diff --git "a/checkpoint-2154/trainer_state.json" "b/checkpoint-2154/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-2154/trainer_state.json"
@@ -0,0 +1,15195 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.965889314305604,
+  "eval_steps": 180,
+  "global_step": 2154,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.08854816108942032,
+      "learning_rate": 2e-05,
+      "loss": 1.1956,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "eval_loss": 1.1464052200317383,
+      "eval_runtime": 96.6552,
+      "eval_samples_per_second": 51.73,
+      "eval_steps_per_second": 12.933,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.10053814947605133,
+      "learning_rate": 4e-05,
+      "loss": 1.1428,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.09509150683879852,
+      "learning_rate": 6e-05,
+      "loss": 1.1963,
+      "step": 3
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0889890268445015,
+      "learning_rate": 8e-05,
+      "loss": 1.2185,
+      "step": 4
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0988571047782898,
+      "learning_rate": 0.0001,
+      "loss": 1.1157,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.089688241481781,
+      "learning_rate": 0.00012,
+      "loss": 1.2244,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.08509726822376251,
+      "learning_rate": 0.00014,
+      "loss": 1.2232,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.07318675518035889,
+      "learning_rate": 0.00016,
+      "loss": 1.131,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.07840697467327118,
+      "learning_rate": 0.00018,
+      "loss": 1.2006,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.08998328447341919,
+      "learning_rate": 0.0002,
+      "loss": 1.0947,
+      "step": 10
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.110952228307724,
+      "learning_rate": 0.0001999998926455355,
+      "loss": 1.0462,
+      "step": 11
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.08632127940654755,
+      "learning_rate": 0.0001999995705823725,
+      "loss": 1.1094,
+      "step": 12
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.06929520517587662,
+      "learning_rate": 0.00019999903381120245,
+      "loss": 1.1388,
+      "step": 13
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.058088913559913635,
+      "learning_rate": 0.0001999982823331779,
+      "loss": 1.0518,
+      "step": 14
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.05951884761452675,
+      "learning_rate": 0.0001999973161499123,
+      "loss": 1.1296,
+      "step": 15
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.06432141363620758,
+      "learning_rate": 0.00019999613526348019,
+      "loss": 1.2231,
+      "step": 16
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.05998155102133751,
+      "learning_rate": 0.00019999473967641696,
+      "loss": 1.1548,
+      "step": 17
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.06434385478496552,
+      "learning_rate": 0.00019999312939171914,
+      "loss": 1.2023,
+      "step": 18
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.059631019830703735,
+      "learning_rate": 0.00019999130441284408,
+      "loss": 1.176,
+      "step": 19
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.06641547381877899,
+      "learning_rate": 0.00019998926474371022,
+      "loss": 1.1081,
+      "step": 20
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.05724109336733818,
+      "learning_rate": 0.0001999870103886969,
+      "loss": 1.1644,
+      "step": 21
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.05426528677344322,
+      "learning_rate": 0.00019998454135264444,
+      "loss": 1.0424,
+      "step": 22
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.06212550774216652,
+      "learning_rate": 0.00019998185764085404,
+      "loss": 1.1966,
+      "step": 23
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.05805288627743721,
+      "learning_rate": 0.0001999789592590879,
+      "loss": 1.0189,
+      "step": 24
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.06322558224201202,
+      "learning_rate": 0.00019997584621356907,
+      "loss": 1.1605,
+      "step": 25
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.05367085337638855,
+      "learning_rate": 0.0001999725185109816,
+      "loss": 1.1008,
+      "step": 26
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.06441543996334076,
+      "learning_rate": 0.00019996897615847032,
+      "loss": 1.1974,
+      "step": 27
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.06023159250617027,
+      "learning_rate": 0.00019996521916364096,
+      "loss": 1.0898,
+      "step": 28
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.05423330143094063,
+      "learning_rate": 0.00019996124753456017,
+      "loss": 1.1021,
+      "step": 29
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.05826232209801674,
+      "learning_rate": 0.00019995706127975537,
+      "loss": 1.1573,
+      "step": 30
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.06136321276426315,
+      "learning_rate": 0.00019995266040821483,
+      "loss": 1.0987,
+      "step": 31
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.057079821825027466,
+      "learning_rate": 0.0001999480449293876,
+      "loss": 1.1706,
+      "step": 32
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0606551319360733,
+      "learning_rate": 0.00019994321485318354,
+      "loss": 1.1549,
+      "step": 33
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.059561554342508316,
+      "learning_rate": 0.00019993817018997323,
+      "loss": 1.1983,
+      "step": 34
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.05861816182732582,
+      "learning_rate": 0.00019993291095058804,
+      "loss": 1.1644,
+      "step": 35
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.05660370737314224,
+      "learning_rate": 0.00019992743714632,
+      "loss": 1.1671,
+      "step": 36
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.05508609116077423,
+      "learning_rate": 0.00019992174878892186,
+      "loss": 1.2054,
+      "step": 37
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.058200910687446594,
+      "learning_rate": 0.0001999158458906071,
+      "loss": 1.1693,
+      "step": 38
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.05510352924466133,
+      "learning_rate": 0.00019990972846404967,
+      "loss": 1.0654,
+      "step": 39
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.05602727457880974,
+      "learning_rate": 0.0001999033965223843,
+      "loss": 1.0574,
+      "step": 40
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.06220879405736923,
+      "learning_rate": 0.0001998968500792062,
+      "loss": 1.1197,
+      "step": 41
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0568472184240818,
+      "learning_rate": 0.00019989008914857116,
+      "loss": 1.0929,
+      "step": 42
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.06510251760482788,
+      "learning_rate": 0.0001998831137449955,
+      "loss": 1.1372,
+      "step": 43
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.05792449042201042,
+      "learning_rate": 0.00019987592388345611,
+      "loss": 1.1506,
+      "step": 44
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.05337774381041527,
+      "learning_rate": 0.00019986851957939016,
+      "loss": 1.1419,
+      "step": 45
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.05266134813427925,
+      "learning_rate": 0.00019986090084869545,
+      "loss": 1.1284,
+      "step": 46
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.05520001798868179,
+      "learning_rate": 0.00019985306770773,
+      "loss": 1.1934,
+      "step": 47
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.056133098900318146,
+      "learning_rate": 0.00019984502017331225,
+      "loss": 1.1348,
+      "step": 48
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.05897597223520279,
+      "learning_rate": 0.00019983675826272106,
+      "loss": 1.0986,
+      "step": 49
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0529349111020565,
+      "learning_rate": 0.00019982828199369541,
+      "loss": 1.0987,
+      "step": 50
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0615835078060627,
+      "learning_rate": 0.00019981959138443467,
+      "loss": 1.1104,
+      "step": 51
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.059136223047971725,
+      "learning_rate": 0.0001998106864535983,
+      "loss": 1.1611,
+      "step": 52
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.05952458456158638,
+      "learning_rate": 0.000199801567220306,
+      "loss": 1.0697,
+      "step": 53
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.059737298637628555,
+      "learning_rate": 0.00019979223370413763,
+      "loss": 1.1294,
+      "step": 54
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.05678507685661316,
+      "learning_rate": 0.00019978268592513296,
+      "loss": 1.1584,
+      "step": 55
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.05472412705421448,
+      "learning_rate": 0.00019977292390379207,
+      "loss": 1.1416,
+      "step": 56
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.05907364934682846,
+      "learning_rate": 0.00019976294766107478,
+      "loss": 1.141,
+      "step": 57
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.05549243837594986,
+      "learning_rate": 0.00019975275721840103,
+      "loss": 1.0759,
+      "step": 58
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.062153417617082596,
+      "learning_rate": 0.0001997423525976506,
+      "loss": 1.1715,
+      "step": 59
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.05978319048881531,
+      "learning_rate": 0.0001997317338211631,
+      "loss": 1.1452,
+      "step": 60
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.053922947496175766,
+      "learning_rate": 0.00019972090091173805,
+      "loss": 1.204,
+      "step": 61
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.055929865688085556,
+      "learning_rate": 0.00019970985389263467,
+      "loss": 1.1136,
+      "step": 62
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.05876065790653229,
+      "learning_rate": 0.00019969859278757185,
+      "loss": 1.1364,
+      "step": 63
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.06039357930421829,
+      "learning_rate": 0.0001996871176207282,
+      "loss": 1.0784,
+      "step": 64
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.055668413639068604,
+      "learning_rate": 0.00019967542841674196,
+      "loss": 1.0948,
+      "step": 65
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.054827943444252014,
+      "learning_rate": 0.0001996635252007109,
+      "loss": 1.1803,
+      "step": 66
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.05121610313653946,
+      "learning_rate": 0.00019965140799819224,
+      "loss": 1.0376,
+      "step": 67
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.05753542110323906,
+      "learning_rate": 0.00019963907683520274,
+      "loss": 1.1206,
+      "step": 68
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.055351294577121735,
+      "learning_rate": 0.00019962653173821843,
+      "loss": 1.1842,
+      "step": 69
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.061469532549381256,
+      "learning_rate": 0.00019961377273417487,
+      "loss": 1.0909,
+      "step": 70
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.054115746170282364,
+      "learning_rate": 0.0001996007998504667,
+      "loss": 1.0843,
+      "step": 71
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.05452101677656174,
+      "learning_rate": 0.0001995876131149479,
+      "loss": 1.1006,
+      "step": 72
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.055213477462530136,
+      "learning_rate": 0.00019957421255593154,
+      "loss": 1.1685,
+      "step": 73
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.055633626878261566,
+      "learning_rate": 0.00019956059820218982,
+      "loss": 1.1772,
+      "step": 74
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.05437927320599556,
+      "learning_rate": 0.000199546770082954,
+      "loss": 1.1043,
+      "step": 75
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.058048781007528305,
+      "learning_rate": 0.00019953272822791424,
+      "loss": 1.1045,
+      "step": 76
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.06041900813579559,
+      "learning_rate": 0.0001995184726672197,
+      "loss": 1.1975,
+      "step": 77
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.060865532606840134,
+      "learning_rate": 0.00019950400343147833,
+      "loss": 1.1731,
+      "step": 78
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.056755781173706055,
+      "learning_rate": 0.00019948932055175686,
+      "loss": 1.0745,
+      "step": 79
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.058669883757829666,
+      "learning_rate": 0.00019947442405958074,
+      "loss": 1.0734,
+      "step": 80
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.058248650282621384,
+      "learning_rate": 0.00019945931398693408,
+      "loss": 1.1543,
+      "step": 81
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.060298267751932144,
+      "learning_rate": 0.0001994439903662596,
+      "loss": 1.1762,
+      "step": 82
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.05813004449009895,
+      "learning_rate": 0.00019942845323045837,
+      "loss": 1.1652,
+      "step": 83
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.056509047746658325,
+      "learning_rate": 0.00019941270261289012,
+      "loss": 1.1391,
+      "step": 84
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.05784199759364128,
+      "learning_rate": 0.00019939673854737277,
+      "loss": 1.1592,
+      "step": 85
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.05908602103590965,
+      "learning_rate": 0.00019938056106818261,
+      "loss": 1.1867,
+      "step": 86
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.05213363468647003,
+      "learning_rate": 0.00019936417021005414,
+      "loss": 1.0747,
+      "step": 87
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0627497211098671,
+      "learning_rate": 0.00019934756600817997,
+      "loss": 1.1086,
+      "step": 88
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.05405977740883827,
+      "learning_rate": 0.00019933074849821084,
+      "loss": 0.9786,
+      "step": 89
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.057349707931280136,
+      "learning_rate": 0.00019931371771625544,
+      "loss": 1.1537,
+      "step": 90
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.05519171431660652,
+      "learning_rate": 0.00019929647369888034,
+      "loss": 1.1859,
+      "step": 91
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.05836858972907066,
+      "learning_rate": 0.00019927901648311003,
+      "loss": 1.1702,
+      "step": 92
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.058726366609334946,
+      "learning_rate": 0.00019926134610642667,
+      "loss": 1.0712,
+      "step": 93
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.06261308491230011,
+      "learning_rate": 0.0001992434626067702,
+      "loss": 1.0476,
+      "step": 94
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.05999467894434929,
+      "learning_rate": 0.00019922536602253802,
+      "loss": 1.0721,
+      "step": 95
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.06016852334141731,
+      "learning_rate": 0.00019920705639258517,
+      "loss": 1.1634,
+      "step": 96
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.055001530796289444,
+      "learning_rate": 0.00019918853375622402,
+      "loss": 1.1175,
+      "step": 97
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.05680612102150917,
+      "learning_rate": 0.00019916979815322433,
+      "loss": 1.1508,
+      "step": 98
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.061409611254930496,
+      "learning_rate": 0.0001991508496238131,
+      "loss": 1.0693,
+      "step": 99
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.06438402086496353,
+      "learning_rate": 0.00019913168820867458,
+      "loss": 1.1216,
+      "step": 100
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.06238175556063652,
+      "learning_rate": 0.00019911231394894997,
+      "loss": 1.1668,
+      "step": 101
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.057234276086091995,
+      "learning_rate": 0.00019909272688623756,
+      "loss": 1.1387,
+      "step": 102
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.07677511870861053,
+      "learning_rate": 0.00019907292706259253,
+      "loss": 1.2115,
+      "step": 103
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.059833280742168427,
+      "learning_rate": 0.00019905291452052687,
+      "loss": 1.1444,
+      "step": 104
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.05804947763681412,
+      "learning_rate": 0.00019903268930300926,
+      "loss": 1.1475,
+      "step": 105
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.06504326313734055,
+      "learning_rate": 0.0001990122514534651,
+      "loss": 1.1432,
+      "step": 106
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.057522453367710114,
+      "learning_rate": 0.00019899160101577624,
+      "loss": 1.1549,
+      "step": 107
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.06902632117271423,
+      "learning_rate": 0.00019897073803428104,
+      "loss": 1.1719,
+      "step": 108
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.05860123783349991,
+      "learning_rate": 0.00019894966255377416,
+      "loss": 1.0437,
+      "step": 109
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.05878465995192528,
+      "learning_rate": 0.00019892837461950652,
+      "loss": 1.1757,
+      "step": 110
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.05310744047164917,
+      "learning_rate": 0.00019890687427718528,
+      "loss": 1.0382,
+      "step": 111
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.05617301166057587,
+      "learning_rate": 0.00019888516157297358,
+      "loss": 1.0004,
+      "step": 112
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.05957025662064552,
+      "learning_rate": 0.00019886323655349047,
+      "loss": 1.0967,
+      "step": 113
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.05660326033830643,
+      "learning_rate": 0.00019884109926581096,
+      "loss": 1.1104,
+      "step": 114
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0532040037214756,
+      "learning_rate": 0.00019881874975746582,
+      "loss": 1.1462,
+      "step": 115
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.054993633180856705,
+      "learning_rate": 0.00019879618807644138,
+      "loss": 1.1964,
+      "step": 116
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.06195160746574402,
+      "learning_rate": 0.00019877341427117966,
+      "loss": 1.1926,
+      "step": 117
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0648837760090828,
+      "learning_rate": 0.00019875042839057798,
+      "loss": 1.1438,
+      "step": 118
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.055000726133584976,
+      "learning_rate": 0.00019872723048398912,
+      "loss": 1.1134,
+      "step": 119
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.06383106857538223,
+      "learning_rate": 0.000198703820601221,
+      "loss": 1.1825,
+      "step": 120
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.059399884194135666,
+      "learning_rate": 0.00019868019879253684,
+      "loss": 1.1175,
+      "step": 121
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.061064597219228745,
+      "learning_rate": 0.00019865636510865467,
+      "loss": 1.1299,
+      "step": 122
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.05886513367295265,
+      "learning_rate": 0.00019863231960074758,
+      "loss": 1.1096,
+      "step": 123
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.05704960227012634,
+      "learning_rate": 0.00019860806232044337,
+      "loss": 1.1557,
+      "step": 124
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.05626309663057327,
+      "learning_rate": 0.00019858359331982467,
+      "loss": 1.1348,
+      "step": 125
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.05755053088068962,
+      "learning_rate": 0.0001985589126514286,
+      "loss": 1.2132,
+      "step": 126
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.05787818506360054,
+      "learning_rate": 0.0001985340203682467,
+      "loss": 1.146,
+      "step": 127
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.05755281075835228,
+      "learning_rate": 0.0001985089165237249,
+      "loss": 1.1248,
+      "step": 128
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.05296372249722481,
+      "learning_rate": 0.00019848360117176352,
+      "loss": 1.1296,
+      "step": 129
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.05723942816257477,
+      "learning_rate": 0.0001984580743667168,
+      "loss": 1.0936,
+      "step": 130
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.057743217796087265,
+      "learning_rate": 0.00019843233616339306,
+      "loss": 1.1184,
+      "step": 131
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.05883733928203583,
+      "learning_rate": 0.00019840638661705454,
+      "loss": 1.1413,
+      "step": 132
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.06587976962327957,
+      "learning_rate": 0.00019838022578341725,
+      "loss": 1.1704,
+      "step": 133
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.053919676691293716,
+      "learning_rate": 0.0001983538537186508,
+      "loss": 1.08,
+      "step": 134
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.05607760325074196,
+      "learning_rate": 0.0001983272704793784,
+      "loss": 1.095,
+      "step": 135
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.055869605392217636,
+      "learning_rate": 0.00019830047612267663,
+      "loss": 1.1441,
+      "step": 136
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.058360978960990906,
+      "learning_rate": 0.00019827347070607536,
+      "loss": 1.1664,
+      "step": 137
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0630207508802414,
+      "learning_rate": 0.0001982462542875576,
+      "loss": 1.2234,
+      "step": 138
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.05601505562663078,
+      "learning_rate": 0.0001982188269255595,
+      "loss": 1.1665,
+      "step": 139
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.06406079232692719,
+      "learning_rate": 0.00019819118867897003,
+      "loss": 1.21,
+      "step": 140
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.05720619857311249,
+      "learning_rate": 0.00019816333960713094,
+      "loss": 1.0835,
+      "step": 141
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.05414717271924019,
+      "learning_rate": 0.0001981352797698367,
+      "loss": 1.0968,
+      "step": 142
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.060023292899131775,
+      "learning_rate": 0.00019810700922733428,
+      "loss": 1.12,
+      "step": 143
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.06230263411998749,
+      "learning_rate": 0.00019807852804032305,
+      "loss": 1.2331,
+      "step": 144
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.05766064673662186,
+      "learning_rate": 0.00019804983626995468,
+      "loss": 1.1895,
+      "step": 145
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.06487365067005157,
+      "learning_rate": 0.00019802093397783296,
+      "loss": 1.116,
+      "step": 146
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.05351267755031586,
+      "learning_rate": 0.0001979918212260137,
+      "loss": 1.1481,
+      "step": 147
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.06439248472452164,
+      "learning_rate": 0.00019796249807700457,
+      "loss": 1.2206,
+      "step": 148
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.06001568213105202,
+      "learning_rate": 0.00019793296459376494,
+      "loss": 1.1055,
+      "step": 149
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.055722612887620926,
+      "learning_rate": 0.0001979032208397059,
+      "loss": 1.0998,
+      "step": 150
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.06338882446289062,
+      "learning_rate": 0.00019787326687868993,
+      "loss": 1.0981,
+      "step": 151
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.053316015750169754,
+      "learning_rate": 0.00019784310277503085,
+      "loss": 1.0689,
+      "step": 152
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.055426474660634995,
+      "learning_rate": 0.00019781272859349368,
+      "loss": 1.1028,
+      "step": 153
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08009687811136246,
+      "learning_rate": 0.00019778214439929452,
+      "loss": 1.1934,
+      "step": 154
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.06281601637601852,
+      "learning_rate": 0.00019775135025810035,
+      "loss": 1.1871,
+      "step": 155
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.05907437577843666,
+      "learning_rate": 0.00019772034623602894,
+      "loss": 1.1801,
+      "step": 156
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.05395223945379257,
+      "learning_rate": 0.00019768913239964872,
+      "loss": 1.1755,
+      "step": 157
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.05565226450562477,
+      "learning_rate": 0.00019765770881597855,
+      "loss": 1.1147,
+      "step": 158
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.06353943049907684,
+      "learning_rate": 0.0001976260755524877,
+      "loss": 1.148,
+      "step": 159
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.056039877235889435,
+      "learning_rate": 0.00019759423267709555,
+      "loss": 1.1285,
+      "step": 160
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.06256404519081116,
+      "learning_rate": 0.00019756218025817165,
+      "loss": 1.1148,
+      "step": 161
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.06320231407880783,
+      "learning_rate": 0.00019752991836453543,
+      "loss": 1.1897,
+      "step": 162
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.06373140960931778,
+      "learning_rate": 0.00019749744706545598,
+      "loss": 1.147,
+      "step": 163
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.061453696340322495,
+      "learning_rate": 0.00019746476643065216,
+      "loss": 1.1333,
+      "step": 164
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.06347192078828812,
+      "learning_rate": 0.00019743187653029214,
+      "loss": 1.1834,
+      "step": 165
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.058654770255088806,
+      "learning_rate": 0.00019739877743499352,
+      "loss": 1.1774,
+      "step": 166
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.05696241557598114,
+      "learning_rate": 0.000197365469215823,
+      "loss": 1.1321,
+      "step": 167
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.057179030030965805,
+      "learning_rate": 0.00019733195194429628,
+      "loss": 1.0855,
+      "step": 168
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.058308880776166916,
+      "learning_rate": 0.00019729822569237797,
+      "loss": 1.0362,
+      "step": 169
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.06666514277458191,
+      "learning_rate": 0.0001972642905324813,
+      "loss": 1.1118,
+      "step": 170
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.05894763022661209,
+      "learning_rate": 0.00019723014653746815,
+      "loss": 1.1454,
+      "step": 171
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.06241476908326149,
+      "learning_rate": 0.00019719579378064869,
+      "loss": 1.1657,
+      "step": 172
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.05953037366271019,
+      "learning_rate": 0.00019716123233578134,
+      "loss": 1.1378,
+      "step": 173
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.05746786668896675,
+      "learning_rate": 0.00019712646227707263,
+      "loss": 1.1115,
+      "step": 174
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.05923418328166008,
+      "learning_rate": 0.00019709148367917697,
+      "loss": 1.0896,
+      "step": 175
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.05635738745331764,
+      "learning_rate": 0.00019705629661719652,
+      "loss": 1.1049,
+      "step": 176
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.05885479226708412,
+      "learning_rate": 0.0001970209011666811,
+      "loss": 1.1252,
+      "step": 177
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.057710710912942886,
+      "learning_rate": 0.00019698529740362785,
+      "loss": 1.1097,
+      "step": 178
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.06445460021495819,
+      "learning_rate": 0.00019694948540448123,
+      "loss": 1.2045,
+      "step": 179
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.05737968161702156,
+      "learning_rate": 0.00019691346524613286,
+      "loss": 1.1544,
+      "step": 180
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 1.0905169248580933,
+      "eval_runtime": 97.1012,
+      "eval_samples_per_second": 51.493,
+      "eval_steps_per_second": 12.873,
+      "step": 180
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.06000959873199463,
+      "learning_rate": 0.00019687723700592116,
+      "loss": 1.14,
+      "step": 181
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.06186012923717499,
+      "learning_rate": 0.00019684080076163142,
+      "loss": 1.1284,
+      "step": 182
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.061752576380968094,
+      "learning_rate": 0.00019680415659149554,
+      "loss": 1.1719,
+      "step": 183
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.05509311705827713,
+      "learning_rate": 0.00019676730457419178,
+      "loss": 1.1155,
+      "step": 184
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.06091266870498657,
+      "learning_rate": 0.00019673024478884473,
+      "loss": 1.0099,
+      "step": 185
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.065452940762043,
+      "learning_rate": 0.00019669297731502507,
+      "loss": 1.1272,
+      "step": 186
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.05998782813549042,
+      "learning_rate": 0.00019665550223274937,
+      "loss": 1.0957,
+      "step": 187
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.06403583288192749,
+      "learning_rate": 0.00019661781962248003,
+      "loss": 1.137,
+      "step": 188
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.06362204998731613,
+      "learning_rate": 0.0001965799295651249,
+      "loss": 1.1025,
+      "step": 189
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.06712298840284348,
+      "learning_rate": 0.0001965418321420374,
+      "loss": 1.1069,
+      "step": 190
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.059146981686353683,
+      "learning_rate": 0.00019650352743501604,
+      "loss": 1.1502,
+      "step": 191
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0600997731089592,
+      "learning_rate": 0.00019646501552630444,
+      "loss": 1.1693,
+      "step": 192
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.06209671124815941,
+      "learning_rate": 0.0001964262964985912,
+      "loss": 1.1384,
+      "step": 193
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.061546314507722855,
+      "learning_rate": 0.0001963873704350094,
+      "loss": 1.1485,
+      "step": 194
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.06550047546625137,
+      "learning_rate": 0.00019634823741913687,
+      "loss": 1.1136,
+      "step": 195
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.05800144001841545,
+      "learning_rate": 0.0001963088975349956,
+      "loss": 1.1783,
+      "step": 196
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.06252887099981308,
+      "learning_rate": 0.00019626935086705194,
+      "loss": 1.1337,
+      "step": 197
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.06455714255571365,
+      "learning_rate": 0.00019622959750021605,
+      "loss": 1.1107,
+      "step": 198
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.06395559012889862,
+      "learning_rate": 0.00019618963751984195,
+      "loss": 1.0746,
+      "step": 199
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.05962133780121803,
+      "learning_rate": 0.00019614947101172732,
+      "loss": 1.2018,
+      "step": 200
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.06970924139022827,
+      "learning_rate": 0.00019610909806211323,
+      "loss": 1.2098,
+      "step": 201
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.06366313248872757,
+      "learning_rate": 0.000196068518757684,
+      "loss": 1.0914,
+      "step": 202
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.05959942191839218,
+      "learning_rate": 0.00019602773318556707,
+      "loss": 1.1557,
+      "step": 203
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.06882765144109726,
+      "learning_rate": 0.00019598674143333263,
+      "loss": 1.2375,
+      "step": 204
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.06210785359144211,
+      "learning_rate": 0.00019594554358899366,
+      "loss": 1.1748,
+      "step": 205
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.06589924544095993,
+      "learning_rate": 0.0001959041397410056,
+      "loss": 1.1242,
+      "step": 206
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.05691949278116226,
+      "learning_rate": 0.00019586252997826626,
+      "loss": 1.1532,
+      "step": 207
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.061684492975473404,
+      "learning_rate": 0.00019582071439011546,
+      "loss": 1.1318,
+      "step": 208
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.06065653637051582,
+      "learning_rate": 0.00019577869306633503,
+      "loss": 1.1281,
+      "step": 209
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.057913534343242645,
+      "learning_rate": 0.0001957364660971485,
+      "loss": 1.1318,
+      "step": 210
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.06302294880151749,
+      "learning_rate": 0.0001956940335732209,
+      "loss": 1.1971,
+      "step": 211
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0630715936422348,
+      "learning_rate": 0.0001956513955856587,
+      "loss": 1.215,
+      "step": 212
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.067079097032547,
+      "learning_rate": 0.00019560855222600947,
+      "loss": 1.1484,
+      "step": 213
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.05915733054280281,
+      "learning_rate": 0.0001955655035862617,
+      "loss": 1.0748,
+      "step": 214
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.07278767973184586,
+      "learning_rate": 0.00019552224975884466,
+      "loss": 1.0966,
+      "step": 215
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.05778686702251434,
+      "learning_rate": 0.00019547879083662819,
+      "loss": 1.0561,
+      "step": 216
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.06521858274936676,
+      "learning_rate": 0.00019543512691292251,
+      "loss": 1.1249,
+      "step": 217
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.058512818068265915,
+      "learning_rate": 0.0001953912580814779,
+      "loss": 1.0463,
+      "step": 218
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.06136133521795273,
+      "learning_rate": 0.00019534718443648473,
+      "loss": 1.1914,
+      "step": 219
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.05871862918138504,
+      "learning_rate": 0.000195302906072573,
+      "loss": 1.0887,
+      "step": 220
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.05922750383615494,
+      "learning_rate": 0.00019525842308481227,
+      "loss": 1.0888,
+      "step": 221
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.06273704022169113,
+      "learning_rate": 0.0001952137355687116,
+      "loss": 1.1218,
+      "step": 222
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.05759681016206741,
+      "learning_rate": 0.000195168843620219,
+      "loss": 1.1278,
+      "step": 223
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.060312505811452866,
+      "learning_rate": 0.00019512374733572153,
+      "loss": 1.1688,
+      "step": 224
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.06145263463258743,
+      "learning_rate": 0.0001950784468120449,
+      "loss": 1.1055,
+      "step": 225
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.06581958383321762,
+      "learning_rate": 0.00019503294214645337,
+      "loss": 1.2683,
+      "step": 226
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.06561966240406036,
+      "learning_rate": 0.00019498723343664955,
+      "loss": 1.2038,
+      "step": 227
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.06790629774332047,
+      "learning_rate": 0.00019494132078077414,
+      "loss": 1.1307,
+      "step": 228
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.06045442819595337,
+      "learning_rate": 0.0001948952042774057,
+      "loss": 1.1119,
+      "step": 229
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.06775711476802826,
+      "learning_rate": 0.00019484888402556045,
+      "loss": 1.1636,
+      "step": 230
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.06277008354663849,
+      "learning_rate": 0.00019480236012469213,
+      "loss": 1.1908,
+      "step": 231
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0631483793258667,
+      "learning_rate": 0.00019475563267469173,
+      "loss": 1.1828,
+      "step": 232
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.06697548180818558,
+      "learning_rate": 0.0001947087017758872,
+      "loss": 1.1107,
+      "step": 233
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.06013527512550354,
+      "learning_rate": 0.00019466156752904343,
+      "loss": 1.1176,
+      "step": 234
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.06709557771682739,
+      "learning_rate": 0.00019461423003536182,
+      "loss": 1.1575,
+      "step": 235
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.057948894798755646,
+      "learning_rate": 0.0001945666893964802,
+      "loss": 1.0455,
+      "step": 236
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.057657647877931595,
+      "learning_rate": 0.00019451894571447258,
+      "loss": 1.1221,
+      "step": 237
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.06317263096570969,
+      "learning_rate": 0.0001944709990918489,
+      "loss": 1.1305,
+      "step": 238
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.06375168263912201,
+      "learning_rate": 0.00019442284963155487,
+      "loss": 1.206,
+      "step": 239
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.06475285440683365,
+      "learning_rate": 0.00019437449743697164,
+      "loss": 1.204,
+      "step": 240
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.057945527136325836,
+      "learning_rate": 0.00019432594261191568,
+      "loss": 0.9143,
+      "step": 241
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.05976686626672745,
+      "learning_rate": 0.00019427718526063856,
+      "loss": 1.1177,
+      "step": 242
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.059271346777677536,
+      "learning_rate": 0.00019422822548782666,
+      "loss": 1.1089,
+      "step": 243
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.05958615615963936,
+      "learning_rate": 0.00019417906339860098,
+      "loss": 1.1451,
+      "step": 244
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.06455380469560623,
+      "learning_rate": 0.00019412969909851695,
+      "loss": 1.2423,
+      "step": 245
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0598178505897522,
+      "learning_rate": 0.00019408013269356408,
+      "loss": 1.081,
+      "step": 246
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.06671278178691864,
+      "learning_rate": 0.0001940303642901659,
+      "loss": 1.1094,
+      "step": 247
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.06148644909262657,
+      "learning_rate": 0.0001939803939951796,
+      "loss": 1.1045,
+      "step": 248
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.06747152656316757,
+      "learning_rate": 0.00019393022191589588,
+      "loss": 1.1443,
+      "step": 249
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.060876958072185516,
+      "learning_rate": 0.00019387984816003867,
+      "loss": 1.1341,
+      "step": 250
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.05667860805988312,
+      "learning_rate": 0.00019382927283576488,
+      "loss": 1.2032,
+      "step": 251
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.06507498770952225,
+      "learning_rate": 0.0001937784960516643,
+      "loss": 1.1355,
+      "step": 252
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.05810081586241722,
+      "learning_rate": 0.0001937275179167592,
+      "loss": 1.0911,
+      "step": 253
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.06210170313715935,
+      "learning_rate": 0.00019367633854050422,
+      "loss": 1.2247,
+      "step": 254
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.060430753976106644,
+      "learning_rate": 0.00019362495803278598,
+      "loss": 1.1633,
+      "step": 255
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0607365183532238,
+      "learning_rate": 0.0001935733765039231,
+      "loss": 1.1253,
+      "step": 256
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.062377460300922394,
+      "learning_rate": 0.00019352159406466564,
+      "loss": 1.1525,
+      "step": 257
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.06079326942563057,
+      "learning_rate": 0.00019346961082619522,
+      "loss": 1.067,
+      "step": 258
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.06413824111223221,
+      "learning_rate": 0.00019341742690012442,
+      "loss": 1.0536,
+      "step": 259
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.06453590095043182,
+      "learning_rate": 0.00019336504239849677,
+      "loss": 1.1118,
+      "step": 260
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.06608355045318604,
+      "learning_rate": 0.00019331245743378658,
+      "loss": 1.2306,
+      "step": 261
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.06034080311655998,
+      "learning_rate": 0.00019325967211889834,
+      "loss": 1.0643,
+      "step": 262
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.06985899806022644,
+      "learning_rate": 0.0001932066865671669,
+      "loss": 1.1907,
+      "step": 263
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.06518577039241791,
+      "learning_rate": 0.000193153500892357,
+      "loss": 1.0499,
+      "step": 264
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.05966721102595329,
+      "learning_rate": 0.00019310011520866298,
+      "loss": 1.2111,
+      "step": 265
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0628088116645813,
+      "learning_rate": 0.0001930465296307087,
+      "loss": 1.2063,
+      "step": 266
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.06214950606226921,
+      "learning_rate": 0.00019299274427354713,
+      "loss": 1.1141,
+      "step": 267
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.05994017794728279,
+      "learning_rate": 0.00019293875925266028,
+      "loss": 1.0661,
+      "step": 268
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0646560862660408,
+      "learning_rate": 0.0001928845746839588,
+      "loss": 1.1208,
+      "step": 269
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0596354603767395,
