.6268,
+      "step": 6932
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1047064832640023,
+      "learning_rate": 1.7987721516381056e-06,
+      "loss": 0.6004,
+      "step": 6933
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4994668971667364,
+      "learning_rate": 1.7974631924664533e-06,
+      "loss": 0.6405,
+      "step": 6934
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.388551949225777,
+      "learning_rate": 1.7961546053562684e-06,
+      "loss": 0.647,
+      "step": 6935
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.348701907039752,
+      "learning_rate": 1.7948463904595826e-06,
+      "loss": 0.5663,
+      "step": 6936
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3266351864310137,
+      "learning_rate": 1.793538547928378e-06,
+      "loss": 0.604,
+      "step": 6937
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.150907502807278,
+      "learning_rate": 1.7922310779145941e-06,
+      "loss": 0.5916,
+      "step": 6938
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.2575410386870502,
+      "learning_rate": 1.7909239805701307e-06,
+      "loss": 0.6358,
+      "step": 6939
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0203290093238349,
+      "learning_rate": 1.7896172560468427e-06,
+      "loss": 0.5392,
+      "step": 6940
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.455935523599585,
+      "learning_rate": 1.7883109044965452e-06,
+      "loss": 0.6663,
+      "step": 6941
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.292636831216748,
+      "learning_rate": 1.7870049260709992e-06,
+      "loss": 0.6031,
+      "step": 6942
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.740746619986824,
+      "learning_rate": 1.785699320921933e-06,
+      "loss": 0.5899,
+      "step": 6943
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.438088048263127,
+      "learning_rate": 1.784394089201028e-06,
+      "loss": 0.609,
+      "step": 6944
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.2919583772455385,
+      "learning_rate": 1.7830892310599245e-06,
+      "loss": 0.5254,
+      "step": 6945
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.0375360511034284,
+      "learning_rate": 1.7817847466502146e-06,
+      "loss": 0.5886,
+      "step": 6946
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.278371165535222,
+      "learning_rate": 1.780480636123449e-06,
+      "loss": 0.6285,
+      "step": 6947
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6474986848344844,
+      "learning_rate": 1.7791768996311355e-06,
+      "loss": 0.6188,
+      "step": 6948
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.053270599492303,
+      "learning_rate": 1.7778735373247414e-06,
+      "loss": 0.6231,
+      "step": 6949
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6023055630386622,
+      "learning_rate": 1.7765705493556857e-06,
+      "loss": 0.5683,
+      "step": 6950
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.180205593322918,
+      "learning_rate": 1.7752679358753433e-06,
+      "loss": 0.6944,
+      "step": 6951
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.3913259162961666,
+      "learning_rate": 1.7739656970350505e-06,
+      "loss": 0.6438,
+      "step": 6952
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.504035889002391,
+      "learning_rate": 1.7726638329860978e-06,
+      "loss": 0.6856,
+      "step": 6953
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.764765266914022,
+      "learning_rate": 1.7713623438797335e-06,
+      "loss": 0.5888,
+      "step": 6954
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.2754835937489934,
+      "learning_rate": 1.7700612298671587e-06,
+      "loss": 0.6877,
+      "step": 6955
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.449529265194134,
+      "learning_rate": 1.7687604910995321e-06,
+      "loss": 0.6649,
+      "step": 6956
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4212481065052907,
+      "learning_rate": 1.7674601277279707e-06,
+      "loss": 0.618,
+      "step": 6957
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.736474279576478,
+      "learning_rate": 1.7661601399035494e-06,
+      "loss": 0.5901,
+      "step": 6958
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.691986553144693,
+      "learning_rate": 1.7648605277772945e-06,
+      "loss": 0.6313,
+      "step": 6959
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6214863033679943,
+      "learning_rate": 1.7635612915001903e-06,
+      "loss": 0.7215,
+      "step": 6960
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.9600672604153533,
+      "learning_rate": 1.7622624312231795e-06,
+      "loss": 0.6593,
+      "step": 6961
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.8520335479613363,
+      "learning_rate": 1.7609639470971618e-06,
+      "loss": 0.5712,
+      "step": 6962
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2886996974373814,
+      "learning_rate": 1.7596658392729897e-06,
+      "loss": 0.5834,
+      "step": 6963
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.8110720875652033,
+      "learning_rate": 1.7583681079014713e-06,
+      "loss": 0.6194,
+      "step": 6964
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.229998449327696,
+      "learning_rate": 1.7570707531333763e-06,
+      "loss": 0.5953,
+      "step": 6965
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6339048670728284,
+      "learning_rate": 1.7557737751194264e-06,
+      "loss": 0.6276,
+      "step": 6966
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.126572883969871,
+      "learning_rate": 1.7544771740103034e-06,
+      "loss": 0.6219,
+      "step": 6967
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.532068840298241,
+      "learning_rate": 1.7531809499566399e-06,
+      "loss": 0.6469,
+      "step": 6968
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.1935164634899,
+      "learning_rate": 1.7518851031090267e-06,
+      "loss": 0.612,
+      "step": 6969
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.541841612435397,
+      "learning_rate": 1.7505896336180128e-06,
+      "loss": 0.5458,
+      "step": 6970
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3737229858880187,
+      "learning_rate": 1.7492945416341034e-06,
+      "loss": 0.6253,
+      "step": 6971
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.2573282697164827,
+      "learning_rate": 1.7479998273077581e-06,
+      "loss": 0.633,
+      "step": 6972
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7028931132859264,
+      "learning_rate": 1.74670549078939e-06,
+      "loss": 0.6755,
+      "step": 6973
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.702801368704138,
+      "learning_rate": 1.7454115322293735e-06,
+      "loss": 0.5617,
+      "step": 6974
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3197807420820613,
+      "learning_rate": 1.7441179517780376e-06,
+      "loss": 0.6572,
+      "step": 6975
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.674381378155758,
+      "learning_rate": 1.7428247495856699e-06,
+      "loss": 0.6158,
+      "step": 6976
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.8755462439432047,
+      "learning_rate": 1.7415319258025032e-06,
+      "loss": 0.697,
+      "step": 6977
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.0812031503014983,
+      "learning_rate": 1.7402394805787388e-06,
+      "loss": 0.6277,
+      "step": 6978
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 11.5464107755622,
+      "learning_rate": 1.7389474140645279e-06,
+      "loss": 0.5804,
+      "step": 6979
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.338776528858335,
+      "learning_rate": 1.7376557264099813e-06,
+      "loss": 0.5742,
+      "step": 6980
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5226545710029087,
+      "learning_rate": 1.7363644177651623e-06,
+      "loss": 0.5308,
+      "step": 6981
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.538706474138878,
+      "learning_rate": 1.7350734882800891e-06,
+      "loss": 0.6669,
+      "step": 6982
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.553740639868528,
+      "learning_rate": 1.7337829381047405e-06,
+      "loss": 0.6483,
+      "step": 6983
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4491443310423726,
+      "learning_rate": 1.7324927673890495e-06,
+      "loss": 0.5876,
+      "step": 6984
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3165156688725244,
+      "learning_rate": 1.7312029762829042e-06,
+      "loss": 0.5532,
+      "step": 6985
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3703050030038884,
+      "learning_rate": 1.729913564936146e-06,
+      "loss": 0.646,
+      "step": 6986
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.457421486560212,
+      "learning_rate": 1.728624533498577e-06,
+      "loss": 0.6851,
+      "step": 6987
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.7152345384682492,
+      "learning_rate": 1.7273358821199527e-06,
+      "loss": 0.6291,
+      "step": 6988
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.928945773179569,
+      "learning_rate": 1.7260476109499885e-06,
+      "loss": 0.6238,
+      "step": 6989
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0786975252518975,
+      "learning_rate": 1.7247597201383459e-06,
+      "loss": 0.6212,
+      "step": 6990
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.2322246793230462,
+      "learning_rate": 1.7234722098346512e-06,
+      "loss": 0.5755,
+      "step": 6991
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.385974734862681,
+      "learning_rate": 1.7221850801884838e-06,
+      "loss": 0.6729,
+      "step": 6992
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2783917987848175,
+      "learning_rate": 1.7208983313493804e-06,
+      "loss": 0.5833,
+      "step": 6993
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.1121321239656368,
+      "learning_rate": 1.7196119634668296e-06,
+      "loss": 0.6074,
+      "step": 6994
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4956465433518087,
+      "learning_rate": 1.7183259766902765e-06,
+      "loss": 0.5979,
+      "step": 6995
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2925119083577363,
+      "learning_rate": 1.7170403711691252e-06,
+      "loss": 0.5927,
+      "step": 6996
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.1901529302786145,
+      "learning_rate": 1.7157551470527356e-06,
+      "loss": 0.5835,
+      "step": 6997
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1103141114003794,
+      "learning_rate": 1.7144703044904186e-06,
+      "loss": 0.647,
+      "step": 6998
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2812632686596874,
+      "learning_rate": 1.7131858436314431e-06,
+      "loss": 0.5589,
+      "step": 6999
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.893898766295147,
+      "learning_rate": 1.7119017646250346e-06,
+      "loss": 0.5332,
+      "step": 7000
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.333401918284876,
+      "learning_rate": 1.7106180676203743e-06,
+      "loss": 0.644,
+      "step": 7001
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2776944296611905,
+      "learning_rate": 1.7093347527666e-06,
+      "loss": 0.6315,
+      "step": 7002
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4878470361517535,
+      "learning_rate": 1.708051820212801e-06,
+      "loss": 0.5557,
+      "step": 7003
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9538072931011493,
+      "learning_rate": 1.7067692701080247e-06,
+      "loss": 0.556,
+      "step": 7004
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.936334663090163,
+      "learning_rate": 1.7054871026012748e-06,
+      "loss": 0.6332,
+      "step": 7005
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.569749185859605,
+      "learning_rate": 1.7042053178415114e-06,
+      "loss": 0.597,
+      "step": 7006
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2112215878359844,
+      "learning_rate": 1.7029239159776468e-06,
+      "loss": 0.6083,
+      "step": 7007
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.9065918760961362,
+      "learning_rate": 1.7016428971585491e-06,
+      "loss": 0.6281,
+      "step": 7008
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0826011656455226,
+      "learning_rate": 1.700362261533045e-06,
+      "loss": 0.5959,
+      "step": 7009
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3233212826323664,
+      "learning_rate": 1.699082009249915e-06,
+      "loss": 0.6065,
+      "step": 7010
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9508372060278683,
+      "learning_rate": 1.6978021404578986e-06,
+      "loss": 0.5127,
+      "step": 7011
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3456337360777826,
+      "learning_rate": 1.6965226553056807e-06,
+      "loss": 0.6536,
+      "step": 7012
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.740110007869177,
+      "learning_rate": 1.6952435539419114e-06,
+      "loss": 0.6519,
+      "step": 7013
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.285157779061375,
+      "learning_rate": 1.6939648365151929e-06,
+      "loss": 0.6668,
+      "step": 7014
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.286468423276322,
+      "learning_rate": 1.692686503174084e-06,
+      "loss": 0.5346,
+      "step": 7015
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.544160329777678,
+      "learning_rate": 1.6914085540670972e-06,
+      "loss": 0.6256,
+      "step": 7016
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6105527595546683,
+      "learning_rate": 1.6901309893426987e-06,
+      "loss": 0.5112,
+      "step": 7017
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.824511640017726,
+      "learning_rate": 1.688853809149314e-06,
+      "loss": 0.6006,
+      "step": 7018
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.421375917201986,
+      "learning_rate": 1.6875770136353237e-06,
+      "loss": 0.5986,
+      "step": 7019
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.775614276256507,
+      "learning_rate": 1.686300602949061e-06,
+      "loss": 0.641,
+      "step": 7020
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8464697028833026,
+      "learning_rate": 1.6850245772388136e-06,
+      "loss": 0.6148,
+      "step": 7021
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6951000358839816,
+      "learning_rate": 1.6837489366528275e-06,
+      "loss": 0.5996,
+      "step": 7022
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2776028181000036,
+      "learning_rate": 1.6824736813393044e-06,
+      "loss": 0.5767,
+      "step": 7023
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.049024430013574,
+      "learning_rate": 1.6811988114464024e-06,
+      "loss": 0.5973,
+      "step": 7024
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.259489650346381,
+      "learning_rate": 1.6799243271222248e-06,
+      "loss": 0.6279,
+      "step": 7025
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.7237784454005047,
+      "learning_rate": 1.678650228514842e-06,
+      "loss": 0.6341,
+      "step": 7026
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.492400316742001,
+      "learning_rate": 1.677376515772275e-06,
+      "loss": 0.5853,
+      "step": 7027
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.370329142260077,
+      "learning_rate": 1.6761031890425007e-06,
+      "loss": 0.6075,
+      "step": 7028
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1958100915780823,
+      "learning_rate": 1.6748302484734496e-06,
+      "loss": 0.5913,
+      "step": 7029
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.685029763578393,
+      "learning_rate": 1.6735576942130066e-06,
+      "loss": 0.5991,
+      "step": 7030
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0884424565510606,
+      "learning_rate": 1.672285526409015e-06,
+      "loss": 0.5875,
+      "step": 7031
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.244543977179305,
+      "learning_rate": 1.6710137452092728e-06,
+      "loss": 0.647,
+      "step": 7032
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8555347187741154,
+      "learning_rate": 1.6697423507615307e-06,
+      "loss": 0.6764,
+      "step": 7033
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.943847721173569,
+      "learning_rate": 1.6684713432134935e-06,
+      "loss": 0.637,
+      "step": 7034
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.9179340893616694,
+      "learning_rate": 1.6672007227128256e-06,
+      "loss": 0.6128,
+      "step": 7035
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3128758180540188,
+      "learning_rate": 1.6659304894071437e-06,
+      "loss": 0.5963,
+      "step": 7036
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.279530033918136,
+      "learning_rate": 1.6646606434440216e-06,
+      "loss": 0.5943,
+      "step": 7037
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4218365961921924,
+      "learning_rate": 1.6633911849709838e-06,
+      "loss": 0.5705,
+      "step": 7038
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4252321627282565,
+      "learning_rate": 1.6621221141355114e-06,
+      "loss": 0.6823,
+      "step": 7039
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.815209527008716,
+      "learning_rate": 1.6608534310850432e-06,
+      "loss": 0.6023,
+      "step": 7040
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.2696793670858457,
+      "learning_rate": 1.6595851359669723e-06,
+      "loss": 0.6344,
+      "step": 7041
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.048168501203653,
+      "learning_rate": 1.6583172289286447e-06,
+      "loss": 0.6739,
+      "step": 7042
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.345623717215056,
+      "learning_rate": 1.6570497101173595e-06,
+      "loss": 0.5041,
+      "step": 7043
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.656636229228406,
+      "learning_rate": 1.6557825796803755e-06,
+      "loss": 0.5783,
+      "step": 7044
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.047758363020559,
+      "learning_rate": 1.6545158377649063e-06,
+      "loss": 0.6101,
+      "step": 7045
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4301367175325446,
+      "learning_rate": 1.6532494845181157e-06,
+      "loss": 0.6579,
+      "step": 7046
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4798435707976627,
+      "learning_rate": 1.6519835200871243e-06,
+      "loss": 0.6268,
+      "step": 7047
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2104318045494717,
+      "learning_rate": 1.6507179446190091e-06,
+      "loss": 0.5432,
+      "step": 7048
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.720179561419772,
+      "learning_rate": 1.649452758260801e-06,
+      "loss": 0.5763,
+      "step": 7049
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8768611462675713,
+      "learning_rate": 1.648187961159488e-06,
+      "loss": 0.5999,
+      "step": 7050
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.9574492500850096,
+      "learning_rate": 1.6469235534620087e-06,
+      "loss": 0.6756,
+      "step": 7051
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4478650282919654,
+      "learning_rate": 1.6456595353152566e-06,
+      "loss": 0.6178,
+      "step": 7052
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.4521440344822647,
+      "learning_rate": 1.644395906866083e-06,
+      "loss": 0.633,
+      "step": 7053
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3845856771271645,
+      "learning_rate": 1.6431326682612947e-06,
+      "loss": 0.6037,
+      "step": 7054
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9893993557633762,
+      "learning_rate": 1.6418698196476497e-06,
+      "loss": 0.5822,
+      "step": 7055
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.502449539506666,
+      "learning_rate": 1.6406073611718593e-06,
+      "loss": 0.6611,
+      "step": 7056
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.886805930275283,
+      "learning_rate": 1.6393452929805947e-06,
+      "loss": 0.6163,
+      "step": 7057
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.324217731618782,
+      "learning_rate": 1.63808361522048e-06,
+      "loss": 0.6323,
+      "step": 7058
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4841967196947907,
+      "learning_rate": 1.6368223280380951e-06,
+      "loss": 0.6249,
+      "step": 7059
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.491514087882754,
+      "learning_rate": 1.6355614315799673e-06,
+      "loss": 0.5953,
+      "step": 7060
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.09150523009862,
+      "learning_rate": 1.6343009259925863e-06,
+      "loss": 0.663,
+      "step": 7061
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.642962343711459,
+      "learning_rate": 1.633040811422395e-06,
+      "loss": 0.6604,
+      "step": 7062
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.222557968930457,
+      "learning_rate": 1.6317810880157908e-06,
+      "loss": 0.5329,
+      "step": 7063
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.7336550978700083,
+      "learning_rate": 1.6305217559191232e-06,
+      "loss": 0.5325,
+      "step": 7064
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1083210565732484,
+      "learning_rate": 1.629262815278696e-06,
+      "loss": 0.5392,
+      "step": 7065
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.720575901544298,
+      "learning_rate": 1.6280042662407714e-06,
+      "loss": 0.6113,
+      "step": 7066
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.16404863181636,
+      "learning_rate": 1.626746108951565e-06,
+      "loss": 0.5831,
+      "step": 7067
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4115471190550117,
+      "learning_rate": 1.6254883435572449e-06,
+      "loss": 0.59,
+      "step": 7068
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.0583452239282183,
+      "learning_rate": 1.6242309702039327e-06,
+      "loss": 0.6492,
+      "step": 7069
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.2718246037744843,
+      "learning_rate": 1.6229739890377084e-06,
+      "loss": 0.5989,
+      "step": 7070
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.7419306661734812,
+      "learning_rate": 1.6217174002046032e-06,
+      "loss": 0.5111,
+      "step": 7071
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.0816295246885725,
+      "learning_rate": 1.6204612038506068e-06,
+      "loss": 0.639,
+      "step": 7072
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2970090731001003,
+      "learning_rate": 1.6192054001216585e-06,
+      "loss": 0.6099,
+      "step": 7073
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.7790697280686962,
+      "learning_rate": 1.6179499891636524e-06,
+      "loss": 0.6486,
+      "step": 7074
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6588858988791624,
+      "learning_rate": 1.61669497112244e-06,
+      "loss": 0.6991,
+      "step": 7075
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4764296283414584,
+      "learning_rate": 1.6154403461438273e-06,
+      "loss": 0.6273,
+      "step": 7076
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.658819160311856,
+      "learning_rate": 1.6141861143735716e-06,
+      "loss": 0.5699,
+      "step": 7077
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.315234814191497,
+      "learning_rate": 1.612932275957384e-06,
+      "loss": 0.6531,
+      "step": 7078
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.409181763534464,
+      "learning_rate": 1.6116788310409332e-06,
+      "loss": 0.5654,
+      "step": 7079
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.573872111742514,
+      "learning_rate": 1.6104257797698431e-06,
+      "loss": 0.6174,
+      "step": 7080
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1867723143982847,
+      "learning_rate": 1.6091731222896877e-06,
+      "loss": 0.6744,
+      "step": 7081
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.436773932857918,
+      "learning_rate": 1.6079208587459954e-06,
+      "loss": 0.6701,
+      "step": 7082
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6539143787518165,
+      "learning_rate": 1.6066689892842525e-06,
+      "loss": 0.5861,
+      "step": 7083
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9851718926572968,
+      "learning_rate": 1.6054175140498967e-06,
+      "loss": 0.6667,
+      "step": 7084
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.130011313816518,
+      "learning_rate": 1.6041664331883233e-06,
+      "loss": 0.6442,
+      "step": 7085
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.8378492504021393,
+      "learning_rate": 1.6029157468448775e-06,
+      "loss": 0.6244,
+      "step": 7086
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2783353748879542,
+      "learning_rate": 1.601665455164858e-06,
+      "loss": 0.6466,
+      "step": 7087
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.5318917740950577,
+      "learning_rate": 1.6004155582935232e-06,
+      "loss": 0.7144,
+      "step": 7088
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.405042680383137,
+      "learning_rate": 1.599166056376083e-06,
+      "loss": 0.554,
+      "step": 7089
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.537663239731998,
+      "learning_rate": 1.5979169495576991e-06,
+      "loss": 0.6213,
+      "step": 7090
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.34724684013334,
+      "learning_rate": 1.5966682379834887e-06,
+      "loss": 0.5951,
+      "step": 7091
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.683042293367126,
+      "learning_rate": 1.5954199217985233e-06,
+      "loss": 0.6249,
+      "step": 7092
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.962353374507347,
+      "learning_rate": 1.5941720011478323e-06,
+      "loss": 0.5715,
+      "step": 7093
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6239346592286377,
+      "learning_rate": 1.5929244761763924e-06,
+      "loss": 0.5831,
+      "step": 7094
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.338124980340621,
+      "learning_rate": 1.591677347029137e-06,
+      "loss": 0.576,
+      "step": 7095
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.8481926089677714,
+      "learning_rate": 1.5904306138509545e-06,
+      "loss": 0.6755,
+      "step": 7096
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.540089155983873,
+      "learning_rate": 1.5891842767866872e-06,
+      "loss": 0.5622,
+      "step": 7097
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4998455761237133,
+      "learning_rate": 1.587938335981133e-06,
+      "loss": 0.5839,
+      "step": 7098
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.281273941072614,
+      "learning_rate": 1.5866927915790391e-06,
+      "loss": 0.6369,
+      "step": 7099
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.401257413792379,
+      "learning_rate": 1.585447643725108e-06,
+      "loss": 0.6927,
+      "step": 7100
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.885694164271074,
+      "learning_rate": 1.5842028925640002e-06,
+      "loss": 0.5983,
+      "step": 7101
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.249464206134762,
+      "learning_rate": 1.5829585382403273e-06,
+      "loss": 0.552,
+      "step": 7102
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2266693760673335,
+      "learning_rate": 1.5817145808986534e-06,
+      "loss": 0.6486,
+      "step": 7103
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6048570897327545,
+      "learning_rate": 1.5804710206834972e-06,
+      "loss": 0.6163,
+      "step": 7104
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.52361431641286,
+      "learning_rate": 1.5792278577393327e-06,
+      "loss": 0.5478,
+      "step": 7105
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.0440511092050317,
+      "learning_rate": 1.577985092210587e-06,
+      "loss": 0.648,
+      "step": 7106
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5613872240202467,
+      "learning_rate": 1.5767427242416433e-06,
+      "loss": 0.6007,
+      "step": 7107
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4150239405907303,
+      "learning_rate": 1.575500753976834e-06,
+      "loss": 0.5187,
+      "step": 7108
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.236040550601962,
+      "learning_rate": 1.5742591815604463e-06,
+      "loss": 0.6156,
+      "step": 7109
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.8872460453894053,
+      "learning_rate": 1.5730180071367247e-06,
+      "loss": 0.6328,
+      "step": 7110
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.329877740414494,
+      "learning_rate": 1.5717772308498651e-06,
+      "loss": 0.6179,
+      "step": 7111
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.635293780550099,
+      "learning_rate": 1.5705368528440178e-06,
+      "loss": 0.6769,
+      "step": 7112
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5346783855495905,
+      "learning_rate": 1.569296873263283e-06,
+      "loss": 0.4746,
+      "step": 7113
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1425884362151524,
+      "learning_rate": 1.5680572922517206e-06,
+      "loss": 0.5694,
+      "step": 7114
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1871638635512443,
+      "learning_rate": 1.5668181099533431e-06,
+      "loss": 0.5599,
+      "step": 7115
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6126133591306533,
+      "learning_rate": 1.5655793265121132e-06,
+      "loss": 0.6576,
+      "step": 7116
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.6620967020652584,
+      "learning_rate": 1.5643409420719475e-06,
+      "loss": 0.7142,
+      "step": 7117
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.539598366314506,
+      "learning_rate": 1.5631029567767197e-06,
+      "loss": 0.6241,
+      "step": 7118
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.873095508009267,
+      "learning_rate": 1.5618653707702553e-06,
+      "loss": 0.6067,
+      "step": 7119
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.2962610827522454,
+      "learning_rate": 1.560628184196335e-06,
+      "loss": 0.5866,
+      "step": 7120
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.418665449353965,
+      "learning_rate": 1.55939139719869e-06,
+      "loss": 0.5084,
+      "step": 7121
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 54.21175226317471,
+      "learning_rate": 1.5581550099210053e-06,
+      "loss": 0.5915,
+      "step": 7122
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.5068547209902965,
+      "learning_rate": 1.5569190225069226e-06,
+      "loss": 0.6405,
+      "step": 7123
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.7309517743093017,
+      "learning_rate": 1.5556834351000356e-06,
+      "loss": 0.6173,
+      "step": 7124
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0711077498083168,
+      "learning_rate": 1.554448247843891e-06,
+      "loss": 0.5749,
+      "step": 7125
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4429328414022162,
+      "learning_rate": 1.5532134608819876e-06,
+      "loss": 0.5991,
+      "step": 7126
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.816964217476705,
+      "learning_rate": 1.55197907435778e-06,
+      "loss": 0.6666,
+      "step": 7127
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.535197145555227,
+      "learning_rate": 1.5507450884146784e-06,
+      "loss": 0.5548,
+      "step": 7128
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.319066286953146,
+      "learning_rate": 1.5495115031960418e-06,
+      "loss": 0.6441,
+      "step": 7129
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.8047667310373727,
+      "learning_rate": 1.5482783188451822e-06,
+      "loss": 0.6417,
+      "step": 7130
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.9106777141475417,
+      "learning_rate": 1.5470455355053704e-06,
+      "loss": 0.6389,
+      "step": 7131
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6056116741953828,
+      "learning_rate": 1.5458131533198261e-06,
+      "loss": 0.6165,
+      "step": 7132
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9784665393315755,
+      "learning_rate": 1.5445811724317267e-06,
+      "loss": 0.584,
+      "step": 7133
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.057328186390915,
+      "learning_rate": 1.5433495929841979e-06,
+      "loss": 0.538,
+      "step": 7134
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.899774453962374,
+      "learning_rate": 1.5421184151203194e-06,
+      "loss": 0.6381,
+      "step": 7135
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.302231817534724,
+      "learning_rate": 1.5408876389831278e-06,
+      "loss": 0.5251,
+      "step": 7136
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.779434622195697,
+      "learning_rate": 1.5396572647156126e-06,
+      "loss": 0.6052,
+      "step": 7137
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6994213766514155,
+      "learning_rate": 1.538427292460714e-06,
+      "loss": 0.6085,
+      "step": 7138
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.504683013737572,
+      "learning_rate": 1.5371977223613238e-06,
+      "loss": 0.6257,
+      "step": 7139
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2855288271378664,
+      "learning_rate": 1.535968554560293e-06,
+      "loss": 0.6142,
+      "step": 7140
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4654873006023608,
+      "learning_rate": 1.5347397892004234e-06,
+      "loss": 0.6748,
+      "step": 7141
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2166955270138478,
+      "learning_rate": 1.533511426424466e-06,
+      "loss": 0.5971,
+      "step": 7142
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5878170941910934,
+      "learning_rate": 1.532283466375133e-06,
+      "loss": 0.5781,
+      "step": 7143
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.088438430805986,
+      "learning_rate": 1.5310559091950805e-06,
+      "loss": 0.5924,
+      "step": 7144
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.455905487244043,
+      "learning_rate": 1.5298287550269248e-06,
+      "loss": 0.6324,
+      "step": 7145
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.296776703964561,
+      "learning_rate": 1.5286020040132344e-06,
+      "loss": 0.6318,
+      "step": 7146
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.502122104084021,
+      "learning_rate": 1.5273756562965286e-06,
+      "loss": 0.5917,
+      "step": 7147
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.813679350238381,
+      "learning_rate": 1.5261497120192791e-06,
+      "loss": 0.7005,
+      "step": 7148
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.386900010136105,
+      "learning_rate": 1.5249241713239148e-06,
+      "loss": 0.6208,
+      "step": 7149
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.097816570449662,
+      "learning_rate": 1.5236990343528157e-06,
+      "loss": 0.5714,
+      "step": 7150
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.8641415092943023,
+      "learning_rate": 1.522474301248314e-06,
+      "loss": 0.6228,
+      "step": 7151
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.953458669418631,
+      "learning_rate": 1.5212499721526942e-06,
+      "loss": 0.635,
+      "step": 7152
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.293638293253605,
+      "learning_rate": 1.5200260472081962e-06,
+      "loss": 0.6122,
+      "step": 7153
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.165393072045299,
+      "learning_rate": 1.5188025265570127e-06,
+      "loss": 0.5967,
+      "step": 7154
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6653448901727215,
+      "learning_rate": 1.51757941034129e-06,
+      "loss": 0.5992,
+      "step": 7155
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.590367390269278,
+      "learning_rate": 1.5163566987031246e-06,
+      "loss": 0.5784,
+      "step": 7156
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.274011024448419,
+      "learning_rate": 1.5151343917845662e-06,
+      "loss": 0.5486,
+      "step": 7157
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.640547755917543,
+      "learning_rate": 1.513912489727621e-06,
+      "loss": 0.6622,
+      "step": 7158
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.435856702185333,
+      "learning_rate": 1.5126909926742461e-06,
+      "loss": 0.5952,
+      "step": 7159
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.327306617131621,
+      "learning_rate": 1.511469900766352e-06,
+      "loss": 0.5737,
+      "step": 7160
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.954132593391894,
+      "learning_rate": 1.510249214145798e-06,
+      "loss": 0.6354,
+      "step": 7161
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.435417507636399,
+      "learning_rate": 1.5090289329544028e-06,
+      "loss": 0.658,
+      "step": 7162
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9854856200768136,
+      "learning_rate": 1.5078090573339365e-06,
+      "loss": 0.6282,
+      "step": 7163
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.106673158220274,
+      "learning_rate": 1.506589587426119e-06,
+      "loss": 0.5409,
+      "step": 7164
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4494465926620497,
+      "learning_rate": 1.5053705233726228e-06,
+      "loss": 0.6402,
+      "step": 7165
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1863048410990804,
+      "learning_rate": 1.5041518653150777e-06,
+      "loss": 0.6386,
+      "step": 7166
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2078949677416864,
+      "learning_rate": 1.5029336133950635e-06,
+      "loss": 0.5456,
+      "step": 7167
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.738221773697226,
+      "learning_rate": 1.5017157677541144e-06,
+      "loss": 0.6146,
+      "step": 7168
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.250673304439142,
+      "learning_rate": 1.500498328533715e-06,
+      "loss": 0.5573,
+      "step": 7169
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5150783835660797,
+      "learning_rate": 1.4992812958753023e-06,
+      "loss": 0.5967,
+      "step": 7170
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9829915284477175,
+      "learning_rate": 1.4980646699202684e-06,
+      "loss": 0.5587,
+      "step": 7171
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1533475910931696,
+      "learning_rate": 1.4968484508099606e-06,
+      "loss": 0.6385,
+      "step": 7172
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.851990435700211,
+      "learning_rate": 1.4956326386856723e-06,
+      "loss": 0.5869,
+      "step": 7173
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.7719426441646275,
+      "learning_rate": 1.494417233688653e-06,
+      "loss": 0.5874,
+      "step": 7174
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.102882961445121,
+      "learning_rate": 1.4932022359601056e-06,
+      "loss": 0.6421,
+      "step": 7175
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6384157210222536,
+      "learning_rate": 1.4919876456411875e-06,
+      "loss": 0.5523,
+      "step": 7176
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.378591334938083,
+      "learning_rate": 1.490773462873002e-06,
+      "loss": 0.544,
+      "step": 7177
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 20.641360117403863,
+      "learning_rate": 1.4895596877966128e-06,
+      "loss": 0.636,
+      "step": 7178
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4236725068487304,
+      "learning_rate": 1.4883463205530302e-06,
+      "loss": 0.6411,
+      "step": 7179
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2183036139750714,
+      "learning_rate": 1.4871333612832206e-06,
+      "loss": 0.5712,
+      "step": 7180
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.050796166901124,
+      "learning_rate": 1.4859208101281041e-06,
+      "loss": 0.6366,
+      "step": 7181
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4152995939569166,
+      "learning_rate": 1.48470866722855e-06,
+      "loss": 0.5498,
+      "step": 7182
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.362918250869478,
+      "learning_rate": 1.4834969327253795e-06,
+      "loss": 0.5972,
+      "step": 7183
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0541545853510184,
+      "learning_rate": 1.48228560675937e-06,
+      "loss": 0.5475,
+      "step": 7184
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.0605689354251995,
+      "learning_rate": 1.481074689471252e-06,
+      "loss": 0.5439,
+      "step": 7185
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1804727796434964,
+      "learning_rate": 1.479864181001704e-06,
+      "loss": 0.6155,
+      "step": 7186
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.353882130124197,
+      "learning_rate": 1.4786540814913586e-06,
+      "loss": 0.6792,
+      "step": 7187
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4609607660129162,
+      "learning_rate": 1.4774443910808023e-06,
+      "loss": 0.6776,
+      "step": 7188
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6666982763784812,
+      "learning_rate": 1.476235109910576e-06,
+      "loss": 0.6191,
+      "step": 7189
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.56167731467775,
+      "learning_rate": 1.4750262381211665e-06,
+      "loss": 0.5887,
+      "step": 7190
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.44278557698414,
+      "learning_rate": 1.4738177758530208e-06,
+      "loss": 0.6259,
+      "step": 7191
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5250241812484906,
+      "learning_rate": 1.4726097232465314e-06,
+      "loss": 0.6089,
+      "step": 7192
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.2400336695589744,
+      "learning_rate": 1.471402080442047e-06,
+      "loss": 0.5666,
+      "step": 7193
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7151462386306657,
+      "learning_rate": 1.470194847579871e-06,
+      "loss": 0.6301,
+      "step": 7194
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.698877450197644,
+      "learning_rate": 1.4689880248002537e-06,
+      "loss": 0.7084,
+      "step": 7195
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6410464096856665,
+      "learning_rate": 1.467781612243399e-06,
+      "loss": 0.6245,
+      "step": 7196
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.612323546310781,
+      "learning_rate": 1.466575610049466e-06,
+      "loss": 0.6783,
+      "step": 7197
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7178474296193564,
+      "learning_rate": 1.4653700183585663e-06,
+      "loss": 0.6122,
+      "step": 7198
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.512024045937146,
+      "learning_rate": 1.4641648373107598e-06,
+      "loss": 0.6147,
+      "step": 7199
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0834058077279685,
+      "learning_rate": 1.4629600670460603e-06,
+      "loss": 0.5514,
+      "step": 7200
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.216935664488474,
+      "learning_rate": 1.4617557077044352e-06,
+      "loss": 0.6227,
+      "step": 7201
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.46166081455518,
+      "learning_rate": 1.4605517594258046e-06,
+      "loss": 0.5871,
+      "step": 7202
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1948644668696375,
+      "learning_rate": 1.4593482223500406e-06,
+      "loss": 0.6353,
+      "step": 7203
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.33193407924704,
+      "learning_rate": 1.4581450966169648e-06,
+      "loss": 0.6246,
+      "step": 7204
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.624255396763174,
+      "learning_rate": 1.4569423823663515e-06,
+      "loss": 0.6982,
+      "step": 7205
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3813097303259028,
+      "learning_rate": 1.4557400797379306e-06,
+      "loss": 0.5593,
+      "step": 7206
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8170072633637067,
+      "learning_rate": 1.4545381888713833e-06,
+      "loss": 0.622,
+      "step": 7207
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0227331919960505,
+      "learning_rate": 1.45333670990634e-06,
+      "loss": 0.4883,
+      "step": 7208
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.9590588104321056,
+      "learning_rate": 1.452135642982384e-06,
+      "loss": 0.6183,
+      "step": 7209
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.33117855582943,
+      "learning_rate": 1.450934988239054e-06,
+      "loss": 0.6134,
+      "step": 7210
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4557426055942786,
+      "learning_rate": 1.4497347458158384e-06,
+      "loss": 0.62,
+      "step": 7211
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.926982164970039,
+      "learning_rate": 1.4485349158521756e-06,
+      "loss": 0.6036,
+      "step": 7212
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2092246121938164,
+      "learning_rate": 1.447335498487462e-06,
+      "loss": 0.6053,
+      "step": 7213
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4017310344462994,
+      "learning_rate": 1.446136493861039e-06,
+      "loss": 0.6659,
+      "step": 7214
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1199053957284346,
+      "learning_rate": 1.4449379021122045e-06,
+      "loss": 0.6202,
+      "step": 7215
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3367095552631816,
+      "learning_rate": 1.4437397233802098e-06,
+      "loss": 0.629,
+      "step": 7216
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.0011471861240357,
+      "learning_rate": 1.4425419578042538e-06,
+      "loss": 0.6302,
+      "step": 7217
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.595981973547213,
+      "learning_rate": 1.4413446055234882e-06,
+      "loss": 0.6064,
+      "step": 7218
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.0100131890890243,
+      "learning_rate": 1.4401476666770191e-06,
+      "loss": 0.6153,
+      "step": 7219
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.068773777984486,
+      "learning_rate": 1.4389511414039053e-06,
+      "loss": 0.5228,
+      "step": 7220
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.183378387130152,
+      "learning_rate": 1.4377550298431536e-06,
+      "loss": 0.6687,
+      "step": 7221
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8674171438993237,
+      "learning_rate": 1.436559332133724e-06,
+      "loss": 0.6141,
+      "step": 7222
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8843731419145637,
+      "learning_rate": 1.4353640484145304e-06,
+      "loss": 0.5749,
+      "step": 7223
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.27701809442643,
+      "learning_rate": 1.4341691788244383e-06,
+      "loss": 0.6422,
+      "step": 7224
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5322604425389112,
+      "learning_rate": 1.4329747235022624e-06,
+      "loss": 0.5397,
+      "step": 7225
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.431501916595479,
+      "learning_rate": 1.431780682586773e-06,
+      "loss": 0.6515,
+      "step": 7226
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3394101864352983,
+      "learning_rate": 1.4305870562166873e-06,
+      "loss": 0.6069,
+      "step": 7227
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2242293964368156,
+      "learning_rate": 1.4293938445306798e-06,
+      "loss": 0.5442,
+      "step": 7228
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5833298123991604,
+      "learning_rate": 1.4282010476673758e-06,
+      "loss": 0.6004,
+      "step": 7229
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6359655886818913,
+      "learning_rate": 1.427008665765348e-06,
+      "loss": 0.6834,
+      "step": 7230
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.585971432337667,
+      "learning_rate": 1.4258166989631245e-06,
+      "loss": 0.6375,
+      "step": 7231
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.257480641444835,
+      "learning_rate": 1.4246251473991845e-06,
+      "loss": 0.6148,
+      "step": 7232
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6802149842143397,
+      "learning_rate": 1.423434011211961e-06,
+      "loss": 0.6633,
+      "step": 7233
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.460163635030121,
+      "learning_rate": 1.4222432905398353e-06,
+      "loss": 0.5807,
+      "step": 7234
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.22647375488727,
+      "learning_rate": 1.4210529855211403e-06,
+      "loss": 0.4943,
+      "step": 7235
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.121150139420541,
+      "learning_rate": 1.4198630962941639e-06,
+      "loss": 0.6249,
+      "step": 7236
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.946918461716052,
+      "learning_rate": 1.4186736229971455e-06,
+      "loss": 0.6211,
+      "step": 7237
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.528975031658922,
+      "learning_rate": 1.4174845657682712e-06,
+      "loss": 0.6419,
+      "step": 7238
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5787193861812496,
+      "learning_rate": 1.4162959247456854e-06,
+      "loss": 0.6206,
+      "step": 7239
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.406544835370247,
+      "learning_rate": 1.4151077000674784e-06,
+      "loss": 0.624,
+      "step": 7240
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.207698162326682,
+      "learning_rate": 1.4139198918716956e-06,
+      "loss": 0.5709,
+      "step": 7241
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2463252191495924,
+      "learning_rate": 1.4127325002963355e-06,
+      "loss": 0.5574,
+      "step": 7242
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3509763553384597,
+      "learning_rate": 1.4115455254793436e-06,
+      "loss": 0.5806,
+      "step": 7243
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.52195463891043,
+      "learning_rate": 1.4103589675586176e-06,
+      "loss": 0.6281,
+      "step": 7244
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6057177351757734,
+      "learning_rate": 1.4091728266720106e-06,
+      "loss": 0.5374,
+      "step": 7245
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.378426724174707,
+      "learning_rate": 1.4079871029573254e-06,
+      "loss": 0.5312,
+      "step": 7246
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4658224720117308,
+      "learning_rate": 1.4068017965523146e-06,
+      "loss": 0.6043,
+      "step": 7247
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.367796137825628,
+      "learning_rate": 1.4056169075946846e-06,
+      "loss": 0.5532,
+      "step": 7248
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4057618387618436,
+      "learning_rate": 1.4044324362220912e-06,
+      "loss": 0.608,
+      "step": 7249
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.7601186688014003,
+      "learning_rate": 1.4032483825721432e-06,
+      "loss": 0.6427,
+      "step": 7250
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5498571521744045,
+      "learning_rate": 1.4020647467824028e-06,
+      "loss": 0.6887,
+      "step": 7251
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9344224790730087,
+      "learning_rate": 1.4008815289903798e-06,
+      "loss": 0.5365,
+      "step": 7252
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.219760527227203,
+      "learning_rate": 1.3996987293335345e-06,
+      "loss": 0.6433,
