diff --git "a/checkpoint-2700/trainer_state.json" "b/checkpoint-2700/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-2700/trainer_state.json"
@@ -0,0 +1,18933 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9042196918955124,
+  "eval_steps": 500,
+  "global_step": 2700,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0003348961821835231,
+      "grad_norm": 18.186641693115234,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 2.0732,
+      "step": 1
+    },
+    {
+      "epoch": 0.0006697923643670462,
+      "grad_norm": 15.181599617004395,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 2.0167,
+      "step": 2
+    },
+    {
+      "epoch": 0.0010046885465505692,
+      "grad_norm": 14.661438941955566,
+      "learning_rate": 6e-06,
+      "loss": 2.0268,
+      "step": 3
+    },
+    {
+      "epoch": 0.0013395847287340924,
+      "grad_norm": 9.589187622070312,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 2.0335,
+      "step": 4
+    },
+    {
+      "epoch": 0.0016744809109176155,
+      "grad_norm": 5.236238479614258,
+      "learning_rate": 1e-05,
+      "loss": 1.8674,
+      "step": 5
+    },
+    {
+      "epoch": 0.0020093770931011385,
+      "grad_norm": 3.8239383697509766,
+      "learning_rate": 1.2e-05,
+      "loss": 1.8922,
+      "step": 6
+    },
+    {
+      "epoch": 0.002344273275284662,
+      "grad_norm": 2.503788709640503,
+      "learning_rate": 1.4000000000000001e-05,
+      "loss": 1.8812,
+      "step": 7
+    },
+    {
+      "epoch": 0.0026791694574681848,
+      "grad_norm": 1.736107349395752,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 1.8371,
+      "step": 8
+    },
+    {
+      "epoch": 0.003014065639651708,
+      "grad_norm": 2.421985149383545,
+      "learning_rate": 1.8e-05,
+      "loss": 1.8273,
+      "step": 9
+    },
+    {
+      "epoch": 0.003348961821835231,
+      "grad_norm": 2.248610496520996,
+      "learning_rate": 2e-05,
+      "loss": 1.7187,
+      "step": 10
+    },
+    {
+      "epoch": 0.003683858004018754,
+      "grad_norm": 1.246288776397705,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 1.8992,
+      "step": 11
+    },
+    {
+      "epoch": 0.004018754186202277,
+      "grad_norm": 1.4386337995529175,
+      "learning_rate": 2.4e-05,
+      "loss": 1.7567,
+      "step": 12
+    },
+    {
+      "epoch": 0.004353650368385801,
+      "grad_norm": 0.9633839726448059,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 1.7409,
+      "step": 13
+    },
+    {
+      "epoch": 0.004688546550569324,
+      "grad_norm": 0.8049834966659546,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 1.669,
+      "step": 14
+    },
+    {
+      "epoch": 0.005023442732752847,
+      "grad_norm": 0.8306618928909302,
+      "learning_rate": 3e-05,
+      "loss": 1.7372,
+      "step": 15
+    },
+    {
+      "epoch": 0.0053583389149363695,
+      "grad_norm": 0.9875802397727966,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 1.7664,
+      "step": 16
+    },
+    {
+      "epoch": 0.0056932350971198925,
+      "grad_norm": 0.9656491875648499,
+      "learning_rate": 3.4000000000000007e-05,
+      "loss": 1.7711,
+      "step": 17
+    },
+    {
+      "epoch": 0.006028131279303416,
+      "grad_norm": 0.8717008829116821,
+      "learning_rate": 3.6e-05,
+      "loss": 1.6949,
+      "step": 18
+    },
+    {
+      "epoch": 0.006363027461486939,
+      "grad_norm": 0.8283945918083191,
+      "learning_rate": 3.8e-05,
+      "loss": 1.7468,
+      "step": 19
+    },
+    {
+      "epoch": 0.006697923643670462,
+      "grad_norm": 0.7493966221809387,
+      "learning_rate": 4e-05,
+      "loss": 1.6974,
+      "step": 20
+    },
+    {
+      "epoch": 0.007032819825853985,
+      "grad_norm": 0.7758145928382874,
+      "learning_rate": 4.2e-05,
+      "loss": 1.7485,
+      "step": 21
+    },
+    {
+      "epoch": 0.007367716008037508,
+      "grad_norm": 0.7642164826393127,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 1.745,
+      "step": 22
+    },
+    {
+      "epoch": 0.007702612190221032,
+      "grad_norm": 0.706461489200592,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 1.6891,
+      "step": 23
+    },
+    {
+      "epoch": 0.008037508372404554,
+      "grad_norm": 0.7163651585578918,
+      "learning_rate": 4.8e-05,
+      "loss": 1.7129,
+      "step": 24
+    },
+    {
+      "epoch": 0.008372404554588079,
+      "grad_norm": 0.8007297515869141,
+      "learning_rate": 5e-05,
+      "loss": 1.6796,
+      "step": 25
+    },
+    {
+      "epoch": 0.008707300736771601,
+      "grad_norm": 0.7303621768951416,
+      "learning_rate": 5.2000000000000004e-05,
+      "loss": 1.5926,
+      "step": 26
+    },
+    {
+      "epoch": 0.009042196918955124,
+      "grad_norm": 0.8409650325775146,
+      "learning_rate": 5.4000000000000005e-05,
+      "loss": 1.7859,
+      "step": 27
+    },
+    {
+      "epoch": 0.009377093101138647,
+      "grad_norm": 0.7095215320587158,
+      "learning_rate": 5.6000000000000006e-05,
+      "loss": 1.654,
+      "step": 28
+    },
+    {
+      "epoch": 0.00971198928332217,
+      "grad_norm": 0.7262499332427979,
+      "learning_rate": 5.8e-05,
+      "loss": 1.7337,
+      "step": 29
+    },
+    {
+      "epoch": 0.010046885465505693,
+      "grad_norm": 0.7079520225524902,
+      "learning_rate": 6e-05,
+      "loss": 1.7591,
+      "step": 30
+    },
+    {
+      "epoch": 0.010381781647689216,
+      "grad_norm": 0.6347719430923462,
+      "learning_rate": 6.2e-05,
+      "loss": 1.6914,
+      "step": 31
+    },
+    {
+      "epoch": 0.010716677829872739,
+      "grad_norm": 0.6208151578903198,
+      "learning_rate": 6.400000000000001e-05,
+      "loss": 1.6545,
+      "step": 32
+    },
+    {
+      "epoch": 0.011051574012056262,
+      "grad_norm": 0.5771459341049194,
+      "learning_rate": 6.6e-05,
+      "loss": 1.6261,
+      "step": 33
+    },
+    {
+      "epoch": 0.011386470194239785,
+      "grad_norm": 0.6044191122055054,
+      "learning_rate": 6.800000000000001e-05,
+      "loss": 1.6579,
+      "step": 34
+    },
+    {
+      "epoch": 0.01172136637642331,
+      "grad_norm": 0.6431082487106323,
+      "learning_rate": 7e-05,
+      "loss": 1.6837,
+      "step": 35
+    },
+    {
+      "epoch": 0.012056262558606833,
+      "grad_norm": 0.6023712158203125,
+      "learning_rate": 7.2e-05,
+      "loss": 1.6224,
+      "step": 36
+    },
+    {
+      "epoch": 0.012391158740790355,
+      "grad_norm": 0.6101360321044922,
+      "learning_rate": 7.4e-05,
+      "loss": 1.6735,
+      "step": 37
+    },
+    {
+      "epoch": 0.012726054922973878,
+      "grad_norm": 0.5465442538261414,
+      "learning_rate": 7.6e-05,
+      "loss": 1.5562,
+      "step": 38
+    },
+    {
+      "epoch": 0.013060951105157401,
+      "grad_norm": 0.6528838872909546,
+      "learning_rate": 7.800000000000001e-05,
+      "loss": 1.7578,
+      "step": 39
+    },
+    {
+      "epoch": 0.013395847287340924,
+      "grad_norm": 0.5717447400093079,
+      "learning_rate": 8e-05,
+      "loss": 1.6505,
+      "step": 40
+    },
+    {
+      "epoch": 0.013730743469524447,
+      "grad_norm": 0.6488905549049377,
+      "learning_rate": 8.2e-05,
+      "loss": 1.6638,
+      "step": 41
+    },
+    {
+      "epoch": 0.01406563965170797,
+      "grad_norm": 0.558308482170105,
+      "learning_rate": 8.4e-05,
+      "loss": 1.6428,
+      "step": 42
+    },
+    {
+      "epoch": 0.014400535833891493,
+      "grad_norm": 0.608172595500946,
+      "learning_rate": 8.6e-05,
+      "loss": 1.6953,
+      "step": 43
+    },
+    {
+      "epoch": 0.014735432016075016,
+      "grad_norm": 0.5436323881149292,
+      "learning_rate": 8.800000000000001e-05,
+      "loss": 1.701,
+      "step": 44
+    },
+    {
+      "epoch": 0.01507032819825854,
+      "grad_norm": 0.5608189105987549,
+      "learning_rate": 9e-05,
+      "loss": 1.6088,
+      "step": 45
+    },
+    {
+      "epoch": 0.015405224380442064,
+      "grad_norm": 0.5891563296318054,
+      "learning_rate": 9.200000000000001e-05,
+      "loss": 1.6878,
+      "step": 46
+    },
+    {
+      "epoch": 0.015740120562625585,
+      "grad_norm": 0.5707667469978333,
+      "learning_rate": 9.4e-05,
+      "loss": 1.6171,
+      "step": 47
+    },
+    {
+      "epoch": 0.016075016744809108,
+      "grad_norm": 0.5519657731056213,
+      "learning_rate": 9.6e-05,
+      "loss": 1.6526,
+      "step": 48
+    },
+    {
+      "epoch": 0.01640991292699263,
+      "grad_norm": 0.5898794531822205,
+      "learning_rate": 9.8e-05,
+      "loss": 1.7157,
+      "step": 49
+    },
+    {
+      "epoch": 0.016744809109176157,
+      "grad_norm": 0.5518242716789246,
+      "learning_rate": 0.0001,
+      "loss": 1.6441,
+      "step": 50
+    },
+    {
+      "epoch": 0.01707970529135968,
+      "grad_norm": 0.5902829766273499,
+      "learning_rate": 0.0001,
+      "loss": 1.6424,
+      "step": 51
+    },
+    {
+      "epoch": 0.017414601473543203,
+      "grad_norm": 0.5596480965614319,
+      "learning_rate": 0.0001,
+      "loss": 1.7147,
+      "step": 52
+    },
+    {
+      "epoch": 0.017749497655726726,
+      "grad_norm": 0.5877808332443237,
+      "learning_rate": 0.0001,
+      "loss": 1.6327,
+      "step": 53
+    },
+    {
+      "epoch": 0.01808439383791025,
+      "grad_norm": 0.543509840965271,
+      "learning_rate": 0.0001,
+      "loss": 1.7011,
+      "step": 54
+    },
+    {
+      "epoch": 0.018419290020093772,
+      "grad_norm": 0.5165537595748901,
+      "learning_rate": 0.0001,
+      "loss": 1.5977,
+      "step": 55
+    },
+    {
+      "epoch": 0.018754186202277295,
+      "grad_norm": 0.5444625616073608,
+      "learning_rate": 0.0001,
+      "loss": 1.6508,
+      "step": 56
+    },
+    {
+      "epoch": 0.019089082384460818,
+      "grad_norm": 0.5317073464393616,
+      "learning_rate": 0.0001,
+      "loss": 1.6175,
+      "step": 57
+    },
+    {
+      "epoch": 0.01942397856664434,
+      "grad_norm": 0.570904016494751,
+      "learning_rate": 0.0001,
+      "loss": 1.7569,
+      "step": 58
+    },
+    {
+      "epoch": 0.019758874748827863,
+      "grad_norm": 0.5219286680221558,
+      "learning_rate": 0.0001,
+      "loss": 1.6522,
+      "step": 59
+    },
+    {
+      "epoch": 0.020093770931011386,
+      "grad_norm": 0.5607163310050964,
+      "learning_rate": 0.0001,
+      "loss": 1.6803,
+      "step": 60
+    },
+    {
+      "epoch": 0.02042866711319491,
+      "grad_norm": 0.5346683859825134,
+      "learning_rate": 0.0001,
+      "loss": 1.6452,
+      "step": 61
+    },
+    {
+      "epoch": 0.020763563295378432,
+      "grad_norm": 0.5436587333679199,
+      "learning_rate": 0.0001,
+      "loss": 1.6682,
+      "step": 62
+    },
+    {
+      "epoch": 0.021098459477561955,
+      "grad_norm": 0.5531876683235168,
+      "learning_rate": 0.0001,
+      "loss": 1.6845,
+      "step": 63
+    },
+    {
+      "epoch": 0.021433355659745478,
+      "grad_norm": 0.5084998607635498,
+      "learning_rate": 0.0001,
+      "loss": 1.682,
+      "step": 64
+    },
+    {
+      "epoch": 0.021768251841929,
+      "grad_norm": 0.5299241542816162,
+      "learning_rate": 0.0001,
+      "loss": 1.5749,
+      "step": 65
+    },
+    {
+      "epoch": 0.022103148024112524,
+      "grad_norm": 0.5054894089698792,
+      "learning_rate": 0.0001,
+      "loss": 1.6374,
+      "step": 66
+    },
+    {
+      "epoch": 0.022438044206296047,
+      "grad_norm": 0.5474890470504761,
+      "learning_rate": 0.0001,
+      "loss": 1.7065,
+      "step": 67
+    },
+    {
+      "epoch": 0.02277294038847957,
+      "grad_norm": 0.5523289442062378,
+      "learning_rate": 0.0001,
+      "loss": 1.8205,
+      "step": 68
+    },
+    {
+      "epoch": 0.023107836570663093,
+      "grad_norm": 0.5250942707061768,
+      "learning_rate": 0.0001,
+      "loss": 1.5829,
+      "step": 69
+    },
+    {
+      "epoch": 0.02344273275284662,
+      "grad_norm": 0.5237876772880554,
+      "learning_rate": 0.0001,
+      "loss": 1.65,
+      "step": 70
+    },
+    {
+      "epoch": 0.023777628935030142,
+      "grad_norm": 0.514022707939148,
+      "learning_rate": 0.0001,
+      "loss": 1.5468,
+      "step": 71
+    },
+    {
+      "epoch": 0.024112525117213665,
+      "grad_norm": 0.503179132938385,
+      "learning_rate": 0.0001,
+      "loss": 1.541,
+      "step": 72
+    },
+    {
+      "epoch": 0.024447421299397188,
+      "grad_norm": 0.5128785371780396,
+      "learning_rate": 0.0001,
+      "loss": 1.6285,
+      "step": 73
+    },
+    {
+      "epoch": 0.02478231748158071,
+      "grad_norm": 0.5173330903053284,
+      "learning_rate": 0.0001,
+      "loss": 1.6583,
+      "step": 74
+    },
+    {
+      "epoch": 0.025117213663764234,
+      "grad_norm": 0.5024128556251526,
+      "learning_rate": 0.0001,
+      "loss": 1.608,
+      "step": 75
+    },
+    {
+      "epoch": 0.025452109845947757,
+      "grad_norm": 0.5092707872390747,
+      "learning_rate": 0.0001,
+      "loss": 1.6552,
+      "step": 76
+    },
+    {
+      "epoch": 0.02578700602813128,
+      "grad_norm": 0.4851345717906952,
+      "learning_rate": 0.0001,
+      "loss": 1.5511,
+      "step": 77
+    },
+    {
+      "epoch": 0.026121902210314803,
+      "grad_norm": 0.5057558417320251,
+      "learning_rate": 0.0001,
+      "loss": 1.6468,
+      "step": 78
+    },
+    {
+      "epoch": 0.026456798392498326,
+      "grad_norm": 0.5011716485023499,
+      "learning_rate": 0.0001,
+      "loss": 1.6291,
+      "step": 79
+    },
+    {
+      "epoch": 0.02679169457468185,
+      "grad_norm": 0.5014387965202332,
+      "learning_rate": 0.0001,
+      "loss": 1.6706,
+      "step": 80
+    },
+    {
+      "epoch": 0.02712659075686537,
+      "grad_norm": 0.5116551518440247,
+      "learning_rate": 0.0001,
+      "loss": 1.6637,
+      "step": 81
+    },
+    {
+      "epoch": 0.027461486939048894,
+      "grad_norm": 0.4828563928604126,
+      "learning_rate": 0.0001,
+      "loss": 1.6061,
+      "step": 82
+    },
+    {
+      "epoch": 0.027796383121232417,
+      "grad_norm": 0.4810364842414856,
+      "learning_rate": 0.0001,
+      "loss": 1.6528,
+      "step": 83
+    },
+    {
+      "epoch": 0.02813127930341594,
+      "grad_norm": 0.49675026535987854,
+      "learning_rate": 0.0001,
+      "loss": 1.6463,
+      "step": 84
+    },
+    {
+      "epoch": 0.028466175485599463,
+      "grad_norm": 0.4998936951160431,
+      "learning_rate": 0.0001,
+      "loss": 1.5671,
+      "step": 85
+    },
+    {
+      "epoch": 0.028801071667782986,
+      "grad_norm": 0.5172111392021179,
+      "learning_rate": 0.0001,
+      "loss": 1.7286,
+      "step": 86
+    },
+    {
+      "epoch": 0.02913596784996651,
+      "grad_norm": 0.48785221576690674,
+      "learning_rate": 0.0001,
+      "loss": 1.5728,
+      "step": 87
+    },
+    {
+      "epoch": 0.029470864032150032,
+      "grad_norm": 0.4777396321296692,
+      "learning_rate": 0.0001,
+      "loss": 1.6219,
+      "step": 88
+    },
+    {
+      "epoch": 0.029805760214333555,
+      "grad_norm": 0.5374659299850464,
+      "learning_rate": 0.0001,
+      "loss": 1.7086,
+      "step": 89
+    },
+    {
+      "epoch": 0.03014065639651708,
+      "grad_norm": 0.4848657548427582,
+      "learning_rate": 0.0001,
+      "loss": 1.6158,
+      "step": 90
+    },
+    {
+      "epoch": 0.030475552578700604,
+      "grad_norm": 0.5165850520133972,
+      "learning_rate": 0.0001,
+      "loss": 1.6394,
+      "step": 91
+    },
+    {
+      "epoch": 0.030810448760884127,
+      "grad_norm": 0.501922607421875,
+      "learning_rate": 0.0001,
+      "loss": 1.6589,
+      "step": 92
+    },
+    {
+      "epoch": 0.03114534494306765,
+      "grad_norm": 0.5083390474319458,
+      "learning_rate": 0.0001,
+      "loss": 1.6182,
+      "step": 93
+    },
+    {
+      "epoch": 0.03148024112525117,
+      "grad_norm": 0.49391478300094604,
+      "learning_rate": 0.0001,
+      "loss": 1.6031,
+      "step": 94
+    },
+    {
+      "epoch": 0.031815137307434696,
+      "grad_norm": 0.5007357597351074,
+      "learning_rate": 0.0001,
+      "loss": 1.7042,
+      "step": 95
+    },
+    {
+      "epoch": 0.032150033489618215,
+      "grad_norm": 0.5300001502037048,
+      "learning_rate": 0.0001,
+      "loss": 1.6702,
+      "step": 96
+    },
+    {
+      "epoch": 0.03248492967180174,
+      "grad_norm": 0.46860334277153015,
+      "learning_rate": 0.0001,
+      "loss": 1.5321,
+      "step": 97
+    },
+    {
+      "epoch": 0.03281982585398526,
+      "grad_norm": 0.5419498682022095,
+      "learning_rate": 0.0001,
+      "loss": 1.6481,
+      "step": 98
+    },
+    {
+      "epoch": 0.03315472203616879,
+      "grad_norm": 0.4763254225254059,
+      "learning_rate": 0.0001,
+      "loss": 1.621,
+      "step": 99
+    },
+    {
+      "epoch": 0.033489618218352314,
+      "grad_norm": 0.4968387484550476,
+      "learning_rate": 0.0001,
+      "loss": 1.6008,
+      "step": 100
+    },
+    {
+      "epoch": 0.033824514400535834,
+      "grad_norm": 0.48966512084007263,
+      "learning_rate": 0.0001,
+      "loss": 1.6324,
+      "step": 101
+    },
+    {
+      "epoch": 0.03415941058271936,
+      "grad_norm": 0.49071934819221497,
+      "learning_rate": 0.0001,
+      "loss": 1.6715,
+      "step": 102
+    },
+    {
+      "epoch": 0.03449430676490288,
+      "grad_norm": 0.5435018539428711,
+      "learning_rate": 0.0001,
+      "loss": 1.7803,
+      "step": 103
+    },
+    {
+      "epoch": 0.034829202947086406,
+      "grad_norm": 0.5261979699134827,
+      "learning_rate": 0.0001,
+      "loss": 1.7211,
+      "step": 104
+    },
+    {
+      "epoch": 0.035164099129269925,
+      "grad_norm": 0.5578387379646301,
+      "learning_rate": 0.0001,
+      "loss": 1.7508,
+      "step": 105
+    },
+    {
+      "epoch": 0.03549899531145345,
+      "grad_norm": 0.5257611274719238,
+      "learning_rate": 0.0001,
+      "loss": 1.5434,
+      "step": 106
+    },
+    {
+      "epoch": 0.03583389149363697,
+      "grad_norm": 0.5152483582496643,
+      "learning_rate": 0.0001,
+      "loss": 1.7124,
+      "step": 107
+    },
+    {
+      "epoch": 0.0361687876758205,
+      "grad_norm": 0.4882405698299408,
+      "learning_rate": 0.0001,
+      "loss": 1.5853,
+      "step": 108
+    },
+    {
+      "epoch": 0.03650368385800402,
+      "grad_norm": 0.5341641306877136,
+      "learning_rate": 0.0001,
+      "loss": 1.63,
+      "step": 109
+    },
+    {
+      "epoch": 0.036838580040187544,
+      "grad_norm": 0.47165918350219727,
+      "learning_rate": 0.0001,
+      "loss": 1.6256,
+      "step": 110
+    },
+    {
+      "epoch": 0.03717347622237106,
+      "grad_norm": 0.48669424653053284,
+      "learning_rate": 0.0001,
+      "loss": 1.6785,
+      "step": 111
+    },
+    {
+      "epoch": 0.03750837240455459,
+      "grad_norm": 0.502474308013916,
+      "learning_rate": 0.0001,
+      "loss": 1.6176,
+      "step": 112
+    },
+    {
+      "epoch": 0.03784326858673811,
+      "grad_norm": 0.46959713101387024,
+      "learning_rate": 0.0001,
+      "loss": 1.5865,
+      "step": 113
+    },
+    {
+      "epoch": 0.038178164768921635,
+      "grad_norm": 0.5385250449180603,
+      "learning_rate": 0.0001,
+      "loss": 1.6965,
+      "step": 114
+    },
+    {
+      "epoch": 0.038513060951105155,
+      "grad_norm": 0.5123361349105835,
+      "learning_rate": 0.0001,
+      "loss": 1.637,
+      "step": 115
+    },
+    {
+      "epoch": 0.03884795713328868,
+      "grad_norm": 0.49659615755081177,
+      "learning_rate": 0.0001,
+      "loss": 1.6763,
+      "step": 116
+    },
+    {
+      "epoch": 0.0391828533154722,
+      "grad_norm": 0.47849777340888977,
+      "learning_rate": 0.0001,
+      "loss": 1.5615,
+      "step": 117
+    },
+    {
+      "epoch": 0.03951774949765573,
+      "grad_norm": 0.4773821234703064,
+      "learning_rate": 0.0001,
+      "loss": 1.6418,
+      "step": 118
+    },
+    {
+      "epoch": 0.039852645679839246,
+      "grad_norm": 0.46657007932662964,
+      "learning_rate": 0.0001,
+      "loss": 1.5907,
+      "step": 119
+    },
+    {
+      "epoch": 0.04018754186202277,
+      "grad_norm": 0.48586177825927734,
+      "learning_rate": 0.0001,
+      "loss": 1.6218,
+      "step": 120
+    },
+    {
+      "epoch": 0.0405224380442063,
+      "grad_norm": 0.49452975392341614,
+      "learning_rate": 0.0001,
+      "loss": 1.6308,
+      "step": 121
+    },
+    {
+      "epoch": 0.04085733422638982,
+      "grad_norm": 0.4575989246368408,
+      "learning_rate": 0.0001,
+      "loss": 1.5618,
+      "step": 122
+    },
+    {
+      "epoch": 0.041192230408573345,
+      "grad_norm": 0.4881477355957031,
+      "learning_rate": 0.0001,
+      "loss": 1.6237,
+      "step": 123
+    },
+    {
+      "epoch": 0.041527126590756865,
+      "grad_norm": 0.4715660512447357,
+      "learning_rate": 0.0001,
+      "loss": 1.6247,
+      "step": 124
+    },
+    {
+      "epoch": 0.04186202277294039,
+      "grad_norm": 0.4727221429347992,
+      "learning_rate": 0.0001,
+      "loss": 1.5856,
+      "step": 125
+    },
+    {
+      "epoch": 0.04219691895512391,
+      "grad_norm": 0.47755780816078186,
+      "learning_rate": 0.0001,
+      "loss": 1.5104,
+      "step": 126
+    },
+    {
+      "epoch": 0.04253181513730744,
+      "grad_norm": 0.4786100387573242,
+      "learning_rate": 0.0001,
+      "loss": 1.536,
+      "step": 127
+    },
+    {
+      "epoch": 0.042866711319490956,
+      "grad_norm": 0.5105636119842529,
+      "learning_rate": 0.0001,
+      "loss": 1.7084,
+      "step": 128
+    },
+    {
+      "epoch": 0.04320160750167448,
+      "grad_norm": 0.5502948760986328,
+      "learning_rate": 0.0001,
+      "loss": 1.7168,
+      "step": 129
+    },
+    {
+      "epoch": 0.043536503683858,
+      "grad_norm": 0.4909312129020691,
+      "learning_rate": 0.0001,
+      "loss": 1.7093,
+      "step": 130
+    },
+    {
+      "epoch": 0.04387139986604153,
+      "grad_norm": 0.46086591482162476,
+      "learning_rate": 0.0001,
+      "loss": 1.5041,
+      "step": 131
+    },
+    {
+      "epoch": 0.04420629604822505,
+      "grad_norm": 0.49003592133522034,
+      "learning_rate": 0.0001,
+      "loss": 1.6106,
+      "step": 132
+    },
+    {
+      "epoch": 0.044541192230408574,
+      "grad_norm": 0.4714739918708801,
+      "learning_rate": 0.0001,
+      "loss": 1.6323,
+      "step": 133
+    },
+    {
+      "epoch": 0.044876088412592094,
+      "grad_norm": 0.46458280086517334,
+      "learning_rate": 0.0001,
+      "loss": 1.5321,
+      "step": 134
+    },
+    {
+      "epoch": 0.04521098459477562,
+      "grad_norm": 0.48216310143470764,
+      "learning_rate": 0.0001,
+      "loss": 1.668,
+      "step": 135
+    },
+    {
+      "epoch": 0.04554588077695914,
+      "grad_norm": 0.49191856384277344,
+      "learning_rate": 0.0001,
+      "loss": 1.5979,
+      "step": 136
+    },
+    {
+      "epoch": 0.045880776959142666,
+      "grad_norm": 0.48731929063796997,
+      "learning_rate": 0.0001,
+      "loss": 1.5877,
+      "step": 137
+    },
+    {
+      "epoch": 0.046215673141326186,
+      "grad_norm": 0.4742424786090851,
+      "learning_rate": 0.0001,
+      "loss": 1.654,
+      "step": 138
+    },
+    {
+      "epoch": 0.04655056932350971,
+      "grad_norm": 0.4645659327507019,
+      "learning_rate": 0.0001,
+      "loss": 1.5754,
+      "step": 139
+    },
+    {
+      "epoch": 0.04688546550569324,
+      "grad_norm": 0.48509088158607483,
+      "learning_rate": 0.0001,
+      "loss": 1.5807,
+      "step": 140
+    },
+    {
+      "epoch": 0.04722036168787676,
+      "grad_norm": 0.45820191502571106,
+      "learning_rate": 0.0001,
+      "loss": 1.6338,
+      "step": 141
+    },
+    {
+      "epoch": 0.047555257870060284,
+      "grad_norm": 0.4655672609806061,
+      "learning_rate": 0.0001,
+      "loss": 1.6341,
+      "step": 142
+    },
+    {
+      "epoch": 0.047890154052243804,
+      "grad_norm": 0.4950549006462097,
+      "learning_rate": 0.0001,
+      "loss": 1.6622,
+      "step": 143
+    },
+    {
+      "epoch": 0.04822505023442733,
+      "grad_norm": 0.46763336658477783,
+      "learning_rate": 0.0001,
+      "loss": 1.5912,
+      "step": 144
+    },
+    {
+      "epoch": 0.04855994641661085,
+      "grad_norm": 0.47102072834968567,
+      "learning_rate": 0.0001,
+      "loss": 1.5721,
+      "step": 145
+    },
+    {
+      "epoch": 0.048894842598794376,
+      "grad_norm": 0.4916516840457916,
+      "learning_rate": 0.0001,
+      "loss": 1.6117,
+      "step": 146
+    },
+    {
+      "epoch": 0.049229738780977896,
+      "grad_norm": 0.4632984399795532,
+      "learning_rate": 0.0001,
+      "loss": 1.6008,
+      "step": 147
+    },
+    {
+      "epoch": 0.04956463496316142,
+      "grad_norm": 0.45667365193367004,
+      "learning_rate": 0.0001,
+      "loss": 1.5391,
+      "step": 148
+    },
+    {
+      "epoch": 0.04989953114534494,
+      "grad_norm": 0.4753417670726776,
+      "learning_rate": 0.0001,
+      "loss": 1.6279,
+      "step": 149
+    },
+    {
+      "epoch": 0.05023442732752847,
+      "grad_norm": 0.48078927397727966,
+      "learning_rate": 0.0001,
+      "loss": 1.6326,
+      "step": 150
+    },
+    {
+      "epoch": 0.05056932350971199,
+      "grad_norm": 0.4711712896823883,
+      "learning_rate": 0.0001,
+      "loss": 1.643,
+      "step": 151
+    },
+    {
+      "epoch": 0.050904219691895514,
+      "grad_norm": 0.5034206509590149,
+      "learning_rate": 0.0001,
+      "loss": 1.6122,
+      "step": 152
+    },
+    {
+      "epoch": 0.05123911587407903,
+      "grad_norm": 0.46327561140060425,
+      "learning_rate": 0.0001,
+      "loss": 1.6089,
+      "step": 153
+    },
+    {
+      "epoch": 0.05157401205626256,
+      "grad_norm": 0.4954410493373871,
+      "learning_rate": 0.0001,
+      "loss": 1.6108,
+      "step": 154
+    },
+    {
+      "epoch": 0.05190890823844608,
+      "grad_norm": 0.5449566841125488,
+      "learning_rate": 0.0001,
+      "loss": 1.7258,
+      "step": 155
+    },
+    {
+      "epoch": 0.052243804420629605,
+      "grad_norm": 0.4606022238731384,
+      "learning_rate": 0.0001,
+      "loss": 1.5401,
+      "step": 156
+    },
+    {
+      "epoch": 0.052578700602813125,
+      "grad_norm": 0.5017722249031067,
+      "learning_rate": 0.0001,
+      "loss": 1.6968,
+      "step": 157
+    },
+    {
+      "epoch": 0.05291359678499665,
+      "grad_norm": 0.4925632178783417,
+      "learning_rate": 0.0001,
+      "loss": 1.6701,
+      "step": 158
+    },
+    {
+      "epoch": 0.05324849296718017,
+      "grad_norm": 0.4624950587749481,
+      "learning_rate": 0.0001,
+      "loss": 1.5978,
+      "step": 159
+    },
+    {
+      "epoch": 0.0535833891493637,
+      "grad_norm": 0.5322679281234741,
+      "learning_rate": 0.0001,
+      "loss": 1.6278,
+      "step": 160
+    },
+    {
+      "epoch": 0.053918285331547224,
+      "grad_norm": 0.4694281816482544,
+      "learning_rate": 0.0001,
+      "loss": 1.6588,
+      "step": 161
+    },
+    {
+      "epoch": 0.05425318151373074,
+      "grad_norm": 0.5067780017852783,
+      "learning_rate": 0.0001,
+      "loss": 1.5661,
+      "step": 162
+    },
+    {
+      "epoch": 0.05458807769591427,
+      "grad_norm": 0.4347548484802246,
+      "learning_rate": 0.0001,
+      "loss": 1.5589,
+      "step": 163
+    },
+    {
+      "epoch": 0.05492297387809779,
+      "grad_norm": 0.5048144459724426,
+      "learning_rate": 0.0001,
+      "loss": 1.6149,
+      "step": 164
+    },
+    {
+      "epoch": 0.055257870060281315,
+      "grad_norm": 0.48803550004959106,
+      "learning_rate": 0.0001,
+      "loss": 1.59,
+      "step": 165
+    },
+    {
+      "epoch": 0.055592766242464835,
+      "grad_norm": 0.4865317940711975,
+      "learning_rate": 0.0001,
+      "loss": 1.5392,
+      "step": 166
+    },
+    {
+      "epoch": 0.05592766242464836,
+      "grad_norm": 0.4582239091396332,
+      "learning_rate": 0.0001,
+      "loss": 1.6371,
+      "step": 167
+    },
+    {
+      "epoch": 0.05626255860683188,
+      "grad_norm": 0.5068498253822327,
+      "learning_rate": 0.0001,
+      "loss": 1.665,
+      "step": 168
+    },
+    {
+      "epoch": 0.05659745478901541,
+      "grad_norm": 0.4794592559337616,
+      "learning_rate": 0.0001,
+      "loss": 1.629,
+      "step": 169
+    },
+    {
+      "epoch": 0.056932350971198926,
+      "grad_norm": 0.4870087802410126,
+      "learning_rate": 0.0001,
+      "loss": 1.5944,
+      "step": 170
+    },
+    {
+      "epoch": 0.05726724715338245,
+      "grad_norm": 0.4849359095096588,
+      "learning_rate": 0.0001,
+      "loss": 1.6089,
+      "step": 171
+    },
+    {
+      "epoch": 0.05760214333556597,
+      "grad_norm": 0.4833835959434509,
+      "learning_rate": 0.0001,
+      "loss": 1.5886,
+      "step": 172
+    },
+    {
+      "epoch": 0.0579370395177495,
+      "grad_norm": 0.46753743290901184,
+      "learning_rate": 0.0001,
+      "loss": 1.6548,
+      "step": 173
+    },
+    {
+      "epoch": 0.05827193569993302,
+      "grad_norm": 0.49789801239967346,
+      "learning_rate": 0.0001,
+      "loss": 1.6392,
+      "step": 174
+    },
+    {
+      "epoch": 0.058606831882116545,
+      "grad_norm": 0.46485090255737305,
+      "learning_rate": 0.0001,
+      "loss": 1.6543,
+      "step": 175
+    },
+    {
+      "epoch": 0.058941728064300064,
+      "grad_norm": 0.49583113193511963,
+      "learning_rate": 0.0001,
+      "loss": 1.6233,
+      "step": 176
+    },
+    {
+      "epoch": 0.05927662424648359,
+      "grad_norm": 0.4523816704750061,
+      "learning_rate": 0.0001,
+      "loss": 1.5896,
+      "step": 177
+    },
+    {
+      "epoch": 0.05961152042866711,
+      "grad_norm": 0.4893614649772644,
+      "learning_rate": 0.0001,
+      "loss": 1.6138,
+      "step": 178
+    },
+    {
+      "epoch": 0.059946416610850636,
+      "grad_norm": 0.47194191813468933,
+      "learning_rate": 0.0001,
+      "loss": 1.6718,
+      "step": 179
+    },
+    {
+      "epoch": 0.06028131279303416,
+      "grad_norm": 0.46762990951538086,
+      "learning_rate": 0.0001,
+      "loss": 1.6478,
+      "step": 180
+    },
+    {
+      "epoch": 0.06061620897521768,
+      "grad_norm": 0.47963568568229675,
+      "learning_rate": 0.0001,
+      "loss": 1.7146,
+      "step": 181
+    },
+    {
+      "epoch": 0.06095110515740121,
+      "grad_norm": 0.47497475147247314,
+      "learning_rate": 0.0001,
+      "loss": 1.598,
+      "step": 182
+    },
+    {
+      "epoch": 0.06128600133958473,
+      "grad_norm": 0.45387428998947144,
+      "learning_rate": 0.0001,
+      "loss": 1.5978,
+      "step": 183
+    },
+    {
+      "epoch": 0.061620897521768254,
+      "grad_norm": 0.45365142822265625,
+      "learning_rate": 0.0001,
+      "loss": 1.598,
+      "step": 184
+    },
+    {
+      "epoch": 0.061955793703951774,
+      "grad_norm": 0.4785883128643036,
+      "learning_rate": 0.0001,
+      "loss": 1.674,
+      "step": 185
+    },
+    {
+      "epoch": 0.0622906898861353,
+      "grad_norm": 0.49143221974372864,
+      "learning_rate": 0.0001,
+      "loss": 1.6021,
+      "step": 186
+    },
+    {
+      "epoch": 0.06262558606831882,
+      "grad_norm": 0.49770817160606384,
+      "learning_rate": 0.0001,
+      "loss": 1.6205,
+      "step": 187
+    },
+    {
+      "epoch": 0.06296048225050234,
+      "grad_norm": 0.4437146782875061,
+      "learning_rate": 0.0001,
+      "loss": 1.4854,
+      "step": 188
+    },
+    {
+      "epoch": 0.06329537843268587,
+      "grad_norm": 0.46671444177627563,
+      "learning_rate": 0.0001,
+      "loss": 1.6339,
+      "step": 189
+    },
+    {
+      "epoch": 0.06363027461486939,
+      "grad_norm": 0.45990461111068726,
+      "learning_rate": 0.0001,
+      "loss": 1.607,
+      "step": 190
+    },
+    {
+      "epoch": 0.06396517079705291,
+      "grad_norm": 0.49323973059654236,
+      "learning_rate": 0.0001,
+      "loss": 1.6806,
+      "step": 191
+    },
+    {
+      "epoch": 0.06430006697923643,
+      "grad_norm": 0.4419143497943878,
+      "learning_rate": 0.0001,
+      "loss": 1.569,
+      "step": 192
+    },
+    {
+      "epoch": 0.06463496316141996,
+      "grad_norm": 0.47556862235069275,
+      "learning_rate": 0.0001,
+      "loss": 1.5985,
+      "step": 193
+    },
+    {
+      "epoch": 0.06496985934360348,
+      "grad_norm": 0.462699294090271,
+      "learning_rate": 0.0001,
+      "loss": 1.6084,
+      "step": 194
+    },
+    {
+      "epoch": 0.065304755525787,
+      "grad_norm": 0.46956902742385864,
+      "learning_rate": 0.0001,
+      "loss": 1.6413,
+      "step": 195
+    },
+    {
+      "epoch": 0.06563965170797052,
+      "grad_norm": 0.4640052020549774,
+      "learning_rate": 0.0001,
+      "loss": 1.6613,
+      "step": 196
+    },
+    {
+      "epoch": 0.06597454789015406,
+      "grad_norm": 0.46174609661102295,
+      "learning_rate": 0.0001,
+      "loss": 1.6004,
+      "step": 197
+    },
+    {
+      "epoch": 0.06630944407233758,
+      "grad_norm": 0.4698941707611084,
+      "learning_rate": 0.0001,
+      "loss": 1.6381,
+      "step": 198
+    },
+    {
+      "epoch": 0.0666443402545211,
+      "grad_norm": 0.4754740297794342,
+      "learning_rate": 0.0001,
+      "loss": 1.6448,
+      "step": 199
+    },
+    {
+      "epoch": 0.06697923643670463,
+      "grad_norm": 0.4478375315666199,
+      "learning_rate": 0.0001,
+      "loss": 1.5242,
+      "step": 200
+    },
+    {
+      "epoch": 0.06731413261888815,
+      "grad_norm": 0.48765239119529724,
+      "learning_rate": 0.0001,
+      "loss": 1.6898,
+      "step": 201
+    },
+    {
+      "epoch": 0.06764902880107167,
+      "grad_norm": 0.4749129116535187,
+      "learning_rate": 0.0001,
+      "loss": 1.7313,
+      "step": 202
+    },
+    {
+      "epoch": 0.06798392498325519,
+      "grad_norm": 0.46355462074279785,
+      "learning_rate": 0.0001,
+      "loss": 1.6475,
+      "step": 203
+    },
+    {
+      "epoch": 0.06831882116543872,
+      "grad_norm": 0.46045640110969543,
+      "learning_rate": 0.0001,
+      "loss": 1.5821,
+      "step": 204
+    },
+    {
+      "epoch": 0.06865371734762224,
+      "grad_norm": 0.46230217814445496,
+      "learning_rate": 0.0001,
+      "loss": 1.532,
+      "step": 205
+    },
+    {
+      "epoch": 0.06898861352980576,
+      "grad_norm": 0.4512956142425537,
+      "learning_rate": 0.0001,
+      "loss": 1.5849,
+      "step": 206
+    },
+    {
+      "epoch": 0.06932350971198928,
+      "grad_norm": 0.4672728180885315,
+      "learning_rate": 0.0001,
+      "loss": 1.6527,
+      "step": 207
+    },
+    {
+      "epoch": 0.06965840589417281,
+      "grad_norm": 0.5102751851081848,
+      "learning_rate": 0.0001,
+      "loss": 1.6008,
+      "step": 208
+    },
+    {
+      "epoch": 0.06999330207635633,
+      "grad_norm": 0.47706693410873413,
+      "learning_rate": 0.0001,
+      "loss": 1.5342,
+      "step": 209
+    },
+    {
+      "epoch": 0.07032819825853985,
+      "grad_norm": 0.4652723968029022,
+      "learning_rate": 0.0001,
+      "loss": 1.6368,
+      "step": 210
+    },
+    {
+      "epoch": 0.07066309444072337,
+      "grad_norm": 0.4909239113330841,
+      "learning_rate": 0.0001,
+      "loss": 1.7325,
+      "step": 211
+    },
+    {
+      "epoch": 0.0709979906229069,
+      "grad_norm": 0.4498032331466675,
+      "learning_rate": 0.0001,
+      "loss": 1.6275,
+      "step": 212
+    },
+    {
+      "epoch": 0.07133288680509042,
+      "grad_norm": 0.47739377617836,
+      "learning_rate": 0.0001,
+      "loss": 1.6555,
+      "step": 213
+    },
+    {
+      "epoch": 0.07166778298727394,
+      "grad_norm": 0.4853709936141968,
+      "learning_rate": 0.0001,
+      "loss": 1.7459,
+      "step": 214
+    },
+    {
+      "epoch": 0.07200267916945746,
+      "grad_norm": 0.46574491262435913,
+      "learning_rate": 0.0001,
+      "loss": 1.6285,
+      "step": 215
+    },
+    {
+      "epoch": 0.072337575351641,
+      "grad_norm": 0.46146032214164734,
+      "learning_rate": 0.0001,
+      "loss": 1.5787,
+      "step": 216
+    },
+    {
+      "epoch": 0.07267247153382451,
+      "grad_norm": 0.4617884159088135,
+      "learning_rate": 0.0001,
+      "loss": 1.5852,
+      "step": 217
+    },
+    {
+      "epoch": 0.07300736771600803,
+      "grad_norm": 0.45636409521102905,
+      "learning_rate": 0.0001,
+      "loss": 1.6063,
+      "step": 218
+    },
+    {
+      "epoch": 0.07334226389819157,
+      "grad_norm": 0.4762187898159027,
+      "learning_rate": 0.0001,
+      "loss": 1.6394,
+      "step": 219
+    },
+    {
+      "epoch": 0.07367716008037509,
+      "grad_norm": 0.4896661937236786,
+      "learning_rate": 0.0001,
+      "loss": 1.6506,
+      "step": 220
+    },
+    {
+      "epoch": 0.0740120562625586,
+      "grad_norm": 0.46752890944480896,
+      "learning_rate": 0.0001,
+      "loss": 1.5646,
+      "step": 221
+    },
+    {
+      "epoch": 0.07434695244474213,
+      "grad_norm": 0.46463754773139954,
+      "learning_rate": 0.0001,
+      "loss": 1.4999,
+      "step": 222
+    },
+    {
+      "epoch": 0.07468184862692566,
+      "grad_norm": 0.548875093460083,
+      "learning_rate": 0.0001,
+      "loss": 1.6708,
+      "step": 223
+    },
+    {
+      "epoch": 0.07501674480910918,
+      "grad_norm": 0.47329181432724,
+      "learning_rate": 0.0001,
+      "loss": 1.6911,
+      "step": 224
+    },
+    {
+      "epoch": 0.0753516409912927,
+      "grad_norm": 0.46423107385635376,
+      "learning_rate": 0.0001,
+      "loss": 1.5469,
+      "step": 225
+    },
+    {
+      "epoch": 0.07568653717347622,
+      "grad_norm": 0.4901514947414398,
+      "learning_rate": 0.0001,
+      "loss": 1.6217,
+      "step": 226
+    },
+    {
+      "epoch": 0.07602143335565975,
+      "grad_norm": 0.5000611543655396,
+      "learning_rate": 0.0001,
+      "loss": 1.6934,
+      "step": 227
+    },
+    {
+      "epoch": 0.07635632953784327,
+      "grad_norm": 0.48425373435020447,
+      "learning_rate": 0.0001,
+      "loss": 1.6358,
+      "step": 228
+    },
+    {
+      "epoch": 0.07669122572002679,
+      "grad_norm": 0.48660895228385925,
+      "learning_rate": 0.0001,
+      "loss": 1.6489,
+      "step": 229
+    },
+    {
+      "epoch": 0.07702612190221031,
+      "grad_norm": 0.5343098640441895,
+      "learning_rate": 0.0001,
+      "loss": 1.7937,
+      "step": 230
+    },
+    {
+      "epoch": 0.07736101808439384,
+      "grad_norm": 0.44870078563690186,
+      "learning_rate": 0.0001,
+      "loss": 1.4956,
+      "step": 231
+    },
+    {
+      "epoch": 0.07769591426657736,
+      "grad_norm": 0.44342944025993347,
+      "learning_rate": 0.0001,
+      "loss": 1.4829,
+      "step": 232
+    },
+    {
+      "epoch": 0.07803081044876088,
+      "grad_norm": 0.5041643381118774,
+      "learning_rate": 0.0001,
+      "loss": 1.6187,
+      "step": 233
+    },
+    {
+      "epoch": 0.0783657066309444,
+      "grad_norm": 0.4803434908390045,
+      "learning_rate": 0.0001,
+      "loss": 1.5341,
+      "step": 234
+    },
+    {
+      "epoch": 0.07870060281312793,
+      "grad_norm": 0.4685923457145691,
+      "learning_rate": 0.0001,
+      "loss": 1.612,
+      "step": 235
+    },
+    {
+      "epoch": 0.07903549899531145,
+      "grad_norm": 0.5002148151397705,
+      "learning_rate": 0.0001,
+      "loss": 1.5885,
+      "step": 236
+    },
+    {
+      "epoch": 0.07937039517749497,
+      "grad_norm": 0.44594261050224304,
+      "learning_rate": 0.0001,
+      "loss": 1.6518,
+      "step": 237
+    },
+    {
+      "epoch": 0.07970529135967849,
+      "grad_norm": 0.4505089819431305,
+      "learning_rate": 0.0001,
+      "loss": 1.6333,
+      "step": 238
+    },
+    {
+      "epoch": 0.08004018754186203,
+      "grad_norm": 0.47187793254852295,
+      "learning_rate": 0.0001,
+      "loss": 1.6145,
+      "step": 239
+    },
+    {
+      "epoch": 0.08037508372404555,
+      "grad_norm": 0.43274542689323425,
+      "learning_rate": 0.0001,
+      "loss": 1.5611,
+      "step": 240
+    },
+    {
+      "epoch": 0.08070997990622907,
+      "grad_norm": 0.4760628938674927,
+      "learning_rate": 0.0001,
+      "loss": 1.6176,
+      "step": 241
+    },
+    {
+      "epoch": 0.0810448760884126,
+      "grad_norm": 0.4943337142467499,
+      "learning_rate": 0.0001,
+      "loss": 1.6503,
+      "step": 242
+    },
+    {
+      "epoch": 0.08137977227059612,
+      "grad_norm": 0.45220503211021423,
+      "learning_rate": 0.0001,
+      "loss": 1.5666,
+      "step": 243
+    },
+    {
+      "epoch": 0.08171466845277964,
+      "grad_norm": 0.4890030026435852,
+      "learning_rate": 0.0001,
+      "loss": 1.6096,
+      "step": 244
+    },
+    {
+      "epoch": 0.08204956463496316,
+      "grad_norm": 0.426581472158432,
+      "learning_rate": 0.0001,
+      "loss": 1.5234,
+      "step": 245
+    },
+    {
+      "epoch": 0.08238446081714669,
+      "grad_norm": 0.5227492451667786,
+      "learning_rate": 0.0001,
+      "loss": 1.7529,
+      "step": 246
+    },
+    {
+      "epoch": 0.08271935699933021,
+      "grad_norm": 0.4621161222457886,
+      "learning_rate": 0.0001,
+      "loss": 1.6967,
+      "step": 247
+    },
+    {
+      "epoch": 0.08305425318151373,
+      "grad_norm": 0.45487985014915466,
+      "learning_rate": 0.0001,
+      "loss": 1.5204,
+      "step": 248
+    },
+    {
+      "epoch": 0.08338914936369725,
+      "grad_norm": 0.4499679505825043,
+      "learning_rate": 0.0001,
+      "loss": 1.6882,
+      "step": 249
+    },
+    {
+      "epoch": 0.08372404554588078,
+      "grad_norm": 0.46019238233566284,
+      "learning_rate": 0.0001,
+      "loss": 1.5759,
+      "step": 250
+    },
+    {
+      "epoch": 0.0840589417280643,
+      "grad_norm": 0.4673830270767212,
+      "learning_rate": 0.0001,
+      "loss": 1.6,
+      "step": 251
+    },
+    {
+      "epoch": 0.08439383791024782,
+      "grad_norm": 0.4210320711135864,
+      "learning_rate": 0.0001,
+      "loss": 1.459,
+      "step": 252
+    },
+    {
+      "epoch": 0.08472873409243134,
+      "grad_norm": 0.49863752722740173,
+      "learning_rate": 0.0001,
+      "loss": 1.7663,
+      "step": 253
+    },
+    {
+      "epoch": 0.08506363027461487,
+      "grad_norm": 0.5102465748786926,
+      "learning_rate": 0.0001,
+      "loss": 1.6699,
+      "step": 254
+    },
+    {
+      "epoch": 0.0853985264567984,
+      "grad_norm": 0.45504051446914673,
+      "learning_rate": 0.0001,
+      "loss": 1.5721,
+      "step": 255
+    },
+    {
+      "epoch": 0.08573342263898191,
+      "grad_norm": 0.48080578446388245,
+      "learning_rate": 0.0001,
+      "loss": 1.5565,
+      "step": 256
+    },
+    {
+      "epoch": 0.08606831882116543,
+      "grad_norm": 0.4657333791255951,
+      "learning_rate": 0.0001,
+      "loss": 1.6291,
+      "step": 257
+    },
+    {
+      "epoch": 0.08640321500334897,
+      "grad_norm": 0.4602212905883789,
+      "learning_rate": 0.0001,
+      "loss": 1.5471,
+      "step": 258
+    },
+    {
+      "epoch": 0.08673811118553248,
+      "grad_norm": 0.4672396183013916,
+      "learning_rate": 0.0001,
+      "loss": 1.6641,
+      "step": 259
+    },
+    {
+      "epoch": 0.087073007367716,
+      "grad_norm": 0.47578176856040955,
+      "learning_rate": 0.0001,
+      "loss": 1.6204,
+      "step": 260
+    },
+    {
+      "epoch": 0.08740790354989954,
+      "grad_norm": 0.45810386538505554,
+      "learning_rate": 0.0001,
+      "loss": 1.6366,
+      "step": 261
+    },
+    {
+      "epoch": 0.08774279973208306,
+      "grad_norm": 0.4805028736591339,
+      "learning_rate": 0.0001,
+      "loss": 1.6073,
+      "step": 262
+    },
+    {
+      "epoch": 0.08807769591426658,
+      "grad_norm": 0.495786190032959,
+      "learning_rate": 0.0001,
+      "loss": 1.6871,
+      "step": 263
+    },
+    {
+      "epoch": 0.0884125920964501,
+      "grad_norm": 0.45714274048805237,
+      "learning_rate": 0.0001,
+      "loss": 1.6176,
+      "step": 264
+    },
+    {
+      "epoch": 0.08874748827863363,
+      "grad_norm": 0.4549003839492798,
+      "learning_rate": 0.0001,
+      "loss": 1.576,
+      "step": 265
+    },
+    {
+      "epoch": 0.08908238446081715,
+      "grad_norm": 0.4851682782173157,
+      "learning_rate": 0.0001,
+      "loss": 1.6564,
+      "step": 266
+    },
+    {
+      "epoch": 0.08941728064300067,
+      "grad_norm": 0.4567672908306122,
+      "learning_rate": 0.0001,
+      "loss": 1.6241,
+      "step": 267
+    },
+    {
+      "epoch": 0.08975217682518419,
+      "grad_norm": 0.45151346921920776,
+      "learning_rate": 0.0001,
+      "loss": 1.5876,
+      "step": 268
+    },
+    {
+      "epoch": 0.09008707300736772,
+      "grad_norm": 0.4828041195869446,
+      "learning_rate": 0.0001,
+      "loss": 1.5486,
+      "step": 269
+    },
+    {
+      "epoch": 0.09042196918955124,
+      "grad_norm": 0.4572749733924866,
+      "learning_rate": 0.0001,
+      "loss": 1.5367,
+      "step": 270
+    },
+    {
+      "epoch": 0.09075686537173476,
+      "grad_norm": 0.49878939986228943,
+      "learning_rate": 0.0001,
+      "loss": 1.6074,
+      "step": 271
+    },
+    {
+      "epoch": 0.09109176155391828,
+      "grad_norm": 0.4592888653278351,
+      "learning_rate": 0.0001,
+      "loss": 1.6461,
+      "step": 272
+    },
+    {
+      "epoch": 0.09142665773610181,
+      "grad_norm": 0.47039178013801575,
+      "learning_rate": 0.0001,
+      "loss": 1.6303,
+      "step": 273
+    },
+    {
+      "epoch": 0.09176155391828533,
+      "grad_norm": 0.4495210349559784,
+      "learning_rate": 0.0001,
+      "loss": 1.5465,
+      "step": 274
+    },
+    {
+      "epoch": 0.09209645010046885,
+      "grad_norm": 0.4704029858112335,
+      "learning_rate": 0.0001,
+      "loss": 1.6109,
+      "step": 275
+    },
+    {
+      "epoch": 0.09243134628265237,
+      "grad_norm": 0.44413095712661743,
+      "learning_rate": 0.0001,
+      "loss": 1.5308,
+      "step": 276
+    },
+    {
+      "epoch": 0.0927662424648359,
+      "grad_norm": 0.4834573566913605,
+      "learning_rate": 0.0001,
+      "loss": 1.6865,
+      "step": 277
+    },
+    {
+      "epoch": 0.09310113864701942,
+      "grad_norm": 0.47912439703941345,
+      "learning_rate": 0.0001,
+      "loss": 1.6228,
+      "step": 278
+    },
+    {
+      "epoch": 0.09343603482920294,
+      "grad_norm": 0.47050291299819946,
+      "learning_rate": 0.0001,
+      "loss": 1.6517,
+      "step": 279
+    },
+    {
+      "epoch": 0.09377093101138648,
+      "grad_norm": 0.4590076208114624,
+      "learning_rate": 0.0001,
+      "loss": 1.6135,
+      "step": 280
+    },
+    {
+      "epoch": 0.09410582719357,
+      "grad_norm": 0.4425355792045593,
+      "learning_rate": 0.0001,
+      "loss": 1.5617,
+      "step": 281
+    },
+    {
+      "epoch": 0.09444072337575352,
+      "grad_norm": 0.4535706043243408,
+      "learning_rate": 0.0001,
+      "loss": 1.6319,
+      "step": 282
+    },
+    {
+      "epoch": 0.09477561955793704,
+      "grad_norm": 0.4804530441761017,
+      "learning_rate": 0.0001,
+      "loss": 1.623,
+      "step": 283
+    },
+    {
+      "epoch": 0.09511051574012057,
+      "grad_norm": 0.4477209448814392,
+      "learning_rate": 0.0001,
+      "loss": 1.5297,
+      "step": 284
+    },
+    {
+      "epoch": 0.09544541192230409,
+      "grad_norm": 0.4630716145038605,
+      "learning_rate": 0.0001,
+      "loss": 1.6254,
+      "step": 285
+    },
+    {
+      "epoch": 0.09578030810448761,
+      "grad_norm": 0.4582269787788391,
+      "learning_rate": 0.0001,
+      "loss": 1.5493,
+      "step": 286
+    },
+    {
+      "epoch": 0.09611520428667113,
+      "grad_norm": 0.460605263710022,
+      "learning_rate": 0.0001,
+      "loss": 1.5757,
+      "step": 287
+    },
+    {
+      "epoch": 0.09645010046885466,
+      "grad_norm": 0.4452902376651764,
+      "learning_rate": 0.0001,
+      "loss": 1.5142,
+      "step": 288
+    },
+    {
+      "epoch": 0.09678499665103818,
+      "grad_norm": 0.46841830015182495,
+      "learning_rate": 0.0001,
+      "loss": 1.6504,
+      "step": 289
+    },
+    {
+      "epoch": 0.0971198928332217,
+      "grad_norm": 0.4708998203277588,
+      "learning_rate": 0.0001,
+      "loss": 1.6167,
+      "step": 290
+    },
+    {
+      "epoch": 0.09745478901540522,
+      "grad_norm": 0.47076770663261414,
+      "learning_rate": 0.0001,
+      "loss": 1.5892,
+      "step": 291
+    },
+    {
+      "epoch": 0.09778968519758875,
+      "grad_norm": 0.4641210436820984,
+      "learning_rate": 0.0001,
+      "loss": 1.6524,
+      "step": 292
+    },
+    {
+      "epoch": 0.09812458137977227,
+      "grad_norm": 0.45254385471343994,
+      "learning_rate": 0.0001,
+      "loss": 1.5052,
+      "step": 293
+    },
+    {
+      "epoch": 0.09845947756195579,
+      "grad_norm": 0.4456687569618225,
+      "learning_rate": 0.0001,
+      "loss": 1.5605,
+      "step": 294
+    },
+    {
+      "epoch": 0.09879437374413931,
+      "grad_norm": 0.4839523136615753,
+      "learning_rate": 0.0001,
+      "loss": 1.7271,
+      "step": 295
+    },
+    {
+      "epoch": 0.09912926992632284,
+      "grad_norm": 0.44768890738487244,
+      "learning_rate": 0.0001,
+      "loss": 1.543,
+      "step": 296
+    },
+    {
+      "epoch": 0.09946416610850636,
+      "grad_norm": 0.4762071669101715,
+      "learning_rate": 0.0001,
+      "loss": 1.5863,
+      "step": 297
+    },
+    {
+      "epoch": 0.09979906229068988,
+      "grad_norm": 0.44802209734916687,
+      "learning_rate": 0.0001,
+      "loss": 1.6379,
+      "step": 298
+    },
+    {
+      "epoch": 0.10013395847287342,
+      "grad_norm": 0.4678582549095154,
+      "learning_rate": 0.0001,
+      "loss": 1.4977,
+      "step": 299
+    },
+    {
+      "epoch": 0.10046885465505694,
+      "grad_norm": 0.44217634201049805,
+      "learning_rate": 0.0001,
+      "loss": 1.5518,
+      "step": 300
+    },
+    {
+      "epoch": 0.10080375083724046,
+      "grad_norm": 0.4706065356731415,
+      "learning_rate": 0.0001,
+      "loss": 1.5462,
+      "step": 301
+    },
+    {
+      "epoch": 0.10113864701942397,
+      "grad_norm": 0.4437488913536072,
+      "learning_rate": 0.0001,
+      "loss": 1.5496,
+      "step": 302
+    },
+    {
+      "epoch": 0.10147354320160751,
+      "grad_norm": 0.5229858160018921,
+      "learning_rate": 0.0001,
+      "loss": 1.6437,
+      "step": 303
+    },
+    {
+      "epoch": 0.10180843938379103,
+      "grad_norm": 0.44042670726776123,
+      "learning_rate": 0.0001,
+      "loss": 1.5646,
+      "step": 304
+    },
+    {
+      "epoch": 0.10214333556597455,
+      "grad_norm": 0.4837404787540436,
+      "learning_rate": 0.0001,
+      "loss": 1.6794,
+      "step": 305
+    },
+    {
+      "epoch": 0.10247823174815807,
+      "grad_norm": 0.4524439871311188,
+      "learning_rate": 0.0001,
+      "loss": 1.6054,
+      "step": 306
+    },
+    {
+      "epoch": 0.1028131279303416,
+      "grad_norm": 0.47352391481399536,
+      "learning_rate": 0.0001,
+      "loss": 1.5684,
+      "step": 307
+    },
+    {
+      "epoch": 0.10314802411252512,
+      "grad_norm": 0.48607951402664185,
+      "learning_rate": 0.0001,
+      "loss": 1.6231,
+      "step": 308
+    },
+    {
+      "epoch": 0.10348292029470864,
+      "grad_norm": 0.49019041657447815,
+      "learning_rate": 0.0001,
+      "loss": 1.6256,
+      "step": 309
+    },
+    {
+      "epoch": 0.10381781647689216,
+      "grad_norm": 0.4620254933834076,
+      "learning_rate": 0.0001,
+      "loss": 1.6367,
+      "step": 310
+    },
+    {
+      "epoch": 0.10415271265907569,
+      "grad_norm": 0.429916113615036,
+      "learning_rate": 0.0001,
+      "loss": 1.5171,
+      "step": 311
+    },
+    {
+      "epoch": 0.10448760884125921,
+      "grad_norm": 0.4744836688041687,
+      "learning_rate": 0.0001,
+      "loss": 1.624,
+      "step": 312
+    },
+    {
+      "epoch": 0.10482250502344273,
+      "grad_norm": 0.43478068709373474,
+      "learning_rate": 0.0001,
+      "loss": 1.5532,
+      "step": 313
+    },
+    {
+      "epoch": 0.10515740120562625,
+      "grad_norm": 0.5063771605491638,
+      "learning_rate": 0.0001,
+      "loss": 1.628,
+      "step": 314
+    },
+    {
+      "epoch": 0.10549229738780978,
+      "grad_norm": 0.4402046799659729,
+      "learning_rate": 0.0001,
+      "loss": 1.5388,
+      "step": 315
+    },
+    {
+      "epoch": 0.1058271935699933,
+      "grad_norm": 0.46563291549682617,
+      "learning_rate": 0.0001,
+      "loss": 1.6603,
+      "step": 316
+    },
+    {
+      "epoch": 0.10616208975217682,
+      "grad_norm": 0.4675760269165039,
+      "learning_rate": 0.0001,
+      "loss": 1.6199,
+      "step": 317
+    },
+    {
+      "epoch": 0.10649698593436034,
+      "grad_norm": 0.46061259508132935,
+      "learning_rate": 0.0001,
+      "loss": 1.6721,
+      "step": 318
+    },
+    {
+      "epoch": 0.10683188211654387,
+      "grad_norm": 0.45733362436294556,
+      "learning_rate": 0.0001,
+      "loss": 1.6237,
+      "step": 319
+    },
+    {
+      "epoch": 0.1071667782987274,
+      "grad_norm": 0.4503006935119629,
+      "learning_rate": 0.0001,
+      "loss": 1.5131,
+      "step": 320
+    },
+    {
+      "epoch": 0.10750167448091091,
+      "grad_norm": 0.4447789788246155,
+      "learning_rate": 0.0001,
+      "loss": 1.4741,
+      "step": 321
+    },
+    {
+      "epoch": 0.10783657066309445,
+      "grad_norm": 0.48432567715644836,
+      "learning_rate": 0.0001,
+      "loss": 1.6338,
+      "step": 322
+    },
+    {
+      "epoch": 0.10817146684527797,
+      "grad_norm": 0.46932923793792725,
+      "learning_rate": 0.0001,
+      "loss": 1.6176,
+      "step": 323
+    },
+    {
+      "epoch": 0.10850636302746149,
+      "grad_norm": 0.4636939764022827,
+      "learning_rate": 0.0001,
+      "loss": 1.5967,
+      "step": 324
+    },
+    {
+      "epoch": 0.108841259209645,
+      "grad_norm": 0.47708818316459656,
+      "learning_rate": 0.0001,
+      "loss": 1.6065,
+      "step": 325
+    },
+    {
+      "epoch": 0.10917615539182854,
+      "grad_norm": 0.43039318919181824,
+      "learning_rate": 0.0001,
+      "loss": 1.586,
+      "step": 326
+    },
+    {
+      "epoch": 0.10951105157401206,
+      "grad_norm": 0.46356332302093506,
+      "learning_rate": 0.0001,
+      "loss": 1.5193,
+      "step": 327
+    },
+    {
+      "epoch": 0.10984594775619558,
+      "grad_norm": 0.4567822515964508,
+      "learning_rate": 0.0001,
+      "loss": 1.607,
+      "step": 328
+    },
+    {
+      "epoch": 0.1101808439383791,
+      "grad_norm": 0.4588714838027954,
+      "learning_rate": 0.0001,
+      "loss": 1.5343,
+      "step": 329
+    },
+    {
+      "epoch": 0.11051574012056263,
+      "grad_norm": 0.48270532488822937,
+      "learning_rate": 0.0001,
+      "loss": 1.6282,
+      "step": 330
+    },
+    {
+      "epoch": 0.11085063630274615,
+      "grad_norm": 0.45466941595077515,
+      "learning_rate": 0.0001,
+      "loss": 1.6338,
+      "step": 331
+    },
+    {
+      "epoch": 0.11118553248492967,
+      "grad_norm": 0.4509162902832031,
+      "learning_rate": 0.0001,
+      "loss": 1.602,
+      "step": 332
+    },
+    {
+      "epoch": 0.11152042866711319,
+      "grad_norm": 0.4739763140678406,
+      "learning_rate": 0.0001,
+      "loss": 1.6797,
+      "step": 333
+    },
+    {
+      "epoch": 0.11185532484929672,
+      "grad_norm": 0.4735789895057678,
+      "learning_rate": 0.0001,
+      "loss": 1.6173,
+      "step": 334
+    },
+    {
+      "epoch": 0.11219022103148024,
+      "grad_norm": 0.4974208176136017,
+      "learning_rate": 0.0001,
+      "loss": 1.6478,
+      "step": 335
+    },
+    {
