diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,18944 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.4833773821115592,
+  "eval_steps": 899,
+  "global_step": 2697,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0005564056196967589,
+      "grad_norm": 1.3203125,
+      "learning_rate": 2e-05,
+      "loss": 1.7806,
+      "step": 1
+    },
+    {
+      "epoch": 0.0005564056196967589,
+      "eval_loss": 1.7791756391525269,
+      "eval_runtime": 1750.0366,
+      "eval_samples_per_second": 2.681,
+      "eval_steps_per_second": 1.341,
+      "step": 1
+    },
+    {
+      "epoch": 0.0011128112393935178,
+      "grad_norm": 1.1875,
+      "learning_rate": 4e-05,
+      "loss": 1.7467,
+      "step": 2
+    },
+    {
+      "epoch": 0.0016692168590902768,
+      "grad_norm": 1.1484375,
+      "learning_rate": 6e-05,
+      "loss": 1.6933,
+      "step": 3
+    },
+    {
+      "epoch": 0.0022256224787870356,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8e-05,
+      "loss": 1.6916,
+      "step": 4
+    },
+    {
+      "epoch": 0.0027820280984837947,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0001,
+      "loss": 1.6973,
+      "step": 5
+    },
+    {
+      "epoch": 0.0033384337181805537,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.00012,
+      "loss": 1.6518,
+      "step": 6
+    },
+    {
+      "epoch": 0.0038948393378773127,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00014,
+      "loss": 1.6383,
+      "step": 7
+    },
+    {
+      "epoch": 0.004451244957574071,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00016,
+      "loss": 1.5714,
+      "step": 8
+    },
+    {
+      "epoch": 0.005007650577270831,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00018,
+      "loss": 1.553,
+      "step": 9
+    },
+    {
+      "epoch": 0.005564056196967589,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5321,
+      "step": 10
+    },
+    {
+      "epoch": 0.006120461816664348,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.0001999999829570899,
+      "loss": 1.5589,
+      "step": 11
+    },
+    {
+      "epoch": 0.006676867436361107,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001999999318283654,
+      "loss": 1.4519,
+      "step": 12
+    },
+    {
+      "epoch": 0.007233273056057866,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001999998466138439,
+      "loss": 1.4035,
+      "step": 13
+    },
+    {
+      "epoch": 0.007789678675754625,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001999997273135545,
+      "loss": 1.4071,
+      "step": 14
+    },
+    {
+      "epoch": 0.008346084295451384,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001999995739275378,
+      "loss": 1.2862,
+      "step": 15
+    },
+    {
+      "epoch": 0.008902489915148143,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019999938645584616,
+      "loss": 1.2801,
+      "step": 16
+    },
+    {
+      "epoch": 0.009458895534844901,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001999991648985434,
+      "loss": 1.2358,
+      "step": 17
+    },
+    {
+      "epoch": 0.010015301154541661,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001999989092557051,
+      "loss": 1.2122,
+      "step": 18
+    },
+    {
+      "epoch": 0.01057170677423842,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019999861952741839,
+      "loss": 1.2294,
+      "step": 19
+    },
+    {
+      "epoch": 0.011128112393935179,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019999829571378197,
+      "loss": 1.2641,
+      "step": 20
+    },
+    {
+      "epoch": 0.011684518013631937,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0001999979378149063,
+      "loss": 1.2373,
+      "step": 21
+    },
+    {
+      "epoch": 0.012240923633328696,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0001999975458309133,
+      "loss": 1.2389,
+      "step": 22
+    },
+    {
+      "epoch": 0.012797329253025456,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00019999711976193663,
+      "loss": 1.2716,
+      "step": 23
+    },
+    {
+      "epoch": 0.013353734872722215,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0001999966596081215,
+      "loss": 1.2038,
+      "step": 24
+    },
+    {
+      "epoch": 0.013910140492418973,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00019999616536962473,
+      "loss": 1.2588,
+      "step": 25
+    },
+    {
+      "epoch": 0.014466546112115732,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00019999563704661483,
+      "loss": 1.198,
+      "step": 26
+    },
+    {
+      "epoch": 0.01502295173181249,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00019999507463927188,
+      "loss": 1.2047,
+      "step": 27
+    },
+    {
+      "epoch": 0.01557935735150925,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00019999447814778755,
+      "loss": 1.1884,
+      "step": 28
+    },
+    {
+      "epoch": 0.01613576297120601,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001999938475723652,
+      "loss": 1.2156,
+      "step": 29
+    },
+    {
+      "epoch": 0.016692168590902768,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00019999318291321968,
+      "loss": 1.1636,
+      "step": 30
+    },
+    {
+      "epoch": 0.017248574210599527,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00019999248417057767,
+      "loss": 1.1857,
+      "step": 31
+    },
+    {
+      "epoch": 0.017804979830296285,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00019999175134467726,
+      "loss": 1.1862,
+      "step": 32
+    },
+    {
+      "epoch": 0.018361385449993044,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00019999098443576827,
+      "loss": 1.2155,
+      "step": 33
+    },
+    {
+      "epoch": 0.018917791069689802,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00019999018344411207,
+      "loss": 1.1009,
+      "step": 34
+    },
+    {
+      "epoch": 0.019474196689386564,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00019998934836998173,
+      "loss": 1.1714,
+      "step": 35
+    },
+    {
+      "epoch": 0.020030602309083323,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00019998847921366187,
+      "loss": 1.0824,
+      "step": 36
+    },
+    {
+      "epoch": 0.02058700792878008,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00019998757597544875,
+      "loss": 1.0703,
+      "step": 37
+    },
+    {
+      "epoch": 0.02114341354847684,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00019998663865565023,
+      "loss": 1.1183,
+      "step": 38
+    },
+    {
+      "epoch": 0.0216998191681736,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00019998566725458586,
+      "loss": 1.1328,
+      "step": 39
+    },
+    {
+      "epoch": 0.022256224787870357,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0001999846617725867,
+      "loss": 1.1076,
+      "step": 40
+    },
+    {
+      "epoch": 0.022812630407567116,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00019998362220999548,
+      "loss": 1.1143,
+      "step": 41
+    },
+    {
+      "epoch": 0.023369036027263874,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00019998254856716657,
+      "loss": 1.1768,
+      "step": 42
+    },
+    {
+      "epoch": 0.023925441646960633,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0001999814408444659,
+      "loss": 1.0872,
+      "step": 43
+    },
+    {
+      "epoch": 0.02448184726665739,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00019998029904227103,
+      "loss": 1.122,
+      "step": 44
+    },
+    {
+      "epoch": 0.025038252886354154,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0001999791231609712,
+      "loss": 1.0025,
+      "step": 45
+    },
+    {
+      "epoch": 0.025594658506050912,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00019997791320096718,
+      "loss": 1.1672,
+      "step": 46
+    },
+    {
+      "epoch": 0.02615106412574767,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00019997666916267144,
+      "loss": 1.1096,
+      "step": 47
+    },
+    {
+      "epoch": 0.02670746974544443,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.000199975391046508,
+      "loss": 1.1446,
+      "step": 48
+    },
+    {
+      "epoch": 0.027263875365141188,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0001999740788529125,
+      "loss": 1.1254,
+      "step": 49
+    },
+    {
+      "epoch": 0.027820280984837947,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00019997273258233223,
+      "loss": 1.1143,
+      "step": 50
+    },
+    {
+      "epoch": 0.028376686604534705,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00019997135223522605,
+      "loss": 1.1123,
+      "step": 51
+    },
+    {
+      "epoch": 0.028933092224231464,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00019996993781206448,
+      "loss": 1.0469,
+      "step": 52
+    },
+    {
+      "epoch": 0.029489497843928222,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00019996848931332965,
+      "loss": 1.1145,
+      "step": 53
+    },
+    {
+      "epoch": 0.03004590346362498,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0001999670067395153,
+      "loss": 1.0489,
+      "step": 54
+    },
+    {
+      "epoch": 0.030602309083321743,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00019996549009112675,
+      "loss": 1.0671,
+      "step": 55
+    },
+    {
+      "epoch": 0.0311587147030185,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00019996393936868098,
+      "loss": 1.0341,
+      "step": 56
+    },
+    {
+      "epoch": 0.03171512032271526,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00019996235457270654,
+      "loss": 1.0725,
+      "step": 57
+    },
+    {
+      "epoch": 0.03227152594241202,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00019996073570374367,
+      "loss": 1.1031,
+      "step": 58
+    },
+    {
+      "epoch": 0.032827931562108774,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0001999590827623441,
+      "loss": 0.9974,
+      "step": 59
+    },
+    {
+      "epoch": 0.033384337181805536,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00019995739574907133,
+      "loss": 1.1398,
+      "step": 60
+    },
+    {
+      "epoch": 0.0339407428015023,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00019995567466450035,
+      "loss": 1.0728,
+      "step": 61
+    },
+    {
+      "epoch": 0.03449714842119905,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00019995391950921782,
+      "loss": 0.9649,
+      "step": 62
+    },
+    {
+      "epoch": 0.035053554040895815,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.000199952130283822,
+      "loss": 1.1536,
+      "step": 63
+    },
+    {
+      "epoch": 0.03560995966059257,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0001999503069889227,
+      "loss": 1.0135,
+      "step": 64
+    },
+    {
+      "epoch": 0.03616636528028933,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0001999484496251415,
+      "loss": 1.0599,
+      "step": 65
+    },
+    {
+      "epoch": 0.03672277089998609,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00019994655819311144,
+      "loss": 1.0521,
+      "step": 66
+    },
+    {
+      "epoch": 0.03727917651968285,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00019994463269347725,
+      "loss": 1.0737,
+      "step": 67
+    },
+    {
+      "epoch": 0.037835582139379605,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00019994267312689525,
+      "loss": 1.0856,
+      "step": 68
+    },
+    {
+      "epoch": 0.03839198775907637,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00019994067949403337,
+      "loss": 0.9674,
+      "step": 69
+    },
+    {
+      "epoch": 0.03894839337877313,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00019993865179557117,
+      "loss": 1.1069,
+      "step": 70
+    },
+    {
+      "epoch": 0.039504798998469884,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00019993659003219978,
+      "loss": 1.0198,
+      "step": 71
+    },
+    {
+      "epoch": 0.040061204618166646,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00019993449420462197,
+      "loss": 1.0864,
+      "step": 72
+    },
+    {
+      "epoch": 0.0406176102378634,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00019993236431355217,
+      "loss": 1.037,
+      "step": 73
+    },
+    {
+      "epoch": 0.04117401585756016,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00019993020035971633,
+      "loss": 1.0724,
+      "step": 74
+    },
+    {
+      "epoch": 0.04173042147725692,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00019992800234385201,
+      "loss": 0.9854,
+      "step": 75
+    },
+    {
+      "epoch": 0.04228682709695368,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00019992577026670854,
+      "loss": 1.0655,
+      "step": 76
+    },
+    {
+      "epoch": 0.042843232716650435,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00019992350412904663,
+      "loss": 1.0487,
+      "step": 77
+    },
+    {
+      "epoch": 0.0433996383363472,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00019992120393163876,
+      "loss": 0.9593,
+      "step": 78
+    },
+    {
+      "epoch": 0.04395604395604396,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00019991886967526897,
+      "loss": 1.1132,
+      "step": 79
+    },
+    {
+      "epoch": 0.044512449575740715,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00019991650136073287,
+      "loss": 0.9343,
+      "step": 80
+    },
+    {
+      "epoch": 0.04506885519543748,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0001999140989888378,
+      "loss": 1.0228,
+      "step": 81
+    },
+    {
+      "epoch": 0.04562526081513423,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00019991166256040256,
+      "loss": 1.0239,
+      "step": 82
+    },
+    {
+      "epoch": 0.046181666434830994,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00019990919207625764,
+      "loss": 1.0706,
+      "step": 83
+    },
+    {
+      "epoch": 0.04673807205452775,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00019990668753724515,
+      "loss": 1.0069,
+      "step": 84
+    },
+    {
+      "epoch": 0.04729447767422451,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00019990414894421876,
+      "loss": 1.0511,
+      "step": 85
+    },
+    {
+      "epoch": 0.047850883293921266,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0001999015762980438,
+      "loss": 0.9398,
+      "step": 86
+    },
+    {
+      "epoch": 0.04840728891361803,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00019989896959959712,
+      "loss": 1.1496,
+      "step": 87
+    },
+    {
+      "epoch": 0.04896369453331478,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00019989632884976727,
+      "loss": 1.0488,
+      "step": 88
+    },
+    {
+      "epoch": 0.049520100153011545,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00019989365404945436,
+      "loss": 1.0843,
+      "step": 89
+    },
+    {
+      "epoch": 0.05007650577270831,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00019989094519957015,
+      "loss": 1.0822,
+      "step": 90
+    },
+    {
+      "epoch": 0.05063291139240506,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00019988820230103796,
+      "loss": 1.0301,
+      "step": 91
+    },
+    {
+      "epoch": 0.051189317012101825,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0001998854253547927,
+      "loss": 1.0098,
+      "step": 92
+    },
+    {
+      "epoch": 0.05174572263179858,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00019988261436178093,
+      "loss": 0.9847,
+      "step": 93
+    },
+    {
+      "epoch": 0.05230212825149534,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0001998797693229608,
+      "loss": 1.0233,
+      "step": 94
+    },
+    {
+      "epoch": 0.0528585338711921,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0001998768902393021,
+      "loss": 1.0747,
+      "step": 95
+    },
+    {
+      "epoch": 0.05341493949088886,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00019987397711178614,
+      "loss": 0.9823,
+      "step": 96
+    },
+    {
+      "epoch": 0.053971345110585614,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0001998710299414059,
+      "loss": 1.0238,
+      "step": 97
+    },
+    {
+      "epoch": 0.054527750730282376,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00019986804872916593,
+      "loss": 0.9832,
+      "step": 98
+    },
+    {
+      "epoch": 0.05508415634997914,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00019986503347608245,
+      "loss": 0.9833,
+      "step": 99
+    },
+    {
+      "epoch": 0.05564056196967589,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0001998619841831832,
+      "loss": 1.0971,
+      "step": 100
+    },
+    {
+      "epoch": 0.056196967589372655,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0001998589008515076,
+      "loss": 1.0077,
+      "step": 101
+    },
+    {
+      "epoch": 0.05675337320906941,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0001998557834821065,
+      "loss": 1.0066,
+      "step": 102
+    },
+    {
+      "epoch": 0.05730977882876617,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00019985263207604264,
+      "loss": 1.0267,
+      "step": 103
+    },
+    {
+      "epoch": 0.05786618444846293,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0001998494466343901,
+      "loss": 0.975,
+      "step": 104
+    },
+    {
+      "epoch": 0.05842259006815969,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00019984622715823474,
+      "loss": 1.0134,
+      "step": 105
+    },
+    {
+      "epoch": 0.058978995687856445,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00019984297364867385,
+      "loss": 1.1247,
+      "step": 106
+    },
+    {
+      "epoch": 0.05953540130755321,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00019983968610681648,
+      "loss": 0.974,
+      "step": 107
+    },
+    {
+      "epoch": 0.06009180692724996,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0001998363645337832,
+      "loss": 1.0099,
+      "step": 108
+    },
+    {
+      "epoch": 0.060648212546946724,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00019983300893070623,
+      "loss": 0.8479,
+      "step": 109
+    },
+    {
+      "epoch": 0.061204618166643486,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00019982961929872927,
+      "loss": 1.0034,
+      "step": 110
+    },
+    {
+      "epoch": 0.06176102378634024,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00019982619563900782,
+      "loss": 1.0165,
+      "step": 111
+    },
+    {
+      "epoch": 0.062317429406037,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00019982273795270876,
+      "loss": 1.0338,
+      "step": 112
+    },
+    {
+      "epoch": 0.06287383502573377,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00019981924624101071,
+      "loss": 1.0465,
+      "step": 113
+    },
+    {
+      "epoch": 0.06343024064543051,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00019981572050510387,
+      "loss": 0.9284,
+      "step": 114
+    },
+    {
+      "epoch": 0.06398664626512728,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00019981216074619,
+      "loss": 1.074,
+      "step": 115
+    },
+    {
+      "epoch": 0.06454305188482404,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00019980856696548242,
+      "loss": 0.8981,
+      "step": 116
+    },
+    {
+      "epoch": 0.0650994575045208,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00019980493916420616,
+      "loss": 1.0605,
+      "step": 117
+    },
+    {
+      "epoch": 0.06565586312421755,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00019980127734359782,
+      "loss": 1.0433,
+      "step": 118
+    },
+    {
+      "epoch": 0.06621226874391431,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00019979758150490546,
+      "loss": 1.0483,
+      "step": 119
+    },
+    {
+      "epoch": 0.06676867436361107,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0001997938516493889,
+      "loss": 1.0543,
+      "step": 120
+    },
+    {
+      "epoch": 0.06732507998330783,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0001997900877783195,
+      "loss": 1.0162,
+      "step": 121
+    },
+    {
+      "epoch": 0.0678814856030046,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0001997862898929802,
+      "loss": 0.8468,
+      "step": 122
+    },
+    {
+      "epoch": 0.06843789122270134,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00019978245799466554,
+      "loss": 1.0417,
+      "step": 123
+    },
+    {
+      "epoch": 0.0689942968423981,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00019977859208468161,
+      "loss": 0.9901,
+      "step": 124
+    },
+    {
+      "epoch": 0.06955070246209487,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00019977469216434619,
+      "loss": 1.0117,
+      "step": 125
+    },
+    {
+      "epoch": 0.07010710808179163,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0001997707582349886,
+      "loss": 0.9719,
+      "step": 126
+    },
+    {
+      "epoch": 0.07066351370148838,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00019976679029794968,
+      "loss": 0.9176,
+      "step": 127
+    },
+    {
+      "epoch": 0.07121991932118514,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00019976278835458203,
+      "loss": 1.0633,
+      "step": 128
+    },
+    {
+      "epoch": 0.0717763249408819,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00019975875240624968,
+      "loss": 0.9845,
+      "step": 129
+    },
+    {
+      "epoch": 0.07233273056057866,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00019975468245432833,
+      "loss": 0.9352,
+      "step": 130
+    },
+    {
+      "epoch": 0.07288913618027543,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0001997505785002053,
+      "loss": 1.0966,
+      "step": 131
+    },
+    {
+      "epoch": 0.07344554179997217,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0001997464405452794,
+      "loss": 0.9278,
+      "step": 132
+    },
+    {
+      "epoch": 0.07400194741966894,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0001997422685909611,
+      "loss": 0.9244,
+      "step": 133
+    },
+    {
+      "epoch": 0.0745583530393657,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00019973806263867247,
+      "loss": 0.8941,
+      "step": 134
+    },
+    {
+      "epoch": 0.07511475865906246,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00019973382268984713,
+      "loss": 0.9702,
+      "step": 135
+    },
+    {
+      "epoch": 0.07567116427875921,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00019972954874593026,
+      "loss": 0.9809,
+      "step": 136
+    },
+    {
+      "epoch": 0.07622756989845597,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00019972524080837873,
+      "loss": 1.0604,
+      "step": 137
+    },
+    {
+      "epoch": 0.07678397551815273,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00019972089887866092,
+      "loss": 0.9702,
+      "step": 138
+    },
+    {
+      "epoch": 0.0773403811378495,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00019971652295825681,
+      "loss": 0.9903,
+      "step": 139
+    },
+    {
+      "epoch": 0.07789678675754626,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00019971211304865795,
+      "loss": 0.979,
+      "step": 140
+    },
+    {
+      "epoch": 0.078453192377243,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00019970766915136746,
+      "loss": 1.1067,
+      "step": 141
+    },
+    {
+      "epoch": 0.07900959799693977,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0001997031912679002,
+      "loss": 0.966,
+      "step": 142
+    },
+    {
+      "epoch": 0.07956600361663653,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00019969867939978236,
+      "loss": 1.0207,
+      "step": 143
+    },
+    {
+      "epoch": 0.08012240923633329,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00019969413354855195,
+      "loss": 0.9127,
+      "step": 144
+    },
+    {
+      "epoch": 0.08067881485603004,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0001996895537157584,
+      "loss": 1.0944,
+      "step": 145
+    },
+    {
+      "epoch": 0.0812352204757268,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00019968493990296277,
+      "loss": 0.948,
+      "step": 146
+    },
+    {
+      "epoch": 0.08179162609542356,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00019968029211173778,
+      "loss": 0.9824,
+      "step": 147
+    },
+    {
+      "epoch": 0.08234803171512033,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00019967561034366763,
+      "loss": 0.9328,
+      "step": 148
+    },
+    {
+      "epoch": 0.08290443733481709,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00019967089460034814,
+      "loss": 1.0162,
+      "step": 149
+    },
+    {
+      "epoch": 0.08346084295451384,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0001996661448833867,
+      "loss": 0.7696,
+      "step": 150
+    },
+    {
+      "epoch": 0.0840172485742106,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00019966136119440233,
+      "loss": 0.9133,
+      "step": 151
+    },
+    {
+      "epoch": 0.08457365419390736,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00019965654353502554,
+      "loss": 1.0071,
+      "step": 152
+    },
+    {
+      "epoch": 0.08513005981360412,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0001996516919068985,
+      "loss": 0.9797,
+      "step": 153
+    },
+    {
+      "epoch": 0.08568646543330087,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00019964680631167492,
+      "loss": 0.9521,
+      "step": 154
+    },
+    {
+      "epoch": 0.08624287105299763,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0001996418867510201,
+      "loss": 1.034,
+      "step": 155
+    },
+    {
+      "epoch": 0.0867992766726944,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00019963693322661087,
+      "loss": 0.9971,
+      "step": 156
+    },
+    {
+      "epoch": 0.08735568229239116,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00019963194574013573,
+      "loss": 1.0262,
+      "step": 157
+    },
+    {
+      "epoch": 0.08791208791208792,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0001996269242932947,
+      "loss": 0.9913,
+      "step": 158
+    },
+    {
+      "epoch": 0.08846849353178467,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00019962186888779936,
+      "loss": 1.0822,
+      "step": 159
+    },
+    {
+      "epoch": 0.08902489915148143,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00019961677952537292,
+      "loss": 0.8287,
+      "step": 160
+    },
+    {
+      "epoch": 0.08958130477117819,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0001996116562077501,
+      "loss": 0.9663,
+      "step": 161
+    },
+    {
+      "epoch": 0.09013771039087495,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00019960649893667722,
+      "loss": 0.7657,
+      "step": 162
+    },
+    {
+      "epoch": 0.0906941160105717,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0001996013077139122,
+      "loss": 0.8837,
+      "step": 163
+    },
+    {
+      "epoch": 0.09125052163026846,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0001995960825412245,
+      "loss": 0.9727,
+      "step": 164
+    },
+    {
+      "epoch": 0.09180692724996523,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0001995908234203952,
+      "loss": 0.9438,
+      "step": 165
+    },
+    {
+      "epoch": 0.09236333286966199,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00019958553035321686,
+      "loss": 0.9364,
+      "step": 166
+    },
+    {
+      "epoch": 0.09291973848935874,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00019958020334149366,
+      "loss": 0.9435,
+      "step": 167
+    },
+    {
+      "epoch": 0.0934761441090555,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00019957484238704139,
+      "loss": 1.034,
+      "step": 168
+    },
+    {
+      "epoch": 0.09403254972875226,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00019956944749168737,
+      "loss": 0.9662,
+      "step": 169
+    },
+    {
+      "epoch": 0.09458895534844902,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00019956401865727053,
+      "loss": 1.0057,
+      "step": 170
+    },
+    {
+      "epoch": 0.09514536096814578,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00019955855588564125,
+      "loss": 0.9569,
+      "step": 171
+    },
+    {
+      "epoch": 0.09570176658784253,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00019955305917866164,
+      "loss": 0.9651,
+      "step": 172
+    },
+    {
+      "epoch": 0.0962581722075393,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00019954752853820523,
+      "loss": 0.8911,
+      "step": 173
+    },
+    {
+      "epoch": 0.09681457782723606,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00019954196396615721,
+      "loss": 0.9656,
+      "step": 174
+    },
+    {
+      "epoch": 0.09737098344693282,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00019953636546441431,
+      "loss": 0.8851,
+      "step": 175
+    },
+    {
+      "epoch": 0.09792738906662957,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00019953073303488486,
+      "loss": 0.91,
+      "step": 176
+    },
+    {
+      "epoch": 0.09848379468632633,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00019952506667948868,
+      "loss": 1.0532,
+      "step": 177
+    },
+    {
+      "epoch": 0.09904020030602309,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00019951936640015723,
+      "loss": 1.052,
+      "step": 178
+    },
+    {
+      "epoch": 0.09959660592571985,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00019951363219883344,
+      "loss": 0.9436,
+      "step": 179
+    },
+    {
+      "epoch": 0.10015301154541661,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0001995078640774719,
+      "loss": 1.0067,
+      "step": 180
+    },
+    {
+      "epoch": 0.10070941716511336,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00019950206203803874,
+      "loss": 0.8923,
+      "step": 181
+    },
+    {
+      "epoch": 0.10126582278481013,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00019949622608251156,
+      "loss": 0.9681,
+      "step": 182
+    },
+    {
+      "epoch": 0.10182222840450689,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00019949035621287967,
+      "loss": 0.9994,
+      "step": 183
+    },
+    {
+      "epoch": 0.10237863402420365,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00019948445243114384,
+      "loss": 1.0139,
+      "step": 184
+    },
+    {
+      "epoch": 0.1029350396439004,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0001994785147393164,
+      "loss": 0.9698,
+      "step": 185
+    },
+    {
+      "epoch": 0.10349144526359716,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0001994725431394213,
+      "loss": 0.9727,
+      "step": 186
+    },
+    {
+      "epoch": 0.10404785088329392,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00019946653763349394,
+      "loss": 0.9109,
+      "step": 187
+    },
+    {
+      "epoch": 0.10460425650299068,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0001994604982235814,
+      "loss": 0.9881,
+      "step": 188
+    },
+    {
+      "epoch": 0.10516066212268745,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00019945442491174227,
+      "loss": 1.0062,
+      "step": 189
+    },
+    {
+      "epoch": 0.1057170677423842,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00019944831770004665,
+      "loss": 0.9374,
+      "step": 190
+    },
+    {
+      "epoch": 0.10627347336208096,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00019944217659057626,
+      "loss": 0.907,
+      "step": 191
+    },
+    {
+      "epoch": 0.10682987898177772,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00019943600158542436,
+      "loss": 0.9287,
+      "step": 192
+    },
+    {
+      "epoch": 0.10738628460147448,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00019942979268669573,
+      "loss": 0.9948,
+      "step": 193
+    },
+    {
+      "epoch": 0.10794269022117123,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0001994235498965067,
+      "loss": 0.9651,
+      "step": 194
+    },
+    {
+      "epoch": 0.10849909584086799,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00019941727321698521,
+      "loss": 0.9267,
+      "step": 195
+    },
+    {
+      "epoch": 0.10905550146056475,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00019941096265027074,
+      "loss": 0.8551,
+      "step": 196
+    },
+    {
+      "epoch": 0.10961190708026151,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00019940461819851425,
+      "loss": 0.9683,
+      "step": 197
+    },
+    {
+      "epoch": 0.11016831269995828,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00019939823986387834,
+      "loss": 0.9127,
+      "step": 198
+    },
+    {
+      "epoch": 0.11072471831965502,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00019939182764853707,
+      "loss": 1.0557,
+      "step": 199
+    },
+    {
+      "epoch": 0.11128112393935179,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00019938538155467615,
+      "loss": 0.9967,
+      "step": 200
+    },
+    {
+      "epoch": 0.11183752955904855,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00019937890158449272,
+      "loss": 0.9391,
+      "step": 201
+    },
+    {
+      "epoch": 0.11239393517874531,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00019937238774019557,
+      "loss": 1.9551,
+      "step": 202
+    },
+    {
+      "epoch": 0.11295034079844206,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.000199365840024005,
+      "loss": 0.9682,
+      "step": 203
+    },
+    {
+      "epoch": 0.11350674641813882,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00019935925843815285,
+      "loss": 0.961,
+      "step": 204
+    },
+    {
+      "epoch": 0.11406315203783558,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0001993526429848825,
+      "loss": 1.0718,
+      "step": 205
+    },
+    {
+      "epoch": 0.11461955765753234,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0001993459936664489,
+      "loss": 0.9232,
+      "step": 206
+    },
+    {
+      "epoch": 0.1151759632772291,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0001993393104851185,
+      "loss": 1.0022,
+      "step": 207
+    },
+    {
+      "epoch": 0.11573236889692586,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00019933259344316934,
+      "loss": 0.9549,
+      "step": 208
+    },
+    {
+      "epoch": 0.11628877451662262,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00019932584254289096,
+      "loss": 0.9472,
+      "step": 209
+    },
+    {
+      "epoch": 0.11684518013631938,
+      "grad_norm": 2.859375,
+      "learning_rate": 0.00019931905778658443,
+      "loss": 1.9077,
+      "step": 210
+    },
+    {
+      "epoch": 0.11740158575601614,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00019931223917656247,
+      "loss": 1.0227,
+      "step": 211
+    },
+    {
+      "epoch": 0.11795799137571289,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00019930538671514923,
+      "loss": 1.0234,
+      "step": 212
+    },
+    {
+      "epoch": 0.11851439699540965,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00019929850040468035,
+      "loss": 0.9573,
+      "step": 213
+    },
+    {
+      "epoch": 0.11907080261510641,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0001992915802475032,
+      "loss": 0.9576,
+      "step": 214
+    },
+    {
+      "epoch": 0.11962720823480318,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0001992846262459765,
+      "loss": 0.9715,
+      "step": 215
+    },
+    {
+      "epoch": 0.12018361385449992,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0001992776384024706,
+      "loss": 0.8743,
+      "step": 216
+    },
+    {
+      "epoch": 0.12074001947419669,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00019927061671936737,
+      "loss": 0.8957,
+      "step": 217
+    },
+    {
+      "epoch": 0.12129642509389345,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0001992635611990602,
+      "loss": 0.9718,
+      "step": 218
+    },
+    {
+      "epoch": 0.12185283071359021,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00019925647184395402,
+      "loss": 0.8916,
+      "step": 219
+    },
+    {
+      "epoch": 0.12240923633328697,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0001992493486564653,
+      "loss": 0.9546,
+      "step": 220
+    },
+    {
+      "epoch": 0.12296564195298372,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00019924219163902204,
+      "loss": 0.9285,
+      "step": 221
+    },
+    {
+      "epoch": 0.12352204757268048,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00019923500079406373,
+      "loss": 0.9095,
+      "step": 222
+    },
+    {
+      "epoch": 0.12407845319237724,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00019922777612404148,
+      "loss": 0.9921,
+      "step": 223
+    },
+    {
+      "epoch": 0.124634858812074,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00019922051763141788,
+      "loss": 0.9756,
+      "step": 224
+    },
+    {
+      "epoch": 0.12519126443177075,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00019921322531866702,
+      "loss": 0.832,
+      "step": 225
+    },
+    {
+      "epoch": 0.12574767005146753,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00019920589918827455,
+      "loss": 0.741,
+      "step": 226
+    },
+    {
+      "epoch": 0.12630407567116428,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0001991985392427376,
+      "loss": 0.8839,
+      "step": 227
+    },
+    {
+      "epoch": 0.12686048129086103,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00019919114548456497,
+      "loss": 1.0329,
+      "step": 228
+    },
+    {
+      "epoch": 0.1274168869105578,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0001991837179162768,
+      "loss": 0.9777,
+      "step": 229
+    },
+    {
+      "epoch": 0.12797329253025455,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00019917625654040485,
+      "loss": 0.9311,
+      "step": 230
+    },
+    {
+      "epoch": 0.12852969814995133,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0001991687613594924,
+      "loss": 0.9973,
+      "step": 231
+    },
+    {
+      "epoch": 0.12908610376964808,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00019916123237609428,
+      "loss": 1.054,
+      "step": 232
+    },
+    {
+      "epoch": 0.12964250938934482,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00019915366959277674,
+      "loss": 0.886,
+      "step": 233
+    },
+    {
+      "epoch": 0.1301989150090416,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00019914607301211765,
+      "loss": 0.9879,
+      "step": 234
+    },
+    {
+      "epoch": 0.13075532062873835,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00019913844263670638,
+      "loss": 0.8272,
+      "step": 235
+    },
+    {
+      "epoch": 0.1313117262484351,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00019913077846914377,
+      "loss": 0.8881,
+      "step": 236
+    },
+    {
+      "epoch": 0.13186813186813187,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00019912308051204226,
+      "loss": 0.9737,
+      "step": 237
+    },
+    {
+      "epoch": 0.13242453748782862,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00019911534876802572,
+      "loss": 1.1547,
+      "step": 238
+    },
+    {
+      "epoch": 0.1329809431075254,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00019910758323972961,
+      "loss": 0.9975,
+      "step": 239
+    },
+    {
+      "epoch": 0.13353734872722214,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00019909978392980087,
+      "loss": 0.9454,
+      "step": 240
+    },
+    {
+      "epoch": 0.1340937543469189,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00019909195084089792,
+      "loss": 0.969,
+      "step": 241
+    },
+    {
+      "epoch": 0.13465015996661567,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00019908408397569075,
+      "loss": 1.0012,
+      "step": 242
+    },
+    {
+      "epoch": 0.13520656558631242,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0001990761833368609,
+      "loss": 0.9292,
+      "step": 243
+    },
+    {
+      "epoch": 0.1357629712060092,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0001990682489271013,
+      "loss": 0.8401,
+      "step": 244
+    },
+    {
+      "epoch": 0.13631937682570594,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00019906028074911645,
+      "loss": 0.9614,
+      "step": 245
+    },
+    {
+      "epoch": 0.1368757824454027,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00019905227880562246,
+      "loss": 0.9454,
+      "step": 246
+    },
+    {
+      "epoch": 0.13743218806509946,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0001990442430993468,
+      "loss": 0.8745,
+      "step": 247
+    },
+    {
+      "epoch": 0.1379885936847962,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00019903617363302846,
+      "loss": 0.9185,
+      "step": 248
+    },
+    {
+      "epoch": 0.138544999304493,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00019902807040941806,
+      "loss": 0.9842,
+      "step": 249
+    },
+    {
+      "epoch": 0.13910140492418974,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00019901993343127763,
+      "loss": 1.0057,
+      "step": 250
+    },
+    {
+      "epoch": 0.13965781054388648,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0001990117627013807,
+      "loss": 0.8891,
+      "step": 251
+    },
+    {
+      "epoch": 0.14021421616358326,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00019900355822251233,
+      "loss": 0.9715,
+      "step": 252
+    },
+    {
+      "epoch": 0.14077062178328,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0001989953199974691,
+      "loss": 0.8475,
+      "step": 253
+    },
+    {
+      "epoch": 0.14132702740297676,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0001989870480290591,
+      "loss": 0.9162,
+      "step": 254
+    },
+    {
+      "epoch": 0.14188343302267353,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00019897874232010186,
+      "loss": 0.8788,
+      "step": 255
+    },
+    {
+      "epoch": 0.14243983864237028,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.0001989704028734285,
+      "loss": 1.8645,
+      "step": 256
+    },
+    {
+      "epoch": 0.14299624426206706,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019896202969188148,
+      "loss": 1.0641,
+      "step": 257
+    },
+    {
+      "epoch": 0.1435526498817638,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.000198953622778315,
+      "loss": 1.0014,
+      "step": 258
+    },
+    {
+      "epoch": 0.14410905550146055,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00019894518213559457,
+      "loss": 1.0273,
+      "step": 259
+    },
+    {
+      "epoch": 0.14466546112115733,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0001989367077665972,
+      "loss": 0.9224,
+      "step": 260
+    },
+    {
+      "epoch": 0.14522186674085408,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00019892819967421154,
+      "loss": 0.9585,
+      "step": 261
+    },
+    {
+      "epoch": 0.14577827236055085,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0001989196578613376,
+      "loss": 0.9612,
+      "step": 262
+    },
+    {
+      "epoch": 0.1463346779802476,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0001989110823308869,
+      "loss": 0.8635,
+      "step": 263
+    },
+    {
+      "epoch": 0.14689108359994435,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0001989024730857825,
+      "loss": 0.9478,
+      "step": 264
+    },
+    {
+      "epoch": 0.14744748921964113,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00019889383012895896,
+      "loss": 1.0103,
+      "step": 265
+    },
+    {
+      "epoch": 0.14800389483933787,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00019888515346336226,
+      "loss": 0.8712,
+      "step": 266
+    },
+    {
+      "epoch": 0.14856030045903465,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00019887644309195,
+      "loss": 0.9338,
+      "step": 267
+    },
+    {
+      "epoch": 0.1491167060787314,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00019886769901769104,
+      "loss": 0.9435,
+      "step": 268
+    },
+    {
+      "epoch": 0.14967311169842815,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00019885892124356597,
+      "loss": 1.0054,
+      "step": 269
+    },
+    {
+      "epoch": 0.15022951731812492,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00019885010977256675,
+      "loss": 0.9563,
+      "step": 270
+    },
+    {
+      "epoch": 0.15078592293782167,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0001988412646076968,
+      "loss": 0.889,
+      "step": 271
+    },
+    {
+      "epoch": 0.15134232855751842,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00019883238575197113,
+      "loss": 0.8797,
+      "step": 272
+    },
+    {
+      "epoch": 0.1518987341772152,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00019882347320841615,
+      "loss": 0.9045,
+      "step": 273
+    },
+    {
+      "epoch": 0.15245513979691194,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00019881452698006973,
+      "loss": 0.8376,
+      "step": 274
+    },
+    {
+      "epoch": 0.15301154541660872,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0001988055470699813,
+      "loss": 0.9139,
+      "step": 275
+    },
+    {
+      "epoch": 0.15356795103630547,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00019879653348121176,
+      "loss": 0.9675,
+      "step": 276
+    },
+    {
+      "epoch": 0.15412435665600221,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00019878748621683344,
+      "loss": 0.8359,
+      "step": 277
+    },
+    {
+      "epoch": 0.154680762275699,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00019877840527993018,
+      "loss": 0.9296,
+      "step": 278
+    },
+    {
+      "epoch": 0.15523716789539574,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00019876929067359726,
+      "loss": 0.9716,
+      "step": 279
+    },
+    {
+      "epoch": 0.15579357351509251,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00019876014240094154,
+      "loss": 0.9313,
+      "step": 280
+    },
+    {
+      "epoch": 0.15634997913478926,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0001987509604650812,
+      "loss": 0.9064,
+      "step": 281
+    },
+    {
+      "epoch": 0.156906384754486,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00019874174486914602,
+      "loss": 1.0199,
+      "step": 282
+    },
+    {
+      "epoch": 0.1574627903741828,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00019873249561627723,
+      "loss": 0.9184,
+      "step": 283
+    },
+    {
+      "epoch": 0.15801919599387954,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001987232127096275,
+      "loss": 0.8171,
+      "step": 284
+    },
+    {
+      "epoch": 0.15857560161357628,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00019871389615236094,
+      "loss": 0.9977,
+      "step": 285
+    },
+    {
+      "epoch": 0.15913200723327306,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00019870454594765323,
+      "loss": 0.9063,
+      "step": 286
+    },
+    {
+      "epoch": 0.1596884128529698,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00019869516209869144,
+      "loss": 1.0768,
+      "step": 287
+    },
+    {
+      "epoch": 0.16024481847266658,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00019868574460867416,
+      "loss": 0.9082,
+      "step": 288
+    },
+    {
+      "epoch": 0.16080122409236333,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00019867629348081138,
+      "loss": 0.9355,
+      "step": 289
+    },
+    {
+      "epoch": 0.16135762971206008,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0001986668087183246,
+      "loss": 0.8966,
+      "step": 290
+    },
+    {
+      "epoch": 0.16191403533175686,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0001986572903244468,
+      "loss": 0.8964,
+      "step": 291
+    },
+    {
+      "epoch": 0.1624704409514536,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00019864773830242244,
+      "loss": 0.9721,
+      "step": 292
+    },
+    {
+      "epoch": 0.16302684657115038,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0001986381526555073,
+      "loss": 0.9809,
+      "step": 293
+    },
+    {
+      "epoch": 0.16358325219084713,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0001986285333869688,
+      "loss": 0.9851,
+      "step": 294
+    },
+    {
+      "epoch": 0.16413965781054388,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00019861888050008575,
+      "loss": 0.8559,
+      "step": 295
+    },
+    {
+      "epoch": 0.16469606343024065,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00019860919399814837,
+      "loss": 0.9342,
+      "step": 296
+    },
+    {
+      "epoch": 0.1652524690499374,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0001985994738844584,
+      "loss": 0.9079,
+      "step": 297
+    },
+    {
+      "epoch": 0.16580887466963418,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00019858972016232907,
+      "loss": 0.8145,
+      "step": 298
+    },
+    {
+      "epoch": 0.16636528028933092,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00019857993283508497,
+      "loss": 0.9047,
+      "step": 299
+    },
+    {
+      "epoch": 0.16692168590902767,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00019857011190606215,
+      "loss": 0.9684,
+      "step": 300
+    },
+    {
+      "epoch": 0.16747809152872445,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00019856025737860828,
+      "loss": 1.0272,
+      "step": 301
+    },
+    {
+      "epoch": 0.1680344971484212,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0001985503692560822,
+      "loss": 0.9,
+      "step": 302
+    },
+    {
+      "epoch": 0.16859090276811795,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0001985404475418545,
+      "loss": 1.0117,
+      "step": 303
+    },
+    {
+      "epoch": 0.16914730838781472,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00019853049223930697,
+      "loss": 0.8626,
+      "step": 304
+    },
+    {
+      "epoch": 0.16970371400751147,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00019852050335183305,
+      "loss": 0.9383,
+      "step": 305
+    },
+    {
+      "epoch": 0.17026011962720825,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00019851048088283742,
+      "loss": 0.9119,
+      "step": 306
+    },
+    {
+      "epoch": 0.170816525246905,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00019850042483573644,
+      "loss": 0.9732,
+      "step": 307
+    },
+    {
+      "epoch": 0.17137293086660174,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00019849033521395773,
+      "loss": 0.9586,
+      "step": 308
+    },
+    {
+      "epoch": 0.17192933648629852,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00019848021202094042,
+      "loss": 0.943,
+      "step": 309
+    },
+    {
+      "epoch": 0.17248574210599527,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0001984700552601351,
+      "loss": 0.9244,
+      "step": 310
+    },
+    {
+      "epoch": 0.17304214772569204,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00019845986493500378,
+      "loss": 1.0124,
+      "step": 311
+    },
+    {
+      "epoch": 0.1735985533453888,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00019844964104901992,
+      "loss": 1.0024,
+      "step": 312
+    },
+    {
+      "epoch": 0.17415495896508554,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00019843938360566843,
+      "loss": 0.856,
+      "step": 313
+    },
+    {
+      "epoch": 0.1747113645847823,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0001984290926084456,
+      "loss": 0.9451,
+      "step": 314
+    },
+    {
+      "epoch": 0.17526777020447906,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00019841876806085923,
+      "loss": 0.9995,
+      "step": 315
+    },
+    {
+      "epoch": 0.17582417582417584,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00019840840996642852,
+      "loss": 0.9539,
+      "step": 316
+    },
+    {
+      "epoch": 0.17638058144387259,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00019839801832868413,
+      "loss": 0.9055,
+      "step": 317
+    },
+    {
+      "epoch": 0.17693698706356933,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0001983875931511681,
+      "loss": 0.8727,
+      "step": 318
+    },
+    {
+      "epoch": 0.1774933926832661,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00019837713443743397,
+      "loss": 0.899,
+      "step": 319
+    },
+    {
+      "epoch": 0.17804979830296286,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00019836664219104666,
+      "loss": 0.8607,
+      "step": 320
+    },
+    {
+      "epoch": 0.1786062039226596,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00019835611641558253,
+      "loss": 1.0124,
+      "step": 321
+    },
+    {
+      "epoch": 0.17916260954235638,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0001983455571146294,
+      "loss": 1.0013,
+      "step": 322
+    },
+    {
+      "epoch": 0.17971901516205313,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00019833496429178652,
+      "loss": 0.8477,
+      "step": 323
+    },
+    {
+      "epoch": 0.1802754207817499,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00019832433795066446,
+      "loss": 0.889,
+      "step": 324
+    },
+    {
+      "epoch": 0.18083182640144665,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00019831367809488537,
+      "loss": 0.789,
+      "step": 325
+    },
+    {
+      "epoch": 0.1813882320211434,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0001983029847280827,
+      "loss": 0.8853,
+      "step": 326
+    },
+    {
+      "epoch": 0.18194463764084018,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0001982922578539014,
+      "loss": 0.9427,
+      "step": 327
+    },
+    {
+      "epoch": 0.18250104326053693,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0001982814974759978,
+      "loss": 0.8221,
+      "step": 328
+    },
+    {
+      "epoch": 0.1830574488802337,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00019827070359803968,
+      "loss": 0.9266,
+      "step": 329
+    },
+    {
+      "epoch": 0.18361385449993045,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0001982598762237062,
+      "loss": 1.0117,
+      "step": 330
+    },
+    {
+      "epoch": 0.1841702601196272,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00019824901535668796,
+      "loss": 0.9188,
+      "step": 331
+    },
+    {
+      "epoch": 0.18472666573932398,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.000198238121000687,
+      "loss": 0.952,
+      "step": 332
+    },
+    {
+      "epoch": 0.18528307135902072,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00019822719315941672,
+      "loss": 0.9938,
+      "step": 333
+    },
+    {
+      "epoch": 0.18583947697871747,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.000198216231836602,
+      "loss": 0.8886,
+      "step": 334
+    },
+    {
+      "epoch": 0.18639588259841425,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00019820523703597908,
+      "loss": 0.8996,
+      "step": 335
+    },
+    {
+      "epoch": 0.186952288218111,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0001981942087612956,
+      "loss": 0.8635,
+      "step": 336
+    },
+    {