+      "learning_rate": 0.00019283019068378182,
+      "loss": 1.0839,
+      "step": 270
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06729655712842941,
+      "learning_rate": 0.00019277560736889656,
+      "loss": 1.1983,
+      "step": 271
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.060334715992212296,
+      "learning_rate": 0.0001927208248564984,
+      "loss": 1.1733,
+      "step": 272
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06451458483934402,
+      "learning_rate": 0.0001926658432642102,
+      "loss": 1.1861,
+      "step": 273
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0642094537615776,
+      "learning_rate": 0.00019261066271008235,
+      "loss": 1.2082,
+      "step": 274
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06414438039064407,
+      "learning_rate": 0.00019255528331259245,
+      "loss": 1.2229,
+      "step": 275
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06006895750761032,
+      "learning_rate": 0.00019249970519064503,
+      "loss": 1.1456,
+      "step": 276
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.06294744461774826,
+      "learning_rate": 0.00019244392846357123,
+      "loss": 1.072,
+      "step": 277
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.06234186887741089,
+      "learning_rate": 0.0001923879532511287,
+      "loss": 1.0956,
+      "step": 278
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.06028866767883301,
+      "learning_rate": 0.00019233177967350117,
+      "loss": 1.1874,
+      "step": 279
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.06322412937879562,
+      "learning_rate": 0.0001922754078512984,
+      "loss": 1.0697,
+      "step": 280
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.059378113597631454,
+      "learning_rate": 0.00019221883790555567,
+      "loss": 1.1663,
+      "step": 281
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.062111396342515945,
+      "learning_rate": 0.00019216206995773373,
+      "loss": 1.1077,
+      "step": 282
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.06687445938587189,
+      "learning_rate": 0.00019210510412971844,
+      "loss": 1.193,
+      "step": 283
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06198667362332344,
+      "learning_rate": 0.00019204794054382052,
+      "loss": 1.2258,
+      "step": 284
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.05913461372256279,
+      "learning_rate": 0.00019199057932277525,
+      "loss": 1.141,
+      "step": 285
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06464271247386932,
+      "learning_rate": 0.00019193302058974232,
+      "loss": 1.0863,
+      "step": 286
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.05979037657380104,
+      "learning_rate": 0.0001918752644683055,
+      "loss": 1.1686,
+      "step": 287
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06707989424467087,
+      "learning_rate": 0.00019181731108247228,
+      "loss": 1.2413,
+      "step": 288
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06815943866968155,
+      "learning_rate": 0.0001917591605566738,
+      "loss": 1.2022,
+      "step": 289
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06593851000070572,
+      "learning_rate": 0.00019170081301576444,
+      "loss": 1.0464,
+      "step": 290
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.06178038939833641,
+      "learning_rate": 0.00019164226858502153,
+      "loss": 1.1011,
+      "step": 291
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.06253184378147125,
+      "learning_rate": 0.00019158352739014523,
+      "loss": 1.0302,
+      "step": 292
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.06402178853750229,
+      "learning_rate": 0.0001915245895572581,
+      "loss": 1.1971,
+      "step": 293
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0673704445362091,
+      "learning_rate": 0.00019146545521290495,
+      "loss": 1.1555,
+      "step": 294
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.06606754660606384,
+      "learning_rate": 0.00019140612448405252,
+      "loss": 1.1057,
+      "step": 295
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.060698218643665314,
+      "learning_rate": 0.00019134659749808913,
+      "loss": 1.1279,
+      "step": 296
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.06506410986185074,
+      "learning_rate": 0.00019128687438282457,
+      "loss": 1.1735,
+      "step": 297
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.061889030039310455,
+      "learning_rate": 0.00019122695526648968,
+      "loss": 1.1158,
+      "step": 298
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06519250571727753,
+      "learning_rate": 0.0001911668402777362,
+      "loss": 1.1934,
+      "step": 299
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06114347651600838,
+      "learning_rate": 0.00019110652954563631,
+      "loss": 1.1105,
+      "step": 300
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06125333905220032,
+      "learning_rate": 0.0001910460231996826,
+      "loss": 1.0982,
+      "step": 301
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06343477964401245,
+      "learning_rate": 0.00019098532136978754,
+      "loss": 1.1342,
+      "step": 302
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.060326121747493744,
+      "learning_rate": 0.00019092442418628343,
+      "loss": 1.1897,
+      "step": 303
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06917954236268997,
+      "learning_rate": 0.00019086333177992191,
+      "loss": 1.0707,
+      "step": 304
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06384435296058655,
+      "learning_rate": 0.00019080204428187388,
+      "loss": 1.0565,
+      "step": 305
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.06388357281684875,
+      "learning_rate": 0.00019074056182372907,
+      "loss": 1.1176,
+      "step": 306
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0598420649766922,
+      "learning_rate": 0.00019067888453749575,
+      "loss": 1.1335,
+      "step": 307
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.05912897735834122,
+      "learning_rate": 0.0001906170125556006,
+      "loss": 1.0945,
+      "step": 308
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0659290999174118,
+      "learning_rate": 0.00019055494601088834,
+      "loss": 1.1361,
+      "step": 309
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0634288564324379,
+      "learning_rate": 0.00019049268503662126,
+      "loss": 1.1754,
+      "step": 310
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.06105343624949455,
+      "learning_rate": 0.00019043022976647934,
+      "loss": 1.1168,
+      "step": 311
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.06428004801273346,
+      "learning_rate": 0.00019036758033455956,
+      "loss": 1.0972,
+      "step": 312
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.061880819499492645,
+      "learning_rate": 0.0001903047368753759,
+      "loss": 1.1545,
+      "step": 313
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07274632155895233,
+      "learning_rate": 0.00019024169952385885,
+      "loss": 1.1257,
+      "step": 314
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06646310538053513,
+      "learning_rate": 0.00019017846841535522,
+      "loss": 1.154,
+      "step": 315
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06078438088297844,
+      "learning_rate": 0.00019011504368562782,
+      "loss": 1.1303,
+      "step": 316
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07181931287050247,
+      "learning_rate": 0.00019005142547085527,
+      "loss": 1.1603,
+      "step": 317
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06724654138088226,
+      "learning_rate": 0.00018998761390763154,
+      "loss": 1.1718,
+      "step": 318
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.05919702351093292,
+      "learning_rate": 0.00018992360913296574,
+      "loss": 1.142,
+      "step": 319
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06586825847625732,
+      "learning_rate": 0.00018985941128428185,
+      "loss": 1.1859,
+      "step": 320
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06922873854637146,
+      "learning_rate": 0.00018979502049941833,
+      "loss": 1.1,
+      "step": 321
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06558050960302353,
+      "learning_rate": 0.00018973043691662803,
+      "loss": 1.1329,
+      "step": 322
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06796015799045563,
+      "learning_rate": 0.0001896656606745776,
+      "loss": 1.1413,
+      "step": 323
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06253591179847717,
+      "learning_rate": 0.00018960069191234746,
+      "loss": 1.1416,
+      "step": 324
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07260395586490631,
+      "learning_rate": 0.00018953553076943134,
+      "loss": 1.2171,
+      "step": 325
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06394602358341217,
+      "learning_rate": 0.000189470177385736,
+      "loss": 1.1857,
+      "step": 326
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.06450013071298599,
+      "learning_rate": 0.000189404631901581,
+      "loss": 1.1278,
+      "step": 327
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.06406421959400177,
+      "learning_rate": 0.00018933889445769836,
+      "loss": 1.2068,
+      "step": 328
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.06315187364816666,
+      "learning_rate": 0.00018927296519523226,
+      "loss": 1.0789,
+      "step": 329
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.06743728369474411,
+      "learning_rate": 0.00018920684425573865,
+      "loss": 1.1594,
+      "step": 330
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.06548736989498138,
+      "learning_rate": 0.0001891405317811852,
+      "loss": 1.0963,
+      "step": 331
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.06221878528594971,
+      "learning_rate": 0.00018907402791395057,
+      "loss": 1.1802,
+      "step": 332
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.06024456024169922,
+      "learning_rate": 0.00018900733279682462,
+      "loss": 1.1534,
+      "step": 333
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0619971826672554,
+      "learning_rate": 0.00018894044657300765,
+      "loss": 1.1287,
+      "step": 334
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.061270926147699356,
+      "learning_rate": 0.0001888733693861104,
+      "loss": 1.1306,
+      "step": 335
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.06278882175683975,
+      "learning_rate": 0.00018880610138015356,
+      "loss": 1.0748,
+      "step": 336
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.061774447560310364,
+      "learning_rate": 0.00018873864269956756,
+      "loss": 1.083,
+      "step": 337
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.06414585560560226,
+      "learning_rate": 0.00018867099348919217,
+      "loss": 1.1351,
+      "step": 338
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.06463301926851273,
+      "learning_rate": 0.00018860315389427633,
+      "loss": 1.1633,
+      "step": 339
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.06183488667011261,
+      "learning_rate": 0.00018853512406047772,
+      "loss": 1.0311,
+      "step": 340
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.06121757999062538,
+      "learning_rate": 0.00018846690413386238,
+      "loss": 1.1787,
+      "step": 341
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06119801104068756,
+      "learning_rate": 0.0001883984942609047,
+      "loss": 1.1699,
+      "step": 342
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06786572933197021,
+      "learning_rate": 0.0001883298945884867,
+      "loss": 1.1863,
+      "step": 343
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06401295214891434,
+      "learning_rate": 0.00018826110526389803,
+      "loss": 1.1483,
+      "step": 344
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06333799660205841,
+      "learning_rate": 0.0001881921264348355,
+      "loss": 1.0793,
+      "step": 345
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06162017583847046,
+      "learning_rate": 0.00018812295824940285,
+      "loss": 1.1106,
+      "step": 346
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.05900746211409569,
+      "learning_rate": 0.00018805360085611032,
+      "loss": 1.1257,
+      "step": 347
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06269056349992752,
+      "learning_rate": 0.00018798405440387445,
+      "loss": 1.1076,
+      "step": 348
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.06586477160453796,
+      "learning_rate": 0.00018791431904201765,
+      "loss": 1.0705,
+      "step": 349
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07007557153701782,
+      "learning_rate": 0.00018784439492026798,
+      "loss": 1.2659,
+      "step": 350
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.06069363281130791,
+      "learning_rate": 0.00018777428218875878,
+      "loss": 1.0526,
+      "step": 351
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.06026845797896385,
+      "learning_rate": 0.00018770398099802836,
+      "loss": 1.1309,
+      "step": 352
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.061220645904541016,
+      "learning_rate": 0.0001876334914990196,
+      "loss": 1.0999,
+      "step": 353
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.06346994638442993,
+      "learning_rate": 0.00018756281384307982,
+      "loss": 1.0391,
+      "step": 354
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.06735444813966751,
+      "learning_rate": 0.00018749194818196022,
+      "loss": 1.1974,
+      "step": 355
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06252043694257736,
+      "learning_rate": 0.0001874208946678157,
+      "loss": 1.1771,
+      "step": 356
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06052526459097862,
+      "learning_rate": 0.0001873496534532045,
+      "loss": 1.1878,
+      "step": 357
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06671708077192307,
+      "learning_rate": 0.0001872782246910879,
+      "loss": 1.1576,
+      "step": 358
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06357080489397049,
+      "learning_rate": 0.00018720660853482977,
+      "loss": 1.164,
+      "step": 359
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06754573434591293,
+      "learning_rate": 0.00018713480513819644,
+      "loss": 1.1932,
+      "step": 360
+    },
+    {
+      "epoch": 0.5,
+      "eval_loss": 1.0889657735824585,
+      "eval_runtime": 97.0368,
+      "eval_samples_per_second": 51.527,
+      "eval_steps_per_second": 12.882,
+      "step": 360
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06117720529437065,
+      "learning_rate": 0.0001870628146553562,
+      "loss": 1.0225,
+      "step": 361
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06431601196527481,
+      "learning_rate": 0.00018699063724087904,
+      "loss": 1.1954,
+      "step": 362
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.06532489508390427,
+      "learning_rate": 0.0001869182730497363,
+      "loss": 1.1475,
+      "step": 363
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.06369902938604355,
+      "learning_rate": 0.00018684572223730045,
+      "loss": 1.0935,
+      "step": 364
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.06582259386777878,
+      "learning_rate": 0.0001867729849593444,
+      "loss": 1.163,
+      "step": 365
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.06448414921760559,
+      "learning_rate": 0.0001867000613720417,
+      "loss": 1.1137,
+      "step": 366
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0670081228017807,
+      "learning_rate": 0.00018662695163196579,
+      "loss": 1.1921,
+      "step": 367
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.06068355217576027,
+      "learning_rate": 0.0001865536558960898,
+      "loss": 0.9705,
+      "step": 368
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.06467951089143753,
+      "learning_rate": 0.0001864801743217862,
+      "loss": 1.0868,
+      "step": 369
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.06502547860145569,
+      "learning_rate": 0.0001864065070668265,
+      "loss": 1.0937,
+      "step": 370
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07280650734901428,
+      "learning_rate": 0.00018633265428938088,
+      "loss": 1.1701,
+      "step": 371
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.05973858758807182,
+      "learning_rate": 0.00018625861614801785,
+      "loss": 1.0324,
+      "step": 372
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.06514280289411545,
+      "learning_rate": 0.00018618439280170392,
+      "loss": 1.1338,
+      "step": 373
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.06471429020166397,
+      "learning_rate": 0.00018610998440980324,
+      "loss": 1.1876,
+      "step": 374
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0652419924736023,
+      "learning_rate": 0.00018603539113207722,
+      "loss": 1.2204,
+      "step": 375
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.06470503658056259,
+      "learning_rate": 0.0001859606131286843,
+      "loss": 1.2483,
+      "step": 376
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.06816300749778748,
+      "learning_rate": 0.00018588565056017965,
+      "loss": 1.1036,
+      "step": 377
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.062083158642053604,
+      "learning_rate": 0.00018581050358751445,
+      "loss": 1.1232,
+      "step": 378
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.06199049949645996,
+      "learning_rate": 0.00018573517237203602,
+      "loss": 1.0623,
+      "step": 379
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.06976150721311569,
+      "learning_rate": 0.0001856596570754872,
+      "loss": 1.0361,
+      "step": 380
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.06474665552377701,
+      "learning_rate": 0.0001855839578600061,
+      "loss": 1.1786,
+      "step": 381
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.06963939964771271,
+      "learning_rate": 0.00018550807488812562,
+      "loss": 1.1381,
+      "step": 382
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.06306161731481552,
+      "learning_rate": 0.00018543200832277337,
+      "loss": 1.1783,
+      "step": 383
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.06826300919055939,
+      "learning_rate": 0.00018535575832727102,
+      "loss": 1.1401,
+      "step": 384
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.061530325561761856,
+      "learning_rate": 0.0001852793250653341,
+      "loss": 1.0992,
+      "step": 385
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.06267255544662476,
+      "learning_rate": 0.00018520270870107166,
+      "loss": 1.0652,
+      "step": 386
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.059738241136074066,
+      "learning_rate": 0.0001851259093989859,
+      "loss": 1.1092,
+      "step": 387
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.062138039618730545,
+      "learning_rate": 0.00018504892732397173,
+      "loss": 1.0607,
+      "step": 388
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.06500416994094849,
+      "learning_rate": 0.00018497176264131656,
+      "loss": 1.0877,
+      "step": 389
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.06768455356359482,
+      "learning_rate": 0.00018489441551669986,
+      "loss": 1.2099,
+      "step": 390
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.06814403086900711,
+      "learning_rate": 0.00018481688611619285,
+      "loss": 1.1743,
+      "step": 391
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.06558533012866974,
+      "learning_rate": 0.00018473917460625798,
+      "loss": 1.0566,
+      "step": 392
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.06515612453222275,
+      "learning_rate": 0.00018466128115374888,
+      "loss": 1.1267,
+      "step": 393
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.06616493314504623,
+      "learning_rate": 0.00018458320592590975,
+      "loss": 1.1703,
+      "step": 394
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.06906623393297195,
+      "learning_rate": 0.00018450494909037502,
+      "loss": 1.2327,
+      "step": 395
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.06478845328092575,
+      "learning_rate": 0.00018442651081516917,
+      "loss": 1.1189,
+      "step": 396
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0638127252459526,
+      "learning_rate": 0.00018434789126870612,
+      "loss": 1.1117,
+      "step": 397
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.06424181908369064,
+      "learning_rate": 0.00018426909061978908,
+      "loss": 1.1469,
+      "step": 398
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.06596982479095459,
+      "learning_rate": 0.0001841901090376101,
+      "loss": 1.1003,
+      "step": 399
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.06202305108308792,
+      "learning_rate": 0.00018411094669174965,
+      "loss": 1.0425,
+      "step": 400
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0672823116183281,
+      "learning_rate": 0.00018403160375217637,
+      "loss": 1.1015,
+      "step": 401
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.06283591687679291,
+      "learning_rate": 0.00018395208038924667,
+      "loss": 1.0863,
+      "step": 402
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.06623289734125137,
+      "learning_rate": 0.00018387237677370427,
+      "loss": 1.0828,
+      "step": 403
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.06633896380662918,
+      "learning_rate": 0.00018379249307667994,
+      "loss": 1.145,
+      "step": 404
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.06487153470516205,
+      "learning_rate": 0.00018371242946969117,
+      "loss": 1.1447,
+      "step": 405
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.06051245704293251,
+      "learning_rate": 0.00018363218612464158,
+      "loss": 1.0637,
+      "step": 406
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.05917385593056679,
+      "learning_rate": 0.00018355176321382087,
+      "loss": 1.1459,
+      "step": 407
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.061461642384529114,
+      "learning_rate": 0.00018347116090990424,
+      "loss": 1.0601,
+      "step": 408
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.06090409681200981,
+      "learning_rate": 0.00018339037938595193,
+      "loss": 1.0531,
+      "step": 409
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.0663984939455986,
+      "learning_rate": 0.00018330941881540915,
+      "loss": 1.1315,
+      "step": 410
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.06271735578775406,
+      "learning_rate": 0.0001832282793721055,
+      "loss": 1.0869,
+      "step": 411
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.06272199004888535,
+      "learning_rate": 0.00018314696123025454,
+      "loss": 1.1414,
+      "step": 412
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.06375264376401901,
+      "learning_rate": 0.0001830654645644536,
+      "loss": 1.0946,
+      "step": 413
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.06219152733683586,
+      "learning_rate": 0.00018298378954968337,
+      "loss": 1.0448,
+      "step": 414
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.06392863392829895,
+      "learning_rate": 0.00018290193636130727,
+      "loss": 1.1606,
+      "step": 415
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.06729146838188171,
+      "learning_rate": 0.00018281990517507156,
+      "loss": 1.1532,
+      "step": 416
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.0799422636628151,
+      "learning_rate": 0.00018273769616710436,
+      "loss": 1.1133,
+      "step": 417
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.06902763992547989,
+      "learning_rate": 0.0001826553095139159,
+      "loss": 1.1404,
+      "step": 418
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.06674634665250778,
+      "learning_rate": 0.00018257274539239756,
+      "loss": 1.2019,
+      "step": 419
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.06378789991140366,
+      "learning_rate": 0.00018249000397982195,
+      "loss": 1.1069,
+      "step": 420
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.06818845123052597,
+      "learning_rate": 0.0001824070854538422,
+      "loss": 1.1844,
+      "step": 421
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07562380284070969,
+      "learning_rate": 0.00018232398999249192,
+      "loss": 1.0978,
+      "step": 422
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07126368582248688,
+      "learning_rate": 0.00018224071777418437,
+      "loss": 1.2169,
+      "step": 423
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.06599973887205124,
+      "learning_rate": 0.00018215726897771248,
+      "loss": 1.1339,
+      "step": 424
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.06484373658895493,
+      "learning_rate": 0.00018207364378224826,
+      "loss": 1.1609,
+      "step": 425
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.062379587441682816,
+      "learning_rate": 0.00018198984236734246,
+      "loss": 1.1661,
+      "step": 426
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0662432610988617,
+      "learning_rate": 0.00018190586491292424,
+      "loss": 1.1372,
+      "step": 427
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.06816010177135468,
+      "learning_rate": 0.00018182171159930065,
+      "loss": 1.1678,
+      "step": 428
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.0649806335568428,
+      "learning_rate": 0.0001817373826071564,
+      "loss": 1.1688,
+      "step": 429
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.06410276144742966,
+      "learning_rate": 0.0001816528781175533,
+      "loss": 1.1723,
+      "step": 430
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.06425531208515167,
+      "learning_rate": 0.00018156819831193012,
+      "loss": 1.148,
+      "step": 431
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.06638933718204498,
+      "learning_rate": 0.00018148334337210193,
+      "loss": 1.0427,
+      "step": 432
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.06576802581548691,
+      "learning_rate": 0.00018139831348025988,
+      "loss": 1.0814,
+      "step": 433
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.06729135662317276,
+      "learning_rate": 0.0001813131088189707,
+      "loss": 1.2238,
+      "step": 434
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.066316619515419,
+      "learning_rate": 0.00018122772957117645,
+      "loss": 1.124,
+      "step": 435
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.0657489225268364,
+      "learning_rate": 0.00018114217592019393,
+      "loss": 1.1616,
+      "step": 436
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.06329985707998276,
+      "learning_rate": 0.00018105644804971454,
+      "loss": 1.1662,
+      "step": 437
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.06630551815032959,
+      "learning_rate": 0.00018097054614380365,
+      "loss": 1.0863,
+      "step": 438
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.0653151199221611,
+      "learning_rate": 0.00018088447038690033,
+      "loss": 1.0568,
+      "step": 439
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.06651636958122253,
+      "learning_rate": 0.00018079822096381688,
+      "loss": 1.1219,
+      "step": 440
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.06870090961456299,
+      "learning_rate": 0.00018071179805973856,
+      "loss": 1.1327,
+      "step": 441
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.06400321424007416,
+      "learning_rate": 0.000180625201860223,
+      "loss": 1.0736,
+      "step": 442
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.06521004438400269,
+      "learning_rate": 0.00018053843255120002,
+      "loss": 1.0366,
+      "step": 443
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.0785718560218811,
+      "learning_rate": 0.0001804514903189711,
+      "loss": 1.0535,
+      "step": 444
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.0732305571436882,
+      "learning_rate": 0.0001803643753502089,
+      "loss": 1.0828,
+      "step": 445
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.06389783322811127,
+      "learning_rate": 0.0001802770878319571,
+      "loss": 1.0966,
+      "step": 446
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.06395277380943298,
+      "learning_rate": 0.00018018962795162977,
+      "loss": 1.1414,
+      "step": 447
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.06490381062030792,
+      "learning_rate": 0.00018010199589701107,
+      "loss": 1.1405,
+      "step": 448
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.06930892169475555,
+      "learning_rate": 0.00018001419185625488,
+      "loss": 1.0145,
+      "step": 449
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07346969097852707,
+      "learning_rate": 0.00017992621601788428,
+      "loss": 1.1745,
+      "step": 450
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07306042313575745,
+      "learning_rate": 0.0001798380685707913,
+      "loss": 1.1333,
+      "step": 451
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.06085991859436035,
+      "learning_rate": 0.00017974974970423632,
+      "loss": 1.0801,
+      "step": 452
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.0631103590130806,
+      "learning_rate": 0.00017966125960784787,
+      "loss": 1.1209,
+      "step": 453
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.06399189680814743,
+      "learning_rate": 0.00017957259847162205,
+      "loss": 1.1683,
+      "step": 454
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.06143182888627052,
+      "learning_rate": 0.00017948376648592232,
+      "loss": 1.0855,
+      "step": 455
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.06295006722211838,
+      "learning_rate": 0.00017939476384147877,
+      "loss": 1.1398,
+      "step": 456
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.06850682944059372,
+      "learning_rate": 0.0001793055907293881,
+      "loss": 1.1245,
+      "step": 457
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.061626896262168884,
+      "learning_rate": 0.00017921624734111292,
+      "loss": 1.0763,
+      "step": 458
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.06570126861333847,
+      "learning_rate": 0.00017912673386848152,
+      "loss": 1.1547,
+      "step": 459
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.06566940248012543,
+      "learning_rate": 0.00017903705050368722,
+      "loss": 1.1218,
+      "step": 460
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.06877337396144867,
+      "learning_rate": 0.00017894719743928827,
+      "loss": 1.2173,
+      "step": 461
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.06467350572347641,
+      "learning_rate": 0.00017885717486820722,
+      "loss": 1.1347,
+      "step": 462
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08330734819173813,
+      "learning_rate": 0.00017876698298373053,
+      "loss": 1.1373,
+      "step": 463
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.06419449299573898,
+      "learning_rate": 0.0001786766219795083,
+      "loss": 1.1285,
+      "step": 464
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07984116673469543,
+      "learning_rate": 0.0001785860920495536,
+      "loss": 1.1309,
+      "step": 465
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07000021636486053,
+      "learning_rate": 0.00017849539338824231,
+      "loss": 1.1193,
+      "step": 466
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07003198564052582,
+      "learning_rate": 0.00017840452619031258,
+      "loss": 1.1596,
+      "step": 467
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07209260761737823,
+      "learning_rate": 0.00017831349065086435,
+      "loss": 1.2199,
+      "step": 468
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.06799673289060593,
+      "learning_rate": 0.00017822228696535907,
+      "loss": 1.1247,
+      "step": 469
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.06913288682699203,
+      "learning_rate": 0.0001781309153296192,
+      "loss": 1.1035,
+      "step": 470
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.06872755289077759,
+      "learning_rate": 0.0001780393759398278,
+      "loss": 1.1613,
+      "step": 471
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.06622165441513062,
+      "learning_rate": 0.00017794766899252812,
+      "loss": 1.0869,
+      "step": 472
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.06728823482990265,
+      "learning_rate": 0.00017785579468462316,
+      "loss": 1.1408,
+      "step": 473
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.06555955857038498,
+      "learning_rate": 0.00017776375321337521,
+      "loss": 1.1584,
+      "step": 474
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.0661938264966011,
+      "learning_rate": 0.0001776715447764056,
+      "loss": 1.1542,
+      "step": 475
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07007063180208206,
+      "learning_rate": 0.00017757916957169404,
+      "loss": 1.1424,
+      "step": 476
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.06441737711429596,
+      "learning_rate": 0.0001774866277975784,
+      "loss": 1.1361,
+      "step": 477
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.06655770540237427,
+      "learning_rate": 0.00017739391965275404,
+      "loss": 1.0968,
+      "step": 478
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07319090515375137,
+      "learning_rate": 0.0001773010453362737,
+      "loss": 1.107,
+      "step": 479
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.06766431778669357,
+      "learning_rate": 0.0001772080050475468,
+      "loss": 1.1571,
+      "step": 480
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.06634720414876938,
+      "learning_rate": 0.00017711479898633914,
+      "loss": 1.175,
+      "step": 481
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.06589143723249435,
+      "learning_rate": 0.00017702142735277247,
+      "loss": 1.1475,
+      "step": 482
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.06698860973119736,
+      "learning_rate": 0.00017692789034732403,
+      "loss": 1.1042,
+      "step": 483
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.06417112797498703,
+      "learning_rate": 0.0001768341881708261,
+      "loss": 1.0588,
+      "step": 484
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.06298335641622543,
+      "learning_rate": 0.00017674032102446563,
+      "loss": 0.9964,
+      "step": 485
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.06435716152191162,
+      "learning_rate": 0.00017664628910978375,
+      "loss": 1.1126,
+      "step": 486
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.06721868366003036,
+      "learning_rate": 0.00017655209262867541,
+      "loss": 1.1541,
+      "step": 487
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.06489331275224686,
+      "learning_rate": 0.00017645773178338886,
+      "loss": 1.0539,
+      "step": 488
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.06617634743452072,
+      "learning_rate": 0.00017636320677652526,
+      "loss": 1.094,
+      "step": 489
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.06776611506938934,
+      "learning_rate": 0.0001762685178110382,
+      "loss": 1.0155,
+      "step": 490
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.06376908719539642,
+      "learning_rate": 0.00017617366509023338,
+      "loss": 1.2292,
+      "step": 491
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.06204572319984436,
+      "learning_rate": 0.00017607864881776807,
+      "loss": 1.0968,
+      "step": 492
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.07098660618066788,
+      "learning_rate": 0.00017598346919765067,
+      "loss": 1.1067,
+      "step": 493
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.06708233058452606,
+      "learning_rate": 0.00017588812643424032,
+      "loss": 1.126,
+      "step": 494
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.0708741545677185,
+      "learning_rate": 0.00017579262073224647,
+      "loss": 1.0234,
+      "step": 495
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.06812895089387894,
+      "learning_rate": 0.00017569695229672835,
+      "loss": 1.14,
+      "step": 496
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.06947667896747589,
+      "learning_rate": 0.00017560112133309467,
+      "loss": 1.1888,
+      "step": 497
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.06499818712472916,
+      "learning_rate": 0.0001755051280471031,
+      "loss": 1.0889,
+      "step": 498
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.063695028424263,
+      "learning_rate": 0.00017540897264485967,
+      "loss": 1.0678,
+      "step": 499
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.06530368328094482,
+      "learning_rate": 0.00017531265533281872,
+      "loss": 1.0348,
+      "step": 500
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07548119127750397,
+      "learning_rate": 0.0001752161763177821,
+      "loss": 1.1479,
+      "step": 501
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.06523353606462479,
+      "learning_rate": 0.00017511953580689888,
+      "loss": 1.1491,
+      "step": 502
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.06849244982004166,
+      "learning_rate": 0.00017502273400766484,
+      "loss": 1.143,
+      "step": 503
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.06976161897182465,
+      "learning_rate": 0.00017492577112792208,
+      "loss": 1.1597,
+      "step": 504
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.06484940648078918,
+      "learning_rate": 0.0001748286473758586,
+      "loss": 1.1013,
+      "step": 505
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.06747712939977646,
+      "learning_rate": 0.00017473136296000772,
+      "loss": 1.1129,
+      "step": 506
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.06816747039556503,
+      "learning_rate": 0.0001746339180892478,
+      "loss": 1.2297,
+      "step": 507
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.06335021555423737,
+      "learning_rate": 0.00017453631297280166,
+      "loss": 1.0683,
+      "step": 508
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.06458504498004913,
+      "learning_rate": 0.00017443854782023624,
+      "loss": 1.0101,
+      "step": 509
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.0690096914768219,
+      "learning_rate": 0.000174340622841462,
+      "loss": 1.1363,
+      "step": 510
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.06535279005765915,
+      "learning_rate": 0.00017424253824673263,
+      "loss": 1.0364,
+      "step": 511
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.0651826336979866,
+      "learning_rate": 0.00017414429424664454,
+      "loss": 1.1102,
+      "step": 512
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.06632277369499207,
+      "learning_rate": 0.00017404589105213632,
+      "loss": 1.1461,
+      "step": 513
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.06560949981212616,
+      "learning_rate": 0.00017394732887448847,
+      "loss": 1.1939,
+      "step": 514
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.06644757837057114,
+      "learning_rate": 0.00017384860792532276,
+      "loss": 1.1603,
+      "step": 515
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.06334047764539719,
+      "learning_rate": 0.00017374972841660186,
+      "loss": 1.1102,
+      "step": 516
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.061144351959228516,
+      "learning_rate": 0.00017365069056062894,
+      "loss": 1.0735,
+      "step": 517
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.0644732415676117,
+      "learning_rate": 0.00017355149457004709,
+      "loss": 1.1449,
+      "step": 518
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.06779829412698746,
+      "learning_rate": 0.00017345214065783897,
+      "loss": 1.0644,
+      "step": 519
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.06311120837926865,
+      "learning_rate": 0.00017335262903732634,
+      "loss": 1.0737,
+      "step": 520
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.070918507874012,
+      "learning_rate": 0.0001732529599221695,
+      "loss": 1.002,
+      "step": 521
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.06275039911270142,
+      "learning_rate": 0.0001731531335263669,
+      "loss": 1.0902,
+      "step": 522
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.06746424734592438,
+      "learning_rate": 0.0001730531500642548,
+      "loss": 1.1495,
+      "step": 523
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07816082239151001,
+      "learning_rate": 0.00017295300975050658,
+      "loss": 1.0652,
+      "step": 524
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.06773856282234192,
+      "learning_rate": 0.00017285271280013245,
+      "loss": 1.1856,
+      "step": 525
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0698360875248909,
+      "learning_rate": 0.0001727522594284789,
+      "loss": 1.201,
+      "step": 526
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07427448779344559,
+      "learning_rate": 0.0001726516498512283,
+      "loss": 1.116,
+      "step": 527
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.06394834816455841,
+      "learning_rate": 0.00017255088428439836,
+      "loss": 1.1401,
+      "step": 528
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.06733064353466034,
+      "learning_rate": 0.0001724499629443418,
+      "loss": 1.1239,
+      "step": 529
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.06488535553216934,
+      "learning_rate": 0.00017234888604774574,
+      "loss": 1.1434,
+      "step": 530
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.0653986781835556,
+      "learning_rate": 0.00017224765381163132,
+      "loss": 1.1392,
+      "step": 531
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07161843031644821,
+      "learning_rate": 0.00017214626645335314,
+      "loss": 1.0808,
+      "step": 532
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.06368163228034973,
+      "learning_rate": 0.00017204472419059895,
+      "loss": 1.1395,
+      "step": 533
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07309909909963608,
+      "learning_rate": 0.00017194302724138903,
+      "loss": 1.1411,
+      "step": 534
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.0663491040468216,
+      "learning_rate": 0.00017184117582407578,
+      "loss": 1.1509,
+      "step": 535
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.06098710745573044,
+      "learning_rate": 0.00017173917015734336,
+      "loss": 1.0276,
+      "step": 536
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.0714559257030487,
+      "learning_rate": 0.00017163701046020707,
+      "loss": 1.2508,
+      "step": 537
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.06662353128194809,
+      "learning_rate": 0.00017153469695201277,
+      "loss": 1.0658,
+      "step": 538
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.0661638155579567,
+      "learning_rate": 0.0001714322298524368,
+      "loss": 1.1066,
+      "step": 539
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07075709104537964,
+      "learning_rate": 0.00017132960938148512,
+      "loss": 1.1827,
+      "step": 540
+    },
+    {