+      "epoch": 0.11252511721366376,
+      "grad_norm": 0.428434282541275,
+      "learning_rate": 0.0001,
+      "loss": 1.4542,
+      "step": 336
+    },
+    {
+      "epoch": 0.11286001339584728,
+      "grad_norm": 0.4694981575012207,
+      "learning_rate": 0.0001,
+      "loss": 1.6672,
+      "step": 337
+    },
+    {
+      "epoch": 0.11319490957803081,
+      "grad_norm": 0.4457933306694031,
+      "learning_rate": 0.0001,
+      "loss": 1.5994,
+      "step": 338
+    },
+    {
+      "epoch": 0.11352980576021433,
+      "grad_norm": 0.4311225414276123,
+      "learning_rate": 0.0001,
+      "loss": 1.4973,
+      "step": 339
+    },
+    {
+      "epoch": 0.11386470194239785,
+      "grad_norm": 0.4609581232070923,
+      "learning_rate": 0.0001,
+      "loss": 1.5811,
+      "step": 340
+    },
+    {
+      "epoch": 0.11419959812458139,
+      "grad_norm": 0.4629519581794739,
+      "learning_rate": 0.0001,
+      "loss": 1.6941,
+      "step": 341
+    },
+    {
+      "epoch": 0.1145344943067649,
+      "grad_norm": 0.4684126377105713,
+      "learning_rate": 0.0001,
+      "loss": 1.6668,
+      "step": 342
+    },
+    {
+      "epoch": 0.11486939048894843,
+      "grad_norm": 0.4475191533565521,
+      "learning_rate": 0.0001,
+      "loss": 1.5619,
+      "step": 343
+    },
+    {
+      "epoch": 0.11520428667113194,
+      "grad_norm": 0.45096057653427124,
+      "learning_rate": 0.0001,
+      "loss": 1.6197,
+      "step": 344
+    },
+    {
+      "epoch": 0.11553918285331548,
+      "grad_norm": 0.44147443771362305,
+      "learning_rate": 0.0001,
+      "loss": 1.556,
+      "step": 345
+    },
+    {
+      "epoch": 0.115874079035499,
+      "grad_norm": 0.4539279639720917,
+      "learning_rate": 0.0001,
+      "loss": 1.6454,
+      "step": 346
+    },
+    {
+      "epoch": 0.11620897521768252,
+      "grad_norm": 0.44900423288345337,
+      "learning_rate": 0.0001,
+      "loss": 1.6007,
+      "step": 347
+    },
+    {
+      "epoch": 0.11654387139986604,
+      "grad_norm": 0.4491700828075409,
+      "learning_rate": 0.0001,
+      "loss": 1.6662,
+      "step": 348
+    },
+    {
+      "epoch": 0.11687876758204957,
+      "grad_norm": 0.4434390962123871,
+      "learning_rate": 0.0001,
+      "loss": 1.6097,
+      "step": 349
+    },
+    {
+      "epoch": 0.11721366376423309,
+      "grad_norm": 0.47392427921295166,
+      "learning_rate": 0.0001,
+      "loss": 1.7676,
+      "step": 350
+    },
+    {
+      "epoch": 0.11754855994641661,
+      "grad_norm": 0.4450167417526245,
+      "learning_rate": 0.0001,
+      "loss": 1.6478,
+      "step": 351
+    },
+    {
+      "epoch": 0.11788345612860013,
+      "grad_norm": 0.44741788506507874,
+      "learning_rate": 0.0001,
+      "loss": 1.4809,
+      "step": 352
+    },
+    {
+      "epoch": 0.11821835231078366,
+      "grad_norm": 0.4440721869468689,
+      "learning_rate": 0.0001,
+      "loss": 1.5977,
+      "step": 353
+    },
+    {
+      "epoch": 0.11855324849296718,
+      "grad_norm": 0.4559003710746765,
+      "learning_rate": 0.0001,
+      "loss": 1.5861,
+      "step": 354
+    },
+    {
+      "epoch": 0.1188881446751507,
+      "grad_norm": 0.4684537649154663,
+      "learning_rate": 0.0001,
+      "loss": 1.6041,
+      "step": 355
+    },
+    {
+      "epoch": 0.11922304085733422,
+      "grad_norm": 0.44140082597732544,
+      "learning_rate": 0.0001,
+      "loss": 1.6321,
+      "step": 356
+    },
+    {
+      "epoch": 0.11955793703951775,
+      "grad_norm": 0.4456671178340912,
+      "learning_rate": 0.0001,
+      "loss": 1.6585,
+      "step": 357
+    },
+    {
+      "epoch": 0.11989283322170127,
+      "grad_norm": 0.4394270181655884,
+      "learning_rate": 0.0001,
+      "loss": 1.6098,
+      "step": 358
+    },
+    {
+      "epoch": 0.12022772940388479,
+      "grad_norm": 0.46983978152275085,
+      "learning_rate": 0.0001,
+      "loss": 1.5997,
+      "step": 359
+    },
+    {
+      "epoch": 0.12056262558606833,
+      "grad_norm": 0.4956510663032532,
+      "learning_rate": 0.0001,
+      "loss": 1.6731,
+      "step": 360
+    },
+    {
+      "epoch": 0.12089752176825184,
+      "grad_norm": 0.47455278038978577,
+      "learning_rate": 0.0001,
+      "loss": 1.6163,
+      "step": 361
+    },
+    {
+      "epoch": 0.12123241795043536,
+      "grad_norm": 0.4317326247692108,
+      "learning_rate": 0.0001,
+      "loss": 1.4407,
+      "step": 362
+    },
+    {
+      "epoch": 0.12156731413261888,
+      "grad_norm": 0.4892672300338745,
+      "learning_rate": 0.0001,
+      "loss": 1.6077,
+      "step": 363
+    },
+    {
+      "epoch": 0.12190221031480242,
+      "grad_norm": 0.44915369153022766,
+      "learning_rate": 0.0001,
+      "loss": 1.5941,
+      "step": 364
+    },
+    {
+      "epoch": 0.12223710649698594,
+      "grad_norm": 0.45726680755615234,
+      "learning_rate": 0.0001,
+      "loss": 1.588,
+      "step": 365
+    },
+    {
+      "epoch": 0.12257200267916946,
+      "grad_norm": 0.4513232409954071,
+      "learning_rate": 0.0001,
+      "loss": 1.6027,
+      "step": 366
+    },
+    {
+      "epoch": 0.12290689886135298,
+      "grad_norm": 0.473565012216568,
+      "learning_rate": 0.0001,
+      "loss": 1.6643,
+      "step": 367
+    },
+    {
+      "epoch": 0.12324179504353651,
+      "grad_norm": 0.48082560300827026,
+      "learning_rate": 0.0001,
+      "loss": 1.5707,
+      "step": 368
+    },
+    {
+      "epoch": 0.12357669122572003,
+      "grad_norm": 0.44344571232795715,
+      "learning_rate": 0.0001,
+      "loss": 1.5762,
+      "step": 369
+    },
+    {
+      "epoch": 0.12391158740790355,
+      "grad_norm": 0.4536781311035156,
+      "learning_rate": 0.0001,
+      "loss": 1.6305,
+      "step": 370
+    },
+    {
+      "epoch": 0.12424648359008707,
+      "grad_norm": 0.4507012367248535,
+      "learning_rate": 0.0001,
+      "loss": 1.6478,
+      "step": 371
+    },
+    {
+      "epoch": 0.1245813797722706,
+      "grad_norm": 0.4527163505554199,
+      "learning_rate": 0.0001,
+      "loss": 1.5439,
+      "step": 372
+    },
+    {
+      "epoch": 0.12491627595445412,
+      "grad_norm": 0.45644611120224,
+      "learning_rate": 0.0001,
+      "loss": 1.6578,
+      "step": 373
+    },
+    {
+      "epoch": 0.12525117213663764,
+      "grad_norm": 0.48362550139427185,
+      "learning_rate": 0.0001,
+      "loss": 1.6283,
+      "step": 374
+    },
+    {
+      "epoch": 0.12558606831882116,
+      "grad_norm": 0.4587193429470062,
+      "learning_rate": 0.0001,
+      "loss": 1.604,
+      "step": 375
+    },
+    {
+      "epoch": 0.12592096450100468,
+      "grad_norm": 0.5075493454933167,
+      "learning_rate": 0.0001,
+      "loss": 1.653,
+      "step": 376
+    },
+    {
+      "epoch": 0.1262558606831882,
+      "grad_norm": 0.4548470675945282,
+      "learning_rate": 0.0001,
+      "loss": 1.6488,
+      "step": 377
+    },
+    {
+      "epoch": 0.12659075686537175,
+      "grad_norm": 0.49109938740730286,
+      "learning_rate": 0.0001,
+      "loss": 1.5358,
+      "step": 378
+    },
+    {
+      "epoch": 0.12692565304755526,
+      "grad_norm": 0.4333401322364807,
+      "learning_rate": 0.0001,
+      "loss": 1.5494,
+      "step": 379
+    },
+    {
+      "epoch": 0.12726054922973878,
+      "grad_norm": 0.4399802088737488,
+      "learning_rate": 0.0001,
+      "loss": 1.4847,
+      "step": 380
+    },
+    {
+      "epoch": 0.1275954454119223,
+      "grad_norm": 0.467268168926239,
+      "learning_rate": 0.0001,
+      "loss": 1.5548,
+      "step": 381
+    },
+    {
+      "epoch": 0.12793034159410582,
+      "grad_norm": 0.4567405581474304,
+      "learning_rate": 0.0001,
+      "loss": 1.5626,
+      "step": 382
+    },
+    {
+      "epoch": 0.12826523777628934,
+      "grad_norm": 0.4765430688858032,
+      "learning_rate": 0.0001,
+      "loss": 1.6036,
+      "step": 383
+    },
+    {
+      "epoch": 0.12860013395847286,
+      "grad_norm": 0.46208590269088745,
+      "learning_rate": 0.0001,
+      "loss": 1.6427,
+      "step": 384
+    },
+    {
+      "epoch": 0.1289350301406564,
+      "grad_norm": 0.47394025325775146,
+      "learning_rate": 0.0001,
+      "loss": 1.562,
+      "step": 385
+    },
+    {
+      "epoch": 0.12926992632283993,
+      "grad_norm": 0.4401998221874237,
+      "learning_rate": 0.0001,
+      "loss": 1.6174,
+      "step": 386
+    },
+    {
+      "epoch": 0.12960482250502345,
+      "grad_norm": 0.4623987078666687,
+      "learning_rate": 0.0001,
+      "loss": 1.5635,
+      "step": 387
+    },
+    {
+      "epoch": 0.12993971868720697,
+      "grad_norm": 0.459249883890152,
+      "learning_rate": 0.0001,
+      "loss": 1.6743,
+      "step": 388
+    },
+    {
+      "epoch": 0.1302746148693905,
+      "grad_norm": 0.4397091567516327,
+      "learning_rate": 0.0001,
+      "loss": 1.5132,
+      "step": 389
+    },
+    {
+      "epoch": 0.130609511051574,
+      "grad_norm": 0.4568580687046051,
+      "learning_rate": 0.0001,
+      "loss": 1.5375,
+      "step": 390
+    },
+    {
+      "epoch": 0.13094440723375753,
+      "grad_norm": 0.42559221386909485,
+      "learning_rate": 0.0001,
+      "loss": 1.5235,
+      "step": 391
+    },
+    {
+      "epoch": 0.13127930341594105,
+      "grad_norm": 0.4833087921142578,
+      "learning_rate": 0.0001,
+      "loss": 1.6442,
+      "step": 392
+    },
+    {
+      "epoch": 0.1316141995981246,
+      "grad_norm": 0.43460625410079956,
+      "learning_rate": 0.0001,
+      "loss": 1.5806,
+      "step": 393
+    },
+    {
+      "epoch": 0.1319490957803081,
+      "grad_norm": 0.44238537549972534,
+      "learning_rate": 0.0001,
+      "loss": 1.5061,
+      "step": 394
+    },
+    {
+      "epoch": 0.13228399196249163,
+      "grad_norm": 0.4920506775379181,
+      "learning_rate": 0.0001,
+      "loss": 1.7213,
+      "step": 395
+    },
+    {
+      "epoch": 0.13261888814467515,
+      "grad_norm": 0.4509977102279663,
+      "learning_rate": 0.0001,
+      "loss": 1.611,
+      "step": 396
+    },
+    {
+      "epoch": 0.13295378432685867,
+      "grad_norm": 0.47244614362716675,
+      "learning_rate": 0.0001,
+      "loss": 1.6666,
+      "step": 397
+    },
+    {
+      "epoch": 0.1332886805090422,
+      "grad_norm": 0.4667363166809082,
+      "learning_rate": 0.0001,
+      "loss": 1.5529,
+      "step": 398
+    },
+    {
+      "epoch": 0.1336235766912257,
+      "grad_norm": 0.4608834683895111,
+      "learning_rate": 0.0001,
+      "loss": 1.6098,
+      "step": 399
+    },
+    {
+      "epoch": 0.13395847287340926,
+      "grad_norm": 0.431717187166214,
+      "learning_rate": 0.0001,
+      "loss": 1.5438,
+      "step": 400
+    },
+    {
+      "epoch": 0.13429336905559278,
+      "grad_norm": 0.5081926584243774,
+      "learning_rate": 0.0001,
+      "loss": 1.7707,
+      "step": 401
+    },
+    {
+      "epoch": 0.1346282652377763,
+      "grad_norm": 0.4610143005847931,
+      "learning_rate": 0.0001,
+      "loss": 1.6355,
+      "step": 402
+    },
+    {
+      "epoch": 0.13496316141995982,
+      "grad_norm": 0.4362616539001465,
+      "learning_rate": 0.0001,
+      "loss": 1.5324,
+      "step": 403
+    },
+    {
+      "epoch": 0.13529805760214333,
+      "grad_norm": 0.4362455904483795,
+      "learning_rate": 0.0001,
+      "loss": 1.6054,
+      "step": 404
+    },
+    {
+      "epoch": 0.13563295378432685,
+      "grad_norm": 0.443097323179245,
+      "learning_rate": 0.0001,
+      "loss": 1.5388,
+      "step": 405
+    },
+    {
+      "epoch": 0.13596784996651037,
+      "grad_norm": 0.4276910424232483,
+      "learning_rate": 0.0001,
+      "loss": 1.501,
+      "step": 406
+    },
+    {
+      "epoch": 0.1363027461486939,
+      "grad_norm": 0.4446798861026764,
+      "learning_rate": 0.0001,
+      "loss": 1.5247,
+      "step": 407
+    },
+    {
+      "epoch": 0.13663764233087744,
+      "grad_norm": 0.4959389567375183,
+      "learning_rate": 0.0001,
+      "loss": 1.6415,
+      "step": 408
+    },
+    {
+      "epoch": 0.13697253851306096,
+      "grad_norm": 0.4616144001483917,
+      "learning_rate": 0.0001,
+      "loss": 1.6574,
+      "step": 409
+    },
+    {
+      "epoch": 0.13730743469524448,
+      "grad_norm": 0.46740710735321045,
+      "learning_rate": 0.0001,
+      "loss": 1.5424,
+      "step": 410
+    },
+    {
+      "epoch": 0.137642330877428,
+      "grad_norm": 0.43289265036582947,
+      "learning_rate": 0.0001,
+      "loss": 1.528,
+      "step": 411
+    },
+    {
+      "epoch": 0.13797722705961152,
+      "grad_norm": 0.4461548924446106,
+      "learning_rate": 0.0001,
+      "loss": 1.5542,
+      "step": 412
+    },
+    {
+      "epoch": 0.13831212324179504,
+      "grad_norm": 0.4627973139286041,
+      "learning_rate": 0.0001,
+      "loss": 1.6688,
+      "step": 413
+    },
+    {
+      "epoch": 0.13864701942397856,
+      "grad_norm": 0.4514913856983185,
+      "learning_rate": 0.0001,
+      "loss": 1.5679,
+      "step": 414
+    },
+    {
+      "epoch": 0.13898191560616208,
+      "grad_norm": 0.4539605975151062,
+      "learning_rate": 0.0001,
+      "loss": 1.5908,
+      "step": 415
+    },
+    {
+      "epoch": 0.13931681178834562,
+      "grad_norm": 0.435198038816452,
+      "learning_rate": 0.0001,
+      "loss": 1.5868,
+      "step": 416
+    },
+    {
+      "epoch": 0.13965170797052914,
+      "grad_norm": 0.47620803117752075,
+      "learning_rate": 0.0001,
+      "loss": 1.537,
+      "step": 417
+    },
+    {
+      "epoch": 0.13998660415271266,
+      "grad_norm": 0.4514625668525696,
+      "learning_rate": 0.0001,
+      "loss": 1.6141,
+      "step": 418
+    },
+    {
+      "epoch": 0.14032150033489618,
+      "grad_norm": 0.49498939514160156,
+      "learning_rate": 0.0001,
+      "loss": 1.5555,
+      "step": 419
+    },
+    {
+      "epoch": 0.1406563965170797,
+      "grad_norm": 0.44335290789604187,
+      "learning_rate": 0.0001,
+      "loss": 1.5191,
+      "step": 420
+    },
+    {
+      "epoch": 0.14099129269926322,
+      "grad_norm": 0.49824535846710205,
+      "learning_rate": 0.0001,
+      "loss": 1.6201,
+      "step": 421
+    },
+    {
+      "epoch": 0.14132618888144674,
+      "grad_norm": 0.4429181218147278,
+      "learning_rate": 0.0001,
+      "loss": 1.6018,
+      "step": 422
+    },
+    {
+      "epoch": 0.1416610850636303,
+      "grad_norm": 0.4757002294063568,
+      "learning_rate": 0.0001,
+      "loss": 1.5436,
+      "step": 423
+    },
+    {
+      "epoch": 0.1419959812458138,
+      "grad_norm": 0.4324745237827301,
+      "learning_rate": 0.0001,
+      "loss": 1.5126,
+      "step": 424
+    },
+    {
+      "epoch": 0.14233087742799733,
+      "grad_norm": 0.4601553678512573,
+      "learning_rate": 0.0001,
+      "loss": 1.6138,
+      "step": 425
+    },
+    {
+      "epoch": 0.14266577361018085,
+      "grad_norm": 0.4711107611656189,
+      "learning_rate": 0.0001,
+      "loss": 1.5933,
+      "step": 426
+    },
+    {
+      "epoch": 0.14300066979236437,
+      "grad_norm": 0.4817027747631073,
+      "learning_rate": 0.0001,
+      "loss": 1.5768,
+      "step": 427
+    },
+    {
+      "epoch": 0.14333556597454788,
+      "grad_norm": 0.44300511479377747,
+      "learning_rate": 0.0001,
+      "loss": 1.553,
+      "step": 428
+    },
+    {
+      "epoch": 0.1436704621567314,
+      "grad_norm": 0.4497964084148407,
+      "learning_rate": 0.0001,
+      "loss": 1.61,
+      "step": 429
+    },
+    {
+      "epoch": 0.14400535833891492,
+      "grad_norm": 0.4525448977947235,
+      "learning_rate": 0.0001,
+      "loss": 1.5809,
+      "step": 430
+    },
+    {
+      "epoch": 0.14434025452109847,
+      "grad_norm": 0.45168110728263855,
+      "learning_rate": 0.0001,
+      "loss": 1.6471,
+      "step": 431
+    },
+    {
+      "epoch": 0.144675150703282,
+      "grad_norm": 0.4560868740081787,
+      "learning_rate": 0.0001,
+      "loss": 1.5933,
+      "step": 432
+    },
+    {
+      "epoch": 0.1450100468854655,
+      "grad_norm": 0.4391075670719147,
+      "learning_rate": 0.0001,
+      "loss": 1.5445,
+      "step": 433
+    },
+    {
+      "epoch": 0.14534494306764903,
+      "grad_norm": 0.44584378600120544,
+      "learning_rate": 0.0001,
+      "loss": 1.587,
+      "step": 434
+    },
+    {
+      "epoch": 0.14567983924983255,
+      "grad_norm": 0.4544006288051605,
+      "learning_rate": 0.0001,
+      "loss": 1.5944,
+      "step": 435
+    },
+    {
+      "epoch": 0.14601473543201607,
+      "grad_norm": 0.41910088062286377,
+      "learning_rate": 0.0001,
+      "loss": 1.5533,
+      "step": 436
+    },
+    {
+      "epoch": 0.1463496316141996,
+      "grad_norm": 0.4487232565879822,
+      "learning_rate": 0.0001,
+      "loss": 1.5474,
+      "step": 437
+    },
+    {
+      "epoch": 0.14668452779638314,
+      "grad_norm": 0.44522252678871155,
+      "learning_rate": 0.0001,
+      "loss": 1.5708,
+      "step": 438
+    },
+    {
+      "epoch": 0.14701942397856665,
+      "grad_norm": 0.4506109058856964,
+      "learning_rate": 0.0001,
+      "loss": 1.6352,
+      "step": 439
+    },
+    {
+      "epoch": 0.14735432016075017,
+      "grad_norm": 0.49480026960372925,
+      "learning_rate": 0.0001,
+      "loss": 1.5882,
+      "step": 440
+    },
+    {
+      "epoch": 0.1476892163429337,
+      "grad_norm": 0.45891493558883667,
+      "learning_rate": 0.0001,
+      "loss": 1.5441,
+      "step": 441
+    },
+    {
+      "epoch": 0.1480241125251172,
+      "grad_norm": 0.46245646476745605,
+      "learning_rate": 0.0001,
+      "loss": 1.5591,
+      "step": 442
+    },
+    {
+      "epoch": 0.14835900870730073,
+      "grad_norm": 0.44219061732292175,
+      "learning_rate": 0.0001,
+      "loss": 1.5168,
+      "step": 443
+    },
+    {
+      "epoch": 0.14869390488948425,
+      "grad_norm": 0.4645799398422241,
+      "learning_rate": 0.0001,
+      "loss": 1.6928,
+      "step": 444
+    },
+    {
+      "epoch": 0.14902880107166777,
+      "grad_norm": 0.48340848088264465,
+      "learning_rate": 0.0001,
+      "loss": 1.6647,
+      "step": 445
+    },
+    {
+      "epoch": 0.14936369725385132,
+      "grad_norm": 0.45820561051368713,
+      "learning_rate": 0.0001,
+      "loss": 1.6125,
+      "step": 446
+    },
+    {
+      "epoch": 0.14969859343603484,
+      "grad_norm": 0.46245497465133667,
+      "learning_rate": 0.0001,
+      "loss": 1.5606,
+      "step": 447
+    },
+    {
+      "epoch": 0.15003348961821836,
+      "grad_norm": 0.4639655351638794,
+      "learning_rate": 0.0001,
+      "loss": 1.62,
+      "step": 448
+    },
+    {
+      "epoch": 0.15036838580040188,
+      "grad_norm": 0.43300825357437134,
+      "learning_rate": 0.0001,
+      "loss": 1.5333,
+      "step": 449
+    },
+    {
+      "epoch": 0.1507032819825854,
+      "grad_norm": 0.44144105911254883,
+      "learning_rate": 0.0001,
+      "loss": 1.6079,
+      "step": 450
+    },
+    {
+      "epoch": 0.15103817816476892,
+      "grad_norm": 0.45080748200416565,
+      "learning_rate": 0.0001,
+      "loss": 1.662,
+      "step": 451
+    },
+    {
+      "epoch": 0.15137307434695244,
+      "grad_norm": 0.4435412287712097,
+      "learning_rate": 0.0001,
+      "loss": 1.5376,
+      "step": 452
+    },
+    {
+      "epoch": 0.15170797052913595,
+      "grad_norm": 0.4712364077568054,
+      "learning_rate": 0.0001,
+      "loss": 1.5347,
+      "step": 453
+    },
+    {
+      "epoch": 0.1520428667113195,
+      "grad_norm": 0.6480394601821899,
+      "learning_rate": 0.0001,
+      "loss": 1.536,
+      "step": 454
+    },
+    {
+      "epoch": 0.15237776289350302,
+      "grad_norm": 0.4520147740840912,
+      "learning_rate": 0.0001,
+      "loss": 1.5105,
+      "step": 455
+    },
+    {
+      "epoch": 0.15271265907568654,
+      "grad_norm": 0.47112977504730225,
+      "learning_rate": 0.0001,
+      "loss": 1.5844,
+      "step": 456
+    },
+    {
+      "epoch": 0.15304755525787006,
+      "grad_norm": 0.42187485098838806,
+      "learning_rate": 0.0001,
+      "loss": 1.4871,
+      "step": 457
+    },
+    {
+      "epoch": 0.15338245144005358,
+      "grad_norm": 0.466097891330719,
+      "learning_rate": 0.0001,
+      "loss": 1.656,
+      "step": 458
+    },
+    {
+      "epoch": 0.1537173476222371,
+      "grad_norm": 0.48392733931541443,
+      "learning_rate": 0.0001,
+      "loss": 1.7219,
+      "step": 459
+    },
+    {
+      "epoch": 0.15405224380442062,
+      "grad_norm": 0.45683392882347107,
+      "learning_rate": 0.0001,
+      "loss": 1.673,
+      "step": 460
+    },
+    {
+      "epoch": 0.15438713998660417,
+      "grad_norm": 0.4448898434638977,
+      "learning_rate": 0.0001,
+      "loss": 1.5867,
+      "step": 461
+    },
+    {
+      "epoch": 0.15472203616878769,
+      "grad_norm": 0.44210925698280334,
+      "learning_rate": 0.0001,
+      "loss": 1.5485,
+      "step": 462
+    },
+    {
+      "epoch": 0.1550569323509712,
+      "grad_norm": 0.4557957649230957,
+      "learning_rate": 0.0001,
+      "loss": 1.6391,
+      "step": 463
+    },
+    {
+      "epoch": 0.15539182853315472,
+      "grad_norm": 0.44259247183799744,
+      "learning_rate": 0.0001,
+      "loss": 1.6136,
+      "step": 464
+    },
+    {
+      "epoch": 0.15572672471533824,
+      "grad_norm": 0.4970155954360962,
+      "learning_rate": 0.0001,
+      "loss": 1.5398,
+      "step": 465
+    },
+    {
+      "epoch": 0.15606162089752176,
+      "grad_norm": 0.4889807403087616,
+      "learning_rate": 0.0001,
+      "loss": 1.6689,
+      "step": 466
+    },
+    {
+      "epoch": 0.15639651707970528,
+      "grad_norm": 0.4557911157608032,
+      "learning_rate": 0.0001,
+      "loss": 1.5573,
+      "step": 467
+    },
+    {
+      "epoch": 0.1567314132618888,
+      "grad_norm": 0.4632352292537689,
+      "learning_rate": 0.0001,
+      "loss": 1.5766,
+      "step": 468
+    },
+    {
+      "epoch": 0.15706630944407235,
+      "grad_norm": 0.4403630793094635,
+      "learning_rate": 0.0001,
+      "loss": 1.6143,
+      "step": 469
+    },
+    {
+      "epoch": 0.15740120562625587,
+      "grad_norm": 0.4789538085460663,
+      "learning_rate": 0.0001,
+      "loss": 1.6662,
+      "step": 470
+    },
+    {
+      "epoch": 0.1577361018084394,
+      "grad_norm": 0.46481096744537354,
+      "learning_rate": 0.0001,
+      "loss": 1.629,
+      "step": 471
+    },
+    {
+      "epoch": 0.1580709979906229,
+      "grad_norm": 0.47119757533073425,
+      "learning_rate": 0.0001,
+      "loss": 1.5943,
+      "step": 472
+    },
+    {
+      "epoch": 0.15840589417280643,
+      "grad_norm": 0.4695896804332733,
+      "learning_rate": 0.0001,
+      "loss": 1.6441,
+      "step": 473
+    },
+    {
+      "epoch": 0.15874079035498995,
+      "grad_norm": 0.46595636010169983,
+      "learning_rate": 0.0001,
+      "loss": 1.6329,
+      "step": 474
+    },
+    {
+      "epoch": 0.15907568653717347,
+      "grad_norm": 0.43559157848358154,
+      "learning_rate": 0.0001,
+      "loss": 1.5618,
+      "step": 475
+    },
+    {
+      "epoch": 0.15941058271935699,
+      "grad_norm": 0.4306503236293793,
+      "learning_rate": 0.0001,
+      "loss": 1.5752,
+      "step": 476
+    },
+    {
+      "epoch": 0.15974547890154053,
+      "grad_norm": 0.4469747543334961,
+      "learning_rate": 0.0001,
+      "loss": 1.6192,
+      "step": 477
+    },
+    {
+      "epoch": 0.16008037508372405,
+      "grad_norm": 0.43521738052368164,
+      "learning_rate": 0.0001,
+      "loss": 1.5409,
+      "step": 478
+    },
+    {
+      "epoch": 0.16041527126590757,
+      "grad_norm": 0.44318079948425293,
+      "learning_rate": 0.0001,
+      "loss": 1.5524,
+      "step": 479
+    },
+    {
+      "epoch": 0.1607501674480911,
+      "grad_norm": 0.42884916067123413,
+      "learning_rate": 0.0001,
+      "loss": 1.5466,
+      "step": 480
+    },
+    {
+      "epoch": 0.1610850636302746,
+      "grad_norm": 0.44447648525238037,
+      "learning_rate": 0.0001,
+      "loss": 1.578,
+      "step": 481
+    },
+    {
+      "epoch": 0.16141995981245813,
+      "grad_norm": 0.45701852440834045,
+      "learning_rate": 0.0001,
+      "loss": 1.6444,
+      "step": 482
+    },
+    {
+      "epoch": 0.16175485599464165,
+      "grad_norm": 0.44687342643737793,
+      "learning_rate": 0.0001,
+      "loss": 1.6858,
+      "step": 483
+    },
+    {
+      "epoch": 0.1620897521768252,
+      "grad_norm": 0.41851550340652466,
+      "learning_rate": 0.0001,
+      "loss": 1.4588,
+      "step": 484
+    },
+    {
+      "epoch": 0.16242464835900872,
+      "grad_norm": 0.4491201937198639,
+      "learning_rate": 0.0001,
+      "loss": 1.5753,
+      "step": 485
+    },
+    {
+      "epoch": 0.16275954454119224,
+      "grad_norm": 0.4423746168613434,
+      "learning_rate": 0.0001,
+      "loss": 1.5222,
+      "step": 486
+    },
+    {
+      "epoch": 0.16309444072337576,
+      "grad_norm": 0.44861212372779846,
+      "learning_rate": 0.0001,
+      "loss": 1.6327,
+      "step": 487
+    },
+    {
+      "epoch": 0.16342933690555927,
+      "grad_norm": 0.4326763451099396,
+      "learning_rate": 0.0001,
+      "loss": 1.5405,
+      "step": 488
+    },
+    {
+      "epoch": 0.1637642330877428,
+      "grad_norm": 0.43661123514175415,
+      "learning_rate": 0.0001,
+      "loss": 1.52,
+      "step": 489
+    },
+    {
+      "epoch": 0.1640991292699263,
+      "grad_norm": 0.4316999614238739,
+      "learning_rate": 0.0001,
+      "loss": 1.516,
+      "step": 490
+    },
+    {
+      "epoch": 0.16443402545210983,
+      "grad_norm": 0.4735428988933563,
+      "learning_rate": 0.0001,
+      "loss": 1.6493,
+      "step": 491
+    },
+    {
+      "epoch": 0.16476892163429338,
+      "grad_norm": 0.4458985924720764,
+      "learning_rate": 0.0001,
+      "loss": 1.5783,
+      "step": 492
+    },
+    {
+      "epoch": 0.1651038178164769,
+      "grad_norm": 0.46412917971611023,
+      "learning_rate": 0.0001,
+      "loss": 1.6794,
+      "step": 493
+    },
+    {
+      "epoch": 0.16543871399866042,
+      "grad_norm": 0.4410945773124695,
+      "learning_rate": 0.0001,
+      "loss": 1.6956,
+      "step": 494
+    },
+    {
+      "epoch": 0.16577361018084394,
+      "grad_norm": 0.44671037793159485,
+      "learning_rate": 0.0001,
+      "loss": 1.6173,
+      "step": 495
+    },
+    {
+      "epoch": 0.16610850636302746,
+      "grad_norm": 0.44258755445480347,
+      "learning_rate": 0.0001,
+      "loss": 1.6461,
+      "step": 496
+    },
+    {
+      "epoch": 0.16644340254521098,
+      "grad_norm": 0.4786889851093292,
+      "learning_rate": 0.0001,
+      "loss": 1.4906,
+      "step": 497
+    },
+    {
+      "epoch": 0.1667782987273945,
+      "grad_norm": 0.4585178792476654,
+      "learning_rate": 0.0001,
+      "loss": 1.6344,
+      "step": 498
+    },
+    {
+      "epoch": 0.16711319490957804,
+      "grad_norm": 0.44208812713623047,
+      "learning_rate": 0.0001,
+      "loss": 1.5837,
+      "step": 499
+    },
+    {
+      "epoch": 0.16744809109176156,
+      "grad_norm": 0.4394063651561737,
+      "learning_rate": 0.0001,
+      "loss": 1.4863,
+      "step": 500
+    },
+    {
+      "epoch": 0.16778298727394508,
+      "grad_norm": 0.4426101744174957,
+      "learning_rate": 0.0001,
+      "loss": 1.5901,
+      "step": 501
+    },
+    {
+      "epoch": 0.1681178834561286,
+      "grad_norm": 0.45466336607933044,
+      "learning_rate": 0.0001,
+      "loss": 1.5928,
+      "step": 502
+    },
+    {
+      "epoch": 0.16845277963831212,
+      "grad_norm": 0.44412124156951904,
+      "learning_rate": 0.0001,
+      "loss": 1.6238,
+      "step": 503
+    },
+    {
+      "epoch": 0.16878767582049564,
+      "grad_norm": 0.45599251985549927,
+      "learning_rate": 0.0001,
+      "loss": 1.6092,
+      "step": 504
+    },
+    {
+      "epoch": 0.16912257200267916,
+      "grad_norm": 0.4290856719017029,
+      "learning_rate": 0.0001,
+      "loss": 1.5464,
+      "step": 505
+    },
+    {
+      "epoch": 0.16945746818486268,
+      "grad_norm": 0.45510977506637573,
+      "learning_rate": 0.0001,
+      "loss": 1.5688,
+      "step": 506
+    },
+    {
+      "epoch": 0.16979236436704623,
+      "grad_norm": 0.44688305258750916,
+      "learning_rate": 0.0001,
+      "loss": 1.5343,
+      "step": 507
+    },
+    {
+      "epoch": 0.17012726054922975,
+      "grad_norm": 0.4349426031112671,
+      "learning_rate": 0.0001,
+      "loss": 1.6256,
+      "step": 508
+    },
+    {
+      "epoch": 0.17046215673141327,
+      "grad_norm": 0.45415475964546204,
+      "learning_rate": 0.0001,
+      "loss": 1.643,
+      "step": 509
+    },
+    {
+      "epoch": 0.1707970529135968,
+      "grad_norm": 0.4510926604270935,
+      "learning_rate": 0.0001,
+      "loss": 1.5533,
+      "step": 510
+    },
+    {
+      "epoch": 0.1711319490957803,
+      "grad_norm": 0.43210574984550476,
+      "learning_rate": 0.0001,
+      "loss": 1.5273,
+      "step": 511
+    },
+    {
+      "epoch": 0.17146684527796383,
+      "grad_norm": 0.49950575828552246,
+      "learning_rate": 0.0001,
+      "loss": 1.6401,
+      "step": 512
+    },
+    {
+      "epoch": 0.17180174146014734,
+      "grad_norm": 0.4558428227901459,
+      "learning_rate": 0.0001,
+      "loss": 1.5263,
+      "step": 513
+    },
+    {
+      "epoch": 0.17213663764233086,
+      "grad_norm": 0.46560290455818176,
+      "learning_rate": 0.0001,
+      "loss": 1.6968,
+      "step": 514
+    },
+    {
+      "epoch": 0.1724715338245144,
+      "grad_norm": 0.4596928656101227,
+      "learning_rate": 0.0001,
+      "loss": 1.6222,
+      "step": 515
+    },
+    {
+      "epoch": 0.17280643000669793,
+      "grad_norm": 0.43341562151908875,
+      "learning_rate": 0.0001,
+      "loss": 1.5597,
+      "step": 516
+    },
+    {
+      "epoch": 0.17314132618888145,
+      "grad_norm": 0.4746050536632538,
+      "learning_rate": 0.0001,
+      "loss": 1.5563,
+      "step": 517
+    },
+    {
+      "epoch": 0.17347622237106497,
+      "grad_norm": 0.44967103004455566,
+      "learning_rate": 0.0001,
+      "loss": 1.5534,
+      "step": 518
+    },
+    {
+      "epoch": 0.1738111185532485,
+      "grad_norm": 0.43267977237701416,
+      "learning_rate": 0.0001,
+      "loss": 1.4994,
+      "step": 519
+    },
+    {
+      "epoch": 0.174146014735432,
+      "grad_norm": 0.45056331157684326,
+      "learning_rate": 0.0001,
+      "loss": 1.5279,
+      "step": 520
+    },
+    {
+      "epoch": 0.17448091091761553,
+      "grad_norm": 0.4300089478492737,
+      "learning_rate": 0.0001,
+      "loss": 1.5114,
+      "step": 521
+    },
+    {
+      "epoch": 0.17481580709979908,
+      "grad_norm": 0.48487889766693115,
+      "learning_rate": 0.0001,
+      "loss": 1.5379,
+      "step": 522
+    },
+    {
+      "epoch": 0.1751507032819826,
+      "grad_norm": 0.4652538299560547,
+      "learning_rate": 0.0001,
+      "loss": 1.5751,
+      "step": 523
+    },
+    {
+      "epoch": 0.17548559946416611,
+      "grad_norm": 0.5223555564880371,
+      "learning_rate": 0.0001,
+      "loss": 1.6954,
+      "step": 524
+    },
+    {
+      "epoch": 0.17582049564634963,
+      "grad_norm": 0.4837857186794281,
+      "learning_rate": 0.0001,
+      "loss": 1.6461,
+      "step": 525
+    },
+    {
+      "epoch": 0.17615539182853315,
+      "grad_norm": 0.4681553542613983,
+      "learning_rate": 0.0001,
+      "loss": 1.5406,
+      "step": 526
+    },
+    {
+      "epoch": 0.17649028801071667,
+      "grad_norm": 0.44726818799972534,
+      "learning_rate": 0.0001,
+      "loss": 1.5826,
+      "step": 527
+    },
+    {
+      "epoch": 0.1768251841929002,
+      "grad_norm": 0.4904516637325287,
+      "learning_rate": 0.0001,
+      "loss": 1.6685,
+      "step": 528
+    },
+    {
+      "epoch": 0.1771600803750837,
+      "grad_norm": 0.4621741771697998,
+      "learning_rate": 0.0001,
+      "loss": 1.6141,
+      "step": 529
+    },
+    {
+      "epoch": 0.17749497655726726,
+      "grad_norm": 0.46151936054229736,
+      "learning_rate": 0.0001,
+      "loss": 1.497,
+      "step": 530
+    },
+    {
+      "epoch": 0.17782987273945078,
+      "grad_norm": 0.4495275914669037,
+      "learning_rate": 0.0001,
+      "loss": 1.5489,
+      "step": 531
+    },
+    {
+      "epoch": 0.1781647689216343,
+      "grad_norm": 0.44468608498573303,
+      "learning_rate": 0.0001,
+      "loss": 1.5458,
+      "step": 532
+    },
+    {
+      "epoch": 0.17849966510381782,
+      "grad_norm": 0.4433906078338623,
+      "learning_rate": 0.0001,
+      "loss": 1.6035,
+      "step": 533
+    },
+    {
+      "epoch": 0.17883456128600134,
+      "grad_norm": 0.4223592281341553,
+      "learning_rate": 0.0001,
+      "loss": 1.5338,
+      "step": 534
+    },
+    {
+      "epoch": 0.17916945746818486,
+      "grad_norm": 0.46216294169425964,
+      "learning_rate": 0.0001,
+      "loss": 1.6656,
+      "step": 535
+    },
+    {
+      "epoch": 0.17950435365036838,
+      "grad_norm": 0.46588945388793945,
+      "learning_rate": 0.0001,
+      "loss": 1.5895,
+      "step": 536
+    },
+    {
+      "epoch": 0.1798392498325519,
+      "grad_norm": 0.44910311698913574,
+      "learning_rate": 0.0001,
+      "loss": 1.6307,
+      "step": 537
+    },
+    {
+      "epoch": 0.18017414601473544,
+      "grad_norm": 0.4371938407421112,
+      "learning_rate": 0.0001,
+      "loss": 1.5129,
+      "step": 538
+    },
+    {
+      "epoch": 0.18050904219691896,
+      "grad_norm": 0.4528622031211853,
+      "learning_rate": 0.0001,
+      "loss": 1.6958,
+      "step": 539
+    },
+    {
+      "epoch": 0.18084393837910248,
+      "grad_norm": 0.46079522371292114,
+      "learning_rate": 0.0001,
+      "loss": 1.642,
+      "step": 540
+    },
+    {
+      "epoch": 0.181178834561286,
+      "grad_norm": 0.4528695344924927,
+      "learning_rate": 0.0001,
+      "loss": 1.5441,
+      "step": 541
+    },
+    {
+      "epoch": 0.18151373074346952,
+      "grad_norm": 0.44676560163497925,
+      "learning_rate": 0.0001,
+      "loss": 1.548,
+      "step": 542
+    },
+    {
+      "epoch": 0.18184862692565304,
+      "grad_norm": 0.42788952589035034,
+      "learning_rate": 0.0001,
+      "loss": 1.5221,
+      "step": 543
+    },
+    {
+      "epoch": 0.18218352310783656,
+      "grad_norm": 0.4401197135448456,
+      "learning_rate": 0.0001,
+      "loss": 1.5169,
+      "step": 544
+    },
+    {
+      "epoch": 0.1825184192900201,
+      "grad_norm": 0.4610879719257355,
+      "learning_rate": 0.0001,
+      "loss": 1.6219,
+      "step": 545
+    },
+    {
+      "epoch": 0.18285331547220363,
+      "grad_norm": 0.45374858379364014,
+      "learning_rate": 0.0001,
+      "loss": 1.5726,
+      "step": 546
+    },
+    {
+      "epoch": 0.18318821165438715,
+      "grad_norm": 0.45622724294662476,
+      "learning_rate": 0.0001,
+      "loss": 1.6732,
+      "step": 547
+    },
+    {
+      "epoch": 0.18352310783657066,
+      "grad_norm": 0.45034128427505493,
+      "learning_rate": 0.0001,
+      "loss": 1.6309,
+      "step": 548
+    },
+    {
+      "epoch": 0.18385800401875418,
+      "grad_norm": 0.46395254135131836,
+      "learning_rate": 0.0001,
+      "loss": 1.5214,
+      "step": 549
+    },
+    {
+      "epoch": 0.1841929002009377,
+      "grad_norm": 0.44303593039512634,
+      "learning_rate": 0.0001,
+      "loss": 1.546,
+      "step": 550
+    },
+    {
+      "epoch": 0.18452779638312122,
+      "grad_norm": 0.4510901868343353,
+      "learning_rate": 0.0001,
+      "loss": 1.5942,
+      "step": 551
+    },
+    {
+      "epoch": 0.18486269256530474,
+      "grad_norm": 0.44481050968170166,
+      "learning_rate": 0.0001,
+      "loss": 1.5295,
+      "step": 552
+    },
+    {
+      "epoch": 0.1851975887474883,
+      "grad_norm": 0.46099281311035156,
+      "learning_rate": 0.0001,
+      "loss": 1.5907,
+      "step": 553
+    },
+    {
+      "epoch": 0.1855324849296718,
+      "grad_norm": 0.4374411106109619,
+      "learning_rate": 0.0001,
+      "loss": 1.5278,
+      "step": 554
+    },
+    {
+      "epoch": 0.18586738111185533,
+      "grad_norm": 0.4573977291584015,
+      "learning_rate": 0.0001,
+      "loss": 1.4811,
+      "step": 555
+    },
+    {
+      "epoch": 0.18620227729403885,
+      "grad_norm": 0.4958246052265167,
+      "learning_rate": 0.0001,
+      "loss": 1.6173,
+      "step": 556
+    },
+    {
+      "epoch": 0.18653717347622237,
+      "grad_norm": 0.44467490911483765,
+      "learning_rate": 0.0001,
+      "loss": 1.5038,
+      "step": 557
+    },
+    {
+      "epoch": 0.1868720696584059,
+      "grad_norm": 0.45773619413375854,
+      "learning_rate": 0.0001,
+      "loss": 1.5535,
+      "step": 558
+    },
+    {
+      "epoch": 0.1872069658405894,
+      "grad_norm": 0.47369417548179626,
+      "learning_rate": 0.0001,
+      "loss": 1.6407,
+      "step": 559
+    },
+    {
+      "epoch": 0.18754186202277295,
+      "grad_norm": 0.4539751708507538,
+      "learning_rate": 0.0001,
+      "loss": 1.5265,
+      "step": 560
+    },
+    {
+      "epoch": 0.18787675820495647,
+      "grad_norm": 0.45979294180870056,
+      "learning_rate": 0.0001,
+      "loss": 1.7121,
+      "step": 561
+    },
+    {
+      "epoch": 0.18821165438714,
+      "grad_norm": 0.4300471544265747,
+      "learning_rate": 0.0001,
+      "loss": 1.5071,
+      "step": 562
+    },
+    {
+      "epoch": 0.1885465505693235,
+      "grad_norm": 0.41900163888931274,
+      "learning_rate": 0.0001,
+      "loss": 1.5415,
+      "step": 563
+    },
+    {
+      "epoch": 0.18888144675150703,
+      "grad_norm": 0.4393864870071411,
+      "learning_rate": 0.0001,
+      "loss": 1.5997,
+      "step": 564
+    },
+    {
+      "epoch": 0.18921634293369055,
+      "grad_norm": 0.42667317390441895,
+      "learning_rate": 0.0001,
+      "loss": 1.5435,
+      "step": 565
+    },
+    {
+      "epoch": 0.18955123911587407,
+      "grad_norm": 0.4409005045890808,
+      "learning_rate": 0.0001,
+      "loss": 1.6134,
+      "step": 566
+    },
+    {
+      "epoch": 0.1898861352980576,
+      "grad_norm": 0.4696614444255829,
+      "learning_rate": 0.0001,
+      "loss": 1.6321,
+      "step": 567
+    },
+    {
+      "epoch": 0.19022103148024114,
+      "grad_norm": 0.4674033522605896,
+      "learning_rate": 0.0001,
+      "loss": 1.5435,
+      "step": 568
+    },
+    {
+      "epoch": 0.19055592766242466,
+      "grad_norm": 0.46117159724235535,
+      "learning_rate": 0.0001,
+      "loss": 1.5953,
+      "step": 569
+    },
+    {
+      "epoch": 0.19089082384460818,
+      "grad_norm": 0.48608630895614624,
+      "learning_rate": 0.0001,
+      "loss": 1.5979,
+      "step": 570
+    },
+    {
+      "epoch": 0.1912257200267917,
+      "grad_norm": 0.4258008897304535,
+      "learning_rate": 0.0001,
+      "loss": 1.5425,
+      "step": 571
+    },
+    {
+      "epoch": 0.19156061620897522,
+      "grad_norm": 0.4251329302787781,
+      "learning_rate": 0.0001,
+      "loss": 1.4931,
+      "step": 572
+    },
+    {
+      "epoch": 0.19189551239115873,
+      "grad_norm": 0.4929272532463074,
+      "learning_rate": 0.0001,
+      "loss": 1.7021,
+      "step": 573
+    },
+    {
+      "epoch": 0.19223040857334225,
+      "grad_norm": 0.4487670958042145,
+      "learning_rate": 0.0001,
+      "loss": 1.6679,
+      "step": 574
+    },
+    {
+      "epoch": 0.19256530475552577,
+      "grad_norm": 0.4457535147666931,
+      "learning_rate": 0.0001,
+      "loss": 1.6178,
+      "step": 575
+    },
+    {
+      "epoch": 0.19290020093770932,
+      "grad_norm": 0.42327451705932617,
+      "learning_rate": 0.0001,
+      "loss": 1.5191,
+      "step": 576
+    },
+    {
+      "epoch": 0.19323509711989284,
+      "grad_norm": 0.43987563252449036,
+      "learning_rate": 0.0001,
+      "loss": 1.6082,
+      "step": 577
+    },
+    {
+      "epoch": 0.19356999330207636,
+      "grad_norm": 0.43453678488731384,
+      "learning_rate": 0.0001,
+      "loss": 1.561,
+      "step": 578
+    },
+    {
+      "epoch": 0.19390488948425988,
+      "grad_norm": 0.4256730377674103,
+      "learning_rate": 0.0001,
+      "loss": 1.5515,
+      "step": 579
+    },
+    {
+      "epoch": 0.1942397856664434,
+      "grad_norm": 0.4502320885658264,
+      "learning_rate": 0.0001,
+      "loss": 1.6113,
+      "step": 580
+    },
+    {
+      "epoch": 0.19457468184862692,
+      "grad_norm": 0.444354772567749,
+      "learning_rate": 0.0001,
+      "loss": 1.5668,
+      "step": 581
+    },
+    {
+      "epoch": 0.19490957803081044,
+      "grad_norm": 0.43178266286849976,
+      "learning_rate": 0.0001,
+      "loss": 1.5433,
+      "step": 582
+    },
+    {
+      "epoch": 0.19524447421299398,
+      "grad_norm": 0.4282374382019043,
+      "learning_rate": 0.0001,
+      "loss": 1.5646,
+      "step": 583
+    },
+    {
+      "epoch": 0.1955793703951775,
+      "grad_norm": 0.46302083134651184,
+      "learning_rate": 0.0001,
+      "loss": 1.547,
+      "step": 584
+    },
+    {
+      "epoch": 0.19591426657736102,
+      "grad_norm": 0.4419693052768707,
+      "learning_rate": 0.0001,
+      "loss": 1.6002,
+      "step": 585
+    },
+    {
+      "epoch": 0.19624916275954454,
+      "grad_norm": 0.4477298855781555,
+      "learning_rate": 0.0001,
+      "loss": 1.6908,
+      "step": 586
+    },
+    {
+      "epoch": 0.19658405894172806,
+      "grad_norm": 0.44686245918273926,
+      "learning_rate": 0.0001,
+      "loss": 1.541,
+      "step": 587
+    },
+    {
+      "epoch": 0.19691895512391158,
+      "grad_norm": 0.44238075613975525,
+      "learning_rate": 0.0001,
+      "loss": 1.6098,
+      "step": 588
+    },
+    {
+      "epoch": 0.1972538513060951,
+      "grad_norm": 0.44104936718940735,
+      "learning_rate": 0.0001,
+      "loss": 1.5786,
+      "step": 589
+    },
+    {
+      "epoch": 0.19758874748827862,
+      "grad_norm": 0.44523486495018005,
+      "learning_rate": 0.0001,
+      "loss": 1.5821,
+      "step": 590
+    },
+    {
+      "epoch": 0.19792364367046217,
+      "grad_norm": 0.4347565770149231,
+      "learning_rate": 0.0001,
+      "loss": 1.531,
+      "step": 591
+    },
+    {
+      "epoch": 0.1982585398526457,
+      "grad_norm": 0.4408382177352905,
+      "learning_rate": 0.0001,
+      "loss": 1.522,
+      "step": 592
+    },
+    {
+      "epoch": 0.1985934360348292,
+      "grad_norm": 0.45242637395858765,
+      "learning_rate": 0.0001,
+      "loss": 1.5678,
+      "step": 593
+    },
+    {
+      "epoch": 0.19892833221701273,
+      "grad_norm": 0.42536839842796326,
+      "learning_rate": 0.0001,
+      "loss": 1.389,
+      "step": 594
+    },
+    {
+      "epoch": 0.19926322839919625,
+      "grad_norm": 0.4437471628189087,
+      "learning_rate": 0.0001,
+      "loss": 1.5923,
+      "step": 595
+    },
+    {
+      "epoch": 0.19959812458137977,
+      "grad_norm": 0.46893227100372314,
+      "learning_rate": 0.0001,
+      "loss": 1.7063,
+      "step": 596
+    },
+    {
+      "epoch": 0.19993302076356329,
+      "grad_norm": 0.46792319416999817,
+      "learning_rate": 0.0001,
+      "loss": 1.6279,
+      "step": 597
+    },
+    {
+      "epoch": 0.20026791694574683,
+      "grad_norm": 0.43383094668388367,
+      "learning_rate": 0.0001,
+      "loss": 1.5666,
+      "step": 598
+    },
+    {
+      "epoch": 0.20060281312793035,
+      "grad_norm": 0.4612673819065094,
+      "learning_rate": 0.0001,
+      "loss": 1.657,
+      "step": 599
+    },
+    {
+      "epoch": 0.20093770931011387,
+      "grad_norm": 0.45714735984802246,
+      "learning_rate": 0.0001,
+      "loss": 1.5166,
+      "step": 600
+    },
+    {
+      "epoch": 0.2012726054922974,
+      "grad_norm": 0.4980306625366211,
+      "learning_rate": 0.0001,
+      "loss": 1.6338,
+      "step": 601
+    },
+    {
+      "epoch": 0.2016075016744809,
+      "grad_norm": 0.4374399483203888,
+      "learning_rate": 0.0001,
+      "loss": 1.4887,
+      "step": 602
+    },
+    {
+      "epoch": 0.20194239785666443,
+      "grad_norm": 0.4663514792919159,
+      "learning_rate": 0.0001,
+      "loss": 1.583,
+      "step": 603
+    },
+    {
+      "epoch": 0.20227729403884795,
+      "grad_norm": 0.4445118308067322,
+      "learning_rate": 0.0001,
+      "loss": 1.5925,
+      "step": 604
+    },
+    {
+      "epoch": 0.20261219022103147,
+      "grad_norm": 0.47844797372817993,
+      "learning_rate": 0.0001,
+      "loss": 1.6224,
+      "step": 605
+    },
+    {
+      "epoch": 0.20294708640321502,
+      "grad_norm": 0.43679067492485046,
+      "learning_rate": 0.0001,
+      "loss": 1.5509,
+      "step": 606
+    },
+    {
+      "epoch": 0.20328198258539854,
+      "grad_norm": 0.4718396067619324,
+      "learning_rate": 0.0001,
+      "loss": 1.5242,
+      "step": 607
+    },
+    {
+      "epoch": 0.20361687876758205,
+      "grad_norm": 0.43233776092529297,
+      "learning_rate": 0.0001,
+      "loss": 1.5069,
+      "step": 608
+    },
+    {
+      "epoch": 0.20395177494976557,
+      "grad_norm": 0.45062628388404846,
+      "learning_rate": 0.0001,
+      "loss": 1.5507,
+      "step": 609
+    },
+    {
+      "epoch": 0.2042866711319491,
+      "grad_norm": 0.43045535683631897,
+      "learning_rate": 0.0001,
+      "loss": 1.4764,
+      "step": 610
+    },
+    {
+      "epoch": 0.2046215673141326,
+      "grad_norm": 0.4338122010231018,
+      "learning_rate": 0.0001,
+      "loss": 1.4395,
+      "step": 611
+    },
+    {
+      "epoch": 0.20495646349631613,
+      "grad_norm": 0.4495809078216553,
+      "learning_rate": 0.0001,
+      "loss": 1.5663,
+      "step": 612
+    },
+    {
+      "epoch": 0.20529135967849965,
+      "grad_norm": 0.4504551291465759,
+      "learning_rate": 0.0001,
+      "loss": 1.5791,
+      "step": 613
+    },
+    {
+      "epoch": 0.2056262558606832,
+      "grad_norm": 0.4504256546497345,
+      "learning_rate": 0.0001,
+      "loss": 1.5453,
+      "step": 614
+    },
+    {
+      "epoch": 0.20596115204286672,
+      "grad_norm": 0.44566693902015686,
+      "learning_rate": 0.0001,
+      "loss": 1.4733,
+      "step": 615
+    },
+    {
+      "epoch": 0.20629604822505024,
+      "grad_norm": 0.45763394236564636,
+      "learning_rate": 0.0001,
+      "loss": 1.5266,
+      "step": 616
+    },
+    {
+      "epoch": 0.20663094440723376,
+      "grad_norm": 0.4807845652103424,
+      "learning_rate": 0.0001,
+      "loss": 1.5362,
+      "step": 617
+    },
+    {
+      "epoch": 0.20696584058941728,
+      "grad_norm": 0.44216951727867126,
+      "learning_rate": 0.0001,
+      "loss": 1.6271,
+      "step": 618
+    },
+    {
+      "epoch": 0.2073007367716008,
+      "grad_norm": 0.4926336109638214,
+      "learning_rate": 0.0001,
+      "loss": 1.5567,
+      "step": 619
+    },
+    {
+      "epoch": 0.20763563295378432,
+      "grad_norm": 0.4597328305244446,
+      "learning_rate": 0.0001,
+      "loss": 1.5826,
+      "step": 620
+    },
+    {
+      "epoch": 0.20797052913596786,
+      "grad_norm": 0.5120271444320679,
+      "learning_rate": 0.0001,
+      "loss": 1.4821,
+      "step": 621
+    },
+    {
+      "epoch": 0.20830542531815138,
+      "grad_norm": 0.44934895634651184,
+      "learning_rate": 0.0001,
+      "loss": 1.6378,
+      "step": 622
+    },
+    {
+      "epoch": 0.2086403215003349,
+      "grad_norm": 0.5311879515647888,
+      "learning_rate": 0.0001,
+      "loss": 1.6859,
+      "step": 623
+    },
+    {
+      "epoch": 0.20897521768251842,
+      "grad_norm": 0.4356757402420044,
+      "learning_rate": 0.0001,
+      "loss": 1.5312,
+      "step": 624
+    },
+    {
+      "epoch": 0.20931011386470194,
+      "grad_norm": 0.4556974470615387,
+      "learning_rate": 0.0001,
+      "loss": 1.5898,
+      "step": 625
+    },
+    {
+      "epoch": 0.20964501004688546,
+      "grad_norm": 0.45119380950927734,
+      "learning_rate": 0.0001,
+      "loss": 1.6319,
+      "step": 626
+    },
+    {
+      "epoch": 0.20997990622906898,
+      "grad_norm": 0.44451865553855896,
+      "learning_rate": 0.0001,
+      "loss": 1.5044,
+      "step": 627
+    },
+    {
+      "epoch": 0.2103148024112525,
+      "grad_norm": 0.4557413160800934,
+      "learning_rate": 0.0001,
+      "loss": 1.5661,
+      "step": 628
+    },
+    {
+      "epoch": 0.21064969859343605,
+      "grad_norm": 0.45663654804229736,
+      "learning_rate": 0.0001,
+      "loss": 1.5894,
+      "step": 629
+    },
+    {
+      "epoch": 0.21098459477561957,
+      "grad_norm": 0.47488924860954285,
+      "learning_rate": 0.0001,
+      "loss": 1.598,
+      "step": 630
+    },
+    {
+      "epoch": 0.21131949095780309,
+      "grad_norm": 0.4461064338684082,
+      "learning_rate": 0.0001,
+      "loss": 1.4725,
+      "step": 631
+    },
+    {
+      "epoch": 0.2116543871399866,
+      "grad_norm": 0.4668349027633667,
+      "learning_rate": 0.0001,
+      "loss": 1.4965,
+      "step": 632
+    },
+    {
+      "epoch": 0.21198928332217012,
+      "grad_norm": 0.45846477150917053,
+      "learning_rate": 0.0001,
+      "loss": 1.5392,
+      "step": 633
+    },
+    {
+      "epoch": 0.21232417950435364,
+      "grad_norm": 0.456134557723999,
+      "learning_rate": 0.0001,
+      "loss": 1.5775,
+      "step": 634
+    },
+    {
+      "epoch": 0.21265907568653716,
+      "grad_norm": 0.4363207221031189,
+      "learning_rate": 0.0001,
+      "loss": 1.5577,
+      "step": 635
+    },
+    {
+      "epoch": 0.21299397186872068,
+      "grad_norm": 0.4299388825893402,
+      "learning_rate": 0.0001,
+      "loss": 1.5528,
+      "step": 636
+    },
+    {
+      "epoch": 0.21332886805090423,
+      "grad_norm": 0.4551451504230499,
+      "learning_rate": 0.0001,
+      "loss": 1.6204,
+      "step": 637
+    },
+    {
+      "epoch": 0.21366376423308775,
+      "grad_norm": 0.459564745426178,
+      "learning_rate": 0.0001,
+      "loss": 1.5634,
+      "step": 638
+    },
+    {
+      "epoch": 0.21399866041527127,
+      "grad_norm": 0.4449961483478546,
+      "learning_rate": 0.0001,
+      "loss": 1.6278,
+      "step": 639
+    },
+    {
+      "epoch": 0.2143335565974548,
+      "grad_norm": 0.42216700315475464,
+      "learning_rate": 0.0001,
+      "loss": 1.5006,
+      "step": 640
+    },
+    {
+      "epoch": 0.2146684527796383,
+      "grad_norm": 0.4538195729255676,
+      "learning_rate": 0.0001,
+      "loss": 1.5815,
+      "step": 641
+    },
+    {
+      "epoch": 0.21500334896182183,
+      "grad_norm": 0.4515959918498993,
+      "learning_rate": 0.0001,
+      "loss": 1.5853,
+      "step": 642
+    },
+    {
+      "epoch": 0.21533824514400535,
+      "grad_norm": 0.4696963429450989,
+      "learning_rate": 0.0001,
+      "loss": 1.6153,
+      "step": 643
+    },
+    {
+      "epoch": 0.2156731413261889,
+      "grad_norm": 0.45178598165512085,
+      "learning_rate": 0.0001,
+      "loss": 1.6019,
+      "step": 644
+    },
+    {
+      "epoch": 0.2160080375083724,
+      "grad_norm": 0.4643690884113312,
+      "learning_rate": 0.0001,
+      "loss": 1.608,
+      "step": 645
+    },
+    {
+      "epoch": 0.21634293369055593,
+      "grad_norm": 0.4362979531288147,
+      "learning_rate": 0.0001,
+      "loss": 1.6127,
+      "step": 646
+    },
+    {
+      "epoch": 0.21667782987273945,
+      "grad_norm": 0.4543468952178955,
+      "learning_rate": 0.0001,
+      "loss": 1.6253,
+      "step": 647
+    },
+    {
+      "epoch": 0.21701272605492297,
+      "grad_norm": 0.44180887937545776,
+      "learning_rate": 0.0001,
+      "loss": 1.5105,
+      "step": 648
+    },
+    {
+      "epoch": 0.2173476222371065,
+      "grad_norm": 0.45792412757873535,
+      "learning_rate": 0.0001,
+      "loss": 1.6662,
+      "step": 649
+    },
+    {
+      "epoch": 0.21768251841929,
+      "grad_norm": 0.4656217694282532,
+      "learning_rate": 0.0001,
+      "loss": 1.7252,
+      "step": 650
+    },
+    {
+      "epoch": 0.21801741460147353,
+      "grad_norm": 0.44460582733154297,
+      "learning_rate": 0.0001,
+      "loss": 1.5543,
+      "step": 651
+    },
+    {
+      "epoch": 0.21835231078365708,
+      "grad_norm": 0.45652541518211365,
+      "learning_rate": 0.0001,
+      "loss": 1.5782,
+      "step": 652
+    },
+    {
+      "epoch": 0.2186872069658406,
+      "grad_norm": 0.449512779712677,
+      "learning_rate": 0.0001,
+      "loss": 1.5156,
+      "step": 653
+    },
+    {
+      "epoch": 0.21902210314802412,
+      "grad_norm": 0.4760141670703888,
+      "learning_rate": 0.0001,
+      "loss": 1.5886,
+      "step": 654
+    },
+    {
+      "epoch": 0.21935699933020764,
+      "grad_norm": 0.45350316166877747,
+      "learning_rate": 0.0001,
+      "loss": 1.5902,
+      "step": 655
+    },
+    {
+      "epoch": 0.21969189551239116,
+      "grad_norm": 0.4670565128326416,
+      "learning_rate": 0.0001,
+      "loss": 1.509,
+      "step": 656
+    },
+    {
+      "epoch": 0.22002679169457467,
+      "grad_norm": 0.462492436170578,
+      "learning_rate": 0.0001,
+      "loss": 1.5283,
+      "step": 657
+    },
+    {
+      "epoch": 0.2203616878767582,
+      "grad_norm": 0.4635310173034668,
+      "learning_rate": 0.0001,
+      "loss": 1.6162,
+      "step": 658
+    },
+    {
+      "epoch": 0.22069658405894174,
+      "grad_norm": 0.48554205894470215,
+      "learning_rate": 0.0001,
+      "loss": 1.5574,
+      "step": 659
+    },
+    {
+      "epoch": 0.22103148024112526,
+      "grad_norm": 0.4405190050601959,
+      "learning_rate": 0.0001,
+      "loss": 1.5643,
+      "step": 660
+    },
+    {
+      "epoch": 0.22136637642330878,
+      "grad_norm": 0.506333589553833,
+      "learning_rate": 0.0001,
+      "loss": 1.5499,
+      "step": 661
+    },
+    {
+      "epoch": 0.2217012726054923,
+      "grad_norm": 0.4472103714942932,
+      "learning_rate": 0.0001,
+      "loss": 1.4229,
+      "step": 662
+    },
+    {
+      "epoch": 0.22203616878767582,
+      "grad_norm": 0.46667394042015076,
+      "learning_rate": 0.0001,
+      "loss": 1.5524,
+      "step": 663
+    },
+    {
+      "epoch": 0.22237106496985934,
+      "grad_norm": 0.4534272253513336,
+      "learning_rate": 0.0001,
+      "loss": 1.6028,
+      "step": 664
+    },
+    {
+      "epoch": 0.22270596115204286,
+      "grad_norm": 0.49894607067108154,
+      "learning_rate": 0.0001,
+      "loss": 1.6009,
+      "step": 665
+    },
+    {
+      "epoch": 0.22304085733422638,
+      "grad_norm": 0.4377092123031616,
+      "learning_rate": 0.0001,
+      "loss": 1.5558,
+      "step": 666
+    },
+    {
+      "epoch": 0.22337575351640993,
+      "grad_norm": 0.47780877351760864,
+      "learning_rate": 0.0001,
+      "loss": 1.5691,
+      "step": 667
+    },
+    {
+      "epoch": 0.22371064969859344,
+      "grad_norm": 0.4473220705986023,
+      "learning_rate": 0.0001,
+      "loss": 1.63,
+      "step": 668
+    },
+    {
+      "epoch": 0.22404554588077696,
+      "grad_norm": 0.44755011796951294,
+      "learning_rate": 0.0001,
+      "loss": 1.5047,
+      "step": 669
+    },
+    {
+      "epoch": 0.22438044206296048,
+      "grad_norm": 0.4666520059108734,
+      "learning_rate": 0.0001,
+      "loss": 1.5957,
+      "step": 670
+    },
+    {
+      "epoch": 0.224715338245144,
+      "grad_norm": 0.4471485912799835,
+      "learning_rate": 0.0001,
+      "loss": 1.5304,
+      "step": 671
+    },
+    {
+      "epoch": 0.22505023442732752,
+      "grad_norm": 0.4512856900691986,
+      "learning_rate": 0.0001,
+      "loss": 1.6551,
+      "step": 672
+    },
+    {
+      "epoch": 0.22538513060951104,
+      "grad_norm": 0.4558720290660858,
+      "learning_rate": 0.0001,
+      "loss": 1.5348,
+      "step": 673
+    },
+    {
+      "epoch": 0.22572002679169456,
+      "grad_norm": 0.44984716176986694,