+      "epoch": 0.18750869383780777,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0001981831470163107,
+      "loss": 1.029,
+      "step": 337
+    },
+    {
+      "epoch": 0.18806509945750452,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0001981720518047948,
+      "loss": 0.9771,
+      "step": 338
+    },
+    {
+      "epoch": 0.18862150507720127,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0001981609231305298,
+      "loss": 0.919,
+      "step": 339
+    },
+    {
+      "epoch": 0.18917791069689804,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00019814976099730904,
+      "loss": 0.9555,
+      "step": 340
+    },
+    {
+      "epoch": 0.1897343163165948,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00019813856540893722,
+      "loss": 1.0713,
+      "step": 341
+    },
+    {
+      "epoch": 0.19029072193629157,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00019812733636923044,
+      "loss": 0.9679,
+      "step": 342
+    },
+    {
+      "epoch": 0.19084712755598832,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0001981160738820162,
+      "loss": 0.9542,
+      "step": 343
+    },
+    {
+      "epoch": 0.19140353317568506,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0001981047779511334,
+      "loss": 0.991,
+      "step": 344
+    },
+    {
+      "epoch": 0.19195993879538184,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0001980934485804324,
+      "loss": 0.9589,
+      "step": 345
+    },
+    {
+      "epoch": 0.1925163444150786,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00019808208577377486,
+      "loss": 0.9242,
+      "step": 346
+    },
+    {
+      "epoch": 0.19307275003477536,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0001980706895350339,
+      "loss": 0.9109,
+      "step": 347
+    },
+    {
+      "epoch": 0.1936291556544721,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00019805925986809403,
+      "loss": 1.0326,
+      "step": 348
+    },
+    {
+      "epoch": 0.19418556127416886,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00019804779677685112,
+      "loss": 0.9841,
+      "step": 349
+    },
+    {
+      "epoch": 0.19474196689386564,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0001980363002652125,
+      "loss": 0.9203,
+      "step": 350
+    },
+    {
+      "epoch": 0.19529837251356238,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0001980247703370968,
+      "loss": 0.8824,
+      "step": 351
+    },
+    {
+      "epoch": 0.19585477813325913,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00019801320699643416,
+      "loss": 0.7956,
+      "step": 352
+    },
+    {
+      "epoch": 0.1964111837529559,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00019800161024716598,
+      "loss": 0.922,
+      "step": 353
+    },
+    {
+      "epoch": 0.19696758937265266,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00019798998009324512,
+      "loss": 0.9232,
+      "step": 354
+    },
+    {
+      "epoch": 0.19752399499234943,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00019797831653863582,
+      "loss": 0.8868,
+      "step": 355
+    },
+    {
+      "epoch": 0.19808040061204618,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0001979666195873137,
+      "loss": 0.9103,
+      "step": 356
+    },
+    {
+      "epoch": 0.19863680623174293,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00019795488924326574,
+      "loss": 0.8998,
+      "step": 357
+    },
+    {
+      "epoch": 0.1991932118514397,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00019794312551049039,
+      "loss": 0.8973,
+      "step": 358
+    },
+    {
+      "epoch": 0.19974961747113645,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00019793132839299732,
+      "loss": 0.9458,
+      "step": 359
+    },
+    {
+      "epoch": 0.20030602309083323,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00019791949789480775,
+      "loss": 1.0083,
+      "step": 360
+    },
+    {
+      "epoch": 0.20086242871052998,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00019790763401995415,
+      "loss": 0.9661,
+      "step": 361
+    },
+    {
+      "epoch": 0.20141883433022673,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00019789573677248047,
+      "loss": 0.8475,
+      "step": 362
+    },
+    {
+      "epoch": 0.2019752399499235,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00019788380615644196,
+      "loss": 0.8632,
+      "step": 363
+    },
+    {
+      "epoch": 0.20253164556962025,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00019787184217590523,
+      "loss": 0.9522,
+      "step": 364
+    },
+    {
+      "epoch": 0.20308805118931703,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00019785984483494838,
+      "loss": 0.8749,
+      "step": 365
+    },
+    {
+      "epoch": 0.20364445680901377,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00019784781413766073,
+      "loss": 0.8714,
+      "step": 366
+    },
+    {
+      "epoch": 0.20420086242871052,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0001978357500881431,
+      "loss": 0.9581,
+      "step": 367
+    },
+    {
+      "epoch": 0.2047572680484073,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00019782365269050756,
+      "loss": 1.0496,
+      "step": 368
+    },
+    {
+      "epoch": 0.20531367366810405,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00019781152194887764,
+      "loss": 0.8804,
+      "step": 369
+    },
+    {
+      "epoch": 0.2058700792878008,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00019779935786738822,
+      "loss": 0.8474,
+      "step": 370
+    },
+    {
+      "epoch": 0.20642648490749757,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0001977871604501855,
+      "loss": 0.8413,
+      "step": 371
+    },
+    {
+      "epoch": 0.20698289052719432,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00019777492970142707,
+      "loss": 0.8636,
+      "step": 372
+    },
+    {
+      "epoch": 0.2075392961468911,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00019776266562528192,
+      "loss": 0.8351,
+      "step": 373
+    },
+    {
+      "epoch": 0.20809570176658784,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001977503682259303,
+      "loss": 0.8751,
+      "step": 374
+    },
+    {
+      "epoch": 0.2086521073862846,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00019773803750756393,
+      "loss": 0.9113,
+      "step": 375
+    },
+    {
+      "epoch": 0.20920851300598137,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0001977256734743858,
+      "loss": 0.9433,
+      "step": 376
+    },
+    {
+      "epoch": 0.20976491862567812,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0001977132761306103,
+      "loss": 0.8863,
+      "step": 377
+    },
+    {
+      "epoch": 0.2103213242453749,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00019770084548046318,
+      "loss": 0.8073,
+      "step": 378
+    },
+    {
+      "epoch": 0.21087772986507164,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00019768838152818153,
+      "loss": 0.9198,
+      "step": 379
+    },
+    {
+      "epoch": 0.2114341354847684,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0001976758842780138,
+      "loss": 0.9204,
+      "step": 380
+    },
+    {
+      "epoch": 0.21199054110446516,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00019766335373421973,
+      "loss": 0.9714,
+      "step": 381
+    },
+    {
+      "epoch": 0.2125469467241619,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00019765078990107048,
+      "loss": 1.0368,
+      "step": 382
+    },
+    {
+      "epoch": 0.21310335234385866,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0001976381927828486,
+      "loss": 0.8881,
+      "step": 383
+    },
+    {
+      "epoch": 0.21365975796355544,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00019762556238384784,
+      "loss": 1.0236,
+      "step": 384
+    },
+    {
+      "epoch": 0.21421616358325218,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0001976128987083734,
+      "loss": 0.9628,
+      "step": 385
+    },
+    {
+      "epoch": 0.21477256920294896,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0001976002017607418,
+      "loss": 0.8861,
+      "step": 386
+    },
+    {
+      "epoch": 0.2153289748226457,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00019758747154528092,
+      "loss": 1.0946,
+      "step": 387
+    },
+    {
+      "epoch": 0.21588538044234246,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00019757470806632994,
+      "loss": 1.0047,
+      "step": 388
+    },
+    {
+      "epoch": 0.21644178606203923,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00019756191132823937,
+      "loss": 0.8831,
+      "step": 389
+    },
+    {
+      "epoch": 0.21699819168173598,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00019754908133537113,
+      "loss": 0.8999,
+      "step": 390
+    },
+    {
+      "epoch": 0.21755459730143276,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00019753621809209842,
+      "loss": 0.8368,
+      "step": 391
+    },
+    {
+      "epoch": 0.2181110029211295,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00019752332160280576,
+      "loss": 0.9262,
+      "step": 392
+    },
+    {
+      "epoch": 0.21866740854082625,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.000197510391871889,
+      "loss": 0.9496,
+      "step": 393
+    },
+    {
+      "epoch": 0.21922381416052303,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0001974974289037554,
+      "loss": 0.9303,
+      "step": 394
+    },
+    {
+      "epoch": 0.21978021978021978,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0001974844327028235,
+      "loss": 0.924,
+      "step": 395
+    },
+    {
+      "epoch": 0.22033662539991655,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00019747140327352306,
+      "loss": 0.9763,
+      "step": 396
+    },
+    {
+      "epoch": 0.2208930310196133,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001974583406202954,
+      "loss": 0.7589,
+      "step": 397
+    },
+    {
+      "epoch": 0.22144943663931005,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00019744524474759296,
+      "loss": 0.9391,
+      "step": 398
+    },
+    {
+      "epoch": 0.22200584225900682,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0001974321156598796,
+      "loss": 0.9961,
+      "step": 399
+    },
+    {
+      "epoch": 0.22256224787870357,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00019741895336163046,
+      "loss": 0.9214,
+      "step": 400
+    },
+    {
+      "epoch": 0.22311865349840032,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.000197405757857332,
+      "loss": 0.8915,
+      "step": 401
+    },
+    {
+      "epoch": 0.2236750591180971,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00019739252915148208,
+      "loss": 0.8334,
+      "step": 402
+    },
+    {
+      "epoch": 0.22423146473779385,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00019737926724858976,
+      "loss": 0.9007,
+      "step": 403
+    },
+    {
+      "epoch": 0.22478787035749062,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0001973659721531755,
+      "loss": 0.8694,
+      "step": 404
+    },
+    {
+      "epoch": 0.22534427597718737,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00019735264386977098,
+      "loss": 0.9033,
+      "step": 405
+    },
+    {
+      "epoch": 0.22590068159688412,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00019733928240291932,
+      "loss": 0.8603,
+      "step": 406
+    },
+    {
+      "epoch": 0.2264570872165809,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00019732588775717486,
+      "loss": 1.024,
+      "step": 407
+    },
+    {
+      "epoch": 0.22701349283627764,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00019731245993710327,
+      "loss": 0.8826,
+      "step": 408
+    },
+    {
+      "epoch": 0.22756989845597442,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00019729899894728158,
+      "loss": 0.8364,
+      "step": 409
+    },
+    {
+      "epoch": 0.22812630407567117,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.000197285504792298,
+      "loss": 1.0239,
+      "step": 410
+    },
+    {
+      "epoch": 0.22868270969536791,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00019727197747675218,
+      "loss": 0.9286,
+      "step": 411
+    },
+    {
+      "epoch": 0.2292391153150647,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.000197258417005255,
+      "loss": 0.8623,
+      "step": 412
+    },
+    {
+      "epoch": 0.22979552093476144,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00019724482338242866,
+      "loss": 0.8571,
+      "step": 413
+    },
+    {
+      "epoch": 0.2303519265544582,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00019723119661290664,
+      "loss": 1.0023,
+      "step": 414
+    },
+    {
+      "epoch": 0.23090833217415496,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00019721753670133376,
+      "loss": 0.8866,
+      "step": 415
+    },
+    {
+      "epoch": 0.2314647377938517,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001972038436523661,
+      "loss": 0.7956,
+      "step": 416
+    },
+    {
+      "epoch": 0.2320211434135485,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00019719011747067108,
+      "loss": 0.9645,
+      "step": 417
+    },
+    {
+      "epoch": 0.23257754903324523,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00019717635816092732,
+      "loss": 0.9324,
+      "step": 418
+    },
+    {
+      "epoch": 0.23313395465294198,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00019716256572782482,
+      "loss": 0.9626,
+      "step": 419
+    },
+    {
+      "epoch": 0.23369036027263876,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0001971487401760649,
+      "loss": 0.8419,
+      "step": 420
+    },
+    {
+      "epoch": 0.2342467658923355,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00019713488151036003,
+      "loss": 0.8471,
+      "step": 421
+    },
+    {
+      "epoch": 0.23480317151203228,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00019712098973543408,
+      "loss": 1.0177,
+      "step": 422
+    },
+    {
+      "epoch": 0.23535957713172903,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0001971070648560222,
+      "loss": 0.944,
+      "step": 423
+    },
+    {
+      "epoch": 0.23591598275142578,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001970931068768708,
+      "loss": 0.8007,
+      "step": 424
+    },
+    {
+      "epoch": 0.23647238837112255,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0001970791158027375,
+      "loss": 0.9658,
+      "step": 425
+    },
+    {
+      "epoch": 0.2370287939908193,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00019706509163839136,
+      "loss": 0.8966,
+      "step": 426
+    },
+    {
+      "epoch": 0.23758519961051608,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0001970510343886126,
+      "loss": 1.0448,
+      "step": 427
+    },
+    {
+      "epoch": 0.23814160523021283,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0001970369440581927,
+      "loss": 0.937,
+      "step": 428
+    },
+    {
+      "epoch": 0.23869801084990958,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00019702282065193455,
+      "loss": 0.9181,
+      "step": 429
+    },
+    {
+      "epoch": 0.23925441646960635,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00019700866417465216,
+      "loss": 0.9101,
+      "step": 430
+    },
+    {
+      "epoch": 0.2398108220893031,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00019699447463117093,
+      "loss": 0.9438,
+      "step": 431
+    },
+    {
+      "epoch": 0.24036722770899985,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00019698025202632743,
+      "loss": 0.8315,
+      "step": 432
+    },
+    {
+      "epoch": 0.24092363332869662,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00019696599636496962,
+      "loss": 0.8385,
+      "step": 433
+    },
+    {
+      "epoch": 0.24148003894839337,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0001969517076519566,
+      "loss": 0.8667,
+      "step": 434
+    },
+    {
+      "epoch": 0.24203644456809015,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00019693738589215883,
+      "loss": 0.881,
+      "step": 435
+    },
+    {
+      "epoch": 0.2425928501877869,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00019692303109045795,
+      "loss": 0.9509,
+      "step": 436
+    },
+    {
+      "epoch": 0.24314925580748364,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00019690864325174699,
+      "loss": 0.9065,
+      "step": 437
+    },
+    {
+      "epoch": 0.24370566142718042,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0001968942223809301,
+      "loss": 1.072,
+      "step": 438
+    },
+    {
+      "epoch": 0.24426206704687717,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0001968797684829228,
+      "loss": 0.9068,
+      "step": 439
+    },
+    {
+      "epoch": 0.24481847266657394,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00019686528156265175,
+      "loss": 0.8168,
+      "step": 440
+    },
+    {
+      "epoch": 0.2453748782862707,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.000196850761625055,
+      "loss": 0.8748,
+      "step": 441
+    },
+    {
+      "epoch": 0.24593128390596744,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00019683620867508176,
+      "loss": 0.7686,
+      "step": 442
+    },
+    {
+      "epoch": 0.24648768952566422,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00019682162271769256,
+      "loss": 0.9036,
+      "step": 443
+    },
+    {
+      "epoch": 0.24704409514536096,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0001968070037578591,
+      "loss": 0.9472,
+      "step": 444
+    },
+    {
+      "epoch": 0.24760050076505774,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00019679235180056437,
+      "loss": 0.8882,
+      "step": 445
+    },
+    {
+      "epoch": 0.2481569063847545,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00019677766685080264,
+      "loss": 0.9693,
+      "step": 446
+    },
+    {
+      "epoch": 0.24871331200445124,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00019676294891357937,
+      "loss": 0.8764,
+      "step": 447
+    },
+    {
+      "epoch": 0.249269717624148,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00019674819799391129,
+      "loss": 0.9671,
+      "step": 448
+    },
+    {
+      "epoch": 0.24982612324384476,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0001967334140968264,
+      "loss": 0.9398,
+      "step": 449
+    },
+    {
+      "epoch": 0.2503825288635415,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00019671859722736389,
+      "loss": 0.9572,
+      "step": 450
+    },
+    {
+      "epoch": 0.2509389344832383,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0001967037473905742,
+      "loss": 0.958,
+      "step": 451
+    },
+    {
+      "epoch": 0.25149534010293506,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00019668886459151908,
+      "loss": 0.8959,
+      "step": 452
+    },
+    {
+      "epoch": 0.2520517457226318,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00019667394883527139,
+      "loss": 0.9529,
+      "step": 453
+    },
+    {
+      "epoch": 0.25260815134232856,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00019665900012691529,
+      "loss": 0.9213,
+      "step": 454
+    },
+    {
+      "epoch": 0.25316455696202533,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0001966440184715462,
+      "loss": 0.952,
+      "step": 455
+    },
+    {
+      "epoch": 0.25372096258172205,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00019662900387427075,
+      "loss": 0.9548,
+      "step": 456
+    },
+    {
+      "epoch": 0.25427736820141883,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00019661395634020673,
+      "loss": 0.8359,
+      "step": 457
+    },
+    {
+      "epoch": 0.2548337738211156,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00019659887587448327,
+      "loss": 0.9204,
+      "step": 458
+    },
+    {
+      "epoch": 0.2553901794408123,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00019658376248224063,
+      "loss": 0.9216,
+      "step": 459
+    },
+    {
+      "epoch": 0.2559465850605091,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00019656861616863038,
+      "loss": 0.9383,
+      "step": 460
+    },
+    {
+      "epoch": 0.2565029906802059,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00019655343693881526,
+      "loss": 0.9428,
+      "step": 461
+    },
+    {
+      "epoch": 0.25705939629990265,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0001965382247979692,
+      "loss": 0.9357,
+      "step": 462
+    },
+    {
+      "epoch": 0.2576158019195994,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00019652297975127737,
+      "loss": 0.967,
+      "step": 463
+    },
+    {
+      "epoch": 0.25817220753929615,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0001965077018039362,
+      "loss": 0.8993,
+      "step": 464
+    },
+    {
+      "epoch": 0.2587286131589929,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00019649239096115336,
+      "loss": 1.0113,
+      "step": 465
+    },
+    {
+      "epoch": 0.25928501877868965,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00019647704722814755,
+      "loss": 0.9117,
+      "step": 466
+    },
+    {
+      "epoch": 0.2598414243983864,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00019646167061014892,
+      "loss": 0.9318,
+      "step": 467
+    },
+    {
+      "epoch": 0.2603978300180832,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00019644626111239863,
+      "loss": 0.9237,
+      "step": 468
+    },
+    {
+      "epoch": 0.2609542356377799,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0001964308187401492,
+      "loss": 0.9926,
+      "step": 469
+    },
+    {
+      "epoch": 0.2615106412574767,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001964153434986643,
+      "loss": 0.8647,
+      "step": 470
+    },
+    {
+      "epoch": 0.26206704687717347,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001963998353932187,
+      "loss": 0.7973,
+      "step": 471
+    },
+    {
+      "epoch": 0.2626234524968702,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00019638429442909853,
+      "loss": 0.8922,
+      "step": 472
+    },
+    {
+      "epoch": 0.26317985811656697,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00019636872061160109,
+      "loss": 0.9267,
+      "step": 473
+    },
+    {
+      "epoch": 0.26373626373626374,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00019635311394603476,
+      "loss": 0.8743,
+      "step": 474
+    },
+    {
+      "epoch": 0.2642926693559605,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00019633747443771924,
+      "loss": 0.7997,
+      "step": 475
+    },
+    {
+      "epoch": 0.26484907497565724,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00019632180209198542,
+      "loss": 0.9695,
+      "step": 476
+    },
+    {
+      "epoch": 0.265405480595354,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00019630609691417527,
+      "loss": 0.7992,
+      "step": 477
+    },
+    {
+      "epoch": 0.2659618862150508,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00019629035890964208,
+      "loss": 0.7188,
+      "step": 478
+    },
+    {
+      "epoch": 0.2665182918347475,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0001962745880837503,
+      "loss": 0.93,
+      "step": 479
+    },
+    {
+      "epoch": 0.2670746974544443,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00019625878444187551,
+      "loss": 0.8121,
+      "step": 480
+    },
+    {
+      "epoch": 0.26763110307414106,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00019624294798940451,
+      "loss": 0.9366,
+      "step": 481
+    },
+    {
+      "epoch": 0.2681875086938378,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0001962270787317353,
+      "loss": 0.94,
+      "step": 482
+    },
+    {
+      "epoch": 0.26874391431353456,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.000196211176674277,
+      "loss": 0.9317,
+      "step": 483
+    },
+    {
+      "epoch": 0.26930031993323134,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00019619524182245,
+      "loss": 0.8751,
+      "step": 484
+    },
+    {
+      "epoch": 0.2698567255529281,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00019617927418168587,
+      "loss": 0.9093,
+      "step": 485
+    },
+    {
+      "epoch": 0.27041313117262483,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00019616327375742722,
+      "loss": 0.8399,
+      "step": 486
+    },
+    {
+      "epoch": 0.2709695367923216,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00019614724055512799,
+      "loss": 0.9025,
+      "step": 487
+    },
+    {
+      "epoch": 0.2715259424120184,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0001961311745802532,
+      "loss": 0.8661,
+      "step": 488
+    },
+    {
+      "epoch": 0.2720823480317151,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00019611507583827903,
+      "loss": 0.8429,
+      "step": 489
+    },
+    {
+      "epoch": 0.2726387536514119,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00019609894433469295,
+      "loss": 0.9272,
+      "step": 490
+    },
+    {
+      "epoch": 0.27319515927110866,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0001960827800749935,
+      "loss": 0.87,
+      "step": 491
+    },
+    {
+      "epoch": 0.2737515648908054,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00019606658306469034,
+      "loss": 0.865,
+      "step": 492
+    },
+    {
+      "epoch": 0.27430797051050215,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00019605035330930442,
+      "loss": 0.8939,
+      "step": 493
+    },
+    {
+      "epoch": 0.27486437613019893,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00019603409081436773,
+      "loss": 0.9009,
+      "step": 494
+    },
+    {
+      "epoch": 0.27542078174989565,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0001960177955854235,
+      "loss": 0.8828,
+      "step": 495
+    },
+    {
+      "epoch": 0.2759771873695924,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00019600146762802613,
+      "loss": 0.9509,
+      "step": 496
+    },
+    {
+      "epoch": 0.2765335929892892,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00019598510694774108,
+      "loss": 0.7998,
+      "step": 497
+    },
+    {
+      "epoch": 0.277089998608986,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00019596871355014506,
+      "loss": 0.8909,
+      "step": 498
+    },
+    {
+      "epoch": 0.2776464042286827,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00019595228744082586,
+      "loss": 0.9643,
+      "step": 499
+    },
+    {
+      "epoch": 0.2782028098483795,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00019593582862538248,
+      "loss": 0.9311,
+      "step": 500
+    },
+    {
+      "epoch": 0.27875921546807625,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00019591933710942503,
+      "loss": 0.8344,
+      "step": 501
+    },
+    {
+      "epoch": 0.27931562108777297,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00019590281289857478,
+      "loss": 0.9109,
+      "step": 502
+    },
+    {
+      "epoch": 0.27987202670746975,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00019588625599846417,
+      "loss": 0.994,
+      "step": 503
+    },
+    {
+      "epoch": 0.2804284323271665,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00019586966641473673,
+      "loss": 0.8917,
+      "step": 504
+    },
+    {
+      "epoch": 0.28098483794686324,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00019585304415304711,
+      "loss": 0.8417,
+      "step": 505
+    },
+    {
+      "epoch": 0.28154124356656,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00019583638921906124,
+      "loss": 0.9425,
+      "step": 506
+    },
+    {
+      "epoch": 0.2820976491862568,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00019581970161845603,
+      "loss": 0.8792,
+      "step": 507
+    },
+    {
+      "epoch": 0.2826540548059535,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0001958029813569196,
+      "loss": 0.9791,
+      "step": 508
+    },
+    {
+      "epoch": 0.2832104604256503,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00019578622844015117,
+      "loss": 0.8546,
+      "step": 509
+    },
+    {
+      "epoch": 0.28376686604534707,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00019576944287386113,
+      "loss": 0.8836,
+      "step": 510
+    },
+    {
+      "epoch": 0.28432327166504384,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.000195752624663771,
+      "loss": 0.8764,
+      "step": 511
+    },
+    {
+      "epoch": 0.28487967728474056,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00019573577381561336,
+      "loss": 0.9411,
+      "step": 512
+    },
+    {
+      "epoch": 0.28543608290443734,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00019571889033513198,
+      "loss": 0.9222,
+      "step": 513
+    },
+    {
+      "epoch": 0.2859924885241341,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00019570197422808173,
+      "loss": 0.9914,
+      "step": 514
+    },
+    {
+      "epoch": 0.28654889414383083,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00019568502550022858,
+      "loss": 0.8788,
+      "step": 515
+    },
+    {
+      "epoch": 0.2871052997635276,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001956680441573497,
+      "loss": 0.811,
+      "step": 516
+    },
+    {
+      "epoch": 0.2876617053832244,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00019565103020523328,
+      "loss": 0.9191,
+      "step": 517
+    },
+    {
+      "epoch": 0.2882181110029211,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00019563398364967868,
+      "loss": 0.9129,
+      "step": 518
+    },
+    {
+      "epoch": 0.2887745166226179,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00019561690449649636,
+      "loss": 0.9591,
+      "step": 519
+    },
+    {
+      "epoch": 0.28933092224231466,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00019559979275150783,
+      "loss": 0.8129,
+      "step": 520
+    },
+    {
+      "epoch": 0.2898873278620114,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00019558264842054585,
+      "loss": 0.8881,
+      "step": 521
+    },
+    {
+      "epoch": 0.29044373348170816,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00019556547150945418,
+      "loss": 0.8941,
+      "step": 522
+    },
+    {
+      "epoch": 0.29100013910140493,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0001955482620240877,
+      "loss": 0.8872,
+      "step": 523
+    },
+    {
+      "epoch": 0.2915565447211017,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001955310199703124,
+      "loss": 0.7887,
+      "step": 524
+    },
+    {
+      "epoch": 0.2921129503407984,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00019551374535400543,
+      "loss": 0.8684,
+      "step": 525
+    },
+    {
+      "epoch": 0.2926693559604952,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001954964381810549,
+      "loss": 0.7739,
+      "step": 526
+    },
+    {
+      "epoch": 0.293225761580192,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00019547909845736013,
+      "loss": 0.8568,
+      "step": 527
+    },
+    {
+      "epoch": 0.2937821671998887,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00019546172618883155,
+      "loss": 0.9188,
+      "step": 528
+    },
+    {
+      "epoch": 0.2943385728195855,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00019544432138139062,
+      "loss": 0.8119,
+      "step": 529
+    },
+    {
+      "epoch": 0.29489497843928225,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00019542688404096986,
+      "loss": 0.9192,
+      "step": 530
+    },
+    {
+      "epoch": 0.29545138405897897,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.000195409414173513,
+      "loss": 0.8616,
+      "step": 531
+    },
+    {
+      "epoch": 0.29600778967867575,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00019539191178497472,
+      "loss": 0.9191,
+      "step": 532
+    },
+    {
+      "epoch": 0.2965641952983725,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00019537437688132094,
+      "loss": 0.7765,
+      "step": 533
+    },
+    {
+      "epoch": 0.2971206009180693,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00019535680946852852,
+      "loss": 0.9185,
+      "step": 534
+    },
+    {
+      "epoch": 0.297677006537766,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00019533920955258546,
+      "loss": 0.8514,
+      "step": 535
+    },
+    {
+      "epoch": 0.2982334121574628,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00019532157713949083,
+      "loss": 1.8071,
+      "step": 536
+    },
+    {
+      "epoch": 0.29878981777715957,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00019530391223525478,
+      "loss": 0.9006,
+      "step": 537
+    },
+    {
+      "epoch": 0.2993462233968563,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00019528621484589859,
+      "loss": 0.9186,
+      "step": 538
+    },
+    {
+      "epoch": 0.29990262901655307,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00019526848497745448,
+      "loss": 0.8292,
+      "step": 539
+    },
+    {
+      "epoch": 0.30045903463624984,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00019525072263596585,
+      "loss": 0.7303,
+      "step": 540
+    },
+    {
+      "epoch": 0.30101544025594656,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0001952329278274872,
+      "loss": 0.9214,
+      "step": 541
+    },
+    {
+      "epoch": 0.30157184587564334,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00019521510055808396,
+      "loss": 0.9173,
+      "step": 542
+    },
+    {
+      "epoch": 0.3021282514953401,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00019519724083383273,
+      "loss": 0.8349,
+      "step": 543
+    },
+    {
+      "epoch": 0.30268465711503684,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00019517934866082116,
+      "loss": 0.9247,
+      "step": 544
+    },
+    {
+      "epoch": 0.3032410627347336,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0001951614240451479,
+      "loss": 0.8788,
+      "step": 545
+    },
+    {
+      "epoch": 0.3037974683544304,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00019514346699292274,
+      "loss": 0.8521,
+      "step": 546
+    },
+    {
+      "epoch": 0.30435387397412716,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00019512547751026648,
+      "loss": 0.8488,
+      "step": 547
+    },
+    {
+      "epoch": 0.3049102795938239,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00019510745560331097,
+      "loss": 0.9655,
+      "step": 548
+    },
+    {
+      "epoch": 0.30546668521352066,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00019508940127819913,
+      "loss": 1.0282,
+      "step": 549
+    },
+    {
+      "epoch": 0.30602309083321744,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00019507131454108497,
+      "loss": 1.0152,
+      "step": 550
+    },
+    {
+      "epoch": 0.30657949645291416,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001950531953981334,
+      "loss": 0.8472,
+      "step": 551
+    },
+    {
+      "epoch": 0.30713590207261093,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0001950350438555206,
+      "loss": 0.8322,
+      "step": 552
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00019501685991943357,
+      "loss": 0.9437,
+      "step": 553
+    },
+    {
+      "epoch": 0.30824871331200443,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001949986435960705,
+      "loss": 0.817,
+      "step": 554
+    },
+    {
+      "epoch": 0.3088051189317012,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019498039489164058,
+      "loss": 0.9954,
+      "step": 555
+    },
+    {
+      "epoch": 0.309361524551398,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00019496211381236401,
+      "loss": 0.8284,
+      "step": 556
+    },
+    {
+      "epoch": 0.3099179301710947,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0001949438003644721,
+      "loss": 0.8699,
+      "step": 557
+    },
+    {
+      "epoch": 0.3104743357907915,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00019492545455420703,
+      "loss": 0.8738,
+      "step": 558
+    },
+    {
+      "epoch": 0.31103074141048825,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0001949070763878222,
+      "loss": 0.894,
+      "step": 559
+    },
+    {
+      "epoch": 0.31158714703018503,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00019488866587158196,
+      "loss": 0.8258,
+      "step": 560
+    },
+    {
+      "epoch": 0.31214355264988175,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00019487022301176168,
+      "loss": 0.9347,
+      "step": 561
+    },
+    {
+      "epoch": 0.3126999582695785,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00019485174781464773,
+      "loss": 0.9083,
+      "step": 562
+    },
+    {
+      "epoch": 0.3132563638892753,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00019483324028653754,
+      "loss": 0.784,
+      "step": 563
+    },
+    {
+      "epoch": 0.313812769508972,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0001948147004337396,
+      "loss": 0.8954,
+      "step": 564
+    },
+    {
+      "epoch": 0.3143691751286688,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0001947961282625733,
+      "loss": 0.8873,
+      "step": 565
+    },
+    {
+      "epoch": 0.3149255807483656,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00019477752377936917,
+      "loss": 0.9064,
+      "step": 566
+    },
+    {
+      "epoch": 0.3154819863680623,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00019475888699046866,
+      "loss": 0.8424,
+      "step": 567
+    },
+    {
+      "epoch": 0.31603839198775907,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0001947402179022243,
+      "loss": 0.9908,
+      "step": 568
+    },
+    {
+      "epoch": 0.31659479760745585,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00019472151652099963,
+      "loss": 0.9116,
+      "step": 569
+    },
+    {
+      "epoch": 0.31715120322715257,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001947027828531691,
+      "loss": 0.9074,
+      "step": 570
+    },
+    {
+      "epoch": 0.31770760884684934,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00019468401690511832,
+      "loss": 0.9643,
+      "step": 571
+    },
+    {
+      "epoch": 0.3182640144665461,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00019466521868324375,
+      "loss": 0.9002,
+      "step": 572
+    },
+    {
+      "epoch": 0.3188204200862429,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0001946463881939529,
+      "loss": 0.9181,
+      "step": 573
+    },
+    {
+      "epoch": 0.3193768257059396,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00019462752544366436,
+      "loss": 0.9275,
+      "step": 574
+    },
+    {
+      "epoch": 0.3199332313256364,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0001946086304388076,
+      "loss": 0.8565,
+      "step": 575
+    },
+    {
+      "epoch": 0.32048963694533317,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0001945897031858232,
+      "loss": 0.9334,
+      "step": 576
+    },
+    {
+      "epoch": 0.3210460425650299,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0001945707436911626,
+      "loss": 0.9068,
+      "step": 577
+    },
+    {
+      "epoch": 0.32160244818472666,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00019455175196128838,
+      "loss": 0.8463,
+      "step": 578
+    },
+    {
+      "epoch": 0.32215885380442344,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00019453272800267393,
+      "loss": 0.8119,
+      "step": 579
+    },
+    {
+      "epoch": 0.32271525942412016,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00019451367182180378,
+      "loss": 0.9386,
+      "step": 580
+    },
+    {
+      "epoch": 0.32327166504381694,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00019449458342517338,
+      "loss": 0.9527,
+      "step": 581
+    },
+    {
+      "epoch": 0.3238280706635137,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00019447546281928916,
+      "loss": 0.9448,
+      "step": 582
+    },
+    {
+      "epoch": 0.3243844762832105,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00019445631001066853,
+      "loss": 0.8528,
+      "step": 583
+    },
+    {
+      "epoch": 0.3249408819029072,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00019443712500583993,
+      "loss": 0.9443,
+      "step": 584
+    },
+    {
+      "epoch": 0.325497287522604,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00019441790781134265,
+      "loss": 0.9369,
+      "step": 585
+    },
+    {
+      "epoch": 0.32605369314230076,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0001943986584337271,
+      "loss": 0.8631,
+      "step": 586
+    },
+    {
+      "epoch": 0.3266100987619975,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0001943793768795545,
+      "loss": 0.921,
+      "step": 587
+    },
+    {
+      "epoch": 0.32716650438169426,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0001943600631553972,
+      "loss": 0.8827,
+      "step": 588
+    },
+    {
+      "epoch": 0.32772291000139103,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00019434071726783843,
+      "loss": 0.9897,
+      "step": 589
+    },
+    {
+      "epoch": 0.32827931562108775,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0001943213392234724,
+      "loss": 0.7868,
+      "step": 590
+    },
+    {
+      "epoch": 0.32883572124078453,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00019430192902890423,
+      "loss": 0.8787,
+      "step": 591
+    },
+    {
+      "epoch": 0.3293921268604813,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0001942824866907501,
+      "loss": 0.861,
+      "step": 592
+    },
+    {
+      "epoch": 0.329948532480178,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00019426301221563702,
+      "loss": 0.8415,
+      "step": 593
+    },
+    {
+      "epoch": 0.3305049380998748,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00019424350561020312,
+      "loss": 0.9665,
+      "step": 594
+    },
+    {
+      "epoch": 0.3310613437195716,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0001942239668810973,
+      "loss": 0.934,
+      "step": 595
+    },
+    {
+      "epoch": 0.33161774933926835,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00019420439603497956,
+      "loss": 0.9518,
+      "step": 596
+    },
+    {
+      "epoch": 0.3321741549589651,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00019418479307852074,
+      "loss": 0.8112,
+      "step": 597
+    },
+    {
+      "epoch": 0.33273056057866185,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001941651580184027,
+      "loss": 0.7649,
+      "step": 598
+    },
+    {
+      "epoch": 0.3332869661983586,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00019414549086131815,
+      "loss": 0.9321,
+      "step": 599
+    },
+    {
+      "epoch": 0.33384337181805535,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0001941257916139709,
+      "loss": 0.9537,
+      "step": 600
+    },
+    {
+      "epoch": 0.3343997774377521,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0001941060602830755,
+      "loss": 0.961,
+      "step": 601
+    },
+    {
+      "epoch": 0.3349561830574489,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00019408629687535763,
+      "loss": 0.9001,
+      "step": 602
+    },
+    {
+      "epoch": 0.3355125886771456,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00019406650139755375,
+      "loss": 0.8772,
+      "step": 603
+    },
+    {
+      "epoch": 0.3360689942968424,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00019404667385641128,
+      "loss": 0.9444,
+      "step": 604
+    },
+    {
+      "epoch": 0.33662539991653917,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001940268142586887,
+      "loss": 0.8729,
+      "step": 605
+    },
+    {
+      "epoch": 0.3371818055362359,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0001940069226111552,
+      "loss": 0.8825,
+      "step": 606
+    },
+    {
+      "epoch": 0.33773821115593267,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00019398699892059112,
+      "loss": 0.9166,
+      "step": 607
+    },
+    {
+      "epoch": 0.33829461677562944,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00019396704319378754,
+      "loss": 0.9447,
+      "step": 608
+    },
+    {
+      "epoch": 0.3388510223953262,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0001939470554375466,
+      "loss": 0.932,
+      "step": 609
+    },
+    {
+      "epoch": 0.33940742801502294,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0001939270356586812,
+      "loss": 0.8981,
+      "step": 610
+    },
+    {
+      "epoch": 0.3399638336347197,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00019390698386401533,
+      "loss": 0.8561,
+      "step": 611
+    },
+    {
+      "epoch": 0.3405202392544165,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00019388690006038375,
+      "loss": 0.8992,
+      "step": 612
+    },
+    {
+      "epoch": 0.3410766448741132,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00019386678425463221,
+      "loss": 0.952,
+      "step": 613
+    },
+    {
+      "epoch": 0.34163305049381,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00019384663645361736,
+      "loss": 0.8244,
+      "step": 614
+    },
+    {
+      "epoch": 0.34218945611350676,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00019382645666420674,
+      "loss": 0.9347,
+      "step": 615
+    },
+    {
+      "epoch": 0.3427458617332035,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00019380624489327876,
+      "loss": 0.9098,
+      "step": 616
+    },
+    {
+      "epoch": 0.34330226735290026,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00019378600114772283,
+      "loss": 1.9243,
+      "step": 617
+    },
+    {
+      "epoch": 0.34385867297259703,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00019376572543443915,
+      "loss": 0.8078,
+      "step": 618
+    },
+    {
+      "epoch": 0.34441507859229376,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00019374541776033888,
+      "loss": 0.8494,
+      "step": 619
+    },
+    {
+      "epoch": 0.34497148421199053,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00019372507813234405,
+      "loss": 0.8841,
+      "step": 620
+    },
+    {
+      "epoch": 0.3455278898316873,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0001937047065573876,
+      "loss": 0.8424,
+      "step": 621
+    },
+    {
+      "epoch": 0.3460842954513841,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0001936843030424133,
+      "loss": 0.8028,
+      "step": 622
+    },
+    {
+      "epoch": 0.3466407010710808,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00019366386759437596,
+      "loss": 0.8883,
+      "step": 623
+    },
+    {
+      "epoch": 0.3471971066907776,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001936434002202411,
+      "loss": 0.7923,
+      "step": 624
+    },
+    {
+      "epoch": 0.34775351231047436,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00019362290092698515,
+      "loss": 0.8398,
+      "step": 625
+    },
+    {
+      "epoch": 0.3483099179301711,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001936023697215956,
+      "loss": 0.866,
+      "step": 626
+    },
+    {
+      "epoch": 0.34886632354986785,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00019358180661107051,
+      "loss": 0.8925,
+      "step": 627
+    },
+    {
+      "epoch": 0.3494227291695646,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0001935612116024191,
+      "loss": 0.8689,
+      "step": 628
+    },
+    {
+      "epoch": 0.34997913478926135,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0001935405847026613,
+      "loss": 0.8397,
+      "step": 629
+    },
+    {
+      "epoch": 0.3505355404089581,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00019351992591882798,
+      "loss": 0.8053,
+      "step": 630
+    },
+    {
+      "epoch": 0.3510919460286549,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00019349923525796084,
+      "loss": 0.8547,
+      "step": 631
+    },
+    {
+      "epoch": 0.3516483516483517,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00019347851272711247,
+      "loss": 0.944,
+      "step": 632
+    },
+    {
+      "epoch": 0.3522047572680484,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00019345775833334633,
+      "loss": 0.8962,
+      "step": 633
+    },
+    {
+      "epoch": 0.35276116288774517,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001934369720837367,
+      "loss": 0.8379,
+      "step": 634
+    },
+    {
+      "epoch": 0.35331756850744195,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00019341615398536877,
+      "loss": 0.9129,
+      "step": 635
+    },
+    {
+      "epoch": 0.35387397412713867,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00019339530404533854,
+      "loss": 0.8244,
+      "step": 636
+    },
+    {
+      "epoch": 0.35443037974683544,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00019337442227075285,
+      "loss": 0.8405,
+      "step": 637
+    },
+    {
+      "epoch": 0.3549867853665322,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0001933535086687295,
+      "loss": 0.906,
+      "step": 638
+    },
+    {
+      "epoch": 0.35554319098622894,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.000193332563246397,
+      "loss": 0.9176,
+      "step": 639
+    },
+    {
+      "epoch": 0.3560995966059257,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001933115860108948,
+      "loss": 0.8164,
+      "step": 640
+    },
+    {
+      "epoch": 0.3566560022256225,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00019329057696937317,
+      "loss": 0.8414,
+      "step": 641
+    },
+    {
+      "epoch": 0.3572124078453192,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00019326953612899318,
+      "loss": 0.9027,
+      "step": 642
+    },
+    {
+      "epoch": 0.357768813465016,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0001932484634969268,
+      "loss": 0.8986,
+      "step": 643
+    },
+    {
+      "epoch": 0.35832521908471276,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00019322735908035678,
+      "loss": 0.891,
+      "step": 644
+    },
+    {
+      "epoch": 0.35888162470440954,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0001932062228864768,
+      "loss": 0.9787,
+      "step": 645
+    },
+    {
+      "epoch": 0.35943803032410626,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0001931850549224912,
+      "loss": 0.8145,
+      "step": 646
+    },
+    {
+      "epoch": 0.35999443594380304,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00019316385519561538,
+      "loss": 0.8913,
+      "step": 647
+    },
+    {
+      "epoch": 0.3605508415634998,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00019314262371307532,
+      "loss": 0.7981,
+      "step": 648
+    },
+    {
+      "epoch": 0.36110724718319653,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00019312136048210801,
+      "loss": 0.8772,
+      "step": 649
+    },
+    {
+      "epoch": 0.3616636528028933,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001931000655099612,
+      "loss": 0.8837,
+      "step": 650
+    },
+    {
+      "epoch": 0.3622200584225901,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00019307873880389345,
+      "loss": 0.923,
+      "step": 651
+    },
+    {
+      "epoch": 0.3627764640422868,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00019305738037117412,
+      "loss": 0.8609,
+      "step": 652
+    },
+    {
+      "epoch": 0.3633328696619836,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00019303599021908342,
+      "loss": 0.8668,
+      "step": 653
+    },
+    {
+      "epoch": 0.36388927528168036,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00019301456835491237,
+      "loss": 0.9394,
+      "step": 654
+    },
+    {
+      "epoch": 0.3644456809013771,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001929931147859628,
+      "loss": 0.7789,
+      "step": 655
+    },
+    {
+      "epoch": 0.36500208652107385,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00019297162951954728,
+      "loss": 0.9009,
+      "step": 656
+    },
+    {
+      "epoch": 0.36555849214077063,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00019295011256298929,
+      "loss": 0.7917,
+      "step": 657
+    },
+    {
+      "epoch": 0.3661148977604674,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00019292856392362303,
+      "loss": 0.9266,
+      "step": 658
+    },
+    {
+      "epoch": 0.3666713033801641,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00019290698360879355,
+      "loss": 0.849,
+      "step": 659
+    },
+    {
+      "epoch": 0.3672277089998609,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001928853716258567,
+      "loss": 0.8509,
+      "step": 660
+    },
+    {
+      "epoch": 0.3677841146195577,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00019286372798217905,
+      "loss": 0.8732,
+      "step": 661
+    },
+    {
+      "epoch": 0.3683405202392544,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00019284205268513803,
+      "loss": 0.8759,
+      "step": 662
+    },
+    {
+      "epoch": 0.3688969258589512,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00019282034574212185,
+      "loss": 1.8908,
+      "step": 663
+    },
+    {
+      "epoch": 0.36945333147864795,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00019279860716052947,
+      "loss": 0.8435,
+      "step": 664
+    },
+    {
+      "epoch": 0.37000973709834467,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00019277683694777074,
+      "loss": 0.8151,
+      "step": 665
+    },
+    {
+      "epoch": 0.37056614271804145,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00019275503511126618,
+      "loss": 0.9638,
+      "step": 666
+    },
+    {
+      "epoch": 0.3711225483377382,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00019273320165844707,
+      "loss": 0.9098,
+      "step": 667
+    },
+    {
+      "epoch": 0.37167895395743494,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00019271133659675563,
+      "loss": 0.9032,
+      "step": 668
+    },
+    {
+      "epoch": 0.3722353595771317,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00019268943993364462,
+      "loss": 0.7979,
+      "step": 669
+    },
+    {
+      "epoch": 0.3727917651968285,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00019266751167657778,
+      "loss": 0.9499,
+      "step": 670
+    },
+    {
+      "epoch": 0.37334817081652527,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001926455518330295,
+      "loss": 0.8741,
+      "step": 671
+    },
+    {
+      "epoch": 0.373904576436222,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.000192623560410485,
+      "loss": 0.9162,
+      "step": 672
+    },
+    {
+      "epoch": 0.37446098205591877,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00019260153741644021,
+      "loss": 0.8135,
+      "step": 673
+    },
+    {
+      "epoch": 0.37501738767561554,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00019257948285840188,
+      "loss": 0.6664,
+      "step": 674
+    },
+    {
+      "epoch": 0.37557379329531226,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00019255739674388746,