+      "epoch": 0.75,
+      "eval_loss": 1.0872660875320435,
+      "eval_runtime": 96.8426,
+      "eval_samples_per_second": 51.63,
+      "eval_steps_per_second": 12.908,
+      "step": 540
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07708127051591873,
+      "learning_rate": 0.00017122683575949307,
+      "loss": 1.2037,
+      "step": 541
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.06289788335561752,
+      "learning_rate": 0.0001711239092071248,
+      "loss": 1.0755,
+      "step": 542
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0677611231803894,
+      "learning_rate": 0.00017102082994537282,
+      "loss": 1.1438,
+      "step": 543
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.06381526589393616,
+      "learning_rate": 0.00017091759819555744,
+      "loss": 1.1003,
+      "step": 544
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.06795594096183777,
+      "learning_rate": 0.00017081421417932652,
+      "loss": 1.1063,
+      "step": 545
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.06619235128164291,
+      "learning_rate": 0.00017071067811865476,
+      "loss": 1.1499,
+      "step": 546
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.06958078593015671,
+      "learning_rate": 0.0001706069902358433,
+      "loss": 1.0918,
+      "step": 547
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.06839771568775177,
+      "learning_rate": 0.0001705031507535193,
+      "loss": 1.0789,
+      "step": 548
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0677015408873558,
+      "learning_rate": 0.00017039915989463538,
+      "loss": 1.1419,
+      "step": 549
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.0648641586303711,
+      "learning_rate": 0.00017029501788246924,
+      "loss": 1.1461,
+      "step": 550
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.06705692410469055,
+      "learning_rate": 0.00017019072494062304,
+      "loss": 1.1288,
+      "step": 551
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.06862998753786087,
+      "learning_rate": 0.00017008628129302307,
+      "loss": 1.107,
+      "step": 552
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07801419496536255,
+      "learning_rate": 0.00016998168716391915,
+      "loss": 1.097,
+      "step": 553
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07198420912027359,
+      "learning_rate": 0.00016987694277788417,
+      "loss": 1.1328,
+      "step": 554
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.06974530965089798,
+      "learning_rate": 0.00016977204835981374,
+      "loss": 1.233,
+      "step": 555
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.06750082969665527,
+      "learning_rate": 0.00016966700413492556,
+      "loss": 1.1373,
+      "step": 556
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.06531986594200134,
+      "learning_rate": 0.00016956181032875894,
+      "loss": 1.0542,
+      "step": 557
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.06485234200954437,
+      "learning_rate": 0.0001694564671671743,
+      "loss": 1.0888,
+      "step": 558
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.06844603270292282,
+      "learning_rate": 0.0001693509748763529,
+      "loss": 1.1358,
+      "step": 559
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.06846935302019119,
+      "learning_rate": 0.00016924533368279607,
+      "loss": 1.186,
+      "step": 560
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07385078817605972,
+      "learning_rate": 0.00016913954381332492,
+      "loss": 1.137,
+      "step": 561
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.06992754340171814,
+      "learning_rate": 0.0001690336054950797,
+      "loss": 1.1749,
+      "step": 562
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.06728782504796982,
+      "learning_rate": 0.0001689275189555195,
+      "loss": 1.041,
+      "step": 563
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.0680394098162651,
+      "learning_rate": 0.00016882128442242156,
+      "loss": 1.171,
+      "step": 564
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.06516265124082565,
+      "learning_rate": 0.00016871490212388087,
+      "loss": 1.126,
+      "step": 565
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.06674471497535706,
+      "learning_rate": 0.00016860837228830974,
+      "loss": 1.1474,
+      "step": 566
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.06331098824739456,
+      "learning_rate": 0.00016850169514443728,
+      "loss": 1.1349,
+      "step": 567
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.06607010960578918,
+      "learning_rate": 0.00016839487092130883,
+      "loss": 1.1848,
+      "step": 568
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.06833096593618393,
+      "learning_rate": 0.0001682878998482855,
+      "loss": 1.1646,
+      "step": 569
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.06246986240148544,
+      "learning_rate": 0.0001681807821550438,
+      "loss": 1.0467,
+      "step": 570
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.06146036833524704,
+      "learning_rate": 0.0001680735180715749,
+      "loss": 0.9574,
+      "step": 571
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07327894866466522,
+      "learning_rate": 0.00016796610782818442,
+      "loss": 1.1735,
+      "step": 572
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.06582464277744293,
+      "learning_rate": 0.00016785855165549175,
+      "loss": 1.0437,
+      "step": 573
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.06511417776346207,
+      "learning_rate": 0.00016775084978442955,
+      "loss": 1.0813,
+      "step": 574
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.06808367371559143,
+      "learning_rate": 0.00016764300244624338,
+      "loss": 1.0771,
+      "step": 575
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.06845704466104507,
+      "learning_rate": 0.0001675350098724911,
+      "loss": 1.1645,
+      "step": 576
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.0673985630273819,
+      "learning_rate": 0.00016742687229504246,
+      "loss": 1.0358,
+      "step": 577
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.0681719034910202,
+      "learning_rate": 0.00016731858994607838,
+      "loss": 1.0971,
+      "step": 578
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.00016721016305809084,
+      "loss": 1.242,
+      "step": 579
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07134019583463669,
+      "learning_rate": 0.00016710159186388203,
+      "loss": 1.0486,
+      "step": 580
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07230845093727112,
+      "learning_rate": 0.00016699287659656395,
+      "loss": 1.1678,
+      "step": 581
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.06907068192958832,
+      "learning_rate": 0.00016688401748955802,
+      "loss": 1.1183,
+      "step": 582
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.06549103558063507,
+      "learning_rate": 0.00016677501477659445,
+      "loss": 1.1073,
+      "step": 583
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.06864264607429504,
+      "learning_rate": 0.0001666658686917118,
+      "loss": 1.1528,
+      "step": 584
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.06810259073972702,
+      "learning_rate": 0.00016655657946925646,
+      "loss": 1.1308,
+      "step": 585
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.06780505925416946,
+      "learning_rate": 0.00016644714734388217,
+      "loss": 1.16,
+      "step": 586
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.06738951057195663,
+      "learning_rate": 0.00016633757255054947,
+      "loss": 1.1706,
+      "step": 587
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.06679335236549377,
+      "learning_rate": 0.0001662278553245252,
+      "loss": 1.1704,
+      "step": 588
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.06979308277368546,
+      "learning_rate": 0.00016611799590138203,
+      "loss": 1.1482,
+      "step": 589
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.0626835972070694,
+      "learning_rate": 0.00016600799451699802,
+      "loss": 1.1115,
+      "step": 590
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.06404287368059158,
+      "learning_rate": 0.0001658978514075559,
+      "loss": 1.175,
+      "step": 591
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.07459414005279541,
+      "learning_rate": 0.00016578756680954277,
+      "loss": 1.1302,
+      "step": 592
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.0666908472776413,
+      "learning_rate": 0.00016567714095974952,
+      "loss": 1.1636,
+      "step": 593
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.06724884361028671,
+      "learning_rate": 0.0001655665740952703,
+      "loss": 1.1355,
+      "step": 594
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.0689745619893074,
+      "learning_rate": 0.00016545586645350205,
+      "loss": 1.1162,
+      "step": 595
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.06672396510839462,
+      "learning_rate": 0.000165345018272144,
+      "loss": 1.0862,
+      "step": 596
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.06393066048622131,
+      "learning_rate": 0.00016523402978919704,
+      "loss": 1.1315,
+      "step": 597
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.07766008377075195,
+      "learning_rate": 0.00016512290124296336,
+      "loss": 1.1024,
+      "step": 598
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.07702287286520004,
+      "learning_rate": 0.00016501163287204588,
+      "loss": 1.1509,
+      "step": 599
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.06490621715784073,
+      "learning_rate": 0.00016490022491534768,
+      "loss": 1.0167,
+      "step": 600
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.06604059785604477,
+      "learning_rate": 0.0001647886776120717,
+      "loss": 1.106,
+      "step": 601
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.06731244921684265,
+      "learning_rate": 0.00016467699120171987,
+      "loss": 1.0898,
+      "step": 602
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.07821813970804214,
+      "learning_rate": 0.00016456516592409294,
+      "loss": 1.1176,
+      "step": 603
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.06816477328538895,
+      "learning_rate": 0.0001644532020192897,
+      "loss": 1.1786,
+      "step": 604
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.06874791532754898,
+      "learning_rate": 0.0001643410997277067,
+      "loss": 1.0926,
+      "step": 605
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.06932558119297028,
+      "learning_rate": 0.00016422885929003758,
+      "loss": 1.1679,
+      "step": 606
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06875893473625183,
+      "learning_rate": 0.00016411648094727253,
+      "loss": 1.1935,
+      "step": 607
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06305741518735886,
+      "learning_rate": 0.00016400396494069792,
+      "loss": 1.0654,
+      "step": 608
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06925616413354874,
+      "learning_rate": 0.00016389131151189567,
+      "loss": 1.1218,
+      "step": 609
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.07191671431064606,
+      "learning_rate": 0.00016377852090274276,
+      "loss": 1.1875,
+      "step": 610
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06721782684326172,
+      "learning_rate": 0.00016366559335541066,
+      "loss": 1.2297,
+      "step": 611
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.07050056755542755,
+      "learning_rate": 0.00016355252911236492,
+      "loss": 1.2071,
+      "step": 612
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.0658205896615982,
+      "learning_rate": 0.00016343932841636456,
+      "loss": 1.1652,
+      "step": 613
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.07456327974796295,
+      "learning_rate": 0.0001633259915104616,
+      "loss": 1.0894,
+      "step": 614
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06406401097774506,
+      "learning_rate": 0.00016321251863800045,
+      "loss": 1.0864,
+      "step": 615
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06928929686546326,
+      "learning_rate": 0.00016309891004261755,
+      "loss": 1.0215,
+      "step": 616
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06436607241630554,
+      "learning_rate": 0.00016298516596824068,
+      "loss": 1.1855,
+      "step": 617
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06957562267780304,
+      "learning_rate": 0.0001628712866590885,
+      "loss": 1.2085,
+      "step": 618
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.07085030525922775,
+      "learning_rate": 0.0001627572723596701,
+      "loss": 1.1079,
+      "step": 619
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06608451902866364,
+      "learning_rate": 0.0001626431233147843,
+      "loss": 1.1456,
+      "step": 620
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.0686989277601242,
+      "learning_rate": 0.00016252883976951938,
+      "loss": 1.1112,
+      "step": 621
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06871567666530609,
+      "learning_rate": 0.00016241442196925223,
+      "loss": 1.1612,
+      "step": 622
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06559213250875473,
+      "learning_rate": 0.00016229987015964814,
+      "loss": 1.1588,
+      "step": 623
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06943194568157196,
+      "learning_rate": 0.00016218518458666008,
+      "loss": 1.1108,
+      "step": 624
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06725773215293884,
+      "learning_rate": 0.00016207036549652814,
+      "loss": 1.1942,
+      "step": 625
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06677956134080887,
+      "learning_rate": 0.00016195541313577923,
+      "loss": 1.129,
+      "step": 626
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06949102878570557,
+      "learning_rate": 0.00016184032775122631,
+      "loss": 1.0889,
+      "step": 627
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06555166095495224,
+      "learning_rate": 0.00016172510958996795,
+      "loss": 1.1484,
+      "step": 628
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.07066816091537476,
+      "learning_rate": 0.0001616097588993879,
+      "loss": 1.0807,
+      "step": 629
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06895067542791367,
+      "learning_rate": 0.00016149427592715432,
+      "loss": 1.1586,
+      "step": 630
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.07003293186426163,
+      "learning_rate": 0.00016137866092121953,
+      "loss": 1.1059,
+      "step": 631
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06466325372457504,
+      "learning_rate": 0.0001612629141298192,
+      "loss": 1.1759,
+      "step": 632
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06705661863088608,
+      "learning_rate": 0.00016114703580147202,
+      "loss": 1.1766,
+      "step": 633
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.07274851948022842,
+      "learning_rate": 0.00016103102618497922,
+      "loss": 1.211,
+      "step": 634
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06761591881513596,
+      "learning_rate": 0.0001609148855294237,
+      "loss": 1.1049,
+      "step": 635
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06395487487316132,
+      "learning_rate": 0.00016079861408416985,
+      "loss": 1.0404,
+      "step": 636
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.07156820595264435,
+      "learning_rate": 0.00016068221209886288,
+      "loss": 1.1569,
+      "step": 637
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06414105743169785,
+      "learning_rate": 0.00016056567982342817,
+      "loss": 1.1568,
+      "step": 638
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06608510762453079,
+      "learning_rate": 0.00016044901750807098,
+      "loss": 1.1529,
+      "step": 639
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06756897270679474,
+      "learning_rate": 0.00016033222540327567,
+      "loss": 1.0845,
+      "step": 640
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06797973066568375,
+      "learning_rate": 0.00016021530375980535,
+      "loss": 1.0768,
+      "step": 641
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06994897872209549,
+      "learning_rate": 0.00016009825282870126,
+      "loss": 1.1013,
+      "step": 642
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06616318970918655,
+      "learning_rate": 0.00015998107286128215,
+      "loss": 1.0895,
+      "step": 643
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.07222306728363037,
+      "learning_rate": 0.00015986376410914388,
+      "loss": 1.1811,
+      "step": 644
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06701509654521942,
+      "learning_rate": 0.00015974632682415885,
+      "loss": 1.1747,
+      "step": 645
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.0668727234005928,
+      "learning_rate": 0.00015962876125847535,
+      "loss": 1.1241,
+      "step": 646
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06632225960493088,
+      "learning_rate": 0.00015951106766451715,
+      "loss": 1.1301,
+      "step": 647
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06712299585342407,
+      "learning_rate": 0.00015939324629498294,
+      "loss": 1.1646,
+      "step": 648
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06921354681253433,
+      "learning_rate": 0.0001592752974028457,
+      "loss": 1.1075,
+      "step": 649
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06251943111419678,
+      "learning_rate": 0.00015915722124135227,
+      "loss": 1.139,
+      "step": 650
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06602328270673752,
+      "learning_rate": 0.00015903901806402264,
+      "loss": 1.1455,
+      "step": 651
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06831764429807663,
+      "learning_rate": 0.00015892068812464963,
+      "loss": 1.1503,
+      "step": 652
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.0706949532032013,
+      "learning_rate": 0.00015880223167729818,
+      "loss": 1.1477,
+      "step": 653
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.07188212126493454,
+      "learning_rate": 0.0001586836489763049,
+      "loss": 1.0946,
+      "step": 654
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.0642421543598175,
+      "learning_rate": 0.00015856494027627734,
+      "loss": 1.1365,
+      "step": 655
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.07110840082168579,
+      "learning_rate": 0.00015844610583209373,
+      "loss": 1.1358,
+      "step": 656
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06591419875621796,
+      "learning_rate": 0.0001583271458989023,
+      "loss": 1.1176,
+      "step": 657
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.0683559700846672,
+      "learning_rate": 0.00015820806073212055,
+      "loss": 1.0383,
+      "step": 658
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.06749717891216278,
+      "learning_rate": 0.00015808885058743498,
+      "loss": 1.1389,
+      "step": 659
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.07007017731666565,
+      "learning_rate": 0.00015796951572080047,
+      "loss": 1.1519,
+      "step": 660
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.06702763587236404,
+      "learning_rate": 0.00015785005638843956,
+      "loss": 1.1068,
+      "step": 661
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.07349605113267899,
+      "learning_rate": 0.0001577304728468422,
+      "loss": 1.2148,
+      "step": 662
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.06870365887880325,
+      "learning_rate": 0.00015761076535276486,
+      "loss": 1.1951,
+      "step": 663
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.07068511843681335,
+      "learning_rate": 0.00015749093416323024,
+      "loss": 1.1536,
+      "step": 664
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.0644054263830185,
+      "learning_rate": 0.00015737097953552658,
+      "loss": 1.1183,
+      "step": 665
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.06498972326517105,
+      "learning_rate": 0.0001572509017272072,
+      "loss": 1.0997,
+      "step": 666
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.06673012673854828,
+      "learning_rate": 0.00015713070099608988,
+      "loss": 1.1471,
+      "step": 667
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.06598661094903946,
+      "learning_rate": 0.0001570103776002563,
+      "loss": 1.0986,
+      "step": 668
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.06478138267993927,
+      "learning_rate": 0.00015688993179805154,
+      "loss": 1.0817,
+      "step": 669
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.06719911098480225,
+      "learning_rate": 0.00015676936384808354,
+      "loss": 1.1287,
+      "step": 670
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.06300705671310425,
+      "learning_rate": 0.00015664867400922239,
+      "loss": 1.0937,
+      "step": 671
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.06910447776317596,
+      "learning_rate": 0.00015652786254059998,
+      "loss": 1.0682,
+      "step": 672
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.06596586108207703,
+      "learning_rate": 0.0001564069297016093,
+      "loss": 1.0551,
+      "step": 673
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.06589323282241821,
+      "learning_rate": 0.00015628587575190395,
+      "loss": 1.1609,
+      "step": 674
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.07057927548885345,
+      "learning_rate": 0.00015616470095139762,
+      "loss": 1.1117,
+      "step": 675
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.07004386186599731,
+      "learning_rate": 0.0001560434055602634,
+      "loss": 1.1485,
+      "step": 676
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.06499215215444565,
+      "learning_rate": 0.00015592198983893329,
+      "loss": 1.0118,
+      "step": 677
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.06642549484968185,
+      "learning_rate": 0.00015580045404809772,
+      "loss": 1.1556,
+      "step": 678
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.06729692965745926,
+      "learning_rate": 0.0001556787984487049,
+      "loss": 1.1552,
+      "step": 679
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.069004587829113,
+      "learning_rate": 0.00015555702330196023,
+      "loss": 1.1373,
+      "step": 680
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.07328642904758453,
+      "learning_rate": 0.00015543512886932582,
+      "loss": 1.1266,
+      "step": 681
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.06573385000228882,
+      "learning_rate": 0.00015531311541251995,
+      "loss": 1.1094,
+      "step": 682
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.0658872127532959,
+      "learning_rate": 0.00015519098319351636,
+      "loss": 1.087,
+      "step": 683
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.0687699168920517,
+      "learning_rate": 0.00015506873247454384,
+      "loss": 1.0914,
+      "step": 684
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.06617342680692673,
+      "learning_rate": 0.00015494636351808563,
+      "loss": 1.1377,
+      "step": 685
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.06719239801168442,
+      "learning_rate": 0.00015482387658687875,
+      "loss": 1.2187,
+      "step": 686
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.0662359967827797,
+      "learning_rate": 0.0001547012719439136,
+      "loss": 1.0325,
+      "step": 687
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.0648040920495987,
+      "learning_rate": 0.0001545785498524333,
+      "loss": 1.0903,
+      "step": 688
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.06930810958147049,
+      "learning_rate": 0.00015445571057593315,
+      "loss": 1.128,
+      "step": 689
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.06874281167984009,
+      "learning_rate": 0.00015433275437816004,
+      "loss": 1.2245,
+      "step": 690
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.06701396405696869,
+      "learning_rate": 0.0001542096815231119,
+      "loss": 1.1042,
+      "step": 691
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.06753288954496384,
+      "learning_rate": 0.00015408649227503714,
+      "loss": 1.1593,
+      "step": 692
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.06974251568317413,
+      "learning_rate": 0.00015396318689843408,
+      "loss": 1.1662,
+      "step": 693
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.07186184823513031,
+      "learning_rate": 0.00015383976565805035,
+      "loss": 1.2292,
+      "step": 694
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.06782180815935135,
+      "learning_rate": 0.0001537162288188824,
+      "loss": 1.1006,
+      "step": 695
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.06984351575374603,
+      "learning_rate": 0.00015359257664617485,
+      "loss": 1.0839,
+      "step": 696
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.0696905329823494,
+      "learning_rate": 0.00015346880940541997,
+      "loss": 1.0847,
+      "step": 697
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.06678467988967896,
+      "learning_rate": 0.00015334492736235705,
+      "loss": 1.145,
+      "step": 698
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.06899551302194595,
+      "learning_rate": 0.0001532209307829719,
+      "loss": 1.0262,
+      "step": 699
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.06538818776607513,
+      "learning_rate": 0.00015309681993349626,
+      "loss": 1.1179,
+      "step": 700
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.0666823759675026,
+      "learning_rate": 0.00015297259508040718,
+      "loss": 1.1654,
+      "step": 701
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.0722886174917221,
+      "learning_rate": 0.00015284825649042655,
+      "loss": 1.1446,
+      "step": 702
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.0693565309047699,
+      "learning_rate": 0.0001527238044305204,
+      "loss": 1.2072,
+      "step": 703
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.07026789337396622,
+      "learning_rate": 0.00015259923916789844,
+      "loss": 1.1308,
+      "step": 704
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.07177083194255829,
+      "learning_rate": 0.00015247456097001338,
+      "loss": 1.1002,
+      "step": 705
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.0645269900560379,
+      "learning_rate": 0.00015234977010456047,
+      "loss": 1.1355,
+      "step": 706
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.0640435442328453,
+      "learning_rate": 0.00015222486683947678,
+      "loss": 1.1648,
+      "step": 707
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.06458823382854462,
+      "learning_rate": 0.0001520998514429409,
+      "loss": 1.0977,
+      "step": 708
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.06381873786449432,
+      "learning_rate": 0.0001519747241833719,
+      "loss": 1.0494,
+      "step": 709
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.07047054916620255,
+      "learning_rate": 0.00015184948532942928,
+      "loss": 1.0674,
+      "step": 710
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.07138516753911972,
+      "learning_rate": 0.000151724135150012,
+      "loss": 1.2869,
+      "step": 711
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.06848011165857315,
+      "learning_rate": 0.0001515986739142581,
+      "loss": 1.1813,
+      "step": 712
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.0663144439458847,
+      "learning_rate": 0.00015147310189154406,
+      "loss": 1.2019,
+      "step": 713
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.07065600156784058,
+      "learning_rate": 0.0001513474193514842,
+      "loss": 1.1387,
+      "step": 714
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.06766260415315628,
+      "learning_rate": 0.0001512216265639302,
+      "loss": 1.1266,
+      "step": 715
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.06984108686447144,
+      "learning_rate": 0.00015109572379897035,
+      "loss": 1.2299,
+      "step": 716
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.06617981195449829,
+      "learning_rate": 0.00015096971132692917,
+      "loss": 1.1855,
+      "step": 717
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.07162600755691528,
+      "learning_rate": 0.0001508435894183667,
+      "loss": 1.0786,
+      "step": 718
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.06593747437000275,
+      "learning_rate": 0.0001507173583440779,
+      "loss": 1.2103,
+      "step": 719
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.06227962672710419,
+      "learning_rate": 0.0001505910183750922,
+      "loss": 1.0816,
+      "step": 720
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 1.0861259698867798,
+      "eval_runtime": 97.3461,
+      "eval_samples_per_second": 51.363,
+      "eval_steps_per_second": 12.841,
+      "step": 720
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.06885813921689987,
+      "learning_rate": 0.00015046456978267276,
+      "loss": 1.0982,
+      "step": 721
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.06474511325359344,
+      "learning_rate": 0.00015033801283831596,
+      "loss": 1.1012,
+      "step": 722
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.06354521960020065,
+      "learning_rate": 0.00015021134781375097,
+      "loss": 1.1416,
+      "step": 723
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.06504935026168823,
+      "learning_rate": 0.00015008457498093882,
+      "loss": 1.0888,
+      "step": 724
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.06544163823127747,
+      "learning_rate": 0.00014995769461207214,
+      "loss": 1.0266,
+      "step": 725
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.06988748162984848,
+      "learning_rate": 0.00014983070697957438,
+      "loss": 1.071,
+      "step": 726
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.0684196725487709,
+      "learning_rate": 0.00014970361235609937,
+      "loss": 1.1538,
+      "step": 727
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.06629320979118347,
+      "learning_rate": 0.00014957641101453055,
+      "loss": 1.0883,
+      "step": 728
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.06252582371234894,
+      "learning_rate": 0.0001494491032279806,
+      "loss": 1.0837,
+      "step": 729
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.06974456459283829,
+      "learning_rate": 0.00014932168926979074,
+      "loss": 1.0883,
+      "step": 730
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.06840938329696655,
+      "learning_rate": 0.0001491941694135301,
+      "loss": 1.0915,
+      "step": 731
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.0694158524274826,
+      "learning_rate": 0.0001490665439329951,
+      "loss": 1.1775,
+      "step": 732
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.07047273218631744,
+      "learning_rate": 0.00014893881310220914,
+      "loss": 1.1565,
+      "step": 733
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.06564071029424667,
+      "learning_rate": 0.00014881097719542173,
+      "loss": 1.0357,
+      "step": 734
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.0719495341181755,
+      "learning_rate": 0.00014868303648710792,
+      "loss": 1.0986,
+      "step": 735
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.0710771381855011,
+      "learning_rate": 0.00014855499125196784,
+      "loss": 1.1424,
+      "step": 736
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.06624703109264374,
+      "learning_rate": 0.00014842684176492608,
+      "loss": 1.0868,
+      "step": 737
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.07482437789440155,
+      "learning_rate": 0.000148298588301131,
+      "loss": 1.0954,
+      "step": 738
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.06619468331336975,
+      "learning_rate": 0.00014817023113595427,
+      "loss": 1.0935,
+      "step": 739
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.06574519723653793,
+      "learning_rate": 0.00014804177054499016,
+      "loss": 1.17,
+      "step": 740
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.06976670771837234,
+      "learning_rate": 0.00014791320680405503,
+      "loss": 1.0459,
+      "step": 741
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.07105021923780441,
+      "learning_rate": 0.0001477845401891867,
+      "loss": 1.0618,
+      "step": 742
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.067878857254982,
+      "learning_rate": 0.00014765577097664392,
+      "loss": 1.074,
+      "step": 743
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.07623715698719025,
+      "learning_rate": 0.00014752689944290564,
+      "loss": 1.1404,
+      "step": 744
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.06776713579893112,
+      "learning_rate": 0.00014739792586467061,
+      "loss": 1.0972,
+      "step": 745
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.06891727447509766,
+      "learning_rate": 0.00014726885051885653,
+      "loss": 1.0612,
+      "step": 746
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.07326652109622955,
+      "learning_rate": 0.0001471396736825998,
+      "loss": 1.057,
+      "step": 747
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.07193190604448318,
+      "learning_rate": 0.00014701039563325453,
+      "loss": 1.044,
+      "step": 748
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.0713285356760025,
+      "learning_rate": 0.00014688101664839224,
+      "loss": 1.0477,
+      "step": 749
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.07577778398990631,
+      "learning_rate": 0.00014675153700580124,
+      "loss": 1.1469,
+      "step": 750
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.0669977068901062,
+      "learning_rate": 0.00014662195698348585,
+      "loss": 1.0286,
+      "step": 751
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.07274546474218369,
+      "learning_rate": 0.00014649227685966588,
+      "loss": 1.1343,
+      "step": 752
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.070924311876297,
+      "learning_rate": 0.00014636249691277623,
+      "loss": 1.0737,
+      "step": 753
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.07124372571706772,
+      "learning_rate": 0.00014623261742146602,
+      "loss": 1.059,
+      "step": 754
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.07382449507713318,
+      "learning_rate": 0.00014610263866459805,
+      "loss": 1.1133,
+      "step": 755
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.08469633758068085,
+      "learning_rate": 0.00014597256092124832,
+      "loss": 1.1482,
+      "step": 756
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.07403711974620819,
+      "learning_rate": 0.00014584238447070547,
+      "loss": 1.0972,
+      "step": 757
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.07138115912675858,
+      "learning_rate": 0.00014571210959246988,
+      "loss": 1.1579,
+      "step": 758
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.07389011234045029,
+      "learning_rate": 0.00014558173656625332,
+      "loss": 1.0797,
+      "step": 759
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.07290019094944,
+      "learning_rate": 0.00014545126567197834,
+      "loss": 1.0503,
+      "step": 760
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.07357808947563171,
+      "learning_rate": 0.00014532069718977764,
+      "loss": 1.1109,
+      "step": 761
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.0705307200551033,
+      "learning_rate": 0.00014519003139999337,
+      "loss": 1.1023,
+      "step": 762
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.0704856589436531,
+      "learning_rate": 0.00014505926858317668,
+      "loss": 1.1139,
+      "step": 763
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.07422898709774017,
+      "learning_rate": 0.000144928409020087,
+      "loss": 1.1164,
+      "step": 764
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.07562410831451416,
+      "learning_rate": 0.00014479745299169147,
+      "loss": 1.0434,
+      "step": 765
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.07194274663925171,
+      "learning_rate": 0.0001446664007791644,
+      "loss": 0.9754,
+      "step": 766
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.06834331154823303,
+      "learning_rate": 0.0001445352526638866,
+      "loss": 1.1614,
+      "step": 767
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.06937012821435928,
+      "learning_rate": 0.00014440400892744477,
+      "loss": 1.063,
+      "step": 768
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.07618941366672516,
+      "learning_rate": 0.0001442726698516309,
+      "loss": 1.0918,
+      "step": 769
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.07851953059434891,
+      "learning_rate": 0.00014414123571844178,
+      "loss": 0.9882,
+      "step": 770
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.07552509754896164,
+      "learning_rate": 0.0001440097068100782,
+      "loss": 1.1529,
+      "step": 771
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.06892319023609161,
+      "learning_rate": 0.00014387808340894444,
+      "loss": 1.1288,
+      "step": 772
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.07122083753347397,
+      "learning_rate": 0.00014374636579764772,
+      "loss": 1.1284,
+      "step": 773
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.07426360249519348,
+      "learning_rate": 0.00014361455425899756,
+      "loss": 1.1505,
+      "step": 774
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.07406499236822128,
+      "learning_rate": 0.000143482649076005,
+      "loss": 1.0639,
+      "step": 775
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.07255791127681732,
+      "learning_rate": 0.0001433506505318823,
+      "loss": 1.0874,
+      "step": 776
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.08859973400831223,
+      "learning_rate": 0.00014321855891004216,
+      "loss": 1.0444,
+      "step": 777
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.06971631199121475,
+      "learning_rate": 0.00014308637449409706,
+      "loss": 1.0736,
+      "step": 778
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.07728686183691025,
+      "learning_rate": 0.0001429540975678587,
+      "loss": 1.1992,
+      "step": 779
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.07209434360265732,
+      "learning_rate": 0.0001428217284153375,
+      "loss": 1.0556,
+      "step": 780
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.08733708411455154,
+      "learning_rate": 0.0001426892673207418,
+      "loss": 1.097,
+      "step": 781
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.07391902059316635,
+      "learning_rate": 0.00014255671456847748,
+      "loss": 1.1503,
+      "step": 782
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.08149562031030655,
+      "learning_rate": 0.00014242407044314707,
+      "loss": 1.1151,
+      "step": 783
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.07579551637172699,
+      "learning_rate": 0.00014229133522954937,
+      "loss": 1.1052,
+      "step": 784
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.07156646996736526,
+      "learning_rate": 0.00014215850921267873,
+      "loss": 1.0769,
+      "step": 785
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.07255309820175171,
+      "learning_rate": 0.00014202559267772444,
+      "loss": 1.0048,
+      "step": 786
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.07417687028646469,
+      "learning_rate": 0.00014189258591007023,
+      "loss": 1.0496,
+      "step": 787
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.07492946088314056,
+      "learning_rate": 0.0001417594891952935,
+      "loss": 1.1562,
+      "step": 788
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.07493904232978821,
+      "learning_rate": 0.0001416263028191647,
+      "loss": 1.1239,
+      "step": 789
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.0770529955625534,
+      "learning_rate": 0.00014149302706764697,
+      "loss": 1.0968,
+      "step": 790
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.10481111705303192,
+      "learning_rate": 0.00014135966222689523,
+      "loss": 1.17,
+      "step": 791
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.08004539459943771,
+      "learning_rate": 0.00014122620858325564,
+      "loss": 1.1182,
+      "step": 792
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.08175237476825714,
+      "learning_rate": 0.00014109266642326515,
+      "loss": 1.0617,
+      "step": 793
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.07505473494529724,
+      "learning_rate": 0.00014095903603365066,
+      "loss": 1.1622,
+      "step": 794
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.07435812056064606,
+      "learning_rate": 0.00014082531770132856,
+      "loss": 1.0833,
+      "step": 795
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.07598494738340378,
+      "learning_rate": 0.0001406915117134041,
+      "loss": 1.0753,
+      "step": 796
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.08518858253955841,
+      "learning_rate": 0.0001405576183571706,
+      "loss": 1.0947,
+      "step": 797
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.0808837041258812,
+      "learning_rate": 0.0001404236379201091,
+      "loss": 1.1801,
+      "step": 798
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.08147124946117401,
+      "learning_rate": 0.00014028957068988756,
+      "loss": 1.0991,
+      "step": 799
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.07758703827857971,
+      "learning_rate": 0.00014015541695436027,
+      "loss": 1.1374,
+      "step": 800
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.07507652044296265,
+      "learning_rate": 0.0001400211770015673,
+      "loss": 1.0599,
+      "step": 801
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.07711515575647354,
+      "learning_rate": 0.00013988685111973384,
+      "loss": 1.1197,
+      "step": 802
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.07404826581478119,
+      "learning_rate": 0.0001397524395972695,
+      "loss": 1.0463,
+      "step": 803
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.07590384036302567,
+      "learning_rate": 0.00013961794272276788,
+      "loss": 1.0114,
+      "step": 804
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.08123478293418884,
+      "learning_rate": 0.0001394833607850057,
+      "loss": 1.1701,
+      "step": 805
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.08145097643136978,
+      "learning_rate": 0.00013934869407294245,
+      "loss": 1.1554,
+      "step": 806
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.07815194875001907,
+      "learning_rate": 0.0001392139428757196,
+      "loss": 1.0891,
+      "step": 807
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.07877809554338455,
+      "learning_rate": 0.0001390791074826599,
+      "loss": 1.1764,
+      "step": 808
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.07488015294075012,
+      "learning_rate": 0.00013894418818326712,
+      "loss": 1.0886,
+      "step": 809
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.0857730582356453,
+      "learning_rate": 0.00013880918526722497,
+      "loss": 1.0871,