+      "step": 7253
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.310538446761164,
+      "learning_rate": 1.3985163479492842e-06,
+      "loss": 0.6225,
+      "step": 7254
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8517117856482517,
+      "learning_rate": 1.3973343849749948e-06,
+      "loss": 0.6371,
+      "step": 7255
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.661140852133978,
+      "learning_rate": 1.3961528405479824e-06,
+      "loss": 0.6281,
+      "step": 7256
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2805811042491286,
+      "learning_rate": 1.3949717148055136e-06,
+      "loss": 0.5957,
+      "step": 7257
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.300483005629338,
+      "learning_rate": 1.39379100788481e-06,
+      "loss": 0.6551,
+      "step": 7258
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.485309336234544,
+      "learning_rate": 1.3926107199230442e-06,
+      "loss": 0.6081,
+      "step": 7259
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.341174494875231,
+      "learning_rate": 1.3914308510573354e-06,
+      "loss": 0.6515,
+      "step": 7260
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8732963511207283,
+      "learning_rate": 1.3902514014247608e-06,
+      "loss": 0.5541,
+      "step": 7261
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2043168478650896,
+      "learning_rate": 1.3890723711623421e-06,
+      "loss": 0.588,
+      "step": 7262
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.853557533816863,
+      "learning_rate": 1.3878937604070568e-06,
+      "loss": 0.5689,
+      "step": 7263
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8253444135108845,
+      "learning_rate": 1.386715569295835e-06,
+      "loss": 0.5745,
+      "step": 7264
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9796152518485929,
+      "learning_rate": 1.3855377979655533e-06,
+      "loss": 0.5749,
+      "step": 7265
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.6160882171024125,
+      "learning_rate": 1.3843604465530398e-06,
+      "loss": 0.5622,
+      "step": 7266
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0041115107011476,
+      "learning_rate": 1.3831835151950784e-06,
+      "loss": 0.5528,
+      "step": 7267
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.16044720788834,
+      "learning_rate": 1.3820070040284023e-06,
+      "loss": 0.6886,
+      "step": 7268
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8645110276055927,
+      "learning_rate": 1.380830913189694e-06,
+      "loss": 0.6323,
+      "step": 7269
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0345813564664548,
+      "learning_rate": 1.3796552428155868e-06,
+      "loss": 0.5463,
+      "step": 7270
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4861411344231317,
+      "learning_rate": 1.378479993042668e-06,
+      "loss": 0.6813,
+      "step": 7271
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5548781593021594,
+      "learning_rate": 1.3773051640074764e-06,
+      "loss": 0.6686,
+      "step": 7272
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7242847584411165,
+      "learning_rate": 1.3761307558464975e-06,
+      "loss": 0.635,
+      "step": 7273
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.079673277666802,
+      "learning_rate": 1.3749567686961728e-06,
+      "loss": 0.6109,
+      "step": 7274
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9119319275996398,
+      "learning_rate": 1.3737832026928905e-06,
+      "loss": 0.5083,
+      "step": 7275
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.645571565188374,
+      "learning_rate": 1.3726100579729935e-06,
+      "loss": 0.6791,
+      "step": 7276
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.280058640830768,
+      "learning_rate": 1.3714373346727754e-06,
+      "loss": 0.6213,
+      "step": 7277
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3684177527062755,
+      "learning_rate": 1.3702650329284794e-06,
+      "loss": 0.6239,
+      "step": 7278
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.5713000588210955,
+      "learning_rate": 1.3690931528762974e-06,
+      "loss": 0.6513,
+      "step": 7279
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.191606619177731,
+      "learning_rate": 1.3679216946523771e-06,
+      "loss": 0.6179,
+      "step": 7280
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.520087458009411,
+      "learning_rate": 1.3667506583928163e-06,
+      "loss": 0.6273,
+      "step": 7281
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.8211254894412083,
+      "learning_rate": 1.3655800442336597e-06,
+      "loss": 0.6629,
+      "step": 7282
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5218498030614986,
+      "learning_rate": 1.3644098523109096e-06,
+      "loss": 0.5631,
+      "step": 7283
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.187549795477403,
+      "learning_rate": 1.3632400827605113e-06,
+      "loss": 0.552,
+      "step": 7284
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.067246657340486,
+      "learning_rate": 1.3620707357183694e-06,
+      "loss": 0.6166,
+      "step": 7285
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 15.55712958329288,
+      "learning_rate": 1.3609018113203314e-06,
+      "loss": 0.6316,
+      "step": 7286
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5699486322531566,
+      "learning_rate": 1.3597333097022031e-06,
+      "loss": 0.6579,
+      "step": 7287
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.606631331064434,
+      "learning_rate": 1.3585652309997344e-06,
+      "loss": 0.6924,
+      "step": 7288
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4639655618590646,
+      "learning_rate": 1.3573975753486313e-06,
+      "loss": 0.6486,
+      "step": 7289
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.494505254711063,
+      "learning_rate": 1.356230342884549e-06,
+      "loss": 0.6138,
+      "step": 7290
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6073670766020154,
+      "learning_rate": 1.3550635337430928e-06,
+      "loss": 0.5951,
+      "step": 7291
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.2580430689462734,
+      "learning_rate": 1.3538971480598167e-06,
+      "loss": 0.6027,
+      "step": 7292
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9517406444625436,
+      "learning_rate": 1.3527311859702308e-06,
+      "loss": 0.5322,
+      "step": 7293
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.860206476602887,
+      "learning_rate": 1.3515656476097937e-06,
+      "loss": 0.5832,
+      "step": 7294
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.9771138110934405,
+      "learning_rate": 1.350400533113912e-06,
+      "loss": 0.543,
+      "step": 7295
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.582876236962234,
+      "learning_rate": 1.3492358426179475e-06,
+      "loss": 0.6746,
+      "step": 7296
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0636423377041937,
+      "learning_rate": 1.3480715762572078e-06,
+      "loss": 0.657,
+      "step": 7297
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.220396192129797,
+      "learning_rate": 1.3469077341669579e-06,
+      "loss": 0.6404,
+      "step": 7298
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.8263887939999264,
+      "learning_rate": 1.3457443164824053e-06,
+      "loss": 0.6004,
+      "step": 7299
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.70882407444034,
+      "learning_rate": 1.3445813233387167e-06,
+      "loss": 0.7198,
+      "step": 7300
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.3517602576307355,
+      "learning_rate": 1.3434187548710014e-06,
+      "loss": 0.6133,
+      "step": 7301
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.8385464767154356,
+      "learning_rate": 1.3422566112143248e-06,
+      "loss": 0.527,
+      "step": 7302
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7137520222257563,
+      "learning_rate": 1.3410948925037037e-06,
+      "loss": 0.7291,
+      "step": 7303
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.238965834669248,
+      "learning_rate": 1.3399335988741007e-06,
+      "loss": 0.6578,
+      "step": 7304
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9827149232414178,
+      "learning_rate": 1.338772730460431e-06,
+      "loss": 0.5438,
+      "step": 7305
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.982687189717289,
+      "learning_rate": 1.3376122873975616e-06,
+      "loss": 0.5661,
+      "step": 7306
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.394229778749516,
+      "learning_rate": 1.3364522698203114e-06,
+      "loss": 0.5627,
+      "step": 7307
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5676561438034358,
+      "learning_rate": 1.3352926778634446e-06,
+      "loss": 0.5664,
+      "step": 7308
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.69224823273842,
+      "learning_rate": 1.3341335116616822e-06,
+      "loss": 0.5727,
+      "step": 7309
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.8984898812420727,
+      "learning_rate": 1.3329747713496904e-06,
+      "loss": 0.6014,
+      "step": 7310
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5133785385178657,
+      "learning_rate": 1.331816457062089e-06,
+      "loss": 0.6487,
+      "step": 7311
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.550922385813062,
+      "learning_rate": 1.3306585689334494e-06,
+      "loss": 0.605,
+      "step": 7312
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6330342158074487,
+      "learning_rate": 1.3295011070982906e-06,
+      "loss": 0.5916,
+      "step": 7313
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.906335153096584,
+      "learning_rate": 1.3283440716910812e-06,
+      "loss": 0.6149,
+      "step": 7314
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.095159349488047,
+      "learning_rate": 1.327187462846244e-06,
+      "loss": 0.5979,
+      "step": 7315
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.8365869518598976,
+      "learning_rate": 1.3260312806981517e-06,
+      "loss": 0.4815,
+      "step": 7316
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.072771877951047,
+      "learning_rate": 1.3248755253811236e-06,
+      "loss": 0.5718,
+      "step": 7317
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.5814898012658896,
+      "learning_rate": 1.3237201970294344e-06,
+      "loss": 0.5606,
+      "step": 7318
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.290088791951938,
+      "learning_rate": 1.3225652957773044e-06,
+      "loss": 0.6147,
+      "step": 7319
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.485404940783879,
+      "learning_rate": 1.3214108217589095e-06,
+      "loss": 0.5559,
+      "step": 7320
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.890814837354043,
+      "learning_rate": 1.3202567751083701e-06,
+      "loss": 0.5864,
+      "step": 7321
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3304011547001933,
+      "learning_rate": 1.3191031559597628e-06,
+      "loss": 0.686,
+      "step": 7322
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.939817245282438,
+      "learning_rate": 1.3179499644471088e-06,
+      "loss": 0.7083,
+      "step": 7323
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9322448250042434,
+      "learning_rate": 1.3167972007043844e-06,
+      "loss": 0.617,
+      "step": 7324
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2730836852993885,
+      "learning_rate": 1.3156448648655163e-06,
+      "loss": 0.6175,
+      "step": 7325
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.216691599440806,
+      "learning_rate": 1.3144929570643767e-06,
+      "loss": 0.5699,
+      "step": 7326
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2709102711271103,
+      "learning_rate": 1.3133414774347903e-06,
+      "loss": 0.6783,
+      "step": 7327
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.7008961073983926,
+      "learning_rate": 1.3121904261105339e-06,
+      "loss": 0.5866,
+      "step": 7328
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0182432463584763,
+      "learning_rate": 1.3110398032253346e-06,
+      "loss": 0.5087,
+      "step": 7329
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0497274511645123,
+      "learning_rate": 1.3098896089128666e-06,
+      "loss": 0.5744,
+      "step": 7330
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2533986618656225,
+      "learning_rate": 1.3087398433067577e-06,
+      "loss": 0.5694,
+      "step": 7331
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4413558422766193,
+      "learning_rate": 1.307590506540582e-06,
+      "loss": 0.616,
+      "step": 7332
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3644581867426897,
+      "learning_rate": 1.3064415987478691e-06,
+      "loss": 0.6208,
+      "step": 7333
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9694715957645057,
+      "learning_rate": 1.3052931200620926e-06,
+      "loss": 0.5448,
+      "step": 7334
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.989251780651652,
+      "learning_rate": 1.3041450706166831e-06,
+      "loss": 0.5686,
+      "step": 7335
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.568697403594937,
+      "learning_rate": 1.3029974505450137e-06,
+      "loss": 0.6258,
+      "step": 7336
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4006514658503697,
+      "learning_rate": 1.301850259980414e-06,
+      "loss": 0.587,
+      "step": 7337
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4391008775614704,
+      "learning_rate": 1.3007034990561619e-06,
+      "loss": 0.6487,
+      "step": 7338
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.27176792056776,
+      "learning_rate": 1.2995571679054835e-06,
+      "loss": 0.6879,
+      "step": 7339
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.332459416723728,
+      "learning_rate": 1.2984112666615555e-06,
+      "loss": 0.5647,
+      "step": 7340
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0350097122071702,
+      "learning_rate": 1.2972657954575064e-06,
+      "loss": 0.541,
+      "step": 7341
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3128014660034273,
+      "learning_rate": 1.2961207544264149e-06,
+      "loss": 0.6494,
+      "step": 7342
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.281808135510517,
+      "learning_rate": 1.2949761437013059e-06,
+      "loss": 0.6048,
+      "step": 7343
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.495322499540691,
+      "learning_rate": 1.2938319634151597e-06,
+      "loss": 0.5942,
+      "step": 7344
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.914464189765621,
+      "learning_rate": 1.2926882137009012e-06,
+      "loss": 0.66,
+      "step": 7345
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.098291543647439,
+      "learning_rate": 1.2915448946914106e-06,
+      "loss": 0.5978,
+      "step": 7346
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.435837683057105,
+      "learning_rate": 1.2904020065195127e-06,
+      "loss": 0.5946,
+      "step": 7347
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4918968126223633,
+      "learning_rate": 1.2892595493179876e-06,
+      "loss": 0.6562,
+      "step": 7348
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2303147723923877,
+      "learning_rate": 1.2881175232195604e-06,
+      "loss": 0.6037,
+      "step": 7349
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4085569825882223,
+      "learning_rate": 1.2869759283569088e-06,
+      "loss": 0.5633,
+      "step": 7350
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.9401916376736774,
+      "learning_rate": 1.2858347648626623e-06,
+      "loss": 0.6793,
+      "step": 7351
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.620281409590404,
+      "learning_rate": 1.2846940328693952e-06,
+      "loss": 0.5221,
+      "step": 7352
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.460645653228923,
+      "learning_rate": 1.2835537325096364e-06,
+      "loss": 0.6427,
+      "step": 7353
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1645773174699605,
+      "learning_rate": 1.2824138639158607e-06,
+      "loss": 0.5951,
+      "step": 7354
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.940681938546237,
+      "learning_rate": 1.2812744272204969e-06,
+      "loss": 0.6173,
+      "step": 7355
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9852641803573394,
+      "learning_rate": 1.2801354225559194e-06,
+      "loss": 0.634,
+      "step": 7356
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.408422665342694,
+      "learning_rate": 1.2789968500544563e-06,
+      "loss": 0.5832,
+      "step": 7357
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.54563268747217,
+      "learning_rate": 1.277858709848382e-06,
+      "loss": 0.5927,
+      "step": 7358
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.892292877622211,
+      "learning_rate": 1.2767210020699234e-06,
+      "loss": 0.6332,
+      "step": 7359
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9690186601279351,
+      "learning_rate": 1.2755837268512566e-06,
+      "loss": 0.5541,
+      "step": 7360
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3973108957180673,
+      "learning_rate": 1.2744468843245066e-06,
+      "loss": 0.6973,
+      "step": 7361
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.445703072300839,
+      "learning_rate": 1.2733104746217468e-06,
+      "loss": 0.5102,
+      "step": 7362
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4460880394351956,
+      "learning_rate": 1.2721744978750028e-06,
+      "loss": 0.6275,
+      "step": 7363
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5731172106909193,
+      "learning_rate": 1.271038954216251e-06,
+      "loss": 0.567,
+      "step": 7364
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.312904188551475,
+      "learning_rate": 1.269903843777413e-06,
+      "loss": 0.6004,
+      "step": 7365
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3261426877154916,
+      "learning_rate": 1.2687691666903657e-06,
+      "loss": 0.5648,
+      "step": 7366
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.634024092091831,
+      "learning_rate": 1.2676349230869283e-06,
+      "loss": 0.6585,
+      "step": 7367
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5573244615411075,
+      "learning_rate": 1.2665011130988786e-06,
+      "loss": 0.5732,
+      "step": 7368
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2354267402034234,
+      "learning_rate": 1.2653677368579354e-06,
+      "loss": 0.5843,
+      "step": 7369
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2289656447356867,
+      "learning_rate": 1.2642347944957744e-06,
+      "loss": 0.7049,
+      "step": 7370
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.1454417786703353,
+      "learning_rate": 1.2631022861440145e-06,
+      "loss": 0.5499,
+      "step": 7371
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.367273641722864,
+      "learning_rate": 1.2619702119342286e-06,
+      "loss": 0.5656,
+      "step": 7372
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.156216839264307,
+      "learning_rate": 1.2608385719979394e-06,
+      "loss": 0.6229,
+      "step": 7373
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.707241733582545,
+      "learning_rate": 1.2597073664666159e-06,
+      "loss": 0.5974,
+      "step": 7374
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.589860918286414,
+      "learning_rate": 1.2585765954716773e-06,
+      "loss": 0.6479,
+      "step": 7375
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1972502473528763,
+      "learning_rate": 1.257446259144494e-06,
+      "loss": 0.6352,
+      "step": 7376
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.8739108933647244,
+      "learning_rate": 1.2563163576163879e-06,
+      "loss": 0.6169,
+      "step": 7377
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.599760260919812,
+      "learning_rate": 1.2551868910186238e-06,
+      "loss": 0.5686,
+      "step": 7378
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0036011302793741,
+      "learning_rate": 1.2540578594824226e-06,
+      "loss": 0.5261,
+      "step": 7379
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.1369364911691604,
+      "learning_rate": 1.25292926313895e-06,
+      "loss": 0.5366,
+      "step": 7380
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.733863227114797,
+      "learning_rate": 1.251801102119325e-06,
+      "loss": 0.5542,
+      "step": 7381
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.392211563561417,
+      "learning_rate": 1.2506733765546115e-06,
+      "loss": 0.5557,
+      "step": 7382
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.2289899376136315,
+      "learning_rate": 1.2495460865758286e-06,
+      "loss": 0.6201,
+      "step": 7383
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.389324115301021,
+      "learning_rate": 1.2484192323139382e-06,
+      "loss": 0.5614,
+      "step": 7384
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4415323240042524,
+      "learning_rate": 1.2472928138998569e-06,
+      "loss": 0.5961,
+      "step": 7385
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.376325421905973,
+      "learning_rate": 1.2461668314644499e-06,
+      "loss": 0.551,
+      "step": 7386
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6372801647944795,
+      "learning_rate": 1.2450412851385275e-06,
+      "loss": 0.6422,
+      "step": 7387
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0428132195188833,
+      "learning_rate": 1.2439161750528555e-06,
+      "loss": 0.5497,
+      "step": 7388
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.5630278208472985,
+      "learning_rate": 1.2427915013381436e-06,
+      "loss": 0.5869,
+      "step": 7389
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0559100223708637,
+      "learning_rate": 1.2416672641250548e-06,
+      "loss": 0.5977,
+      "step": 7390
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4409035745001746,
+      "learning_rate": 1.2405434635441982e-06,
+      "loss": 0.559,
+      "step": 7391
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.831933762565664,
+      "learning_rate": 1.2394200997261358e-06,
+      "loss": 0.6682,
+      "step": 7392
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8825046392899807,
+      "learning_rate": 1.2382971728013742e-06,
+      "loss": 0.5093,
+      "step": 7393
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5797992352654386,
+      "learning_rate": 1.2371746829003745e-06,
+      "loss": 0.6392,
+      "step": 7394
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.9414008384348906,
+      "learning_rate": 1.2360526301535408e-06,
+      "loss": 0.5538,
+      "step": 7395
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.37560198530972,
+      "learning_rate": 1.234931014691234e-06,
+      "loss": 0.6542,
+      "step": 7396
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.760017872274412,
+      "learning_rate": 1.2338098366437574e-06,
+      "loss": 0.6122,
+      "step": 7397
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.8422640961336607,
+      "learning_rate": 1.2326890961413663e-06,
+      "loss": 0.5873,
+      "step": 7398
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.1453980578084426,
+      "learning_rate": 1.2315687933142672e-06,
+      "loss": 0.6239,
+      "step": 7399
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.749851183004622,
+      "learning_rate": 1.2304489282926109e-06,
+      "loss": 0.5232,
+      "step": 7400
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.472329629666702,
+      "learning_rate": 1.2293295012065032e-06,
+      "loss": 0.6043,
+      "step": 7401
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.7136693979225757,
+      "learning_rate": 1.228210512185992e-06,
+      "loss": 0.572,
+      "step": 7402
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3539836790068907,
+      "learning_rate": 1.2270919613610828e-06,
+      "loss": 0.5805,
+      "step": 7403
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4768244757191487,
+      "learning_rate": 1.2259738488617211e-06,
+      "loss": 0.6012,
+      "step": 7404
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6680180975475234,
+      "learning_rate": 1.2248561748178094e-06,
+      "loss": 0.647,
+      "step": 7405
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2801324593756784,
+      "learning_rate": 1.2237389393591931e-06,
+      "loss": 0.688,
+      "step": 7406
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.0717420409196436,
+      "learning_rate": 1.222622142615671e-06,
+      "loss": 0.6413,
+      "step": 7407
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3216939444158515,
+      "learning_rate": 1.2215057847169904e-06,
+      "loss": 0.5048,
+      "step": 7408
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.7933790970927896,
+      "learning_rate": 1.2203898657928453e-06,
+      "loss": 0.6127,
+      "step": 7409
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6973298552671556,
+      "learning_rate": 1.2192743859728784e-06,
+      "loss": 0.5957,
+      "step": 7410
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.010078673364939,
+      "learning_rate": 1.2181593453866841e-06,
+      "loss": 0.5948,
+      "step": 7411
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0077226242761792,
+      "learning_rate": 1.2170447441638067e-06,
+      "loss": 0.5568,
+      "step": 7412
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3706862467692345,
+      "learning_rate": 1.2159305824337337e-06,
+      "loss": 0.5863,
+      "step": 7413
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.0863837485718935,
+      "learning_rate": 1.2148168603259086e-06,
+      "loss": 0.6177,
+      "step": 7414
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.246542506000711,
+      "learning_rate": 1.213703577969717e-06,
+      "loss": 0.5643,
+      "step": 7415
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2507630179043505,
+      "learning_rate": 1.2125907354945004e-06,
+      "loss": 0.5588,
+      "step": 7416
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.077759944294835,
+      "learning_rate": 1.2114783330295426e-06,
+      "loss": 0.6429,
+      "step": 7417
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1001739169623956,
+      "learning_rate": 1.210366370704082e-06,
+      "loss": 0.6044,
+      "step": 7418
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0635599866105245,
+      "learning_rate": 1.2092548486473e-06,
+      "loss": 0.5829,
+      "step": 7419
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.7552402862941308,
+      "learning_rate": 1.2081437669883323e-06,
+      "loss": 0.611,
+      "step": 7420
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6069981960534463,
+      "learning_rate": 1.2070331258562612e-06,
+      "loss": 0.6104,
+      "step": 7421
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.173648068395166,
+      "learning_rate": 1.2059229253801164e-06,
+      "loss": 0.6458,
+      "step": 7422
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.606347301566492,
+      "learning_rate": 1.2048131656888801e-06,
+      "loss": 0.6129,
+      "step": 7423
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.002157820915137,
+      "learning_rate": 1.2037038469114775e-06,
+      "loss": 0.6574,
+      "step": 7424
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1433237595069006,
+      "learning_rate": 1.2025949691767895e-06,
+      "loss": 0.6796,
+      "step": 7425
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6482892199703247,
+      "learning_rate": 1.2014865326136393e-06,
+      "loss": 0.6542,
+      "step": 7426
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.8554690646087244,
+      "learning_rate": 1.2003785373508054e-06,
+      "loss": 0.5909,
+      "step": 7427
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1151193571586755,
+      "learning_rate": 1.1992709835170075e-06,
+      "loss": 0.6143,
+      "step": 7428
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3281163582456483,
+      "learning_rate": 1.198163871240921e-06,
+      "loss": 0.4967,
+      "step": 7429
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3181484526166045,
+      "learning_rate": 1.197057200651165e-06,
+      "loss": 0.6032,
+      "step": 7430
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.4330393599119065,
+      "learning_rate": 1.195950971876312e-06,
+      "loss": 0.5914,
+      "step": 7431
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.740984634407769,
+      "learning_rate": 1.1948451850448767e-06,
+      "loss": 0.7403,
+      "step": 7432
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3670282194811905,
+      "learning_rate": 1.1937398402853283e-06,
+      "loss": 0.5251,
+      "step": 7433
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0276430189430175,
+      "learning_rate": 1.1926349377260843e-06,
+      "loss": 0.538,
+      "step": 7434
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4380874996957984,
+      "learning_rate": 1.1915304774955054e-06,
+      "loss": 0.6245,
+      "step": 7435
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.242941226659693,
+      "learning_rate": 1.1904264597219078e-06,
+      "loss": 0.5773,
+      "step": 7436
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.993175401734412,
+      "learning_rate": 1.189322884533551e-06,
+      "loss": 0.6,
+      "step": 7437
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9803046786575148,
+      "learning_rate": 1.1882197520586464e-06,
+      "loss": 0.5364,
+      "step": 7438
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.157862409696275,
+      "learning_rate": 1.1871170624253515e-06,
+      "loss": 0.5523,
+      "step": 7439
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5580589814696273,
+      "learning_rate": 1.1860148157617757e-06,
+      "loss": 0.6233,
+      "step": 7440
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4718486982211854,
+      "learning_rate": 1.1849130121959717e-06,
+      "loss": 0.5901,
+      "step": 7441
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4594217860273098,
+      "learning_rate": 1.1838116518559474e-06,
+      "loss": 0.6861,
+      "step": 7442
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9338351636324551,
+      "learning_rate": 1.1827107348696526e-06,
+      "loss": 0.5445,
+      "step": 7443
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3059234955489814,
+      "learning_rate": 1.181610261364991e-06,
+      "loss": 0.5499,
+      "step": 7444
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.581067104733737,
+      "learning_rate": 1.1805102314698103e-06,
+      "loss": 0.6374,
+      "step": 7445
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.9256095254675536,
+      "learning_rate": 1.1794106453119098e-06,
+      "loss": 0.5952,
+      "step": 7446
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5727809288583776,
+      "learning_rate": 1.1783115030190378e-06,
+      "loss": 0.6924,
+      "step": 7447
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.922966924396616,
+      "learning_rate": 1.1772128047188864e-06,
+      "loss": 0.6865,
+      "step": 7448
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.39429256222939,
+      "learning_rate": 1.1761145505391025e-06,
+      "loss": 0.6303,
+      "step": 7449
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.7885079254569667,
+      "learning_rate": 1.1750167406072743e-06,
+      "loss": 0.5501,
+      "step": 7450
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4497210304886603,
+      "learning_rate": 1.1739193750509465e-06,
+      "loss": 0.5928,
+      "step": 7451
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4756214907759198,
+      "learning_rate": 1.1728224539976035e-06,
+      "loss": 0.5829,
+      "step": 7452
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.503009767146594,
+      "learning_rate": 1.1717259775746865e-06,
+      "loss": 0.5743,
+      "step": 7453
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4617513372135003,
+      "learning_rate": 1.1706299459095776e-06,
+      "loss": 0.6251,
+      "step": 7454
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2715662368230274,
+      "learning_rate": 1.1695343591296115e-06,
+      "loss": 0.6545,
+      "step": 7455
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.2007462619830287,
+      "learning_rate": 1.1684392173620729e-06,
+      "loss": 0.5534,
+      "step": 7456
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9869030111837869,
+      "learning_rate": 1.1673445207341882e-06,
+      "loss": 0.534,
+      "step": 7457
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9843494999894192,
+      "learning_rate": 1.1662502693731393e-06,
+      "loss": 0.5375,
+      "step": 7458
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.964320138816884,
+      "learning_rate": 1.1651564634060509e-06,
+      "loss": 0.5889,
+      "step": 7459
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3924454080709854,
+      "learning_rate": 1.1640631029600002e-06,
+      "loss": 0.5073,
+      "step": 7460
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.8236225380832143,
+      "learning_rate": 1.1629701881620086e-06,
+      "loss": 0.6282,
+      "step": 7461
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.172480547369297,
+      "learning_rate": 1.1618777191390502e-06,
+      "loss": 0.5964,
+      "step": 7462
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.810127828045121,
+      "learning_rate": 1.1607856960180413e-06,
+      "loss": 0.6193,
+      "step": 7463
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6027058379471204,
+      "learning_rate": 1.1596941189258542e-06,
+      "loss": 0.5957,
+      "step": 7464
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5799914326081077,
+      "learning_rate": 1.1586029879893018e-06,
+      "loss": 0.5961,
+      "step": 7465
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.272129728569525,
+      "learning_rate": 1.1575123033351514e-06,
+      "loss": 0.574,
+      "step": 7466
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.353854643372349,
+      "learning_rate": 1.1564220650901126e-06,
+      "loss": 0.548,
+      "step": 7467
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.284598981438445,
+      "learning_rate": 1.1553322733808474e-06,
+      "loss": 0.6057,
+      "step": 7468
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5638147489723737,
+      "learning_rate": 1.1542429283339669e-06,
+      "loss": 0.625,
+      "step": 7469
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.859248521334551,
+      "learning_rate": 1.153154030076024e-06,
+      "loss": 0.5896,
+      "step": 7470
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1507209034830352,
+      "learning_rate": 1.1520655787335272e-06,
+      "loss": 0.5909,
+      "step": 7471
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2804744847150293,
+      "learning_rate": 1.150977574432927e-06,
+      "loss": 0.6216,
+      "step": 7472
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2755713818694026,
+      "learning_rate": 1.1498900173006271e-06,
+      "loss": 0.5953,
+      "step": 7473
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4375823375991676,
+      "learning_rate": 1.1488029074629742e-06,
+      "loss": 0.5324,
+      "step": 7474
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6695360497471374,
+      "learning_rate": 1.1477162450462681e-06,
+      "loss": 0.5473,
+      "step": 7475
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.055784414516847,
+      "learning_rate": 1.1466300301767513e-06,
+      "loss": 0.6207,
+      "step": 7476
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3379616900640507,
+      "learning_rate": 1.1455442629806208e-06,
+      "loss": 0.57,
+      "step": 7477
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3050676485829817,
+      "learning_rate": 1.1444589435840136e-06,
+      "loss": 0.5941,
+      "step": 7478
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2327610274880056,
+      "learning_rate": 1.1433740721130227e-06,
+      "loss": 0.6705,
+      "step": 7479
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.318030102730509,
+      "learning_rate": 1.1422896486936819e-06,
+      "loss": 0.5924,
+      "step": 7480
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3743209801016847,
+      "learning_rate": 1.1412056734519788e-06,
+      "loss": 0.6173,
+      "step": 7481
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.34093180724465,
+      "learning_rate": 1.1401221465138468e-06,
+      "loss": 0.6462,
+      "step": 7482
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6932639467851116,
+      "learning_rate": 1.1390390680051649e-06,
+      "loss": 0.6229,
+      "step": 7483
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.346396013718988,
+      "learning_rate": 1.1379564380517648e-06,
+      "loss": 0.6471,
+      "step": 7484
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5191557971368876,
+      "learning_rate": 1.1368742567794199e-06,
+      "loss": 0.5747,
+      "step": 7485
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.187304261864126,
+      "learning_rate": 1.1357925243138585e-06,
+      "loss": 0.6561,
+      "step": 7486
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 11.808798126077034,
+      "learning_rate": 1.1347112407807499e-06,
+      "loss": 0.61,
+      "step": 7487
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5589240903441346,
+      "learning_rate": 1.1336304063057169e-06,
+      "loss": 0.5923,
+      "step": 7488
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2691294777604085,
+      "learning_rate": 1.1325500210143253e-06,
+      "loss": 0.6155,
+      "step": 7489
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.901817891190817,
+      "learning_rate": 1.1314700850320948e-06,
+      "loss": 0.6281,
+      "step": 7490
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.711266040481578,
+      "learning_rate": 1.1303905984844848e-06,
+      "loss": 0.6484,
+      "step": 7491
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2630100169064113,
+      "learning_rate": 1.1293115614969109e-06,
+      "loss": 0.5749,
+      "step": 7492
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.207926785870351,
+      "learning_rate": 1.1282329741947295e-06,
+      "loss": 0.584,
+      "step": 7493
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3024880371784513,
+      "learning_rate": 1.1271548367032487e-06,
+      "loss": 0.6285,
+      "step": 7494
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.259781922595136,
+      "learning_rate": 1.1260771491477252e-06,
+      "loss": 0.534,
+      "step": 7495
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3229570086330815,
+      "learning_rate": 1.1249999116533589e-06,
+      "loss": 0.5503,
+      "step": 7496
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.553935192730057,
+      "learning_rate": 1.1239231243453025e-06,
+      "loss": 0.664,
+      "step": 7497
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.92848847705716,
+      "learning_rate": 1.122846787348652e-06,
+      "loss": 0.5382,
+      "step": 7498
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.25087708292023,
+      "learning_rate": 1.1217709007884548e-06,
+      "loss": 0.6478,
+      "step": 7499
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.887135255747007,
+      "learning_rate": 1.1206954647897023e-06,
+      "loss": 0.6988,
+      "step": 7500
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.881071494956094,
+      "learning_rate": 1.1196204794773385e-06,
+      "loss": 0.6754,
+      "step": 7501
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.8124395050447877,
+      "learning_rate": 1.1185459449762486e-06,
+      "loss": 0.6218,
+      "step": 7502
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.5635427523252625,
+      "learning_rate": 1.1174718614112711e-06,
+      "loss": 0.6018,
+      "step": 7503
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1176840263880585,
+      "learning_rate": 1.1163982289071907e-06,
+      "loss": 0.5724,
+      "step": 7504
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.389675591616161,
+      "learning_rate": 1.1153250475887362e-06,
+      "loss": 0.658,
+      "step": 7505
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.2712660526620176,
+      "learning_rate": 1.1142523175805896e-06,
+      "loss": 0.6494,
+      "step": 7506
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3498472808911988,
+      "learning_rate": 1.113180039007375e-06,
+      "loss": 0.6334,
+      "step": 7507
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.183640363375537,
+      "learning_rate": 1.112108211993669e-06,
+      "loss": 0.6067,
+      "step": 7508
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.308029778683292,
+      "learning_rate": 1.1110368366639906e-06,
+      "loss": 0.6066,
+      "step": 7509
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.528977099429406,
+      "learning_rate": 1.109965913142812e-06,
+      "loss": 0.6271,
+      "step": 7510
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.7928225891739076,
+      "learning_rate": 1.1088954415545478e-06,
+      "loss": 0.5326,
+      "step": 7511
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4456570967777926,
+      "learning_rate": 1.107825422023564e-06,
+      "loss": 0.6375,
+      "step": 7512
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.0183023237081814,
+      "learning_rate": 1.1067558546741708e-06,
+      "loss": 0.6086,
+      "step": 7513
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.757754038229962,
+      "learning_rate": 1.1056867396306293e-06,
+      "loss": 0.5815,
+      "step": 7514
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5086536878121533,
+      "learning_rate": 1.1046180770171433e-06,
+      "loss": 0.5206,
+      "step": 7515
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.022738679492978,
+      "learning_rate": 1.1035498669578693e-06,
+      "loss": 0.5553,
+      "step": 7516
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.9716402092398586,
+      "learning_rate": 1.1024821095769089e-06,
+      "loss": 0.5598,
+      "step": 7517
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.7359639039165096,
+      "learning_rate": 1.1014148049983097e-06,
+      "loss": 0.6813,
+      "step": 7518
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3089260079854244,
+      "learning_rate": 1.1003479533460698e-06,
+      "loss": 0.5112,
+      "step": 7519
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.623336530654308,
+      "learning_rate": 1.099281554744131e-06,
+      "loss": 0.6542,
+      "step": 7520
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.247425868920514,
+      "learning_rate": 1.0982156093163864e-06,
+      "loss": 0.571,
+      "step": 7521
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.231031748647357,
+      "learning_rate": 1.0971501171866717e-06,
+      "loss": 0.5311,
+      "step": 7522
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.718419525290093,
+      "learning_rate": 1.0960850784787763e-06,
+      "loss": 0.5205,
+      "step": 7523
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.0286666822872843,
+      "learning_rate": 1.09502049331643e-06,
+      "loss": 0.611,
+      "step": 7524
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.9672246988207207,
+      "learning_rate": 1.0939563618233156e-06,
+      "loss": 0.5913,
+      "step": 7525
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6418538967198884,
+      "learning_rate": 1.0928926841230585e-06,
+      "loss": 0.59,
+      "step": 7526
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2332248750501242,
+      "learning_rate": 1.0918294603392371e-06,
+      "loss": 0.5947,
+      "step": 7527
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5309030658613185,
+      "learning_rate": 1.0907666905953696e-06,
+      "loss": 0.5941,
+      "step": 7528
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3540324184287456,
+      "learning_rate": 1.0897043750149277e-06,
+      "loss": 0.6044,
+      "step": 7529
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.770490477689187,
+      "learning_rate": 1.0886425137213297e-06,
+      "loss": 0.594,
+      "step": 7530
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6612418290795157,
+      "learning_rate": 1.087581106837936e-06,
+      "loss": 0.5736,
+      "step": 7531
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6742845066681014,
+      "learning_rate": 1.086520154488061e-06,
+      "loss": 0.6691,
+      "step": 7532
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.9951587604725614,
+      "learning_rate": 1.0854596567949605e-06,
+      "loss": 0.5694,
+      "step": 7533
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5789976556897116,
+      "learning_rate": 1.084399613881843e-06,
+      "loss": 0.6256,
+      "step": 7534
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.7119280648765347,
+      "learning_rate": 1.0833400258718579e-06,
+      "loss": 0.5858,
+      "step": 7535
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.597761699649087,
+      "learning_rate": 1.0822808928881078e-06,
+      "loss": 0.5949,
+      "step": 7536
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.7367475679821016,
+      "learning_rate": 1.0812222150536379e-06,
+      "loss": 0.4645,
+      "step": 7537
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.775373037805519,
+      "learning_rate": 1.0801639924914437e-06,
+      "loss": 0.6371,
+      "step": 7538
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.673800251379294,
+      "learning_rate": 1.0791062253244644e-06,
+      "loss": 0.562,
+      "step": 7539
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.733905978311876,
+      "learning_rate": 1.07804891367559e-06,
+      "loss": 0.6194,
+      "step": 7540
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.189114404251597,
+      "learning_rate": 1.0769920576676569e-06,
+      "loss": 0.5761,
+      "step": 7541
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.283883510957929,
+      "learning_rate": 1.0759356574234447e-06,
+      "loss": 0.6039,
+      "step": 7542
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.347107417686752,
+      "learning_rate": 1.0748797130656862e-06,
+      "loss": 0.6732,
+      "step": 7543
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.3757009779681755,
+      "learning_rate": 1.0738242247170549e-06,
+      "loss": 0.5663,
+      "step": 7544
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1237347595673874,
+      "learning_rate": 1.0727691925001765e-06,
+      "loss": 0.6165,
+      "step": 7545
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.0239588731460882,
+      "learning_rate": 1.0717146165376202e-06,
+      "loss": 0.6566,
+      "step": 7546
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.560705932085416,
+      "learning_rate": 1.0706604969519052e-06,
+      "loss": 0.6015,
+      "step": 7547
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.331047433053311,
+      "learning_rate": 1.069606833865494e-06,
+      "loss": 0.6184,
+      "step": 7548
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.347877312775048,
+      "learning_rate": 1.0685536274008002e-06,
+      "loss": 0.5979,
+      "step": 7549
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.246678913685663,
+      "learning_rate": 1.0675008776801804e-06,
+      "loss": 0.5749,
+      "step": 7550
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.0685585261266795,
+      "learning_rate": 1.066448584825942e-06,
+      "loss": 0.6043,
+      "step": 7551
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.204865595293493,
+      "learning_rate": 1.065396748960335e-06,
+      "loss": 0.6199,
+      "step": 7552
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0913674075484185,
+      "learning_rate": 1.06434537020556e-06,
+      "loss": 0.5869,
+      "step": 7553
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.7478603902570047,
+      "learning_rate": 1.0632944486837642e-06,
+      "loss": 0.5817,
+      "step": 7554
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.327839999993541,
+      "learning_rate": 1.0622439845170385e-06,
+      "loss": 0.6111,
+      "step": 7555
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.0045359616694594,
+      "learning_rate": 1.061193977827425e-06,
+      "loss": 0.609,
+      "step": 7556
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.062618698945312,
+      "learning_rate": 1.0601444287369073e-06,
+      "loss": 0.6086,
+      "step": 7557
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4008049526313604,
+      "learning_rate": 1.0590953373674229e-06,
+      "loss": 0.6595,
+      "step": 7558
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.1811579275109,
+      "learning_rate": 1.0580467038408487e-06,
+      "loss": 0.5707,
+      "step": 7559
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3155227686288713,
+      "learning_rate": 1.0569985282790145e-06,
+      "loss": 0.6299,
+      "step": 7560
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.5347215291469682,
+      "learning_rate": 1.0559508108036926e-06,
+      "loss": 0.5662,
+      "step": 7561
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.007580421779567,
+      "learning_rate": 1.0549035515366052e-06,
+      "loss": 0.5217,
+      "step": 7562
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.308048869247993,
+      "learning_rate": 1.0538567505994175e-06,
+      "loss": 0.6871,
+      "step": 7563
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.425142794535516,
+      "learning_rate": 1.052810408113746e-06,
+      "loss": 0.643,
+      "step": 7564
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.639780916922588,
+      "learning_rate": 1.051764524201152e-06,
+      "loss": 0.5547,
+      "step": 7565
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.474431995538235,
+      "learning_rate": 1.0507190989831412e-06,
+      "loss": 0.632,
+      "step": 7566
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.2149333641540214,
+      "learning_rate": 1.0496741325811705e-06,
+      "loss": 0.6219,
+      "step": 7567
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0743039032781865,
+      "learning_rate": 1.0486296251166383e-06,
+      "loss": 0.5175,
+      "step": 7568
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.633639411660315,
+      "learning_rate": 1.0475855767108956e-06,
+      "loss": 0.6497,
+      "step": 7569
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.457721991255091,
+      "learning_rate": 1.0465419874852338e-06,
+      "loss": 0.5951,
+      "step": 7570
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.239650791365524,
+      "learning_rate": 1.0454988575608976e-06,
+      "loss": 0.5779,
+      "step": 7571
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.0137305763833675,
+      "learning_rate": 1.0444561870590707e-06,
+      "loss": 0.6572,
+      "step": 7572
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.832898876677455,
+      "learning_rate": 1.0434139761008915e-06,
+      "loss": 0.592,
+      "step": 7573
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.8742909520681805,
+      "learning_rate": 1.042372224807438e-06,