+      "loss": 0.9409,
+      "step": 675
+    },
+    {
+      "epoch": 0.37613019891500904,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00019253527908042522,
+      "loss": 0.9325,
+      "step": 676
+    },
+    {
+      "epoch": 0.3766866045347058,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00019251312987555408,
+      "loss": 0.8681,
+      "step": 677
+    },
+    {
+      "epoch": 0.37724301015440254,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00019249094913682382,
+      "loss": 0.9045,
+      "step": 678
+    },
+    {
+      "epoch": 0.3777994157740993,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00019246873687179495,
+      "loss": 0.8807,
+      "step": 679
+    },
+    {
+      "epoch": 0.3783558213937961,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00019244649308803867,
+      "loss": 0.9102,
+      "step": 680
+    },
+    {
+      "epoch": 0.37891222701349286,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00019242421779313692,
+      "loss": 0.8124,
+      "step": 681
+    },
+    {
+      "epoch": 0.3794686326331896,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00019240191099468253,
+      "loss": 0.9337,
+      "step": 682
+    },
+    {
+      "epoch": 0.38002503825288636,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.00019237957270027884,
+      "loss": 1.8244,
+      "step": 683
+    },
+    {
+      "epoch": 0.38058144387258314,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0001923572029175401,
+      "loss": 0.8975,
+      "step": 684
+    },
+    {
+      "epoch": 0.38113784949227986,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0001923348016540912,
+      "loss": 0.8755,
+      "step": 685
+    },
+    {
+      "epoch": 0.38169425511197663,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00019231236891756787,
+      "loss": 0.8125,
+      "step": 686
+    },
+    {
+      "epoch": 0.3822506607316734,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00019228990471561636,
+      "loss": 0.8673,
+      "step": 687
+    },
+    {
+      "epoch": 0.38280706635137013,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001922674090558939,
+      "loss": 0.8462,
+      "step": 688
+    },
+    {
+      "epoch": 0.3833634719710669,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00019224488194606826,
+      "loss": 1.0229,
+      "step": 689
+    },
+    {
+      "epoch": 0.3839198775907637,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00019222232339381802,
+      "loss": 0.8654,
+      "step": 690
+    },
+    {
+      "epoch": 0.3844762832104604,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00019219973340683244,
+      "loss": 0.877,
+      "step": 691
+    },
+    {
+      "epoch": 0.3850326888301572,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00019217711199281148,
+      "loss": 0.9084,
+      "step": 692
+    },
+    {
+      "epoch": 0.38558909444985395,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00019215445915946584,
+      "loss": 0.9667,
+      "step": 693
+    },
+    {
+      "epoch": 0.38614550006955073,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00019213177491451692,
+      "loss": 0.8251,
+      "step": 694
+    },
+    {
+      "epoch": 0.38670190568924745,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00019210905926569688,
+      "loss": 0.8764,
+      "step": 695
+    },
+    {
+      "epoch": 0.3872583113089442,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0001920863122207485,
+      "loss": 0.8484,
+      "step": 696
+    },
+    {
+      "epoch": 0.387814716928641,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00019206353378742528,
+      "loss": 1.0047,
+      "step": 697
+    },
+    {
+      "epoch": 0.3883711225483377,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00019204072397349144,
+      "loss": 0.8996,
+      "step": 698
+    },
+    {
+      "epoch": 0.3889275281680345,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00019201788278672193,
+      "loss": 0.9355,
+      "step": 699
+    },
+    {
+      "epoch": 0.3894839337877313,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00019199501023490232,
+      "loss": 0.8971,
+      "step": 700
+    },
+    {
+      "epoch": 0.390040339407428,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0001919721063258289,
+      "loss": 0.8437,
+      "step": 701
+    },
+    {
+      "epoch": 0.39059674502712477,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0001919491710673087,
+      "loss": 0.9162,
+      "step": 702
+    },
+    {
+      "epoch": 0.39115315064682155,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00019192620446715933,
+      "loss": 0.9021,
+      "step": 703
+    },
+    {
+      "epoch": 0.39170955626651827,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00019190320653320918,
+      "loss": 0.8099,
+      "step": 704
+    },
+    {
+      "epoch": 0.39226596188621504,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0001918801772732973,
+      "loss": 0.9569,
+      "step": 705
+    },
+    {
+      "epoch": 0.3928223675059118,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0001918571166952734,
+      "loss": 0.9673,
+      "step": 706
+    },
+    {
+      "epoch": 0.3933787731256086,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001918340248069978,
+      "loss": 0.8443,
+      "step": 707
+    },
+    {
+      "epoch": 0.3939351787453053,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00019181090161634167,
+      "loss": 0.9339,
+      "step": 708
+    },
+    {
+      "epoch": 0.3944915843650021,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00019178774713118663,
+      "loss": 0.834,
+      "step": 709
+    },
+    {
+      "epoch": 0.39504798998469887,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00019176456135942515,
+      "loss": 0.9933,
+      "step": 710
+    },
+    {
+      "epoch": 0.3956043956043956,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00019174134430896027,
+      "loss": 0.924,
+      "step": 711
+    },
+    {
+      "epoch": 0.39616080122409236,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0001917180959877057,
+      "loss": 0.8976,
+      "step": 712
+    },
+    {
+      "epoch": 0.39671720684378914,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00019169481640358583,
+      "loss": 0.8488,
+      "step": 713
+    },
+    {
+      "epoch": 0.39727361246348586,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00019167150556453569,
+      "loss": 0.8426,
+      "step": 714
+    },
+    {
+      "epoch": 0.39783001808318263,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00019164816347850096,
+      "loss": 1.0289,
+      "step": 715
+    },
+    {
+      "epoch": 0.3983864237028794,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00019162479015343803,
+      "loss": 0.8953,
+      "step": 716
+    },
+    {
+      "epoch": 0.39894282932257613,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00019160138559731386,
+      "loss": 0.9065,
+      "step": 717
+    },
+    {
+      "epoch": 0.3994992349422729,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00019157794981810607,
+      "loss": 0.908,
+      "step": 718
+    },
+    {
+      "epoch": 0.4000556405619697,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00019155448282380294,
+      "loss": 0.796,
+      "step": 719
+    },
+    {
+      "epoch": 0.40061204618166646,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00019153098462240342,
+      "loss": 0.9603,
+      "step": 720
+    },
+    {
+      "epoch": 0.4011684518013632,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00019150745522191701,
+      "loss": 0.9096,
+      "step": 721
+    },
+    {
+      "epoch": 0.40172485742105996,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00019148389463036395,
+      "loss": 1.0184,
+      "step": 722
+    },
+    {
+      "epoch": 0.40228126304075673,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00019146030285577506,
+      "loss": 0.8274,
+      "step": 723
+    },
+    {
+      "epoch": 0.40283766866045345,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00019143667990619177,
+      "loss": 0.9454,
+      "step": 724
+    },
+    {
+      "epoch": 0.4033940742801502,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00019141302578966614,
+      "loss": 0.8516,
+      "step": 725
+    },
+    {
+      "epoch": 0.403950479899847,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00019138934051426092,
+      "loss": 0.9234,
+      "step": 726
+    },
+    {
+      "epoch": 0.4045068855195437,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00019136562408804934,
+      "loss": 0.8836,
+      "step": 727
+    },
+    {
+      "epoch": 0.4050632911392405,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00019134187651911546,
+      "loss": 0.9283,
+      "step": 728
+    },
+    {
+      "epoch": 0.4056196967589373,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00019131809781555373,
+      "loss": 0.9731,
+      "step": 729
+    },
+    {
+      "epoch": 0.40617610237863405,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00019129428798546938,
+      "loss": 0.7996,
+      "step": 730
+    },
+    {
+      "epoch": 0.4067325079983308,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00019127044703697816,
+      "loss": 0.8759,
+      "step": 731
+    },
+    {
+      "epoch": 0.40728891361802755,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00019124657497820644,
+      "loss": 0.9137,
+      "step": 732
+    },
+    {
+      "epoch": 0.4078453192377243,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00019122267181729123,
+      "loss": 0.8381,
+      "step": 733
+    },
+    {
+      "epoch": 0.40840172485742104,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0001911987375623801,
+      "loss": 0.9209,
+      "step": 734
+    },
+    {
+      "epoch": 0.4089581304771178,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00019117477222163127,
+      "loss": 0.8937,
+      "step": 735
+    },
+    {
+      "epoch": 0.4095145360968146,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0001911507758032135,
+      "loss": 0.9494,
+      "step": 736
+    },
+    {
+      "epoch": 0.4100709417165113,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00019112674831530616,
+      "loss": 0.9375,
+      "step": 737
+    },
+    {
+      "epoch": 0.4106273473362081,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00019110268976609923,
+      "loss": 0.8973,
+      "step": 738
+    },
+    {
+      "epoch": 0.41118375295590487,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00019107860016379324,
+      "loss": 0.9579,
+      "step": 739
+    },
+    {
+      "epoch": 0.4117401585756016,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00019105447951659935,
+      "loss": 0.7406,
+      "step": 740
+    },
+    {
+      "epoch": 0.41229656419529837,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0001910303278327393,
+      "loss": 0.9315,
+      "step": 741
+    },
+    {
+      "epoch": 0.41285296981499514,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00019100614512044536,
+      "loss": 0.8731,
+      "step": 742
+    },
+    {
+      "epoch": 0.4134093754346919,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0001909819313879604,
+      "loss": 0.8679,
+      "step": 743
+    },
+    {
+      "epoch": 0.41396578105438864,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00019095768664353785,
+      "loss": 0.8881,
+      "step": 744
+    },
+    {
+      "epoch": 0.4145221866740854,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001909334108954418,
+      "loss": 0.8616,
+      "step": 745
+    },
+    {
+      "epoch": 0.4150785922937822,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00019090910415194682,
+      "loss": 0.8466,
+      "step": 746
+    },
+    {
+      "epoch": 0.4156349979134789,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.000190884766421338,
+      "loss": 0.963,
+      "step": 747
+    },
+    {
+      "epoch": 0.4161914035331757,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00019086039771191112,
+      "loss": 0.8879,
+      "step": 748
+    },
+    {
+      "epoch": 0.41674780915287246,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0001908359980319724,
+      "loss": 0.9345,
+      "step": 749
+    },
+    {
+      "epoch": 0.4173042147725692,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00019081156738983874,
+      "loss": 0.9356,
+      "step": 750
+    },
+    {
+      "epoch": 0.41786062039226596,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00019078710579383746,
+      "loss": 0.8295,
+      "step": 751
+    },
+    {
+      "epoch": 0.41841702601196273,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00019076261325230655,
+      "loss": 0.9092,
+      "step": 752
+    },
+    {
+      "epoch": 0.41897343163165945,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00019073808977359446,
+      "loss": 0.83,
+      "step": 753
+    },
+    {
+      "epoch": 0.41952983725135623,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001907135353660602,
+      "loss": 0.7945,
+      "step": 754
+    },
+    {
+      "epoch": 0.420086242871053,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00019068895003807339,
+      "loss": 0.959,
+      "step": 755
+    },
+    {
+      "epoch": 0.4206426484907498,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0001906643337980141,
+      "loss": 0.8404,
+      "step": 756
+    },
+    {
+      "epoch": 0.4211990541104465,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.000190639686654273,
+      "loss": 0.811,
+      "step": 757
+    },
+    {
+      "epoch": 0.4217554597301433,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00019061500861525127,
+      "loss": 0.9065,
+      "step": 758
+    },
+    {
+      "epoch": 0.42231186534984005,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00019059029968936062,
+      "loss": 0.8662,
+      "step": 759
+    },
+    {
+      "epoch": 0.4228682709695368,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001905655598850233,
+      "loss": 0.7609,
+      "step": 760
+    },
+    {
+      "epoch": 0.42342467658923355,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00019054078921067205,
+      "loss": 0.8826,
+      "step": 761
+    },
+    {
+      "epoch": 0.4239810822089303,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00019051598767475016,
+      "loss": 0.8253,
+      "step": 762
+    },
+    {
+      "epoch": 0.42453748782862705,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00019049115528571147,
+      "loss": 0.7537,
+      "step": 763
+    },
+    {
+      "epoch": 0.4250938934483238,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00019046629205202027,
+      "loss": 0.9363,
+      "step": 764
+    },
+    {
+      "epoch": 0.4256502990680206,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00019044139798215143,
+      "loss": 0.8248,
+      "step": 765
+    },
+    {
+      "epoch": 0.4262067046877173,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00019041647308459022,
+      "loss": 0.7364,
+      "step": 766
+    },
+    {
+      "epoch": 0.4267631103074141,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00019039151736783262,
+      "loss": 0.9075,
+      "step": 767
+    },
+    {
+      "epoch": 0.42731951592711087,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00019036653084038486,
+      "loss": 0.966,
+      "step": 768
+    },
+    {
+      "epoch": 0.42787592154680765,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00019034151351076387,
+      "loss": 0.8645,
+      "step": 769
+    },
+    {
+      "epoch": 0.42843232716650437,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00019031646538749707,
+      "loss": 0.824,
+      "step": 770
+    },
+    {
+      "epoch": 0.42898873278620114,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00019029138647912218,
+      "loss": 0.8858,
+      "step": 771
+    },
+    {
+      "epoch": 0.4295451384058979,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00019026627679418767,
+      "loss": 0.8066,
+      "step": 772
+    },
+    {
+      "epoch": 0.43010154402559464,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00019024113634125234,
+      "loss": 0.8256,
+      "step": 773
+    },
+    {
+      "epoch": 0.4306579496452914,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0001902159651288855,
+      "loss": 0.8785,
+      "step": 774
+    },
+    {
+      "epoch": 0.4312143552649882,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.000190190763165667,
+      "loss": 0.8643,
+      "step": 775
+    },
+    {
+      "epoch": 0.4317707608846849,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00019016553046018712,
+      "loss": 0.9032,
+      "step": 776
+    },
+    {
+      "epoch": 0.4323271665043817,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00019014026702104662,
+      "loss": 0.8785,
+      "step": 777
+    },
+    {
+      "epoch": 0.43288357212407846,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00019011497285685678,
+      "loss": 0.8607,
+      "step": 778
+    },
+    {
+      "epoch": 0.4334399777437752,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0001900896479762393,
+      "loss": 0.8355,
+      "step": 779
+    },
+    {
+      "epoch": 0.43399638336347196,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00019006429238782637,
+      "loss": 0.9727,
+      "step": 780
+    },
+    {
+      "epoch": 0.43455278898316874,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00019003890610026067,
+      "loss": 0.8807,
+      "step": 781
+    },
+    {
+      "epoch": 0.4351091946028655,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00019001348912219533,
+      "loss": 0.9325,
+      "step": 782
+    },
+    {
+      "epoch": 0.43566560022256223,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00018998804146229387,
+      "loss": 0.8559,
+      "step": 783
+    },
+    {
+      "epoch": 0.436222005842259,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00018996256312923043,
+      "loss": 0.8304,
+      "step": 784
+    },
+    {
+      "epoch": 0.4367784114619558,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00018993705413168944,
+      "loss": 0.9767,
+      "step": 785
+    },
+    {
+      "epoch": 0.4373348170816525,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0001899115144783659,
+      "loss": 0.8711,
+      "step": 786
+    },
+    {
+      "epoch": 0.4378912227013493,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00018988594417796516,
+      "loss": 0.8525,
+      "step": 787
+    },
+    {
+      "epoch": 0.43844762832104606,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00018986034323920312,
+      "loss": 0.83,
+      "step": 788
+    },
+    {
+      "epoch": 0.4390040339407428,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.000189834711670806,
+      "loss": 0.8088,
+      "step": 789
+    },
+    {
+      "epoch": 0.43956043956043955,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00018980904948151062,
+      "loss": 0.784,
+      "step": 790
+    },
+    {
+      "epoch": 0.44011684518013633,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00018978335668006407,
+      "loss": 0.9129,
+      "step": 791
+    },
+    {
+      "epoch": 0.4406732507998331,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.000189757633275224,
+      "loss": 1.0234,
+      "step": 792
+    },
+    {
+      "epoch": 0.4412296564195298,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0001897318792757584,
+      "loss": 0.8711,
+      "step": 793
+    },
+    {
+      "epoch": 0.4417860620392266,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00018970609469044577,
+      "loss": 0.9124,
+      "step": 794
+    },
+    {
+      "epoch": 0.4423424676589234,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00018968027952807494,
+      "loss": 0.7081,
+      "step": 795
+    },
+    {
+      "epoch": 0.4428988732786201,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00018965443379744535,
+      "loss": 0.7776,
+      "step": 796
+    },
+    {
+      "epoch": 0.4434552788983169,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00018962855750736656,
+      "loss": 0.8216,
+      "step": 797
+    },
+    {
+      "epoch": 0.44401168451801365,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00018960265066665882,
+      "loss": 0.8406,
+      "step": 798
+    },
+    {
+      "epoch": 0.44456809013771037,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0001895767132841527,
+      "loss": 0.9074,
+      "step": 799
+    },
+    {
+      "epoch": 0.44512449575740715,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00018955074536868913,
+      "loss": 0.88,
+      "step": 800
+    },
+    {
+      "epoch": 0.4456809013771039,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0001895247469291195,
+      "loss": 0.9454,
+      "step": 801
+    },
+    {
+      "epoch": 0.44623730699680064,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00018949871797430557,
+      "loss": 0.7475,
+      "step": 802
+    },
+    {
+      "epoch": 0.4467937126164974,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00018947265851311954,
+      "loss": 0.9466,
+      "step": 803
+    },
+    {
+      "epoch": 0.4473501182361942,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.000189446568554444,
+      "loss": 0.791,
+      "step": 804
+    },
+    {
+      "epoch": 0.44790652385589097,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00018942044810717188,
+      "loss": 0.898,
+      "step": 805
+    },
+    {
+      "epoch": 0.4484629294755877,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00018939429718020661,
+      "loss": 0.7666,
+      "step": 806
+    },
+    {
+      "epoch": 0.44901933509528447,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0001893681157824619,
+      "loss": 0.9343,
+      "step": 807
+    },
+    {
+      "epoch": 0.44957574071498124,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00018934190392286198,
+      "loss": 0.8074,
+      "step": 808
+    },
+    {
+      "epoch": 0.45013214633467796,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00018931566161034126,
+      "loss": 0.8394,
+      "step": 809
+    },
+    {
+      "epoch": 0.45068855195437474,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00018928938885384472,
+      "loss": 0.8655,
+      "step": 810
+    },
+    {
+      "epoch": 0.4512449575740715,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00018926308566232763,
+      "loss": 0.792,
+      "step": 811
+    },
+    {
+      "epoch": 0.45180136319376824,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00018923675204475564,
+      "loss": 0.8638,
+      "step": 812
+    },
+    {
+      "epoch": 0.452357768813465,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00018921038801010477,
+      "loss": 0.8151,
+      "step": 813
+    },
+    {
+      "epoch": 0.4529141744331618,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00018918399356736146,
+      "loss": 0.8198,
+      "step": 814
+    },
+    {
+      "epoch": 0.4534705800528585,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00018915756872552242,
+      "loss": 0.6904,
+      "step": 815
+    },
+    {
+      "epoch": 0.4540269856725553,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0001891311134935948,
+      "loss": 0.8762,
+      "step": 816
+    },
+    {
+      "epoch": 0.45458339129225206,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001891046278805961,
+      "loss": 0.7527,
+      "step": 817
+    },
+    {
+      "epoch": 0.45513979691194884,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00018907811189555412,
+      "loss": 0.7777,
+      "step": 818
+    },
+    {
+      "epoch": 0.45569620253164556,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00018905156554750708,
+      "loss": 0.8556,
+      "step": 819
+    },
+    {
+      "epoch": 0.45625260815134233,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001890249888455035,
+      "loss": 0.8273,
+      "step": 820
+    },
+    {
+      "epoch": 0.4568090137710391,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001889983817986023,
+      "loss": 0.8405,
+      "step": 821
+    },
+    {
+      "epoch": 0.45736541939073583,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00018897174441587266,
+      "loss": 0.9006,
+      "step": 822
+    },
+    {
+      "epoch": 0.4579218250104326,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00018894507670639418,
+      "loss": 0.9191,
+      "step": 823
+    },
+    {
+      "epoch": 0.4584782306301294,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00018891837867925678,
+      "loss": 0.9777,
+      "step": 824
+    },
+    {
+      "epoch": 0.4590346362498261,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00018889165034356072,
+      "loss": 0.892,
+      "step": 825
+    },
+    {
+      "epoch": 0.4595910418695229,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00018886489170841648,
+      "loss": 0.8615,
+      "step": 826
+    },
+    {
+      "epoch": 0.46014744748921965,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00018883810278294503,
+      "loss": 0.8765,
+      "step": 827
+    },
+    {
+      "epoch": 0.4607038531089164,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00018881128357627763,
+      "loss": 0.8803,
+      "step": 828
+    },
+    {
+      "epoch": 0.46126025872861315,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00018878443409755575,
+      "loss": 0.9017,
+      "step": 829
+    },
+    {
+      "epoch": 0.4618166643483099,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00018875755435593129,
+      "loss": 0.8139,
+      "step": 830
+    },
+    {
+      "epoch": 0.4623730699680067,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0001887306443605664,
+      "loss": 0.9611,
+      "step": 831
+    },
+    {
+      "epoch": 0.4629294755877034,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00018870370412063365,
+      "loss": 0.9311,
+      "step": 832
+    },
+    {
+      "epoch": 0.4634858812074002,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00018867673364531578,
+      "loss": 0.9108,
+      "step": 833
+    },
+    {
+      "epoch": 0.464042286827097,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00018864973294380587,
+      "loss": 1.0142,
+      "step": 834
+    },
+    {
+      "epoch": 0.4645986924467937,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00018862270202530738,
+      "loss": 0.744,
+      "step": 835
+    },
+    {
+      "epoch": 0.46515509806649047,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00018859564089903404,
+      "loss": 0.8419,
+      "step": 836
+    },
+    {
+      "epoch": 0.46571150368618724,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00018856854957420983,
+      "loss": 0.7806,
+      "step": 837
+    },
+    {
+      "epoch": 0.46626790930588397,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00018854142806006902,
+      "loss": 0.9135,
+      "step": 838
+    },
+    {
+      "epoch": 0.46682431492558074,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00018851427636585623,
+      "loss": 0.836,
+      "step": 839
+    },
+    {
+      "epoch": 0.4673807205452775,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00018848709450082637,
+      "loss": 0.8926,
+      "step": 840
+    },
+    {
+      "epoch": 0.4679371261649743,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0001884598824742445,
+      "loss": 0.8962,
+      "step": 841
+    },
+    {
+      "epoch": 0.468493531784671,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001884326402953862,
+      "loss": 0.8185,
+      "step": 842
+    },
+    {
+      "epoch": 0.4690499374043678,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0001884053679735371,
+      "loss": 0.948,
+      "step": 843
+    },
+    {
+      "epoch": 0.46960634302406457,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0001883780655179932,
+      "loss": 1.0652,
+      "step": 844
+    },
+    {
+      "epoch": 0.4701627486437613,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00018835073293806075,
+      "loss": 0.9291,
+      "step": 845
+    },
+    {
+      "epoch": 0.47071915426345806,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00018832337024305638,
+      "loss": 0.9503,
+      "step": 846
+    },
+    {
+      "epoch": 0.47127555988315484,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00018829597744230678,
+      "loss": 0.8685,
+      "step": 847
+    },
+    {
+      "epoch": 0.47183196550285156,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00018826855454514907,
+      "loss": 0.8796,
+      "step": 848
+    },
+    {
+      "epoch": 0.47238837112254833,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00018824110156093053,
+      "loss": 0.7906,
+      "step": 849
+    },
+    {
+      "epoch": 0.4729447767422451,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00018821361849900877,
+      "loss": 0.9085,
+      "step": 850
+    },
+    {
+      "epoch": 0.47350118236194183,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00018818610536875164,
+      "loss": 0.9018,
+      "step": 851
+    },
+    {
+      "epoch": 0.4740575879816386,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00018815856217953714,
+      "loss": 0.8632,
+      "step": 852
+    },
+    {
+      "epoch": 0.4746139936013354,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00018813098894075365,
+      "loss": 0.9047,
+      "step": 853
+    },
+    {
+      "epoch": 0.47517039922103216,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00018810338566179973,
+      "loss": 0.9046,
+      "step": 854
+    },
+    {
+      "epoch": 0.4757268048407289,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00018807575235208415,
+      "loss": 0.8683,
+      "step": 855
+    },
+    {
+      "epoch": 0.47628321046042565,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00018804808902102597,
+      "loss": 0.8541,
+      "step": 856
+    },
+    {
+      "epoch": 0.47683961608012243,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00018802039567805448,
+      "loss": 0.8484,
+      "step": 857
+    },
+    {
+      "epoch": 0.47739602169981915,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00018799267233260918,
+      "loss": 0.8559,
+      "step": 858
+    },
+    {
+      "epoch": 0.4779524273195159,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00018796491899413976,
+      "loss": 0.8646,
+      "step": 859
+    },
+    {
+      "epoch": 0.4785088329392127,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0001879371356721062,
+      "loss": 0.9625,
+      "step": 860
+    },
+    {
+      "epoch": 0.4790652385589094,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001879093223759787,
+      "loss": 0.7927,
+      "step": 861
+    },
+    {
+      "epoch": 0.4796216441786062,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00018788147911523762,
+      "loss": 0.8716,
+      "step": 862
+    },
+    {
+      "epoch": 0.480178049798303,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00018785360589937354,
+      "loss": 0.7913,
+      "step": 863
+    },
+    {
+      "epoch": 0.4807344554179997,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00018782570273788736,
+      "loss": 0.8636,
+      "step": 864
+    },
+    {
+      "epoch": 0.48129086103769647,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00018779776964029,
+      "loss": 0.8082,
+      "step": 865
+    },
+    {
+      "epoch": 0.48184726665739325,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00018776980661610273,
+      "loss": 0.8549,
+      "step": 866
+    },
+    {
+      "epoch": 0.48240367227709,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.000187741813674857,
+      "loss": 0.8543,
+      "step": 867
+    },
+    {
+      "epoch": 0.48296007789678674,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00018771379082609436,
+      "loss": 0.8735,
+      "step": 868
+    },
+    {
+      "epoch": 0.4835164835164835,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0001876857380793667,
+      "loss": 0.8338,
+      "step": 869
+    },
+    {
+      "epoch": 0.4840728891361803,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00018765765544423606,
+      "loss": 0.8904,
+      "step": 870
+    },
+    {
+      "epoch": 0.484629294755877,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00018762954293027453,
+      "loss": 0.7981,
+      "step": 871
+    },
+    {
+      "epoch": 0.4851857003755738,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00018760140054706453,
+      "loss": 0.9093,
+      "step": 872
+    },
+    {
+      "epoch": 0.48574210599527057,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00018757322830419867,
+      "loss": 0.9216,
+      "step": 873
+    },
+    {
+      "epoch": 0.4862985116149673,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00018754502621127966,
+      "loss": 0.7949,
+      "step": 874
+    },
+    {
+      "epoch": 0.48685491723466406,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00018751679427792038,
+      "loss": 0.9682,
+      "step": 875
+    },
+    {
+      "epoch": 0.48741132285436084,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00018748853251374396,
+      "loss": 0.792,
+      "step": 876
+    },
+    {
+      "epoch": 0.48796772847405756,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00018746024092838366,
+      "loss": 0.8901,
+      "step": 877
+    },
+    {
+      "epoch": 0.48852413409375434,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00018743191953148286,
+      "loss": 0.9955,
+      "step": 878
+    },
+    {
+      "epoch": 0.4890805397134511,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00018740356833269516,
+      "loss": 0.9246,
+      "step": 879
+    },
+    {
+      "epoch": 0.4896369453331479,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00018737518734168428,
+      "loss": 0.9122,
+      "step": 880
+    },
+    {
+      "epoch": 0.4901933509528446,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001873467765681242,
+      "loss": 0.9119,
+      "step": 881
+    },
+    {
+      "epoch": 0.4907497565725414,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00018731833602169884,
+      "loss": 0.7738,
+      "step": 882
+    },
+    {
+      "epoch": 0.49130616219223816,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00018728986571210244,
+      "loss": 0.8866,
+      "step": 883
+    },
+    {
+      "epoch": 0.4918625678119349,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00018726136564903938,
+      "loss": 0.9322,
+      "step": 884
+    },
+    {
+      "epoch": 0.49241897343163166,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001872328358422241,
+      "loss": 0.7841,
+      "step": 885
+    },
+    {
+      "epoch": 0.49297537905132843,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00018720427630138122,
+      "loss": 0.9459,
+      "step": 886
+    },
+    {
+      "epoch": 0.49353178467102515,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001871756870362455,
+      "loss": 0.8372,
+      "step": 887
+    },
+    {
+      "epoch": 0.49408819029072193,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00018714706805656185,
+      "loss": 0.8801,
+      "step": 888
+    },
+    {
+      "epoch": 0.4946445959104187,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00018711841937208527,
+      "loss": 0.7583,
+      "step": 889
+    },
+    {
+      "epoch": 0.4952010015301155,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00018708974099258086,
+      "loss": 0.8007,
+      "step": 890
+    },
+    {
+      "epoch": 0.4957574071498122,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0001870610329278239,
+      "loss": 0.9087,
+      "step": 891
+    },
+    {
+      "epoch": 0.496313812769509,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00018703229518759985,
+      "loss": 0.8809,
+      "step": 892
+    },
+    {
+      "epoch": 0.49687021838920575,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00018700352778170407,
+      "loss": 0.8517,
+      "step": 893
+    },
+    {
+      "epoch": 0.4974266240089025,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00018697473071994225,
+      "loss": 0.9004,
+      "step": 894
+    },
+    {
+      "epoch": 0.49798302962859925,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001869459040121301,
+      "loss": 0.8786,
+      "step": 895
+    },
+    {
+      "epoch": 0.498539435248296,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0001869170476680934,
+      "loss": 0.911,
+      "step": 896
+    },
+    {
+      "epoch": 0.49909584086799275,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0001868881616976681,
+      "loss": 0.8994,
+      "step": 897
+    },
+    {
+      "epoch": 0.4996522464876895,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00018685924611070025,
+      "loss": 0.8162,
+      "step": 898
+    },
+    {
+      "epoch": 0.5002086521073863,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00018683030091704592,
+      "loss": 0.891,
+      "step": 899
+    },
+    {
+      "epoch": 0.5002086521073863,
+      "eval_loss": 0.8054929971694946,
+      "eval_runtime": 1766.0961,
+      "eval_samples_per_second": 2.656,
+      "eval_steps_per_second": 1.328,
+      "step": 899
+    },
+    {
+      "epoch": 0.500765057727083,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001868013261265713,
+      "loss": 0.8493,
+      "step": 900
+    },
+    {
+      "epoch": 0.5013214633467798,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00018677232174915275,
+      "loss": 0.8253,
+      "step": 901
+    },
+    {
+      "epoch": 0.5018778689664766,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0001867432877946766,
+      "loss": 0.8701,
+      "step": 902
+    },
+    {
+      "epoch": 0.5024342745861733,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0001867142242730393,
+      "loss": 0.9443,
+      "step": 903
+    },
+    {
+      "epoch": 0.5029906802058701,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00018668513119414744,
+      "loss": 0.9508,
+      "step": 904
+    },
+    {
+      "epoch": 0.5035470858255668,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00018665600856791764,
+      "loss": 0.7881,
+      "step": 905
+    },
+    {
+      "epoch": 0.5041034914452636,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00018662685640427653,
+      "loss": 0.8958,
+      "step": 906
+    },
+    {
+      "epoch": 0.5046598970649604,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00018659767471316087,
+      "loss": 0.8597,
+      "step": 907
+    },
+    {
+      "epoch": 0.5052163026846571,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0001865684635045175,
+      "loss": 0.8495,
+      "step": 908
+    },
+    {
+      "epoch": 0.5057727083043538,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00018653922278830335,
+      "loss": 0.8809,
+      "step": 909
+    },
+    {
+      "epoch": 0.5063291139240507,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00018650995257448526,
+      "loss": 0.8809,
+      "step": 910
+    },
+    {
+      "epoch": 0.5068855195437474,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001864806528730403,
+      "loss": 0.894,
+      "step": 911
+    },
+    {
+      "epoch": 0.5074419251634441,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00018645132369395545,
+      "loss": 0.8067,
+      "step": 912
+    },
+    {
+      "epoch": 0.5079983307831409,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00018642196504722785,
+      "loss": 0.8951,
+      "step": 913
+    },
+    {
+      "epoch": 0.5085547364028377,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00018639257694286462,
+      "loss": 0.846,
+      "step": 914
+    },
+    {
+      "epoch": 0.5091111420225344,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00018636315939088292,
+      "loss": 0.8076,
+      "step": 915
+    },
+    {
+      "epoch": 0.5096675476422312,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00018633371240131,
+      "loss": 0.7998,
+      "step": 916
+    },
+    {
+      "epoch": 0.5102239532619279,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00018630423598418308,
+      "loss": 0.7803,
+      "step": 917
+    },
+    {
+      "epoch": 0.5107803588816247,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00018627473014954946,
+      "loss": 0.9279,
+      "step": 918
+    },
+    {
+      "epoch": 0.5113367645013215,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0001862451949074664,
+      "loss": 0.9039,
+      "step": 919
+    },
+    {
+      "epoch": 0.5118931701210182,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00018621563026800127,
+      "loss": 0.7871,
+      "step": 920
+    },
+    {
+      "epoch": 0.512449575740715,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00018618603624123145,
+      "loss": 0.8431,
+      "step": 921
+    },
+    {
+      "epoch": 0.5130059813604118,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00018615641283724425,
+      "loss": 0.8864,
+      "step": 922
+    },
+    {
+      "epoch": 0.5135623869801085,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00018612676006613706,
+      "loss": 0.8569,
+      "step": 923
+    },
+    {
+      "epoch": 0.5141187925998053,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00018609707793801726,
+      "loss": 0.9234,
+      "step": 924
+    },
+    {
+      "epoch": 0.514675198219502,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0001860673664630023,
+      "loss": 0.8782,
+      "step": 925
+    },
+    {
+      "epoch": 0.5152316038391987,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00018603762565121953,
+      "loss": 0.8889,
+      "step": 926
+    },
+    {
+      "epoch": 0.5157880094588956,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00018600785551280637,
+      "loss": 0.9133,
+      "step": 927
+    },
+    {
+      "epoch": 0.5163444150785923,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001859780560579102,
+      "loss": 0.8214,
+      "step": 928
+    },
+    {
+      "epoch": 0.516900820698289,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00018594822729668843,
+      "loss": 0.7542,
+      "step": 929
+    },
+    {
+      "epoch": 0.5174572263179859,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00018591836923930843,
+      "loss": 0.8442,
+      "step": 930
+    },
+    {
+      "epoch": 0.5180136319376826,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00018588848189594756,
+      "loss": 0.9669,
+      "step": 931
+    },
+    {
+      "epoch": 0.5185700375573793,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00018585856527679316,
+      "loss": 0.8215,
+      "step": 932
+    },
+    {
+      "epoch": 0.5191264431770761,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00018582861939204256,
+      "loss": 0.8499,
+      "step": 933
+    },
+    {
+      "epoch": 0.5196828487967728,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00018579864425190309,
+      "loss": 0.8397,
+      "step": 934
+    },
+    {
+      "epoch": 0.5202392544164696,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00018576863986659198,
+      "loss": 0.8204,
+      "step": 935
+    },
+    {
+      "epoch": 0.5207956600361664,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001857386062463365,
+      "loss": 0.765,
+      "step": 936
+    },
+    {
+      "epoch": 0.5213520656558631,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0001857085434013738,
+      "loss": 0.8809,
+      "step": 937
+    },
+    {
+      "epoch": 0.5219084712755598,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001856784513419511,
+      "loss": 0.8837,
+      "step": 938
+    },
+    {
+      "epoch": 0.5224648768952567,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00018564833007832555,
+      "loss": 0.8835,
+      "step": 939
+    },
+    {
+      "epoch": 0.5230212825149534,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00018561817962076414,
+      "loss": 0.8856,
+      "step": 940
+    },
+    {
+      "epoch": 0.5235776881346501,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00018558799997954402,
+      "loss": 0.8138,
+      "step": 941
+    },
+    {
+      "epoch": 0.5241340937543469,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00018555779116495206,
+      "loss": 0.8903,
+      "step": 942
+    },
+    {
+      "epoch": 0.5246904993740437,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00018552755318728523,
+      "loss": 0.7882,
+      "step": 943
+    },
+    {
+      "epoch": 0.5252469049937404,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00018549728605685042,
+      "loss": 0.8532,
+      "step": 944
+    },
+    {
+      "epoch": 0.5258033106134372,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00018546698978396436,
+      "loss": 0.7801,
+      "step": 945
+    },
+    {
+      "epoch": 0.5263597162331339,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00018543666437895385,
+      "loss": 0.8937,
+      "step": 946
+    },
+    {
+      "epoch": 0.5269161218528308,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00018540630985215548,
+      "loss": 0.9237,
+      "step": 947
+    },
+    {
+      "epoch": 0.5274725274725275,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00018537592621391591,
+      "loss": 0.8728,
+      "step": 948
+    },
+    {
+      "epoch": 0.5280289330922242,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00018534551347459163,
+      "loss": 0.8142,
+      "step": 949
+    },
+    {
+      "epoch": 0.528585338711921,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001853150716445491,
+      "loss": 0.9059,
+      "step": 950
+    },
+    {
+      "epoch": 0.5291417443316178,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00018528460073416456,
+      "loss": 0.8826,
+      "step": 951
+    },
+    {
+      "epoch": 0.5296981499513145,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00018525410075382438,
+      "loss": 0.9263,
+      "step": 952
+    },
+    {
+      "epoch": 0.5302545555710113,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0001852235717139247,
+      "loss": 0.8751,
+      "step": 953
+    },
+    {
+      "epoch": 0.530810961190708,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00018519301362487158,
+      "loss": 0.833,
+      "step": 954
+    },
+    {
+      "epoch": 0.5313673668104048,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00018516242649708103,
+      "loss": 0.9195,
+      "step": 955
+    },
+    {
+      "epoch": 0.5319237724301016,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00018513181034097886,
+      "loss": 0.961,
+      "step": 956
+    },
+    {
+      "epoch": 0.5324801780497983,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001851011651670009,
+      "loss": 0.8375,
+      "step": 957
+    },
+    {
+      "epoch": 0.533036583669495,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00018507049098559275,
+      "loss": 0.8556,
+      "step": 958
+    },
+    {
+      "epoch": 0.5335929892891919,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00018503978780721,
+      "loss": 0.9757,
+      "step": 959
+    },
+    {
+      "epoch": 0.5341493949088886,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00018500905564231807,
+      "loss": 0.9229,
+      "step": 960
+    },
+    {
+      "epoch": 0.5347058005285853,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001849782945013923,
+      "loss": 0.8092,
+      "step": 961
+    },
+    {
+      "epoch": 0.5352622061482821,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00018494750439491782,
+      "loss": 0.8918,
+      "step": 962
+    },
+    {
+      "epoch": 0.5358186117679788,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00018491668533338973,
+      "loss": 0.8153,
+      "step": 963
+    },
+    {
+      "epoch": 0.5363750173876756,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00018488583732731293,
+      "loss": 0.8872,
+      "step": 964
+    },
+    {
+      "epoch": 0.5369314230073724,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00018485496038720222,
+      "loss": 0.9895,
+      "step": 965
+    },
+    {
+      "epoch": 0.5374878286270691,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00018482405452358228,
+      "loss": 0.8771,
+      "step": 966
+    },
+    {
+      "epoch": 0.5380442342467658,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00018479311974698765,
+      "loss": 0.9115,
+      "step": 967
+    },
+    {
+      "epoch": 0.5386006398664627,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00018476215606796263,
+      "loss": 0.8777,
+      "step": 968
+    },
+    {
+      "epoch": 0.5391570454861594,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00018473116349706153,
+      "loss": 0.9589,
+      "step": 969
+    },
+    {
+      "epoch": 0.5397134511058562,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00018470014204484834,
+      "loss": 0.7294,
+      "step": 970
+    },
+    {
+      "epoch": 0.5402698567255529,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.000184669091721897,
+      "loss": 0.8296,
+      "step": 971
+    },
+    {
+      "epoch": 0.5408262623452497,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00018463801253879128,
+      "loss": 0.9549,
+      "step": 972
+    },
+    {
+      "epoch": 0.5413826679649465,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001846069045061248,
+      "loss": 0.8513,
+      "step": 973
+    },
+    {
+      "epoch": 0.5419390735846432,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00018457576763450097,
+      "loss": 0.879,
+      "step": 974
+    },
+    {
+      "epoch": 0.5424954792043399,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00018454460193453303,
+      "loss": 0.8198,
+      "step": 975
+    },
+    {
+      "epoch": 0.5430518848240368,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00018451340741684408,
+      "loss": 0.8208,
+      "step": 976
+    },
+    {
+      "epoch": 0.5436082904437335,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.000184482184092067,
+      "loss": 0.9505,
+      "step": 977
+    },
+    {
+      "epoch": 0.5441646960634302,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00018445093197084456,
+      "loss": 0.8958,
+      "step": 978
+    },
+    {
+      "epoch": 0.544721101683127,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00018441965106382927,
+      "loss": 0.8845,
+      "step": 979
+    },
+    {
+      "epoch": 0.5452775073028238,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001843883413816835,
+      "loss": 0.7969,
+      "step": 980
+    },
+    {
+      "epoch": 0.5458339129225205,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0001843570029350794,
+      "loss": 0.8301,
+      "step": 981
+    },
+    {
+      "epoch": 0.5463903185422173,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00018432563573469895,
+      "loss": 0.7873,
+      "step": 982
+    },
+    {
+      "epoch": 0.546946724161914,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0001842942397912339,
+      "loss": 0.8894,
+      "step": 983
+    },
+    {
+      "epoch": 0.5475031297816108,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00018426281511538584,
+      "loss": 0.91,
+      "step": 984
+    },
+    {
+      "epoch": 0.5480595354013076,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00018423136171786611,
+      "loss": 0.823,
+      "step": 985
+    },
+    {
+      "epoch": 0.5486159410210043,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00018419987960939586,
+      "loss": 0.8033,
+      "step": 986
+    },
+    {
+      "epoch": 0.549172346640701,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00018416836880070603,
+      "loss": 0.8735,
+      "step": 987
+    },
+    {
+      "epoch": 0.5497287522603979,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00018413682930253734,
+      "loss": 0.8009,
+      "step": 988
+    },
+    {
+      "epoch": 0.5502851578800946,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00018410526112564026,
+      "loss": 0.9164,
+      "step": 989
+    },
+    {
+      "epoch": 0.5508415634997913,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001840736642807751,
+      "loss": 0.869,
+      "step": 990
+    },
+    {
+      "epoch": 0.5513979691194881,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00018404203877871186,
+      "loss": 0.7933,
+      "step": 991
+    },
+    {
+      "epoch": 0.5519543747391848,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001840103846302304,
+      "loss": 0.8246,
+      "step": 992
+    },
+    {
+      "epoch": 0.5525107803588816,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001839787018461203,
+      "loss": 0.8753,
+      "step": 993
+    },
+    {
+      "epoch": 0.5530671859785784,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00018394699043718085,
+      "loss": 0.8591,
+      "step": 994
+    },
+    {
+      "epoch": 0.5536235915982751,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00018391525041422114,
+      "loss": 0.8757,
+      "step": 995
+    },
+    {
+      "epoch": 0.554179997217972,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00018388348178806008,
+      "loss": 0.8981,
+      "step": 996
+    },
+    {
+      "epoch": 0.5547364028376687,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00018385168456952621,
+      "loss": 0.7889,
+      "step": 997
+    },
+    {
+      "epoch": 0.5552928084573654,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0001838198587694579,
+      "loss": 0.8176,
+      "step": 998
+    },
+    {
+      "epoch": 0.5558492140770622,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0001837880043987032,
+      "loss": 0.9225,
+      "step": 999
+    },
+    {
+      "epoch": 0.556405619696759,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00018375612146812,
+      "loss": 0.8243,
+      "step": 1000
+    },
+    {
+      "epoch": 0.5569620253164557,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00018372420998857584,
+      "loss": 0.7683,
+      "step": 1001
+    },
+    {
+      "epoch": 0.5575184309361525,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00018369226997094792,
+      "loss": 0.906,
+      "step": 1002
+    },
+    {
+      "epoch": 0.5580748365558492,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0001836603014261234,
+      "loss": 0.8942,
+      "step": 1003
+    },
+    {
+      "epoch": 0.5586312421755459,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00018362830436499886,
+      "loss": 0.8148,
+      "step": 1004
+    },
+    {
+      "epoch": 0.5591876477952428,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0001835962787984809,
+      "loss": 0.9291,
+      "step": 1005
+    },
+    {
+      "epoch": 0.5597440534149395,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00018356422473748564,
+      "loss": 0.7764,
+      "step": 1006
+    },
+    {
+      "epoch": 0.5603004590346362,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00018353214219293896,
+      "loss": 0.8792,
+      "step": 1007
+    },
+    {
+      "epoch": 0.560856864654333,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0001835000311757765,
+      "loss": 0.8608,
+      "step": 1008
+    },
+    {
+      "epoch": 0.5614132702740298,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0001834678916969435,
+      "loss": 0.927,
+      "step": 1009
+    },
+    {
+      "epoch": 0.5619696758937265,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00018343572376739505,
+      "loss": 0.7639,
+      "step": 1010
+    },
+    {
+      "epoch": 0.5625260815134233,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00018340352739809577,
+      "loss": 0.8458,
+      "step": 1011
+    },
+    {
+      "epoch": 0.56308248713312,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0001833713026000201,
+      "loss": 0.9023,
+      "step": 1012
+    },