+      "step": 810
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.07474977523088455,
+      "learning_rate": 0.0001386740990243968,
+      "loss": 1.0571,
+      "step": 811
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.09196992218494415,
+      "learning_rate": 0.00013853892974482476,
+      "loss": 0.9109,
+      "step": 812
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.0746723860502243,
+      "learning_rate": 0.00013840367771872947,
+      "loss": 1.159,
+      "step": 813
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.07313646376132965,
+      "learning_rate": 0.000138268343236509,
+      "loss": 1.007,
+      "step": 814
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.11828210949897766,
+      "learning_rate": 0.0001381329265887386,
+      "loss": 1.0908,
+      "step": 815
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.07454833388328552,
+      "learning_rate": 0.00013799742806616994,
+      "loss": 1.1658,
+      "step": 816
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.07726167887449265,
+      "learning_rate": 0.00013786184795973045,
+      "loss": 1.1124,
+      "step": 817
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.07478827983140945,
+      "learning_rate": 0.0001377261865605227,
+      "loss": 1.0835,
+      "step": 818
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.07675985991954803,
+      "learning_rate": 0.00013759044415982382,
+      "loss": 1.0954,
+      "step": 819
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.08556641638278961,
+      "learning_rate": 0.00013745462104908487,
+      "loss": 1.0492,
+      "step": 820
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.07620813697576523,
+      "learning_rate": 0.00013731871751993018,
+      "loss": 1.1234,
+      "step": 821
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.07860769331455231,
+      "learning_rate": 0.0001371827338641568,
+      "loss": 1.0359,
+      "step": 822
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.08416145294904709,
+      "learning_rate": 0.00013704667037373375,
+      "loss": 0.999,
+      "step": 823
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.08307678997516632,
+      "learning_rate": 0.00013691052734080152,
+      "loss": 1.1396,
+      "step": 824
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.07701461762189865,
+      "learning_rate": 0.00013677430505767136,
+      "loss": 1.0699,
+      "step": 825
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.08459299057722092,
+      "learning_rate": 0.00013663800381682464,
+      "loss": 1.0399,
+      "step": 826
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.07567258924245834,
+      "learning_rate": 0.0001365016239109123,
+      "loss": 1.0648,
+      "step": 827
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.07730862498283386,
+      "learning_rate": 0.00013636516563275418,
+      "loss": 1.0743,
+      "step": 828
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.07540915906429291,
+      "learning_rate": 0.00013622862927533845,
+      "loss": 1.0239,
+      "step": 829
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.0840970054268837,
+      "learning_rate": 0.00013609201513182075,
+      "loss": 1.0338,
+      "step": 830
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.08027436584234238,
+      "learning_rate": 0.00013595532349552393,
+      "loss": 1.0949,
+      "step": 831
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.07819013297557831,
+      "learning_rate": 0.0001358185546599371,
+      "loss": 1.1237,
+      "step": 832
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.08244457095861435,
+      "learning_rate": 0.0001356817089187152,
+      "loss": 1.2162,
+      "step": 833
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.07645738124847412,
+      "learning_rate": 0.00013554478656567818,
+      "loss": 1.2098,
+      "step": 834
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.07934603840112686,
+      "learning_rate": 0.0001354077878948106,
+      "loss": 1.0631,
+      "step": 835
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.08178233355283737,
+      "learning_rate": 0.0001352707132002609,
+      "loss": 1.1594,
+      "step": 836
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.07968655228614807,
+      "learning_rate": 0.00013513356277634057,
+      "loss": 0.9803,
+      "step": 837
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.07808240503072739,
+      "learning_rate": 0.0001349963369175239,
+      "loss": 1.1199,
+      "step": 838
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.07478972524404526,
+      "learning_rate": 0.0001348590359184471,
+      "loss": 1.0029,
+      "step": 839
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.07492795586585999,
+      "learning_rate": 0.00013472166007390753,
+      "loss": 1.0261,
+      "step": 840
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.08683785796165466,
+      "learning_rate": 0.0001345842096788635,
+      "loss": 1.1995,
+      "step": 841
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.07922057062387466,
+      "learning_rate": 0.0001344466850284333,
+      "loss": 1.0421,
+      "step": 842
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.0820443406701088,
+      "learning_rate": 0.00013430908641789456,
+      "loss": 1.0757,
+      "step": 843
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.07989905029535294,
+      "learning_rate": 0.00013417141414268384,
+      "loss": 1.097,
+      "step": 844
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.08366917073726654,
+      "learning_rate": 0.00013403366849839578,
+      "loss": 1.1318,
+      "step": 845
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.08655679225921631,
+      "learning_rate": 0.00013389584978078258,
+      "loss": 1.0963,
+      "step": 846
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.08200956881046295,
+      "learning_rate": 0.00013375795828575332,
+      "loss": 1.0947,
+      "step": 847
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.07856731861829758,
+      "learning_rate": 0.00013361999430937338,
+      "loss": 1.044,
+      "step": 848
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.07763627171516418,
+      "learning_rate": 0.00013348195814786372,
+      "loss": 1.0379,
+      "step": 849
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.08052556961774826,
+      "learning_rate": 0.00013334385009760032,
+      "loss": 1.1328,
+      "step": 850
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.07773949205875397,
+      "learning_rate": 0.00013320567045511347,
+      "loss": 1.0755,
+      "step": 851
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.07407914102077484,
+      "learning_rate": 0.00013306741951708723,
+      "loss": 1.0117,
+      "step": 852
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.07695198059082031,
+      "learning_rate": 0.00013292909758035867,
+      "loss": 1.0461,
+      "step": 853
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.07831941545009613,
+      "learning_rate": 0.00013279070494191737,
+      "loss": 1.039,
+      "step": 854
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.08240978419780731,
+      "learning_rate": 0.0001326522418989047,
+      "loss": 1.1131,
+      "step": 855
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.07506600022315979,
+      "learning_rate": 0.0001325137087486131,
+      "loss": 1.0739,
+      "step": 856
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.08516227453947067,
+      "learning_rate": 0.0001323751057884857,
+      "loss": 1.1227,
+      "step": 857
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.08774415403604507,
+      "learning_rate": 0.00013223643331611537,
+      "loss": 1.1147,
+      "step": 858
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.0774305909872055,
+      "learning_rate": 0.00013209769162924437,
+      "loss": 1.0748,
+      "step": 859
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.08648118376731873,
+      "learning_rate": 0.00013195888102576336,
+      "loss": 1.1234,
+      "step": 860
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.08546962589025497,
+      "learning_rate": 0.0001318200018037112,
+      "loss": 1.1716,
+      "step": 861
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.07781863957643509,
+      "learning_rate": 0.000131681054261274,
+      "loss": 1.035,
+      "step": 862
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.07804084569215775,
+      "learning_rate": 0.00013154203869678442,
+      "loss": 1.0362,
+      "step": 863
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.07699594646692276,
+      "learning_rate": 0.0001314029554087214,
+      "loss": 1.0995,
+      "step": 864
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.08640976995229721,
+      "learning_rate": 0.0001312638046957092,
+      "loss": 1.1351,
+      "step": 865
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.08521472662687302,
+      "learning_rate": 0.00013112458685651668,
+      "loss": 1.0974,
+      "step": 866
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.08203436434268951,
+      "learning_rate": 0.00013098530219005712,
+      "loss": 1.046,
+      "step": 867
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.0792430117726326,
+      "learning_rate": 0.00013084595099538705,
+      "loss": 1.0843,
+      "step": 868
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.07786545902490616,
+      "learning_rate": 0.00013070653357170593,
+      "loss": 1.0435,
+      "step": 869
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.07653090357780457,
+      "learning_rate": 0.00013056705021835546,
+      "loss": 1.0472,
+      "step": 870
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.07724118232727051,
+      "learning_rate": 0.00013042750123481889,
+      "loss": 1.1262,
+      "step": 871
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.08276543021202087,
+      "learning_rate": 0.00013028788692072025,
+      "loss": 1.0581,
+      "step": 872
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.07761494815349579,
+      "learning_rate": 0.00013014820757582395,
+      "loss": 1.1223,
+      "step": 873
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.08198413997888565,
+      "learning_rate": 0.0001300084635000341,
+      "loss": 1.1038,
+      "step": 874
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.07821369171142578,
+      "learning_rate": 0.0001298686549933936,
+      "loss": 1.0962,
+      "step": 875
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.07938680052757263,
+      "learning_rate": 0.0001297287823560839,
+      "loss": 1.1057,
+      "step": 876
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.07824011892080307,
+      "learning_rate": 0.00012958884588842398,
+      "loss": 1.0612,
+      "step": 877
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.07857441157102585,
+      "learning_rate": 0.00012944884589086993,
+      "loss": 1.0409,
+      "step": 878
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.07374000549316406,
+      "learning_rate": 0.0001293087826640143,
+      "loss": 1.1007,
+      "step": 879
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.07789342105388641,
+      "learning_rate": 0.00012916865650858527,
+      "loss": 1.1156,
+      "step": 880
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.07925521582365036,
+      "learning_rate": 0.00012902846772544624,
+      "loss": 1.0786,
+      "step": 881
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.07993773370981216,
+      "learning_rate": 0.00012888821661559508,
+      "loss": 1.1056,
+      "step": 882
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.08116604387760162,
+      "learning_rate": 0.00012874790348016337,
+      "loss": 1.1644,
+      "step": 883
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.08475354313850403,
+      "learning_rate": 0.00012860752862041603,
+      "loss": 1.1275,
+      "step": 884
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.07930230349302292,
+      "learning_rate": 0.00012846709233775035,
+      "loss": 1.0778,
+      "step": 885
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.08094102889299393,
+      "learning_rate": 0.00012832659493369557,
+      "loss": 1.0287,
+      "step": 886
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.08016285300254822,
+      "learning_rate": 0.0001281860367099122,
+      "loss": 1.0219,
+      "step": 887
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.08391551673412323,
+      "learning_rate": 0.0001280454179681913,
+      "loss": 1.107,
+      "step": 888
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.08297763764858246,
+      "learning_rate": 0.0001279047390104538,
+      "loss": 1.091,
+      "step": 889
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.0829230397939682,
+      "learning_rate": 0.00012776400013875006,
+      "loss": 1.172,
+      "step": 890
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.0828934758901596,
+      "learning_rate": 0.00012762320165525897,
+      "loss": 1.0247,
+      "step": 891
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.07887378334999084,
+      "learning_rate": 0.00012748234386228746,
+      "loss": 0.9858,
+      "step": 892
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.07972092926502228,
+      "learning_rate": 0.00012734142706226972,
+      "loss": 1.0359,
+      "step": 893
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.08054916560649872,
+      "learning_rate": 0.0001272004515577668,
+      "loss": 1.1457,
+      "step": 894
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.07671314477920532,
+      "learning_rate": 0.00012705941765146566,
+      "loss": 1.0979,
+      "step": 895
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.07783249765634537,
+      "learning_rate": 0.00012691832564617865,
+      "loss": 1.1148,
+      "step": 896
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.07776234298944473,
+      "learning_rate": 0.00012677717584484296,
+      "loss": 1.0295,
+      "step": 897
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.07843487709760666,
+      "learning_rate": 0.0001266359685505198,
+      "loss": 1.0823,
+      "step": 898
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.08097175508737564,
+      "learning_rate": 0.0001264947040663938,
+      "loss": 1.0815,
+      "step": 899
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.08583438396453857,
+      "learning_rate": 0.00012635338269577248,
+      "loss": 1.0741,
+      "step": 900
+    },
+    {
+      "epoch": 1.24,
+      "eval_loss": 1.0887025594711304,
+      "eval_runtime": 96.9356,
+      "eval_samples_per_second": 51.581,
+      "eval_steps_per_second": 12.895,
+      "step": 900
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.08100762218236923,
+      "learning_rate": 0.00012621200474208544,
+      "loss": 1.121,
+      "step": 901
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.08114492893218994,
+      "learning_rate": 0.0001260705705088837,
+      "loss": 1.0742,
+      "step": 902
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.08283861726522446,
+      "learning_rate": 0.00012592908029983923,
+      "loss": 1.1485,
+      "step": 903
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.08205985277891159,
+      "learning_rate": 0.00012578753441874416,
+      "loss": 1.0767,
+      "step": 904
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.08423864096403122,
+      "learning_rate": 0.00012564593316951014,
+      "loss": 1.0282,
+      "step": 905
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.07910145074129105,
+      "learning_rate": 0.00012550427685616765,
+      "loss": 1.1672,
+      "step": 906
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.08530287444591522,
+      "learning_rate": 0.00012536256578286547,
+      "loss": 1.0889,
+      "step": 907
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.08540420234203339,
+      "learning_rate": 0.00012522080025386995,
+      "loss": 1.1112,
+      "step": 908
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.0812099501490593,
+      "learning_rate": 0.00012507898057356428,
+      "loss": 1.1563,
+      "step": 909
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.08489541709423065,
+      "learning_rate": 0.00012493710704644805,
+      "loss": 1.1648,
+      "step": 910
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.07855423539876938,
+      "learning_rate": 0.00012479517997713636,
+      "loss": 1.0266,
+      "step": 911
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.0805073156952858,
+      "learning_rate": 0.00012465319967035925,
+      "loss": 1.1197,
+      "step": 912
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.07947047799825668,
+      "learning_rate": 0.0001245111664309612,
+      "loss": 1.1188,
+      "step": 913
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.08117293566465378,
+      "learning_rate": 0.0001243690805639002,
+      "loss": 1.0104,
+      "step": 914
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.0896577462553978,
+      "learning_rate": 0.0001242269423742473,
+      "loss": 1.0753,
+      "step": 915
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.08124979585409164,
+      "learning_rate": 0.0001240847521671859,
+      "loss": 1.0907,
+      "step": 916
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.08808252960443497,
+      "learning_rate": 0.0001239425102480111,
+      "loss": 1.1121,
+      "step": 917
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.08195436745882034,
+      "learning_rate": 0.00012380021692212894,
+      "loss": 1.1001,
+      "step": 918
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.08645926415920258,
+      "learning_rate": 0.0001236578724950559,
+      "loss": 1.1597,
+      "step": 919
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.08274907618761063,
+      "learning_rate": 0.00012351547727241824,
+      "loss": 1.1069,
+      "step": 920
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.08482358604669571,
+      "learning_rate": 0.00012337303155995116,
+      "loss": 1.055,
+      "step": 921
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.09005696326494217,
+      "learning_rate": 0.00012323053566349834,
+      "loss": 1.1091,
+      "step": 922
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.08296368271112442,
+      "learning_rate": 0.00012308798988901117,
+      "loss": 1.1234,
+      "step": 923
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.08951117843389511,
+      "learning_rate": 0.00012294539454254822,
+      "loss": 1.1522,
+      "step": 924
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.0824238732457161,
+      "learning_rate": 0.00012280274993027435,
+      "loss": 1.05,
+      "step": 925
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.08390199393033981,
+      "learning_rate": 0.00012266005635846037,
+      "loss": 1.0136,
+      "step": 926
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.07979277521371841,
+      "learning_rate": 0.00012251731413348205,
+      "loss": 1.068,
+      "step": 927
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.08345918357372284,
+      "learning_rate": 0.00012237452356181968,
+      "loss": 1.0642,
+      "step": 928
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.09497953951358795,
+      "learning_rate": 0.0001222316849500574,
+      "loss": 1.1254,
+      "step": 929
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.08493612706661224,
+      "learning_rate": 0.0001220887986048825,
+      "loss": 1.1122,
+      "step": 930
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.08732904493808746,
+      "learning_rate": 0.00012194586483308466,
+      "loss": 1.0784,
+      "step": 931
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.08586254715919495,
+      "learning_rate": 0.00012180288394155547,
+      "loss": 1.1624,
+      "step": 932
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.09152837097644806,
+      "learning_rate": 0.00012165985623728768,
+      "loss": 1.0859,
+      "step": 933
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.08018858730792999,
+      "learning_rate": 0.00012151678202737456,
+      "loss": 1.0137,
+      "step": 934
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.08106279373168945,
+      "learning_rate": 0.00012137366161900916,
+      "loss": 1.0519,
+      "step": 935
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.09281496703624725,
+      "learning_rate": 0.00012123049531948381,
+      "loss": 1.0764,
+      "step": 936
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.08181937783956528,
+      "learning_rate": 0.00012108728343618934,
+      "loss": 1.0972,
+      "step": 937
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.0799027532339096,
+      "learning_rate": 0.00012094402627661447,
+      "loss": 1.1055,
+      "step": 938
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.08846897631883621,
+      "learning_rate": 0.00012080072414834504,
+      "loss": 1.1138,
+      "step": 939
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.09898900985717773,
+      "learning_rate": 0.00012065737735906361,
+      "loss": 1.0617,
+      "step": 940
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.08367432653903961,
+      "learning_rate": 0.00012051398621654845,
+      "loss": 1.0993,
+      "step": 941
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.08614856749773026,
+      "learning_rate": 0.00012037055102867321,
+      "loss": 1.1265,
+      "step": 942
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.0842709168791771,
+      "learning_rate": 0.00012022707210340605,
+      "loss": 1.1385,
+      "step": 943
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.09475839883089066,
+      "learning_rate": 0.00012008354974880896,
+      "loss": 1.1809,
+      "step": 944
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.09907343983650208,
+      "learning_rate": 0.00011993998427303732,
+      "loss": 1.0824,
+      "step": 945
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.08275450766086578,
+      "learning_rate": 0.00011979637598433899,
+      "loss": 1.042,
+      "step": 946
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.08368710428476334,
+      "learning_rate": 0.00011965272519105381,
+      "loss": 1.0565,
+      "step": 947
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.08711770921945572,
+      "learning_rate": 0.00011950903220161285,
+      "loss": 1.1091,
+      "step": 948
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.07989653199911118,
+      "learning_rate": 0.00011936529732453776,
+      "loss": 1.0853,
+      "step": 949
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.08757849782705307,
+      "learning_rate": 0.00011922152086844023,
+      "loss": 1.0667,
+      "step": 950
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.08539264649152756,
+      "learning_rate": 0.00011907770314202107,
+      "loss": 1.1241,
+      "step": 951
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.08794459700584412,
+      "learning_rate": 0.00011893384445406983,
+      "loss": 1.1133,
+      "step": 952
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.08636278659105301,
+      "learning_rate": 0.00011878994511346393,
+      "loss": 1.0675,
+      "step": 953
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.08531929552555084,
+      "learning_rate": 0.00011864600542916813,
+      "loss": 1.1387,
+      "step": 954
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.08436352014541626,
+      "learning_rate": 0.00011850202571023374,
+      "loss": 1.052,
+      "step": 955
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.08548779040575027,
+      "learning_rate": 0.00011835800626579814,
+      "loss": 1.1393,
+      "step": 956
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.08386053889989853,
+      "learning_rate": 0.0001182139474050839,
+      "loss": 1.0996,
+      "step": 957
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.08456090092658997,
+      "learning_rate": 0.00011806984943739821,
+      "loss": 1.0834,
+      "step": 958
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.09080865234136581,
+      "learning_rate": 0.00011792571267213235,
+      "loss": 1.1442,
+      "step": 959
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.08809977024793625,
+      "learning_rate": 0.00011778153741876081,
+      "loss": 1.1091,
+      "step": 960
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.08135165274143219,
+      "learning_rate": 0.00011763732398684069,
+      "loss": 1.0534,
+      "step": 961
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.08232025802135468,
+      "learning_rate": 0.00011749307268601111,
+      "loss": 1.0976,
+      "step": 962
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.08592227846384048,
+      "learning_rate": 0.00011734878382599257,
+      "loss": 1.0806,
+      "step": 963
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.08487439900636673,
+      "learning_rate": 0.00011720445771658604,
+      "loss": 1.1325,
+      "step": 964
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.08509836345911026,
+      "learning_rate": 0.00011706009466767262,
+      "loss": 1.0668,
+      "step": 965
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.08378329873085022,
+      "learning_rate": 0.00011691569498921264,
+      "loss": 1.0719,
+      "step": 966
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.0843590572476387,
+      "learning_rate": 0.00011677125899124513,
+      "loss": 1.0378,
+      "step": 967
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.08136268705129623,
+      "learning_rate": 0.00011662678698388702,
+      "loss": 1.0342,
+      "step": 968
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.08490864932537079,
+      "learning_rate": 0.00011648227927733268,
+      "loss": 1.1118,
+      "step": 969
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.10053803771734238,
+      "learning_rate": 0.00011633773618185302,
+      "loss": 1.1212,
+      "step": 970
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.07749824970960617,
+      "learning_rate": 0.00011619315800779498,
+      "loss": 1.0743,
+      "step": 971
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.09433187544345856,
+      "learning_rate": 0.00011604854506558083,
+      "loss": 1.1005,
+      "step": 972
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.08602363616228104,
+      "learning_rate": 0.00011590389766570742,
+      "loss": 1.0752,
+      "step": 973
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.08477716147899628,
+      "learning_rate": 0.00011575921611874565,
+      "loss": 1.1001,
+      "step": 974
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.08645256608724594,
+      "learning_rate": 0.00011561450073533974,
+      "loss": 1.0843,
+      "step": 975
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.09470678120851517,
+      "learning_rate": 0.00011546975182620656,
+      "loss": 1.1211,
+      "step": 976
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.08098270744085312,
+      "learning_rate": 0.00011532496970213487,
+      "loss": 1.1356,
+      "step": 977
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.07967498898506165,
+      "learning_rate": 0.00011518015467398489,
+      "loss": 1.0511,
+      "step": 978
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.09172879159450531,
+      "learning_rate": 0.00011503530705268737,
+      "loss": 1.0819,
+      "step": 979
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.08532454818487167,
+      "learning_rate": 0.00011489042714924312,
+      "loss": 1.1357,
+      "step": 980
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.08550993353128433,
+      "learning_rate": 0.0001147455152747222,
+      "loss": 1.0743,
+      "step": 981
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.08236683160066605,
+      "learning_rate": 0.00011460057174026335,
+      "loss": 1.1301,
+      "step": 982
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.08636253327131271,
+      "learning_rate": 0.0001144555968570733,
+      "loss": 1.0674,
+      "step": 983
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.08238498866558075,
+      "learning_rate": 0.00011431059093642605,
+      "loss": 1.105,
+      "step": 984
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.08513858914375305,
+      "learning_rate": 0.00011416555428966227,
+      "loss": 1.0903,
+      "step": 985
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.08221293985843658,
+      "learning_rate": 0.0001140204872281886,
+      "loss": 0.9784,
+      "step": 986
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.08562242239713669,
+      "learning_rate": 0.00011387539006347694,
+      "loss": 1.0485,
+      "step": 987
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.0831090584397316,
+      "learning_rate": 0.00011373026310706386,
+      "loss": 0.9691,
+      "step": 988
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.0824708342552185,
+      "learning_rate": 0.00011358510667054996,
+      "loss": 1.0012,
+      "step": 989
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.08758291602134705,
+      "learning_rate": 0.00011343992106559898,
+      "loss": 1.0687,
+      "step": 990
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.08582205325365067,
+      "learning_rate": 0.00011329470660393742,
+      "loss": 1.1252,
+      "step": 991
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.08085180073976517,
+      "learning_rate": 0.00011314946359735373,
+      "loss": 1.078,
+      "step": 992
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.08840853720903397,
+      "learning_rate": 0.00011300419235769756,
+      "loss": 1.0812,
+      "step": 993
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.08567378669977188,
+      "learning_rate": 0.00011285889319687923,
+      "loss": 1.0892,
+      "step": 994
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.09124870598316193,
+      "learning_rate": 0.00011271356642686905,
+      "loss": 1.1502,
+      "step": 995
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.08820147812366486,
+      "learning_rate": 0.00011256821235969657,
+      "loss": 1.0375,
+      "step": 996
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.09900910407304764,
+      "learning_rate": 0.00011242283130744992,
+      "loss": 1.1585,
+      "step": 997
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.0887092798948288,
+      "learning_rate": 0.00011227742358227522,
+      "loss": 1.1317,
+      "step": 998
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.08710745722055435,
+      "learning_rate": 0.00011213198949637587,
+      "loss": 1.1126,
+      "step": 999
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.08236543834209442,
+      "learning_rate": 0.00011198652936201175,
+      "loss": 1.0186,
+      "step": 1000
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.08354160934686661,
+      "learning_rate": 0.00011184104349149884,
+      "loss": 1.1367,
+      "step": 1001
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.08492964506149292,
+      "learning_rate": 0.00011169553219720828,
+      "loss": 1.1186,
+      "step": 1002
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.11172673851251602,
+      "learning_rate": 0.00011154999579156575,
+      "loss": 1.0217,
+      "step": 1003
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.08548691868782043,
+      "learning_rate": 0.00011140443458705096,
+      "loss": 1.1381,
+      "step": 1004
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.09090396016836166,
+      "learning_rate": 0.00011125884889619684,
+      "loss": 1.1466,
+      "step": 1005
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.08847847580909729,
+      "learning_rate": 0.00011111323903158885,
+      "loss": 1.1136,
+      "step": 1006
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.08943888545036316,
+      "learning_rate": 0.00011096760530586432,
+      "loss": 1.177,
+      "step": 1007
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.0783921629190445,
+      "learning_rate": 0.0001108219480317119,
+      "loss": 1.0227,
+      "step": 1008
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.08215797692537308,
+      "learning_rate": 0.0001106762675218708,
+      "loss": 1.1419,
+      "step": 1009
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.08204173296689987,
+      "learning_rate": 0.00011053056408913001,
+      "loss": 1.085,
+      "step": 1010
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.0815238356590271,
+      "learning_rate": 0.00011038483804632784,
+      "loss": 1.0832,
+      "step": 1011
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.079544298350811,
+      "learning_rate": 0.00011023908970635115,
+      "loss": 1.0038,
+      "step": 1012
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.08273560553789139,
+      "learning_rate": 0.00011009331938213461,
+      "loss": 1.1052,
+      "step": 1013
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.08768191933631897,
+      "learning_rate": 0.0001099475273866601,
+      "loss": 1.1363,
+      "step": 1014
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.08958650380373001,
+      "learning_rate": 0.0001098017140329561,
+      "loss": 1.02,
+      "step": 1015
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.08500074595212936,
+      "learning_rate": 0.00010965587963409682,
+      "loss": 1.0458,
+      "step": 1016
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.09041952341794968,
+      "learning_rate": 0.00010951002450320182,
+      "loss": 1.2018,
+      "step": 1017
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.08596602082252502,
+      "learning_rate": 0.0001093641489534351,
+      "loss": 1.1136,
+      "step": 1018
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.08457526564598083,
+      "learning_rate": 0.00010921825329800441,
+      "loss": 1.0642,
+      "step": 1019
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.09035574644804001,
+      "learning_rate": 0.0001090723378501608,
+      "loss": 1.1282,
+      "step": 1020
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.0817820355296135,
+      "learning_rate": 0.00010892640292319776,
+      "loss": 1.1539,
+      "step": 1021
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.08362513035535812,
+      "learning_rate": 0.0001087804488304506,
+      "loss": 1.1689,
+      "step": 1022
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.08480046689510345,
+      "learning_rate": 0.00010863447588529579,
+      "loss": 1.1198,
+      "step": 1023
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.08355570584535599,
+      "learning_rate": 0.00010848848440115028,
+      "loss": 1.1016,
+      "step": 1024
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.08374227583408356,
+      "learning_rate": 0.00010834247469147082,
+      "loss": 1.0682,
+      "step": 1025
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.09010465443134308,
+      "learning_rate": 0.00010819644706975332,
+      "loss": 1.0946,
+      "step": 1026
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.11306039243936539,
+      "learning_rate": 0.00010805040184953208,
+      "loss": 1.1514,
+      "step": 1027
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.08693739026784897,
+      "learning_rate": 0.00010790433934437922,
+      "loss": 1.0189,
+      "step": 1028
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.08382602035999298,
+      "learning_rate": 0.00010775825986790407,
+      "loss": 1.0565,
+      "step": 1029
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.08611752837896347,
+      "learning_rate": 0.00010761216373375221,
+      "loss": 1.0991,
+      "step": 1030
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.08273041248321533,
+      "learning_rate": 0.00010746605125560511,
+      "loss": 1.099,
+      "step": 1031
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.0907968133687973,
+      "learning_rate": 0.00010731992274717937,
+      "loss": 1.1583,
+      "step": 1032
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.0841410830616951,
+      "learning_rate": 0.00010717377852222585,
+      "loss": 1.092,
+      "step": 1033
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.09076603502035141,
+      "learning_rate": 0.0001070276188945293,
+      "loss": 1.1218,
+      "step": 1034
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.08737336099147797,
+      "learning_rate": 0.0001068814441779075,
+      "loss": 1.0521,
+      "step": 1035
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.07990612834692001,
+      "learning_rate": 0.0001067352546862106,
+      "loss": 1.0102,
+      "step": 1036
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.08482470363378525,
+      "learning_rate": 0.00010658905073332048,
+      "loss": 1.0712,
+      "step": 1037
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.08994842320680618,
+      "learning_rate": 0.00010644283263315014,
+      "loss": 1.126,
+      "step": 1038
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.0877973660826683,
+      "learning_rate": 0.00010629660069964285,
+      "loss": 1.1181,
+      "step": 1039
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.08806651085615158,
+      "learning_rate": 0.0001061503552467716,
+      "loss": 1.1333,
+      "step": 1040
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.08447300642728806,
+      "learning_rate": 0.00010600409658853847,
+      "loss": 1.1155,
+      "step": 1041
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.08353561162948608,
+      "learning_rate": 0.00010585782503897388,
+      "loss": 1.1167,
+      "step": 1042
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.08206984400749207,
+      "learning_rate": 0.00010571154091213586,
+      "loss": 1.021,
+      "step": 1043
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.08564114570617676,
+      "learning_rate": 0.00010556524452210952,
+      "loss": 1.1222,
+      "step": 1044
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.08629685640335083,
+      "learning_rate": 0.00010541893618300629,
+      "loss": 1.0748,
+      "step": 1045
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.08371461182832718,
+      "learning_rate": 0.00010527261620896323,
+      "loss": 1.1113,
+      "step": 1046
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.08749718964099884,
+      "learning_rate": 0.00010512628491414236,
+      "loss": 1.0958,
+      "step": 1047
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.08116140961647034,
+      "learning_rate": 0.00010497994261273006,
+      "loss": 1.0732,
+      "step": 1048
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.08554930984973907,
+      "learning_rate": 0.00010483358961893628,
+      "loss": 1.0564,
+      "step": 1049
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.09141719341278076,
+      "learning_rate": 0.00010468722624699401,
+      "loss": 1.0413,
+      "step": 1050
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.09536071121692657,
+      "learning_rate": 0.00010454085281115848,
+      "loss": 1.0658,
+      "step": 1051
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.09226777404546738,
+      "learning_rate": 0.00010439446962570652,
+      "loss": 1.1553,
+      "step": 1052
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.08684992790222168,
+      "learning_rate": 0.00010424807700493586,
+      "loss": 1.1536,
+      "step": 1053
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.09266629815101624,
+      "learning_rate": 0.00010410167526316457,
+      "loss": 1.0939,
+      "step": 1054
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.08212045580148697,
+      "learning_rate": 0.00010395526471473026,
+      "loss": 1.0389,
+      "step": 1055
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.09039530903100967,
+      "learning_rate": 0.00010380884567398943,
+      "loss": 1.1239,
+      "step": 1056
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.08995039761066437,
+      "learning_rate": 0.00010366241845531684,
+      "loss": 1.082,
+      "step": 1057
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.08430516719818115,
+      "learning_rate": 0.00010351598337310482,
+      "loss": 1.12,
+      "step": 1058
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.08436471223831177,
+      "learning_rate": 0.00010336954074176257,
+      "loss": 1.059,
+      "step": 1059
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.08221795409917831,
+      "learning_rate": 0.00010322309087571544,
+      "loss": 1.0816,
+      "step": 1060
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.08930781483650208,
+      "learning_rate": 0.00010307663408940446,
+      "loss": 1.0842,
+      "step": 1061
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.08792873471975327,
+      "learning_rate": 0.00010293017069728535,
+      "loss": 1.1566,
+      "step": 1062
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.08898869156837463,
+      "learning_rate": 0.0001027837010138281,
+      "loss": 1.006,
+      "step": 1063
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.08551101386547089,
+      "learning_rate": 0.00010263722535351626,
+      "loss": 1.1351,
+      "step": 1064
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.08300764113664627,
+      "learning_rate": 0.00010249074403084606,
+      "loss": 1.0498,
+      "step": 1065
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.08598140627145767,
+      "learning_rate": 0.00010234425736032607,
+      "loss": 1.117,
+      "step": 1066
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.08814927190542221,
+      "learning_rate": 0.00010219776565647617,
+      "loss": 1.1414,
+      "step": 1067
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.08437316864728928,
+      "learning_rate": 0.0001020512692338272,
+      "loss": 1.0666,
+      "step": 1068
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.0825188085436821,
+      "learning_rate": 0.00010190476840692001,
+      "loss": 1.0752,
+      "step": 1069
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.08613229542970657,
+      "learning_rate": 0.00010175826349030496,
+      "loss": 1.1396,
+      "step": 1070
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.08590491116046906,
+      "learning_rate": 0.00010161175479854121,
+      "loss": 1.0884,
+      "step": 1071
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.08533625304698944,
+      "learning_rate": 0.00010146524264619601,
+      "loss": 1.133,
+      "step": 1072
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.09502088278532028,
+      "learning_rate": 0.000101318727347844,
+      "loss": 1.0839,
+      "step": 1073
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.09577028453350067,
+      "learning_rate": 0.00010117220921806664,
+      "loss": 1.1456,
+      "step": 1074
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.0864022895693779,
+      "learning_rate": 0.00010102568857145143,
+      "loss": 1.0561,
+      "step": 1075
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.08769258111715317,
+      "learning_rate": 0.00010087916572259124,
+      "loss": 1.1448,
+      "step": 1076
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.08523321151733398,
+      "learning_rate": 0.00010073264098608377,
+      "loss": 1.1586,
+      "step": 1077
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.0859287679195404,
+      "learning_rate": 0.00010058611467653066,
+      "loss": 1.0636,
+      "step": 1078
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.09155905246734619,
+      "learning_rate": 0.00010043958710853703,
+      "loss": 1.0443,
+      "step": 1079
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.08811119943857193,
+      "learning_rate": 0.00010029305859671061,