+      "learning_rate": 0.0001,
+      "loss": 1.5811,
+      "step": 674
+    },
+    {
+      "epoch": 0.2260549229738781,
+      "grad_norm": 0.4666685163974762,
+      "learning_rate": 0.0001,
+      "loss": 1.6274,
+      "step": 675
+    },
+    {
+      "epoch": 0.22638981915606163,
+      "grad_norm": 0.4678918719291687,
+      "learning_rate": 0.0001,
+      "loss": 1.6061,
+      "step": 676
+    },
+    {
+      "epoch": 0.22672471533824515,
+      "grad_norm": 0.4496474266052246,
+      "learning_rate": 0.0001,
+      "loss": 1.4882,
+      "step": 677
+    },
+    {
+      "epoch": 0.22705961152042867,
+      "grad_norm": 0.455914169549942,
+      "learning_rate": 0.0001,
+      "loss": 1.6567,
+      "step": 678
+    },
+    {
+      "epoch": 0.2273945077026122,
+      "grad_norm": 0.43080762028694153,
+      "learning_rate": 0.0001,
+      "loss": 1.4999,
+      "step": 679
+    },
+    {
+      "epoch": 0.2277294038847957,
+      "grad_norm": 0.44386598467826843,
+      "learning_rate": 0.0001,
+      "loss": 1.5976,
+      "step": 680
+    },
+    {
+      "epoch": 0.22806430006697923,
+      "grad_norm": 0.44304677844047546,
+      "learning_rate": 0.0001,
+      "loss": 1.5395,
+      "step": 681
+    },
+    {
+      "epoch": 0.22839919624916277,
+      "grad_norm": 0.44651317596435547,
+      "learning_rate": 0.0001,
+      "loss": 1.5642,
+      "step": 682
+    },
+    {
+      "epoch": 0.2287340924313463,
+      "grad_norm": 0.4273903965950012,
+      "learning_rate": 0.0001,
+      "loss": 1.4957,
+      "step": 683
+    },
+    {
+      "epoch": 0.2290689886135298,
+      "grad_norm": 0.45144903659820557,
+      "learning_rate": 0.0001,
+      "loss": 1.6001,
+      "step": 684
+    },
+    {
+      "epoch": 0.22940388479571333,
+      "grad_norm": 0.4881639778614044,
+      "learning_rate": 0.0001,
+      "loss": 1.5736,
+      "step": 685
+    },
+    {
+      "epoch": 0.22973878097789685,
+      "grad_norm": 0.4803936183452606,
+      "learning_rate": 0.0001,
+      "loss": 1.6289,
+      "step": 686
+    },
+    {
+      "epoch": 0.23007367716008037,
+      "grad_norm": 0.44501784443855286,
+      "learning_rate": 0.0001,
+      "loss": 1.5637,
+      "step": 687
+    },
+    {
+      "epoch": 0.2304085733422639,
+      "grad_norm": 0.44302091002464294,
+      "learning_rate": 0.0001,
+      "loss": 1.3864,
+      "step": 688
+    },
+    {
+      "epoch": 0.2307434695244474,
+      "grad_norm": 0.4551304280757904,
+      "learning_rate": 0.0001,
+      "loss": 1.606,
+      "step": 689
+    },
+    {
+      "epoch": 0.23107836570663096,
+      "grad_norm": 0.5015878677368164,
+      "learning_rate": 0.0001,
+      "loss": 1.645,
+      "step": 690
+    },
+    {
+      "epoch": 0.23141326188881448,
+      "grad_norm": 0.45502328872680664,
+      "learning_rate": 0.0001,
+      "loss": 1.5172,
+      "step": 691
+    },
+    {
+      "epoch": 0.231748158070998,
+      "grad_norm": 0.45328280329704285,
+      "learning_rate": 0.0001,
+      "loss": 1.5835,
+      "step": 692
+    },
+    {
+      "epoch": 0.23208305425318151,
+      "grad_norm": 0.44391489028930664,
+      "learning_rate": 0.0001,
+      "loss": 1.5525,
+      "step": 693
+    },
+    {
+      "epoch": 0.23241795043536503,
+      "grad_norm": 0.4328315258026123,
+      "learning_rate": 0.0001,
+      "loss": 1.4947,
+      "step": 694
+    },
+    {
+      "epoch": 0.23275284661754855,
+      "grad_norm": 0.4365626871585846,
+      "learning_rate": 0.0001,
+      "loss": 1.5518,
+      "step": 695
+    },
+    {
+      "epoch": 0.23308774279973207,
+      "grad_norm": 0.4440605640411377,
+      "learning_rate": 0.0001,
+      "loss": 1.5446,
+      "step": 696
+    },
+    {
+      "epoch": 0.23342263898191562,
+      "grad_norm": 0.4284314811229706,
+      "learning_rate": 0.0001,
+      "loss": 1.4681,
+      "step": 697
+    },
+    {
+      "epoch": 0.23375753516409914,
+      "grad_norm": 0.4485720992088318,
+      "learning_rate": 0.0001,
+      "loss": 1.5829,
+      "step": 698
+    },
+    {
+      "epoch": 0.23409243134628266,
+      "grad_norm": 0.45565420389175415,
+      "learning_rate": 0.0001,
+      "loss": 1.6203,
+      "step": 699
+    },
+    {
+      "epoch": 0.23442732752846618,
+      "grad_norm": 0.45044267177581787,
+      "learning_rate": 0.0001,
+      "loss": 1.5372,
+      "step": 700
+    },
+    {
+      "epoch": 0.2347622237106497,
+      "grad_norm": 0.46250998973846436,
+      "learning_rate": 0.0001,
+      "loss": 1.5617,
+      "step": 701
+    },
+    {
+      "epoch": 0.23509711989283322,
+      "grad_norm": 0.47578829526901245,
+      "learning_rate": 0.0001,
+      "loss": 1.6007,
+      "step": 702
+    },
+    {
+      "epoch": 0.23543201607501674,
+      "grad_norm": 0.4471472501754761,
+      "learning_rate": 0.0001,
+      "loss": 1.4711,
+      "step": 703
+    },
+    {
+      "epoch": 0.23576691225720026,
+      "grad_norm": 0.4580185115337372,
+      "learning_rate": 0.0001,
+      "loss": 1.5707,
+      "step": 704
+    },
+    {
+      "epoch": 0.2361018084393838,
+      "grad_norm": 0.4444526135921478,
+      "learning_rate": 0.0001,
+      "loss": 1.5372,
+      "step": 705
+    },
+    {
+      "epoch": 0.23643670462156732,
+      "grad_norm": 0.4543907046318054,
+      "learning_rate": 0.0001,
+      "loss": 1.6054,
+      "step": 706
+    },
+    {
+      "epoch": 0.23677160080375084,
+      "grad_norm": 0.4625626802444458,
+      "learning_rate": 0.0001,
+      "loss": 1.6129,
+      "step": 707
+    },
+    {
+      "epoch": 0.23710649698593436,
+      "grad_norm": 0.45688140392303467,
+      "learning_rate": 0.0001,
+      "loss": 1.6426,
+      "step": 708
+    },
+    {
+      "epoch": 0.23744139316811788,
+      "grad_norm": 0.4441126883029938,
+      "learning_rate": 0.0001,
+      "loss": 1.5791,
+      "step": 709
+    },
+    {
+      "epoch": 0.2377762893503014,
+      "grad_norm": 0.4567664861679077,
+      "learning_rate": 0.0001,
+      "loss": 1.4215,
+      "step": 710
+    },
+    {
+      "epoch": 0.23811118553248492,
+      "grad_norm": 0.47155311703681946,
+      "learning_rate": 0.0001,
+      "loss": 1.6239,
+      "step": 711
+    },
+    {
+      "epoch": 0.23844608171466844,
+      "grad_norm": 0.4321557581424713,
+      "learning_rate": 0.0001,
+      "loss": 1.493,
+      "step": 712
+    },
+    {
+      "epoch": 0.238780977896852,
+      "grad_norm": 0.43683552742004395,
+      "learning_rate": 0.0001,
+      "loss": 1.5824,
+      "step": 713
+    },
+    {
+      "epoch": 0.2391158740790355,
+      "grad_norm": 0.4407794773578644,
+      "learning_rate": 0.0001,
+      "loss": 1.5692,
+      "step": 714
+    },
+    {
+      "epoch": 0.23945077026121903,
+      "grad_norm": 0.43725642561912537,
+      "learning_rate": 0.0001,
+      "loss": 1.5868,
+      "step": 715
+    },
+    {
+      "epoch": 0.23978566644340255,
+      "grad_norm": 0.4484725296497345,
+      "learning_rate": 0.0001,
+      "loss": 1.6105,
+      "step": 716
+    },
+    {
+      "epoch": 0.24012056262558606,
+      "grad_norm": 0.47937703132629395,
+      "learning_rate": 0.0001,
+      "loss": 1.5828,
+      "step": 717
+    },
+    {
+      "epoch": 0.24045545880776958,
+      "grad_norm": 0.45690029859542847,
+      "learning_rate": 0.0001,
+      "loss": 1.542,
+      "step": 718
+    },
+    {
+      "epoch": 0.2407903549899531,
+      "grad_norm": 0.4617394208908081,
+      "learning_rate": 0.0001,
+      "loss": 1.6432,
+      "step": 719
+    },
+    {
+      "epoch": 0.24112525117213665,
+      "grad_norm": 0.46813878417015076,
+      "learning_rate": 0.0001,
+      "loss": 1.5294,
+      "step": 720
+    },
+    {
+      "epoch": 0.24146014735432017,
+      "grad_norm": 0.44215312600135803,
+      "learning_rate": 0.0001,
+      "loss": 1.5601,
+      "step": 721
+    },
+    {
+      "epoch": 0.2417950435365037,
+      "grad_norm": 0.44908544421195984,
+      "learning_rate": 0.0001,
+      "loss": 1.5717,
+      "step": 722
+    },
+    {
+      "epoch": 0.2421299397186872,
+      "grad_norm": 0.46838417649269104,
+      "learning_rate": 0.0001,
+      "loss": 1.5559,
+      "step": 723
+    },
+    {
+      "epoch": 0.24246483590087073,
+      "grad_norm": 0.4515591859817505,
+      "learning_rate": 0.0001,
+      "loss": 1.6158,
+      "step": 724
+    },
+    {
+      "epoch": 0.24279973208305425,
+      "grad_norm": 0.4613933563232422,
+      "learning_rate": 0.0001,
+      "loss": 1.5708,
+      "step": 725
+    },
+    {
+      "epoch": 0.24313462826523777,
+      "grad_norm": 0.4435577392578125,
+      "learning_rate": 0.0001,
+      "loss": 1.5237,
+      "step": 726
+    },
+    {
+      "epoch": 0.2434695244474213,
+      "grad_norm": 0.43679118156433105,
+      "learning_rate": 0.0001,
+      "loss": 1.5505,
+      "step": 727
+    },
+    {
+      "epoch": 0.24380442062960483,
+      "grad_norm": 0.42168736457824707,
+      "learning_rate": 0.0001,
+      "loss": 1.5065,
+      "step": 728
+    },
+    {
+      "epoch": 0.24413931681178835,
+      "grad_norm": 0.46262645721435547,
+      "learning_rate": 0.0001,
+      "loss": 1.5761,
+      "step": 729
+    },
+    {
+      "epoch": 0.24447421299397187,
+      "grad_norm": 0.4420361816883087,
+      "learning_rate": 0.0001,
+      "loss": 1.4914,
+      "step": 730
+    },
+    {
+      "epoch": 0.2448091091761554,
+      "grad_norm": 0.4631895124912262,
+      "learning_rate": 0.0001,
+      "loss": 1.6061,
+      "step": 731
+    },
+    {
+      "epoch": 0.2451440053583389,
+      "grad_norm": 0.4412582814693451,
+      "learning_rate": 0.0001,
+      "loss": 1.5394,
+      "step": 732
+    },
+    {
+      "epoch": 0.24547890154052243,
+      "grad_norm": 0.461233913898468,
+      "learning_rate": 0.0001,
+      "loss": 1.6477,
+      "step": 733
+    },
+    {
+      "epoch": 0.24581379772270595,
+      "grad_norm": 0.44835495948791504,
+      "learning_rate": 0.0001,
+      "loss": 1.5493,
+      "step": 734
+    },
+    {
+      "epoch": 0.24614869390488947,
+      "grad_norm": 0.4717530310153961,
+      "learning_rate": 0.0001,
+      "loss": 1.5904,
+      "step": 735
+    },
+    {
+      "epoch": 0.24648359008707302,
+      "grad_norm": 0.44697943329811096,
+      "learning_rate": 0.0001,
+      "loss": 1.5974,
+      "step": 736
+    },
+    {
+      "epoch": 0.24681848626925654,
+      "grad_norm": 0.4591861665248871,
+      "learning_rate": 0.0001,
+      "loss": 1.5437,
+      "step": 737
+    },
+    {
+      "epoch": 0.24715338245144006,
+      "grad_norm": 0.4309515357017517,
+      "learning_rate": 0.0001,
+      "loss": 1.5082,
+      "step": 738
+    },
+    {
+      "epoch": 0.24748827863362358,
+      "grad_norm": 0.48085346817970276,
+      "learning_rate": 0.0001,
+      "loss": 1.5783,
+      "step": 739
+    },
+    {
+      "epoch": 0.2478231748158071,
+      "grad_norm": 0.4605758786201477,
+      "learning_rate": 0.0001,
+      "loss": 1.6161,
+      "step": 740
+    },
+    {
+      "epoch": 0.24815807099799062,
+      "grad_norm": 0.449176162481308,
+      "learning_rate": 0.0001,
+      "loss": 1.6114,
+      "step": 741
+    },
+    {
+      "epoch": 0.24849296718017413,
+      "grad_norm": 0.4406338632106781,
+      "learning_rate": 0.0001,
+      "loss": 1.5702,
+      "step": 742
+    },
+    {
+      "epoch": 0.24882786336235768,
+      "grad_norm": 0.444603830575943,
+      "learning_rate": 0.0001,
+      "loss": 1.6437,
+      "step": 743
+    },
+    {
+      "epoch": 0.2491627595445412,
+      "grad_norm": 0.4413153827190399,
+      "learning_rate": 0.0001,
+      "loss": 1.5557,
+      "step": 744
+    },
+    {
+      "epoch": 0.24949765572672472,
+      "grad_norm": 0.4562162756919861,
+      "learning_rate": 0.0001,
+      "loss": 1.7314,
+      "step": 745
+    },
+    {
+      "epoch": 0.24983255190890824,
+      "grad_norm": 0.4683936536312103,
+      "learning_rate": 0.0001,
+      "loss": 1.5625,
+      "step": 746
+    },
+    {
+      "epoch": 0.25016744809109176,
+      "grad_norm": 0.43468979001045227,
+      "learning_rate": 0.0001,
+      "loss": 1.5982,
+      "step": 747
+    },
+    {
+      "epoch": 0.2505023442732753,
+      "grad_norm": 0.45922327041625977,
+      "learning_rate": 0.0001,
+      "loss": 1.5583,
+      "step": 748
+    },
+    {
+      "epoch": 0.2508372404554588,
+      "grad_norm": 0.45437920093536377,
+      "learning_rate": 0.0001,
+      "loss": 1.5728,
+      "step": 749
+    },
+    {
+      "epoch": 0.2511721366376423,
+      "grad_norm": 0.43223872780799866,
+      "learning_rate": 0.0001,
+      "loss": 1.5178,
+      "step": 750
+    },
+    {
+      "epoch": 0.25150703281982584,
+      "grad_norm": 0.4360998868942261,
+      "learning_rate": 0.0001,
+      "loss": 1.5791,
+      "step": 751
+    },
+    {
+      "epoch": 0.25184192900200936,
+      "grad_norm": 0.4442369043827057,
+      "learning_rate": 0.0001,
+      "loss": 1.5894,
+      "step": 752
+    },
+    {
+      "epoch": 0.2521768251841929,
+      "grad_norm": 0.4531276226043701,
+      "learning_rate": 0.0001,
+      "loss": 1.5669,
+      "step": 753
+    },
+    {
+      "epoch": 0.2525117213663764,
+      "grad_norm": 0.4386443495750427,
+      "learning_rate": 0.0001,
+      "loss": 1.4815,
+      "step": 754
+    },
+    {
+      "epoch": 0.25284661754855997,
+      "grad_norm": 0.45330408215522766,
+      "learning_rate": 0.0001,
+      "loss": 1.6762,
+      "step": 755
+    },
+    {
+      "epoch": 0.2531815137307435,
+      "grad_norm": 0.4780266582965851,
+      "learning_rate": 0.0001,
+      "loss": 1.5828,
+      "step": 756
+    },
+    {
+      "epoch": 0.253516409912927,
+      "grad_norm": 0.45670267939567566,
+      "learning_rate": 0.0001,
+      "loss": 1.5561,
+      "step": 757
+    },
+    {
+      "epoch": 0.25385130609511053,
+      "grad_norm": 0.4919813573360443,
+      "learning_rate": 0.0001,
+      "loss": 1.6987,
+      "step": 758
+    },
+    {
+      "epoch": 0.25418620227729405,
+      "grad_norm": 0.45109012722969055,
+      "learning_rate": 0.0001,
+      "loss": 1.578,
+      "step": 759
+    },
+    {
+      "epoch": 0.25452109845947757,
+      "grad_norm": 0.46791893243789673,
+      "learning_rate": 0.0001,
+      "loss": 1.5083,
+      "step": 760
+    },
+    {
+      "epoch": 0.2548559946416611,
+      "grad_norm": 0.44861626625061035,
+      "learning_rate": 0.0001,
+      "loss": 1.5325,
+      "step": 761
+    },
+    {
+      "epoch": 0.2551908908238446,
+      "grad_norm": 0.4484095275402069,
+      "learning_rate": 0.0001,
+      "loss": 1.5581,
+      "step": 762
+    },
+    {
+      "epoch": 0.2555257870060281,
+      "grad_norm": 0.4307587146759033,
+      "learning_rate": 0.0001,
+      "loss": 1.4857,
+      "step": 763
+    },
+    {
+      "epoch": 0.25586068318821165,
+      "grad_norm": 0.4829903542995453,
+      "learning_rate": 0.0001,
+      "loss": 1.5391,
+      "step": 764
+    },
+    {
+      "epoch": 0.25619557937039517,
+      "grad_norm": 0.4534643292427063,
+      "learning_rate": 0.0001,
+      "loss": 1.6835,
+      "step": 765
+    },
+    {
+      "epoch": 0.2565304755525787,
+      "grad_norm": 0.48684337735176086,
+      "learning_rate": 0.0001,
+      "loss": 1.6099,
+      "step": 766
+    },
+    {
+      "epoch": 0.2568653717347622,
+      "grad_norm": 0.45247527956962585,
+      "learning_rate": 0.0001,
+      "loss": 1.4907,
+      "step": 767
+    },
+    {
+      "epoch": 0.2572002679169457,
+      "grad_norm": 0.48305273056030273,
+      "learning_rate": 0.0001,
+      "loss": 1.5884,
+      "step": 768
+    },
+    {
+      "epoch": 0.25753516409912924,
+      "grad_norm": 0.45419007539749146,
+      "learning_rate": 0.0001,
+      "loss": 1.5194,
+      "step": 769
+    },
+    {
+      "epoch": 0.2578700602813128,
+      "grad_norm": 0.4592842161655426,
+      "learning_rate": 0.0001,
+      "loss": 1.5247,
+      "step": 770
+    },
+    {
+      "epoch": 0.25820495646349634,
+      "grad_norm": 0.4464816451072693,
+      "learning_rate": 0.0001,
+      "loss": 1.5456,
+      "step": 771
+    },
+    {
+      "epoch": 0.25853985264567986,
+      "grad_norm": 0.44528427720069885,
+      "learning_rate": 0.0001,
+      "loss": 1.4692,
+      "step": 772
+    },
+    {
+      "epoch": 0.2588747488278634,
+      "grad_norm": 0.4575278162956238,
+      "learning_rate": 0.0001,
+      "loss": 1.5124,
+      "step": 773
+    },
+    {
+      "epoch": 0.2592096450100469,
+      "grad_norm": 0.44354987144470215,
+      "learning_rate": 0.0001,
+      "loss": 1.5577,
+      "step": 774
+    },
+    {
+      "epoch": 0.2595445411922304,
+      "grad_norm": 0.4407951831817627,
+      "learning_rate": 0.0001,
+      "loss": 1.637,
+      "step": 775
+    },
+    {
+      "epoch": 0.25987943737441394,
+      "grad_norm": 0.4313081204891205,
+      "learning_rate": 0.0001,
+      "loss": 1.533,
+      "step": 776
+    },
+    {
+      "epoch": 0.26021433355659745,
+      "grad_norm": 0.4363669157028198,
+      "learning_rate": 0.0001,
+      "loss": 1.6094,
+      "step": 777
+    },
+    {
+      "epoch": 0.260549229738781,
+      "grad_norm": 0.44643253087997437,
+      "learning_rate": 0.0001,
+      "loss": 1.5489,
+      "step": 778
+    },
+    {
+      "epoch": 0.2608841259209645,
+      "grad_norm": 0.460085928440094,
+      "learning_rate": 0.0001,
+      "loss": 1.5721,
+      "step": 779
+    },
+    {
+      "epoch": 0.261219022103148,
+      "grad_norm": 0.44179096817970276,
+      "learning_rate": 0.0001,
+      "loss": 1.5302,
+      "step": 780
+    },
+    {
+      "epoch": 0.26155391828533153,
+      "grad_norm": 0.44657012820243835,
+      "learning_rate": 0.0001,
+      "loss": 1.4765,
+      "step": 781
+    },
+    {
+      "epoch": 0.26188881446751505,
+      "grad_norm": 0.458996444940567,
+      "learning_rate": 0.0001,
+      "loss": 1.6837,
+      "step": 782
+    },
+    {
+      "epoch": 0.26222371064969857,
+      "grad_norm": 0.4423884451389313,
+      "learning_rate": 0.0001,
+      "loss": 1.5736,
+      "step": 783
+    },
+    {
+      "epoch": 0.2625586068318821,
+      "grad_norm": 0.46296098828315735,
+      "learning_rate": 0.0001,
+      "loss": 1.6311,
+      "step": 784
+    },
+    {
+      "epoch": 0.26289350301406567,
+      "grad_norm": 0.43383488059043884,
+      "learning_rate": 0.0001,
+      "loss": 1.6063,
+      "step": 785
+    },
+    {
+      "epoch": 0.2632283991962492,
+      "grad_norm": 0.43596151471138,
+      "learning_rate": 0.0001,
+      "loss": 1.4607,
+      "step": 786
+    },
+    {
+      "epoch": 0.2635632953784327,
+      "grad_norm": 0.4386815130710602,
+      "learning_rate": 0.0001,
+      "loss": 1.5185,
+      "step": 787
+    },
+    {
+      "epoch": 0.2638981915606162,
+      "grad_norm": 0.47889411449432373,
+      "learning_rate": 0.0001,
+      "loss": 1.5469,
+      "step": 788
+    },
+    {
+      "epoch": 0.26423308774279974,
+      "grad_norm": 0.4255765974521637,
+      "learning_rate": 0.0001,
+      "loss": 1.4188,
+      "step": 789
+    },
+    {
+      "epoch": 0.26456798392498326,
+      "grad_norm": 0.5282460451126099,
+      "learning_rate": 0.0001,
+      "loss": 1.4845,
+      "step": 790
+    },
+    {
+      "epoch": 0.2649028801071668,
+      "grad_norm": 0.4410010576248169,
+      "learning_rate": 0.0001,
+      "loss": 1.5932,
+      "step": 791
+    },
+    {
+      "epoch": 0.2652377762893503,
+      "grad_norm": 0.4921286106109619,
+      "learning_rate": 0.0001,
+      "loss": 1.7258,
+      "step": 792
+    },
+    {
+      "epoch": 0.2655726724715338,
+      "grad_norm": 0.46794795989990234,
+      "learning_rate": 0.0001,
+      "loss": 1.6228,
+      "step": 793
+    },
+    {
+      "epoch": 0.26590756865371734,
+      "grad_norm": 0.4620145261287689,
+      "learning_rate": 0.0001,
+      "loss": 1.4615,
+      "step": 794
+    },
+    {
+      "epoch": 0.26624246483590086,
+      "grad_norm": 0.4698029160499573,
+      "learning_rate": 0.0001,
+      "loss": 1.6091,
+      "step": 795
+    },
+    {
+      "epoch": 0.2665773610180844,
+      "grad_norm": 0.4392451047897339,
+      "learning_rate": 0.0001,
+      "loss": 1.5311,
+      "step": 796
+    },
+    {
+      "epoch": 0.2669122572002679,
+      "grad_norm": 0.5082414746284485,
+      "learning_rate": 0.0001,
+      "loss": 1.5202,
+      "step": 797
+    },
+    {
+      "epoch": 0.2672471533824514,
+      "grad_norm": 0.4562722444534302,
+      "learning_rate": 0.0001,
+      "loss": 1.6005,
+      "step": 798
+    },
+    {
+      "epoch": 0.26758204956463494,
+      "grad_norm": 0.473054438829422,
+      "learning_rate": 0.0001,
+      "loss": 1.6171,
+      "step": 799
+    },
+    {
+      "epoch": 0.2679169457468185,
+      "grad_norm": 0.4447856843471527,
+      "learning_rate": 0.0001,
+      "loss": 1.5041,
+      "step": 800
+    },
+    {
+      "epoch": 0.26825184192900203,
+      "grad_norm": 0.4536293148994446,
+      "learning_rate": 0.0001,
+      "loss": 1.5749,
+      "step": 801
+    },
+    {
+      "epoch": 0.26858673811118555,
+      "grad_norm": 0.46840107440948486,
+      "learning_rate": 0.0001,
+      "loss": 1.5392,
+      "step": 802
+    },
+    {
+      "epoch": 0.26892163429336907,
+      "grad_norm": 0.4613092839717865,
+      "learning_rate": 0.0001,
+      "loss": 1.6126,
+      "step": 803
+    },
+    {
+      "epoch": 0.2692565304755526,
+      "grad_norm": 0.46580761671066284,
+      "learning_rate": 0.0001,
+      "loss": 1.5783,
+      "step": 804
+    },
+    {
+      "epoch": 0.2695914266577361,
+      "grad_norm": 0.43385007977485657,
+      "learning_rate": 0.0001,
+      "loss": 1.5475,
+      "step": 805
+    },
+    {
+      "epoch": 0.26992632283991963,
+      "grad_norm": 0.44625353813171387,
+      "learning_rate": 0.0001,
+      "loss": 1.5473,
+      "step": 806
+    },
+    {
+      "epoch": 0.27026121902210315,
+      "grad_norm": 0.43871554732322693,
+      "learning_rate": 0.0001,
+      "loss": 1.5818,
+      "step": 807
+    },
+    {
+      "epoch": 0.27059611520428667,
+      "grad_norm": 0.44773444533348083,
+      "learning_rate": 0.0001,
+      "loss": 1.6033,
+      "step": 808
+    },
+    {
+      "epoch": 0.2709310113864702,
+      "grad_norm": 0.4612893760204315,
+      "learning_rate": 0.0001,
+      "loss": 1.597,
+      "step": 809
+    },
+    {
+      "epoch": 0.2712659075686537,
+      "grad_norm": 0.44292980432510376,
+      "learning_rate": 0.0001,
+      "loss": 1.5572,
+      "step": 810
+    },
+    {
+      "epoch": 0.2716008037508372,
+      "grad_norm": 0.44339171051979065,
+      "learning_rate": 0.0001,
+      "loss": 1.569,
+      "step": 811
+    },
+    {
+      "epoch": 0.27193569993302075,
+      "grad_norm": 0.46169915795326233,
+      "learning_rate": 0.0001,
+      "loss": 1.6067,
+      "step": 812
+    },
+    {
+      "epoch": 0.27227059611520427,
+      "grad_norm": 0.4542236924171448,
+      "learning_rate": 0.0001,
+      "loss": 1.5463,
+      "step": 813
+    },
+    {
+      "epoch": 0.2726054922973878,
+      "grad_norm": 0.45044103264808655,
+      "learning_rate": 0.0001,
+      "loss": 1.515,
+      "step": 814
+    },
+    {
+      "epoch": 0.2729403884795713,
+      "grad_norm": 0.4549798369407654,
+      "learning_rate": 0.0001,
+      "loss": 1.5613,
+      "step": 815
+    },
+    {
+      "epoch": 0.2732752846617549,
+      "grad_norm": 0.4387785792350769,
+      "learning_rate": 0.0001,
+      "loss": 1.6267,
+      "step": 816
+    },
+    {
+      "epoch": 0.2736101808439384,
+      "grad_norm": 0.44183585047721863,
+      "learning_rate": 0.0001,
+      "loss": 1.584,
+      "step": 817
+    },
+    {
+      "epoch": 0.2739450770261219,
+      "grad_norm": 0.4401913285255432,
+      "learning_rate": 0.0001,
+      "loss": 1.4822,
+      "step": 818
+    },
+    {
+      "epoch": 0.27427997320830544,
+      "grad_norm": 0.44586020708084106,
+      "learning_rate": 0.0001,
+      "loss": 1.5685,
+      "step": 819
+    },
+    {
+      "epoch": 0.27461486939048896,
+      "grad_norm": 0.44778352975845337,
+      "learning_rate": 0.0001,
+      "loss": 1.6375,
+      "step": 820
+    },
+    {
+      "epoch": 0.2749497655726725,
+      "grad_norm": 0.4404751658439636,
+      "learning_rate": 0.0001,
+      "loss": 1.5043,
+      "step": 821
+    },
+    {
+      "epoch": 0.275284661754856,
+      "grad_norm": 0.4170089066028595,
+      "learning_rate": 0.0001,
+      "loss": 1.4737,
+      "step": 822
+    },
+    {
+      "epoch": 0.2756195579370395,
+      "grad_norm": 0.43222057819366455,
+      "learning_rate": 0.0001,
+      "loss": 1.5244,
+      "step": 823
+    },
+    {
+      "epoch": 0.27595445411922304,
+      "grad_norm": 0.43604615330696106,
+      "learning_rate": 0.0001,
+      "loss": 1.5028,
+      "step": 824
+    },
+    {
+      "epoch": 0.27628935030140656,
+      "grad_norm": 0.46478959918022156,
+      "learning_rate": 0.0001,
+      "loss": 1.5599,
+      "step": 825
+    },
+    {
+      "epoch": 0.2766242464835901,
+      "grad_norm": 0.4629174768924713,
+      "learning_rate": 0.0001,
+      "loss": 1.5553,
+      "step": 826
+    },
+    {
+      "epoch": 0.2769591426657736,
+      "grad_norm": 0.4615054130554199,
+      "learning_rate": 0.0001,
+      "loss": 1.6052,
+      "step": 827
+    },
+    {
+      "epoch": 0.2772940388479571,
+      "grad_norm": 0.4389117360115051,
+      "learning_rate": 0.0001,
+      "loss": 1.5926,
+      "step": 828
+    },
+    {
+      "epoch": 0.27762893503014063,
+      "grad_norm": 0.4443889558315277,
+      "learning_rate": 0.0001,
+      "loss": 1.5341,
+      "step": 829
+    },
+    {
+      "epoch": 0.27796383121232415,
+      "grad_norm": 0.43649813532829285,
+      "learning_rate": 0.0001,
+      "loss": 1.5323,
+      "step": 830
+    },
+    {
+      "epoch": 0.27829872739450773,
+      "grad_norm": 0.4427342414855957,
+      "learning_rate": 0.0001,
+      "loss": 1.5615,
+      "step": 831
+    },
+    {
+      "epoch": 0.27863362357669125,
+      "grad_norm": 0.44088149070739746,
+      "learning_rate": 0.0001,
+      "loss": 1.5227,
+      "step": 832
+    },
+    {
+      "epoch": 0.27896851975887477,
+      "grad_norm": 0.478352427482605,
+      "learning_rate": 0.0001,
+      "loss": 1.6994,
+      "step": 833
+    },
+    {
+      "epoch": 0.2793034159410583,
+      "grad_norm": 0.45829662680625916,
+      "learning_rate": 0.0001,
+      "loss": 1.6766,
+      "step": 834
+    },
+    {
+      "epoch": 0.2796383121232418,
+      "grad_norm": 0.44967886805534363,
+      "learning_rate": 0.0001,
+      "loss": 1.4908,
+      "step": 835
+    },
+    {
+      "epoch": 0.2799732083054253,
+      "grad_norm": 0.45656687021255493,
+      "learning_rate": 0.0001,
+      "loss": 1.5392,
+      "step": 836
+    },
+    {
+      "epoch": 0.28030810448760884,
+      "grad_norm": 0.4756939113140106,
+      "learning_rate": 0.0001,
+      "loss": 1.5648,
+      "step": 837
+    },
+    {
+      "epoch": 0.28064300066979236,
+      "grad_norm": 0.4533655047416687,
+      "learning_rate": 0.0001,
+      "loss": 1.5131,
+      "step": 838
+    },
+    {
+      "epoch": 0.2809778968519759,
+      "grad_norm": 0.4716136157512665,
+      "learning_rate": 0.0001,
+      "loss": 1.6867,
+      "step": 839
+    },
+    {
+      "epoch": 0.2813127930341594,
+      "grad_norm": 0.43292513489723206,
+      "learning_rate": 0.0001,
+      "loss": 1.5307,
+      "step": 840
+    },
+    {
+      "epoch": 0.2816476892163429,
+      "grad_norm": 0.45670297741889954,
+      "learning_rate": 0.0001,
+      "loss": 1.5685,
+      "step": 841
+    },
+    {
+      "epoch": 0.28198258539852644,
+      "grad_norm": 0.43542346358299255,
+      "learning_rate": 0.0001,
+      "loss": 1.532,
+      "step": 842
+    },
+    {
+      "epoch": 0.28231748158070996,
+      "grad_norm": 0.44551143050193787,
+      "learning_rate": 0.0001,
+      "loss": 1.5195,
+      "step": 843
+    },
+    {
+      "epoch": 0.2826523777628935,
+      "grad_norm": 0.4447404444217682,
+      "learning_rate": 0.0001,
+      "loss": 1.4835,
+      "step": 844
+    },
+    {
+      "epoch": 0.282987273945077,
+      "grad_norm": 0.45286113023757935,
+      "learning_rate": 0.0001,
+      "loss": 1.6399,
+      "step": 845
+    },
+    {
+      "epoch": 0.2833221701272606,
+      "grad_norm": 0.4515502154827118,
+      "learning_rate": 0.0001,
+      "loss": 1.5199,
+      "step": 846
+    },
+    {
+      "epoch": 0.2836570663094441,
+      "grad_norm": 0.4420303404331207,
+      "learning_rate": 0.0001,
+      "loss": 1.4901,
+      "step": 847
+    },
+    {
+      "epoch": 0.2839919624916276,
+      "grad_norm": 0.4807213544845581,
+      "learning_rate": 0.0001,
+      "loss": 1.6002,
+      "step": 848
+    },
+    {
+      "epoch": 0.28432685867381113,
+      "grad_norm": 0.4393681585788727,
+      "learning_rate": 0.0001,
+      "loss": 1.595,
+      "step": 849
+    },
+    {
+      "epoch": 0.28466175485599465,
+      "grad_norm": 0.4498080313205719,
+      "learning_rate": 0.0001,
+      "loss": 1.5361,
+      "step": 850
+    },
+    {
+      "epoch": 0.2849966510381782,
+      "grad_norm": 0.4682947099208832,
+      "learning_rate": 0.0001,
+      "loss": 1.6209,
+      "step": 851
+    },
+    {
+      "epoch": 0.2853315472203617,
+      "grad_norm": 0.43859928846359253,
+      "learning_rate": 0.0001,
+      "loss": 1.526,
+      "step": 852
+    },
+    {
+      "epoch": 0.2856664434025452,
+      "grad_norm": 0.4339533746242523,
+      "learning_rate": 0.0001,
+      "loss": 1.5533,
+      "step": 853
+    },
+    {
+      "epoch": 0.28600133958472873,
+      "grad_norm": 0.45164206624031067,
+      "learning_rate": 0.0001,
+      "loss": 1.6037,
+      "step": 854
+    },
+    {
+      "epoch": 0.28633623576691225,
+      "grad_norm": 0.4741949737071991,
+      "learning_rate": 0.0001,
+      "loss": 1.594,
+      "step": 855
+    },
+    {
+      "epoch": 0.28667113194909577,
+      "grad_norm": 0.46638429164886475,
+      "learning_rate": 0.0001,
+      "loss": 1.5615,
+      "step": 856
+    },
+    {
+      "epoch": 0.2870060281312793,
+      "grad_norm": 0.46086496114730835,
+      "learning_rate": 0.0001,
+      "loss": 1.5066,
+      "step": 857
+    },
+    {
+      "epoch": 0.2873409243134628,
+      "grad_norm": 0.4441535174846649,
+      "learning_rate": 0.0001,
+      "loss": 1.5619,
+      "step": 858
+    },
+    {
+      "epoch": 0.28767582049564633,
+      "grad_norm": 0.44475072622299194,
+      "learning_rate": 0.0001,
+      "loss": 1.6023,
+      "step": 859
+    },
+    {
+      "epoch": 0.28801071667782985,
+      "grad_norm": 0.45497584342956543,
+      "learning_rate": 0.0001,
+      "loss": 1.5075,
+      "step": 860
+    },
+    {
+      "epoch": 0.2883456128600134,
+      "grad_norm": 0.4760246276855469,
+      "learning_rate": 0.0001,
+      "loss": 1.5075,
+      "step": 861
+    },
+    {
+      "epoch": 0.28868050904219694,
+      "grad_norm": 0.46818268299102783,
+      "learning_rate": 0.0001,
+      "loss": 1.6088,
+      "step": 862
+    },
+    {
+      "epoch": 0.28901540522438046,
+      "grad_norm": 0.45602530241012573,
+      "learning_rate": 0.0001,
+      "loss": 1.5326,
+      "step": 863
+    },
+    {
+      "epoch": 0.289350301406564,
+      "grad_norm": 0.4454081654548645,
+      "learning_rate": 0.0001,
+      "loss": 1.5957,
+      "step": 864
+    },
+    {
+      "epoch": 0.2896851975887475,
+      "grad_norm": 0.45742931962013245,
+      "learning_rate": 0.0001,
+      "loss": 1.5309,
+      "step": 865
+    },
+    {
+      "epoch": 0.290020093770931,
+      "grad_norm": 0.45301052927970886,
+      "learning_rate": 0.0001,
+      "loss": 1.6096,
+      "step": 866
+    },
+    {
+      "epoch": 0.29035498995311454,
+      "grad_norm": 0.45820847153663635,
+      "learning_rate": 0.0001,
+      "loss": 1.5232,
+      "step": 867
+    },
+    {
+      "epoch": 0.29068988613529806,
+      "grad_norm": 0.47826266288757324,
+      "learning_rate": 0.0001,
+      "loss": 1.6533,
+      "step": 868
+    },
+    {
+      "epoch": 0.2910247823174816,
+      "grad_norm": 0.43897736072540283,
+      "learning_rate": 0.0001,
+      "loss": 1.5567,
+      "step": 869
+    },
+    {
+      "epoch": 0.2913596784996651,
+      "grad_norm": 0.43555381894111633,
+      "learning_rate": 0.0001,
+      "loss": 1.5425,
+      "step": 870
+    },
+    {
+      "epoch": 0.2916945746818486,
+      "grad_norm": 0.4390638768672943,
+      "learning_rate": 0.0001,
+      "loss": 1.6117,
+      "step": 871
+    },
+    {
+      "epoch": 0.29202947086403214,
+      "grad_norm": 0.4563729465007782,
+      "learning_rate": 0.0001,
+      "loss": 1.676,
+      "step": 872
+    },
+    {
+      "epoch": 0.29236436704621566,
+      "grad_norm": 0.44380465149879456,
+      "learning_rate": 0.0001,
+      "loss": 1.4597,
+      "step": 873
+    },
+    {
+      "epoch": 0.2926992632283992,
+      "grad_norm": 0.5023241639137268,
+      "learning_rate": 0.0001,
+      "loss": 1.5869,
+      "step": 874
+    },
+    {
+      "epoch": 0.2930341594105827,
+      "grad_norm": 0.4469534158706665,
+      "learning_rate": 0.0001,
+      "loss": 1.4953,
+      "step": 875
+    },
+    {
+      "epoch": 0.29336905559276627,
+      "grad_norm": 0.4469916820526123,
+      "learning_rate": 0.0001,
+      "loss": 1.5673,
+      "step": 876
+    },
+    {
+      "epoch": 0.2937039517749498,
+      "grad_norm": 0.46404793858528137,
+      "learning_rate": 0.0001,
+      "loss": 1.5858,
+      "step": 877
+    },
+    {
+      "epoch": 0.2940388479571333,
+      "grad_norm": 0.45173484086990356,
+      "learning_rate": 0.0001,
+      "loss": 1.6339,
+      "step": 878
+    },
+    {
+      "epoch": 0.29437374413931683,
+      "grad_norm": 0.45506271719932556,
+      "learning_rate": 0.0001,
+      "loss": 1.6059,
+      "step": 879
+    },
+    {
+      "epoch": 0.29470864032150035,
+      "grad_norm": 0.46061256527900696,
+      "learning_rate": 0.0001,
+      "loss": 1.5897,
+      "step": 880
+    },
+    {
+      "epoch": 0.29504353650368387,
+      "grad_norm": 0.46344825625419617,
+      "learning_rate": 0.0001,
+      "loss": 1.6338,
+      "step": 881
+    },
+    {
+      "epoch": 0.2953784326858674,
+      "grad_norm": 0.43154940009117126,
+      "learning_rate": 0.0001,
+      "loss": 1.4563,
+      "step": 882
+    },
+    {
+      "epoch": 0.2957133288680509,
+      "grad_norm": 0.43573182821273804,
+      "learning_rate": 0.0001,
+      "loss": 1.4755,
+      "step": 883
+    },
+    {
+      "epoch": 0.2960482250502344,
+      "grad_norm": 0.4424009919166565,
+      "learning_rate": 0.0001,
+      "loss": 1.6018,
+      "step": 884
+    },
+    {
+      "epoch": 0.29638312123241795,
+      "grad_norm": 0.4376654326915741,
+      "learning_rate": 0.0001,
+      "loss": 1.5291,
+      "step": 885
+    },
+    {
+      "epoch": 0.29671801741460146,
+      "grad_norm": 0.44850626587867737,
+      "learning_rate": 0.0001,
+      "loss": 1.5544,
+      "step": 886
+    },
+    {
+      "epoch": 0.297052913596785,
+      "grad_norm": 0.44172677397727966,
+      "learning_rate": 0.0001,
+      "loss": 1.5138,
+      "step": 887
+    },
+    {
+      "epoch": 0.2973878097789685,
+      "grad_norm": 0.4760349690914154,
+      "learning_rate": 0.0001,
+      "loss": 1.6774,
+      "step": 888
+    },
+    {
+      "epoch": 0.297722705961152,
+      "grad_norm": 0.4951341152191162,
+      "learning_rate": 0.0001,
+      "loss": 1.5995,
+      "step": 889
+    },
+    {
+      "epoch": 0.29805760214333554,
+      "grad_norm": 0.46811017394065857,
+      "learning_rate": 0.0001,
+      "loss": 1.6005,
+      "step": 890
+    },
+    {
+      "epoch": 0.29839249832551906,
+      "grad_norm": 0.46500176191329956,
+      "learning_rate": 0.0001,
+      "loss": 1.61,
+      "step": 891
+    },
+    {
+      "epoch": 0.29872739450770264,
+      "grad_norm": 0.4412824511528015,
+      "learning_rate": 0.0001,
+      "loss": 1.56,
+      "step": 892
+    },
+    {
+      "epoch": 0.29906229068988616,
+      "grad_norm": 0.4398655891418457,
+      "learning_rate": 0.0001,
+      "loss": 1.5329,
+      "step": 893
+    },
+    {
+      "epoch": 0.2993971868720697,
+      "grad_norm": 0.4386952817440033,
+      "learning_rate": 0.0001,
+      "loss": 1.5272,
+      "step": 894
+    },
+    {
+      "epoch": 0.2997320830542532,
+      "grad_norm": 0.47550323605537415,
+      "learning_rate": 0.0001,
+      "loss": 1.5591,
+      "step": 895
+    },
+    {
+      "epoch": 0.3000669792364367,
+      "grad_norm": 0.43544065952301025,
+      "learning_rate": 0.0001,
+      "loss": 1.5343,
+      "step": 896
+    },
+    {
+      "epoch": 0.30040187541862023,
+      "grad_norm": 0.43268048763275146,
+      "learning_rate": 0.0001,
+      "loss": 1.4944,
+      "step": 897
+    },
+    {
+      "epoch": 0.30073677160080375,
+      "grad_norm": 0.4282139837741852,
+      "learning_rate": 0.0001,
+      "loss": 1.4313,
+      "step": 898
+    },
+    {
+      "epoch": 0.3010716677829873,
+      "grad_norm": 0.46300849318504333,
+      "learning_rate": 0.0001,
+      "loss": 1.6772,
+      "step": 899
+    },
+    {
+      "epoch": 0.3014065639651708,
+      "grad_norm": 0.4673476815223694,
+      "learning_rate": 0.0001,
+      "loss": 1.6028,
+      "step": 900
+    },
+    {
+      "epoch": 0.3017414601473543,
+      "grad_norm": 0.43645891547203064,
+      "learning_rate": 0.0001,
+      "loss": 1.5157,
+      "step": 901
+    },
+    {
+      "epoch": 0.30207635632953783,
+      "grad_norm": 0.44053512811660767,
+      "learning_rate": 0.0001,
+      "loss": 1.5987,
+      "step": 902
+    },
+    {
+      "epoch": 0.30241125251172135,
+      "grad_norm": 0.43275871872901917,
+      "learning_rate": 0.0001,
+      "loss": 1.4311,
+      "step": 903
+    },
+    {
+      "epoch": 0.30274614869390487,
+      "grad_norm": 0.45569944381713867,
+      "learning_rate": 0.0001,
+      "loss": 1.6145,
+      "step": 904
+    },
+    {
+      "epoch": 0.3030810448760884,
+      "grad_norm": 0.45299553871154785,
+      "learning_rate": 0.0001,
+      "loss": 1.5617,
+      "step": 905
+    },
+    {
+      "epoch": 0.3034159410582719,
+      "grad_norm": 0.4409105181694031,
+      "learning_rate": 0.0001,
+      "loss": 1.5418,
+      "step": 906
+    },
+    {
+      "epoch": 0.3037508372404555,
+      "grad_norm": 0.4408958852291107,
+      "learning_rate": 0.0001,
+      "loss": 1.4375,
+      "step": 907
+    },
+    {
+      "epoch": 0.304085733422639,
+      "grad_norm": 0.4480845332145691,
+      "learning_rate": 0.0001,
+      "loss": 1.5399,
+      "step": 908
+    },
+    {
+      "epoch": 0.3044206296048225,
+      "grad_norm": 0.4365403950214386,
+      "learning_rate": 0.0001,
+      "loss": 1.4934,
+      "step": 909
+    },
+    {
+      "epoch": 0.30475552578700604,
+      "grad_norm": 0.45147716999053955,
+      "learning_rate": 0.0001,
+      "loss": 1.5526,
+      "step": 910
+    },
+    {
+      "epoch": 0.30509042196918956,
+      "grad_norm": 0.45969611406326294,
+      "learning_rate": 0.0001,
+      "loss": 1.5609,
+      "step": 911
+    },
+    {
+      "epoch": 0.3054253181513731,
+      "grad_norm": 0.4570698142051697,
+      "learning_rate": 0.0001,
+      "loss": 1.626,
+      "step": 912
+    },
+    {
+      "epoch": 0.3057602143335566,
+      "grad_norm": 0.45453599095344543,
+      "learning_rate": 0.0001,
+      "loss": 1.6268,
+      "step": 913
+    },
+    {
+      "epoch": 0.3060951105157401,
+      "grad_norm": 0.4364335536956787,
+      "learning_rate": 0.0001,
+      "loss": 1.5523,
+      "step": 914
+    },
+    {
+      "epoch": 0.30643000669792364,
+      "grad_norm": 0.4353067874908447,
+      "learning_rate": 0.0001,
+      "loss": 1.5668,
+      "step": 915
+    },
+    {
+      "epoch": 0.30676490288010716,
+      "grad_norm": 0.439081072807312,
+      "learning_rate": 0.0001,
+      "loss": 1.4408,
+      "step": 916
+    },
+    {
+      "epoch": 0.3070997990622907,
+      "grad_norm": 0.4517746567726135,
+      "learning_rate": 0.0001,
+      "loss": 1.514,
+      "step": 917
+    },
+    {
+      "epoch": 0.3074346952444742,
+      "grad_norm": 0.4414094090461731,
+      "learning_rate": 0.0001,
+      "loss": 1.4736,
+      "step": 918
+    },
+    {
+      "epoch": 0.3077695914266577,
+      "grad_norm": 0.4406594932079315,
+      "learning_rate": 0.0001,
+      "loss": 1.576,
+      "step": 919
+    },
+    {
+      "epoch": 0.30810448760884124,
+      "grad_norm": 0.46079543232917786,
+      "learning_rate": 0.0001,
+      "loss": 1.5106,
+      "step": 920
+    },
+    {
+      "epoch": 0.30843938379102476,
+      "grad_norm": 0.4498567283153534,
+      "learning_rate": 0.0001,
+      "loss": 1.5298,
+      "step": 921
+    },
+    {
+      "epoch": 0.30877427997320833,
+      "grad_norm": 0.45379623770713806,
+      "learning_rate": 0.0001,
+      "loss": 1.5265,
+      "step": 922
+    },
+    {
+      "epoch": 0.30910917615539185,
+      "grad_norm": 0.4834693372249603,
+      "learning_rate": 0.0001,
+      "loss": 1.6934,
+      "step": 923
+    },
+    {
+      "epoch": 0.30944407233757537,
+      "grad_norm": 0.46280190348625183,
+      "learning_rate": 0.0001,
+      "loss": 1.552,
+      "step": 924
+    },
+    {
+      "epoch": 0.3097789685197589,
+      "grad_norm": 0.4772225320339203,
+      "learning_rate": 0.0001,
+      "loss": 1.5628,
+      "step": 925
+    },
+    {
+      "epoch": 0.3101138647019424,
+      "grad_norm": 0.45620399713516235,
+      "learning_rate": 0.0001,
+      "loss": 1.5828,
+      "step": 926
+    },
+    {
+      "epoch": 0.31044876088412593,
+      "grad_norm": 0.45849597454071045,
+      "learning_rate": 0.0001,
+      "loss": 1.4951,
+      "step": 927
+    },
+    {
+      "epoch": 0.31078365706630945,
+      "grad_norm": 0.46801209449768066,
+      "learning_rate": 0.0001,
+      "loss": 1.5744,
+      "step": 928
+    },
+    {
+      "epoch": 0.31111855324849297,
+      "grad_norm": 0.4555777311325073,
+      "learning_rate": 0.0001,
+      "loss": 1.4925,
+      "step": 929
+    },
+    {
+      "epoch": 0.3114534494306765,
+      "grad_norm": 0.4439859688282013,
+      "learning_rate": 0.0001,
+      "loss": 1.5912,
+      "step": 930
+    },
+    {
+      "epoch": 0.31178834561286,
+      "grad_norm": 0.4438050389289856,
+      "learning_rate": 0.0001,
+      "loss": 1.5269,
+      "step": 931
+    },
+    {
+      "epoch": 0.3121232417950435,
+      "grad_norm": 0.4439409077167511,
+      "learning_rate": 0.0001,
+      "loss": 1.6237,
+      "step": 932
+    },
+    {
+      "epoch": 0.31245813797722705,
+      "grad_norm": 0.43066197633743286,
+      "learning_rate": 0.0001,
+      "loss": 1.5733,
+      "step": 933
+    },
+    {
+      "epoch": 0.31279303415941057,
+      "grad_norm": 0.42914170026779175,
+      "learning_rate": 0.0001,
+      "loss": 1.5052,
+      "step": 934
+    },
+    {
+      "epoch": 0.3131279303415941,
+      "grad_norm": 0.4268810749053955,
+      "learning_rate": 0.0001,
+      "loss": 1.544,
+      "step": 935
+    },
+    {
+      "epoch": 0.3134628265237776,
+      "grad_norm": 0.4368149936199188,
+      "learning_rate": 0.0001,
+      "loss": 1.622,
+      "step": 936
+    },
+    {
+      "epoch": 0.3137977227059612,
+      "grad_norm": 0.441754013299942,
+      "learning_rate": 0.0001,
+      "loss": 1.4592,
+      "step": 937
+    },
+    {
+      "epoch": 0.3141326188881447,
+      "grad_norm": 0.43533989787101746,
+      "learning_rate": 0.0001,
+      "loss": 1.5468,
+      "step": 938
+    },
+    {
+      "epoch": 0.3144675150703282,
+      "grad_norm": 0.4400700032711029,
+      "learning_rate": 0.0001,
+      "loss": 1.5719,
+      "step": 939
+    },
+    {
+      "epoch": 0.31480241125251174,
+      "grad_norm": 0.4473206400871277,
+      "learning_rate": 0.0001,
+      "loss": 1.5889,
+      "step": 940
+    },
+    {
+      "epoch": 0.31513730743469526,
+      "grad_norm": 0.4529035687446594,
+      "learning_rate": 0.0001,
+      "loss": 1.5555,
+      "step": 941
+    },
+    {
+      "epoch": 0.3154722036168788,
+      "grad_norm": 0.45281359553337097,
+      "learning_rate": 0.0001,
+      "loss": 1.6016,
+      "step": 942
+    },
+    {
+      "epoch": 0.3158070997990623,
+      "grad_norm": 0.4597030282020569,
+      "learning_rate": 0.0001,
+      "loss": 1.5947,
+      "step": 943
+    },
+    {
+      "epoch": 0.3161419959812458,
+      "grad_norm": 0.4365510642528534,
+      "learning_rate": 0.0001,
+      "loss": 1.4508,
+      "step": 944
+    },
+    {
+      "epoch": 0.31647689216342934,
+      "grad_norm": 0.44290444254875183,
+      "learning_rate": 0.0001,
+      "loss": 1.5718,
+      "step": 945
+    },
+    {
+      "epoch": 0.31681178834561285,
+      "grad_norm": 0.4928835332393646,
+      "learning_rate": 0.0001,
+      "loss": 1.5904,
+      "step": 946
+    },
+    {
+      "epoch": 0.3171466845277964,
+      "grad_norm": 0.44557881355285645,
+      "learning_rate": 0.0001,
+      "loss": 1.5851,
+      "step": 947
+    },
+    {
+      "epoch": 0.3174815807099799,
+      "grad_norm": 0.5136463046073914,
+      "learning_rate": 0.0001,
+      "loss": 1.5845,
+      "step": 948
+    },
+    {
+      "epoch": 0.3178164768921634,
+      "grad_norm": 0.4590359032154083,
+      "learning_rate": 0.0001,
+      "loss": 1.5975,
+      "step": 949
+    },
+    {
+      "epoch": 0.31815137307434693,
+      "grad_norm": 0.44247373938560486,
+      "learning_rate": 0.0001,
+      "loss": 1.458,
+      "step": 950
+    },
+    {
+      "epoch": 0.31848626925653045,
+      "grad_norm": 0.43431544303894043,
+      "learning_rate": 0.0001,
+      "loss": 1.4779,
+      "step": 951
+    },
+    {
+      "epoch": 0.31882116543871397,
+      "grad_norm": 0.4587078392505646,
+      "learning_rate": 0.0001,
+      "loss": 1.5128,
+      "step": 952
+    },
+    {
+      "epoch": 0.31915606162089755,
+      "grad_norm": 0.44057053327560425,
+      "learning_rate": 0.0001,
+      "loss": 1.603,
+      "step": 953
+    },
+    {
+      "epoch": 0.31949095780308107,
+      "grad_norm": 0.43187084794044495,
+      "learning_rate": 0.0001,
+      "loss": 1.5044,
+      "step": 954
+    },
+    {
+      "epoch": 0.3198258539852646,
+      "grad_norm": 0.4155219793319702,
+      "learning_rate": 0.0001,
+      "loss": 1.471,
+      "step": 955
+    },
+    {
+      "epoch": 0.3201607501674481,
+      "grad_norm": 0.47326022386550903,
+      "learning_rate": 0.0001,
+      "loss": 1.5794,
+      "step": 956
+    },
+    {
+      "epoch": 0.3204956463496316,
+      "grad_norm": 0.45905107259750366,
+      "learning_rate": 0.0001,
+      "loss": 1.5353,
+      "step": 957
+    },
+    {
+      "epoch": 0.32083054253181514,
+      "grad_norm": 0.44687706232070923,
+      "learning_rate": 0.0001,
+      "loss": 1.5249,
+      "step": 958
+    },
+    {
+      "epoch": 0.32116543871399866,
+      "grad_norm": 0.46414482593536377,
+      "learning_rate": 0.0001,
+      "loss": 1.5599,
+      "step": 959
+    },
+    {
+      "epoch": 0.3215003348961822,
+      "grad_norm": 0.44584473967552185,
+      "learning_rate": 0.0001,
+      "loss": 1.5236,
+      "step": 960
+    },
+    {
+      "epoch": 0.3218352310783657,
+      "grad_norm": 0.49396583437919617,
+      "learning_rate": 0.0001,
+      "loss": 1.6187,
+      "step": 961
+    },
+    {
+      "epoch": 0.3221701272605492,
+      "grad_norm": 0.45838162302970886,
+      "learning_rate": 0.0001,
+      "loss": 1.5558,
+      "step": 962
+    },
+    {
+      "epoch": 0.32250502344273274,
+      "grad_norm": 0.4484620988368988,
+      "learning_rate": 0.0001,
+      "loss": 1.5179,
+      "step": 963
+    },
+    {
+      "epoch": 0.32283991962491626,
+      "grad_norm": 0.45024028420448303,
+      "learning_rate": 0.0001,
+      "loss": 1.5206,
+      "step": 964
+    },
+    {
+      "epoch": 0.3231748158070998,
+      "grad_norm": 0.43294331431388855,
+      "learning_rate": 0.0001,
+      "loss": 1.6291,
+      "step": 965
+    },
+    {
+      "epoch": 0.3235097119892833,
+      "grad_norm": 0.4585472345352173,
+      "learning_rate": 0.0001,
+      "loss": 1.545,
+      "step": 966
+    },
+    {
+      "epoch": 0.3238446081714668,
+      "grad_norm": 0.44397369027137756,
+      "learning_rate": 0.0001,
+      "loss": 1.5551,
+      "step": 967
+    },
+    {
+      "epoch": 0.3241795043536504,
+      "grad_norm": 0.462065190076828,
+      "learning_rate": 0.0001,
+      "loss": 1.5943,
+      "step": 968
+    },
+    {
+      "epoch": 0.3245144005358339,
+      "grad_norm": 0.4261050820350647,
+      "learning_rate": 0.0001,
+      "loss": 1.4429,
+      "step": 969
+    },
+    {
+      "epoch": 0.32484929671801743,
+      "grad_norm": 0.44713786244392395,
+      "learning_rate": 0.0001,
+      "loss": 1.6157,
+      "step": 970
+    },
+    {
+      "epoch": 0.32518419290020095,
+      "grad_norm": 0.46949875354766846,
+      "learning_rate": 0.0001,
+      "loss": 1.579,
+      "step": 971
+    },
+    {
+      "epoch": 0.32551908908238447,
+      "grad_norm": 0.4351710379123688,
+      "learning_rate": 0.0001,
+      "loss": 1.4521,
+      "step": 972
+    },
+    {
+      "epoch": 0.325853985264568,
+      "grad_norm": 0.5096984505653381,
+      "learning_rate": 0.0001,
+      "loss": 1.6964,
+      "step": 973
+    },
+    {
+      "epoch": 0.3261888814467515,
+      "grad_norm": 0.44196903705596924,
+      "learning_rate": 0.0001,
+      "loss": 1.5528,
+      "step": 974
+    },
+    {
+      "epoch": 0.32652377762893503,
+      "grad_norm": 0.4667568802833557,
+      "learning_rate": 0.0001,
+      "loss": 1.5643,
+      "step": 975
+    },
+    {
+      "epoch": 0.32685867381111855,
+      "grad_norm": 0.46379634737968445,
+      "learning_rate": 0.0001,
+      "loss": 1.6277,
+      "step": 976
+    },
+    {
+      "epoch": 0.32719356999330207,
+      "grad_norm": 0.451865553855896,
+      "learning_rate": 0.0001,
+      "loss": 1.6356,
+      "step": 977
+    },
+    {
+      "epoch": 0.3275284661754856,
+      "grad_norm": 0.42902424931526184,
+      "learning_rate": 0.0001,
+      "loss": 1.5017,
+      "step": 978
+    },
+    {
+      "epoch": 0.3278633623576691,
+      "grad_norm": 0.46481791138648987,
+      "learning_rate": 0.0001,
+      "loss": 1.5955,
+      "step": 979
+    },
+    {
+      "epoch": 0.3281982585398526,
+      "grad_norm": 0.45368990302085876,
+      "learning_rate": 0.0001,
+      "loss": 1.6469,
+      "step": 980
+    },
+    {
+      "epoch": 0.32853315472203615,
+      "grad_norm": 0.46452459692955017,
+      "learning_rate": 0.0001,
+      "loss": 1.5121,
+      "step": 981
+    },
+    {
+      "epoch": 0.32886805090421967,
+      "grad_norm": 0.4299396276473999,
+      "learning_rate": 0.0001,
+      "loss": 1.4892,
+      "step": 982
+    },
+    {
+      "epoch": 0.32920294708640324,
+      "grad_norm": 0.46773141622543335,
+      "learning_rate": 0.0001,
+      "loss": 1.6213,
+      "step": 983
+    },
+    {
+      "epoch": 0.32953784326858676,
+      "grad_norm": 0.4707593023777008,
+      "learning_rate": 0.0001,
+      "loss": 1.672,
+      "step": 984
+    },
+    {
+      "epoch": 0.3298727394507703,
+      "grad_norm": 0.46293628215789795,
+      "learning_rate": 0.0001,
+      "loss": 1.6151,
+      "step": 985
+    },
+    {
+      "epoch": 0.3302076356329538,
+      "grad_norm": 0.4385524392127991,
+      "learning_rate": 0.0001,
+      "loss": 1.4511,
+      "step": 986
+    },
+    {
+      "epoch": 0.3305425318151373,
+      "grad_norm": 0.43595242500305176,
+      "learning_rate": 0.0001,
+      "loss": 1.4555,
+      "step": 987
+    },
+    {
+      "epoch": 0.33087742799732084,
+      "grad_norm": 0.4468819499015808,
+      "learning_rate": 0.0001,
+      "loss": 1.6023,
+      "step": 988
+    },
+    {
+      "epoch": 0.33121232417950436,
+      "grad_norm": 0.45761582255363464,
+      "learning_rate": 0.0001,
+      "loss": 1.5441,
+      "step": 989
+    },
+    {
+      "epoch": 0.3315472203616879,
+      "grad_norm": 0.4671020209789276,
+      "learning_rate": 0.0001,
+      "loss": 1.633,
+      "step": 990
+    },
+    {
+      "epoch": 0.3318821165438714,
+      "grad_norm": 0.4528156518936157,
+      "learning_rate": 0.0001,
+      "loss": 1.4952,
+      "step": 991
+    },
+    {
+      "epoch": 0.3322170127260549,
+      "grad_norm": 0.4482688009738922,
+      "learning_rate": 0.0001,
+      "loss": 1.537,
+      "step": 992
+    },
+    {
+      "epoch": 0.33255190890823844,
+      "grad_norm": 0.43465572595596313,
+      "learning_rate": 0.0001,
+      "loss": 1.4921,
+      "step": 993
+    },
+    {
+      "epoch": 0.33288680509042196,
+      "grad_norm": 0.4700695872306824,
+      "learning_rate": 0.0001,
+      "loss": 1.6168,
+      "step": 994
+    },
+    {
+      "epoch": 0.3332217012726055,
+      "grad_norm": 0.4559037387371063,
+      "learning_rate": 0.0001,
+      "loss": 1.6104,
+      "step": 995
+    },
+    {
+      "epoch": 0.333556597454789,
+      "grad_norm": 0.44127675890922546,
+      "learning_rate": 0.0001,
+      "loss": 1.5354,
+      "step": 996
+    },
+    {
+      "epoch": 0.3338914936369725,
+      "grad_norm": 0.4611609876155853,
+      "learning_rate": 0.0001,
+      "loss": 1.6717,
+      "step": 997
+    },
+    {
+      "epoch": 0.3342263898191561,
+      "grad_norm": 0.4240245223045349,
+      "learning_rate": 0.0001,
+      "loss": 1.4703,
+      "step": 998
+    },
+    {
+      "epoch": 0.3345612860013396,
+      "grad_norm": 0.4530922472476959,
+      "learning_rate": 0.0001,
+      "loss": 1.5208,
+      "step": 999
+    },
+    {
+      "epoch": 0.33489618218352313,
+      "grad_norm": 0.4416406452655792,
+      "learning_rate": 0.0001,
+      "loss": 1.5127,
+      "step": 1000
+    },
+    {
+      "epoch": 0.33523107836570665,
+      "grad_norm": 0.45270246267318726,
+      "learning_rate": 0.0001,
+      "loss": 1.5718,
+      "step": 1001
+    },
+    {
+      "epoch": 0.33556597454789017,
+      "grad_norm": 0.439411997795105,
+      "learning_rate": 0.0001,
+      "loss": 1.517,
+      "step": 1002
+    },
+    {
+      "epoch": 0.3359008707300737,
+      "grad_norm": 0.46239691972732544,
+      "learning_rate": 0.0001,
+      "loss": 1.6039,
+      "step": 1003
+    },
+    {
+      "epoch": 0.3362357669122572,
+      "grad_norm": 0.43868276476860046,
+      "learning_rate": 0.0001,
+      "loss": 1.6153,
+      "step": 1004
+    },
+    {
+      "epoch": 0.3365706630944407,
+      "grad_norm": 0.48331135511398315,
+      "learning_rate": 0.0001,
+      "loss": 1.5955,
+      "step": 1005
+    },
+    {
+      "epoch": 0.33690555927662424,
+      "grad_norm": 0.45195087790489197,
+      "learning_rate": 0.0001,
+      "loss": 1.6155,
+      "step": 1006
+    },
+    {
+      "epoch": 0.33724045545880776,
+      "grad_norm": 0.4425548315048218,
+      "learning_rate": 0.0001,
+      "loss": 1.4241,
+      "step": 1007
+    },
+    {
+      "epoch": 0.3375753516409913,
+      "grad_norm": 0.43767687678337097,
+      "learning_rate": 0.0001,
+      "loss": 1.5117,
+      "step": 1008
+    },
+    {
+      "epoch": 0.3379102478231748,
+      "grad_norm": 0.43844422698020935,
+      "learning_rate": 0.0001,
+      "loss": 1.5459,
+      "step": 1009
+    },
+    {
+      "epoch": 0.3382451440053583,
+      "grad_norm": 0.4390891492366791,
+      "learning_rate": 0.0001,
+      "loss": 1.5435,
+      "step": 1010
+    },
+    {
+      "epoch": 0.33858004018754184,
+      "grad_norm": 0.4437672793865204,
+      "learning_rate": 0.0001,
+      "loss": 1.4874,
+      "step": 1011
+    },
+    {
+      "epoch": 0.33891493636972536,
+      "grad_norm": 0.4345516264438629,
+      "learning_rate": 0.0001,
+      "loss": 1.4398,
+      "step": 1012
+    },
+    {
+      "epoch": 0.3392498325519089,