+    {
+      "epoch": 0.5636388927528168,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001833390493841521,
+      "loss": 0.8104,
+      "step": 1013
+    },
+    {
+      "epoch": 0.5641952983725136,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00018330676776148557,
+      "loss": 0.8363,
+      "step": 1014
+    },
+    {
+      "epoch": 0.5647517039922103,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00018327445774302397,
+      "loss": 0.8123,
+      "step": 1015
+    },
+    {
+      "epoch": 0.565308109611907,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00018324211933978043,
+      "loss": 0.8886,
+      "step": 1016
+    },
+    {
+      "epoch": 0.5658645152316039,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001832097525627777,
+      "loss": 0.7553,
+      "step": 1017
+    },
+    {
+      "epoch": 0.5664209208513006,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00018317735742304832,
+      "loss": 0.8144,
+      "step": 1018
+    },
+    {
+      "epoch": 0.5669773264709974,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00018314493393163444,
+      "loss": 0.8803,
+      "step": 1019
+    },
+    {
+      "epoch": 0.5675337320906941,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001831124820995879,
+      "loss": 0.8016,
+      "step": 1020
+    },
+    {
+      "epoch": 0.5680901377103909,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00018308000193797012,
+      "loss": 0.7583,
+      "step": 1021
+    },
+    {
+      "epoch": 0.5686465433300877,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00018304749345785223,
+      "loss": 0.8315,
+      "step": 1022
+    },
+    {
+      "epoch": 0.5692029489497844,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00018301495667031505,
+      "loss": 0.7795,
+      "step": 1023
+    },
+    {
+      "epoch": 0.5697593545694811,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.000182982391586449,
+      "loss": 0.7642,
+      "step": 1024
+    },
+    {
+      "epoch": 0.570315760189178,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00018294979821735413,
+      "loss": 0.8609,
+      "step": 1025
+    },
+    {
+      "epoch": 0.5708721658088747,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00018291717657414013,
+      "loss": 0.9518,
+      "step": 1026
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00018288452666792644,
+      "loss": 0.8299,
+      "step": 1027
+    },
+    {
+      "epoch": 0.5719849770482682,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.000182851848509842,
+      "loss": 0.9225,
+      "step": 1028
+    },
+    {
+      "epoch": 0.572541382667965,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00018281914211102542,
+      "loss": 0.7879,
+      "step": 1029
+    },
+    {
+      "epoch": 0.5730977882876617,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00018278640748262498,
+      "loss": 0.6934,
+      "step": 1030
+    },
+    {
+      "epoch": 0.5736541939073585,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00018275364463579853,
+      "loss": 0.9117,
+      "step": 1031
+    },
+    {
+      "epoch": 0.5742105995270552,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00018272085358171354,
+      "loss": 0.8398,
+      "step": 1032
+    },
+    {
+      "epoch": 0.5747670051467519,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00018268803433154712,
+      "loss": 0.7672,
+      "step": 1033
+    },
+    {
+      "epoch": 0.5753234107664488,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00018265518689648596,
+      "loss": 0.7822,
+      "step": 1034
+    },
+    {
+      "epoch": 0.5758798163861455,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00018262231128772643,
+      "loss": 0.7824,
+      "step": 1035
+    },
+    {
+      "epoch": 0.5764362220058422,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00018258940751647443,
+      "loss": 0.9375,
+      "step": 1036
+    },
+    {
+      "epoch": 0.576992627625539,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00018255647559394542,
+      "loss": 0.8067,
+      "step": 1037
+    },
+    {
+      "epoch": 0.5775490332452358,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0001825235155313646,
+      "loss": 0.912,
+      "step": 1038
+    },
+    {
+      "epoch": 0.5781054388649325,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00018249052733996664,
+      "loss": 0.9291,
+      "step": 1039
+    },
+    {
+      "epoch": 0.5786618444846293,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0001824575110309958,
+      "loss": 0.8022,
+      "step": 1040
+    },
+    {
+      "epoch": 0.579218250104326,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00018242446661570605,
+      "loss": 0.8948,
+      "step": 1041
+    },
+    {
+      "epoch": 0.5797746557240228,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00018239139410536077,
+      "loss": 0.9978,
+      "step": 1042
+    },
+    {
+      "epoch": 0.5803310613437196,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00018235829351123304,
+      "loss": 0.8297,
+      "step": 1043
+    },
+    {
+      "epoch": 0.5808874669634163,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00018232516484460541,
+      "loss": 0.8944,
+      "step": 1044
+    },
+    {
+      "epoch": 0.5814438725831131,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00018229200811677016,
+      "loss": 0.9035,
+      "step": 1045
+    },
+    {
+      "epoch": 0.5820002782028099,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00018225882333902896,
+      "loss": 0.7308,
+      "step": 1046
+    },
+    {
+      "epoch": 0.5825566838225066,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0001822256105226931,
+      "loss": 0.8482,
+      "step": 1047
+    },
+    {
+      "epoch": 0.5831130894422034,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00018219236967908348,
+      "loss": 0.7428,
+      "step": 1048
+    },
+    {
+      "epoch": 0.5836694950619001,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00018215910081953053,
+      "loss": 0.8924,
+      "step": 1049
+    },
+    {
+      "epoch": 0.5842259006815969,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00018212580395537417,
+      "loss": 0.7806,
+      "step": 1050
+    },
+    {
+      "epoch": 0.5847823063012937,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0001820924790979639,
+      "loss": 0.8435,
+      "step": 1051
+    },
+    {
+      "epoch": 0.5853387119209904,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00018205912625865881,
+      "loss": 0.9288,
+      "step": 1052
+    },
+    {
+      "epoch": 0.5858951175406871,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001820257454488275,
+      "loss": 0.7463,
+      "step": 1053
+    },
+    {
+      "epoch": 0.586451523160384,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00018199233667984802,
+      "loss": 0.7873,
+      "step": 1054
+    },
+    {
+      "epoch": 0.5870079287800807,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00018195889996310813,
+      "loss": 0.8771,
+      "step": 1055
+    },
+    {
+      "epoch": 0.5875643343997774,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001819254353100049,
+      "loss": 0.8398,
+      "step": 1056
+    },
+    {
+      "epoch": 0.5881207400194742,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00018189194273194513,
+      "loss": 0.9242,
+      "step": 1057
+    },
+    {
+      "epoch": 0.588677145639171,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00018185842224034495,
+      "loss": 0.889,
+      "step": 1058
+    },
+    {
+      "epoch": 0.5892335512588677,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00018182487384663017,
+      "loss": 0.9042,
+      "step": 1059
+    },
+    {
+      "epoch": 0.5897899568785645,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.000181791297562236,
+      "loss": 0.8742,
+      "step": 1060
+    },
+    {
+      "epoch": 0.5903463624982612,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00018175769339860714,
+      "loss": 0.8162,
+      "step": 1061
+    },
+    {
+      "epoch": 0.5909027681179579,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00018172406136719794,
+      "loss": 0.9006,
+      "step": 1062
+    },
+    {
+      "epoch": 0.5914591737376548,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00018169040147947212,
+      "loss": 0.7795,
+      "step": 1063
+    },
+    {
+      "epoch": 0.5920155793573515,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00018165671374690293,
+      "loss": 0.8475,
+      "step": 1064
+    },
+    {
+      "epoch": 0.5925719849770482,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00018162299818097303,
+      "loss": 0.803,
+      "step": 1065
+    },
+    {
+      "epoch": 0.593128390596745,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001815892547931748,
+      "loss": 0.8489,
+      "step": 1066
+    },
+    {
+      "epoch": 0.5936847962164418,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00018155548359500981,
+      "loss": 0.8463,
+      "step": 1067
+    },
+    {
+      "epoch": 0.5942412018361386,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00018152168459798935,
+      "loss": 0.8165,
+      "step": 1068
+    },
+    {
+      "epoch": 0.5947976074558353,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00018148785781363401,
+      "loss": 0.8304,
+      "step": 1069
+    },
+    {
+      "epoch": 0.595354013075532,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.000181454003253474,
+      "loss": 0.7645,
+      "step": 1070
+    },
+    {
+      "epoch": 0.5959104186952289,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00018142012092904885,
+      "loss": 0.8276,
+      "step": 1071
+    },
+    {
+      "epoch": 0.5964668243149256,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00018138621085190767,
+      "loss": 0.889,
+      "step": 1072
+    },
+    {
+      "epoch": 0.5970232299346223,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.000181352273033609,
+      "loss": 0.83,
+      "step": 1073
+    },
+    {
+      "epoch": 0.5975796355543191,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00018131830748572078,
+      "loss": 0.9765,
+      "step": 1074
+    },
+    {
+      "epoch": 0.5981360411740159,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00018128431421982047,
+      "loss": 0.8316,
+      "step": 1075
+    },
+    {
+      "epoch": 0.5986924467937126,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00018125029324749499,
+      "loss": 0.867,
+      "step": 1076
+    },
+    {
+      "epoch": 0.5992488524134094,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00018121624458034062,
+      "loss": 0.7405,
+      "step": 1077
+    },
+    {
+      "epoch": 0.5998052580331061,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0001811821682299631,
+      "loss": 0.9073,
+      "step": 1078
+    },
+    {
+      "epoch": 0.6003616636528029,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0001811480642079777,
+      "loss": 0.8368,
+      "step": 1079
+    },
+    {
+      "epoch": 0.6009180692724997,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00018111393252600902,
+      "loss": 0.7902,
+      "step": 1080
+    },
+    {
+      "epoch": 0.6014744748921964,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00018107977319569115,
+      "loss": 0.8296,
+      "step": 1081
+    },
+    {
+      "epoch": 0.6020308805118931,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00018104558622866756,
+      "loss": 0.8123,
+      "step": 1082
+    },
+    {
+      "epoch": 0.60258728613159,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00018101137163659115,
+      "loss": 0.8452,
+      "step": 1083
+    },
+    {
+      "epoch": 0.6031436917512867,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0001809771294311242,
+      "loss": 0.8303,
+      "step": 1084
+    },
+    {
+      "epoch": 0.6037000973709834,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00018094285962393856,
+      "loss": 0.7941,
+      "step": 1085
+    },
+    {
+      "epoch": 0.6042565029906802,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0001809085622267153,
+      "loss": 0.9412,
+      "step": 1086
+    },
+    {
+      "epoch": 0.604812908610377,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00018087423725114498,
+      "loss": 0.8406,
+      "step": 1087
+    },
+    {
+      "epoch": 0.6053693142300737,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00018083988470892756,
+      "loss": 0.9588,
+      "step": 1088
+    },
+    {
+      "epoch": 0.6059257198497705,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0001808055046117724,
+      "loss": 0.9362,
+      "step": 1089
+    },
+    {
+      "epoch": 0.6064821254694672,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00018077109697139818,
+      "loss": 0.8893,
+      "step": 1090
+    },
+    {
+      "epoch": 0.607038531089164,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00018073666179953306,
+      "loss": 0.8135,
+      "step": 1091
+    },
+    {
+      "epoch": 0.6075949367088608,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00018070219910791457,
+      "loss": 0.8649,
+      "step": 1092
+    },
+    {
+      "epoch": 0.6081513423285575,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001806677089082896,
+      "loss": 0.8343,
+      "step": 1093
+    },
+    {
+      "epoch": 0.6087077479482543,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00018063319121241434,
+      "loss": 0.8991,
+      "step": 1094
+    },
+    {
+      "epoch": 0.609264153567951,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00018059864603205456,
+      "loss": 0.8552,
+      "step": 1095
+    },
+    {
+      "epoch": 0.6098205591876478,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00018056407337898515,
+      "loss": 0.7787,
+      "step": 1096
+    },
+    {
+      "epoch": 0.6103769648073446,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00018052947326499056,
+      "loss": 0.8752,
+      "step": 1097
+    },
+    {
+      "epoch": 0.6109333704270413,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001804948457018645,
+      "loss": 0.851,
+      "step": 1098
+    },
+    {
+      "epoch": 0.611489776046738,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00018046019070141006,
+      "loss": 0.8067,
+      "step": 1099
+    },
+    {
+      "epoch": 0.6120461816664349,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00018042550827543962,
+      "loss": 0.8336,
+      "step": 1100
+    },
+    {
+      "epoch": 0.6126025872861316,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00018039079843577507,
+      "loss": 0.8959,
+      "step": 1101
+    },
+    {
+      "epoch": 0.6131589929058283,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001803560611942475,
+      "loss": 0.8092,
+      "step": 1102
+    },
+    {
+      "epoch": 0.6137153985255251,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00018032129656269738,
+      "loss": 0.8899,
+      "step": 1103
+    },
+    {
+      "epoch": 0.6142718041452219,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0001802865045529745,
+      "loss": 0.9702,
+      "step": 1104
+    },
+    {
+      "epoch": 0.6148282097649186,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00018025168517693803,
+      "loss": 0.7923,
+      "step": 1105
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00018021683844645644,
+      "loss": 0.9082,
+      "step": 1106
+    },
+    {
+      "epoch": 0.6159410210043121,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001801819643734075,
+      "loss": 0.8512,
+      "step": 1107
+    },
+    {
+      "epoch": 0.6164974266240089,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00018014706296967838,
+      "loss": 0.8654,
+      "step": 1108
+    },
+    {
+      "epoch": 0.6170538322437057,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00018011213424716545,
+      "loss": 0.9283,
+      "step": 1109
+    },
+    {
+      "epoch": 0.6176102378634024,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00018007717821777445,
+      "loss": 0.7825,
+      "step": 1110
+    },
+    {
+      "epoch": 0.6181666434830991,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00018004219489342046,
+      "loss": 0.8412,
+      "step": 1111
+    },
+    {
+      "epoch": 0.618723049102796,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00018000718428602786,
+      "loss": 0.8454,
+      "step": 1112
+    },
+    {
+      "epoch": 0.6192794547224927,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00017997214640753023,
+      "loss": 0.9071,
+      "step": 1113
+    },
+    {
+      "epoch": 0.6198358603421894,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00017993708126987058,
+      "loss": 0.9327,
+      "step": 1114
+    },
+    {
+      "epoch": 0.6203922659618862,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001799019888850011,
+      "loss": 0.7861,
+      "step": 1115
+    },
+    {
+      "epoch": 0.620948671581583,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00017986686926488337,
+      "loss": 0.8135,
+      "step": 1116
+    },
+    {
+      "epoch": 0.6215050772012798,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00017983172242148814,
+      "loss": 0.9299,
+      "step": 1117
+    },
+    {
+      "epoch": 0.6220614828209765,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00017979654836679555,
+      "loss": 0.8861,
+      "step": 1118
+    },
+    {
+      "epoch": 0.6226178884406732,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00017976134711279496,
+      "loss": 0.7301,
+      "step": 1119
+    },
+    {
+      "epoch": 0.6231742940603701,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00017972611867148497,
+      "loss": 0.8426,
+      "step": 1120
+    },
+    {
+      "epoch": 0.6237306996800668,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00017969086305487354,
+      "loss": 0.9108,
+      "step": 1121
+    },
+    {
+      "epoch": 0.6242871052997635,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00017965558027497775,
+      "loss": 0.9061,
+      "step": 1122
+    },
+    {
+      "epoch": 0.6248435109194603,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00017962027034382411,
+      "loss": 0.7835,
+      "step": 1123
+    },
+    {
+      "epoch": 0.625399916539157,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00017958493327344826,
+      "loss": 0.92,
+      "step": 1124
+    },
+    {
+      "epoch": 0.6259563221588538,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00017954956907589517,
+      "loss": 0.9074,
+      "step": 1125
+    },
+    {
+      "epoch": 0.6265127277785506,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00017951417776321894,
+      "loss": 0.8155,
+      "step": 1126
+    },
+    {
+      "epoch": 0.6270691333982473,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00017947875934748306,
+      "loss": 0.8642,
+      "step": 1127
+    },
+    {
+      "epoch": 0.627625539017944,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00017944331384076015,
+      "loss": 0.9221,
+      "step": 1128
+    },
+    {
+      "epoch": 0.6281819446376409,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00017940784125513212,
+      "loss": 0.8576,
+      "step": 1129
+    },
+    {
+      "epoch": 0.6287383502573376,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0001793723416026901,
+      "loss": 0.8981,
+      "step": 1130
+    },
+    {
+      "epoch": 0.6292947558770343,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0001793368148955344,
+      "loss": 0.8868,
+      "step": 1131
+    },
+    {
+      "epoch": 0.6298511614967311,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00017930126114577464,
+      "loss": 0.9415,
+      "step": 1132
+    },
+    {
+      "epoch": 0.6304075671164279,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00017926568036552953,
+      "loss": 0.7919,
+      "step": 1133
+    },
+    {
+      "epoch": 0.6309639727361246,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00017923007256692713,
+      "loss": 0.9498,
+      "step": 1134
+    },
+    {
+      "epoch": 0.6315203783558214,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0001791944377621047,
+      "loss": 0.9427,
+      "step": 1135
+    },
+    {
+      "epoch": 0.6320767839755181,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00017915877596320855,
+      "loss": 0.8621,
+      "step": 1136
+    },
+    {
+      "epoch": 0.6326331895952149,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00017912308718239433,
+      "loss": 0.7246,
+      "step": 1137
+    },
+    {
+      "epoch": 0.6331895952149117,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0001790873714318269,
+      "loss": 0.9493,
+      "step": 1138
+    },
+    {
+      "epoch": 0.6337460008346084,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00017905162872368022,
+      "loss": 0.8226,
+      "step": 1139
+    },
+    {
+      "epoch": 0.6343024064543051,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0001790158590701375,
+      "loss": 0.8976,
+      "step": 1140
+    },
+    {
+      "epoch": 0.634858812074002,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0001789800624833911,
+      "loss": 0.9134,
+      "step": 1141
+    },
+    {
+      "epoch": 0.6354152176936987,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001789442389756426,
+      "loss": 0.8021,
+      "step": 1142
+    },
+    {
+      "epoch": 0.6359716233133955,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00017890838855910276,
+      "loss": 0.8757,
+      "step": 1143
+    },
+    {
+      "epoch": 0.6365280289330922,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00017887251124599143,
+      "loss": 0.8675,
+      "step": 1144
+    },
+    {
+      "epoch": 0.637084434552789,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00017883660704853775,
+      "loss": 0.811,
+      "step": 1145
+    },
+    {
+      "epoch": 0.6376408401724858,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0001788006759789799,
+      "loss": 0.8422,
+      "step": 1146
+    },
+    {
+      "epoch": 0.6381972457921825,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00017876471804956535,
+      "loss": 0.751,
+      "step": 1147
+    },
+    {
+      "epoch": 0.6387536514118792,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001787287332725506,
+      "loss": 0.793,
+      "step": 1148
+    },
+    {
+      "epoch": 0.6393100570315761,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00017869272166020134,
+      "loss": 0.7835,
+      "step": 1149
+    },
+    {
+      "epoch": 0.6398664626512728,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001786566832247925,
+      "loss": 0.8919,
+      "step": 1150
+    },
+    {
+      "epoch": 0.6404228682709695,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00017862061797860798,
+      "loss": 0.8714,
+      "step": 1151
+    },
+    {
+      "epoch": 0.6409792738906663,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.000178584525933941,
+      "loss": 0.9351,
+      "step": 1152
+    },
+    {
+      "epoch": 0.641535679510363,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001785484071030938,
+      "loss": 0.92,
+      "step": 1153
+    },
+    {
+      "epoch": 0.6420920851300598,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00017851226149837778,
+      "loss": 0.8925,
+      "step": 1154
+    },
+    {
+      "epoch": 0.6426484907497566,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00017847608913211342,
+      "loss": 0.9021,
+      "step": 1155
+    },
+    {
+      "epoch": 0.6432048963694533,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00017843989001663043,
+      "loss": 0.8723,
+      "step": 1156
+    },
+    {
+      "epoch": 0.64376130198915,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00017840366416426755,
+      "loss": 0.9426,
+      "step": 1157
+    },
+    {
+      "epoch": 0.6443177076088469,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00017836741158737266,
+      "loss": 0.8317,
+      "step": 1158
+    },
+    {
+      "epoch": 0.6448741132285436,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001783311322983028,
+      "loss": 0.906,
+      "step": 1159
+    },
+    {
+      "epoch": 0.6454305188482403,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00017829482630942394,
+      "loss": 0.8644,
+      "step": 1160
+    },
+    {
+      "epoch": 0.6459869244679372,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001782584936331114,
+      "loss": 0.7903,
+      "step": 1161
+    },
+    {
+      "epoch": 0.6465433300876339,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001782221342817494,
+      "loss": 0.8909,
+      "step": 1162
+    },
+    {
+      "epoch": 0.6470997357073306,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00017818574826773133,
+      "loss": 0.8217,
+      "step": 1163
+    },
+    {
+      "epoch": 0.6476561413270274,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001781493356034597,
+      "loss": 0.7876,
+      "step": 1164
+    },
+    {
+      "epoch": 0.6482125469467241,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00017811289630134601,
+      "loss": 0.8962,
+      "step": 1165
+    },
+    {
+      "epoch": 0.648768952566421,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00017807643037381094,
+      "loss": 0.8133,
+      "step": 1166
+    },
+    {
+      "epoch": 0.6493253581861177,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0001780399378332842,
+      "loss": 0.6899,
+      "step": 1167
+    },
+    {
+      "epoch": 0.6498817638058144,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00017800341869220456,
+      "loss": 0.8604,
+      "step": 1168
+    },
+    {
+      "epoch": 0.6504381694255112,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00017796687296301985,
+      "loss": 0.7746,
+      "step": 1169
+    },
+    {
+      "epoch": 0.650994575045208,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00017793030065818699,
+      "loss": 0.8599,
+      "step": 1170
+    },
+    {
+      "epoch": 0.6515509806649047,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00017789370179017194,
+      "loss": 0.8365,
+      "step": 1171
+    },
+    {
+      "epoch": 0.6521073862846015,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00017785707637144976,
+      "loss": 0.8358,
+      "step": 1172
+    },
+    {
+      "epoch": 0.6526637919042982,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001778204244145045,
+      "loss": 0.8689,
+      "step": 1173
+    },
+    {
+      "epoch": 0.653220197523995,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00017778374593182931,
+      "loss": 0.8538,
+      "step": 1174
+    },
+    {
+      "epoch": 0.6537766031436918,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00017774704093592632,
+      "loss": 0.8759,
+      "step": 1175
+    },
+    {
+      "epoch": 0.6543330087633885,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001777103094393067,
+      "loss": 0.902,
+      "step": 1176
+    },
+    {
+      "epoch": 0.6548894143830852,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00017767355145449073,
+      "loss": 0.8503,
+      "step": 1177
+    },
+    {
+      "epoch": 0.6554458200027821,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001776367669940077,
+      "loss": 0.8096,
+      "step": 1178
+    },
+    {
+      "epoch": 0.6560022256224788,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001775999560703958,
+      "loss": 0.7987,
+      "step": 1179
+    },
+    {
+      "epoch": 0.6565586312421755,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0001775631186962024,
+      "loss": 0.8789,
+      "step": 1180
+    },
+    {
+      "epoch": 0.6571150368618723,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00017752625488398383,
+      "loss": 0.8525,
+      "step": 1181
+    },
+    {
+      "epoch": 0.6576714424815691,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001774893646463054,
+      "loss": 0.8729,
+      "step": 1182
+    },
+    {
+      "epoch": 0.6582278481012658,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00017745244799574145,
+      "loss": 0.9459,
+      "step": 1183
+    },
+    {
+      "epoch": 0.6587842537209626,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00017741550494487532,
+      "loss": 0.8822,
+      "step": 1184
+    },
+    {
+      "epoch": 0.6593406593406593,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00017737853550629936,
+      "loss": 0.9891,
+      "step": 1185
+    },
+    {
+      "epoch": 0.659897064960356,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001773415396926149,
+      "loss": 0.7885,
+      "step": 1186
+    },
+    {
+      "epoch": 0.6604534705800529,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00017730451751643228,
+      "loss": 0.8582,
+      "step": 1187
+    },
+    {
+      "epoch": 0.6610098761997496,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00017726746899037079,
+      "loss": 0.9027,
+      "step": 1188
+    },
+    {
+      "epoch": 0.6615662818194463,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00017723039412705877,
+      "loss": 0.802,
+      "step": 1189
+    },
+    {
+      "epoch": 0.6621226874391432,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00017719329293913342,
+      "loss": 0.8545,
+      "step": 1190
+    },
+    {
+      "epoch": 0.6626790930588399,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00017715616543924104,
+      "loss": 0.8496,
+      "step": 1191
+    },
+    {
+      "epoch": 0.6632354986785367,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001771190116400368,
+      "loss": 0.7292,
+      "step": 1192
+    },
+    {
+      "epoch": 0.6637919042982334,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00017708183155418493,
+      "loss": 0.7995,
+      "step": 1193
+    },
+    {
+      "epoch": 0.6643483099179301,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00017704462519435853,
+      "loss": 0.8858,
+      "step": 1194
+    },
+    {
+      "epoch": 0.664904715537627,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001770073925732397,
+      "loss": 0.8412,
+      "step": 1195
+    },
+    {
+      "epoch": 0.6654611211573237,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001769701337035195,
+      "loss": 0.8375,
+      "step": 1196
+    },
+    {
+      "epoch": 0.6660175267770204,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001769328485978979,
+      "loss": 0.8676,
+      "step": 1197
+    },
+    {
+      "epoch": 0.6665739323967173,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00017689553726908387,
+      "loss": 0.9652,
+      "step": 1198
+    },
+    {
+      "epoch": 0.667130338016414,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00017685819972979524,
+      "loss": 0.7994,
+      "step": 1199
+    },
+    {
+      "epoch": 0.6676867436361107,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0001768208359927588,
+      "loss": 0.7993,
+      "step": 1200
+    },
+    {
+      "epoch": 0.6682431492558075,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00017678344607071034,
+      "loss": 0.7557,
+      "step": 1201
+    },
+    {
+      "epoch": 0.6687995548755042,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001767460299763945,
+      "loss": 0.8166,
+      "step": 1202
+    },
+    {
+      "epoch": 0.669355960495201,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00017670858772256487,
+      "loss": 0.8687,
+      "step": 1203
+    },
+    {
+      "epoch": 0.6699123661148978,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00017667111932198392,
+      "loss": 0.8069,
+      "step": 1204
+    },
+    {
+      "epoch": 0.6704687717345945,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0001766336247874231,
+      "loss": 0.8764,
+      "step": 1205
+    },
+    {
+      "epoch": 0.6710251773542912,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00017659610413166269,
+      "loss": 0.7519,
+      "step": 1206
+    },
+    {
+      "epoch": 0.6715815829739881,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00017655855736749198,
+      "loss": 0.945,
+      "step": 1207
+    },
+    {
+      "epoch": 0.6721379885936848,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00017652098450770902,
+      "loss": 0.8002,
+      "step": 1208
+    },
+    {
+      "epoch": 0.6726943942133815,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00017648338556512086,
+      "loss": 0.8359,
+      "step": 1209
+    },
+    {
+      "epoch": 0.6732507998330783,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00017644576055254337,
+      "loss": 0.8427,
+      "step": 1210
+    },
+    {
+      "epoch": 0.6738072054527751,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00017640810948280143,
+      "loss": 0.8537,
+      "step": 1211
+    },
+    {
+      "epoch": 0.6743636110724718,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00017637043236872865,
+      "loss": 0.924,
+      "step": 1212
+    },
+    {
+      "epoch": 0.6749200166921686,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00017633272922316757,
+      "loss": 0.9067,
+      "step": 1213
+    },
+    {
+      "epoch": 0.6754764223118653,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00017629500005896968,
+      "loss": 0.9031,
+      "step": 1214
+    },
+    {
+      "epoch": 0.6760328279315622,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001762572448889952,
+      "loss": 0.7696,
+      "step": 1215
+    },
+    {
+      "epoch": 0.6765892335512589,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00017621946372611336,
+      "loss": 0.7062,
+      "step": 1216
+    },
+    {
+      "epoch": 0.6771456391709556,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00017618165658320214,
+      "loss": 0.8407,
+      "step": 1217
+    },
+    {
+      "epoch": 0.6777020447906524,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00017614382347314842,
+      "loss": 0.8517,
+      "step": 1218
+    },
+    {
+      "epoch": 0.6782584504103492,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00017610596440884792,
+      "loss": 0.8495,
+      "step": 1219
+    },
+    {
+      "epoch": 0.6788148560300459,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00017606807940320524,
+      "loss": 0.8023,
+      "step": 1220
+    },
+    {
+      "epoch": 0.6793712616497427,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00017603016846913375,
+      "loss": 0.8952,
+      "step": 1221
+    },
+    {
+      "epoch": 0.6799276672694394,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00017599223161955573,
+      "loss": 0.8634,
+      "step": 1222
+    },
+    {
+      "epoch": 0.6804840728891361,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0001759542688674023,
+      "loss": 0.9523,
+      "step": 1223
+    },
+    {
+      "epoch": 0.681040478508833,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001759162802256133,
+      "loss": 0.7981,
+      "step": 1224
+    },
+    {
+      "epoch": 0.6815968841285297,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00017587826570713752,
+      "loss": 0.9015,
+      "step": 1225
+    },
+    {
+      "epoch": 0.6821532897482264,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00017584022532493254,
+      "loss": 0.8965,
+      "step": 1226
+    },
+    {
+      "epoch": 0.6827096953679233,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00017580215909196467,
+      "loss": 0.8798,
+      "step": 1227
+    },
+    {
+      "epoch": 0.68326610098762,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00017576406702120916,
+      "loss": 0.8244,
+      "step": 1228
+    },
+    {
+      "epoch": 0.6838225066073167,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00017572594912564996,
+      "loss": 0.8822,
+      "step": 1229
+    },
+    {
+      "epoch": 0.6843789122270135,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00017568780541827987,
+      "loss": 0.8287,
+      "step": 1230
+    },
+    {
+      "epoch": 0.6849353178467102,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00017564963591210054,
+      "loss": 0.8288,
+      "step": 1231
+    },
+    {
+      "epoch": 0.685491723466407,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0001756114406201223,
+      "loss": 0.9075,
+      "step": 1232
+    },
+    {
+      "epoch": 0.6860481290861038,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00017557321955536436,
+      "loss": 0.9005,
+      "step": 1233
+    },
+    {
+      "epoch": 0.6866045347058005,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00017553497273085466,
+      "loss": 0.8201,
+      "step": 1234
+    },
+    {
+      "epoch": 0.6871609403254972,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00017549670015962998,
+      "loss": 0.8144,
+      "step": 1235
+    },
+    {
+      "epoch": 0.6877173459451941,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00017545840185473582,
+      "loss": 0.8734,
+      "step": 1236
+    },
+    {
+      "epoch": 0.6882737515648908,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00017542007782922646,
+      "loss": 0.8758,
+      "step": 1237
+    },
+    {
+      "epoch": 0.6888301571845875,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00017538172809616494,
+      "loss": 0.7767,
+      "step": 1238
+    },
+    {
+      "epoch": 0.6893865628042843,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00017534335266862315,
+      "loss": 0.7504,
+      "step": 1239
+    },
+    {
+      "epoch": 0.6899429684239811,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001753049515596816,
+      "loss": 0.8237,
+      "step": 1240
+    },
+    {
+      "epoch": 0.6904993740436779,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00017526652478242966,
+      "loss": 0.7673,
+      "step": 1241
+    },
+    {
+      "epoch": 0.6910557796633746,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00017522807234996538,
+      "loss": 0.8724,
+      "step": 1242
+    },
+    {
+      "epoch": 0.6916121852830713,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00017518959427539562,
+      "loss": 0.8233,
+      "step": 1243
+    },
+    {
+      "epoch": 0.6921685909027682,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00017515109057183596,
+      "loss": 0.6871,
+      "step": 1244
+    },
+    {
+      "epoch": 0.6927249965224649,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00017511256125241068,
+      "loss": 0.7386,
+      "step": 1245
+    },
+    {
+      "epoch": 0.6932814021421616,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00017507400633025278,
+      "loss": 0.7724,
+      "step": 1246
+    },
+    {
+      "epoch": 0.6938378077618584,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00017503542581850408,
+      "loss": 0.81,
+      "step": 1247
+    },
+    {
+      "epoch": 0.6943942133815552,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00017499681973031503,
+      "loss": 1.8259,
+      "step": 1248
+    },
+    {
+      "epoch": 0.6949506190012519,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00017495818807884486,
+      "loss": 0.7351,
+      "step": 1249
+    },
+    {
+      "epoch": 0.6955070246209487,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00017491953087726142,
+      "loss": 0.9045,
+      "step": 1250
+    },
+    {
+      "epoch": 0.6960634302406454,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00017488084813874143,
+      "loss": 0.6979,
+      "step": 1251
+    },
+    {
+      "epoch": 0.6966198358603422,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00017484213987647014,
+      "loss": 0.8801,
+      "step": 1252
+    },
+    {
+      "epoch": 0.697176241480039,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00017480340610364161,
+      "loss": 0.8701,
+      "step": 1253
+    },
+    {
+      "epoch": 0.6977326470997357,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0001747646468334586,
+      "loss": 0.8198,
+      "step": 1254
+    },
+    {
+      "epoch": 0.6982890527194324,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00017472586207913245,
+      "loss": 0.8907,
+      "step": 1255
+    },
+    {
+      "epoch": 0.6988454583391293,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00017468705185388333,
+      "loss": 0.8712,
+      "step": 1256
+    },
+    {
+      "epoch": 0.699401863958826,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00017464821617093997,
+      "loss": 1.0164,
+      "step": 1257
+    },
+    {
+      "epoch": 0.6999582695785227,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00017460935504353984,
+      "loss": 0.9493,
+      "step": 1258
+    },
+    {
+      "epoch": 0.7005146751982195,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001745704684849291,
+      "loss": 0.7055,
+      "step": 1259
+    },
+    {
+      "epoch": 0.7010710808179162,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00017453155650836255,
+      "loss": 0.8317,
+      "step": 1260
+    },
+    {
+      "epoch": 0.701627486437613,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00017449261912710364,
+      "loss": 0.8239,
+      "step": 1261
+    },
+    {
+      "epoch": 0.7021838920573098,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001744536563544245,
+      "loss": 0.771,
+      "step": 1262
+    },
+    {
+      "epoch": 0.7027402976770065,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00017441466820360596,
+      "loss": 0.9036,
+      "step": 1263
+    },
+    {
+      "epoch": 0.7032967032967034,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00017437565468793732,
+      "loss": 0.7954,
+      "step": 1264
+    },
+    {
+      "epoch": 0.7038531089164001,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00017433661582071677,
+      "loss": 0.8681,
+      "step": 1265
+    },
+    {
+      "epoch": 0.7044095145360968,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00017429755161525105,
+      "loss": 0.8394,
+      "step": 1266
+    },
+    {
+      "epoch": 0.7049659201557936,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00017425846208485543,
+      "loss": 0.8418,
+      "step": 1267
+    },
+    {
+      "epoch": 0.7055223257754903,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00017421934724285394,
+      "loss": 0.9183,
+      "step": 1268
+    },
+    {
+      "epoch": 0.7060787313951871,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00017418020710257916,
+      "loss": 0.8866,
+      "step": 1269
+    },
+    {
+      "epoch": 0.7066351370148839,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00017414104167737238,
+      "loss": 0.8109,
+      "step": 1270
+    },
+    {
+      "epoch": 0.7071915426345806,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0001741018509805834,
+      "loss": 0.922,
+      "step": 1271
+    },
+    {
+      "epoch": 0.7077479482542773,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00017406263502557073,
+      "loss": 0.9158,
+      "step": 1272
+    },
+    {
+      "epoch": 0.7083043538739742,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00017402339382570147,
+      "loss": 0.8446,
+      "step": 1273
+    },
+    {
+      "epoch": 0.7088607594936709,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00017398412739435125,
+      "loss": 0.881,
+      "step": 1274
+    },
+    {
+      "epoch": 0.7094171651133676,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00017394483574490437,
+      "loss": 0.9032,
+      "step": 1275
+    },
+    {
+      "epoch": 0.7099735707330644,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00017390551889075374,
+      "loss": 0.8782,
+      "step": 1276
+    },
+    {
+      "epoch": 0.7105299763527612,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00017386617684530083,
+      "loss": 0.8744,
+      "step": 1277
+    },
+    {
+      "epoch": 0.7110863819724579,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00017382680962195566,
+      "loss": 0.8058,
+      "step": 1278
+    },
+    {
+      "epoch": 0.7116427875921547,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001737874172341369,
+      "loss": 0.6973,
+      "step": 1279
+    },
+    {
+      "epoch": 0.7121991932118514,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00017374799969527174,
+      "loss": 0.8964,
+      "step": 1280
+    },
+    {
+      "epoch": 0.7127555988315482,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.000173708557018796,
+      "loss": 0.8244,
+      "step": 1281
+    },
+    {
+      "epoch": 0.713312004451245,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00017366908921815404,
+      "loss": 0.8864,
+      "step": 1282
+    },
+    {
+      "epoch": 0.7138684100709417,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001736295963067988,
+      "loss": 0.8349,
+      "step": 1283
+    },
+    {
+      "epoch": 0.7144248156906384,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0001735900782981917,
+      "loss": 0.8759,
+      "step": 1284
+    },
+    {
+      "epoch": 0.7149812213103353,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00017355053520580284,
+      "loss": 0.7247,
+      "step": 1285
+    },
+    {
+      "epoch": 0.715537626930032,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001735109670431108,
+      "loss": 0.8187,
+      "step": 1286
+    },
+    {
+      "epoch": 0.7160940325497287,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00017347137382360268,
+      "loss": 0.8287,
+      "step": 1287
+    },
+    {
+      "epoch": 0.7166504381694255,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00017343175556077417,
+      "loss": 0.8427,
+      "step": 1288
+    },
+    {
+      "epoch": 0.7172068437891223,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00017339211226812947,
+      "loss": 0.7638,
+      "step": 1289
+    },
+    {
+      "epoch": 0.7177632494088191,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00017335244395918136,
+      "loss": 0.8827,
+      "step": 1290
+    },
+    {
+      "epoch": 0.7183196550285158,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00017331275064745105,
+      "loss": 0.925,
+      "step": 1291
+    },
+    {
+      "epoch": 0.7188760606482125,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00017327303234646838,
+      "loss": 0.7816,
+      "step": 1292
+    },
+    {
+      "epoch": 0.7194324662679094,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00017323328906977166,
+      "loss": 0.9068,
+      "step": 1293
+    },
+    {
+      "epoch": 0.7199888718876061,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00017319352083090765,
+      "loss": 0.8168,
+      "step": 1294
+    },
+    {
+      "epoch": 0.7205452775073028,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00017315372764343173,
+      "loss": 0.8039,
+      "step": 1295
+    },
+    {
+      "epoch": 0.7211016831269996,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00017311390952090775,
+      "loss": 0.7918,
+      "step": 1296
+    },
+    {
+      "epoch": 0.7216580887466963,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.000173074066476908,
+      "loss": 0.8398,
+      "step": 1297
+    },
+    {
+      "epoch": 0.7222144943663931,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00017303419852501332,
+      "loss": 0.7868,
+      "step": 1298
+    },
+    {
+      "epoch": 0.7227708999860899,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00017299430567881306,
+      "loss": 0.7343,
+      "step": 1299
+    },
+    {
+      "epoch": 0.7233273056057866,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.000172954387951905,
+      "loss": 0.8459,
+      "step": 1300
+    },
+    {
+      "epoch": 0.7238837112254833,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0001729144453578954,
+      "loss": 0.8439,
+      "step": 1301
+    },
+    {
+      "epoch": 0.7244401168451802,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00017287447791039905,
+      "loss": 0.8974,
+      "step": 1302
+    },
+    {
+      "epoch": 0.7249965224648769,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00017283448562303922,
+      "loss": 0.8415,
+      "step": 1303
+    },
+    {
+      "epoch": 0.7255529280845736,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00017279446850944753,
+      "loss": 0.8327,
+      "step": 1304
+    },
+    {
+      "epoch": 0.7261093337042704,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0001727544265832642,
+      "loss": 0.9265,
+      "step": 1305
+    },
+    {
+      "epoch": 0.7266657393239672,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001727143598581378,
+      "loss": 0.8761,
+      "step": 1306
+    },
+    {
+      "epoch": 0.7272221449436639,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00017267426834772546,
+      "loss": 0.8675,
+      "step": 1307
+    },
+    {
+      "epoch": 0.7277785505633607,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00017263415206569268,
+      "loss": 0.8685,
+      "step": 1308
+    },
+    {
+      "epoch": 0.7283349561830574,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00017259401102571336,
+      "loss": 0.882,
+      "step": 1309
+    },
+    {
+      "epoch": 0.7288913618027542,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00017255384524146998,
+      "loss": 0.7924,
+      "step": 1310
+    },
+    {
+      "epoch": 0.729447767422451,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001725136547266534,
+      "loss": 0.8117,
+      "step": 1311
+    },
+    {
+      "epoch": 0.7300041730421477,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001724734394949628,
+      "loss": 0.8178,
+      "step": 1312
+    },
+    {
+      "epoch": 0.7305605786618445,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00017243319956010587,
+      "loss": 0.7766,
+      "step": 1313
+    },
+    {
+      "epoch": 0.7311169842815413,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00017239293493579884,
+      "loss": 0.8458,
+      "step": 1314
+    },
+    {
+      "epoch": 0.731673389901238,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00017235264563576612,
+      "loss": 0.8981,
+      "step": 1315
+    },
+    {
+      "epoch": 0.7322297955209348,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001723123316737407,
+      "loss": 0.7909,
+      "step": 1316
+    },
+    {
+      "epoch": 0.7327862011406315,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00017227199306346392,
+      "loss": 0.9362,
+      "step": 1317
+    },
+    {
+      "epoch": 0.7333426067603283,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001722316298186855,
+      "loss": 0.8683,
+      "step": 1318
+    },
+    {
+      "epoch": 0.7338990123800251,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00017219124195316362,
+      "loss": 0.8612,
+      "step": 1319
+    },
+    {
+      "epoch": 0.7344554179997218,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001721508294806648,
+      "loss": 0.862,
+      "step": 1320
+    },
+    {
+      "epoch": 0.7350118236194185,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00017211039241496394,
+      "loss": 0.8873,
+      "step": 1321
+    },
+    {
+      "epoch": 0.7355682292391154,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001720699307698444,
+      "loss": 0.9028,
+      "step": 1322
+    },
+    {
+      "epoch": 0.7361246348588121,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001720294445590978,
+      "loss": 0.8096,
+      "step": 1323
+    },
+    {
+      "epoch": 0.7366810404785088,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00017198893379652424,
+      "loss": 0.8278,
+      "step": 1324
+    },
+    {
+      "epoch": 0.7372374460982056,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00017194839849593214,
+      "loss": 0.8649,
+      "step": 1325
+    },
+    {
+      "epoch": 0.7377938517179023,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00017190783867113827,
+      "loss": 0.8515,
+      "step": 1326
+    },
+    {
+      "epoch": 0.7383502573375991,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00017186725433596779,
+      "loss": 0.7985,
+      "step": 1327
+    },
+    {
+      "epoch": 0.7389066629572959,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00017182664550425422,
+      "loss": 0.7828,
+      "step": 1328
+    },
+    {
+      "epoch": 0.7394630685769926,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00017178601218983941,
+      "loss": 0.7932,
+      "step": 1329
+    },
+    {
+      "epoch": 0.7400194741966893,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0001717453544065735,
+      "loss": 0.8869,
+      "step": 1330
+    },
+    {
+      "epoch": 0.7405758798163862,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00017170467216831513,
+      "loss": 0.8317,
+      "step": 1331
+    },
+    {
+      "epoch": 0.7411322854360829,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00017166396548893106,
+      "loss": 0.7774,
+      "step": 1332
+    },
+    {
+      "epoch": 0.7416886910557796,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00017162323438229662,
+      "loss": 0.8914,
+      "step": 1333
+    },
+    {
+      "epoch": 0.7422450966754764,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00017158247886229523,
+      "loss": 0.8156,
+      "step": 1334
+    },
+    {
+      "epoch": 0.7428015022951732,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001715416989428188,
+      "loss": 0.829,
+      "step": 1335
+    },
+    {
+      "epoch": 0.7433579079148699,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00017150089463776754,
+      "loss": 0.8551,
+      "step": 1336
+    },
+    {
+      "epoch": 0.7439143135345667,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00017146006596104982,
+      "loss": 0.8098,
+      "step": 1337
+    },
+    {
+      "epoch": 0.7444707191542634,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00017141921292658253,
+      "loss": 0.702,
+      "step": 1338
+    },
+    {
+      "epoch": 0.7450271247739603,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001713783355482907,
+      "loss": 0.805,
+      "step": 1339
+    },
+    {
+      "epoch": 0.745583530393657,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00017133743384010778,
+      "loss": 0.7702,
+      "step": 1340
+    },
+    {
+      "epoch": 0.7461399360133537,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001712965078159754,
+      "loss": 0.8625,
+      "step": 1341
+    },
+    {
+      "epoch": 0.7466963416330505,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00017125555748984353,
+      "loss": 0.7743,
+      "step": 1342
+    },
+    {
+      "epoch": 0.7472527472527473,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00017121458287567044,
+      "loss": 0.8225,
+      "step": 1343
+    },
+    {
+      "epoch": 0.747809152872444,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00017117358398742268,
+      "loss": 0.8579,
+      "step": 1344
+    },
+    {
+      "epoch": 0.7483655584921408,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00017113256083907505,
+      "loss": 0.8478,
+      "step": 1345
+    },
+    {
+      "epoch": 0.7489219641118375,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00017109151344461062,
+      "loss": 0.8209,
+      "step": 1346
+    },
+    {
+      "epoch": 0.7494783697315343,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001710504418180207,
+      "loss": 0.7909,
+      "step": 1347
+    },
+    {
+      "epoch": 0.7500347753512311,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00017100934597330495,
+      "loss": 0.8893,
+      "step": 1348
+    },
+    {
+      "epoch": 0.7505911809709278,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001709682259244712,
+      "loss": 0.7629,
+      "step": 1349
+    },
+    {
+      "epoch": 0.7511475865906245,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00017092708168553556,
+      "loss": 0.8391,
+      "step": 1350
+    },
+    {