+      "loss": 1.0849,
+      "step": 1080
+    },
+    {
+      "epoch": 1.49,
+      "eval_loss": 1.0885454416275024,
+      "eval_runtime": 97.1092,
+      "eval_samples_per_second": 51.488,
+      "eval_steps_per_second": 12.872,
+      "step": 1080
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.09313588589429855,
+      "learning_rate": 0.00010014652945566122,
+      "loss": 1.1097,
+      "step": 1081
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.08620044589042664,
+      "learning_rate": 0.0001,
+      "loss": 1.0977,
+      "step": 1082
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.08679936081171036,
+      "learning_rate": 9.985347054433878e-05,
+      "loss": 1.0991,
+      "step": 1083
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.0829780250787735,
+      "learning_rate": 9.970694140328941e-05,
+      "loss": 0.967,
+      "step": 1084
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.08080025017261505,
+      "learning_rate": 9.956041289146299e-05,
+      "loss": 1.1039,
+      "step": 1085
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.08560314029455185,
+      "learning_rate": 9.941388532346934e-05,
+      "loss": 1.0779,
+      "step": 1086
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.0891140028834343,
+      "learning_rate": 9.926735901391626e-05,
+      "loss": 1.0807,
+      "step": 1087
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.08984285593032837,
+      "learning_rate": 9.912083427740877e-05,
+      "loss": 1.1227,
+      "step": 1088
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.09170664846897125,
+      "learning_rate": 9.897431142854863e-05,
+      "loss": 1.1369,
+      "step": 1089
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.09122758358716965,
+      "learning_rate": 9.882779078193339e-05,
+      "loss": 1.0513,
+      "step": 1090
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.08716864883899689,
+      "learning_rate": 9.868127265215603e-05,
+      "loss": 1.1559,
+      "step": 1091
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.08301733434200287,
+      "learning_rate": 9.853475735380402e-05,
+      "loss": 1.1214,
+      "step": 1092
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.09429632127285004,
+      "learning_rate": 9.83882452014588e-05,
+      "loss": 1.0856,
+      "step": 1093
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.08227617293596268,
+      "learning_rate": 9.824173650969507e-05,
+      "loss": 1.0423,
+      "step": 1094
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.0875801220536232,
+      "learning_rate": 9.809523159308003e-05,
+      "loss": 1.0848,
+      "step": 1095
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.08982393890619278,
+      "learning_rate": 9.794873076617283e-05,
+      "loss": 1.0228,
+      "step": 1096
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.08761537820100784,
+      "learning_rate": 9.780223434352386e-05,
+      "loss": 1.1135,
+      "step": 1097
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.09077221900224686,
+      "learning_rate": 9.765574263967396e-05,
+      "loss": 1.0473,
+      "step": 1098
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.08941144496202469,
+      "learning_rate": 9.750925596915394e-05,
+      "loss": 1.0445,
+      "step": 1099
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.09036858379840851,
+      "learning_rate": 9.736277464648378e-05,
+      "loss": 1.037,
+      "step": 1100
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.08361994475126266,
+      "learning_rate": 9.72162989861719e-05,
+      "loss": 1.0222,
+      "step": 1101
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.08968660235404968,
+      "learning_rate": 9.706982930271465e-05,
+      "loss": 1.1361,
+      "step": 1102
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.08964897692203522,
+      "learning_rate": 9.692336591059559e-05,
+      "loss": 1.1318,
+      "step": 1103
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.08598687499761581,
+      "learning_rate": 9.677690912428458e-05,
+      "loss": 1.0837,
+      "step": 1104
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.09063093364238739,
+      "learning_rate": 9.663045925823747e-05,
+      "loss": 1.1359,
+      "step": 1105
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.08562393486499786,
+      "learning_rate": 9.648401662689521e-05,
+      "loss": 1.0762,
+      "step": 1106
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.08244912326335907,
+      "learning_rate": 9.633758154468319e-05,
+      "loss": 1.0983,
+      "step": 1107
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.0850737988948822,
+      "learning_rate": 9.61911543260106e-05,
+      "loss": 1.1197,
+      "step": 1108
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.08195479214191437,
+      "learning_rate": 9.604473528526975e-05,
+      "loss": 1.0302,
+      "step": 1109
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.0881543979048729,
+      "learning_rate": 9.589832473683547e-05,
+      "loss": 1.1176,
+      "step": 1110
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.08444536477327347,
+      "learning_rate": 9.575192299506415e-05,
+      "loss": 0.9164,
+      "step": 1111
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.08216176182031631,
+      "learning_rate": 9.56055303742935e-05,
+      "loss": 1.1116,
+      "step": 1112
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.09001285582780838,
+      "learning_rate": 9.545914718884154e-05,
+      "loss": 1.108,
+      "step": 1113
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.08972419053316116,
+      "learning_rate": 9.531277375300599e-05,
+      "loss": 1.1412,
+      "step": 1114
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.08583509176969528,
+      "learning_rate": 9.516641038106372e-05,
+      "loss": 1.0117,
+      "step": 1115
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.08382473886013031,
+      "learning_rate": 9.502005738727e-05,
+      "loss": 1.0248,
+      "step": 1116
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.08485911041498184,
+      "learning_rate": 9.48737150858577e-05,
+      "loss": 1.0615,
+      "step": 1117
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.08868833631277084,
+      "learning_rate": 9.472738379103682e-05,
+      "loss": 1.0809,
+      "step": 1118
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.08792780339717865,
+      "learning_rate": 9.458106381699372e-05,
+      "loss": 1.0657,
+      "step": 1119
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.09843221306800842,
+      "learning_rate": 9.44347554778905e-05,
+      "loss": 1.2098,
+      "step": 1120
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.08840066194534302,
+      "learning_rate": 9.428845908786416e-05,
+      "loss": 1.0306,
+      "step": 1121
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.08681873232126236,
+      "learning_rate": 9.414217496102614e-05,
+      "loss": 1.0544,
+      "step": 1122
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.08229317516088486,
+      "learning_rate": 9.399590341146156e-05,
+      "loss": 1.0268,
+      "step": 1123
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.08595220744609833,
+      "learning_rate": 9.384964475322843e-05,
+      "loss": 1.05,
+      "step": 1124
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.08658715337514877,
+      "learning_rate": 9.370339930035719e-05,
+      "loss": 1.0756,
+      "step": 1125
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.08395317196846008,
+      "learning_rate": 9.355716736684988e-05,
+      "loss": 1.0733,
+      "step": 1126
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.08322521299123764,
+      "learning_rate": 9.341094926667952e-05,
+      "loss": 1.0246,
+      "step": 1127
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.09082307666540146,
+      "learning_rate": 9.32647453137894e-05,
+      "loss": 1.1378,
+      "step": 1128
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.09710096567869186,
+      "learning_rate": 9.311855582209253e-05,
+      "loss": 1.1753,
+      "step": 1129
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.09195496141910553,
+      "learning_rate": 9.297238110547074e-05,
+      "loss": 1.0528,
+      "step": 1130
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.08516698330640793,
+      "learning_rate": 9.282622147777419e-05,
+      "loss": 1.0357,
+      "step": 1131
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.08785136789083481,
+      "learning_rate": 9.268007725282068e-05,
+      "loss": 1.0915,
+      "step": 1132
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.08511301130056381,
+      "learning_rate": 9.253394874439491e-05,
+      "loss": 1.0572,
+      "step": 1133
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.0908292606472969,
+      "learning_rate": 9.238783626624781e-05,
+      "loss": 1.0818,
+      "step": 1134
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.08461970835924149,
+      "learning_rate": 9.224174013209595e-05,
+      "loss": 1.0678,
+      "step": 1135
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.0929746925830841,
+      "learning_rate": 9.209566065562079e-05,
+      "loss": 1.0649,
+      "step": 1136
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.08492594957351685,
+      "learning_rate": 9.194959815046796e-05,
+      "loss": 1.1076,
+      "step": 1137
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.08972315490245819,
+      "learning_rate": 9.180355293024669e-05,
+      "loss": 1.1237,
+      "step": 1138
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.08630712330341339,
+      "learning_rate": 9.165752530852919e-05,
+      "loss": 1.0868,
+      "step": 1139
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.09331818670034409,
+      "learning_rate": 9.151151559884973e-05,
+      "loss": 1.1403,
+      "step": 1140
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.08604791760444641,
+      "learning_rate": 9.136552411470421e-05,
+      "loss": 1.0971,
+      "step": 1141
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.08526983112096786,
+      "learning_rate": 9.121955116954942e-05,
+      "loss": 1.0536,
+      "step": 1142
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.08930695056915283,
+      "learning_rate": 9.107359707680228e-05,
+      "loss": 1.021,
+      "step": 1143
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.09311593323945999,
+      "learning_rate": 9.092766214983924e-05,
+      "loss": 1.1028,
+      "step": 1144
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.08786014467477798,
+      "learning_rate": 9.078174670199562e-05,
+      "loss": 1.1535,
+      "step": 1145
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.0856819599866867,
+      "learning_rate": 9.063585104656493e-05,
+      "loss": 1.1133,
+      "step": 1146
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.07982165366411209,
+      "learning_rate": 9.04899754967982e-05,
+      "loss": 1.0017,
+      "step": 1147
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.08650309592485428,
+      "learning_rate": 9.03441203659032e-05,
+      "loss": 1.0352,
+      "step": 1148
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.09129110723733902,
+      "learning_rate": 9.019828596704394e-05,
+      "loss": 1.1385,
+      "step": 1149
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.08559451252222061,
+      "learning_rate": 9.005247261333993e-05,
+      "loss": 1.1037,
+      "step": 1150
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.0937194898724556,
+      "learning_rate": 8.990668061786543e-05,
+      "loss": 1.2672,
+      "step": 1151
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.0885760709643364,
+      "learning_rate": 8.976091029364888e-05,
+      "loss": 1.2034,
+      "step": 1152
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.08901859074831009,
+      "learning_rate": 8.961516195367217e-05,
+      "loss": 1.0484,
+      "step": 1153
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.08576090633869171,
+      "learning_rate": 8.946943591087e-05,
+      "loss": 1.0771,
+      "step": 1154
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.09184170514345169,
+      "learning_rate": 8.932373247812923e-05,
+      "loss": 1.1334,
+      "step": 1155
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.08978337049484253,
+      "learning_rate": 8.917805196828812e-05,
+      "loss": 0.994,
+      "step": 1156
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.08952417224645615,
+      "learning_rate": 8.903239469413573e-05,
+      "loss": 1.0762,
+      "step": 1157
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.09208005666732788,
+      "learning_rate": 8.88867609684112e-05,
+      "loss": 1.1774,
+      "step": 1158
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.08697770535945892,
+      "learning_rate": 8.874115110380317e-05,
+      "loss": 1.135,
+      "step": 1159
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.08731523901224136,
+      "learning_rate": 8.859556541294905e-05,
+      "loss": 1.0671,
+      "step": 1160
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.0901203379034996,
+      "learning_rate": 8.845000420843427e-05,
+      "loss": 1.0757,
+      "step": 1161
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.08368512243032455,
+      "learning_rate": 8.830446780279176e-05,
+      "loss": 1.1406,
+      "step": 1162
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.08632200956344604,
+      "learning_rate": 8.81589565085012e-05,
+      "loss": 1.0722,
+      "step": 1163
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.08953368663787842,
+      "learning_rate": 8.801347063798826e-05,
+      "loss": 1.0973,
+      "step": 1164
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.08895053714513779,
+      "learning_rate": 8.786801050362415e-05,
+      "loss": 1.0827,
+      "step": 1165
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.08292561024427414,
+      "learning_rate": 8.772257641772479e-05,
+      "loss": 1.034,
+      "step": 1166
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.08720581233501434,
+      "learning_rate": 8.757716869255009e-05,
+      "loss": 1.1181,
+      "step": 1167
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.087997205555439,
+      "learning_rate": 8.743178764030343e-05,
+      "loss": 1.1329,
+      "step": 1168
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.09110581129789352,
+      "learning_rate": 8.728643357313096e-05,
+      "loss": 1.1038,
+      "step": 1169
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.08588361740112305,
+      "learning_rate": 8.71411068031208e-05,
+      "loss": 1.1219,
+      "step": 1170
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.08964413404464722,
+      "learning_rate": 8.699580764230249e-05,
+      "loss": 1.1152,
+      "step": 1171
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.08481792360544205,
+      "learning_rate": 8.685053640264631e-05,
+      "loss": 1.1076,
+      "step": 1172
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.08725820481777191,
+      "learning_rate": 8.67052933960626e-05,
+      "loss": 1.0864,
+      "step": 1173
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.09291683882474899,
+      "learning_rate": 8.656007893440106e-05,
+      "loss": 1.1034,
+      "step": 1174
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.09966468065977097,
+      "learning_rate": 8.641489332945008e-05,
+      "loss": 1.0532,
+      "step": 1175
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.08579923957586288,
+      "learning_rate": 8.626973689293615e-05,
+      "loss": 1.0846,
+      "step": 1176
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.09335643798112869,
+      "learning_rate": 8.61246099365231e-05,
+      "loss": 1.0757,
+      "step": 1177
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.08344405144453049,
+      "learning_rate": 8.597951277181142e-05,
+      "loss": 1.0643,
+      "step": 1178
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.08903339505195618,
+      "learning_rate": 8.583444571033774e-05,
+      "loss": 1.0853,
+      "step": 1179
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.08619026094675064,
+      "learning_rate": 8.568940906357396e-05,
+      "loss": 1.1152,
+      "step": 1180
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.08517514914274216,
+      "learning_rate": 8.554440314292668e-05,
+      "loss": 1.0853,
+      "step": 1181
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.0899636372923851,
+      "learning_rate": 8.539942825973666e-05,
+      "loss": 1.1488,
+      "step": 1182
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.0933493822813034,
+      "learning_rate": 8.525448472527784e-05,
+      "loss": 1.1207,
+      "step": 1183
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.08925125002861023,
+      "learning_rate": 8.510957285075692e-05,
+      "loss": 1.0764,
+      "step": 1184
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.0878109335899353,
+      "learning_rate": 8.496469294731264e-05,
+      "loss": 1.0107,
+      "step": 1185
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.1002742201089859,
+      "learning_rate": 8.481984532601515e-05,
+      "loss": 1.2072,
+      "step": 1186
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.08254406601190567,
+      "learning_rate": 8.467503029786514e-05,
+      "loss": 1.0487,
+      "step": 1187
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.08793588727712631,
+      "learning_rate": 8.453024817379347e-05,
+      "loss": 1.0915,
+      "step": 1188
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.08795217424631119,
+      "learning_rate": 8.438549926466027e-05,
+      "loss": 1.0778,
+      "step": 1189
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.08860927820205688,
+      "learning_rate": 8.424078388125436e-05,
+      "loss": 1.0931,
+      "step": 1190
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.09497710317373276,
+      "learning_rate": 8.409610233429259e-05,
+      "loss": 1.1505,
+      "step": 1191
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.08706358820199966,
+      "learning_rate": 8.395145493441921e-05,
+      "loss": 1.1076,
+      "step": 1192
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.08769765496253967,
+      "learning_rate": 8.380684199220502e-05,
+      "loss": 1.0434,
+      "step": 1193
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.09465378522872925,
+      "learning_rate": 8.366226381814697e-05,
+      "loss": 1.1677,
+      "step": 1194
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.08951201289892197,
+      "learning_rate": 8.351772072266733e-05,
+      "loss": 1.133,
+      "step": 1195
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.08738221973180771,
+      "learning_rate": 8.337321301611301e-05,
+      "loss": 1.0187,
+      "step": 1196
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.08495625853538513,
+      "learning_rate": 8.322874100875492e-05,
+      "loss": 1.053,
+      "step": 1197
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.09371285885572433,
+      "learning_rate": 8.308430501078739e-05,
+      "loss": 1.2348,
+      "step": 1198
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.09792681038379669,
+      "learning_rate": 8.293990533232743e-05,
+      "loss": 1.1307,
+      "step": 1199
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.0853876918554306,
+      "learning_rate": 8.2795542283414e-05,
+      "loss": 1.1236,
+      "step": 1200
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.08959940820932388,
+      "learning_rate": 8.265121617400747e-05,
+      "loss": 1.0187,
+      "step": 1201
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.08905159682035446,
+      "learning_rate": 8.25069273139889e-05,
+      "loss": 1.0772,
+      "step": 1202
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.08399379253387451,
+      "learning_rate": 8.236267601315935e-05,
+      "loss": 1.0727,
+      "step": 1203
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.08472379297018051,
+      "learning_rate": 8.221846258123921e-05,
+      "loss": 1.0514,
+      "step": 1204
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.09165279567241669,
+      "learning_rate": 8.207428732786766e-05,
+      "loss": 1.0687,
+      "step": 1205
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.09237626940011978,
+      "learning_rate": 8.19301505626018e-05,
+      "loss": 1.0487,
+      "step": 1206
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.08367465436458588,
+      "learning_rate": 8.178605259491611e-05,
+      "loss": 1.1032,
+      "step": 1207
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.08903051912784576,
+      "learning_rate": 8.164199373420187e-05,
+      "loss": 1.1018,
+      "step": 1208
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.08853001892566681,
+      "learning_rate": 8.149797428976624e-05,
+      "loss": 1.1208,
+      "step": 1209
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.08579990267753601,
+      "learning_rate": 8.13539945708319e-05,
+      "loss": 1.0659,
+      "step": 1210
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.085668183863163,
+      "learning_rate": 8.121005488653609e-05,
+      "loss": 1.1048,
+      "step": 1211
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.09721795469522476,
+      "learning_rate": 8.106615554593021e-05,
+      "loss": 1.1408,
+      "step": 1212
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.08581605553627014,
+      "learning_rate": 8.092229685797895e-05,
+      "loss": 1.0879,
+      "step": 1213
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.08966419845819473,
+      "learning_rate": 8.07784791315598e-05,
+      "loss": 1.1583,
+      "step": 1214
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.08896504342556,
+      "learning_rate": 8.063470267546226e-05,
+      "loss": 1.1903,
+      "step": 1215
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.09235654026269913,
+      "learning_rate": 8.049096779838719e-05,
+      "loss": 1.0966,
+      "step": 1216
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.08318577706813812,
+      "learning_rate": 8.034727480894621e-05,
+      "loss": 1.0831,
+      "step": 1217
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.08445431292057037,
+      "learning_rate": 8.020362401566103e-05,
+      "loss": 1.073,
+      "step": 1218
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.08299394696950912,
+      "learning_rate": 8.00600157269627e-05,
+      "loss": 1.016,
+      "step": 1219
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.08932309597730637,
+      "learning_rate": 7.991645025119104e-05,
+      "loss": 1.1142,
+      "step": 1220
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.09079217910766602,
+      "learning_rate": 7.977292789659399e-05,
+      "loss": 1.1535,
+      "step": 1221
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.09069611877202988,
+      "learning_rate": 7.962944897132678e-05,
+      "loss": 1.1488,
+      "step": 1222
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.08451555669307709,
+      "learning_rate": 7.948601378345157e-05,
+      "loss": 1.0251,
+      "step": 1223
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.08608248829841614,
+      "learning_rate": 7.934262264093642e-05,
+      "loss": 1.1113,
+      "step": 1224
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.08254078775644302,
+      "learning_rate": 7.919927585165498e-05,
+      "loss": 1.0436,
+      "step": 1225
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.08391185849905014,
+      "learning_rate": 7.905597372338558e-05,
+      "loss": 1.0443,
+      "step": 1226
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.0895843505859375,
+      "learning_rate": 7.891271656381067e-05,
+      "loss": 1.134,
+      "step": 1227
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.0923517569899559,
+      "learning_rate": 7.876950468051623e-05,
+      "loss": 1.1015,
+      "step": 1228
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.08910176157951355,
+      "learning_rate": 7.862633838099087e-05,
+      "loss": 1.0813,
+      "step": 1229
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.09368553757667542,
+      "learning_rate": 7.848321797262548e-05,
+      "loss": 1.1199,
+      "step": 1230
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.09111451357603073,
+      "learning_rate": 7.834014376271235e-05,
+      "loss": 1.1747,
+      "step": 1231
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.08973907679319382,
+      "learning_rate": 7.819711605844454e-05,
+      "loss": 1.0631,
+      "step": 1232
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.09037698060274124,
+      "learning_rate": 7.805413516691535e-05,
+      "loss": 1.1312,
+      "step": 1233
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.0851440504193306,
+      "learning_rate": 7.791120139511752e-05,
+      "loss": 1.0225,
+      "step": 1234
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.08332479000091553,
+      "learning_rate": 7.77683150499426e-05,
+      "loss": 1.0809,
+      "step": 1235
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.0928303524851799,
+      "learning_rate": 7.762547643818032e-05,
+      "loss": 1.1005,
+      "step": 1236
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.08913961797952652,
+      "learning_rate": 7.748268586651801e-05,
+      "loss": 1.0543,
+      "step": 1237
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.09041823446750641,
+      "learning_rate": 7.733994364153969e-05,
+      "loss": 1.1725,
+      "step": 1238
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.08820322155952454,
+      "learning_rate": 7.719725006972567e-05,
+      "loss": 1.099,
+      "step": 1239
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.0889720693230629,
+      "learning_rate": 7.705460545745182e-05,
+      "loss": 1.1158,
+      "step": 1240
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.0891098827123642,
+      "learning_rate": 7.691201011098885e-05,
+      "loss": 1.1931,
+      "step": 1241
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.0883520245552063,
+      "learning_rate": 7.67694643365017e-05,
+      "loss": 1.1366,
+      "step": 1242
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.08913884311914444,
+      "learning_rate": 7.662696844004887e-05,
+      "loss": 1.0712,
+      "step": 1243
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.08595810830593109,
+      "learning_rate": 7.64845227275818e-05,
+      "loss": 1.1192,
+      "step": 1244
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.08838467299938202,
+      "learning_rate": 7.634212750494411e-05,
+      "loss": 1.0921,
+      "step": 1245
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.08665958791971207,
+      "learning_rate": 7.619978307787108e-05,
+      "loss": 1.0475,
+      "step": 1246
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.09563381224870682,
+      "learning_rate": 7.605748975198893e-05,
+      "loss": 1.0999,
+      "step": 1247
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.08659172803163528,
+      "learning_rate": 7.591524783281409e-05,
+      "loss": 1.074,
+      "step": 1248
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.08777296543121338,
+      "learning_rate": 7.57730576257527e-05,
+      "loss": 1.0894,
+      "step": 1249
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.09524232149124146,
+      "learning_rate": 7.563091943609984e-05,
+      "loss": 1.0493,
+      "step": 1250
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.09214726835489273,
+      "learning_rate": 7.548883356903885e-05,
+      "loss": 1.0895,
+      "step": 1251
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.09514345973730087,
+      "learning_rate": 7.534680032964078e-05,
+      "loss": 1.0622,
+      "step": 1252
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.08954614400863647,
+      "learning_rate": 7.520482002286368e-05,
+      "loss": 1.1818,
+      "step": 1253
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.08538757264614105,
+      "learning_rate": 7.506289295355198e-05,
+      "loss": 1.1286,
+      "step": 1254
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.08991134911775589,
+      "learning_rate": 7.492101942643572e-05,
+      "loss": 1.1396,
+      "step": 1255
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.09406349062919617,
+      "learning_rate": 7.477919974613007e-05,
+      "loss": 1.1755,
+      "step": 1256
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.08960568159818649,
+      "learning_rate": 7.463743421713454e-05,
+      "loss": 1.0782,
+      "step": 1257
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.08744941651821136,
+      "learning_rate": 7.449572314383237e-05,
+      "loss": 1.0558,
+      "step": 1258
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.08898703008890152,
+      "learning_rate": 7.435406683048987e-05,
+      "loss": 1.183,
+      "step": 1259
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.09725111722946167,
+      "learning_rate": 7.421246558125585e-05,
+      "loss": 1.0629,
+      "step": 1260
+    },
+    {
+      "epoch": 1.74,
+      "eval_loss": 1.087813377380371,
+      "eval_runtime": 97.0335,
+      "eval_samples_per_second": 51.529,
+      "eval_steps_per_second": 12.882,
+      "step": 1260
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.08966143429279327,
+      "learning_rate": 7.407091970016078e-05,
+      "loss": 1.0665,
+      "step": 1261
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.0915473997592926,
+      "learning_rate": 7.392942949111633e-05,
+      "loss": 1.0731,
+      "step": 1262
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.08808726817369461,
+      "learning_rate": 7.378799525791462e-05,
+      "loss": 1.1236,
+      "step": 1263
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.0844448134303093,
+      "learning_rate": 7.364661730422756e-05,
+      "loss": 1.037,
+      "step": 1264
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.08915546536445618,
+      "learning_rate": 7.350529593360623e-05,
+      "loss": 1.1232,
+      "step": 1265
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.09011270105838776,
+      "learning_rate": 7.336403144948022e-05,
+      "loss": 1.0891,
+      "step": 1266
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.08984367549419403,
+      "learning_rate": 7.322282415515708e-05,
+      "loss": 1.0993,
+      "step": 1267
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.0926932767033577,
+      "learning_rate": 7.308167435382137e-05,
+      "loss": 1.1139,
+      "step": 1268
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.09094691276550293,
+      "learning_rate": 7.294058234853437e-05,
+      "loss": 0.9993,
+      "step": 1269
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.08597652614116669,
+      "learning_rate": 7.279954844223323e-05,
+      "loss": 1.0795,
+      "step": 1270
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.08908293396234512,
+      "learning_rate": 7.265857293773028e-05,
+      "loss": 1.0497,
+      "step": 1271
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.0965980812907219,
+      "learning_rate": 7.251765613771257e-05,
+      "loss": 1.0227,
+      "step": 1272
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.0919419601559639,
+      "learning_rate": 7.237679834474105e-05,
+      "loss": 1.0878,
+      "step": 1273
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.09026243537664413,
+      "learning_rate": 7.223599986124994e-05,
+      "loss": 0.998,
+      "step": 1274
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.08920803666114807,
+      "learning_rate": 7.209526098954618e-05,
+      "loss": 1.1311,
+      "step": 1275
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.08237479627132416,
+      "learning_rate": 7.195458203180872e-05,
+      "loss": 1.028,
+      "step": 1276
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.09907656908035278,
+      "learning_rate": 7.181396329008782e-05,
+      "loss": 1.0628,
+      "step": 1277
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.08868470042943954,
+      "learning_rate": 7.167340506630445e-05,
+      "loss": 1.0957,
+      "step": 1278
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.08448733389377594,
+      "learning_rate": 7.153290766224969e-05,
+      "loss": 1.0563,
+      "step": 1279
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.09597695618867874,
+      "learning_rate": 7.1392471379584e-05,
+      "loss": 1.1293,
+      "step": 1280
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.08930587768554688,
+      "learning_rate": 7.125209651983664e-05,
+      "loss": 1.073,
+      "step": 1281
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.09344474226236343,
+      "learning_rate": 7.111178338440496e-05,
+      "loss": 1.165,
+      "step": 1282
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.0895252525806427,
+      "learning_rate": 7.097153227455379e-05,
+      "loss": 1.0997,
+      "step": 1283
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.08774872869253159,
+      "learning_rate": 7.083134349141475e-05,
+      "loss": 1.122,
+      "step": 1284
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.09199096262454987,
+      "learning_rate": 7.069121733598572e-05,
+      "loss": 1.083,
+      "step": 1285
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.09524177759885788,
+      "learning_rate": 7.055115410913009e-05,
+      "loss": 1.0968,
+      "step": 1286
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.09341250360012054,
+      "learning_rate": 7.041115411157604e-05,
+      "loss": 1.144,
+      "step": 1287
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.08840780705213547,
+      "learning_rate": 7.02712176439161e-05,
+      "loss": 1.0985,
+      "step": 1288
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.09169577807188034,
+      "learning_rate": 7.013134500660641e-05,
+      "loss": 1.1169,
+      "step": 1289
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.08526241779327393,
+      "learning_rate": 6.999153649996595e-05,
+      "loss": 1.0351,
+      "step": 1290
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.09697016328573227,
+      "learning_rate": 6.985179242417607e-05,
+      "loss": 1.1353,
+      "step": 1291
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.09291993826627731,
+      "learning_rate": 6.971211307927981e-05,
+      "loss": 1.1357,
+      "step": 1292
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.0878898948431015,
+      "learning_rate": 6.957249876518115e-05,
+      "loss": 1.1034,
+      "step": 1293
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.0865277647972107,
+      "learning_rate": 6.943294978164454e-05,
+      "loss": 1.0824,
+      "step": 1294
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.0858265608549118,
+      "learning_rate": 6.929346642829408e-05,
+      "loss": 1.1242,
+      "step": 1295
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.09510330855846405,
+      "learning_rate": 6.915404900461297e-05,
+      "loss": 1.0961,
+      "step": 1296
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.08849316090345383,
+      "learning_rate": 6.901469780994291e-05,
+      "loss": 1.1239,
+      "step": 1297
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.09125662595033646,
+      "learning_rate": 6.887541314348333e-05,
+      "loss": 1.056,
+      "step": 1298
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.08722524344921112,
+      "learning_rate": 6.873619530429083e-05,
+      "loss": 1.0212,
+      "step": 1299
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.09114434570074081,
+      "learning_rate": 6.85970445912786e-05,
+      "loss": 1.0584,
+      "step": 1300
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.0878918468952179,
+      "learning_rate": 6.845796130321556e-05,
+      "loss": 1.0961,
+      "step": 1301
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.09036703407764435,
+      "learning_rate": 6.831894573872601e-05,
+      "loss": 1.0384,
+      "step": 1302
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.0844586044549942,
+      "learning_rate": 6.81799981962888e-05,
+      "loss": 1.0002,
+      "step": 1303
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.09075265377759933,
+      "learning_rate": 6.804111897423667e-05,
+      "loss": 1.001,
+      "step": 1304
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.08699437230825424,
+      "learning_rate": 6.790230837075569e-05,
+      "loss": 1.1198,
+      "step": 1305
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.08844509720802307,
+      "learning_rate": 6.776356668388464e-05,
+      "loss": 1.0886,
+      "step": 1306
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.08665040880441666,
+      "learning_rate": 6.762489421151434e-05,
+      "loss": 1.0571,
+      "step": 1307
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.08629337698221207,
+      "learning_rate": 6.748629125138691e-05,
+      "loss": 1.0885,
+      "step": 1308
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.09220010042190552,
+      "learning_rate": 6.734775810109533e-05,
+      "loss": 1.0952,
+      "step": 1309
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.09041441231966019,
+      "learning_rate": 6.720929505808265e-05,
+      "loss": 1.0501,
+      "step": 1310
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.08775871247053146,
+      "learning_rate": 6.707090241964134e-05,
+      "loss": 1.1374,
+      "step": 1311
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.0879032239317894,
+      "learning_rate": 6.693258048291277e-05,
+      "loss": 1.1375,
+      "step": 1312
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.08759757876396179,
+      "learning_rate": 6.679432954488655e-05,
+      "loss": 1.0403,
+      "step": 1313
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.0876547172665596,
+      "learning_rate": 6.665614990239969e-05,
+      "loss": 1.0792,
+      "step": 1314
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.10128232836723328,
+      "learning_rate": 6.651804185213627e-05,
+      "loss": 1.1643,
+      "step": 1315
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.0852789506316185,
+      "learning_rate": 6.638000569062664e-05,
+      "loss": 1.0715,
+      "step": 1316
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.0875835195183754,
+      "learning_rate": 6.624204171424673e-05,
+      "loss": 1.1001,
+      "step": 1317
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.09133362025022507,
+      "learning_rate": 6.610415021921747e-05,
+      "loss": 1.0942,
+      "step": 1318
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.08645410090684891,
+      "learning_rate": 6.596633150160426e-05,
+      "loss": 1.026,
+      "step": 1319
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.0897056832909584,
+      "learning_rate": 6.58285858573162e-05,
+      "loss": 1.1362,
+      "step": 1320
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.10235519707202911,
+      "learning_rate": 6.569091358210546e-05,
+      "loss": 1.144,
+      "step": 1321
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.0917138010263443,
+      "learning_rate": 6.555331497156672e-05,
+      "loss": 1.1132,
+      "step": 1322
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.09157434105873108,
+      "learning_rate": 6.54157903211365e-05,
+      "loss": 1.0438,
+      "step": 1323
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.09228910505771637,
+      "learning_rate": 6.527833992609248e-05,
+      "loss": 1.0564,
+      "step": 1324
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.0902438685297966,
+      "learning_rate": 6.514096408155294e-05,
+      "loss": 1.1119,
+      "step": 1325
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.09249547868967056,
+      "learning_rate": 6.50036630824761e-05,
+      "loss": 1.2069,
+      "step": 1326
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.08891062438488007,
+      "learning_rate": 6.486643722365943e-05,
+      "loss": 1.1228,
+      "step": 1327
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.09137904644012451,
+      "learning_rate": 6.472928679973912e-05,
+      "loss": 1.1202,
+      "step": 1328
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.08870182186365128,
+      "learning_rate": 6.45922121051894e-05,
+      "loss": 0.9974,
+      "step": 1329
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.09010548889636993,
+      "learning_rate": 6.445521343432188e-05,
+      "loss": 1.0843,
+      "step": 1330
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.08569913357496262,
+      "learning_rate": 6.431829108128487e-05,
+      "loss": 1.0746,
+      "step": 1331
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.09511716663837433,
+      "learning_rate": 6.418144534006292e-05,
+      "loss": 1.1999,
+      "step": 1332
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.08663185685873032,
+      "learning_rate": 6.404467650447612e-05,
+      "loss": 1.0887,
+      "step": 1333
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.0993635505437851,
+      "learning_rate": 6.390798486817929e-05,
+      "loss": 1.1503,
+      "step": 1334
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.08961442857980728,
+      "learning_rate": 6.377137072466159e-05,
+      "loss": 1.0608,
+      "step": 1335
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.0859212800860405,
+      "learning_rate": 6.363483436724583e-05,
+      "loss": 1.073,
+      "step": 1336
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.08427256345748901,
+      "learning_rate": 6.349837608908774e-05,
+      "loss": 1.0334,
+      "step": 1337
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.08796001970767975,
+      "learning_rate": 6.336199618317537e-05,
+      "loss": 1.0962,
+      "step": 1338
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.08590462803840637,
+      "learning_rate": 6.322569494232868e-05,
+      "loss": 1.059,
+      "step": 1339
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.08501654863357544,
+      "learning_rate": 6.308947265919849e-05,
+      "loss": 1.0245,
+      "step": 1340
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.08513573557138443,
+      "learning_rate": 6.295332962626623e-05,
+      "loss": 0.9842,
+      "step": 1341
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.08738056570291519,
+      "learning_rate": 6.281726613584321e-05,
+      "loss": 1.0686,
+      "step": 1342
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.09028667956590652,
+      "learning_rate": 6.268128248006982e-05,
+      "loss": 1.1008,
+      "step": 1343
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.08641254156827927,
+      "learning_rate": 6.254537895091518e-05,
+      "loss": 1.0192,
+      "step": 1344
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.08913418650627136,
+      "learning_rate": 6.240955584017622e-05,
+      "loss": 1.1069,
+      "step": 1345
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.08536715060472488,
+      "learning_rate": 6.227381343947733e-05,
+      "loss": 1.0832,
+      "step": 1346
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.08903218060731888,
+      "learning_rate": 6.213815204026957e-05,
+      "loss": 1.0197,
+      "step": 1347
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.08824066072702408,
+      "learning_rate": 6.200257193383006e-05,
+      "loss": 1.1067,
+      "step": 1348
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.08809547126293182,
+      "learning_rate": 6.186707341126141e-05,