+      "loss": 0.6899,
+      "step": 7574
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.538021201102594,
+      "learning_rate": 1.0413309332997385e-06,
+      "loss": 0.533,
+      "step": 7575
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.9699203259048996,
+      "learning_rate": 1.0402901016987694e-06,
+      "loss": 0.6227,
+      "step": 7576
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0629324280519654,
+      "learning_rate": 1.0392497301254489e-06,
+      "loss": 0.5473,
+      "step": 7577
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.150602467486783,
+      "learning_rate": 1.0382098187006463e-06,
+      "loss": 0.6386,
+      "step": 7578
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.012171485780843,
+      "learning_rate": 1.0371703675451732e-06,
+      "loss": 0.5258,
+      "step": 7579
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.144392113247383,
+      "learning_rate": 1.0361313767797932e-06,
+      "loss": 0.5706,
+      "step": 7580
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4348220328637216,
+      "learning_rate": 1.0350928465252103e-06,
+      "loss": 0.6039,
+      "step": 7581
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.6562323332446294,
+      "learning_rate": 1.0340547769020798e-06,
+      "loss": 0.5633,
+      "step": 7582
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9865343111730549,
+      "learning_rate": 1.0330171680309996e-06,
+      "loss": 0.5352,
+      "step": 7583
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.3099068356431776,
+      "learning_rate": 1.0319800200325193e-06,
+      "loss": 0.5597,
+      "step": 7584
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.519735515375185,
+      "learning_rate": 1.0309433330271285e-06,
+      "loss": 0.5921,
+      "step": 7585
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.187601261956032,
+      "learning_rate": 1.029907107135269e-06,
+      "loss": 0.5567,
+      "step": 7586
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.348573049735444,
+      "learning_rate": 1.0288713424773238e-06,
+      "loss": 0.6187,
+      "step": 7587
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.701801842597667,
+      "learning_rate": 1.027836039173627e-06,
+      "loss": 0.5421,
+      "step": 7588
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.717130793307138,
+      "learning_rate": 1.026801197344458e-06,
+      "loss": 0.6378,
+      "step": 7589
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3303124612250543,
+      "learning_rate": 1.0257668171100393e-06,
+      "loss": 0.6418,
+      "step": 7590
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7662913464027623,
+      "learning_rate": 1.0247328985905446e-06,
+      "loss": 0.5441,
+      "step": 7591
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.544677485406088,
+      "learning_rate": 1.0236994419060892e-06,
+      "loss": 0.568,
+      "step": 7592
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.899566327372865,
+      "learning_rate": 1.0226664471767401e-06,
+      "loss": 0.6502,
+      "step": 7593
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.506201535469169,
+      "learning_rate": 1.021633914522504e-06,
+      "loss": 0.6079,
+      "step": 7594
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.518907467485238,
+      "learning_rate": 1.0206018440633408e-06,
+      "loss": 0.5106,
+      "step": 7595
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3943093369268227,
+      "learning_rate": 1.0195702359191507e-06,
+      "loss": 0.579,
+      "step": 7596
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4145954273830927,
+      "learning_rate": 1.0185390902097857e-06,
+      "loss": 0.6422,
+      "step": 7597
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.921201826914954,
+      "learning_rate": 1.017508407055039e-06,
+      "loss": 0.6023,
+      "step": 7598
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.937076600981269,
+      "learning_rate": 1.0164781865746542e-06,
+      "loss": 0.5655,
+      "step": 7599
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.760440801599633,
+      "learning_rate": 1.0154484288883177e-06,
+      "loss": 0.6897,
+      "step": 7600
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4696937629349898,
+      "learning_rate": 1.014419134115664e-06,
+      "loss": 0.6095,
+      "step": 7601
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.811722355164242,
+      "learning_rate": 1.0133903023762758e-06,
+      "loss": 0.6984,
+      "step": 7602
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.115947584954783,
+      "learning_rate": 1.0123619337896767e-06,
+      "loss": 0.5696,
+      "step": 7603
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.564549686867802,
+      "learning_rate": 1.0113340284753425e-06,
+      "loss": 0.655,
+      "step": 7604
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 12.618329274026946,
+      "learning_rate": 1.0103065865526895e-06,
+      "loss": 0.5778,
+      "step": 7605
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.8969821170985206,
+      "learning_rate": 1.0092796081410856e-06,
+      "loss": 0.5542,
+      "step": 7606
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.5199683999249087,
+      "learning_rate": 1.0082530933598388e-06,
+      "loss": 0.6197,
+      "step": 7607
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.435215141719612,
+      "learning_rate": 1.0072270423282104e-06,
+      "loss": 0.5886,
+      "step": 7608
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1869076971904104,
+      "learning_rate": 1.0062014551654015e-06,
+      "loss": 0.5902,
+      "step": 7609
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.5157709264478054,
+      "learning_rate": 1.0051763319905622e-06,
+      "loss": 0.5485,
+      "step": 7610
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3517546543055747,
+      "learning_rate": 1.0041516729227902e-06,
+      "loss": 0.5734,
+      "step": 7611
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.447965128681716,
+      "learning_rate": 1.0031274780811245e-06,
+      "loss": 0.6666,
+      "step": 7612
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.976761782472005,
+      "learning_rate": 1.0021037475845557e-06,
+      "loss": 0.5641,
+      "step": 7613
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.698568711787718,
+      "learning_rate": 1.0010804815520159e-06,
+      "loss": 0.6378,
+      "step": 7614
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.3363776215524092,
+      "learning_rate": 1.0000576801023876e-06,
+      "loss": 0.6686,
+      "step": 7615
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.7415611338554813,
+      "learning_rate": 9.990353433544935e-07,
+      "loss": 0.6204,
+      "step": 7616
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.136268007867056,
+      "learning_rate": 9.980134714271088e-07,
+      "loss": 0.7083,
+      "step": 7617
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.186125465023069,
+      "learning_rate": 9.969920644389498e-07,
+      "loss": 0.583,
+      "step": 7618
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.6754021896396454,
+      "learning_rate": 9.959711225086822e-07,
+      "loss": 0.6414,
+      "step": 7619
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.3008931024230703,
+      "learning_rate": 9.949506457549135e-07,
+      "loss": 0.5452,
+      "step": 7620
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.6881102400726142,
+      "learning_rate": 9.93930634296203e-07,
+      "loss": 0.5804,
+      "step": 7621
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.510590511858801,
+      "learning_rate": 9.929110882510496e-07,
+      "loss": 0.5921,
+      "step": 7622
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.218316814824091,
+      "learning_rate": 9.91892007737903e-07,
+      "loss": 0.6482,
+      "step": 7623
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3200140091226107,
+      "learning_rate": 9.908733928751574e-07,
+      "loss": 0.5587,
+      "step": 7624
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7818896360301317,
+      "learning_rate": 9.89855243781151e-07,
+      "loss": 0.6569,
+      "step": 7625
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.972728123297002,
+      "learning_rate": 9.888375605741713e-07,
+      "loss": 0.587,
+      "step": 7626
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.467181443995698,
+      "learning_rate": 9.87820343372447e-07,
+      "loss": 0.6413,
+      "step": 7627
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.771873420545783,
+      "learning_rate": 9.868035922941594e-07,
+      "loss": 0.6314,
+      "step": 7628
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.073309761721186,
+      "learning_rate": 9.857873074574276e-07,
+      "loss": 0.6316,
+      "step": 7629
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3110538134158447,
+      "learning_rate": 9.847714889803233e-07,
+      "loss": 0.6234,
+      "step": 7630
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3475214347720086,
+      "learning_rate": 9.8375613698086e-07,
+      "loss": 0.5873,
+      "step": 7631
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.338962974057601,
+      "learning_rate": 9.827412515770003e-07,
+      "loss": 0.6285,
+      "step": 7632
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4511140443396053,
+      "learning_rate": 9.817268328866474e-07,
+      "loss": 0.5206,
+      "step": 7633
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.6141500186568076,
+      "learning_rate": 9.807128810276573e-07,
+      "loss": 0.6173,
+      "step": 7634
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.6588577517347343,
+      "learning_rate": 9.796993961178247e-07,
+      "loss": 0.5779,
+      "step": 7635
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2544484813883354,
+      "learning_rate": 9.786863782748946e-07,
+      "loss": 0.6226,
+      "step": 7636
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7154960137258373,
+      "learning_rate": 9.776738276165576e-07,
+      "loss": 0.5301,
+      "step": 7637
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.8331620353922093,
+      "learning_rate": 9.76661744260447e-07,
+      "loss": 0.6592,
+      "step": 7638
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.8596209122623604,
+      "learning_rate": 9.75650128324146e-07,
+      "loss": 0.6592,
+      "step": 7639
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9617449759052116,
+      "learning_rate": 9.746389799251783e-07,
+      "loss": 0.5056,
+      "step": 7640
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.6375463130140555,
+      "learning_rate": 9.736282991810191e-07,
+      "loss": 0.649,
+      "step": 7641
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.170328479172141,
+      "learning_rate": 9.72618086209084e-07,
+      "loss": 0.6085,
+      "step": 7642
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0826307503241073,
+      "learning_rate": 9.71608341126739e-07,
+      "loss": 0.5132,
+      "step": 7643
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.426019291056326,
+      "learning_rate": 9.705990640512909e-07,
+      "loss": 0.5804,
+      "step": 7644
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.8657766408804495,
+      "learning_rate": 9.695902550999953e-07,
+      "loss": 0.6634,
+      "step": 7645
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1192272871905335,
+      "learning_rate": 9.685819143900544e-07,
+      "loss": 0.575,
+      "step": 7646
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.9730510980964873,
+      "learning_rate": 9.675740420386132e-07,
+      "loss": 0.6602,
+      "step": 7647
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.8204604472554924,
+      "learning_rate": 9.66566638162762e-07,
+      "loss": 0.6495,
+      "step": 7648
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.539741388640035,
+      "learning_rate": 9.655597028795394e-07,
+      "loss": 0.6173,
+      "step": 7649
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5656949751155484,
+      "learning_rate": 9.64553236305929e-07,
+      "loss": 0.6095,
+      "step": 7650
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.450376476628125,
+      "learning_rate": 9.635472385588573e-07,
+      "loss": 0.5739,
+      "step": 7651
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6482970996607156,
+      "learning_rate": 9.625417097552003e-07,
+      "loss": 0.6081,
+      "step": 7652
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6667414141134174,
+      "learning_rate": 9.615366500117757e-07,
+      "loss": 0.6578,
+      "step": 7653
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.74349389755696,
+      "learning_rate": 9.6053205944535e-07,
+      "loss": 0.5548,
+      "step": 7654
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.010896829422771,
+      "learning_rate": 9.595279381726308e-07,
+      "loss": 0.5475,
+      "step": 7655
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4405802174170725,
+      "learning_rate": 9.58524286310278e-07,
+      "loss": 0.6354,
+      "step": 7656
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.919810726968023,
+      "learning_rate": 9.575211039748893e-07,
+      "loss": 0.5448,
+      "step": 7657
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.363605925097881,
+      "learning_rate": 9.565183912830134e-07,
+      "loss": 0.6191,
+      "step": 7658
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4814246714387385,
+      "learning_rate": 9.555161483511434e-07,
+      "loss": 0.6056,
+      "step": 7659
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3442144571403394,
+      "learning_rate": 9.545143752957143e-07,
+      "loss": 0.5799,
+      "step": 7660
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9844963618684487,
+      "learning_rate": 9.535130722331121e-07,
+      "loss": 0.5501,
+      "step": 7661
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.841148309195625,
+      "learning_rate": 9.525122392796632e-07,
+      "loss": 0.5798,
+      "step": 7662
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6946034136889114,
+      "learning_rate": 9.515118765516429e-07,
+      "loss": 0.624,
+      "step": 7663
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.614312745110826,
+      "learning_rate": 9.505119841652688e-07,
+      "loss": 0.6012,
+      "step": 7664
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2341716999178733,
+      "learning_rate": 9.495125622367079e-07,
+      "loss": 0.699,
+      "step": 7665
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.454492464172891,
+      "learning_rate": 9.485136108820675e-07,
+      "loss": 0.5752,
+      "step": 7666
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7299044786206266,
+      "learning_rate": 9.475151302174052e-07,
+      "loss": 0.6267,
+      "step": 7667
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.46712121179473,
+      "learning_rate": 9.465171203587192e-07,
+      "loss": 0.69,
+      "step": 7668
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.8581757438763984,
+      "learning_rate": 9.455195814219581e-07,
+      "loss": 0.6329,
+      "step": 7669
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7749337082880636,
+      "learning_rate": 9.445225135230102e-07,
+      "loss": 0.5863,
+      "step": 7670
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.9200575440191554,
+      "learning_rate": 9.435259167777139e-07,
+      "loss": 0.6594,
+      "step": 7671
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.1961551877877343,
+      "learning_rate": 9.425297913018517e-07,
+      "loss": 0.6262,
+      "step": 7672
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.621987087951769,
+      "learning_rate": 9.41534137211148e-07,
+      "loss": 0.6806,
+      "step": 7673
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.37550460624206,
+      "learning_rate": 9.405389546212779e-07,
+      "loss": 0.4877,
+      "step": 7674
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1856374832479073,
+      "learning_rate": 9.395442436478558e-07,
+      "loss": 0.6366,
+      "step": 7675
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3120929072875986,
+      "learning_rate": 9.385500044064472e-07,
+      "loss": 0.5728,
+      "step": 7676
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 11.267589648078419,
+      "learning_rate": 9.375562370125574e-07,
+      "loss": 0.6151,
+      "step": 7677
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.3911562826517248,
+      "learning_rate": 9.365629415816418e-07,
+      "loss": 0.5745,
+      "step": 7678
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0931728619715138,
+      "learning_rate": 9.355701182290961e-07,
+      "loss": 0.5957,
+      "step": 7679
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4056655342238416,
+      "learning_rate": 9.345777670702649e-07,
+      "loss": 0.6034,
+      "step": 7680
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.31101973582715,
+      "learning_rate": 9.335858882204385e-07,
+      "loss": 0.5717,
+      "step": 7681
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.2633888627628336,
+      "learning_rate": 9.325944817948485e-07,
+      "loss": 0.6403,
+      "step": 7682
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.507966888011093,
+      "learning_rate": 9.316035479086727e-07,
+      "loss": 0.6023,
+      "step": 7683
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6707445301357104,
+      "learning_rate": 9.306130866770364e-07,
+      "loss": 0.694,
+      "step": 7684
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3698584587195777,
+      "learning_rate": 9.296230982150095e-07,
+      "loss": 0.499,
+      "step": 7685
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.646352948765854,
+      "learning_rate": 9.28633582637603e-07,
+      "loss": 0.5998,
+      "step": 7686
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3308088558522724,
+      "learning_rate": 9.276445400597795e-07,
+      "loss": 0.6442,
+      "step": 7687
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7184765380374314,
+      "learning_rate": 9.266559705964401e-07,
+      "loss": 0.7093,
+      "step": 7688
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.623044957668708,
+      "learning_rate": 9.256678743624364e-07,
+      "loss": 0.6417,
+      "step": 7689
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.079400937660165,
+      "learning_rate": 9.246802514725601e-07,
+      "loss": 0.6116,
+      "step": 7690
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2731708710443503,
+      "learning_rate": 9.236931020415529e-07,
+      "loss": 0.5756,
+      "step": 7691
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.3136227736096924,
+      "learning_rate": 9.227064261840962e-07,
+      "loss": 0.6093,
+      "step": 7692
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1475285005337974,
+      "learning_rate": 9.21720224014821e-07,
+      "loss": 0.6814,
+      "step": 7693
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.9968897146706133,
+      "learning_rate": 9.207344956483022e-07,
+      "loss": 0.6033,
+      "step": 7694
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6350679079149866,
+      "learning_rate": 9.197492411990571e-07,
+      "loss": 0.626,
+      "step": 7695
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.72376089543527,
+      "learning_rate": 9.187644607815499e-07,
+      "loss": 0.6423,
+      "step": 7696
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.8530684492710217,
+      "learning_rate": 9.1778015451019e-07,
+      "loss": 0.6458,
+      "step": 7697
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0539321212137107,
+      "learning_rate": 9.16796322499332e-07,
+      "loss": 0.6108,
+      "step": 7698
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7055060163420066,
+      "learning_rate": 9.15812964863273e-07,
+      "loss": 0.5557,
+      "step": 7699
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.334028049190322,
+      "learning_rate": 9.148300817162587e-07,
+      "loss": 0.5686,
+      "step": 7700
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8483172248492421,
+      "learning_rate": 9.138476731724749e-07,
+      "loss": 0.5497,
+      "step": 7701
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.586655116930661,
+      "learning_rate": 9.128657393460583e-07,
+      "loss": 0.5979,
+      "step": 7702
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.314047969656968,
+      "learning_rate": 9.118842803510841e-07,
+      "loss": 0.6292,
+      "step": 7703
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.410214810373217,
+      "learning_rate": 9.10903296301578e-07,
+      "loss": 0.6381,
+      "step": 7704
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.104473643611443,
+      "learning_rate": 9.099227873115047e-07,
+      "loss": 0.6241,
+      "step": 7705
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.4899505415847676,
+      "learning_rate": 9.089427534947792e-07,
+      "loss": 0.6409,
+      "step": 7706
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1792155269408617,
+      "learning_rate": 9.0796319496526e-07,
+      "loss": 0.6243,
+      "step": 7707
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.648707594215365,
+      "learning_rate": 9.069841118367462e-07,
+      "loss": 0.6185,
+      "step": 7708
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.801759382084992,
+      "learning_rate": 9.060055042229881e-07,
+      "loss": 0.6844,
+      "step": 7709
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.8450393356650685,
+      "learning_rate": 9.050273722376746e-07,
+      "loss": 0.535,
+      "step": 7710
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.237580480646928,
+      "learning_rate": 9.04049715994445e-07,
+      "loss": 0.6118,
+      "step": 7711
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.029884410505219,
+      "learning_rate": 9.030725356068781e-07,
+      "loss": 0.642,
+      "step": 7712
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5945698031312086,
+      "learning_rate": 9.020958311885019e-07,
+      "loss": 0.6349,
+      "step": 7713
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.455581536260283,
+      "learning_rate": 9.011196028527853e-07,
+      "loss": 0.6013,
+      "step": 7714
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9942779258867724,
+      "learning_rate": 9.001438507131444e-07,
+      "loss": 0.527,
+      "step": 7715
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.700359868586655,
+      "learning_rate": 8.991685748829404e-07,
+      "loss": 0.6477,
+      "step": 7716
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.104403745631631,
+      "learning_rate": 8.981937754754777e-07,
+      "loss": 0.6163,
+      "step": 7717
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5972794158820482,
+      "learning_rate": 8.972194526040034e-07,
+      "loss": 0.5271,
+      "step": 7718
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4719633270189583,
+      "learning_rate": 8.962456063817132e-07,
+      "loss": 0.6623,
+      "step": 7719
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9667484827218678,
+      "learning_rate": 8.95272236921747e-07,
+      "loss": 0.5675,
+      "step": 7720
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.425564331369107,
+      "learning_rate": 8.942993443371856e-07,
+      "loss": 0.609,
+      "step": 7721
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.947734739772372,
+      "learning_rate": 8.93326928741059e-07,
+      "loss": 0.6371,
+      "step": 7722
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.9080706900479045,
+      "learning_rate": 8.923549902463374e-07,
+      "loss": 0.6628,
+      "step": 7723
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.8252938892041635,
+      "learning_rate": 8.913835289659406e-07,
+      "loss": 0.5489,
+      "step": 7724
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5458974669966805,
+      "learning_rate": 8.904125450127272e-07,
+      "loss": 0.6616,
+      "step": 7725
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.184591112903663,
+      "learning_rate": 8.894420384995056e-07,
+      "loss": 0.6075,
+      "step": 7726
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.9011418508765647,
+      "learning_rate": 8.884720095390248e-07,
+      "loss": 0.6888,
+      "step": 7727
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.454534274847728,
+      "learning_rate": 8.875024582439801e-07,
+      "loss": 0.6525,
+      "step": 7728
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3930270904791335,
+      "learning_rate": 8.865333847270135e-07,
+      "loss": 0.6685,
+      "step": 7729
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5600293308303783,
+      "learning_rate": 8.855647891007075e-07,
+      "loss": 0.5978,
+      "step": 7730
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.437056971277586,
+      "learning_rate": 8.845966714775894e-07,
+      "loss": 0.6665,
+      "step": 7731
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 21.97578049995487,
+      "learning_rate": 8.836290319701335e-07,
+      "loss": 0.5903,
+      "step": 7732
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.878994513782418,
+      "learning_rate": 8.826618706907585e-07,
+      "loss": 0.6547,
+      "step": 7733
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6089834334746893,
+      "learning_rate": 8.816951877518243e-07,
+      "loss": 0.6462,
+      "step": 7734
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.5891885718164533,
+      "learning_rate": 8.807289832656396e-07,
+      "loss": 0.6704,
+      "step": 7735
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.53450443321136,
+      "learning_rate": 8.797632573444526e-07,
+      "loss": 0.5234,
+      "step": 7736
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4367976647015843,
+      "learning_rate": 8.787980101004612e-07,
+      "loss": 0.5515,
+      "step": 7737
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.393038737753617,
+      "learning_rate": 8.77833241645803e-07,
+      "loss": 0.6213,
+      "step": 7738
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.942620580800737,
+      "learning_rate": 8.768689520925638e-07,
+      "loss": 0.5855,
+      "step": 7739
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2895413377993816,
+      "learning_rate": 8.759051415527697e-07,
+      "loss": 0.6244,
+      "step": 7740
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.1158010305913275,
+      "learning_rate": 8.749418101383944e-07,
+      "loss": 0.5865,
+      "step": 7741
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6974129522038877,
+      "learning_rate": 8.739789579613572e-07,
+      "loss": 0.5408,
+      "step": 7742
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.895780450687752,
+      "learning_rate": 8.73016585133517e-07,
+      "loss": 0.6004,
+      "step": 7743
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0182236513579874,
+      "learning_rate": 8.720546917666789e-07,
+      "loss": 0.6118,
+      "step": 7744
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.370869512524687,
+      "learning_rate": 8.710932779725939e-07,
+      "loss": 0.6114,
+      "step": 7745
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3023012937961846,
+      "learning_rate": 8.701323438629577e-07,
+      "loss": 0.5891,
+      "step": 7746
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7894697251608,
+      "learning_rate": 8.691718895494067e-07,
+      "loss": 0.5835,
+      "step": 7747
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2253869804344077,
+      "learning_rate": 8.682119151435258e-07,
+      "loss": 0.657,
+      "step": 7748
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1968761019330536,
+      "learning_rate": 8.672524207568389e-07,
+      "loss": 0.5731,
+      "step": 7749
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0451951033529774,
+      "learning_rate": 8.6629340650082e-07,
+      "loss": 0.6249,
+      "step": 7750
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.290396646074176,
+      "learning_rate": 8.653348724868843e-07,
+      "loss": 0.6095,
+      "step": 7751
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4998002085053828,
+      "learning_rate": 8.643768188263918e-07,
+      "loss": 0.5901,
+      "step": 7752
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.927706275807229,
+      "learning_rate": 8.63419245630644e-07,
+      "loss": 0.6667,
+      "step": 7753
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.514050334470586,
+      "learning_rate": 8.624621530108901e-07,
+      "loss": 0.5498,
+      "step": 7754
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.00269176738163,
+      "learning_rate": 8.615055410783246e-07,
+      "loss": 0.5119,
+      "step": 7755
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2181227482346895,
+      "learning_rate": 8.605494099440808e-07,
+      "loss": 0.6086,
+      "step": 7756
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9630978359237016,
+      "learning_rate": 8.595937597192422e-07,
+      "loss": 0.5444,
+      "step": 7757
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.040120236257973,
+      "learning_rate": 8.586385905148304e-07,
+      "loss": 0.5648,
+      "step": 7758
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7334403291410077,
+      "learning_rate": 8.576839024418165e-07,
+      "loss": 0.6385,
+      "step": 7759
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2473246156394096,
+      "learning_rate": 8.567296956111121e-07,
+      "loss": 0.5395,
+      "step": 7760
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5453764969411057,
+      "learning_rate": 8.557759701335755e-07,
+      "loss": 0.6596,
+      "step": 7761
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.755171473082308,
+      "learning_rate": 8.54822726120006e-07,
+      "loss": 0.5935,
+      "step": 7762
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3621820404124114,
+      "learning_rate": 8.538699636811493e-07,
+      "loss": 0.6247,
+      "step": 7763
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5573182046892384,
+      "learning_rate": 8.529176829276964e-07,
+      "loss": 0.5769,
+      "step": 7764
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.452128633867755,
+      "learning_rate": 8.519658839702787e-07,
+      "loss": 0.5634,
+      "step": 7765
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.1439091682655964,
+      "learning_rate": 8.51014566919473e-07,
+      "loss": 0.6145,
+      "step": 7766
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.479175976072859,
+      "learning_rate": 8.500637318858018e-07,
+      "loss": 0.6538,
+      "step": 7767
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.430856695338035,
+      "learning_rate": 8.491133789797307e-07,
+      "loss": 0.6279,
+      "step": 7768
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.770347480171204,
+      "learning_rate": 8.481635083116668e-07,
+      "loss": 0.6057,
+      "step": 7769
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.069965846489253,
+      "learning_rate": 8.472141199919664e-07,
+      "loss": 0.5681,
+      "step": 7770
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.9992990174535383,
+      "learning_rate": 8.462652141309242e-07,
+      "loss": 0.6433,
+      "step": 7771
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.79447885278482,
+      "learning_rate": 8.453167908387827e-07,
+      "loss": 0.6942,
+      "step": 7772
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3694158265158682,
+      "learning_rate": 8.443688502257253e-07,
+      "loss": 0.5695,
+      "step": 7773
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.6521967702602134,
+      "learning_rate": 8.434213924018836e-07,
+      "loss": 0.6175,
+      "step": 7774
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.450142386712109,
+      "learning_rate": 8.424744174773281e-07,
+      "loss": 0.5917,
+      "step": 7775
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.60990893147203,
+      "learning_rate": 8.415279255620762e-07,
+      "loss": 0.7291,
+      "step": 7776
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3162030741562702,
+      "learning_rate": 8.405819167660906e-07,
+      "loss": 0.5496,
+      "step": 7777
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.454400577973904,
+      "learning_rate": 8.396363911992739e-07,
+      "loss": 0.6948,
+      "step": 7778
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.833872698534407,
+      "learning_rate": 8.386913489714737e-07,
+      "loss": 0.5477,
+      "step": 7779
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.445942443952145,
+      "learning_rate": 8.377467901924835e-07,
+      "loss": 0.5796,
+      "step": 7780
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.537918156805052,
+      "learning_rate": 8.368027149720404e-07,
+      "loss": 0.6614,
+      "step": 7781
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.100483149802904,
+      "learning_rate": 8.358591234198221e-07,
+      "loss": 0.6177,
+      "step": 7782
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7198326768529277,
+      "learning_rate": 8.34916015645455e-07,
+      "loss": 0.625,
+      "step": 7783
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.8763253745346042,
+      "learning_rate": 8.339733917585041e-07,
+      "loss": 0.504,
+      "step": 7784
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.0339142550402896,
+      "learning_rate": 8.330312518684813e-07,
+      "loss": 0.6544,
+      "step": 7785
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.372358625907128,
+      "learning_rate": 8.320895960848435e-07,
+      "loss": 0.662,
+      "step": 7786
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 11.12693531528549,
+      "learning_rate": 8.311484245169888e-07,
+      "loss": 0.5841,
+      "step": 7787
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.0258617844542615,
+      "learning_rate": 8.302077372742573e-07,
+      "loss": 0.5996,
+      "step": 7788
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.149848595276636,
+      "learning_rate": 8.292675344659374e-07,
+      "loss": 0.5624,
+      "step": 7789
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.921586434700963,
+      "learning_rate": 8.283278162012604e-07,
+      "loss": 0.6565,
+      "step": 7790
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.223681029796389,
+      "learning_rate": 8.273885825893984e-07,
+      "loss": 0.572,
+      "step": 7791
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 12.307302243450632,
+      "learning_rate": 8.264498337394683e-07,
+      "loss": 0.7158,
+      "step": 7792
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.185259212276455,
+      "learning_rate": 8.255115697605315e-07,
+      "loss": 0.6205,
+      "step": 7793
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3742771690682867,
+      "learning_rate": 8.245737907615948e-07,
+      "loss": 0.571,
+      "step": 7794
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0306399507949875,
+      "learning_rate": 8.236364968516036e-07,
+      "loss": 0.574,
+      "step": 7795
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.95185086622032,
+      "learning_rate": 8.226996881394533e-07,
+      "loss": 0.5969,
+      "step": 7796
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.83828313253317,
+      "learning_rate": 8.217633647339762e-07,
+      "loss": 0.6308,
+      "step": 7797
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.740058288843735,
+      "learning_rate": 8.208275267439536e-07,
+      "loss": 0.6409,
+      "step": 7798
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.205986939831031,
+      "learning_rate": 8.19892174278109e-07,
+      "loss": 0.638,
+      "step": 7799
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.133784418145159,
+      "learning_rate": 8.189573074451084e-07,
+      "loss": 0.5896,
+      "step": 7800
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.651287297043312,
+      "learning_rate": 8.180229263535605e-07,
+      "loss": 0.6102,
+      "step": 7801
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8083107733624537,
+      "learning_rate": 8.1708903111202e-07,
+      "loss": 0.5911,
+      "step": 7802
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.697503873505146,
+      "learning_rate": 8.161556218289857e-07,
+      "loss": 0.5624,
+      "step": 7803
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3094841871089065,
+      "learning_rate": 8.15222698612897e-07,
+      "loss": 0.5943,
+      "step": 7804
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.791054703314953,
+      "learning_rate": 8.142902615721371e-07,
+      "loss": 0.5893,
+      "step": 7805
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2449207941075198,
+      "learning_rate": 8.133583108150345e-07,
+      "loss": 0.5883,
+      "step": 7806
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.264533076641818,
+      "learning_rate": 8.124268464498625e-07,
+      "loss": 0.5802,
+      "step": 7807
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.397265245054088,
+      "learning_rate": 8.114958685848334e-07,
+      "loss": 0.7004,
+      "step": 7808
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.687783198325523,
+      "learning_rate": 8.105653773281074e-07,
+      "loss": 0.633,
+      "step": 7809
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.572173865987845,
+      "learning_rate": 8.096353727877843e-07,
+      "loss": 0.6158,
+      "step": 7810
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.084933298498584,
+      "learning_rate": 8.087058550719107e-07,
+      "loss": 0.6373,
+      "step": 7811
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.117019095691375,
+      "learning_rate": 8.077768242884759e-07,
+      "loss": 0.5337,
+      "step": 7812
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.526067117482907,
+      "learning_rate": 8.068482805454115e-07,
+      "loss": 0.5935,
+      "step": 7813
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.770744913286671,
+      "learning_rate": 8.059202239505915e-07,
+      "loss": 0.6824,
+      "step": 7814
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3876150794603417,
+      "learning_rate": 8.049926546118359e-07,
+      "loss": 0.6197,
+      "step": 7815
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.301040153613851,
+      "learning_rate": 8.040655726369079e-07,
+      "loss": 0.6231,
+      "step": 7816
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1640120127578,
+      "learning_rate": 8.031389781335119e-07,
+      "loss": 0.5915,
+      "step": 7817
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0216851076508051,
+      "learning_rate": 8.022128712092986e-07,
+      "loss": 0.5629,
+      "step": 7818
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3407168223787758,
+      "learning_rate": 8.012872519718578e-07,
+      "loss": 0.5779,
+      "step": 7819
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.240628782070062,
+      "learning_rate": 8.003621205287271e-07,
+      "loss": 0.6247,
+      "step": 7820
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5968226751951202,
+      "learning_rate": 7.994374769873864e-07,
+      "loss": 0.5631,
+      "step": 7821
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.9225493166351844,
+      "learning_rate": 7.98513321455257e-07,
+      "loss": 0.5619,
+      "step": 7822
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5510204032518633,
+      "learning_rate": 7.975896540397038e-07,
+      "loss": 0.5647,
+      "step": 7823
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.2618088874821436,
+      "learning_rate": 7.966664748480362e-07,
+      "loss": 0.681,
+      "step": 7824
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.432627062148813,
+      "learning_rate": 7.957437839875088e-07,
+      "loss": 0.619,
+      "step": 7825
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.881899570004556,
+      "learning_rate": 7.948215815653149e-07,
+      "loss": 0.5353,
+      "step": 7826
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.211030969729491,
+      "learning_rate": 7.938998676885922e-07,
+      "loss": 0.5992,
+      "step": 7827
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7189882982651374,
+      "learning_rate": 7.929786424644248e-07,
+      "loss": 0.5802,
+      "step": 7828
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.6023452660932165,
+      "learning_rate": 7.920579059998384e-07,
+      "loss": 0.575,
+      "step": 7829
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.417297217989513,
+      "learning_rate": 7.911376584017993e-07,
+      "loss": 0.5887,
+      "step": 7830
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.636618821203334,
+      "learning_rate": 7.90217899777222e-07,
+      "loss": 0.5957,
+      "step": 7831
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.783745505915478,
+      "learning_rate": 7.89298630232958e-07,
+      "loss": 0.6066,
+      "step": 7832
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.211994261833092,
+      "learning_rate": 7.883798498758077e-07,
+      "loss": 0.5615,
+      "step": 7833
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.316992802552742,
+      "learning_rate": 7.874615588125128e-07,
+      "loss": 0.5865,
+      "step": 7834
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.540384973156056,
+      "learning_rate": 7.865437571497569e-07,
+      "loss": 0.5444,
+      "step": 7835
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2642580877781904,
+      "learning_rate": 7.856264449941664e-07,
+      "loss": 0.5456,
+      "step": 7836
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5525408933985214,
+      "learning_rate": 7.847096224523132e-07,
+      "loss": 0.607,
+      "step": 7837
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3245711806217613,
+      "learning_rate": 7.837932896307116e-07,
+      "loss": 0.5573,
+      "step": 7838
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.05579417550974,
+      "learning_rate": 7.82877446635818e-07,
+      "loss": 0.5761,
+      "step": 7839
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5025364694067136,
+      "learning_rate": 7.819620935740313e-07,
+      "loss": 0.5728,
+      "step": 7840
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.541642302021767,
+      "learning_rate": 7.810472305516947e-07,
+      "loss": 0.6021,
+      "step": 7841
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6527868881900747,
+      "learning_rate": 7.801328576750971e-07,
+      "loss": 0.5955,
+      "step": 7842
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.035882397492675,
+      "learning_rate": 7.792189750504642e-07,
+      "loss": 0.559,
+      "step": 7843
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.4225984060139196,
+      "learning_rate": 7.783055827839709e-07,
+      "loss": 0.594,
+      "step": 7844
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.435724626217545,
+      "learning_rate": 7.7739268098173e-07,
+      "loss": 0.593,
+      "step": 7845
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4904962113373412,
+      "learning_rate": 7.764802697498009e-07,
+      "loss": 0.615,
+      "step": 7846
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.464037403138296,
+      "learning_rate": 7.755683491941867e-07,
+      "loss": 0.7109,
+      "step": 7847
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.9748316199128824,
+      "learning_rate": 7.746569194208298e-07,
+      "loss": 0.6551,
+      "step": 7848
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2358191510721923,
+      "learning_rate": 7.737459805356168e-07,
+      "loss": 0.5977,
+      "step": 7849
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.419490109419338,
+      "learning_rate": 7.728355326443792e-07,
+      "loss": 0.5004,
+      "step": 7850
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.911759699027262,
+      "learning_rate": 7.719255758528904e-07,
+      "loss": 0.604,
+      "step": 7851
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.859440106403186,
+      "learning_rate": 7.710161102668667e-07,
+      "loss": 0.5881,
+      "step": 7852
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.133615283383733,
+      "learning_rate": 7.701071359919654e-07,
+      "loss": 0.5945,
+      "step": 7853
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.754614469699022,
+      "learning_rate": 7.691986531337891e-07,
+      "loss": 0.5677,
+      "step": 7854
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7443708202298573,
+      "learning_rate": 7.682906617978836e-07,
+      "loss": 0.6404,
+      "step": 7855
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.087205618369263,
+      "learning_rate": 7.673831620897376e-07,
+      "loss": 0.6038,
+      "step": 7856
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.5096389980291365,
+      "learning_rate": 7.664761541147803e-07,
+      "loss": 0.6013,
+      "step": 7857
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4275211367126226,
+      "learning_rate": 7.65569637978385e-07,
+      "loss": 0.6306,
+      "step": 7858
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0724412487922104,
+      "learning_rate": 7.646636137858682e-07,
+      "loss": 0.5641,
+      "step": 7859
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6908765435507656,
+      "learning_rate": 7.637580816424906e-07,
+      "loss": 0.5556,
+      "step": 7860
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2336816700244158,
+      "learning_rate": 7.628530416534536e-07,
+      "loss": 0.6658,
+      "step": 7861
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.534474436786471,
+      "learning_rate": 7.619484939239008e-07,
+      "loss": 0.5065,
+      "step": 7862
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.25517760727134,
+      "learning_rate": 7.610444385589206e-07,
+      "loss": 0.5669,
+      "step": 7863
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4664200875331077,
+      "learning_rate": 7.601408756635454e-07,
+      "loss": 0.5449,
+      "step": 7864
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3983166993158402,
+      "learning_rate": 7.592378053427463e-07,
+      "loss": 0.631,
+      "step": 7865
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.44965513205932,
+      "learning_rate": 7.583352277014405e-07,
+      "loss": 0.6007,
+      "step": 7866
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.783794296236303,
+      "learning_rate": 7.574331428444859e-07,
+      "loss": 0.6112,
+      "step": 7867
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.858796558481055,
+      "learning_rate": 7.565315508766846e-07,
+      "loss": 0.6879,
+      "step": 7868
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3392675349173246,
+      "learning_rate": 7.556304519027824e-07,
+      "loss": 0.535,
+      "step": 7869
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.345381147919508,
+      "learning_rate": 7.547298460274655e-07,
+      "loss": 0.5671,
+      "step": 7870
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.960829913288072,
+      "learning_rate": 7.538297333553613e-07,
+      "loss": 0.5243,
+      "step": 7871
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4386836992879974,
+      "learning_rate": 7.529301139910444e-07,
+      "loss": 0.5946,
+      "step": 7872
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1914556637515674,
+      "learning_rate": 7.520309880390314e-07,
+      "loss": 0.5832,
+      "step": 7873
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.241007701198296,
+      "learning_rate": 7.51132355603778e-07,
+      "loss": 0.5416,
+      "step": 7874
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0724515989906513,
+      "learning_rate": 7.502342167896847e-07,
+      "loss": 0.5964,
+      "step": 7875
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5100584726157886,
+      "learning_rate": 7.493365717010947e-07,
+      "loss": 0.631,
+      "step": 7876
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7356056595469536,
+      "learning_rate": 7.484394204422962e-07,
+      "loss": 0.5424,
+      "step": 7877
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0885569649811786,
+      "learning_rate": 7.475427631175141e-07,
+      "loss": 0.5181,
+      "step": 7878
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.304547742686197,
+      "learning_rate": 7.466465998309225e-07,
+      "loss": 0.6151,
+      "step": 7879
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2979784331851913,
+      "learning_rate": 7.457509306866329e-07,
+      "loss": 0.569,
+      "step": 7880
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0618039183537324,
+      "learning_rate": 7.448557557887021e-07,
+      "loss": 0.586,
+      "step": 7881
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.5231994277364835,
+      "learning_rate": 7.439610752411303e-07,
+      "loss": 0.6385,
+      "step": 7882
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5269036735363617,
+      "learning_rate": 7.430668891478576e-07,
+      "loss": 0.5901,
+      "step": 7883
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3674099325580866,
+      "learning_rate": 7.421731976127672e-07,
+      "loss": 0.7775,
+      "step": 7884
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.671376901489743,
+      "learning_rate": 7.41280000739687e-07,
+      "loss": 0.5261,
+      "step": 7885
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.820411629914039,
+      "learning_rate": 7.403872986323862e-07,
+      "loss": 0.569,
+      "step": 7886
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.577175781646432,
+      "learning_rate": 7.394950913945759e-07,
+      "loss": 0.5512,
+      "step": 7887
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4642031464494174,
+      "learning_rate": 7.386033791299091e-07,
+      "loss": 0.6127,
+      "step": 7888
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.434672247809207,
+      "learning_rate": 7.37712161941983e-07,
+      "loss": 0.6554,
+      "step": 7889
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5103511043710465,
+      "learning_rate": 7.368214399343371e-07,
+      "loss": 0.5331,
+      "step": 7890
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7655999882645057,
+      "learning_rate": 7.35931213210454e-07,
+      "loss": 0.7108,
+      "step": 7891
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.2183982961814324,
+      "learning_rate": 7.350414818737562e-07,
+      "loss": 0.5502,
+      "step": 7892
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.586664475078197,
+      "learning_rate": 7.34152246027609e-07,
+      "loss": 0.5924,
+      "step": 7893
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.366201926305009,
+      "learning_rate": 7.332635057753224e-07,
+      "loss": 0.6236,
+      "step": 7894
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3175961493215635,
+      "learning_rate": 7.323752612201491e-07,
+      "loss": 0.5758,
+      "step": 7895
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 16.41576198449689,
+      "learning_rate": 7.314875124652815e-07,
+      "loss": 0.6657,
+      "step": 7896
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.516520516533135,