+      "epoch": 0.7517039922103214,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00017088591327052237,
+      "loss": 0.8319,
+      "step": 1351
+    },
+    {
+      "epoch": 0.7522603978300181,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001708447206934642,
+      "loss": 0.8358,
+      "step": 1352
+    },
+    {
+      "epoch": 0.7528168034497148,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00017080350396840192,
+      "loss": 0.8513,
+      "step": 1353
+    },
+    {
+      "epoch": 0.7533732090694116,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001707622631093846,
+      "loss": 0.8373,
+      "step": 1354
+    },
+    {
+      "epoch": 0.7539296146891084,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00017072099813046944,
+      "loss": 0.874,
+      "step": 1355
+    },
+    {
+      "epoch": 0.7544860203088051,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00017067970904572206,
+      "loss": 0.7982,
+      "step": 1356
+    },
+    {
+      "epoch": 0.7550424259285019,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001706383958692161,
+      "loss": 0.8553,
+      "step": 1357
+    },
+    {
+      "epoch": 0.7555988315481986,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00017059705861503351,
+      "loss": 0.8443,
+      "step": 1358
+    },
+    {
+      "epoch": 0.7561552371678953,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001705556972972645,
+      "loss": 0.8124,
+      "step": 1359
+    },
+    {
+      "epoch": 0.7567116427875922,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0001705143119300073,
+      "loss": 0.9886,
+      "step": 1360
+    },
+    {
+      "epoch": 0.7572680484072889,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00017047290252736853,
+      "loss": 0.8099,
+      "step": 1361
+    },
+    {
+      "epoch": 0.7578244540269857,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00017043146910346294,
+      "loss": 0.767,
+      "step": 1362
+    },
+    {
+      "epoch": 0.7583808596466824,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00017039001167241336,
+      "loss": 0.9191,
+      "step": 1363
+    },
+    {
+      "epoch": 0.7589372652663792,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.000170348530248351,
+      "loss": 0.8611,
+      "step": 1364
+    },
+    {
+      "epoch": 0.759493670886076,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00017030702484541505,
+      "loss": 0.7678,
+      "step": 1365
+    },
+    {
+      "epoch": 0.7600500765057727,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00017026549547775302,
+      "loss": 0.8689,
+      "step": 1366
+    },
+    {
+      "epoch": 0.7606064821254694,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0001702239421595206,
+      "loss": 0.8887,
+      "step": 1367
+    },
+    {
+      "epoch": 0.7611628877451663,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00017018236490488144,
+      "loss": 0.8028,
+      "step": 1368
+    },
+    {
+      "epoch": 0.761719293364863,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00017014076372800761,
+      "loss": 0.8483,
+      "step": 1369
+    },
+    {
+      "epoch": 0.7622756989845597,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001700991386430791,
+      "loss": 0.776,
+      "step": 1370
+    },
+    {
+      "epoch": 0.7628321046042565,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001700574896642843,
+      "loss": 0.9116,
+      "step": 1371
+    },
+    {
+      "epoch": 0.7633885102239533,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00017001581680581944,
+      "loss": 0.7911,
+      "step": 1372
+    },
+    {
+      "epoch": 0.76394491584365,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00016997412008188918,
+      "loss": 0.828,
+      "step": 1373
+    },
+    {
+      "epoch": 0.7645013214633468,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00016993239950670614,
+      "loss": 0.8273,
+      "step": 1374
+    },
+    {
+      "epoch": 0.7650577270830435,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00016989065509449115,
+      "loss": 0.789,
+      "step": 1375
+    },
+    {
+      "epoch": 0.7656141327027403,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00016984888685947312,
+      "loss": 0.8335,
+      "step": 1376
+    },
+    {
+      "epoch": 0.7661705383224371,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0001698070948158891,
+      "loss": 0.9027,
+      "step": 1377
+    },
+    {
+      "epoch": 0.7667269439421338,
+      "grad_norm": 0.921875,
+      "learning_rate": 0.00016976527897798424,
+      "loss": 1.8603,
+      "step": 1378
+    },
+    {
+      "epoch": 0.7672833495618305,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00016972343936001182,
+      "loss": 0.8506,
+      "step": 1379
+    },
+    {
+      "epoch": 0.7678397551815274,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001696815759762332,
+      "loss": 0.8211,
+      "step": 1380
+    },
+    {
+      "epoch": 0.7683961608012241,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001696396888409179,
+      "loss": 0.8601,
+      "step": 1381
+    },
+    {
+      "epoch": 0.7689525664209208,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00016959777796834343,
+      "loss": 0.8136,
+      "step": 1382
+    },
+    {
+      "epoch": 0.7695089720406176,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001695558433727955,
+      "loss": 0.8032,
+      "step": 1383
+    },
+    {
+      "epoch": 0.7700653776603144,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00016951388506856785,
+      "loss": 0.9557,
+      "step": 1384
+    },
+    {
+      "epoch": 0.7706217832800111,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00016947190306996235,
+      "loss": 0.9532,
+      "step": 1385
+    },
+    {
+      "epoch": 0.7711781888997079,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00016942989739128885,
+      "loss": 0.9016,
+      "step": 1386
+    },
+    {
+      "epoch": 0.7717345945194046,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001693878680468653,
+      "loss": 0.7211,
+      "step": 1387
+    },
+    {
+      "epoch": 0.7722910001391015,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00016934581505101786,
+      "loss": 0.8251,
+      "step": 1388
+    },
+    {
+      "epoch": 0.7728474057587982,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00016930373841808055,
+      "loss": 0.9217,
+      "step": 1389
+    },
+    {
+      "epoch": 0.7734038113784949,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00016926163816239557,
+      "loss": 0.798,
+      "step": 1390
+    },
+    {
+      "epoch": 0.7739602169981917,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0001692195142983131,
+      "loss": 0.9148,
+      "step": 1391
+    },
+    {
+      "epoch": 0.7745166226178885,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00016917736684019148,
+      "loss": 0.7434,
+      "step": 1392
+    },
+    {
+      "epoch": 0.7750730282375852,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00016913519580239697,
+      "loss": 0.7526,
+      "step": 1393
+    },
+    {
+      "epoch": 0.775629433857282,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00016909300119930388,
+      "loss": 0.7844,
+      "step": 1394
+    },
+    {
+      "epoch": 0.7761858394769787,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00016905078304529464,
+      "loss": 0.8256,
+      "step": 1395
+    },
+    {
+      "epoch": 0.7767422450966754,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00016900854135475966,
+      "loss": 0.8872,
+      "step": 1396
+    },
+    {
+      "epoch": 0.7772986507163723,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001689662761420973,
+      "loss": 0.8541,
+      "step": 1397
+    },
+    {
+      "epoch": 0.777855056336069,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00016892398742171408,
+      "loss": 0.8455,
+      "step": 1398
+    },
+    {
+      "epoch": 0.7784114619557657,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001688816752080244,
+      "loss": 0.9086,
+      "step": 1399
+    },
+    {
+      "epoch": 0.7789678675754625,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00016883933951545075,
+      "loss": 0.7954,
+      "step": 1400
+    },
+    {
+      "epoch": 0.7795242731951593,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0001687969803584236,
+      "loss": 0.8604,
+      "step": 1401
+    },
+    {
+      "epoch": 0.780080678814856,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00016875459775138142,
+      "loss": 0.7879,
+      "step": 1402
+    },
+    {
+      "epoch": 0.7806370844345528,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00016871219170877066,
+      "loss": 0.6737,
+      "step": 1403
+    },
+    {
+      "epoch": 0.7811934900542495,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00016866976224504576,
+      "loss": 0.8641,
+      "step": 1404
+    },
+    {
+      "epoch": 0.7817498956739463,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00016862730937466916,
+      "loss": 0.7594,
+      "step": 1405
+    },
+    {
+      "epoch": 0.7823063012936431,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001685848331121113,
+      "loss": 0.7111,
+      "step": 1406
+    },
+    {
+      "epoch": 0.7828627069133398,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00016854233347185052,
+      "loss": 0.7913,
+      "step": 1407
+    },
+    {
+      "epoch": 0.7834191125330365,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00016849981046837317,
+      "loss": 0.8784,
+      "step": 1408
+    },
+    {
+      "epoch": 0.7839755181527334,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001684572641161736,
+      "loss": 0.8471,
+      "step": 1409
+    },
+    {
+      "epoch": 0.7845319237724301,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00016841469442975405,
+      "loss": 0.8297,
+      "step": 1410
+    },
+    {
+      "epoch": 0.7850883293921269,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00016837210142362473,
+      "loss": 0.907,
+      "step": 1411
+    },
+    {
+      "epoch": 0.7856447350118236,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00016832948511230386,
+      "loss": 0.7822,
+      "step": 1412
+    },
+    {
+      "epoch": 0.7862011406315204,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00016828684551031757,
+      "loss": 0.8785,
+      "step": 1413
+    },
+    {
+      "epoch": 0.7867575462512172,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00016824418263219987,
+      "loss": 0.9351,
+      "step": 1414
+    },
+    {
+      "epoch": 0.7873139518709139,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00016820149649249277,
+      "loss": 0.8879,
+      "step": 1415
+    },
+    {
+      "epoch": 0.7878703574906106,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00016815878710574618,
+      "loss": 0.8392,
+      "step": 1416
+    },
+    {
+      "epoch": 0.7884267631103075,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.000168116054486518,
+      "loss": 0.812,
+      "step": 1417
+    },
+    {
+      "epoch": 0.7889831687300042,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00016807329864937392,
+      "loss": 0.8525,
+      "step": 1418
+    },
+    {
+      "epoch": 0.7895395743497009,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00016803051960888766,
+      "loss": 0.9073,
+      "step": 1419
+    },
+    {
+      "epoch": 0.7900959799693977,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00016798771737964077,
+      "loss": 0.7898,
+      "step": 1420
+    },
+    {
+      "epoch": 0.7906523855890945,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00016794489197622279,
+      "loss": 0.8647,
+      "step": 1421
+    },
+    {
+      "epoch": 0.7912087912087912,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001679020434132311,
+      "loss": 0.8098,
+      "step": 1422
+    },
+    {
+      "epoch": 0.791765196828488,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00016785917170527095,
+      "loss": 0.8732,
+      "step": 1423
+    },
+    {
+      "epoch": 0.7923216024481847,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00016781627686695552,
+      "loss": 0.8421,
+      "step": 1424
+    },
+    {
+      "epoch": 0.7928780080678814,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001677733589129059,
+      "loss": 0.9202,
+      "step": 1425
+    },
+    {
+      "epoch": 0.7934344136875783,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.000167730417857751,
+      "loss": 0.8072,
+      "step": 1426
+    },
+    {
+      "epoch": 0.793990819307275,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00016768745371612765,
+      "loss": 0.8307,
+      "step": 1427
+    },
+    {
+      "epoch": 0.7945472249269717,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00016764446650268053,
+      "loss": 0.8375,
+      "step": 1428
+    },
+    {
+      "epoch": 0.7951036305466685,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.0001676014562320622,
+      "loss": 1.8024,
+      "step": 1429
+    },
+    {
+      "epoch": 0.7956600361663653,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.000167558422918933,
+      "loss": 0.6899,
+      "step": 1430
+    },
+    {
+      "epoch": 0.796216441786062,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00016751536657796123,
+      "loss": 0.8481,
+      "step": 1431
+    },
+    {
+      "epoch": 0.7967728474057588,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00016747228722382298,
+      "loss": 0.8259,
+      "step": 1432
+    },
+    {
+      "epoch": 0.7973292530254555,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00016742918487120222,
+      "loss": 0.8527,
+      "step": 1433
+    },
+    {
+      "epoch": 0.7978856586451523,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00016738605953479075,
+      "loss": 0.6968,
+      "step": 1434
+    },
+    {
+      "epoch": 0.7984420642648491,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00016734291122928816,
+      "loss": 0.9613,
+      "step": 1435
+    },
+    {
+      "epoch": 0.7989984698845458,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00016729973996940193,
+      "loss": 0.8458,
+      "step": 1436
+    },
+    {
+      "epoch": 0.7995548755042426,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001672565457698473,
+      "loss": 0.8313,
+      "step": 1437
+    },
+    {
+      "epoch": 0.8001112811239394,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00016721332864534744,
+      "loss": 0.7054,
+      "step": 1438
+    },
+    {
+      "epoch": 0.8006676867436361,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00016717008861063322,
+      "loss": 0.9309,
+      "step": 1439
+    },
+    {
+      "epoch": 0.8012240923633329,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00016712682568044334,
+      "loss": 0.7426,
+      "step": 1440
+    },
+    {
+      "epoch": 0.8017804979830296,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00016708353986952433,
+      "loss": 0.8588,
+      "step": 1441
+    },
+    {
+      "epoch": 0.8023369036027264,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00016704023119263058,
+      "loss": 0.9493,
+      "step": 1442
+    },
+    {
+      "epoch": 0.8028933092224232,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001669968996645241,
+      "loss": 0.838,
+      "step": 1443
+    },
+    {
+      "epoch": 0.8034497148421199,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00016695354529997486,
+      "loss": 0.7359,
+      "step": 1444
+    },
+    {
+      "epoch": 0.8040061204618166,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00016691016811376058,
+      "loss": 0.905,
+      "step": 1445
+    },
+    {
+      "epoch": 0.8045625260815135,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00016686676812066668,
+      "loss": 0.8376,
+      "step": 1446
+    },
+    {
+      "epoch": 0.8051189317012102,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00016682334533548642,
+      "loss": 0.7095,
+      "step": 1447
+    },
+    {
+      "epoch": 0.8056753373209069,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00016677989977302078,
+      "loss": 0.8286,
+      "step": 1448
+    },
+    {
+      "epoch": 0.8062317429406037,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0001667364314480786,
+      "loss": 0.8728,
+      "step": 1449
+    },
+    {
+      "epoch": 0.8067881485603005,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00016669294037547635,
+      "loss": 0.8136,
+      "step": 1450
+    },
+    {
+      "epoch": 0.8073445541799972,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00016664942657003838,
+      "loss": 0.8813,
+      "step": 1451
+    },
+    {
+      "epoch": 0.807900959799694,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00016660589004659663,
+      "loss": 0.7414,
+      "step": 1452
+    },
+    {
+      "epoch": 0.8084573654193907,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.000166562330819991,
+      "loss": 0.8392,
+      "step": 1453
+    },
+    {
+      "epoch": 0.8090137710390874,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00016651874890506897,
+      "loss": 0.8908,
+      "step": 1454
+    },
+    {
+      "epoch": 0.8095701766587843,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0001664751443166858,
+      "loss": 0.8836,
+      "step": 1455
+    },
+    {
+      "epoch": 0.810126582278481,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001664315170697044,
+      "loss": 0.8345,
+      "step": 1456
+    },
+    {
+      "epoch": 0.8106829878981777,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00016638786717899551,
+      "loss": 0.7562,
+      "step": 1457
+    },
+    {
+      "epoch": 0.8112393935178746,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0001663441946594376,
+      "loss": 0.8809,
+      "step": 1458
+    },
+    {
+      "epoch": 0.8117957991375713,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0001663004995259168,
+      "loss": 0.8177,
+      "step": 1459
+    },
+    {
+      "epoch": 0.8123522047572681,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00016625678179332692,
+      "loss": 0.8005,
+      "step": 1460
+    },
+    {
+      "epoch": 0.8129086103769648,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00016621304147656948,
+      "loss": 0.8027,
+      "step": 1461
+    },
+    {
+      "epoch": 0.8134650159966615,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001661692785905538,
+      "loss": 0.8325,
+      "step": 1462
+    },
+    {
+      "epoch": 0.8140214216163584,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001661254931501968,
+      "loss": 0.7392,
+      "step": 1463
+    },
+    {
+      "epoch": 0.8145778272360551,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00016608168517042305,
+      "loss": 0.819,
+      "step": 1464
+    },
+    {
+      "epoch": 0.8151342328557518,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00016603785466616492,
+      "loss": 0.7979,
+      "step": 1465
+    },
+    {
+      "epoch": 0.8156906384754486,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001659940016523624,
+      "loss": 0.7334,
+      "step": 1466
+    },
+    {
+      "epoch": 0.8162470440951454,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0001659501261439631,
+      "loss": 0.6871,
+      "step": 1467
+    },
+    {
+      "epoch": 0.8168034497148421,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00016590622815592236,
+      "loss": 0.902,
+      "step": 1468
+    },
+    {
+      "epoch": 0.8173598553345389,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001658623077032032,
+      "loss": 0.896,
+      "step": 1469
+    },
+    {
+      "epoch": 0.8179162609542356,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00016581836480077623,
+      "loss": 0.8791,
+      "step": 1470
+    },
+    {
+      "epoch": 0.8184726665739324,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00016577439946361978,
+      "loss": 0.9967,
+      "step": 1471
+    },
+    {
+      "epoch": 0.8190290721936292,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001657304117067198,
+      "loss": 0.8007,
+      "step": 1472
+    },
+    {
+      "epoch": 0.8195854778133259,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00016568640154506983,
+      "loss": 0.9419,
+      "step": 1473
+    },
+    {
+      "epoch": 0.8201418834330226,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00016564236899367115,
+      "loss": 0.8375,
+      "step": 1474
+    },
+    {
+      "epoch": 0.8206982890527195,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00016559831406753257,
+      "loss": 0.7735,
+      "step": 1475
+    },
+    {
+      "epoch": 0.8212546946724162,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00016555423678167063,
+      "loss": 0.7933,
+      "step": 1476
+    },
+    {
+      "epoch": 0.8218111002921129,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00016551013715110932,
+      "loss": 0.8085,
+      "step": 1477
+    },
+    {
+      "epoch": 0.8223675059118097,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00016546601519088053,
+      "loss": 0.7552,
+      "step": 1478
+    },
+    {
+      "epoch": 0.8229239115315065,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00016542187091602345,
+      "loss": 0.7744,
+      "step": 1479
+    },
+    {
+      "epoch": 0.8234803171512032,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001653777043415851,
+      "loss": 0.8328,
+      "step": 1480
+    },
+    {
+      "epoch": 0.8240367227709,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00016533351548261998,
+      "loss": 0.8444,
+      "step": 1481
+    },
+    {
+      "epoch": 0.8245931283905967,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001652893043541902,
+      "loss": 0.8248,
+      "step": 1482
+    },
+    {
+      "epoch": 0.8251495340102935,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00016524507097136554,
+      "loss": 0.8688,
+      "step": 1483
+    },
+    {
+      "epoch": 0.8257059396299903,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001652008153492233,
+      "loss": 0.8503,
+      "step": 1484
+    },
+    {
+      "epoch": 0.826262345249687,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00016515653750284837,
+      "loss": 0.8137,
+      "step": 1485
+    },
+    {
+      "epoch": 0.8268187508693838,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001651122374473332,
+      "loss": 0.7943,
+      "step": 1486
+    },
+    {
+      "epoch": 0.8273751564890806,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00016506791519777786,
+      "loss": 0.7787,
+      "step": 1487
+    },
+    {
+      "epoch": 0.8279315621087773,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00016502357076928989,
+      "loss": 0.856,
+      "step": 1488
+    },
+    {
+      "epoch": 0.8284879677284741,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001649792041769845,
+      "loss": 0.7854,
+      "step": 1489
+    },
+    {
+      "epoch": 0.8290443733481708,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00016493481543598443,
+      "loss": 0.7728,
+      "step": 1490
+    },
+    {
+      "epoch": 0.8296007789678675,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00016489040456141988,
+      "loss": 0.7262,
+      "step": 1491
+    },
+    {
+      "epoch": 0.8301571845875644,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001648459715684287,
+      "loss": 0.721,
+      "step": 1492
+    },
+    {
+      "epoch": 0.8307135902072611,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00016480151647215626,
+      "loss": 0.8713,
+      "step": 1493
+    },
+    {
+      "epoch": 0.8312699958269578,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00016475703928775542,
+      "loss": 0.8769,
+      "step": 1494
+    },
+    {
+      "epoch": 0.8318264014466547,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00016471254003038656,
+      "loss": 0.8117,
+      "step": 1495
+    },
+    {
+      "epoch": 0.8323828070663514,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001646680187152177,
+      "loss": 0.8064,
+      "step": 1496
+    },
+    {
+      "epoch": 0.8329392126860481,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00016462347535742419,
+      "loss": 0.783,
+      "step": 1497
+    },
+    {
+      "epoch": 0.8334956183057449,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00016457890997218907,
+      "loss": 0.8462,
+      "step": 1498
+    },
+    {
+      "epoch": 0.8340520239254416,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00016453432257470282,
+      "loss": 0.8125,
+      "step": 1499
+    },
+    {
+      "epoch": 0.8346084295451384,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001644897131801634,
+      "loss": 0.812,
+      "step": 1500
+    },
+    {
+      "epoch": 0.8351648351648352,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00016444508180377628,
+      "loss": 0.722,
+      "step": 1501
+    },
+    {
+      "epoch": 0.8357212407845319,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00016440042846075446,
+      "loss": 0.8175,
+      "step": 1502
+    },
+    {
+      "epoch": 0.8362776464042286,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00016435575316631833,
+      "loss": 0.9485,
+      "step": 1503
+    },
+    {
+      "epoch": 0.8368340520239255,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001643110559356959,
+      "loss": 0.7406,
+      "step": 1504
+    },
+    {
+      "epoch": 0.8373904576436222,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00016426633678412263,
+      "loss": 0.7846,
+      "step": 1505
+    },
+    {
+      "epoch": 0.8379468632633189,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00016422159572684127,
+      "loss": 0.8702,
+      "step": 1506
+    },
+    {
+      "epoch": 0.8385032688830157,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00016417683277910228,
+      "loss": 0.8541,
+      "step": 1507
+    },
+    {
+      "epoch": 0.8390596745027125,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00016413204795616345,
+      "loss": 0.8826,
+      "step": 1508
+    },
+    {
+      "epoch": 0.8396160801224093,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00016408724127329005,
+      "loss": 0.7689,
+      "step": 1509
+    },
+    {
+      "epoch": 0.840172485742106,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001640424127457548,
+      "loss": 0.9144,
+      "step": 1510
+    },
+    {
+      "epoch": 0.8407288913618027,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001639975623888379,
+      "loss": 0.8451,
+      "step": 1511
+    },
+    {
+      "epoch": 0.8412852969814996,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00016395269021782696,
+      "loss": 0.8054,
+      "step": 1512
+    },
+    {
+      "epoch": 0.8418417026011963,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00016390779624801699,
+      "loss": 0.8637,
+      "step": 1513
+    },
+    {
+      "epoch": 0.842398108220893,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00016386288049471048,
+      "loss": 0.8536,
+      "step": 1514
+    },
+    {
+      "epoch": 0.8429545138405898,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00016381794297321735,
+      "loss": 0.8799,
+      "step": 1515
+    },
+    {
+      "epoch": 0.8435109194602866,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001637729836988549,
+      "loss": 0.814,
+      "step": 1516
+    },
+    {
+      "epoch": 0.8440673250799833,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00016372800268694792,
+      "loss": 0.7759,
+      "step": 1517
+    },
+    {
+      "epoch": 0.8446237306996801,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00016368299995282848,
+      "loss": 0.851,
+      "step": 1518
+    },
+    {
+      "epoch": 0.8451801363193768,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0001636379755118362,
+      "loss": 0.8151,
+      "step": 1519
+    },
+    {
+      "epoch": 0.8457365419390735,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00016359292937931797,
+      "loss": 0.7879,
+      "step": 1520
+    },
+    {
+      "epoch": 0.8462929475587704,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00016354786157062818,
+      "loss": 0.825,
+      "step": 1521
+    },
+    {
+      "epoch": 0.8468493531784671,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00016350277210112855,
+      "loss": 0.7414,
+      "step": 1522
+    },
+    {
+      "epoch": 0.8474057587981638,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001634576609861882,
+      "loss": 0.7679,
+      "step": 1523
+    },
+    {
+      "epoch": 0.8479621644178607,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00016341252824118358,
+      "loss": 0.8039,
+      "step": 1524
+    },
+    {
+      "epoch": 0.8485185700375574,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00016336737388149863,
+      "loss": 0.6974,
+      "step": 1525
+    },
+    {
+      "epoch": 0.8490749756572541,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001633221979225245,
+      "loss": 0.852,
+      "step": 1526
+    },
+    {
+      "epoch": 0.8496313812769509,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00016327700037965987,
+      "loss": 0.9108,
+      "step": 1527
+    },
+    {
+      "epoch": 0.8501877868966476,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00016323178126831066,
+      "loss": 0.8568,
+      "step": 1528
+    },
+    {
+      "epoch": 0.8507441925163444,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00016318654060389016,
+      "loss": 0.8606,
+      "step": 1529
+    },
+    {
+      "epoch": 0.8513005981360412,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00016314127840181902,
+      "loss": 0.7187,
+      "step": 1530
+    },
+    {
+      "epoch": 0.8518570037557379,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00016309599467752522,
+      "loss": 0.8346,
+      "step": 1531
+    },
+    {
+      "epoch": 0.8524134093754346,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00016305068944644417,
+      "loss": 0.9214,
+      "step": 1532
+    },
+    {
+      "epoch": 0.8529698149951315,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00016300536272401844,
+      "loss": 0.7717,
+      "step": 1533
+    },
+    {
+      "epoch": 0.8535262206148282,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00016296001452569808,
+      "loss": 0.8721,
+      "step": 1534
+    },
+    {
+      "epoch": 0.854082626234525,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001629146448669403,
+      "loss": 0.8538,
+      "step": 1535
+    },
+    {
+      "epoch": 0.8546390318542217,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00016286925376320986,
+      "loss": 0.9144,
+      "step": 1536
+    },
+    {
+      "epoch": 0.8551954374739185,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00016282384122997856,
+      "loss": 0.8563,
+      "step": 1537
+    },
+    {
+      "epoch": 0.8557518430936153,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001627784072827257,
+      "loss": 0.8297,
+      "step": 1538
+    },
+    {
+      "epoch": 0.856308248713312,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001627329519369378,
+      "loss": 0.7608,
+      "step": 1539
+    },
+    {
+      "epoch": 0.8568646543330087,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00016268747520810869,
+      "loss": 0.8589,
+      "step": 1540
+    },
+    {
+      "epoch": 0.8574210599527056,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00016264197711173946,
+      "loss": 0.8031,
+      "step": 1541
+    },
+    {
+      "epoch": 0.8579774655724023,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00016259645766333858,
+      "loss": 0.7716,
+      "step": 1542
+    },
+    {
+      "epoch": 0.858533871192099,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00016255091687842165,
+      "loss": 0.7499,
+      "step": 1543
+    },
+    {
+      "epoch": 0.8590902768117958,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00016250535477251168,
+      "loss": 0.8882,
+      "step": 1544
+    },
+    {
+      "epoch": 0.8596466824314926,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00016245977136113883,
+      "loss": 0.7705,
+      "step": 1545
+    },
+    {
+      "epoch": 0.8602030880511893,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00016241416665984063,
+      "loss": 0.8502,
+      "step": 1546
+    },
+    {
+      "epoch": 0.8607594936708861,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00016236854068416178,
+      "loss": 0.774,
+      "step": 1547
+    },
+    {
+      "epoch": 0.8613158992905828,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00016232289344965428,
+      "loss": 0.9005,
+      "step": 1548
+    },
+    {
+      "epoch": 0.8618723049102796,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0001622772249718774,
+      "loss": 0.8641,
+      "step": 1549
+    },
+    {
+      "epoch": 0.8624287105299764,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00016223153526639754,
+      "loss": 0.9873,
+      "step": 1550
+    },
+    {
+      "epoch": 0.8629851161496731,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00016218582434878847,
+      "loss": 0.8727,
+      "step": 1551
+    },
+    {
+      "epoch": 0.8635415217693698,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001621400922346311,
+      "loss": 0.802,
+      "step": 1552
+    },
+    {
+      "epoch": 0.8640979273890667,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00016209433893951366,
+      "loss": 0.8282,
+      "step": 1553
+    },
+    {
+      "epoch": 0.8646543330087634,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00016204856447903145,
+      "loss": 0.9093,
+      "step": 1554
+    },
+    {
+      "epoch": 0.8652107386284601,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0001620027688687871,
+      "loss": 0.9119,
+      "step": 1555
+    },
+    {
+      "epoch": 0.8657671442481569,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00016195695212439044,
+      "loss": 0.8263,
+      "step": 1556
+    },
+    {
+      "epoch": 0.8663235498678536,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00016191111426145846,
+      "loss": 0.9783,
+      "step": 1557
+    },
+    {
+      "epoch": 0.8668799554875504,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00016186525529561537,
+      "loss": 0.8366,
+      "step": 1558
+    },
+    {
+      "epoch": 0.8674363611072472,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00016181937524249263,
+      "loss": 0.7817,
+      "step": 1559
+    },
+    {
+      "epoch": 0.8679927667269439,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00016177347411772876,
+      "loss": 0.7799,
+      "step": 1560
+    },
+    {
+      "epoch": 0.8685491723466408,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00016172755193696956,
+      "loss": 0.8473,
+      "step": 1561
+    },
+    {
+      "epoch": 0.8691055779663375,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.000161681608715868,
+      "loss": 0.8397,
+      "step": 1562
+    },
+    {
+      "epoch": 0.8696619835860342,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00016163564447008416,
+      "loss": 0.8296,
+      "step": 1563
+    },
+    {
+      "epoch": 0.870218389205731,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00016158965921528537,
+      "loss": 0.7982,
+      "step": 1564
+    },
+    {
+      "epoch": 0.8707747948254277,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001615436529671461,
+      "loss": 0.7385,
+      "step": 1565
+    },
+    {
+      "epoch": 0.8713312004451245,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001614976257413479,
+      "loss": 0.8126,
+      "step": 1566
+    },
+    {
+      "epoch": 0.8718876060648213,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00016145157755357957,
+      "loss": 0.756,
+      "step": 1567
+    },
+    {
+      "epoch": 0.872444011684518,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00016140550841953698,
+      "loss": 0.8072,
+      "step": 1568
+    },
+    {
+      "epoch": 0.8730004173042147,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00016135941835492318,
+      "loss": 0.9268,
+      "step": 1569
+    },
+    {
+      "epoch": 0.8735568229239116,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001613133073754484,
+      "loss": 0.8406,
+      "step": 1570
+    },
+    {
+      "epoch": 0.8741132285436083,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0001612671754968299,
+      "loss": 0.8901,
+      "step": 1571
+    },
+    {
+      "epoch": 0.874669634163305,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00016122102273479207,
+      "loss": 0.8701,
+      "step": 1572
+    },
+    {
+      "epoch": 0.8752260397830018,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00016117484910506652,
+      "loss": 0.7697,
+      "step": 1573
+    },
+    {
+      "epoch": 0.8757824454026986,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0001611286546233919,
+      "loss": 0.9444,
+      "step": 1574
+    },
+    {
+      "epoch": 0.8763388510223953,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00016108243930551392,
+      "loss": 0.8555,
+      "step": 1575
+    },
+    {
+      "epoch": 0.8768952566420921,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00016103620316718553,
+      "loss": 0.7459,
+      "step": 1576
+    },
+    {
+      "epoch": 0.8774516622617888,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00016098994622416665,
+      "loss": 0.8395,
+      "step": 1577
+    },
+    {
+      "epoch": 0.8780080678814856,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00016094366849222437,
+      "loss": 0.799,
+      "step": 1578
+    },
+    {
+      "epoch": 0.8785644735011824,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001608973699871328,
+      "loss": 0.9022,
+      "step": 1579
+    },
+    {
+      "epoch": 0.8791208791208791,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00016085105072467318,
+      "loss": 0.7485,
+      "step": 1580
+    },
+    {
+      "epoch": 0.8796772847405758,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00016080471072063386,
+      "loss": 0.8096,
+      "step": 1581
+    },
+    {
+      "epoch": 0.8802336903602727,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00016075834999081012,
+      "loss": 0.8698,
+      "step": 1582
+    },
+    {
+      "epoch": 0.8807900959799694,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00016071196855100444,
+      "loss": 0.7957,
+      "step": 1583
+    },
+    {
+      "epoch": 0.8813465015996662,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001606655664170263,
+      "loss": 0.8163,
+      "step": 1584
+    },
+    {
+      "epoch": 0.8819029072193629,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00016061914360469227,
+      "loss": 0.8946,
+      "step": 1585
+    },
+    {
+      "epoch": 0.8824593128390597,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00016057270012982596,
+      "loss": 0.8084,
+      "step": 1586
+    },
+    {
+      "epoch": 0.8830157184587565,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00016052623600825797,
+      "loss": 0.8569,
+      "step": 1587
+    },
+    {
+      "epoch": 0.8835721240784532,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.000160479751255826,
+      "loss": 0.771,
+      "step": 1588
+    },
+    {
+      "epoch": 0.8841285296981499,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00016043324588837474,
+      "loss": 0.7227,
+      "step": 1589
+    },
+    {
+      "epoch": 0.8846849353178468,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00016038671992175594,
+      "loss": 0.8204,
+      "step": 1590
+    },
+    {
+      "epoch": 0.8852413409375435,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00016034017337182834,
+      "loss": 0.9365,
+      "step": 1591
+    },
+    {
+      "epoch": 0.8857977465572402,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00016029360625445773,
+      "loss": 0.7505,
+      "step": 1592
+    },
+    {
+      "epoch": 0.886354152176937,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00016024701858551694,
+      "loss": 0.8051,
+      "step": 1593
+    },
+    {
+      "epoch": 0.8869105577966337,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00016020041038088568,
+      "loss": 0.8519,
+      "step": 1594
+    },
+    {
+      "epoch": 0.8874669634163305,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00016015378165645076,
+      "loss": 0.9041,
+      "step": 1595
+    },
+    {
+      "epoch": 0.8880233690360273,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.000160107132428106,
+      "loss": 0.7955,
+      "step": 1596
+    },
+    {
+      "epoch": 0.888579774655724,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001600604627117521,
+      "loss": 0.7547,
+      "step": 1597
+    },
+    {
+      "epoch": 0.8891361802754207,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001600137725232969,
+      "loss": 0.8384,
+      "step": 1598
+    },
+    {
+      "epoch": 0.8896925858951176,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001599670618786551,
+      "loss": 0.7619,
+      "step": 1599
+    },
+    {
+      "epoch": 0.8902489915148143,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00015992033079374838,
+      "loss": 0.8609,
+      "step": 1600
+    },
+    {
+      "epoch": 0.890805397134511,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00015987357928450542,
+      "loss": 0.9125,
+      "step": 1601
+    },
+    {
+      "epoch": 0.8913618027542078,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0001598268073668619,
+      "loss": 0.8257,
+      "step": 1602
+    },
+    {
+      "epoch": 0.8919182083739046,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00015978001505676035,
+      "loss": 0.766,
+      "step": 1603
+    },
+    {
+      "epoch": 0.8924746139936013,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0001597332023701504,
+      "loss": 0.8824,
+      "step": 1604
+    },
+    {
+      "epoch": 0.8930310196132981,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00015968636932298843,
+      "loss": 0.8271,
+      "step": 1605
+    },
+    {
+      "epoch": 0.8935874252329948,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00015963951593123797,
+      "loss": 0.8136,
+      "step": 1606
+    },
+    {
+      "epoch": 0.8941438308526916,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001595926422108693,
+      "loss": 0.7455,
+      "step": 1607
+    },
+    {
+      "epoch": 0.8947002364723884,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00015954574817785973,
+      "loss": 0.8478,
+      "step": 1608
+    },
+    {
+      "epoch": 0.8952566420920851,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00015949883384819353,
+      "loss": 0.7687,
+      "step": 1609
+    },
+    {
+      "epoch": 0.8958130477117819,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00015945189923786176,
+      "loss": 0.8229,
+      "step": 1610
+    },
+    {
+      "epoch": 0.8963694533314787,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00015940494436286252,
+      "loss": 0.8361,
+      "step": 1611
+    },
+    {
+      "epoch": 0.8969258589511754,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00015935796923920078,
+      "loss": 0.7942,
+      "step": 1612
+    },
+    {
+      "epoch": 0.8974822645708722,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001593109738828883,
+      "loss": 0.79,
+      "step": 1613
+    },
+    {
+      "epoch": 0.8980386701905689,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00015926395830994394,
+      "loss": 0.8377,
+      "step": 1614
+    },
+    {
+      "epoch": 0.8985950758102657,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00015921692253639328,
+      "loss": 0.796,
+      "step": 1615
+    },
+    {
+      "epoch": 0.8991514814299625,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00015916986657826885,
+      "loss": 0.7916,
+      "step": 1616
+    },
+    {
+      "epoch": 0.8997078870496592,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001591227904516101,
+      "loss": 0.7697,
+      "step": 1617
+    },
+    {
+      "epoch": 0.9002642926693559,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00015907569417246333,
+      "loss": 0.8299,
+      "step": 1618
+    },
+    {
+      "epoch": 0.9008206982890528,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0001590285777568816,
+      "loss": 0.7568,
+      "step": 1619
+    },
+    {
+      "epoch": 0.9013771039087495,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00015898144122092505,
+      "loss": 0.8744,
+      "step": 1620
+    },
+    {
+      "epoch": 0.9019335095284462,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00015893428458066042,
+      "loss": 0.7819,
+      "step": 1621
+    },
+    {
+      "epoch": 0.902489915148143,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00015888710785216153,
+      "loss": 0.8183,
+      "step": 1622
+    },
+    {
+      "epoch": 0.9030463207678397,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001588399110515089,
+      "loss": 0.7904,
+      "step": 1623
+    },
+    {
+      "epoch": 0.9036027263875365,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00015879269419479002,
+      "loss": 0.8461,
+      "step": 1624
+    },
+    {
+      "epoch": 0.9041591320072333,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00015874545729809906,
+      "loss": 0.8309,
+      "step": 1625
+    },
+    {
+      "epoch": 0.90471553762693,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00015869820037753716,
+      "loss": 0.6568,
+      "step": 1626
+    },
+    {
+      "epoch": 0.9052719432466267,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0001586509234492122,
+      "loss": 0.8832,
+      "step": 1627
+    },
+    {
+      "epoch": 0.9058283488663236,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00015860362652923894,
+      "loss": 0.7721,
+      "step": 1628
+    },
+    {
+      "epoch": 0.9063847544860203,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001585563096337389,
+      "loss": 0.8272,
+      "step": 1629
+    },
+    {
+      "epoch": 0.906941160105717,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00015850897277884044,
+      "loss": 0.9139,
+      "step": 1630
+    },
+    {
+      "epoch": 0.9074975657254138,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001584616159806787,
+      "loss": 0.8265,
+      "step": 1631
+    },
+    {
+      "epoch": 0.9080539713451106,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00015841423925539564,
+      "loss": 0.7456,
+      "step": 1632
+    },
+    {
+      "epoch": 0.9086103769648074,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00015836684261914,
+      "loss": 0.7247,
+      "step": 1633
+    },
+    {
+      "epoch": 0.9091667825845041,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00015831942608806734,
+      "loss": 0.7684,
+      "step": 1634
+    },
+    {
+      "epoch": 0.9097231882042008,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00015827198967833993,
+      "loss": 0.7793,
+      "step": 1635
+    },
+    {
+      "epoch": 0.9102795938238977,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00015822453340612693,
+      "loss": 0.8197,
+      "step": 1636
+    },
+    {
+      "epoch": 0.9108359994435944,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0001581770572876041,
+      "loss": 0.8746,
+      "step": 1637
+    },
+    {
+      "epoch": 0.9113924050632911,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00015812956133895412,
+      "loss": 0.8732,
+      "step": 1638
+    },
+    {
+      "epoch": 0.9119488106829879,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00015808204557636642,
+      "loss": 0.8501,
+      "step": 1639
+    },
+    {
+      "epoch": 0.9125052163026847,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00015803451001603706,
+      "loss": 0.7846,
+      "step": 1640
+    },
+    {
+      "epoch": 0.9130616219223814,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00015798695467416897,
+      "loss": 0.871,
+      "step": 1641
+    },
+    {
+      "epoch": 0.9136180275420782,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00015793937956697175,
+      "loss": 0.7881,
+      "step": 1642
+    },
+    {
+      "epoch": 0.9141744331617749,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00015789178471066178,
+      "loss": 0.7453,
+      "step": 1643
+    },
+    {
+      "epoch": 0.9147308387814717,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00015784417012146217,
+      "loss": 0.8814,
+      "step": 1644
+    },
+    {
+      "epoch": 0.9152872444011685,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0001577965358156027,
+      "loss": 0.8565,
+      "step": 1645
+    },
+    {
+      "epoch": 0.9158436500208652,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00015774888180931993,
+      "loss": 0.8185,
+      "step": 1646
+    },
+    {
+      "epoch": 0.9164000556405619,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0001577012081188572,
+      "loss": 0.8507,
+      "step": 1647
+    },
+    {
+      "epoch": 0.9169564612602588,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00015765351476046435,
+      "loss": 0.721,
+      "step": 1648
+    },
+    {
+      "epoch": 0.9175128668799555,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00015760580175039812,
+      "loss": 0.8975,
+      "step": 1649
+    },
+    {
+      "epoch": 0.9180692724996522,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00015755806910492187,
+      "loss": 0.7393,
+      "step": 1650
+    },
+    {
+      "epoch": 0.918625678119349,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00015751031684030568,
+      "loss": 0.8464,
+      "step": 1651
+    },
+    {
+      "epoch": 0.9191820837390458,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001574625449728263,
+      "loss": 0.7609,
+      "step": 1652
+    },
+    {
+      "epoch": 0.9197384893587425,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00015741475351876717,
+      "loss": 0.8405,
+      "step": 1653
+    },
+    {
+      "epoch": 0.9202948949784393,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00015736694249441832,
+      "loss": 0.8624,
+      "step": 1654
+    },
+    {
+      "epoch": 0.920851300598136,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00015731911191607661,
+      "loss": 0.7953,
+      "step": 1655
+    },
+    {
+      "epoch": 0.9214077062178327,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001572712618000455,
+      "loss": 0.7722,
+      "step": 1656
+    },
+    {
+      "epoch": 0.9219641118375296,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00015722339216263503,
+      "loss": 0.8173,
+      "step": 1657
+    },
+    {
+      "epoch": 0.9225205174572263,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00015717550302016197,
+      "loss": 0.7558,
+      "step": 1658
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00015712759438894978,
+      "loss": 0.83,
+      "step": 1659
+    },
+    {
+      "epoch": 0.9236333286966198,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00015707966628532847,
+      "loss": 0.8345,
+      "step": 1660
+    },
+    {
+      "epoch": 0.9241897343163166,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00015703171872563472,
+      "loss": 0.7731,
+      "step": 1661
+    },
+    {
+      "epoch": 0.9247461399360134,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00015698375172621188,
+      "loss": 0.7117,
+      "step": 1662
+    },
+    {
+      "epoch": 0.9253025455557101,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00015693576530340987,
+      "loss": 0.7655,
+      "step": 1663
+    },
+    {
+      "epoch": 0.9258589511754068,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00015688775947358524,
+      "loss": 0.8557,
+      "step": 1664
+    },
+    {
+      "epoch": 0.9264153567951037,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0001568397342531012,
+      "loss": 0.7155,
+      "step": 1665
+    },
+    {
+      "epoch": 0.9269717624148004,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001567916896583275,
+      "loss": 0.7699,
+      "step": 1666
+    },
+    {
+      "epoch": 0.9275281680344971,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00015674362570564063,
+      "loss": 0.9217,
+      "step": 1667
+    },
+    {
+      "epoch": 0.928084573654194,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001566955424114235,
+      "loss": 0.8761,
+      "step": 1668
+    },
+    {
+      "epoch": 0.9286409792738907,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00015664743979206574,
+      "loss": 0.854,
+      "step": 1669
+    },
+    {
+      "epoch": 0.9291973848935874,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00015659931786396345,
+      "loss": 0.9302,
+      "step": 1670
+    },
+    {
+      "epoch": 0.9297537905132842,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00015655117664351945,
+      "loss": 0.6646,
+      "step": 1671
+    },
+    {
+      "epoch": 0.9303101961329809,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00015650301614714303,
+      "loss": 0.6723,
+      "step": 1672
+    },
+    {
+      "epoch": 0.9308666017526777,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00015645483639125014,
+      "loss": 0.8624,
+      "step": 1673
+    },
+    {
+      "epoch": 0.9314230073723745,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0001564066373922632,
+      "loss": 0.8976,
+      "step": 1674
+    },
+    {
+      "epoch": 0.9319794129920712,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00015635841916661127,
+      "loss": 0.8596,
+      "step": 1675
+    },
+    {
+      "epoch": 0.9325358186117679,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00015631018173072988,
+      "loss": 0.9422,
+      "step": 1676
+    },
+    {
+      "epoch": 0.9330922242314648,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00015626192510106123,
+      "loss": 0.7931,
+      "step": 1677
+    },
+    {
+      "epoch": 0.9336486298511615,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001562136492940539,
+      "loss": 0.8655,
+      "step": 1678
+    },
+    {
+      "epoch": 0.9342050354708582,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00015616535432616316,
+      "loss": 0.7388,
+      "step": 1679
+    },
+    {
+      "epoch": 0.934761441090555,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001561170402138507,
+      "loss": 0.8805,
+      "step": 1680
+    },
+    {
+      "epoch": 0.9353178467102518,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00015606870697358483,
+      "loss": 0.7576,
+      "step": 1681
+    },
+    {
+      "epoch": 0.9358742523299486,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0001560203546218403,
+      "loss": 0.8602,
+      "step": 1682
+    },
+    {
+      "epoch": 0.9364306579496453,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00015597198317509838,
+      "loss": 0.8555,
+      "step": 1683
+    },
+    {
+      "epoch": 0.936987063569342,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00015592359264984693,
+      "loss": 0.8234,
+      "step": 1684
+    },
+    {
+      "epoch": 0.9375434691890389,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00015587518306258024,
+      "loss": 0.8587,
+      "step": 1685
+    },
+    {
+      "epoch": 0.9380998748087356,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001558267544297991,
+      "loss": 0.7904,
+      "step": 1686
+    },
+    {
+      "epoch": 0.9386562804284323,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00015577830676801079,
+      "loss": 0.8887,
+      "step": 1687
+    },
+    {
+      "epoch": 0.9392126860481291,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00015572984009372916,
+      "loss": 0.8488,
+      "step": 1688