+      "loss": 1.0564,
+      "step": 1349
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.0945824459195137,
+      "learning_rate": 6.173165676349103e-05,
+      "loss": 1.115,
+      "step": 1350
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.0924530029296875,
+      "learning_rate": 6.159632228127057e-05,
+      "loss": 1.0992,
+      "step": 1351
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.09352656453847885,
+      "learning_rate": 6.146107025517524e-05,
+      "loss": 1.131,
+      "step": 1352
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.09185071289539337,
+      "learning_rate": 6.132590097560322e-05,
+      "loss": 1.1045,
+      "step": 1353
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.08831333369016647,
+      "learning_rate": 6.119081473277501e-05,
+      "loss": 1.1655,
+      "step": 1354
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.08583110570907593,
+      "learning_rate": 6.105581181673287e-05,
+      "loss": 1.061,
+      "step": 1355
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.08521490544080734,
+      "learning_rate": 6.0920892517340077e-05,
+      "loss": 1.1266,
+      "step": 1356
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.1055929958820343,
+      "learning_rate": 6.078605712428047e-05,
+      "loss": 1.1286,
+      "step": 1357
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.08971645683050156,
+      "learning_rate": 6.065130592705759e-05,
+      "loss": 1.1861,
+      "step": 1358
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.08825929462909698,
+      "learning_rate": 6.051663921499435e-05,
+      "loss": 0.9411,
+      "step": 1359
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.0894852951169014,
+      "learning_rate": 6.0382057277232184e-05,
+      "loss": 1.1486,
+      "step": 1360
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.0972943976521492,
+      "learning_rate": 6.024756040273052e-05,
+      "loss": 1.1913,
+      "step": 1361
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.08465013653039932,
+      "learning_rate": 6.01131488802662e-05,
+      "loss": 1.1442,
+      "step": 1362
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.08698281645774841,
+      "learning_rate": 5.997882299843272e-05,
+      "loss": 1.0575,
+      "step": 1363
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.08935410529375076,
+      "learning_rate": 5.9844583045639734e-05,
+      "loss": 1.1568,
+      "step": 1364
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.08780406415462494,
+      "learning_rate": 5.971042931011248e-05,
+      "loss": 1.1071,
+      "step": 1365
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.08981499820947647,
+      "learning_rate": 5.9576362079890925e-05,
+      "loss": 1.1244,
+      "step": 1366
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.08927271515130997,
+      "learning_rate": 5.9442381642829424e-05,
+      "loss": 1.0942,
+      "step": 1367
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.08475492149591446,
+      "learning_rate": 5.930848828659594e-05,
+      "loss": 1.0602,
+      "step": 1368
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.08721037954092026,
+      "learning_rate": 5.9174682298671446e-05,
+      "loss": 1.1014,
+      "step": 1369
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.08866467326879501,
+      "learning_rate": 5.904096396634935e-05,
+      "loss": 1.0992,
+      "step": 1370
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.09010431915521622,
+      "learning_rate": 5.89073335767349e-05,
+      "loss": 1.139,
+      "step": 1371
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.08755801618099213,
+      "learning_rate": 5.87737914167444e-05,
+      "loss": 1.0865,
+      "step": 1372
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.08487731218338013,
+      "learning_rate": 5.8640337773104816e-05,
+      "loss": 1.0546,
+      "step": 1373
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.09027353674173355,
+      "learning_rate": 5.8506972932353035e-05,
+      "loss": 1.0377,
+      "step": 1374
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.09060210734605789,
+      "learning_rate": 5.8373697180835305e-05,
+      "loss": 1.1405,
+      "step": 1375
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.08755095303058624,
+      "learning_rate": 5.824051080470654e-05,
+      "loss": 1.1054,
+      "step": 1376
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.09296034276485443,
+      "learning_rate": 5.810741408992977e-05,
+      "loss": 1.0837,
+      "step": 1377
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.09099520742893219,
+      "learning_rate": 5.797440732227555e-05,
+      "loss": 1.1317,
+      "step": 1378
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.09965849667787552,
+      "learning_rate": 5.784149078732128e-05,
+      "loss": 1.109,
+      "step": 1379
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.08717842400074005,
+      "learning_rate": 5.770866477045067e-05,
+      "loss": 1.1144,
+      "step": 1380
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.08823736757040024,
+      "learning_rate": 5.7575929556852956e-05,
+      "loss": 1.0415,
+      "step": 1381
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.09152673929929733,
+      "learning_rate": 5.744328543152253e-05,
+      "loss": 1.1866,
+      "step": 1382
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.09454263746738434,
+      "learning_rate": 5.7310732679258185e-05,
+      "loss": 1.1066,
+      "step": 1383
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.08985551446676254,
+      "learning_rate": 5.7178271584662535e-05,
+      "loss": 1.0604,
+      "step": 1384
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.09177891910076141,
+      "learning_rate": 5.7045902432141316e-05,
+      "loss": 1.1235,
+      "step": 1385
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.09092294424772263,
+      "learning_rate": 5.691362550590297e-05,
+      "loss": 1.1185,
+      "step": 1386
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.08297441899776459,
+      "learning_rate": 5.6781441089957864e-05,
+      "loss": 1.0643,
+      "step": 1387
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.09762497991323471,
+      "learning_rate": 5.6649349468117706e-05,
+      "loss": 1.1674,
+      "step": 1388
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.27256688475608826,
+      "learning_rate": 5.651735092399503e-05,
+      "loss": 1.04,
+      "step": 1389
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.08811064064502716,
+      "learning_rate": 5.638544574100249e-05,
+      "loss": 1.0693,
+      "step": 1390
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.08900339156389236,
+      "learning_rate": 5.625363420235228e-05,
+      "loss": 1.0742,
+      "step": 1391
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.08699032664299011,
+      "learning_rate": 5.6121916591055565e-05,
+      "loss": 1.1009,
+      "step": 1392
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.09272560477256775,
+      "learning_rate": 5.599029318992184e-05,
+      "loss": 1.0207,
+      "step": 1393
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.0854259729385376,
+      "learning_rate": 5.585876428155824e-05,
+      "loss": 1.0796,
+      "step": 1394
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.10661471635103226,
+      "learning_rate": 5.57273301483691e-05,
+      "loss": 1.1077,
+      "step": 1395
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.09196636080741882,
+      "learning_rate": 5.559599107255524e-05,
+      "loss": 1.0705,
+      "step": 1396
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.09632512927055359,
+      "learning_rate": 5.546474733611344e-05,
+      "loss": 1.1419,
+      "step": 1397
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.09986668825149536,
+      "learning_rate": 5.533359922083562e-05,
+      "loss": 1.1917,
+      "step": 1398
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.09595406800508499,
+      "learning_rate": 5.520254700830853e-05,
+      "loss": 1.0723,
+      "step": 1399
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.10813736915588379,
+      "learning_rate": 5.5071590979913046e-05,
+      "loss": 1.0976,
+      "step": 1400
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.0890180766582489,
+      "learning_rate": 5.494073141682334e-05,
+      "loss": 1.0728,
+      "step": 1401
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.09232466667890549,
+      "learning_rate": 5.4809968600006635e-05,
+      "loss": 1.1543,
+      "step": 1402
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.09320066124200821,
+      "learning_rate": 5.4679302810222355e-05,
+      "loss": 1.1332,
+      "step": 1403
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.09119054675102234,
+      "learning_rate": 5.4548734328021655e-05,
+      "loss": 1.0136,
+      "step": 1404
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.08590566366910934,
+      "learning_rate": 5.44182634337467e-05,
+      "loss": 1.1004,
+      "step": 1405
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.09079393744468689,
+      "learning_rate": 5.4287890407530175e-05,
+      "loss": 1.0667,
+      "step": 1406
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.09354889392852783,
+      "learning_rate": 5.415761552929455e-05,
+      "loss": 1.0768,
+      "step": 1407
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.09370460361242294,
+      "learning_rate": 5.4027439078751666e-05,
+      "loss": 1.0527,
+      "step": 1408
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.09006518125534058,
+      "learning_rate": 5.389736133540198e-05,
+      "loss": 1.1209,
+      "step": 1409
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.08552158623933792,
+      "learning_rate": 5.3767382578534e-05,
+      "loss": 1.073,
+      "step": 1410
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.09304409474134445,
+      "learning_rate": 5.3637503087223784e-05,
+      "loss": 1.1205,
+      "step": 1411
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.09761402755975723,
+      "learning_rate": 5.350772314033412e-05,
+      "loss": 1.0624,
+      "step": 1412
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.090894415974617,
+      "learning_rate": 5.337804301651421e-05,
+      "loss": 1.0048,
+      "step": 1413
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.08948024362325668,
+      "learning_rate": 5.324846299419879e-05,
+      "loss": 1.1133,
+      "step": 1414
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.09624718874692917,
+      "learning_rate": 5.3118983351607765e-05,
+      "loss": 1.0775,
+      "step": 1415
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.10124653577804565,
+      "learning_rate": 5.2989604366745516e-05,
+      "loss": 1.1361,
+      "step": 1416
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.08634812384843826,
+      "learning_rate": 5.286032631740023e-05,
+      "loss": 1.0195,
+      "step": 1417
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.09139488637447357,
+      "learning_rate": 5.273114948114346e-05,
+      "loss": 1.0741,
+      "step": 1418
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.08641989529132843,
+      "learning_rate": 5.260207413532945e-05,
+      "loss": 1.0297,
+      "step": 1419
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.08730223029851913,
+      "learning_rate": 5.247310055709438e-05,
+      "loss": 1.0814,
+      "step": 1420
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.09232880175113678,
+      "learning_rate": 5.234422902335609e-05,
+      "loss": 1.1526,
+      "step": 1421
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.09333455562591553,
+      "learning_rate": 5.2215459810813306e-05,
+      "loss": 1.1131,
+      "step": 1422
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.09269339591264725,
+      "learning_rate": 5.208679319594497e-05,
+      "loss": 1.1363,
+      "step": 1423
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.08907824754714966,
+      "learning_rate": 5.1958229455009876e-05,
+      "loss": 1.072,
+      "step": 1424
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.08229704946279526,
+      "learning_rate": 5.182976886404575e-05,
+      "loss": 1.0926,
+      "step": 1425
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.08884652704000473,
+      "learning_rate": 5.170141169886904e-05,
+      "loss": 1.072,
+      "step": 1426
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.09383190423250198,
+      "learning_rate": 5.157315823507396e-05,
+      "loss": 1.1201,
+      "step": 1427
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.08669448643922806,
+      "learning_rate": 5.1445008748032185e-05,
+      "loss": 1.0882,
+      "step": 1428
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.08610597997903824,
+      "learning_rate": 5.131696351289211e-05,
+      "loss": 1.0827,
+      "step": 1429
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.09162358194589615,
+      "learning_rate": 5.118902280457829e-05,
+      "loss": 1.142,
+      "step": 1430
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.08728539943695068,
+      "learning_rate": 5.106118689779086e-05,
+      "loss": 1.0363,
+      "step": 1431
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.12429453432559967,
+      "learning_rate": 5.093345606700494e-05,
+      "loss": 1.1241,
+      "step": 1432
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.09186503291130066,
+      "learning_rate": 5.080583058646997e-05,
+      "loss": 1.0076,
+      "step": 1433
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.08748605847358704,
+      "learning_rate": 5.0678310730209275e-05,
+      "loss": 1.1003,
+      "step": 1434
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.08991532772779465,
+      "learning_rate": 5.0550896772019384e-05,
+      "loss": 1.0619,
+      "step": 1435
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.09335019439458847,
+      "learning_rate": 5.042358898546944e-05,
+      "loss": 1.1002,
+      "step": 1436
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.09859992563724518,
+      "learning_rate": 5.0296387643900636e-05,
+      "loss": 1.2158,
+      "step": 1437
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.09363356232643127,
+      "learning_rate": 5.016929302042563e-05,
+      "loss": 1.0424,
+      "step": 1438
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.09643103927373886,
+      "learning_rate": 5.004230538792791e-05,
+      "loss": 1.1151,
+      "step": 1439
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.09008517116308212,
+      "learning_rate": 4.9915425019061224e-05,
+      "loss": 1.0165,
+      "step": 1440
+    },
+    {
+      "epoch": 1.99,
+      "eval_loss": 1.0866000652313232,
+      "eval_runtime": 96.8477,
+      "eval_samples_per_second": 51.627,
+      "eval_steps_per_second": 12.907,
+      "step": 1440
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.09026389569044113,
+      "learning_rate": 4.978865218624906e-05,
+      "loss": 1.0368,
+      "step": 1441
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.09662413597106934,
+      "learning_rate": 4.9661987161684045e-05,
+      "loss": 1.1799,
+      "step": 1442
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.08834712952375412,
+      "learning_rate": 4.953543021732727e-05,
+      "loss": 1.1138,
+      "step": 1443
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.08902858942747116,
+      "learning_rate": 4.9408981624907815e-05,
+      "loss": 1.0431,
+      "step": 1444
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.0973421111702919,
+      "learning_rate": 4.928264165592212e-05,
+      "loss": 1.1493,
+      "step": 1445
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.09019078314304352,
+      "learning_rate": 4.9156410581633317e-05,
+      "loss": 1.0703,
+      "step": 1446
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.09086113423109055,
+      "learning_rate": 4.903028867307082e-05,
+      "loss": 1.0874,
+      "step": 1447
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.09136053919792175,
+      "learning_rate": 4.890427620102964e-05,
+      "loss": 1.1472,
+      "step": 1448
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.08980842679738998,
+      "learning_rate": 4.87783734360698e-05,
+      "loss": 1.0863,
+      "step": 1449
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.08701515942811966,
+      "learning_rate": 4.865258064851579e-05,
+      "loss": 1.0549,
+      "step": 1450
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.08598434925079346,
+      "learning_rate": 4.852689810845597e-05,
+      "loss": 1.1522,
+      "step": 1451
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.09614142775535583,
+      "learning_rate": 4.840132608574195e-05,
+      "loss": 1.1458,
+      "step": 1452
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.10245929658412933,
+      "learning_rate": 4.827586484998804e-05,
+      "loss": 1.164,
+      "step": 1453
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.09107328951358795,
+      "learning_rate": 4.8150514670570754e-05,
+      "loss": 1.1194,
+      "step": 1454
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.09416957944631577,
+      "learning_rate": 4.8025275816628124e-05,
+      "loss": 1.1264,
+      "step": 1455
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.08844498544931412,
+      "learning_rate": 4.790014855705913e-05,
+      "loss": 1.0895,
+      "step": 1456
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.08474775403738022,
+      "learning_rate": 4.77751331605232e-05,
+      "loss": 1.0535,
+      "step": 1457
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.08969343453645706,
+      "learning_rate": 4.765022989543958e-05,
+      "loss": 1.0941,
+      "step": 1458
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.0917932316660881,
+      "learning_rate": 4.752543902998665e-05,
+      "loss": 1.0095,
+      "step": 1459
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.09174329042434692,
+      "learning_rate": 4.740076083210158e-05,
+      "loss": 1.1048,
+      "step": 1460
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.10104726254940033,
+      "learning_rate": 4.7276195569479596e-05,
+      "loss": 1.0,
+      "step": 1461
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.08537132292985916,
+      "learning_rate": 4.7151743509573444e-05,
+      "loss": 1.069,
+      "step": 1462
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.08783391863107681,
+      "learning_rate": 4.702740491959281e-05,
+      "loss": 1.067,
+      "step": 1463
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.08943789452314377,
+      "learning_rate": 4.690318006650377e-05,
+      "loss": 1.0054,
+      "step": 1464
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.08996465802192688,
+      "learning_rate": 4.677906921702816e-05,
+      "loss": 1.1619,
+      "step": 1465
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.08470418304204941,
+      "learning_rate": 4.665507263764299e-05,
+      "loss": 1.0453,
+      "step": 1466
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.08629050850868225,
+      "learning_rate": 4.653119059458007e-05,
+      "loss": 1.0335,
+      "step": 1467
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.08558977395296097,
+      "learning_rate": 4.6407423353825166e-05,
+      "loss": 0.9751,
+      "step": 1468
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.09226896613836288,
+      "learning_rate": 4.62837711811176e-05,
+      "loss": 1.0276,
+      "step": 1469
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.09151601791381836,
+      "learning_rate": 4.6160234341949646e-05,
+      "loss": 1.0749,
+      "step": 1470
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.10254024714231491,
+      "learning_rate": 4.6036813101565954e-05,
+      "loss": 1.0346,
+      "step": 1471
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.08552608639001846,
+      "learning_rate": 4.591350772496289e-05,
+      "loss": 1.0482,
+      "step": 1472
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.09262850880622864,
+      "learning_rate": 4.579031847688812e-05,
+      "loss": 1.0869,
+      "step": 1473
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.09947109967470169,
+      "learning_rate": 4.5667245621839974e-05,
+      "loss": 1.0823,
+      "step": 1474
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.09800399839878082,
+      "learning_rate": 4.554428942406684e-05,
+      "loss": 1.04,
+      "step": 1475
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.0967416986823082,
+      "learning_rate": 4.5421450147566694e-05,
+      "loss": 1.1243,
+      "step": 1476
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.10723847895860672,
+      "learning_rate": 4.529872805608644e-05,
+      "loss": 1.0438,
+      "step": 1477
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.09573391079902649,
+      "learning_rate": 4.5176123413121284e-05,
+      "loss": 1.075,
+      "step": 1478
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.10541462153196335,
+      "learning_rate": 4.505363648191443e-05,
+      "loss": 1.1675,
+      "step": 1479
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.10641545057296753,
+      "learning_rate": 4.493126752545619e-05,
+      "loss": 1.1008,
+      "step": 1480
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.092914879322052,
+      "learning_rate": 4.480901680648366e-05,
+      "loss": 1.0668,
+      "step": 1481
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.09610269963741302,
+      "learning_rate": 4.468688458748006e-05,
+      "loss": 0.9789,
+      "step": 1482
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.0949723944067955,
+      "learning_rate": 4.456487113067417e-05,
+      "loss": 1.0127,
+      "step": 1483
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.09450149536132812,
+      "learning_rate": 4.444297669803981e-05,
+      "loss": 1.0141,
+      "step": 1484
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.10026988387107849,
+      "learning_rate": 4.4321201551295135e-05,
+      "loss": 1.0976,
+      "step": 1485
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.09817048907279968,
+      "learning_rate": 4.4199545951902286e-05,
+      "loss": 1.0906,
+      "step": 1486
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.10230293869972229,
+      "learning_rate": 4.407801016106672e-05,
+      "loss": 1.0158,
+      "step": 1487
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.09846383333206177,
+      "learning_rate": 4.395659443973661e-05,
+      "loss": 0.9991,
+      "step": 1488
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.1006157398223877,
+      "learning_rate": 4.383529904860237e-05,
+      "loss": 1.0784,
+      "step": 1489
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.0998954325914383,
+      "learning_rate": 4.3714124248096067e-05,
+      "loss": 1.0527,
+      "step": 1490
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.1004033014178276,
+      "learning_rate": 4.3593070298390726e-05,
+      "loss": 0.9772,
+      "step": 1491
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.09377897530794144,
+      "learning_rate": 4.3472137459400084e-05,
+      "loss": 0.9641,
+      "step": 1492
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.10157407820224762,
+      "learning_rate": 4.335132599077766e-05,
+      "loss": 1.0356,
+      "step": 1493
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.09858424961566925,
+      "learning_rate": 4.3230636151916495e-05,
+      "loss": 1.0707,
+      "step": 1494
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.09798162430524826,
+      "learning_rate": 4.311006820194846e-05,
+      "loss": 1.1141,
+      "step": 1495
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.09100697189569473,
+      "learning_rate": 4.2989622399743714e-05,
+      "loss": 1.0671,
+      "step": 1496
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.09760253131389618,
+      "learning_rate": 4.2869299003910125e-05,
+      "loss": 1.0623,
+      "step": 1497
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.09561190009117126,
+      "learning_rate": 4.274909827279283e-05,
+      "loss": 0.9892,
+      "step": 1498
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.1004791259765625,
+      "learning_rate": 4.262902046447345e-05,
+      "loss": 1.0243,
+      "step": 1499
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.10229208320379257,
+      "learning_rate": 4.250906583676978e-05,
+      "loss": 1.0374,
+      "step": 1500
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.09850307554006577,
+      "learning_rate": 4.2389234647235145e-05,
+      "loss": 1.0424,
+      "step": 1501
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.09601115435361862,
+      "learning_rate": 4.226952715315779e-05,
+      "loss": 1.0787,
+      "step": 1502
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.09920283406972885,
+      "learning_rate": 4.2149943611560395e-05,
+      "loss": 0.9795,
+      "step": 1503
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.10054267197847366,
+      "learning_rate": 4.203048427919954e-05,
+      "loss": 1.0457,
+      "step": 1504
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.09651023894548416,
+      "learning_rate": 4.1911149412565056e-05,
+      "loss": 1.0578,
+      "step": 1505
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.0934935063123703,
+      "learning_rate": 4.17919392678795e-05,
+      "loss": 0.9551,
+      "step": 1506
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.09893841296434402,
+      "learning_rate": 4.1672854101097746e-05,
+      "loss": 1.0836,
+      "step": 1507
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.10160830616950989,
+      "learning_rate": 4.155389416790627e-05,
+      "loss": 1.1085,
+      "step": 1508
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.09854742884635925,
+      "learning_rate": 4.143505972372268e-05,
+      "loss": 1.0675,
+      "step": 1509
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.09768696129322052,
+      "learning_rate": 4.131635102369513e-05,
+      "loss": 1.0242,
+      "step": 1510
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.0980030745267868,
+      "learning_rate": 4.1197768322701846e-05,
+      "loss": 1.0789,
+      "step": 1511
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.10240290313959122,
+      "learning_rate": 4.10793118753504e-05,
+      "loss": 1.0962,
+      "step": 1512
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.10006707161664963,
+      "learning_rate": 4.096098193597738e-05,
+      "loss": 1.022,
+      "step": 1513
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.10069313645362854,
+      "learning_rate": 4.084277875864776e-05,
+      "loss": 1.1021,
+      "step": 1514
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.09589628130197525,
+      "learning_rate": 4.0724702597154296e-05,
+      "loss": 1.0483,
+      "step": 1515
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.09331901371479034,
+      "learning_rate": 4.0606753705017056e-05,
+      "loss": 0.9933,
+      "step": 1516
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.09794598817825317,
+      "learning_rate": 4.048893233548287e-05,
+      "loss": 1.0095,
+      "step": 1517
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.09837094694375992,
+      "learning_rate": 4.037123874152472e-05,
+      "loss": 1.0672,
+      "step": 1518
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.0900319516658783,
+      "learning_rate": 4.025367317584121e-05,
+      "loss": 1.0234,
+      "step": 1519
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.10217685997486115,
+      "learning_rate": 4.0136235890856155e-05,
+      "loss": 1.059,
+      "step": 1520
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.09992237389087677,
+      "learning_rate": 4.001892713871788e-05,
+      "loss": 1.0578,
+      "step": 1521
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.10689819604158401,
+      "learning_rate": 3.9901747171298766e-05,
+      "loss": 0.9973,
+      "step": 1522
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.10196186602115631,
+      "learning_rate": 3.978469624019464e-05,
+      "loss": 1.0871,
+      "step": 1523
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.09724418818950653,
+      "learning_rate": 3.966777459672437e-05,
+      "loss": 1.0746,
+      "step": 1524
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.10112694650888443,
+      "learning_rate": 3.955098249192907e-05,
+      "loss": 1.0705,
+      "step": 1525
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.09966527670621872,
+      "learning_rate": 3.943432017657186e-05,
+      "loss": 0.9555,
+      "step": 1526
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.102567657828331,
+      "learning_rate": 3.9317787901137146e-05,
+      "loss": 1.0423,
+      "step": 1527
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.10698642581701279,
+      "learning_rate": 3.920138591583015e-05,
+      "loss": 0.9851,
+      "step": 1528
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.09708695113658905,
+      "learning_rate": 3.908511447057629e-05,
+      "loss": 1.0606,
+      "step": 1529
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.09802483767271042,
+      "learning_rate": 3.8968973815020806e-05,
+      "loss": 0.9938,
+      "step": 1530
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.09706152975559235,
+      "learning_rate": 3.8852964198528e-05,
+      "loss": 1.0398,
+      "step": 1531
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.09986060112714767,
+      "learning_rate": 3.873708587018086e-05,
+      "loss": 0.9545,
+      "step": 1532
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.1054854691028595,
+      "learning_rate": 3.8621339078780524e-05,
+      "loss": 1.0501,
+      "step": 1533
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.09385277330875397,
+      "learning_rate": 3.850572407284569e-05,
+      "loss": 0.9677,
+      "step": 1534
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.10004087537527084,
+      "learning_rate": 3.839024110061212e-05,
+      "loss": 1.1073,
+      "step": 1535
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.09722103178501129,
+      "learning_rate": 3.8274890410032035e-05,
+      "loss": 1.0592,
+      "step": 1536
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.10227985680103302,
+      "learning_rate": 3.815967224877373e-05,
+      "loss": 1.0686,
+      "step": 1537
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.0962626114487648,
+      "learning_rate": 3.80445868642208e-05,
+      "loss": 0.9858,
+      "step": 1538
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.10224612802267075,
+      "learning_rate": 3.792963450347188e-05,
+      "loss": 1.048,
+      "step": 1539
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.10373633354902267,
+      "learning_rate": 3.7814815413339946e-05,
+      "loss": 0.9976,
+      "step": 1540
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.10008113086223602,
+      "learning_rate": 3.770012984035185e-05,
+      "loss": 0.996,
+      "step": 1541
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.10487159341573715,
+      "learning_rate": 3.7585578030747744e-05,
+      "loss": 1.0956,
+      "step": 1542
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.09948190301656723,
+      "learning_rate": 3.747116023048064e-05,
+      "loss": 1.115,
+      "step": 1543
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.10538452863693237,
+      "learning_rate": 3.7356876685215694e-05,
+      "loss": 1.1052,
+      "step": 1544
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.10581420361995697,
+      "learning_rate": 3.7242727640329954e-05,
+      "loss": 0.9876,
+      "step": 1545
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.09604854136705399,
+      "learning_rate": 3.7128713340911535e-05,
+      "loss": 0.9929,
+      "step": 1546
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.10109111666679382,
+      "learning_rate": 3.701483403175936e-05,
+      "loss": 1.0175,
+      "step": 1547
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.10949523001909256,
+      "learning_rate": 3.690108995738247e-05,
+      "loss": 1.052,
+      "step": 1548
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.10243970155715942,
+      "learning_rate": 3.678748136199956e-05,
+      "loss": 1.1381,
+      "step": 1549
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.09962951391935349,
+      "learning_rate": 3.667400848953845e-05,
+      "loss": 1.0583,
+      "step": 1550
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.10279103368520737,
+      "learning_rate": 3.6560671583635467e-05,
+      "loss": 1.0773,
+      "step": 1551
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.09922298043966293,
+      "learning_rate": 3.6447470887635096e-05,
+      "loss": 1.0402,
+      "step": 1552
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.10650550574064255,
+      "learning_rate": 3.633440664458935e-05,
+      "loss": 1.1341,
+      "step": 1553
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.09875325113534927,
+      "learning_rate": 3.622147909725724e-05,
+      "loss": 0.9874,
+      "step": 1554
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.1077369898557663,
+      "learning_rate": 3.6108688488104304e-05,
+      "loss": 1.0326,
+      "step": 1555
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.10243602097034454,
+      "learning_rate": 3.599603505930208e-05,
+      "loss": 1.0043,
+      "step": 1556
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.09883877635002136,
+      "learning_rate": 3.588351905272748e-05,
+      "loss": 1.0513,
+      "step": 1557
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.1076011061668396,
+      "learning_rate": 3.577114070996247e-05,
+      "loss": 1.0576,
+      "step": 1558
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.09689098596572876,
+      "learning_rate": 3.565890027229333e-05,
+      "loss": 1.0427,
+      "step": 1559
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.10068055987358093,
+      "learning_rate": 3.554679798071032e-05,
+      "loss": 1.0892,
+      "step": 1560
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.09878088533878326,
+      "learning_rate": 3.5434834075907096e-05,
+      "loss": 1.0769,
+      "step": 1561
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.10241078585386276,
+      "learning_rate": 3.532300879828013e-05,
+      "loss": 1.0827,
+      "step": 1562
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.1007351204752922,
+      "learning_rate": 3.5211322387928337e-05,
+      "loss": 1.0447,
+      "step": 1563
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.10430484265089035,
+      "learning_rate": 3.509977508465232e-05,
+      "loss": 1.0556,
+      "step": 1564
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.10023441165685654,
+      "learning_rate": 3.498836712795416e-05,
+      "loss": 1.0338,
+      "step": 1565
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.09625580161809921,
+      "learning_rate": 3.4877098757036665e-05,
+      "loss": 0.9924,
+      "step": 1566
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.09926914423704147,
+      "learning_rate": 3.476597021080297e-05,
+      "loss": 1.0404,
+      "step": 1567
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.10817915946245193,
+      "learning_rate": 3.4654981727855995e-05,
+      "loss": 1.0177,
+      "step": 1568
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.09937068074941635,
+      "learning_rate": 3.454413354649795e-05,
+      "loss": 1.0682,
+      "step": 1569
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.09850066900253296,
+      "learning_rate": 3.44334259047297e-05,
+      "loss": 1.094,
+      "step": 1570
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.1019996702671051,
+      "learning_rate": 3.432285904025049e-05,
+      "loss": 1.0583,
+      "step": 1571
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.09256354719400406,
+      "learning_rate": 3.421243319045727e-05,
+      "loss": 0.9321,
+      "step": 1572
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.09737757593393326,
+      "learning_rate": 3.410214859244414e-05,
+      "loss": 1.067,
+      "step": 1573
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.10051169991493225,
+      "learning_rate": 3.3992005483002e-05,
+      "loss": 1.1126,
+      "step": 1574
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.10207143425941467,
+      "learning_rate": 3.388200409861796e-05,
+      "loss": 0.9893,
+      "step": 1575
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.10268302261829376,
+      "learning_rate": 3.3772144675474835e-05,
+      "loss": 1.0936,
+      "step": 1576
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.10778101533651352,
+      "learning_rate": 3.366242744945056e-05,
+      "loss": 1.0589,
+      "step": 1577
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.10165120661258698,
+      "learning_rate": 3.355285265611784e-05,
+      "loss": 1.0399,
+      "step": 1578
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.10599108040332794,
+      "learning_rate": 3.344342053074354e-05,
+      "loss": 1.0713,
+      "step": 1579
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.108592189848423,
+      "learning_rate": 3.333413130828821e-05,
+      "loss": 1.0695,
+      "step": 1580
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.11459142714738846,
+      "learning_rate": 3.322498522340556e-05,
+      "loss": 1.163,
+      "step": 1581
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.09888488054275513,
+      "learning_rate": 3.3115982510442014e-05,
+      "loss": 1.0263,
+      "step": 1582
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.10739021003246307,
+      "learning_rate": 3.300712340343608e-05,
+      "loss": 0.9891,
+      "step": 1583
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.09790048748254776,
+      "learning_rate": 3.289840813611798e-05,
+      "loss": 1.0021,
+      "step": 1584
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.09580394625663757,
+      "learning_rate": 3.278983694190918e-05,
+      "loss": 1.0975,
+      "step": 1585
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.10121367871761322,
+      "learning_rate": 3.268141005392163e-05,
+      "loss": 1.034,
+      "step": 1586
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.09889926761388779,
+      "learning_rate": 3.257312770495757e-05,
+      "loss": 0.9765,
+      "step": 1587
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.10181820392608643,
+      "learning_rate": 3.24649901275089e-05,
+      "loss": 1.0281,
+      "step": 1588
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.09702855348587036,
+      "learning_rate": 3.235699755375666e-05,
+      "loss": 1.0771,
+      "step": 1589
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.1054118201136589,
+      "learning_rate": 3.224915021557049e-05,
+      "loss": 1.0678,
+      "step": 1590
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.10467803478240967,
+      "learning_rate": 3.21414483445083e-05,
+      "loss": 1.0878,
+      "step": 1591
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.10275258868932724,
+      "learning_rate": 3.2033892171815595e-05,
+      "loss": 1.0583,
+      "step": 1592
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.10269324481487274,
+      "learning_rate": 3.192648192842511e-05,
+      "loss": 1.0571,
+      "step": 1593
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.0981050431728363,
+      "learning_rate": 3.1819217844956214e-05,
+      "loss": 0.9605,
+      "step": 1594
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.0961243063211441,
+      "learning_rate": 3.171210015171452e-05,
+      "loss": 1.0872,
+      "step": 1595
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.10179060697555542,
+      "learning_rate": 3.1605129078691196e-05,
+      "loss": 1.0642,
+      "step": 1596
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.10174161940813065,
+      "learning_rate": 3.149830485556273e-05,
+      "loss": 1.1075,
+      "step": 1597
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.10251487046480179,
+      "learning_rate": 3.1391627711690296e-05,
+      "loss": 1.1247,
+      "step": 1598
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.09439977258443832,
+      "learning_rate": 3.128509787611917e-05,
+      "loss": 1.0369,
+      "step": 1599
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.10113759338855743,
+      "learning_rate": 3.1178715577578486e-05,
+      "loss": 1.061,
+      "step": 1600
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.10200539976358414,
+      "learning_rate": 3.107248104448052e-05,
+      "loss": 1.0871,
+      "step": 1601
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.10500266402959824,
+      "learning_rate": 3.0966394504920317e-05,
+      "loss": 1.0978,
+      "step": 1602
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.10838335007429123,
+      "learning_rate": 3.0860456186675104e-05,
+      "loss": 0.9892,
+      "step": 1603
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.09842187911272049,
+      "learning_rate": 3.075466631720394e-05,
+      "loss": 0.9836,
+      "step": 1604
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.09797590970993042,
+      "learning_rate": 3.0649025123647124e-05,
+      "loss": 0.9682,
+      "step": 1605
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.10447612404823303,
+      "learning_rate": 3.0543532832825715e-05,
+      "loss": 1.0135,
+      "step": 1606
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.11107984930276871,
+      "learning_rate": 3.043818967124109e-05,
+      "loss": 1.0436,
+      "step": 1607
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.10223361104726791,
+      "learning_rate": 3.0332995865074467e-05,
+      "loss": 1.0308,
+      "step": 1608
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.10475568473339081,
+      "learning_rate": 3.022795164018626e-05,
+      "loss": 1.0287,
+      "step": 1609
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.10647712647914886,
+      "learning_rate": 3.0123057222115836e-05,
+      "loss": 1.0913,
+      "step": 1610
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.10135121643543243,
+      "learning_rate": 3.0018312836080885e-05,
+      "loss": 1.0081,
+      "step": 1611
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.10377095639705658,
+      "learning_rate": 2.9913718706976968e-05,
+      "loss": 1.0498,
+      "step": 1612
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.1093953549861908,
+      "learning_rate": 2.9809275059376983e-05,
+      "loss": 1.0901,
+      "step": 1613
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.10123565793037415,
+      "learning_rate": 2.9704982117530777e-05,
+      "loss": 1.0527,
+      "step": 1614
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.09855428338050842,
+      "learning_rate": 2.9600840105364613e-05,
+      "loss": 1.0272,
+      "step": 1615
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.10368844121694565,
+      "learning_rate": 2.949684924648073e-05,
+      "loss": 1.1009,
+      "step": 1616
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.10128173232078552,
+      "learning_rate": 2.939300976415672e-05,
+      "loss": 1.0852,
+      "step": 1617
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.10072508454322815,
+      "learning_rate": 2.9289321881345254e-05,
+      "loss": 1.0419,
+      "step": 1618