+      "grad_norm": 0.47101935744285583,
+      "learning_rate": 0.0001,
+      "loss": 1.7193,
+      "step": 1013
+    },
+    {
+      "epoch": 0.33958472873409246,
+      "grad_norm": 0.4873376190662384,
+      "learning_rate": 0.0001,
+      "loss": 1.6006,
+      "step": 1014
+    },
+    {
+      "epoch": 0.339919624916276,
+      "grad_norm": 0.4573456943035126,
+      "learning_rate": 0.0001,
+      "loss": 1.4895,
+      "step": 1015
+    },
+    {
+      "epoch": 0.3402545210984595,
+      "grad_norm": 0.4560162425041199,
+      "learning_rate": 0.0001,
+      "loss": 1.6158,
+      "step": 1016
+    },
+    {
+      "epoch": 0.340589417280643,
+      "grad_norm": 0.45291417837142944,
+      "learning_rate": 0.0001,
+      "loss": 1.5367,
+      "step": 1017
+    },
+    {
+      "epoch": 0.34092431346282653,
+      "grad_norm": 0.46238189935684204,
+      "learning_rate": 0.0001,
+      "loss": 1.4627,
+      "step": 1018
+    },
+    {
+      "epoch": 0.34125920964501005,
+      "grad_norm": 0.47080227732658386,
+      "learning_rate": 0.0001,
+      "loss": 1.5742,
+      "step": 1019
+    },
+    {
+      "epoch": 0.3415941058271936,
+      "grad_norm": 0.46811938285827637,
+      "learning_rate": 0.0001,
+      "loss": 1.6345,
+      "step": 1020
+    },
+    {
+      "epoch": 0.3419290020093771,
+      "grad_norm": 0.46588802337646484,
+      "learning_rate": 0.0001,
+      "loss": 1.5783,
+      "step": 1021
+    },
+    {
+      "epoch": 0.3422638981915606,
+      "grad_norm": 0.4479082524776459,
+      "learning_rate": 0.0001,
+      "loss": 1.6295,
+      "step": 1022
+    },
+    {
+      "epoch": 0.34259879437374413,
+      "grad_norm": 0.45164626836776733,
+      "learning_rate": 0.0001,
+      "loss": 1.5263,
+      "step": 1023
+    },
+    {
+      "epoch": 0.34293369055592765,
+      "grad_norm": 0.45851123332977295,
+      "learning_rate": 0.0001,
+      "loss": 1.5631,
+      "step": 1024
+    },
+    {
+      "epoch": 0.34326858673811117,
+      "grad_norm": 0.4423898756504059,
+      "learning_rate": 0.0001,
+      "loss": 1.6015,
+      "step": 1025
+    },
+    {
+      "epoch": 0.3436034829202947,
+      "grad_norm": 0.46684497594833374,
+      "learning_rate": 0.0001,
+      "loss": 1.5274,
+      "step": 1026
+    },
+    {
+      "epoch": 0.3439383791024782,
+      "grad_norm": 0.42938143014907837,
+      "learning_rate": 0.0001,
+      "loss": 1.4798,
+      "step": 1027
+    },
+    {
+      "epoch": 0.34427327528466173,
+      "grad_norm": 0.45635873079299927,
+      "learning_rate": 0.0001,
+      "loss": 1.6065,
+      "step": 1028
+    },
+    {
+      "epoch": 0.3446081714668453,
+      "grad_norm": 0.4600565731525421,
+      "learning_rate": 0.0001,
+      "loss": 1.523,
+      "step": 1029
+    },
+    {
+      "epoch": 0.3449430676490288,
+      "grad_norm": 0.4675713777542114,
+      "learning_rate": 0.0001,
+      "loss": 1.5928,
+      "step": 1030
+    },
+    {
+      "epoch": 0.34527796383121234,
+      "grad_norm": 0.45423001050949097,
+      "learning_rate": 0.0001,
+      "loss": 1.6305,
+      "step": 1031
+    },
+    {
+      "epoch": 0.34561286001339586,
+      "grad_norm": 0.44853097200393677,
+      "learning_rate": 0.0001,
+      "loss": 1.5195,
+      "step": 1032
+    },
+    {
+      "epoch": 0.3459477561955794,
+      "grad_norm": 0.4483386278152466,
+      "learning_rate": 0.0001,
+      "loss": 1.677,
+      "step": 1033
+    },
+    {
+      "epoch": 0.3462826523777629,
+      "grad_norm": 0.43875786662101746,
+      "learning_rate": 0.0001,
+      "loss": 1.5086,
+      "step": 1034
+    },
+    {
+      "epoch": 0.3466175485599464,
+      "grad_norm": 0.44167378544807434,
+      "learning_rate": 0.0001,
+      "loss": 1.6012,
+      "step": 1035
+    },
+    {
+      "epoch": 0.34695244474212994,
+      "grad_norm": 0.4552042782306671,
+      "learning_rate": 0.0001,
+      "loss": 1.6003,
+      "step": 1036
+    },
+    {
+      "epoch": 0.34728734092431346,
+      "grad_norm": 0.4757082462310791,
+      "learning_rate": 0.0001,
+      "loss": 1.6217,
+      "step": 1037
+    },
+    {
+      "epoch": 0.347622237106497,
+      "grad_norm": 0.4523075819015503,
+      "learning_rate": 0.0001,
+      "loss": 1.5876,
+      "step": 1038
+    },
+    {
+      "epoch": 0.3479571332886805,
+      "grad_norm": 0.4553813636302948,
+      "learning_rate": 0.0001,
+      "loss": 1.5238,
+      "step": 1039
+    },
+    {
+      "epoch": 0.348292029470864,
+      "grad_norm": 0.4430896043777466,
+      "learning_rate": 0.0001,
+      "loss": 1.5503,
+      "step": 1040
+    },
+    {
+      "epoch": 0.34862692565304754,
+      "grad_norm": 0.4415845274925232,
+      "learning_rate": 0.0001,
+      "loss": 1.5227,
+      "step": 1041
+    },
+    {
+      "epoch": 0.34896182183523106,
+      "grad_norm": 0.4520629942417145,
+      "learning_rate": 0.0001,
+      "loss": 1.6111,
+      "step": 1042
+    },
+    {
+      "epoch": 0.3492967180174146,
+      "grad_norm": 0.4249700605869293,
+      "learning_rate": 0.0001,
+      "loss": 1.496,
+      "step": 1043
+    },
+    {
+      "epoch": 0.34963161419959815,
+      "grad_norm": 0.4444321095943451,
+      "learning_rate": 0.0001,
+      "loss": 1.5424,
+      "step": 1044
+    },
+    {
+      "epoch": 0.34996651038178167,
+      "grad_norm": 0.41880741715431213,
+      "learning_rate": 0.0001,
+      "loss": 1.4905,
+      "step": 1045
+    },
+    {
+      "epoch": 0.3503014065639652,
+      "grad_norm": 0.48384132981300354,
+      "learning_rate": 0.0001,
+      "loss": 1.6709,
+      "step": 1046
+    },
+    {
+      "epoch": 0.3506363027461487,
+      "grad_norm": 0.4757600724697113,
+      "learning_rate": 0.0001,
+      "loss": 1.5308,
+      "step": 1047
+    },
+    {
+      "epoch": 0.35097119892833223,
+      "grad_norm": 0.4434678554534912,
+      "learning_rate": 0.0001,
+      "loss": 1.6136,
+      "step": 1048
+    },
+    {
+      "epoch": 0.35130609511051575,
+      "grad_norm": 0.43935686349868774,
+      "learning_rate": 0.0001,
+      "loss": 1.5569,
+      "step": 1049
+    },
+    {
+      "epoch": 0.35164099129269927,
+      "grad_norm": 0.43453824520111084,
+      "learning_rate": 0.0001,
+      "loss": 1.5371,
+      "step": 1050
+    },
+    {
+      "epoch": 0.3519758874748828,
+      "grad_norm": 0.43408510088920593,
+      "learning_rate": 0.0001,
+      "loss": 1.4975,
+      "step": 1051
+    },
+    {
+      "epoch": 0.3523107836570663,
+      "grad_norm": 0.437828004360199,
+      "learning_rate": 0.0001,
+      "loss": 1.4563,
+      "step": 1052
+    },
+    {
+      "epoch": 0.3526456798392498,
+      "grad_norm": 0.43769609928131104,
+      "learning_rate": 0.0001,
+      "loss": 1.5476,
+      "step": 1053
+    },
+    {
+      "epoch": 0.35298057602143335,
+      "grad_norm": 0.4404858350753784,
+      "learning_rate": 0.0001,
+      "loss": 1.5501,
+      "step": 1054
+    },
+    {
+      "epoch": 0.35331547220361686,
+      "grad_norm": 0.44895392656326294,
+      "learning_rate": 0.0001,
+      "loss": 1.543,
+      "step": 1055
+    },
+    {
+      "epoch": 0.3536503683858004,
+      "grad_norm": 0.4335348904132843,
+      "learning_rate": 0.0001,
+      "loss": 1.4981,
+      "step": 1056
+    },
+    {
+      "epoch": 0.3539852645679839,
+      "grad_norm": 0.43414801359176636,
+      "learning_rate": 0.0001,
+      "loss": 1.4583,
+      "step": 1057
+    },
+    {
+      "epoch": 0.3543201607501674,
+      "grad_norm": 0.44667840003967285,
+      "learning_rate": 0.0001,
+      "loss": 1.5666,
+      "step": 1058
+    },
+    {
+      "epoch": 0.354655056932351,
+      "grad_norm": 0.43654167652130127,
+      "learning_rate": 0.0001,
+      "loss": 1.4808,
+      "step": 1059
+    },
+    {
+      "epoch": 0.3549899531145345,
+      "grad_norm": 0.4739477336406708,
+      "learning_rate": 0.0001,
+      "loss": 1.5895,
+      "step": 1060
+    },
+    {
+      "epoch": 0.35532484929671804,
+      "grad_norm": 0.46855276823043823,
+      "learning_rate": 0.0001,
+      "loss": 1.6228,
+      "step": 1061
+    },
+    {
+      "epoch": 0.35565974547890156,
+      "grad_norm": 0.4664900004863739,
+      "learning_rate": 0.0001,
+      "loss": 1.523,
+      "step": 1062
+    },
+    {
+      "epoch": 0.3559946416610851,
+      "grad_norm": 0.4334528148174286,
+      "learning_rate": 0.0001,
+      "loss": 1.5027,
+      "step": 1063
+    },
+    {
+      "epoch": 0.3563295378432686,
+      "grad_norm": 0.4692411422729492,
+      "learning_rate": 0.0001,
+      "loss": 1.5308,
+      "step": 1064
+    },
+    {
+      "epoch": 0.3566644340254521,
+      "grad_norm": 0.4434930086135864,
+      "learning_rate": 0.0001,
+      "loss": 1.5424,
+      "step": 1065
+    },
+    {
+      "epoch": 0.35699933020763563,
+      "grad_norm": 0.43848127126693726,
+      "learning_rate": 0.0001,
+      "loss": 1.523,
+      "step": 1066
+    },
+    {
+      "epoch": 0.35733422638981915,
+      "grad_norm": 0.4805077612400055,
+      "learning_rate": 0.0001,
+      "loss": 1.5774,
+      "step": 1067
+    },
+    {
+      "epoch": 0.3576691225720027,
+      "grad_norm": 0.46720483899116516,
+      "learning_rate": 0.0001,
+      "loss": 1.5654,
+      "step": 1068
+    },
+    {
+      "epoch": 0.3580040187541862,
+      "grad_norm": 0.47015318274497986,
+      "learning_rate": 0.0001,
+      "loss": 1.6036,
+      "step": 1069
+    },
+    {
+      "epoch": 0.3583389149363697,
+      "grad_norm": 0.46188637614250183,
+      "learning_rate": 0.0001,
+      "loss": 1.5884,
+      "step": 1070
+    },
+    {
+      "epoch": 0.35867381111855323,
+      "grad_norm": 0.46755450963974,
+      "learning_rate": 0.0001,
+      "loss": 1.6366,
+      "step": 1071
+    },
+    {
+      "epoch": 0.35900870730073675,
+      "grad_norm": 0.44284823536872864,
+      "learning_rate": 0.0001,
+      "loss": 1.5063,
+      "step": 1072
+    },
+    {
+      "epoch": 0.35934360348292027,
+      "grad_norm": 0.452816903591156,
+      "learning_rate": 0.0001,
+      "loss": 1.5715,
+      "step": 1073
+    },
+    {
+      "epoch": 0.3596784996651038,
+      "grad_norm": 0.44831642508506775,
+      "learning_rate": 0.0001,
+      "loss": 1.5484,
+      "step": 1074
+    },
+    {
+      "epoch": 0.36001339584728737,
+      "grad_norm": 0.45156675577163696,
+      "learning_rate": 0.0001,
+      "loss": 1.5377,
+      "step": 1075
+    },
+    {
+      "epoch": 0.3603482920294709,
+      "grad_norm": 0.46498456597328186,
+      "learning_rate": 0.0001,
+      "loss": 1.5168,
+      "step": 1076
+    },
+    {
+      "epoch": 0.3606831882116544,
+      "grad_norm": 0.4552898705005646,
+      "learning_rate": 0.0001,
+      "loss": 1.5627,
+      "step": 1077
+    },
+    {
+      "epoch": 0.3610180843938379,
+      "grad_norm": 0.44832298159599304,
+      "learning_rate": 0.0001,
+      "loss": 1.5315,
+      "step": 1078
+    },
+    {
+      "epoch": 0.36135298057602144,
+      "grad_norm": 0.4634014070034027,
+      "learning_rate": 0.0001,
+      "loss": 1.6449,
+      "step": 1079
+    },
+    {
+      "epoch": 0.36168787675820496,
+      "grad_norm": 0.4542264938354492,
+      "learning_rate": 0.0001,
+      "loss": 1.5183,
+      "step": 1080
+    },
+    {
+      "epoch": 0.3620227729403885,
+      "grad_norm": 0.46849897503852844,
+      "learning_rate": 0.0001,
+      "loss": 1.4831,
+      "step": 1081
+    },
+    {
+      "epoch": 0.362357669122572,
+      "grad_norm": 0.43632426857948303,
+      "learning_rate": 0.0001,
+      "loss": 1.4944,
+      "step": 1082
+    },
+    {
+      "epoch": 0.3626925653047555,
+      "grad_norm": 0.4542003273963928,
+      "learning_rate": 0.0001,
+      "loss": 1.5185,
+      "step": 1083
+    },
+    {
+      "epoch": 0.36302746148693904,
+      "grad_norm": 0.45889243483543396,
+      "learning_rate": 0.0001,
+      "loss": 1.6152,
+      "step": 1084
+    },
+    {
+      "epoch": 0.36336235766912256,
+      "grad_norm": 0.4599902629852295,
+      "learning_rate": 0.0001,
+      "loss": 1.5389,
+      "step": 1085
+    },
+    {
+      "epoch": 0.3636972538513061,
+      "grad_norm": 0.4660051763057709,
+      "learning_rate": 0.0001,
+      "loss": 1.5831,
+      "step": 1086
+    },
+    {
+      "epoch": 0.3640321500334896,
+      "grad_norm": 0.4395544230937958,
+      "learning_rate": 0.0001,
+      "loss": 1.5524,
+      "step": 1087
+    },
+    {
+      "epoch": 0.3643670462156731,
+      "grad_norm": 0.4700053930282593,
+      "learning_rate": 0.0001,
+      "loss": 1.5226,
+      "step": 1088
+    },
+    {
+      "epoch": 0.36470194239785664,
+      "grad_norm": 0.4484608471393585,
+      "learning_rate": 0.0001,
+      "loss": 1.5259,
+      "step": 1089
+    },
+    {
+      "epoch": 0.3650368385800402,
+      "grad_norm": 0.4695565104484558,
+      "learning_rate": 0.0001,
+      "loss": 1.5009,
+      "step": 1090
+    },
+    {
+      "epoch": 0.36537173476222373,
+      "grad_norm": 0.447613388299942,
+      "learning_rate": 0.0001,
+      "loss": 1.5532,
+      "step": 1091
+    },
+    {
+      "epoch": 0.36570663094440725,
+      "grad_norm": 0.48899951577186584,
+      "learning_rate": 0.0001,
+      "loss": 1.6002,
+      "step": 1092
+    },
+    {
+      "epoch": 0.36604152712659077,
+      "grad_norm": 0.45602288842201233,
+      "learning_rate": 0.0001,
+      "loss": 1.6501,
+      "step": 1093
+    },
+    {
+      "epoch": 0.3663764233087743,
+      "grad_norm": 0.449912965297699,
+      "learning_rate": 0.0001,
+      "loss": 1.5524,
+      "step": 1094
+    },
+    {
+      "epoch": 0.3667113194909578,
+      "grad_norm": 0.4617339074611664,
+      "learning_rate": 0.0001,
+      "loss": 1.5734,
+      "step": 1095
+    },
+    {
+      "epoch": 0.36704621567314133,
+      "grad_norm": 0.4346519708633423,
+      "learning_rate": 0.0001,
+      "loss": 1.4779,
+      "step": 1096
+    },
+    {
+      "epoch": 0.36738111185532485,
+      "grad_norm": 0.4436924457550049,
+      "learning_rate": 0.0001,
+      "loss": 1.5651,
+      "step": 1097
+    },
+    {
+      "epoch": 0.36771600803750837,
+      "grad_norm": 0.4515039324760437,
+      "learning_rate": 0.0001,
+      "loss": 1.5545,
+      "step": 1098
+    },
+    {
+      "epoch": 0.3680509042196919,
+      "grad_norm": 0.44696515798568726,
+      "learning_rate": 0.0001,
+      "loss": 1.542,
+      "step": 1099
+    },
+    {
+      "epoch": 0.3683858004018754,
+      "grad_norm": 0.4515751600265503,
+      "learning_rate": 0.0001,
+      "loss": 1.5397,
+      "step": 1100
+    },
+    {
+      "epoch": 0.3687206965840589,
+      "grad_norm": 0.4478546977043152,
+      "learning_rate": 0.0001,
+      "loss": 1.5181,
+      "step": 1101
+    },
+    {
+      "epoch": 0.36905559276624245,
+      "grad_norm": 0.46708935499191284,
+      "learning_rate": 0.0001,
+      "loss": 1.4895,
+      "step": 1102
+    },
+    {
+      "epoch": 0.36939048894842597,
+      "grad_norm": 0.43863463401794434,
+      "learning_rate": 0.0001,
+      "loss": 1.5248,
+      "step": 1103
+    },
+    {
+      "epoch": 0.3697253851306095,
+      "grad_norm": 0.4555443227291107,
+      "learning_rate": 0.0001,
+      "loss": 1.5217,
+      "step": 1104
+    },
+    {
+      "epoch": 0.37006028131279306,
+      "grad_norm": 0.4301038086414337,
+      "learning_rate": 0.0001,
+      "loss": 1.4424,
+      "step": 1105
+    },
+    {
+      "epoch": 0.3703951774949766,
+      "grad_norm": 0.44846969842910767,
+      "learning_rate": 0.0001,
+      "loss": 1.523,
+      "step": 1106
+    },
+    {
+      "epoch": 0.3707300736771601,
+      "grad_norm": 0.45033925771713257,
+      "learning_rate": 0.0001,
+      "loss": 1.5273,
+      "step": 1107
+    },
+    {
+      "epoch": 0.3710649698593436,
+      "grad_norm": 0.4505354166030884,
+      "learning_rate": 0.0001,
+      "loss": 1.6396,
+      "step": 1108
+    },
+    {
+      "epoch": 0.37139986604152714,
+      "grad_norm": 0.4425196647644043,
+      "learning_rate": 0.0001,
+      "loss": 1.585,
+      "step": 1109
+    },
+    {
+      "epoch": 0.37173476222371066,
+      "grad_norm": 0.45227015018463135,
+      "learning_rate": 0.0001,
+      "loss": 1.6102,
+      "step": 1110
+    },
+    {
+      "epoch": 0.3720696584058942,
+      "grad_norm": 0.4454893469810486,
+      "learning_rate": 0.0001,
+      "loss": 1.6097,
+      "step": 1111
+    },
+    {
+      "epoch": 0.3724045545880777,
+      "grad_norm": 0.43742790818214417,
+      "learning_rate": 0.0001,
+      "loss": 1.5027,
+      "step": 1112
+    },
+    {
+      "epoch": 0.3727394507702612,
+      "grad_norm": 0.4249231517314911,
+      "learning_rate": 0.0001,
+      "loss": 1.5145,
+      "step": 1113
+    },
+    {
+      "epoch": 0.37307434695244474,
+      "grad_norm": 0.4511743485927582,
+      "learning_rate": 0.0001,
+      "loss": 1.625,
+      "step": 1114
+    },
+    {
+      "epoch": 0.37340924313462825,
+      "grad_norm": 0.4412231147289276,
+      "learning_rate": 0.0001,
+      "loss": 1.5336,
+      "step": 1115
+    },
+    {
+      "epoch": 0.3737441393168118,
+      "grad_norm": 0.4392704963684082,
+      "learning_rate": 0.0001,
+      "loss": 1.4292,
+      "step": 1116
+    },
+    {
+      "epoch": 0.3740790354989953,
+      "grad_norm": 0.43994712829589844,
+      "learning_rate": 0.0001,
+      "loss": 1.4992,
+      "step": 1117
+    },
+    {
+      "epoch": 0.3744139316811788,
+      "grad_norm": 0.44654780626296997,
+      "learning_rate": 0.0001,
+      "loss": 1.5576,
+      "step": 1118
+    },
+    {
+      "epoch": 0.37474882786336233,
+      "grad_norm": 0.4374602138996124,
+      "learning_rate": 0.0001,
+      "loss": 1.473,
+      "step": 1119
+    },
+    {
+      "epoch": 0.3750837240455459,
+      "grad_norm": 0.43292462825775146,
+      "learning_rate": 0.0001,
+      "loss": 1.5379,
+      "step": 1120
+    },
+    {
+      "epoch": 0.3754186202277294,
+      "grad_norm": 0.46613624691963196,
+      "learning_rate": 0.0001,
+      "loss": 1.5969,
+      "step": 1121
+    },
+    {
+      "epoch": 0.37575351640991295,
+      "grad_norm": 0.4482156038284302,
+      "learning_rate": 0.0001,
+      "loss": 1.6308,
+      "step": 1122
+    },
+    {
+      "epoch": 0.37608841259209647,
+      "grad_norm": 0.4478096067905426,
+      "learning_rate": 0.0001,
+      "loss": 1.5234,
+      "step": 1123
+    },
+    {
+      "epoch": 0.37642330877428,
+      "grad_norm": 0.4612812399864197,
+      "learning_rate": 0.0001,
+      "loss": 1.6014,
+      "step": 1124
+    },
+    {
+      "epoch": 0.3767582049564635,
+      "grad_norm": 0.45624470710754395,
+      "learning_rate": 0.0001,
+      "loss": 1.6727,
+      "step": 1125
+    },
+    {
+      "epoch": 0.377093101138647,
+      "grad_norm": 0.45116880536079407,
+      "learning_rate": 0.0001,
+      "loss": 1.5213,
+      "step": 1126
+    },
+    {
+      "epoch": 0.37742799732083054,
+      "grad_norm": 0.45031848549842834,
+      "learning_rate": 0.0001,
+      "loss": 1.5688,
+      "step": 1127
+    },
+    {
+      "epoch": 0.37776289350301406,
+      "grad_norm": 0.4345763921737671,
+      "learning_rate": 0.0001,
+      "loss": 1.5191,
+      "step": 1128
+    },
+    {
+      "epoch": 0.3780977896851976,
+      "grad_norm": 0.4490206837654114,
+      "learning_rate": 0.0001,
+      "loss": 1.5108,
+      "step": 1129
+    },
+    {
+      "epoch": 0.3784326858673811,
+      "grad_norm": 0.46614694595336914,
+      "learning_rate": 0.0001,
+      "loss": 1.5587,
+      "step": 1130
+    },
+    {
+      "epoch": 0.3787675820495646,
+      "grad_norm": 0.4247433543205261,
+      "learning_rate": 0.0001,
+      "loss": 1.4486,
+      "step": 1131
+    },
+    {
+      "epoch": 0.37910247823174814,
+      "grad_norm": 0.45231062173843384,
+      "learning_rate": 0.0001,
+      "loss": 1.5752,
+      "step": 1132
+    },
+    {
+      "epoch": 0.37943737441393166,
+      "grad_norm": 0.4566837251186371,
+      "learning_rate": 0.0001,
+      "loss": 1.6133,
+      "step": 1133
+    },
+    {
+      "epoch": 0.3797722705961152,
+      "grad_norm": 0.457768052816391,
+      "learning_rate": 0.0001,
+      "loss": 1.5689,
+      "step": 1134
+    },
+    {
+      "epoch": 0.38010716677829876,
+      "grad_norm": 0.4402942955493927,
+      "learning_rate": 0.0001,
+      "loss": 1.6279,
+      "step": 1135
+    },
+    {
+      "epoch": 0.3804420629604823,
+      "grad_norm": 0.4514990448951721,
+      "learning_rate": 0.0001,
+      "loss": 1.5486,
+      "step": 1136
+    },
+    {
+      "epoch": 0.3807769591426658,
+      "grad_norm": 0.45938149094581604,
+      "learning_rate": 0.0001,
+      "loss": 1.4448,
+      "step": 1137
+    },
+    {
+      "epoch": 0.3811118553248493,
+      "grad_norm": 0.4579198360443115,
+      "learning_rate": 0.0001,
+      "loss": 1.5811,
+      "step": 1138
+    },
+    {
+      "epoch": 0.38144675150703283,
+      "grad_norm": 0.43243205547332764,
+      "learning_rate": 0.0001,
+      "loss": 1.5189,
+      "step": 1139
+    },
+    {
+      "epoch": 0.38178164768921635,
+      "grad_norm": 0.45426496863365173,
+      "learning_rate": 0.0001,
+      "loss": 1.5627,
+      "step": 1140
+    },
+    {
+      "epoch": 0.38211654387139987,
+      "grad_norm": 0.4345931112766266,
+      "learning_rate": 0.0001,
+      "loss": 1.4466,
+      "step": 1141
+    },
+    {
+      "epoch": 0.3824514400535834,
+      "grad_norm": 0.44259166717529297,
+      "learning_rate": 0.0001,
+      "loss": 1.5348,
+      "step": 1142
+    },
+    {
+      "epoch": 0.3827863362357669,
+      "grad_norm": 0.4565470516681671,
+      "learning_rate": 0.0001,
+      "loss": 1.5584,
+      "step": 1143
+    },
+    {
+      "epoch": 0.38312123241795043,
+      "grad_norm": 0.44665977358818054,
+      "learning_rate": 0.0001,
+      "loss": 1.589,
+      "step": 1144
+    },
+    {
+      "epoch": 0.38345612860013395,
+      "grad_norm": 0.44836845993995667,
+      "learning_rate": 0.0001,
+      "loss": 1.5347,
+      "step": 1145
+    },
+    {
+      "epoch": 0.38379102478231747,
+      "grad_norm": 0.44463881850242615,
+      "learning_rate": 0.0001,
+      "loss": 1.476,
+      "step": 1146
+    },
+    {
+      "epoch": 0.384125920964501,
+      "grad_norm": 0.457633376121521,
+      "learning_rate": 0.0001,
+      "loss": 1.559,
+      "step": 1147
+    },
+    {
+      "epoch": 0.3844608171466845,
+      "grad_norm": 0.4234335720539093,
+      "learning_rate": 0.0001,
+      "loss": 1.4449,
+      "step": 1148
+    },
+    {
+      "epoch": 0.384795713328868,
+      "grad_norm": 0.4380183815956116,
+      "learning_rate": 0.0001,
+      "loss": 1.4661,
+      "step": 1149
+    },
+    {
+      "epoch": 0.38513060951105155,
+      "grad_norm": 0.4707789421081543,
+      "learning_rate": 0.0001,
+      "loss": 1.6386,
+      "step": 1150
+    },
+    {
+      "epoch": 0.3854655056932351,
+      "grad_norm": 0.4421239197254181,
+      "learning_rate": 0.0001,
+      "loss": 1.4635,
+      "step": 1151
+    },
+    {
+      "epoch": 0.38580040187541864,
+      "grad_norm": 0.47190219163894653,
+      "learning_rate": 0.0001,
+      "loss": 1.4755,
+      "step": 1152
+    },
+    {
+      "epoch": 0.38613529805760216,
+      "grad_norm": 0.4488900899887085,
+      "learning_rate": 0.0001,
+      "loss": 1.5538,
+      "step": 1153
+    },
+    {
+      "epoch": 0.3864701942397857,
+      "grad_norm": 0.46024009585380554,
+      "learning_rate": 0.0001,
+      "loss": 1.4989,
+      "step": 1154
+    },
+    {
+      "epoch": 0.3868050904219692,
+      "grad_norm": 0.45907169580459595,
+      "learning_rate": 0.0001,
+      "loss": 1.6402,
+      "step": 1155
+    },
+    {
+      "epoch": 0.3871399866041527,
+      "grad_norm": 0.4615232050418854,
+      "learning_rate": 0.0001,
+      "loss": 1.6111,
+      "step": 1156
+    },
+    {
+      "epoch": 0.38747488278633624,
+      "grad_norm": 0.4274507164955139,
+      "learning_rate": 0.0001,
+      "loss": 1.4771,
+      "step": 1157
+    },
+    {
+      "epoch": 0.38780977896851976,
+      "grad_norm": 0.46233120560646057,
+      "learning_rate": 0.0001,
+      "loss": 1.511,
+      "step": 1158
+    },
+    {
+      "epoch": 0.3881446751507033,
+      "grad_norm": 0.46118366718292236,
+      "learning_rate": 0.0001,
+      "loss": 1.5729,
+      "step": 1159
+    },
+    {
+      "epoch": 0.3884795713328868,
+      "grad_norm": 0.4777587056159973,
+      "learning_rate": 0.0001,
+      "loss": 1.656,
+      "step": 1160
+    },
+    {
+      "epoch": 0.3888144675150703,
+      "grad_norm": 0.4458993971347809,
+      "learning_rate": 0.0001,
+      "loss": 1.5991,
+      "step": 1161
+    },
+    {
+      "epoch": 0.38914936369725384,
+      "grad_norm": 0.4479345679283142,
+      "learning_rate": 0.0001,
+      "loss": 1.5968,
+      "step": 1162
+    },
+    {
+      "epoch": 0.38948425987943736,
+      "grad_norm": 0.45050573348999023,
+      "learning_rate": 0.0001,
+      "loss": 1.5441,
+      "step": 1163
+    },
+    {
+      "epoch": 0.3898191560616209,
+      "grad_norm": 0.45146119594573975,
+      "learning_rate": 0.0001,
+      "loss": 1.5685,
+      "step": 1164
+    },
+    {
+      "epoch": 0.3901540522438044,
+      "grad_norm": 0.46487465500831604,
+      "learning_rate": 0.0001,
+      "loss": 1.6563,
+      "step": 1165
+    },
+    {
+      "epoch": 0.39048894842598797,
+      "grad_norm": 0.45339757204055786,
+      "learning_rate": 0.0001,
+      "loss": 1.5304,
+      "step": 1166
+    },
+    {
+      "epoch": 0.3908238446081715,
+      "grad_norm": 0.4561135172843933,
+      "learning_rate": 0.0001,
+      "loss": 1.5212,
+      "step": 1167
+    },
+    {
+      "epoch": 0.391158740790355,
+      "grad_norm": 0.471395343542099,
+      "learning_rate": 0.0001,
+      "loss": 1.6656,
+      "step": 1168
+    },
+    {
+      "epoch": 0.39149363697253853,
+      "grad_norm": 0.4500054121017456,
+      "learning_rate": 0.0001,
+      "loss": 1.5474,
+      "step": 1169
+    },
+    {
+      "epoch": 0.39182853315472205,
+      "grad_norm": 0.4495108425617218,
+      "learning_rate": 0.0001,
+      "loss": 1.4906,
+      "step": 1170
+    },
+    {
+      "epoch": 0.39216342933690557,
+      "grad_norm": 0.4497644007205963,
+      "learning_rate": 0.0001,
+      "loss": 1.6199,
+      "step": 1171
+    },
+    {
+      "epoch": 0.3924983255190891,
+      "grad_norm": 0.47592660784721375,
+      "learning_rate": 0.0001,
+      "loss": 1.5653,
+      "step": 1172
+    },
+    {
+      "epoch": 0.3928332217012726,
+      "grad_norm": 0.4495856761932373,
+      "learning_rate": 0.0001,
+      "loss": 1.5728,
+      "step": 1173
+    },
+    {
+      "epoch": 0.3931681178834561,
+      "grad_norm": 0.4661826193332672,
+      "learning_rate": 0.0001,
+      "loss": 1.5641,
+      "step": 1174
+    },
+    {
+      "epoch": 0.39350301406563964,
+      "grad_norm": 0.4343692660331726,
+      "learning_rate": 0.0001,
+      "loss": 1.5582,
+      "step": 1175
+    },
+    {
+      "epoch": 0.39383791024782316,
+      "grad_norm": 0.48545411229133606,
+      "learning_rate": 0.0001,
+      "loss": 1.5754,
+      "step": 1176
+    },
+    {
+      "epoch": 0.3941728064300067,
+      "grad_norm": 0.44631296396255493,
+      "learning_rate": 0.0001,
+      "loss": 1.6401,
+      "step": 1177
+    },
+    {
+      "epoch": 0.3945077026121902,
+      "grad_norm": 0.46297982335090637,
+      "learning_rate": 0.0001,
+      "loss": 1.4894,
+      "step": 1178
+    },
+    {
+      "epoch": 0.3948425987943737,
+      "grad_norm": 0.435683935880661,
+      "learning_rate": 0.0001,
+      "loss": 1.5418,
+      "step": 1179
+    },
+    {
+      "epoch": 0.39517749497655724,
+      "grad_norm": 0.47171199321746826,
+      "learning_rate": 0.0001,
+      "loss": 1.5518,
+      "step": 1180
+    },
+    {
+      "epoch": 0.3955123911587408,
+      "grad_norm": 0.44476407766342163,
+      "learning_rate": 0.0001,
+      "loss": 1.5165,
+      "step": 1181
+    },
+    {
+      "epoch": 0.39584728734092434,
+      "grad_norm": 0.45694172382354736,
+      "learning_rate": 0.0001,
+      "loss": 1.6215,
+      "step": 1182
+    },
+    {
+      "epoch": 0.39618218352310786,
+      "grad_norm": 0.4570925533771515,
+      "learning_rate": 0.0001,
+      "loss": 1.543,
+      "step": 1183
+    },
+    {
+      "epoch": 0.3965170797052914,
+      "grad_norm": 0.4366303086280823,
+      "learning_rate": 0.0001,
+      "loss": 1.521,
+      "step": 1184
+    },
+    {
+      "epoch": 0.3968519758874749,
+      "grad_norm": 0.44927310943603516,
+      "learning_rate": 0.0001,
+      "loss": 1.5209,
+      "step": 1185
+    },
+    {
+      "epoch": 0.3971868720696584,
+      "grad_norm": 0.4475070536136627,
+      "learning_rate": 0.0001,
+      "loss": 1.53,
+      "step": 1186
+    },
+    {
+      "epoch": 0.39752176825184193,
+      "grad_norm": 0.4301213026046753,
+      "learning_rate": 0.0001,
+      "loss": 1.5712,
+      "step": 1187
+    },
+    {
+      "epoch": 0.39785666443402545,
+      "grad_norm": 0.4296334385871887,
+      "learning_rate": 0.0001,
+      "loss": 1.5227,
+      "step": 1188
+    },
+    {
+      "epoch": 0.398191560616209,
+      "grad_norm": 0.4464989900588989,
+      "learning_rate": 0.0001,
+      "loss": 1.553,
+      "step": 1189
+    },
+    {
+      "epoch": 0.3985264567983925,
+      "grad_norm": 0.4370761513710022,
+      "learning_rate": 0.0001,
+      "loss": 1.5375,
+      "step": 1190
+    },
+    {
+      "epoch": 0.398861352980576,
+      "grad_norm": 0.43171000480651855,
+      "learning_rate": 0.0001,
+      "loss": 1.4762,
+      "step": 1191
+    },
+    {
+      "epoch": 0.39919624916275953,
+      "grad_norm": 0.457441508769989,
+      "learning_rate": 0.0001,
+      "loss": 1.5028,
+      "step": 1192
+    },
+    {
+      "epoch": 0.39953114534494305,
+      "grad_norm": 0.4684367775917053,
+      "learning_rate": 0.0001,
+      "loss": 1.6496,
+      "step": 1193
+    },
+    {
+      "epoch": 0.39986604152712657,
+      "grad_norm": 0.4578244686126709,
+      "learning_rate": 0.0001,
+      "loss": 1.573,
+      "step": 1194
+    },
+    {
+      "epoch": 0.4002009377093101,
+      "grad_norm": 0.46412548422813416,
+      "learning_rate": 0.0001,
+      "loss": 1.4879,
+      "step": 1195
+    },
+    {
+      "epoch": 0.40053583389149366,
+      "grad_norm": 0.4387846291065216,
+      "learning_rate": 0.0001,
+      "loss": 1.5545,
+      "step": 1196
+    },
+    {
+      "epoch": 0.4008707300736772,
+      "grad_norm": 0.44225457310676575,
+      "learning_rate": 0.0001,
+      "loss": 1.5465,
+      "step": 1197
+    },
+    {
+      "epoch": 0.4012056262558607,
+      "grad_norm": 0.44463130831718445,
+      "learning_rate": 0.0001,
+      "loss": 1.5255,
+      "step": 1198
+    },
+    {
+      "epoch": 0.4015405224380442,
+      "grad_norm": 0.4457281231880188,
+      "learning_rate": 0.0001,
+      "loss": 1.5522,
+      "step": 1199
+    },
+    {
+      "epoch": 0.40187541862022774,
+      "grad_norm": 0.4476866126060486,
+      "learning_rate": 0.0001,
+      "loss": 1.4994,
+      "step": 1200
+    },
+    {
+      "epoch": 0.40221031480241126,
+      "grad_norm": 0.44953233003616333,
+      "learning_rate": 0.0001,
+      "loss": 1.5804,
+      "step": 1201
+    },
+    {
+      "epoch": 0.4025452109845948,
+      "grad_norm": 0.46008914709091187,
+      "learning_rate": 0.0001,
+      "loss": 1.679,
+      "step": 1202
+    },
+    {
+      "epoch": 0.4028801071667783,
+      "grad_norm": 0.4389602243900299,
+      "learning_rate": 0.0001,
+      "loss": 1.4782,
+      "step": 1203
+    },
+    {
+      "epoch": 0.4032150033489618,
+      "grad_norm": 0.43227729201316833,
+      "learning_rate": 0.0001,
+      "loss": 1.469,
+      "step": 1204
+    },
+    {
+      "epoch": 0.40354989953114534,
+      "grad_norm": 0.4598828852176666,
+      "learning_rate": 0.0001,
+      "loss": 1.6371,
+      "step": 1205
+    },
+    {
+      "epoch": 0.40388479571332886,
+      "grad_norm": 0.4646576941013336,
+      "learning_rate": 0.0001,
+      "loss": 1.597,
+      "step": 1206
+    },
+    {
+      "epoch": 0.4042196918955124,
+      "grad_norm": 0.444198876619339,
+      "learning_rate": 0.0001,
+      "loss": 1.5285,
+      "step": 1207
+    },
+    {
+      "epoch": 0.4045545880776959,
+      "grad_norm": 0.468888521194458,
+      "learning_rate": 0.0001,
+      "loss": 1.5654,
+      "step": 1208
+    },
+    {
+      "epoch": 0.4048894842598794,
+      "grad_norm": 0.4509347677230835,
+      "learning_rate": 0.0001,
+      "loss": 1.5588,
+      "step": 1209
+    },
+    {
+      "epoch": 0.40522438044206294,
+      "grad_norm": 0.463221937417984,
+      "learning_rate": 0.0001,
+      "loss": 1.6426,
+      "step": 1210
+    },
+    {
+      "epoch": 0.40555927662424646,
+      "grad_norm": 0.44848695397377014,
+      "learning_rate": 0.0001,
+      "loss": 1.5946,
+      "step": 1211
+    },
+    {
+      "epoch": 0.40589417280643003,
+      "grad_norm": 0.45708489418029785,
+      "learning_rate": 0.0001,
+      "loss": 1.6386,
+      "step": 1212
+    },
+    {
+      "epoch": 0.40622906898861355,
+      "grad_norm": 0.4351506233215332,
+      "learning_rate": 0.0001,
+      "loss": 1.4849,
+      "step": 1213
+    },
+    {
+      "epoch": 0.40656396517079707,
+      "grad_norm": 0.43319088220596313,
+      "learning_rate": 0.0001,
+      "loss": 1.6073,
+      "step": 1214
+    },
+    {
+      "epoch": 0.4068988613529806,
+      "grad_norm": 0.4341161549091339,
+      "learning_rate": 0.0001,
+      "loss": 1.5351,
+      "step": 1215
+    },
+    {
+      "epoch": 0.4072337575351641,
+      "grad_norm": 0.4327983558177948,
+      "learning_rate": 0.0001,
+      "loss": 1.5076,
+      "step": 1216
+    },
+    {
+      "epoch": 0.40756865371734763,
+      "grad_norm": 0.4561294615268707,
+      "learning_rate": 0.0001,
+      "loss": 1.5733,
+      "step": 1217
+    },
+    {
+      "epoch": 0.40790354989953115,
+      "grad_norm": 0.4407605230808258,
+      "learning_rate": 0.0001,
+      "loss": 1.5913,
+      "step": 1218
+    },
+    {
+      "epoch": 0.40823844608171467,
+      "grad_norm": 0.4305534362792969,
+      "learning_rate": 0.0001,
+      "loss": 1.5868,
+      "step": 1219
+    },
+    {
+      "epoch": 0.4085733422638982,
+      "grad_norm": 0.4355212152004242,
+      "learning_rate": 0.0001,
+      "loss": 1.5317,
+      "step": 1220
+    },
+    {
+      "epoch": 0.4089082384460817,
+      "grad_norm": 0.4315885007381439,
+      "learning_rate": 0.0001,
+      "loss": 1.5013,
+      "step": 1221
+    },
+    {
+      "epoch": 0.4092431346282652,
+      "grad_norm": 0.4414379596710205,
+      "learning_rate": 0.0001,
+      "loss": 1.4476,
+      "step": 1222
+    },
+    {
+      "epoch": 0.40957803081044875,
+      "grad_norm": 0.45077255368232727,
+      "learning_rate": 0.0001,
+      "loss": 1.4619,
+      "step": 1223
+    },
+    {
+      "epoch": 0.40991292699263226,
+      "grad_norm": 0.5059629678726196,
+      "learning_rate": 0.0001,
+      "loss": 1.5868,
+      "step": 1224
+    },
+    {
+      "epoch": 0.4102478231748158,
+      "grad_norm": 0.4635983109474182,
+      "learning_rate": 0.0001,
+      "loss": 1.6556,
+      "step": 1225
+    },
+    {
+      "epoch": 0.4105827193569993,
+      "grad_norm": 0.4328989088535309,
+      "learning_rate": 0.0001,
+      "loss": 1.4919,
+      "step": 1226
+    },
+    {
+      "epoch": 0.4109176155391829,
+      "grad_norm": 0.4631066918373108,
+      "learning_rate": 0.0001,
+      "loss": 1.3765,
+      "step": 1227
+    },
+    {
+      "epoch": 0.4112525117213664,
+      "grad_norm": 0.4278043508529663,
+      "learning_rate": 0.0001,
+      "loss": 1.4469,
+      "step": 1228
+    },
+    {
+      "epoch": 0.4115874079035499,
+      "grad_norm": 0.5048599243164062,
+      "learning_rate": 0.0001,
+      "loss": 1.5859,
+      "step": 1229
+    },
+    {
+      "epoch": 0.41192230408573344,
+      "grad_norm": 0.4466727674007416,
+      "learning_rate": 0.0001,
+      "loss": 1.5656,
+      "step": 1230
+    },
+    {
+      "epoch": 0.41225720026791696,
+      "grad_norm": 0.48620685935020447,
+      "learning_rate": 0.0001,
+      "loss": 1.5889,
+      "step": 1231
+    },
+    {
+      "epoch": 0.4125920964501005,
+      "grad_norm": 0.47321709990501404,
+      "learning_rate": 0.0001,
+      "loss": 1.6017,
+      "step": 1232
+    },
+    {
+      "epoch": 0.412926992632284,
+      "grad_norm": 0.48233163356781006,
+      "learning_rate": 0.0001,
+      "loss": 1.5773,
+      "step": 1233
+    },
+    {
+      "epoch": 0.4132618888144675,
+      "grad_norm": 0.4501434564590454,
+      "learning_rate": 0.0001,
+      "loss": 1.5323,
+      "step": 1234
+    },
+    {
+      "epoch": 0.41359678499665103,
+      "grad_norm": 0.45282772183418274,
+      "learning_rate": 0.0001,
+      "loss": 1.5687,
+      "step": 1235
+    },
+    {
+      "epoch": 0.41393168117883455,
+      "grad_norm": 0.4586126506328583,
+      "learning_rate": 0.0001,
+      "loss": 1.5614,
+      "step": 1236
+    },
+    {
+      "epoch": 0.4142665773610181,
+      "grad_norm": 0.4504382312297821,
+      "learning_rate": 0.0001,
+      "loss": 1.5353,
+      "step": 1237
+    },
+    {
+      "epoch": 0.4146014735432016,
+      "grad_norm": 0.46755629777908325,
+      "learning_rate": 0.0001,
+      "loss": 1.5325,
+      "step": 1238
+    },
+    {
+      "epoch": 0.4149363697253851,
+      "grad_norm": 0.45872950553894043,
+      "learning_rate": 0.0001,
+      "loss": 1.5826,
+      "step": 1239
+    },
+    {
+      "epoch": 0.41527126590756863,
+      "grad_norm": 0.47714489698410034,
+      "learning_rate": 0.0001,
+      "loss": 1.574,
+      "step": 1240
+    },
+    {
+      "epoch": 0.41560616208975215,
+      "grad_norm": 0.449758380651474,
+      "learning_rate": 0.0001,
+      "loss": 1.5127,
+      "step": 1241
+    },
+    {
+      "epoch": 0.4159410582719357,
+      "grad_norm": 0.4618908166885376,
+      "learning_rate": 0.0001,
+      "loss": 1.7078,
+      "step": 1242
+    },
+    {
+      "epoch": 0.41627595445411925,
+      "grad_norm": 0.43264150619506836,
+      "learning_rate": 0.0001,
+      "loss": 1.5085,
+      "step": 1243
+    },
+    {
+      "epoch": 0.41661085063630277,
+      "grad_norm": 0.4599592387676239,
+      "learning_rate": 0.0001,
+      "loss": 1.553,
+      "step": 1244
+    },
+    {
+      "epoch": 0.4169457468184863,
+      "grad_norm": 0.43902966380119324,
+      "learning_rate": 0.0001,
+      "loss": 1.5751,
+      "step": 1245
+    },
+    {
+      "epoch": 0.4172806430006698,
+      "grad_norm": 0.4508312940597534,
+      "learning_rate": 0.0001,
+      "loss": 1.5752,
+      "step": 1246
+    },
+    {
+      "epoch": 0.4176155391828533,
+      "grad_norm": 0.4534173309803009,
+      "learning_rate": 0.0001,
+      "loss": 1.5037,
+      "step": 1247
+    },
+    {
+      "epoch": 0.41795043536503684,
+      "grad_norm": 0.46478843688964844,
+      "learning_rate": 0.0001,
+      "loss": 1.6001,
+      "step": 1248
+    },
+    {
+      "epoch": 0.41828533154722036,
+      "grad_norm": 0.47987523674964905,
+      "learning_rate": 0.0001,
+      "loss": 1.564,
+      "step": 1249
+    },
+    {
+      "epoch": 0.4186202277294039,
+      "grad_norm": 0.46117353439331055,
+      "learning_rate": 0.0001,
+      "loss": 1.5695,
+      "step": 1250
+    },
+    {
+      "epoch": 0.4189551239115874,
+      "grad_norm": 0.46787649393081665,
+      "learning_rate": 0.0001,
+      "loss": 1.4925,
+      "step": 1251
+    },
+    {
+      "epoch": 0.4192900200937709,
+      "grad_norm": 0.4669802784919739,
+      "learning_rate": 0.0001,
+      "loss": 1.5902,
+      "step": 1252
+    },
+    {
+      "epoch": 0.41962491627595444,
+      "grad_norm": 0.4593794047832489,
+      "learning_rate": 0.0001,
+      "loss": 1.5026,
+      "step": 1253
+    },
+    {
+      "epoch": 0.41995981245813796,
+      "grad_norm": 0.4430108964443207,
+      "learning_rate": 0.0001,
+      "loss": 1.6007,
+      "step": 1254
+    },
+    {
+      "epoch": 0.4202947086403215,
+      "grad_norm": 0.48188650608062744,
+      "learning_rate": 0.0001,
+      "loss": 1.6276,
+      "step": 1255
+    },
+    {
+      "epoch": 0.420629604822505,
+      "grad_norm": 0.44512104988098145,
+      "learning_rate": 0.0001,
+      "loss": 1.476,
+      "step": 1256
+    },
+    {
+      "epoch": 0.4209645010046886,
+      "grad_norm": 0.46867334842681885,
+      "learning_rate": 0.0001,
+      "loss": 1.5168,
+      "step": 1257
+    },
+    {
+      "epoch": 0.4212993971868721,
+      "grad_norm": 0.4648856222629547,
+      "learning_rate": 0.0001,
+      "loss": 1.6594,
+      "step": 1258
+    },
+    {
+      "epoch": 0.4216342933690556,
+      "grad_norm": 0.47854331135749817,
+      "learning_rate": 0.0001,
+      "loss": 1.567,
+      "step": 1259
+    },
+    {
+      "epoch": 0.42196918955123913,
+      "grad_norm": 0.48470669984817505,
+      "learning_rate": 0.0001,
+      "loss": 1.6363,
+      "step": 1260
+    },
+    {
+      "epoch": 0.42230408573342265,
+      "grad_norm": 0.44366371631622314,
+      "learning_rate": 0.0001,
+      "loss": 1.5321,
+      "step": 1261
+    },
+    {
+      "epoch": 0.42263898191560617,
+      "grad_norm": 0.43455782532691956,
+      "learning_rate": 0.0001,
+      "loss": 1.4234,
+      "step": 1262
+    },
+    {
+      "epoch": 0.4229738780977897,
+      "grad_norm": 0.44860178232192993,
+      "learning_rate": 0.0001,
+      "loss": 1.4949,
+      "step": 1263
+    },
+    {
+      "epoch": 0.4233087742799732,
+      "grad_norm": 0.47595492005348206,
+      "learning_rate": 0.0001,
+      "loss": 1.6015,
+      "step": 1264
+    },
+    {
+      "epoch": 0.42364367046215673,
+      "grad_norm": 0.4563322961330414,
+      "learning_rate": 0.0001,
+      "loss": 1.5518,
+      "step": 1265
+    },
+    {
+      "epoch": 0.42397856664434025,
+      "grad_norm": 0.45181918144226074,
+      "learning_rate": 0.0001,
+      "loss": 1.6027,
+      "step": 1266
+    },
+    {
+      "epoch": 0.42431346282652377,
+      "grad_norm": 0.43826577067375183,
+      "learning_rate": 0.0001,
+      "loss": 1.5372,
+      "step": 1267
+    },
+    {
+      "epoch": 0.4246483590087073,
+      "grad_norm": 0.46303653717041016,
+      "learning_rate": 0.0001,
+      "loss": 1.6133,
+      "step": 1268
+    },
+    {
+      "epoch": 0.4249832551908908,
+      "grad_norm": 0.4417208731174469,
+      "learning_rate": 0.0001,
+      "loss": 1.6145,
+      "step": 1269
+    },
+    {
+      "epoch": 0.4253181513730743,
+      "grad_norm": 0.4449644684791565,
+      "learning_rate": 0.0001,
+      "loss": 1.531,
+      "step": 1270
+    },
+    {
+      "epoch": 0.42565304755525785,
+      "grad_norm": 0.4542076289653778,
+      "learning_rate": 0.0001,
+      "loss": 1.5358,
+      "step": 1271
+    },
+    {
+      "epoch": 0.42598794373744137,
+      "grad_norm": 0.4694463610649109,
+      "learning_rate": 0.0001,
+      "loss": 1.6805,
+      "step": 1272
+    },
+    {
+      "epoch": 0.42632283991962494,
+      "grad_norm": 0.4595602750778198,
+      "learning_rate": 0.0001,
+      "loss": 1.5962,
+      "step": 1273
+    },
+    {
+      "epoch": 0.42665773610180846,
+      "grad_norm": 0.4681914746761322,
+      "learning_rate": 0.0001,
+      "loss": 1.591,
+      "step": 1274
+    },
+    {
+      "epoch": 0.426992632283992,
+      "grad_norm": 0.4733419120311737,
+      "learning_rate": 0.0001,
+      "loss": 1.4529,
+      "step": 1275
+    },
+    {
+      "epoch": 0.4273275284661755,
+      "grad_norm": 0.43556496500968933,
+      "learning_rate": 0.0001,
+      "loss": 1.5598,
+      "step": 1276
+    },
+    {
+      "epoch": 0.427662424648359,
+      "grad_norm": 0.4413219094276428,
+      "learning_rate": 0.0001,
+      "loss": 1.4457,
+      "step": 1277
+    },
+    {
+      "epoch": 0.42799732083054254,
+      "grad_norm": 0.4566291272640228,
+      "learning_rate": 0.0001,
+      "loss": 1.5364,
+      "step": 1278
+    },
+    {
+      "epoch": 0.42833221701272606,
+      "grad_norm": 0.43846362829208374,
+      "learning_rate": 0.0001,
+      "loss": 1.5502,
+      "step": 1279
+    },
+    {
+      "epoch": 0.4286671131949096,
+      "grad_norm": 0.44239136576652527,
+      "learning_rate": 0.0001,
+      "loss": 1.5821,
+      "step": 1280
+    },
+    {
+      "epoch": 0.4290020093770931,
+      "grad_norm": 0.41309061646461487,
+      "learning_rate": 0.0001,
+      "loss": 1.3608,
+      "step": 1281
+    },
+    {
+      "epoch": 0.4293369055592766,
+      "grad_norm": 0.4513871371746063,
+      "learning_rate": 0.0001,
+      "loss": 1.6097,
+      "step": 1282
+    },
+    {
+      "epoch": 0.42967180174146014,
+      "grad_norm": 0.44489651918411255,
+      "learning_rate": 0.0001,
+      "loss": 1.5062,
+      "step": 1283
+    },
+    {
+      "epoch": 0.43000669792364365,
+      "grad_norm": 0.4357435405254364,
+      "learning_rate": 0.0001,
+      "loss": 1.4767,
+      "step": 1284
+    },
+    {
+      "epoch": 0.4303415941058272,
+      "grad_norm": 0.44091832637786865,
+      "learning_rate": 0.0001,
+      "loss": 1.5169,
+      "step": 1285
+    },
+    {
+      "epoch": 0.4306764902880107,
+      "grad_norm": 0.4526561498641968,
+      "learning_rate": 0.0001,
+      "loss": 1.5086,
+      "step": 1286
+    },
+    {
+      "epoch": 0.4310113864701942,
+      "grad_norm": 0.43664780259132385,
+      "learning_rate": 0.0001,
+      "loss": 1.5234,
+      "step": 1287
+    },
+    {
+      "epoch": 0.4313462826523778,
+      "grad_norm": 0.4524122178554535,
+      "learning_rate": 0.0001,
+      "loss": 1.5574,
+      "step": 1288
+    },
+    {
+      "epoch": 0.4316811788345613,
+      "grad_norm": 0.4236125946044922,
+      "learning_rate": 0.0001,
+      "loss": 1.5119,
+      "step": 1289
+    },
+    {
+      "epoch": 0.4320160750167448,
+      "grad_norm": 0.44944438338279724,
+      "learning_rate": 0.0001,
+      "loss": 1.6364,
+      "step": 1290
+    },
+    {
+      "epoch": 0.43235097119892835,
+      "grad_norm": 0.44755467772483826,
+      "learning_rate": 0.0001,
+      "loss": 1.5394,
+      "step": 1291
+    },
+    {
+      "epoch": 0.43268586738111187,
+      "grad_norm": 0.4241560697555542,
+      "learning_rate": 0.0001,
+      "loss": 1.4207,
+      "step": 1292
+    },
+    {
+      "epoch": 0.4330207635632954,
+      "grad_norm": 0.45893594622612,
+      "learning_rate": 0.0001,
+      "loss": 1.581,
+      "step": 1293
+    },
+    {
+      "epoch": 0.4333556597454789,
+      "grad_norm": 0.4530966877937317,
+      "learning_rate": 0.0001,
+      "loss": 1.6222,
+      "step": 1294
+    },
+    {
+      "epoch": 0.4336905559276624,
+      "grad_norm": 0.4758310616016388,
+      "learning_rate": 0.0001,
+      "loss": 1.651,
+      "step": 1295
+    },
+    {
+      "epoch": 0.43402545210984594,
+      "grad_norm": 0.442427396774292,
+      "learning_rate": 0.0001,
+      "loss": 1.6638,
+      "step": 1296
+    },
+    {
+      "epoch": 0.43436034829202946,
+      "grad_norm": 0.4493705928325653,
+      "learning_rate": 0.0001,
+      "loss": 1.4638,
+      "step": 1297
+    },
+    {
+      "epoch": 0.434695244474213,
+      "grad_norm": 0.44565629959106445,
+      "learning_rate": 0.0001,
+      "loss": 1.5027,
+      "step": 1298
+    },
+    {
+      "epoch": 0.4350301406563965,
+      "grad_norm": 0.44541066884994507,
+      "learning_rate": 0.0001,
+      "loss": 1.6266,
+      "step": 1299
+    },
+    {
+      "epoch": 0.43536503683858,
+      "grad_norm": 0.4457719624042511,
+      "learning_rate": 0.0001,
+      "loss": 1.5075,
+      "step": 1300
+    },
+    {
+      "epoch": 0.43569993302076354,
+      "grad_norm": 0.4656354486942291,
+      "learning_rate": 0.0001,
+      "loss": 1.6376,
+      "step": 1301
+    },
+    {
+      "epoch": 0.43603482920294706,
+      "grad_norm": 0.44514432549476624,
+      "learning_rate": 0.0001,
+      "loss": 1.5267,
+      "step": 1302
+    },
+    {
+      "epoch": 0.43636972538513064,
+      "grad_norm": 0.43724554777145386,
+      "learning_rate": 0.0001,
+      "loss": 1.5427,
+      "step": 1303
+    },
+    {
+      "epoch": 0.43670462156731416,
+      "grad_norm": 0.44671183824539185,
+      "learning_rate": 0.0001,
+      "loss": 1.5544,
+      "step": 1304
+    },
+    {
+      "epoch": 0.4370395177494977,
+      "grad_norm": 0.4274207353591919,
+      "learning_rate": 0.0001,
+      "loss": 1.4558,
+      "step": 1305
+    },
+    {
+      "epoch": 0.4373744139316812,
+      "grad_norm": 0.45309945940971375,
+      "learning_rate": 0.0001,
+      "loss": 1.5629,
+      "step": 1306
+    },
+    {
+      "epoch": 0.4377093101138647,
+      "grad_norm": 0.5383068323135376,
+      "learning_rate": 0.0001,
+      "loss": 1.5105,
+      "step": 1307
+    },
+    {
+      "epoch": 0.43804420629604823,
+      "grad_norm": 0.4550372064113617,
+      "learning_rate": 0.0001,
+      "loss": 1.587,
+      "step": 1308
+    },
+    {
+      "epoch": 0.43837910247823175,
+      "grad_norm": 0.4549542963504791,
+      "learning_rate": 0.0001,
+      "loss": 1.5023,
+      "step": 1309
+    },
+    {
+      "epoch": 0.43871399866041527,
+      "grad_norm": 0.49144187569618225,
+      "learning_rate": 0.0001,
+      "loss": 1.5724,
+      "step": 1310
+    },
+    {
+      "epoch": 0.4390488948425988,
+      "grad_norm": 0.4485637843608856,
+      "learning_rate": 0.0001,
+      "loss": 1.59,
+      "step": 1311
+    },
+    {
+      "epoch": 0.4393837910247823,
+      "grad_norm": 0.47622618079185486,
+      "learning_rate": 0.0001,
+      "loss": 1.6332,
+      "step": 1312
+    },
+    {
+      "epoch": 0.43971868720696583,
+      "grad_norm": 0.4740828573703766,
+      "learning_rate": 0.0001,
+      "loss": 1.5848,
+      "step": 1313
+    },
+    {
+      "epoch": 0.44005358338914935,
+      "grad_norm": 0.4865962862968445,
+      "learning_rate": 0.0001,
+      "loss": 1.6034,
+      "step": 1314
+    },
+    {
+      "epoch": 0.44038847957133287,
+      "grad_norm": 0.4601993262767792,
+      "learning_rate": 0.0001,
+      "loss": 1.5108,
+      "step": 1315
+    },
+    {
+      "epoch": 0.4407233757535164,
+      "grad_norm": 0.4529930055141449,
+      "learning_rate": 0.0001,
+      "loss": 1.5559,
+      "step": 1316
+    },
+    {
+      "epoch": 0.4410582719356999,
+      "grad_norm": 0.45661240816116333,
+      "learning_rate": 0.0001,
+      "loss": 1.5116,
+      "step": 1317
+    },
+    {
+      "epoch": 0.4413931681178835,
+      "grad_norm": 0.46067968010902405,
+      "learning_rate": 0.0001,
+      "loss": 1.5718,
+      "step": 1318
+    },
+    {
+      "epoch": 0.441728064300067,
+      "grad_norm": 0.4406924545764923,
+      "learning_rate": 0.0001,
+      "loss": 1.5766,
+      "step": 1319
+    },
+    {
+      "epoch": 0.4420629604822505,
+      "grad_norm": 0.41819578409194946,
+      "learning_rate": 0.0001,
+      "loss": 1.5329,
+      "step": 1320
+    },
+    {
+      "epoch": 0.44239785666443404,
+      "grad_norm": 0.4444178342819214,
+      "learning_rate": 0.0001,
+      "loss": 1.501,
+      "step": 1321
+    },
+    {
+      "epoch": 0.44273275284661756,
+      "grad_norm": 0.47189411520957947,
+      "learning_rate": 0.0001,
+      "loss": 1.5189,
+      "step": 1322
+    },
+    {
+      "epoch": 0.4430676490288011,
+      "grad_norm": 0.44008877873420715,
+      "learning_rate": 0.0001,
+      "loss": 1.5558,
+      "step": 1323
+    },
+    {
+      "epoch": 0.4434025452109846,
+      "grad_norm": 0.44241729378700256,
+      "learning_rate": 0.0001,
+      "loss": 1.4663,
+      "step": 1324
+    },
+    {
+      "epoch": 0.4437374413931681,
+      "grad_norm": 0.4437888264656067,
+      "learning_rate": 0.0001,
+      "loss": 1.5171,
+      "step": 1325
+    },
+    {
+      "epoch": 0.44407233757535164,
+      "grad_norm": 0.4380750358104706,
+      "learning_rate": 0.0001,
+      "loss": 1.4767,
+      "step": 1326
+    },
+    {
+      "epoch": 0.44440723375753516,
+      "grad_norm": 0.44850581884384155,
+      "learning_rate": 0.0001,
+      "loss": 1.6045,
+      "step": 1327
+    },
+    {
+      "epoch": 0.4447421299397187,
+      "grad_norm": 0.47014400362968445,
+      "learning_rate": 0.0001,
+      "loss": 1.5321,
+      "step": 1328
+    },
+    {
+      "epoch": 0.4450770261219022,
+      "grad_norm": 0.4461109936237335,
+      "learning_rate": 0.0001,
+      "loss": 1.5916,
+      "step": 1329
+    },
+    {
+      "epoch": 0.4454119223040857,
+      "grad_norm": 0.4577700197696686,
+      "learning_rate": 0.0001,
+      "loss": 1.5366,
+      "step": 1330
+    },
+    {
+      "epoch": 0.44574681848626924,
+      "grad_norm": 0.47515833377838135,
+      "learning_rate": 0.0001,
+      "loss": 1.6021,
+      "step": 1331
+    },
+    {
+      "epoch": 0.44608171466845276,
+      "grad_norm": 0.4587557911872864,
+      "learning_rate": 0.0001,
+      "loss": 1.5462,
+      "step": 1332
+    },
+    {
+      "epoch": 0.4464166108506363,
+      "grad_norm": 0.45300590991973877,
+      "learning_rate": 0.0001,
+      "loss": 1.4961,
+      "step": 1333
+    },
+    {
+      "epoch": 0.44675150703281985,
+      "grad_norm": 0.42595532536506653,
+      "learning_rate": 0.0001,
+      "loss": 1.4913,
+      "step": 1334
+    },
+    {
+      "epoch": 0.44708640321500337,
+      "grad_norm": 0.4388653635978699,
+      "learning_rate": 0.0001,
+      "loss": 1.4831,
+      "step": 1335
+    },
+    {
+      "epoch": 0.4474212993971869,
+      "grad_norm": 0.4430214464664459,
+      "learning_rate": 0.0001,
+      "loss": 1.5763,
+      "step": 1336
+    },
+    {
+      "epoch": 0.4477561955793704,
+      "grad_norm": 0.4435592591762543,
+      "learning_rate": 0.0001,
+      "loss": 1.5364,
+      "step": 1337
+    },
+    {
+      "epoch": 0.44809109176155393,
+      "grad_norm": 0.4518854320049286,
+      "learning_rate": 0.0001,
+      "loss": 1.5821,
+      "step": 1338
+    },
+    {
+      "epoch": 0.44842598794373745,
+      "grad_norm": 0.4529475271701813,
+      "learning_rate": 0.0001,
+      "loss": 1.5938,
+      "step": 1339
+    },
+    {
+      "epoch": 0.44876088412592097,
+      "grad_norm": 0.4599657654762268,
+      "learning_rate": 0.0001,
+      "loss": 1.5055,
+      "step": 1340
+    },
+    {
+      "epoch": 0.4490957803081045,
+      "grad_norm": 0.46153104305267334,
+      "learning_rate": 0.0001,
+      "loss": 1.5245,
+      "step": 1341
+    },
+    {
+      "epoch": 0.449430676490288,
+      "grad_norm": 0.4551970064640045,
+      "learning_rate": 0.0001,
+      "loss": 1.5615,
+      "step": 1342
+    },
+    {
+      "epoch": 0.4497655726724715,
+      "grad_norm": 0.4437796473503113,
+      "learning_rate": 0.0001,
+      "loss": 1.5732,
+      "step": 1343
+    },
+    {
+      "epoch": 0.45010046885465504,
+      "grad_norm": 0.4721779227256775,
+      "learning_rate": 0.0001,
+      "loss": 1.63,
+      "step": 1344
+    },
+    {
+      "epoch": 0.45043536503683856,
+      "grad_norm": 0.45753568410873413,
+      "learning_rate": 0.0001,
+      "loss": 1.5609,
+      "step": 1345
+    },
+    {
+      "epoch": 0.4507702612190221,
+      "grad_norm": 0.4599592983722687,
+      "learning_rate": 0.0001,
+      "loss": 1.5068,
+      "step": 1346
+    },
+    {
+      "epoch": 0.4511051574012056,
+      "grad_norm": 0.45269471406936646,
+      "learning_rate": 0.0001,
+      "loss": 1.5497,
+      "step": 1347
+    },
+    {
+      "epoch": 0.4514400535833891,
+      "grad_norm": 0.44746312499046326,
+      "learning_rate": 0.0001,
+      "loss": 1.4903,
+      "step": 1348
+    },
+    {
+      "epoch": 0.4517749497655727,
+      "grad_norm": 0.4894854724407196,
+      "learning_rate": 0.0001,
+      "loss": 1.635,
+      "step": 1349
+    },
+    {
+      "epoch": 0.4521098459477562,