+      "learning_rate": 7.306002596138551e-07,
+      "loss": 0.6438,
+      "step": 7897
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.9227517959801372,
+      "learning_rate": 7.297135027689484e-07,
+      "loss": 0.5942,
+      "step": 7898
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6069484447619486,
+      "learning_rate": 7.288272420335841e-07,
+      "loss": 0.5786,
+      "step": 7899
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5621275414127007,
+      "learning_rate": 7.279414775107241e-07,
+      "loss": 0.6239,
+      "step": 7900
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5478933874139176,
+      "learning_rate": 7.270562093032724e-07,
+      "loss": 0.6154,
+      "step": 7901
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5146219884970686,
+      "learning_rate": 7.261714375140788e-07,
+      "loss": 0.6525,
+      "step": 7902
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.26095835769191,
+      "learning_rate": 7.252871622459335e-07,
+      "loss": 0.6109,
+      "step": 7903
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7915685895483926,
+      "learning_rate": 7.244033836015696e-07,
+      "loss": 0.7238,
+      "step": 7904
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.44562542266984,
+      "learning_rate": 7.235201016836613e-07,
+      "loss": 0.5962,
+      "step": 7905
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.6869459471637755,
+      "learning_rate": 7.226373165948241e-07,
+      "loss": 0.5836,
+      "step": 7906
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.282445234624132,
+      "learning_rate": 7.21755028437619e-07,
+      "loss": 0.6254,
+      "step": 7907
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.521493455628609,
+      "learning_rate": 7.208732373145483e-07,
+      "loss": 0.6589,
+      "step": 7908
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.232695529398515,
+      "learning_rate": 7.199919433280555e-07,
+      "loss": 0.6094,
+      "step": 7909
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.962644679645622,
+      "learning_rate": 7.191111465805256e-07,
+      "loss": 0.5981,
+      "step": 7910
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4471859782636165,
+      "learning_rate": 7.182308471742877e-07,
+      "loss": 0.5816,
+      "step": 7911
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.702379481712408,
+      "learning_rate": 7.173510452116139e-07,
+      "loss": 0.68,
+      "step": 7912
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0632897742865417,
+      "learning_rate": 7.164717407947142e-07,
+      "loss": 0.5708,
+      "step": 7913
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.423696780817971,
+      "learning_rate": 7.155929340257467e-07,
+      "loss": 0.5654,
+      "step": 7914
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0545652099506952,
+      "learning_rate": 7.14714625006806e-07,
+      "loss": 0.549,
+      "step": 7915
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.240068457011979,
+      "learning_rate": 7.138368138399327e-07,
+      "loss": 0.575,
+      "step": 7916
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.354272324791377,
+      "learning_rate": 7.129595006271095e-07,
+      "loss": 0.6664,
+      "step": 7917
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.1167965022837523,
+      "learning_rate": 7.120826854702589e-07,
+      "loss": 0.5909,
+      "step": 7918
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8860947312539493,
+      "learning_rate": 7.112063684712456e-07,
+      "loss": 0.5764,
+      "step": 7919
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5639149644592307,
+      "learning_rate": 7.103305497318786e-07,
+      "loss": 0.5729,
+      "step": 7920
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.157567164040736,
+      "learning_rate": 7.094552293539098e-07,
+      "loss": 0.6691,
+      "step": 7921
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.017279987323207,
+      "learning_rate": 7.08580407439029e-07,
+      "loss": 0.7188,
+      "step": 7922
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.180206068529602,
+      "learning_rate": 7.077060840888705e-07,
+      "loss": 0.5835,
+      "step": 7923
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.2183678021512128,
+      "learning_rate": 7.068322594050114e-07,
+      "loss": 0.5853,
+      "step": 7924
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.730991122722091,
+      "learning_rate": 7.059589334889705e-07,
+      "loss": 0.5394,
+      "step": 7925
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7372841955338085,
+      "learning_rate": 7.050861064422087e-07,
+      "loss": 0.6588,
+      "step": 7926
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8242014928415333,
+      "learning_rate": 7.042137783661273e-07,
+      "loss": 0.6495,
+      "step": 7927
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.3717042988457577,
+      "learning_rate": 7.033419493620708e-07,
+      "loss": 0.6167,
+      "step": 7928
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9434034854373674,
+      "learning_rate": 7.024706195313258e-07,
+      "loss": 0.5459,
+      "step": 7929
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.948265264022775,
+      "learning_rate": 7.015997889751225e-07,
+      "loss": 0.5389,
+      "step": 7930
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.464153456017488,
+      "learning_rate": 7.007294577946306e-07,
+      "loss": 0.5651,
+      "step": 7931
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0196234432409446,
+      "learning_rate": 6.998596260909607e-07,
+      "loss": 0.6672,
+      "step": 7932
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2187036742506003,
+      "learning_rate": 6.989902939651694e-07,
+      "loss": 0.6165,
+      "step": 7933
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4128874884249774,
+      "learning_rate": 6.981214615182541e-07,
+      "loss": 0.5248,
+      "step": 7934
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.714711543409674,
+      "learning_rate": 6.972531288511514e-07,
+      "loss": 0.6234,
+      "step": 7935
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.566568280987358,
+      "learning_rate": 6.963852960647416e-07,
+      "loss": 0.6142,
+      "step": 7936
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5912361052965105,
+      "learning_rate": 6.955179632598475e-07,
+      "loss": 0.6019,
+      "step": 7937
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.0046865938581657,
+      "learning_rate": 6.946511305372327e-07,
+      "loss": 0.6551,
+      "step": 7938
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.62750432152481,
+      "learning_rate": 6.937847979976059e-07,
+      "loss": 0.7307,
+      "step": 7939
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.463425594035302,
+      "learning_rate": 6.929189657416136e-07,
+      "loss": 0.6263,
+      "step": 7940
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.293961976197948,
+      "learning_rate": 6.920536338698436e-07,
+      "loss": 0.7305,
+      "step": 7941
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.245455408558447,
+      "learning_rate": 6.911888024828295e-07,
+      "loss": 0.6554,
+      "step": 7942
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0647694096526474,
+      "learning_rate": 6.903244716810459e-07,
+      "loss": 0.5978,
+      "step": 7943
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2770572793711135,
+      "learning_rate": 6.894606415649074e-07,
+      "loss": 0.6284,
+      "step": 7944
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.590717718772522,
+      "learning_rate": 6.8859731223477e-07,
+      "loss": 0.6585,
+      "step": 7945
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1415058085079095,
+      "learning_rate": 6.877344837909334e-07,
+      "loss": 0.6856,
+      "step": 7946
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5987535704327875,
+      "learning_rate": 6.868721563336406e-07,
+      "loss": 0.6328,
+      "step": 7947
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5233039406914495,
+      "learning_rate": 6.860103299630722e-07,
+      "loss": 0.6124,
+      "step": 7948
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.524754669248327,
+      "learning_rate": 6.851490047793524e-07,
+      "loss": 0.6072,
+      "step": 7949
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.540427405947607,
+      "learning_rate": 6.84288180882548e-07,
+      "loss": 0.6113,
+      "step": 7950
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9908048969702398,
+      "learning_rate": 6.834278583726677e-07,
+      "loss": 0.532,
+      "step": 7951
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.2038346124338073,
+      "learning_rate": 6.825680373496618e-07,
+      "loss": 0.5439,
+      "step": 7952
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9377925721646436,
+      "learning_rate": 6.817087179134208e-07,
+      "loss": 0.5117,
+      "step": 7953
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.407003844113903,
+      "learning_rate": 6.80849900163777e-07,
+      "loss": 0.5797,
+      "step": 7954
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2839557052539923,
+      "learning_rate": 6.799915842005062e-07,
+      "loss": 0.5703,
+      "step": 7955
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3947698210437935,
+      "learning_rate": 6.791337701233269e-07,
+      "loss": 0.5937,
+      "step": 7956
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6739236325422904,
+      "learning_rate": 6.782764580318951e-07,
+      "loss": 0.7199,
+      "step": 7957
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1116863967686004,
+      "learning_rate": 6.774196480258111e-07,
+      "loss": 0.6494,
+      "step": 7958
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1683016397788997,
+      "learning_rate": 6.765633402046168e-07,
+      "loss": 0.5955,
+      "step": 7959
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.032305057994704,
+      "learning_rate": 6.757075346677961e-07,
+      "loss": 0.603,
+      "step": 7960
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2604613175721164,
+      "learning_rate": 6.748522315147744e-07,
+      "loss": 0.6187,
+      "step": 7961
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.969636492342722,
+      "learning_rate": 6.739974308449176e-07,
+      "loss": 0.6081,
+      "step": 7962
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.0339603599056373,
+      "learning_rate": 6.731431327575339e-07,
+      "loss": 0.6901,
+      "step": 7963
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.2703468846615533,
+      "learning_rate": 6.722893373518724e-07,
+      "loss": 0.5622,
+      "step": 7964
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4646080011429485,
+      "learning_rate": 6.714360447271273e-07,
+      "loss": 0.522,
+      "step": 7965
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.364242035976487,
+      "learning_rate": 6.705832549824293e-07,
+      "loss": 0.6273,
+      "step": 7966
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.7352098181866817,
+      "learning_rate": 6.69730968216853e-07,
+      "loss": 0.6764,
+      "step": 7967
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.616311936586499,
+      "learning_rate": 6.688791845294151e-07,
+      "loss": 0.6525,
+      "step": 7968
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4255233825081826,
+      "learning_rate": 6.680279040190745e-07,
+      "loss": 0.6588,
+      "step": 7969
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5378364716116764,
+      "learning_rate": 6.671771267847299e-07,
+      "loss": 0.5982,
+      "step": 7970
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.323293588729379,
+      "learning_rate": 6.663268529252209e-07,
+      "loss": 0.6346,
+      "step": 7971
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.164418560400295,
+      "learning_rate": 6.654770825393303e-07,
+      "loss": 0.6584,
+      "step": 7972
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.9789326506768754,
+      "learning_rate": 6.646278157257824e-07,
+      "loss": 0.5408,
+      "step": 7973
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4473343292404257,
+      "learning_rate": 6.637790525832438e-07,
+      "loss": 0.6776,
+      "step": 7974
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.333751171962327,
+      "learning_rate": 6.629307932103201e-07,
+      "loss": 0.6228,
+      "step": 7975
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.194575988760013,
+      "learning_rate": 6.620830377055587e-07,
+      "loss": 0.5451,
+      "step": 7976
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1276202217427196,
+      "learning_rate": 6.612357861674501e-07,
+      "loss": 0.5791,
+      "step": 7977
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5647125903291776,
+      "learning_rate": 6.603890386944273e-07,
+      "loss": 0.5679,
+      "step": 7978
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3644218883143338,
+      "learning_rate": 6.59542795384861e-07,
+      "loss": 0.5646,
+      "step": 7979
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.3454428465301946,
+      "learning_rate": 6.586970563370649e-07,
+      "loss": 0.6593,
+      "step": 7980
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.2957829519224,
+      "learning_rate": 6.578518216492951e-07,
+      "loss": 0.6074,
+      "step": 7981
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.989238602602019,
+      "learning_rate": 6.570070914197496e-07,
+      "loss": 0.6612,
+      "step": 7982
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6361300624581046,
+      "learning_rate": 6.561628657465663e-07,
+      "loss": 0.6422,
+      "step": 7983
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3784169240024085,
+      "learning_rate": 6.553191447278234e-07,
+      "loss": 0.6598,
+      "step": 7984
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.116665012500234,
+      "learning_rate": 6.544759284615431e-07,
+      "loss": 0.5628,
+      "step": 7985
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.866416810480354,
+      "learning_rate": 6.536332170456877e-07,
+      "loss": 0.5985,
+      "step": 7986
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.122808635192976,
+      "learning_rate": 6.527910105781626e-07,
+      "loss": 0.5883,
+      "step": 7987
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.9397977174918566,
+      "learning_rate": 6.519493091568108e-07,
+      "loss": 0.6253,
+      "step": 7988
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.2026307795457765,
+      "learning_rate": 6.511081128794183e-07,
+      "loss": 0.6072,
+      "step": 7989
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.391514496115847,
+      "learning_rate": 6.502674218437144e-07,
+      "loss": 0.5603,
+      "step": 7990
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.156056886121653,
+      "learning_rate": 6.494272361473681e-07,
+      "loss": 0.6524,
+      "step": 7991
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.732431840720779,
+      "learning_rate": 6.485875558879895e-07,
+      "loss": 0.6481,
+      "step": 7992
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9873437823454094,
+      "learning_rate": 6.477483811631291e-07,
+      "loss": 0.5987,
+      "step": 7993
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.613588104761939,
+      "learning_rate": 6.469097120702805e-07,
+      "loss": 0.6542,
+      "step": 7994
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.288884381877685,
+      "learning_rate": 6.460715487068781e-07,
+      "loss": 0.5694,
+      "step": 7995
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.0335170876836783,
+      "learning_rate": 6.452338911702994e-07,
+      "loss": 0.7001,
+      "step": 7996
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.7744780023995363,
+      "learning_rate": 6.443967395578565e-07,
+      "loss": 0.6253,
+      "step": 7997
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6347984361148904,
+      "learning_rate": 6.435600939668096e-07,
+      "loss": 0.592,
+      "step": 7998
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.011072253449798,
+      "learning_rate": 6.42723954494358e-07,
+      "loss": 0.6553,
+      "step": 7999
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.0584864481962364,
+      "learning_rate": 6.418883212376431e-07,
+      "loss": 0.5839,
+      "step": 8000
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.296221820496963,
+      "learning_rate": 6.410531942937448e-07,
+      "loss": 0.5928,
+      "step": 8001
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.406097985573007,
+      "learning_rate": 6.402185737596844e-07,
+      "loss": 0.6986,
+      "step": 8002
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2589114977154625,
+      "learning_rate": 6.393844597324278e-07,
+      "loss": 0.5709,
+      "step": 8003
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.7836938958633284,
+      "learning_rate": 6.385508523088801e-07,
+      "loss": 0.5831,
+      "step": 8004
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6551153221400194,
+      "learning_rate": 6.377177515858874e-07,
+      "loss": 0.6167,
+      "step": 8005
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6555647926872816,
+      "learning_rate": 6.368851576602347e-07,
+      "loss": 0.5478,
+      "step": 8006
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.0795265737705124,
+      "learning_rate": 6.360530706286516e-07,
+      "loss": 0.6234,
+      "step": 8007
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.01122414638164,
+      "learning_rate": 6.352214905878085e-07,
+      "loss": 0.5999,
+      "step": 8008
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.8498907965644706,
+      "learning_rate": 6.343904176343169e-07,
+      "loss": 0.6705,
+      "step": 8009
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4166442373218096,
+      "learning_rate": 6.335598518647251e-07,
+      "loss": 0.5388,
+      "step": 8010
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6880105299523067,
+      "learning_rate": 6.327297933755272e-07,
+      "loss": 0.5941,
+      "step": 8011
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.846835922064916,
+      "learning_rate": 6.319002422631582e-07,
+      "loss": 0.6664,
+      "step": 8012
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.031689173749022,
+      "learning_rate": 6.310711986239926e-07,
+      "loss": 0.5696,
+      "step": 8013
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.73606226733396,
+      "learning_rate": 6.302426625543457e-07,
+      "loss": 0.5136,
+      "step": 8014
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.9278610126869204,
+      "learning_rate": 6.294146341504742e-07,
+      "loss": 0.6988,
+      "step": 8015
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6587658045083002,
+      "learning_rate": 6.285871135085758e-07,
+      "loss": 0.586,
+      "step": 8016
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5101848696418028,
+      "learning_rate": 6.277601007247913e-07,
+      "loss": 0.6072,
+      "step": 8017
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.094591388241789,
+      "learning_rate": 6.269335958951995e-07,
+      "loss": 0.6101,
+      "step": 8018
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1190721327022017,
+      "learning_rate": 6.2610759911582e-07,
+      "loss": 0.5325,
+      "step": 8019
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.394049016837729,
+      "learning_rate": 6.252821104826163e-07,
+      "loss": 0.6149,
+      "step": 8020
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.185048534616482,
+      "learning_rate": 6.244571300914909e-07,
+      "loss": 0.5355,
+      "step": 8021
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9946377288465473,
+      "learning_rate": 6.23632658038289e-07,
+      "loss": 0.5187,
+      "step": 8022
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.8433532638524928,
+      "learning_rate": 6.228086944187939e-07,
+      "loss": 0.6301,
+      "step": 8023
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.239562343118617,
+      "learning_rate": 6.219852393287302e-07,
+      "loss": 0.6108,
+      "step": 8024
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.19755867950319,
+      "learning_rate": 6.211622928637662e-07,
+      "loss": 0.5238,
+      "step": 8025
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1376360439152657,
+      "learning_rate": 6.2033985511951e-07,
+      "loss": 0.6205,
+      "step": 8026
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.563081310031987,
+      "learning_rate": 6.19517926191509e-07,
+      "loss": 0.6733,
+      "step": 8027
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.266713807274647,
+      "learning_rate": 6.186965061752515e-07,
+      "loss": 0.5803,
+      "step": 8028
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.33107338075643,
+      "learning_rate": 6.178755951661692e-07,
+      "loss": 0.542,
+      "step": 8029
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1249280857293966,
+      "learning_rate": 6.170551932596336e-07,
+      "loss": 0.6044,
+      "step": 8030
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.222877055640638,
+      "learning_rate": 6.162353005509558e-07,
+      "loss": 0.5909,
+      "step": 8031
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.466145340026216,
+      "learning_rate": 6.154159171353879e-07,
+      "loss": 0.6029,
+      "step": 8032
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5315597654614805,
+      "learning_rate": 6.145970431081238e-07,
+      "loss": 0.5896,
+      "step": 8033
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.769252012832659,
+      "learning_rate": 6.137786785642985e-07,
+      "loss": 0.6102,
+      "step": 8034
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4509615775141977,
+      "learning_rate": 6.129608235989881e-07,
+      "loss": 0.5448,
+      "step": 8035
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.141998734885848,
+      "learning_rate": 6.121434783072077e-07,
+      "loss": 0.6219,
+      "step": 8036
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.090304082790666,
+      "learning_rate": 6.113266427839126e-07,
+      "loss": 0.5533,
+      "step": 8037
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.290451018341766,
+      "learning_rate": 6.105103171240018e-07,
+      "loss": 0.6181,
+      "step": 8038
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.270211096142163,
+      "learning_rate": 6.096945014223149e-07,
+      "loss": 0.614,
+      "step": 8039
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.9037913388153433,
+      "learning_rate": 6.088791957736301e-07,
+      "loss": 0.6391,
+      "step": 8040
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3997821521809874,
+      "learning_rate": 6.080644002726655e-07,
+      "loss": 0.6159,
+      "step": 8041
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.436406794174699,
+      "learning_rate": 6.072501150140824e-07,
+      "loss": 0.6379,
+      "step": 8042
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.1984050878329087,
+      "learning_rate": 6.064363400924839e-07,
+      "loss": 0.6351,
+      "step": 8043
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.9565521370159096,
+      "learning_rate": 6.056230756024123e-07,
+      "loss": 0.5821,
+      "step": 8044
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.4666008560066732,
+      "learning_rate": 6.048103216383472e-07,
+      "loss": 0.5522,
+      "step": 8045
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3667556756148342,
+      "learning_rate": 6.03998078294713e-07,
+      "loss": 0.5512,
+      "step": 8046
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6465130816201254,
+      "learning_rate": 6.031863456658754e-07,
+      "loss": 0.5849,
+      "step": 8047
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.036938049317216,
+      "learning_rate": 6.023751238461389e-07,
+      "loss": 0.5304,
+      "step": 8048
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.075249746973687,
+      "learning_rate": 6.015644129297482e-07,
+      "loss": 0.6031,
+      "step": 8049
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8124802810968377,
+      "learning_rate": 6.007542130108885e-07,
+      "loss": 0.6553,
+      "step": 8050
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.474131988219833,
+      "learning_rate": 5.999445241836877e-07,
+      "loss": 0.5124,
+      "step": 8051
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.668817739499102,
+      "learning_rate": 5.991353465422134e-07,
+      "loss": 0.5076,
+      "step": 8052
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.411857351900381,
+      "learning_rate": 5.983266801804732e-07,
+      "loss": 0.6072,
+      "step": 8053
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4264765584589165,
+      "learning_rate": 5.975185251924143e-07,
+      "loss": 0.6211,
+      "step": 8054
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.102878662708729,
+      "learning_rate": 5.967108816719264e-07,
+      "loss": 0.6705,
+      "step": 8055
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.401655319504964,
+      "learning_rate": 5.959037497128401e-07,
+      "loss": 0.6786,
+      "step": 8056
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.152666465422462,
+      "learning_rate": 5.950971294089258e-07,
+      "loss": 0.6148,
+      "step": 8057
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.675848160084293,
+      "learning_rate": 5.942910208538943e-07,
+      "loss": 0.595,
+      "step": 8058
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.2338755878945475,
+      "learning_rate": 5.934854241413951e-07,
+      "loss": 0.62,
+      "step": 8059
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.7539432304617,
+      "learning_rate": 5.926803393650215e-07,
+      "loss": 0.563,
+      "step": 8060
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.418072027591026,
+      "learning_rate": 5.918757666183067e-07,
+      "loss": 0.5881,
+      "step": 8061
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4209012407142225,
+      "learning_rate": 5.91071705994723e-07,
+      "loss": 0.5809,
+      "step": 8062
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.302902584346649,
+      "learning_rate": 5.902681575876822e-07,
+      "loss": 0.5901,
+      "step": 8063
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3794113470801763,
+      "learning_rate": 5.894651214905395e-07,
+      "loss": 0.5332,
+      "step": 8064
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4361869128303573,
+      "learning_rate": 5.88662597796591e-07,
+      "loss": 0.5329,
+      "step": 8065
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.815110075655275,
+      "learning_rate": 5.878605865990694e-07,
+      "loss": 0.6081,
+      "step": 8066
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.448042678880235,
+      "learning_rate": 5.870590879911498e-07,
+      "loss": 0.5438,
+      "step": 8067
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6312265037825413,
+      "learning_rate": 5.862581020659491e-07,
+      "loss": 0.6526,
+      "step": 8068
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.162596770198441,
+      "learning_rate": 5.854576289165232e-07,
+      "loss": 0.6465,
+      "step": 8069
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5578744017745305,
+      "learning_rate": 5.846576686358696e-07,
+      "loss": 0.6271,
+      "step": 8070
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8100954598723407,
+      "learning_rate": 5.838582213169247e-07,
+      "loss": 0.6043,
+      "step": 8071
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.2170401537564057,
+      "learning_rate": 5.830592870525647e-07,
+      "loss": 0.6683,
+      "step": 8072
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.379729669844667,
+      "learning_rate": 5.822608659356093e-07,
+      "loss": 0.5669,
+      "step": 8073
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.759833858853163,
+      "learning_rate": 5.814629580588165e-07,
+      "loss": 0.679,
+      "step": 8074
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.379428037812773,
+      "learning_rate": 5.80665563514885e-07,
+      "loss": 0.5477,
+      "step": 8075
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9686713188136623,
+      "learning_rate": 5.798686823964517e-07,
+      "loss": 0.5656,
+      "step": 8076
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.208901084034373,
+      "learning_rate": 5.79072314796098e-07,
+      "loss": 0.6539,
+      "step": 8077
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.319119693164557,
+      "learning_rate": 5.78276460806343e-07,
+      "loss": 0.5599,
+      "step": 8078
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.486494110324032,
+      "learning_rate": 5.77481120519649e-07,
+      "loss": 0.6557,
+      "step": 8079
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.356169785828175,
+      "learning_rate": 5.766862940284124e-07,
+      "loss": 0.5505,
+      "step": 8080
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.308551249984606,
+      "learning_rate": 5.758919814249753e-07,
+      "loss": 0.6307,
+      "step": 8081
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.1735857620440453,
+      "learning_rate": 5.750981828016189e-07,
+      "loss": 0.5625,
+      "step": 8082
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.0910390532362695,
+      "learning_rate": 5.743048982505656e-07,
+      "loss": 0.5263,
+      "step": 8083
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.0386446705294796,
+      "learning_rate": 5.73512127863976e-07,
+      "loss": 0.5617,
+      "step": 8084
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.7279151957638255,
+      "learning_rate": 5.727198717339511e-07,
+      "loss": 0.6422,
+      "step": 8085
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.677171114462176,
+      "learning_rate": 5.719281299525331e-07,
+      "loss": 0.5548,
+      "step": 8086
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.209112258602768,
+      "learning_rate": 5.711369026117053e-07,
+      "loss": 0.6056,
+      "step": 8087
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.9912493406236726,
+      "learning_rate": 5.703461898033902e-07,
+      "loss": 0.6632,
+      "step": 8088
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.8418625332520318,
+      "learning_rate": 5.695559916194488e-07,
+      "loss": 0.6912,
+      "step": 8089
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.516446973072753,
+      "learning_rate": 5.687663081516853e-07,
+      "loss": 0.6293,
+      "step": 8090
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.7840898044343776,
+      "learning_rate": 5.679771394918427e-07,
+      "loss": 0.5642,
+      "step": 8091
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6457572435732244,
+      "learning_rate": 5.671884857316051e-07,
+      "loss": 0.559,
+      "step": 8092
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3271517036024933,
+      "learning_rate": 5.66400346962595e-07,
+      "loss": 0.7182,
+      "step": 8093
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.7706715409747082,
+      "learning_rate": 5.656127232763759e-07,
+      "loss": 0.6649,
+      "step": 8094
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.34603048762334,
+      "learning_rate": 5.64825614764452e-07,
+      "loss": 0.6284,
+      "step": 8095
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0215689336292253,
+      "learning_rate": 5.640390215182683e-07,
+      "loss": 0.6201,
+      "step": 8096
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.53100525886137,
+      "learning_rate": 5.632529436292083e-07,
+      "loss": 0.5778,
+      "step": 8097
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.126583426550784,
+      "learning_rate": 5.624673811885945e-07,
+      "loss": 0.61,
+      "step": 8098
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4029139519399867,
+      "learning_rate": 5.616823342876932e-07,
+      "loss": 0.5994,
+      "step": 8099
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3048343587136615,
+      "learning_rate": 5.608978030177087e-07,
+      "loss": 0.5854,
+      "step": 8100
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.276243991774919,
+      "learning_rate": 5.601137874697859e-07,
+      "loss": 0.6745,
+      "step": 8101
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5942109454123052,
+      "learning_rate": 5.593302877350076e-07,
+      "loss": 0.6314,
+      "step": 8102
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.9318122332058856,
+      "learning_rate": 5.585473039044004e-07,
+      "loss": 0.5933,
+      "step": 8103
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4441412822757806,
+      "learning_rate": 5.577648360689281e-07,
+      "loss": 0.6418,
+      "step": 8104
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.2034597908955575,
+      "learning_rate": 5.569828843194969e-07,
+      "loss": 0.5708,
+      "step": 8105
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6115667573813166,
+      "learning_rate": 5.562014487469502e-07,
+      "loss": 0.6829,
+      "step": 8106
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6541755315222266,
+      "learning_rate": 5.554205294420733e-07,
+      "loss": 0.5536,
+      "step": 8107
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.9752658226736735,
+      "learning_rate": 5.546401264955909e-07,
+      "loss": 0.5814,
+      "step": 8108
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.8346292608939656,
+      "learning_rate": 5.538602399981696e-07,
+      "loss": 0.593,
+      "step": 8109
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.279904270343523,
+      "learning_rate": 5.530808700404128e-07,
+      "loss": 0.6239,
+      "step": 8110
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4794073490515975,
+      "learning_rate": 5.523020167128651e-07,
+      "loss": 0.5705,
+      "step": 8111
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.4668680618890395,
+      "learning_rate": 5.51523680106012e-07,
+      "loss": 0.6668,
+      "step": 8112
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.484931444733449,
+      "learning_rate": 5.507458603102783e-07,
+      "loss": 0.6922,
+      "step": 8113
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.674438977252659,
+      "learning_rate": 5.499685574160312e-07,
+      "loss": 0.5529,
+      "step": 8114
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.284875736493241,
+      "learning_rate": 5.491917715135719e-07,
+      "loss": 0.5868,
+      "step": 8115
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.901045185802944,
+      "learning_rate": 5.484155026931459e-07,
+      "loss": 0.6206,
+      "step": 8116
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.196587431457302,
+      "learning_rate": 5.476397510449389e-07,
+      "loss": 0.6526,
+      "step": 8117
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.1539318616200513,
+      "learning_rate": 5.468645166590758e-07,
+      "loss": 0.6617,
+      "step": 8118
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4155197419898746,
+      "learning_rate": 5.46089799625621e-07,
+      "loss": 0.5818,
+      "step": 8119
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6201210125396237,
+      "learning_rate": 5.453156000345772e-07,
+      "loss": 0.5792,
+      "step": 8120
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.457503317771027,
+      "learning_rate": 5.445419179758893e-07,
+      "loss": 0.5982,
+      "step": 8121
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.551730796116664,
+      "learning_rate": 5.437687535394431e-07,
+      "loss": 0.5919,
+      "step": 8122
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.726936420412654,
+      "learning_rate": 5.429961068150619e-07,
+      "loss": 0.5739,
+      "step": 8123
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.536441146532936,
+      "learning_rate": 5.422239778925076e-07,
+      "loss": 0.6811,
+      "step": 8124
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.9590736750871485,
+      "learning_rate": 5.414523668614857e-07,
+      "loss": 0.5417,
+      "step": 8125
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.7180436990498595,
+      "learning_rate": 5.406812738116396e-07,
+      "loss": 0.6244,
+      "step": 8126
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2060907410932096,
+      "learning_rate": 5.399106988325543e-07,
+      "loss": 0.5769,
+      "step": 8127
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9386958343089704,
+      "learning_rate": 5.39140642013749e-07,
+      "loss": 0.5423,
+      "step": 8128
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5672726543123683,
+      "learning_rate": 5.383711034446892e-07,
+      "loss": 0.5791,
+      "step": 8129
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6077162998901873,
+      "learning_rate": 5.376020832147777e-07,
+      "loss": 0.6193,
+      "step": 8130
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.6993050614435203,
+      "learning_rate": 5.368335814133569e-07,
+      "loss": 0.5434,
+      "step": 8131
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2676443993494813,
+      "learning_rate": 5.360655981297097e-07,
+      "loss": 0.5061,
+      "step": 8132
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.204912156598437,
+      "learning_rate": 5.352981334530555e-07,
+      "loss": 0.5548,
+      "step": 8133
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.3583543994309633,
+      "learning_rate": 5.345311874725584e-07,
+      "loss": 0.5469,
+      "step": 8134
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.73031391634573,
+      "learning_rate": 5.337647602773211e-07,
+      "loss": 0.6302,
+      "step": 8135
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6740573030486434,
+      "learning_rate": 5.329988519563828e-07,
+      "loss": 0.5692,
+      "step": 8136
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9923350695353295,
+      "learning_rate": 5.322334625987241e-07,
+      "loss": 0.5938,
+      "step": 8137
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8706603816339067,
+      "learning_rate": 5.314685922932666e-07,
+      "loss": 0.5283,
+      "step": 8138
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4156359250535178,
+      "learning_rate": 5.30704241128871e-07,
+      "loss": 0.6129,
+      "step": 8139
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.789191292597911,
+      "learning_rate": 5.299404091943383e-07,
+      "loss": 0.6423,
+      "step": 8140
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3542836512039425,
+      "learning_rate": 5.291770965784076e-07,
+      "loss": 0.6534,
+      "step": 8141
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6693420950157534,
+      "learning_rate": 5.284143033697565e-07,
+      "loss": 0.5847,
+      "step": 8142
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4801931250434497,
+      "learning_rate": 5.276520296570053e-07,
+      "loss": 0.6162,
+      "step": 8143
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6330298094643676,
+      "learning_rate": 5.268902755287148e-07,
+      "loss": 0.5991,
+      "step": 8144
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.514352693873068,
+      "learning_rate": 5.26129041073381e-07,
+      "loss": 0.5545,
+      "step": 8145
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.334792126153137,
+      "learning_rate": 5.253683263794418e-07,
+      "loss": 0.5464,
+      "step": 8146
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.269857305345398,
+      "learning_rate": 5.246081315352758e-07,
+      "loss": 0.6648,
+      "step": 8147
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5865095430846656,
+      "learning_rate": 5.238484566292002e-07,
+      "loss": 0.7269,
+      "step": 8148
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4282223527075573,
+      "learning_rate": 5.230893017494731e-07,
+      "loss": 0.5743,
+      "step": 8149
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0285336151160749,
+      "learning_rate": 5.223306669842876e-07,
+      "loss": 0.5332,
+      "step": 8150
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5241337376310526,
+      "learning_rate": 5.215725524217818e-07,
+      "loss": 0.555,
+      "step": 8151
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.159756791738227,
+      "learning_rate": 5.20814958150031e-07,
+      "loss": 0.6031,
+      "step": 8152
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.682488901407614,
+      "learning_rate": 5.200578842570508e-07,
+      "loss": 0.5975,
+      "step": 8153
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9504110630710011,
+      "learning_rate": 5.19301330830796e-07,
+      "loss": 0.559,
+      "step": 8154
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.283295404157167,
+      "learning_rate": 5.185452979591593e-07,
+      "loss": 0.647,
+      "step": 8155
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.895291731557172,
+      "learning_rate": 5.177897857299752e-07,
+      "loss": 0.5969,
+      "step": 8156
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0204477401700256,
+      "learning_rate": 5.170347942310177e-07,
+      "loss": 0.4857,
+      "step": 8157
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.281207098958877,
+      "learning_rate": 5.162803235499992e-07,
+      "loss": 0.5998,
+      "step": 8158
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9244566758863595,
+      "learning_rate": 5.155263737745703e-07,
+      "loss": 0.5635,
+      "step": 8159
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0701940095232954,
+      "learning_rate": 5.147729449923244e-07,
+      "loss": 0.5843,
+      "step": 8160
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5067250985588987,
+      "learning_rate": 5.140200372907921e-07,
+      "loss": 0.6793,
+      "step": 8161
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.105940578606609,
+      "learning_rate": 5.132676507574463e-07,
+      "loss": 0.6469,
+      "step": 8162
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.241058072116501,
+      "learning_rate": 5.125157854796925e-07,
+      "loss": 0.4969,
+      "step": 8163
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.792486187126245,
+      "learning_rate": 5.11764441544883e-07,
+      "loss": 0.6391,
+      "step": 8164
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.829466955623266,
+      "learning_rate": 5.11013619040307e-07,
+      "loss": 0.7384,
+      "step": 8165
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4147912110071825,
+      "learning_rate": 5.10263318053193e-07,
+      "loss": 0.6223,
+      "step": 8166
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.362586700546461,
+      "learning_rate": 5.095135386707084e-07,
+      "loss": 0.588,
+      "step": 8167
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.445979909568467,
+      "learning_rate": 5.087642809799587e-07,
+      "loss": 0.6189,
+      "step": 8168
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.732920831150975,
+      "learning_rate": 5.080155450679924e-07,
+      "loss": 0.5372,
+      "step": 8169
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.705560498251637,
+      "learning_rate": 5.072673310217957e-07,
+      "loss": 0.6203,
+      "step": 8170
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.31980030969563,
+      "learning_rate": 5.065196389282939e-07,
+      "loss": 0.5589,
+      "step": 8171
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5656538295462887,
+      "learning_rate": 5.057724688743498e-07,
+      "loss": 0.6339,
+      "step": 8172
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6519050302952736,
+      "learning_rate": 5.050258209467684e-07,
+      "loss": 0.7224,
+      "step": 8173
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.591412532091134,
+      "learning_rate": 5.042796952322943e-07,
+      "loss": 0.5866,
+      "step": 8174
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.212827161480501,
+      "learning_rate": 5.035340918176096e-07,
+      "loss": 0.7674,
+      "step": 8175
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.487115162626594,
+      "learning_rate": 5.027890107893368e-07,
+      "loss": 0.6352,
+      "step": 8176
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.360221538239416,
+      "learning_rate": 5.020444522340351e-07,
+      "loss": 0.5827,
+      "step": 8177
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.6272603981979445,
+      "learning_rate": 5.013004162382068e-07,
+      "loss": 0.5703,
+      "step": 8178
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.297902416786526,
+      "learning_rate": 5.005569028882928e-07,
+      "loss": 0.6296,
+      "step": 8179
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.272681457850102,
+      "learning_rate": 4.998139122706713e-07,
+      "loss": 0.6528,
+      "step": 8180
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8422597029287604,
+      "learning_rate": 4.990714444716594e-07,
+      "loss": 0.6148,
+      "step": 8181
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.645227694382434,
+      "learning_rate": 4.983294995775167e-07,
+      "loss": 0.6103,
+      "step": 8182
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.283592322689147,
+      "learning_rate": 4.975880776744397e-07,
+      "loss": 0.5401,
+      "step": 8183
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1952627320177975,
+      "learning_rate": 4.968471788485663e-07,
+      "loss": 0.5755,
+      "step": 8184
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4356967067289714,
+      "learning_rate": 4.961068031859684e-07,
+      "loss": 0.6096,
+      "step": 8185
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8567686330899567,
+      "learning_rate": 4.953669507726633e-07,
+      "loss": 0.6889,
+      "step": 8186
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.5528236481312305,
+      "learning_rate": 4.946276216946034e-07,
+      "loss": 0.6052,
+      "step": 8187
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.0251800181776676,
+      "learning_rate": 4.938888160376842e-07,
+      "loss": 0.6357,
+      "step": 8188
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.306752775047198,
+      "learning_rate": 4.931505338877363e-07,
+      "loss": 0.4966,
+      "step": 8189
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.467733910306402,
+      "learning_rate": 4.924127753305308e-07,
+      "loss": 0.5819,
+      "step": 8190
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5766576901577807,
+      "learning_rate": 4.916755404517787e-07,
+      "loss": 0.6238,
+      "step": 8191
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.477146410617243,
+      "learning_rate": 4.909388293371309e-07,
+      "loss": 0.5792,
+      "step": 8192
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.725439865413354,
+      "learning_rate": 4.902026420721756e-07,
+      "loss": 0.589,
+      "step": 8193
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.1642328795374106,
+      "learning_rate": 4.894669787424399e-07,
+      "loss": 0.5731,
+      "step": 8194
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7228885313758933,
+      "learning_rate": 4.887318394333923e-07,
+      "loss": 0.6069,
+      "step": 8195
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.187726267604148,
+      "learning_rate": 4.879972242304382e-07,
+      "loss": 0.5925,
+      "step": 8196
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4064432279219896,
+      "learning_rate": 4.872631332189259e-07,
+      "loss": 0.6251,
+      "step": 8197
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0589344787984993,
+      "learning_rate": 4.865295664841363e-07,
+      "loss": 0.5597,
+      "step": 8198
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3406278362981645,
+      "learning_rate": 4.857965241112938e-07,
+      "loss": 0.5722,
+      "step": 8199
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.988444067366749,
+      "learning_rate": 4.850640061855627e-07,
+      "loss": 0.6008,
+      "step": 8200
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7834751537926765,
+      "learning_rate": 4.843320127920442e-07,
+      "loss": 0.5944,
+      "step": 8201
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5609186141419173,
+      "learning_rate": 4.836005440157798e-07,
+      "loss": 0.6557,
+      "step": 8202
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3038313600115945,
+      "learning_rate": 4.828695999417471e-07,
+      "loss": 0.6159,
+      "step": 8203
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3290658242492674,
+      "learning_rate": 4.821391806548664e-07,
+      "loss": 0.6091,
+      "step": 8204
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7246422374040185,
+      "learning_rate": 4.814092862399971e-07,
+      "loss": 0.6781,
+      "step": 8205
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.543683504573766,
+      "learning_rate": 4.806799167819354e-07,
+      "loss": 0.5268,
+      "step": 8206
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.248828186134925,
+      "learning_rate": 4.799510723654154e-07,
+      "loss": 0.5292,
+      "step": 8207
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.487968354722907,
+      "learning_rate": 4.792227530751137e-07,
+      "loss": 0.618,
+      "step": 8208
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.112110087489705,
+      "learning_rate": 4.784949589956444e-07,
+      "loss": 0.6297,
+      "step": 8209
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.553846046938456,
+      "learning_rate": 4.777676902115613e-07,
+      "loss": 0.6159,
+      "step": 8210
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8837009978624635,
+      "learning_rate": 4.770409468073562e-07,
+      "loss": 0.6905,
+      "step": 8211
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.1032584634222546,
+      "learning_rate": 4.7631472886745746e-07,
+      "loss": 0.6244,
+      "step": 8212
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.367855236507304,
+      "learning_rate": 4.755890364762372e-07,
+      "loss": 0.5847,
+      "step": 8213
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9490535768407505,
+      "learning_rate": 4.748638697180052e-07,
+      "loss": 0.6026,
+      "step": 8214
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.6375758139751895,
+      "learning_rate": 4.741392286770075e-07,
+      "loss": 0.5925,
+      "step": 8215
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8159200769119366,
+      "learning_rate": 4.734151134374304e-07,
+      "loss": 0.5985,
+      "step": 8216
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.1013992053905923,
+      "learning_rate": 4.7269152408340067e-07,
+      "loss": 0.718,
+      "step": 8217
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2854215939813325,
+      "learning_rate": 4.7196846069898216e-07,
+      "loss": 0.5956,
+      "step": 8218
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3465869398994483,
+      "learning_rate": 4.71245923368181e-07,
+      "loss": 0.535,
+      "step": 8219