+    },
+    {
+      "epoch": 0.9397690916678259,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0001556813544234744,
+      "loss": 0.7988,
+      "step": 1689
+    },
+    {
+      "epoch": 0.9403254972875226,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001556328497737733,
+      "loss": 0.8036,
+      "step": 1690
+    },
+    {
+      "epoch": 0.9408819029072194,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00015558432616115901,
+      "loss": 0.8355,
+      "step": 1691
+    },
+    {
+      "epoch": 0.9414383085269161,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00015553578360217126,
+      "loss": 0.8544,
+      "step": 1692
+    },
+    {
+      "epoch": 0.9419947141466128,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0001554872221133562,
+      "loss": 0.8712,
+      "step": 1693
+    },
+    {
+      "epoch": 0.9425511197663097,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00015543864171126633,
+      "loss": 0.696,
+      "step": 1694
+    },
+    {
+      "epoch": 0.9431075253860064,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001553900424124607,
+      "loss": 0.8413,
+      "step": 1695
+    },
+    {
+      "epoch": 0.9436639310057031,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00015534142423350483,
+      "loss": 0.8712,
+      "step": 1696
+    },
+    {
+      "epoch": 0.9442203366254,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00015529278719097057,
+      "loss": 0.8315,
+      "step": 1697
+    },
+    {
+      "epoch": 0.9447767422450967,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00015524413130143632,
+      "loss": 0.8684,
+      "step": 1698
+    },
+    {
+      "epoch": 0.9453331478647934,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00015519545658148674,
+      "loss": 1.8628,
+      "step": 1699
+    },
+    {
+      "epoch": 0.9458895534844902,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001551467630477131,
+      "loss": 0.8537,
+      "step": 1700
+    },
+    {
+      "epoch": 0.9464459591041869,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00015509805071671292,
+      "loss": 0.8349,
+      "step": 1701
+    },
+    {
+      "epoch": 0.9470023647238837,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001550493196050903,
+      "loss": 0.7872,
+      "step": 1702
+    },
+    {
+      "epoch": 0.9475587703435805,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0001550005697294555,
+      "loss": 0.9317,
+      "step": 1703
+    },
+    {
+      "epoch": 0.9481151759632772,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00015495180110642543,
+      "loss": 0.909,
+      "step": 1704
+    },
+    {
+      "epoch": 0.9486715815829739,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001549030137526232,
+      "loss": 0.8417,
+      "step": 1705
+    },
+    {
+      "epoch": 0.9492279872026708,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00015485420768467843,
+      "loss": 0.7273,
+      "step": 1706
+    },
+    {
+      "epoch": 0.9497843928223675,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00015480538291922703,
+      "loss": 0.8405,
+      "step": 1707
+    },
+    {
+      "epoch": 0.9503407984420643,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00015475653947291136,
+      "loss": 0.9213,
+      "step": 1708
+    },
+    {
+      "epoch": 0.950897204061761,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00015470767736238007,
+      "loss": 0.8166,
+      "step": 1709
+    },
+    {
+      "epoch": 0.9514536096814578,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001546587966042883,
+      "loss": 0.8303,
+      "step": 1710
+    },
+    {
+      "epoch": 0.9520100153011546,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001546098972152973,
+      "loss": 0.8412,
+      "step": 1711
+    },
+    {
+      "epoch": 0.9525664209208513,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00015456097921207496,
+      "loss": 0.7055,
+      "step": 1712
+    },
+    {
+      "epoch": 0.953122826540548,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00015451204261129534,
+      "loss": 0.925,
+      "step": 1713
+    },
+    {
+      "epoch": 0.9536792321602449,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0001544630874296389,
+      "loss": 0.8671,
+      "step": 1714
+    },
+    {
+      "epoch": 0.9542356377799416,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00015441411368379235,
+      "loss": 0.7963,
+      "step": 1715
+    },
+    {
+      "epoch": 0.9547920433996383,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00015436512139044886,
+      "loss": 0.7544,
+      "step": 1716
+    },
+    {
+      "epoch": 0.9553484490193351,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00015431611056630784,
+      "loss": 0.7664,
+      "step": 1717
+    },
+    {
+      "epoch": 0.9559048546390319,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.000154267081228075,
+      "loss": 0.8951,
+      "step": 1718
+    },
+    {
+      "epoch": 0.9564612602587286,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00015421803339246245,
+      "loss": 0.8817,
+      "step": 1719
+    },
+    {
+      "epoch": 0.9570176658784254,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00015416896707618852,
+      "loss": 0.8008,
+      "step": 1720
+    },
+    {
+      "epoch": 0.9575740714981221,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00015411988229597782,
+      "loss": 0.9524,
+      "step": 1721
+    },
+    {
+      "epoch": 0.9581304771178188,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00015407077906856137,
+      "loss": 0.7774,
+      "step": 1722
+    },
+    {
+      "epoch": 0.9586868827375157,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001540216574106764,
+      "loss": 0.7962,
+      "step": 1723
+    },
+    {
+      "epoch": 0.9592432883572124,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00015397251733906636,
+      "loss": 0.8501,
+      "step": 1724
+    },
+    {
+      "epoch": 0.9597996939769091,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00015392335887048114,
+      "loss": 0.7948,
+      "step": 1725
+    },
+    {
+      "epoch": 0.960356099596606,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00015387418202167673,
+      "loss": 0.8632,
+      "step": 1726
+    },
+    {
+      "epoch": 0.9609125052163027,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0001538249868094155,
+      "loss": 0.9704,
+      "step": 1727
+    },
+    {
+      "epoch": 0.9614689108359994,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00015377577325046606,
+      "loss": 0.7435,
+      "step": 1728
+    },
+    {
+      "epoch": 0.9620253164556962,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00015372654136160322,
+      "loss": 0.8626,
+      "step": 1729
+    },
+    {
+      "epoch": 0.9625817220753929,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0001536772911596081,
+      "loss": 0.7785,
+      "step": 1730
+    },
+    {
+      "epoch": 0.9631381276950898,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.000153628022661268,
+      "loss": 0.7472,
+      "step": 1731
+    },
+    {
+      "epoch": 0.9636945333147865,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001535787358833765,
+      "loss": 0.854,
+      "step": 1732
+    },
+    {
+      "epoch": 0.9642509389344832,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00015352943084273344,
+      "loss": 0.8813,
+      "step": 1733
+    },
+    {
+      "epoch": 0.96480734455418,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0001534801075561448,
+      "loss": 0.9251,
+      "step": 1734
+    },
+    {
+      "epoch": 0.9653637501738768,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00015343076604042285,
+      "loss": 0.7338,
+      "step": 1735
+    },
+    {
+      "epoch": 0.9659201557935735,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00015338140631238604,
+      "loss": 0.761,
+      "step": 1736
+    },
+    {
+      "epoch": 0.9664765614132703,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00015333202838885908,
+      "loss": 0.8448,
+      "step": 1737
+    },
+    {
+      "epoch": 0.967032967032967,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001532826322866728,
+      "loss": 0.7478,
+      "step": 1738
+    },
+    {
+      "epoch": 0.9675893726526638,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00015323321802266428,
+      "loss": 0.8116,
+      "step": 1739
+    },
+    {
+      "epoch": 0.9681457782723606,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00015318378561367672,
+      "loss": 0.8851,
+      "step": 1740
+    },
+    {
+      "epoch": 0.9687021838920573,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00015313433507655968,
+      "loss": 0.8524,
+      "step": 1741
+    },
+    {
+      "epoch": 0.969258589511754,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00015308486642816869,
+      "loss": 0.7722,
+      "step": 1742
+    },
+    {
+      "epoch": 0.9698149951314509,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00015303537968536557,
+      "loss": 0.9203,
+      "step": 1743
+    },
+    {
+      "epoch": 0.9703714007511476,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00015298587486501828,
+      "loss": 0.8607,
+      "step": 1744
+    },
+    {
+      "epoch": 0.9709278063708443,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00015293635198400095,
+      "loss": 0.7585,
+      "step": 1745
+    },
+    {
+      "epoch": 0.9714842119905411,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00015288681105919385,
+      "loss": 0.7475,
+      "step": 1746
+    },
+    {
+      "epoch": 0.9720406176102379,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001528372521074834,
+      "loss": 0.7884,
+      "step": 1747
+    },
+    {
+      "epoch": 0.9725970232299346,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00015278767514576223,
+      "loss": 0.7938,
+      "step": 1748
+    },
+    {
+      "epoch": 0.9731534288496314,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00015273808019092896,
+      "loss": 0.8201,
+      "step": 1749
+    },
+    {
+      "epoch": 0.9737098344693281,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00015268846725988855,
+      "loss": 0.7848,
+      "step": 1750
+    },
+    {
+      "epoch": 0.9742662400890248,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001526388363695519,
+      "loss": 0.8647,
+      "step": 1751
+    },
+    {
+      "epoch": 0.9748226457087217,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00015258918753683612,
+      "loss": 0.8837,
+      "step": 1752
+    },
+    {
+      "epoch": 0.9753790513284184,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00015253952077866437,
+      "loss": 0.72,
+      "step": 1753
+    },
+    {
+      "epoch": 0.9759354569481151,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0001524898361119661,
+      "loss": 0.9329,
+      "step": 1754
+    },
+    {
+      "epoch": 0.976491862567812,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00015244013355367663,
+      "loss": 0.7853,
+      "step": 1755
+    },
+    {
+      "epoch": 0.9770482681875087,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00015239041312073754,
+      "loss": 0.7884,
+      "step": 1756
+    },
+    {
+      "epoch": 0.9776046738072055,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001523406748300964,
+      "loss": 0.8844,
+      "step": 1757
+    },
+    {
+      "epoch": 0.9781610794269022,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00015229091869870697,
+      "loss": 0.8389,
+      "step": 1758
+    },
+    {
+      "epoch": 0.978717485046599,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00015224114474352897,
+      "loss": 0.9325,
+      "step": 1759
+    },
+    {
+      "epoch": 0.9792738906662958,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00015219135298152835,
+      "loss": 0.8748,
+      "step": 1760
+    },
+    {
+      "epoch": 0.9798302962859925,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00015214154342967691,
+      "loss": 0.8733,
+      "step": 1761
+    },
+    {
+      "epoch": 0.9803867019056892,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00015209171610495277,
+      "loss": 0.8178,
+      "step": 1762
+    },
+    {
+      "epoch": 0.980943107525386,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001520418710243399,
+      "loss": 0.7557,
+      "step": 1763
+    },
+    {
+      "epoch": 0.9814995131450828,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001519920082048285,
+      "loss": 0.8423,
+      "step": 1764
+    },
+    {
+      "epoch": 0.9820559187647795,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00015194212766341458,
+      "loss": 0.8604,
+      "step": 1765
+    },
+    {
+      "epoch": 0.9826123243844763,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00015189222941710043,
+      "loss": 0.849,
+      "step": 1766
+    },
+    {
+      "epoch": 0.983168730004173,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001518423134828943,
+      "loss": 0.8249,
+      "step": 1767
+    },
+    {
+      "epoch": 0.9837251356238698,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00015179237987781037,
+      "loss": 0.8117,
+      "step": 1768
+    },
+    {
+      "epoch": 0.9842815412435666,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00015174242861886892,
+      "loss": 0.7628,
+      "step": 1769
+    },
+    {
+      "epoch": 0.9848379468632633,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00015169245972309633,
+      "loss": 0.7333,
+      "step": 1770
+    },
+    {
+      "epoch": 0.98539435248296,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00015164247320752488,
+      "loss": 0.8816,
+      "step": 1771
+    },
+    {
+      "epoch": 0.9859507581026569,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001515924690891928,
+      "loss": 0.814,
+      "step": 1772
+    },
+    {
+      "epoch": 0.9865071637223536,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00015154244738514448,
+      "loss": 0.7076,
+      "step": 1773
+    },
+    {
+      "epoch": 0.9870635693420503,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00015149240811243022,
+      "loss": 0.8046,
+      "step": 1774
+    },
+    {
+      "epoch": 0.9876199749617471,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001514423512881063,
+      "loss": 0.7358,
+      "step": 1775
+    },
+    {
+      "epoch": 0.9881763805814439,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.000151392276929235,
+      "loss": 0.7509,
+      "step": 1776
+    },
+    {
+      "epoch": 0.9887327862011406,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00015134218505288458,
+      "loss": 0.719,
+      "step": 1777
+    },
+    {
+      "epoch": 0.9892891918208374,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00015129207567612932,
+      "loss": 0.945,
+      "step": 1778
+    },
+    {
+      "epoch": 0.9898455974405341,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00015124194881604929,
+      "loss": 0.8144,
+      "step": 1779
+    },
+    {
+      "epoch": 0.990402003060231,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00015119180448973073,
+      "loss": 0.7926,
+      "step": 1780
+    },
+    {
+      "epoch": 0.9909584086799277,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001511416427142658,
+      "loss": 0.9045,
+      "step": 1781
+    },
+    {
+      "epoch": 0.9915148142996244,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00015109146350675236,
+      "loss": 0.7689,
+      "step": 1782
+    },
+    {
+      "epoch": 0.9920712199193212,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00015104126688429462,
+      "loss": 0.8457,
+      "step": 1783
+    },
+    {
+      "epoch": 0.992627625539018,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00015099105286400238,
+      "loss": 0.8154,
+      "step": 1784
+    },
+    {
+      "epoch": 0.9931840311587147,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00015094082146299153,
+      "loss": 0.7393,
+      "step": 1785
+    },
+    {
+      "epoch": 0.9937404367784115,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001508905726983839,
+      "loss": 0.7191,
+      "step": 1786
+    },
+    {
+      "epoch": 0.9942968423981082,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001508403065873071,
+      "loss": 0.7611,
+      "step": 1787
+    },
+    {
+      "epoch": 0.994853248017805,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00015079002314689484,
+      "loss": 0.7972,
+      "step": 1788
+    },
+    {
+      "epoch": 0.9954096536375018,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001507397223942866,
+      "loss": 0.7863,
+      "step": 1789
+    },
+    {
+      "epoch": 0.9959660592571985,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00015068940434662776,
+      "loss": 0.801,
+      "step": 1790
+    },
+    {
+      "epoch": 0.9965224648768952,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00015063906902106975,
+      "loss": 0.7198,
+      "step": 1791
+    },
+    {
+      "epoch": 0.997078870496592,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0001505887164347697,
+      "loss": 0.6595,
+      "step": 1792
+    },
+    {
+      "epoch": 0.9976352761162888,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00015053834660489073,
+      "loss": 0.8524,
+      "step": 1793
+    },
+    {
+      "epoch": 0.9981916817359855,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00015048795954860172,
+      "loss": 0.8229,
+      "step": 1794
+    },
+    {
+      "epoch": 0.9987480873556823,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00015043755528307767,
+      "loss": 0.8114,
+      "step": 1795
+    },
+    {
+      "epoch": 0.999304492975379,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001503871338254992,
+      "loss": 0.868,
+      "step": 1796
+    },
+    {
+      "epoch": 0.9998608985950758,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00015033669519305286,
+      "loss": 0.771,
+      "step": 1797
+    },
+    {
+      "epoch": 1.0004173042147726,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001502862394029311,
+      "loss": 0.7104,
+      "step": 1798
+    },
+    {
+      "epoch": 1.0004173042147726,
+      "eval_loss": 0.7584885954856873,
+      "eval_runtime": 1763.8079,
+      "eval_samples_per_second": 2.66,
+      "eval_steps_per_second": 1.33,
+      "step": 1798
+    },
+    {
+      "epoch": 1.0009737098344693,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00015023576647233216,
+      "loss": 0.8586,
+      "step": 1799
+    },
+    {
+      "epoch": 1.001530115454166,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00015018527641846022,
+      "loss": 0.7438,
+      "step": 1800
+    },
+    {
+      "epoch": 1.0020865210738628,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001501347692585252,
+      "loss": 0.8,
+      "step": 1801
+    },
+    {
+      "epoch": 1.0026429266935597,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00015008424500974284,
+      "loss": 0.783,
+      "step": 1802
+    },
+    {
+      "epoch": 1.0031993323132564,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00015003370368933477,
+      "loss": 0.7659,
+      "step": 1803
+    },
+    {
+      "epoch": 1.0037557379329531,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001499831453145284,
+      "loss": 0.7565,
+      "step": 1804
+    },
+    {
+      "epoch": 1.0043121435526499,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00014993256990255703,
+      "loss": 0.8467,
+      "step": 1805
+    },
+    {
+      "epoch": 1.0048685491723466,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001498819774706596,
+      "loss": 0.8653,
+      "step": 1806
+    },
+    {
+      "epoch": 1.0054249547920433,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00014983136803608103,
+      "loss": 0.8532,
+      "step": 1807
+    },
+    {
+      "epoch": 1.0059813604117402,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00014978074161607196,
+      "loss": 0.8039,
+      "step": 1808
+    },
+    {
+      "epoch": 1.006537766031437,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00014973009822788876,
+      "loss": 0.8743,
+      "step": 1809
+    },
+    {
+      "epoch": 1.0070941716511337,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00014967943788879372,
+      "loss": 0.8014,
+      "step": 1810
+    },
+    {
+      "epoch": 1.0076505772708304,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001496287606160548,
+      "loss": 0.7635,
+      "step": 1811
+    },
+    {
+      "epoch": 1.0082069828905271,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00014957806642694577,
+      "loss": 0.7415,
+      "step": 1812
+    },
+    {
+      "epoch": 1.0087633885102238,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00014952735533874613,
+      "loss": 0.8067,
+      "step": 1813
+    },
+    {
+      "epoch": 1.0093197941299208,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00014947662736874116,
+      "loss": 0.8203,
+      "step": 1814
+    },
+    {
+      "epoch": 1.0098761997496175,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00014942588253422204,
+      "loss": 0.7599,
+      "step": 1815
+    },
+    {
+      "epoch": 1.0104326053693142,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001493751208524854,
+      "loss": 0.75,
+      "step": 1816
+    },
+    {
+      "epoch": 1.010989010989011,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00014932434234083384,
+      "loss": 0.7379,
+      "step": 1817
+    },
+    {
+      "epoch": 1.0115454166087077,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00014927354701657562,
+      "loss": 0.869,
+      "step": 1818
+    },
+    {
+      "epoch": 1.0121018222284046,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00014922273489702473,
+      "loss": 0.8448,
+      "step": 1819
+    },
+    {
+      "epoch": 1.0126582278481013,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00014917190599950098,
+      "loss": 0.6718,
+      "step": 1820
+    },
+    {
+      "epoch": 1.013214633467798,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00014912106034132972,
+      "loss": 0.8731,
+      "step": 1821
+    },
+    {
+      "epoch": 1.0137710390874948,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00014907019793984215,
+      "loss": 0.8285,
+      "step": 1822
+    },
+    {
+      "epoch": 1.0143274447071915,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001490193188123751,
+      "loss": 0.8005,
+      "step": 1823
+    },
+    {
+      "epoch": 1.0148838503268882,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001489684229762712,
+      "loss": 0.8463,
+      "step": 1824
+    },
+    {
+      "epoch": 1.0154402559465852,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0001489175104488787,
+      "loss": 0.7568,
+      "step": 1825
+    },
+    {
+      "epoch": 1.0159966615662819,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001488665812475515,
+      "loss": 0.8092,
+      "step": 1826
+    },
+    {
+      "epoch": 1.0165530671859786,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001488156353896493,
+      "loss": 0.7904,
+      "step": 1827
+    },
+    {
+      "epoch": 1.0171094728056753,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00014876467289253736,
+      "loss": 0.7924,
+      "step": 1828
+    },
+    {
+      "epoch": 1.0004173042147726,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00014871369377358673,
+      "loss": 0.765,
+      "step": 1829
+    },
+    {
+      "epoch": 1.0009737098344693,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.000148662698050174,
+      "loss": 0.6763,
+      "step": 1830
+    },
+    {
+      "epoch": 1.001530115454166,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001486116857396815,
+      "loss": 0.7167,
+      "step": 1831
+    },
+    {
+      "epoch": 1.0020865210738628,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001485606568594972,
+      "loss": 0.7272,
+      "step": 1832
+    },
+    {
+      "epoch": 1.0026429266935597,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001485096114270147,
+      "loss": 0.646,
+      "step": 1833
+    },
+    {
+      "epoch": 1.0031993323132564,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00014845854945963328,
+      "loss": 0.7645,
+      "step": 1834
+    },
+    {
+      "epoch": 1.0037557379329531,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001484074709747578,
+      "loss": 0.6898,
+      "step": 1835
+    },
+    {
+      "epoch": 1.0043121435526499,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001483563759897988,
+      "loss": 0.7485,
+      "step": 1836
+    },
+    {
+      "epoch": 1.0048685491723466,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001483052645221724,
+      "loss": 0.7517,
+      "step": 1837
+    },
+    {
+      "epoch": 1.0054249547920433,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001482541365893004,
+      "loss": 0.8237,
+      "step": 1838
+    },
+    {
+      "epoch": 1.0059813604117402,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00014820299220861014,
+      "loss": 0.7087,
+      "step": 1839
+    },
+    {
+      "epoch": 1.006537766031437,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00014815183139753462,
+      "loss": 0.7483,
+      "step": 1840
+    },
+    {
+      "epoch": 1.0070941716511337,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00014810065417351245,
+      "loss": 0.7024,
+      "step": 1841
+    },
+    {
+      "epoch": 1.0076505772708304,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00014804946055398777,
+      "loss": 0.747,
+      "step": 1842
+    },
+    {
+      "epoch": 1.0082069828905271,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00014799825055641032,
+      "loss": 0.7064,
+      "step": 1843
+    },
+    {
+      "epoch": 1.0087633885102238,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0001479470241982355,
+      "loss": 0.7605,
+      "step": 1844
+    },
+    {
+      "epoch": 1.0093197941299208,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00014789578149692424,
+      "loss": 0.7385,
+      "step": 1845
+    },
+    {
+      "epoch": 1.0098761997496175,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.000147844522469943,
+      "loss": 0.7465,
+      "step": 1846
+    },
+    {
+      "epoch": 1.0104326053693142,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001477932471347638,
+      "loss": 0.6345,
+      "step": 1847
+    },
+    {
+      "epoch": 1.010989010989011,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00014774195550886437,
+      "loss": 0.7599,
+      "step": 1848
+    },
+    {
+      "epoch": 1.0115454166087077,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001476906476097278,
+      "loss": 0.7338,
+      "step": 1849
+    },
+    {
+      "epoch": 1.0121018222284046,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00014763932345484283,
+      "loss": 0.7355,
+      "step": 1850
+    },
+    {
+      "epoch": 1.0126582278481013,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00014758798306170368,
+      "loss": 0.6857,
+      "step": 1851
+    },
+    {
+      "epoch": 1.013214633467798,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00014753662644781022,
+      "loss": 0.7606,
+      "step": 1852
+    },
+    {
+      "epoch": 1.0137710390874948,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00014748525363066773,
+      "loss": 0.7229,
+      "step": 1853
+    },
+    {
+      "epoch": 1.0143274447071915,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00014743386462778702,
+      "loss": 0.6471,
+      "step": 1854
+    },
+    {
+      "epoch": 1.0148838503268882,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00014738245945668455,
+      "loss": 0.6658,
+      "step": 1855
+    },
+    {
+      "epoch": 1.0154402559465852,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00014733103813488208,
+      "loss": 0.6872,
+      "step": 1856
+    },
+    {
+      "epoch": 1.0159966615662819,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001472796006799071,
+      "loss": 0.7166,
+      "step": 1857
+    },
+    {
+      "epoch": 1.0165530671859786,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001472281471092924,
+      "loss": 0.6374,
+      "step": 1858
+    },
+    {
+      "epoch": 1.0171094728056753,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00014717667744057643,
+      "loss": 0.6689,
+      "step": 1859
+    },
+    {
+      "epoch": 1.017665878425372,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00014712519169130298,
+      "loss": 0.7737,
+      "step": 1860
+    },
+    {
+      "epoch": 1.0182222840450688,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001470736898790214,
+      "loss": 0.6229,
+      "step": 1861
+    },
+    {
+      "epoch": 1.0187786896647657,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00014702217202128655,
+      "loss": 0.7961,
+      "step": 1862
+    },
+    {
+      "epoch": 1.0193350952844624,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00014697063813565866,
+      "loss": 0.8566,
+      "step": 1863
+    },
+    {
+      "epoch": 1.0198915009041591,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001469190882397035,
+      "loss": 0.6613,
+      "step": 1864
+    },
+    {
+      "epoch": 1.0204479065238559,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00014686752235099228,
+      "loss": 0.6908,
+      "step": 1865
+    },
+    {
+      "epoch": 1.0210043121435526,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00014681594048710165,
+      "loss": 0.7401,
+      "step": 1866
+    },
+    {
+      "epoch": 1.0215607177632493,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00014676434266561373,
+      "loss": 0.6622,
+      "step": 1867
+    },
+    {
+      "epoch": 1.0221171233829462,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00014671272890411603,
+      "loss": 0.7471,
+      "step": 1868
+    },
+    {
+      "epoch": 1.022673529002643,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00014666109922020155,
+      "loss": 0.6932,
+      "step": 1869
+    },
+    {
+      "epoch": 1.0232299346223397,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001466094536314687,
+      "loss": 0.7359,
+      "step": 1870
+    },
+    {
+      "epoch": 1.0237863402420364,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00014655779215552126,
+      "loss": 0.8022,
+      "step": 1871
+    },
+    {
+      "epoch": 1.0243427458617331,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00014650611480996844,
+      "loss": 0.6607,
+      "step": 1872
+    },
+    {
+      "epoch": 1.02489915148143,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00014645442161242503,
+      "loss": 0.6856,
+      "step": 1873
+    },
+    {
+      "epoch": 1.0254555571011268,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00014640271258051094,
+      "loss": 0.7469,
+      "step": 1874
+    },
+    {
+      "epoch": 1.0260119627208235,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001463509877318517,
+      "loss": 0.7641,
+      "step": 1875
+    },
+    {
+      "epoch": 1.0265683683405202,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00014629924708407807,
+      "loss": 0.6705,
+      "step": 1876
+    },
+    {
+      "epoch": 1.027124773960217,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00014624749065482635,
+      "loss": 0.6511,
+      "step": 1877
+    },
+    {
+      "epoch": 1.0276811795799137,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001461957184617381,
+      "loss": 0.7156,
+      "step": 1878
+    },
+    {
+      "epoch": 1.0282375851996106,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00014614393052246033,
+      "loss": 0.7368,
+      "step": 1879
+    },
+    {
+      "epoch": 1.0287939908193073,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00014609212685464531,
+      "loss": 0.7679,
+      "step": 1880
+    },
+    {
+      "epoch": 1.029350396439004,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00014604030747595082,
+      "loss": 0.7735,
+      "step": 1881
+    },
+    {
+      "epoch": 1.0299068020587008,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00014598847240403996,
+      "loss": 0.7116,
+      "step": 1882
+    },
+    {
+      "epoch": 1.0304632076783975,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.000145936621656581,
+      "loss": 0.6661,
+      "step": 1883
+    },
+    {
+      "epoch": 1.0310196132980942,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0001458847552512478,
+      "loss": 0.8101,
+      "step": 1884
+    },
+    {
+      "epoch": 1.0315760189177912,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001458328732057194,
+      "loss": 0.669,
+      "step": 1885
+    },
+    {
+      "epoch": 1.0321324245374879,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00014578097553768026,
+      "loss": 0.7535,
+      "step": 1886
+    },
+    {
+      "epoch": 1.0326888301571846,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00014572906226482013,
+      "loss": 0.7153,
+      "step": 1887
+    },
+    {
+      "epoch": 1.0332452357768813,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.000145677133404834,
+      "loss": 0.7,
+      "step": 1888
+    },
+    {
+      "epoch": 1.033801641396578,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00014562518897542236,
+      "loss": 0.642,
+      "step": 1889
+    },
+    {
+      "epoch": 1.0343580470162748,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001455732289942908,
+      "loss": 0.695,
+      "step": 1890
+    },
+    {
+      "epoch": 1.0349144526359717,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00014552125347915037,
+      "loss": 0.7717,
+      "step": 1891
+    },
+    {
+      "epoch": 1.0354708582556684,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00014546926244771727,
+      "loss": 0.6881,
+      "step": 1892
+    },
+    {
+      "epoch": 1.0360272638753651,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001454172559177131,
+      "loss": 0.7123,
+      "step": 1893
+    },
+    {
+      "epoch": 1.0365836694950619,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001453652339068648,
+      "loss": 0.7429,
+      "step": 1894
+    },
+    {
+      "epoch": 1.0371400751147586,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00014531319643290442,
+      "loss": 0.7497,
+      "step": 1895
+    },
+    {
+      "epoch": 1.0376964807344553,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00014526114351356935,
+      "loss": 0.7219,
+      "step": 1896
+    },
+    {
+      "epoch": 1.0382528863541522,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00014520907516660228,
+      "loss": 0.6278,
+      "step": 1897
+    },
+    {
+      "epoch": 1.038809291973849,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00014515699140975113,
+      "loss": 0.7991,
+      "step": 1898
+    },
+    {
+      "epoch": 1.0393656975935457,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00014510489226076908,
+      "loss": 0.774,
+      "step": 1899
+    },
+    {
+      "epoch": 1.0399221032132424,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00014505277773741454,
+      "loss": 0.7075,
+      "step": 1900
+    },
+    {
+      "epoch": 1.0404785088329391,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00014500064785745117,
+      "loss": 0.7846,
+      "step": 1901
+    },
+    {
+      "epoch": 1.041034914452636,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001449485026386479,
+      "loss": 0.6826,
+      "step": 1902
+    },
+    {
+      "epoch": 1.0415913200723328,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00014489634209877882,
+      "loss": 0.5808,
+      "step": 1903
+    },
+    {
+      "epoch": 1.0421477256920295,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00014484416625562332,
+      "loss": 0.7508,
+      "step": 1904
+    },
+    {
+      "epoch": 1.0427041313117262,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00014479197512696587,
+      "loss": 0.7421,
+      "step": 1905
+    },
+    {
+      "epoch": 1.043260536931423,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00014473976873059636,
+      "loss": 0.6702,
+      "step": 1906
+    },
+    {
+      "epoch": 1.0438169425511197,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001446875470843097,
+      "loss": 0.7361,
+      "step": 1907
+    },
+    {
+      "epoch": 1.0443733481708166,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001446353102059061,
+      "loss": 0.6914,
+      "step": 1908
+    },
+    {
+      "epoch": 1.0449297537905133,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00014458305811319088,
+      "loss": 0.8452,
+      "step": 1909
+    },
+    {
+      "epoch": 1.04548615941021,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00014453079082397464,
+      "loss": 0.7407,
+      "step": 1910
+    },
+    {
+      "epoch": 1.0460425650299068,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001444785083560731,
+      "loss": 0.686,
+      "step": 1911
+    },
+    {
+      "epoch": 1.0465989706496035,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00014442621072730717,
+      "loss": 0.7235,
+      "step": 1912
+    },
+    {
+      "epoch": 1.0471553762693002,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00014437389795550292,
+      "loss": 0.7048,
+      "step": 1913
+    },
+    {
+      "epoch": 1.0477117818889972,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001443215700584916,
+      "loss": 0.7442,
+      "step": 1914
+    },
+    {
+      "epoch": 1.0482681875086939,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001442692270541096,
+      "loss": 0.7133,
+      "step": 1915
+    },
+    {
+      "epoch": 1.0488245931283906,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00014421686896019846,
+      "loss": 0.8122,
+      "step": 1916
+    },
+    {
+      "epoch": 1.0493809987480873,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00014416449579460484,
+      "loss": 0.7727,
+      "step": 1917
+    },
+    {
+      "epoch": 1.049937404367784,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00014411210757518062,
+      "loss": 0.6545,
+      "step": 1918
+    },
+    {
+      "epoch": 1.050493809987481,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00014405970431978268,
+      "loss": 0.7588,
+      "step": 1919
+    },
+    {
+      "epoch": 1.0510502156071777,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00014400728604627315,
+      "loss": 0.7321,
+      "step": 1920
+    },
+    {
+      "epoch": 1.0516066212268744,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00014395485277251922,
+      "loss": 0.6813,
+      "step": 1921
+    },
+    {
+      "epoch": 1.0521630268465711,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001439024045163932,
+      "loss": 0.7022,
+      "step": 1922
+    },
+    {
+      "epoch": 1.0527194324662679,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001438499412957725,
+      "loss": 0.7135,
+      "step": 1923
+    },
+    {
+      "epoch": 1.0532758380859646,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00014379746312853967,
+      "loss": 0.6793,
+      "step": 1924
+    },
+    {
+      "epoch": 1.0538322437056615,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00014374497003258232,
+      "loss": 0.6833,
+      "step": 1925
+    },
+    {
+      "epoch": 1.0543886493253583,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00014369246202579306,
+      "loss": 0.8443,
+      "step": 1926
+    },
+    {
+      "epoch": 1.054945054945055,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00014363993912606978,
+      "loss": 0.6652,
+      "step": 1927
+    },
+    {
+      "epoch": 1.0555014605647517,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00014358740135131533,
+      "loss": 0.7908,
+      "step": 1928
+    },
+    {
+      "epoch": 1.0560578661844484,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00014353484871943762,
+      "loss": 0.7398,
+      "step": 1929
+    },
+    {
+      "epoch": 1.0566142718041451,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00014348228124834962,
+      "loss": 0.6743,
+      "step": 1930
+    },
+    {
+      "epoch": 1.057170677423842,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00014342969895596942,
+      "loss": 0.7547,
+      "step": 1931
+    },
+    {
+      "epoch": 1.0577270830435388,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00014337710186022013,
+      "loss": 0.7303,
+      "step": 1932
+    },
+    {
+      "epoch": 1.0582834886632355,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00014332448997902988,
+      "loss": 0.7287,
+      "step": 1933
+    },
+    {
+      "epoch": 1.0588398942829322,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00014327186333033186,
+      "loss": 0.7541,
+      "step": 1934
+    },
+    {
+      "epoch": 1.059396299902629,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0001432192219320643,
+      "loss": 0.7516,
+      "step": 1935
+    },
+    {
+      "epoch": 1.0599527055223257,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00014316656580217046,
+      "loss": 0.778,
+      "step": 1936
+    },
+    {
+      "epoch": 1.0605091111420226,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00014311389495859863,
+      "loss": 0.7498,
+      "step": 1937
+    },
+    {
+      "epoch": 1.0610655167617193,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.000143061209419302,
+      "loss": 0.7228,
+      "step": 1938
+    },
+    {
+      "epoch": 1.061621922381416,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00014300850920223902,
+      "loss": 0.7578,
+      "step": 1939
+    },
+    {
+      "epoch": 1.0621783280011128,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001429557943253729,
+      "loss": 0.6644,
+      "step": 1940
+    },
+    {
+      "epoch": 1.0627347336208095,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00014290306480667196,
+      "loss": 0.7054,
+      "step": 1941
+    },
+    {
+      "epoch": 1.0632911392405062,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00014285032066410947,
+      "loss": 0.6552,
+      "step": 1942
+    },
+    {
+      "epoch": 1.0638475448602032,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00014279756191566372,
+      "loss": 0.6962,
+      "step": 1943
+    },
+    {
+      "epoch": 1.0644039504798999,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00014274478857931796,
+      "loss": 0.6713,
+      "step": 1944
+    },
+    {
+      "epoch": 1.0649603560995966,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00014269200067306044,
+      "loss": 0.7131,
+      "step": 1945
+    },
+    {
+      "epoch": 1.0655167617192933,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001426391982148843,
+      "loss": 0.7393,
+      "step": 1946
+    },
+    {
+      "epoch": 1.06607316733899,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0001425863812227877,
+      "loss": 0.6188,
+      "step": 1947
+    },
+    {
+      "epoch": 1.066629572958687,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001425335497147738,
+      "loss": 0.707,
+      "step": 1948
+    },
+    {
+      "epoch": 1.0671859785783837,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001424807037088506,
+      "loss": 0.6628,
+      "step": 1949
+    },
+    {
+      "epoch": 1.0677423841980804,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001424278432230311,
+      "loss": 0.644,
+      "step": 1950
+    },
+    {
+      "epoch": 1.0682987898177772,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00014237496827533325,
+      "loss": 0.7098,
+      "step": 1951
+    },
+    {
+      "epoch": 1.0688551954374739,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001423220788837799,
+      "loss": 0.7083,
+      "step": 1952
+    },
+    {
+      "epoch": 1.0694116010571706,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00014226917506639884,
+      "loss": 0.7144,
+      "step": 1953
+    },
+    {
+      "epoch": 1.0699680066768675,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0001422162568412227,
+      "loss": 0.8027,
+      "step": 1954
+    },
+    {
+      "epoch": 1.0705244122965643,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0001421633242262892,
+      "loss": 0.7031,
+      "step": 1955
+    },
+    {
+      "epoch": 1.071080817916261,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00014211037723964082,
+      "loss": 0.7143,
+      "step": 1956
+    },
+    {
+      "epoch": 1.0716372235359577,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00014205741589932495,
+      "loss": 0.6698,
+      "step": 1957
+    },
+    {
+      "epoch": 1.0721936291556544,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001420044402233939,
+      "loss": 0.6744,
+      "step": 1958
+    },
+    {
+      "epoch": 1.0727500347753511,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001419514502299049,
+      "loss": 0.6424,
+      "step": 1959
+    },
+    {
+      "epoch": 1.073306440395048,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00014189844593691996,
+      "loss": 0.7172,
+      "step": 1960
+    },
+    {
+      "epoch": 1.0738628460147448,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001418454273625061,
+      "loss": 0.6733,
+      "step": 1961
+    },
+    {
+      "epoch": 1.0744192516344415,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001417923945247351,
+      "loss": 0.6761,
+      "step": 1962
+    },
+    {
+      "epoch": 1.0749756572541382,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00014173934744168365,
+      "loss": 0.7845,
+      "step": 1963
+    },
+    {
+      "epoch": 1.075532062873835,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00014168628613143324,
+      "loss": 0.7739,
+      "step": 1964
+    },
+    {
+      "epoch": 1.076088468493532,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00014163321061207032,
+      "loss": 0.7574,
+      "step": 1965
+    },
+    {
+      "epoch": 1.0766448741132286,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00014158012090168605,
+      "loss": 0.7267,
+      "step": 1966
+    },
+    {
+      "epoch": 1.0772012797329253,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00014152701701837652,
+      "loss": 0.7494,
+      "step": 1967
+    },
+    {
+      "epoch": 1.077757685352622,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00014147389898024266,
+      "loss": 0.7303,
+      "step": 1968
+    },
+    {
+      "epoch": 1.0783140909723188,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00014142076680539014,
+      "loss": 0.6702,
+      "step": 1969
+    },
+    {
+      "epoch": 1.0788704965920155,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00014136762051192954,
+      "loss": 0.8128,
+      "step": 1970
+    },
+    {
+      "epoch": 1.0794269022117122,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00014131446011797613,
+      "loss": 0.8153,
+      "step": 1971
+    },
+    {
+      "epoch": 1.0799833078314092,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00014126128564165017,
+      "loss": 0.7381,
+      "step": 1972
+    },
+    {
+      "epoch": 1.0805397134511059,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00014120809710107656,
+      "loss": 0.7896,
+      "step": 1973
+    },
+    {
+      "epoch": 1.0810961190708026,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00014115489451438506,
+      "loss": 0.7006,
+      "step": 1974
+    },
+    {
+      "epoch": 1.0816525246904993,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001411016778997102,
+      "loss": 0.6817,
+      "step": 1975
+    },
+    {
+      "epoch": 1.082208930310196,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00014104844727519134,
+      "loss": 0.7037,
+      "step": 1976
+    },
+    {
+      "epoch": 1.082765335929893,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001409952026589725,
+      "loss": 0.6873,
+      "step": 1977
+    },
+    {
+      "epoch": 1.0833217415495897,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001409419440692026,
+      "loss": 0.6717,
+      "step": 1978
+    },
+    {
+      "epoch": 1.0838781471692864,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00014088867152403525,
+      "loss": 0.8306,
+      "step": 1979
+    },
+    {
+      "epoch": 1.0844345527889832,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00014083538504162883,
+      "loss": 0.7378,
+      "step": 1980
+    },
+    {
+      "epoch": 1.0849909584086799,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001407820846401465,
+      "loss": 0.7589,
+      "step": 1981
+    },
+    {
+      "epoch": 1.0855473640283766,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001407287703377561,
+      "loss": 0.7173,
+      "step": 1982
+    },
+    {
+      "epoch": 1.0861037696480735,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00014067544215263025,
+      "loss": 0.7238,
+      "step": 1983
+    },
+    {
+      "epoch": 1.0866601752677703,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00014062210010294635,
+      "loss": 0.7191,
+      "step": 1984
+    },
+    {
+      "epoch": 1.087216580887467,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00014056874420688645,
+      "loss": 0.7664,
+      "step": 1985
+    },
+    {
+      "epoch": 1.0877729865071637,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00014051537448263733,
+      "loss": 0.7795,
+      "step": 1986
+    },
+    {
+      "epoch": 1.0883293921268604,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001404619909483905,
+      "loss": 0.7233,
+      "step": 1987
+    },
+    {
+      "epoch": 1.0888857977465571,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001404085936223422,
+      "loss": 0.772,
+      "step": 1988
+    },
+    {
+      "epoch": 1.089442203366254,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001403551825226933,
+      "loss": 0.8023,
+      "step": 1989
+    },
+    {
+      "epoch": 1.0899986089859508,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00014030175766764944,
+      "loss": 0.6747,
+      "step": 1990
+    },
+    {
+      "epoch": 1.0905550146056475,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00014024831907542094,
+      "loss": 0.8103,
+      "step": 1991
+    },
+    {
+      "epoch": 1.0911114202253442,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00014019486676422274,
+      "loss": 0.5994,
+      "step": 1992
+    },
+    {
+      "epoch": 1.091667825845041,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00014014140075227453,
+      "loss": 0.7553,
+      "step": 1993
+    },
+    {
+      "epoch": 1.092224231464738,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00014008792105780064,
+      "loss": 0.646,
+      "step": 1994
+    },
+    {
+      "epoch": 1.0927806370844346,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00014003442769903,
+      "loss": 0.7307,
+      "step": 1995
+    },
+    {
+      "epoch": 1.0933370427041313,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00013998092069419636,
+      "loss": 0.7022,
+      "step": 1996
+    },
+    {
+      "epoch": 1.093893448323828,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00013992740006153796,
+      "loss": 0.8088,
+      "step": 1997
+    },
+    {
+      "epoch": 1.0944498539435248,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001398738658192978,
+      "loss": 0.652,
+      "step": 1998
+    },
+    {
+      "epoch": 1.0950062595632215,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00013982031798572336,
+      "loss": 0.756,
+      "step": 1999
+    },
+    {
+      "epoch": 1.0955626651829184,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00013976675657906694,
+      "loss": 0.6802,
+      "step": 2000
+    },
+    {
+      "epoch": 1.0961190708026152,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001397131816175854,
+      "loss": 0.6098,
+      "step": 2001
+    },
+    {
+      "epoch": 1.096675476422312,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00013965959311954012,
+      "loss": 0.7077,
+      "step": 2002
+    },
+    {
+      "epoch": 1.0972318820420086,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00013960599110319726,
+      "loss": 0.7186,
+      "step": 2003
+    },
+    {
+      "epoch": 1.0977882876617053,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001395523755868275,
+      "loss": 0.6666,
+      "step": 2004
+    },
+    {
+      "epoch": 1.098344693281402,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00013949874658870608,
+      "loss": 0.7174,
+      "step": 2005
+    },
+    {
+      "epoch": 1.098901098901099,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00013944510412711294,
+      "loss": 0.7461,
+      "step": 2006
+    },
+    {
+      "epoch": 1.0994575045207957,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00013939144822033253,
+      "loss": 0.701,
+      "step": 2007
+    },
+    {
+      "epoch": 1.1000139101404924,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00013933777888665383,
+      "loss": 0.6507,
+      "step": 2008
+    },
+    {
+      "epoch": 1.1005703157601892,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001392840961443706,
+      "loss": 0.6878,
+      "step": 2009
+    },
+    {
+      "epoch": 1.1011267213798859,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00013923040001178098,
+      "loss": 0.7909,
+      "step": 2010
+    },
+    {
+      "epoch": 1.1016831269995826,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00013917669050718775,
+      "loss": 0.7602,
+      "step": 2011
+    },
+    {
+      "epoch": 1.1022395326192795,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001391229676488982,
+      "loss": 0.7296,
+      "step": 2012
+    },
+    {
+      "epoch": 1.1027959382389763,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00013906923145522426,
+      "loss": 0.7202,
+      "step": 2013
+    },
+    {
+      "epoch": 1.103352343858673,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0001390154819444823,
+      "loss": 0.6599,
+      "step": 2014
+    },
+    {
+      "epoch": 1.1039087494783697,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00013896171913499336,
+      "loss": 0.663,
+      "step": 2015
+    },
+    {
+      "epoch": 1.1044651550980664,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00013890794304508282,
+      "loss": 0.6925,
+      "step": 2016
+    },
+    {
+      "epoch": 1.1050215607177631,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001388541536930808,
+      "loss": 0.7559,
+      "step": 2017
+    },
+    {
+      "epoch": 1.10557796633746,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001388003510973218,
+      "loss": 0.8143,
+      "step": 2018
+    },
+    {
+      "epoch": 1.1061343719571568,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00013874653527614492,
+      "loss": 0.6369,
+      "step": 2019
+    },
+    {
+      "epoch": 1.1066907775768535,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00013869270624789366,
+      "loss": 0.6989,
+      "step": 2020
+    },
+    {
+      "epoch": 1.1072471831965502,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0001386388640309161,
+      "loss": 0.7568,
+      "step": 2021
+    },
+    {
+      "epoch": 1.107803588816247,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00013858500864356483,
+      "loss": 0.7158,
+      "step": 2022
+    },
+    {
+      "epoch": 1.108359994435944,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00013853114010419689,
+      "loss": 0.6169,
+      "step": 2023
+    },
+    {
+      "epoch": 1.1089164000556406,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00013847725843117375,
+      "loss": 1.592,
+      "step": 2024
+    },
+    {
+      "epoch": 1.1094728056753373,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00013842336364286155,