+      "grad_norm": 0.4542970359325409,
+      "learning_rate": 0.0001,
+      "loss": 1.5604,
+      "step": 1350
+    },
+    {
+      "epoch": 0.45244474212993974,
+      "grad_norm": 0.4588306248188019,
+      "learning_rate": 0.0001,
+      "loss": 1.5299,
+      "step": 1351
+    },
+    {
+      "epoch": 0.45277963831212326,
+      "grad_norm": 0.44788745045661926,
+      "learning_rate": 0.0001,
+      "loss": 1.5458,
+      "step": 1352
+    },
+    {
+      "epoch": 0.4531145344943068,
+      "grad_norm": 0.45654165744781494,
+      "learning_rate": 0.0001,
+      "loss": 1.555,
+      "step": 1353
+    },
+    {
+      "epoch": 0.4534494306764903,
+      "grad_norm": 0.44824379682540894,
+      "learning_rate": 0.0001,
+      "loss": 1.5576,
+      "step": 1354
+    },
+    {
+      "epoch": 0.4537843268586738,
+      "grad_norm": 0.446819543838501,
+      "learning_rate": 0.0001,
+      "loss": 1.4559,
+      "step": 1355
+    },
+    {
+      "epoch": 0.45411922304085733,
+      "grad_norm": 0.4417940378189087,
+      "learning_rate": 0.0001,
+      "loss": 1.3853,
+      "step": 1356
+    },
+    {
+      "epoch": 0.45445411922304085,
+      "grad_norm": 0.43769514560699463,
+      "learning_rate": 0.0001,
+      "loss": 1.5094,
+      "step": 1357
+    },
+    {
+      "epoch": 0.4547890154052244,
+      "grad_norm": 0.4573988914489746,
+      "learning_rate": 0.0001,
+      "loss": 1.5601,
+      "step": 1358
+    },
+    {
+      "epoch": 0.4551239115874079,
+      "grad_norm": 0.4574551582336426,
+      "learning_rate": 0.0001,
+      "loss": 1.6729,
+      "step": 1359
+    },
+    {
+      "epoch": 0.4554588077695914,
+      "grad_norm": 0.4588263928890228,
+      "learning_rate": 0.0001,
+      "loss": 1.5916,
+      "step": 1360
+    },
+    {
+      "epoch": 0.45579370395177493,
+      "grad_norm": 0.4549427032470703,
+      "learning_rate": 0.0001,
+      "loss": 1.5437,
+      "step": 1361
+    },
+    {
+      "epoch": 0.45612860013395845,
+      "grad_norm": 0.4670909643173218,
+      "learning_rate": 0.0001,
+      "loss": 1.5125,
+      "step": 1362
+    },
+    {
+      "epoch": 0.45646349631614197,
+      "grad_norm": 0.45247015357017517,
+      "learning_rate": 0.0001,
+      "loss": 1.5788,
+      "step": 1363
+    },
+    {
+      "epoch": 0.45679839249832555,
+      "grad_norm": 0.4637114703655243,
+      "learning_rate": 0.0001,
+      "loss": 1.6297,
+      "step": 1364
+    },
+    {
+      "epoch": 0.45713328868050906,
+      "grad_norm": 0.4805709719657898,
+      "learning_rate": 0.0001,
+      "loss": 1.6032,
+      "step": 1365
+    },
+    {
+      "epoch": 0.4574681848626926,
+      "grad_norm": 0.44155606627464294,
+      "learning_rate": 0.0001,
+      "loss": 1.486,
+      "step": 1366
+    },
+    {
+      "epoch": 0.4578030810448761,
+      "grad_norm": 0.46776220202445984,
+      "learning_rate": 0.0001,
+      "loss": 1.5869,
+      "step": 1367
+    },
+    {
+      "epoch": 0.4581379772270596,
+      "grad_norm": 0.47851744294166565,
+      "learning_rate": 0.0001,
+      "loss": 1.5728,
+      "step": 1368
+    },
+    {
+      "epoch": 0.45847287340924314,
+      "grad_norm": 0.4378662407398224,
+      "learning_rate": 0.0001,
+      "loss": 1.5544,
+      "step": 1369
+    },
+    {
+      "epoch": 0.45880776959142666,
+      "grad_norm": 0.4366649389266968,
+      "learning_rate": 0.0001,
+      "loss": 1.455,
+      "step": 1370
+    },
+    {
+      "epoch": 0.4591426657736102,
+      "grad_norm": 0.4515934884548187,
+      "learning_rate": 0.0001,
+      "loss": 1.545,
+      "step": 1371
+    },
+    {
+      "epoch": 0.4594775619557937,
+      "grad_norm": 0.4429531395435333,
+      "learning_rate": 0.0001,
+      "loss": 1.4451,
+      "step": 1372
+    },
+    {
+      "epoch": 0.4598124581379772,
+      "grad_norm": 0.4400520920753479,
+      "learning_rate": 0.0001,
+      "loss": 1.4532,
+      "step": 1373
+    },
+    {
+      "epoch": 0.46014735432016074,
+      "grad_norm": 0.4494495987892151,
+      "learning_rate": 0.0001,
+      "loss": 1.4642,
+      "step": 1374
+    },
+    {
+      "epoch": 0.46048225050234426,
+      "grad_norm": 0.447348415851593,
+      "learning_rate": 0.0001,
+      "loss": 1.5466,
+      "step": 1375
+    },
+    {
+      "epoch": 0.4608171466845278,
+      "grad_norm": 0.4689560532569885,
+      "learning_rate": 0.0001,
+      "loss": 1.62,
+      "step": 1376
+    },
+    {
+      "epoch": 0.4611520428667113,
+      "grad_norm": 0.4379405379295349,
+      "learning_rate": 0.0001,
+      "loss": 1.4164,
+      "step": 1377
+    },
+    {
+      "epoch": 0.4614869390488948,
+      "grad_norm": 0.4504340589046478,
+      "learning_rate": 0.0001,
+      "loss": 1.5672,
+      "step": 1378
+    },
+    {
+      "epoch": 0.4618218352310784,
+      "grad_norm": 0.466804176568985,
+      "learning_rate": 0.0001,
+      "loss": 1.4973,
+      "step": 1379
+    },
+    {
+      "epoch": 0.4621567314132619,
+      "grad_norm": 0.45532482862472534,
+      "learning_rate": 0.0001,
+      "loss": 1.5122,
+      "step": 1380
+    },
+    {
+      "epoch": 0.46249162759544543,
+      "grad_norm": 0.4531165361404419,
+      "learning_rate": 0.0001,
+      "loss": 1.5097,
+      "step": 1381
+    },
+    {
+      "epoch": 0.46282652377762895,
+      "grad_norm": 0.44292378425598145,
+      "learning_rate": 0.0001,
+      "loss": 1.5498,
+      "step": 1382
+    },
+    {
+      "epoch": 0.46316141995981247,
+      "grad_norm": 0.4559877812862396,
+      "learning_rate": 0.0001,
+      "loss": 1.5847,
+      "step": 1383
+    },
+    {
+      "epoch": 0.463496316141996,
+      "grad_norm": 0.4499688744544983,
+      "learning_rate": 0.0001,
+      "loss": 1.5498,
+      "step": 1384
+    },
+    {
+      "epoch": 0.4638312123241795,
+      "grad_norm": 0.46169179677963257,
+      "learning_rate": 0.0001,
+      "loss": 1.5752,
+      "step": 1385
+    },
+    {
+      "epoch": 0.46416610850636303,
+      "grad_norm": 0.4544617235660553,
+      "learning_rate": 0.0001,
+      "loss": 1.5089,
+      "step": 1386
+    },
+    {
+      "epoch": 0.46450100468854655,
+      "grad_norm": 0.4471500515937805,
+      "learning_rate": 0.0001,
+      "loss": 1.5842,
+      "step": 1387
+    },
+    {
+      "epoch": 0.46483590087073007,
+      "grad_norm": 0.490632027387619,
+      "learning_rate": 0.0001,
+      "loss": 1.6259,
+      "step": 1388
+    },
+    {
+      "epoch": 0.4651707970529136,
+      "grad_norm": 0.46217668056488037,
+      "learning_rate": 0.0001,
+      "loss": 1.5832,
+      "step": 1389
+    },
+    {
+      "epoch": 0.4655056932350971,
+      "grad_norm": 0.45851930975914,
+      "learning_rate": 0.0001,
+      "loss": 1.4738,
+      "step": 1390
+    },
+    {
+      "epoch": 0.4658405894172806,
+      "grad_norm": 0.43563809990882874,
+      "learning_rate": 0.0001,
+      "loss": 1.4767,
+      "step": 1391
+    },
+    {
+      "epoch": 0.46617548559946415,
+      "grad_norm": 0.4546947181224823,
+      "learning_rate": 0.0001,
+      "loss": 1.5412,
+      "step": 1392
+    },
+    {
+      "epoch": 0.46651038178164767,
+      "grad_norm": 0.456242173910141,
+      "learning_rate": 0.0001,
+      "loss": 1.5476,
+      "step": 1393
+    },
+    {
+      "epoch": 0.46684527796383124,
+      "grad_norm": 0.4377875328063965,
+      "learning_rate": 0.0001,
+      "loss": 1.4826,
+      "step": 1394
+    },
+    {
+      "epoch": 0.46718017414601476,
+      "grad_norm": 0.4969845414161682,
+      "learning_rate": 0.0001,
+      "loss": 1.6719,
+      "step": 1395
+    },
+    {
+      "epoch": 0.4675150703281983,
+      "grad_norm": 0.4470406770706177,
+      "learning_rate": 0.0001,
+      "loss": 1.5464,
+      "step": 1396
+    },
+    {
+      "epoch": 0.4678499665103818,
+      "grad_norm": 0.47641536593437195,
+      "learning_rate": 0.0001,
+      "loss": 1.6137,
+      "step": 1397
+    },
+    {
+      "epoch": 0.4681848626925653,
+      "grad_norm": 0.44054463505744934,
+      "learning_rate": 0.0001,
+      "loss": 1.4965,
+      "step": 1398
+    },
+    {
+      "epoch": 0.46851975887474884,
+      "grad_norm": 0.4915468096733093,
+      "learning_rate": 0.0001,
+      "loss": 1.5505,
+      "step": 1399
+    },
+    {
+      "epoch": 0.46885465505693236,
+      "grad_norm": 0.43659910559654236,
+      "learning_rate": 0.0001,
+      "loss": 1.415,
+      "step": 1400
+    },
+    {
+      "epoch": 0.4691895512391159,
+      "grad_norm": 0.47145307064056396,
+      "learning_rate": 0.0001,
+      "loss": 1.5463,
+      "step": 1401
+    },
+    {
+      "epoch": 0.4695244474212994,
+      "grad_norm": 0.470012366771698,
+      "learning_rate": 0.0001,
+      "loss": 1.632,
+      "step": 1402
+    },
+    {
+      "epoch": 0.4698593436034829,
+      "grad_norm": 0.4900912642478943,
+      "learning_rate": 0.0001,
+      "loss": 1.5168,
+      "step": 1403
+    },
+    {
+      "epoch": 0.47019423978566643,
+      "grad_norm": 0.5177119374275208,
+      "learning_rate": 0.0001,
+      "loss": 1.5512,
+      "step": 1404
+    },
+    {
+      "epoch": 0.47052913596784995,
+      "grad_norm": 0.5017977952957153,
+      "learning_rate": 0.0001,
+      "loss": 1.4807,
+      "step": 1405
+    },
+    {
+      "epoch": 0.4708640321500335,
+      "grad_norm": 0.4529106616973877,
+      "learning_rate": 0.0001,
+      "loss": 1.5158,
+      "step": 1406
+    },
+    {
+      "epoch": 0.471198928332217,
+      "grad_norm": 0.4319479763507843,
+      "learning_rate": 0.0001,
+      "loss": 1.5249,
+      "step": 1407
+    },
+    {
+      "epoch": 0.4715338245144005,
+      "grad_norm": 0.4647986590862274,
+      "learning_rate": 0.0001,
+      "loss": 1.4653,
+      "step": 1408
+    },
+    {
+      "epoch": 0.47186872069658403,
+      "grad_norm": 0.458469420671463,
+      "learning_rate": 0.0001,
+      "loss": 1.4954,
+      "step": 1409
+    },
+    {
+      "epoch": 0.4722036168787676,
+      "grad_norm": 0.50237637758255,
+      "learning_rate": 0.0001,
+      "loss": 1.6159,
+      "step": 1410
+    },
+    {
+      "epoch": 0.4725385130609511,
+      "grad_norm": 0.46038538217544556,
+      "learning_rate": 0.0001,
+      "loss": 1.5783,
+      "step": 1411
+    },
+    {
+      "epoch": 0.47287340924313465,
+      "grad_norm": 0.46736159920692444,
+      "learning_rate": 0.0001,
+      "loss": 1.6369,
+      "step": 1412
+    },
+    {
+      "epoch": 0.47320830542531817,
+      "grad_norm": 0.47504475712776184,
+      "learning_rate": 0.0001,
+      "loss": 1.5477,
+      "step": 1413
+    },
+    {
+      "epoch": 0.4735432016075017,
+      "grad_norm": 0.4639604687690735,
+      "learning_rate": 0.0001,
+      "loss": 1.4814,
+      "step": 1414
+    },
+    {
+      "epoch": 0.4738780977896852,
+      "grad_norm": 0.4924888014793396,
+      "learning_rate": 0.0001,
+      "loss": 1.7246,
+      "step": 1415
+    },
+    {
+      "epoch": 0.4742129939718687,
+      "grad_norm": 0.4379710257053375,
+      "learning_rate": 0.0001,
+      "loss": 1.4826,
+      "step": 1416
+    },
+    {
+      "epoch": 0.47454789015405224,
+      "grad_norm": 0.4734916687011719,
+      "learning_rate": 0.0001,
+      "loss": 1.5407,
+      "step": 1417
+    },
+    {
+      "epoch": 0.47488278633623576,
+      "grad_norm": 0.4327939748764038,
+      "learning_rate": 0.0001,
+      "loss": 1.5339,
+      "step": 1418
+    },
+    {
+      "epoch": 0.4752176825184193,
+      "grad_norm": 0.4371609091758728,
+      "learning_rate": 0.0001,
+      "loss": 1.5058,
+      "step": 1419
+    },
+    {
+      "epoch": 0.4755525787006028,
+      "grad_norm": 0.43390345573425293,
+      "learning_rate": 0.0001,
+      "loss": 1.5108,
+      "step": 1420
+    },
+    {
+      "epoch": 0.4758874748827863,
+      "grad_norm": 0.44707202911376953,
+      "learning_rate": 0.0001,
+      "loss": 1.5266,
+      "step": 1421
+    },
+    {
+      "epoch": 0.47622237106496984,
+      "grad_norm": 0.4597282111644745,
+      "learning_rate": 0.0001,
+      "loss": 1.5125,
+      "step": 1422
+    },
+    {
+      "epoch": 0.47655726724715336,
+      "grad_norm": 0.495181143283844,
+      "learning_rate": 0.0001,
+      "loss": 1.5363,
+      "step": 1423
+    },
+    {
+      "epoch": 0.4768921634293369,
+      "grad_norm": 0.447334349155426,
+      "learning_rate": 0.0001,
+      "loss": 1.5074,
+      "step": 1424
+    },
+    {
+      "epoch": 0.47722705961152045,
+      "grad_norm": 0.450639545917511,
+      "learning_rate": 0.0001,
+      "loss": 1.5462,
+      "step": 1425
+    },
+    {
+      "epoch": 0.477561955793704,
+      "grad_norm": 0.4501533806324005,
+      "learning_rate": 0.0001,
+      "loss": 1.4815,
+      "step": 1426
+    },
+    {
+      "epoch": 0.4778968519758875,
+      "grad_norm": 0.4699714779853821,
+      "learning_rate": 0.0001,
+      "loss": 1.5042,
+      "step": 1427
+    },
+    {
+      "epoch": 0.478231748158071,
+      "grad_norm": 0.4734054505825043,
+      "learning_rate": 0.0001,
+      "loss": 1.6161,
+      "step": 1428
+    },
+    {
+      "epoch": 0.47856664434025453,
+      "grad_norm": 0.46738970279693604,
+      "learning_rate": 0.0001,
+      "loss": 1.5816,
+      "step": 1429
+    },
+    {
+      "epoch": 0.47890154052243805,
+      "grad_norm": 0.43620428442955017,
+      "learning_rate": 0.0001,
+      "loss": 1.5605,
+      "step": 1430
+    },
+    {
+      "epoch": 0.47923643670462157,
+      "grad_norm": 0.48260974884033203,
+      "learning_rate": 0.0001,
+      "loss": 1.6668,
+      "step": 1431
+    },
+    {
+      "epoch": 0.4795713328868051,
+      "grad_norm": 0.4639464318752289,
+      "learning_rate": 0.0001,
+      "loss": 1.5394,
+      "step": 1432
+    },
+    {
+      "epoch": 0.4799062290689886,
+      "grad_norm": 0.456023633480072,
+      "learning_rate": 0.0001,
+      "loss": 1.5617,
+      "step": 1433
+    },
+    {
+      "epoch": 0.48024112525117213,
+      "grad_norm": 0.4547133147716522,
+      "learning_rate": 0.0001,
+      "loss": 1.5456,
+      "step": 1434
+    },
+    {
+      "epoch": 0.48057602143335565,
+      "grad_norm": 0.4591653048992157,
+      "learning_rate": 0.0001,
+      "loss": 1.641,
+      "step": 1435
+    },
+    {
+      "epoch": 0.48091091761553917,
+      "grad_norm": 0.4386260211467743,
+      "learning_rate": 0.0001,
+      "loss": 1.391,
+      "step": 1436
+    },
+    {
+      "epoch": 0.4812458137977227,
+      "grad_norm": 0.48630258440971375,
+      "learning_rate": 0.0001,
+      "loss": 1.5811,
+      "step": 1437
+    },
+    {
+      "epoch": 0.4815807099799062,
+      "grad_norm": 0.46854186058044434,
+      "learning_rate": 0.0001,
+      "loss": 1.6036,
+      "step": 1438
+    },
+    {
+      "epoch": 0.4819156061620897,
+      "grad_norm": 0.450717955827713,
+      "learning_rate": 0.0001,
+      "loss": 1.4546,
+      "step": 1439
+    },
+    {
+      "epoch": 0.4822505023442733,
+      "grad_norm": 0.454939067363739,
+      "learning_rate": 0.0001,
+      "loss": 1.5829,
+      "step": 1440
+    },
+    {
+      "epoch": 0.4825853985264568,
+      "grad_norm": 0.45046836137771606,
+      "learning_rate": 0.0001,
+      "loss": 1.4901,
+      "step": 1441
+    },
+    {
+      "epoch": 0.48292029470864034,
+      "grad_norm": 0.4793626666069031,
+      "learning_rate": 0.0001,
+      "loss": 1.5315,
+      "step": 1442
+    },
+    {
+      "epoch": 0.48325519089082386,
+      "grad_norm": 0.4342884421348572,
+      "learning_rate": 0.0001,
+      "loss": 1.4576,
+      "step": 1443
+    },
+    {
+      "epoch": 0.4835900870730074,
+      "grad_norm": 0.47625261545181274,
+      "learning_rate": 0.0001,
+      "loss": 1.4951,
+      "step": 1444
+    },
+    {
+      "epoch": 0.4839249832551909,
+      "grad_norm": 0.44386520981788635,
+      "learning_rate": 0.0001,
+      "loss": 1.5058,
+      "step": 1445
+    },
+    {
+      "epoch": 0.4842598794373744,
+      "grad_norm": 0.4521249532699585,
+      "learning_rate": 0.0001,
+      "loss": 1.5013,
+      "step": 1446
+    },
+    {
+      "epoch": 0.48459477561955794,
+      "grad_norm": 0.46549564599990845,
+      "learning_rate": 0.0001,
+      "loss": 1.499,
+      "step": 1447
+    },
+    {
+      "epoch": 0.48492967180174146,
+      "grad_norm": 0.4664047956466675,
+      "learning_rate": 0.0001,
+      "loss": 1.4905,
+      "step": 1448
+    },
+    {
+      "epoch": 0.485264567983925,
+      "grad_norm": 0.45398756861686707,
+      "learning_rate": 0.0001,
+      "loss": 1.5065,
+      "step": 1449
+    },
+    {
+      "epoch": 0.4855994641661085,
+      "grad_norm": 0.46122342348098755,
+      "learning_rate": 0.0001,
+      "loss": 1.4866,
+      "step": 1450
+    },
+    {
+      "epoch": 0.485934360348292,
+      "grad_norm": 0.43852072954177856,
+      "learning_rate": 0.0001,
+      "loss": 1.4706,
+      "step": 1451
+    },
+    {
+      "epoch": 0.48626925653047554,
+      "grad_norm": 0.4522353410720825,
+      "learning_rate": 0.0001,
+      "loss": 1.5468,
+      "step": 1452
+    },
+    {
+      "epoch": 0.48660415271265905,
+      "grad_norm": 0.455821692943573,
+      "learning_rate": 0.0001,
+      "loss": 1.61,
+      "step": 1453
+    },
+    {
+      "epoch": 0.4869390488948426,
+      "grad_norm": 0.4561960697174072,
+      "learning_rate": 0.0001,
+      "loss": 1.5701,
+      "step": 1454
+    },
+    {
+      "epoch": 0.48727394507702615,
+      "grad_norm": 0.45001938939094543,
+      "learning_rate": 0.0001,
+      "loss": 1.4908,
+      "step": 1455
+    },
+    {
+      "epoch": 0.48760884125920967,
+      "grad_norm": 0.4559260904788971,
+      "learning_rate": 0.0001,
+      "loss": 1.578,
+      "step": 1456
+    },
+    {
+      "epoch": 0.4879437374413932,
+      "grad_norm": 0.4333287477493286,
+      "learning_rate": 0.0001,
+      "loss": 1.4492,
+      "step": 1457
+    },
+    {
+      "epoch": 0.4882786336235767,
+      "grad_norm": 0.4796982407569885,
+      "learning_rate": 0.0001,
+      "loss": 1.6042,
+      "step": 1458
+    },
+    {
+      "epoch": 0.4886135298057602,
+      "grad_norm": 0.43639075756073,
+      "learning_rate": 0.0001,
+      "loss": 1.4885,
+      "step": 1459
+    },
+    {
+      "epoch": 0.48894842598794375,
+      "grad_norm": 0.4537317454814911,
+      "learning_rate": 0.0001,
+      "loss": 1.5781,
+      "step": 1460
+    },
+    {
+      "epoch": 0.48928332217012727,
+      "grad_norm": 0.4368795156478882,
+      "learning_rate": 0.0001,
+      "loss": 1.4805,
+      "step": 1461
+    },
+    {
+      "epoch": 0.4896182183523108,
+      "grad_norm": 0.43467339873313904,
+      "learning_rate": 0.0001,
+      "loss": 1.4288,
+      "step": 1462
+    },
+    {
+      "epoch": 0.4899531145344943,
+      "grad_norm": 0.49252694845199585,
+      "learning_rate": 0.0001,
+      "loss": 1.5542,
+      "step": 1463
+    },
+    {
+      "epoch": 0.4902880107166778,
+      "grad_norm": 0.4549628496170044,
+      "learning_rate": 0.0001,
+      "loss": 1.5405,
+      "step": 1464
+    },
+    {
+      "epoch": 0.49062290689886134,
+      "grad_norm": 0.4875522255897522,
+      "learning_rate": 0.0001,
+      "loss": 1.6199,
+      "step": 1465
+    },
+    {
+      "epoch": 0.49095780308104486,
+      "grad_norm": 0.4471842348575592,
+      "learning_rate": 0.0001,
+      "loss": 1.5369,
+      "step": 1466
+    },
+    {
+      "epoch": 0.4912926992632284,
+      "grad_norm": 0.45362889766693115,
+      "learning_rate": 0.0001,
+      "loss": 1.4544,
+      "step": 1467
+    },
+    {
+      "epoch": 0.4916275954454119,
+      "grad_norm": 0.4525300860404968,
+      "learning_rate": 0.0001,
+      "loss": 1.4715,
+      "step": 1468
+    },
+    {
+      "epoch": 0.4919624916275954,
+      "grad_norm": 0.46428170800209045,
+      "learning_rate": 0.0001,
+      "loss": 1.5975,
+      "step": 1469
+    },
+    {
+      "epoch": 0.49229738780977894,
+      "grad_norm": 0.4399421811103821,
+      "learning_rate": 0.0001,
+      "loss": 1.4805,
+      "step": 1470
+    },
+    {
+      "epoch": 0.4926322839919625,
+      "grad_norm": 0.4486546516418457,
+      "learning_rate": 0.0001,
+      "loss": 1.4835,
+      "step": 1471
+    },
+    {
+      "epoch": 0.49296718017414604,
+      "grad_norm": 0.4494679868221283,
+      "learning_rate": 0.0001,
+      "loss": 1.507,
+      "step": 1472
+    },
+    {
+      "epoch": 0.49330207635632956,
+      "grad_norm": 0.49521908164024353,
+      "learning_rate": 0.0001,
+      "loss": 1.6279,
+      "step": 1473
+    },
+    {
+      "epoch": 0.4936369725385131,
+      "grad_norm": 0.4433446526527405,
+      "learning_rate": 0.0001,
+      "loss": 1.5223,
+      "step": 1474
+    },
+    {
+      "epoch": 0.4939718687206966,
+      "grad_norm": 0.47312355041503906,
+      "learning_rate": 0.0001,
+      "loss": 1.5241,
+      "step": 1475
+    },
+    {
+      "epoch": 0.4943067649028801,
+      "grad_norm": 0.42530277371406555,
+      "learning_rate": 0.0001,
+      "loss": 1.4212,
+      "step": 1476
+    },
+    {
+      "epoch": 0.49464166108506363,
+      "grad_norm": 0.4690539240837097,
+      "learning_rate": 0.0001,
+      "loss": 1.4736,
+      "step": 1477
+    },
+    {
+      "epoch": 0.49497655726724715,
+      "grad_norm": 0.44447723031044006,
+      "learning_rate": 0.0001,
+      "loss": 1.5079,
+      "step": 1478
+    },
+    {
+      "epoch": 0.49531145344943067,
+      "grad_norm": 0.4463139772415161,
+      "learning_rate": 0.0001,
+      "loss": 1.4826,
+      "step": 1479
+    },
+    {
+      "epoch": 0.4956463496316142,
+      "grad_norm": 0.45554107427597046,
+      "learning_rate": 0.0001,
+      "loss": 1.5286,
+      "step": 1480
+    },
+    {
+      "epoch": 0.4959812458137977,
+      "grad_norm": 0.45664557814598083,
+      "learning_rate": 0.0001,
+      "loss": 1.5602,
+      "step": 1481
+    },
+    {
+      "epoch": 0.49631614199598123,
+      "grad_norm": 0.46041151881217957,
+      "learning_rate": 0.0001,
+      "loss": 1.5732,
+      "step": 1482
+    },
+    {
+      "epoch": 0.49665103817816475,
+      "grad_norm": 0.46366506814956665,
+      "learning_rate": 0.0001,
+      "loss": 1.497,
+      "step": 1483
+    },
+    {
+      "epoch": 0.49698593436034827,
+      "grad_norm": 0.46002379059791565,
+      "learning_rate": 0.0001,
+      "loss": 1.5219,
+      "step": 1484
+    },
+    {
+      "epoch": 0.4973208305425318,
+      "grad_norm": 0.4685385525226593,
+      "learning_rate": 0.0001,
+      "loss": 1.5126,
+      "step": 1485
+    },
+    {
+      "epoch": 0.49765572672471536,
+      "grad_norm": 0.4530273973941803,
+      "learning_rate": 0.0001,
+      "loss": 1.5677,
+      "step": 1486
+    },
+    {
+      "epoch": 0.4979906229068989,
+      "grad_norm": 0.450147807598114,
+      "learning_rate": 0.0001,
+      "loss": 1.5109,
+      "step": 1487
+    },
+    {
+      "epoch": 0.4983255190890824,
+      "grad_norm": 0.4494372010231018,
+      "learning_rate": 0.0001,
+      "loss": 1.5489,
+      "step": 1488
+    },
+    {
+      "epoch": 0.4986604152712659,
+      "grad_norm": 0.43944913148880005,
+      "learning_rate": 0.0001,
+      "loss": 1.5326,
+      "step": 1489
+    },
+    {
+      "epoch": 0.49899531145344944,
+      "grad_norm": 0.4616487920284271,
+      "learning_rate": 0.0001,
+      "loss": 1.4999,
+      "step": 1490
+    },
+    {
+      "epoch": 0.49933020763563296,
+      "grad_norm": 0.4538317918777466,
+      "learning_rate": 0.0001,
+      "loss": 1.4909,
+      "step": 1491
+    },
+    {
+      "epoch": 0.4996651038178165,
+      "grad_norm": 0.45227381587028503,
+      "learning_rate": 0.0001,
+      "loss": 1.4275,
+      "step": 1492
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.44475820660591125,
+      "learning_rate": 0.0001,
+      "loss": 1.4533,
+      "step": 1493
+    },
+    {
+      "epoch": 0.5003348961821835,
+      "grad_norm": 0.46907156705856323,
+      "learning_rate": 0.0001,
+      "loss": 1.5041,
+      "step": 1494
+    },
+    {
+      "epoch": 0.500669792364367,
+      "grad_norm": 0.4697141647338867,
+      "learning_rate": 0.0001,
+      "loss": 1.5366,
+      "step": 1495
+    },
+    {
+      "epoch": 0.5010046885465506,
+      "grad_norm": 0.4554111361503601,
+      "learning_rate": 0.0001,
+      "loss": 1.5005,
+      "step": 1496
+    },
+    {
+      "epoch": 0.5013395847287341,
+      "grad_norm": 0.44460198283195496,
+      "learning_rate": 0.0001,
+      "loss": 1.5039,
+      "step": 1497
+    },
+    {
+      "epoch": 0.5016744809109176,
+      "grad_norm": 0.46140316128730774,
+      "learning_rate": 0.0001,
+      "loss": 1.5757,
+      "step": 1498
+    },
+    {
+      "epoch": 0.5020093770931011,
+      "grad_norm": 0.4617026746273041,
+      "learning_rate": 0.0001,
+      "loss": 1.5548,
+      "step": 1499
+    },
+    {
+      "epoch": 0.5023442732752846,
+      "grad_norm": 0.4561994671821594,
+      "learning_rate": 0.0001,
+      "loss": 1.4872,
+      "step": 1500
+    },
+    {
+      "epoch": 0.5026791694574682,
+      "grad_norm": 0.45759817957878113,
+      "learning_rate": 0.0001,
+      "loss": 1.5807,
+      "step": 1501
+    },
+    {
+      "epoch": 0.5030140656396517,
+      "grad_norm": 0.45219355821609497,
+      "learning_rate": 0.0001,
+      "loss": 1.4733,
+      "step": 1502
+    },
+    {
+      "epoch": 0.5033489618218352,
+      "grad_norm": 0.4509015679359436,
+      "learning_rate": 0.0001,
+      "loss": 1.5254,
+      "step": 1503
+    },
+    {
+      "epoch": 0.5036838580040187,
+      "grad_norm": 0.4564638137817383,
+      "learning_rate": 0.0001,
+      "loss": 1.5299,
+      "step": 1504
+    },
+    {
+      "epoch": 0.5040187541862022,
+      "grad_norm": 0.45715585350990295,
+      "learning_rate": 0.0001,
+      "loss": 1.5694,
+      "step": 1505
+    },
+    {
+      "epoch": 0.5043536503683858,
+      "grad_norm": 0.45385318994522095,
+      "learning_rate": 0.0001,
+      "loss": 1.4992,
+      "step": 1506
+    },
+    {
+      "epoch": 0.5046885465505693,
+      "grad_norm": 0.44898638129234314,
+      "learning_rate": 0.0001,
+      "loss": 1.4665,
+      "step": 1507
+    },
+    {
+      "epoch": 0.5050234427327528,
+      "grad_norm": 0.46023446321487427,
+      "learning_rate": 0.0001,
+      "loss": 1.4715,
+      "step": 1508
+    },
+    {
+      "epoch": 0.5053583389149364,
+      "grad_norm": 0.44878068566322327,
+      "learning_rate": 0.0001,
+      "loss": 1.492,
+      "step": 1509
+    },
+    {
+      "epoch": 0.5056932350971199,
+      "grad_norm": 0.46955686807632446,
+      "learning_rate": 0.0001,
+      "loss": 1.5671,
+      "step": 1510
+    },
+    {
+      "epoch": 0.5060281312793035,
+      "grad_norm": 0.46982455253601074,
+      "learning_rate": 0.0001,
+      "loss": 1.4894,
+      "step": 1511
+    },
+    {
+      "epoch": 0.506363027461487,
+      "grad_norm": 0.45919641852378845,
+      "learning_rate": 0.0001,
+      "loss": 1.5646,
+      "step": 1512
+    },
+    {
+      "epoch": 0.5066979236436705,
+      "grad_norm": 0.47328880429267883,
+      "learning_rate": 0.0001,
+      "loss": 1.6072,
+      "step": 1513
+    },
+    {
+      "epoch": 0.507032819825854,
+      "grad_norm": 0.4526023268699646,
+      "learning_rate": 0.0001,
+      "loss": 1.501,
+      "step": 1514
+    },
+    {
+      "epoch": 0.5073677160080375,
+      "grad_norm": 0.46821823716163635,
+      "learning_rate": 0.0001,
+      "loss": 1.5246,
+      "step": 1515
+    },
+    {
+      "epoch": 0.5077026121902211,
+      "grad_norm": 0.45301198959350586,
+      "learning_rate": 0.0001,
+      "loss": 1.5364,
+      "step": 1516
+    },
+    {
+      "epoch": 0.5080375083724046,
+      "grad_norm": 0.46978890895843506,
+      "learning_rate": 0.0001,
+      "loss": 1.5137,
+      "step": 1517
+    },
+    {
+      "epoch": 0.5083724045545881,
+      "grad_norm": 0.4836040735244751,
+      "learning_rate": 0.0001,
+      "loss": 1.5262,
+      "step": 1518
+    },
+    {
+      "epoch": 0.5087073007367716,
+      "grad_norm": 0.4945816099643707,
+      "learning_rate": 0.0001,
+      "loss": 1.5042,
+      "step": 1519
+    },
+    {
+      "epoch": 0.5090421969189551,
+      "grad_norm": 0.46678590774536133,
+      "learning_rate": 0.0001,
+      "loss": 1.4932,
+      "step": 1520
+    },
+    {
+      "epoch": 0.5093770931011387,
+      "grad_norm": 0.4569990038871765,
+      "learning_rate": 0.0001,
+      "loss": 1.5965,
+      "step": 1521
+    },
+    {
+      "epoch": 0.5097119892833222,
+      "grad_norm": 0.4823521375656128,
+      "learning_rate": 0.0001,
+      "loss": 1.5114,
+      "step": 1522
+    },
+    {
+      "epoch": 0.5100468854655057,
+      "grad_norm": 0.44416478276252747,
+      "learning_rate": 0.0001,
+      "loss": 1.4804,
+      "step": 1523
+    },
+    {
+      "epoch": 0.5103817816476892,
+      "grad_norm": 0.46118879318237305,
+      "learning_rate": 0.0001,
+      "loss": 1.5135,
+      "step": 1524
+    },
+    {
+      "epoch": 0.5107166778298727,
+      "grad_norm": 0.4893559515476227,
+      "learning_rate": 0.0001,
+      "loss": 1.6105,
+      "step": 1525
+    },
+    {
+      "epoch": 0.5110515740120563,
+      "grad_norm": 0.4696842133998871,
+      "learning_rate": 0.0001,
+      "loss": 1.5501,
+      "step": 1526
+    },
+    {
+      "epoch": 0.5113864701942398,
+      "grad_norm": 0.5141232013702393,
+      "learning_rate": 0.0001,
+      "loss": 1.6102,
+      "step": 1527
+    },
+    {
+      "epoch": 0.5117213663764233,
+      "grad_norm": 0.47689032554626465,
+      "learning_rate": 0.0001,
+      "loss": 1.6108,
+      "step": 1528
+    },
+    {
+      "epoch": 0.5120562625586068,
+      "grad_norm": 0.47305381298065186,
+      "learning_rate": 0.0001,
+      "loss": 1.6228,
+      "step": 1529
+    },
+    {
+      "epoch": 0.5123911587407903,
+      "grad_norm": 0.4634510576725006,
+      "learning_rate": 0.0001,
+      "loss": 1.5599,
+      "step": 1530
+    },
+    {
+      "epoch": 0.5127260549229739,
+      "grad_norm": 0.466594398021698,
+      "learning_rate": 0.0001,
+      "loss": 1.5349,
+      "step": 1531
+    },
+    {
+      "epoch": 0.5130609511051574,
+      "grad_norm": 0.47930100560188293,
+      "learning_rate": 0.0001,
+      "loss": 1.5236,
+      "step": 1532
+    },
+    {
+      "epoch": 0.5133958472873409,
+      "grad_norm": 0.44495946168899536,
+      "learning_rate": 0.0001,
+      "loss": 1.5767,
+      "step": 1533
+    },
+    {
+      "epoch": 0.5137307434695244,
+      "grad_norm": 0.4560399651527405,
+      "learning_rate": 0.0001,
+      "loss": 1.512,
+      "step": 1534
+    },
+    {
+      "epoch": 0.5140656396517079,
+      "grad_norm": 0.47690311074256897,
+      "learning_rate": 0.0001,
+      "loss": 1.5891,
+      "step": 1535
+    },
+    {
+      "epoch": 0.5144005358338914,
+      "grad_norm": 0.4491158425807953,
+      "learning_rate": 0.0001,
+      "loss": 1.5635,
+      "step": 1536
+    },
+    {
+      "epoch": 0.514735432016075,
+      "grad_norm": 0.4559384882450104,
+      "learning_rate": 0.0001,
+      "loss": 1.6068,
+      "step": 1537
+    },
+    {
+      "epoch": 0.5150703281982585,
+      "grad_norm": 0.4365287721157074,
+      "learning_rate": 0.0001,
+      "loss": 1.4712,
+      "step": 1538
+    },
+    {
+      "epoch": 0.5154052243804421,
+      "grad_norm": 0.48256635665893555,
+      "learning_rate": 0.0001,
+      "loss": 1.6654,
+      "step": 1539
+    },
+    {
+      "epoch": 0.5157401205626256,
+      "grad_norm": 0.4454759955406189,
+      "learning_rate": 0.0001,
+      "loss": 1.5364,
+      "step": 1540
+    },
+    {
+      "epoch": 0.5160750167448092,
+      "grad_norm": 0.46772095561027527,
+      "learning_rate": 0.0001,
+      "loss": 1.442,
+      "step": 1541
+    },
+    {
+      "epoch": 0.5164099129269927,
+      "grad_norm": 0.4505743682384491,
+      "learning_rate": 0.0001,
+      "loss": 1.452,
+      "step": 1542
+    },
+    {
+      "epoch": 0.5167448091091762,
+      "grad_norm": 0.516400158405304,
+      "learning_rate": 0.0001,
+      "loss": 1.5716,
+      "step": 1543
+    },
+    {
+      "epoch": 0.5170797052913597,
+      "grad_norm": 0.48185044527053833,
+      "learning_rate": 0.0001,
+      "loss": 1.6428,
+      "step": 1544
+    },
+    {
+      "epoch": 0.5174146014735432,
+      "grad_norm": 0.5100380778312683,
+      "learning_rate": 0.0001,
+      "loss": 1.6228,
+      "step": 1545
+    },
+    {
+      "epoch": 0.5177494976557268,
+      "grad_norm": 0.4634498655796051,
+      "learning_rate": 0.0001,
+      "loss": 1.5288,
+      "step": 1546
+    },
+    {
+      "epoch": 0.5180843938379103,
+      "grad_norm": 0.466386616230011,
+      "learning_rate": 0.0001,
+      "loss": 1.6542,
+      "step": 1547
+    },
+    {
+      "epoch": 0.5184192900200938,
+      "grad_norm": 0.483235627412796,
+      "learning_rate": 0.0001,
+      "loss": 1.6404,
+      "step": 1548
+    },
+    {
+      "epoch": 0.5187541862022773,
+      "grad_norm": 0.4495801627635956,
+      "learning_rate": 0.0001,
+      "loss": 1.4987,
+      "step": 1549
+    },
+    {
+      "epoch": 0.5190890823844608,
+      "grad_norm": 0.4899047017097473,
+      "learning_rate": 0.0001,
+      "loss": 1.5573,
+      "step": 1550
+    },
+    {
+      "epoch": 0.5194239785666444,
+      "grad_norm": 0.43485280871391296,
+      "learning_rate": 0.0001,
+      "loss": 1.5082,
+      "step": 1551
+    },
+    {
+      "epoch": 0.5197588747488279,
+      "grad_norm": 0.47759371995925903,
+      "learning_rate": 0.0001,
+      "loss": 1.5424,
+      "step": 1552
+    },
+    {
+      "epoch": 0.5200937709310114,
+      "grad_norm": 0.44978150725364685,
+      "learning_rate": 0.0001,
+      "loss": 1.5328,
+      "step": 1553
+    },
+    {
+      "epoch": 0.5204286671131949,
+      "grad_norm": 0.46513211727142334,
+      "learning_rate": 0.0001,
+      "loss": 1.6004,
+      "step": 1554
+    },
+    {
+      "epoch": 0.5207635632953784,
+      "grad_norm": 0.45478302240371704,
+      "learning_rate": 0.0001,
+      "loss": 1.5036,
+      "step": 1555
+    },
+    {
+      "epoch": 0.521098459477562,
+      "grad_norm": 0.46547824144363403,
+      "learning_rate": 0.0001,
+      "loss": 1.5354,
+      "step": 1556
+    },
+    {
+      "epoch": 0.5214333556597455,
+      "grad_norm": 0.44873952865600586,
+      "learning_rate": 0.0001,
+      "loss": 1.5017,
+      "step": 1557
+    },
+    {
+      "epoch": 0.521768251841929,
+      "grad_norm": 0.430286705493927,
+      "learning_rate": 0.0001,
+      "loss": 1.4757,
+      "step": 1558
+    },
+    {
+      "epoch": 0.5221031480241125,
+      "grad_norm": 0.44566696882247925,
+      "learning_rate": 0.0001,
+      "loss": 1.5695,
+      "step": 1559
+    },
+    {
+      "epoch": 0.522438044206296,
+      "grad_norm": 0.4625225067138672,
+      "learning_rate": 0.0001,
+      "loss": 1.6006,
+      "step": 1560
+    },
+    {
+      "epoch": 0.5227729403884795,
+      "grad_norm": 0.45360034704208374,
+      "learning_rate": 0.0001,
+      "loss": 1.5741,
+      "step": 1561
+    },
+    {
+      "epoch": 0.5231078365706631,
+      "grad_norm": 0.4490211009979248,
+      "learning_rate": 0.0001,
+      "loss": 1.5536,
+      "step": 1562
+    },
+    {
+      "epoch": 0.5234427327528466,
+      "grad_norm": 0.4657514691352844,
+      "learning_rate": 0.0001,
+      "loss": 1.4871,
+      "step": 1563
+    },
+    {
+      "epoch": 0.5237776289350301,
+      "grad_norm": 0.44376838207244873,
+      "learning_rate": 0.0001,
+      "loss": 1.5381,
+      "step": 1564
+    },
+    {
+      "epoch": 0.5241125251172136,
+      "grad_norm": 0.47033995389938354,
+      "learning_rate": 0.0001,
+      "loss": 1.619,
+      "step": 1565
+    },
+    {
+      "epoch": 0.5244474212993971,
+      "grad_norm": 0.4549630582332611,
+      "learning_rate": 0.0001,
+      "loss": 1.5554,
+      "step": 1566
+    },
+    {
+      "epoch": 0.5247823174815807,
+      "grad_norm": 0.446274071931839,
+      "learning_rate": 0.0001,
+      "loss": 1.4818,
+      "step": 1567
+    },
+    {
+      "epoch": 0.5251172136637642,
+      "grad_norm": 0.45129215717315674,
+      "learning_rate": 0.0001,
+      "loss": 1.5083,
+      "step": 1568
+    },
+    {
+      "epoch": 0.5254521098459477,
+      "grad_norm": 0.4612513780593872,
+      "learning_rate": 0.0001,
+      "loss": 1.5301,
+      "step": 1569
+    },
+    {
+      "epoch": 0.5257870060281313,
+      "grad_norm": 0.4741304814815521,
+      "learning_rate": 0.0001,
+      "loss": 1.5815,
+      "step": 1570
+    },
+    {
+      "epoch": 0.5261219022103149,
+      "grad_norm": 0.45285868644714355,
+      "learning_rate": 0.0001,
+      "loss": 1.5875,
+      "step": 1571
+    },
+    {
+      "epoch": 0.5264567983924984,
+      "grad_norm": 0.4500559866428375,
+      "learning_rate": 0.0001,
+      "loss": 1.5457,
+      "step": 1572
+    },
+    {
+      "epoch": 0.5267916945746819,
+      "grad_norm": 0.45833030343055725,
+      "learning_rate": 0.0001,
+      "loss": 1.6693,
+      "step": 1573
+    },
+    {
+      "epoch": 0.5271265907568654,
+      "grad_norm": 0.4745745062828064,
+      "learning_rate": 0.0001,
+      "loss": 1.5703,
+      "step": 1574
+    },
+    {
+      "epoch": 0.5274614869390489,
+      "grad_norm": 0.44691261649131775,
+      "learning_rate": 0.0001,
+      "loss": 1.5275,
+      "step": 1575
+    },
+    {
+      "epoch": 0.5277963831212324,
+      "grad_norm": 0.473972350358963,
+      "learning_rate": 0.0001,
+      "loss": 1.5378,
+      "step": 1576
+    },
+    {
+      "epoch": 0.528131279303416,
+      "grad_norm": 0.4648163318634033,
+      "learning_rate": 0.0001,
+      "loss": 1.5336,
+      "step": 1577
+    },
+    {
+      "epoch": 0.5284661754855995,
+      "grad_norm": 0.4657084047794342,
+      "learning_rate": 0.0001,
+      "loss": 1.5296,
+      "step": 1578
+    },
+    {
+      "epoch": 0.528801071667783,
+      "grad_norm": 0.43905413150787354,
+      "learning_rate": 0.0001,
+      "loss": 1.46,
+      "step": 1579
+    },
+    {
+      "epoch": 0.5291359678499665,
+      "grad_norm": 0.4439171850681305,
+      "learning_rate": 0.0001,
+      "loss": 1.4725,
+      "step": 1580
+    },
+    {
+      "epoch": 0.52947086403215,
+      "grad_norm": 0.47205954790115356,
+      "learning_rate": 0.0001,
+      "loss": 1.6402,
+      "step": 1581
+    },
+    {
+      "epoch": 0.5298057602143336,
+      "grad_norm": 0.4546120762825012,
+      "learning_rate": 0.0001,
+      "loss": 1.5361,
+      "step": 1582
+    },
+    {
+      "epoch": 0.5301406563965171,
+      "grad_norm": 0.44714152812957764,
+      "learning_rate": 0.0001,
+      "loss": 1.5889,
+      "step": 1583
+    },
+    {
+      "epoch": 0.5304755525787006,
+      "grad_norm": 0.4529459476470947,
+      "learning_rate": 0.0001,
+      "loss": 1.5373,
+      "step": 1584
+    },
+    {
+      "epoch": 0.5308104487608841,
+      "grad_norm": 0.45140767097473145,
+      "learning_rate": 0.0001,
+      "loss": 1.5747,
+      "step": 1585
+    },
+    {
+      "epoch": 0.5311453449430676,
+      "grad_norm": 0.4440329968929291,
+      "learning_rate": 0.0001,
+      "loss": 1.5368,
+      "step": 1586
+    },
+    {
+      "epoch": 0.5314802411252512,
+      "grad_norm": 0.45267820358276367,
+      "learning_rate": 0.0001,
+      "loss": 1.5345,
+      "step": 1587
+    },
+    {
+      "epoch": 0.5318151373074347,
+      "grad_norm": 0.45014524459838867,
+      "learning_rate": 0.0001,
+      "loss": 1.5113,
+      "step": 1588
+    },
+    {
+      "epoch": 0.5321500334896182,
+      "grad_norm": 0.46034717559814453,
+      "learning_rate": 0.0001,
+      "loss": 1.5748,
+      "step": 1589
+    },
+    {
+      "epoch": 0.5324849296718017,
+      "grad_norm": 0.4627227187156677,
+      "learning_rate": 0.0001,
+      "loss": 1.5218,
+      "step": 1590
+    },
+    {
+      "epoch": 0.5328198258539852,
+      "grad_norm": 0.45107871294021606,
+      "learning_rate": 0.0001,
+      "loss": 1.5449,
+      "step": 1591
+    },
+    {
+      "epoch": 0.5331547220361688,
+      "grad_norm": 0.45086443424224854,
+      "learning_rate": 0.0001,
+      "loss": 1.5803,
+      "step": 1592
+    },
+    {
+      "epoch": 0.5334896182183523,
+      "grad_norm": 0.4722006916999817,
+      "learning_rate": 0.0001,
+      "loss": 1.5947,
+      "step": 1593
+    },
+    {
+      "epoch": 0.5338245144005358,
+      "grad_norm": 0.44285401701927185,
+      "learning_rate": 0.0001,
+      "loss": 1.4987,
+      "step": 1594
+    },
+    {
+      "epoch": 0.5341594105827193,
+      "grad_norm": 0.4450331926345825,
+      "learning_rate": 0.0001,
+      "loss": 1.4929,
+      "step": 1595
+    },
+    {
+      "epoch": 0.5344943067649028,
+      "grad_norm": 0.4595484733581543,
+      "learning_rate": 0.0001,
+      "loss": 1.6151,
+      "step": 1596
+    },
+    {
+      "epoch": 0.5348292029470864,
+      "grad_norm": 0.44541820883750916,
+      "learning_rate": 0.0001,
+      "loss": 1.5107,
+      "step": 1597
+    },
+    {
+      "epoch": 0.5351640991292699,
+      "grad_norm": 0.4498431980609894,
+      "learning_rate": 0.0001,
+      "loss": 1.4339,
+      "step": 1598
+    },
+    {
+      "epoch": 0.5354989953114534,
+      "grad_norm": 0.47547462582588196,
+      "learning_rate": 0.0001,
+      "loss": 1.5636,
+      "step": 1599
+    },
+    {
+      "epoch": 0.535833891493637,
+      "grad_norm": 0.4520166516304016,
+      "learning_rate": 0.0001,
+      "loss": 1.515,
+      "step": 1600
+    },
+    {
+      "epoch": 0.5361687876758205,
+      "grad_norm": 0.47776710987091064,
+      "learning_rate": 0.0001,
+      "loss": 1.5845,
+      "step": 1601
+    },
+    {
+      "epoch": 0.5365036838580041,
+      "grad_norm": 0.4761412441730499,
+      "learning_rate": 0.0001,
+      "loss": 1.5393,
+      "step": 1602
+    },
+    {
+      "epoch": 0.5368385800401876,
+      "grad_norm": 0.45610764622688293,
+      "learning_rate": 0.0001,
+      "loss": 1.4664,
+      "step": 1603
+    },
+    {
+      "epoch": 0.5371734762223711,
+      "grad_norm": 0.4365233778953552,
+      "learning_rate": 0.0001,
+      "loss": 1.5039,
+      "step": 1604
+    },
+    {
+      "epoch": 0.5375083724045546,
+      "grad_norm": 0.4999910891056061,
+      "learning_rate": 0.0001,
+      "loss": 1.6004,
+      "step": 1605
+    },
+    {
+      "epoch": 0.5378432685867381,
+      "grad_norm": 0.44889217615127563,
+      "learning_rate": 0.0001,
+      "loss": 1.4564,
+      "step": 1606
+    },
+    {
+      "epoch": 0.5381781647689217,
+      "grad_norm": 0.44685789942741394,
+      "learning_rate": 0.0001,
+      "loss": 1.5223,
+      "step": 1607
+    },
+    {
+      "epoch": 0.5385130609511052,
+      "grad_norm": 0.4566046893596649,
+      "learning_rate": 0.0001,
+      "loss": 1.5094,
+      "step": 1608
+    },
+    {
+      "epoch": 0.5388479571332887,
+      "grad_norm": 0.4515156149864197,
+      "learning_rate": 0.0001,
+      "loss": 1.4626,
+      "step": 1609
+    },
+    {
+      "epoch": 0.5391828533154722,
+      "grad_norm": 0.4659847021102905,
+      "learning_rate": 0.0001,
+      "loss": 1.4892,
+      "step": 1610
+    },
+    {
+      "epoch": 0.5395177494976557,
+      "grad_norm": 0.46315014362335205,
+      "learning_rate": 0.0001,
+      "loss": 1.6071,
+      "step": 1611
+    },
+    {
+      "epoch": 0.5398526456798393,
+      "grad_norm": 0.4561232924461365,
+      "learning_rate": 0.0001,
+      "loss": 1.4692,
+      "step": 1612
+    },
+    {
+      "epoch": 0.5401875418620228,
+      "grad_norm": 0.4539635181427002,
+      "learning_rate": 0.0001,
+      "loss": 1.5656,
+      "step": 1613
+    },
+    {
+      "epoch": 0.5405224380442063,
+      "grad_norm": 0.48085281252861023,
+      "learning_rate": 0.0001,
+      "loss": 1.6298,
+      "step": 1614
+    },
+    {
+      "epoch": 0.5408573342263898,
+      "grad_norm": 0.44261398911476135,
+      "learning_rate": 0.0001,
+      "loss": 1.4697,
+      "step": 1615
+    },
+    {
+      "epoch": 0.5411922304085733,
+      "grad_norm": 0.46685075759887695,
+      "learning_rate": 0.0001,
+      "loss": 1.5277,
+      "step": 1616
+    },
+    {
+      "epoch": 0.5415271265907569,
+      "grad_norm": 0.46934112906455994,
+      "learning_rate": 0.0001,
+      "loss": 1.5227,
+      "step": 1617
+    },
+    {
+      "epoch": 0.5418620227729404,
+      "grad_norm": 0.4731907844543457,
+      "learning_rate": 0.0001,
+      "loss": 1.6273,
+      "step": 1618
+    },
+    {
+      "epoch": 0.5421969189551239,
+      "grad_norm": 0.4367448687553406,
+      "learning_rate": 0.0001,
+      "loss": 1.5548,
+      "step": 1619
+    },
+    {
+      "epoch": 0.5425318151373074,
+      "grad_norm": 0.47099941968917847,
+      "learning_rate": 0.0001,
+      "loss": 1.617,
+      "step": 1620
+    },
+    {
+      "epoch": 0.5428667113194909,
+      "grad_norm": 0.46735990047454834,
+      "learning_rate": 0.0001,
+      "loss": 1.5671,
+      "step": 1621
+    },
+    {
+      "epoch": 0.5432016075016745,
+      "grad_norm": 0.45181360840797424,
+      "learning_rate": 0.0001,
+      "loss": 1.5791,
+      "step": 1622
+    },
+    {
+      "epoch": 0.543536503683858,
+      "grad_norm": 0.464851975440979,
+      "learning_rate": 0.0001,
+      "loss": 1.4774,
+      "step": 1623
+    },
+    {
+      "epoch": 0.5438713998660415,
+      "grad_norm": 0.4605390429496765,
+      "learning_rate": 0.0001,
+      "loss": 1.5034,
+      "step": 1624
+    },
+    {
+      "epoch": 0.544206296048225,
+      "grad_norm": 0.46630990505218506,
+      "learning_rate": 0.0001,
+      "loss": 1.5039,
+      "step": 1625
+    },
+    {
+      "epoch": 0.5445411922304085,
+      "grad_norm": 0.4575473964214325,
+      "learning_rate": 0.0001,
+      "loss": 1.4877,
+      "step": 1626
+    },
+    {
+      "epoch": 0.544876088412592,
+      "grad_norm": 0.46546870470046997,
+      "learning_rate": 0.0001,
+      "loss": 1.5551,
+      "step": 1627
+    },
+    {
+      "epoch": 0.5452109845947756,
+      "grad_norm": 0.464974969625473,
+      "learning_rate": 0.0001,
+      "loss": 1.6319,
+      "step": 1628
+    },
+    {
+      "epoch": 0.5455458807769591,
+      "grad_norm": 0.4432606101036072,
+      "learning_rate": 0.0001,
+      "loss": 1.552,
+      "step": 1629
+    },
+    {
+      "epoch": 0.5458807769591426,
+      "grad_norm": 0.43254828453063965,
+      "learning_rate": 0.0001,
+      "loss": 1.4958,
+      "step": 1630
+    },
+    {
+      "epoch": 0.5462156731413262,
+      "grad_norm": 0.43968817591667175,
+      "learning_rate": 0.0001,
+      "loss": 1.5392,
+      "step": 1631
+    },
+    {
+      "epoch": 0.5465505693235098,
+      "grad_norm": 0.4545215964317322,
+      "learning_rate": 0.0001,
+      "loss": 1.4829,
+      "step": 1632
+    },
+    {
+      "epoch": 0.5468854655056933,
+      "grad_norm": 0.4330245852470398,
+      "learning_rate": 0.0001,
+      "loss": 1.4775,
+      "step": 1633
+    },
+    {
+      "epoch": 0.5472203616878768,
+      "grad_norm": 0.45169565081596375,
+      "learning_rate": 0.0001,
+      "loss": 1.5474,
+      "step": 1634
+    },
+    {
+      "epoch": 0.5475552578700603,
+      "grad_norm": 0.472061425447464,
+      "learning_rate": 0.0001,
+      "loss": 1.4483,
+      "step": 1635
+    },
+    {
+      "epoch": 0.5478901540522438,
+      "grad_norm": 0.4540085792541504,
+      "learning_rate": 0.0001,
+      "loss": 1.5259,
+      "step": 1636
+    },
+    {
+      "epoch": 0.5482250502344274,
+      "grad_norm": 0.4446408152580261,
+      "learning_rate": 0.0001,
+      "loss": 1.4931,
+      "step": 1637
+    },
+    {
+      "epoch": 0.5485599464166109,
+      "grad_norm": 0.4567773640155792,
+      "learning_rate": 0.0001,
+      "loss": 1.5267,
+      "step": 1638
+    },
+    {
+      "epoch": 0.5488948425987944,
+      "grad_norm": 0.447899729013443,
+      "learning_rate": 0.0001,
+      "loss": 1.5575,
+      "step": 1639
+    },
+    {
+      "epoch": 0.5492297387809779,
+      "grad_norm": 0.432354599237442,
+      "learning_rate": 0.0001,
+      "loss": 1.4557,
+      "step": 1640
+    },
+    {
+      "epoch": 0.5495646349631614,
+      "grad_norm": 0.4684544503688812,
+      "learning_rate": 0.0001,
+      "loss": 1.5363,
+      "step": 1641
+    },
+    {
+      "epoch": 0.549899531145345,
+      "grad_norm": 0.4516255855560303,
+      "learning_rate": 0.0001,
+      "loss": 1.6221,
+      "step": 1642
+    },
+    {
+      "epoch": 0.5502344273275285,
+      "grad_norm": 0.4939975142478943,
+      "learning_rate": 0.0001,
+      "loss": 1.5669,
+      "step": 1643
+    },
+    {
+      "epoch": 0.550569323509712,
+      "grad_norm": 0.4381793141365051,
+      "learning_rate": 0.0001,
+      "loss": 1.4645,
+      "step": 1644
+    },
+    {
+      "epoch": 0.5509042196918955,
+      "grad_norm": 0.4454663097858429,
+      "learning_rate": 0.0001,
+      "loss": 1.5221,
+      "step": 1645
+    },
+    {
+      "epoch": 0.551239115874079,
+      "grad_norm": 0.46727123856544495,
+      "learning_rate": 0.0001,
+      "loss": 1.5181,
+      "step": 1646
+    },
+    {
+      "epoch": 0.5515740120562626,
+      "grad_norm": 0.4579801857471466,
+      "learning_rate": 0.0001,
+      "loss": 1.5446,
+      "step": 1647
+    },
+    {
+      "epoch": 0.5519089082384461,
+      "grad_norm": 0.4584062099456787,
+      "learning_rate": 0.0001,
+      "loss": 1.5231,
+      "step": 1648
+    },
+    {
+      "epoch": 0.5522438044206296,
+      "grad_norm": 0.4621238112449646,
+      "learning_rate": 0.0001,
+      "loss": 1.486,
+      "step": 1649
+    },
+    {
+      "epoch": 0.5525787006028131,
+      "grad_norm": 0.4448164403438568,
+      "learning_rate": 0.0001,
+      "loss": 1.5596,
+      "step": 1650
+    },
+    {
+      "epoch": 0.5529135967849966,
+      "grad_norm": 0.4601452946662903,
+      "learning_rate": 0.0001,
+      "loss": 1.5746,
+      "step": 1651
+    },
+    {
+      "epoch": 0.5532484929671801,
+      "grad_norm": 0.49113860726356506,
+      "learning_rate": 0.0001,
+      "loss": 1.5183,
+      "step": 1652
+    },
+    {
+      "epoch": 0.5535833891493637,
+      "grad_norm": 0.46025919914245605,
+      "learning_rate": 0.0001,
+      "loss": 1.4989,
+      "step": 1653
+    },
+    {
+      "epoch": 0.5539182853315472,
+      "grad_norm": 0.46592333912849426,
+      "learning_rate": 0.0001,
+      "loss": 1.5384,
+      "step": 1654
+    },
+    {
+      "epoch": 0.5542531815137307,
+      "grad_norm": 0.44121989607810974,
+      "learning_rate": 0.0001,
+      "loss": 1.5274,
+      "step": 1655
+    },
+    {
+      "epoch": 0.5545880776959142,
+      "grad_norm": 0.454671710729599,
+      "learning_rate": 0.0001,
+      "loss": 1.5095,
+      "step": 1656
+    },
+    {
+      "epoch": 0.5549229738780977,
+      "grad_norm": 0.46223875880241394,
+      "learning_rate": 0.0001,
+      "loss": 1.4922,
+      "step": 1657
+    },
+    {
+      "epoch": 0.5552578700602813,
+      "grad_norm": 0.4409008324146271,
+      "learning_rate": 0.0001,
+      "loss": 1.5563,
+      "step": 1658
+    },
+    {
+      "epoch": 0.5555927662424648,
+      "grad_norm": 0.4422995150089264,
+      "learning_rate": 0.0001,
+      "loss": 1.4886,
+      "step": 1659
+    },
+    {
+      "epoch": 0.5559276624246483,
+      "grad_norm": 0.4750559329986572,
+      "learning_rate": 0.0001,
+      "loss": 1.5231,
+      "step": 1660
+    },
+    {
+      "epoch": 0.5562625586068319,
+      "grad_norm": 0.46889910101890564,
+      "learning_rate": 0.0001,
+      "loss": 1.5685,
+      "step": 1661
+    },
+    {
+      "epoch": 0.5565974547890155,
+      "grad_norm": 0.46044906973838806,
+      "learning_rate": 0.0001,
+      "loss": 1.5453,
+      "step": 1662
+    },
+    {
+      "epoch": 0.556932350971199,
+      "grad_norm": 0.4508223533630371,
+      "learning_rate": 0.0001,
+      "loss": 1.4371,
+      "step": 1663
+    },
+    {
+      "epoch": 0.5572672471533825,
+      "grad_norm": 0.4626295864582062,
+      "learning_rate": 0.0001,
+      "loss": 1.4826,
+      "step": 1664
+    },
+    {
+      "epoch": 0.557602143335566,
+      "grad_norm": 0.46173542737960815,
+      "learning_rate": 0.0001,
+      "loss": 1.5468,
+      "step": 1665
+    },
+    {
+      "epoch": 0.5579370395177495,
+      "grad_norm": 0.45359376072883606,
+      "learning_rate": 0.0001,
+      "loss": 1.5793,
+      "step": 1666
+    },
+    {
+      "epoch": 0.558271935699933,
+      "grad_norm": 0.4820610582828522,
+      "learning_rate": 0.0001,
+      "loss": 1.5675,
+      "step": 1667
+    },
+    {
+      "epoch": 0.5586068318821166,
+      "grad_norm": 0.44977930188179016,
+      "learning_rate": 0.0001,
+      "loss": 1.452,
+      "step": 1668
+    },
+    {
+      "epoch": 0.5589417280643001,
+      "grad_norm": 0.4835660755634308,
+      "learning_rate": 0.0001,
+      "loss": 1.583,
+      "step": 1669
+    },
+    {
+      "epoch": 0.5592766242464836,
+      "grad_norm": 0.4514458179473877,
+      "learning_rate": 0.0001,
+      "loss": 1.5142,
+      "step": 1670
+    },
+    {
+      "epoch": 0.5596115204286671,
+      "grad_norm": 0.48170676827430725,
+      "learning_rate": 0.0001,
+      "loss": 1.6334,
+      "step": 1671
+    },
+    {
+      "epoch": 0.5599464166108507,
+      "grad_norm": 0.44558587670326233,
+      "learning_rate": 0.0001,
+      "loss": 1.5629,
+      "step": 1672
+    },
+    {
+      "epoch": 0.5602813127930342,
+      "grad_norm": 0.476109117269516,
+      "learning_rate": 0.0001,
+      "loss": 1.5928,
+      "step": 1673
+    },
+    {
+      "epoch": 0.5606162089752177,
+      "grad_norm": 0.4857975244522095,
+      "learning_rate": 0.0001,
+      "loss": 1.5988,
+      "step": 1674
+    },
+    {
+      "epoch": 0.5609511051574012,
+      "grad_norm": 0.49178189039230347,
+      "learning_rate": 0.0001,
+      "loss": 1.6398,
+      "step": 1675
+    },
+    {
+      "epoch": 0.5612860013395847,
+      "grad_norm": 0.45551440119743347,
+      "learning_rate": 0.0001,
+      "loss": 1.473,
+      "step": 1676
+    },
+    {
+      "epoch": 0.5616208975217682,
+      "grad_norm": 0.4543136954307556,
+      "learning_rate": 0.0001,
+      "loss": 1.4731,
+      "step": 1677
+    },
+    {
+      "epoch": 0.5619557937039518,
+      "grad_norm": 0.44175422191619873,
+      "learning_rate": 0.0001,
+      "loss": 1.5275,
+      "step": 1678
+    },
+    {
+      "epoch": 0.5622906898861353,
+      "grad_norm": 0.45219117403030396,
+      "learning_rate": 0.0001,
+      "loss": 1.4795,
+      "step": 1679
+    },
+    {
+      "epoch": 0.5626255860683188,
+      "grad_norm": 0.47150030732154846,
+      "learning_rate": 0.0001,
+      "loss": 1.5716,
+      "step": 1680
+    },
+    {
+      "epoch": 0.5629604822505023,
+      "grad_norm": 0.47483593225479126,
+      "learning_rate": 0.0001,
+      "loss": 1.6408,
+      "step": 1681
+    },
+    {
+      "epoch": 0.5632953784326858,
+      "grad_norm": 0.44375649094581604,
+      "learning_rate": 0.0001,
+      "loss": 1.4819,
+      "step": 1682
+    },
+    {
+      "epoch": 0.5636302746148694,
+      "grad_norm": 0.4467259347438812,
+      "learning_rate": 0.0001,
+      "loss": 1.446,
+      "step": 1683
+    },
+    {
+      "epoch": 0.5639651707970529,
+      "grad_norm": 0.4565868079662323,
+      "learning_rate": 0.0001,
+      "loss": 1.4954,
+      "step": 1684
+    },
+    {
+      "epoch": 0.5643000669792364,
+      "grad_norm": 0.4438111484050751,
+      "learning_rate": 0.0001,
+      "loss": 1.4772,
+      "step": 1685
+    },
+    {
+      "epoch": 0.5646349631614199,
+      "grad_norm": 0.48216649889945984,
+      "learning_rate": 0.0001,
+      "loss": 1.5377,
+      "step": 1686
+    },
+    {
+      "epoch": 0.5649698593436034,
+      "grad_norm": 0.44630300998687744,
+      "learning_rate": 0.0001,
+      "loss": 1.4529,
+      "step": 1687
+    },
+    {