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.521050340647012,
+      "learning_rate": 4.7052391217493497e-07,
+      "loss": 0.596,
+      "step": 8220
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4595843214175392,
+      "learning_rate": 4.698024272031276e-07,
+      "loss": 0.5647,
+      "step": 8221
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.029712654205638,
+      "learning_rate": 4.690814685365791e-07,
+      "loss": 0.6448,
+      "step": 8222
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.341790414077542,
+      "learning_rate": 4.683610362590485e-07,
+      "loss": 0.5764,
+      "step": 8223
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.5428852112483105,
+      "learning_rate": 4.6764113045423274e-07,
+      "loss": 0.6887,
+      "step": 8224
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1809278528567937,
+      "learning_rate": 4.6692175120576834e-07,
+      "loss": 0.6309,
+      "step": 8225
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.412349436289653,
+      "learning_rate": 4.6620289859723114e-07,
+      "loss": 0.583,
+      "step": 8226
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6851844120083275,
+      "learning_rate": 4.65484572712136e-07,
+      "loss": 0.5324,
+      "step": 8227
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.350365679203484,
+      "learning_rate": 4.6476677363393507e-07,
+      "loss": 0.6305,
+      "step": 8228
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.74850582630496,
+      "learning_rate": 4.6404950144602e-07,
+      "loss": 0.6442,
+      "step": 8229
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.48312204654485,
+      "learning_rate": 4.6333275623172137e-07,
+      "loss": 0.6308,
+      "step": 8230
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.447477016671054,
+      "learning_rate": 4.626165380743086e-07,
+      "loss": 0.5315,
+      "step": 8231
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6758503396323916,
+      "learning_rate": 4.6190084705699243e-07,
+      "loss": 0.6075,
+      "step": 8232
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.7690378845704013,
+      "learning_rate": 4.6118568326291577e-07,
+      "loss": 0.5348,
+      "step": 8233
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.288950693928659,
+      "learning_rate": 4.604710467751661e-07,
+      "loss": 0.5712,
+      "step": 8234
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.882335822482473,
+      "learning_rate": 4.5975693767676746e-07,
+      "loss": 0.5979,
+      "step": 8235
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.0747373794610313,
+      "learning_rate": 4.590433560506841e-07,
+      "loss": 0.6,
+      "step": 8236
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6607131094292322,
+      "learning_rate": 4.583303019798174e-07,
+      "loss": 0.6083,
+      "step": 8237
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.6597203389257147,
+      "learning_rate": 4.576177755470068e-07,
+      "loss": 0.6753,
+      "step": 8238
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.4666690631206585,
+      "learning_rate": 4.5690577683503214e-07,
+      "loss": 0.5823,
+      "step": 8239
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.426851608730652,
+      "learning_rate": 4.561943059266122e-07,
+      "loss": 0.6137,
+      "step": 8240
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.5313147718873426,
+      "learning_rate": 4.554833629044031e-07,
+      "loss": 0.5988,
+      "step": 8241
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4878565766187477,
+      "learning_rate": 4.547729478509993e-07,
+      "loss": 0.6371,
+      "step": 8242
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 12.485806184499575,
+      "learning_rate": 4.540630608489355e-07,
+      "loss": 0.5358,
+      "step": 8243
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.51891715467682,
+      "learning_rate": 4.533537019806844e-07,
+      "loss": 0.5765,
+      "step": 8244
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.680356743825882,
+      "learning_rate": 4.52644871328658e-07,
+      "loss": 0.6421,
+      "step": 8245
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0197526172225977,
+      "learning_rate": 4.5193656897520534e-07,
+      "loss": 0.5282,
+      "step": 8246
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.456701500945603,
+      "learning_rate": 4.5122879500261396e-07,
+      "loss": 0.5945,
+      "step": 8247
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.368043410381294,
+      "learning_rate": 4.50521549493112e-07,
+      "loss": 0.6154,
+      "step": 8248
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.207890875247668,
+      "learning_rate": 4.498148325288665e-07,
+      "loss": 0.6075,
+      "step": 8249
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2518464473290987,
+      "learning_rate": 4.491086441919801e-07,
+      "loss": 0.6067,
+      "step": 8250
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2804388518048135,
+      "learning_rate": 4.484029845644955e-07,
+      "loss": 0.6258,
+      "step": 8251
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0120465903949816,
+      "learning_rate": 4.4769785372839493e-07,
+      "loss": 0.533,
+      "step": 8252
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.7135204524309904,
+      "learning_rate": 4.469932517655978e-07,
+      "loss": 0.5617,
+      "step": 8253
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.9167866927413715,
+      "learning_rate": 4.462891787579654e-07,
+      "loss": 0.6386,
+      "step": 8254
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0215478723122295,
+      "learning_rate": 4.4558563478729113e-07,
+      "loss": 0.5989,
+      "step": 8255
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.910071063555147,
+      "learning_rate": 4.4488261993531233e-07,
+      "loss": 0.5422,
+      "step": 8256
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 11.796143039706289,
+      "learning_rate": 4.441801342837027e-07,
+      "loss": 0.502,
+      "step": 8257
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.342733719631572,
+      "learning_rate": 4.4347817791407677e-07,
+      "loss": 0.567,
+      "step": 8258
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.3267678983826565,
+      "learning_rate": 4.4277675090798445e-07,
+      "loss": 0.5816,
+      "step": 8259
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.157334595406685,
+      "learning_rate": 4.4207585334691493e-07,
+      "loss": 0.5783,
+      "step": 8260
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.343720857313059,
+      "learning_rate": 4.41375485312297e-07,
+      "loss": 0.5595,
+      "step": 8261
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.109175745260641,
+      "learning_rate": 4.406756468854989e-07,
+      "loss": 0.5158,
+      "step": 8262
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3196868345115775,
+      "learning_rate": 4.3997633814782393e-07,
+      "loss": 0.5602,
+      "step": 8263
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.7303164750460387,
+      "learning_rate": 4.392775591805154e-07,
+      "loss": 0.5038,
+      "step": 8264
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.74241779032744,
+      "learning_rate": 4.385793100647567e-07,
+      "loss": 0.6497,
+      "step": 8265
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.430609319723213,
+      "learning_rate": 4.378815908816675e-07,
+      "loss": 0.5677,
+      "step": 8266
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.7793948165517435,
+      "learning_rate": 4.371844017123095e-07,
+      "loss": 0.5426,
+      "step": 8267
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.4548513277509145,
+      "learning_rate": 4.3648774263767624e-07,
+      "loss": 0.5881,
+      "step": 8268
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2805067121865363,
+      "learning_rate": 4.3579161373870526e-07,
+      "loss": 0.6534,
+      "step": 8269
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.521977095171458,
+      "learning_rate": 4.350960150962702e-07,
+      "loss": 0.5821,
+      "step": 8270
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.045332855637481,
+      "learning_rate": 4.344009467911858e-07,
+      "loss": 0.5343,
+      "step": 8271
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.071925089409981,
+      "learning_rate": 4.3370640890420145e-07,
+      "loss": 0.5714,
+      "step": 8272
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3770918532643552,
+      "learning_rate": 4.3301240151600587e-07,
+      "loss": 0.6368,
+      "step": 8273
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0377462716813706,
+      "learning_rate": 4.3231892470722736e-07,
+      "loss": 0.5848,
+      "step": 8274
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3436477279860513,
+      "learning_rate": 4.316259785584337e-07,
+      "loss": 0.524,
+      "step": 8275
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.7364690734810475,
+      "learning_rate": 4.309335631501277e-07,
+      "loss": 0.635,
+      "step": 8276
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5489141180850075,
+      "learning_rate": 4.3024167856275166e-07,
+      "loss": 0.5364,
+      "step": 8277
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.3087010035902105,
+      "learning_rate": 4.2955032487668745e-07,
+      "loss": 0.5727,
+      "step": 8278
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3875544612935014,
+      "learning_rate": 4.2885950217225525e-07,
+      "loss": 0.573,
+      "step": 8279
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.997641790782197,
+      "learning_rate": 4.281692105297125e-07,
+      "loss": 0.5424,
+      "step": 8280
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.5825231781643305,
+      "learning_rate": 4.2747945002925507e-07,
+      "loss": 0.604,
+      "step": 8281
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.751380772733463,
+      "learning_rate": 4.267902207510166e-07,
+      "loss": 0.6387,
+      "step": 8282
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.7858717540005844,
+      "learning_rate": 4.261015227750709e-07,
+      "loss": 0.5677,
+      "step": 8283
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.001147795122434,
+      "learning_rate": 4.254133561814289e-07,
+      "loss": 0.5422,
+      "step": 8284
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4257602264136264,
+      "learning_rate": 4.247257210500394e-07,
+      "loss": 0.5679,
+      "step": 8285
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.98287549752436,
+      "learning_rate": 4.240386174607891e-07,
+      "loss": 0.5719,
+      "step": 8286
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0556012748341397,
+      "learning_rate": 4.2335204549350415e-07,
+      "loss": 0.5993,
+      "step": 8287
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1184763996018456,
+      "learning_rate": 4.226660052279491e-07,
+      "loss": 0.5858,
+      "step": 8288
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4200580318072165,
+      "learning_rate": 4.219804967438279e-07,
+      "loss": 0.5608,
+      "step": 8289
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2140173131580885,
+      "learning_rate": 4.2129552012077636e-07,
+      "loss": 0.577,
+      "step": 8290
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8022432950471288,
+      "learning_rate": 4.2061107543837633e-07,
+      "loss": 0.5129,
+      "step": 8291
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3001424430333746,
+      "learning_rate": 4.1992716277614365e-07,
+      "loss": 0.5429,
+      "step": 8292
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0628372093439378,
+      "learning_rate": 4.1924378221353425e-07,
+      "loss": 0.5649,
+      "step": 8293
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.0938803545878013,
+      "learning_rate": 4.185609338299407e-07,
+      "loss": 0.6533,
+      "step": 8294
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.2890771263913123,
+      "learning_rate": 4.178786177046934e-07,
+      "loss": 0.633,
+      "step": 8295
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1087154492100217,
+      "learning_rate": 4.1719683391706235e-07,
+      "loss": 0.5208,
+      "step": 8296
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.20130117340025,
+      "learning_rate": 4.165155825462569e-07,
+      "loss": 0.66,
+      "step": 8297
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.38259009410489,
+      "learning_rate": 4.158348636714216e-07,
+      "loss": 0.698,
+      "step": 8298
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.955775034898537,
+      "learning_rate": 4.151546773716392e-07,
+      "loss": 0.6124,
+      "step": 8299
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.598068006883137,
+      "learning_rate": 4.1447502372593316e-07,
+      "loss": 0.5196,
+      "step": 8300
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1646914007025493,
+      "learning_rate": 4.137959028132632e-07,
+      "loss": 0.631,
+      "step": 8301
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3381901988980673,
+      "learning_rate": 4.1311731471253e-07,
+      "loss": 0.5472,
+      "step": 8302
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.41287496079573,
+      "learning_rate": 4.1243925950256616e-07,
+      "loss": 0.6113,
+      "step": 8303
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.709280007140201,
+      "learning_rate": 4.117617372621474e-07,
+      "loss": 0.5998,
+      "step": 8304
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2689051954483386,
+      "learning_rate": 4.11084748069987e-07,
+      "loss": 0.6169,
+      "step": 8305
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3500449701909503,
+      "learning_rate": 4.1040829200473643e-07,
+      "loss": 0.6554,
+      "step": 8306
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4784819446542996,
+      "learning_rate": 4.0973236914498284e-07,
+      "loss": 0.6293,
+      "step": 8307
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.513599880145634,
+      "learning_rate": 4.090569795692528e-07,
+      "loss": 0.5934,
+      "step": 8308
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8994935140490363,
+      "learning_rate": 4.08382123356012e-07,
+      "loss": 0.6651,
+      "step": 8309
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.715503190088068,
+      "learning_rate": 4.077078005836638e-07,
+      "loss": 0.5931,
+      "step": 8310
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8979413858053675,
+      "learning_rate": 4.070340113305482e-07,
+      "loss": 0.6289,
+      "step": 8311
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.316723946930344,
+      "learning_rate": 4.0636075567494384e-07,
+      "loss": 0.6424,
+      "step": 8312
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6873395112200575,
+      "learning_rate": 4.056880336950675e-07,
+      "loss": 0.6267,
+      "step": 8313
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.2114274473336453,
+      "learning_rate": 4.05015845469075e-07,
+      "loss": 0.6438,
+      "step": 8314
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1479506466214247,
+      "learning_rate": 4.043441910750595e-07,
+      "loss": 0.6171,
+      "step": 8315
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.941279814979407,
+      "learning_rate": 4.036730705910513e-07,
+      "loss": 0.6145,
+      "step": 8316
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.467418133816867,
+      "learning_rate": 4.030024840950181e-07,
+      "loss": 0.592,
+      "step": 8317
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.429232092167091,
+      "learning_rate": 4.0233243166486804e-07,
+      "loss": 0.5985,
+      "step": 8318
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.965427954109061,
+      "learning_rate": 4.016629133784461e-07,
+      "loss": 0.6265,
+      "step": 8319
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.148171593670399,
+      "learning_rate": 4.0099392931353454e-07,
+      "loss": 0.5878,
+      "step": 8320
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.435832787059743,
+      "learning_rate": 4.0032547954785286e-07,
+      "loss": 0.6319,
+      "step": 8321
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6890143714422567,
+      "learning_rate": 3.996575641590611e-07,
+      "loss": 0.5508,
+      "step": 8322
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2991123215639337,
+      "learning_rate": 3.9899018322475503e-07,
+      "loss": 0.5276,
+      "step": 8323
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.492027408747583,
+      "learning_rate": 3.983233368224709e-07,
+      "loss": 0.639,
+      "step": 8324
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.436217435669819,
+      "learning_rate": 3.9765702502967795e-07,
+      "loss": 0.5844,
+      "step": 8325
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2735941889345,
+      "learning_rate": 3.969912479237875e-07,
+      "loss": 0.5295,
+      "step": 8326
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4581911213333942,
+      "learning_rate": 3.963260055821477e-07,
+      "loss": 0.5177,
+      "step": 8327
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4753926504162496,
+      "learning_rate": 3.9566129808204624e-07,
+      "loss": 0.6626,
+      "step": 8328
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.324542564189604,
+      "learning_rate": 3.9499712550070513e-07,
+      "loss": 0.6273,
+      "step": 8329
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.14806628673995,
+      "learning_rate": 3.943334879152849e-07,
+      "loss": 0.6657,
+      "step": 8330
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2230261640469764,
+      "learning_rate": 3.936703854028873e-07,
+      "loss": 0.5797,
+      "step": 8331
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3480123532810095,
+      "learning_rate": 3.9300781804054887e-07,
+      "loss": 0.5624,
+      "step": 8332
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.982281191390795,
+      "learning_rate": 3.9234578590524486e-07,
+      "loss": 0.4974,
+      "step": 8333
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5837995450453755,
+      "learning_rate": 3.9168428907388755e-07,
+      "loss": 0.6192,
+      "step": 8334
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.346120506226917,
+      "learning_rate": 3.9102332762332775e-07,
+      "loss": 0.5529,
+      "step": 8335
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.9988359051893787,
+      "learning_rate": 3.903629016303551e-07,
+      "loss": 0.6231,
+      "step": 8336
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.929097624377366,
+      "learning_rate": 3.897030111716971e-07,
+      "loss": 0.5565,
+      "step": 8337
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.490972492353745,
+      "learning_rate": 3.890436563240141e-07,
+      "loss": 0.5473,
+      "step": 8338
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.123463476780392,
+      "learning_rate": 3.883848371639104e-07,
+      "loss": 0.6492,
+      "step": 8339
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1499837297094007,
+      "learning_rate": 3.8772655376792535e-07,
+      "loss": 0.5906,
+      "step": 8340
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.667246941203882,
+      "learning_rate": 3.870688062125377e-07,
+      "loss": 0.6599,
+      "step": 8341
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3272993650955405,
+      "learning_rate": 3.864115945741609e-07,
+      "loss": 0.601,
+      "step": 8342
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.81091418829116,
+      "learning_rate": 3.8575491892914816e-07,
+      "loss": 0.6082,
+      "step": 8343
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9611755295529303,
+      "learning_rate": 3.8509877935379083e-07,
+      "loss": 0.5979,
+      "step": 8344
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.61950855812886,
+      "learning_rate": 3.8444317592431724e-07,
+      "loss": 0.6358,
+      "step": 8345
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9406115494374441,
+      "learning_rate": 3.837881087168932e-07,
+      "loss": 0.4899,
+      "step": 8346
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.38460628767397,
+      "learning_rate": 3.8313357780762227e-07,
+      "loss": 0.6332,
+      "step": 8347
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7558226853136016,
+      "learning_rate": 3.8247958327254586e-07,
+      "loss": 0.5917,
+      "step": 8348
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5358732790727547,
+      "learning_rate": 3.8182612518764374e-07,
+      "loss": 0.545,
+      "step": 8349
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3595312762742418,
+      "learning_rate": 3.811732036288335e-07,
+      "loss": 0.608,
+      "step": 8350
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7712731901406173,
+      "learning_rate": 3.805208186719689e-07,
+      "loss": 0.6207,
+      "step": 8351
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9142414884157021,
+      "learning_rate": 3.7986897039284043e-07,
+      "loss": 0.5117,
+      "step": 8352
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.316330933057498,
+      "learning_rate": 3.792176588671803e-07,
+      "loss": 0.6112,
+      "step": 8353
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.9741114811382747,
+      "learning_rate": 3.785668841706558e-07,
+      "loss": 0.6341,
+      "step": 8354
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0963466583939874,
+      "learning_rate": 3.7791664637887137e-07,
+      "loss": 0.5715,
+      "step": 8355
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.1371953655685396,
+      "learning_rate": 3.7726694556736943e-07,
+      "loss": 0.4608,
+      "step": 8356
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2052960552688363,
+      "learning_rate": 3.7661778181163067e-07,
+      "loss": 0.5932,
+      "step": 8357
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7273903186451576,
+      "learning_rate": 3.759691551870737e-07,
+      "loss": 0.6266,
+      "step": 8358
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5969037148869796,
+      "learning_rate": 3.753210657690537e-07,
+      "loss": 0.5756,
+      "step": 8359
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5347492281922293,
+      "learning_rate": 3.746735136328633e-07,
+      "loss": 0.6715,
+      "step": 8360
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.8106795258984225,
+      "learning_rate": 3.740264988537329e-07,
+      "loss": 0.5104,
+      "step": 8361
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.00381965510081,
+      "learning_rate": 3.7338002150683174e-07,
+      "loss": 0.6463,
+      "step": 8362
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.173783144313059,
+      "learning_rate": 3.727340816672664e-07,
+      "loss": 0.6332,
+      "step": 8363
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.9002556770934373,
+      "learning_rate": 3.7208867941007974e-07,
+      "loss": 0.6431,
+      "step": 8364
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7468354426920008,
+      "learning_rate": 3.7144381481025114e-07,
+      "loss": 0.5672,
+      "step": 8365
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4532097840078,
+      "learning_rate": 3.707994879427007e-07,
+      "loss": 0.6065,
+      "step": 8366
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.664239854523862,
+      "learning_rate": 3.7015569888228464e-07,
+      "loss": 0.5785,
+      "step": 8367
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.1037887455483544,
+      "learning_rate": 3.6951244770379593e-07,
+      "loss": 0.5904,
+      "step": 8368
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4768409598228636,
+      "learning_rate": 3.6886973448196475e-07,
+      "loss": 0.5523,
+      "step": 8369
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7756714612946563,
+      "learning_rate": 3.682275592914608e-07,
+      "loss": 0.6883,
+      "step": 8370
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9591826331531919,
+      "learning_rate": 3.675859222068895e-07,
+      "loss": 0.5301,
+      "step": 8371
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9503987597108043,
+      "learning_rate": 3.669448233027967e-07,
+      "loss": 0.5432,
+      "step": 8372
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2208818742361625,
+      "learning_rate": 3.6630426265366003e-07,
+      "loss": 0.5948,
+      "step": 8373
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9397750897472155,
+      "learning_rate": 3.6566424033389947e-07,
+      "loss": 0.5438,
+      "step": 8374
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.765113073654002,
+      "learning_rate": 3.6502475641787107e-07,
+      "loss": 0.6899,
+      "step": 8375
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.389118275586379,
+      "learning_rate": 3.6438581097986867e-07,
+      "loss": 0.7196,
+      "step": 8376
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1693703595965332,
+      "learning_rate": 3.637474040941225e-07,
+      "loss": 0.5825,
+      "step": 8377
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.064136002731287,
+      "learning_rate": 3.6310953583480024e-07,
+      "loss": 0.6088,
+      "step": 8378
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.293347710145248,
+      "learning_rate": 3.6247220627600833e-07,
+      "loss": 0.6058,
+      "step": 8379
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6363547175045237,
+      "learning_rate": 3.6183541549179025e-07,
+      "loss": 0.5608,
+      "step": 8380
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2724907939257672,
+      "learning_rate": 3.6119916355612627e-07,
+      "loss": 0.5671,
+      "step": 8381
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.166605602109891,
+      "learning_rate": 3.6056345054293283e-07,
+      "loss": 0.6426,
+      "step": 8382
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.242029102578855,
+      "learning_rate": 3.59928276526067e-07,
+      "loss": 0.5707,
+      "step": 8383
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7693849737510656,
+      "learning_rate": 3.592936415793208e-07,
+      "loss": 0.6228,
+      "step": 8384
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.104105928630141,
+      "learning_rate": 3.586595457764247e-07,
+      "loss": 0.5438,
+      "step": 8385
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.368575714906391,
+      "learning_rate": 3.580259891910465e-07,
+      "loss": 0.6578,
+      "step": 8386
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.8815306378727747,
+      "learning_rate": 3.573929718967889e-07,
+      "loss": 0.6015,
+      "step": 8387
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2975996264972705,
+      "learning_rate": 3.567604939671959e-07,
+      "loss": 0.6111,
+      "step": 8388
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.8232534667191658,
+      "learning_rate": 3.561285554757471e-07,
+      "loss": 0.567,
+      "step": 8389
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.616180660257726,
+      "learning_rate": 3.554971564958587e-07,
+      "loss": 0.5183,
+      "step": 8390
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.4538629426171275,
+      "learning_rate": 3.548662971008837e-07,
+      "loss": 0.6621,
+      "step": 8391
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.688169563692595,
+      "learning_rate": 3.5423597736411463e-07,
+      "loss": 0.6208,
+      "step": 8392
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.685828822376977,
+      "learning_rate": 3.536061973587812e-07,
+      "loss": 0.7017,
+      "step": 8393
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.36466383731313,
+      "learning_rate": 3.5297695715804825e-07,
+      "loss": 0.6273,
+      "step": 8394
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.105461376785732,
+      "learning_rate": 3.523482568350184e-07,
+      "loss": 0.5799,
+      "step": 8395
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.2543324276401413,
+      "learning_rate": 3.517200964627332e-07,
+      "loss": 0.6232,
+      "step": 8396
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.139641724312683,
+      "learning_rate": 3.510924761141704e-07,
+      "loss": 0.6816,
+      "step": 8397
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.1532567391885493,
+      "learning_rate": 3.504653958622456e-07,
+      "loss": 0.5811,
+      "step": 8398
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7533862384085492,
+      "learning_rate": 3.49838855779811e-07,
+      "loss": 0.5851,
+      "step": 8399
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2806278164148184,
+      "learning_rate": 3.492128559396552e-07,
+      "loss": 0.5411,
+      "step": 8400
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.786894398505118,
+      "learning_rate": 3.485873964145053e-07,
+      "loss": 0.5874,
+      "step": 8401
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.223189773086896,
+      "learning_rate": 3.479624772770268e-07,
+      "loss": 0.6098,
+      "step": 8402
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.559630254288751,
+      "learning_rate": 3.4733809859982037e-07,
+      "loss": 0.5536,
+      "step": 8403
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.389108579984836,
+      "learning_rate": 3.46714260455423e-07,
+      "loss": 0.517,
+      "step": 8404
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5393552430851085,
+      "learning_rate": 3.460909629163117e-07,
+      "loss": 0.6811,
+      "step": 8405
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7176090256610452,
+      "learning_rate": 3.4546820605489974e-07,
+      "loss": 0.5495,
+      "step": 8406
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.11043021583628,
+      "learning_rate": 3.448459899435369e-07,
+      "loss": 0.6564,
+      "step": 8407
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5053404923404297,
+      "learning_rate": 3.442243146545093e-07,
+      "loss": 0.5676,
+      "step": 8408
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.246167043873439,
+      "learning_rate": 3.436031802600426e-07,
+      "loss": 0.6413,
+      "step": 8409
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.9518371356414472,
+      "learning_rate": 3.4298258683229836e-07,
+      "loss": 0.6431,
+      "step": 8410
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5332175625722124,
+      "learning_rate": 3.423625344433756e-07,
+      "loss": 0.6081,
+      "step": 8411
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9693277831137703,
+      "learning_rate": 3.417430231653096e-07,
+      "loss": 0.5281,
+      "step": 8412
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1510911382680074,
+      "learning_rate": 3.4112405307007266e-07,
+      "loss": 0.6343,
+      "step": 8413
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.465031585748283,
+      "learning_rate": 3.4050562422957624e-07,
+      "loss": 0.5676,
+      "step": 8414
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.664072247681827,
+      "learning_rate": 3.3988773671566777e-07,
+      "loss": 0.6398,
+      "step": 8415
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.75888208434279,
+      "learning_rate": 3.3927039060013045e-07,
+      "loss": 0.6638,
+      "step": 8416
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.388691306338053,
+      "learning_rate": 3.3865358595468635e-07,
+      "loss": 0.5692,
+      "step": 8417
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0916358970739397,
+      "learning_rate": 3.380373228509937e-07,
+      "loss": 0.5645,
+      "step": 8418
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.745762404593912,
+      "learning_rate": 3.374216013606485e-07,
+      "loss": 0.5756,
+      "step": 8419
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1892760583248885,
+      "learning_rate": 3.368064215551842e-07,
+      "loss": 0.6245,
+      "step": 8420
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.857345470679988,
+      "learning_rate": 3.3619178350607016e-07,
+      "loss": 0.6549,
+      "step": 8421
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.379512536999841,
+      "learning_rate": 3.355776872847122e-07,
+      "loss": 0.5277,
+      "step": 8422
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.321362797949176,
+      "learning_rate": 3.3496413296245536e-07,
+      "loss": 0.6036,
+      "step": 8423
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.245567667052969,
+      "learning_rate": 3.343511206105804e-07,
+      "loss": 0.5352,
+      "step": 8424
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.045437582667061,
+      "learning_rate": 3.3373865030030536e-07,
+      "loss": 0.5313,
+      "step": 8425
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.440444633467484,
+      "learning_rate": 3.331267221027845e-07,
+      "loss": 0.602,
+      "step": 8426
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9346701755245185,
+      "learning_rate": 3.325153360891109e-07,
+      "loss": 0.5728,
+      "step": 8427
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.964299003618511,
+      "learning_rate": 3.319044923303133e-07,
+      "loss": 0.5423,
+      "step": 8428
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.7005210221802867,
+      "learning_rate": 3.3129419089735825e-07,
+      "loss": 0.6404,
+      "step": 8429
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6961944982240307,
+      "learning_rate": 3.306844318611474e-07,
+      "loss": 0.6103,
+      "step": 8430
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.9859468059962073,
+      "learning_rate": 3.300752152925213e-07,
+      "loss": 0.616,
+      "step": 8431
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.782849535399388,
+      "learning_rate": 3.2946654126225776e-07,
+      "loss": 0.6134,
+      "step": 8432
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3605109038050074,
+      "learning_rate": 3.288584098410708e-07,
+      "loss": 0.578,
+      "step": 8433
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.0653052962653273,
+      "learning_rate": 3.282508210996105e-07,
+      "loss": 0.5938,
+      "step": 8434
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.320847704691404,
+      "learning_rate": 3.276437751084649e-07,
+      "loss": 0.6857,
+      "step": 8435
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.171177977533934,
+      "learning_rate": 3.270372719381587e-07,
+      "loss": 0.5834,
+      "step": 8436
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.886349099099324,
+      "learning_rate": 3.264313116591555e-07,
+      "loss": 0.6321,
+      "step": 8437
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.8574070263501765,
+      "learning_rate": 3.2582589434185184e-07,
+      "loss": 0.593,
+      "step": 8438
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9494414704404982,
+      "learning_rate": 3.252210200565842e-07,
+      "loss": 0.5359,
+      "step": 8439
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.1811163106053844,
+      "learning_rate": 3.2461668887362407e-07,
+      "loss": 0.6684,
+      "step": 8440
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.381250101023003,
+      "learning_rate": 3.2401290086318315e-07,
+      "loss": 0.5573,
+      "step": 8441
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4346013172983683,
+      "learning_rate": 3.2340965609540643e-07,
+      "loss": 0.6883,
+      "step": 8442
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3133127483274003,
+      "learning_rate": 3.228069546403767e-07,
+      "loss": 0.5346,
+      "step": 8443
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.184411233315387,
+      "learning_rate": 3.222047965681141e-07,
+      "loss": 0.5935,
+      "step": 8444
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.0278851604452073,
+      "learning_rate": 3.2160318194857655e-07,
+      "loss": 0.57,
+      "step": 8445
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.526714549043835,
+      "learning_rate": 3.210021108516581e-07,
+      "loss": 0.5528,
+      "step": 8446
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.168643812080329,
+      "learning_rate": 3.204015833471885e-07,
+      "loss": 0.5877,
+      "step": 8447
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.058474171768321,
+      "learning_rate": 3.1980159950493526e-07,
+      "loss": 0.6024,
+      "step": 8448
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4688281862366876,
+      "learning_rate": 3.1920215939460263e-07,
+      "loss": 0.5965,
+      "step": 8449
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.718194185512959,
+      "learning_rate": 3.186032630858332e-07,
+      "loss": 0.5933,
+      "step": 8450
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4889377896135856,
+      "learning_rate": 3.180049106482047e-07,
+      "loss": 0.5516,
+      "step": 8451
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 12.775983633424683,
+      "learning_rate": 3.1740710215122985e-07,
+      "loss": 0.5611,
+      "step": 8452
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2721038696367137,
+      "learning_rate": 3.1680983766436244e-07,
+      "loss": 0.6155,
+      "step": 8453
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.1139767872960693,
+      "learning_rate": 3.16213117256991e-07,
+      "loss": 0.5942,
+      "step": 8454
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.630434059537754,
+      "learning_rate": 3.1561694099843885e-07,
+      "loss": 0.6442,
+      "step": 8455
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.577929607649873,
+      "learning_rate": 3.1502130895797066e-07,
+      "loss": 0.6169,
+      "step": 8456
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.8393452527215897,
+      "learning_rate": 3.144262212047833e-07,
+      "loss": 0.6314,
+      "step": 8457
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.113366300161247,
+      "learning_rate": 3.138316778080125e-07,
+      "loss": 0.6043,
+      "step": 8458
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6875673979733223,
+      "learning_rate": 3.1323767883673193e-07,
+      "loss": 0.5829,
+      "step": 8459
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.46249726044324,
+      "learning_rate": 3.1264422435994977e-07,
+      "loss": 0.6048,
+      "step": 8460
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3009561213241656,
+      "learning_rate": 3.120513144466109e-07,
+      "loss": 0.6559,
+      "step": 8461
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.97192574369102,
+      "learning_rate": 3.114589491655989e-07,
+      "loss": 0.5194,
+      "step": 8462
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9106568447644855,
+      "learning_rate": 3.1086712858573396e-07,
+      "loss": 0.546,
+      "step": 8463
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.8679076537404593,
+      "learning_rate": 3.10275852775771e-07,
+      "loss": 0.6296,
+      "step": 8464
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.538688766180127,
+      "learning_rate": 3.0968512180440225e-07,
+      "loss": 0.5939,
+      "step": 8465
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.185856983077277,
+      "learning_rate": 3.090949357402573e-07,
+      "loss": 0.5949,
+      "step": 8466
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1121250006231236,
+      "learning_rate": 3.0850529465190295e-07,
+      "loss": 0.6447,
+      "step": 8467
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.842408960010343,
+      "learning_rate": 3.079161986078427e-07,
+      "loss": 0.7068,
+      "step": 8468
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2491385619430364,
+      "learning_rate": 3.073276476765147e-07,
+      "loss": 0.6113,
+      "step": 8469
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.786879233199517,
+      "learning_rate": 3.0673964192629466e-07,
+      "loss": 0.5902,
+      "step": 8470
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.7520700236563482,
+      "learning_rate": 3.061521814254964e-07,
+      "loss": 0.6111,
+      "step": 8471
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.037239255860453,
+      "learning_rate": 3.0556526624237025e-07,
+      "loss": 0.5664,
+      "step": 8472
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6847363520434486,
+      "learning_rate": 3.049788964451006e-07,
+      "loss": 0.6382,
+      "step": 8473
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.446565844450911,
+      "learning_rate": 3.043930721018107e-07,
+      "loss": 0.5745,
+      "step": 8474
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.727557086897107,
+      "learning_rate": 3.0380779328055945e-07,
+      "loss": 0.5672,
+      "step": 8475
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.677310040784338,
+      "learning_rate": 3.0322306004934467e-07,
+      "loss": 0.6836,
+      "step": 8476
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 12.712228333925863,
+      "learning_rate": 3.026388724760976e-07,
+      "loss": 0.5878,
+      "step": 8477
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.7959771086323078,
+      "learning_rate": 3.020552306286867e-07,
+      "loss": 0.5857,
+      "step": 8478
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3823323020029257,
+      "learning_rate": 3.0147213457491887e-07,
+      "loss": 0.6022,
+      "step": 8479
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.488707780332215,
+      "learning_rate": 3.0088958438253656e-07,
+      "loss": 0.5982,
+      "step": 8480
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3382438488043396,
+      "learning_rate": 3.00307580119219e-07,
+      "loss": 0.543,
+      "step": 8481
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.034614711972979,
+      "learning_rate": 2.9972612185258155e-07,
+      "loss": 0.5225,
+      "step": 8482
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.30925472166408,
+      "learning_rate": 2.9914520965017515e-07,
+      "loss": 0.6756,
+      "step": 8483
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0845217256514865,
+      "learning_rate": 2.985648435794897e-07,
+      "loss": 0.5391,
+      "step": 8484
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0947266192143617,
+      "learning_rate": 2.9798502370795123e-07,
+      "loss": 0.6038,
+      "step": 8485
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.143425501436348,
+      "learning_rate": 2.974057501029204e-07,
+      "loss": 0.6399,
+      "step": 8486
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.9656804898570805,
+      "learning_rate": 2.968270228316944e-07,
+      "loss": 0.5797,
+      "step": 8487
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.1462792820258367,
+      "learning_rate": 2.9624884196151003e-07,
+      "loss": 0.689,
+      "step": 8488
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3954574670266324,
+      "learning_rate": 2.956712075595386e-07,
+      "loss": 0.6849,
+      "step": 8489
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4138440964653873,
+      "learning_rate": 2.950941196928869e-07,
+      "loss": 0.6271,
+      "step": 8490
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0756281900786377,
+      "learning_rate": 2.945175784286003e-07,
+      "loss": 0.4975,
+      "step": 8491
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6184403922706543,
+      "learning_rate": 2.93941583833659e-07,
+      "loss": 0.6096,
+      "step": 8492
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.0901832655249137,
+      "learning_rate": 2.933661359749801e-07,
+      "loss": 0.6738,
+      "step": 8493
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.0034179637882055,
+      "learning_rate": 2.9279123491941895e-07,
+      "loss": 0.5927,
+      "step": 8494
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4572881737070493,
+      "learning_rate": 2.9221688073376497e-07,
+      "loss": 0.5141,
+      "step": 8495
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.101626346769178,
+      "learning_rate": 2.916430734847442e-07,
+      "loss": 0.5403,
+      "step": 8496
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.641326366393812,
+      "learning_rate": 2.910698132390211e-07,
+      "loss": 0.5564,
+      "step": 8497
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.471920861357954,
+      "learning_rate": 2.904971000631951e-07,
+      "loss": 0.6577,
+      "step": 8498
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4266070572352922,
+      "learning_rate": 2.899249340238025e-07,
+      "loss": 0.5911,
+      "step": 8499
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.536560963334578,
+      "learning_rate": 2.893533151873146e-07,
+      "loss": 0.5689,
+      "step": 8500
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6887191325980644,
+      "learning_rate": 2.887822436201415e-07,
+      "loss": 0.634,
+      "step": 8501
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.3999765388204106,
+      "learning_rate": 2.882117193886297e-07,
+      "loss": 0.5534,
+      "step": 8502
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2917807653579576,
+      "learning_rate": 2.8764174255905886e-07,
+      "loss": 0.6365,
+      "step": 8503
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.318274584316322,
+      "learning_rate": 2.870723131976494e-07,
+      "loss": 0.5574,
+      "step": 8504
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.890909130662085,
+      "learning_rate": 2.865034313705539e-07,
+      "loss": 0.5967,
+      "step": 8505
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0997713123306756,
+      "learning_rate": 2.8593509714386456e-07,
+      "loss": 0.5588,
+      "step": 8506
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.617809589685534,
+      "learning_rate": 2.853673105836091e-07,
+      "loss": 0.5374,
+      "step": 8507
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.68605129954623,
+      "learning_rate": 2.8480007175575144e-07,
+      "loss": 0.5403,
+      "step": 8508
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2348488550547074,
+      "learning_rate": 2.842333807261899e-07,
+      "loss": 0.575,
+      "step": 8509
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.701909082687133,
+      "learning_rate": 2.836672375607624e-07,
+      "loss": 0.5603,
+      "step": 8510
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.982475752321344,
+      "learning_rate": 2.831016423252425e-07,
+      "loss": 0.5673,
+      "step": 8511
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.9635176976288626,
+      "learning_rate": 2.825365950853387e-07,
+      "loss": 0.6621,
+      "step": 8512
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 12.14753234673202,
+      "learning_rate": 2.8197209590669573e-07,
+      "loss": 0.5816,
+      "step": 8513
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4088842765153546,
+      "learning_rate": 2.814081448548961e-07,
+      "loss": 0.5515,
+      "step": 8514
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5924726034885115,
+      "learning_rate": 2.8084474199545907e-07,
+      "loss": 0.6995,
+      "step": 8515
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.319648207905272,
+      "learning_rate": 2.802818873938373e-07,
+      "loss": 0.5947,
+      "step": 8516
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.9901729743477024,
+      "learning_rate": 2.79719581115423e-07,
+      "loss": 0.6025,
+      "step": 8517
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1532085821221703,
+      "learning_rate": 2.7915782322554265e-07,
+      "loss": 0.574,
+      "step": 8518
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.386635632359815,
+      "learning_rate": 2.7859661378945966e-07,
+      "loss": 0.5769,
+      "step": 8519
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.495915367797323,
+      "learning_rate": 2.7803595287237416e-07,
+      "loss": 0.6033,
+      "step": 8520
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2480787603907597,
+      "learning_rate": 2.7747584053942236e-07,
+      "loss": 0.5373,
+      "step": 8521
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.755749524388031,
+      "learning_rate": 2.7691627685567545e-07,
+      "loss": 0.6319,
+      "step": 8522
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.8563418049060942,
+      "learning_rate": 2.763572618861421e-07,
+      "loss": 0.6336,
+      "step": 8523
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9803488628787727,
+      "learning_rate": 2.7579879569576805e-07,
+      "loss": 0.5288,
+      "step": 8524
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.158913363300771,
+      "learning_rate": 2.7524087834943257e-07,
+      "loss": 0.5508,
+      "step": 8525
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2990257664705704,
+      "learning_rate": 2.746835099119555e-07,
+      "loss": 0.5405,
+      "step": 8526
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.722742879522637,
+      "learning_rate": 2.7412669044808714e-07,
+      "loss": 0.59,
+      "step": 8527
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.7098085005125347,
+      "learning_rate": 2.7357042002251977e-07,
+      "loss": 0.6599,
+      "step": 8528
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.414604589485784,
+      "learning_rate": 2.730146986998783e-07,
+      "loss": 0.5914,
+      "step": 8529
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.9999875277016037,
+      "learning_rate": 2.7245952654472495e-07,
+      "loss": 0.6089,
+      "step": 8530
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4698756738709617,
+      "learning_rate": 2.7190490362155706e-07,
+      "loss": 0.7083,
+      "step": 8531
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.0071257148376604,
+      "learning_rate": 2.7135082999481033e-07,
+      "loss": 0.6586,
+      "step": 8532
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4350296312759276,
+      "learning_rate": 2.707973057288554e-07,
+      "loss": 0.5523,
+      "step": 8533
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9785381511638156,
+      "learning_rate": 2.7024433088799874e-07,
+      "loss": 0.5391,
+      "step": 8534
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.7213298482636703,
+      "learning_rate": 2.696919055364827e-07,
+      "loss": 0.6217,
+      "step": 8535
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.7426731105858804,
+      "learning_rate": 2.691400297384872e-07,
+      "loss": 0.622,
+      "step": 8536
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.9682834230681583,
+      "learning_rate": 2.6858870355812807e-07,
+      "loss": 0.6019,
+      "step": 8537
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5959681300857804,
+      "learning_rate": 2.6803792705945574e-07,
+      "loss": 0.5257,
+      "step": 8538
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.117284087962713,
+      "learning_rate": 2.674877003064591e-07,
+      "loss": 0.6142,
+      "step": 8539
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1014683352430614,
+      "learning_rate": 2.669380233630603e-07,
+      "loss": 0.6515,
+      "step": 8540
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.665551926914057,
+      "learning_rate": 2.6638889629311994e-07,
+      "loss": 0.5391,