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.09912049025297165,
+      "learning_rate": 2.918578582067347e-05,
+      "loss": 1.0564,
+      "step": 1619
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.10717380046844482,
+      "learning_rate": 2.908240180444255e-05,
+      "loss": 1.1012,
+      "step": 1620
+    },
+    {
+      "epoch": 2.22,
+      "eval_loss": 1.0938183069229126,
+      "eval_runtime": 96.9019,
+      "eval_samples_per_second": 51.599,
+      "eval_steps_per_second": 12.9,
+      "step": 1620
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.09515704959630966,
+      "learning_rate": 2.8979170054627225e-05,
+      "loss": 0.949,
+      "step": 1621
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.0976545512676239,
+      "learning_rate": 2.887609079287521e-05,
+      "loss": 1.0251,
+      "step": 1622
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.09744174033403397,
+      "learning_rate": 2.8773164240506934e-05,
+      "loss": 1.0421,
+      "step": 1623
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.10808070749044418,
+      "learning_rate": 2.8670390618514898e-05,
+      "loss": 1.0256,
+      "step": 1624
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.10681386291980743,
+      "learning_rate": 2.8567770147563255e-05,
+      "loss": 1.1412,
+      "step": 1625
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.10412570089101791,
+      "learning_rate": 2.8465303047987267e-05,
+      "loss": 1.0952,
+      "step": 1626
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.09856940805912018,
+      "learning_rate": 2.836298953979297e-05,
+      "loss": 1.0283,
+      "step": 1627
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.1043558195233345,
+      "learning_rate": 2.826082984265662e-05,
+      "loss": 1.0441,
+      "step": 1628
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.10388113558292389,
+      "learning_rate": 2.8158824175924225e-05,
+      "loss": 1.0257,
+      "step": 1629
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.10083478689193726,
+      "learning_rate": 2.805697275861101e-05,
+      "loss": 1.0298,
+      "step": 1630
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.10164723545312881,
+      "learning_rate": 2.7955275809401084e-05,
+      "loss": 1.0155,
+      "step": 1631
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.10648273676633835,
+      "learning_rate": 2.7853733546646866e-05,
+      "loss": 1.0383,
+      "step": 1632
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.10495666414499283,
+      "learning_rate": 2.7752346188368683e-05,
+      "loss": 1.109,
+      "step": 1633
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.10313493758440018,
+      "learning_rate": 2.765111395225424e-05,
+      "loss": 1.0275,
+      "step": 1634
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.1055515930056572,
+      "learning_rate": 2.7550037055658206e-05,
+      "loss": 1.0626,
+      "step": 1635
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.1033952459692955,
+      "learning_rate": 2.744911571560165e-05,
+      "loss": 0.9854,
+      "step": 1636
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.09736838936805725,
+      "learning_rate": 2.734835014877173e-05,
+      "loss": 1.0379,
+      "step": 1637
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.10089022666215897,
+      "learning_rate": 2.7247740571521118e-05,
+      "loss": 0.9855,
+      "step": 1638
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.10304795950651169,
+      "learning_rate": 2.714728719986759e-05,
+      "loss": 1.0937,
+      "step": 1639
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.09536028653383255,
+      "learning_rate": 2.7046990249493443e-05,
+      "loss": 0.9948,
+      "step": 1640
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.10237076878547668,
+      "learning_rate": 2.694684993574521e-05,
+      "loss": 1.0429,
+      "step": 1641
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.10333292186260223,
+      "learning_rate": 2.6846866473633125e-05,
+      "loss": 1.0919,
+      "step": 1642
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.10692143440246582,
+      "learning_rate": 2.6747040077830543e-05,
+      "loss": 1.1033,
+      "step": 1643
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.10679848492145538,
+      "learning_rate": 2.664737096267368e-05,
+      "loss": 1.0972,
+      "step": 1644
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.11273592710494995,
+      "learning_rate": 2.654785934216103e-05,
+      "loss": 1.0613,
+      "step": 1645
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.10312017053365707,
+      "learning_rate": 2.6448505429952917e-05,
+      "loss": 1.0305,
+      "step": 1646
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.10544013231992722,
+      "learning_rate": 2.6349309439371074e-05,
+      "loss": 1.1043,
+      "step": 1647
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.09748465567827225,
+      "learning_rate": 2.6250271583398166e-05,
+      "loss": 0.9812,
+      "step": 1648
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.106238454580307,
+      "learning_rate": 2.6151392074677274e-05,
+      "loss": 1.0556,
+      "step": 1649
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.10319123417139053,
+      "learning_rate": 2.605267112551154e-05,
+      "loss": 1.0464,
+      "step": 1650
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.10551314800977707,
+      "learning_rate": 2.5954108947863676e-05,
+      "loss": 1.0929,
+      "step": 1651
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.10372096300125122,
+      "learning_rate": 2.58557057533555e-05,
+      "loss": 1.0971,
+      "step": 1652
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.10320667177438736,
+      "learning_rate": 2.5757461753267388e-05,
+      "loss": 1.0405,
+      "step": 1653
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.09582614153623581,
+      "learning_rate": 2.5659377158538012e-05,
+      "loss": 0.9764,
+      "step": 1654
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.1082243099808693,
+      "learning_rate": 2.55614521797638e-05,
+      "loss": 1.0881,
+      "step": 1655
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.1074262261390686,
+      "learning_rate": 2.5463687027198356e-05,
+      "loss": 1.0416,
+      "step": 1656
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.10129725188016891,
+      "learning_rate": 2.5366081910752216e-05,
+      "loss": 1.042,
+      "step": 1657
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.10456307232379913,
+      "learning_rate": 2.5268637039992293e-05,
+      "loss": 1.0227,
+      "step": 1658
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.11030813306570053,
+      "learning_rate": 2.517135262414142e-05,
+      "loss": 1.0421,
+      "step": 1659
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.10014370828866959,
+      "learning_rate": 2.507422887207792e-05,
+      "loss": 1.0462,
+      "step": 1660
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.10228565335273743,
+      "learning_rate": 2.4977265992335184e-05,
+      "loss": 0.9831,
+      "step": 1661
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.10230918228626251,
+      "learning_rate": 2.488046419310114e-05,
+      "loss": 1.0371,
+      "step": 1662
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.1046641543507576,
+      "learning_rate": 2.47838236822179e-05,
+      "loss": 1.0389,
+      "step": 1663
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.1080273985862732,
+      "learning_rate": 2.4687344667181276e-05,
+      "loss": 0.9772,
+      "step": 1664
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.10280653834342957,
+      "learning_rate": 2.459102735514035e-05,
+      "loss": 1.105,
+      "step": 1665
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.09807223826646805,
+      "learning_rate": 2.4494871952896947e-05,
+      "loss": 1.0822,
+      "step": 1666
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.1035659983754158,
+      "learning_rate": 2.4398878666905334e-05,
+      "loss": 1.0347,
+      "step": 1667
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.10742058604955673,
+      "learning_rate": 2.4303047703271676e-05,
+      "loss": 1.0499,
+      "step": 1668
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.10122169554233551,
+      "learning_rate": 2.420737926775356e-05,
+      "loss": 1.0037,
+      "step": 1669
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.10576605051755905,
+      "learning_rate": 2.411187356575969e-05,
+      "loss": 1.0703,
+      "step": 1670
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.10527858138084412,
+      "learning_rate": 2.4016530802349345e-05,
+      "loss": 1.0792,
+      "step": 1671
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.109225332736969,
+      "learning_rate": 2.3921351182231932e-05,
+      "loss": 1.1268,
+      "step": 1672
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.10001665353775024,
+      "learning_rate": 2.3826334909766612e-05,
+      "loss": 0.9729,
+      "step": 1673
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.10480818897485733,
+      "learning_rate": 2.3731482188961818e-05,
+      "loss": 1.0545,
+      "step": 1674
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.10467101633548737,
+      "learning_rate": 2.3636793223474764e-05,
+      "loss": 1.0794,
+      "step": 1675
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.10887391865253448,
+      "learning_rate": 2.354226821661114e-05,
+      "loss": 1.0804,
+      "step": 1676
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.11070802062749863,
+      "learning_rate": 2.3447907371324572e-05,
+      "loss": 1.0831,
+      "step": 1677
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.09851560741662979,
+      "learning_rate": 2.335371089021623e-05,
+      "loss": 1.0583,
+      "step": 1678
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.10580728203058243,
+      "learning_rate": 2.3259678975534395e-05,
+      "loss": 1.0339,
+      "step": 1679
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.09998568892478943,
+      "learning_rate": 2.3165811829173923e-05,
+      "loss": 1.0827,
+      "step": 1680
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.10592679679393768,
+      "learning_rate": 2.307210965267601e-05,
+      "loss": 1.0697,
+      "step": 1681
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.10456671565771103,
+      "learning_rate": 2.297857264722756e-05,
+      "loss": 1.0762,
+      "step": 1682
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.10421790182590485,
+      "learning_rate": 2.2885201013660885e-05,
+      "loss": 1.0143,
+      "step": 1683
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.10303252190351486,
+      "learning_rate": 2.2791994952453223e-05,
+      "loss": 1.0281,
+      "step": 1684
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.10463783890008926,
+      "learning_rate": 2.26989546637263e-05,
+      "loss": 1.0449,
+      "step": 1685
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.10058246552944183,
+      "learning_rate": 2.260608034724595e-05,
+      "loss": 1.0591,
+      "step": 1686
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.10279016941785812,
+      "learning_rate": 2.2513372202421634e-05,
+      "loss": 0.9799,
+      "step": 1687
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.10460835695266724,
+      "learning_rate": 2.242083042830595e-05,
+      "loss": 1.0517,
+      "step": 1688
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.09564954787492752,
+      "learning_rate": 2.2328455223594402e-05,
+      "loss": 0.9831,
+      "step": 1689
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.10919241607189178,
+      "learning_rate": 2.2236246786624792e-05,
+      "loss": 1.0128,
+      "step": 1690
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.10170120745897293,
+      "learning_rate": 2.2144205315376866e-05,
+      "loss": 0.9818,
+      "step": 1691
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.09706525504589081,
+      "learning_rate": 2.2052331007471915e-05,
+      "loss": 1.0277,
+      "step": 1692
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.0990089401602745,
+      "learning_rate": 2.1960624060172208e-05,
+      "loss": 0.9989,
+      "step": 1693
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.10192267596721649,
+      "learning_rate": 2.1869084670380835e-05,
+      "loss": 1.0065,
+      "step": 1694
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.10314077138900757,
+      "learning_rate": 2.1777713034640956e-05,
+      "loss": 1.0964,
+      "step": 1695
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.09723629057407379,
+      "learning_rate": 2.1686509349135666e-05,
+      "loss": 1.0611,
+      "step": 1696
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.12461608648300171,
+      "learning_rate": 2.1595473809687437e-05,
+      "loss": 1.1006,
+      "step": 1697
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.10097838193178177,
+      "learning_rate": 2.150460661175768e-05,
+      "loss": 1.0402,
+      "step": 1698
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.10562681406736374,
+      "learning_rate": 2.1413907950446398e-05,
+      "loss": 1.0224,
+      "step": 1699
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.11095453053712845,
+      "learning_rate": 2.1323378020491736e-05,
+      "loss": 1.0098,
+      "step": 1700
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.09808256477117538,
+      "learning_rate": 2.1233017016269475e-05,
+      "loss": 1.0223,
+      "step": 1701
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.0997125506401062,
+      "learning_rate": 2.114282513179281e-05,
+      "loss": 0.9614,
+      "step": 1702
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.10099100321531296,
+      "learning_rate": 2.1052802560711737e-05,
+      "loss": 1.0061,
+      "step": 1703
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.10371831804513931,
+      "learning_rate": 2.096294949631278e-05,
+      "loss": 1.0301,
+      "step": 1704
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.09722629934549332,
+      "learning_rate": 2.0873266131518486e-05,
+      "loss": 0.9791,
+      "step": 1705
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.09912237524986267,
+      "learning_rate": 2.0783752658887066e-05,
+      "loss": 1.0268,
+      "step": 1706
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.10983093827962875,
+      "learning_rate": 2.0694409270611927e-05,
+      "loss": 1.0207,
+      "step": 1707
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.10646209865808487,
+      "learning_rate": 2.0605236158521256e-05,
+      "loss": 1.0364,
+      "step": 1708
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.10662958770990372,
+      "learning_rate": 2.051623351407772e-05,
+      "loss": 1.0191,
+      "step": 1709
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.10977033525705338,
+      "learning_rate": 2.0427401528377953e-05,
+      "loss": 1.0829,
+      "step": 1710
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.10580970346927643,
+      "learning_rate": 2.0338740392152146e-05,
+      "loss": 1.0717,
+      "step": 1711
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.10238678753376007,
+      "learning_rate": 2.0250250295763683e-05,
+      "loss": 1.0813,
+      "step": 1712
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.10537730902433395,
+      "learning_rate": 2.0161931429208737e-05,
+      "loss": 1.1302,
+      "step": 1713
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.10271787643432617,
+      "learning_rate": 2.0073783982115723e-05,
+      "loss": 0.9561,
+      "step": 1714
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.10018088668584824,
+      "learning_rate": 1.9985808143745133e-05,
+      "loss": 1.0742,
+      "step": 1715
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.10974530130624771,
+      "learning_rate": 1.9898004102988933e-05,
+      "loss": 1.0671,
+      "step": 1716
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.10251577943563461,
+      "learning_rate": 1.981037204837024e-05,
+      "loss": 0.9763,
+      "step": 1717
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.10576394200325012,
+      "learning_rate": 1.9722912168042897e-05,
+      "loss": 1.0521,
+      "step": 1718
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.10241660475730896,
+      "learning_rate": 1.9635624649791118e-05,
+      "loss": 1.0231,
+      "step": 1719
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.10159710049629211,
+      "learning_rate": 1.954850968102895e-05,
+      "loss": 1.0593,
+      "step": 1720
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.10333461314439774,
+      "learning_rate": 1.946156744880001e-05,
+      "loss": 1.0883,
+      "step": 1721
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.10093347728252411,
+      "learning_rate": 1.937479813977703e-05,
+      "loss": 1.0596,
+      "step": 1722
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.11626313626766205,
+      "learning_rate": 1.9288201940261485e-05,
+      "loss": 1.0937,
+      "step": 1723
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.10415246337652206,
+      "learning_rate": 1.9201779036183142e-05,
+      "loss": 1.0947,
+      "step": 1724
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.10219456255435944,
+      "learning_rate": 1.9115529613099702e-05,
+      "loss": 1.0337,
+      "step": 1725
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.10417490452528,
+      "learning_rate": 1.9029453856196376e-05,
+      "loss": 1.0306,
+      "step": 1726
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.10196002572774887,
+      "learning_rate": 1.8943551950285476e-05,
+      "loss": 1.0822,
+      "step": 1727
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.10779150575399399,
+      "learning_rate": 1.8857824079806086e-05,
+      "loss": 1.0773,
+      "step": 1728
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.10047119855880737,
+      "learning_rate": 1.877227042882358e-05,
+      "loss": 0.9733,
+      "step": 1729
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.10384653508663177,
+      "learning_rate": 1.868689118102931e-05,
+      "loss": 1.0503,
+      "step": 1730
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.1088796854019165,
+      "learning_rate": 1.8601686519740125e-05,
+      "loss": 1.1055,
+      "step": 1731
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.10231135785579681,
+      "learning_rate": 1.8516656627898078e-05,
+      "loss": 1.0271,
+      "step": 1732
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.10645679384469986,
+      "learning_rate": 1.8431801688069906e-05,
+      "loss": 1.0387,
+      "step": 1733
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.10481590777635574,
+      "learning_rate": 1.8347121882446717e-05,
+      "loss": 1.0175,
+      "step": 1734
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.11759142577648163,
+      "learning_rate": 1.826261739284365e-05,
+      "loss": 0.9371,
+      "step": 1735
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.1081736758351326,
+      "learning_rate": 1.8178288400699373e-05,
+      "loss": 1.1234,
+      "step": 1736
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.10870080441236496,
+      "learning_rate": 1.8094135087075782e-05,
+      "loss": 1.0893,
+      "step": 1737
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.10015043616294861,
+      "learning_rate": 1.8010157632657543e-05,
+      "loss": 1.0067,
+      "step": 1738
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.10474923253059387,
+      "learning_rate": 1.7926356217751772e-05,
+      "loss": 1.1385,
+      "step": 1739
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.10636482387781143,
+      "learning_rate": 1.784273102228754e-05,
+      "loss": 1.0852,
+      "step": 1740
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.10842975974082947,
+      "learning_rate": 1.775928222581564e-05,
+      "loss": 1.0417,
+      "step": 1741
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.09994855523109436,
+      "learning_rate": 1.7676010007508092e-05,
+      "loss": 1.0393,
+      "step": 1742
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.10009664297103882,
+      "learning_rate": 1.7592914546157768e-05,
+      "loss": 0.9942,
+      "step": 1743
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.10713919997215271,
+      "learning_rate": 1.7509996020178065e-05,
+      "loss": 1.1683,
+      "step": 1744
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.10401186347007751,
+      "learning_rate": 1.7427254607602472e-05,
+      "loss": 1.0572,
+      "step": 1745
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.09964354336261749,
+      "learning_rate": 1.7344690486084137e-05,
+      "loss": 1.0159,
+      "step": 1746
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.10515418648719788,
+      "learning_rate": 1.7262303832895666e-05,
+      "loss": 1.0135,
+      "step": 1747
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.10769481956958771,
+      "learning_rate": 1.7180094824928493e-05,
+      "loss": 1.1535,
+      "step": 1748
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.10604380071163177,
+      "learning_rate": 1.709806363869273e-05,
+      "loss": 1.05,
+      "step": 1749
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.10077527910470963,
+      "learning_rate": 1.701621045031666e-05,
+      "loss": 1.0095,
+      "step": 1750
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.1054917722940445,
+      "learning_rate": 1.6934535435546396e-05,
+      "loss": 1.1184,
+      "step": 1751
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.10609609633684158,
+      "learning_rate": 1.6853038769745467e-05,
+      "loss": 1.0737,
+      "step": 1752
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.111763596534729,
+      "learning_rate": 1.6771720627894538e-05,
+      "loss": 1.0675,
+      "step": 1753
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.11631076782941818,
+      "learning_rate": 1.6690581184590858e-05,
+      "loss": 1.0265,
+      "step": 1754
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.10097277164459229,
+      "learning_rate": 1.6609620614048084e-05,
+      "loss": 1.022,
+      "step": 1755
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.11081673949956894,
+      "learning_rate": 1.652883909009578e-05,
+      "loss": 1.1163,
+      "step": 1756
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.10136005282402039,
+      "learning_rate": 1.6448236786179107e-05,
+      "loss": 0.9825,
+      "step": 1757
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.10328476130962372,
+      "learning_rate": 1.636781387535843e-05,
+      "loss": 0.9773,
+      "step": 1758
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.10568514466285706,
+      "learning_rate": 1.628757053030887e-05,
+      "loss": 1.0143,
+      "step": 1759
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.10391801595687866,
+      "learning_rate": 1.6207506923320092e-05,
+      "loss": 0.9856,
+      "step": 1760
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.11014340072870255,
+      "learning_rate": 1.6127623226295773e-05,
+      "loss": 1.0941,
+      "step": 1761
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.10594548285007477,
+      "learning_rate": 1.604791961075336e-05,
+      "loss": 0.9972,
+      "step": 1762
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.10003345459699631,
+      "learning_rate": 1.5968396247823636e-05,
+      "loss": 0.9757,
+      "step": 1763
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.10709632933139801,
+      "learning_rate": 1.5889053308250368e-05,
+      "loss": 1.0431,
+      "step": 1764
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.10228748619556427,
+      "learning_rate": 1.5809890962389916e-05,
+      "loss": 1.0635,
+      "step": 1765
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.10298215597867966,
+      "learning_rate": 1.5730909380210945e-05,
+      "loss": 1.0033,
+      "step": 1766
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.10248055309057236,
+      "learning_rate": 1.5652108731293902e-05,
+      "loss": 0.9939,
+      "step": 1767
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.10340028256177902,
+      "learning_rate": 1.557348918483086e-05,
+      "loss": 0.947,
+      "step": 1768
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.10001228749752045,
+      "learning_rate": 1.5495050909624975e-05,
+      "loss": 0.9718,
+      "step": 1769
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.10652384907007217,
+      "learning_rate": 1.5416794074090258e-05,
+      "loss": 1.0975,
+      "step": 1770
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.10320189595222473,
+      "learning_rate": 1.533871884625111e-05,
+      "loss": 1.0403,
+      "step": 1771
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.11049370467662811,
+      "learning_rate": 1.526082539374203e-05,
+      "loss": 1.0683,
+      "step": 1772
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.10843852162361145,
+      "learning_rate": 1.5183113883807209e-05,
+      "loss": 1.0721,
+      "step": 1773
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.09895328432321548,
+      "learning_rate": 1.5105584483300162e-05,
+      "loss": 0.9994,
+      "step": 1774
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.10595796257257462,
+      "learning_rate": 1.5028237358683473e-05,
+      "loss": 1.0459,
+      "step": 1775
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.10822641104459763,
+      "learning_rate": 1.495107267602831e-05,
+      "loss": 1.0297,
+      "step": 1776
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.11243944615125656,
+      "learning_rate": 1.4874090601014135e-05,
+      "loss": 1.0415,
+      "step": 1777
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.09978371858596802,
+      "learning_rate": 1.479729129892835e-05,
+      "loss": 1.0404,
+      "step": 1778
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.1077616810798645,
+      "learning_rate": 1.4720674934665934e-05,
+      "loss": 1.0784,
+      "step": 1779
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.10674630850553513,
+      "learning_rate": 1.4644241672729008e-05,
+      "loss": 1.1027,
+      "step": 1780
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.10672060400247574,
+      "learning_rate": 1.4567991677226645e-05,
+      "loss": 1.025,
+      "step": 1781
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.10774379223585129,
+      "learning_rate": 1.4491925111874383e-05,
+      "loss": 1.0723,
+      "step": 1782
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.11169103533029556,
+      "learning_rate": 1.4416042139993923e-05,
+      "loss": 1.0912,
+      "step": 1783
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.1060202419757843,
+      "learning_rate": 1.4340342924512806e-05,
+      "loss": 1.0525,
+      "step": 1784
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.11215674132108688,
+      "learning_rate": 1.4264827627964006e-05,
+      "loss": 1.0881,
+      "step": 1785
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.10207799822092056,
+      "learning_rate": 1.4189496412485592e-05,
+      "loss": 1.0384,
+      "step": 1786
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.10249175131320953,
+      "learning_rate": 1.4114349439820396e-05,
+      "loss": 1.0276,
+      "step": 1787
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.11130068451166153,
+      "learning_rate": 1.4039386871315696e-05,
+      "loss": 1.0286,
+      "step": 1788
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.09945540875196457,
+      "learning_rate": 1.3964608867922813e-05,
+      "loss": 1.0028,
+      "step": 1789
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.10365299880504608,
+      "learning_rate": 1.3890015590196803e-05,
+      "loss": 1.0792,
+      "step": 1790
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.10548333078622818,
+      "learning_rate": 1.38156071982961e-05,
+      "loss": 0.9682,
+      "step": 1791
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.10640215873718262,
+      "learning_rate": 1.374138385198217e-05,
+      "loss": 1.0625,
+      "step": 1792
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.10771811008453369,
+      "learning_rate": 1.3667345710619128e-05,
+      "loss": 1.0784,
+      "step": 1793
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.10678724944591522,
+      "learning_rate": 1.3593492933173512e-05,
+      "loss": 1.0566,
+      "step": 1794
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.1097477376461029,
+      "learning_rate": 1.3519825678213815e-05,
+      "loss": 1.0878,
+      "step": 1795
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.10955114662647247,
+      "learning_rate": 1.3446344103910203e-05,
+      "loss": 1.0258,
+      "step": 1796
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.10596105456352234,
+      "learning_rate": 1.3373048368034203e-05,
+      "loss": 1.0732,
+      "step": 1797
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.11166518926620483,
+      "learning_rate": 1.3299938627958297e-05,
+      "loss": 1.1145,
+      "step": 1798
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.10216694325208664,
+      "learning_rate": 1.3227015040655621e-05,
+      "loss": 0.9794,
+      "step": 1799
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.10172343254089355,
+      "learning_rate": 1.3154277762699607e-05,
+      "loss": 1.0574,
+      "step": 1800
+    },
+    {
+      "epoch": 2.47,
+      "eval_loss": 1.0942561626434326,
+      "eval_runtime": 96.9933,
+      "eval_samples_per_second": 51.55,
+      "eval_steps_per_second": 12.887,
+      "step": 1800
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.10276013612747192,
+      "learning_rate": 1.3081726950263707e-05,
+      "loss": 1.1027,
+      "step": 1801
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.10411372780799866,
+      "learning_rate": 1.300936275912098e-05,
+      "loss": 1.0387,
+      "step": 1802
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.10722065716981888,
+      "learning_rate": 1.293718534464381e-05,
+      "loss": 1.0912,
+      "step": 1803
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.10419046878814697,
+      "learning_rate": 1.2865194861803564e-05,
+      "loss": 1.0129,
+      "step": 1804
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.10487858951091766,
+      "learning_rate": 1.2793391465170246e-05,
+      "loss": 0.9972,
+      "step": 1805
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.10990484803915024,
+      "learning_rate": 1.2721775308912132e-05,
+      "loss": 1.0202,
+      "step": 1806
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.10575643181800842,
+      "learning_rate": 1.26503465467955e-05,
+      "loss": 1.0184,
+      "step": 1807
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.10020840167999268,
+      "learning_rate": 1.2579105332184304e-05,
+      "loss": 1.0572,
+      "step": 1808
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.10312289744615555,
+      "learning_rate": 1.250805181803978e-05,
+      "loss": 1.0754,
+      "step": 1809
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.10634654015302658,
+      "learning_rate": 1.2437186156920167e-05,
+      "loss": 1.0571,
+      "step": 1810
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.10992537438869476,
+      "learning_rate": 1.2366508500980389e-05,
+      "loss": 1.0793,
+      "step": 1811
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.1019081249833107,
+      "learning_rate": 1.229601900197166e-05,
+      "loss": 0.9911,
+      "step": 1812
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.1067473292350769,
+      "learning_rate": 1.2225717811241243e-05,
+      "loss": 1.0698,
+      "step": 1813
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.10666471719741821,
+      "learning_rate": 1.2155605079732046e-05,
+      "loss": 1.1006,
+      "step": 1814
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.1086270660161972,
+      "learning_rate": 1.208568095798237e-05,
+      "loss": 1.1359,
+      "step": 1815
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.10846429318189621,
+      "learning_rate": 1.2015945596125567e-05,
+      "loss": 1.0907,
+      "step": 1816
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.1090848371386528,
+      "learning_rate": 1.1946399143889687e-05,
+      "loss": 1.1315,
+      "step": 1817
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.10526415705680847,
+      "learning_rate": 1.1877041750597173e-05,
+      "loss": 1.0716,
+      "step": 1818
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.10600166767835617,
+      "learning_rate": 1.1807873565164506e-05,
+      "loss": 1.044,
+      "step": 1819
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.11075668781995773,
+      "learning_rate": 1.1738894736101991e-05,
+      "loss": 1.1191,
+      "step": 1820
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.10121472924947739,
+      "learning_rate": 1.1670105411513321e-05,
+      "loss": 1.0161,
+      "step": 1821
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.10279441624879837,
+      "learning_rate": 1.160150573909532e-05,
+      "loss": 1.0316,
+      "step": 1822
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.10803600400686264,
+      "learning_rate": 1.153309586613761e-05,
+      "loss": 1.0707,
+      "step": 1823
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.10342013835906982,
+      "learning_rate": 1.1464875939522312e-05,
+      "loss": 1.0384,
+      "step": 1824
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.10714070498943329,
+      "learning_rate": 1.1396846105723669e-05,
+      "loss": 0.9685,
+      "step": 1825
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.10151127725839615,
+      "learning_rate": 1.132900651080785e-05,
+      "loss": 1.0001,
+      "step": 1826
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.10048438608646393,
+      "learning_rate": 1.1261357300432474e-05,
+      "loss": 1.0087,
+      "step": 1827
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.09962281584739685,
+      "learning_rate": 1.1193898619846465e-05,
+      "loss": 1.0417,
+      "step": 1828
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.10164935141801834,
+      "learning_rate": 1.1126630613889622e-05,
+      "loss": 1.0189,
+      "step": 1829
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.1068333312869072,
+      "learning_rate": 1.1059553426992365e-05,
+      "loss": 1.0857,
+      "step": 1830
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.10585252195596695,
+      "learning_rate": 1.0992667203175422e-05,
+      "loss": 1.0171,
+      "step": 1831
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.10877200961112976,
+      "learning_rate": 1.0925972086049452e-05,
+      "loss": 1.0372,
+      "step": 1832
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.1089455857872963,
+      "learning_rate": 1.085946821881485e-05,
+      "loss": 1.0978,
+      "step": 1833
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.10706791281700134,
+      "learning_rate": 1.0793155744261351e-05,
+      "loss": 1.0645,
+      "step": 1834
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.10820597410202026,
+      "learning_rate": 1.0727034804767755e-05,
+      "loss": 1.0319,
+      "step": 1835
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.09909602999687195,
+      "learning_rate": 1.0661105542301642e-05,
+      "loss": 0.8632,
+      "step": 1836
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.10533516108989716,
+      "learning_rate": 1.0595368098419022e-05,
+      "loss": 1.1273,
+      "step": 1837
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.10292701423168182,
+      "learning_rate": 1.0529822614264018e-05,
+      "loss": 1.0939,
+      "step": 1838
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.1019676923751831,
+      "learning_rate": 1.0464469230568675e-05,
+      "loss": 1.0197,
+      "step": 1839
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.10791152715682983,
+      "learning_rate": 1.039930808765256e-05,
+      "loss": 0.9698,
+      "step": 1840
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.10098548978567123,
+      "learning_rate": 1.0334339325422405e-05,
+      "loss": 0.9904,
+      "step": 1841
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.11264815926551819,
+      "learning_rate": 1.026956308337199e-05,
+      "loss": 1.0869,
+      "step": 1842
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.10908129811286926,
+      "learning_rate": 1.0204979500581668e-05,
+      "loss": 1.0789,
+      "step": 1843
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.11052219569683075,
+      "learning_rate": 1.0140588715718191e-05,
+      "loss": 1.0466,
+      "step": 1844
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.11108071357011795,
+      "learning_rate": 1.007639086703428e-05,
+      "loss": 0.9978,
+      "step": 1845
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.10201063007116318,
+      "learning_rate": 1.0012386092368475e-05,
+      "loss": 1.0463,
+      "step": 1846
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.10083018988370895,
+      "learning_rate": 9.948574529144727e-06,
+      "loss": 1.0405,
+      "step": 1847
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.10655052959918976,
+      "learning_rate": 9.884956314372174e-06,
+      "loss": 0.9986,
+      "step": 1848
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.10638248175382614,
+      "learning_rate": 9.821531584644794e-06,
+      "loss": 1.0571,
+      "step": 1849
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.10480670630931854,
+      "learning_rate": 9.75830047614117e-06,
+      "loss": 0.9988,
+      "step": 1850
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.10650740563869476,
+      "learning_rate": 9.695263124624099e-06,
+      "loss": 1.0778,
+      "step": 1851
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.1115521788597107,
+      "learning_rate": 9.632419665440428e-06,
+      "loss": 1.0638,
+      "step": 1852
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.11067399382591248,
+      "learning_rate": 9.56977023352068e-06,
+      "loss": 1.072,
+      "step": 1853
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.10589943826198578,
+      "learning_rate": 9.507314963378745e-06,
+      "loss": 0.991,
+      "step": 1854
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.10115532577037811,
+      "learning_rate": 9.445053989111696e-06,
+      "loss": 0.9984,
+      "step": 1855
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.11026972532272339,
+      "learning_rate": 9.382987444399394e-06,
+      "loss": 1.0212,
+      "step": 1856
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.1195150837302208,
+      "learning_rate": 9.321115462504271e-06,
+      "loss": 1.0621,
+      "step": 1857
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.10526645928621292,
+      "learning_rate": 9.259438176270962e-06,
+      "loss": 1.1279,
+      "step": 1858
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.0987568348646164,
+      "learning_rate": 9.197955718126128e-06,
+      "loss": 0.964,
+      "step": 1859
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.10304315388202667,
+      "learning_rate": 9.136668220078094e-06,
+      "loss": 1.0034,
+      "step": 1860
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.10379643738269806,
+      "learning_rate": 9.075575813716585e-06,
+      "loss": 1.0065,
+      "step": 1861
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.10735524445772171,
+      "learning_rate": 9.014678630212469e-06,
+      "loss": 0.9876,
+      "step": 1862
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.10403770208358765,
+      "learning_rate": 8.953976800317432e-06,
+      "loss": 1.0212,
+      "step": 1863
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.11137989163398743,
+      "learning_rate": 8.893470454363706e-06,
+      "loss": 1.105,
+      "step": 1864
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.11065743118524551,
+      "learning_rate": 8.833159722263829e-06,
+      "loss": 1.0291,
+      "step": 1865
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.1015489399433136,
+      "learning_rate": 8.773044733510338e-06,
+      "loss": 1.049,
+      "step": 1866
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.10459042340517044,
+      "learning_rate": 8.713125617175454e-06,
+      "loss": 1.0266,
+      "step": 1867
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.10275865346193314,
+      "learning_rate": 8.65340250191089e-06,
+      "loss": 1.0305,
+      "step": 1868
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.10716346651315689,
+      "learning_rate": 8.593875515947513e-06,
+      "loss": 0.9619,
+      "step": 1869
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.10621901601552963,
+      "learning_rate": 8.534544787095078e-06,
+      "loss": 1.0288,
+      "step": 1870
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.10612958669662476,
+      "learning_rate": 8.475410442741927e-06,
+      "loss": 0.9937,
+      "step": 1871
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.1043461263179779,
+      "learning_rate": 8.416472609854808e-06,
+      "loss": 1.0225,
+      "step": 1872
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.1087508499622345,
+      "learning_rate": 8.357731414978498e-06,
+      "loss": 0.985,
+      "step": 1873
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.10143236070871353,
+      "learning_rate": 8.299186984235585e-06,
+      "loss": 1.0092,
+      "step": 1874
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.10356243699789047,
+      "learning_rate": 8.240839443326198e-06,
+      "loss": 1.0168,
+      "step": 1875
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.12115638703107834,
+      "learning_rate": 8.182688917527726e-06,
+      "loss": 1.0554,
+      "step": 1876
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.10988619178533554,
+      "learning_rate": 8.124735531694517e-06,
+      "loss": 1.0756,
+      "step": 1877
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.10726433247327805,
+      "learning_rate": 8.06697941025768e-06,
+      "loss": 1.0353,
+      "step": 1878
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.10422394424676895,
+      "learning_rate": 8.009420677224755e-06,
+      "loss": 0.9666,
+      "step": 1879
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.1098356544971466,
+      "learning_rate": 7.952059456179507e-06,
+      "loss": 1.1157,
+      "step": 1880
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.1013469323515892,
+      "learning_rate": 7.89489587028156e-06,
+      "loss": 1.0883,
+      "step": 1881
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.10621859133243561,
+      "learning_rate": 7.837930042266262e-06,
+      "loss": 1.0191,
+      "step": 1882
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.1011110171675682,
+      "learning_rate": 7.781162094444328e-06,
+      "loss": 1.0573,
+      "step": 1883
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.09822903573513031,
+      "learning_rate": 7.724592148701615e-06,
+      "loss": 0.9579,
+      "step": 1884
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.11107337474822998,
+      "learning_rate": 7.668220326498831e-06,
+      "loss": 1.0309,
+      "step": 1885
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.10874823480844498,
+      "learning_rate": 7.612046748871327e-06,