+      "epoch": 0.565304755525787,
+      "grad_norm": 0.44514498114585876,
+      "learning_rate": 0.0001,
+      "loss": 1.466,
+      "step": 1688
+    },
+    {
+      "epoch": 0.5656396517079705,
+      "grad_norm": 0.44218650460243225,
+      "learning_rate": 0.0001,
+      "loss": 1.5396,
+      "step": 1689
+    },
+    {
+      "epoch": 0.565974547890154,
+      "grad_norm": 0.45768415927886963,
+      "learning_rate": 0.0001,
+      "loss": 1.5937,
+      "step": 1690
+    },
+    {
+      "epoch": 0.5663094440723375,
+      "grad_norm": 0.4722628891468048,
+      "learning_rate": 0.0001,
+      "loss": 1.577,
+      "step": 1691
+    },
+    {
+      "epoch": 0.5666443402545212,
+      "grad_norm": 0.45134085416793823,
+      "learning_rate": 0.0001,
+      "loss": 1.5307,
+      "step": 1692
+    },
+    {
+      "epoch": 0.5669792364367047,
+      "grad_norm": 0.48546385765075684,
+      "learning_rate": 0.0001,
+      "loss": 1.5803,
+      "step": 1693
+    },
+    {
+      "epoch": 0.5673141326188882,
+      "grad_norm": 0.4477085769176483,
+      "learning_rate": 0.0001,
+      "loss": 1.5038,
+      "step": 1694
+    },
+    {
+      "epoch": 0.5676490288010717,
+      "grad_norm": 0.45485320687294006,
+      "learning_rate": 0.0001,
+      "loss": 1.5179,
+      "step": 1695
+    },
+    {
+      "epoch": 0.5679839249832552,
+      "grad_norm": 0.4585912227630615,
+      "learning_rate": 0.0001,
+      "loss": 1.4936,
+      "step": 1696
+    },
+    {
+      "epoch": 0.5683188211654387,
+      "grad_norm": 0.453314870595932,
+      "learning_rate": 0.0001,
+      "loss": 1.5132,
+      "step": 1697
+    },
+    {
+      "epoch": 0.5686537173476223,
+      "grad_norm": 0.4477619528770447,
+      "learning_rate": 0.0001,
+      "loss": 1.4687,
+      "step": 1698
+    },
+    {
+      "epoch": 0.5689886135298058,
+      "grad_norm": 0.4389646351337433,
+      "learning_rate": 0.0001,
+      "loss": 1.5626,
+      "step": 1699
+    },
+    {
+      "epoch": 0.5693235097119893,
+      "grad_norm": 0.45276832580566406,
+      "learning_rate": 0.0001,
+      "loss": 1.4914,
+      "step": 1700
+    },
+    {
+      "epoch": 0.5696584058941728,
+      "grad_norm": 0.46535176038742065,
+      "learning_rate": 0.0001,
+      "loss": 1.5462,
+      "step": 1701
+    },
+    {
+      "epoch": 0.5699933020763563,
+      "grad_norm": 0.4585205018520355,
+      "learning_rate": 0.0001,
+      "loss": 1.5687,
+      "step": 1702
+    },
+    {
+      "epoch": 0.5703281982585399,
+      "grad_norm": 0.4479345381259918,
+      "learning_rate": 0.0001,
+      "loss": 1.49,
+      "step": 1703
+    },
+    {
+      "epoch": 0.5706630944407234,
+      "grad_norm": 0.4654242694377899,
+      "learning_rate": 0.0001,
+      "loss": 1.5405,
+      "step": 1704
+    },
+    {
+      "epoch": 0.5709979906229069,
+      "grad_norm": 0.45105135440826416,
+      "learning_rate": 0.0001,
+      "loss": 1.4867,
+      "step": 1705
+    },
+    {
+      "epoch": 0.5713328868050904,
+      "grad_norm": 0.47442853450775146,
+      "learning_rate": 0.0001,
+      "loss": 1.5458,
+      "step": 1706
+    },
+    {
+      "epoch": 0.5716677829872739,
+      "grad_norm": 0.4526834785938263,
+      "learning_rate": 0.0001,
+      "loss": 1.5531,
+      "step": 1707
+    },
+    {
+      "epoch": 0.5720026791694575,
+      "grad_norm": 0.464555561542511,
+      "learning_rate": 0.0001,
+      "loss": 1.5832,
+      "step": 1708
+    },
+    {
+      "epoch": 0.572337575351641,
+      "grad_norm": 0.461290180683136,
+      "learning_rate": 0.0001,
+      "loss": 1.5108,
+      "step": 1709
+    },
+    {
+      "epoch": 0.5726724715338245,
+      "grad_norm": 0.4585516154766083,
+      "learning_rate": 0.0001,
+      "loss": 1.4864,
+      "step": 1710
+    },
+    {
+      "epoch": 0.573007367716008,
+      "grad_norm": 0.4530777335166931,
+      "learning_rate": 0.0001,
+      "loss": 1.4983,
+      "step": 1711
+    },
+    {
+      "epoch": 0.5733422638981915,
+      "grad_norm": 0.4393008053302765,
+      "learning_rate": 0.0001,
+      "loss": 1.471,
+      "step": 1712
+    },
+    {
+      "epoch": 0.5736771600803751,
+      "grad_norm": 0.456870436668396,
+      "learning_rate": 0.0001,
+      "loss": 1.4866,
+      "step": 1713
+    },
+    {
+      "epoch": 0.5740120562625586,
+      "grad_norm": 0.4635424017906189,
+      "learning_rate": 0.0001,
+      "loss": 1.4987,
+      "step": 1714
+    },
+    {
+      "epoch": 0.5743469524447421,
+      "grad_norm": 0.4464293122291565,
+      "learning_rate": 0.0001,
+      "loss": 1.5718,
+      "step": 1715
+    },
+    {
+      "epoch": 0.5746818486269256,
+      "grad_norm": 0.45753639936447144,
+      "learning_rate": 0.0001,
+      "loss": 1.5706,
+      "step": 1716
+    },
+    {
+      "epoch": 0.5750167448091091,
+      "grad_norm": 0.44542786478996277,
+      "learning_rate": 0.0001,
+      "loss": 1.585,
+      "step": 1717
+    },
+    {
+      "epoch": 0.5753516409912927,
+      "grad_norm": 0.4802192151546478,
+      "learning_rate": 0.0001,
+      "loss": 1.6014,
+      "step": 1718
+    },
+    {
+      "epoch": 0.5756865371734762,
+      "grad_norm": 0.4398585259914398,
+      "learning_rate": 0.0001,
+      "loss": 1.4779,
+      "step": 1719
+    },
+    {
+      "epoch": 0.5760214333556597,
+      "grad_norm": 0.4635072350502014,
+      "learning_rate": 0.0001,
+      "loss": 1.5682,
+      "step": 1720
+    },
+    {
+      "epoch": 0.5763563295378432,
+      "grad_norm": 0.45373332500457764,
+      "learning_rate": 0.0001,
+      "loss": 1.5157,
+      "step": 1721
+    },
+    {
+      "epoch": 0.5766912257200268,
+      "grad_norm": 0.46268123388290405,
+      "learning_rate": 0.0001,
+      "loss": 1.6094,
+      "step": 1722
+    },
+    {
+      "epoch": 0.5770261219022104,
+      "grad_norm": 0.4583353102207184,
+      "learning_rate": 0.0001,
+      "loss": 1.5139,
+      "step": 1723
+    },
+    {
+      "epoch": 0.5773610180843939,
+      "grad_norm": 0.4540995955467224,
+      "learning_rate": 0.0001,
+      "loss": 1.5147,
+      "step": 1724
+    },
+    {
+      "epoch": 0.5776959142665774,
+      "grad_norm": 0.44238102436065674,
+      "learning_rate": 0.0001,
+      "loss": 1.5002,
+      "step": 1725
+    },
+    {
+      "epoch": 0.5780308104487609,
+      "grad_norm": 0.4580625891685486,
+      "learning_rate": 0.0001,
+      "loss": 1.5356,
+      "step": 1726
+    },
+    {
+      "epoch": 0.5783657066309444,
+      "grad_norm": 0.4413588047027588,
+      "learning_rate": 0.0001,
+      "loss": 1.5364,
+      "step": 1727
+    },
+    {
+      "epoch": 0.578700602813128,
+      "grad_norm": 0.4560210704803467,
+      "learning_rate": 0.0001,
+      "loss": 1.5458,
+      "step": 1728
+    },
+    {
+      "epoch": 0.5790354989953115,
+      "grad_norm": 0.4587970972061157,
+      "learning_rate": 0.0001,
+      "loss": 1.5131,
+      "step": 1729
+    },
+    {
+      "epoch": 0.579370395177495,
+      "grad_norm": 0.4495079219341278,
+      "learning_rate": 0.0001,
+      "loss": 1.4704,
+      "step": 1730
+    },
+    {
+      "epoch": 0.5797052913596785,
+      "grad_norm": 0.4411208927631378,
+      "learning_rate": 0.0001,
+      "loss": 1.4685,
+      "step": 1731
+    },
+    {
+      "epoch": 0.580040187541862,
+      "grad_norm": 0.4805608093738556,
+      "learning_rate": 0.0001,
+      "loss": 1.5575,
+      "step": 1732
+    },
+    {
+      "epoch": 0.5803750837240456,
+      "grad_norm": 0.4375392198562622,
+      "learning_rate": 0.0001,
+      "loss": 1.4592,
+      "step": 1733
+    },
+    {
+      "epoch": 0.5807099799062291,
+      "grad_norm": 0.46086931228637695,
+      "learning_rate": 0.0001,
+      "loss": 1.5548,
+      "step": 1734
+    },
+    {
+      "epoch": 0.5810448760884126,
+      "grad_norm": 0.4644346535205841,
+      "learning_rate": 0.0001,
+      "loss": 1.499,
+      "step": 1735
+    },
+    {
+      "epoch": 0.5813797722705961,
+      "grad_norm": 0.46922892332077026,
+      "learning_rate": 0.0001,
+      "loss": 1.4893,
+      "step": 1736
+    },
+    {
+      "epoch": 0.5817146684527796,
+      "grad_norm": 0.496118426322937,
+      "learning_rate": 0.0001,
+      "loss": 1.5636,
+      "step": 1737
+    },
+    {
+      "epoch": 0.5820495646349632,
+      "grad_norm": 0.4836883842945099,
+      "learning_rate": 0.0001,
+      "loss": 1.5376,
+      "step": 1738
+    },
+    {
+      "epoch": 0.5823844608171467,
+      "grad_norm": 0.4585513472557068,
+      "learning_rate": 0.0001,
+      "loss": 1.4839,
+      "step": 1739
+    },
+    {
+      "epoch": 0.5827193569993302,
+      "grad_norm": 0.44239330291748047,
+      "learning_rate": 0.0001,
+      "loss": 1.4958,
+      "step": 1740
+    },
+    {
+      "epoch": 0.5830542531815137,
+      "grad_norm": 0.4673005938529968,
+      "learning_rate": 0.0001,
+      "loss": 1.564,
+      "step": 1741
+    },
+    {
+      "epoch": 0.5833891493636972,
+      "grad_norm": 0.47267746925354004,
+      "learning_rate": 0.0001,
+      "loss": 1.5387,
+      "step": 1742
+    },
+    {
+      "epoch": 0.5837240455458808,
+      "grad_norm": 0.4444943964481354,
+      "learning_rate": 0.0001,
+      "loss": 1.5379,
+      "step": 1743
+    },
+    {
+      "epoch": 0.5840589417280643,
+      "grad_norm": 0.46939796209335327,
+      "learning_rate": 0.0001,
+      "loss": 1.56,
+      "step": 1744
+    },
+    {
+      "epoch": 0.5843938379102478,
+      "grad_norm": 0.4565601646900177,
+      "learning_rate": 0.0001,
+      "loss": 1.6041,
+      "step": 1745
+    },
+    {
+      "epoch": 0.5847287340924313,
+      "grad_norm": 0.4543992280960083,
+      "learning_rate": 0.0001,
+      "loss": 1.5446,
+      "step": 1746
+    },
+    {
+      "epoch": 0.5850636302746148,
+      "grad_norm": 0.4492827355861664,
+      "learning_rate": 0.0001,
+      "loss": 1.4757,
+      "step": 1747
+    },
+    {
+      "epoch": 0.5853985264567984,
+      "grad_norm": 0.4561556279659271,
+      "learning_rate": 0.0001,
+      "loss": 1.5098,
+      "step": 1748
+    },
+    {
+      "epoch": 0.5857334226389819,
+      "grad_norm": 0.45513200759887695,
+      "learning_rate": 0.0001,
+      "loss": 1.5315,
+      "step": 1749
+    },
+    {
+      "epoch": 0.5860683188211654,
+      "grad_norm": 0.4617581367492676,
+      "learning_rate": 0.0001,
+      "loss": 1.4389,
+      "step": 1750
+    },
+    {
+      "epoch": 0.5864032150033489,
+      "grad_norm": 0.4772202968597412,
+      "learning_rate": 0.0001,
+      "loss": 1.4551,
+      "step": 1751
+    },
+    {
+      "epoch": 0.5867381111855325,
+      "grad_norm": 0.4441646933555603,
+      "learning_rate": 0.0001,
+      "loss": 1.4341,
+      "step": 1752
+    },
+    {
+      "epoch": 0.5870730073677161,
+      "grad_norm": 0.4554411470890045,
+      "learning_rate": 0.0001,
+      "loss": 1.5049,
+      "step": 1753
+    },
+    {
+      "epoch": 0.5874079035498996,
+      "grad_norm": 0.44894352555274963,
+      "learning_rate": 0.0001,
+      "loss": 1.4413,
+      "step": 1754
+    },
+    {
+      "epoch": 0.5877427997320831,
+      "grad_norm": 0.4450191855430603,
+      "learning_rate": 0.0001,
+      "loss": 1.4164,
+      "step": 1755
+    },
+    {
+      "epoch": 0.5880776959142666,
+      "grad_norm": 0.4631238281726837,
+      "learning_rate": 0.0001,
+      "loss": 1.5182,
+      "step": 1756
+    },
+    {
+      "epoch": 0.5884125920964501,
+      "grad_norm": 0.46651196479797363,
+      "learning_rate": 0.0001,
+      "loss": 1.636,
+      "step": 1757
+    },
+    {
+      "epoch": 0.5887474882786337,
+      "grad_norm": 0.43748751282691956,
+      "learning_rate": 0.0001,
+      "loss": 1.4548,
+      "step": 1758
+    },
+    {
+      "epoch": 0.5890823844608172,
+      "grad_norm": 0.4362008273601532,
+      "learning_rate": 0.0001,
+      "loss": 1.4005,
+      "step": 1759
+    },
+    {
+      "epoch": 0.5894172806430007,
+      "grad_norm": 0.4466918408870697,
+      "learning_rate": 0.0001,
+      "loss": 1.4687,
+      "step": 1760
+    },
+    {
+      "epoch": 0.5897521768251842,
+      "grad_norm": 0.46769535541534424,
+      "learning_rate": 0.0001,
+      "loss": 1.6474,
+      "step": 1761
+    },
+    {
+      "epoch": 0.5900870730073677,
+      "grad_norm": 0.44609808921813965,
+      "learning_rate": 0.0001,
+      "loss": 1.5352,
+      "step": 1762
+    },
+    {
+      "epoch": 0.5904219691895513,
+      "grad_norm": 0.47041457891464233,
+      "learning_rate": 0.0001,
+      "loss": 1.7184,
+      "step": 1763
+    },
+    {
+      "epoch": 0.5907568653717348,
+      "grad_norm": 0.4481906592845917,
+      "learning_rate": 0.0001,
+      "loss": 1.5418,
+      "step": 1764
+    },
+    {
+      "epoch": 0.5910917615539183,
+      "grad_norm": 0.45170557498931885,
+      "learning_rate": 0.0001,
+      "loss": 1.5844,
+      "step": 1765
+    },
+    {
+      "epoch": 0.5914266577361018,
+      "grad_norm": 0.47625941038131714,
+      "learning_rate": 0.0001,
+      "loss": 1.548,
+      "step": 1766
+    },
+    {
+      "epoch": 0.5917615539182853,
+      "grad_norm": 0.4283701181411743,
+      "learning_rate": 0.0001,
+      "loss": 1.5318,
+      "step": 1767
+    },
+    {
+      "epoch": 0.5920964501004689,
+      "grad_norm": 0.4596593677997589,
+      "learning_rate": 0.0001,
+      "loss": 1.5102,
+      "step": 1768
+    },
+    {
+      "epoch": 0.5924313462826524,
+      "grad_norm": 0.45534801483154297,
+      "learning_rate": 0.0001,
+      "loss": 1.5354,
+      "step": 1769
+    },
+    {
+      "epoch": 0.5927662424648359,
+      "grad_norm": 0.4619196355342865,
+      "learning_rate": 0.0001,
+      "loss": 1.5106,
+      "step": 1770
+    },
+    {
+      "epoch": 0.5931011386470194,
+      "grad_norm": 0.47275838255882263,
+      "learning_rate": 0.0001,
+      "loss": 1.5252,
+      "step": 1771
+    },
+    {
+      "epoch": 0.5934360348292029,
+      "grad_norm": 0.4409637451171875,
+      "learning_rate": 0.0001,
+      "loss": 1.4677,
+      "step": 1772
+    },
+    {
+      "epoch": 0.5937709310113864,
+      "grad_norm": 0.4811408519744873,
+      "learning_rate": 0.0001,
+      "loss": 1.6534,
+      "step": 1773
+    },
+    {
+      "epoch": 0.59410582719357,
+      "grad_norm": 0.4599914252758026,
+      "learning_rate": 0.0001,
+      "loss": 1.5346,
+      "step": 1774
+    },
+    {
+      "epoch": 0.5944407233757535,
+      "grad_norm": 0.461795449256897,
+      "learning_rate": 0.0001,
+      "loss": 1.6085,
+      "step": 1775
+    },
+    {
+      "epoch": 0.594775619557937,
+      "grad_norm": 0.48418354988098145,
+      "learning_rate": 0.0001,
+      "loss": 1.6454,
+      "step": 1776
+    },
+    {
+      "epoch": 0.5951105157401205,
+      "grad_norm": 0.4726675748825073,
+      "learning_rate": 0.0001,
+      "loss": 1.6204,
+      "step": 1777
+    },
+    {
+      "epoch": 0.595445411922304,
+      "grad_norm": 0.4598071873188019,
+      "learning_rate": 0.0001,
+      "loss": 1.5689,
+      "step": 1778
+    },
+    {
+      "epoch": 0.5957803081044876,
+      "grad_norm": 0.4411622881889343,
+      "learning_rate": 0.0001,
+      "loss": 1.459,
+      "step": 1779
+    },
+    {
+      "epoch": 0.5961152042866711,
+      "grad_norm": 0.4531992971897125,
+      "learning_rate": 0.0001,
+      "loss": 1.539,
+      "step": 1780
+    },
+    {
+      "epoch": 0.5964501004688546,
+      "grad_norm": 0.4620833694934845,
+      "learning_rate": 0.0001,
+      "loss": 1.5204,
+      "step": 1781
+    },
+    {
+      "epoch": 0.5967849966510381,
+      "grad_norm": 0.44788408279418945,
+      "learning_rate": 0.0001,
+      "loss": 1.4709,
+      "step": 1782
+    },
+    {
+      "epoch": 0.5971198928332218,
+      "grad_norm": 0.4573745131492615,
+      "learning_rate": 0.0001,
+      "loss": 1.6093,
+      "step": 1783
+    },
+    {
+      "epoch": 0.5974547890154053,
+      "grad_norm": 0.45187342166900635,
+      "learning_rate": 0.0001,
+      "loss": 1.4622,
+      "step": 1784
+    },
+    {
+      "epoch": 0.5977896851975888,
+      "grad_norm": 0.4585152268409729,
+      "learning_rate": 0.0001,
+      "loss": 1.533,
+      "step": 1785
+    },
+    {
+      "epoch": 0.5981245813797723,
+      "grad_norm": 0.45203301310539246,
+      "learning_rate": 0.0001,
+      "loss": 1.4723,
+      "step": 1786
+    },
+    {
+      "epoch": 0.5984594775619558,
+      "grad_norm": 0.48862025141716003,
+      "learning_rate": 0.0001,
+      "loss": 1.5242,
+      "step": 1787
+    },
+    {
+      "epoch": 0.5987943737441394,
+      "grad_norm": 0.4569394290447235,
+      "learning_rate": 0.0001,
+      "loss": 1.496,
+      "step": 1788
+    },
+    {
+      "epoch": 0.5991292699263229,
+      "grad_norm": 0.45889320969581604,
+      "learning_rate": 0.0001,
+      "loss": 1.5281,
+      "step": 1789
+    },
+    {
+      "epoch": 0.5994641661085064,
+      "grad_norm": 0.45484670996665955,
+      "learning_rate": 0.0001,
+      "loss": 1.521,
+      "step": 1790
+    },
+    {
+      "epoch": 0.5997990622906899,
+      "grad_norm": 0.4456666111946106,
+      "learning_rate": 0.0001,
+      "loss": 1.4718,
+      "step": 1791
+    },
+    {
+      "epoch": 0.6001339584728734,
+      "grad_norm": 0.45055705308914185,
+      "learning_rate": 0.0001,
+      "loss": 1.4933,
+      "step": 1792
+    },
+    {
+      "epoch": 0.600468854655057,
+      "grad_norm": 0.4577826261520386,
+      "learning_rate": 0.0001,
+      "loss": 1.5425,
+      "step": 1793
+    },
+    {
+      "epoch": 0.6008037508372405,
+      "grad_norm": 0.4573463201522827,
+      "learning_rate": 0.0001,
+      "loss": 1.5396,
+      "step": 1794
+    },
+    {
+      "epoch": 0.601138647019424,
+      "grad_norm": 0.4672832489013672,
+      "learning_rate": 0.0001,
+      "loss": 1.6364,
+      "step": 1795
+    },
+    {
+      "epoch": 0.6014735432016075,
+      "grad_norm": 0.46467325091362,
+      "learning_rate": 0.0001,
+      "loss": 1.5913,
+      "step": 1796
+    },
+    {
+      "epoch": 0.601808439383791,
+      "grad_norm": 0.4452815651893616,
+      "learning_rate": 0.0001,
+      "loss": 1.4723,
+      "step": 1797
+    },
+    {
+      "epoch": 0.6021433355659745,
+      "grad_norm": 0.46799108386039734,
+      "learning_rate": 0.0001,
+      "loss": 1.421,
+      "step": 1798
+    },
+    {
+      "epoch": 0.6024782317481581,
+      "grad_norm": 0.48382145166397095,
+      "learning_rate": 0.0001,
+      "loss": 1.4459,
+      "step": 1799
+    },
+    {
+      "epoch": 0.6028131279303416,
+      "grad_norm": 0.46710142493247986,
+      "learning_rate": 0.0001,
+      "loss": 1.4924,
+      "step": 1800
+    },
+    {
+      "epoch": 0.6031480241125251,
+      "grad_norm": 0.48632174730300903,
+      "learning_rate": 0.0001,
+      "loss": 1.6276,
+      "step": 1801
+    },
+    {
+      "epoch": 0.6034829202947086,
+      "grad_norm": 0.4617525339126587,
+      "learning_rate": 0.0001,
+      "loss": 1.4878,
+      "step": 1802
+    },
+    {
+      "epoch": 0.6038178164768921,
+      "grad_norm": 0.4950006306171417,
+      "learning_rate": 0.0001,
+      "loss": 1.5309,
+      "step": 1803
+    },
+    {
+      "epoch": 0.6041527126590757,
+      "grad_norm": 0.4577718675136566,
+      "learning_rate": 0.0001,
+      "loss": 1.4796,
+      "step": 1804
+    },
+    {
+      "epoch": 0.6044876088412592,
+      "grad_norm": 0.47661927342414856,
+      "learning_rate": 0.0001,
+      "loss": 1.4687,
+      "step": 1805
+    },
+    {
+      "epoch": 0.6048225050234427,
+      "grad_norm": 0.44819697737693787,
+      "learning_rate": 0.0001,
+      "loss": 1.4981,
+      "step": 1806
+    },
+    {
+      "epoch": 0.6051574012056262,
+      "grad_norm": 0.4891256093978882,
+      "learning_rate": 0.0001,
+      "loss": 1.5688,
+      "step": 1807
+    },
+    {
+      "epoch": 0.6054922973878097,
+      "grad_norm": 0.452263742685318,
+      "learning_rate": 0.0001,
+      "loss": 1.5526,
+      "step": 1808
+    },
+    {
+      "epoch": 0.6058271935699933,
+      "grad_norm": 0.4358367621898651,
+      "learning_rate": 0.0001,
+      "loss": 1.5021,
+      "step": 1809
+    },
+    {
+      "epoch": 0.6061620897521768,
+      "grad_norm": 0.4441038966178894,
+      "learning_rate": 0.0001,
+      "loss": 1.5329,
+      "step": 1810
+    },
+    {
+      "epoch": 0.6064969859343603,
+      "grad_norm": 0.4507153034210205,
+      "learning_rate": 0.0001,
+      "loss": 1.4325,
+      "step": 1811
+    },
+    {
+      "epoch": 0.6068318821165438,
+      "grad_norm": 0.45769181847572327,
+      "learning_rate": 0.0001,
+      "loss": 1.5259,
+      "step": 1812
+    },
+    {
+      "epoch": 0.6071667782987274,
+      "grad_norm": 0.4539431631565094,
+      "learning_rate": 0.0001,
+      "loss": 1.5086,
+      "step": 1813
+    },
+    {
+      "epoch": 0.607501674480911,
+      "grad_norm": 0.4860522449016571,
+      "learning_rate": 0.0001,
+      "loss": 1.5724,
+      "step": 1814
+    },
+    {
+      "epoch": 0.6078365706630945,
+      "grad_norm": 0.43865716457366943,
+      "learning_rate": 0.0001,
+      "loss": 1.5729,
+      "step": 1815
+    },
+    {
+      "epoch": 0.608171466845278,
+      "grad_norm": 0.43174469470977783,
+      "learning_rate": 0.0001,
+      "loss": 1.4406,
+      "step": 1816
+    },
+    {
+      "epoch": 0.6085063630274615,
+      "grad_norm": 0.44520848989486694,
+      "learning_rate": 0.0001,
+      "loss": 1.4226,
+      "step": 1817
+    },
+    {
+      "epoch": 0.608841259209645,
+      "grad_norm": 0.46049627661705017,
+      "learning_rate": 0.0001,
+      "loss": 1.5524,
+      "step": 1818
+    },
+    {
+      "epoch": 0.6091761553918286,
+      "grad_norm": 0.4523634910583496,
+      "learning_rate": 0.0001,
+      "loss": 1.5514,
+      "step": 1819
+    },
+    {
+      "epoch": 0.6095110515740121,
+      "grad_norm": 0.44883349537849426,
+      "learning_rate": 0.0001,
+      "loss": 1.515,
+      "step": 1820
+    },
+    {
+      "epoch": 0.6098459477561956,
+      "grad_norm": 0.4615044891834259,
+      "learning_rate": 0.0001,
+      "loss": 1.5721,
+      "step": 1821
+    },
+    {
+      "epoch": 0.6101808439383791,
+      "grad_norm": 0.4608284831047058,
+      "learning_rate": 0.0001,
+      "loss": 1.4685,
+      "step": 1822
+    },
+    {
+      "epoch": 0.6105157401205626,
+      "grad_norm": 0.45228102803230286,
+      "learning_rate": 0.0001,
+      "loss": 1.5517,
+      "step": 1823
+    },
+    {
+      "epoch": 0.6108506363027462,
+      "grad_norm": 0.4694874882698059,
+      "learning_rate": 0.0001,
+      "loss": 1.6284,
+      "step": 1824
+    },
+    {
+      "epoch": 0.6111855324849297,
+      "grad_norm": 0.4595600962638855,
+      "learning_rate": 0.0001,
+      "loss": 1.5509,
+      "step": 1825
+    },
+    {
+      "epoch": 0.6115204286671132,
+      "grad_norm": 0.4391246736049652,
+      "learning_rate": 0.0001,
+      "loss": 1.4557,
+      "step": 1826
+    },
+    {
+      "epoch": 0.6118553248492967,
+      "grad_norm": 0.4490840435028076,
+      "learning_rate": 0.0001,
+      "loss": 1.5605,
+      "step": 1827
+    },
+    {
+      "epoch": 0.6121902210314802,
+      "grad_norm": 0.434456467628479,
+      "learning_rate": 0.0001,
+      "loss": 1.519,
+      "step": 1828
+    },
+    {
+      "epoch": 0.6125251172136638,
+      "grad_norm": 0.4547174274921417,
+      "learning_rate": 0.0001,
+      "loss": 1.3821,
+      "step": 1829
+    },
+    {
+      "epoch": 0.6128600133958473,
+      "grad_norm": 0.46523991227149963,
+      "learning_rate": 0.0001,
+      "loss": 1.5759,
+      "step": 1830
+    },
+    {
+      "epoch": 0.6131949095780308,
+      "grad_norm": 0.4709138870239258,
+      "learning_rate": 0.0001,
+      "loss": 1.5246,
+      "step": 1831
+    },
+    {
+      "epoch": 0.6135298057602143,
+      "grad_norm": 0.45439472794532776,
+      "learning_rate": 0.0001,
+      "loss": 1.5276,
+      "step": 1832
+    },
+    {
+      "epoch": 0.6138647019423978,
+      "grad_norm": 0.4664701521396637,
+      "learning_rate": 0.0001,
+      "loss": 1.5766,
+      "step": 1833
+    },
+    {
+      "epoch": 0.6141995981245814,
+      "grad_norm": 0.44643253087997437,
+      "learning_rate": 0.0001,
+      "loss": 1.4825,
+      "step": 1834
+    },
+    {
+      "epoch": 0.6145344943067649,
+      "grad_norm": 0.45350247621536255,
+      "learning_rate": 0.0001,
+      "loss": 1.5316,
+      "step": 1835
+    },
+    {
+      "epoch": 0.6148693904889484,
+      "grad_norm": 0.5003080368041992,
+      "learning_rate": 0.0001,
+      "loss": 1.5022,
+      "step": 1836
+    },
+    {
+      "epoch": 0.6152042866711319,
+      "grad_norm": 0.4404720366001129,
+      "learning_rate": 0.0001,
+      "loss": 1.5069,
+      "step": 1837
+    },
+    {
+      "epoch": 0.6155391828533154,
+      "grad_norm": 0.4526624083518982,
+      "learning_rate": 0.0001,
+      "loss": 1.5446,
+      "step": 1838
+    },
+    {
+      "epoch": 0.615874079035499,
+      "grad_norm": 0.4543852210044861,
+      "learning_rate": 0.0001,
+      "loss": 1.4988,
+      "step": 1839
+    },
+    {
+      "epoch": 0.6162089752176825,
+      "grad_norm": 0.4725828766822815,
+      "learning_rate": 0.0001,
+      "loss": 1.5402,
+      "step": 1840
+    },
+    {
+      "epoch": 0.616543871399866,
+      "grad_norm": 0.4534866511821747,
+      "learning_rate": 0.0001,
+      "loss": 1.6049,
+      "step": 1841
+    },
+    {
+      "epoch": 0.6168787675820495,
+      "grad_norm": 0.46109819412231445,
+      "learning_rate": 0.0001,
+      "loss": 1.467,
+      "step": 1842
+    },
+    {
+      "epoch": 0.617213663764233,
+      "grad_norm": 0.4313134253025055,
+      "learning_rate": 0.0001,
+      "loss": 1.473,
+      "step": 1843
+    },
+    {
+      "epoch": 0.6175485599464167,
+      "grad_norm": 0.4366667866706848,
+      "learning_rate": 0.0001,
+      "loss": 1.4491,
+      "step": 1844
+    },
+    {
+      "epoch": 0.6178834561286002,
+      "grad_norm": 0.44986891746520996,
+      "learning_rate": 0.0001,
+      "loss": 1.5261,
+      "step": 1845
+    },
+    {
+      "epoch": 0.6182183523107837,
+      "grad_norm": 0.4566178619861603,
+      "learning_rate": 0.0001,
+      "loss": 1.4325,
+      "step": 1846
+    },
+    {
+      "epoch": 0.6185532484929672,
+      "grad_norm": 0.452364444732666,
+      "learning_rate": 0.0001,
+      "loss": 1.4646,
+      "step": 1847
+    },
+    {
+      "epoch": 0.6188881446751507,
+      "grad_norm": 0.5066172480583191,
+      "learning_rate": 0.0001,
+      "loss": 1.5345,
+      "step": 1848
+    },
+    {
+      "epoch": 0.6192230408573343,
+      "grad_norm": 0.48632514476776123,
+      "learning_rate": 0.0001,
+      "loss": 1.5451,
+      "step": 1849
+    },
+    {
+      "epoch": 0.6195579370395178,
+      "grad_norm": 0.49846601486206055,
+      "learning_rate": 0.0001,
+      "loss": 1.6758,
+      "step": 1850
+    },
+    {
+      "epoch": 0.6198928332217013,
+      "grad_norm": 0.4605197012424469,
+      "learning_rate": 0.0001,
+      "loss": 1.4672,
+      "step": 1851
+    },
+    {
+      "epoch": 0.6202277294038848,
+      "grad_norm": 0.5055616497993469,
+      "learning_rate": 0.0001,
+      "loss": 1.5908,
+      "step": 1852
+    },
+    {
+      "epoch": 0.6205626255860683,
+      "grad_norm": 0.43290427327156067,
+      "learning_rate": 0.0001,
+      "loss": 1.4314,
+      "step": 1853
+    },
+    {
+      "epoch": 0.6208975217682519,
+      "grad_norm": 0.4735543131828308,
+      "learning_rate": 0.0001,
+      "loss": 1.5928,
+      "step": 1854
+    },
+    {
+      "epoch": 0.6212324179504354,
+      "grad_norm": 0.45503896474838257,
+      "learning_rate": 0.0001,
+      "loss": 1.4838,
+      "step": 1855
+    },
+    {
+      "epoch": 0.6215673141326189,
+      "grad_norm": 0.47979220747947693,
+      "learning_rate": 0.0001,
+      "loss": 1.5964,
+      "step": 1856
+    },
+    {
+      "epoch": 0.6219022103148024,
+      "grad_norm": 0.4965604245662689,
+      "learning_rate": 0.0001,
+      "loss": 1.6738,
+      "step": 1857
+    },
+    {
+      "epoch": 0.6222371064969859,
+      "grad_norm": 0.46204668283462524,
+      "learning_rate": 0.0001,
+      "loss": 1.5834,
+      "step": 1858
+    },
+    {
+      "epoch": 0.6225720026791695,
+      "grad_norm": 0.4727908968925476,
+      "learning_rate": 0.0001,
+      "loss": 1.5032,
+      "step": 1859
+    },
+    {
+      "epoch": 0.622906898861353,
+      "grad_norm": 0.4483385980129242,
+      "learning_rate": 0.0001,
+      "loss": 1.5106,
+      "step": 1860
+    },
+    {
+      "epoch": 0.6232417950435365,
+      "grad_norm": 0.4823533594608307,
+      "learning_rate": 0.0001,
+      "loss": 1.5182,
+      "step": 1861
+    },
+    {
+      "epoch": 0.62357669122572,
+      "grad_norm": 0.444034606218338,
+      "learning_rate": 0.0001,
+      "loss": 1.4752,
+      "step": 1862
+    },
+    {
+      "epoch": 0.6239115874079035,
+      "grad_norm": 0.4599262773990631,
+      "learning_rate": 0.0001,
+      "loss": 1.6058,
+      "step": 1863
+    },
+    {
+      "epoch": 0.624246483590087,
+      "grad_norm": 0.46915575861930847,
+      "learning_rate": 0.0001,
+      "loss": 1.5703,
+      "step": 1864
+    },
+    {
+      "epoch": 0.6245813797722706,
+      "grad_norm": 0.4495287835597992,
+      "learning_rate": 0.0001,
+      "loss": 1.5697,
+      "step": 1865
+    },
+    {
+      "epoch": 0.6249162759544541,
+      "grad_norm": 0.4550332725048065,
+      "learning_rate": 0.0001,
+      "loss": 1.593,
+      "step": 1866
+    },
+    {
+      "epoch": 0.6252511721366376,
+      "grad_norm": 0.45989227294921875,
+      "learning_rate": 0.0001,
+      "loss": 1.5264,
+      "step": 1867
+    },
+    {
+      "epoch": 0.6255860683188211,
+      "grad_norm": 0.4636523723602295,
+      "learning_rate": 0.0001,
+      "loss": 1.5075,
+      "step": 1868
+    },
+    {
+      "epoch": 0.6259209645010047,
+      "grad_norm": 0.47894105315208435,
+      "learning_rate": 0.0001,
+      "loss": 1.5339,
+      "step": 1869
+    },
+    {
+      "epoch": 0.6262558606831882,
+      "grad_norm": 0.48035570979118347,
+      "learning_rate": 0.0001,
+      "loss": 1.6396,
+      "step": 1870
+    },
+    {
+      "epoch": 0.6265907568653717,
+      "grad_norm": 0.4636676013469696,
+      "learning_rate": 0.0001,
+      "loss": 1.5379,
+      "step": 1871
+    },
+    {
+      "epoch": 0.6269256530475552,
+      "grad_norm": 0.45790281891822815,
+      "learning_rate": 0.0001,
+      "loss": 1.5592,
+      "step": 1872
+    },
+    {
+      "epoch": 0.6272605492297387,
+      "grad_norm": 0.4582197964191437,
+      "learning_rate": 0.0001,
+      "loss": 1.4942,
+      "step": 1873
+    },
+    {
+      "epoch": 0.6275954454119224,
+      "grad_norm": 0.46599966287612915,
+      "learning_rate": 0.0001,
+      "loss": 1.4827,
+      "step": 1874
+    },
+    {
+      "epoch": 0.6279303415941059,
+      "grad_norm": 0.46609312295913696,
+      "learning_rate": 0.0001,
+      "loss": 1.4927,
+      "step": 1875
+    },
+    {
+      "epoch": 0.6282652377762894,
+      "grad_norm": 0.4750870168209076,
+      "learning_rate": 0.0001,
+      "loss": 1.5363,
+      "step": 1876
+    },
+    {
+      "epoch": 0.6286001339584729,
+      "grad_norm": 0.450873464345932,
+      "learning_rate": 0.0001,
+      "loss": 1.4469,
+      "step": 1877
+    },
+    {
+      "epoch": 0.6289350301406564,
+      "grad_norm": 0.4762636423110962,
+      "learning_rate": 0.0001,
+      "loss": 1.5899,
+      "step": 1878
+    },
+    {
+      "epoch": 0.62926992632284,
+      "grad_norm": 0.4646572172641754,
+      "learning_rate": 0.0001,
+      "loss": 1.5064,
+      "step": 1879
+    },
+    {
+      "epoch": 0.6296048225050235,
+      "grad_norm": 0.45741555094718933,
+      "learning_rate": 0.0001,
+      "loss": 1.4961,
+      "step": 1880
+    },
+    {
+      "epoch": 0.629939718687207,
+      "grad_norm": 0.466714471578598,
+      "learning_rate": 0.0001,
+      "loss": 1.544,
+      "step": 1881
+    },
+    {
+      "epoch": 0.6302746148693905,
+      "grad_norm": 0.4649389684200287,
+      "learning_rate": 0.0001,
+      "loss": 1.5034,
+      "step": 1882
+    },
+    {
+      "epoch": 0.630609511051574,
+      "grad_norm": 0.4543673098087311,
+      "learning_rate": 0.0001,
+      "loss": 1.5047,
+      "step": 1883
+    },
+    {
+      "epoch": 0.6309444072337576,
+      "grad_norm": 0.4476313889026642,
+      "learning_rate": 0.0001,
+      "loss": 1.5352,
+      "step": 1884
+    },
+    {
+      "epoch": 0.6312793034159411,
+      "grad_norm": 0.45701703429222107,
+      "learning_rate": 0.0001,
+      "loss": 1.4728,
+      "step": 1885
+    },
+    {
+      "epoch": 0.6316141995981246,
+      "grad_norm": 0.47941985726356506,
+      "learning_rate": 0.0001,
+      "loss": 1.55,
+      "step": 1886
+    },
+    {
+      "epoch": 0.6319490957803081,
+      "grad_norm": 0.4696800112724304,
+      "learning_rate": 0.0001,
+      "loss": 1.5414,
+      "step": 1887
+    },
+    {
+      "epoch": 0.6322839919624916,
+      "grad_norm": 0.46519196033477783,
+      "learning_rate": 0.0001,
+      "loss": 1.535,
+      "step": 1888
+    },
+    {
+      "epoch": 0.6326188881446752,
+      "grad_norm": 0.48627668619155884,
+      "learning_rate": 0.0001,
+      "loss": 1.5944,
+      "step": 1889
+    },
+    {
+      "epoch": 0.6329537843268587,
+      "grad_norm": 0.48074227571487427,
+      "learning_rate": 0.0001,
+      "loss": 1.6139,
+      "step": 1890
+    },
+    {
+      "epoch": 0.6332886805090422,
+      "grad_norm": 0.46736252307891846,
+      "learning_rate": 0.0001,
+      "loss": 1.4678,
+      "step": 1891
+    },
+    {
+      "epoch": 0.6336235766912257,
+      "grad_norm": 0.4569910168647766,
+      "learning_rate": 0.0001,
+      "loss": 1.4618,
+      "step": 1892
+    },
+    {
+      "epoch": 0.6339584728734092,
+      "grad_norm": 0.47562307119369507,
+      "learning_rate": 0.0001,
+      "loss": 1.5823,
+      "step": 1893
+    },
+    {
+      "epoch": 0.6342933690555927,
+      "grad_norm": 0.466452956199646,
+      "learning_rate": 0.0001,
+      "loss": 1.5113,
+      "step": 1894
+    },
+    {
+      "epoch": 0.6346282652377763,
+      "grad_norm": 0.46311044692993164,
+      "learning_rate": 0.0001,
+      "loss": 1.5501,
+      "step": 1895
+    },
+    {
+      "epoch": 0.6349631614199598,
+      "grad_norm": 0.4702926278114319,
+      "learning_rate": 0.0001,
+      "loss": 1.5268,
+      "step": 1896
+    },
+    {
+      "epoch": 0.6352980576021433,
+      "grad_norm": 0.45443615317344666,
+      "learning_rate": 0.0001,
+      "loss": 1.4828,
+      "step": 1897
+    },
+    {
+      "epoch": 0.6356329537843268,
+      "grad_norm": 0.4715770184993744,
+      "learning_rate": 0.0001,
+      "loss": 1.4838,
+      "step": 1898
+    },
+    {
+      "epoch": 0.6359678499665103,
+      "grad_norm": 0.4504530429840088,
+      "learning_rate": 0.0001,
+      "loss": 1.5517,
+      "step": 1899
+    },
+    {
+      "epoch": 0.6363027461486939,
+      "grad_norm": 0.44113561511039734,
+      "learning_rate": 0.0001,
+      "loss": 1.5354,
+      "step": 1900
+    },
+    {
+      "epoch": 0.6366376423308774,
+      "grad_norm": 0.4682255685329437,
+      "learning_rate": 0.0001,
+      "loss": 1.5009,
+      "step": 1901
+    },
+    {
+      "epoch": 0.6369725385130609,
+      "grad_norm": 0.4606624245643616,
+      "learning_rate": 0.0001,
+      "loss": 1.5275,
+      "step": 1902
+    },
+    {
+      "epoch": 0.6373074346952444,
+      "grad_norm": 0.4960009455680847,
+      "learning_rate": 0.0001,
+      "loss": 1.6387,
+      "step": 1903
+    },
+    {
+      "epoch": 0.6376423308774279,
+      "grad_norm": 0.4498940706253052,
+      "learning_rate": 0.0001,
+      "loss": 1.3875,
+      "step": 1904
+    },
+    {
+      "epoch": 0.6379772270596116,
+      "grad_norm": 0.4744248688220978,
+      "learning_rate": 0.0001,
+      "loss": 1.5715,
+      "step": 1905
+    },
+    {
+      "epoch": 0.6383121232417951,
+      "grad_norm": 0.4565998911857605,
+      "learning_rate": 0.0001,
+      "loss": 1.6134,
+      "step": 1906
+    },
+    {
+      "epoch": 0.6386470194239786,
+      "grad_norm": 0.4789954423904419,
+      "learning_rate": 0.0001,
+      "loss": 1.5371,
+      "step": 1907
+    },
+    {
+      "epoch": 0.6389819156061621,
+      "grad_norm": 0.4668721556663513,
+      "learning_rate": 0.0001,
+      "loss": 1.6295,
+      "step": 1908
+    },
+    {
+      "epoch": 0.6393168117883457,
+      "grad_norm": 0.4449285566806793,
+      "learning_rate": 0.0001,
+      "loss": 1.4532,
+      "step": 1909
+    },
+    {
+      "epoch": 0.6396517079705292,
+      "grad_norm": 0.46385762095451355,
+      "learning_rate": 0.0001,
+      "loss": 1.5756,
+      "step": 1910
+    },
+    {
+      "epoch": 0.6399866041527127,
+      "grad_norm": 0.4670940339565277,
+      "learning_rate": 0.0001,
+      "loss": 1.5893,
+      "step": 1911
+    },
+    {
+      "epoch": 0.6403215003348962,
+      "grad_norm": 0.4583105742931366,
+      "learning_rate": 0.0001,
+      "loss": 1.4533,
+      "step": 1912
+    },
+    {
+      "epoch": 0.6406563965170797,
+      "grad_norm": 0.43470409512519836,
+      "learning_rate": 0.0001,
+      "loss": 1.417,
+      "step": 1913
+    },
+    {
+      "epoch": 0.6409912926992632,
+      "grad_norm": 0.4683259427547455,
+      "learning_rate": 0.0001,
+      "loss": 1.5425,
+      "step": 1914
+    },
+    {
+      "epoch": 0.6413261888814468,
+      "grad_norm": 0.43709784746170044,
+      "learning_rate": 0.0001,
+      "loss": 1.498,
+      "step": 1915
+    },
+    {
+      "epoch": 0.6416610850636303,
+      "grad_norm": 0.4760400950908661,
+      "learning_rate": 0.0001,
+      "loss": 1.5995,
+      "step": 1916
+    },
+    {
+      "epoch": 0.6419959812458138,
+      "grad_norm": 0.47921547293663025,
+      "learning_rate": 0.0001,
+      "loss": 1.5829,
+      "step": 1917
+    },
+    {
+      "epoch": 0.6423308774279973,
+      "grad_norm": 0.4762752056121826,
+      "learning_rate": 0.0001,
+      "loss": 1.6675,
+      "step": 1918
+    },
+    {
+      "epoch": 0.6426657736101808,
+      "grad_norm": 0.46576380729675293,
+      "learning_rate": 0.0001,
+      "loss": 1.5385,
+      "step": 1919
+    },
+    {
+      "epoch": 0.6430006697923644,
+      "grad_norm": 0.4593994915485382,
+      "learning_rate": 0.0001,
+      "loss": 1.4759,
+      "step": 1920
+    },
+    {
+      "epoch": 0.6433355659745479,
+      "grad_norm": 0.45551782846450806,
+      "learning_rate": 0.0001,
+      "loss": 1.4443,
+      "step": 1921
+    },
+    {
+      "epoch": 0.6436704621567314,
+      "grad_norm": 0.45817625522613525,
+      "learning_rate": 0.0001,
+      "loss": 1.4591,
+      "step": 1922
+    },
+    {
+      "epoch": 0.6440053583389149,
+      "grad_norm": 0.4861002564430237,
+      "learning_rate": 0.0001,
+      "loss": 1.5437,
+      "step": 1923
+    },
+    {
+      "epoch": 0.6443402545210984,
+      "grad_norm": 0.4582726061344147,
+      "learning_rate": 0.0001,
+      "loss": 1.4162,
+      "step": 1924
+    },
+    {
+      "epoch": 0.644675150703282,
+      "grad_norm": 0.45267635583877563,
+      "learning_rate": 0.0001,
+      "loss": 1.4411,
+      "step": 1925
+    },
+    {
+      "epoch": 0.6450100468854655,
+      "grad_norm": 0.4521283805370331,
+      "learning_rate": 0.0001,
+      "loss": 1.5681,
+      "step": 1926
+    },
+    {
+      "epoch": 0.645344943067649,
+      "grad_norm": 0.4693661332130432,
+      "learning_rate": 0.0001,
+      "loss": 1.5153,
+      "step": 1927
+    },
+    {
+      "epoch": 0.6456798392498325,
+      "grad_norm": 0.4437626302242279,
+      "learning_rate": 0.0001,
+      "loss": 1.3758,
+      "step": 1928
+    },
+    {
+      "epoch": 0.646014735432016,
+      "grad_norm": 0.4534982740879059,
+      "learning_rate": 0.0001,
+      "loss": 1.5104,
+      "step": 1929
+    },
+    {
+      "epoch": 0.6463496316141996,
+      "grad_norm": 0.4561862051486969,
+      "learning_rate": 0.0001,
+      "loss": 1.5212,
+      "step": 1930
+    },
+    {
+      "epoch": 0.6466845277963831,
+      "grad_norm": 0.46854645013809204,
+      "learning_rate": 0.0001,
+      "loss": 1.495,
+      "step": 1931
+    },
+    {
+      "epoch": 0.6470194239785666,
+      "grad_norm": 0.4689598083496094,
+      "learning_rate": 0.0001,
+      "loss": 1.6433,
+      "step": 1932
+    },
+    {
+      "epoch": 0.6473543201607501,
+      "grad_norm": 0.474661260843277,
+      "learning_rate": 0.0001,
+      "loss": 1.4658,
+      "step": 1933
+    },
+    {
+      "epoch": 0.6476892163429336,
+      "grad_norm": 0.48324620723724365,
+      "learning_rate": 0.0001,
+      "loss": 1.5707,
+      "step": 1934
+    },
+    {
+      "epoch": 0.6480241125251173,
+      "grad_norm": 0.4637855291366577,
+      "learning_rate": 0.0001,
+      "loss": 1.5954,
+      "step": 1935
+    },
+    {
+      "epoch": 0.6483590087073008,
+      "grad_norm": 0.4889835715293884,
+      "learning_rate": 0.0001,
+      "loss": 1.4605,
+      "step": 1936
+    },
+    {
+      "epoch": 0.6486939048894843,
+      "grad_norm": 0.4517784118652344,
+      "learning_rate": 0.0001,
+      "loss": 1.4816,
+      "step": 1937
+    },
+    {
+      "epoch": 0.6490288010716678,
+      "grad_norm": 0.47071799635887146,
+      "learning_rate": 0.0001,
+      "loss": 1.5544,
+      "step": 1938
+    },
+    {
+      "epoch": 0.6493636972538513,
+      "grad_norm": 0.45107269287109375,
+      "learning_rate": 0.0001,
+      "loss": 1.5698,
+      "step": 1939
+    },
+    {
+      "epoch": 0.6496985934360349,
+      "grad_norm": 0.47279268503189087,
+      "learning_rate": 0.0001,
+      "loss": 1.6169,
+      "step": 1940
+    },
+    {
+      "epoch": 0.6500334896182184,
+      "grad_norm": 0.4569358825683594,
+      "learning_rate": 0.0001,
+      "loss": 1.5735,
+      "step": 1941
+    },
+    {
+      "epoch": 0.6503683858004019,
+      "grad_norm": 0.49081119894981384,
+      "learning_rate": 0.0001,
+      "loss": 1.5683,
+      "step": 1942
+    },
+    {
+      "epoch": 0.6507032819825854,
+      "grad_norm": 0.45732182264328003,
+      "learning_rate": 0.0001,
+      "loss": 1.5146,
+      "step": 1943
+    },
+    {
+      "epoch": 0.6510381781647689,
+      "grad_norm": 0.48255446553230286,
+      "learning_rate": 0.0001,
+      "loss": 1.5982,
+      "step": 1944
+    },
+    {
+      "epoch": 0.6513730743469525,
+      "grad_norm": 0.4548420310020447,
+      "learning_rate": 0.0001,
+      "loss": 1.5714,
+      "step": 1945
+    },
+    {
+      "epoch": 0.651707970529136,
+      "grad_norm": 0.46165576577186584,
+      "learning_rate": 0.0001,
+      "loss": 1.6128,
+      "step": 1946
+    },
+    {
+      "epoch": 0.6520428667113195,
+      "grad_norm": 0.4421144425868988,
+      "learning_rate": 0.0001,
+      "loss": 1.4837,
+      "step": 1947
+    },
+    {
+      "epoch": 0.652377762893503,
+      "grad_norm": 0.4581396281719208,
+      "learning_rate": 0.0001,
+      "loss": 1.4513,
+      "step": 1948
+    },
+    {
+      "epoch": 0.6527126590756865,
+      "grad_norm": 0.45781993865966797,
+      "learning_rate": 0.0001,
+      "loss": 1.4742,
+      "step": 1949
+    },
+    {
+      "epoch": 0.6530475552578701,
+      "grad_norm": 0.4650152325630188,
+      "learning_rate": 0.0001,
+      "loss": 1.5343,
+      "step": 1950
+    },
+    {
+      "epoch": 0.6533824514400536,
+      "grad_norm": 0.47808071970939636,
+      "learning_rate": 0.0001,
+      "loss": 1.5066,
+      "step": 1951
+    },
+    {
+      "epoch": 0.6537173476222371,
+      "grad_norm": 0.4686528146266937,
+      "learning_rate": 0.0001,
+      "loss": 1.5331,
+      "step": 1952
+    },
+    {
+      "epoch": 0.6540522438044206,
+      "grad_norm": 0.497200071811676,
+      "learning_rate": 0.0001,
+      "loss": 1.591,
+      "step": 1953
+    },
+    {
+      "epoch": 0.6543871399866041,
+      "grad_norm": 0.4465658664703369,
+      "learning_rate": 0.0001,
+      "loss": 1.5553,
+      "step": 1954
+    },
+    {
+      "epoch": 0.6547220361687877,
+      "grad_norm": 0.44191426038742065,
+      "learning_rate": 0.0001,
+      "loss": 1.4675,
+      "step": 1955
+    },
+    {
+      "epoch": 0.6550569323509712,
+      "grad_norm": 0.4407011568546295,
+      "learning_rate": 0.0001,
+      "loss": 1.4161,
+      "step": 1956
+    },
+    {
+      "epoch": 0.6553918285331547,
+      "grad_norm": 0.4675586521625519,
+      "learning_rate": 0.0001,
+      "loss": 1.476,
+      "step": 1957
+    },
+    {
+      "epoch": 0.6557267247153382,
+      "grad_norm": 0.46932560205459595,
+      "learning_rate": 0.0001,
+      "loss": 1.5244,
+      "step": 1958
+    },
+    {
+      "epoch": 0.6560616208975217,
+      "grad_norm": 0.4337617754936218,
+      "learning_rate": 0.0001,
+      "loss": 1.5247,
+      "step": 1959
+    },
+    {
+      "epoch": 0.6563965170797053,
+      "grad_norm": 0.48785457015037537,
+      "learning_rate": 0.0001,
+      "loss": 1.5425,
+      "step": 1960
+    },
+    {
+      "epoch": 0.6567314132618888,
+      "grad_norm": 0.4481433033943176,
+      "learning_rate": 0.0001,
+      "loss": 1.4779,
+      "step": 1961
+    },
+    {
+      "epoch": 0.6570663094440723,
+      "grad_norm": 0.4887458384037018,
+      "learning_rate": 0.0001,
+      "loss": 1.4865,
+      "step": 1962
+    },
+    {
+      "epoch": 0.6574012056262558,
+      "grad_norm": 0.4612896740436554,
+      "learning_rate": 0.0001,
+      "loss": 1.5068,
+      "step": 1963
+    },
+    {
+      "epoch": 0.6577361018084393,
+      "grad_norm": 0.4455910623073578,
+      "learning_rate": 0.0001,
+      "loss": 1.4341,
+      "step": 1964
+    },
+    {
+      "epoch": 0.6580709979906229,
+      "grad_norm": 0.4743984043598175,
+      "learning_rate": 0.0001,
+      "loss": 1.5445,
+      "step": 1965
+    },
+    {
+      "epoch": 0.6584058941728065,
+      "grad_norm": 0.46120843291282654,
+      "learning_rate": 0.0001,
+      "loss": 1.4902,
+      "step": 1966
+    },
+    {
+      "epoch": 0.65874079035499,
+      "grad_norm": 0.4710855782032013,
+      "learning_rate": 0.0001,
+      "loss": 1.5625,
+      "step": 1967
+    },
+    {
+      "epoch": 0.6590756865371735,
+      "grad_norm": 0.44773563742637634,
+      "learning_rate": 0.0001,
+      "loss": 1.5546,
+      "step": 1968
+    },
+    {
+      "epoch": 0.659410582719357,
+      "grad_norm": 0.4674702286720276,
+      "learning_rate": 0.0001,
+      "loss": 1.5137,
+      "step": 1969
+    },
+    {
+      "epoch": 0.6597454789015406,
+      "grad_norm": 0.4413098096847534,
+      "learning_rate": 0.0001,
+      "loss": 1.4792,
+      "step": 1970
+    },
+    {
+      "epoch": 0.6600803750837241,
+      "grad_norm": 0.4673274755477905,
+      "learning_rate": 0.0001,
+      "loss": 1.5341,
+      "step": 1971
+    },
+    {
+      "epoch": 0.6604152712659076,
+      "grad_norm": 0.46262720227241516,
+      "learning_rate": 0.0001,
+      "loss": 1.4812,
+      "step": 1972
+    },
+    {
+      "epoch": 0.6607501674480911,
+      "grad_norm": 0.4727884829044342,
+      "learning_rate": 0.0001,
+      "loss": 1.6146,
+      "step": 1973
+    },
+    {
+      "epoch": 0.6610850636302746,
+      "grad_norm": 0.4771195650100708,
+      "learning_rate": 0.0001,
+      "loss": 1.5431,
+      "step": 1974
+    },
+    {
+      "epoch": 0.6614199598124582,
+      "grad_norm": 0.48989588022232056,
+      "learning_rate": 0.0001,
+      "loss": 1.5296,
+      "step": 1975
+    },
+    {
+      "epoch": 0.6617548559946417,
+      "grad_norm": 0.4625195264816284,
+      "learning_rate": 0.0001,
+      "loss": 1.4064,
+      "step": 1976
+    },
+    {
+      "epoch": 0.6620897521768252,
+      "grad_norm": 0.48750412464141846,
+      "learning_rate": 0.0001,
+      "loss": 1.4338,
+      "step": 1977
+    },
+    {
+      "epoch": 0.6624246483590087,
+      "grad_norm": 0.47945865988731384,
+      "learning_rate": 0.0001,
+      "loss": 1.5737,
+      "step": 1978
+    },
+    {
+      "epoch": 0.6627595445411922,
+      "grad_norm": 0.4766711890697479,
+      "learning_rate": 0.0001,
+      "loss": 1.5785,
+      "step": 1979
+    },
+    {
+      "epoch": 0.6630944407233758,
+      "grad_norm": 0.46893224120140076,
+      "learning_rate": 0.0001,
+      "loss": 1.5609,
+      "step": 1980
+    },
+    {
+      "epoch": 0.6634293369055593,
+      "grad_norm": 0.4679139256477356,
+      "learning_rate": 0.0001,
+      "loss": 1.5144,
+      "step": 1981
+    },
+    {
+      "epoch": 0.6637642330877428,
+      "grad_norm": 0.451107919216156,
+      "learning_rate": 0.0001,
+      "loss": 1.5252,
+      "step": 1982
+    },
+    {
+      "epoch": 0.6640991292699263,
+      "grad_norm": 0.4496656358242035,
+      "learning_rate": 0.0001,
+      "loss": 1.3826,
+      "step": 1983
+    },
+    {
+      "epoch": 0.6644340254521098,
+      "grad_norm": 0.46544507145881653,
+      "learning_rate": 0.0001,
+      "loss": 1.5085,
+      "step": 1984
+    },
+    {
+      "epoch": 0.6647689216342934,
+      "grad_norm": 0.4583112597465515,
+      "learning_rate": 0.0001,
+      "loss": 1.5213,
+      "step": 1985
+    },
+    {
+      "epoch": 0.6651038178164769,
+      "grad_norm": 0.4679916799068451,
+      "learning_rate": 0.0001,
+      "loss": 1.5077,
+      "step": 1986
+    },
+    {
+      "epoch": 0.6654387139986604,
+      "grad_norm": 0.4852266311645508,
+      "learning_rate": 0.0001,
+      "loss": 1.5449,
+      "step": 1987
+    },
+    {
+      "epoch": 0.6657736101808439,
+      "grad_norm": 0.46231263875961304,
+      "learning_rate": 0.0001,
+      "loss": 1.5846,
+      "step": 1988
+    },
+    {
+      "epoch": 0.6661085063630274,
+      "grad_norm": 0.4823322892189026,
+      "learning_rate": 0.0001,
+      "loss": 1.5246,
+      "step": 1989
+    },
+    {
+      "epoch": 0.666443402545211,
+      "grad_norm": 0.45604240894317627,
+      "learning_rate": 0.0001,
+      "loss": 1.4723,
+      "step": 1990
+    },
+    {
+      "epoch": 0.6667782987273945,
+      "grad_norm": 0.47065314650535583,
+      "learning_rate": 0.0001,
+      "loss": 1.4831,
+      "step": 1991
+    },
+    {
+      "epoch": 0.667113194909578,
+      "grad_norm": 0.46221935749053955,
+      "learning_rate": 0.0001,
+      "loss": 1.4796,
+      "step": 1992
+    },
+    {
+      "epoch": 0.6674480910917615,
+      "grad_norm": 0.475827157497406,
+      "learning_rate": 0.0001,
+      "loss": 1.5351,
+      "step": 1993
+    },
+    {
+      "epoch": 0.667782987273945,
+      "grad_norm": 0.45660120248794556,
+      "learning_rate": 0.0001,
+      "loss": 1.5737,
+      "step": 1994
+    },
+    {
+      "epoch": 0.6681178834561285,
+      "grad_norm": 0.4607517123222351,
+      "learning_rate": 0.0001,
+      "loss": 1.5358,
+      "step": 1995
+    },
+    {
+      "epoch": 0.6684527796383122,
+      "grad_norm": 0.4408629536628723,
+      "learning_rate": 0.0001,
+      "loss": 1.5249,
+      "step": 1996
+    },
+    {
+      "epoch": 0.6687876758204957,
+      "grad_norm": 0.44046109914779663,
+      "learning_rate": 0.0001,
+      "loss": 1.4979,
+      "step": 1997
+    },
+    {
+      "epoch": 0.6691225720026792,
+      "grad_norm": 0.45813706517219543,
+      "learning_rate": 0.0001,
+      "loss": 1.4541,
+      "step": 1998
+    },
+    {
+      "epoch": 0.6694574681848627,
+      "grad_norm": 0.4600253105163574,
+      "learning_rate": 0.0001,
+      "loss": 1.4795,
+      "step": 1999
+    },
+    {
+      "epoch": 0.6697923643670463,
+      "grad_norm": 0.4730280935764313,
+      "learning_rate": 0.0001,
+      "loss": 1.5925,
+      "step": 2000
+    },
+    {
+      "epoch": 0.6701272605492298,
+      "grad_norm": 0.4675295054912567,
+      "learning_rate": 0.0001,
+      "loss": 1.5452,
+      "step": 2001
+    },
+    {
+      "epoch": 0.6704621567314133,
+      "grad_norm": 0.4651346206665039,
+      "learning_rate": 0.0001,
+      "loss": 1.5475,
+      "step": 2002
+    },
+    {
+      "epoch": 0.6707970529135968,
+      "grad_norm": 0.4588277041912079,
+      "learning_rate": 0.0001,
+      "loss": 1.5906,
+      "step": 2003
+    },
+    {
+      "epoch": 0.6711319490957803,
+      "grad_norm": 0.4812864363193512,
+      "learning_rate": 0.0001,
+      "loss": 1.5754,
+      "step": 2004
+    },
+    {
+      "epoch": 0.6714668452779639,
+      "grad_norm": 0.45574623346328735,
+      "learning_rate": 0.0001,
+      "loss": 1.4828,
+      "step": 2005
+    },
+    {
+      "epoch": 0.6718017414601474,
+      "grad_norm": 0.46802258491516113,
+      "learning_rate": 0.0001,
+      "loss": 1.5108,
+      "step": 2006
+    },
+    {
+      "epoch": 0.6721366376423309,
+      "grad_norm": 0.47356149554252625,
+      "learning_rate": 0.0001,
+      "loss": 1.577,
+      "step": 2007
+    },
+    {
+      "epoch": 0.6724715338245144,
+      "grad_norm": 0.46529653668403625,
+      "learning_rate": 0.0001,
+      "loss": 1.478,
+      "step": 2008
+    },
+    {
+      "epoch": 0.6728064300066979,
+      "grad_norm": 0.4784877896308899,
+      "learning_rate": 0.0001,
+      "loss": 1.5498,
+      "step": 2009
+    },
+    {
+      "epoch": 0.6731413261888815,
+      "grad_norm": 0.47873613238334656,
+      "learning_rate": 0.0001,
+      "loss": 1.5988,
+      "step": 2010
+    },
+    {
+      "epoch": 0.673476222371065,
+      "grad_norm": 0.4695528447628021,
+      "learning_rate": 0.0001,
+      "loss": 1.5282,
+      "step": 2011
+    },
+    {
+      "epoch": 0.6738111185532485,
+      "grad_norm": 0.44687068462371826,
+      "learning_rate": 0.0001,
+      "loss": 1.5051,
+      "step": 2012
+    },
+    {
+      "epoch": 0.674146014735432,
+      "grad_norm": 0.4465655982494354,
+      "learning_rate": 0.0001,
+      "loss": 1.5711,
+      "step": 2013
+    },
+    {
+      "epoch": 0.6744809109176155,
+      "grad_norm": 0.5006896257400513,
+      "learning_rate": 0.0001,
+      "loss": 1.5201,
+      "step": 2014
+    },
+    {
+      "epoch": 0.674815807099799,
+      "grad_norm": 0.4736153185367584,
+      "learning_rate": 0.0001,
+      "loss": 1.5697,
+      "step": 2015
+    },
+    {
+      "epoch": 0.6751507032819826,
+      "grad_norm": 0.4669327735900879,
+      "learning_rate": 0.0001,
+      "loss": 1.5168,
+      "step": 2016
+    },
+    {
+      "epoch": 0.6754855994641661,
+      "grad_norm": 0.44387710094451904,
+      "learning_rate": 0.0001,
+      "loss": 1.4377,
+      "step": 2017
+    },
+    {
+      "epoch": 0.6758204956463496,
+      "grad_norm": 0.4410373568534851,
+      "learning_rate": 0.0001,
+      "loss": 1.4047,
+      "step": 2018
+    },
+    {
+      "epoch": 0.6761553918285331,
+      "grad_norm": 0.45553281903266907,
+      "learning_rate": 0.0001,
+      "loss": 1.5127,
+      "step": 2019
+    },
+    {
+      "epoch": 0.6764902880107166,
+      "grad_norm": 0.4672766327857971,
+      "learning_rate": 0.0001,
+      "loss": 1.5614,
+      "step": 2020
+    },
+    {
+      "epoch": 0.6768251841929002,
+      "grad_norm": 0.46312415599823,
+      "learning_rate": 0.0001,
+      "loss": 1.4514,
+      "step": 2021
+    },
+    {
+      "epoch": 0.6771600803750837,
+      "grad_norm": 0.4592866599559784,
+      "learning_rate": 0.0001,
+      "loss": 1.47,
+      "step": 2022
+    },
+    {
+      "epoch": 0.6774949765572672,
+      "grad_norm": 0.46353161334991455,
+      "learning_rate": 0.0001,
+      "loss": 1.5574,
+      "step": 2023
+    },
+    {
+      "epoch": 0.6778298727394507,
+      "grad_norm": 0.45665234327316284,
+      "learning_rate": 0.0001,
+      "loss": 1.5377,
+      "step": 2024
+    },
+    {
+      "epoch": 0.6781647689216342,
+      "grad_norm": 0.47426679730415344,
+      "learning_rate": 0.0001,
+      "loss": 1.5885,
+      "step": 2025
+    },
+    {
+      "epoch": 0.6784996651038178,