+      "loss": 0.7245,
+      "step": 2025
+    },
+    {
+      "epoch": 1.110029211295034,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00013836945575763067,
+      "loss": 0.648,
+      "step": 2026
+    },
+    {
+      "epoch": 1.1105856169147308,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0001383155347938561,
+      "loss": 0.8117,
+      "step": 2027
+    },
+    {
+      "epoch": 1.1111420225344275,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001382616007699172,
+      "loss": 0.7578,
+      "step": 2028
+    },
+    {
+      "epoch": 1.1116984281541245,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00013820765370419787,
+      "loss": 0.6699,
+      "step": 2029
+    },
+    {
+      "epoch": 1.1122548337738212,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00013815369361508636,
+      "loss": 0.6344,
+      "step": 2030
+    },
+    {
+      "epoch": 1.112811239393518,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0001380997205209755,
+      "loss": 0.7591,
+      "step": 2031
+    },
+    {
+      "epoch": 1.1133676450132146,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00013804573444026237,
+      "loss": 0.7,
+      "step": 2032
+    },
+    {
+      "epoch": 1.1139240506329113,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0001379917353913486,
+      "loss": 0.7596,
+      "step": 2033
+    },
+    {
+      "epoch": 1.114480456252608,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001379377233926402,
+      "loss": 0.7626,
+      "step": 2034
+    },
+    {
+      "epoch": 1.115036861872305,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00013788369846254764,
+      "loss": 0.8207,
+      "step": 2035
+    },
+    {
+      "epoch": 1.1155932674920017,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0001378296606194857,
+      "loss": 0.8149,
+      "step": 2036
+    },
+    {
+      "epoch": 1.1161496731116984,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00013777560988187363,
+      "loss": 0.7041,
+      "step": 2037
+    },
+    {
+      "epoch": 1.1167060787313952,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00013772154626813513,
+      "loss": 0.7434,
+      "step": 2038
+    },
+    {
+      "epoch": 1.1172624843510919,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001376674697966982,
+      "loss": 0.7522,
+      "step": 2039
+    },
+    {
+      "epoch": 1.1178188899707888,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00013761338048599525,
+      "loss": 0.771,
+      "step": 2040
+    },
+    {
+      "epoch": 1.1183752955904855,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00013755927835446298,
+      "loss": 0.761,
+      "step": 2041
+    },
+    {
+      "epoch": 1.1189317012101823,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00013750516342054265,
+      "loss": 0.7439,
+      "step": 2042
+    },
+    {
+      "epoch": 1.119488106829879,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00013745103570267977,
+      "loss": 0.7558,
+      "step": 2043
+    },
+    {
+      "epoch": 1.1200445124495757,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00013739689521932418,
+      "loss": 0.7231,
+      "step": 2044
+    },
+    {
+      "epoch": 1.1206009180692724,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00013734274198893008,
+      "loss": 0.7869,
+      "step": 2045
+    },
+    {
+      "epoch": 1.1211573236889691,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00013728857602995613,
+      "loss": 0.6867,
+      "step": 2046
+    },
+    {
+      "epoch": 1.121713729308666,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0001372343973608652,
+      "loss": 0.744,
+      "step": 2047
+    },
+    {
+      "epoch": 1.1222701349283628,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00013718020600012452,
+      "loss": 0.7727,
+      "step": 2048
+    },
+    {
+      "epoch": 1.1228265405480595,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00013712600196620566,
+      "loss": 0.6544,
+      "step": 2049
+    },
+    {
+      "epoch": 1.1233829461677562,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00013707178527758452,
+      "loss": 0.725,
+      "step": 2050
+    },
+    {
+      "epoch": 1.123939351787453,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001370175559527413,
+      "loss": 0.5613,
+      "step": 2051
+    },
+    {
+      "epoch": 1.12449575740715,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00013696331401016052,
+      "loss": 0.8063,
+      "step": 2052
+    },
+    {
+      "epoch": 1.1250521630268466,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00013690905946833098,
+      "loss": 0.6506,
+      "step": 2053
+    },
+    {
+      "epoch": 1.1256085686465434,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00013685479234574577,
+      "loss": 0.7368,
+      "step": 2054
+    },
+    {
+      "epoch": 1.12616497426624,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001368005126609023,
+      "loss": 0.7023,
+      "step": 2055
+    },
+    {
+      "epoch": 1.1267213798859368,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00013674622043230228,
+      "loss": 0.7071,
+      "step": 2056
+    },
+    {
+      "epoch": 1.1272777855056335,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00013669191567845158,
+      "loss": 0.6503,
+      "step": 2057
+    },
+    {
+      "epoch": 1.1278341911253305,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001366375984178605,
+      "loss": 0.6353,
+      "step": 2058
+    },
+    {
+      "epoch": 1.1283905967450272,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001365832686690435,
+      "loss": 0.7651,
+      "step": 2059
+    },
+    {
+      "epoch": 1.128947002364724,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0001365289264505193,
+      "loss": 0.6411,
+      "step": 2060
+    },
+    {
+      "epoch": 1.1295034079844206,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0001364745717808109,
+      "loss": 0.593,
+      "step": 2061
+    },
+    {
+      "epoch": 1.1300598136041173,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00013642020467844555,
+      "loss": 0.657,
+      "step": 2062
+    },
+    {
+      "epoch": 1.130616219223814,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00013636582516195472,
+      "loss": 0.704,
+      "step": 2063
+    },
+    {
+      "epoch": 1.131172624843511,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0001363114332498741,
+      "loss": 0.8242,
+      "step": 2064
+    },
+    {
+      "epoch": 1.1317290304632077,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00013625702896074361,
+      "loss": 0.749,
+      "step": 2065
+    },
+    {
+      "epoch": 1.1322854360829044,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00013620261231310743,
+      "loss": 0.7413,
+      "step": 2066
+    },
+    {
+      "epoch": 1.1328418417026012,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001361481833255139,
+      "loss": 0.7288,
+      "step": 2067
+    },
+    {
+      "epoch": 1.1333982473222979,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0001360937420165156,
+      "loss": 0.7728,
+      "step": 2068
+    },
+    {
+      "epoch": 1.1339546529419948,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00013603928840466928,
+      "loss": 0.7667,
+      "step": 2069
+    },
+    {
+      "epoch": 1.1345110585616915,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00013598482250853588,
+      "loss": 0.7391,
+      "step": 2070
+    },
+    {
+      "epoch": 1.1350674641813883,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00013593034434668063,
+      "loss": 0.6652,
+      "step": 2071
+    },
+    {
+      "epoch": 1.135623869801085,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00013587585393767278,
+      "loss": 0.7477,
+      "step": 2072
+    },
+    {
+      "epoch": 1.1361802754207817,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00013582135130008584,
+      "loss": 0.7554,
+      "step": 2073
+    },
+    {
+      "epoch": 1.1367366810404784,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0001357668364524975,
+      "loss": 0.8427,
+      "step": 2074
+    },
+    {
+      "epoch": 1.1372930866601751,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001357123094134896,
+      "loss": 0.7391,
+      "step": 2075
+    },
+    {
+      "epoch": 1.137849492279872,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00013565777020164814,
+      "loss": 0.7641,
+      "step": 2076
+    },
+    {
+      "epoch": 1.1384058978995688,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001356032188355632,
+      "loss": 0.6439,
+      "step": 2077
+    },
+    {
+      "epoch": 1.1389623035192655,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00013554865533382915,
+      "loss": 0.5995,
+      "step": 2078
+    },
+    {
+      "epoch": 1.1395187091389622,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001354940797150443,
+      "loss": 0.7362,
+      "step": 2079
+    },
+    {
+      "epoch": 1.140075114758659,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001354394919978113,
+      "loss": 0.7123,
+      "step": 2080
+    },
+    {
+      "epoch": 1.140631520378356,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00013538489220073677,
+      "loss": 0.7947,
+      "step": 2081
+    },
+    {
+      "epoch": 1.1411879259980526,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00013533028034243145,
+      "loss": 0.7087,
+      "step": 2082
+    },
+    {
+      "epoch": 1.1417443316177494,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00013527565644151033,
+      "loss": 0.7273,
+      "step": 2083
+    },
+    {
+      "epoch": 1.142300737237446,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00013522102051659234,
+      "loss": 0.8265,
+      "step": 2084
+    },
+    {
+      "epoch": 1.1428571428571428,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00013516637258630067,
+      "loss": 0.696,
+      "step": 2085
+    },
+    {
+      "epoch": 1.1434135484768397,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00013511171266926243,
+      "loss": 0.708,
+      "step": 2086
+    },
+    {
+      "epoch": 1.1439699540965365,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00013505704078410898,
+      "loss": 0.7611,
+      "step": 2087
+    },
+    {
+      "epoch": 1.1445263597162332,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00013500235694947557,
+      "loss": 0.7434,
+      "step": 2088
+    },
+    {
+      "epoch": 1.14508276533593,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00013494766118400176,
+      "loss": 0.7102,
+      "step": 2089
+    },
+    {
+      "epoch": 1.1456391709556266,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00013489295350633095,
+      "loss": 0.6826,
+      "step": 2090
+    },
+    {
+      "epoch": 1.1461955765753233,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00013483823393511076,
+      "loss": 0.6801,
+      "step": 2091
+    },
+    {
+      "epoch": 1.14675198219502,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001347835024889928,
+      "loss": 0.7401,
+      "step": 2092
+    },
+    {
+      "epoch": 1.147308387814717,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00013472875918663272,
+      "loss": 0.6547,
+      "step": 2093
+    },
+    {
+      "epoch": 1.1478647934344137,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00013467400404669017,
+      "loss": 0.7809,
+      "step": 2094
+    },
+    {
+      "epoch": 1.1484211990541104,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00013461923708782896,
+      "loss": 0.7175,
+      "step": 2095
+    },
+    {
+      "epoch": 1.1489776046738072,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00013456445832871688,
+      "loss": 0.7707,
+      "step": 2096
+    },
+    {
+      "epoch": 1.1495340102935039,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00013450966778802564,
+      "loss": 0.6497,
+      "step": 2097
+    },
+    {
+      "epoch": 1.1500904159132008,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00013445486548443108,
+      "loss": 0.6757,
+      "step": 2098
+    },
+    {
+      "epoch": 1.1506468215328975,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00013440005143661304,
+      "loss": 0.7418,
+      "step": 2099
+    },
+    {
+      "epoch": 1.1512032271525943,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001343452256632553,
+      "loss": 0.7288,
+      "step": 2100
+    },
+    {
+      "epoch": 1.151759632772291,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0001342903881830457,
+      "loss": 0.7445,
+      "step": 2101
+    },
+    {
+      "epoch": 1.1523160383919877,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00013423553901467598,
+      "loss": 0.7464,
+      "step": 2102
+    },
+    {
+      "epoch": 1.1528724440116844,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00013418067817684204,
+      "loss": 0.6812,
+      "step": 2103
+    },
+    {
+      "epoch": 1.1534288496313814,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00013412580568824357,
+      "loss": 0.6988,
+      "step": 2104
+    },
+    {
+      "epoch": 1.153985255251078,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00013407092156758434,
+      "loss": 0.7441,
+      "step": 2105
+    },
+    {
+      "epoch": 1.1545416608707748,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00013401602583357206,
+      "loss": 0.762,
+      "step": 2106
+    },
+    {
+      "epoch": 1.1550980664904715,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00013396111850491832,
+      "loss": 0.7637,
+      "step": 2107
+    },
+    {
+      "epoch": 1.1556544721101683,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00013390619960033882,
+      "loss": 0.7193,
+      "step": 2108
+    },
+    {
+      "epoch": 1.156210877729865,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00013385126913855307,
+      "loss": 0.7295,
+      "step": 2109
+    },
+    {
+      "epoch": 1.156767283349562,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00013379632713828458,
+      "loss": 0.7448,
+      "step": 2110
+    },
+    {
+      "epoch": 1.1573236889692586,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00013374137361826075,
+      "loss": 0.7631,
+      "step": 2111
+    },
+    {
+      "epoch": 1.1578800945889554,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00013368640859721302,
+      "loss": 0.7028,
+      "step": 2112
+    },
+    {
+      "epoch": 1.158436500208652,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001336314320938766,
+      "loss": 0.7438,
+      "step": 2113
+    },
+    {
+      "epoch": 1.1589929058283488,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00013357644412699071,
+      "loss": 0.6364,
+      "step": 2114
+    },
+    {
+      "epoch": 1.1595493114480457,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001335214447152984,
+      "loss": 0.7231,
+      "step": 2115
+    },
+    {
+      "epoch": 1.1601057170677425,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0001334664338775468,
+      "loss": 0.6903,
+      "step": 2116
+    },
+    {
+      "epoch": 1.1606621226874392,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013341141163248664,
+      "loss": 1.7995,
+      "step": 2117
+    },
+    {
+      "epoch": 1.161218528307136,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00013335637799887282,
+      "loss": 0.7381,
+      "step": 2118
+    },
+    {
+      "epoch": 1.1617749339268326,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00013330133299546394,
+      "loss": 0.698,
+      "step": 2119
+    },
+    {
+      "epoch": 1.1623313395465293,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00013324627664102257,
+      "loss": 0.6403,
+      "step": 2120
+    },
+    {
+      "epoch": 1.162887745166226,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00013319120895431517,
+      "loss": 0.7535,
+      "step": 2121
+    },
+    {
+      "epoch": 1.163444150785923,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001331361299541119,
+      "loss": 0.7806,
+      "step": 2122
+    },
+    {
+      "epoch": 1.1640005564056197,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00013308103965918694,
+      "loss": 0.6097,
+      "step": 2123
+    },
+    {
+      "epoch": 1.1645569620253164,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001330259380883183,
+      "loss": 0.7872,
+      "step": 2124
+    },
+    {
+      "epoch": 1.1651133676450132,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00013297082526028778,
+      "loss": 0.7253,
+      "step": 2125
+    },
+    {
+      "epoch": 1.1656697732647099,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00013291570119388106,
+      "loss": 0.7474,
+      "step": 2126
+    },
+    {
+      "epoch": 1.1662261788844068,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00013286056590788755,
+      "loss": 0.7563,
+      "step": 2127
+    },
+    {
+      "epoch": 1.1667825845041035,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001328054194211007,
+      "loss": 0.7449,
+      "step": 2128
+    },
+    {
+      "epoch": 1.1673389901238003,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00013275026175231752,
+      "loss": 0.7136,
+      "step": 2129
+    },
+    {
+      "epoch": 1.167895395743497,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00013269509292033901,
+      "loss": 0.6464,
+      "step": 2130
+    },
+    {
+      "epoch": 1.1684518013631937,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001326399129439699,
+      "loss": 0.6748,
+      "step": 2131
+    },
+    {
+      "epoch": 1.1690082069828907,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00013258472184201877,
+      "loss": 0.7183,
+      "step": 2132
+    },
+    {
+      "epoch": 1.1695646126025874,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001325295196332979,
+      "loss": 0.7433,
+      "step": 2133
+    },
+    {
+      "epoch": 1.170121018222284,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0001324743063366235,
+      "loss": 0.8233,
+      "step": 2134
+    },
+    {
+      "epoch": 1.1706774238419808,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00013241908197081538,
+      "loss": 0.6533,
+      "step": 2135
+    },
+    {
+      "epoch": 1.1712338294616775,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00013236384655469728,
+      "loss": 0.6984,
+      "step": 2136
+    },
+    {
+      "epoch": 1.1717902350813743,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00013230860010709665,
+      "loss": 0.6996,
+      "step": 2137
+    },
+    {
+      "epoch": 1.172346640701071,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00013225334264684466,
+      "loss": 0.6575,
+      "step": 2138
+    },
+    {
+      "epoch": 1.172903046320768,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00013219807419277626,
+      "loss": 0.8134,
+      "step": 2139
+    },
+    {
+      "epoch": 1.1734594519404646,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001321427947637302,
+      "loss": 0.6798,
+      "step": 2140
+    },
+    {
+      "epoch": 1.1740158575601614,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0001320875043785489,
+      "loss": 0.776,
+      "step": 2141
+    },
+    {
+      "epoch": 1.174572263179858,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00013203220305607854,
+      "loss": 0.7181,
+      "step": 2142
+    },
+    {
+      "epoch": 1.1751286687995548,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00013197689081516906,
+      "loss": 0.7336,
+      "step": 2143
+    },
+    {
+      "epoch": 1.1756850744192517,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00013192156767467405,
+      "loss": 0.721,
+      "step": 2144
+    },
+    {
+      "epoch": 1.1762414800389485,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00013186623365345088,
+      "loss": 0.615,
+      "step": 2145
+    },
+    {
+      "epoch": 1.1767978856586452,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00013181088877036058,
+      "loss": 0.6619,
+      "step": 2146
+    },
+    {
+      "epoch": 1.177354291278342,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00013175553304426796,
+      "loss": 0.6747,
+      "step": 2147
+    },
+    {
+      "epoch": 1.1779106968980386,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001317001664940414,
+      "loss": 0.6285,
+      "step": 2148
+    },
+    {
+      "epoch": 1.1784671025177353,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001316447891385531,
+      "loss": 0.6129,
+      "step": 2149
+    },
+    {
+      "epoch": 1.179023508137432,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00013158940099667889,
+      "loss": 0.6575,
+      "step": 2150
+    },
+    {
+      "epoch": 1.179579913757129,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00013153400208729823,
+      "loss": 0.7835,
+      "step": 2151
+    },
+    {
+      "epoch": 1.1801363193768257,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00013147859242929432,
+      "loss": 0.8569,
+      "step": 2152
+    },
+    {
+      "epoch": 1.1806927249965224,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00013142317204155397,
+      "loss": 0.7776,
+      "step": 2153
+    },
+    {
+      "epoch": 1.1812491306162192,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001313677409429677,
+      "loss": 0.6162,
+      "step": 2154
+    },
+    {
+      "epoch": 1.1818055362359159,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00013131229915242965,
+      "loss": 0.7546,
+      "step": 2155
+    },
+    {
+      "epoch": 1.1823619418556128,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00013125684668883762,
+      "loss": 0.7137,
+      "step": 2156
+    },
+    {
+      "epoch": 1.1829183474753096,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00013120138357109303,
+      "loss": 0.7243,
+      "step": 2157
+    },
+    {
+      "epoch": 1.1834747530950063,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001311459098181009,
+      "loss": 0.7115,
+      "step": 2158
+    },
+    {
+      "epoch": 1.184031158714703,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00013109042544876996,
+      "loss": 0.7396,
+      "step": 2159
+    },
+    {
+      "epoch": 1.1845875643343997,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00013103493048201246,
+      "loss": 0.7101,
+      "step": 2160
+    },
+    {
+      "epoch": 1.1851439699540967,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0001309794249367444,
+      "loss": 0.8095,
+      "step": 2161
+    },
+    {
+      "epoch": 1.1857003755737934,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00013092390883188526,
+      "loss": 0.6576,
+      "step": 2162
+    },
+    {
+      "epoch": 1.18625678119349,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00013086838218635813,
+      "loss": 0.7822,
+      "step": 2163
+    },
+    {
+      "epoch": 1.1868131868131868,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00013081284501908976,
+      "loss": 0.6946,
+      "step": 2164
+    },
+    {
+      "epoch": 1.1873695924328835,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00013075729734901044,
+      "loss": 0.6942,
+      "step": 2165
+    },
+    {
+      "epoch": 1.1879259980525803,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.000130701739195054,
+      "loss": 0.7864,
+      "step": 2166
+    },
+    {
+      "epoch": 1.188482403672277,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.000130646170576158,
+      "loss": 0.6483,
+      "step": 2167
+    },
+    {
+      "epoch": 1.189038809291974,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00013059059151126335,
+      "loss": 0.674,
+      "step": 2168
+    },
+    {
+      "epoch": 1.1895952149116706,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00013053500201931468,
+      "loss": 0.7591,
+      "step": 2169
+    },
+    {
+      "epoch": 1.1901516205313674,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00013047940211926014,
+      "loss": 0.7159,
+      "step": 2170
+    },
+    {
+      "epoch": 1.190708026151064,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00013042379183005135,
+      "loss": 0.7877,
+      "step": 2171
+    },
+    {
+      "epoch": 1.1912644317707608,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00013036817117064361,
+      "loss": 0.7207,
+      "step": 2172
+    },
+    {
+      "epoch": 1.1918208373904577,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00013031254015999564,
+      "loss": 0.7245,
+      "step": 2173
+    },
+    {
+      "epoch": 1.1923772430101545,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00013025689881706972,
+      "loss": 0.7071,
+      "step": 2174
+    },
+    {
+      "epoch": 1.1929336486298512,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0001302012471608317,
+      "loss": 0.7959,
+      "step": 2175
+    },
+    {
+      "epoch": 1.193490054249548,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0001301455852102508,
+      "loss": 0.7475,
+      "step": 2176
+    },
+    {
+      "epoch": 1.1940464598692446,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00013008991298429999,
+      "loss": 0.6675,
+      "step": 2177
+    },
+    {
+      "epoch": 1.1946028654889413,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001300342305019555,
+      "loss": 0.693,
+      "step": 2178
+    },
+    {
+      "epoch": 1.1951592711086383,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00012997853778219721,
+      "loss": 0.7605,
+      "step": 2179
+    },
+    {
+      "epoch": 1.195715676728335,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00012992283484400847,
+      "loss": 0.7103,
+      "step": 2180
+    },
+    {
+      "epoch": 1.1962720823480317,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00012986712170637597,
+      "loss": 0.6464,
+      "step": 2181
+    },
+    {
+      "epoch": 1.1968284879677284,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001298113983882901,
+      "loss": 0.7043,
+      "step": 2182
+    },
+    {
+      "epoch": 1.1973848935874252,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00012975566490874458,
+      "loss": 0.7208,
+      "step": 2183
+    },
+    {
+      "epoch": 1.197941299207122,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00012969992128673662,
+      "loss": 0.6785,
+      "step": 2184
+    },
+    {
+      "epoch": 1.1984977048268188,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00012964416754126685,
+      "loss": 0.7649,
+      "step": 2185
+    },
+    {
+      "epoch": 1.1990541104465156,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00012958840369133945,
+      "loss": 0.7386,
+      "step": 2186
+    },
+    {
+      "epoch": 1.1996105160662123,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00012953262975596197,
+      "loss": 0.6977,
+      "step": 2187
+    },
+    {
+      "epoch": 1.200166921685909,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0001294768457541454,
+      "loss": 0.806,
+      "step": 2188
+    },
+    {
+      "epoch": 1.2007233273056057,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00012942105170490417,
+      "loss": 0.8397,
+      "step": 2189
+    },
+    {
+      "epoch": 1.2012797329253027,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00012936524762725618,
+      "loss": 0.6611,
+      "step": 2190
+    },
+    {
+      "epoch": 1.2018361385449994,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00012930943354022263,
+      "loss": 0.787,
+      "step": 2191
+    },
+    {
+      "epoch": 1.202392544164696,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0001292536094628283,
+      "loss": 0.7341,
+      "step": 2192
+    },
+    {
+      "epoch": 1.2029489497843928,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00012919777541410126,
+      "loss": 0.8072,
+      "step": 2193
+    },
+    {
+      "epoch": 1.2035053554040895,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.000129141931413073,
+      "loss": 0.7513,
+      "step": 2194
+    },
+    {
+      "epoch": 1.2040617610237863,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00012908607747877832,
+      "loss": 0.7565,
+      "step": 2195
+    },
+    {
+      "epoch": 1.204618166643483,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001290302136302556,
+      "loss": 0.7509,
+      "step": 2196
+    },
+    {
+      "epoch": 1.20517457226318,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00012897433988654644,
+      "loss": 0.6535,
+      "step": 2197
+    },
+    {
+      "epoch": 1.2057309778828766,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001289184562666959,
+      "loss": 0.6429,
+      "step": 2198
+    },
+    {
+      "epoch": 1.2062873835025734,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00012886256278975231,
+      "loss": 0.735,
+      "step": 2199
+    },
+    {
+      "epoch": 1.20684378912227,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00012880665947476752,
+      "loss": 0.8336,
+      "step": 2200
+    },
+    {
+      "epoch": 1.2074001947419668,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00012875074634079652,
+      "loss": 0.6905,
+      "step": 2201
+    },
+    {
+      "epoch": 1.2079566003616637,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001286948234068978,
+      "loss": 0.7417,
+      "step": 2202
+    },
+    {
+      "epoch": 1.2085130059813605,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00012863889069213323,
+      "loss": 0.6903,
+      "step": 2203
+    },
+    {
+      "epoch": 1.2090694116010572,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00012858294821556786,
+      "loss": 0.7601,
+      "step": 2204
+    },
+    {
+      "epoch": 1.209625817220754,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001285269959962701,
+      "loss": 0.7129,
+      "step": 2205
+    },
+    {
+      "epoch": 1.2101822228404506,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00012847103405331178,
+      "loss": 0.6655,
+      "step": 2206
+    },
+    {
+      "epoch": 1.2107386284601476,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00012841506240576802,
+      "loss": 0.722,
+      "step": 2207
+    },
+    {
+      "epoch": 1.2112950340798443,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00012835908107271715,
+      "loss": 0.8153,
+      "step": 2208
+    },
+    {
+      "epoch": 1.211851439699541,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00012830309007324088,
+      "loss": 0.7669,
+      "step": 2209
+    },
+    {
+      "epoch": 1.2124078453192377,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00012824708942642423,
+      "loss": 1.6954,
+      "step": 2210
+    },
+    {
+      "epoch": 1.2129642509389345,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00012819107915135547,
+      "loss": 0.6916,
+      "step": 2211
+    },
+    {
+      "epoch": 1.2135206565586312,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00012813505926712615,
+      "loss": 0.778,
+      "step": 2212
+    },
+    {
+      "epoch": 1.214077062178328,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00012807902979283112,
+      "loss": 0.7034,
+      "step": 2213
+    },
+    {
+      "epoch": 1.2146334677980248,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00012802299074756842,
+      "loss": 0.7273,
+      "step": 2214
+    },
+    {
+      "epoch": 1.2151898734177216,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0001279669421504395,
+      "loss": 0.8011,
+      "step": 2215
+    },
+    {
+      "epoch": 1.2157462790374183,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.000127910884020549,
+      "loss": 0.6606,
+      "step": 2216
+    },
+    {
+      "epoch": 1.216302684657115,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00012785481637700473,
+      "loss": 0.6426,
+      "step": 2217
+    },
+    {
+      "epoch": 1.2168590902768117,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001277987392389178,
+      "loss": 0.6805,
+      "step": 2218
+    },
+    {
+      "epoch": 1.2174154958965087,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00012774265262540264,
+      "loss": 0.7354,
+      "step": 2219
+    },
+    {
+      "epoch": 1.2179719015162054,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00012768655655557675,
+      "loss": 0.6821,
+      "step": 2220
+    },
+    {
+      "epoch": 1.218528307135902,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.000127630451048561,
+      "loss": 0.7067,
+      "step": 2221
+    },
+    {
+      "epoch": 1.2190847127555988,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00012757433612347937,
+      "loss": 0.7222,
+      "step": 2222
+    },
+    {
+      "epoch": 1.2196411183752955,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001275182117994591,
+      "loss": 0.7561,
+      "step": 2223
+    },
+    {
+      "epoch": 1.2201975239949923,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00012746207809563064,
+      "loss": 0.7554,
+      "step": 2224
+    },
+    {
+      "epoch": 1.2207539296146892,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00012740593503112762,
+      "loss": 0.7669,
+      "step": 2225
+    },
+    {
+      "epoch": 1.221310335234386,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00012734978262508682,
+      "loss": 0.6811,
+      "step": 2226
+    },
+    {
+      "epoch": 1.2218667408540826,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00012729362089664828,
+      "loss": 0.6752,
+      "step": 2227
+    },
+    {
+      "epoch": 1.2224231464737794,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00012723744986495525,
+      "loss": 0.7698,
+      "step": 2228
+    },
+    {
+      "epoch": 1.222979552093476,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.000127181269549154,
+      "loss": 0.7404,
+      "step": 2229
+    },
+    {
+      "epoch": 1.2235359577131728,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00012712507996839404,
+      "loss": 0.7231,
+      "step": 2230
+    },
+    {
+      "epoch": 1.2240923633328697,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00012706888114182813,
+      "loss": 0.6616,
+      "step": 2231
+    },
+    {
+      "epoch": 1.2246487689525665,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00012701267308861208,
+      "loss": 0.7432,
+      "step": 2232
+    },
+    {
+      "epoch": 1.2252051745722632,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001269564558279048,
+      "loss": 0.6797,
+      "step": 2233
+    },
+    {
+      "epoch": 1.22576158019196,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0001269002293788685,
+      "loss": 0.7114,
+      "step": 2234
+    },
+    {
+      "epoch": 1.2263179858116566,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00012684399376066834,
+      "loss": 0.7546,
+      "step": 2235
+    },
+    {
+      "epoch": 1.2268743914313536,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00012678774899247273,
+      "loss": 0.7231,
+      "step": 2236
+    },
+    {
+      "epoch": 1.2274307970510503,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00012673149509345318,
+      "loss": 0.6955,
+      "step": 2237
+    },
+    {
+      "epoch": 1.227987202670747,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00012667523208278425,
+      "loss": 0.6672,
+      "step": 2238
+    },
+    {
+      "epoch": 1.2285436082904437,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00012661895997964368,
+      "loss": 0.7162,
+      "step": 2239
+    },
+    {
+      "epoch": 1.2291000139101405,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00012656267880321227,
+      "loss": 0.6993,
+      "step": 2240
+    },
+    {
+      "epoch": 1.2296564195298372,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00012650638857267393,
+      "loss": 0.6886,
+      "step": 2241
+    },
+    {
+      "epoch": 1.230212825149534,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00012645008930721563,
+      "loss": 0.7082,
+      "step": 2242
+    },
+    {
+      "epoch": 1.2307692307692308,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00012639378102602744,
+      "loss": 0.6476,
+      "step": 2243
+    },
+    {
+      "epoch": 1.2313256363889276,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001263374637483025,
+      "loss": 0.8026,
+      "step": 2244
+    },
+    {
+      "epoch": 1.2318820420086243,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00012628113749323702,
+      "loss": 0.7189,
+      "step": 2245
+    },
+    {
+      "epoch": 1.232438447628321,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00012622480228003025,
+      "loss": 0.6837,
+      "step": 2246
+    },
+    {
+      "epoch": 1.2329948532480177,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00012616845812788455,
+      "loss": 0.7705,
+      "step": 2247
+    },
+    {
+      "epoch": 1.2335512588677147,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00012611210505600524,
+      "loss": 0.7483,
+      "step": 2248
+    },
+    {
+      "epoch": 1.2341076644874114,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00012605574308360078,
+      "loss": 0.6461,
+      "step": 2249
+    },
+    {
+      "epoch": 1.234664070107108,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00012599937222988253,
+      "loss": 0.7263,
+      "step": 2250
+    },
+    {
+      "epoch": 1.2352204757268048,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00012594299251406503,
+      "loss": 0.7418,
+      "step": 2251
+    },
+    {
+      "epoch": 1.2357768813465015,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00012588660395536575,
+      "loss": 0.7898,
+      "step": 2252
+    },
+    {
+      "epoch": 1.2363332869661983,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00012583020657300518,
+      "loss": 0.8093,
+      "step": 2253
+    },
+    {
+      "epoch": 1.2368896925858952,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00012577380038620686,
+      "loss": 0.6003,
+      "step": 2254
+    },
+    {
+      "epoch": 1.237446098205592,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001257173854141972,
+      "loss": 0.7314,
+      "step": 2255
+    },
+    {
+      "epoch": 1.2380025038252886,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00012566096167620586,
+      "loss": 0.84,
+      "step": 2256
+    },
+    {
+      "epoch": 1.2385589094449854,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00012560452919146522,
+      "loss": 0.7701,
+      "step": 2257
+    },
+    {
+      "epoch": 1.239115315064682,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00012554808797921084,
+      "loss": 0.7251,
+      "step": 2258
+    },
+    {
+      "epoch": 1.2396717206843788,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00012549163805868107,
+      "loss": 0.6146,
+      "step": 2259
+    },
+    {
+      "epoch": 1.2402281263040758,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001254351794491174,
+      "loss": 0.6198,
+      "step": 2260
+    },
+    {
+      "epoch": 1.2407845319237725,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00012537871216976423,
+      "loss": 0.7025,
+      "step": 2261
+    },
+    {
+      "epoch": 1.2413409375434692,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00012532223623986882,
+      "loss": 0.7221,
+      "step": 2262
+    },
+    {
+      "epoch": 1.241897343163166,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00012526575167868146,
+      "loss": 0.7848,
+      "step": 2263
+    },
+    {
+      "epoch": 1.2424537487828626,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00012520925850545547,
+      "loss": 0.738,
+      "step": 2264
+    },
+    {
+      "epoch": 1.2430101544025596,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00012515275673944691,
+      "loss": 0.7456,
+      "step": 2265
+    },
+    {
+      "epoch": 1.2435665600222563,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00012509624639991493,
+      "loss": 0.7308,
+      "step": 2266
+    },
+    {
+      "epoch": 1.244122965641953,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00012503972750612153,
+      "loss": 0.727,
+      "step": 2267
+    },
+    {
+      "epoch": 1.2446793712616497,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001249832000773316,
+      "loss": 0.688,
+      "step": 2268
+    },
+    {
+      "epoch": 1.2452357768813465,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.000124926664132813,
+      "loss": 0.7602,
+      "step": 2269
+    },
+    {
+      "epoch": 1.2457921825010432,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00012487011969183654,
+      "loss": 0.694,
+      "step": 2270
+    },
+    {
+      "epoch": 1.24634858812074,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00012481356677367572,
+      "loss": 0.6497,
+      "step": 2271
+    },
+    {
+      "epoch": 1.2469049937404368,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001247570053976072,
+      "loss": 0.7003,
+      "step": 2272
+    },
+    {
+      "epoch": 1.2474613993601336,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001247004355829103,
+      "loss": 0.6852,
+      "step": 2273
+    },
+    {
+      "epoch": 1.2480178049798303,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00012464385734886735,
+      "loss": 0.6751,
+      "step": 2274
+    },
+    {
+      "epoch": 1.248574210599527,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00012458727071476346,
+      "loss": 0.7585,
+      "step": 2275
+    },
+    {
+      "epoch": 1.2491306162192237,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00012453067569988672,
+      "loss": 0.6429,
+      "step": 2276
+    },
+    {
+      "epoch": 1.2496870218389207,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00012447407232352796,
+      "loss": 0.7197,
+      "step": 2277
+    },
+    {
+      "epoch": 1.2502434274586174,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001244174606049809,
+      "loss": 0.7096,
+      "step": 2278
+    },
+    {
+      "epoch": 1.250799833078314,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00012436084056354212,
+      "loss": 0.6943,
+      "step": 2279
+    },
+    {
+      "epoch": 1.2513562386980108,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00012430421221851101,
+      "loss": 0.7557,
+      "step": 2280
+    },
+    {
+      "epoch": 1.2519126443177075,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00012424757558918983,
+      "loss": 0.7623,
+      "step": 2281
+    },
+    {
+      "epoch": 1.2524690499374045,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00012419093069488364,
+      "loss": 0.7014,
+      "step": 2282
+    },
+    {
+      "epoch": 1.2530254555571012,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00012413427755490026,
+      "loss": 0.6064,
+      "step": 2283
+    },
+    {
+      "epoch": 1.253581861176798,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00012407761618855046,
+      "loss": 0.7065,
+      "step": 2284
+    },
+    {
+      "epoch": 1.2541382667964946,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001240209466151477,
+      "loss": 0.7589,
+      "step": 2285
+    },
+    {
+      "epoch": 1.2546946724161914,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00012396426885400824,
+      "loss": 0.6115,
+      "step": 2286
+    },
+    {
+      "epoch": 1.255251078035888,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0001239075829244512,
+      "loss": 0.6757,
+      "step": 2287
+    },
+    {
+      "epoch": 1.2558074836555848,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001238508888457984,
+      "loss": 0.71,
+      "step": 2288
+    },
+    {
+      "epoch": 1.2563638892752818,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00012379418663737454,
+      "loss": 0.7367,
+      "step": 2289
+    },
+    {
+      "epoch": 1.2569202948949785,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00012373747631850697,
+      "loss": 0.7003,
+      "step": 2290
+    },
+    {
+      "epoch": 1.2574767005146752,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00012368075790852596,
+      "loss": 0.8202,
+      "step": 2291
+    },
+    {
+      "epoch": 1.258033106134372,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00012362403142676432,
+      "loss": 0.6014,
+      "step": 2292
+    },
+    {
+      "epoch": 1.2585895117540686,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001235672968925578,
+      "loss": 0.7264,
+      "step": 2293
+    },
+    {
+      "epoch": 1.2591459173737656,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00012351055432524487,
+      "loss": 0.7665,
+      "step": 2294
+    },
+    {
+      "epoch": 1.2597023229934623,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00012345380374416667,
+      "loss": 0.7192,
+      "step": 2295
+    },
+    {
+      "epoch": 1.260258728613159,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0001233970451686671,
+      "loss": 0.534,
+      "step": 2296
+    },
+    {
+      "epoch": 1.2608151342328557,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00012334027861809273,
+      "loss": 0.6783,
+      "step": 2297
+    },
+    {
+      "epoch": 1.2613715398525525,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.000123283504111793,
+      "loss": 0.6038,
+      "step": 2298
+    },
+    {
+      "epoch": 1.2619279454722494,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00012322672166911988,
+      "loss": 0.7733,
+      "step": 2299
+    },
+    {
+      "epoch": 1.262484351091946,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00012316993130942816,
+      "loss": 0.6894,
+      "step": 2300
+    },
+    {
+      "epoch": 1.2630407567116428,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00012311313305207535,
+      "loss": 0.7264,
+      "step": 2301
+    },
+    {
+      "epoch": 1.2635971623313396,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00012305632691642155,
+      "loss": 0.6621,
+      "step": 2302
+    },
+    {
+      "epoch": 1.2641535679510363,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0001229995129218296,
+      "loss": 0.8092,
+      "step": 2303
+    },
+    {
+      "epoch": 1.264709973570733,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00012294269108766503,
+      "loss": 0.6325,
+      "step": 2304
+    },
+    {
+      "epoch": 1.2652663791904297,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00012288586143329605,
+      "loss": 0.6751,
+      "step": 2305
+    },
+    {
+      "epoch": 1.2658227848101267,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00012282902397809344,
+      "loss": 0.6358,
+      "step": 2306
+    },
+    {
+      "epoch": 1.2663791904298234,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00012277217874143078,
+      "loss": 0.669,
+      "step": 2307
+    },
+    {
+      "epoch": 1.26693559604952,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00012271532574268423,
+      "loss": 0.6807,
+      "step": 2308
+    },
+    {
+      "epoch": 1.2674920016692168,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001226584650012326,
+      "loss": 0.7123,
+      "step": 2309
+    },
+    {
+      "epoch": 1.2680484072889135,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001226015965364573,
+      "loss": 0.6789,
+      "step": 2310
+    },
+    {
+      "epoch": 1.2686048129086105,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00012254472036774244,
+      "loss": 0.8336,
+      "step": 2311
+    },
+    {
+      "epoch": 1.2691612185283072,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0001224878365144747,
+      "loss": 0.7687,
+      "step": 2312
+    },
+    {
+      "epoch": 1.269717624148004,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00012243094499604347,
+      "loss": 0.8454,
+      "step": 2313
+    },
+    {
+      "epoch": 1.2702740297677007,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00012237404583184067,
+      "loss": 0.682,
+      "step": 2314
+    },
+    {
+      "epoch": 1.2708304353873974,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00012231713904126083,
+      "loss": 0.7317,
+      "step": 2315
+    },
+    {
+      "epoch": 1.271386841007094,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00012226022464370106,
+      "loss": 0.7626,
+      "step": 2316
+    },
+    {
+      "epoch": 1.2719432466267908,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00012220330265856117,
+      "loss": 0.6676,
+      "step": 2317
+    },
+    {
+      "epoch": 1.2724996522464878,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00012214637310524346,
+      "loss": 0.6812,
+      "step": 2318
+    },
+    {
+      "epoch": 1.2730560578661845,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001220894360031528,
+      "loss": 0.6828,
+      "step": 2319
+    },
+    {
+      "epoch": 1.2736124634858812,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00012203249137169672,
+      "loss": 0.7557,
+      "step": 2320
+    },
+    {
+      "epoch": 1.274168869105578,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00012197553923028524,
+      "loss": 0.7348,
+      "step": 2321
+    },
+    {
+      "epoch": 1.2747252747252746,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00012191857959833097,
+      "loss": 0.704,
+      "step": 2322
+    },
+    {
+      "epoch": 1.2752816803449716,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00012186161249524908,
+      "loss": 0.7566,
+      "step": 2323
+    },
+    {
+      "epoch": 1.2758380859646683,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00012180463794045724,
+      "loss": 0.6965,
+      "step": 2324
+    },
+    {
+      "epoch": 1.276394491584365,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00012174765595337572,
+      "loss": 0.7508,
+      "step": 2325
+    },
+    {
+      "epoch": 1.2769508972040617,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00012169066655342729,
+      "loss": 0.7656,
+      "step": 2326
+    },
+    {
+      "epoch": 1.2775073028237585,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00012163366976003727,
+      "loss": 0.7445,
+      "step": 2327
+    },
+    {
+      "epoch": 1.2780637084434554,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00012157666559263345,
+      "loss": 0.6856,
+      "step": 2328
+    },
+    {
+      "epoch": 1.278620114063152,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00012151965407064618,
+      "loss": 0.731,
+      "step": 2329
+    },
+    {
+      "epoch": 1.2791765196828488,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00012146263521350831,
+      "loss": 0.6502,
+      "step": 2330
+    },
+    {
+      "epoch": 1.2797329253025456,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00012140560904065521,
+      "loss": 0.733,
+      "step": 2331
+    },
+    {
+      "epoch": 1.2802893309222423,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001213485755715247,
+      "loss": 0.6601,
+      "step": 2332
+    },
+    {
+      "epoch": 1.280845736541939,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00012129153482555708,
+      "loss": 0.8235,
+      "step": 2333
+    },
+    {
+      "epoch": 1.2814021421616357,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001212344868221952,
+      "loss": 0.7273,
+      "step": 2334
+    },
+    {
+      "epoch": 1.2819585477813327,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00012117743158088431,
+      "loss": 0.7548,
+      "step": 2335
+    },
+    {
+      "epoch": 1.2825149534010294,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00012112036912107216,
+      "loss": 0.7111,
+      "step": 2336
+    },
+    {
+      "epoch": 1.283071359020726,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00012106329946220898,
+      "loss": 0.7864,
+      "step": 2337
+    },
+    {
+      "epoch": 1.2836277646404228,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00012100622262374741,
+      "loss": 0.6583,
+      "step": 2338
+    },
+    {
+      "epoch": 1.2841841702601196,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00012094913862514257,
+      "loss": 0.7486,
+      "step": 2339
+    },
+    {
+      "epoch": 1.2847405758798165,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00012089204748585197,
+      "loss": 0.6553,
+      "step": 2340
+    },
+    {
+      "epoch": 1.2852969814995132,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00012083494922533562,
+      "loss": 0.7516,
+      "step": 2341
+    },
+    {
+      "epoch": 1.28585338711921,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00012077784386305597,
+      "loss": 0.7478,
+      "step": 2342
+    },
+    {
+      "epoch": 1.2864097927389067,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00012072073141847782,
+      "loss": 0.8059,
+      "step": 2343
+    },
+    {
+      "epoch": 1.2869661983586034,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00012066361191106841,
+      "loss": 0.7272,
+      "step": 2344
+    },
+    {
+      "epoch": 1.2875226039783003,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00012060648536029735,
+      "loss": 0.7071,
+      "step": 2345
+    },
+    {
+      "epoch": 1.2880790095979968,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00012054935178563678,
+      "loss": 0.7131,
+      "step": 2346
+    },
+    {
+      "epoch": 1.2886354152176938,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001204922112065611,
+      "loss": 0.691,
+      "step": 2347
+    },
+    {
+      "epoch": 1.2891918208373905,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00012043506364254715,
+      "loss": 0.761,
+      "step": 2348
+    },
+    {
+      "epoch": 1.2897482264570872,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00012037790911307412,
+      "loss": 0.7459,
+      "step": 2349
+    },
+    {
+      "epoch": 1.290304632076784,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00012032074763762367,
+      "loss": 0.7152,
+      "step": 2350
+    },
+    {
+      "epoch": 1.2908610376964806,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00012026357923567968,
+      "loss": 0.7651,
+      "step": 2351
+    },
+    {
+      "epoch": 1.2914174433161776,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00012020640392672853,
+      "loss": 0.708,
+      "step": 2352
+    },
+    {
+      "epoch": 1.2919738489358743,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00012014922173025885,
+      "loss": 0.7738,
+      "step": 2353
+    },
+    {
+      "epoch": 1.292530254555571,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001200920326657617,
+      "loss": 0.6697,
+      "step": 2354
+    },
+    {
+      "epoch": 1.2930866601752677,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001200348367527304,
+      "loss": 0.7044,
+      "step": 2355
+    },
+    {
+      "epoch": 1.2936430657949645,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00011997763401066067,
+      "loss": 0.6637,
+      "step": 2356
+    },
+    {
+      "epoch": 1.2941994714146614,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001199204244590505,
+      "loss": 0.7396,
+      "step": 2357
+    },
+    {
+      "epoch": 1.2947558770343581,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00011986320811740028,
+      "loss": 0.7384,
+      "step": 2358
+    },
+    {
+      "epoch": 1.2953122826540548,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00011980598500521267,
+      "loss": 0.6891,
+      "step": 2359
+    },
+    {
+      "epoch": 1.2958686882737516,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001197487551419926,
+      "loss": 0.6816,
+      "step": 2360
+    },
+    {
+      "epoch": 1.2964250938934483,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00011969151854724737,
+      "loss": 0.7689,
+      "step": 2361
+    },
+    {
+      "epoch": 1.296981499513145,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00011963427524048652,
+      "loss": 0.7527,
+      "step": 2362
+    },
+    {
+      "epoch": 1.2975379051328417,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00011957702524122192,