+      "loss": 1.0779,
+      "step": 1886
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.1034318134188652,
+      "learning_rate": 7.556071536428788e-06,
+      "loss": 1.0626,
+      "step": 1887
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.10825161635875702,
+      "learning_rate": 7.50029480935498e-06,
+      "loss": 0.9903,
+      "step": 1888
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.10360877960920334,
+      "learning_rate": 7.44471668740756e-06,
+      "loss": 1.0593,
+      "step": 1889
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.10969631373882294,
+      "learning_rate": 7.389337289917652e-06,
+      "loss": 1.0776,
+      "step": 1890
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.09470348805189133,
+      "learning_rate": 7.33415673578981e-06,
+      "loss": 1.0452,
+      "step": 1891
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.10235817730426788,
+      "learning_rate": 7.279175143501604e-06,
+      "loss": 1.0492,
+      "step": 1892
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.10740025341510773,
+      "learning_rate": 7.224392631103438e-06,
+      "loss": 0.9625,
+      "step": 1893
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.10129448026418686,
+      "learning_rate": 7.1698093162182125e-06,
+      "loss": 1.0315,
+      "step": 1894
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.10872422158718109,
+      "learning_rate": 7.115425316041213e-06,
+      "loss": 1.0688,
+      "step": 1895
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.10519392788410187,
+      "learning_rate": 7.061240747339737e-06,
+      "loss": 0.8834,
+      "step": 1896
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.1014443039894104,
+      "learning_rate": 7.0072557264529106e-06,
+      "loss": 1.0605,
+      "step": 1897
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.10586616396903992,
+      "learning_rate": 6.953470369291348e-06,
+      "loss": 1.1053,
+      "step": 1898
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.10356176644563675,
+      "learning_rate": 6.899884791337041e-06,
+      "loss": 1.0357,
+      "step": 1899
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.11087659746408463,
+      "learning_rate": 6.846499107643012e-06,
+      "loss": 1.0281,
+      "step": 1900
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.10022734850645065,
+      "learning_rate": 6.793313432833093e-06,
+      "loss": 0.9744,
+      "step": 1901
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.10357245057821274,
+      "learning_rate": 6.74032788110166e-06,
+      "loss": 0.9348,
+      "step": 1902
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.11496304720640182,
+      "learning_rate": 6.687542566213456e-06,
+      "loss": 1.1014,
+      "step": 1903
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.1084774062037468,
+      "learning_rate": 6.634957601503233e-06,
+      "loss": 1.0714,
+      "step": 1904
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.10514631867408752,
+      "learning_rate": 6.58257309987561e-06,
+      "loss": 0.988,
+      "step": 1905
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.10369277000427246,
+      "learning_rate": 6.530389173804807e-06,
+      "loss": 1.0955,
+      "step": 1906
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.0993787869811058,
+      "learning_rate": 6.478405935334364e-06,
+      "loss": 1.0161,
+      "step": 1907
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.1040748804807663,
+      "learning_rate": 6.4266234960769226e-06,
+      "loss": 1.0028,
+      "step": 1908
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.10609594732522964,
+      "learning_rate": 6.3750419672140235e-06,
+      "loss": 1.0572,
+      "step": 1909
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.10241368412971497,
+      "learning_rate": 6.323661459495811e-06,
+      "loss": 1.0838,
+      "step": 1910
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.10709729045629501,
+      "learning_rate": 6.272482083240805e-06,
+      "loss": 0.9935,
+      "step": 1911
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.1096770241856575,
+      "learning_rate": 6.221503948335705e-06,
+      "loss": 1.0257,
+      "step": 1912
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.1070781946182251,
+      "learning_rate": 6.1707271642351346e-06,
+      "loss": 1.0326,
+      "step": 1913
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.10433980822563171,
+      "learning_rate": 6.1201518399613635e-06,
+      "loss": 1.0826,
+      "step": 1914
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.10418420284986496,
+      "learning_rate": 6.069778084104139e-06,
+      "loss": 1.0385,
+      "step": 1915
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.10366898775100708,
+      "learning_rate": 6.019606004820422e-06,
+      "loss": 1.0292,
+      "step": 1916
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.10579089820384979,
+      "learning_rate": 5.96963570983412e-06,
+      "loss": 1.0832,
+      "step": 1917
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.11261386424303055,
+      "learning_rate": 5.919867306435934e-06,
+      "loss": 1.0616,
+      "step": 1918
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.10631542652845383,
+      "learning_rate": 5.870300901483061e-06,
+      "loss": 1.0443,
+      "step": 1919
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.10314253717660904,
+      "learning_rate": 5.820936601399029e-06,
+      "loss": 1.0152,
+      "step": 1920
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.10799778252840042,
+      "learning_rate": 5.771774512173356e-06,
+      "loss": 0.9673,
+      "step": 1921
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.10690139979124069,
+      "learning_rate": 5.722814739361459e-06,
+      "loss": 1.0681,
+      "step": 1922
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.10518421232700348,
+      "learning_rate": 5.674057388084364e-06,
+      "loss": 1.0419,
+      "step": 1923
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.10788912326097488,
+      "learning_rate": 5.625502563028407e-06,
+      "loss": 1.0525,
+      "step": 1924
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.11075406521558762,
+      "learning_rate": 5.577150368445161e-06,
+      "loss": 1.0889,
+      "step": 1925
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.10486439615488052,
+      "learning_rate": 5.529000908151105e-06,
+      "loss": 1.0081,
+      "step": 1926
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.14317287504673004,
+      "learning_rate": 5.48105428552742e-06,
+      "loss": 1.0659,
+      "step": 1927
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.10721174627542496,
+      "learning_rate": 5.4333106035198035e-06,
+      "loss": 0.9646,
+      "step": 1928
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.11699029058218002,
+      "learning_rate": 5.385769964638199e-06,
+      "loss": 1.0394,
+      "step": 1929
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.10632190853357315,
+      "learning_rate": 5.338432470956589e-06,
+      "loss": 1.0635,
+      "step": 1930
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.10536311566829681,
+      "learning_rate": 5.291298224112806e-06,
+      "loss": 0.9977,
+      "step": 1931
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.1122971922159195,
+      "learning_rate": 5.244367325308286e-06,
+      "loss": 1.0825,
+      "step": 1932
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.11183328926563263,
+      "learning_rate": 5.19763987530788e-06,
+      "loss": 0.9779,
+      "step": 1933
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.10649974644184113,
+      "learning_rate": 5.151115974439569e-06,
+      "loss": 1.0045,
+      "step": 1934
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.1091967299580574,
+      "learning_rate": 5.104795722594324e-06,
+      "loss": 1.045,
+      "step": 1935
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.1042511835694313,
+      "learning_rate": 5.058679219225881e-06,
+      "loss": 0.9877,
+      "step": 1936
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.10467984527349472,
+      "learning_rate": 5.012766563350457e-06,
+      "loss": 1.0256,
+      "step": 1937
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.11423452943563461,
+      "learning_rate": 4.967057853546653e-06,
+      "loss": 1.086,
+      "step": 1938
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.10454947501420975,
+      "learning_rate": 4.921553187955142e-06,
+      "loss": 0.9912,
+      "step": 1939
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.10362736135721207,
+      "learning_rate": 4.876252664278502e-06,
+      "loss": 1.0706,
+      "step": 1940
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.10764990746974945,
+      "learning_rate": 4.83115637978101e-06,
+      "loss": 1.0801,
+      "step": 1941
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.10532201081514359,
+      "learning_rate": 4.786264431288423e-06,
+      "loss": 0.9978,
+      "step": 1942
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.09887377172708511,
+      "learning_rate": 4.741576915187729e-06,
+      "loss": 0.9894,
+      "step": 1943
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.1074475422501564,
+      "learning_rate": 4.697093927427032e-06,
+      "loss": 1.0981,
+      "step": 1944
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.10976172238588333,
+      "learning_rate": 4.6528155635152895e-06,
+      "loss": 1.1017,
+      "step": 1945
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.10972798615694046,
+      "learning_rate": 4.608741918522097e-06,
+      "loss": 0.9771,
+      "step": 1946
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.10937816649675369,
+      "learning_rate": 4.564873087077515e-06,
+      "loss": 1.1545,
+      "step": 1947
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.10842366516590118,
+      "learning_rate": 4.521209163371809e-06,
+      "loss": 1.0523,
+      "step": 1948
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.10242139548063278,
+      "learning_rate": 4.477750241155365e-06,
+      "loss": 1.0496,
+      "step": 1949
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.10286231338977814,
+      "learning_rate": 4.434496413738332e-06,
+      "loss": 1.0318,
+      "step": 1950
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.11701862514019012,
+      "learning_rate": 4.391447773990554e-06,
+      "loss": 1.1051,
+      "step": 1951
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.09992954134941101,
+      "learning_rate": 4.348604414341306e-06,
+      "loss": 1.0207,
+      "step": 1952
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.104942686855793,
+      "learning_rate": 4.305966426779118e-06,
+      "loss": 1.0615,
+      "step": 1953
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.10691747814416885,
+      "learning_rate": 4.263533902851535e-06,
+      "loss": 1.0697,
+      "step": 1954
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.1073504313826561,
+      "learning_rate": 4.221306933664992e-06,
+      "loss": 1.0303,
+      "step": 1955
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.10924269258975983,
+      "learning_rate": 4.179285609884554e-06,
+      "loss": 1.0619,
+      "step": 1956
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.11341916024684906,
+      "learning_rate": 4.137470021733747e-06,
+      "loss": 1.0762,
+      "step": 1957
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.10274948179721832,
+      "learning_rate": 4.095860258994388e-06,
+      "loss": 1.005,
+      "step": 1958
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.10694161802530289,
+      "learning_rate": 4.054456411006335e-06,
+      "loss": 1.0104,
+      "step": 1959
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.10656246542930603,
+      "learning_rate": 4.013258566667388e-06,
+      "loss": 1.0581,
+      "step": 1960
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.10152103751897812,
+      "learning_rate": 3.972266814432946e-06,
+      "loss": 1.0847,
+      "step": 1961
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.10767954587936401,
+      "learning_rate": 3.931481242315993e-06,
+      "loss": 0.9821,
+      "step": 1962
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.11468133330345154,
+      "learning_rate": 3.890901937886771e-06,
+      "loss": 1.0606,
+      "step": 1963
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.10704988241195679,
+      "learning_rate": 3.850528988272684e-06,
+      "loss": 1.0763,
+      "step": 1964
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.10794585198163986,
+      "learning_rate": 3.8103624801580583e-06,
+      "loss": 1.1199,
+      "step": 1965
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.1016378179192543,
+      "learning_rate": 3.770402499783976e-06,
+      "loss": 1.042,
+      "step": 1966
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.10717840492725372,
+      "learning_rate": 3.730649132948083e-06,
+      "loss": 1.1171,
+      "step": 1967
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.11034273356199265,
+      "learning_rate": 3.691102465004415e-06,
+      "loss": 0.9966,
+      "step": 1968
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.10772895067930222,
+      "learning_rate": 3.6517625808631716e-06,
+      "loss": 1.0718,
+      "step": 1969
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.10465323179960251,
+      "learning_rate": 3.6126295649906216e-06,
+      "loss": 1.027,
+      "step": 1970
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.10131707042455673,
+      "learning_rate": 3.5737035014088315e-06,
+      "loss": 0.9879,
+      "step": 1971
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.11240613460540771,
+      "learning_rate": 3.534984473695535e-06,
+      "loss": 0.9972,
+      "step": 1972
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.09870439022779465,
+      "learning_rate": 3.496472564983955e-06,
+      "loss": 1.0783,
+      "step": 1973
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.10683514922857285,
+      "learning_rate": 3.458167857962613e-06,
+      "loss": 1.0772,
+      "step": 1974
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.10112517327070236,
+      "learning_rate": 3.4200704348751e-06,
+      "loss": 1.0094,
+      "step": 1975
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.11188273131847382,
+      "learning_rate": 3.3821803775199832e-06,
+      "loss": 1.0631,
+      "step": 1976
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.11307306587696075,
+      "learning_rate": 3.3444977672506205e-06,
+      "loss": 1.0625,
+      "step": 1977
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.11079946905374527,
+      "learning_rate": 3.3070226849749366e-06,
+      "loss": 1.0223,
+      "step": 1978
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.11021959781646729,
+      "learning_rate": 3.2697552111552788e-06,
+      "loss": 1.1104,
+      "step": 1979
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.10750070959329605,
+      "learning_rate": 3.2326954258082408e-06,
+      "loss": 1.033,
+      "step": 1980
+    },
+    {
+      "epoch": 2.72,
+      "eval_loss": 1.0945501327514648,
+      "eval_runtime": 97.0973,
+      "eval_samples_per_second": 51.495,
+      "eval_steps_per_second": 12.874,
+      "step": 1980
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.10753518342971802,
+      "learning_rate": 3.1958434085044952e-06,
+      "loss": 1.0447,
+      "step": 1981
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.10855768620967865,
+      "learning_rate": 3.159199238368593e-06,
+      "loss": 1.0569,
+      "step": 1982
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.10650381445884705,
+      "learning_rate": 3.122762994078854e-06,
+      "loss": 1.03,
+      "step": 1983
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.10415911674499512,
+      "learning_rate": 3.086534753867154e-06,
+      "loss": 1.0624,
+      "step": 1984
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.10523323714733124,
+      "learning_rate": 3.0505145955187496e-06,
+      "loss": 1.0409,
+      "step": 1985
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.10439461469650269,
+      "learning_rate": 3.0147025963721433e-06,
+      "loss": 1.0045,
+      "step": 1986
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.10498375445604324,
+      "learning_rate": 2.9790988333189052e-06,
+      "loss": 1.0429,
+      "step": 1987
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.11714166402816772,
+      "learning_rate": 2.9437033828034866e-06,
+      "loss": 1.0077,
+      "step": 1988
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.10382110625505447,
+      "learning_rate": 2.9085163208230514e-06,
+      "loss": 1.0324,
+      "step": 1989
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.1161193996667862,
+      "learning_rate": 2.8735377229273998e-06,
+      "loss": 1.112,
+      "step": 1990
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.10029222071170807,
+      "learning_rate": 2.838767664218689e-06,
+      "loss": 0.96,
+      "step": 1991
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.10440904647111893,
+      "learning_rate": 2.8042062193513353e-06,
+      "loss": 1.0381,
+      "step": 1992
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.10433690249919891,
+      "learning_rate": 2.7698534625318574e-06,
+      "loss": 1.0061,
+      "step": 1993
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.10676665604114532,
+      "learning_rate": 2.735709467518699e-06,
+      "loss": 1.0403,
+      "step": 1994
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.10396602749824524,
+      "learning_rate": 2.7017743076220515e-06,
+      "loss": 1.0634,
+      "step": 1995
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.10540228337049484,
+      "learning_rate": 2.668048055703731e-06,
+      "loss": 0.996,
+      "step": 1996
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.10654503107070923,
+      "learning_rate": 2.634530784177014e-06,
+      "loss": 1.003,
+      "step": 1997
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.10922779887914658,
+      "learning_rate": 2.6012225650064893e-06,
+      "loss": 1.0265,
+      "step": 1998
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.1074921190738678,
+      "learning_rate": 2.5681234697078616e-06,
+      "loss": 1.0352,
+      "step": 1999
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.099807508289814,
+      "learning_rate": 2.535233569347861e-06,
+      "loss": 0.9606,
+      "step": 2000
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.11240819096565247,
+      "learning_rate": 2.5025529345440333e-06,
+      "loss": 1.0062,
+      "step": 2001
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.10550540685653687,
+      "learning_rate": 2.470081635464594e-06,
+      "loss": 1.1057,
+      "step": 2002
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.10838490724563599,
+      "learning_rate": 2.437819741828351e-06,
+      "loss": 1.1138,
+      "step": 2003
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.1086932122707367,
+      "learning_rate": 2.405767322904462e-06,
+      "loss": 0.9938,
+      "step": 2004
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.11865808814764023,
+      "learning_rate": 2.3739244475123322e-06,
+      "loss": 1.0886,
+      "step": 2005
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.10906819254159927,
+      "learning_rate": 2.342291184021461e-06,
+      "loss": 1.0745,
+      "step": 2006
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.10215233266353607,
+      "learning_rate": 2.3108676003512963e-06,
+      "loss": 1.0591,
+      "step": 2007
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.11253557354211807,
+      "learning_rate": 2.279653763971057e-06,
+      "loss": 1.1196,
+      "step": 2008
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.10783706605434418,
+      "learning_rate": 2.2486497418996556e-06,
+      "loss": 1.1494,
+      "step": 2009
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.11097768694162369,
+      "learning_rate": 2.2178556007054872e-06,
+      "loss": 1.0627,
+      "step": 2010
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.1037830114364624,
+      "learning_rate": 2.1872714065063195e-06,
+      "loss": 0.9815,
+      "step": 2011
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.1082891896367073,
+      "learning_rate": 2.1568972249691566e-06,
+      "loss": 1.0154,
+      "step": 2012
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.10485997051000595,
+      "learning_rate": 2.126733121310087e-06,
+      "loss": 1.0383,
+      "step": 2013
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.11726600676774979,
+      "learning_rate": 2.0967791602941154e-06,
+      "loss": 1.0678,
+      "step": 2014
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.10593971610069275,
+      "learning_rate": 2.067035406235085e-06,
+      "loss": 1.0557,
+      "step": 2015
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.10308150947093964,
+      "learning_rate": 2.037501922995477e-06,
+      "loss": 1.1061,
+      "step": 2016
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.1087673008441925,
+      "learning_rate": 2.0081787739863246e-06,
+      "loss": 1.1489,
+      "step": 2017
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.11029636859893799,
+      "learning_rate": 1.979066022167042e-06,
+      "loss": 0.9823,
+      "step": 2018
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.11021058261394501,
+      "learning_rate": 1.950163730045318e-06,
+      "loss": 1.0475,
+      "step": 2019
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.10561564564704895,
+      "learning_rate": 1.921471959676957e-06,
+      "loss": 1.0393,
+      "step": 2020
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.10535459220409393,
+      "learning_rate": 1.892990772665737e-06,
+      "loss": 0.9632,
+      "step": 2021
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.1089758425951004,
+      "learning_rate": 1.8647202301633194e-06,
+      "loss": 1.0655,
+      "step": 2022
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.11145118623971939,
+      "learning_rate": 1.8366603928690717e-06,
+      "loss": 1.002,
+      "step": 2023
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.105910524725914,
+      "learning_rate": 1.8088113210299795e-06,
+      "loss": 0.9339,
+      "step": 2024
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.10975908488035202,
+      "learning_rate": 1.7811730744404897e-06,
+      "loss": 0.9882,
+      "step": 2025
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.11369116604328156,
+      "learning_rate": 1.7537457124423895e-06,
+      "loss": 1.0856,
+      "step": 2026
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.10238272696733475,
+      "learning_rate": 1.7265292939246501e-06,
+      "loss": 1.055,
+      "step": 2027
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.10706526786088943,
+      "learning_rate": 1.6995238773233835e-06,
+      "loss": 1.1143,
+      "step": 2028
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.10359346121549606,
+      "learning_rate": 1.6727295206216077e-06,
+      "loss": 1.0772,
+      "step": 2029
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.10358715057373047,
+      "learning_rate": 1.6461462813492034e-06,
+      "loss": 0.9611,
+      "step": 2030
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.11153654009103775,
+      "learning_rate": 1.6197742165827589e-06,
+      "loss": 1.069,
+      "step": 2031
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.11218453198671341,
+      "learning_rate": 1.5936133829454691e-06,
+      "loss": 1.0573,
+      "step": 2032
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.11008001118898392,
+      "learning_rate": 1.5676638366069473e-06,
+      "loss": 1.0173,
+      "step": 2033
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.10740575194358826,
+      "learning_rate": 1.5419256332832255e-06,
+      "loss": 1.0578,
+      "step": 2034
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.10864772647619247,
+      "learning_rate": 1.5163988282364871e-06,
+      "loss": 1.024,
+      "step": 2035
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.11233103275299072,
+      "learning_rate": 1.4910834762750902e-06,
+      "loss": 0.966,
+      "step": 2036
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.10624489933252335,
+      "learning_rate": 1.4659796317533337e-06,
+      "loss": 1.0366,
+      "step": 2037
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.10481962561607361,
+      "learning_rate": 1.4410873485714238e-06,
+      "loss": 1.1202,
+      "step": 2038
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.10571899265050888,
+      "learning_rate": 1.4164066801753195e-06,
+      "loss": 1.0358,
+      "step": 2039
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.10203468054533005,
+      "learning_rate": 1.3919376795566209e-06,
+      "loss": 1.0009,
+      "step": 2040
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.10660209506750107,
+      "learning_rate": 1.3676803992524577e-06,
+      "loss": 1.0102,
+      "step": 2041
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.10626547783613205,
+      "learning_rate": 1.3436348913453578e-06,
+      "loss": 1.0325,
+      "step": 2042
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.10563212633132935,
+      "learning_rate": 1.3198012074631782e-06,
+      "loss": 1.0129,
+      "step": 2043
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.1028657853603363,
+      "learning_rate": 1.296179398778985e-06,
+      "loss": 1.0263,
+      "step": 2044
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.10210205614566803,
+      "learning_rate": 1.2727695160108965e-06,
+      "loss": 1.0477,
+      "step": 2045
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.1045612022280693,
+      "learning_rate": 1.249571609422029e-06,
+      "loss": 1.0531,
+      "step": 2046
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.11034145206212997,
+      "learning_rate": 1.226585728820362e-06,
+      "loss": 1.1266,
+      "step": 2047
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.10719690471887589,
+      "learning_rate": 1.2038119235586176e-06,
+      "loss": 0.9949,
+      "step": 2048
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.10602158308029175,
+      "learning_rate": 1.1812502425341931e-06,
+      "loss": 1.0609,
+      "step": 2049
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.10623074322938919,
+      "learning_rate": 1.158900734189039e-06,
+      "loss": 0.9451,
+      "step": 2050
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.10773053020238876,
+      "learning_rate": 1.1367634465095477e-06,
+      "loss": 1.002,
+      "step": 2051
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.10590239614248276,
+      "learning_rate": 1.1148384270264544e-06,
+      "loss": 1.038,
+      "step": 2052
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.10404451936483383,
+      "learning_rate": 1.093125722814725e-06,
+      "loss": 1.0407,
+      "step": 2053
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.11180682480335236,
+      "learning_rate": 1.0716253804934795e-06,
+      "loss": 1.0726,
+      "step": 2054
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.10890422761440277,
+      "learning_rate": 1.0503374462258575e-06,
+      "loss": 1.0762,
+      "step": 2055
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.10686809569597244,
+      "learning_rate": 1.0292619657189751e-06,
+      "loss": 1.0254,
+      "step": 2056
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.1056630089879036,
+      "learning_rate": 1.0083989842237684e-06,
+      "loss": 1.1076,
+      "step": 2057
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.10333152115345001,
+      "learning_rate": 9.877485465349058e-07,
+      "loss": 1.0505,
+      "step": 2058
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.1093318983912468,
+      "learning_rate": 9.67310696990731e-07,
+      "loss": 1.0654,
+      "step": 2059
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.10000356286764145,
+      "learning_rate": 9.470854794731421e-07,
+      "loss": 0.943,
+      "step": 2060
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.11062807589769363,
+      "learning_rate": 9.270729374074694e-07,
+      "loss": 1.0802,
+      "step": 2061
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.11128256469964981,
+      "learning_rate": 9.072731137624413e-07,
+      "loss": 1.0551,
+      "step": 2062
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.10686544328927994,
+      "learning_rate": 8.876860510500296e-07,
+      "loss": 1.0359,
+      "step": 2063
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.10752417892217636,
+      "learning_rate": 8.683117913254268e-07,
+      "loss": 0.9596,
+      "step": 2064
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.11450395733118057,
+      "learning_rate": 8.491503761868802e-07,
+      "loss": 1.1113,
+      "step": 2065
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.10330711305141449,
+      "learning_rate": 8.3020184677568e-07,
+      "loss": 1.0886,
+      "step": 2066
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.10227303951978683,
+      "learning_rate": 8.114662437760046e-07,
+      "loss": 0.9385,
+      "step": 2067
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.10366334021091461,
+      "learning_rate": 7.929436074148533e-07,
+      "loss": 0.9858,
+      "step": 2068
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.10696310549974442,
+      "learning_rate": 7.746339774619804e-07,
+      "loss": 1.0575,
+      "step": 2069
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.10279658436775208,
+      "learning_rate": 7.56537393229817e-07,
+      "loss": 1.0597,
+      "step": 2070
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.10408645123243332,
+      "learning_rate": 7.386538935733267e-07,
+      "loss": 1.035,
+      "step": 2071
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.10752113163471222,
+      "learning_rate": 7.209835168899836e-07,
+      "loss": 1.0076,
+      "step": 2072
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.10591957718133926,
+      "learning_rate": 7.035263011196724e-07,
+      "loss": 1.0706,
+      "step": 2073
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.10105755925178528,
+      "learning_rate": 6.862822837445881e-07,
+      "loss": 1.0207,
+      "step": 2074
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.10948655754327774,
+      "learning_rate": 6.692515017891698e-07,
+      "loss": 0.9719,
+      "step": 2075
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.10324559360742569,
+      "learning_rate": 6.524339918200339e-07,
+      "loss": 1.0106,
+      "step": 2076
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.10212616622447968,
+      "learning_rate": 6.358297899458743e-07,
+      "loss": 1.0089,
+      "step": 2077
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.10476096719503403,
+      "learning_rate": 6.194389318173954e-07,
+      "loss": 1.0263,
+      "step": 2078
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.1066427081823349,
+      "learning_rate": 6.032614526272462e-07,
+      "loss": 1.0655,
+      "step": 2079
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.11500557512044907,
+      "learning_rate": 5.872973871098975e-07,
+      "loss": 1.0229,
+      "step": 2080
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.11618941277265549,
+      "learning_rate": 5.71546769541631e-07,
+      "loss": 1.1361,
+      "step": 2081
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.10675172507762909,
+      "learning_rate": 5.560096337404397e-07,
+      "loss": 1.0,
+      "step": 2082
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.11477717012166977,
+      "learning_rate": 5.406860130659164e-07,
+      "loss": 1.1056,
+      "step": 2083
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.1035357117652893,
+      "learning_rate": 5.255759404192651e-07,
+      "loss": 1.0419,
+      "step": 2084
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.11062569171190262,
+      "learning_rate": 5.106794482431565e-07,
+      "loss": 1.064,
+      "step": 2085
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.1111309677362442,
+      "learning_rate": 4.959965685216949e-07,
+      "loss": 1.0308,
+      "step": 2086
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.10526414215564728,
+      "learning_rate": 4.815273327803182e-07,
+      "loss": 1.0458,
+      "step": 2087
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.11080475896596909,
+      "learning_rate": 4.6727177208577553e-07,
+      "loss": 1.1074,
+      "step": 2088
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.10295668244361877,
+      "learning_rate": 4.5322991704602746e-07,
+      "loss": 1.0114,
+      "step": 2089
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.11288978159427643,
+      "learning_rate": 4.3940179781019055e-07,
+      "loss": 1.0604,
+      "step": 2090
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.10728578269481659,
+      "learning_rate": 4.257874440684706e-07,
+      "loss": 1.0807,
+      "step": 2091
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.10228927433490753,
+      "learning_rate": 4.1238688505211845e-07,
+      "loss": 1.026,
+      "step": 2092
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.10391142219305038,
+      "learning_rate": 3.9920014953330755e-07,
+      "loss": 1.1084,
+      "step": 2093
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.11541945487260818,
+      "learning_rate": 3.8622726582514537e-07,
+      "loss": 1.0383,
+      "step": 2094
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.09570682048797607,
+      "learning_rate": 3.7346826178156213e-07,
+      "loss": 0.9372,
+      "step": 2095
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.10449600219726562,
+      "learning_rate": 3.609231647972999e-07,
+      "loss": 1.0205,
+      "step": 2096
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.1027015745639801,
+      "learning_rate": 3.485920018077793e-07,
+      "loss": 0.9693,
+      "step": 2097
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.11769603192806244,
+      "learning_rate": 3.364747992891215e-07,
+      "loss": 1.073,
+      "step": 2098
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.10148919373750687,
+      "learning_rate": 3.245715832580376e-07,
+      "loss": 1.0234,
+      "step": 2099
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.10759225487709045,
+      "learning_rate": 3.128823792718061e-07,
+      "loss": 1.0337,
+      "step": 2100
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.0999385192990303,
+      "learning_rate": 3.0140721242817303e-07,
+      "loss": 0.9694,
+      "step": 2101
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.104294553399086,
+      "learning_rate": 2.901461073653633e-07,
+      "loss": 1.0124,
+      "step": 2102
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.09916951507329941,
+      "learning_rate": 2.790990882619582e-07,
+      "loss": 0.9932,
+      "step": 2103
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.11422614753246307,
+      "learning_rate": 2.6826617883690673e-07,
+      "loss": 1.0595,
+      "step": 2104
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.10451515018939972,
+      "learning_rate": 2.5764740234943686e-07,
+      "loss": 1.0154,
+      "step": 2105
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.11038151383399963,
+      "learning_rate": 2.472427815989886e-07,
+      "loss": 1.1035,
+      "step": 2106
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.1085631251335144,
+      "learning_rate": 2.370523389252366e-07,
+      "loss": 1.0039,
+      "step": 2107
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.10915839672088623,
+      "learning_rate": 2.2707609620795656e-07,
+      "loss": 1.0662,
+      "step": 2108
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.1062905415892601,
+      "learning_rate": 2.173140748670366e-07,
+      "loss": 1.0583,
+      "step": 2109
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.1031581237912178,
+      "learning_rate": 2.0776629586239936e-07,
+      "loss": 1.0645,
+      "step": 2110
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.10656368732452393,
+      "learning_rate": 1.984327796940022e-07,
+      "loss": 1.0608,
+      "step": 2111
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.09900179505348206,
+      "learning_rate": 1.8931354640171484e-07,
+      "loss": 0.9876,
+      "step": 2112
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.10410574078559875,
+      "learning_rate": 1.8040861556535282e-07,
+      "loss": 0.975,
+      "step": 2113
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.10652648657560349,
+      "learning_rate": 1.7171800630458868e-07,
+      "loss": 1.0494,
+      "step": 2114
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.10137167572975159,
+      "learning_rate": 1.6324173727896297e-07,
+      "loss": 0.9864,
+      "step": 2115
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.10980936139822006,
+      "learning_rate": 1.5497982668775113e-07,
+      "loss": 1.0137,
+      "step": 2116
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.10673750936985016,
+      "learning_rate": 1.4693229227003004e-07,
+      "loss": 1.0688,
+      "step": 2117
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.10257307440042496,
+      "learning_rate": 1.3909915130457808e-07,
+      "loss": 1.0204,
+      "step": 2118
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.11068095266819,
+      "learning_rate": 1.314804206098308e-07,
+      "loss": 1.0531,
+      "step": 2119
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.10518340766429901,
+      "learning_rate": 1.2407611654390305e-07,
+      "loss": 1.01,
+      "step": 2120
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.11482223868370056,
+      "learning_rate": 1.1688625500448913e-07,
+      "loss": 1.1255,
+      "step": 2121
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.1066737249493599,
+      "learning_rate": 1.0991085142886271e-07,
+      "loss": 1.0579,
+      "step": 2122
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.10214506089687347,
+      "learning_rate": 1.0314992079383246e-07,
+      "loss": 1.0828,
+      "step": 2123
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.10915733128786087,
+      "learning_rate": 9.660347761571986e-08,
+      "loss": 1.0076,
+      "step": 2124
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.0965559184551239,
+      "learning_rate": 9.027153595033699e-08,
+      "loss": 0.9414,
+      "step": 2125
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.10989803820848465,
+      "learning_rate": 8.41541093929199e-08,
+      "loss": 1.0125,
+      "step": 2126
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.11001843214035034,
+      "learning_rate": 7.82512110781286e-08,
+      "loss": 1.1175,
+      "step": 2127
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.10287085175514221,
+      "learning_rate": 7.256285368001381e-08,
+      "loss": 1.0601,
+      "step": 2128
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.10772088915109634,
+      "learning_rate": 6.708904941198357e-08,
+      "loss": 1.0688,
+      "step": 2129
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.1085272952914238,
+      "learning_rate": 6.182981002679223e-08,
+      "loss": 1.1043,
+      "step": 2130
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.10850529372692108,
+      "learning_rate": 5.678514681648484e-08,
+      "loss": 1.0376,
+      "step": 2131
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.10329233855009079,
+      "learning_rate": 5.195507061240834e-08,
+      "loss": 1.0923,
+      "step": 2132
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.10155212134122849,
+      "learning_rate": 4.733959178517822e-08,
+      "loss": 1.0032,
+      "step": 2133
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.10105917602777481,
+      "learning_rate": 4.293872024463408e-08,
+      "loss": 0.9926,
+      "step": 2134
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.10948021709918976,
+      "learning_rate": 3.8752465439828576e-08,
+      "loss": 1.1263,
+      "step": 2135
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.11103621870279312,
+      "learning_rate": 3.4780836359038505e-08,
+      "loss": 1.0837,
+      "step": 2136
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.1053047850728035,
+      "learning_rate": 3.1023841529698174e-08,
+      "loss": 1.0355,
+      "step": 2137
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.10603896528482437,
+      "learning_rate": 2.7481489018410523e-08,
+      "loss": 1.0354,
+      "step": 2138
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.10304504632949829,
+      "learning_rate": 2.415378643092492e-08,
+      "loss": 1.0612,
+      "step": 2139
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.10404917597770691,
+      "learning_rate": 2.1040740912126046e-08,
+      "loss": 1.0248,
+      "step": 2140
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.10357488691806793,
+      "learning_rate": 1.8142359145978393e-08,
+      "loss": 1.0345,
+      "step": 2141
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.10982201248407364,
+      "learning_rate": 1.545864735558178e-08,
+      "loss": 1.0902,
+      "step": 2142
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.10491571575403214,
+      "learning_rate": 1.2989611303115823e-08,
+      "loss": 0.9883,
+      "step": 2143
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.11486451327800751,
+      "learning_rate": 1.0735256289795548e-08,
+      "loss": 1.0909,
+      "step": 2144
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.1067158430814743,
+      "learning_rate": 8.69558715593799e-09,
+      "loss": 1.0663,
+      "step": 2145
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.10688413679599762,
+      "learning_rate": 6.8706082808955855e-09,
+      "loss": 1.0889,
+      "step": 2146
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.10961759090423584,
+      "learning_rate": 5.260323583056171e-09,
+      "loss": 1.0789,
+      "step": 2147
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.10168945044279099,
+      "learning_rate": 3.8647365198429815e-09,
+      "loss": 0.9485,
+      "step": 2148
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.1018710508942604,
+      "learning_rate": 2.6838500877035545e-09,
+      "loss": 0.9652,
+      "step": 2149
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.1057569831609726,
+      "learning_rate": 1.7176668221208225e-09,
+      "loss": 1.0747,
+      "step": 2150
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.10160427540540695,
+      "learning_rate": 9.661887975687123e-10,
+      "loss": 0.989,
+      "step": 2151
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.11478938907384872,
+      "learning_rate": 4.294176275232431e-10,
+      "loss": 1.0292,
+      "step": 2152
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.11082255095243454,
+      "learning_rate": 1.0735446450693686e-10,
+      "loss": 1.0586,
+      "step": 2153
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.10913219302892685,
+      "learning_rate": 0.0,
+      "loss": 1.0931,
+      "step": 2154
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2154,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 718,
+  "total_flos": 6.427626055850263e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}