+      "grad_norm": 0.464711457490921,
+      "learning_rate": 0.0001,
+      "loss": 1.5344,
+      "step": 2026
+    },
+    {
+      "epoch": 0.6788345612860014,
+      "grad_norm": 0.4656915068626404,
+      "learning_rate": 0.0001,
+      "loss": 1.5718,
+      "step": 2027
+    },
+    {
+      "epoch": 0.6791694574681849,
+      "grad_norm": 0.48867180943489075,
+      "learning_rate": 0.0001,
+      "loss": 1.5578,
+      "step": 2028
+    },
+    {
+      "epoch": 0.6795043536503684,
+      "grad_norm": 0.43709492683410645,
+      "learning_rate": 0.0001,
+      "loss": 1.4615,
+      "step": 2029
+    },
+    {
+      "epoch": 0.679839249832552,
+      "grad_norm": 0.4825074374675751,
+      "learning_rate": 0.0001,
+      "loss": 1.5181,
+      "step": 2030
+    },
+    {
+      "epoch": 0.6801741460147355,
+      "grad_norm": 0.45096608996391296,
+      "learning_rate": 0.0001,
+      "loss": 1.6186,
+      "step": 2031
+    },
+    {
+      "epoch": 0.680509042196919,
+      "grad_norm": 0.449068158864975,
+      "learning_rate": 0.0001,
+      "loss": 1.4584,
+      "step": 2032
+    },
+    {
+      "epoch": 0.6808439383791025,
+      "grad_norm": 0.46091821789741516,
+      "learning_rate": 0.0001,
+      "loss": 1.5739,
+      "step": 2033
+    },
+    {
+      "epoch": 0.681178834561286,
+      "grad_norm": 0.4668620824813843,
+      "learning_rate": 0.0001,
+      "loss": 1.6237,
+      "step": 2034
+    },
+    {
+      "epoch": 0.6815137307434695,
+      "grad_norm": 0.4587905704975128,
+      "learning_rate": 0.0001,
+      "loss": 1.5535,
+      "step": 2035
+    },
+    {
+      "epoch": 0.6818486269256531,
+      "grad_norm": 0.4671448767185211,
+      "learning_rate": 0.0001,
+      "loss": 1.5589,
+      "step": 2036
+    },
+    {
+      "epoch": 0.6821835231078366,
+      "grad_norm": 0.45884931087493896,
+      "learning_rate": 0.0001,
+      "loss": 1.4523,
+      "step": 2037
+    },
+    {
+      "epoch": 0.6825184192900201,
+      "grad_norm": 0.46959635615348816,
+      "learning_rate": 0.0001,
+      "loss": 1.5474,
+      "step": 2038
+    },
+    {
+      "epoch": 0.6828533154722036,
+      "grad_norm": 0.45907384157180786,
+      "learning_rate": 0.0001,
+      "loss": 1.4985,
+      "step": 2039
+    },
+    {
+      "epoch": 0.6831882116543871,
+      "grad_norm": 0.4683312177658081,
+      "learning_rate": 0.0001,
+      "loss": 1.5353,
+      "step": 2040
+    },
+    {
+      "epoch": 0.6835231078365707,
+      "grad_norm": 0.4766319990158081,
+      "learning_rate": 0.0001,
+      "loss": 1.4801,
+      "step": 2041
+    },
+    {
+      "epoch": 0.6838580040187542,
+      "grad_norm": 0.4715334475040436,
+      "learning_rate": 0.0001,
+      "loss": 1.4394,
+      "step": 2042
+    },
+    {
+      "epoch": 0.6841929002009377,
+      "grad_norm": 0.4581170678138733,
+      "learning_rate": 0.0001,
+      "loss": 1.5153,
+      "step": 2043
+    },
+    {
+      "epoch": 0.6845277963831212,
+      "grad_norm": 0.469703733921051,
+      "learning_rate": 0.0001,
+      "loss": 1.5181,
+      "step": 2044
+    },
+    {
+      "epoch": 0.6848626925653047,
+      "grad_norm": 0.45137694478034973,
+      "learning_rate": 0.0001,
+      "loss": 1.4877,
+      "step": 2045
+    },
+    {
+      "epoch": 0.6851975887474883,
+      "grad_norm": 0.4811345636844635,
+      "learning_rate": 0.0001,
+      "loss": 1.6566,
+      "step": 2046
+    },
+    {
+      "epoch": 0.6855324849296718,
+      "grad_norm": 0.4660329222679138,
+      "learning_rate": 0.0001,
+      "loss": 1.5094,
+      "step": 2047
+    },
+    {
+      "epoch": 0.6858673811118553,
+      "grad_norm": 0.45194128155708313,
+      "learning_rate": 0.0001,
+      "loss": 1.5157,
+      "step": 2048
+    },
+    {
+      "epoch": 0.6862022772940388,
+      "grad_norm": 0.45705869793891907,
+      "learning_rate": 0.0001,
+      "loss": 1.5073,
+      "step": 2049
+    },
+    {
+      "epoch": 0.6865371734762223,
+      "grad_norm": 0.4675770103931427,
+      "learning_rate": 0.0001,
+      "loss": 1.5067,
+      "step": 2050
+    },
+    {
+      "epoch": 0.6868720696584059,
+      "grad_norm": 0.45620042085647583,
+      "learning_rate": 0.0001,
+      "loss": 1.5607,
+      "step": 2051
+    },
+    {
+      "epoch": 0.6872069658405894,
+      "grad_norm": 0.5035691261291504,
+      "learning_rate": 0.0001,
+      "loss": 1.5534,
+      "step": 2052
+    },
+    {
+      "epoch": 0.6875418620227729,
+      "grad_norm": 0.4658897817134857,
+      "learning_rate": 0.0001,
+      "loss": 1.5647,
+      "step": 2053
+    },
+    {
+      "epoch": 0.6878767582049564,
+      "grad_norm": 0.4606130123138428,
+      "learning_rate": 0.0001,
+      "loss": 1.5304,
+      "step": 2054
+    },
+    {
+      "epoch": 0.6882116543871399,
+      "grad_norm": 0.4701388478279114,
+      "learning_rate": 0.0001,
+      "loss": 1.506,
+      "step": 2055
+    },
+    {
+      "epoch": 0.6885465505693235,
+      "grad_norm": 0.4724370837211609,
+      "learning_rate": 0.0001,
+      "loss": 1.5547,
+      "step": 2056
+    },
+    {
+      "epoch": 0.6888814467515071,
+      "grad_norm": 0.44684672355651855,
+      "learning_rate": 0.0001,
+      "loss": 1.4687,
+      "step": 2057
+    },
+    {
+      "epoch": 0.6892163429336906,
+      "grad_norm": 0.5058313608169556,
+      "learning_rate": 0.0001,
+      "loss": 1.5251,
+      "step": 2058
+    },
+    {
+      "epoch": 0.6895512391158741,
+      "grad_norm": 0.4328201711177826,
+      "learning_rate": 0.0001,
+      "loss": 1.4719,
+      "step": 2059
+    },
+    {
+      "epoch": 0.6898861352980576,
+      "grad_norm": 0.4703109860420227,
+      "learning_rate": 0.0001,
+      "loss": 1.5043,
+      "step": 2060
+    },
+    {
+      "epoch": 0.6902210314802412,
+      "grad_norm": 0.46181461215019226,
+      "learning_rate": 0.0001,
+      "loss": 1.4884,
+      "step": 2061
+    },
+    {
+      "epoch": 0.6905559276624247,
+      "grad_norm": 0.4495771527290344,
+      "learning_rate": 0.0001,
+      "loss": 1.4585,
+      "step": 2062
+    },
+    {
+      "epoch": 0.6908908238446082,
+      "grad_norm": 0.5008009672164917,
+      "learning_rate": 0.0001,
+      "loss": 1.5687,
+      "step": 2063
+    },
+    {
+      "epoch": 0.6912257200267917,
+      "grad_norm": 0.4631938636302948,
+      "learning_rate": 0.0001,
+      "loss": 1.5049,
+      "step": 2064
+    },
+    {
+      "epoch": 0.6915606162089752,
+      "grad_norm": 0.5008888244628906,
+      "learning_rate": 0.0001,
+      "loss": 1.5,
+      "step": 2065
+    },
+    {
+      "epoch": 0.6918955123911588,
+      "grad_norm": 0.4508407413959503,
+      "learning_rate": 0.0001,
+      "loss": 1.4246,
+      "step": 2066
+    },
+    {
+      "epoch": 0.6922304085733423,
+      "grad_norm": 0.48627397418022156,
+      "learning_rate": 0.0001,
+      "loss": 1.4769,
+      "step": 2067
+    },
+    {
+      "epoch": 0.6925653047555258,
+      "grad_norm": 0.49048787355422974,
+      "learning_rate": 0.0001,
+      "loss": 1.5519,
+      "step": 2068
+    },
+    {
+      "epoch": 0.6929002009377093,
+      "grad_norm": 0.4490218758583069,
+      "learning_rate": 0.0001,
+      "loss": 1.4597,
+      "step": 2069
+    },
+    {
+      "epoch": 0.6932350971198928,
+      "grad_norm": 0.4886975884437561,
+      "learning_rate": 0.0001,
+      "loss": 1.5541,
+      "step": 2070
+    },
+    {
+      "epoch": 0.6935699933020764,
+      "grad_norm": 0.4652385115623474,
+      "learning_rate": 0.0001,
+      "loss": 1.5426,
+      "step": 2071
+    },
+    {
+      "epoch": 0.6939048894842599,
+      "grad_norm": 0.48018738627433777,
+      "learning_rate": 0.0001,
+      "loss": 1.556,
+      "step": 2072
+    },
+    {
+      "epoch": 0.6942397856664434,
+      "grad_norm": 0.45230361819267273,
+      "learning_rate": 0.0001,
+      "loss": 1.5265,
+      "step": 2073
+    },
+    {
+      "epoch": 0.6945746818486269,
+      "grad_norm": 0.5099328756332397,
+      "learning_rate": 0.0001,
+      "loss": 1.5578,
+      "step": 2074
+    },
+    {
+      "epoch": 0.6949095780308104,
+      "grad_norm": 0.45329004526138306,
+      "learning_rate": 0.0001,
+      "loss": 1.4306,
+      "step": 2075
+    },
+    {
+      "epoch": 0.695244474212994,
+      "grad_norm": 0.4592442810535431,
+      "learning_rate": 0.0001,
+      "loss": 1.5326,
+      "step": 2076
+    },
+    {
+      "epoch": 0.6955793703951775,
+      "grad_norm": 0.46230995655059814,
+      "learning_rate": 0.0001,
+      "loss": 1.603,
+      "step": 2077
+    },
+    {
+      "epoch": 0.695914266577361,
+      "grad_norm": 0.47789496183395386,
+      "learning_rate": 0.0001,
+      "loss": 1.5768,
+      "step": 2078
+    },
+    {
+      "epoch": 0.6962491627595445,
+      "grad_norm": 0.46773889660835266,
+      "learning_rate": 0.0001,
+      "loss": 1.5329,
+      "step": 2079
+    },
+    {
+      "epoch": 0.696584058941728,
+      "grad_norm": 0.46148133277893066,
+      "learning_rate": 0.0001,
+      "loss": 1.5334,
+      "step": 2080
+    },
+    {
+      "epoch": 0.6969189551239116,
+      "grad_norm": 0.46110233664512634,
+      "learning_rate": 0.0001,
+      "loss": 1.5567,
+      "step": 2081
+    },
+    {
+      "epoch": 0.6972538513060951,
+      "grad_norm": 0.4659191370010376,
+      "learning_rate": 0.0001,
+      "loss": 1.5402,
+      "step": 2082
+    },
+    {
+      "epoch": 0.6975887474882786,
+      "grad_norm": 0.45607033371925354,
+      "learning_rate": 0.0001,
+      "loss": 1.4262,
+      "step": 2083
+    },
+    {
+      "epoch": 0.6979236436704621,
+      "grad_norm": 0.4767032563686371,
+      "learning_rate": 0.0001,
+      "loss": 1.5247,
+      "step": 2084
+    },
+    {
+      "epoch": 0.6982585398526456,
+      "grad_norm": 0.4534342288970947,
+      "learning_rate": 0.0001,
+      "loss": 1.4731,
+      "step": 2085
+    },
+    {
+      "epoch": 0.6985934360348292,
+      "grad_norm": 0.46111854910850525,
+      "learning_rate": 0.0001,
+      "loss": 1.614,
+      "step": 2086
+    },
+    {
+      "epoch": 0.6989283322170127,
+      "grad_norm": 0.4589134156703949,
+      "learning_rate": 0.0001,
+      "loss": 1.5565,
+      "step": 2087
+    },
+    {
+      "epoch": 0.6992632283991963,
+      "grad_norm": 0.4603373110294342,
+      "learning_rate": 0.0001,
+      "loss": 1.4874,
+      "step": 2088
+    },
+    {
+      "epoch": 0.6995981245813798,
+      "grad_norm": 0.466277539730072,
+      "learning_rate": 0.0001,
+      "loss": 1.5705,
+      "step": 2089
+    },
+    {
+      "epoch": 0.6999330207635633,
+      "grad_norm": 0.4790922701358795,
+      "learning_rate": 0.0001,
+      "loss": 1.4459,
+      "step": 2090
+    },
+    {
+      "epoch": 0.7002679169457469,
+      "grad_norm": 0.47170254588127136,
+      "learning_rate": 0.0001,
+      "loss": 1.5706,
+      "step": 2091
+    },
+    {
+      "epoch": 0.7006028131279304,
+      "grad_norm": 0.4699598252773285,
+      "learning_rate": 0.0001,
+      "loss": 1.5987,
+      "step": 2092
+    },
+    {
+      "epoch": 0.7009377093101139,
+      "grad_norm": 0.46038177609443665,
+      "learning_rate": 0.0001,
+      "loss": 1.5629,
+      "step": 2093
+    },
+    {
+      "epoch": 0.7012726054922974,
+      "grad_norm": 0.4709053635597229,
+      "learning_rate": 0.0001,
+      "loss": 1.5994,
+      "step": 2094
+    },
+    {
+      "epoch": 0.7016075016744809,
+      "grad_norm": 0.4605868458747864,
+      "learning_rate": 0.0001,
+      "loss": 1.5267,
+      "step": 2095
+    },
+    {
+      "epoch": 0.7019423978566645,
+      "grad_norm": 0.47352883219718933,
+      "learning_rate": 0.0001,
+      "loss": 1.5299,
+      "step": 2096
+    },
+    {
+      "epoch": 0.702277294038848,
+      "grad_norm": 0.4487178921699524,
+      "learning_rate": 0.0001,
+      "loss": 1.4862,
+      "step": 2097
+    },
+    {
+      "epoch": 0.7026121902210315,
+      "grad_norm": 0.4435498118400574,
+      "learning_rate": 0.0001,
+      "loss": 1.434,
+      "step": 2098
+    },
+    {
+      "epoch": 0.702947086403215,
+      "grad_norm": 0.4770961403846741,
+      "learning_rate": 0.0001,
+      "loss": 1.5706,
+      "step": 2099
+    },
+    {
+      "epoch": 0.7032819825853985,
+      "grad_norm": 0.5055303573608398,
+      "learning_rate": 0.0001,
+      "loss": 1.6134,
+      "step": 2100
+    },
+    {
+      "epoch": 0.703616878767582,
+      "grad_norm": 0.46639499068260193,
+      "learning_rate": 0.0001,
+      "loss": 1.5706,
+      "step": 2101
+    },
+    {
+      "epoch": 0.7039517749497656,
+      "grad_norm": 0.47837772965431213,
+      "learning_rate": 0.0001,
+      "loss": 1.5261,
+      "step": 2102
+    },
+    {
+      "epoch": 0.7042866711319491,
+      "grad_norm": 0.4420396089553833,
+      "learning_rate": 0.0001,
+      "loss": 1.554,
+      "step": 2103
+    },
+    {
+      "epoch": 0.7046215673141326,
+      "grad_norm": 0.4579044282436371,
+      "learning_rate": 0.0001,
+      "loss": 1.496,
+      "step": 2104
+    },
+    {
+      "epoch": 0.7049564634963161,
+      "grad_norm": 0.4740173816680908,
+      "learning_rate": 0.0001,
+      "loss": 1.6011,
+      "step": 2105
+    },
+    {
+      "epoch": 0.7052913596784997,
+      "grad_norm": 0.46156203746795654,
+      "learning_rate": 0.0001,
+      "loss": 1.4938,
+      "step": 2106
+    },
+    {
+      "epoch": 0.7056262558606832,
+      "grad_norm": 0.47833749651908875,
+      "learning_rate": 0.0001,
+      "loss": 1.4794,
+      "step": 2107
+    },
+    {
+      "epoch": 0.7059611520428667,
+      "grad_norm": 0.45706647634506226,
+      "learning_rate": 0.0001,
+      "loss": 1.4822,
+      "step": 2108
+    },
+    {
+      "epoch": 0.7062960482250502,
+      "grad_norm": 0.4594823122024536,
+      "learning_rate": 0.0001,
+      "loss": 1.4496,
+      "step": 2109
+    },
+    {
+      "epoch": 0.7066309444072337,
+      "grad_norm": 0.4742507040500641,
+      "learning_rate": 0.0001,
+      "loss": 1.4566,
+      "step": 2110
+    },
+    {
+      "epoch": 0.7069658405894172,
+      "grad_norm": 0.46597233414649963,
+      "learning_rate": 0.0001,
+      "loss": 1.54,
+      "step": 2111
+    },
+    {
+      "epoch": 0.7073007367716008,
+      "grad_norm": 0.4715932011604309,
+      "learning_rate": 0.0001,
+      "loss": 1.5444,
+      "step": 2112
+    },
+    {
+      "epoch": 0.7076356329537843,
+      "grad_norm": 0.45990297198295593,
+      "learning_rate": 0.0001,
+      "loss": 1.4125,
+      "step": 2113
+    },
+    {
+      "epoch": 0.7079705291359678,
+      "grad_norm": 0.4889219105243683,
+      "learning_rate": 0.0001,
+      "loss": 1.5741,
+      "step": 2114
+    },
+    {
+      "epoch": 0.7083054253181513,
+      "grad_norm": 0.46439915895462036,
+      "learning_rate": 0.0001,
+      "loss": 1.5618,
+      "step": 2115
+    },
+    {
+      "epoch": 0.7086403215003348,
+      "grad_norm": 0.4761582612991333,
+      "learning_rate": 0.0001,
+      "loss": 1.5833,
+      "step": 2116
+    },
+    {
+      "epoch": 0.7089752176825184,
+      "grad_norm": 0.46475279331207275,
+      "learning_rate": 0.0001,
+      "loss": 1.5725,
+      "step": 2117
+    },
+    {
+      "epoch": 0.709310113864702,
+      "grad_norm": 0.48302775621414185,
+      "learning_rate": 0.0001,
+      "loss": 1.5543,
+      "step": 2118
+    },
+    {
+      "epoch": 0.7096450100468855,
+      "grad_norm": 0.45327621698379517,
+      "learning_rate": 0.0001,
+      "loss": 1.5391,
+      "step": 2119
+    },
+    {
+      "epoch": 0.709979906229069,
+      "grad_norm": 0.4502553939819336,
+      "learning_rate": 0.0001,
+      "loss": 1.4403,
+      "step": 2120
+    },
+    {
+      "epoch": 0.7103148024112526,
+      "grad_norm": 0.47124046087265015,
+      "learning_rate": 0.0001,
+      "loss": 1.5433,
+      "step": 2121
+    },
+    {
+      "epoch": 0.7106496985934361,
+      "grad_norm": 0.4871177077293396,
+      "learning_rate": 0.0001,
+      "loss": 1.5333,
+      "step": 2122
+    },
+    {
+      "epoch": 0.7109845947756196,
+      "grad_norm": 0.45931074023246765,
+      "learning_rate": 0.0001,
+      "loss": 1.4978,
+      "step": 2123
+    },
+    {
+      "epoch": 0.7113194909578031,
+      "grad_norm": 0.46881625056266785,
+      "learning_rate": 0.0001,
+      "loss": 1.518,
+      "step": 2124
+    },
+    {
+      "epoch": 0.7116543871399866,
+      "grad_norm": 0.47189566493034363,
+      "learning_rate": 0.0001,
+      "loss": 1.5414,
+      "step": 2125
+    },
+    {
+      "epoch": 0.7119892833221702,
+      "grad_norm": 0.49166449904441833,
+      "learning_rate": 0.0001,
+      "loss": 1.5914,
+      "step": 2126
+    },
+    {
+      "epoch": 0.7123241795043537,
+      "grad_norm": 0.4626203775405884,
+      "learning_rate": 0.0001,
+      "loss": 1.5008,
+      "step": 2127
+    },
+    {
+      "epoch": 0.7126590756865372,
+      "grad_norm": 0.5054082274436951,
+      "learning_rate": 0.0001,
+      "loss": 1.5388,
+      "step": 2128
+    },
+    {
+      "epoch": 0.7129939718687207,
+      "grad_norm": 0.46761465072631836,
+      "learning_rate": 0.0001,
+      "loss": 1.6087,
+      "step": 2129
+    },
+    {
+      "epoch": 0.7133288680509042,
+      "grad_norm": 0.46342405676841736,
+      "learning_rate": 0.0001,
+      "loss": 1.5171,
+      "step": 2130
+    },
+    {
+      "epoch": 0.7136637642330877,
+      "grad_norm": 0.43117156624794006,
+      "learning_rate": 0.0001,
+      "loss": 1.5104,
+      "step": 2131
+    },
+    {
+      "epoch": 0.7139986604152713,
+      "grad_norm": 0.4568924307823181,
+      "learning_rate": 0.0001,
+      "loss": 1.4607,
+      "step": 2132
+    },
+    {
+      "epoch": 0.7143335565974548,
+      "grad_norm": 0.47409340739250183,
+      "learning_rate": 0.0001,
+      "loss": 1.5482,
+      "step": 2133
+    },
+    {
+      "epoch": 0.7146684527796383,
+      "grad_norm": 0.44349172711372375,
+      "learning_rate": 0.0001,
+      "loss": 1.5406,
+      "step": 2134
+    },
+    {
+      "epoch": 0.7150033489618218,
+      "grad_norm": 0.4581291973590851,
+      "learning_rate": 0.0001,
+      "loss": 1.5583,
+      "step": 2135
+    },
+    {
+      "epoch": 0.7153382451440053,
+      "grad_norm": 0.46366629004478455,
+      "learning_rate": 0.0001,
+      "loss": 1.4944,
+      "step": 2136
+    },
+    {
+      "epoch": 0.7156731413261889,
+      "grad_norm": 0.4668786823749542,
+      "learning_rate": 0.0001,
+      "loss": 1.4843,
+      "step": 2137
+    },
+    {
+      "epoch": 0.7160080375083724,
+      "grad_norm": 0.473763644695282,
+      "learning_rate": 0.0001,
+      "loss": 1.5267,
+      "step": 2138
+    },
+    {
+      "epoch": 0.7163429336905559,
+      "grad_norm": 0.4776982069015503,
+      "learning_rate": 0.0001,
+      "loss": 1.4722,
+      "step": 2139
+    },
+    {
+      "epoch": 0.7166778298727394,
+      "grad_norm": 0.4626188576221466,
+      "learning_rate": 0.0001,
+      "loss": 1.5093,
+      "step": 2140
+    },
+    {
+      "epoch": 0.7170127260549229,
+      "grad_norm": 0.4587743580341339,
+      "learning_rate": 0.0001,
+      "loss": 1.4735,
+      "step": 2141
+    },
+    {
+      "epoch": 0.7173476222371065,
+      "grad_norm": 0.48433372378349304,
+      "learning_rate": 0.0001,
+      "loss": 1.5957,
+      "step": 2142
+    },
+    {
+      "epoch": 0.71768251841929,
+      "grad_norm": 0.46150970458984375,
+      "learning_rate": 0.0001,
+      "loss": 1.5504,
+      "step": 2143
+    },
+    {
+      "epoch": 0.7180174146014735,
+      "grad_norm": 0.46284884214401245,
+      "learning_rate": 0.0001,
+      "loss": 1.4929,
+      "step": 2144
+    },
+    {
+      "epoch": 0.718352310783657,
+      "grad_norm": 0.46241495013237,
+      "learning_rate": 0.0001,
+      "loss": 1.5442,
+      "step": 2145
+    },
+    {
+      "epoch": 0.7186872069658405,
+      "grad_norm": 0.46660250425338745,
+      "learning_rate": 0.0001,
+      "loss": 1.593,
+      "step": 2146
+    },
+    {
+      "epoch": 0.7190221031480241,
+      "grad_norm": 0.45663195848464966,
+      "learning_rate": 0.0001,
+      "loss": 1.5503,
+      "step": 2147
+    },
+    {
+      "epoch": 0.7193569993302076,
+      "grad_norm": 0.43789803981781006,
+      "learning_rate": 0.0001,
+      "loss": 1.3862,
+      "step": 2148
+    },
+    {
+      "epoch": 0.7196918955123912,
+      "grad_norm": 0.5114313364028931,
+      "learning_rate": 0.0001,
+      "loss": 1.6686,
+      "step": 2149
+    },
+    {
+      "epoch": 0.7200267916945747,
+      "grad_norm": 0.4642086625099182,
+      "learning_rate": 0.0001,
+      "loss": 1.5078,
+      "step": 2150
+    },
+    {
+      "epoch": 0.7203616878767582,
+      "grad_norm": 0.48922207951545715,
+      "learning_rate": 0.0001,
+      "loss": 1.4943,
+      "step": 2151
+    },
+    {
+      "epoch": 0.7206965840589418,
+      "grad_norm": 0.4588397145271301,
+      "learning_rate": 0.0001,
+      "loss": 1.5204,
+      "step": 2152
+    },
+    {
+      "epoch": 0.7210314802411253,
+      "grad_norm": 0.4893192648887634,
+      "learning_rate": 0.0001,
+      "loss": 1.4953,
+      "step": 2153
+    },
+    {
+      "epoch": 0.7213663764233088,
+      "grad_norm": 0.4586998224258423,
+      "learning_rate": 0.0001,
+      "loss": 1.5461,
+      "step": 2154
+    },
+    {
+      "epoch": 0.7217012726054923,
+      "grad_norm": 0.4510471522808075,
+      "learning_rate": 0.0001,
+      "loss": 1.4525,
+      "step": 2155
+    },
+    {
+      "epoch": 0.7220361687876758,
+      "grad_norm": 0.45169365406036377,
+      "learning_rate": 0.0001,
+      "loss": 1.4289,
+      "step": 2156
+    },
+    {
+      "epoch": 0.7223710649698594,
+      "grad_norm": 0.46228212118148804,
+      "learning_rate": 0.0001,
+      "loss": 1.4531,
+      "step": 2157
+    },
+    {
+      "epoch": 0.7227059611520429,
+      "grad_norm": 0.4868308901786804,
+      "learning_rate": 0.0001,
+      "loss": 1.5175,
+      "step": 2158
+    },
+    {
+      "epoch": 0.7230408573342264,
+      "grad_norm": 0.4654078483581543,
+      "learning_rate": 0.0001,
+      "loss": 1.5008,
+      "step": 2159
+    },
+    {
+      "epoch": 0.7233757535164099,
+      "grad_norm": 0.4278988838195801,
+      "learning_rate": 0.0001,
+      "loss": 1.4565,
+      "step": 2160
+    },
+    {
+      "epoch": 0.7237106496985934,
+      "grad_norm": 0.46262645721435547,
+      "learning_rate": 0.0001,
+      "loss": 1.6055,
+      "step": 2161
+    },
+    {
+      "epoch": 0.724045545880777,
+      "grad_norm": 0.44615933299064636,
+      "learning_rate": 0.0001,
+      "loss": 1.4298,
+      "step": 2162
+    },
+    {
+      "epoch": 0.7243804420629605,
+      "grad_norm": 0.4570693373680115,
+      "learning_rate": 0.0001,
+      "loss": 1.5527,
+      "step": 2163
+    },
+    {
+      "epoch": 0.724715338245144,
+      "grad_norm": 0.4507233202457428,
+      "learning_rate": 0.0001,
+      "loss": 1.4555,
+      "step": 2164
+    },
+    {
+      "epoch": 0.7250502344273275,
+      "grad_norm": 0.48417288064956665,
+      "learning_rate": 0.0001,
+      "loss": 1.6364,
+      "step": 2165
+    },
+    {
+      "epoch": 0.725385130609511,
+      "grad_norm": 0.45867475867271423,
+      "learning_rate": 0.0001,
+      "loss": 1.5087,
+      "step": 2166
+    },
+    {
+      "epoch": 0.7257200267916946,
+      "grad_norm": 0.47540226578712463,
+      "learning_rate": 0.0001,
+      "loss": 1.657,
+      "step": 2167
+    },
+    {
+      "epoch": 0.7260549229738781,
+      "grad_norm": 0.48101478815078735,
+      "learning_rate": 0.0001,
+      "loss": 1.534,
+      "step": 2168
+    },
+    {
+      "epoch": 0.7263898191560616,
+      "grad_norm": 0.47601044178009033,
+      "learning_rate": 0.0001,
+      "loss": 1.6616,
+      "step": 2169
+    },
+    {
+      "epoch": 0.7267247153382451,
+      "grad_norm": 0.45919114351272583,
+      "learning_rate": 0.0001,
+      "loss": 1.582,
+      "step": 2170
+    },
+    {
+      "epoch": 0.7270596115204286,
+      "grad_norm": 0.4518474042415619,
+      "learning_rate": 0.0001,
+      "loss": 1.3507,
+      "step": 2171
+    },
+    {
+      "epoch": 0.7273945077026122,
+      "grad_norm": 0.4650135636329651,
+      "learning_rate": 0.0001,
+      "loss": 1.445,
+      "step": 2172
+    },
+    {
+      "epoch": 0.7277294038847957,
+      "grad_norm": 0.45777347683906555,
+      "learning_rate": 0.0001,
+      "loss": 1.5559,
+      "step": 2173
+    },
+    {
+      "epoch": 0.7280643000669792,
+      "grad_norm": 0.4657188653945923,
+      "learning_rate": 0.0001,
+      "loss": 1.5193,
+      "step": 2174
+    },
+    {
+      "epoch": 0.7283991962491627,
+      "grad_norm": 0.4442221224308014,
+      "learning_rate": 0.0001,
+      "loss": 1.4896,
+      "step": 2175
+    },
+    {
+      "epoch": 0.7287340924313462,
+      "grad_norm": 0.4653560519218445,
+      "learning_rate": 0.0001,
+      "loss": 1.4669,
+      "step": 2176
+    },
+    {
+      "epoch": 0.7290689886135298,
+      "grad_norm": 0.4594500660896301,
+      "learning_rate": 0.0001,
+      "loss": 1.5145,
+      "step": 2177
+    },
+    {
+      "epoch": 0.7294038847957133,
+      "grad_norm": 0.4747893214225769,
+      "learning_rate": 0.0001,
+      "loss": 1.6847,
+      "step": 2178
+    },
+    {
+      "epoch": 0.7297387809778969,
+      "grad_norm": 0.47377869486808777,
+      "learning_rate": 0.0001,
+      "loss": 1.5533,
+      "step": 2179
+    },
+    {
+      "epoch": 0.7300736771600804,
+      "grad_norm": 0.4621736407279968,
+      "learning_rate": 0.0001,
+      "loss": 1.4768,
+      "step": 2180
+    },
+    {
+      "epoch": 0.730408573342264,
+      "grad_norm": 0.45624223351478577,
+      "learning_rate": 0.0001,
+      "loss": 1.5619,
+      "step": 2181
+    },
+    {
+      "epoch": 0.7307434695244475,
+      "grad_norm": 0.46043768525123596,
+      "learning_rate": 0.0001,
+      "loss": 1.4666,
+      "step": 2182
+    },
+    {
+      "epoch": 0.731078365706631,
+      "grad_norm": 0.4694119095802307,
+      "learning_rate": 0.0001,
+      "loss": 1.5557,
+      "step": 2183
+    },
+    {
+      "epoch": 0.7314132618888145,
+      "grad_norm": 0.46328192949295044,
+      "learning_rate": 0.0001,
+      "loss": 1.5293,
+      "step": 2184
+    },
+    {
+      "epoch": 0.731748158070998,
+      "grad_norm": 0.4788912236690521,
+      "learning_rate": 0.0001,
+      "loss": 1.5616,
+      "step": 2185
+    },
+    {
+      "epoch": 0.7320830542531815,
+      "grad_norm": 0.4542524814605713,
+      "learning_rate": 0.0001,
+      "loss": 1.5318,
+      "step": 2186
+    },
+    {
+      "epoch": 0.7324179504353651,
+      "grad_norm": 0.457428514957428,
+      "learning_rate": 0.0001,
+      "loss": 1.4104,
+      "step": 2187
+    },
+    {
+      "epoch": 0.7327528466175486,
+      "grad_norm": 0.4589458405971527,
+      "learning_rate": 0.0001,
+      "loss": 1.5151,
+      "step": 2188
+    },
+    {
+      "epoch": 0.7330877427997321,
+      "grad_norm": 0.48504841327667236,
+      "learning_rate": 0.0001,
+      "loss": 1.5749,
+      "step": 2189
+    },
+    {
+      "epoch": 0.7334226389819156,
+      "grad_norm": 0.47457072138786316,
+      "learning_rate": 0.0001,
+      "loss": 1.5697,
+      "step": 2190
+    },
+    {
+      "epoch": 0.7337575351640991,
+      "grad_norm": 0.4697708785533905,
+      "learning_rate": 0.0001,
+      "loss": 1.5097,
+      "step": 2191
+    },
+    {
+      "epoch": 0.7340924313462827,
+      "grad_norm": 0.48357653617858887,
+      "learning_rate": 0.0001,
+      "loss": 1.5399,
+      "step": 2192
+    },
+    {
+      "epoch": 0.7344273275284662,
+      "grad_norm": 0.45093438029289246,
+      "learning_rate": 0.0001,
+      "loss": 1.5439,
+      "step": 2193
+    },
+    {
+      "epoch": 0.7347622237106497,
+      "grad_norm": 0.5057450532913208,
+      "learning_rate": 0.0001,
+      "loss": 1.4819,
+      "step": 2194
+    },
+    {
+      "epoch": 0.7350971198928332,
+      "grad_norm": 0.4725852310657501,
+      "learning_rate": 0.0001,
+      "loss": 1.4298,
+      "step": 2195
+    },
+    {
+      "epoch": 0.7354320160750167,
+      "grad_norm": 0.48134663701057434,
+      "learning_rate": 0.0001,
+      "loss": 1.5225,
+      "step": 2196
+    },
+    {
+      "epoch": 0.7357669122572003,
+      "grad_norm": 0.4459664821624756,
+      "learning_rate": 0.0001,
+      "loss": 1.446,
+      "step": 2197
+    },
+    {
+      "epoch": 0.7361018084393838,
+      "grad_norm": 0.44949108362197876,
+      "learning_rate": 0.0001,
+      "loss": 1.534,
+      "step": 2198
+    },
+    {
+      "epoch": 0.7364367046215673,
+      "grad_norm": 0.4862693250179291,
+      "learning_rate": 0.0001,
+      "loss": 1.5543,
+      "step": 2199
+    },
+    {
+      "epoch": 0.7367716008037508,
+      "grad_norm": 0.45357629656791687,
+      "learning_rate": 0.0001,
+      "loss": 1.5667,
+      "step": 2200
+    },
+    {
+      "epoch": 0.7371064969859343,
+      "grad_norm": 0.47933268547058105,
+      "learning_rate": 0.0001,
+      "loss": 1.5888,
+      "step": 2201
+    },
+    {
+      "epoch": 0.7374413931681179,
+      "grad_norm": 0.4429778754711151,
+      "learning_rate": 0.0001,
+      "loss": 1.5209,
+      "step": 2202
+    },
+    {
+      "epoch": 0.7377762893503014,
+      "grad_norm": 0.4456193745136261,
+      "learning_rate": 0.0001,
+      "loss": 1.4034,
+      "step": 2203
+    },
+    {
+      "epoch": 0.7381111855324849,
+      "grad_norm": 0.494873583316803,
+      "learning_rate": 0.0001,
+      "loss": 1.581,
+      "step": 2204
+    },
+    {
+      "epoch": 0.7384460817146684,
+      "grad_norm": 0.4704873859882355,
+      "learning_rate": 0.0001,
+      "loss": 1.4645,
+      "step": 2205
+    },
+    {
+      "epoch": 0.7387809778968519,
+      "grad_norm": 0.5142254829406738,
+      "learning_rate": 0.0001,
+      "loss": 1.5792,
+      "step": 2206
+    },
+    {
+      "epoch": 0.7391158740790355,
+      "grad_norm": 0.46415936946868896,
+      "learning_rate": 0.0001,
+      "loss": 1.5007,
+      "step": 2207
+    },
+    {
+      "epoch": 0.739450770261219,
+      "grad_norm": 0.5375964641571045,
+      "learning_rate": 0.0001,
+      "loss": 1.5214,
+      "step": 2208
+    },
+    {
+      "epoch": 0.7397856664434025,
+      "grad_norm": 0.47598183155059814,
+      "learning_rate": 0.0001,
+      "loss": 1.466,
+      "step": 2209
+    },
+    {
+      "epoch": 0.7401205626255861,
+      "grad_norm": 0.49870026111602783,
+      "learning_rate": 0.0001,
+      "loss": 1.5136,
+      "step": 2210
+    },
+    {
+      "epoch": 0.7404554588077696,
+      "grad_norm": 0.49020954966545105,
+      "learning_rate": 0.0001,
+      "loss": 1.595,
+      "step": 2211
+    },
+    {
+      "epoch": 0.7407903549899532,
+      "grad_norm": 0.4574756324291229,
+      "learning_rate": 0.0001,
+      "loss": 1.5294,
+      "step": 2212
+    },
+    {
+      "epoch": 0.7411252511721367,
+      "grad_norm": 0.4988451898097992,
+      "learning_rate": 0.0001,
+      "loss": 1.6423,
+      "step": 2213
+    },
+    {
+      "epoch": 0.7414601473543202,
+      "grad_norm": 0.46195825934410095,
+      "learning_rate": 0.0001,
+      "loss": 1.5234,
+      "step": 2214
+    },
+    {
+      "epoch": 0.7417950435365037,
+      "grad_norm": 0.47354868054389954,
+      "learning_rate": 0.0001,
+      "loss": 1.4892,
+      "step": 2215
+    },
+    {
+      "epoch": 0.7421299397186872,
+      "grad_norm": 0.45956993103027344,
+      "learning_rate": 0.0001,
+      "loss": 1.5841,
+      "step": 2216
+    },
+    {
+      "epoch": 0.7424648359008708,
+      "grad_norm": 0.46149998903274536,
+      "learning_rate": 0.0001,
+      "loss": 1.5191,
+      "step": 2217
+    },
+    {
+      "epoch": 0.7427997320830543,
+      "grad_norm": 0.46643468737602234,
+      "learning_rate": 0.0001,
+      "loss": 1.5481,
+      "step": 2218
+    },
+    {
+      "epoch": 0.7431346282652378,
+      "grad_norm": 0.457994282245636,
+      "learning_rate": 0.0001,
+      "loss": 1.4564,
+      "step": 2219
+    },
+    {
+      "epoch": 0.7434695244474213,
+      "grad_norm": 0.48355191946029663,
+      "learning_rate": 0.0001,
+      "loss": 1.4359,
+      "step": 2220
+    },
+    {
+      "epoch": 0.7438044206296048,
+      "grad_norm": 0.4696873724460602,
+      "learning_rate": 0.0001,
+      "loss": 1.5165,
+      "step": 2221
+    },
+    {
+      "epoch": 0.7441393168117884,
+      "grad_norm": 0.4898888170719147,
+      "learning_rate": 0.0001,
+      "loss": 1.5201,
+      "step": 2222
+    },
+    {
+      "epoch": 0.7444742129939719,
+      "grad_norm": 0.47065985202789307,
+      "learning_rate": 0.0001,
+      "loss": 1.5632,
+      "step": 2223
+    },
+    {
+      "epoch": 0.7448091091761554,
+      "grad_norm": 0.47417351603507996,
+      "learning_rate": 0.0001,
+      "loss": 1.4845,
+      "step": 2224
+    },
+    {
+      "epoch": 0.7451440053583389,
+      "grad_norm": 0.4841914176940918,
+      "learning_rate": 0.0001,
+      "loss": 1.5638,
+      "step": 2225
+    },
+    {
+      "epoch": 0.7454789015405224,
+      "grad_norm": 0.47083160281181335,
+      "learning_rate": 0.0001,
+      "loss": 1.5412,
+      "step": 2226
+    },
+    {
+      "epoch": 0.745813797722706,
+      "grad_norm": 0.4515422582626343,
+      "learning_rate": 0.0001,
+      "loss": 1.4791,
+      "step": 2227
+    },
+    {
+      "epoch": 0.7461486939048895,
+      "grad_norm": 0.4730348587036133,
+      "learning_rate": 0.0001,
+      "loss": 1.4716,
+      "step": 2228
+    },
+    {
+      "epoch": 0.746483590087073,
+      "grad_norm": 0.4598521888256073,
+      "learning_rate": 0.0001,
+      "loss": 1.5684,
+      "step": 2229
+    },
+    {
+      "epoch": 0.7468184862692565,
+      "grad_norm": 0.45444709062576294,
+      "learning_rate": 0.0001,
+      "loss": 1.5607,
+      "step": 2230
+    },
+    {
+      "epoch": 0.74715338245144,
+      "grad_norm": 0.4803113639354706,
+      "learning_rate": 0.0001,
+      "loss": 1.6007,
+      "step": 2231
+    },
+    {
+      "epoch": 0.7474882786336235,
+      "grad_norm": 0.515312671661377,
+      "learning_rate": 0.0001,
+      "loss": 1.5927,
+      "step": 2232
+    },
+    {
+      "epoch": 0.7478231748158071,
+      "grad_norm": 0.44947949051856995,
+      "learning_rate": 0.0001,
+      "loss": 1.4926,
+      "step": 2233
+    },
+    {
+      "epoch": 0.7481580709979906,
+      "grad_norm": 0.5019223093986511,
+      "learning_rate": 0.0001,
+      "loss": 1.5951,
+      "step": 2234
+    },
+    {
+      "epoch": 0.7484929671801741,
+      "grad_norm": 0.4727962017059326,
+      "learning_rate": 0.0001,
+      "loss": 1.5391,
+      "step": 2235
+    },
+    {
+      "epoch": 0.7488278633623576,
+      "grad_norm": 0.45799902081489563,
+      "learning_rate": 0.0001,
+      "loss": 1.5248,
+      "step": 2236
+    },
+    {
+      "epoch": 0.7491627595445411,
+      "grad_norm": 0.4660114049911499,
+      "learning_rate": 0.0001,
+      "loss": 1.5082,
+      "step": 2237
+    },
+    {
+      "epoch": 0.7494976557267247,
+      "grad_norm": 0.4819839298725128,
+      "learning_rate": 0.0001,
+      "loss": 1.559,
+      "step": 2238
+    },
+    {
+      "epoch": 0.7498325519089082,
+      "grad_norm": 0.47706669569015503,
+      "learning_rate": 0.0001,
+      "loss": 1.5178,
+      "step": 2239
+    },
+    {
+      "epoch": 0.7501674480910918,
+      "grad_norm": 0.48049312829971313,
+      "learning_rate": 0.0001,
+      "loss": 1.5086,
+      "step": 2240
+    },
+    {
+      "epoch": 0.7505023442732753,
+      "grad_norm": 0.4732329547405243,
+      "learning_rate": 0.0001,
+      "loss": 1.4876,
+      "step": 2241
+    },
+    {
+      "epoch": 0.7508372404554589,
+      "grad_norm": 0.4817315936088562,
+      "learning_rate": 0.0001,
+      "loss": 1.5802,
+      "step": 2242
+    },
+    {
+      "epoch": 0.7511721366376424,
+      "grad_norm": 0.44466477632522583,
+      "learning_rate": 0.0001,
+      "loss": 1.4521,
+      "step": 2243
+    },
+    {
+      "epoch": 0.7515070328198259,
+      "grad_norm": 0.4766438603401184,
+      "learning_rate": 0.0001,
+      "loss": 1.559,
+      "step": 2244
+    },
+    {
+      "epoch": 0.7518419290020094,
+      "grad_norm": 0.5050947070121765,
+      "learning_rate": 0.0001,
+      "loss": 1.5825,
+      "step": 2245
+    },
+    {
+      "epoch": 0.7521768251841929,
+      "grad_norm": 0.4919584393501282,
+      "learning_rate": 0.0001,
+      "loss": 1.6224,
+      "step": 2246
+    },
+    {
+      "epoch": 0.7525117213663765,
+      "grad_norm": 0.4567074179649353,
+      "learning_rate": 0.0001,
+      "loss": 1.476,
+      "step": 2247
+    },
+    {
+      "epoch": 0.75284661754856,
+      "grad_norm": 0.490669846534729,
+      "learning_rate": 0.0001,
+      "loss": 1.4554,
+      "step": 2248
+    },
+    {
+      "epoch": 0.7531815137307435,
+      "grad_norm": 0.4712928533554077,
+      "learning_rate": 0.0001,
+      "loss": 1.4727,
+      "step": 2249
+    },
+    {
+      "epoch": 0.753516409912927,
+      "grad_norm": 0.45836925506591797,
+      "learning_rate": 0.0001,
+      "loss": 1.479,
+      "step": 2250
+    },
+    {
+      "epoch": 0.7538513060951105,
+      "grad_norm": 0.4567963480949402,
+      "learning_rate": 0.0001,
+      "loss": 1.5333,
+      "step": 2251
+    },
+    {
+      "epoch": 0.754186202277294,
+      "grad_norm": 0.4772929847240448,
+      "learning_rate": 0.0001,
+      "loss": 1.564,
+      "step": 2252
+    },
+    {
+      "epoch": 0.7545210984594776,
+      "grad_norm": 0.45997437834739685,
+      "learning_rate": 0.0001,
+      "loss": 1.5597,
+      "step": 2253
+    },
+    {
+      "epoch": 0.7548559946416611,
+      "grad_norm": 0.46403026580810547,
+      "learning_rate": 0.0001,
+      "loss": 1.5363,
+      "step": 2254
+    },
+    {
+      "epoch": 0.7551908908238446,
+      "grad_norm": 0.4640863239765167,
+      "learning_rate": 0.0001,
+      "loss": 1.6155,
+      "step": 2255
+    },
+    {
+      "epoch": 0.7555257870060281,
+      "grad_norm": 0.46999940276145935,
+      "learning_rate": 0.0001,
+      "loss": 1.459,
+      "step": 2256
+    },
+    {
+      "epoch": 0.7558606831882116,
+      "grad_norm": 0.4665103852748871,
+      "learning_rate": 0.0001,
+      "loss": 1.4763,
+      "step": 2257
+    },
+    {
+      "epoch": 0.7561955793703952,
+      "grad_norm": 0.471189945936203,
+      "learning_rate": 0.0001,
+      "loss": 1.4699,
+      "step": 2258
+    },
+    {
+      "epoch": 0.7565304755525787,
+      "grad_norm": 0.467632532119751,
+      "learning_rate": 0.0001,
+      "loss": 1.5045,
+      "step": 2259
+    },
+    {
+      "epoch": 0.7568653717347622,
+      "grad_norm": 0.44308537244796753,
+      "learning_rate": 0.0001,
+      "loss": 1.4245,
+      "step": 2260
+    },
+    {
+      "epoch": 0.7572002679169457,
+      "grad_norm": 0.46536949276924133,
+      "learning_rate": 0.0001,
+      "loss": 1.4407,
+      "step": 2261
+    },
+    {
+      "epoch": 0.7575351640991292,
+      "grad_norm": 0.45661500096321106,
+      "learning_rate": 0.0001,
+      "loss": 1.4973,
+      "step": 2262
+    },
+    {
+      "epoch": 0.7578700602813128,
+      "grad_norm": 0.48596176505088806,
+      "learning_rate": 0.0001,
+      "loss": 1.5173,
+      "step": 2263
+    },
+    {
+      "epoch": 0.7582049564634963,
+      "grad_norm": 0.47135865688323975,
+      "learning_rate": 0.0001,
+      "loss": 1.5616,
+      "step": 2264
+    },
+    {
+      "epoch": 0.7585398526456798,
+      "grad_norm": 0.48448967933654785,
+      "learning_rate": 0.0001,
+      "loss": 1.5076,
+      "step": 2265
+    },
+    {
+      "epoch": 0.7588747488278633,
+      "grad_norm": 0.4838770627975464,
+      "learning_rate": 0.0001,
+      "loss": 1.4286,
+      "step": 2266
+    },
+    {
+      "epoch": 0.7592096450100468,
+      "grad_norm": 0.4586263597011566,
+      "learning_rate": 0.0001,
+      "loss": 1.505,
+      "step": 2267
+    },
+    {
+      "epoch": 0.7595445411922304,
+      "grad_norm": 0.4685727655887604,
+      "learning_rate": 0.0001,
+      "loss": 1.5812,
+      "step": 2268
+    },
+    {
+      "epoch": 0.7598794373744139,
+      "grad_norm": 0.47122806310653687,
+      "learning_rate": 0.0001,
+      "loss": 1.5494,
+      "step": 2269
+    },
+    {
+      "epoch": 0.7602143335565975,
+      "grad_norm": 0.449631929397583,
+      "learning_rate": 0.0001,
+      "loss": 1.5016,
+      "step": 2270
+    },
+    {
+      "epoch": 0.760549229738781,
+      "grad_norm": 0.46288001537323,
+      "learning_rate": 0.0001,
+      "loss": 1.5631,
+      "step": 2271
+    },
+    {
+      "epoch": 0.7608841259209645,
+      "grad_norm": 0.462016224861145,
+      "learning_rate": 0.0001,
+      "loss": 1.5226,
+      "step": 2272
+    },
+    {
+      "epoch": 0.7612190221031481,
+      "grad_norm": 0.4567805230617523,
+      "learning_rate": 0.0001,
+      "loss": 1.4918,
+      "step": 2273
+    },
+    {
+      "epoch": 0.7615539182853316,
+      "grad_norm": 0.4616073668003082,
+      "learning_rate": 0.0001,
+      "loss": 1.5156,
+      "step": 2274
+    },
+    {
+      "epoch": 0.7618888144675151,
+      "grad_norm": 0.4843860864639282,
+      "learning_rate": 0.0001,
+      "loss": 1.5841,
+      "step": 2275
+    },
+    {
+      "epoch": 0.7622237106496986,
+      "grad_norm": 0.4428068995475769,
+      "learning_rate": 0.0001,
+      "loss": 1.4995,
+      "step": 2276
+    },
+    {
+      "epoch": 0.7625586068318821,
+      "grad_norm": 0.4614906311035156,
+      "learning_rate": 0.0001,
+      "loss": 1.5775,
+      "step": 2277
+    },
+    {
+      "epoch": 0.7628935030140657,
+      "grad_norm": 0.4830222725868225,
+      "learning_rate": 0.0001,
+      "loss": 1.6047,
+      "step": 2278
+    },
+    {
+      "epoch": 0.7632283991962492,
+      "grad_norm": 0.4605547785758972,
+      "learning_rate": 0.0001,
+      "loss": 1.5153,
+      "step": 2279
+    },
+    {
+      "epoch": 0.7635632953784327,
+      "grad_norm": 0.4535585641860962,
+      "learning_rate": 0.0001,
+      "loss": 1.4303,
+      "step": 2280
+    },
+    {
+      "epoch": 0.7638981915606162,
+      "grad_norm": 0.4725355803966522,
+      "learning_rate": 0.0001,
+      "loss": 1.584,
+      "step": 2281
+    },
+    {
+      "epoch": 0.7642330877427997,
+      "grad_norm": 0.45823484659194946,
+      "learning_rate": 0.0001,
+      "loss": 1.464,
+      "step": 2282
+    },
+    {
+      "epoch": 0.7645679839249833,
+      "grad_norm": 0.47183576226234436,
+      "learning_rate": 0.0001,
+      "loss": 1.5424,
+      "step": 2283
+    },
+    {
+      "epoch": 0.7649028801071668,
+      "grad_norm": 0.46699899435043335,
+      "learning_rate": 0.0001,
+      "loss": 1.4754,
+      "step": 2284
+    },
+    {
+      "epoch": 0.7652377762893503,
+      "grad_norm": 0.4842059016227722,
+      "learning_rate": 0.0001,
+      "loss": 1.5127,
+      "step": 2285
+    },
+    {
+      "epoch": 0.7655726724715338,
+      "grad_norm": 0.46355974674224854,
+      "learning_rate": 0.0001,
+      "loss": 1.4563,
+      "step": 2286
+    },
+    {
+      "epoch": 0.7659075686537173,
+      "grad_norm": 0.44617757201194763,
+      "learning_rate": 0.0001,
+      "loss": 1.4338,
+      "step": 2287
+    },
+    {
+      "epoch": 0.7662424648359009,
+      "grad_norm": 0.4712759852409363,
+      "learning_rate": 0.0001,
+      "loss": 1.5535,
+      "step": 2288
+    },
+    {
+      "epoch": 0.7665773610180844,
+      "grad_norm": 0.44124874472618103,
+      "learning_rate": 0.0001,
+      "loss": 1.5111,
+      "step": 2289
+    },
+    {
+      "epoch": 0.7669122572002679,
+      "grad_norm": 0.45401328802108765,
+      "learning_rate": 0.0001,
+      "loss": 1.5066,
+      "step": 2290
+    },
+    {
+      "epoch": 0.7672471533824514,
+      "grad_norm": 0.4824156165122986,
+      "learning_rate": 0.0001,
+      "loss": 1.5143,
+      "step": 2291
+    },
+    {
+      "epoch": 0.7675820495646349,
+      "grad_norm": 0.4732379913330078,
+      "learning_rate": 0.0001,
+      "loss": 1.5505,
+      "step": 2292
+    },
+    {
+      "epoch": 0.7679169457468185,
+      "grad_norm": 0.4697463810443878,
+      "learning_rate": 0.0001,
+      "loss": 1.5309,
+      "step": 2293
+    },
+    {
+      "epoch": 0.768251841929002,
+      "grad_norm": 0.45176729559898376,
+      "learning_rate": 0.0001,
+      "loss": 1.4686,
+      "step": 2294
+    },
+    {
+      "epoch": 0.7685867381111855,
+      "grad_norm": 0.44826486706733704,
+      "learning_rate": 0.0001,
+      "loss": 1.5225,
+      "step": 2295
+    },
+    {
+      "epoch": 0.768921634293369,
+      "grad_norm": 0.45865607261657715,
+      "learning_rate": 0.0001,
+      "loss": 1.4993,
+      "step": 2296
+    },
+    {
+      "epoch": 0.7692565304755525,
+      "grad_norm": 0.4361908733844757,
+      "learning_rate": 0.0001,
+      "loss": 1.4904,
+      "step": 2297
+    },
+    {
+      "epoch": 0.769591426657736,
+      "grad_norm": 0.47969165444374084,
+      "learning_rate": 0.0001,
+      "loss": 1.5102,
+      "step": 2298
+    },
+    {
+      "epoch": 0.7699263228399196,
+      "grad_norm": 0.48042747378349304,
+      "learning_rate": 0.0001,
+      "loss": 1.5363,
+      "step": 2299
+    },
+    {
+      "epoch": 0.7702612190221031,
+      "grad_norm": 0.452140212059021,
+      "learning_rate": 0.0001,
+      "loss": 1.4566,
+      "step": 2300
+    },
+    {
+      "epoch": 0.7705961152042867,
+      "grad_norm": 0.48098182678222656,
+      "learning_rate": 0.0001,
+      "loss": 1.5807,
+      "step": 2301
+    },
+    {
+      "epoch": 0.7709310113864702,
+      "grad_norm": 0.45571669936180115,
+      "learning_rate": 0.0001,
+      "loss": 1.5332,
+      "step": 2302
+    },
+    {
+      "epoch": 0.7712659075686538,
+      "grad_norm": 0.4881724715232849,
+      "learning_rate": 0.0001,
+      "loss": 1.566,
+      "step": 2303
+    },
+    {
+      "epoch": 0.7716008037508373,
+      "grad_norm": 0.4681125283241272,
+      "learning_rate": 0.0001,
+      "loss": 1.5146,
+      "step": 2304
+    },
+    {
+      "epoch": 0.7719356999330208,
+      "grad_norm": 0.5105034708976746,
+      "learning_rate": 0.0001,
+      "loss": 1.5381,
+      "step": 2305
+    },
+    {
+      "epoch": 0.7722705961152043,
+      "grad_norm": 0.4545895457267761,
+      "learning_rate": 0.0001,
+      "loss": 1.4814,
+      "step": 2306
+    },
+    {
+      "epoch": 0.7726054922973878,
+      "grad_norm": 0.4493445158004761,
+      "learning_rate": 0.0001,
+      "loss": 1.5643,
+      "step": 2307
+    },
+    {
+      "epoch": 0.7729403884795714,
+      "grad_norm": 0.4737282395362854,
+      "learning_rate": 0.0001,
+      "loss": 1.4749,
+      "step": 2308
+    },
+    {
+      "epoch": 0.7732752846617549,
+      "grad_norm": 0.47303324937820435,
+      "learning_rate": 0.0001,
+      "loss": 1.5203,
+      "step": 2309
+    },
+    {
+      "epoch": 0.7736101808439384,
+      "grad_norm": 0.478311687707901,
+      "learning_rate": 0.0001,
+      "loss": 1.5582,
+      "step": 2310
+    },
+    {
+      "epoch": 0.7739450770261219,
+      "grad_norm": 0.5018264651298523,
+      "learning_rate": 0.0001,
+      "loss": 1.6107,
+      "step": 2311
+    },
+    {
+      "epoch": 0.7742799732083054,
+      "grad_norm": 0.4647297263145447,
+      "learning_rate": 0.0001,
+      "loss": 1.5217,
+      "step": 2312
+    },
+    {
+      "epoch": 0.774614869390489,
+      "grad_norm": 0.4564414918422699,
+      "learning_rate": 0.0001,
+      "loss": 1.5125,
+      "step": 2313
+    },
+    {
+      "epoch": 0.7749497655726725,
+      "grad_norm": 0.4897812008857727,
+      "learning_rate": 0.0001,
+      "loss": 1.5987,
+      "step": 2314
+    },
+    {
+      "epoch": 0.775284661754856,
+      "grad_norm": 0.47572436928749084,
+      "learning_rate": 0.0001,
+      "loss": 1.3898,
+      "step": 2315
+    },
+    {
+      "epoch": 0.7756195579370395,
+      "grad_norm": 0.4664483964443207,
+      "learning_rate": 0.0001,
+      "loss": 1.5198,
+      "step": 2316
+    },
+    {
+      "epoch": 0.775954454119223,
+      "grad_norm": 0.4727928936481476,
+      "learning_rate": 0.0001,
+      "loss": 1.5286,
+      "step": 2317
+    },
+    {
+      "epoch": 0.7762893503014066,
+      "grad_norm": 0.48919540643692017,
+      "learning_rate": 0.0001,
+      "loss": 1.539,
+      "step": 2318
+    },
+    {
+      "epoch": 0.7766242464835901,
+      "grad_norm": 0.46668481826782227,
+      "learning_rate": 0.0001,
+      "loss": 1.5559,
+      "step": 2319
+    },
+    {
+      "epoch": 0.7769591426657736,
+      "grad_norm": 0.46336182951927185,
+      "learning_rate": 0.0001,
+      "loss": 1.5859,
+      "step": 2320
+    },
+    {
+      "epoch": 0.7772940388479571,
+      "grad_norm": 0.45855921506881714,
+      "learning_rate": 0.0001,
+      "loss": 1.4619,
+      "step": 2321
+    },
+    {
+      "epoch": 0.7776289350301406,
+      "grad_norm": 0.46973612904548645,
+      "learning_rate": 0.0001,
+      "loss": 1.5,
+      "step": 2322
+    },
+    {
+      "epoch": 0.7779638312123242,
+      "grad_norm": 0.44328898191452026,
+      "learning_rate": 0.0001,
+      "loss": 1.4844,
+      "step": 2323
+    },
+    {
+      "epoch": 0.7782987273945077,
+      "grad_norm": 0.4648163914680481,
+      "learning_rate": 0.0001,
+      "loss": 1.5545,
+      "step": 2324
+    },
+    {
+      "epoch": 0.7786336235766912,
+      "grad_norm": 0.459846556186676,
+      "learning_rate": 0.0001,
+      "loss": 1.4493,
+      "step": 2325
+    },
+    {
+      "epoch": 0.7789685197588747,
+      "grad_norm": 0.45208072662353516,
+      "learning_rate": 0.0001,
+      "loss": 1.5098,
+      "step": 2326
+    },
+    {
+      "epoch": 0.7793034159410582,
+      "grad_norm": 0.4754399061203003,
+      "learning_rate": 0.0001,
+      "loss": 1.5381,
+      "step": 2327
+    },
+    {
+      "epoch": 0.7796383121232418,
+      "grad_norm": 0.48218896985054016,
+      "learning_rate": 0.0001,
+      "loss": 1.6072,
+      "step": 2328
+    },
+    {
+      "epoch": 0.7799732083054253,
+      "grad_norm": 0.49892452359199524,
+      "learning_rate": 0.0001,
+      "loss": 1.517,
+      "step": 2329
+    },
+    {
+      "epoch": 0.7803081044876088,
+      "grad_norm": 0.49093908071517944,
+      "learning_rate": 0.0001,
+      "loss": 1.5653,
+      "step": 2330
+    },
+    {
+      "epoch": 0.7806430006697924,
+      "grad_norm": 0.46677327156066895,
+      "learning_rate": 0.0001,
+      "loss": 1.5399,
+      "step": 2331
+    },
+    {
+      "epoch": 0.7809778968519759,
+      "grad_norm": 0.45744234323501587,
+      "learning_rate": 0.0001,
+      "loss": 1.4072,
+      "step": 2332
+    },
+    {
+      "epoch": 0.7813127930341595,
+      "grad_norm": 0.48118555545806885,
+      "learning_rate": 0.0001,
+      "loss": 1.4989,
+      "step": 2333
+    },
+    {
+      "epoch": 0.781647689216343,
+      "grad_norm": 0.4569566547870636,
+      "learning_rate": 0.0001,
+      "loss": 1.5025,
+      "step": 2334
+    },
+    {
+      "epoch": 0.7819825853985265,
+      "grad_norm": 0.4544038772583008,
+      "learning_rate": 0.0001,
+      "loss": 1.5297,
+      "step": 2335
+    },
+    {
+      "epoch": 0.78231748158071,
+      "grad_norm": 0.4806601107120514,
+      "learning_rate": 0.0001,
+      "loss": 1.5866,
+      "step": 2336
+    },
+    {
+      "epoch": 0.7826523777628935,
+      "grad_norm": 0.47200947999954224,
+      "learning_rate": 0.0001,
+      "loss": 1.5548,
+      "step": 2337
+    },
+    {
+      "epoch": 0.7829872739450771,
+      "grad_norm": 0.4612894654273987,
+      "learning_rate": 0.0001,
+      "loss": 1.544,
+      "step": 2338
+    },
+    {
+      "epoch": 0.7833221701272606,
+      "grad_norm": 0.4631419777870178,
+      "learning_rate": 0.0001,
+      "loss": 1.527,
+      "step": 2339
+    },
+    {
+      "epoch": 0.7836570663094441,
+      "grad_norm": 0.4725373685359955,
+      "learning_rate": 0.0001,
+      "loss": 1.5301,
+      "step": 2340
+    },
+    {
+      "epoch": 0.7839919624916276,
+      "grad_norm": 0.4709610641002655,
+      "learning_rate": 0.0001,
+      "loss": 1.5305,
+      "step": 2341
+    },
+    {
+      "epoch": 0.7843268586738111,
+      "grad_norm": 0.45424067974090576,
+      "learning_rate": 0.0001,
+      "loss": 1.4967,
+      "step": 2342
+    },
+    {
+      "epoch": 0.7846617548559947,
+      "grad_norm": 0.4698200225830078,
+      "learning_rate": 0.0001,
+      "loss": 1.5697,
+      "step": 2343
+    },
+    {
+      "epoch": 0.7849966510381782,
+      "grad_norm": 0.474651962518692,
+      "learning_rate": 0.0001,
+      "loss": 1.4652,
+      "step": 2344
+    },
+    {
+      "epoch": 0.7853315472203617,
+      "grad_norm": 0.4748631417751312,
+      "learning_rate": 0.0001,
+      "loss": 1.6113,
+      "step": 2345
+    },
+    {
+      "epoch": 0.7856664434025452,
+      "grad_norm": 0.4516560137271881,
+      "learning_rate": 0.0001,
+      "loss": 1.4746,
+      "step": 2346
+    },
+    {
+      "epoch": 0.7860013395847287,
+      "grad_norm": 0.47162577509880066,
+      "learning_rate": 0.0001,
+      "loss": 1.4741,
+      "step": 2347
+    },
+    {
+      "epoch": 0.7863362357669123,
+      "grad_norm": 0.4563868045806885,
+      "learning_rate": 0.0001,
+      "loss": 1.5713,
+      "step": 2348
+    },
+    {
+      "epoch": 0.7866711319490958,
+      "grad_norm": 0.48005253076553345,
+      "learning_rate": 0.0001,
+      "loss": 1.5462,
+      "step": 2349
+    },
+    {
+      "epoch": 0.7870060281312793,
+      "grad_norm": 0.47020062804222107,
+      "learning_rate": 0.0001,
+      "loss": 1.4656,
+      "step": 2350
+    },
+    {
+      "epoch": 0.7873409243134628,
+      "grad_norm": 0.4505898058414459,
+      "learning_rate": 0.0001,
+      "loss": 1.4406,
+      "step": 2351
+    },
+    {
+      "epoch": 0.7876758204956463,
+      "grad_norm": 0.4679824709892273,
+      "learning_rate": 0.0001,
+      "loss": 1.5187,
+      "step": 2352
+    },
+    {
+      "epoch": 0.7880107166778298,
+      "grad_norm": 0.4632118344306946,
+      "learning_rate": 0.0001,
+      "loss": 1.4943,
+      "step": 2353
+    },
+    {
+      "epoch": 0.7883456128600134,
+      "grad_norm": 0.4546247720718384,
+      "learning_rate": 0.0001,
+      "loss": 1.4485,
+      "step": 2354
+    },
+    {
+      "epoch": 0.7886805090421969,
+      "grad_norm": 0.48107215762138367,
+      "learning_rate": 0.0001,
+      "loss": 1.4792,
+      "step": 2355
+    },
+    {
+      "epoch": 0.7890154052243804,
+      "grad_norm": 0.473503977060318,
+      "learning_rate": 0.0001,
+      "loss": 1.5215,
+      "step": 2356
+    },
+    {
+      "epoch": 0.7893503014065639,
+      "grad_norm": 0.4569712281227112,
+      "learning_rate": 0.0001,
+      "loss": 1.4349,
+      "step": 2357
+    },
+    {
+      "epoch": 0.7896851975887474,
+      "grad_norm": 0.44963976740837097,
+      "learning_rate": 0.0001,
+      "loss": 1.464,
+      "step": 2358
+    },
+    {
+      "epoch": 0.790020093770931,
+      "grad_norm": 0.4830678701400757,
+      "learning_rate": 0.0001,
+      "loss": 1.6487,
+      "step": 2359
+    },
+    {
+      "epoch": 0.7903549899531145,
+      "grad_norm": 0.4756050407886505,
+      "learning_rate": 0.0001,
+      "loss": 1.3998,
+      "step": 2360
+    },
+    {
+      "epoch": 0.790689886135298,
+      "grad_norm": 0.46634572744369507,
+      "learning_rate": 0.0001,
+      "loss": 1.5183,
+      "step": 2361
+    },
+    {
+      "epoch": 0.7910247823174816,
+      "grad_norm": 0.4650283753871918,
+      "learning_rate": 0.0001,
+      "loss": 1.4784,
+      "step": 2362
+    },
+    {
+      "epoch": 0.7913596784996652,
+      "grad_norm": 0.45666956901550293,
+      "learning_rate": 0.0001,
+      "loss": 1.3785,
+      "step": 2363
+    },
+    {
+      "epoch": 0.7916945746818487,