+      "step": 8541
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.222214668746664,
+      "learning_rate": 2.6584031916043476e-07,
+      "loss": 0.6326,
+      "step": 8542
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6557840721881734,
+      "learning_rate": 2.652922920287365e-07,
+      "loss": 0.557,
+      "step": 8543
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4193798372931403,
+      "learning_rate": 2.647448149616921e-07,
+      "loss": 0.5985,
+      "step": 8544
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.689574105729332,
+      "learning_rate": 2.641978880229074e-07,
+      "loss": 0.608,
+      "step": 8545
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.743675378873042,
+      "learning_rate": 2.636515112759225e-07,
+      "loss": 0.5973,
+      "step": 8546
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5656395322683725,
+      "learning_rate": 2.631056847842134e-07,
+      "loss": 0.5468,
+      "step": 8547
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.1810462861004862,
+      "learning_rate": 2.625604086111927e-07,
+      "loss": 0.5399,
+      "step": 8548
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5217278501998113,
+      "learning_rate": 2.620156828202092e-07,
+      "loss": 0.6972,
+      "step": 8549
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4534326162271425,
+      "learning_rate": 2.6147150747454776e-07,
+      "loss": 0.6776,
+      "step": 8550
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.7071917479562653,
+      "learning_rate": 2.609278826374284e-07,
+      "loss": 0.6411,
+      "step": 8551
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.581422750056675,
+      "learning_rate": 2.6038480837200896e-07,
+      "loss": 0.6482,
+      "step": 8552
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.335476941637362,
+      "learning_rate": 2.5984228474138115e-07,
+      "loss": 0.535,
+      "step": 8553
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4914563871633013,
+      "learning_rate": 2.593003118085746e-07,
+      "loss": 0.539,
+      "step": 8554
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.477999239774418,
+      "learning_rate": 2.5875888963655396e-07,
+      "loss": 0.5921,
+      "step": 8555
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.0840194634488522,
+      "learning_rate": 2.582180182882205e-07,
+      "loss": 0.579,
+      "step": 8556
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.206238807030622,
+      "learning_rate": 2.576776978264095e-07,
+      "loss": 0.6189,
+      "step": 8557
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.937734436218125,
+      "learning_rate": 2.5713792831389473e-07,
+      "loss": 0.6302,
+      "step": 8558
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.714921870739376,
+      "learning_rate": 2.565987098133865e-07,
+      "loss": 0.597,
+      "step": 8559
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.5143184938458027,
+      "learning_rate": 2.56060042387527e-07,
+      "loss": 0.6705,
+      "step": 8560
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.617749300044612,
+      "learning_rate": 2.5552192609890004e-07,
+      "loss": 0.6626,
+      "step": 8561
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.422359435875593,
+      "learning_rate": 2.5498436101001946e-07,
+      "loss": 0.5862,
+      "step": 8562
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.87556986453871,
+      "learning_rate": 2.544473471833403e-07,
+      "loss": 0.5421,
+      "step": 8563
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5521477189588757,
+      "learning_rate": 2.5391088468124934e-07,
+      "loss": 0.6127,
+      "step": 8564
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.626589882214039,
+      "learning_rate": 2.533749735660729e-07,
+      "loss": 0.6057,
+      "step": 8565
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.272089311491973,
+      "learning_rate": 2.528396139000705e-07,
+      "loss": 0.5462,
+      "step": 8566
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.497924231522892,
+      "learning_rate": 2.5230480574543914e-07,
+      "loss": 0.6173,
+      "step": 8567
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5946958140611014,
+      "learning_rate": 2.5177054916431186e-07,
+      "loss": 0.5676,
+      "step": 8568
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.622437188040537,
+      "learning_rate": 2.5123684421875627e-07,
+      "loss": 0.611,
+      "step": 8569
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4619075515236317,
+      "learning_rate": 2.507036909707766e-07,
+      "loss": 0.597,
+      "step": 8570
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.130051490960946,
+      "learning_rate": 2.5017108948231284e-07,
+      "loss": 0.571,
+      "step": 8571
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3145413732517084,
+      "learning_rate": 2.4963903981524265e-07,
+      "loss": 0.6216,
+      "step": 8572
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6281633995236575,
+      "learning_rate": 2.4910754203137597e-07,
+      "loss": 0.5933,
+      "step": 8573
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9962463423913068,
+      "learning_rate": 2.4857659619246246e-07,
+      "loss": 0.4788,
+      "step": 8574
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.509203947561383,
+      "learning_rate": 2.4804620236018376e-07,
+      "loss": 0.6344,
+      "step": 8575
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3901552696802058,
+      "learning_rate": 2.475163605961617e-07,
+      "loss": 0.6128,
+      "step": 8576
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.851401165089648,
+      "learning_rate": 2.4698707096195094e-07,
+      "loss": 0.5852,
+      "step": 8577
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.7359676704634306,
+      "learning_rate": 2.4645833351904235e-07,
+      "loss": 0.5583,
+      "step": 8578
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.636845767665651,
+      "learning_rate": 2.4593014832886344e-07,
+      "loss": 0.6295,
+      "step": 8579
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.772427246864486,
+      "learning_rate": 2.4540251545277726e-07,
+      "loss": 0.6169,
+      "step": 8580
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3697955879151564,
+      "learning_rate": 2.448754349520832e-07,
+      "loss": 0.5962,
+      "step": 8581
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.7219924370672866,
+      "learning_rate": 2.4434890688801504e-07,
+      "loss": 0.6225,
+      "step": 8582
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.566350444650246,
+      "learning_rate": 2.4382293132174384e-07,
+      "loss": 0.5551,
+      "step": 8583
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5372826204659056,
+      "learning_rate": 2.4329750831437514e-07,
+      "loss": 0.6557,
+      "step": 8584
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8353964126790068,
+      "learning_rate": 2.427726379269524e-07,
+      "loss": 0.5587,
+      "step": 8585
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1888874583305187,
+      "learning_rate": 2.422483202204523e-07,
+      "loss": 0.692,
+      "step": 8586
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4217812609046585,
+      "learning_rate": 2.417245552557901e-07,
+      "loss": 0.5887,
+      "step": 8587
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.31796627980692,
+      "learning_rate": 2.4120134309381315e-07,
+      "loss": 0.7178,
+      "step": 8588
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.213415629562026,
+      "learning_rate": 2.406786837953079e-07,
+      "loss": 0.5923,
+      "step": 8589
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.941272484421341,
+      "learning_rate": 2.401565774209963e-07,
+      "loss": 0.5444,
+      "step": 8590
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.0359025963461446,
+      "learning_rate": 2.396350240315337e-07,
+      "loss": 0.5498,
+      "step": 8591
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6306647881917153,
+      "learning_rate": 2.391140236875128e-07,
+      "loss": 0.6251,
+      "step": 8592
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6487766622684172,
+      "learning_rate": 2.3859357644946233e-07,
+      "loss": 0.6471,
+      "step": 8593
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.645864611230044,
+      "learning_rate": 2.3807368237784735e-07,
+      "loss": 0.5693,
+      "step": 8594
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.212215219692308,
+      "learning_rate": 2.3755434153306555e-07,
+      "loss": 0.5178,
+      "step": 8595
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5525173270216794,
+      "learning_rate": 2.370355539754543e-07,
+      "loss": 0.5043,
+      "step": 8596
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.973557160116561,
+      "learning_rate": 2.3651731976528314e-07,
+      "loss": 0.5824,
+      "step": 8597
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.368358688574046,
+      "learning_rate": 2.3599963896276113e-07,
+      "loss": 0.6403,
+      "step": 8598
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.541437487586155,
+      "learning_rate": 2.354825116280285e-07,
+      "loss": 0.6646,
+      "step": 8599
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6360131144611247,
+      "learning_rate": 2.3496593782116607e-07,
+      "loss": 0.5936,
+      "step": 8600
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4729013605037418,
+      "learning_rate": 2.3444991760218526e-07,
+      "loss": 0.6654,
+      "step": 8601
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4321075434936485,
+      "learning_rate": 2.3393445103103762e-07,
+      "loss": 0.6181,
+      "step": 8602
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.557292183237137,
+      "learning_rate": 2.334195381676091e-07,
+      "loss": 0.6118,
+      "step": 8603
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7334084120232323,
+      "learning_rate": 2.3290517907171962e-07,
+      "loss": 0.5789,
+      "step": 8604
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.2292498265081755,
+      "learning_rate": 2.3239137380312526e-07,
+      "loss": 0.6254,
+      "step": 8605
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2190579271215136,
+      "learning_rate": 2.3187812242151996e-07,
+      "loss": 0.6202,
+      "step": 8606
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6556837855127435,
+      "learning_rate": 2.3136542498653103e-07,
+      "loss": 0.5697,
+      "step": 8607
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.923481136322459,
+      "learning_rate": 2.308532815577219e-07,
+      "loss": 0.5777,
+      "step": 8608
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1026727157977447,
+      "learning_rate": 2.3034169219459336e-07,
+      "loss": 0.5759,
+      "step": 8609
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2566630684707336,
+      "learning_rate": 2.2983065695657835e-07,
+      "loss": 0.605,
+      "step": 8610
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.0101849840783994,
+      "learning_rate": 2.2932017590304945e-07,
+      "loss": 0.6228,
+      "step": 8611
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.9371051066532257,
+      "learning_rate": 2.2881024909331084e-07,
+      "loss": 0.5492,
+      "step": 8612
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.594191038842626,
+      "learning_rate": 2.2830087658660626e-07,
+      "loss": 0.6286,
+      "step": 8613
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2603072217979463,
+      "learning_rate": 2.2779205844211115e-07,
+      "loss": 0.704,
+      "step": 8614
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8989627424235347,
+      "learning_rate": 2.2728379471893992e-07,
+      "loss": 0.4864,
+      "step": 8615
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 16.729220556087142,
+      "learning_rate": 2.2677608547614195e-07,
+      "loss": 0.5918,
+      "step": 8616
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2488176924383954,
+      "learning_rate": 2.2626893077269952e-07,
+      "loss": 0.582,
+      "step": 8617
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4475963095140343,
+      "learning_rate": 2.2576233066753328e-07,
+      "loss": 0.6073,
+      "step": 8618
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.50017796036233,
+      "learning_rate": 2.252562852194984e-07,
+      "loss": 0.6705,
+      "step": 8619
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0355653284721114,
+      "learning_rate": 2.2475079448738667e-07,
+      "loss": 0.5797,
+      "step": 8620
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.740952195511575,
+      "learning_rate": 2.2424585852992287e-07,
+      "loss": 0.6276,
+      "step": 8621
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.624621427504004,
+      "learning_rate": 2.2374147740577058e-07,
+      "loss": 0.5743,
+      "step": 8622
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4340730079804325,
+      "learning_rate": 2.2323765117352625e-07,
+      "loss": 0.5585,
+      "step": 8623
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1286673679109693,
+      "learning_rate": 2.2273437989172308e-07,
+      "loss": 0.6019,
+      "step": 8624
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.029911691462066,
+      "learning_rate": 2.2223166361883096e-07,
+      "loss": 0.5091,
+      "step": 8625
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.1602463814705746,
+      "learning_rate": 2.217295024132532e-07,
+      "loss": 0.627,
+      "step": 8626
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.55085504048545,
+      "learning_rate": 2.2122789633332808e-07,
+      "loss": 0.6616,
+      "step": 8627
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.3530586363040547,
+      "learning_rate": 2.2072684543733236e-07,
+      "loss": 0.6542,
+      "step": 8628
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.8871688456843403,
+      "learning_rate": 2.2022634978347668e-07,
+      "loss": 0.7099,
+      "step": 8629
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5020918168833344,
+      "learning_rate": 2.197264094299062e-07,
+      "loss": 0.6461,
+      "step": 8630
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.9739866771027983,
+      "learning_rate": 2.192270244347039e-07,
+      "loss": 0.6444,
+      "step": 8631
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.453684270588019,
+      "learning_rate": 2.1872819485588504e-07,
+      "loss": 0.5701,
+      "step": 8632
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.9298917878750905,
+      "learning_rate": 2.1822992075140382e-07,
+      "loss": 0.6017,
+      "step": 8633
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.21769668421678,
+      "learning_rate": 2.177322021791478e-07,
+      "loss": 0.552,
+      "step": 8634
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.333398707424515,
+      "learning_rate": 2.1723503919694022e-07,
+      "loss": 0.5818,
+      "step": 8635
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9600492185451742,
+      "learning_rate": 2.167384318625404e-07,
+      "loss": 0.5513,
+      "step": 8636
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1619655902215342,
+      "learning_rate": 2.1624238023364164e-07,
+      "loss": 0.6071,
+      "step": 8637
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.30009030782602,
+      "learning_rate": 2.1574688436787616e-07,
+      "loss": 0.5676,
+      "step": 8638
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.261626619968899,
+      "learning_rate": 2.152519443228074e-07,
+      "loss": 0.6164,
+      "step": 8639
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5969295591844355,
+      "learning_rate": 2.1475756015593597e-07,
+      "loss": 0.577,
+      "step": 8640
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.8285213912360896,
+      "learning_rate": 2.142637319246982e-07,
+      "loss": 0.5747,
+      "step": 8641
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6663798524618434,
+      "learning_rate": 2.1377045968646648e-07,
+      "loss": 0.6167,
+      "step": 8642
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.875625734126703,
+      "learning_rate": 2.1327774349854669e-07,
+      "loss": 0.642,
+      "step": 8643
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0023913927338757,
+      "learning_rate": 2.1278558341818245e-07,
+      "loss": 0.549,
+      "step": 8644
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.492821927184673,
+      "learning_rate": 2.1229397950254971e-07,
+      "loss": 0.6158,
+      "step": 8645
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.954417059877932,
+      "learning_rate": 2.1180293180876333e-07,
+      "loss": 0.5902,
+      "step": 8646
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4328009320824684,
+      "learning_rate": 2.11312440393871e-07,
+      "loss": 0.6403,
+      "step": 8647
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6647018614968547,
+      "learning_rate": 2.1082250531485658e-07,
+      "loss": 0.5715,
+      "step": 8648
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0060574812970433,
+      "learning_rate": 2.1033312662863902e-07,
+      "loss": 0.5627,
+      "step": 8649
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.641443035192098,
+      "learning_rate": 2.0984430439207337e-07,
+      "loss": 0.5118,
+      "step": 8650
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.9361538566284255,
+      "learning_rate": 2.0935603866194975e-07,
+      "loss": 0.6679,
+      "step": 8651
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.39373289597829,
+      "learning_rate": 2.0886832949499337e-07,
+      "loss": 0.5668,
+      "step": 8652
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.3861532699457686,
+      "learning_rate": 2.083811769478644e-07,
+      "loss": 0.6525,
+      "step": 8653
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.936995441518683,
+      "learning_rate": 2.0789458107715876e-07,
+      "loss": 0.7025,
+      "step": 8654
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.349046763306768,
+      "learning_rate": 2.0740854193940896e-07,
+      "loss": 0.6309,
+      "step": 8655
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2596999928277253,
+      "learning_rate": 2.0692305959107982e-07,
+      "loss": 0.6011,
+      "step": 8656
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9228352521725202,
+      "learning_rate": 2.0643813408857516e-07,
+      "loss": 0.4917,
+      "step": 8657
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6385919262613746,
+      "learning_rate": 2.05953765488231e-07,
+      "loss": 0.5755,
+      "step": 8658
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2137952255821753,
+      "learning_rate": 2.0546995384632008e-07,
+      "loss": 0.6005,
+      "step": 8659
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5803538966845636,
+      "learning_rate": 2.0498669921905024e-07,
+      "loss": 0.7106,
+      "step": 8660
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.9792364157241105,
+      "learning_rate": 2.045040016625649e-07,
+      "loss": 0.5886,
+      "step": 8661
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3666959842671464,
+      "learning_rate": 2.04021861232942e-07,
+      "loss": 0.5773,
+      "step": 8662
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.6824064705468786,
+      "learning_rate": 2.0354027798619557e-07,
+      "loss": 0.6144,
+      "step": 8663
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.435875779491401,
+      "learning_rate": 2.030592519782748e-07,
+      "loss": 0.6403,
+      "step": 8664
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6128038565557805,
+      "learning_rate": 2.0257878326506386e-07,
+      "loss": 0.6337,
+      "step": 8665
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.282841908523947,
+      "learning_rate": 2.020988719023814e-07,
+      "loss": 0.5942,
+      "step": 8666
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.370234830147251,
+      "learning_rate": 2.0161951794598233e-07,
+      "loss": 0.5383,
+      "step": 8667
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1849633361416223,
+      "learning_rate": 2.011407214515576e-07,
+      "loss": 0.6604,
+      "step": 8668
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9975177789546006,
+      "learning_rate": 2.0066248247473108e-07,
+      "loss": 0.6303,
+      "step": 8669
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.522737370998732,
+      "learning_rate": 2.0018480107106496e-07,
+      "loss": 0.5688,
+      "step": 8670
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4993421173578714,
+      "learning_rate": 1.9970767729605268e-07,
+      "loss": 0.5487,
+      "step": 8671
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.527990417897828,
+      "learning_rate": 1.992311112051265e-07,
+      "loss": 0.5992,
+      "step": 8672
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.350490216687445,
+      "learning_rate": 1.9875510285365273e-07,
+      "loss": 0.5894,
+      "step": 8673
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5923844574515673,
+      "learning_rate": 1.9827965229693215e-07,
+      "loss": 0.6643,
+      "step": 8674
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.8259634397156805,
+      "learning_rate": 1.978047595902005e-07,
+      "loss": 0.5955,
+      "step": 8675
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.233787620175592,
+      "learning_rate": 1.973304247886304e-07,
+      "loss": 0.5663,
+      "step": 8676
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.9852555325806147,
+      "learning_rate": 1.9685664794732884e-07,
+      "loss": 0.6144,
+      "step": 8677
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.705513555230461,
+      "learning_rate": 1.963834291213368e-07,
+      "loss": 0.5355,
+      "step": 8678
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.025851027664062,
+      "learning_rate": 1.959107683656325e-07,
+      "loss": 0.5568,
+      "step": 8679
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5826049637506676,
+      "learning_rate": 1.954386657351276e-07,
+      "loss": 0.6396,
+      "step": 8680
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.650551552519513,
+      "learning_rate": 1.9496712128467043e-07,
+      "loss": 0.6257,
+      "step": 8681
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.075805993953122,
+      "learning_rate": 1.9449613506904275e-07,
+      "loss": 0.6136,
+      "step": 8682
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6518292689904186,
+      "learning_rate": 1.9402570714296353e-07,
+      "loss": 0.5471,
+      "step": 8683
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6510281697671734,
+      "learning_rate": 1.9355583756108408e-07,
+      "loss": 0.5416,
+      "step": 8684
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7279535357511,
+      "learning_rate": 1.9308652637799352e-07,
+      "loss": 0.5669,
+      "step": 8685
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.359794373937808,
+      "learning_rate": 1.9261777364821542e-07,
+      "loss": 0.6428,
+      "step": 8686
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9251054541397337,
+      "learning_rate": 1.9214957942620738e-07,
+      "loss": 0.5481,
+      "step": 8687
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.718945841396238,
+      "learning_rate": 1.9168194376636308e-07,
+      "loss": 0.6115,
+      "step": 8688
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0986880391335343,
+      "learning_rate": 1.912148667230107e-07,
+      "loss": 0.5174,
+      "step": 8689
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4561034728267597,
+      "learning_rate": 1.9074834835041523e-07,
+      "loss": 0.6184,
+      "step": 8690
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3130570678521942,
+      "learning_rate": 1.9028238870277383e-07,
+      "loss": 0.5663,
+      "step": 8691
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.233247914946738,
+      "learning_rate": 1.8981698783422154e-07,
+      "loss": 0.5151,
+      "step": 8692
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.142511538639313,
+      "learning_rate": 1.8935214579882622e-07,
+      "loss": 0.6681,
+      "step": 8693
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3532758793115693,
+      "learning_rate": 1.88887862650593e-07,
+      "loss": 0.5731,
+      "step": 8694
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.9857413005512314,
+      "learning_rate": 1.8842413844345986e-07,
+      "loss": 0.5648,
+      "step": 8695
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.9566011801313214,
+      "learning_rate": 1.8796097323130202e-07,
+      "loss": 0.5285,
+      "step": 8696
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.808602535252724,
+      "learning_rate": 1.8749836706792758e-07,
+      "loss": 0.6272,
+      "step": 8697
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.524622740136097,
+      "learning_rate": 1.8703632000708128e-07,
+      "loss": 0.5818,
+      "step": 8698
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 12.660611398706083,
+      "learning_rate": 1.8657483210244298e-07,
+      "loss": 0.6365,
+      "step": 8699
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.387834624346501,
+      "learning_rate": 1.8611390340762647e-07,
+      "loss": 0.6027,
+      "step": 8700
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.142755060247217,
+      "learning_rate": 1.8565353397618057e-07,
+      "loss": 0.6635,
+      "step": 8701
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.134961455919104,
+      "learning_rate": 1.8519372386159028e-07,
+      "loss": 0.5493,
+      "step": 8702
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8210863113852738,
+      "learning_rate": 1.8473447311727567e-07,
+      "loss": 0.6306,
+      "step": 8703
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.286001395669524,
+      "learning_rate": 1.8427578179658957e-07,
+      "loss": 0.5625,
+      "step": 8704
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1320659323325755,
+      "learning_rate": 1.8381764995282269e-07,
+      "loss": 0.6272,
+      "step": 8705
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.786721797593951,
+      "learning_rate": 1.8336007763919916e-07,
+      "loss": 0.5459,
+      "step": 8706
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.9741697703618373,
+      "learning_rate": 1.8290306490887866e-07,
+      "loss": 0.6213,
+      "step": 8707
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.392970288943625,
+      "learning_rate": 1.8244661181495426e-07,
+      "loss": 0.5805,
+      "step": 8708
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9781416987342864,
+      "learning_rate": 1.8199071841045746e-07,
+      "loss": 0.575,
+      "step": 8709
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9655678631701408,
+      "learning_rate": 1.8153538474835086e-07,
+      "loss": 0.6157,
+      "step": 8710
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1680563446589924,
+      "learning_rate": 1.810806108815344e-07,
+      "loss": 0.5988,
+      "step": 8711
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4250575483212824,
+      "learning_rate": 1.80626396862843e-07,
+      "loss": 0.5865,
+      "step": 8712
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1845565551153676,
+      "learning_rate": 1.801727427450445e-07,
+      "loss": 0.6155,
+      "step": 8713
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.311124360025313,
+      "learning_rate": 1.79719648580845e-07,
+      "loss": 0.626,
+      "step": 8714
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9616951924685692,
+      "learning_rate": 1.7926711442288247e-07,
+      "loss": 0.5346,
+      "step": 8715
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.549218920031044,
+      "learning_rate": 1.7881514032373147e-07,
+      "loss": 0.6247,
+      "step": 8716
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8740535822228344,
+      "learning_rate": 1.7836372633590005e-07,
+      "loss": 0.5935,
+      "step": 8717
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.583700320203049,
+      "learning_rate": 1.7791287251183398e-07,
+      "loss": 0.6242,
+      "step": 8718
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.7565126848470003,
+      "learning_rate": 1.7746257890391027e-07,
+      "loss": 0.5967,
+      "step": 8719
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.684123664351955,
+      "learning_rate": 1.7701284556444377e-07,
+      "loss": 0.5584,
+      "step": 8720
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.7777890746410563,
+      "learning_rate": 1.7656367254568374e-07,
+      "loss": 0.6293,
+      "step": 8721
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4453226597465716,
+      "learning_rate": 1.7611505989981293e-07,
+      "loss": 0.6327,
+      "step": 8722
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.0727605055674445,
+      "learning_rate": 1.7566700767894906e-07,
+      "loss": 0.6328,
+      "step": 8723
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.726441601765811,
+      "learning_rate": 1.7521951593514718e-07,
+      "loss": 0.5472,
+      "step": 8724
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.130050707112883,
+      "learning_rate": 1.7477258472039517e-07,
+      "loss": 0.5057,
+      "step": 8725
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2032936513711974,
+      "learning_rate": 1.7432621408661532e-07,
+      "loss": 0.5802,
+      "step": 8726
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.577694971816813,
+      "learning_rate": 1.7388040408566674e-07,
+      "loss": 0.6737,
+      "step": 8727
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.571554032274808,
+      "learning_rate": 1.7343515476934136e-07,
+      "loss": 0.5762,
+      "step": 8728
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2675681775342817,
+      "learning_rate": 1.729904661893683e-07,
+      "loss": 0.5929,
+      "step": 8729
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4825069172397907,
+      "learning_rate": 1.725463383974091e-07,
+      "loss": 0.5898,
+      "step": 8730
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9735035948128292,
+      "learning_rate": 1.7210277144506182e-07,
+      "loss": 0.5335,
+      "step": 8731
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.631721683378093,
+      "learning_rate": 1.7165976538385753e-07,
+      "loss": 0.5517,
+      "step": 8732
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0076500063785254,
+      "learning_rate": 1.7121732026526506e-07,
+      "loss": 0.5407,
+      "step": 8733
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.581508418879825,
+      "learning_rate": 1.7077543614068604e-07,
+      "loss": 0.6803,
+      "step": 8734
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4861781722789957,
+      "learning_rate": 1.703341130614572e-07,
+      "loss": 0.6355,
+      "step": 8735
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.515627205530569,
+      "learning_rate": 1.6989335107884863e-07,
+      "loss": 0.6033,
+      "step": 8736
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3777058580194343,
+      "learning_rate": 1.6945315024406883e-07,
+      "loss": 0.513,
+      "step": 8737
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8876260556728341,
+      "learning_rate": 1.6901351060825854e-07,
+      "loss": 0.5056,
+      "step": 8738
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.7599871753116862,
+      "learning_rate": 1.685744322224936e-07,
+      "loss": 0.5685,
+      "step": 8739
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.2326213858288804,
+      "learning_rate": 1.681359151377848e-07,
+      "loss": 0.5538,
+      "step": 8740
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.901673864970551,
+      "learning_rate": 1.676979594050776e-07,
+      "loss": 0.6152,
+      "step": 8741
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8826659096790682,
+      "learning_rate": 1.6726056507525347e-07,
+      "loss": 0.5152,
+      "step": 8742
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.665982233334478,
+      "learning_rate": 1.668237321991262e-07,
+      "loss": 0.5803,
+      "step": 8743
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5198209538212146,
+      "learning_rate": 1.6638746082744684e-07,
+      "loss": 0.6382,
+      "step": 8744
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.568498917912977,
+      "learning_rate": 1.6595175101089877e-07,
+      "loss": 0.6567,
+      "step": 8745
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.409410318358604,
+      "learning_rate": 1.6551660280010316e-07,
+      "loss": 0.6195,
+      "step": 8746
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.719189581656458,
+      "learning_rate": 1.6508201624561404e-07,
+      "loss": 0.6318,
+      "step": 8747
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.214568272301674,
+      "learning_rate": 1.6464799139791877e-07,
+      "loss": 0.5042,
+      "step": 8748
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.086249960723928,
+      "learning_rate": 1.6421452830744366e-07,
+      "loss": 0.6093,
+      "step": 8749
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.992792950152544,
+      "learning_rate": 1.6378162702454458e-07,
+      "loss": 0.5714,
+      "step": 8750
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.417350418321379,
+      "learning_rate": 1.6334928759951684e-07,
+      "loss": 0.6329,
+      "step": 8751
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3306711526974686,
+      "learning_rate": 1.6291751008258693e-07,
+      "loss": 0.5744,
+      "step": 8752
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8025552214715774,
+      "learning_rate": 1.6248629452391862e-07,
+      "loss": 0.5275,
+      "step": 8753
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.15974372047649,
+      "learning_rate": 1.620556409736085e-07,
+      "loss": 0.6049,
+      "step": 8754
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.655812836017327,
+      "learning_rate": 1.616255494816893e-07,
+      "loss": 0.5968,
+      "step": 8755
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.7698264187396346,
+      "learning_rate": 1.6119602009812663e-07,
+      "loss": 0.5531,
+      "step": 8756
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.424978511207146,
+      "learning_rate": 1.6076705287282336e-07,
+      "loss": 0.6074,
+      "step": 8757
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5086511137119807,
+      "learning_rate": 1.6033864785561515e-07,
+      "loss": 0.567,
+      "step": 8758
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.707797137818906,
+      "learning_rate": 1.5991080509627222e-07,
+      "loss": 0.6671,
+      "step": 8759
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3155332443393775,
+      "learning_rate": 1.5948352464450146e-07,
+      "loss": 0.5503,
+      "step": 8760
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.2510331858276005,
+      "learning_rate": 1.590568065499415e-07,
+      "loss": 0.6343,
+      "step": 8761
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6772676116951213,
+      "learning_rate": 1.5863065086216878e-07,
+      "loss": 0.5982,
+      "step": 8762
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.394546302211017,
+      "learning_rate": 1.582050576306915e-07,
+      "loss": 0.6014,
+      "step": 8763
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0972459059941415,
+      "learning_rate": 1.5778002690495453e-07,
+      "loss": 0.5364,
+      "step": 8764
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.9104609457881114,
+      "learning_rate": 1.5735555873433673e-07,
+      "loss": 0.5498,
+      "step": 8765
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.1530370674231816,
+      "learning_rate": 1.569316531681514e-07,
+      "loss": 0.6073,
+      "step": 8766
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.348747898326347,
+      "learning_rate": 1.565083102556464e-07,
+      "loss": 0.5991,
+      "step": 8767
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.64211735957492,
+      "learning_rate": 1.560855300460057e-07,
+      "loss": 0.5962,
+      "step": 8768
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5983785487904294,
+      "learning_rate": 1.5566331258834498e-07,
+      "loss": 0.6021,
+      "step": 8769
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.48505144340559,
+      "learning_rate": 1.552416579317173e-07,
+      "loss": 0.5457,
+      "step": 8770
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.118622729954412,
+      "learning_rate": 1.5482056612510898e-07,
+      "loss": 0.5525,
+      "step": 8771
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.453328716580977,
+      "learning_rate": 1.544000372174409e-07,
+      "loss": 0.5662,
+      "step": 8772
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.715493793911031,
+      "learning_rate": 1.539800712575701e-07,
+      "loss": 0.6442,
+      "step": 8773
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.534362203603708,
+      "learning_rate": 1.5356066829428529e-07,
+      "loss": 0.5944,
+      "step": 8774
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.7191866220592633,
+      "learning_rate": 1.531418283763131e-07,
+      "loss": 0.605,
+      "step": 8775
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.6751125356427465,
+      "learning_rate": 1.5272355155231233e-07,
+      "loss": 0.5287,
+      "step": 8776
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6076140742413605,
+      "learning_rate": 1.5230583787087693e-07,
+      "loss": 0.6233,
+      "step": 8777
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.2526568139523055,
+      "learning_rate": 1.5188868738053643e-07,
+      "loss": 0.6007,
+      "step": 8778
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.556172706540122,
+      "learning_rate": 1.5147210012975366e-07,
+      "loss": 0.5975,
+      "step": 8779
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.317968435237044,
+      "learning_rate": 1.5105607616692665e-07,
+      "loss": 0.5119,
+      "step": 8780
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.364058422346575,
+      "learning_rate": 1.5064061554038723e-07,
+      "loss": 0.6071,
+      "step": 8781
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.123856213413403,
+      "learning_rate": 1.5022571829840404e-07,
+      "loss": 0.5481,
+      "step": 8782
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5488265654803324,
+      "learning_rate": 1.4981138448917686e-07,
+      "loss": 0.5901,
+      "step": 8783
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.9357992577530965,
+      "learning_rate": 1.4939761416084274e-07,
+      "loss": 0.5701,
+      "step": 8784
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0256367627019287,
+      "learning_rate": 1.4898440736147213e-07,
+      "loss": 0.6026,
+      "step": 8785
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.257081804924682,
+      "learning_rate": 1.4857176413907048e-07,
+      "loss": 0.5129,
+      "step": 8786
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.7139299347714076,
+      "learning_rate": 1.481596845415767e-07,
+      "loss": 0.5782,
+      "step": 8787
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4990484620871873,
+      "learning_rate": 1.4774816861686636e-07,
+      "loss": 0.6671,
+      "step": 8788
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.9970898191679884,
+      "learning_rate": 1.4733721641274677e-07,
+      "loss": 0.653,
+      "step": 8789
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.3568839354768514,
+      "learning_rate": 1.4692682797696201e-07,
+      "loss": 0.5761,
+      "step": 8790
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.61930351344336,
+      "learning_rate": 1.4651700335718887e-07,
+      "loss": 0.6616,
+      "step": 8791
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9479674201636344,
+      "learning_rate": 1.4610774260104155e-07,
+      "loss": 0.5103,
+      "step": 8792
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5330814701077635,
+      "learning_rate": 1.456990457560642e-07,
+      "loss": 0.5497,
+      "step": 8793
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.309748805294827,
+      "learning_rate": 1.4529091286973994e-07,
+      "loss": 0.6107,
+      "step": 8794
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.547398975416799,
+      "learning_rate": 1.4488334398948424e-07,
+      "loss": 0.6341,
+      "step": 8795
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0790737085851827,
+      "learning_rate": 1.444763391626458e-07,
+      "loss": 0.5955,
+      "step": 8796
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7131868065415845,
+      "learning_rate": 1.4406989843651186e-07,
+      "loss": 0.6069,
+      "step": 8797
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1351827973252804,
+      "learning_rate": 1.4366402185829852e-07,
+      "loss": 0.5892,
+      "step": 8798
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.532366496256515,
+      "learning_rate": 1.4325870947516195e-07,
+      "loss": 0.6275,
+      "step": 8799
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.5352178877025415,
+      "learning_rate": 1.4285396133418894e-07,
+      "loss": 0.6247,
+      "step": 8800
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3725771799289634,
+      "learning_rate": 1.424497774824024e-07,
+      "loss": 0.5422,
+      "step": 8801
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.453635363952836,
+      "learning_rate": 1.4204615796675813e-07,
+      "loss": 0.6249,
+      "step": 8802
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.6864136459930834,
+      "learning_rate": 1.4164310283414917e-07,
+      "loss": 0.6058,
+      "step": 8803
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5101928312119233,
+      "learning_rate": 1.4124061213139973e-07,
+      "loss": 0.674,
+      "step": 8804
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5816393554893198,
+      "learning_rate": 1.4083868590527128e-07,
+      "loss": 0.56,
+      "step": 8805
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.008204586126537,
+      "learning_rate": 1.4043732420245703e-07,
+      "loss": 0.5047,
+      "step": 8806
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6435612166868547,
+      "learning_rate": 1.400365270695875e-07,
+      "loss": 0.6431,
+      "step": 8807
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3837590071708403,
+      "learning_rate": 1.3963629455322536e-07,
+      "loss": 0.6289,
+      "step": 8808
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2165814241637776,
+      "learning_rate": 1.3923662669986847e-07,
+      "loss": 0.6364,
+      "step": 8809
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.8103240249394665,
+      "learning_rate": 1.388375235559497e-07,
+      "loss": 0.6198,
+      "step": 8810
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.745817288882361,
+      "learning_rate": 1.3843898516783528e-07,
+      "loss": 0.6645,
+      "step": 8811
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9439152142723546,
+      "learning_rate": 1.3804101158182592e-07,
+      "loss": 0.5582,
+      "step": 8812
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2299912321829636,
+      "learning_rate": 1.3764360284415745e-07,
+      "loss": 0.5961,
+      "step": 8813
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.274213469206632,
+      "learning_rate": 1.372467590009996e-07,
+      "loss": 0.6478,
+      "step": 8814
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.736445144043836,
+      "learning_rate": 1.3685048009845602e-07,
+      "loss": 0.5455,
+      "step": 8815
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3511706665739376,
+      "learning_rate": 1.3645476618256658e-07,
+      "loss": 0.6518,
+      "step": 8816
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.9759636145549075,
+      "learning_rate": 1.3605961729930283e-07,
+      "loss": 0.6208,
+      "step": 8817
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6395425552448972,
+      "learning_rate": 1.3566503349457193e-07,
+      "loss": 0.5986,
+      "step": 8818
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5256137921850588,
+      "learning_rate": 1.3527101481421722e-07,
+      "loss": 0.5586,
+      "step": 8819
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.365812363582891,
+      "learning_rate": 1.3487756130401264e-07,
+      "loss": 0.5482,
+      "step": 8820
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.41157315584283,
+      "learning_rate": 1.3448467300966995e-07,
+      "loss": 0.5167,
+      "step": 8821
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.350934489365909,
+      "learning_rate": 1.3409234997683262e-07,
+      "loss": 0.5623,
+      "step": 8822
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.8590303256325225,
+      "learning_rate": 1.3370059225108088e-07,
+      "loss": 0.615,
+      "step": 8823
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.477586205368679,
+      "learning_rate": 1.3330939987792668e-07,
+      "loss": 0.6072,
+      "step": 8824
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.401994252587328,
+      "learning_rate": 1.3291877290281864e-07,
+      "loss": 0.574,
+      "step": 8825
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.7781870354990517,
+      "learning_rate": 1.3252871137113764e-07,
+      "loss": 0.56,
+      "step": 8826
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.367412583097113,
+      "learning_rate": 1.3213921532820084e-07,
+      "loss": 0.678,
+      "step": 8827
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9230400310337836,
+      "learning_rate": 1.3175028481925865e-07,
+      "loss": 0.6137,
+      "step": 8828
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6899938815586157,
+      "learning_rate": 1.3136191988949498e-07,
+      "loss": 0.5719,
+      "step": 8829
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.313782561671406,
+      "learning_rate": 1.3097412058403036e-07,
+      "loss": 0.6413,
+      "step": 8830
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3968960300093514,
+      "learning_rate": 1.305868869479171e-07,
+      "loss": 0.6002,
+      "step": 8831
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9061105754688217,
+      "learning_rate": 1.3020021902614366e-07,
+      "loss": 0.5138,
+      "step": 8832
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.361235492021561,
+      "learning_rate": 1.2981411686363132e-07,
+      "loss": 0.6635,
+      "step": 8833
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.299548642666442,
+      "learning_rate": 1.29428580505237e-07,
+      "loss": 0.574,
+      "step": 8834
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.8233240276159557,
+      "learning_rate": 1.290436099957504e-07,
+      "loss": 0.6384,
+      "step": 8835
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4935098889156753,
+      "learning_rate": 1.2865920537989683e-07,
+      "loss": 0.5546,
+      "step": 8836
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7213656000819872,
+      "learning_rate": 1.2827536670233508e-07,
+      "loss": 0.6305,
+      "step": 8837
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.741318595224126,
+      "learning_rate": 1.2789209400765889e-07,
+      "loss": 0.6613,
+      "step": 8838
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.396722910728486,
+      "learning_rate": 1.2750938734039486e-07,
+      "loss": 0.5485,
+      "step": 8839
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.597262875682383,
+      "learning_rate": 1.2712724674500575e-07,
+      "loss": 0.6165,
+      "step": 8840
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5915988746408636,
+      "learning_rate": 1.2674567226588662e-07,
+      "loss": 0.658,
+      "step": 8841
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 12.953891578992364,
+      "learning_rate": 1.2636466394736758e-07,
+      "loss": 0.5931,
+      "step": 8842
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.102318058166209,
+      "learning_rate": 1.2598422183371484e-07,
+      "loss": 0.5928,
+      "step": 8843
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.2543978164700937,
+      "learning_rate": 1.256043459691253e-07,
+      "loss": 0.6385,
+      "step": 8844
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9749011154837868,
+      "learning_rate": 1.2522503639773254e-07,
+      "loss": 0.5538,
+      "step": 8845
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.8982951468251628,
+      "learning_rate": 1.2484629316360297e-07,
+      "loss": 0.5987,
+      "step": 8846
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.042274517432498,
+      "learning_rate": 1.244681163107392e-07,
+      "loss": 0.5936,
+      "step": 8847
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.543924226290593,
+      "learning_rate": 1.2409050588307547e-07,
+      "loss": 0.6316,
+      "step": 8848
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.275882847499862,
+      "learning_rate": 1.237134619244823e-07,
+      "loss": 0.6165,
+      "step": 8849
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5971023387508803,
+      "learning_rate": 1.2333698447876296e-07,
+      "loss": 0.6973,
+      "step": 8850
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.306532467495735,
+      "learning_rate": 1.229610735896558e-07,
+      "loss": 0.5684,
+      "step": 8851
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5310472966419377,
+      "learning_rate": 1.2258572930083313e-07,
+      "loss": 0.7,
+      "step": 8852
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.187034442784481,
+      "learning_rate": 1.222109516559006e-07,
+      "loss": 0.5667,
+      "step": 8853
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2888506566786493,
+      "learning_rate": 1.2183674069840057e-07,
+      "loss": 0.6023,
+      "step": 8854
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9527027435099331,
+      "learning_rate": 1.2146309647180554e-07,
+      "loss": 0.4499,
+      "step": 8855
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.730891050434837,
+      "learning_rate": 1.2109001901952633e-07,
+      "loss": 0.6274,
+      "step": 8856
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.916048809399537,
+      "learning_rate": 1.2071750838490492e-07,
+      "loss": 0.5672,
+      "step": 8857
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2808207217906893,
+      "learning_rate": 1.2034556461121894e-07,
+      "loss": 0.6706,
+      "step": 8858
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.48748766003941,
+      "learning_rate": 1.1997418774167934e-07,
+      "loss": 0.6039,
+      "step": 8859
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.37457209874063,
+      "learning_rate": 1.196033778194322e-07,
+      "loss": 0.6669,
+      "step": 8860
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1767809049061233,
+      "learning_rate": 1.1923313488755638e-07,
+      "loss": 0.5488,
+      "step": 8861
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2318763739789023,
+      "learning_rate": 1.1886345898906693e-07,
+      "loss": 0.6166,
+      "step": 8862