+      "loss": 0.665,
+      "step": 2363
+    },
+    {
+      "epoch": 1.2980943107525387,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001195197685689677,
+      "loss": 0.7169,
+      "step": 2364
+    },
+    {
+      "epoch": 1.2986507163722354,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00011946250524324023,
+      "loss": 0.7796,
+      "step": 2365
+    },
+    {
+      "epoch": 1.2992071219919321,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00011940523528355825,
+      "loss": 0.6301,
+      "step": 2366
+    },
+    {
+      "epoch": 1.2997635276116288,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00011934795870944264,
+      "loss": 0.7309,
+      "step": 2367
+    },
+    {
+      "epoch": 1.3003199332313256,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00011929067554041662,
+      "loss": 0.7462,
+      "step": 2368
+    },
+    {
+      "epoch": 1.3008763388510225,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00011923338579600563,
+      "loss": 0.6727,
+      "step": 2369
+    },
+    {
+      "epoch": 1.3014327444707192,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00011917608949573729,
+      "loss": 0.7083,
+      "step": 2370
+    },
+    {
+      "epoch": 1.301989150090416,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00011911878665914158,
+      "loss": 0.7418,
+      "step": 2371
+    },
+    {
+      "epoch": 1.3025455557101127,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00011906147730575063,
+      "loss": 0.7167,
+      "step": 2372
+    },
+    {
+      "epoch": 1.3031019613298094,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001190041614550988,
+      "loss": 0.7078,
+      "step": 2373
+    },
+    {
+      "epoch": 1.3036583669495063,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00011894683912672267,
+      "loss": 0.7599,
+      "step": 2374
+    },
+    {
+      "epoch": 1.3042147725692028,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.000118889510340161,
+      "loss": 0.6686,
+      "step": 2375
+    },
+    {
+      "epoch": 1.3047711781888998,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00011883217511495481,
+      "loss": 0.7635,
+      "step": 2376
+    },
+    {
+      "epoch": 1.3053275838085965,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00011877483347064726,
+      "loss": 0.6949,
+      "step": 2377
+    },
+    {
+      "epoch": 1.3058839894282932,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00011871748542678371,
+      "loss": 0.7317,
+      "step": 2378
+    },
+    {
+      "epoch": 1.30644039504799,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00011866013100291176,
+      "loss": 0.8254,
+      "step": 2379
+    },
+    {
+      "epoch": 1.3069968006676866,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00011860277021858111,
+      "loss": 0.7949,
+      "step": 2380
+    },
+    {
+      "epoch": 1.3075532062873836,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00011854540309334363,
+      "loss": 0.6833,
+      "step": 2381
+    },
+    {
+      "epoch": 1.3081096119070803,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0001184880296467534,
+      "loss": 0.8344,
+      "step": 2382
+    },
+    {
+      "epoch": 1.308666017526777,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00011843064989836662,
+      "loss": 0.7631,
+      "step": 2383
+    },
+    {
+      "epoch": 1.3092224231464737,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00011837326386774168,
+      "loss": 0.7569,
+      "step": 2384
+    },
+    {
+      "epoch": 1.3097788287661705,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00011831587157443901,
+      "loss": 0.7218,
+      "step": 2385
+    },
+    {
+      "epoch": 1.3103352343858674,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00011825847303802129,
+      "loss": 0.8403,
+      "step": 2386
+    },
+    {
+      "epoch": 1.3108916400055641,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00011820106827805328,
+      "loss": 0.6815,
+      "step": 2387
+    },
+    {
+      "epoch": 1.3114480456252608,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00011814365731410188,
+      "loss": 0.7281,
+      "step": 2388
+    },
+    {
+      "epoch": 1.3120044512449576,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00011808624016573606,
+      "loss": 0.7367,
+      "step": 2389
+    },
+    {
+      "epoch": 1.3125608568646543,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00011802881685252692,
+      "loss": 0.7303,
+      "step": 2390
+    },
+    {
+      "epoch": 1.313117262484351,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00011797138739404767,
+      "loss": 0.7195,
+      "step": 2391
+    },
+    {
+      "epoch": 1.3136736681040477,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00011791395180987365,
+      "loss": 0.6809,
+      "step": 2392
+    },
+    {
+      "epoch": 1.3142300737237447,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00011785651011958223,
+      "loss": 0.6217,
+      "step": 2393
+    },
+    {
+      "epoch": 1.3147864793434414,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00011779906234275287,
+      "loss": 0.7547,
+      "step": 2394
+    },
+    {
+      "epoch": 1.3153428849631381,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00011774160849896707,
+      "loss": 0.6672,
+      "step": 2395
+    },
+    {
+      "epoch": 1.3158992905828348,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00011768414860780856,
+      "loss": 0.7261,
+      "step": 2396
+    },
+    {
+      "epoch": 1.3164556962025316,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00011762668268886295,
+      "loss": 0.8287,
+      "step": 2397
+    },
+    {
+      "epoch": 1.3170121018222285,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00011756921076171795,
+      "loss": 0.7152,
+      "step": 2398
+    },
+    {
+      "epoch": 1.3175685074419252,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00011751173284596331,
+      "loss": 0.7445,
+      "step": 2399
+    },
+    {
+      "epoch": 1.318124913061622,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00011745424896119095,
+      "loss": 0.7196,
+      "step": 2400
+    },
+    {
+      "epoch": 1.3186813186813187,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00011739675912699466,
+      "loss": 0.708,
+      "step": 2401
+    },
+    {
+      "epoch": 1.3192377243010154,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00011733926336297032,
+      "loss": 0.6392,
+      "step": 2402
+    },
+    {
+      "epoch": 1.3197941299207123,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00011728176168871583,
+      "loss": 0.7001,
+      "step": 2403
+    },
+    {
+      "epoch": 1.3203505355404088,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00011722425412383115,
+      "loss": 0.7052,
+      "step": 2404
+    },
+    {
+      "epoch": 1.3209069411601058,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00011716674068791818,
+      "loss": 0.6777,
+      "step": 2405
+    },
+    {
+      "epoch": 1.3214633467798025,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00011710922140058083,
+      "loss": 0.7806,
+      "step": 2406
+    },
+    {
+      "epoch": 1.3220197523994992,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.000117051696281425,
+      "loss": 0.6538,
+      "step": 2407
+    },
+    {
+      "epoch": 1.322576158019196,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00011699416535005869,
+      "loss": 0.7291,
+      "step": 2408
+    },
+    {
+      "epoch": 1.3231325636388926,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00011693662862609171,
+      "loss": 0.8131,
+      "step": 2409
+    },
+    {
+      "epoch": 1.3236889692585896,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00011687908612913594,
+      "loss": 0.7145,
+      "step": 2410
+    },
+    {
+      "epoch": 1.3242453748782863,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00011682153787880523,
+      "loss": 0.715,
+      "step": 2411
+    },
+    {
+      "epoch": 1.324801780497983,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00011676398389471537,
+      "loss": 0.7063,
+      "step": 2412
+    },
+    {
+      "epoch": 1.3253581861176797,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00011670642419648408,
+      "loss": 0.7678,
+      "step": 2413
+    },
+    {
+      "epoch": 1.3259145917373765,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00011664885880373109,
+      "loss": 0.7466,
+      "step": 2414
+    },
+    {
+      "epoch": 1.3264709973570734,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00011659128773607799,
+      "loss": 0.7625,
+      "step": 2415
+    },
+    {
+      "epoch": 1.3270274029767701,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001165337110131484,
+      "loss": 0.7488,
+      "step": 2416
+    },
+    {
+      "epoch": 1.3275838085964669,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00011647612865456781,
+      "loss": 0.7182,
+      "step": 2417
+    },
+    {
+      "epoch": 1.3281402142161636,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0001164185406799636,
+      "loss": 0.8061,
+      "step": 2418
+    },
+    {
+      "epoch": 1.3286966198358603,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00011636094710896515,
+      "loss": 0.6612,
+      "step": 2419
+    },
+    {
+      "epoch": 1.3292530254555572,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00011630334796120367,
+      "loss": 0.7305,
+      "step": 2420
+    },
+    {
+      "epoch": 1.3298094310752537,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00011624574325631234,
+      "loss": 0.6024,
+      "step": 2421
+    },
+    {
+      "epoch": 1.3303658366949507,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00011618813301392619,
+      "loss": 0.7148,
+      "step": 2422
+    },
+    {
+      "epoch": 1.3309222423146474,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00011613051725368207,
+      "loss": 0.7477,
+      "step": 2423
+    },
+    {
+      "epoch": 1.3314786479343441,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00011607289599521887,
+      "loss": 0.7515,
+      "step": 2424
+    },
+    {
+      "epoch": 1.3320350535540408,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00011601526925817723,
+      "loss": 0.8687,
+      "step": 2425
+    },
+    {
+      "epoch": 1.3325914591737376,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00011595763706219975,
+      "loss": 0.6806,
+      "step": 2426
+    },
+    {
+      "epoch": 1.3331478647934345,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00011589999942693072,
+      "loss": 0.7935,
+      "step": 2427
+    },
+    {
+      "epoch": 1.3337042704131312,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001158423563720165,
+      "loss": 0.6766,
+      "step": 2428
+    },
+    {
+      "epoch": 1.334260676032828,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00011578470791710519,
+      "loss": 0.7671,
+      "step": 2429
+    },
+    {
+      "epoch": 1.3348170816525247,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00011572705408184669,
+      "loss": 1.6101,
+      "step": 2430
+    },
+    {
+      "epoch": 1.3353734872722214,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00011566939488589283,
+      "loss": 0.6784,
+      "step": 2431
+    },
+    {
+      "epoch": 1.3359298928919183,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00011561173034889717,
+      "loss": 0.7173,
+      "step": 2432
+    },
+    {
+      "epoch": 1.336486298511615,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001155540604905152,
+      "loss": 0.6968,
+      "step": 2433
+    },
+    {
+      "epoch": 1.3370427041313118,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00011549638533040412,
+      "loss": 0.8715,
+      "step": 2434
+    },
+    {
+      "epoch": 1.3375991097510085,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00011543870488822302,
+      "loss": 0.7,
+      "step": 2435
+    },
+    {
+      "epoch": 1.3381555153707052,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001153810191836327,
+      "loss": 0.7332,
+      "step": 2436
+    },
+    {
+      "epoch": 1.338711920990402,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00011532332823629583,
+      "loss": 0.6859,
+      "step": 2437
+    },
+    {
+      "epoch": 1.3392683266100986,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00011526563206587688,
+      "loss": 0.6342,
+      "step": 2438
+    },
+    {
+      "epoch": 1.3398247322297956,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00011520793069204201,
+      "loss": 0.7854,
+      "step": 2439
+    },
+    {
+      "epoch": 1.3403811378494923,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00011515022413445924,
+      "loss": 0.7628,
+      "step": 2440
+    },
+    {
+      "epoch": 1.340937543469189,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0001150925124127983,
+      "loss": 0.7843,
+      "step": 2441
+    },
+    {
+      "epoch": 1.3414939490888858,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00011503479554673072,
+      "loss": 0.7167,
+      "step": 2442
+    },
+    {
+      "epoch": 1.3420503547085825,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00011497707355592975,
+      "loss": 0.7199,
+      "step": 2443
+    },
+    {
+      "epoch": 1.3426067603282794,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0001149193464600704,
+      "loss": 0.7374,
+      "step": 2444
+    },
+    {
+      "epoch": 1.3431631659479761,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00011486161427882946,
+      "loss": 0.6563,
+      "step": 2445
+    },
+    {
+      "epoch": 1.3437195715676729,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001148038770318854,
+      "loss": 0.699,
+      "step": 2446
+    },
+    {
+      "epoch": 1.3442759771873696,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00011474613473891842,
+      "loss": 0.5989,
+      "step": 2447
+    },
+    {
+      "epoch": 1.3448323828070663,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00011468838741961045,
+      "loss": 0.6989,
+      "step": 2448
+    },
+    {
+      "epoch": 1.3453887884267632,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00011463063509364517,
+      "loss": 0.6952,
+      "step": 2449
+    },
+    {
+      "epoch": 1.3459451940464597,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00011457287778070793,
+      "loss": 0.7021,
+      "step": 2450
+    },
+    {
+      "epoch": 1.3465015996661567,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00011451511550048577,
+      "loss": 0.7343,
+      "step": 2451
+    },
+    {
+      "epoch": 1.3470580052858534,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00011445734827266738,
+      "loss": 0.8033,
+      "step": 2452
+    },
+    {
+      "epoch": 1.3476144109055501,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00011439957611694332,
+      "loss": 0.6597,
+      "step": 2453
+    },
+    {
+      "epoch": 1.3481708165252468,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00011434179905300559,
+      "loss": 0.7599,
+      "step": 2454
+    },
+    {
+      "epoch": 1.3487272221449436,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00011428401710054806,
+      "loss": 0.7468,
+      "step": 2455
+    },
+    {
+      "epoch": 1.3492836277646405,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00011422623027926612,
+      "loss": 0.712,
+      "step": 2456
+    },
+    {
+      "epoch": 1.3498400333843372,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00011416843860885688,
+      "loss": 0.6616,
+      "step": 2457
+    },
+    {
+      "epoch": 1.350396439004034,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00011411064210901916,
+      "loss": 0.7084,
+      "step": 2458
+    },
+    {
+      "epoch": 1.3509528446237307,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00011405284079945335,
+      "loss": 0.8171,
+      "step": 2459
+    },
+    {
+      "epoch": 1.3515092502434274,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00011399503469986144,
+      "loss": 0.8589,
+      "step": 2460
+    },
+    {
+      "epoch": 1.3520656558631243,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001139372238299472,
+      "loss": 0.6904,
+      "step": 2461
+    },
+    {
+      "epoch": 1.352622061482821,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001138794082094159,
+      "loss": 0.7881,
+      "step": 2462
+    },
+    {
+      "epoch": 1.3531784671025178,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00011382158785797448,
+      "loss": 0.7395,
+      "step": 2463
+    },
+    {
+      "epoch": 1.3537348727222145,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00011376376279533143,
+      "loss": 0.7064,
+      "step": 2464
+    },
+    {
+      "epoch": 1.3542912783419112,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.000113705933041197,
+      "loss": 0.7418,
+      "step": 2465
+    },
+    {
+      "epoch": 1.354847683961608,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00011364809861528283,
+      "loss": 0.585,
+      "step": 2466
+    },
+    {
+      "epoch": 1.3554040895813046,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00011359025953730231,
+      "loss": 0.7384,
+      "step": 2467
+    },
+    {
+      "epoch": 1.3559604952010016,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00011353241582697039,
+      "loss": 0.6822,
+      "step": 2468
+    },
+    {
+      "epoch": 1.3565169008206983,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00011347456750400351,
+      "loss": 0.7019,
+      "step": 2469
+    },
+    {
+      "epoch": 1.357073306440395,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00011341671458811978,
+      "loss": 0.7273,
+      "step": 2470
+    },
+    {
+      "epoch": 1.3576297120600918,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00011335885709903885,
+      "loss": 0.7323,
+      "step": 2471
+    },
+    {
+      "epoch": 1.3581861176797885,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001133009950564819,
+      "loss": 0.704,
+      "step": 2472
+    },
+    {
+      "epoch": 1.3587425232994854,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00011324312848017166,
+      "loss": 0.6183,
+      "step": 2473
+    },
+    {
+      "epoch": 1.3592989289191821,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001131852573898325,
+      "loss": 0.7105,
+      "step": 2474
+    },
+    {
+      "epoch": 1.3598553345388789,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00011312738180519018,
+      "loss": 0.6982,
+      "step": 2475
+    },
+    {
+      "epoch": 1.3604117401585756,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00011306950174597213,
+      "loss": 0.5876,
+      "step": 2476
+    },
+    {
+      "epoch": 1.3609681457782723,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00011301161723190717,
+      "loss": 0.7142,
+      "step": 2477
+    },
+    {
+      "epoch": 1.3615245513979692,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00011295372828272577,
+      "loss": 0.7409,
+      "step": 2478
+    },
+    {
+      "epoch": 1.362080957017666,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00011289583491815985,
+      "loss": 0.7466,
+      "step": 2479
+    },
+    {
+      "epoch": 1.3626373626373627,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00011283793715794282,
+      "loss": 0.6091,
+      "step": 2480
+    },
+    {
+      "epoch": 1.3631937682570594,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001127800350218096,
+      "loss": 0.7041,
+      "step": 2481
+    },
+    {
+      "epoch": 1.3637501738767561,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00011272212852949664,
+      "loss": 0.6647,
+      "step": 2482
+    },
+    {
+      "epoch": 1.3643065794964528,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001126642177007418,
+      "loss": 0.7769,
+      "step": 2483
+    },
+    {
+      "epoch": 1.3648629851161496,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00011260630255528451,
+      "loss": 0.6792,
+      "step": 2484
+    },
+    {
+      "epoch": 1.3654193907358465,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00011254838311286558,
+      "loss": 0.8129,
+      "step": 2485
+    },
+    {
+      "epoch": 1.3659757963555432,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00011249045939322736,
+      "loss": 0.7495,
+      "step": 2486
+    },
+    {
+      "epoch": 1.36653220197524,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00011243253141611358,
+      "loss": 0.7184,
+      "step": 2487
+    },
+    {
+      "epoch": 1.3670886075949367,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00011237459920126951,
+      "loss": 0.6775,
+      "step": 2488
+    },
+    {
+      "epoch": 1.3676450132146334,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00011231666276844177,
+      "loss": 0.6989,
+      "step": 2489
+    },
+    {
+      "epoch": 1.3682014188343303,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00011225872213737854,
+      "loss": 0.6135,
+      "step": 2490
+    },
+    {
+      "epoch": 1.368757824454027,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00011220077732782932,
+      "loss": 0.782,
+      "step": 2491
+    },
+    {
+      "epoch": 1.3693142300737238,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00011214282835954509,
+      "loss": 0.6541,
+      "step": 2492
+    },
+    {
+      "epoch": 1.3698706356934205,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00011208487525227816,
+      "loss": 0.6114,
+      "step": 2493
+    },
+    {
+      "epoch": 1.3704270413131172,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00011202691802578243,
+      "loss": 0.7009,
+      "step": 2494
+    },
+    {
+      "epoch": 1.3709834469328142,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00011196895669981302,
+      "loss": 0.6033,
+      "step": 2495
+    },
+    {
+      "epoch": 1.3715398525525107,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00011191099129412657,
+      "loss": 0.7491,
+      "step": 2496
+    },
+    {
+      "epoch": 1.3720962581722076,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00011185302182848101,
+      "loss": 0.7113,
+      "step": 2497
+    },
+    {
+      "epoch": 1.3726526637919043,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00011179504832263577,
+      "loss": 0.6992,
+      "step": 2498
+    },
+    {
+      "epoch": 1.373209069411601,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00011173707079635156,
+      "loss": 0.7619,
+      "step": 2499
+    },
+    {
+      "epoch": 1.3737654750312978,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0001116790892693905,
+      "loss": 0.7508,
+      "step": 2500
+    },
+    {
+      "epoch": 1.3743218806509945,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00011162110376151604,
+      "loss": 0.7367,
+      "step": 2501
+    },
+    {
+      "epoch": 1.3748782862706914,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00011156311429249306,
+      "loss": 0.7643,
+      "step": 2502
+    },
+    {
+      "epoch": 1.3754346918903881,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00011150512088208776,
+      "loss": 0.793,
+      "step": 2503
+    },
+    {
+      "epoch": 1.3759910975100849,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00011144712355006762,
+      "loss": 0.7647,
+      "step": 2504
+    },
+    {
+      "epoch": 1.3765475031297816,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00011138912231620153,
+      "loss": 0.672,
+      "step": 2505
+    },
+    {
+      "epoch": 1.3771039087494783,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00011133111720025966,
+      "loss": 0.7259,
+      "step": 2506
+    },
+    {
+      "epoch": 1.3776603143691752,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00011127310822201358,
+      "loss": 0.5997,
+      "step": 2507
+    },
+    {
+      "epoch": 1.378216719988872,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00011121509540123609,
+      "loss": 0.7158,
+      "step": 2508
+    },
+    {
+      "epoch": 1.3787731256085687,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00011115707875770134,
+      "loss": 0.7551,
+      "step": 2509
+    },
+    {
+      "epoch": 1.3793295312282654,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001110990583111848,
+      "loss": 0.7265,
+      "step": 2510
+    },
+    {
+      "epoch": 1.3798859368479621,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00011104103408146318,
+      "loss": 0.8133,
+      "step": 2511
+    },
+    {
+      "epoch": 1.3804423424676588,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00011098300608831454,
+      "loss": 0.7369,
+      "step": 2512
+    },
+    {
+      "epoch": 1.3809987480873556,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00011092497435151818,
+      "loss": 0.5932,
+      "step": 2513
+    },
+    {
+      "epoch": 1.3815551537070525,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00011086693889085473,
+      "loss": 0.6975,
+      "step": 2514
+    },
+    {
+      "epoch": 1.3821115593267492,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00011080889972610602,
+      "loss": 0.6966,
+      "step": 2515
+    },
+    {
+      "epoch": 1.382667964946446,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00011075085687705517,
+      "loss": 0.7353,
+      "step": 2516
+    },
+    {
+      "epoch": 1.3832243705661427,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001106928103634866,
+      "loss": 0.6437,
+      "step": 2517
+    },
+    {
+      "epoch": 1.3837807761858394,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00011063476020518588,
+      "loss": 0.661,
+      "step": 2518
+    },
+    {
+      "epoch": 1.3843371818055363,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00011057670642193994,
+      "loss": 0.7933,
+      "step": 2519
+    },
+    {
+      "epoch": 1.384893587425233,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00011051864903353686,
+      "loss": 0.6854,
+      "step": 2520
+    },
+    {
+      "epoch": 1.3854499930449298,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00011046058805976599,
+      "loss": 0.7005,
+      "step": 2521
+    },
+    {
+      "epoch": 1.3860063986646265,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00011040252352041782,
+      "loss": 0.6934,
+      "step": 2522
+    },
+    {
+      "epoch": 1.3865628042843232,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00011034445543528424,
+      "loss": 0.6935,
+      "step": 2523
+    },
+    {
+      "epoch": 1.3871192099040202,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00011028638382415819,
+      "loss": 0.7408,
+      "step": 2524
+    },
+    {
+      "epoch": 1.3876756155237167,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00011022830870683382,
+      "loss": 0.7131,
+      "step": 2525
+    },
+    {
+      "epoch": 1.3882320211434136,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00011017023010310654,
+      "loss": 0.7135,
+      "step": 2526
+    },
+    {
+      "epoch": 1.3887884267631103,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00011011214803277292,
+      "loss": 0.7208,
+      "step": 2527
+    },
+    {
+      "epoch": 1.389344832382807,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00011005406251563068,
+      "loss": 0.7367,
+      "step": 2528
+    },
+    {
+      "epoch": 1.3899012380025038,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001099959735714788,
+      "loss": 0.6008,
+      "step": 2529
+    },
+    {
+      "epoch": 1.3904576436222005,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0001099378812201173,
+      "loss": 0.6261,
+      "step": 2530
+    },
+    {
+      "epoch": 1.3910140492418974,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001098797854813475,
+      "loss": 0.7666,
+      "step": 2531
+    },
+    {
+      "epoch": 1.3915704548615941,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00010982168637497177,
+      "loss": 0.6456,
+      "step": 2532
+    },
+    {
+      "epoch": 1.3921268604812909,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00010976358392079369,
+      "loss": 0.7901,
+      "step": 2533
+    },
+    {
+      "epoch": 1.3926832661009876,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00010970547813861794,
+      "loss": 0.7,
+      "step": 2534
+    },
+    {
+      "epoch": 1.3932396717206843,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00010964736904825035,
+      "loss": 0.6445,
+      "step": 2535
+    },
+    {
+      "epoch": 1.3937960773403812,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00010958925666949791,
+      "loss": 0.6889,
+      "step": 2536
+    },
+    {
+      "epoch": 1.394352482960078,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00010953114102216868,
+      "loss": 0.687,
+      "step": 2537
+    },
+    {
+      "epoch": 1.3949088885797747,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00010947302212607186,
+      "loss": 0.658,
+      "step": 2538
+    },
+    {
+      "epoch": 1.3954652941994714,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00010941490000101773,
+      "loss": 0.7327,
+      "step": 2539
+    },
+    {
+      "epoch": 1.3960216998191681,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00010935677466681773,
+      "loss": 0.6706,
+      "step": 2540
+    },
+    {
+      "epoch": 1.396578105438865,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00010929864614328435,
+      "loss": 0.7794,
+      "step": 2541
+    },
+    {
+      "epoch": 1.3971345110585616,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00010924051445023116,
+      "loss": 0.75,
+      "step": 2542
+    },
+    {
+      "epoch": 1.3976909166782585,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00010918237960747281,
+      "loss": 0.7045,
+      "step": 2543
+    },
+    {
+      "epoch": 1.3982473222979552,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00010912424163482506,
+      "loss": 0.6658,
+      "step": 2544
+    },
+    {
+      "epoch": 1.398803727917652,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00010906610055210473,
+      "loss": 0.8038,
+      "step": 2545
+    },
+    {
+      "epoch": 1.3993601335373487,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00010900795637912967,
+      "loss": 0.6971,
+      "step": 2546
+    },
+    {
+      "epoch": 1.3999165391570454,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00010894980913571874,
+      "loss": 0.7513,
+      "step": 2547
+    },
+    {
+      "epoch": 1.4004729447767423,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00010889165884169199,
+      "loss": 0.7808,
+      "step": 2548
+    },
+    {
+      "epoch": 1.401029350396439,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00010883350551687042,
+      "loss": 0.668,
+      "step": 2549
+    },
+    {
+      "epoch": 1.4015857560161358,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00010877534918107601,
+      "loss": 0.6883,
+      "step": 2550
+    },
+    {
+      "epoch": 1.4021421616358325,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00010871718985413184,
+      "loss": 0.7308,
+      "step": 2551
+    },
+    {
+      "epoch": 1.4026985672555292,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00010865902755586204,
+      "loss": 0.7401,
+      "step": 2552
+    },
+    {
+      "epoch": 1.4032549728752262,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00010860086230609168,
+      "loss": 0.6671,
+      "step": 2553
+    },
+    {
+      "epoch": 1.4038113784949229,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00010854269412464683,
+      "loss": 0.787,
+      "step": 2554
+    },
+    {
+      "epoch": 1.4043677841146196,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00010848452303135464,
+      "loss": 0.7039,
+      "step": 2555
+    },
+    {
+      "epoch": 1.4049241897343163,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00010842634904604319,
+      "loss": 0.7197,
+      "step": 2556
+    },
+    {
+      "epoch": 1.405480595354013,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00010836817218854155,
+      "loss": 0.6888,
+      "step": 2557
+    },
+    {
+      "epoch": 1.4060370009737098,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00010830999247867976,
+      "loss": 0.7686,
+      "step": 2558
+    },
+    {
+      "epoch": 1.4065934065934065,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00010825180993628887,
+      "loss": 0.7198,
+      "step": 2559
+    },
+    {
+      "epoch": 1.4071498122131034,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001081936245812009,
+      "loss": 0.7277,
+      "step": 2560
+    },
+    {
+      "epoch": 1.4077062178328001,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00010813543643324879,
+      "loss": 0.7303,
+      "step": 2561
+    },
+    {
+      "epoch": 1.4082626234524969,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00010807724551226643,
+      "loss": 0.6663,
+      "step": 2562
+    },
+    {
+      "epoch": 1.4088190290721936,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0001080190518380887,
+      "loss": 0.7747,
+      "step": 2563
+    },
+    {
+      "epoch": 1.4093754346918903,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00010796085543055136,
+      "loss": 0.7035,
+      "step": 2564
+    },
+    {
+      "epoch": 1.4099318403115872,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00010790265630949117,
+      "loss": 0.6838,
+      "step": 2565
+    },
+    {
+      "epoch": 1.410488245931284,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00010784445449474576,
+      "loss": 0.725,
+      "step": 2566
+    },
+    {
+      "epoch": 1.4110446515509807,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00010778625000615365,
+      "loss": 0.7595,
+      "step": 2567
+    },
+    {
+      "epoch": 1.4116010571706774,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001077280428635544,
+      "loss": 0.6964,
+      "step": 2568
+    },
+    {
+      "epoch": 1.4121574627903741,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00010766983308678837,
+      "loss": 0.7032,
+      "step": 2569
+    },
+    {
+      "epoch": 1.412713868410071,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00010761162069569681,
+      "loss": 0.7463,
+      "step": 2570
+    },
+    {
+      "epoch": 1.4132702740297676,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00010755340571012189,
+      "loss": 0.6686,
+      "step": 2571
+    },
+    {
+      "epoch": 1.4138266796494645,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00010749518814990669,
+      "loss": 0.7457,
+      "step": 2572
+    },
+    {
+      "epoch": 1.4143830852691612,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00010743696803489514,
+      "loss": 0.6991,
+      "step": 2573
+    },
+    {
+      "epoch": 1.414939490888858,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00010737874538493203,
+      "loss": 0.7904,
+      "step": 2574
+    },
+    {
+      "epoch": 1.4154958965085547,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00010732052021986303,
+      "loss": 0.6694,
+      "step": 2575
+    },
+    {
+      "epoch": 1.4160523021282514,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00010726229255953467,
+      "loss": 0.7395,
+      "step": 2576
+    },
+    {
+      "epoch": 1.4166087077479483,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00010720406242379432,
+      "loss": 0.7902,
+      "step": 2577
+    },
+    {
+      "epoch": 1.417165113367645,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00010714582983249022,
+      "loss": 0.7543,
+      "step": 2578
+    },
+    {
+      "epoch": 1.4177215189873418,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0001070875948054714,
+      "loss": 0.8445,
+      "step": 2579
+    },
+    {
+      "epoch": 1.4182779246070385,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00010702935736258774,
+      "loss": 0.6908,
+      "step": 2580
+    },
+    {
+      "epoch": 1.4188343302267352,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00010697111752368997,
+      "loss": 0.7227,
+      "step": 2581
+    },
+    {
+      "epoch": 1.4193907358464322,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00010691287530862965,
+      "loss": 0.6442,
+      "step": 2582
+    },
+    {
+      "epoch": 1.4199471414661289,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00010685463073725906,
+      "loss": 0.6823,
+      "step": 2583
+    },
+    {
+      "epoch": 1.4205035470858256,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00010679638382943136,
+      "loss": 0.6991,
+      "step": 2584
+    },
+    {
+      "epoch": 1.4210599527055223,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00010673813460500048,
+      "loss": 0.5955,
+      "step": 2585
+    },
+    {
+      "epoch": 1.421616358325219,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00010667988308382117,
+      "loss": 0.7899,
+      "step": 2586
+    },
+    {
+      "epoch": 1.4221727639449158,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00010662162928574892,
+      "loss": 0.6611,
+      "step": 2587
+    },
+    {
+      "epoch": 1.4227291695646125,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00010656337323063999,
+      "loss": 0.7835,
+      "step": 2588
+    },
+    {
+      "epoch": 1.4232855751843094,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00010650511493835149,
+      "loss": 0.6867,
+      "step": 2589
+    },
+    {
+      "epoch": 1.4238419808040061,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00010644685442874118,
+      "loss": 0.7198,
+      "step": 2590
+    },
+    {
+      "epoch": 1.4243983864237029,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00010638859172166769,
+      "loss": 0.6852,
+      "step": 2591
+    },
+    {
+      "epoch": 1.4249547920433996,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00010633032683699029,
+      "loss": 0.6569,
+      "step": 2592
+    },
+    {
+      "epoch": 1.4255111976630963,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00010627205979456906,
+      "loss": 0.6837,
+      "step": 2593
+    },
+    {
+      "epoch": 1.4260676032827933,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00010621379061426485,
+      "loss": 1.7361,
+      "step": 2594
+    },
+    {
+      "epoch": 1.42662400890249,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001061555193159391,
+      "loss": 0.6271,
+      "step": 2595
+    },
+    {
+      "epoch": 1.4271804145221867,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001060972459194541,
+      "loss": 0.6884,
+      "step": 2596
+    },
+    {
+      "epoch": 1.4277368201418834,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00010603897044467284,
+      "loss": 0.6821,
+      "step": 2597
+    },
+    {
+      "epoch": 1.4282932257615801,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00010598069291145896,
+      "loss": 0.7808,
+      "step": 2598
+    },
+    {
+      "epoch": 1.428849631381277,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00010592241333967685,
+      "loss": 0.7096,
+      "step": 2599
+    },
+    {
+      "epoch": 1.4294060370009736,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00010586413174919158,
+      "loss": 0.6982,
+      "step": 2600
+    },
+    {
+      "epoch": 1.4299624426206705,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00010580584815986891,
+      "loss": 0.6833,
+      "step": 2601
+    },
+    {
+      "epoch": 1.4305188482403672,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00010574756259157526,
+      "loss": 0.7476,
+      "step": 2602
+    },
+    {
+      "epoch": 1.431075253860064,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00010568927506417779,
+      "loss": 0.7539,
+      "step": 2603
+    },
+    {
+      "epoch": 1.4316316594797607,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00010563098559754419,
+      "loss": 0.7264,
+      "step": 2604
+    },
+    {
+      "epoch": 1.4321880650994574,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00010557269421154301,
+      "loss": 0.7798,
+      "step": 2605
+    },
+    {
+      "epoch": 1.4327444707191543,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00010551440092604328,
+      "loss": 0.7869,
+      "step": 2606
+    },
+    {
+      "epoch": 1.433300876338851,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00010545610576091478,
+      "loss": 0.6997,
+      "step": 2607
+    },
+    {
+      "epoch": 1.4338572819585478,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00010539780873602785,
+      "loss": 0.7397,
+      "step": 2608
+    },
+    {
+      "epoch": 1.4344136875782445,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00010533950987125356,
+      "loss": 0.6898,
+      "step": 2609
+    },
+    {
+      "epoch": 1.4349700931979412,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00010528120918646353,
+      "loss": 0.6294,
+      "step": 2610
+    },
+    {
+      "epoch": 1.4355264988176382,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00010522290670153005,
+      "loss": 0.604,
+      "step": 2611
+    },
+    {
+      "epoch": 1.4360829044373349,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00010516460243632597,
+      "loss": 0.7194,
+      "step": 2612
+    },
+    {
+      "epoch": 1.4366393100570316,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0001051062964107248,
+      "loss": 0.7471,
+      "step": 2613
+    },
+    {
+      "epoch": 1.4371957156767283,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00010504798864460062,
+      "loss": 0.7736,
+      "step": 2614
+    },
+    {
+      "epoch": 1.437752121296425,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00010498967915782809,
+      "loss": 0.7772,
+      "step": 2615
+    },
+    {
+      "epoch": 1.438308526916122,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00010493136797028249,
+      "loss": 0.7654,
+      "step": 2616
+    },
+    {
+      "epoch": 1.4388649325358185,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00010487305510183967,
+      "loss": 0.7171,
+      "step": 2617
+    },
+    {
+      "epoch": 1.4394213381555154,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00010481474057237604,
+      "loss": 0.7437,
+      "step": 2618
+    },
+    {
+      "epoch": 1.4399777437752121,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00010475642440176861,
+      "loss": 0.6849,
+      "step": 2619
+    },
+    {
+      "epoch": 1.4405341493949089,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00010469810660989492,
+      "loss": 0.6591,
+      "step": 2620
+    },
+    {
+      "epoch": 1.4410905550146056,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00010463978721663301,
+      "loss": 0.7568,
+      "step": 2621
+    },
+    {
+      "epoch": 1.4416469606343023,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00010458146624186163,
+      "loss": 0.6783,
+      "step": 2622
+    },
+    {
+      "epoch": 1.4422033662539993,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00010452314370545987,
+      "loss": 0.7343,
+      "step": 2623
+    },
+    {
+      "epoch": 1.442759771873696,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00010446481962730749,
+      "loss": 0.6946,
+      "step": 2624
+    },
+    {
+      "epoch": 1.4433161774933927,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0001044064940272847,
+      "loss": 0.7157,
+      "step": 2625
+    },
+    {
+      "epoch": 1.4438725831130894,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001043481669252723,
+      "loss": 0.7112,
+      "step": 2626
+    },
+    {
+      "epoch": 1.4444289887327861,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00010428983834115153,
+      "loss": 0.7378,
+      "step": 2627
+    },
+    {
+      "epoch": 1.444985394352483,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0001042315082948042,
+      "loss": 0.6738,
+      "step": 2628
+    },
+    {
+      "epoch": 1.4455417999721798,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00010417317680611254,
+      "loss": 0.692,
+      "step": 2629
+    },
+    {
+      "epoch": 1.4460982055918765,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00010411484389495934,
+      "loss": 0.6794,
+      "step": 2630
+    },
+    {
+      "epoch": 1.4466546112115732,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00010405650958122786,
+      "loss": 0.8288,
+      "step": 2631
+    },
+    {
+      "epoch": 1.44721101683127,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001039981738848018,
+      "loss": 0.6557,
+      "step": 2632
+    },
+    {
+      "epoch": 1.4477674224509667,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00010393983682556537,
+      "loss": 0.6802,
+      "step": 2633
+    },
+    {
+      "epoch": 1.4483238280706634,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00010388149842340325,
+      "loss": 0.7585,
+      "step": 2634
+    },
+    {
+      "epoch": 1.4488802336903603,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00010382315869820056,
+      "loss": 0.7667,
+      "step": 2635
+    },
+    {
+      "epoch": 1.449436639310057,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00010376481766984288,
+      "loss": 0.7618,
+      "step": 2636
+    },
+    {
+      "epoch": 1.4499930449297538,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00010370647535821618,
+      "loss": 0.7417,
+      "step": 2637
+    },
+    {
+      "epoch": 1.4505494505494505,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00010364813178320698,
+      "loss": 0.7083,
+      "step": 2638
+    },
+    {
+      "epoch": 1.4511058561691472,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00010358978696470214,
+      "loss": 0.7089,
+      "step": 2639
+    },
+    {
+      "epoch": 1.4516622617888442,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00010353144092258898,
+      "loss": 0.7021,
+      "step": 2640
+    },
+    {
+      "epoch": 1.4522186674085409,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00010347309367675517,
+      "loss": 0.7072,
+      "step": 2641
+    },
+    {
+      "epoch": 1.4527750730282376,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00010341474524708896,
+      "loss": 0.7622,
+      "step": 2642
+    },
+    {
+      "epoch": 1.4533314786479343,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00010335639565347882,
+      "loss": 0.6601,
+      "step": 2643
+    },
+    {
+      "epoch": 1.453887884267631,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00010329804491581365,
+      "loss": 0.7328,
+      "step": 2644
+    },
+    {
+      "epoch": 1.454444289887328,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00010323969305398282,
+      "loss": 0.7097,
+      "step": 2645
+    },
+    {
+      "epoch": 1.4550006955070245,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00010318134008787606,
+      "loss": 0.6524,
+      "step": 2646
+    },
+    {
+      "epoch": 1.4555571011267214,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00010312298603738342,
+      "loss": 0.6303,
+      "step": 2647
+    },
+    {
+      "epoch": 1.4561135067464182,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001030646309223954,
+      "loss": 0.708,
+      "step": 2648
+    },
+    {
+      "epoch": 1.4566699123661149,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00010300627476280278,
+      "loss": 0.7317,
+      "step": 2649
+    },
+    {
+      "epoch": 1.4572263179858116,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00010294791757849673,
+      "loss": 0.703,
+      "step": 2650
+    },
+    {
+      "epoch": 1.4577827236055083,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00010288955938936882,
+      "loss": 0.8404,
+      "step": 2651
+    },
+    {
+      "epoch": 1.4583391292252053,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00010283120021531087,
+      "loss": 0.7021,
+      "step": 2652
+    },
+    {
+      "epoch": 1.458895534844902,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00010277284007621513,
+      "loss": 0.6848,
+      "step": 2653
+    },
+    {
+      "epoch": 1.4594519404645987,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00010271447899197409,
+      "loss": 0.7605,
+      "step": 2654
+    },
+    {
+      "epoch": 1.4600083460842954,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00010265611698248062,
+      "loss": 0.6646,
+      "step": 2655
+    },
+    {
+      "epoch": 1.4605647517039921,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0001025977540676279,
+      "loss": 0.7369,
+      "step": 2656
+    },
+    {
+      "epoch": 1.461121157323689,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00010253939026730938,
+      "loss": 0.7267,
+      "step": 2657
+    },
+    {
+      "epoch": 1.4616775629433858,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001024810256014189,
+      "loss": 0.7249,
+      "step": 2658
+    },
+    {
+      "epoch": 1.4622339685630825,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00010242266008985046,
+      "loss": 0.7515,
+      "step": 2659
+    },
+    {
+      "epoch": 1.4627903741827792,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00010236429375249848,
+      "loss": 0.8316,
+      "step": 2660
+    },
+    {
+      "epoch": 1.463346779802476,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00010230592660925758,
+      "loss": 0.671,
+      "step": 2661
+    },
+    {
+      "epoch": 1.4639031854221727,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00010224755868002265,
+      "loss": 0.7033,
+      "step": 2662
+    },
+    {
+      "epoch": 1.4644595910418694,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00010218918998468892,
+      "loss": 0.7256,
+      "step": 2663
+    },
+    {
+      "epoch": 1.4650159966615663,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00010213082054315184,
+      "loss": 0.7519,
+      "step": 2664
+    },
+    {
+      "epoch": 1.465572402281263,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00010207245037530711,
+      "loss": 0.7352,
+      "step": 2665
+    },
+    {
+      "epoch": 1.4661288079009598,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00010201407950105064,
+      "loss": 0.6068,
+      "step": 2666
+    },
+    {
+      "epoch": 1.4666852135206565,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00010195570794027866,
+      "loss": 0.6952,
+      "step": 2667
+    },
+    {
+      "epoch": 1.4672416191403532,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00010189733571288761,
+      "loss": 0.7763,
+      "step": 2668
+    },
+    {
+      "epoch": 1.4677980247600502,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00010183896283877412,
+      "loss": 0.776,
+      "step": 2669
+    },
+    {
+      "epoch": 1.4683544303797469,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00010178058933783503,
+      "loss": 0.7237,
+      "step": 2670
+    },
+    {
+      "epoch": 1.4689108359994436,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001017222152299675,
+      "loss": 0.7179,
+      "step": 2671
+    },
+    {
+      "epoch": 1.4694672416191403,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00010166384053506878,
+      "loss": 0.6504,
+      "step": 2672
+    },
+    {
+      "epoch": 1.470023647238837,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00010160546527303635,
+      "loss": 0.6798,
+      "step": 2673
+    },
+    {
+      "epoch": 1.470580052858534,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00010154708946376789,
+      "loss": 0.7528,
+      "step": 2674
+    },
+    {
+      "epoch": 1.4711364584782307,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00010148871312716133,
+      "loss": 0.6758,
+      "step": 2675
+    },
+    {
+      "epoch": 1.4716928640979274,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00010143033628311469,
+      "loss": 0.728,
+      "step": 2676
+    },
+    {
+      "epoch": 1.4722492697176242,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00010137195895152618,
+      "loss": 0.7429,
+      "step": 2677
+    },
+    {
+      "epoch": 1.4728056753373209,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00010131358115229421,
+      "loss": 0.6451,
+      "step": 2678
+    },
+    {
+      "epoch": 1.4733620809570176,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00010125520290531734,
+      "loss": 0.714,
+      "step": 2679
+    },
+    {
+      "epoch": 1.4739184865767143,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00010119682423049425,
+      "loss": 0.6505,
+      "step": 2680
+    },
+    {
+      "epoch": 1.4744748921964113,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00010113844514772384,
+      "loss": 0.7832,
+      "step": 2681
+    },
+    {
+      "epoch": 1.475031297816108,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.000101080065676905,
+      "loss": 0.6569,
+      "step": 2682
+    },
+    {
+      "epoch": 1.4755877034358047,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00010102168583793696,
+      "loss": 0.6913,
+      "step": 2683
+    },
+    {
+      "epoch": 1.4761441090555014,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00010096330565071892,
+      "loss": 0.6483,
+      "step": 2684
+    },
+    {
+      "epoch": 1.4767005146751981,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00010090492513515026,
+      "loss": 0.636,
+      "step": 2685
+    },
+    {
+      "epoch": 1.477256920294895,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00010084654431113043,
+      "loss": 0.7358,
+      "step": 2686
+    },
+    {
+      "epoch": 1.4778133259145918,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00010078816319855903,
+      "loss": 0.7291,
+      "step": 2687
+    },
+    {
+      "epoch": 1.4783697315342885,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00010072978181733574,
+      "loss": 0.695,
+      "step": 2688
+    },
+    {
+      "epoch": 1.4789261371539852,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00010067140018736033,
+      "loss": 0.817,
+      "step": 2689
+    },
+    {
+      "epoch": 1.479482542773682,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00010061301832853267,
+      "loss": 0.7234,
+      "step": 2690
+    },
+    {
+      "epoch": 1.480038948393379,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00010055463626075265,
+      "loss": 0.7691,
+      "step": 2691
+    },
+    {
+      "epoch": 1.4805953540130754,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00010049625400392034,
+      "loss": 0.8378,
+      "step": 2692
+    },
+    {
+      "epoch": 1.4811517596327723,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00010043787157793579,
+      "loss": 0.694,
+      "step": 2693
+    },
+    {
+      "epoch": 1.481708165252469,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00010037948900269909,
+      "loss": 0.6789,
+      "step": 2694
+    },
+    {
+      "epoch": 1.4822645708721658,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00010032110629811044,
+      "loss": 0.7308,
+      "step": 2695
+    },
+    {
+      "epoch": 1.4828209764918625,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00010026272348407011,
+      "loss": 0.7267,
+      "step": 2696
+    },
+    {
+      "epoch": 1.4833773821115592,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00010020434058047833,
+      "loss": 0.788,
+      "step": 2697
+    },
+    {
+      "epoch": 1.4833773821115592,
+      "eval_loss": 0.738091230392456,
+      "eval_runtime": 1753.0991,
+      "eval_samples_per_second": 2.676,
+      "eval_steps_per_second": 1.338,
+      "step": 2697
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 5391,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 899,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.1238739502337884e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}