+      "grad_norm": 0.45927444100379944,
+      "learning_rate": 0.0001,
+      "loss": 1.4673,
+      "step": 2364
+    },
+    {
+      "epoch": 0.7920294708640322,
+      "grad_norm": 0.47485730051994324,
+      "learning_rate": 0.0001,
+      "loss": 1.4951,
+      "step": 2365
+    },
+    {
+      "epoch": 0.7923643670462157,
+      "grad_norm": 0.4480372369289398,
+      "learning_rate": 0.0001,
+      "loss": 1.5052,
+      "step": 2366
+    },
+    {
+      "epoch": 0.7926992632283992,
+      "grad_norm": 0.4613422155380249,
+      "learning_rate": 0.0001,
+      "loss": 1.4325,
+      "step": 2367
+    },
+    {
+      "epoch": 0.7930341594105828,
+      "grad_norm": 0.48185497522354126,
+      "learning_rate": 0.0001,
+      "loss": 1.5527,
+      "step": 2368
+    },
+    {
+      "epoch": 0.7933690555927663,
+      "grad_norm": 0.46273407340049744,
+      "learning_rate": 0.0001,
+      "loss": 1.5327,
+      "step": 2369
+    },
+    {
+      "epoch": 0.7937039517749498,
+      "grad_norm": 0.4855313003063202,
+      "learning_rate": 0.0001,
+      "loss": 1.5299,
+      "step": 2370
+    },
+    {
+      "epoch": 0.7940388479571333,
+      "grad_norm": 0.46928760409355164,
+      "learning_rate": 0.0001,
+      "loss": 1.5184,
+      "step": 2371
+    },
+    {
+      "epoch": 0.7943737441393168,
+      "grad_norm": 0.4650458097457886,
+      "learning_rate": 0.0001,
+      "loss": 1.4791,
+      "step": 2372
+    },
+    {
+      "epoch": 0.7947086403215003,
+      "grad_norm": 0.4902491271495819,
+      "learning_rate": 0.0001,
+      "loss": 1.5162,
+      "step": 2373
+    },
+    {
+      "epoch": 0.7950435365036839,
+      "grad_norm": 0.5136062502861023,
+      "learning_rate": 0.0001,
+      "loss": 1.4842,
+      "step": 2374
+    },
+    {
+      "epoch": 0.7953784326858674,
+      "grad_norm": 0.4830358028411865,
+      "learning_rate": 0.0001,
+      "loss": 1.5568,
+      "step": 2375
+    },
+    {
+      "epoch": 0.7957133288680509,
+      "grad_norm": 0.4639476537704468,
+      "learning_rate": 0.0001,
+      "loss": 1.4308,
+      "step": 2376
+    },
+    {
+      "epoch": 0.7960482250502344,
+      "grad_norm": 0.5048962235450745,
+      "learning_rate": 0.0001,
+      "loss": 1.5455,
+      "step": 2377
+    },
+    {
+      "epoch": 0.796383121232418,
+      "grad_norm": 0.4968639612197876,
+      "learning_rate": 0.0001,
+      "loss": 1.5439,
+      "step": 2378
+    },
+    {
+      "epoch": 0.7967180174146015,
+      "grad_norm": 0.48246777057647705,
+      "learning_rate": 0.0001,
+      "loss": 1.5037,
+      "step": 2379
+    },
+    {
+      "epoch": 0.797052913596785,
+      "grad_norm": 0.46209555864334106,
+      "learning_rate": 0.0001,
+      "loss": 1.6492,
+      "step": 2380
+    },
+    {
+      "epoch": 0.7973878097789685,
+      "grad_norm": 0.4774399995803833,
+      "learning_rate": 0.0001,
+      "loss": 1.4763,
+      "step": 2381
+    },
+    {
+      "epoch": 0.797722705961152,
+      "grad_norm": 0.48918208479881287,
+      "learning_rate": 0.0001,
+      "loss": 1.5102,
+      "step": 2382
+    },
+    {
+      "epoch": 0.7980576021433355,
+      "grad_norm": 0.48564663529396057,
+      "learning_rate": 0.0001,
+      "loss": 1.5395,
+      "step": 2383
+    },
+    {
+      "epoch": 0.7983924983255191,
+      "grad_norm": 0.4565158486366272,
+      "learning_rate": 0.0001,
+      "loss": 1.4679,
+      "step": 2384
+    },
+    {
+      "epoch": 0.7987273945077026,
+      "grad_norm": 0.47012779116630554,
+      "learning_rate": 0.0001,
+      "loss": 1.6424,
+      "step": 2385
+    },
+    {
+      "epoch": 0.7990622906898861,
+      "grad_norm": 0.4518689513206482,
+      "learning_rate": 0.0001,
+      "loss": 1.5085,
+      "step": 2386
+    },
+    {
+      "epoch": 0.7993971868720696,
+      "grad_norm": 0.4934041500091553,
+      "learning_rate": 0.0001,
+      "loss": 1.4546,
+      "step": 2387
+    },
+    {
+      "epoch": 0.7997320830542531,
+      "grad_norm": 0.47564956545829773,
+      "learning_rate": 0.0001,
+      "loss": 1.5534,
+      "step": 2388
+    },
+    {
+      "epoch": 0.8000669792364367,
+      "grad_norm": 0.4532725214958191,
+      "learning_rate": 0.0001,
+      "loss": 1.3697,
+      "step": 2389
+    },
+    {
+      "epoch": 0.8004018754186202,
+      "grad_norm": 0.4541550874710083,
+      "learning_rate": 0.0001,
+      "loss": 1.3951,
+      "step": 2390
+    },
+    {
+      "epoch": 0.8007367716008037,
+      "grad_norm": 0.4611206352710724,
+      "learning_rate": 0.0001,
+      "loss": 1.5683,
+      "step": 2391
+    },
+    {
+      "epoch": 0.8010716677829873,
+      "grad_norm": 0.46378272771835327,
+      "learning_rate": 0.0001,
+      "loss": 1.4767,
+      "step": 2392
+    },
+    {
+      "epoch": 0.8014065639651708,
+      "grad_norm": 0.4524199366569519,
+      "learning_rate": 0.0001,
+      "loss": 1.6076,
+      "step": 2393
+    },
+    {
+      "epoch": 0.8017414601473544,
+      "grad_norm": 0.46128812432289124,
+      "learning_rate": 0.0001,
+      "loss": 1.4649,
+      "step": 2394
+    },
+    {
+      "epoch": 0.8020763563295379,
+      "grad_norm": 0.4794237017631531,
+      "learning_rate": 0.0001,
+      "loss": 1.5022,
+      "step": 2395
+    },
+    {
+      "epoch": 0.8024112525117214,
+      "grad_norm": 0.45057275891304016,
+      "learning_rate": 0.0001,
+      "loss": 1.4823,
+      "step": 2396
+    },
+    {
+      "epoch": 0.8027461486939049,
+      "grad_norm": 0.4925333857536316,
+      "learning_rate": 0.0001,
+      "loss": 1.4393,
+      "step": 2397
+    },
+    {
+      "epoch": 0.8030810448760884,
+      "grad_norm": 0.4900800287723541,
+      "learning_rate": 0.0001,
+      "loss": 1.5873,
+      "step": 2398
+    },
+    {
+      "epoch": 0.803415941058272,
+      "grad_norm": 0.501660168170929,
+      "learning_rate": 0.0001,
+      "loss": 1.4849,
+      "step": 2399
+    },
+    {
+      "epoch": 0.8037508372404555,
+      "grad_norm": 0.4713497459888458,
+      "learning_rate": 0.0001,
+      "loss": 1.5155,
+      "step": 2400
+    },
+    {
+      "epoch": 0.804085733422639,
+      "grad_norm": 0.4639057219028473,
+      "learning_rate": 0.0001,
+      "loss": 1.4363,
+      "step": 2401
+    },
+    {
+      "epoch": 0.8044206296048225,
+      "grad_norm": 0.4746173620223999,
+      "learning_rate": 0.0001,
+      "loss": 1.4313,
+      "step": 2402
+    },
+    {
+      "epoch": 0.804755525787006,
+      "grad_norm": 0.45180779695510864,
+      "learning_rate": 0.0001,
+      "loss": 1.4934,
+      "step": 2403
+    },
+    {
+      "epoch": 0.8050904219691896,
+      "grad_norm": 0.5304993391036987,
+      "learning_rate": 0.0001,
+      "loss": 1.5424,
+      "step": 2404
+    },
+    {
+      "epoch": 0.8054253181513731,
+      "grad_norm": 0.4532136917114258,
+      "learning_rate": 0.0001,
+      "loss": 1.5413,
+      "step": 2405
+    },
+    {
+      "epoch": 0.8057602143335566,
+      "grad_norm": 0.48047250509262085,
+      "learning_rate": 0.0001,
+      "loss": 1.5941,
+      "step": 2406
+    },
+    {
+      "epoch": 0.8060951105157401,
+      "grad_norm": 0.46145549416542053,
+      "learning_rate": 0.0001,
+      "loss": 1.3725,
+      "step": 2407
+    },
+    {
+      "epoch": 0.8064300066979236,
+      "grad_norm": 0.4635370969772339,
+      "learning_rate": 0.0001,
+      "loss": 1.3842,
+      "step": 2408
+    },
+    {
+      "epoch": 0.8067649028801072,
+      "grad_norm": 0.4741126298904419,
+      "learning_rate": 0.0001,
+      "loss": 1.5344,
+      "step": 2409
+    },
+    {
+      "epoch": 0.8070997990622907,
+      "grad_norm": 0.45318159461021423,
+      "learning_rate": 0.0001,
+      "loss": 1.442,
+      "step": 2410
+    },
+    {
+      "epoch": 0.8074346952444742,
+      "grad_norm": 0.46751123666763306,
+      "learning_rate": 0.0001,
+      "loss": 1.5001,
+      "step": 2411
+    },
+    {
+      "epoch": 0.8077695914266577,
+      "grad_norm": 0.47589755058288574,
+      "learning_rate": 0.0001,
+      "loss": 1.5458,
+      "step": 2412
+    },
+    {
+      "epoch": 0.8081044876088412,
+      "grad_norm": 0.4462743103504181,
+      "learning_rate": 0.0001,
+      "loss": 1.4044,
+      "step": 2413
+    },
+    {
+      "epoch": 0.8084393837910248,
+      "grad_norm": 0.4586702585220337,
+      "learning_rate": 0.0001,
+      "loss": 1.5026,
+      "step": 2414
+    },
+    {
+      "epoch": 0.8087742799732083,
+      "grad_norm": 0.4458579123020172,
+      "learning_rate": 0.0001,
+      "loss": 1.4296,
+      "step": 2415
+    },
+    {
+      "epoch": 0.8091091761553918,
+      "grad_norm": 0.46752509474754333,
+      "learning_rate": 0.0001,
+      "loss": 1.4309,
+      "step": 2416
+    },
+    {
+      "epoch": 0.8094440723375753,
+      "grad_norm": 0.4684821665287018,
+      "learning_rate": 0.0001,
+      "loss": 1.5132,
+      "step": 2417
+    },
+    {
+      "epoch": 0.8097789685197588,
+      "grad_norm": 0.48130378127098083,
+      "learning_rate": 0.0001,
+      "loss": 1.4572,
+      "step": 2418
+    },
+    {
+      "epoch": 0.8101138647019424,
+      "grad_norm": 0.45288190245628357,
+      "learning_rate": 0.0001,
+      "loss": 1.3387,
+      "step": 2419
+    },
+    {
+      "epoch": 0.8104487608841259,
+      "grad_norm": 0.47424936294555664,
+      "learning_rate": 0.0001,
+      "loss": 1.4783,
+      "step": 2420
+    },
+    {
+      "epoch": 0.8107836570663094,
+      "grad_norm": 0.4607580006122589,
+      "learning_rate": 0.0001,
+      "loss": 1.4313,
+      "step": 2421
+    },
+    {
+      "epoch": 0.8111185532484929,
+      "grad_norm": 0.4597252905368805,
+      "learning_rate": 0.0001,
+      "loss": 1.5051,
+      "step": 2422
+    },
+    {
+      "epoch": 0.8114534494306765,
+      "grad_norm": 0.4557250738143921,
+      "learning_rate": 0.0001,
+      "loss": 1.5233,
+      "step": 2423
+    },
+    {
+      "epoch": 0.8117883456128601,
+      "grad_norm": 0.46700024604797363,
+      "learning_rate": 0.0001,
+      "loss": 1.5103,
+      "step": 2424
+    },
+    {
+      "epoch": 0.8121232417950436,
+      "grad_norm": 0.46530982851982117,
+      "learning_rate": 0.0001,
+      "loss": 1.4654,
+      "step": 2425
+    },
+    {
+      "epoch": 0.8124581379772271,
+      "grad_norm": 0.4776183068752289,
+      "learning_rate": 0.0001,
+      "loss": 1.4358,
+      "step": 2426
+    },
+    {
+      "epoch": 0.8127930341594106,
+      "grad_norm": 0.4590524733066559,
+      "learning_rate": 0.0001,
+      "loss": 1.5037,
+      "step": 2427
+    },
+    {
+      "epoch": 0.8131279303415941,
+      "grad_norm": 0.47458121180534363,
+      "learning_rate": 0.0001,
+      "loss": 1.5376,
+      "step": 2428
+    },
+    {
+      "epoch": 0.8134628265237777,
+      "grad_norm": 0.4598158001899719,
+      "learning_rate": 0.0001,
+      "loss": 1.5651,
+      "step": 2429
+    },
+    {
+      "epoch": 0.8137977227059612,
+      "grad_norm": 0.4738381505012512,
+      "learning_rate": 0.0001,
+      "loss": 1.5423,
+      "step": 2430
+    },
+    {
+      "epoch": 0.8141326188881447,
+      "grad_norm": 0.5065074563026428,
+      "learning_rate": 0.0001,
+      "loss": 1.554,
+      "step": 2431
+    },
+    {
+      "epoch": 0.8144675150703282,
+      "grad_norm": 0.45927509665489197,
+      "learning_rate": 0.0001,
+      "loss": 1.5128,
+      "step": 2432
+    },
+    {
+      "epoch": 0.8148024112525117,
+      "grad_norm": 0.48357442021369934,
+      "learning_rate": 0.0001,
+      "loss": 1.4922,
+      "step": 2433
+    },
+    {
+      "epoch": 0.8151373074346953,
+      "grad_norm": 0.47746044397354126,
+      "learning_rate": 0.0001,
+      "loss": 1.5573,
+      "step": 2434
+    },
+    {
+      "epoch": 0.8154722036168788,
+      "grad_norm": 0.4867241382598877,
+      "learning_rate": 0.0001,
+      "loss": 1.4885,
+      "step": 2435
+    },
+    {
+      "epoch": 0.8158070997990623,
+      "grad_norm": 0.4920485317707062,
+      "learning_rate": 0.0001,
+      "loss": 1.5621,
+      "step": 2436
+    },
+    {
+      "epoch": 0.8161419959812458,
+      "grad_norm": 0.473763644695282,
+      "learning_rate": 0.0001,
+      "loss": 1.5618,
+      "step": 2437
+    },
+    {
+      "epoch": 0.8164768921634293,
+      "grad_norm": 0.4919544458389282,
+      "learning_rate": 0.0001,
+      "loss": 1.4857,
+      "step": 2438
+    },
+    {
+      "epoch": 0.8168117883456129,
+      "grad_norm": 0.47004011273384094,
+      "learning_rate": 0.0001,
+      "loss": 1.4635,
+      "step": 2439
+    },
+    {
+      "epoch": 0.8171466845277964,
+      "grad_norm": 0.4616757035255432,
+      "learning_rate": 0.0001,
+      "loss": 1.5002,
+      "step": 2440
+    },
+    {
+      "epoch": 0.8174815807099799,
+      "grad_norm": 0.49343329668045044,
+      "learning_rate": 0.0001,
+      "loss": 1.6058,
+      "step": 2441
+    },
+    {
+      "epoch": 0.8178164768921634,
+      "grad_norm": 0.47879213094711304,
+      "learning_rate": 0.0001,
+      "loss": 1.6205,
+      "step": 2442
+    },
+    {
+      "epoch": 0.8181513730743469,
+      "grad_norm": 0.4748070538043976,
+      "learning_rate": 0.0001,
+      "loss": 1.4841,
+      "step": 2443
+    },
+    {
+      "epoch": 0.8184862692565305,
+      "grad_norm": 0.44889041781425476,
+      "learning_rate": 0.0001,
+      "loss": 1.3804,
+      "step": 2444
+    },
+    {
+      "epoch": 0.818821165438714,
+      "grad_norm": 0.47727763652801514,
+      "learning_rate": 0.0001,
+      "loss": 1.5209,
+      "step": 2445
+    },
+    {
+      "epoch": 0.8191560616208975,
+      "grad_norm": 0.4846794307231903,
+      "learning_rate": 0.0001,
+      "loss": 1.5644,
+      "step": 2446
+    },
+    {
+      "epoch": 0.819490957803081,
+      "grad_norm": 0.47240149974823,
+      "learning_rate": 0.0001,
+      "loss": 1.4015,
+      "step": 2447
+    },
+    {
+      "epoch": 0.8198258539852645,
+      "grad_norm": 0.4686112701892853,
+      "learning_rate": 0.0001,
+      "loss": 1.5246,
+      "step": 2448
+    },
+    {
+      "epoch": 0.820160750167448,
+      "grad_norm": 0.4680188000202179,
+      "learning_rate": 0.0001,
+      "loss": 1.4998,
+      "step": 2449
+    },
+    {
+      "epoch": 0.8204956463496316,
+      "grad_norm": 0.4550737142562866,
+      "learning_rate": 0.0001,
+      "loss": 1.5079,
+      "step": 2450
+    },
+    {
+      "epoch": 0.8208305425318151,
+      "grad_norm": 0.4652599096298218,
+      "learning_rate": 0.0001,
+      "loss": 1.4914,
+      "step": 2451
+    },
+    {
+      "epoch": 0.8211654387139986,
+      "grad_norm": 0.46419402956962585,
+      "learning_rate": 0.0001,
+      "loss": 1.5244,
+      "step": 2452
+    },
+    {
+      "epoch": 0.8215003348961822,
+      "grad_norm": 0.4659050703048706,
+      "learning_rate": 0.0001,
+      "loss": 1.5253,
+      "step": 2453
+    },
+    {
+      "epoch": 0.8218352310783658,
+      "grad_norm": 0.4788404405117035,
+      "learning_rate": 0.0001,
+      "loss": 1.6053,
+      "step": 2454
+    },
+    {
+      "epoch": 0.8221701272605493,
+      "grad_norm": 0.4813288450241089,
+      "learning_rate": 0.0001,
+      "loss": 1.5119,
+      "step": 2455
+    },
+    {
+      "epoch": 0.8225050234427328,
+      "grad_norm": 0.47103598713874817,
+      "learning_rate": 0.0001,
+      "loss": 1.4902,
+      "step": 2456
+    },
+    {
+      "epoch": 0.8228399196249163,
+      "grad_norm": 0.47177964448928833,
+      "learning_rate": 0.0001,
+      "loss": 1.5243,
+      "step": 2457
+    },
+    {
+      "epoch": 0.8231748158070998,
+      "grad_norm": 0.4908981919288635,
+      "learning_rate": 0.0001,
+      "loss": 1.5389,
+      "step": 2458
+    },
+    {
+      "epoch": 0.8235097119892834,
+      "grad_norm": 0.46003222465515137,
+      "learning_rate": 0.0001,
+      "loss": 1.4344,
+      "step": 2459
+    },
+    {
+      "epoch": 0.8238446081714669,
+      "grad_norm": 0.4566117227077484,
+      "learning_rate": 0.0001,
+      "loss": 1.4996,
+      "step": 2460
+    },
+    {
+      "epoch": 0.8241795043536504,
+      "grad_norm": 0.47622841596603394,
+      "learning_rate": 0.0001,
+      "loss": 1.5919,
+      "step": 2461
+    },
+    {
+      "epoch": 0.8245144005358339,
+      "grad_norm": 0.4662770926952362,
+      "learning_rate": 0.0001,
+      "loss": 1.4792,
+      "step": 2462
+    },
+    {
+      "epoch": 0.8248492967180174,
+      "grad_norm": 0.44657108187675476,
+      "learning_rate": 0.0001,
+      "loss": 1.4932,
+      "step": 2463
+    },
+    {
+      "epoch": 0.825184192900201,
+      "grad_norm": 0.46892106533050537,
+      "learning_rate": 0.0001,
+      "loss": 1.5541,
+      "step": 2464
+    },
+    {
+      "epoch": 0.8255190890823845,
+      "grad_norm": 0.47214871644973755,
+      "learning_rate": 0.0001,
+      "loss": 1.5707,
+      "step": 2465
+    },
+    {
+      "epoch": 0.825853985264568,
+      "grad_norm": 0.45614370703697205,
+      "learning_rate": 0.0001,
+      "loss": 1.5309,
+      "step": 2466
+    },
+    {
+      "epoch": 0.8261888814467515,
+      "grad_norm": 0.46787700057029724,
+      "learning_rate": 0.0001,
+      "loss": 1.4708,
+      "step": 2467
+    },
+    {
+      "epoch": 0.826523777628935,
+      "grad_norm": 0.466823548078537,
+      "learning_rate": 0.0001,
+      "loss": 1.5455,
+      "step": 2468
+    },
+    {
+      "epoch": 0.8268586738111185,
+      "grad_norm": 0.4862380027770996,
+      "learning_rate": 0.0001,
+      "loss": 1.539,
+      "step": 2469
+    },
+    {
+      "epoch": 0.8271935699933021,
+      "grad_norm": 0.45864713191986084,
+      "learning_rate": 0.0001,
+      "loss": 1.4956,
+      "step": 2470
+    },
+    {
+      "epoch": 0.8275284661754856,
+      "grad_norm": 0.4562831521034241,
+      "learning_rate": 0.0001,
+      "loss": 1.467,
+      "step": 2471
+    },
+    {
+      "epoch": 0.8278633623576691,
+      "grad_norm": 0.4680713415145874,
+      "learning_rate": 0.0001,
+      "loss": 1.5045,
+      "step": 2472
+    },
+    {
+      "epoch": 0.8281982585398526,
+      "grad_norm": 0.4776260554790497,
+      "learning_rate": 0.0001,
+      "loss": 1.4162,
+      "step": 2473
+    },
+    {
+      "epoch": 0.8285331547220361,
+      "grad_norm": 0.4711471199989319,
+      "learning_rate": 0.0001,
+      "loss": 1.5081,
+      "step": 2474
+    },
+    {
+      "epoch": 0.8288680509042197,
+      "grad_norm": 0.4526406526565552,
+      "learning_rate": 0.0001,
+      "loss": 1.5155,
+      "step": 2475
+    },
+    {
+      "epoch": 0.8292029470864032,
+      "grad_norm": 0.48618829250335693,
+      "learning_rate": 0.0001,
+      "loss": 1.5989,
+      "step": 2476
+    },
+    {
+      "epoch": 0.8295378432685867,
+      "grad_norm": 0.4556841552257538,
+      "learning_rate": 0.0001,
+      "loss": 1.4377,
+      "step": 2477
+    },
+    {
+      "epoch": 0.8298727394507702,
+      "grad_norm": 0.47651731967926025,
+      "learning_rate": 0.0001,
+      "loss": 1.4536,
+      "step": 2478
+    },
+    {
+      "epoch": 0.8302076356329537,
+      "grad_norm": 0.4996659755706787,
+      "learning_rate": 0.0001,
+      "loss": 1.6023,
+      "step": 2479
+    },
+    {
+      "epoch": 0.8305425318151373,
+      "grad_norm": 0.46138516068458557,
+      "learning_rate": 0.0001,
+      "loss": 1.4767,
+      "step": 2480
+    },
+    {
+      "epoch": 0.8308774279973208,
+      "grad_norm": 0.4543902277946472,
+      "learning_rate": 0.0001,
+      "loss": 1.5485,
+      "step": 2481
+    },
+    {
+      "epoch": 0.8312123241795043,
+      "grad_norm": 0.4841060936450958,
+      "learning_rate": 0.0001,
+      "loss": 1.4308,
+      "step": 2482
+    },
+    {
+      "epoch": 0.8315472203616878,
+      "grad_norm": 0.4554094076156616,
+      "learning_rate": 0.0001,
+      "loss": 1.4633,
+      "step": 2483
+    },
+    {
+      "epoch": 0.8318821165438715,
+      "grad_norm": 0.4548550844192505,
+      "learning_rate": 0.0001,
+      "loss": 1.4629,
+      "step": 2484
+    },
+    {
+      "epoch": 0.832217012726055,
+      "grad_norm": 0.4791390299797058,
+      "learning_rate": 0.0001,
+      "loss": 1.5586,
+      "step": 2485
+    },
+    {
+      "epoch": 0.8325519089082385,
+      "grad_norm": 0.4800601005554199,
+      "learning_rate": 0.0001,
+      "loss": 1.5139,
+      "step": 2486
+    },
+    {
+      "epoch": 0.832886805090422,
+      "grad_norm": 0.4501197040081024,
+      "learning_rate": 0.0001,
+      "loss": 1.4459,
+      "step": 2487
+    },
+    {
+      "epoch": 0.8332217012726055,
+      "grad_norm": 0.4678644835948944,
+      "learning_rate": 0.0001,
+      "loss": 1.5499,
+      "step": 2488
+    },
+    {
+      "epoch": 0.833556597454789,
+      "grad_norm": 0.47049519419670105,
+      "learning_rate": 0.0001,
+      "loss": 1.5105,
+      "step": 2489
+    },
+    {
+      "epoch": 0.8338914936369726,
+      "grad_norm": 0.4660375416278839,
+      "learning_rate": 0.0001,
+      "loss": 1.514,
+      "step": 2490
+    },
+    {
+      "epoch": 0.8342263898191561,
+      "grad_norm": 0.4684686064720154,
+      "learning_rate": 0.0001,
+      "loss": 1.4937,
+      "step": 2491
+    },
+    {
+      "epoch": 0.8345612860013396,
+      "grad_norm": 0.47460314631462097,
+      "learning_rate": 0.0001,
+      "loss": 1.5526,
+      "step": 2492
+    },
+    {
+      "epoch": 0.8348961821835231,
+      "grad_norm": 0.49566179513931274,
+      "learning_rate": 0.0001,
+      "loss": 1.5733,
+      "step": 2493
+    },
+    {
+      "epoch": 0.8352310783657066,
+      "grad_norm": 0.4774349331855774,
+      "learning_rate": 0.0001,
+      "loss": 1.5573,
+      "step": 2494
+    },
+    {
+      "epoch": 0.8355659745478902,
+      "grad_norm": 0.5103961825370789,
+      "learning_rate": 0.0001,
+      "loss": 1.6416,
+      "step": 2495
+    },
+    {
+      "epoch": 0.8359008707300737,
+      "grad_norm": 0.47637876868247986,
+      "learning_rate": 0.0001,
+      "loss": 1.5274,
+      "step": 2496
+    },
+    {
+      "epoch": 0.8362357669122572,
+      "grad_norm": 0.4631972908973694,
+      "learning_rate": 0.0001,
+      "loss": 1.4233,
+      "step": 2497
+    },
+    {
+      "epoch": 0.8365706630944407,
+      "grad_norm": 0.4680608808994293,
+      "learning_rate": 0.0001,
+      "loss": 1.5582,
+      "step": 2498
+    },
+    {
+      "epoch": 0.8369055592766242,
+      "grad_norm": 0.46085795760154724,
+      "learning_rate": 0.0001,
+      "loss": 1.5227,
+      "step": 2499
+    },
+    {
+      "epoch": 0.8372404554588078,
+      "grad_norm": 0.47325894236564636,
+      "learning_rate": 0.0001,
+      "loss": 1.4911,
+      "step": 2500
+    },
+    {
+      "epoch": 0.8375753516409913,
+      "grad_norm": 0.45909950137138367,
+      "learning_rate": 0.0001,
+      "loss": 1.4819,
+      "step": 2501
+    },
+    {
+      "epoch": 0.8379102478231748,
+      "grad_norm": 0.43451809883117676,
+      "learning_rate": 0.0001,
+      "loss": 1.3889,
+      "step": 2502
+    },
+    {
+      "epoch": 0.8382451440053583,
+      "grad_norm": 0.4574137330055237,
+      "learning_rate": 0.0001,
+      "loss": 1.4686,
+      "step": 2503
+    },
+    {
+      "epoch": 0.8385800401875418,
+      "grad_norm": 0.48103952407836914,
+      "learning_rate": 0.0001,
+      "loss": 1.5808,
+      "step": 2504
+    },
+    {
+      "epoch": 0.8389149363697254,
+      "grad_norm": 0.5158580541610718,
+      "learning_rate": 0.0001,
+      "loss": 1.5612,
+      "step": 2505
+    },
+    {
+      "epoch": 0.8392498325519089,
+      "grad_norm": 0.4528539776802063,
+      "learning_rate": 0.0001,
+      "loss": 1.4579,
+      "step": 2506
+    },
+    {
+      "epoch": 0.8395847287340924,
+      "grad_norm": 0.5151484608650208,
+      "learning_rate": 0.0001,
+      "loss": 1.5334,
+      "step": 2507
+    },
+    {
+      "epoch": 0.8399196249162759,
+      "grad_norm": 0.4524150490760803,
+      "learning_rate": 0.0001,
+      "loss": 1.4369,
+      "step": 2508
+    },
+    {
+      "epoch": 0.8402545210984594,
+      "grad_norm": 0.4746531844139099,
+      "learning_rate": 0.0001,
+      "loss": 1.4458,
+      "step": 2509
+    },
+    {
+      "epoch": 0.840589417280643,
+      "grad_norm": 0.4763810336589813,
+      "learning_rate": 0.0001,
+      "loss": 1.55,
+      "step": 2510
+    },
+    {
+      "epoch": 0.8409243134628265,
+      "grad_norm": 0.4495566189289093,
+      "learning_rate": 0.0001,
+      "loss": 1.5156,
+      "step": 2511
+    },
+    {
+      "epoch": 0.84125920964501,
+      "grad_norm": 0.5042752623558044,
+      "learning_rate": 0.0001,
+      "loss": 1.5954,
+      "step": 2512
+    },
+    {
+      "epoch": 0.8415941058271935,
+      "grad_norm": 0.4725266396999359,
+      "learning_rate": 0.0001,
+      "loss": 1.459,
+      "step": 2513
+    },
+    {
+      "epoch": 0.8419290020093771,
+      "grad_norm": 0.48124271631240845,
+      "learning_rate": 0.0001,
+      "loss": 1.6029,
+      "step": 2514
+    },
+    {
+      "epoch": 0.8422638981915607,
+      "grad_norm": 0.46474096179008484,
+      "learning_rate": 0.0001,
+      "loss": 1.5342,
+      "step": 2515
+    },
+    {
+      "epoch": 0.8425987943737442,
+      "grad_norm": 0.48470795154571533,
+      "learning_rate": 0.0001,
+      "loss": 1.4792,
+      "step": 2516
+    },
+    {
+      "epoch": 0.8429336905559277,
+      "grad_norm": 0.5086537599563599,
+      "learning_rate": 0.0001,
+      "loss": 1.578,
+      "step": 2517
+    },
+    {
+      "epoch": 0.8432685867381112,
+      "grad_norm": 0.4417508542537689,
+      "learning_rate": 0.0001,
+      "loss": 1.4967,
+      "step": 2518
+    },
+    {
+      "epoch": 0.8436034829202947,
+      "grad_norm": 0.474541038274765,
+      "learning_rate": 0.0001,
+      "loss": 1.5432,
+      "step": 2519
+    },
+    {
+      "epoch": 0.8439383791024783,
+      "grad_norm": 0.46175190806388855,
+      "learning_rate": 0.0001,
+      "loss": 1.4789,
+      "step": 2520
+    },
+    {
+      "epoch": 0.8442732752846618,
+      "grad_norm": 0.489646852016449,
+      "learning_rate": 0.0001,
+      "loss": 1.5607,
+      "step": 2521
+    },
+    {
+      "epoch": 0.8446081714668453,
+      "grad_norm": 0.4702965319156647,
+      "learning_rate": 0.0001,
+      "loss": 1.5672,
+      "step": 2522
+    },
+    {
+      "epoch": 0.8449430676490288,
+      "grad_norm": 0.47818228602409363,
+      "learning_rate": 0.0001,
+      "loss": 1.5023,
+      "step": 2523
+    },
+    {
+      "epoch": 0.8452779638312123,
+      "grad_norm": 0.4713198244571686,
+      "learning_rate": 0.0001,
+      "loss": 1.5231,
+      "step": 2524
+    },
+    {
+      "epoch": 0.8456128600133959,
+      "grad_norm": 0.4650212824344635,
+      "learning_rate": 0.0001,
+      "loss": 1.4677,
+      "step": 2525
+    },
+    {
+      "epoch": 0.8459477561955794,
+      "grad_norm": 0.47101250290870667,
+      "learning_rate": 0.0001,
+      "loss": 1.5705,
+      "step": 2526
+    },
+    {
+      "epoch": 0.8462826523777629,
+      "grad_norm": 0.4593891203403473,
+      "learning_rate": 0.0001,
+      "loss": 1.5138,
+      "step": 2527
+    },
+    {
+      "epoch": 0.8466175485599464,
+      "grad_norm": 0.4726957678794861,
+      "learning_rate": 0.0001,
+      "loss": 1.5812,
+      "step": 2528
+    },
+    {
+      "epoch": 0.8469524447421299,
+      "grad_norm": 0.45414677262306213,
+      "learning_rate": 0.0001,
+      "loss": 1.4513,
+      "step": 2529
+    },
+    {
+      "epoch": 0.8472873409243135,
+      "grad_norm": 0.46414148807525635,
+      "learning_rate": 0.0001,
+      "loss": 1.5247,
+      "step": 2530
+    },
+    {
+      "epoch": 0.847622237106497,
+      "grad_norm": 0.4923802614212036,
+      "learning_rate": 0.0001,
+      "loss": 1.5454,
+      "step": 2531
+    },
+    {
+      "epoch": 0.8479571332886805,
+      "grad_norm": 0.4653525650501251,
+      "learning_rate": 0.0001,
+      "loss": 1.4374,
+      "step": 2532
+    },
+    {
+      "epoch": 0.848292029470864,
+      "grad_norm": 0.49471646547317505,
+      "learning_rate": 0.0001,
+      "loss": 1.6124,
+      "step": 2533
+    },
+    {
+      "epoch": 0.8486269256530475,
+      "grad_norm": 0.48391300439834595,
+      "learning_rate": 0.0001,
+      "loss": 1.5271,
+      "step": 2534
+    },
+    {
+      "epoch": 0.8489618218352311,
+      "grad_norm": 0.4469664990901947,
+      "learning_rate": 0.0001,
+      "loss": 1.4229,
+      "step": 2535
+    },
+    {
+      "epoch": 0.8492967180174146,
+      "grad_norm": 0.48336634039878845,
+      "learning_rate": 0.0001,
+      "loss": 1.4498,
+      "step": 2536
+    },
+    {
+      "epoch": 0.8496316141995981,
+      "grad_norm": 0.48580068349838257,
+      "learning_rate": 0.0001,
+      "loss": 1.5695,
+      "step": 2537
+    },
+    {
+      "epoch": 0.8499665103817816,
+      "grad_norm": 0.45114821195602417,
+      "learning_rate": 0.0001,
+      "loss": 1.4471,
+      "step": 2538
+    },
+    {
+      "epoch": 0.8503014065639651,
+      "grad_norm": 0.47201669216156006,
+      "learning_rate": 0.0001,
+      "loss": 1.5152,
+      "step": 2539
+    },
+    {
+      "epoch": 0.8506363027461487,
+      "grad_norm": 0.49140065908432007,
+      "learning_rate": 0.0001,
+      "loss": 1.4929,
+      "step": 2540
+    },
+    {
+      "epoch": 0.8509711989283322,
+      "grad_norm": 0.511402428150177,
+      "learning_rate": 0.0001,
+      "loss": 1.6764,
+      "step": 2541
+    },
+    {
+      "epoch": 0.8513060951105157,
+      "grad_norm": 0.46448859572410583,
+      "learning_rate": 0.0001,
+      "loss": 1.3813,
+      "step": 2542
+    },
+    {
+      "epoch": 0.8516409912926992,
+      "grad_norm": 0.46996361017227173,
+      "learning_rate": 0.0001,
+      "loss": 1.5166,
+      "step": 2543
+    },
+    {
+      "epoch": 0.8519758874748827,
+      "grad_norm": 0.482006698846817,
+      "learning_rate": 0.0001,
+      "loss": 1.6143,
+      "step": 2544
+    },
+    {
+      "epoch": 0.8523107836570664,
+      "grad_norm": 0.46881818771362305,
+      "learning_rate": 0.0001,
+      "loss": 1.501,
+      "step": 2545
+    },
+    {
+      "epoch": 0.8526456798392499,
+      "grad_norm": 0.45807763934135437,
+      "learning_rate": 0.0001,
+      "loss": 1.4859,
+      "step": 2546
+    },
+    {
+      "epoch": 0.8529805760214334,
+      "grad_norm": 0.4654110372066498,
+      "learning_rate": 0.0001,
+      "loss": 1.4679,
+      "step": 2547
+    },
+    {
+      "epoch": 0.8533154722036169,
+      "grad_norm": 0.47218427062034607,
+      "learning_rate": 0.0001,
+      "loss": 1.5526,
+      "step": 2548
+    },
+    {
+      "epoch": 0.8536503683858004,
+      "grad_norm": 0.4757080674171448,
+      "learning_rate": 0.0001,
+      "loss": 1.5045,
+      "step": 2549
+    },
+    {
+      "epoch": 0.853985264567984,
+      "grad_norm": 0.4587678909301758,
+      "learning_rate": 0.0001,
+      "loss": 1.5146,
+      "step": 2550
+    },
+    {
+      "epoch": 0.8543201607501675,
+      "grad_norm": 0.49881237745285034,
+      "learning_rate": 0.0001,
+      "loss": 1.5074,
+      "step": 2551
+    },
+    {
+      "epoch": 0.854655056932351,
+      "grad_norm": 0.4800609350204468,
+      "learning_rate": 0.0001,
+      "loss": 1.5654,
+      "step": 2552
+    },
+    {
+      "epoch": 0.8549899531145345,
+      "grad_norm": 0.4712275564670563,
+      "learning_rate": 0.0001,
+      "loss": 1.5084,
+      "step": 2553
+    },
+    {
+      "epoch": 0.855324849296718,
+      "grad_norm": 0.4913301467895508,
+      "learning_rate": 0.0001,
+      "loss": 1.4832,
+      "step": 2554
+    },
+    {
+      "epoch": 0.8556597454789016,
+      "grad_norm": 0.481899619102478,
+      "learning_rate": 0.0001,
+      "loss": 1.4984,
+      "step": 2555
+    },
+    {
+      "epoch": 0.8559946416610851,
+      "grad_norm": 0.4633455276489258,
+      "learning_rate": 0.0001,
+      "loss": 1.4273,
+      "step": 2556
+    },
+    {
+      "epoch": 0.8563295378432686,
+      "grad_norm": 0.48490163683891296,
+      "learning_rate": 0.0001,
+      "loss": 1.5484,
+      "step": 2557
+    },
+    {
+      "epoch": 0.8566644340254521,
+      "grad_norm": 0.47869372367858887,
+      "learning_rate": 0.0001,
+      "loss": 1.5389,
+      "step": 2558
+    },
+    {
+      "epoch": 0.8569993302076356,
+      "grad_norm": 0.4712234437465668,
+      "learning_rate": 0.0001,
+      "loss": 1.4553,
+      "step": 2559
+    },
+    {
+      "epoch": 0.8573342263898192,
+      "grad_norm": 0.48260584473609924,
+      "learning_rate": 0.0001,
+      "loss": 1.5749,
+      "step": 2560
+    },
+    {
+      "epoch": 0.8576691225720027,
+      "grad_norm": 0.4660404622554779,
+      "learning_rate": 0.0001,
+      "loss": 1.512,
+      "step": 2561
+    },
+    {
+      "epoch": 0.8580040187541862,
+      "grad_norm": 0.47612521052360535,
+      "learning_rate": 0.0001,
+      "loss": 1.4752,
+      "step": 2562
+    },
+    {
+      "epoch": 0.8583389149363697,
+      "grad_norm": 0.4593373239040375,
+      "learning_rate": 0.0001,
+      "loss": 1.5109,
+      "step": 2563
+    },
+    {
+      "epoch": 0.8586738111185532,
+      "grad_norm": 0.4994997978210449,
+      "learning_rate": 0.0001,
+      "loss": 1.6041,
+      "step": 2564
+    },
+    {
+      "epoch": 0.8590087073007368,
+      "grad_norm": 0.46223050355911255,
+      "learning_rate": 0.0001,
+      "loss": 1.4896,
+      "step": 2565
+    },
+    {
+      "epoch": 0.8593436034829203,
+      "grad_norm": 0.48477256298065186,
+      "learning_rate": 0.0001,
+      "loss": 1.5081,
+      "step": 2566
+    },
+    {
+      "epoch": 0.8596784996651038,
+      "grad_norm": 0.4974391758441925,
+      "learning_rate": 0.0001,
+      "loss": 1.6294,
+      "step": 2567
+    },
+    {
+      "epoch": 0.8600133958472873,
+      "grad_norm": 0.48786434531211853,
+      "learning_rate": 0.0001,
+      "loss": 1.4691,
+      "step": 2568
+    },
+    {
+      "epoch": 0.8603482920294708,
+      "grad_norm": 0.4549929201602936,
+      "learning_rate": 0.0001,
+      "loss": 1.4901,
+      "step": 2569
+    },
+    {
+      "epoch": 0.8606831882116543,
+      "grad_norm": 0.5011422634124756,
+      "learning_rate": 0.0001,
+      "loss": 1.5825,
+      "step": 2570
+    },
+    {
+      "epoch": 0.8610180843938379,
+      "grad_norm": 0.475400447845459,
+      "learning_rate": 0.0001,
+      "loss": 1.5564,
+      "step": 2571
+    },
+    {
+      "epoch": 0.8613529805760214,
+      "grad_norm": 0.4625644087791443,
+      "learning_rate": 0.0001,
+      "loss": 1.4707,
+      "step": 2572
+    },
+    {
+      "epoch": 0.8616878767582049,
+      "grad_norm": 0.46159297227859497,
+      "learning_rate": 0.0001,
+      "loss": 1.4398,
+      "step": 2573
+    },
+    {
+      "epoch": 0.8620227729403884,
+      "grad_norm": 0.48057782649993896,
+      "learning_rate": 0.0001,
+      "loss": 1.5164,
+      "step": 2574
+    },
+    {
+      "epoch": 0.8623576691225721,
+      "grad_norm": 0.45620816946029663,
+      "learning_rate": 0.0001,
+      "loss": 1.5166,
+      "step": 2575
+    },
+    {
+      "epoch": 0.8626925653047556,
+      "grad_norm": 0.4661911427974701,
+      "learning_rate": 0.0001,
+      "loss": 1.4998,
+      "step": 2576
+    },
+    {
+      "epoch": 0.8630274614869391,
+      "grad_norm": 0.470892995595932,
+      "learning_rate": 0.0001,
+      "loss": 1.5692,
+      "step": 2577
+    },
+    {
+      "epoch": 0.8633623576691226,
+      "grad_norm": 0.47721078991889954,
+      "learning_rate": 0.0001,
+      "loss": 1.558,
+      "step": 2578
+    },
+    {
+      "epoch": 0.8636972538513061,
+      "grad_norm": 0.5240944027900696,
+      "learning_rate": 0.0001,
+      "loss": 1.5675,
+      "step": 2579
+    },
+    {
+      "epoch": 0.8640321500334897,
+      "grad_norm": 0.48352697491645813,
+      "learning_rate": 0.0001,
+      "loss": 1.513,
+      "step": 2580
+    },
+    {
+      "epoch": 0.8643670462156732,
+      "grad_norm": 0.47838255763053894,
+      "learning_rate": 0.0001,
+      "loss": 1.4973,
+      "step": 2581
+    },
+    {
+      "epoch": 0.8647019423978567,
+      "grad_norm": 0.4820500314235687,
+      "learning_rate": 0.0001,
+      "loss": 1.5632,
+      "step": 2582
+    },
+    {
+      "epoch": 0.8650368385800402,
+      "grad_norm": 0.44904258847236633,
+      "learning_rate": 0.0001,
+      "loss": 1.464,
+      "step": 2583
+    },
+    {
+      "epoch": 0.8653717347622237,
+      "grad_norm": 0.4510500133037567,
+      "learning_rate": 0.0001,
+      "loss": 1.4805,
+      "step": 2584
+    },
+    {
+      "epoch": 0.8657066309444073,
+      "grad_norm": 0.48458313941955566,
+      "learning_rate": 0.0001,
+      "loss": 1.5889,
+      "step": 2585
+    },
+    {
+      "epoch": 0.8660415271265908,
+      "grad_norm": 0.4644976258277893,
+      "learning_rate": 0.0001,
+      "loss": 1.5072,
+      "step": 2586
+    },
+    {
+      "epoch": 0.8663764233087743,
+      "grad_norm": 0.47620436549186707,
+      "learning_rate": 0.0001,
+      "loss": 1.4598,
+      "step": 2587
+    },
+    {
+      "epoch": 0.8667113194909578,
+      "grad_norm": 0.4634052813053131,
+      "learning_rate": 0.0001,
+      "loss": 1.4223,
+      "step": 2588
+    },
+    {
+      "epoch": 0.8670462156731413,
+      "grad_norm": 0.48688220977783203,
+      "learning_rate": 0.0001,
+      "loss": 1.6314,
+      "step": 2589
+    },
+    {
+      "epoch": 0.8673811118553248,
+      "grad_norm": 0.4728447198867798,
+      "learning_rate": 0.0001,
+      "loss": 1.5439,
+      "step": 2590
+    },
+    {
+      "epoch": 0.8677160080375084,
+      "grad_norm": 0.4576449692249298,
+      "learning_rate": 0.0001,
+      "loss": 1.493,
+      "step": 2591
+    },
+    {
+      "epoch": 0.8680509042196919,
+      "grad_norm": 0.47686463594436646,
+      "learning_rate": 0.0001,
+      "loss": 1.5732,
+      "step": 2592
+    },
+    {
+      "epoch": 0.8683858004018754,
+      "grad_norm": 0.49087628722190857,
+      "learning_rate": 0.0001,
+      "loss": 1.4753,
+      "step": 2593
+    },
+    {
+      "epoch": 0.8687206965840589,
+      "grad_norm": 0.45120036602020264,
+      "learning_rate": 0.0001,
+      "loss": 1.4855,
+      "step": 2594
+    },
+    {
+      "epoch": 0.8690555927662424,
+      "grad_norm": 0.458211749792099,
+      "learning_rate": 0.0001,
+      "loss": 1.5443,
+      "step": 2595
+    },
+    {
+      "epoch": 0.869390488948426,
+      "grad_norm": 0.4843875765800476,
+      "learning_rate": 0.0001,
+      "loss": 1.4995,
+      "step": 2596
+    },
+    {
+      "epoch": 0.8697253851306095,
+      "grad_norm": 0.4657866358757019,
+      "learning_rate": 0.0001,
+      "loss": 1.5781,
+      "step": 2597
+    },
+    {
+      "epoch": 0.870060281312793,
+      "grad_norm": 0.4589475691318512,
+      "learning_rate": 0.0001,
+      "loss": 1.4494,
+      "step": 2598
+    },
+    {
+      "epoch": 0.8703951774949765,
+      "grad_norm": 0.4740186035633087,
+      "learning_rate": 0.0001,
+      "loss": 1.6285,
+      "step": 2599
+    },
+    {
+      "epoch": 0.87073007367716,
+      "grad_norm": 0.49049800634384155,
+      "learning_rate": 0.0001,
+      "loss": 1.5859,
+      "step": 2600
+    },
+    {
+      "epoch": 0.8710649698593436,
+      "grad_norm": 0.48565539717674255,
+      "learning_rate": 0.0001,
+      "loss": 1.4981,
+      "step": 2601
+    },
+    {
+      "epoch": 0.8713998660415271,
+      "grad_norm": 0.4501047134399414,
+      "learning_rate": 0.0001,
+      "loss": 1.4426,
+      "step": 2602
+    },
+    {
+      "epoch": 0.8717347622237106,
+      "grad_norm": 0.4414789378643036,
+      "learning_rate": 0.0001,
+      "loss": 1.395,
+      "step": 2603
+    },
+    {
+      "epoch": 0.8720696584058941,
+      "grad_norm": 0.49444058537483215,
+      "learning_rate": 0.0001,
+      "loss": 1.4857,
+      "step": 2604
+    },
+    {
+      "epoch": 0.8724045545880776,
+      "grad_norm": 0.4695194661617279,
+      "learning_rate": 0.0001,
+      "loss": 1.4485,
+      "step": 2605
+    },
+    {
+      "epoch": 0.8727394507702613,
+      "grad_norm": 0.49149829149246216,
+      "learning_rate": 0.0001,
+      "loss": 1.5755,
+      "step": 2606
+    },
+    {
+      "epoch": 0.8730743469524448,
+      "grad_norm": 0.4663974940776825,
+      "learning_rate": 0.0001,
+      "loss": 1.4947,
+      "step": 2607
+    },
+    {
+      "epoch": 0.8734092431346283,
+      "grad_norm": 0.4728209972381592,
+      "learning_rate": 0.0001,
+      "loss": 1.3464,
+      "step": 2608
+    },
+    {
+      "epoch": 0.8737441393168118,
+      "grad_norm": 0.47123968601226807,
+      "learning_rate": 0.0001,
+      "loss": 1.4939,
+      "step": 2609
+    },
+    {
+      "epoch": 0.8740790354989953,
+      "grad_norm": 0.5068251490592957,
+      "learning_rate": 0.0001,
+      "loss": 1.5115,
+      "step": 2610
+    },
+    {
+      "epoch": 0.8744139316811789,
+      "grad_norm": 0.4849814474582672,
+      "learning_rate": 0.0001,
+      "loss": 1.532,
+      "step": 2611
+    },
+    {
+      "epoch": 0.8747488278633624,
+      "grad_norm": 0.5009174346923828,
+      "learning_rate": 0.0001,
+      "loss": 1.6127,
+      "step": 2612
+    },
+    {
+      "epoch": 0.8750837240455459,
+      "grad_norm": 0.46775639057159424,
+      "learning_rate": 0.0001,
+      "loss": 1.4938,
+      "step": 2613
+    },
+    {
+      "epoch": 0.8754186202277294,
+      "grad_norm": 0.4642925560474396,
+      "learning_rate": 0.0001,
+      "loss": 1.5143,
+      "step": 2614
+    },
+    {
+      "epoch": 0.875753516409913,
+      "grad_norm": 0.5046592354774475,
+      "learning_rate": 0.0001,
+      "loss": 1.5009,
+      "step": 2615
+    },
+    {
+      "epoch": 0.8760884125920965,
+      "grad_norm": 0.44560351967811584,
+      "learning_rate": 0.0001,
+      "loss": 1.464,
+      "step": 2616
+    },
+    {
+      "epoch": 0.87642330877428,
+      "grad_norm": 0.4847624599933624,
+      "learning_rate": 0.0001,
+      "loss": 1.5868,
+      "step": 2617
+    },
+    {
+      "epoch": 0.8767582049564635,
+      "grad_norm": 0.4620243310928345,
+      "learning_rate": 0.0001,
+      "loss": 1.478,
+      "step": 2618
+    },
+    {
+      "epoch": 0.877093101138647,
+      "grad_norm": 0.48844829201698303,
+      "learning_rate": 0.0001,
+      "loss": 1.4811,
+      "step": 2619
+    },
+    {
+      "epoch": 0.8774279973208305,
+      "grad_norm": 0.49304163455963135,
+      "learning_rate": 0.0001,
+      "loss": 1.5897,
+      "step": 2620
+    },
+    {
+      "epoch": 0.8777628935030141,
+      "grad_norm": 0.48174530267715454,
+      "learning_rate": 0.0001,
+      "loss": 1.5173,
+      "step": 2621
+    },
+    {
+      "epoch": 0.8780977896851976,
+      "grad_norm": 0.47570034861564636,
+      "learning_rate": 0.0001,
+      "loss": 1.4714,
+      "step": 2622
+    },
+    {
+      "epoch": 0.8784326858673811,
+      "grad_norm": 0.4666658937931061,
+      "learning_rate": 0.0001,
+      "loss": 1.4931,
+      "step": 2623
+    },
+    {
+      "epoch": 0.8787675820495646,
+      "grad_norm": 0.49084752798080444,
+      "learning_rate": 0.0001,
+      "loss": 1.6196,
+      "step": 2624
+    },
+    {
+      "epoch": 0.8791024782317481,
+      "grad_norm": 0.4708821475505829,
+      "learning_rate": 0.0001,
+      "loss": 1.4557,
+      "step": 2625
+    },
+    {
+      "epoch": 0.8794373744139317,
+      "grad_norm": 0.4662223160266876,
+      "learning_rate": 0.0001,
+      "loss": 1.4757,
+      "step": 2626
+    },
+    {
+      "epoch": 0.8797722705961152,
+      "grad_norm": 0.4955317974090576,
+      "learning_rate": 0.0001,
+      "loss": 1.4234,
+      "step": 2627
+    },
+    {
+      "epoch": 0.8801071667782987,
+      "grad_norm": 0.4606055021286011,
+      "learning_rate": 0.0001,
+      "loss": 1.4818,
+      "step": 2628
+    },
+    {
+      "epoch": 0.8804420629604822,
+      "grad_norm": 0.4685207009315491,
+      "learning_rate": 0.0001,
+      "loss": 1.5082,
+      "step": 2629
+    },
+    {
+      "epoch": 0.8807769591426657,
+      "grad_norm": 0.4986380338668823,
+      "learning_rate": 0.0001,
+      "loss": 1.5582,
+      "step": 2630
+    },
+    {
+      "epoch": 0.8811118553248493,
+      "grad_norm": 0.4687456786632538,
+      "learning_rate": 0.0001,
+      "loss": 1.4779,
+      "step": 2631
+    },
+    {
+      "epoch": 0.8814467515070328,
+      "grad_norm": 0.4698921740055084,
+      "learning_rate": 0.0001,
+      "loss": 1.5505,
+      "step": 2632
+    },
+    {
+      "epoch": 0.8817816476892163,
+      "grad_norm": 0.46344539523124695,
+      "learning_rate": 0.0001,
+      "loss": 1.4528,
+      "step": 2633
+    },
+    {
+      "epoch": 0.8821165438713998,
+      "grad_norm": 0.4594753384590149,
+      "learning_rate": 0.0001,
+      "loss": 1.5792,
+      "step": 2634
+    },
+    {
+      "epoch": 0.8824514400535833,
+      "grad_norm": 0.4745759963989258,
+      "learning_rate": 0.0001,
+      "loss": 1.4701,
+      "step": 2635
+    },
+    {
+      "epoch": 0.882786336235767,
+      "grad_norm": 0.4678467810153961,
+      "learning_rate": 0.0001,
+      "loss": 1.509,
+      "step": 2636
+    },
+    {
+      "epoch": 0.8831212324179505,
+      "grad_norm": 0.4778388738632202,
+      "learning_rate": 0.0001,
+      "loss": 1.4904,
+      "step": 2637
+    },
+    {
+      "epoch": 0.883456128600134,
+      "grad_norm": 0.45584356784820557,
+      "learning_rate": 0.0001,
+      "loss": 1.5583,
+      "step": 2638
+    },
+    {
+      "epoch": 0.8837910247823175,
+      "grad_norm": 0.5151787996292114,
+      "learning_rate": 0.0001,
+      "loss": 1.6115,
+      "step": 2639
+    },
+    {
+      "epoch": 0.884125920964501,
+      "grad_norm": 0.4784322679042816,
+      "learning_rate": 0.0001,
+      "loss": 1.6164,
+      "step": 2640
+    },
+    {
+      "epoch": 0.8844608171466846,
+      "grad_norm": 0.4793466329574585,
+      "learning_rate": 0.0001,
+      "loss": 1.5444,
+      "step": 2641
+    },
+    {
+      "epoch": 0.8847957133288681,
+      "grad_norm": 0.477232962846756,
+      "learning_rate": 0.0001,
+      "loss": 1.4949,
+      "step": 2642
+    },
+    {
+      "epoch": 0.8851306095110516,
+      "grad_norm": 0.4777398109436035,
+      "learning_rate": 0.0001,
+      "loss": 1.4061,
+      "step": 2643
+    },
+    {
+      "epoch": 0.8854655056932351,
+      "grad_norm": 0.4389171004295349,
+      "learning_rate": 0.0001,
+      "loss": 1.4163,
+      "step": 2644
+    },
+    {
+      "epoch": 0.8858004018754186,
+      "grad_norm": 0.4764946401119232,
+      "learning_rate": 0.0001,
+      "loss": 1.5869,
+      "step": 2645
+    },
+    {
+      "epoch": 0.8861352980576022,
+      "grad_norm": 0.46828514337539673,
+      "learning_rate": 0.0001,
+      "loss": 1.4037,
+      "step": 2646
+    },
+    {
+      "epoch": 0.8864701942397857,
+      "grad_norm": 0.4633961319923401,
+      "learning_rate": 0.0001,
+      "loss": 1.5485,
+      "step": 2647
+    },
+    {
+      "epoch": 0.8868050904219692,
+      "grad_norm": 0.4534933865070343,
+      "learning_rate": 0.0001,
+      "loss": 1.4125,
+      "step": 2648
+    },
+    {
+      "epoch": 0.8871399866041527,
+      "grad_norm": 0.4641197621822357,
+      "learning_rate": 0.0001,
+      "loss": 1.4686,
+      "step": 2649
+    },
+    {
+      "epoch": 0.8874748827863362,
+      "grad_norm": 0.4668722450733185,
+      "learning_rate": 0.0001,
+      "loss": 1.5477,
+      "step": 2650
+    },
+    {
+      "epoch": 0.8878097789685198,
+      "grad_norm": 0.5055553317070007,
+      "learning_rate": 0.0001,
+      "loss": 1.6096,
+      "step": 2651
+    },
+    {
+      "epoch": 0.8881446751507033,
+      "grad_norm": 0.48708033561706543,
+      "learning_rate": 0.0001,
+      "loss": 1.6343,
+      "step": 2652
+    },
+    {
+      "epoch": 0.8884795713328868,
+      "grad_norm": 0.468066930770874,
+      "learning_rate": 0.0001,
+      "loss": 1.5492,
+      "step": 2653
+    },
+    {
+      "epoch": 0.8888144675150703,
+      "grad_norm": 0.47345656156539917,
+      "learning_rate": 0.0001,
+      "loss": 1.4374,
+      "step": 2654
+    },
+    {
+      "epoch": 0.8891493636972538,
+      "grad_norm": 0.46650683879852295,
+      "learning_rate": 0.0001,
+      "loss": 1.4227,
+      "step": 2655
+    },
+    {
+      "epoch": 0.8894842598794374,
+      "grad_norm": 0.45780879259109497,
+      "learning_rate": 0.0001,
+      "loss": 1.4844,
+      "step": 2656
+    },
+    {
+      "epoch": 0.8898191560616209,
+      "grad_norm": 0.4695964455604553,
+      "learning_rate": 0.0001,
+      "loss": 1.5558,
+      "step": 2657
+    },
+    {
+      "epoch": 0.8901540522438044,
+      "grad_norm": 0.4641098380088806,
+      "learning_rate": 0.0001,
+      "loss": 1.561,
+      "step": 2658
+    },
+    {
+      "epoch": 0.8904889484259879,
+      "grad_norm": 0.4501829147338867,
+      "learning_rate": 0.0001,
+      "loss": 1.362,
+      "step": 2659
+    },
+    {
+      "epoch": 0.8908238446081714,
+      "grad_norm": 0.4778416156768799,
+      "learning_rate": 0.0001,
+      "loss": 1.4539,
+      "step": 2660
+    },
+    {
+      "epoch": 0.891158740790355,
+      "grad_norm": 0.4750802516937256,
+      "learning_rate": 0.0001,
+      "loss": 1.5012,
+      "step": 2661
+    },
+    {
+      "epoch": 0.8914936369725385,
+      "grad_norm": 0.47056931257247925,
+      "learning_rate": 0.0001,
+      "loss": 1.4979,
+      "step": 2662
+    },
+    {
+      "epoch": 0.891828533154722,
+      "grad_norm": 0.4810597002506256,
+      "learning_rate": 0.0001,
+      "loss": 1.4733,
+      "step": 2663
+    },
+    {
+      "epoch": 0.8921634293369055,
+      "grad_norm": 0.5094153881072998,
+      "learning_rate": 0.0001,
+      "loss": 1.6202,
+      "step": 2664
+    },
+    {
+      "epoch": 0.892498325519089,
+      "grad_norm": 0.5079136490821838,
+      "learning_rate": 0.0001,
+      "loss": 1.6265,
+      "step": 2665
+    },
+    {
+      "epoch": 0.8928332217012726,
+      "grad_norm": 0.48242053389549255,
+      "learning_rate": 0.0001,
+      "loss": 1.5071,
+      "step": 2666
+    },
+    {
+      "epoch": 0.8931681178834562,
+      "grad_norm": 0.4616110026836395,
+      "learning_rate": 0.0001,
+      "loss": 1.5125,
+      "step": 2667
+    },
+    {
+      "epoch": 0.8935030140656397,
+      "grad_norm": 0.4890819787979126,
+      "learning_rate": 0.0001,
+      "loss": 1.5626,
+      "step": 2668
+    },
+    {
+      "epoch": 0.8938379102478232,
+      "grad_norm": 0.47077295184135437,
+      "learning_rate": 0.0001,
+      "loss": 1.4627,
+      "step": 2669
+    },
+    {
+      "epoch": 0.8941728064300067,
+      "grad_norm": 0.4646243155002594,
+      "learning_rate": 0.0001,
+      "loss": 1.3951,
+      "step": 2670
+    },
+    {
+      "epoch": 0.8945077026121903,
+      "grad_norm": 0.48134827613830566,
+      "learning_rate": 0.0001,
+      "loss": 1.4382,
+      "step": 2671
+    },
+    {
+      "epoch": 0.8948425987943738,
+      "grad_norm": 0.5001153349876404,
+      "learning_rate": 0.0001,
+      "loss": 1.4841,
+      "step": 2672
+    },
+    {
+      "epoch": 0.8951774949765573,
+      "grad_norm": 0.4859412610530853,
+      "learning_rate": 0.0001,
+      "loss": 1.5493,
+      "step": 2673
+    },
+    {
+      "epoch": 0.8955123911587408,
+      "grad_norm": 0.4802533984184265,
+      "learning_rate": 0.0001,
+      "loss": 1.539,
+      "step": 2674
+    },
+    {
+      "epoch": 0.8958472873409243,
+      "grad_norm": 0.48965784907341003,
+      "learning_rate": 0.0001,
+      "loss": 1.535,
+      "step": 2675
+    },
+    {
+      "epoch": 0.8961821835231079,
+      "grad_norm": 0.44885024428367615,
+      "learning_rate": 0.0001,
+      "loss": 1.5096,
+      "step": 2676
+    },
+    {
+      "epoch": 0.8965170797052914,
+      "grad_norm": 0.4732823371887207,
+      "learning_rate": 0.0001,
+      "loss": 1.4385,
+      "step": 2677
+    },
+    {
+      "epoch": 0.8968519758874749,
+      "grad_norm": 0.4734841585159302,
+      "learning_rate": 0.0001,
+      "loss": 1.4837,
+      "step": 2678
+    },
+    {
+      "epoch": 0.8971868720696584,
+      "grad_norm": 0.46776431798934937,
+      "learning_rate": 0.0001,
+      "loss": 1.4658,
+      "step": 2679
+    },
+    {
+      "epoch": 0.8975217682518419,
+      "grad_norm": 0.4772953391075134,
+      "learning_rate": 0.0001,
+      "loss": 1.4721,
+      "step": 2680
+    },
+    {
+      "epoch": 0.8978566644340255,
+      "grad_norm": 0.4706651568412781,
+      "learning_rate": 0.0001,
+      "loss": 1.4561,
+      "step": 2681
+    },
+    {
+      "epoch": 0.898191560616209,
+      "grad_norm": 0.4698059558868408,
+      "learning_rate": 0.0001,
+      "loss": 1.5312,
+      "step": 2682
+    },
+    {
+      "epoch": 0.8985264567983925,
+      "grad_norm": 0.468711256980896,
+      "learning_rate": 0.0001,
+      "loss": 1.5379,
+      "step": 2683
+    },
+    {
+      "epoch": 0.898861352980576,
+      "grad_norm": 0.479682594537735,
+      "learning_rate": 0.0001,
+      "loss": 1.4411,
+      "step": 2684
+    },
+    {
+      "epoch": 0.8991962491627595,
+      "grad_norm": 0.49723494052886963,
+      "learning_rate": 0.0001,
+      "loss": 1.5347,
+      "step": 2685
+    },
+    {
+      "epoch": 0.899531145344943,
+      "grad_norm": 0.46617788076400757,
+      "learning_rate": 0.0001,
+      "loss": 1.55,
+      "step": 2686
+    },
+    {
+      "epoch": 0.8998660415271266,
+      "grad_norm": 0.44119560718536377,
+      "learning_rate": 0.0001,
+      "loss": 1.5047,
+      "step": 2687
+    },
+    {
+      "epoch": 0.9002009377093101,
+      "grad_norm": 0.4710986614227295,
+      "learning_rate": 0.0001,
+      "loss": 1.4465,
+      "step": 2688
+    },
+    {
+      "epoch": 0.9005358338914936,
+      "grad_norm": 0.4694075286388397,
+      "learning_rate": 0.0001,
+      "loss": 1.5825,
+      "step": 2689
+    },
+    {
+      "epoch": 0.9008707300736771,
+      "grad_norm": 0.4742186963558197,
+      "learning_rate": 0.0001,
+      "loss": 1.5765,
+      "step": 2690
+    },
+    {
+      "epoch": 0.9012056262558606,
+      "grad_norm": 0.47852757573127747,
+      "learning_rate": 0.0001,
+      "loss": 1.4914,
+      "step": 2691
+    },
+    {
+      "epoch": 0.9015405224380442,
+      "grad_norm": 0.47422271966934204,
+      "learning_rate": 0.0001,
+      "loss": 1.5882,
+      "step": 2692
+    },
+    {
+      "epoch": 0.9018754186202277,
+      "grad_norm": 0.4616380035877228,
+      "learning_rate": 0.0001,
+      "loss": 1.5565,
+      "step": 2693
+    },
+    {
+      "epoch": 0.9022103148024112,
+      "grad_norm": 0.46155980229377747,
+      "learning_rate": 0.0001,
+      "loss": 1.5269,
+      "step": 2694
+    },
+    {
+      "epoch": 0.9025452109845947,
+      "grad_norm": 0.48004859685897827,
+      "learning_rate": 0.0001,
+      "loss": 1.5727,
+      "step": 2695
+    },
+    {
+      "epoch": 0.9028801071667782,
+      "grad_norm": 0.5059171915054321,
+      "learning_rate": 0.0001,
+      "loss": 1.5574,
+      "step": 2696
+    },
+    {
+      "epoch": 0.9032150033489619,
+      "grad_norm": 0.4653322994709015,
+      "learning_rate": 0.0001,
+      "loss": 1.5277,
+      "step": 2697
+    },
+    {
+      "epoch": 0.9035498995311454,
+      "grad_norm": 0.4666169285774231,
+      "learning_rate": 0.0001,
+      "loss": 1.4112,
+      "step": 2698
+    },
+    {
+      "epoch": 0.9038847957133289,
+      "grad_norm": 0.5089284777641296,
+      "learning_rate": 0.0001,
+      "loss": 1.5601,
+      "step": 2699
+    },
+    {
+      "epoch": 0.9042196918955124,
+      "grad_norm": 0.4853679835796356,
+      "learning_rate": 0.0001,
+      "loss": 1.5193,
+      "step": 2700
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2986,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 300,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.5327371620830413e+19,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}