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.8605686514104973,
+      "learning_rate": 1.1849435016691003e-07,
+      "loss": 0.606,
+      "step": 8863
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.908693940117577,
+      "learning_rate": 1.1812580846396915e-07,
+      "loss": 0.5751,
+      "step": 8864
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.700097097894084,
+      "learning_rate": 1.1775783392305895e-07,
+      "loss": 0.6018,
+      "step": 8865
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.782107805001388,
+      "learning_rate": 1.1739042658693079e-07,
+      "loss": 0.5801,
+      "step": 8866
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6177299400948395,
+      "learning_rate": 1.1702358649826939e-07,
+      "loss": 0.6479,
+      "step": 8867
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.350114941297819,
+      "learning_rate": 1.166573136996918e-07,
+      "loss": 0.5391,
+      "step": 8868
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5754262051694763,
+      "learning_rate": 1.1629160823375118e-07,
+      "loss": 0.6068,
+      "step": 8869
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.382289379949096,
+      "learning_rate": 1.1592647014293412e-07,
+      "loss": 0.6544,
+      "step": 8870
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3588925482663523,
+      "learning_rate": 1.1556189946966168e-07,
+      "loss": 0.5807,
+      "step": 8871
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.582600577577335,
+      "learning_rate": 1.151978962562883e-07,
+      "loss": 0.6165,
+      "step": 8872
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.245726951542925,
+      "learning_rate": 1.1483446054510294e-07,
+      "loss": 0.5569,
+      "step": 8873
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.638961684798639,
+      "learning_rate": 1.144715923783274e-07,
+      "loss": 0.6425,
+      "step": 8874
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.337475255641479,
+      "learning_rate": 1.1410929179812069e-07,
+      "loss": 0.6021,
+      "step": 8875
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6499250104828334,
+      "learning_rate": 1.1374755884657195e-07,
+      "loss": 0.589,
+      "step": 8876
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.7819394045598353,
+      "learning_rate": 1.1338639356570758e-07,
+      "loss": 0.6157,
+      "step": 8877
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.9903219653791866,
+      "learning_rate": 1.130257959974862e-07,
+      "loss": 0.566,
+      "step": 8878
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9474747256626247,
+      "learning_rate": 1.1266576618380098e-07,
+      "loss": 0.5611,
+      "step": 8879
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.8711918810126424,
+      "learning_rate": 1.1230630416647958e-07,
+      "loss": 0.5494,
+      "step": 8880
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.991121349490518,
+      "learning_rate": 1.1194740998728193e-07,
+      "loss": 0.6804,
+      "step": 8881
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.1006023896574195,
+      "learning_rate": 1.1158908368790523e-07,
+      "loss": 0.7024,
+      "step": 8882
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1579700060562166,
+      "learning_rate": 1.1123132530997727e-07,
+      "loss": 0.6184,
+      "step": 8883
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.466363894718886,
+      "learning_rate": 1.1087413489506205e-07,
+      "loss": 0.6905,
+      "step": 8884
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.6365001531373107,
+      "learning_rate": 1.1051751248465691e-07,
+      "loss": 0.5644,
+      "step": 8885
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.007115123145037,
+      "learning_rate": 1.1016145812019319e-07,
+      "loss": 0.6451,
+      "step": 8886
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.339803693912085,
+      "learning_rate": 1.098059718430361e-07,
+      "loss": 0.6171,
+      "step": 8887
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.514752409454256,
+      "learning_rate": 1.0945105369448483e-07,
+      "loss": 0.6357,
+      "step": 8888
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.1616055453161453,
+      "learning_rate": 1.0909670371577308e-07,
+      "loss": 0.6318,
+      "step": 8889
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3220638444142474,
+      "learning_rate": 1.087429219480679e-07,
+      "loss": 0.6093,
+      "step": 8890
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3976532258477015,
+      "learning_rate": 1.0838970843247143e-07,
+      "loss": 0.6613,
+      "step": 8891
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.661253383224113,
+      "learning_rate": 1.0803706321001805e-07,
+      "loss": 0.6697,
+      "step": 8892
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6683239924659907,
+      "learning_rate": 1.0768498632167779e-07,
+      "loss": 0.6107,
+      "step": 8893
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5718421574140513,
+      "learning_rate": 1.0733347780835346e-07,
+      "loss": 0.5767,
+      "step": 8894
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.146294517128654,
+      "learning_rate": 1.0698253771088241e-07,
+      "loss": 0.591,
+      "step": 8895
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2452315585617613,
+      "learning_rate": 1.0663216607003535e-07,
+      "loss": 0.5671,
+      "step": 8896
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3863559337974825,
+      "learning_rate": 1.0628236292651861e-07,
+      "loss": 0.5705,
+      "step": 8897
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6440910292337767,
+      "learning_rate": 1.0593312832097025e-07,
+      "loss": 0.5802,
+      "step": 8898
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9645714063530803,
+      "learning_rate": 1.055844622939639e-07,
+      "loss": 0.5883,
+      "step": 8899
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.58708264588437,
+      "learning_rate": 1.0523636488600664e-07,
+      "loss": 0.5565,
+      "step": 8900
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.265603984151188,
+      "learning_rate": 1.048888361375383e-07,
+      "loss": 0.6022,
+      "step": 8901
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.357562605848145,
+      "learning_rate": 1.045418760889355e-07,
+      "loss": 0.5128,
+      "step": 8902
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.4292025931619565,
+      "learning_rate": 1.0419548478050601e-07,
+      "loss": 0.5251,
+      "step": 8903
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0323838830808834,
+      "learning_rate": 1.038496622524926e-07,
+      "loss": 0.5835,
+      "step": 8904
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.6168280543265166,
+      "learning_rate": 1.0350440854507205e-07,
+      "loss": 0.636,
+      "step": 8905
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.052977096577669,
+      "learning_rate": 1.0315972369835559e-07,
+      "loss": 0.6365,
+      "step": 8906
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2130917241762083,
+      "learning_rate": 1.0281560775238619e-07,
+      "loss": 0.6581,
+      "step": 8907
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.7050724972147653,
+      "learning_rate": 1.0247206074714411e-07,
+      "loss": 0.6105,
+      "step": 8908
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.090324313422908,
+      "learning_rate": 1.0212908272253963e-07,
+      "loss": 0.6013,
+      "step": 8909
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.5235691542628236,
+      "learning_rate": 1.0178667371842088e-07,
+      "loss": 0.6415,
+      "step": 8910
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4523022912907857,
+      "learning_rate": 1.014448337745666e-07,
+      "loss": 0.6337,
+      "step": 8911
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.318885308337774,
+      "learning_rate": 1.0110356293069168e-07,
+      "loss": 0.6198,
+      "step": 8912
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0757919001418195,
+      "learning_rate": 1.0076286122644274e-07,
+      "loss": 0.6183,
+      "step": 8913
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5362575196035597,
+      "learning_rate": 1.0042272870140258e-07,
+      "loss": 0.6237,
+      "step": 8914
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.31713074091154,
+      "learning_rate": 1.0008316539508733e-07,
+      "loss": 0.6719,
+      "step": 8915
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.767274831301876,
+      "learning_rate": 9.974417134694491e-08,
+      "loss": 0.5058,
+      "step": 8916
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9939390220324551,
+      "learning_rate": 9.940574659635993e-08,
+      "loss": 0.5415,
+      "step": 8917
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.3485962773421636,
+      "learning_rate": 9.90678911826487e-08,
+      "loss": 0.5924,
+      "step": 8918
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.8816111251299037,
+      "learning_rate": 9.873060514506316e-08,
+      "loss": 0.6027,
+      "step": 8919
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0275962573135506,
+      "learning_rate": 9.839388852278752e-08,
+      "loss": 0.5166,
+      "step": 8920
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5666272152464558,
+      "learning_rate": 9.805774135494106e-08,
+      "loss": 0.6445,
+      "step": 8921
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4748907614724827,
+      "learning_rate": 9.772216368057586e-08,
+      "loss": 0.6334,
+      "step": 8922
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.942060956680936,
+      "learning_rate": 9.738715553867851e-08,
+      "loss": 0.6156,
+      "step": 8923
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.9951301127561973,
+      "learning_rate": 9.705271696816954e-08,
+      "loss": 0.5977,
+      "step": 8924
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.777453454552441,
+      "learning_rate": 9.671884800790288e-08,
+      "loss": 0.6252,
+      "step": 8925
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.348283390971774,
+      "learning_rate": 9.638554869666695e-08,
+      "loss": 0.6118,
+      "step": 8926
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4067385375615493,
+      "learning_rate": 9.605281907318243e-08,
+      "loss": 0.6271,
+      "step": 8927
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9552504322728489,
+      "learning_rate": 9.572065917610618e-08,
+      "loss": 0.5533,
+      "step": 8928
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.271928583094003,
+      "learning_rate": 9.538906904402623e-08,
+      "loss": 0.6035,
+      "step": 8929
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.3223263454773297,
+      "learning_rate": 9.505804871546731e-08,
+      "loss": 0.6097,
+      "step": 8930
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.923685715911361,
+      "learning_rate": 9.472759822888478e-08,
+      "loss": 0.576,
+      "step": 8931
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.249066974617624,
+      "learning_rate": 9.439771762267069e-08,
+      "loss": 0.6664,
+      "step": 8932
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2985171963927464,
+      "learning_rate": 9.40684069351483e-08,
+      "loss": 0.6363,
+      "step": 8933
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1586975547776857,
+      "learning_rate": 9.373966620457753e-08,
+      "loss": 0.5921,
+      "step": 8934
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.699332277329513,
+      "learning_rate": 9.341149546914951e-08,
+      "loss": 0.6785,
+      "step": 8935
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.705681753410133,
+      "learning_rate": 9.308389476699043e-08,
+      "loss": 0.6113,
+      "step": 8936
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.299817850591371,
+      "learning_rate": 9.27568641361598e-08,
+      "loss": 0.6317,
+      "step": 8937
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6726846572437575,
+      "learning_rate": 9.243040361465172e-08,
+      "loss": 0.5638,
+      "step": 8938
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0645519294317953,
+      "learning_rate": 9.210451324039304e-08,
+      "loss": 0.6308,
+      "step": 8939
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4527040374452262,
+      "learning_rate": 9.177919305124405e-08,
+      "loss": 0.6753,
+      "step": 8940
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9497404981411139,
+      "learning_rate": 9.145444308500117e-08,
+      "loss": 0.5437,
+      "step": 8941
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.86280566237663,
+      "learning_rate": 9.11302633793909e-08,
+      "loss": 0.5582,
+      "step": 8942
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.231382224332527,
+      "learning_rate": 9.080665397207755e-08,
+      "loss": 0.5663,
+      "step": 8943
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.383310389908959,
+      "learning_rate": 9.048361490065549e-08,
+      "loss": 0.5818,
+      "step": 8944
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.365476329509127,
+      "learning_rate": 9.016114620265526e-08,
+      "loss": 0.6667,
+      "step": 8945
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6118574377559023,
+      "learning_rate": 8.983924791553966e-08,
+      "loss": 0.7023,
+      "step": 8946
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.2390121343870013,
+      "learning_rate": 8.951792007670713e-08,
+      "loss": 0.5439,
+      "step": 8947
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.824711199639862,
+      "learning_rate": 8.919716272348722e-08,
+      "loss": 0.6532,
+      "step": 8948
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.7642587541543717,
+      "learning_rate": 8.88769758931457e-08,
+      "loss": 0.5559,
+      "step": 8949
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.7721849377139645,
+      "learning_rate": 8.855735962288059e-08,
+      "loss": 0.6408,
+      "step": 8950
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 12.581276150709563,
+      "learning_rate": 8.823831394982329e-08,
+      "loss": 0.5789,
+      "step": 8951
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6512425220755578,
+      "learning_rate": 8.791983891104084e-08,
+      "loss": 0.557,
+      "step": 8952
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4054321801371494,
+      "learning_rate": 8.760193454353194e-08,
+      "loss": 0.6467,
+      "step": 8953
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9064106195783275,
+      "learning_rate": 8.728460088422985e-08,
+      "loss": 0.5413,
+      "step": 8954
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5100181773055414,
+      "learning_rate": 8.696783797000174e-08,
+      "loss": 0.6023,
+      "step": 8955
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9356433470805314,
+      "learning_rate": 8.665164583764818e-08,
+      "loss": 0.5343,
+      "step": 8956
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9024523813801144,
+      "learning_rate": 8.633602452390311e-08,
+      "loss": 0.6028,
+      "step": 8957
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9298897593318474,
+      "learning_rate": 8.602097406543442e-08,
+      "loss": 0.6587,
+      "step": 8958
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6001966117983955,
+      "learning_rate": 8.570649449884505e-08,
+      "loss": 0.6481,
+      "step": 8959
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.45160147679471,
+      "learning_rate": 8.539258586066912e-08,
+      "loss": 0.5524,
+      "step": 8960
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.5237192425119686,
+      "learning_rate": 8.507924818737523e-08,
+      "loss": 0.7146,
+      "step": 8961
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5557536371206893,
+      "learning_rate": 8.476648151536704e-08,
+      "loss": 0.6013,
+      "step": 8962
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.195273209399613,
+      "learning_rate": 8.445428588098048e-08,
+      "loss": 0.6442,
+      "step": 8963
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.174929174236636,
+      "learning_rate": 8.414266132048543e-08,
+      "loss": 0.5563,
+      "step": 8964
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.334437374273213,
+      "learning_rate": 8.383160787008627e-08,
+      "loss": 0.6087,
+      "step": 8965
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2982581898665733,
+      "learning_rate": 8.352112556591907e-08,
+      "loss": 0.5159,
+      "step": 8966
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.485884461295592,
+      "learning_rate": 8.321121444405611e-08,
+      "loss": 0.6474,
+      "step": 8967
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9202366397635555,
+      "learning_rate": 8.29018745405008e-08,
+      "loss": 0.4943,
+      "step": 8968
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.242007414517413,
+      "learning_rate": 8.259310589119162e-08,
+      "loss": 0.5891,
+      "step": 8969
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.159210998912373,
+      "learning_rate": 8.2284908532001e-08,
+      "loss": 0.6245,
+      "step": 8970
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5880306782993467,
+      "learning_rate": 8.19772824987336e-08,
+      "loss": 0.5807,
+      "step": 8971
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9458064039539235,
+      "learning_rate": 8.167022782712919e-08,
+      "loss": 0.5854,
+      "step": 8972
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2865336633479743,
+      "learning_rate": 8.136374455286033e-08,
+      "loss": 0.6299,
+      "step": 8973
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9344831427818824,
+      "learning_rate": 8.105783271153356e-08,
+      "loss": 0.5415,
+      "step": 8974
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.63891046584668,
+      "learning_rate": 8.075249233868821e-08,
+      "loss": 0.6369,
+      "step": 8975
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0932455433990051,
+      "learning_rate": 8.044772346979812e-08,
+      "loss": 0.5487,
+      "step": 8976
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8253603754509732,
+      "learning_rate": 8.014352614027054e-08,
+      "loss": 0.531,
+      "step": 8977
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3887697326667325,
+      "learning_rate": 7.983990038544664e-08,
+      "loss": 0.5317,
+      "step": 8978
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.954945688802483,
+      "learning_rate": 7.953684624059987e-08,
+      "loss": 0.5603,
+      "step": 8979
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.949484917320116,
+      "learning_rate": 7.923436374093929e-08,
+      "loss": 0.5994,
+      "step": 8980
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.027629114424706,
+      "learning_rate": 7.893245292160511e-08,
+      "loss": 0.601,
+      "step": 8981
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3171055100942057,
+      "learning_rate": 7.863111381767374e-08,
+      "loss": 0.544,
+      "step": 8982
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5697202512374178,
+      "learning_rate": 7.833034646415272e-08,
+      "loss": 0.6016,
+      "step": 8983
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.502909881767718,
+      "learning_rate": 7.80301508959852e-08,
+      "loss": 0.6382,
+      "step": 8984
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4985875211152675,
+      "learning_rate": 7.773052714804719e-08,
+      "loss": 0.5886,
+      "step": 8985
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.9604096707667176,
+      "learning_rate": 7.743147525514749e-08,
+      "loss": 0.6592,
+      "step": 8986
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.074366972373142,
+      "learning_rate": 7.713299525202944e-08,
+      "loss": 0.5226,
+      "step": 8987
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.844749218296161,
+      "learning_rate": 7.683508717336918e-08,
+      "loss": 0.6071,
+      "step": 8988
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.591141612749998,
+      "learning_rate": 7.653775105377737e-08,
+      "loss": 0.6346,
+      "step": 8989
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.626446005912948,
+      "learning_rate": 7.62409869277969e-08,
+      "loss": 0.6454,
+      "step": 8990
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5111955019235888,
+      "learning_rate": 7.59447948299058e-08,
+      "loss": 0.6752,
+      "step": 8991
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.0568359259106237,
+      "learning_rate": 7.564917479451373e-08,
+      "loss": 0.617,
+      "step": 8992
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.22905556576431,
+      "learning_rate": 7.535412685596599e-08,
+      "loss": 0.658,
+      "step": 8993
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.687646814338487,
+      "learning_rate": 7.505965104854073e-08,
+      "loss": 0.7067,
+      "step": 8994
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1712557178132945,
+      "learning_rate": 7.476574740644838e-08,
+      "loss": 0.5687,
+      "step": 8995
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.390681967447077,
+      "learning_rate": 7.44724159638338e-08,
+      "loss": 0.6088,
+      "step": 8996
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0375781519569047,
+      "learning_rate": 7.417965675477534e-08,
+      "loss": 0.541,
+      "step": 8997
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.695143332239307,
+      "learning_rate": 7.388746981328632e-08,
+      "loss": 0.5657,
+      "step": 8998
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.5000885655908025,
+      "learning_rate": 7.359585517331014e-08,
+      "loss": 0.643,
+      "step": 8999
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.556164468290286,
+      "learning_rate": 7.330481286872749e-08,
+      "loss": 0.5475,
+      "step": 9000
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8838512097484564,
+      "learning_rate": 7.301434293334908e-08,
+      "loss": 0.5353,
+      "step": 9001
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.274766804514225,
+      "learning_rate": 7.272444540092294e-08,
+      "loss": 0.5828,
+      "step": 9002
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1596373340527326,
+      "learning_rate": 7.243512030512656e-08,
+      "loss": 0.6407,
+      "step": 9003
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3934408334922597,
+      "learning_rate": 7.214636767957417e-08,
+      "loss": 0.5951,
+      "step": 9004
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1077351249376233,
+      "learning_rate": 7.18581875578117e-08,
+      "loss": 0.5951,
+      "step": 9005
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.7426426194662987,
+      "learning_rate": 7.157057997331907e-08,
+      "loss": 0.5838,
+      "step": 9006
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.3577955817656235,
+      "learning_rate": 7.128354495951006e-08,
+      "loss": 0.6553,
+      "step": 9007
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.089592673711738,
+      "learning_rate": 7.099708254973136e-08,
+      "loss": 0.542,
+      "step": 9008
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6306257005655937,
+      "learning_rate": 7.071119277726301e-08,
+      "loss": 0.5938,
+      "step": 9009
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.773569073948341,
+      "learning_rate": 7.0425875675319e-08,
+      "loss": 0.6349,
+      "step": 9010
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1860135695945595,
+      "learning_rate": 7.014113127704725e-08,
+      "loss": 0.5982,
+      "step": 9011
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.076079644508135,
+      "learning_rate": 6.985695961552796e-08,
+      "loss": 0.6265,
+      "step": 9012
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9415297800104788,
+      "learning_rate": 6.957336072377586e-08,
+      "loss": 0.5714,
+      "step": 9013
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.7068897773748826,
+      "learning_rate": 6.929033463473789e-08,
+      "loss": 0.5468,
+      "step": 9014
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2176189144911542,
+      "learning_rate": 6.900788138129554e-08,
+      "loss": 0.6505,
+      "step": 9015
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3975049310120897,
+      "learning_rate": 6.872600099626369e-08,
+      "loss": 0.5567,
+      "step": 9016
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9140350212678829,
+      "learning_rate": 6.844469351239003e-08,
+      "loss": 0.5619,
+      "step": 9017
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5388047256734287,
+      "learning_rate": 6.816395896235617e-08,
+      "loss": 0.6404,
+      "step": 9018
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.61129763755357,
+      "learning_rate": 6.78837973787766e-08,
+      "loss": 0.6912,
+      "step": 9019
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.7678872176224463,
+      "learning_rate": 6.760420879420082e-08,
+      "loss": 0.5791,
+      "step": 9020
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.085235281927168,
+      "learning_rate": 6.732519324111009e-08,
+      "loss": 0.6828,
+      "step": 9021
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.045301446382808,
+      "learning_rate": 6.704675075191902e-08,
+      "loss": 0.6762,
+      "step": 9022
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.2248231907993987,
+      "learning_rate": 6.676888135897674e-08,
+      "loss": 0.649,
+      "step": 9023
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.704494653374823,
+      "learning_rate": 6.649158509456576e-08,
+      "loss": 0.5956,
+      "step": 9024
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9722338123614437,
+      "learning_rate": 6.621486199090088e-08,
+      "loss": 0.5335,
+      "step": 9025
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.591910857960895,
+      "learning_rate": 6.593871208013136e-08,
+      "loss": 0.6333,
+      "step": 9026
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.8111496930068176,
+      "learning_rate": 6.566313539433877e-08,
+      "loss": 0.6471,
+      "step": 9027
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.432987898988874,
+      "learning_rate": 6.538813196553973e-08,
+      "loss": 0.5657,
+      "step": 9028
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.439770278419752,
+      "learning_rate": 6.511370182568311e-08,
+      "loss": 0.6174,
+      "step": 9029
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3437842495883623,
+      "learning_rate": 6.483984500665119e-08,
+      "loss": 0.5372,
+      "step": 9030
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.638108993601756,
+      "learning_rate": 6.456656154025964e-08,
+      "loss": 0.6398,
+      "step": 9031
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5417156476529437,
+      "learning_rate": 6.429385145825861e-08,
+      "loss": 0.5727,
+      "step": 9032
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6367489893136846,
+      "learning_rate": 6.402171479233e-08,
+      "loss": 0.6402,
+      "step": 9033
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.001287522657275,
+      "learning_rate": 6.375015157409015e-08,
+      "loss": 0.6013,
+      "step": 9034
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4098090763309523,
+      "learning_rate": 6.347916183508828e-08,
+      "loss": 0.5185,
+      "step": 9035
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6963976959673808,
+      "learning_rate": 6.320874560680757e-08,
+      "loss": 0.7315,
+      "step": 9036
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8986435359967851,
+      "learning_rate": 6.293890292066395e-08,
+      "loss": 0.5098,
+      "step": 9037
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.122976565906962,
+      "learning_rate": 6.266963380800684e-08,
+      "loss": 0.6318,
+      "step": 9038
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4548075643836396,
+      "learning_rate": 6.24009383001195e-08,
+      "loss": 0.6208,
+      "step": 9039
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.0915406479900294,
+      "learning_rate": 6.213281642821811e-08,
+      "loss": 0.562,
+      "step": 9040
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.875966474710966,
+      "learning_rate": 6.186526822345163e-08,
+      "loss": 0.6213,
+      "step": 9041
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.426133515510632,
+      "learning_rate": 6.159829371690407e-08,
+      "loss": 0.5456,
+      "step": 9042
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.21552519795364,
+      "learning_rate": 6.133189293959175e-08,
+      "loss": 0.5443,
+      "step": 9043
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.182803894427842,
+      "learning_rate": 6.106606592246267e-08,
+      "loss": 0.7024,
+      "step": 9044
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.838797575657555,
+      "learning_rate": 6.08008126964016e-08,
+      "loss": 0.5106,
+      "step": 9045
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.464352882434948,
+      "learning_rate": 6.053613329222441e-08,
+      "loss": 0.6797,
+      "step": 9046
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5403063737877813,
+      "learning_rate": 6.027202774068042e-08,
+      "loss": 0.6741,
+      "step": 9047
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.275405839493515,
+      "learning_rate": 6.00084960724534e-08,
+      "loss": 0.5259,
+      "step": 9048
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4141326847728526,
+      "learning_rate": 5.974553831815888e-08,
+      "loss": 0.6703,
+      "step": 9049
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.7803880388763815,
+      "learning_rate": 5.9483154508347406e-08,
+      "loss": 0.6808,
+      "step": 9050
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.32854036522692,
+      "learning_rate": 5.9221344673500714e-08,
+      "loss": 0.6318,
+      "step": 9051
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.7007162566736373,
+      "learning_rate": 5.896010884403669e-08,
+      "loss": 0.6192,
+      "step": 9052
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6424830019293544,
+      "learning_rate": 5.8699447050303284e-08,
+      "loss": 0.6002,
+      "step": 9053
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9699800972674935,
+      "learning_rate": 5.84393593225846e-08,
+      "loss": 0.5143,
+      "step": 9054
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1870238921593677,
+      "learning_rate": 5.817984569109702e-08,
+      "loss": 0.6391,
+      "step": 9055
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3894652143809623,
+      "learning_rate": 5.792090618598922e-08,
+      "loss": 0.5925,
+      "step": 9056
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0806139077721775,
+      "learning_rate": 5.766254083734435e-08,
+      "loss": 0.6226,
+      "step": 9057
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0742443711293004,
+      "learning_rate": 5.740474967517839e-08,
+      "loss": 0.7206,
+      "step": 9058
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.751056686871944,
+      "learning_rate": 5.714753272944129e-08,
+      "loss": 0.63,
+      "step": 9059
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.9924140950273572,
+      "learning_rate": 5.68908900300158e-08,
+      "loss": 0.5352,
+      "step": 9060
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.401100392375077,
+      "learning_rate": 5.6634821606717514e-08,
+      "loss": 0.6337,
+      "step": 9061
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.3084059799813343,
+      "learning_rate": 5.6379327489295424e-08,
+      "loss": 0.5915,
+      "step": 9062
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4362237430967926,
+      "learning_rate": 5.6124407707432436e-08,
+      "loss": 0.6301,
+      "step": 9063
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.974258772040746,
+      "learning_rate": 5.5870062290744876e-08,
+      "loss": 0.6755,
+      "step": 9064
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.330894208256594,
+      "learning_rate": 5.5616291268781875e-08,
+      "loss": 0.6224,
+      "step": 9065
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1719011136515,
+      "learning_rate": 5.53630946710243e-08,
+      "loss": 0.6096,
+      "step": 9066
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.645578213445915,
+      "learning_rate": 5.5110472526889725e-08,
+      "loss": 0.5871,
+      "step": 9067
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.176097941547101,
+      "learning_rate": 5.485842486572579e-08,
+      "loss": 0.6087,
+      "step": 9068
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.84982524682506,
+      "learning_rate": 5.4606951716815735e-08,
+      "loss": 0.6064,
+      "step": 9069
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4460753590123403,
+      "learning_rate": 5.435605310937342e-08,
+      "loss": 0.6023,
+      "step": 9070
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.2859655207695355,
+      "learning_rate": 5.410572907254885e-08,
+      "loss": 0.6332,
+      "step": 9071
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.7744254402541415,
+      "learning_rate": 5.3855979635423774e-08,
+      "loss": 0.6184,
+      "step": 9072
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2908907666964726,
+      "learning_rate": 5.360680482701275e-08,
+      "loss": 0.6022,
+      "step": 9073
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.659397892050746,
+      "learning_rate": 5.3358204676264844e-08,
+      "loss": 0.5827,
+      "step": 9074
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.568905156288951,
+      "learning_rate": 5.3110179212061406e-08,
+      "loss": 0.6192,
+      "step": 9075
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.292649322204462,
+      "learning_rate": 5.286272846321716e-08,
+      "loss": 0.5548,
+      "step": 9076
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.0132313937949267,
+      "learning_rate": 5.2615852458480775e-08,
+      "loss": 0.6589,
+      "step": 9077
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.892679189485966,
+      "learning_rate": 5.23695512265332e-08,
+      "loss": 0.5454,
+      "step": 9078
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2508615371235727,
+      "learning_rate": 5.2123824795988764e-08,
+      "loss": 0.6351,
+      "step": 9079
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.9551130391975455,
+      "learning_rate": 5.187867319539519e-08,
+      "loss": 0.6406,
+      "step": 9080
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.096299527370705,
+      "learning_rate": 5.163409645323414e-08,
+      "loss": 0.6185,
+      "step": 9081
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.0533996853064957,
+      "learning_rate": 5.139009459791955e-08,
+      "loss": 0.5818,
+      "step": 9082
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.9947071352082357,
+      "learning_rate": 5.1146667657798744e-08,
+      "loss": 0.5878,
+      "step": 9083
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6515995229170928,
+      "learning_rate": 5.0903815661152435e-08,
+      "loss": 0.6047,
+      "step": 9084
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.216360216245191,
+      "learning_rate": 5.0661538636194164e-08,
+      "loss": 0.6247,
+      "step": 9085
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.268655449300178,
+      "learning_rate": 5.041983661107142e-08,
+      "loss": 0.622,
+      "step": 9086
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 21.215101180819993,
+      "learning_rate": 5.017870961386451e-08,
+      "loss": 0.568,
+      "step": 9087
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.415400161597065,
+      "learning_rate": 4.9938157672586585e-08,
+      "loss": 0.664,
+      "step": 9088
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.498930091348122,
+      "learning_rate": 4.9698180815183626e-08,
+      "loss": 0.6175,
+      "step": 9089
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2002817993561457,
+      "learning_rate": 4.945877906953722e-08,
+      "loss": 0.5586,
+      "step": 9090
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.936573265428427,
+      "learning_rate": 4.921995246345901e-08,
+      "loss": 0.5832,
+      "step": 9091
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2660732599551445,
+      "learning_rate": 4.898170102469513e-08,
+      "loss": 0.6299,
+      "step": 9092
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3604212255644623,
+      "learning_rate": 4.87440247809251e-08,
+      "loss": 0.6544,
+      "step": 9093
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.441174650138323,
+      "learning_rate": 4.850692375976185e-08,
+      "loss": 0.6546,
+      "step": 9094
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.559330779198589,
+      "learning_rate": 4.827039798875111e-08,
+      "loss": 0.5868,
+      "step": 9095
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6264220920489487,
+      "learning_rate": 4.803444749537145e-08,
+      "loss": 0.592,
+      "step": 9096
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.346201879455138,
+      "learning_rate": 4.7799072307034845e-08,
+      "loss": 0.5913,
+      "step": 9097
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2187162896332735,
+      "learning_rate": 4.756427245108664e-08,
+      "loss": 0.5922,
+      "step": 9098
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.665266994162754,
+      "learning_rate": 4.733004795480556e-08,
+      "loss": 0.5226,
+      "step": 9099
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.309703067456407,
+      "learning_rate": 4.709639884540262e-08,
+      "loss": 0.6754,
+      "step": 9100
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.519489921324912,
+      "learning_rate": 4.686332515002223e-08,
+      "loss": 0.5979,
+      "step": 9101
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.4734030677173644,
+      "learning_rate": 4.663082689574328e-08,
+      "loss": 0.5856,
+      "step": 9102
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9763445326065605,
+      "learning_rate": 4.6398904109575815e-08,
+      "loss": 0.5201,
+      "step": 9103
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2503233196738193,
+      "learning_rate": 4.616755681846441e-08,
+      "loss": 0.543,
+      "step": 9104
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2179469381251273,
+      "learning_rate": 4.593678504928589e-08,
+      "loss": 0.6158,
+      "step": 9105
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.623231471830531,
+      "learning_rate": 4.570658882885104e-08,
+      "loss": 0.6553,
+      "step": 9106
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.50946392413075,
+      "learning_rate": 4.547696818390346e-08,
+      "loss": 0.5586,
+      "step": 9107
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2550240930054586,
+      "learning_rate": 4.524792314111959e-08,
+      "loss": 0.6628,
+      "step": 9108
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.249364536937027,
+      "learning_rate": 4.501945372710925e-08,
+      "loss": 0.5307,
+      "step": 9109
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.641533439490308,
+      "learning_rate": 4.4791559968415664e-08,
+      "loss": 0.581,
+      "step": 9110
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.400735266374335,
+      "learning_rate": 4.456424189151376e-08,
+      "loss": 0.5803,
+      "step": 9111
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.013311376301837,
+      "learning_rate": 4.433749952281463e-08,
+      "loss": 0.5211,
+      "step": 9112
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6401002509833007,
+      "learning_rate": 4.4111332888658876e-08,
+      "loss": 0.6004,
+      "step": 9113
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9931452731017555,
+      "learning_rate": 4.388574201532214e-08,
+      "loss": 0.5143,
+      "step": 9114
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6549315124291772,
+      "learning_rate": 4.366072692901346e-08,
+      "loss": 0.6261,
+      "step": 9115
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.012204151927376,
+      "learning_rate": 4.343628765587471e-08,
+      "loss": 0.6443,
+      "step": 9116
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0512576287150874,
+      "learning_rate": 4.321242422197946e-08,
+      "loss": 0.5194,
+      "step": 9117
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.609812624572654,
+      "learning_rate": 4.298913665333637e-08,
+      "loss": 0.5724,
+      "step": 9118
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.067384788809318,
+      "learning_rate": 4.276642497588579e-08,
+      "loss": 0.5614,
+      "step": 9119
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.4934409570440055,
+      "learning_rate": 4.2544289215502576e-08,
+      "loss": 0.5847,
+      "step": 9120
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.470305736923208,
+      "learning_rate": 4.2322729397992755e-08,
+      "loss": 0.6099,
+      "step": 9121
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.735639182147766,
+      "learning_rate": 4.210174554909796e-08,
+      "loss": 0.6468,
+      "step": 9122
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4294443877199807,
+      "learning_rate": 4.188133769448932e-08,
+      "loss": 0.5507,
+      "step": 9123
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.268281292601878,
+      "learning_rate": 4.1661505859775245e-08,
+      "loss": 0.6073,
+      "step": 9124
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.292116636642608,
+      "learning_rate": 4.1442250070494186e-08,
+      "loss": 0.625,
+      "step": 9125
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.0214437747145553,
+      "learning_rate": 4.1223570352118545e-08,
+      "loss": 0.5964,
+      "step": 9126
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.237561901276038,
+      "learning_rate": 4.100546673005412e-08,
+      "loss": 0.6468,
+      "step": 9127
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.1688911358412177,
+      "learning_rate": 4.078793922963953e-08,
+      "loss": 0.6302,
+      "step": 9128
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.447810752311471,
+      "learning_rate": 4.057098787614677e-08,
+      "loss": 0.5529,
+      "step": 9129
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.636676069560152,
+      "learning_rate": 4.035461269478014e-08,
+      "loss": 0.653,
+      "step": 9130
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8829981767535657,
+      "learning_rate": 4.013881371067841e-08,
+      "loss": 0.617,
+      "step": 9131
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.756084256939002,
+      "learning_rate": 3.992359094891096e-08,
+      "loss": 0.5825,
+      "step": 9132
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2300017538671466,
+      "learning_rate": 3.970894443448281e-08,
+      "loss": 0.5501,
+      "step": 9133
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.0651156169351963,
+      "learning_rate": 3.949487419233122e-08,
+      "loss": 0.5455,
+      "step": 9134
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4150439079741504,
+      "learning_rate": 3.92813802473252e-08,
+      "loss": 0.6961,
+      "step": 9135
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.197514010373594,
+      "learning_rate": 3.906846262426878e-08,
+      "loss": 0.6243,
+      "step": 9136
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.122714920113005,
+      "learning_rate": 3.885612134789718e-08,
+      "loss": 0.6401,
+      "step": 9137
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9331466918924591,
+      "learning_rate": 3.864435644288123e-08,
+      "loss": 0.5596,
+      "step": 9138
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.5512512202448434,
+      "learning_rate": 3.8433167933821234e-08,
+      "loss": 0.6279,
+      "step": 9139
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3384806493446684,
+      "learning_rate": 3.822255584525369e-08,
+      "loss": 0.6029,
+      "step": 9140
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7384044773724057,
+      "learning_rate": 3.8012520201646255e-08,
+      "loss": 0.5762,
+      "step": 9141
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.537480825998291,
+      "learning_rate": 3.780306102740105e-08,
+      "loss": 0.6246,
+      "step": 9142
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.535534515356477,
+      "learning_rate": 3.7594178346851974e-08,
+      "loss": 0.6051,
+      "step": 9143
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6217629251128436,
+      "learning_rate": 3.738587218426626e-08,
+      "loss": 0.5192,
+      "step": 9144
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.930611934484491,
+      "learning_rate": 3.7178142563844e-08,
+      "loss": 0.554,
+      "step": 9145
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4744564645686906,
+      "learning_rate": 3.697098950971922e-08,
+      "loss": 0.6043,
+      "step": 9146
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.800360974943865,
+      "learning_rate": 3.676441304595879e-08,
+      "loss": 0.5262,
+      "step": 9147
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2752239967517283,
+      "learning_rate": 3.655841319656128e-08,
+      "loss": 0.641,
+      "step": 9148
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.9374913960050826,
+      "learning_rate": 3.635298998545922e-08,
+      "loss": 0.62,
+      "step": 9149
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.780995594486591,
+      "learning_rate": 3.614814343651851e-08,
+      "loss": 0.6157,
+      "step": 9150
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.194471742530128,
+      "learning_rate": 3.5943873573537903e-08,
+      "loss": 0.5811,
+      "step": 9151
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3614260129521387,
+      "learning_rate": 3.574018042024785e-08,
+      "loss": 0.5898,
+      "step": 9152
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.281118604888936,
+      "learning_rate": 3.553706400031331e-08,
+      "loss": 0.624,
+      "step": 9153
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.4511839932038293,
+      "learning_rate": 3.533452433733209e-08,
+      "loss": 0.6539,
+      "step": 9154
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.0189891199871286,
+      "learning_rate": 3.513256145483479e-08,
+      "loss": 0.5865,
+      "step": 9155
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.371611401435088,
+      "learning_rate": 3.493117537628432e-08,
+      "loss": 0.6437,
+      "step": 9156
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8768550617990599,
+      "learning_rate": 3.4730366125076966e-08,
+      "loss": 0.5259,
+      "step": 9157
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.802661184621188,
+      "learning_rate": 3.453013372454295e-08,
+      "loss": 0.5858,
+      "step": 9158
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4916523907020736,
+      "learning_rate": 3.433047819794366e-08,
+      "loss": 0.6361,
+      "step": 9159
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3170062905496733,
+      "learning_rate": 3.413139956847611e-08,
+      "loss": 0.6925,
+      "step": 9160
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.102122288265513,
+      "learning_rate": 3.3932897859267346e-08,
+      "loss": 0.6089,
+      "step": 9161
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.212048736855744,
+      "learning_rate": 3.3734973093378367e-08,
+      "loss": 0.678,
+      "step": 9162
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.5526846645578467,
+      "learning_rate": 3.353762529380466e-08,
+      "loss": 0.6322,
+      "step": 9163
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6953009571329827,
+      "learning_rate": 3.334085448347346e-08,
+      "loss": 0.6071,
+      "step": 9164
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.09081032023049,
+      "learning_rate": 3.314466068524425e-08,
+      "loss": 0.5649,
+      "step": 9165
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.8019153487339916,
+      "learning_rate": 3.294904392191045e-08,
+      "loss": 0.6405,
+      "step": 9166
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2938098097617368,
+      "learning_rate": 3.27540042161989e-08,
+      "loss": 0.6646,
+      "step": 9167
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.475187200092886,
+      "learning_rate": 3.255954159076813e-08,
+      "loss": 0.5879,
+      "step": 9168
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.1691062969062194,
+      "learning_rate": 3.236565606821007e-08,
+      "loss": 0.5371,
+      "step": 9169
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4161686022837277,
+      "learning_rate": 3.2172347671050596e-08,
+      "loss": 0.6002,
+      "step": 9170
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.853074919404103,
+      "learning_rate": 3.197961642174674e-08,
+      "loss": 0.5644,
+      "step": 9171
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.210820383910226,
+      "learning_rate": 3.1787462342690036e-08,
+      "loss": 0.6206,
+      "step": 9172
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.628828526440614,
+      "learning_rate": 3.1595885456204845e-08,
+      "loss": 0.6029,
+      "step": 9173
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5487802779308995,
+      "learning_rate": 3.1404885784547256e-08,
+      "loss": 0.645,
+      "step": 9174
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9690656931951284,
+      "learning_rate": 3.1214463349907295e-08,
+      "loss": 0.5252,
+      "step": 9175
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.965446772712665,
+      "learning_rate": 3.102461817440727e-08,
+      "loss": 0.6123,
+      "step": 9176
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6886931228239357,
+      "learning_rate": 3.083535028010343e-08,
+      "loss": 0.6453,
+      "step": 9177
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.884853895703779,
+      "learning_rate": 3.064665968898428e-08,
+      "loss": 0.7053,
+      "step": 9178
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.1113257956144733,
+      "learning_rate": 3.045854642297175e-08,
+      "loss": 0.5508,
+      "step": 9179
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.493207760394977,
+      "learning_rate": 3.0271010503918896e-08,
+      "loss": 0.5488,
+      "step": 9180
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3766479319028724,
+      "learning_rate": 3.0084051953614414e-08,
+      "loss": 0.6282,
+      "step": 9181
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.390922910780254,
+      "learning_rate": 2.98976707937787e-08,
+      "loss": 0.5949,
+      "step": 9182
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.637454787285773,
+      "learning_rate": 2.971186704606388e-08,
+      "loss": 0.6665,
+      "step": 9183
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.1213890226656327,
+      "learning_rate": 2.9526640732056577e-08,
+      "loss": 0.51,
+      "step": 9184
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.490219535861415,
+      "learning_rate": 2.9341991873276244e-08,
+      "loss": 0.6355,
+      "step": 9185
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.218227767506622,
+      "learning_rate": 2.9157920491174606e-08,
+      "loss": 0.683,
+      "step": 9186
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2005059135053537,
+      "learning_rate": 2.8974426607136784e-08,
+      "loss": 0.5855,
+      "step": 9187
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.519798208810904,
+      "learning_rate": 2.8791510242480168e-08,
+      "loss": 0.6008,
+      "step": 9188
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0120446604343019,
+      "learning_rate": 2.8609171418454985e-08,
+      "loss": 0.5326,
+      "step": 9189
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.450219497355183,
+      "learning_rate": 2.842741015624595e-08,
+      "loss": 0.6363,
+      "step": 9190
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.052648319059015,
+      "learning_rate": 2.824622647696895e-08,
+      "loss": 0.5779,
+      "step": 9191
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.335643356301621,
+      "learning_rate": 2.8065620401673823e-08,
+      "loss": 0.5543,
+      "step": 9192
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.8344656986626906,
+      "learning_rate": 2.7885591951342104e-08,
+      "loss": 0.549,
+      "step": 9193
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.599346650297371,
+      "learning_rate": 2.77061411468893e-08,
+      "loss": 0.6037,
+      "step": 9194
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.0866091554980106,
+      "learning_rate": 2.752726800916372e-08,
+      "loss": 0.5579,
+      "step": 9195
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5707910612614655,
+      "learning_rate": 2.734897255894653e-08,
+      "loss": 0.6654,
+      "step": 9196
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5872290746708435,
+      "learning_rate": 2.71712548169506e-08,
+      "loss": 0.6281,
+      "step": 9197
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.2731949836335774,
+      "learning_rate": 2.6994114803823858e-08,
+      "loss": 0.6685,
+      "step": 9198
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.899660903128125,
+      "learning_rate": 2.6817552540144842e-08,
+      "loss": 0.6137,
+      "step": 9199
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.320763158061146,
+      "learning_rate": 2.6641568046427146e-08,
+      "loss": 0.6083,
+      "step": 9200
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.363696210436481,
+      "learning_rate": 2.646616134311497e-08,
+      "loss": 0.6219,
+      "step": 9201
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2956123393115724,
+      "learning_rate": 2.629133245058757e-08,
+      "loss": 0.6388,
+      "step": 9202
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.8842450507583464,
+      "learning_rate": 2.6117081389155362e-08,
+      "loss": 0.6312,
+      "step": 9203
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3338099016770926,
+      "learning_rate": 2.5943408179062713e-08,
+      "loss": 0.5896,
+      "step": 9204
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3320839464277805,
+      "learning_rate": 2.5770312840486255e-08,
+      "loss": 0.6366,
+      "step": 9205
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.984624737573807,
+      "learning_rate": 2.5597795393536017e-08,
+      "loss": 0.6062,
+      "step": 9206
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9651826359687896,
+      "learning_rate": 2.5425855858253744e-08,
+      "loss": 0.5126,
+      "step": 9207
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.288749577144304,
+      "learning_rate": 2.5254494254616236e-08,
+      "loss": 0.6235,
+      "step": 9208
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.813589985884465,
+      "learning_rate": 2.5083710602530897e-08,
+      "loss": 0.6208,
+      "step": 9209
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.407381015638563,
+      "learning_rate": 2.4913504921839084e-08,
+      "loss": 0.5145,
+      "step": 9210
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2924984215742583,
+      "learning_rate": 2.4743877232314416e-08,
+      "loss": 0.5947,
+      "step": 9211
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.1338707036223505,
+      "learning_rate": 2.457482755366447e-08,
+      "loss": 0.5781,
+      "step": 9212
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.277076857886843,
+      "learning_rate": 2.4406355905528534e-08,
+      "loss": 0.6377,
+      "step": 9213
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5947856777870917,
+      "learning_rate": 2.4238462307478727e-08,
+      "loss": 0.6077,
+      "step": 9214
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.869549468783701,
+      "learning_rate": 2.4071146779021116e-08,
+      "loss": 0.5939,
+      "step": 9215
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5062122517145817,
+      "learning_rate": 2.3904409339594036e-08,
+      "loss": 0.5345,
+      "step": 9216
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.9256223760450024,
+      "learning_rate": 2.373825000856811e-08,
+      "loss": 0.5867,
+      "step": 9217
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.919362951808869,
+      "learning_rate": 2.3572668805247335e-08,
+      "loss": 0.6491,
+      "step": 9218
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6991155767439214,
+      "learning_rate": 2.3407665748868548e-08,
+      "loss": 0.5601,
+      "step": 9219
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.246298603916671,
+      "learning_rate": 2.324324085860086e-08,
+      "loss": 0.5792,
+      "step": 9220
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4031521640333504,
+      "learning_rate": 2.3079394153547874e-08,
+      "loss": 0.6562,
+      "step": 9221
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4774885599982843,
+      "learning_rate": 2.2916125652743814e-08,
+      "loss": 0.5821,
+      "step": 9222
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.9008654534964395,
+      "learning_rate": 2.2753435375156284e-08,
+      "loss": 0.6582,
+      "step": 9223
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.023778464208766,
+      "learning_rate": 2.2591323339687387e-08,
+      "loss": 0.6167,
+      "step": 9224
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3713659550667208,
+      "learning_rate": 2.242978956517039e-08,
+      "loss": 0.6183,
+      "step": 9225
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.306655048988526,
+      "learning_rate": 2.2268834070371946e-08,
+      "loss": 0.6889,
+      "step": 9226
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.0952640364000117,
+      "learning_rate": 2.210845687399099e-08,
+      "loss": 0.5602,
+      "step": 9227
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3413653802222494,
+      "learning_rate": 2.1948657994659838e-08,
+      "loss": 0.5923,
+      "step": 9228
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.77564140602996,
+      "learning_rate": 2.1789437450943084e-08,
+      "loss": 0.5727,
+      "step": 9229
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.421552954655445,
+      "learning_rate": 2.163079526133982e-08,
+      "loss": 0.611,
+      "step": 9230
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.169302617388929,
+      "learning_rate": 2.1472731444279193e-08,
+      "loss": 0.5952,
+      "step": 9231
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.101895600351323,
+      "learning_rate": 2.13152460181254e-08,
+      "loss": 0.6372,
+      "step": 9232
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.657299806141878,
+      "learning_rate": 2.115833900117381e-08,
+      "loss": 0.6147,
+      "step": 9233
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.239460891177711,
+      "learning_rate": 2.1002010411654838e-08,
+      "loss": 0.591,
+      "step": 9234
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.488691890857634,
+      "learning_rate": 2.0846260267728957e-08,
+      "loss": 0.6899,
+      "step": 9235
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1238152317575727,
+      "learning_rate": 2.069108858749169e-08,
+      "loss": 0.6646,
+      "step": 9236
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.487320361580777,
+      "learning_rate": 2.0536495388969734e-08,
+      "loss": 0.6297,
+      "step": 9237
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.341768489528419,
+      "learning_rate": 2.0382480690123718e-08,
+      "loss": 0.472,
+      "step": 9238
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.233011462575182,
+      "learning_rate": 2.0229044508845997e-08,
+      "loss": 0.6539,
+      "step": 9239
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.121493412352339,
+      "learning_rate": 2.007618686296342e-08,
+      "loss": 0.5738,
+      "step": 9240
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9581343879954742,
+      "learning_rate": 1.9923907770233453e-08,
+      "loss": 0.5091,
+      "step": 9241
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.403223954181605,
+      "learning_rate": 1.9772207248348607e-08,
+      "loss": 0.6106,
+      "step": 9242
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8777448291544866,
+      "learning_rate": 1.962108531493201e-08,
+      "loss": 0.5469,
+      "step": 9243
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.9765736631382067,
+      "learning_rate": 1.9470541987540727e-08,
+      "loss": 0.5947,
+      "step": 9244
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.5869579127381717,
+      "learning_rate": 1.9320577283664656e-08,
+      "loss": 0.6733,
+      "step": 9245
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7935438970542363,
+      "learning_rate": 1.9171191220726527e-08,
+      "loss": 0.5995,
+      "step": 9246
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.928327785902899,
+      "learning_rate": 1.902238381608079e-08,
+      "loss": 0.5998,
+      "step": 9247
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.4791238402015456,
+      "learning_rate": 1.88741550870164e-08,
+      "loss": 0.5935,
+      "step": 9248
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.863133326433092,
+      "learning_rate": 1.8726505050753464e-08,
+      "loss": 0.5795,
+      "step": 9249
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.0631417802895955,
+      "learning_rate": 1.8579433724446037e-08,
+      "loss": 0.5772,
+      "step": 9250
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2432282328723803,
+      "learning_rate": 1.8432941125179904e-08,
+      "loss": 0.5114,
+      "step": 9251
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5097002070968535,
+      "learning_rate": 1.8287027269974777e-08,
+      "loss": 0.6283,
+      "step": 9252
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.415419955262919,
+      "learning_rate": 1.81416921757821e-08,
+      "loss": 0.6065,
+      "step": 9253
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2888006379470407,
+      "learning_rate": 1.7996935859486143e-08,
+      "loss": 0.5377,
+      "step": 9254
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.95190793167679,
+      "learning_rate": 1.7852758337904564e-08,
+      "loss": 0.5234,
+      "step": 9255
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.928375135216635,
+      "learning_rate": 1.7709159627787853e-08,
+      "loss": 0.5789,
+      "step": 9256
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.772545157709395,
+      "learning_rate": 1.7566139745818778e-08,
+      "loss": 0.6105,
+      "step": 9257
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9706682161569412,
+      "learning_rate": 1.7423698708612935e-08,
+      "loss": 0.5673,
+      "step": 9258
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5143022423925516,
+      "learning_rate": 1.72818365327182e-08,
+      "loss": 0.5667,
+      "step": 9259
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.5152224135253496,
+      "learning_rate": 1.7140553234616385e-08,
+      "loss": 0.5852,
+      "step": 9260
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6502532483318486,
+      "learning_rate": 1.6999848830721033e-08,
+      "loss": 0.585,
+      "step": 9261
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.643359676277189,
+      "learning_rate": 1.6859723337379064e-08,
+      "loss": 0.5682,
+      "step": 9262
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.073844785345583,
+      "learning_rate": 1.6720176770869124e-08,
+      "loss": 0.6263,
+      "step": 9263
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.2055735703872275,
+      "learning_rate": 1.6581209147404355e-08,
+      "loss": 0.6097,
+      "step": 9264
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6043559102032106,
+      "learning_rate": 1.6442820483128508e-08,
+      "loss": 0.6575,
+      "step": 9265
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1311347692762466,
+      "learning_rate": 1.630501079412039e-08,
+      "loss": 0.5999,
+      "step": 9266
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.1904074031842913,
+      "learning_rate": 1.6167780096389417e-08,
+      "loss": 0.6478,
+      "step": 9267
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1605938131759217,
+      "learning_rate": 1.603112840587895e-08,
+      "loss": 0.5674,
+      "step": 9268
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7661197207313495,
+      "learning_rate": 1.589505573846517e-08,
+      "loss": 0.6057,
+      "step": 9269
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6447415105460883,
+      "learning_rate": 1.5759562109955993e-08,
+      "loss": 0.6398,
+      "step": 9270
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.2687696784623625,
+      "learning_rate": 1.562464753609272e-08,
+      "loss": 0.6534,
+      "step": 9271
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6228867844890065,
+      "learning_rate": 1.549031203254947e-08,
+      "loss": 0.627,
+      "step": 9272
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.4312244311880393,
+      "learning_rate": 1.535655561493321e-08,
+      "loss": 0.6083,
+      "step": 9273
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4986020292339344,
+      "learning_rate": 1.5223378298783174e-08,
+      "loss": 0.5576,
+      "step": 9274
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.519385627634563,
+      "learning_rate": 1.5090780099571435e-08,
+      "loss": 0.5539,
+      "step": 9275
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.883104133354655,
+      "learning_rate": 1.4958761032702885e-08,
+      "loss": 0.5969,
+      "step": 9276
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.536493973434565,
+      "learning_rate": 1.4827321113515259e-08,
+      "loss": 0.6409,
+      "step": 9277
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.276130559735301,
+      "learning_rate": 1.4696460357279118e-08,
+      "loss": 0.5695,
+      "step": 9278
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7168915784835974,
+      "learning_rate": 1.4566178779197305e-08,
+      "loss": 0.5578,
+      "step": 9279
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.303485469832306,
+      "learning_rate": 1.443647639440493e-08,
+      "loss": 0.619,
+      "step": 9280
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7280656636349336,
+      "learning_rate": 1.43073532179705e-08,
+      "loss": 0.7012,
+      "step": 9281
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.803606301743524,
+      "learning_rate": 1.4178809264896454e-08,
+      "loss": 0.5857,
+      "step": 9282
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.551430987761346,
+      "learning_rate": 1.4050844550115295e-08,
+      "loss": 0.6231,
+      "step": 9283
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.3909002445584813,
+      "learning_rate": 1.3923459088494574e-08,
+      "loss": 0.5716,
+      "step": 9284
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.489847512775473,
+      "learning_rate": 1.3796652894832452e-08,
+      "loss": 0.6399,
+      "step": 9285
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8687603509839055,
+      "learning_rate": 1.367042598386159e-08,
+      "loss": 0.5983,
+      "step": 9286
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.603423975622383,
+      "learning_rate": 1.3544778370246924e-08,
+      "loss": 0.6098,
+      "step": 9287
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.520998350761481,
+      "learning_rate": 1.3419710068585668e-08,
+      "loss": 0.5934,
+      "step": 9288
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.649199879398089,
+      "learning_rate": 1.3295221093407862e-08,
+      "loss": 0.6294,
+      "step": 9289
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.3152207409028662,
+      "learning_rate": 1.3171311459175829e-08,
+      "loss": 0.6102,
+      "step": 9290
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.91183070336439,
+      "learning_rate": 1.3047981180285274e-08,
+      "loss": 0.5617,
+      "step": 9291
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.536923939716204,
+      "learning_rate": 1.2925230271064736e-08,
+      "loss": 0.6005,
+      "step": 9292
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5935899938613978,
+      "learning_rate": 1.2803058745774477e-08,
+      "loss": 0.5446,
+      "step": 9293
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.67588488131424,
+      "learning_rate": 1.2681466618608696e-08,
+      "loss": 0.6637,
+      "step": 9294
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.298948783588412,
+      "learning_rate": 1.256045390369276e-08,
+      "loss": 0.5459,
+      "step": 9295
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.32146484587026,
+      "learning_rate": 1.2440020615086534e-08,
+      "loss": 0.6503,
+      "step": 9296
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.382979182133977,
+      "learning_rate": 1.232016676678105e-08,
+      "loss": 0.5351,
+      "step": 9297
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.8108107342818927,
+      "learning_rate": 1.2200892372700168e-08,
+      "loss": 0.5628,
+      "step": 9298
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5429167783954405,
+      "learning_rate": 1.2082197446701693e-08,
+      "loss": 0.5136,
+      "step": 9299
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.446935675654442,
+      "learning_rate": 1.196408200257515e-08,
+      "loss": 0.5591,
+      "step": 9300
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4390888642246606,
+      "learning_rate": 1.1846546054042341e-08,
+      "loss": 0.63,
+      "step": 9301
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.182765263784984,
+      "learning_rate": 1.1729589614758452e-08,
+      "loss": 0.6281,
+      "step": 9302
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6515402589211248,
+      "learning_rate": 1.1613212698311504e-08,
+      "loss": 0.5422,
+      "step": 9303
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.006167180494317,
+      "learning_rate": 1.1497415318221239e-08,
+      "loss": 0.5991,
+      "step": 9304
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6423248229678973,
+      "learning_rate": 1.1382197487941337e-08,
+      "loss": 0.5174,
+      "step": 9305
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.713564676817919,
+      "learning_rate": 1.1267559220857204e-08,
+      "loss": 0.6037,
+      "step": 9306
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.653891032493835,
+      "learning_rate": 1.1153500530286521e-08,
+      "loss": 0.5838,
+      "step": 9307
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.3097916904344333,
+      "learning_rate": 1.1040021429480907e-08,
+      "loss": 0.5577,
+      "step": 9308
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.68913204932153,
+      "learning_rate": 1.0927121931624263e-08,
+      "loss": 0.6034,
+      "step": 9309
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.6960977207308368,
+      "learning_rate": 1.0814802049832762e-08,
+      "loss": 0.5788,
+      "step": 9310
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5549598391454404,
+      "learning_rate": 1.0703061797154857e-08,
+      "loss": 0.6632,
+      "step": 9311
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.4964467940181576,
+      "learning_rate": 1.059190118657294e-08,
+      "loss": 0.5849,
+      "step": 9312
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.4191869666444825,
+      "learning_rate": 1.0481320231001124e-08,
+      "loss": 0.6446,
+      "step": 9313
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.233757773597287,
+      "learning_rate": 1.0371318943285802e-08,
+      "loss": 0.5923,
+      "step": 9314
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.207322331157825,
+      "learning_rate": 1.0261897336207305e-08,
+      "loss": 0.617,
+      "step": 9315
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.9505737873093274,
+      "learning_rate": 1.0153055422477686e-08,
+      "loss": 0.6109,
+      "step": 9316
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.9526141318436947,
+      "learning_rate": 1.0044793214742387e-08,
+      "loss": 0.6762,
+      "step": 9317
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.42531517384415,
+      "learning_rate": 9.937110725578015e-09,
+      "loss": 0.6146,
+      "step": 9318
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.143828176250109,
+      "learning_rate": 9.83000796749567e-09,
+      "loss": 0.5419,
+      "step": 9319
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7349892039380914,
+      "learning_rate": 9.723484952937623e-09,
+      "loss": 0.6015,
+      "step": 9320
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.880357843332127,
+      "learning_rate": 9.617541694279532e-09,
+      "loss": 0.5888,
+      "step": 9321
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.046184759076013,
+      "learning_rate": 9.512178203829881e-09,
+      "loss": 0.5733,
+      "step": 9322
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0691863659915253,
+      "learning_rate": 9.407394493829436e-09,
+      "loss": 0.5912,
+      "step": 9323
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.048172174777919,
+      "learning_rate": 9.303190576451237e-09,
+      "loss": 0.641,
+      "step": 9324
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.232433634613707,
+      "learning_rate": 9.199566463801712e-09,
+      "loss": 0.583,
+      "step": 9325
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7792858367309243,
+      "learning_rate": 9.09652216792012e-09,
+      "loss": 0.6,
+      "step": 9326
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.136354810393291,
+      "learning_rate": 8.994057700776881e-09,
+      "loss": 0.5886,
+      "step": 9327
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.469179826762709,
+      "learning_rate": 8.892173074276921e-09,
+      "loss": 0.6292,
+      "step": 9328
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.404099135191639,
+      "learning_rate": 8.790868300255773e-09,
+      "loss": 0.6475,
+      "step": 9329
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.296160070077247,
+      "learning_rate": 8.690143390484018e-09,
+      "loss": 0.6183,
+      "step": 9330
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6296150848981825,
+      "learning_rate": 8.589998356662854e-09,
+      "loss": 0.6401,
+      "step": 9331
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.253263381499738,
+      "learning_rate": 8.490433210426862e-09,
+      "loss": 0.6144,
+      "step": 9332
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.110522658122041,
+      "learning_rate": 8.391447963343457e-09,
+      "loss": 0.5804,
+      "step": 9333
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.067607575219891,
+      "learning_rate": 8.293042626912328e-09,
+      "loss": 0.5936,
+      "step": 9334
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5019568835203203,
+      "learning_rate": 8.195217212565998e-09,
+      "loss": 0.6324,
+      "step": 9335
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.925396010634107,
+      "learning_rate": 8.097971731669263e-09,
+      "loss": 0.583,
+      "step": 9336
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.319061858545403,
+      "learning_rate": 8.001306195520309e-09,
+      "loss": 0.6931,
+      "step": 9337
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2308896693198204,
+      "learning_rate": 7.90522061534904e-09,
+      "loss": 0.5912,
+      "step": 9338
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.557196137943084,
+      "learning_rate": 7.809715002318751e-09,
+      "loss": 0.6286,
+      "step": 9339
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.8420399596277703,
+      "learning_rate": 7.714789367524456e-09,
+      "loss": 0.5345,
+      "step": 9340
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.9739842730635013,
+      "learning_rate": 7.620443721995107e-09,
+      "loss": 0.6429,
+      "step": 9341
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.265940840873313,
+      "learning_rate": 7.52667807669083e-09,
+      "loss": 0.5545,
+      "step": 9342
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.287899726527495,
+      "learning_rate": 7.43349244250513e-09,
+      "loss": 0.6421,
+      "step": 9343
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5574409267349916,
+      "learning_rate": 7.340886830264904e-09,
+      "loss": 0.6416,
+      "step": 9344
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.8843132936950244,
+      "learning_rate": 7.2488612507276564e-09,
+      "loss": 0.6486,
+      "step": 9345
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.5891319217005697,
+      "learning_rate": 7.157415714584836e-09,
+      "loss": 0.6042,
+      "step": 9346
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.5665853233741407,
+      "learning_rate": 7.066550232461278e-09,
+      "loss": 0.5836,
+      "step": 9347
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.156939647962542,
+      "learning_rate": 6.976264814912426e-09,
+      "loss": 0.6158,
+      "step": 9348
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.318404717092306,
+      "learning_rate": 6.886559472427667e-09,
+      "loss": 0.5961,
+      "step": 9349
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9679186604158867,
+      "learning_rate": 6.797434215429222e-09,
+      "loss": 0.5583,
+      "step": 9350
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.882857695190735,
+      "learning_rate": 6.708889054270473e-09,
+      "loss": 0.6257,
+      "step": 9351
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.9855183026376255,
+      "learning_rate": 6.620923999239304e-09,
+      "loss": 0.5986,
+      "step": 9352
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.965482349734234,
+      "learning_rate": 6.533539060554761e-09,
+      "loss": 0.5749,
+      "step": 9353
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4913808973827263,
+      "learning_rate": 6.446734248368725e-09,
+      "loss": 0.6426,
+      "step": 9354
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7932074858752802,
+      "learning_rate": 6.360509572765905e-09,
+      "loss": 0.5945,
+      "step": 9355
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.200283514862744,
+      "learning_rate": 6.2748650437644e-09,
+      "loss": 0.6131,
+      "step": 9356
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.443831408522394,
+      "learning_rate": 6.189800671314028e-09,
+      "loss": 0.6324,
+      "step": 9357
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.722697074400931,
+      "learning_rate": 6.10531646529633e-09,
+      "loss": 0.5599,
+      "step": 9358
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.8796778714377784,
+      "learning_rate": 6.021412435527341e-09,
+      "loss": 0.5826,
+      "step": 9359
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1980873977664355,
+      "learning_rate": 5.938088591754265e-09,
+      "loss": 0.5682,
+      "step": 9360
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.506953681588526,
+      "learning_rate": 5.855344943658248e-09,
+      "loss": 0.6596,
+      "step": 9361
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4320110623826006,
+      "learning_rate": 5.773181500851044e-09,
+      "loss": 0.5329,
+      "step": 9362
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.673093388240049,
+      "learning_rate": 5.691598272878907e-09,
+      "loss": 0.6882,
+      "step": 9363
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.141915304940063,
+      "learning_rate": 5.610595269220364e-09,
+      "loss": 0.6534,
+      "step": 9364
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2742761378178487,
+      "learning_rate": 5.530172499285113e-09,
+      "loss": 0.5644,
+      "step": 9365
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.7078245584599014,
+      "learning_rate": 5.45032997241679e-09,
+      "loss": 0.6237,
+      "step": 9366
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1910604685165675,
+      "learning_rate": 5.371067697891308e-09,
+      "loss": 0.6524,
+      "step": 9367
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.263472973076441,
+      "learning_rate": 5.292385684917411e-09,
+      "loss": 0.5387,
+      "step": 9368
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6018067358751327,
+      "learning_rate": 5.214283942635567e-09,
+      "loss": 0.5654,
+      "step": 9369
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.20137961246332,
+      "learning_rate": 5.136762480120183e-09,
+      "loss": 0.5999,
+      "step": 9370
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5099314930574357,
+      "learning_rate": 5.059821306376833e-09,
+      "loss": 0.5936,
+      "step": 9371
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.70471458958764,
+      "learning_rate": 4.9834604303444774e-09,
+      "loss": 0.631,
+      "step": 9372
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.1223721533213684,
+      "learning_rate": 4.907679860894355e-09,
+      "loss": 0.6049,
+      "step": 9373
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5657179600814275,
+      "learning_rate": 4.832479606831086e-09,
+      "loss": 0.5107,
+      "step": 9374
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.080680659842182,
+      "learning_rate": 4.757859676891019e-09,
+      "loss": 0.5991,
+      "step": 9375
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8969187425122196,
+      "learning_rate": 4.683820079742218e-09,
+      "loss": 0.5855,
+      "step": 9376
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.407260977526988,
+      "learning_rate": 4.610360823987803e-09,
+      "loss": 0.652,
+      "step": 9377
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.990129810345131,
+      "learning_rate": 4.5374819181615015e-09,
+      "loss": 0.5611,
+      "step": 9378
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.437506450033328,
+      "learning_rate": 4.465183370729875e-09,
+      "loss": 0.5923,
+      "step": 9379
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.476756229416724,
+      "learning_rate": 4.393465190092316e-09,
+      "loss": 0.6029,
+      "step": 9380
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.580270764805957,
+      "learning_rate": 4.322327384581604e-09,
+      "loss": 0.6063,
+      "step": 9381
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.062993744196928,
+      "learning_rate": 4.251769962461683e-09,
+      "loss": 0.6294,
+      "step": 9382
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.1334282612359785,
+      "learning_rate": 4.181792931929885e-09,
+      "loss": 0.6693,
+      "step": 9383
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.989553082617997,
+      "learning_rate": 4.1123963011158175e-09,
+      "loss": 0.5787,
+      "step": 9384
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2591858616247964,
+      "learning_rate": 4.043580078081921e-09,
+      "loss": 0.5738,
+      "step": 9385
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6256761692714603,
+      "learning_rate": 3.975344270823467e-09,
+      "loss": 0.5571,
+      "step": 9386
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.9188998285720706,
+      "learning_rate": 3.9076888872668914e-09,
+      "loss": 0.6569,
+      "step": 9387
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4406062966062727,
+      "learning_rate": 3.84061393527313e-09,
+      "loss": 0.5535,
+      "step": 9388
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.499796926984259,
+      "learning_rate": 3.774119422634282e-09,
+      "loss": 0.6678,
+      "step": 9389
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.50651659058804,
+      "learning_rate": 3.7082053570758338e-09,
+      "loss": 0.6311,
+      "step": 9390
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.9483823752948752,
+      "learning_rate": 3.6428717462549944e-09,
+      "loss": 0.5463,
+      "step": 9391
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.723110555766217,
+      "learning_rate": 3.578118597762914e-09,
+      "loss": 0.6034,
+      "step": 9392
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5286683838105293,
+      "learning_rate": 3.5139459191213533e-09,
+      "loss": 0.6347,
+      "step": 9393
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.4531428030517914,
+      "learning_rate": 3.4503537177860145e-09,
+      "loss": 0.6048,
+      "step": 9394
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.04607450890029,
+      "learning_rate": 3.3873420011448778e-09,
+      "loss": 0.5767,
+      "step": 9395
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4392946355168275,
+      "learning_rate": 3.324910776519308e-09,
+      "loss": 0.596,
+      "step": 9396
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.422625198663431,
+      "learning_rate": 3.263060051161282e-09,
+      "loss": 0.5568,
+      "step": 9397
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.418789715138034,
+      "learning_rate": 3.2017898322567185e-09,
+      "loss": 0.6036,
+      "step": 9398
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.929592023564839,
+      "learning_rate": 3.1411001269238127e-09,
+      "loss": 0.6467,
+      "step": 9399
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2696725064713115,
+      "learning_rate": 3.080990942213591e-09,
+      "loss": 0.5624,
+      "step": 9400
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.887084023089943,
+      "learning_rate": 3.0214622851093555e-09,
+      "loss": 0.6042,
+      "step": 9401
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2400748396131984,
+      "learning_rate": 2.9625141625266863e-09,
+      "loss": 0.6244,
+      "step": 9402
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4355210959399383,
+      "learning_rate": 2.9041465813145486e-09,
+      "loss": 0.5651,
+      "step": 9403
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.590028657934874,
+      "learning_rate": 2.8463595482530747e-09,
+      "loss": 0.6597,
+      "step": 9404
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.014180884320301,
+      "learning_rate": 2.7891530700563387e-09,
+      "loss": 0.6173,
+      "step": 9405
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.4571650309323445,
+      "learning_rate": 2.73252715337069e-09,
+      "loss": 0.6504,
+      "step": 9406
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.9475010958623757,
+      "learning_rate": 2.6764818047736453e-09,
+      "loss": 0.5891,
+      "step": 9407
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.3576520108967984,
+      "learning_rate": 2.6210170307777726e-09,
+      "loss": 0.6288,
+      "step": 9408
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4213655278930717,
+      "learning_rate": 2.5661328378262516e-09,
+      "loss": 0.5432,
+      "step": 9409
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.796645905475894,
+      "learning_rate": 2.5118292322950933e-09,
+      "loss": 0.5317,
+      "step": 9410
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7845772144730496,
+      "learning_rate": 2.4581062204931395e-09,
+      "loss": 0.5843,
+      "step": 9411
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.8648265905534904,
+      "learning_rate": 2.404963808662064e-09,
+      "loss": 0.6527,
+      "step": 9412
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7630651447270407,
+      "learning_rate": 2.3524020029758175e-09,
+      "loss": 0.687,
+      "step": 9413
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.873063326108902,
+      "learning_rate": 2.3004208095406268e-09,
+      "loss": 0.6214,
+      "step": 9414
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.581099597260234,
+      "learning_rate": 2.249020234395549e-09,
+      "loss": 0.6127,
+      "step": 9415
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8980029935743684,
+      "learning_rate": 2.198200283512475e-09,
+      "loss": 0.5578,
+      "step": 9416
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.843841317169783,
+      "learning_rate": 2.14796096279557e-09,
+      "loss": 0.6437,
+      "step": 9417
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.669266532272996,
+      "learning_rate": 2.0983022780807217e-09,
+      "loss": 0.5516,
+      "step": 9418
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.9826082921006214,
+      "learning_rate": 2.049224235138314e-09,
+      "loss": 0.6909,
+      "step": 9419
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.724117248063,
+      "learning_rate": 2.0007268396687873e-09,
+      "loss": 0.6637,
+      "step": 9420
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8185604270090447,
+      "learning_rate": 1.9528100973070784e-09,
+      "loss": 0.5829,
+      "step": 9421
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.095495170435446,
+      "learning_rate": 1.9054740136204007e-09,
+      "loss": 0.5911,
+      "step": 9422
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.3241775905169297,
+      "learning_rate": 1.858718594107689e-09,
+      "loss": 0.6762,
+      "step": 9423
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.856463977964485,
+      "learning_rate": 1.8125438442007093e-09,
+      "loss": 0.5795,
+      "step": 9424
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1071167853953825,
+      "learning_rate": 1.766949769264059e-09,
+      "loss": 0.5487,
+      "step": 9425
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.898970735503558,
+      "learning_rate": 1.7219363745946127e-09,
+      "loss": 0.5282,
+      "step": 9426
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.223703639655533,
+      "learning_rate": 1.6775036654226307e-09,
+      "loss": 0.5858,
+      "step": 9427
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6009487272996834,
+      "learning_rate": 1.6336516469089846e-09,
+      "loss": 0.6202,
+      "step": 9428
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.470594389195533,
+      "learning_rate": 1.5903803241490435e-09,
+      "loss": 0.611,
+      "step": 9429
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4607686425198088,
+      "learning_rate": 1.5476897021698968e-09,
+      "loss": 0.6361,
+      "step": 9430
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.015171435256967,
+      "learning_rate": 1.5055797859309108e-09,
+      "loss": 0.6213,
+      "step": 9431
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3552305276446104,
+      "learning_rate": 1.4640505803248384e-09,
+      "loss": 0.5813,
+      "step": 9432
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.0756788209013686,
+      "learning_rate": 1.4231020901755988e-09,
+      "loss": 0.5959,
+      "step": 9433
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8917362317695483,
+      "learning_rate": 1.3827343202410527e-09,
+      "loss": 0.6495,
+      "step": 9434
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4911357145002206,
+      "learning_rate": 1.342947275211337e-09,
+      "loss": 0.6533,
+      "step": 9435
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.243456265519813,
+      "learning_rate": 1.3037409597077555e-09,
+      "loss": 0.5686,
+      "step": 9436
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4091325818193186,
+      "learning_rate": 1.265115378286108e-09,
+      "loss": 0.5949,
+      "step": 9437
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6371286497158675,
+      "learning_rate": 1.2270705354333612e-09,
+      "loss": 0.6817,
+      "step": 9438
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8038456517621673,
+      "learning_rate": 1.1896064355698678e-09,
+      "loss": 0.6074,
+      "step": 9439
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2758737769300783,
+      "learning_rate": 1.152723083047702e-09,
+      "loss": 0.5709,
+      "step": 9440
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5994595563728633,
+      "learning_rate": 1.11642048215177e-09,
+      "loss": 0.6447,
+      "step": 9441
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5885198513951937,
+      "learning_rate": 1.0806986370998086e-09,
+      "loss": 0.6026,
+      "step": 9442
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.56828110164626,
+      "learning_rate": 1.0455575520418315e-09,
+      "loss": 0.6416,
+      "step": 9443
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5864916103094,
+      "learning_rate": 1.0109972310606842e-09,
+      "loss": 0.5943,
+      "step": 9444
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.1320803492939415,
+      "learning_rate": 9.770176781709329e-10,
+      "loss": 0.5664,
+      "step": 9445
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.401636357615803,
+      "learning_rate": 9.436188973210858e-10,
+      "loss": 0.5257,
+      "step": 9446
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.253966978691805,
+      "learning_rate": 9.108008923902623e-10,
+      "loss": 0.5868,
+      "step": 9447
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.506566563109066,
+      "learning_rate": 8.785636671920783e-10,
+      "loss": 0.6921,
+      "step": 9448
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.107145184612608,
+      "learning_rate": 8.469072254713162e-10,
+      "loss": 0.5729,
+      "step": 9449
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.906976575314038,
+      "learning_rate": 8.158315709055897e-10,
+      "loss": 0.4864,
+      "step": 9450
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.135047161405744,
+      "learning_rate": 7.853367071053441e-10,
+      "loss": 0.5252,
+      "step": 9451
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.453216748538149,
+      "learning_rate": 7.554226376133012e-10,
+      "loss": 0.6918,
+      "step": 9452
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.3255864656684557,
+      "learning_rate": 7.26089365905569e-10,
+      "loss": 0.5933,
+      "step": 9453
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.4761913980531283,
+      "learning_rate": 6.97336895388867e-10,
+      "loss": 0.5683,
+      "step": 9454
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.9519019016205963,
+      "learning_rate": 6.691652294038564e-10,
+      "loss": 0.6096,
+      "step": 9455
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.0498036153708985,
+      "learning_rate": 6.415743712240296e-10,
+      "loss": 0.5917,
+      "step": 9456
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0448500807091055,
+      "learning_rate": 6.145643240540456e-10,
+      "loss": 0.5391,
+      "step": 9457
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.6179190080516817,
+      "learning_rate": 5.881350910325046e-10,
+      "loss": 0.5956,
+      "step": 9458
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.3232050763453804,
+      "learning_rate": 5.622866752291734e-10,
+      "loss": 0.6394,
+      "step": 9459
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.7020619835422437,
+      "learning_rate": 5.370190796483155e-10,
+      "loss": 0.6787,
+      "step": 9460
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2927217439792016,
+      "learning_rate": 5.123323072236952e-10,
+      "loss": 0.6097,
+      "step": 9461
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.10178763835819,
+      "learning_rate": 4.88226360824684e-10,
+      "loss": 0.64,
+      "step": 9462
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.5684498862280543,
+      "learning_rate": 4.647012432512643e-10,
+      "loss": 0.4799,
+      "step": 9463
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2494270384655715,
+      "learning_rate": 4.417569572368052e-10,
+      "loss": 0.6441,
+      "step": 9464
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.678614389562663,
+      "learning_rate": 4.1939350544695224e-10,
+      "loss": 0.5611,
+      "step": 9465
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.949351619951574,
+      "learning_rate": 3.9761089047907206e-10,
+      "loss": 0.5365,
+      "step": 9466
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.470566883786651,
+      "learning_rate": 3.764091148650284e-10,
+      "loss": 0.5893,
+      "step": 9467
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.925739305236972,
+      "learning_rate": 3.5578818106674073e-10,
+      "loss": 0.4812,
+      "step": 9468
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.416620880372823,
+      "learning_rate": 3.3574809148062546e-10,
+      "loss": 0.5753,
+      "step": 9469
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.779623019239646,
+      "learning_rate": 3.1628884843537546e-10,
+      "loss": 0.6176,
+      "step": 9470
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.564403061347666,
+      "learning_rate": 2.974104541902945e-10,
+      "loss": 0.539,
+      "step": 9471
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.6305758356659785,
+      "learning_rate": 2.7911291093973835e-10,
+      "loss": 0.6094,
+      "step": 9472
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.6682590701908198,
+      "learning_rate": 2.61396220808674e-10,
+      "loss": 0.6554,
+      "step": 9473
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.690023661401972,
+      "learning_rate": 2.4426038585656507e-10,
+      "loss": 0.5741,
+      "step": 9474
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.037525938734319,
+      "learning_rate": 2.277054080729313e-10,
+      "loss": 0.6665,
+      "step": 9475
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.649606612868773,
+      "learning_rate": 2.117312893817891e-10,
+      "loss": 0.6428,
+      "step": 9476
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.1735139647394646,
+      "learning_rate": 1.9633803163887633e-10,
+      "loss": 0.5668,
+      "step": 9477
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.423065344674815,
+      "learning_rate": 1.8152563663220712e-10,
+      "loss": 0.5754,
+      "step": 9478
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.4791765768968714,
+      "learning_rate": 1.672941060826272e-10,
+      "loss": 0.6408,
+      "step": 9479
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.3385528810246217,
+      "learning_rate": 1.5364344164436885e-10,
+      "loss": 0.557,
+      "step": 9480
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.724307971466079,
+      "learning_rate": 1.4057364490227542e-10,
+      "loss": 0.5809,
+      "step": 9481
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.66578465438902,
+      "learning_rate": 1.2808471737568717e-10,
+      "loss": 0.5427,
+      "step": 9482
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2921319289193,
+      "learning_rate": 1.1617666051455534e-10,
+      "loss": 0.6357,
+      "step": 9483
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.6535598474354454,
+      "learning_rate": 1.0484947570277293e-10,
+      "loss": 0.5376,
+      "step": 9484
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.6190379885009096,
+      "learning_rate": 9.410316425706445e-11,
+      "loss": 0.6051,
+      "step": 9485
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.5315721986313395,
+      "learning_rate": 8.393772742421036e-11,
+      "loss": 0.6413,
+      "step": 9486
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.7299924264341318,
+      "learning_rate": 7.435316638715329e-11,
+      "loss": 0.5898,
+      "step": 9487
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9023624234891299,
+      "learning_rate": 6.53494822577816e-11,
+      "loss": 0.4633,
+      "step": 9488
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.4516895016558786,
+      "learning_rate": 5.6926676083035593e-11,
+      "loss": 0.6467,
+      "step": 9489
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.440327397988931,
+      "learning_rate": 4.908474884102177e-11,
+      "loss": 0.558,
+      "step": 9490
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.9948194582261793,
+      "learning_rate": 4.1823701442678114e-11,
+      "loss": 0.6246,
+      "step": 9491
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.751705490521763,
+      "learning_rate": 3.514353473232923e-11,
+      "loss": 0.5296,
+      "step": 9492
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.4070953890206592,
+      "learning_rate": 2.9044249485465914e-11,
+      "loss": 0.5484,
+      "step": 9493
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.8452438404500078,
+      "learning_rate": 2.3525846410965557e-11,
+      "loss": 0.6063,
+      "step": 9494
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.5585166775524413,
+      "learning_rate": 1.858832614942685e-11,
+      "loss": 0.568,
+      "step": 9495
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.292945274005528,
+      "learning_rate": 1.4231689274835093e-11,
+      "loss": 0.6767,
+      "step": 9496
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.339794191169871,
+      "learning_rate": 1.0455936293451985e-11,
+      "loss": 0.5886,
+      "step": 9497
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.5328995861443517,
+      "learning_rate": 7.261067643815622e-12,
+      "loss": 0.5278,
+      "step": 9498
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.1976850234192837,
+      "learning_rate": 4.647083696740495e-12,
+      "loss": 0.5848,
+      "step": 9499
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.9136231286554226,
+      "learning_rate": 2.613984756427712e-12,
+      "loss": 0.6258,
+      "step": 9500
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.1228901797117725,
+      "learning_rate": 1.161771059354777e-12,
+      "loss": 0.6679,
+      "step": 9501
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.8882038893522686,
+      "learning_rate": 2.90442773165367e-13,
+      "loss": 0.5436,
+      "step": 9502
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.3614569935196172,
+      "learning_rate": 0.0,
+      "loss": 0.5138,
+      "step": 9503
+    },
+    {
+      "epoch": 1.0,
+      "step": 9503,
+      "total_flos": 4.083480469386035e+16,
+      "train_loss": 0.6581587371017812,
+      "train_runtime": 215857.4487,
+      "train_samples_per_second": 8.453,
+      "train_steps_per_second": 0.044
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 9503,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 30000,
+  "total_flos": 4.083480469386035e+16,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}