diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,24255 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 3459,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0002891008962127783,
+      "grad_norm": 10.526118368353698,
+      "learning_rate": 5.714285714285715e-07,
+      "loss": 1.8059,
+      "step": 1
+    },
+    {
+      "epoch": 0.0005782017924255566,
+      "grad_norm": 10.546832449734852,
+      "learning_rate": 1.142857142857143e-06,
+      "loss": 1.8355,
+      "step": 2
+    },
+    {
+      "epoch": 0.0008673026886383347,
+      "grad_norm": 9.678077753237705,
+      "learning_rate": 1.7142857142857145e-06,
+      "loss": 1.8885,
+      "step": 3
+    },
+    {
+      "epoch": 0.0011564035848511131,
+      "grad_norm": 7.133178553946601,
+      "learning_rate": 2.285714285714286e-06,
+      "loss": 1.9111,
+      "step": 4
+    },
+    {
+      "epoch": 0.0014455044810638912,
+      "grad_norm": 4.135673123106036,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 1.7868,
+      "step": 5
+    },
+    {
+      "epoch": 0.0017346053772766695,
+      "grad_norm": 3.0971146670050285,
+      "learning_rate": 3.428571428571429e-06,
+      "loss": 1.8928,
+      "step": 6
+    },
+    {
+      "epoch": 0.002023706273489448,
+      "grad_norm": 2.5830650058945004,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.8522,
+      "step": 7
+    },
+    {
+      "epoch": 0.0023128071697022263,
+      "grad_norm": 2.3055953581941364,
+      "learning_rate": 4.571428571428572e-06,
+      "loss": 1.7831,
+      "step": 8
+    },
+    {
+      "epoch": 0.0026019080659150044,
+      "grad_norm": 1.8810933130960161,
+      "learning_rate": 5.142857142857142e-06,
+      "loss": 1.8335,
+      "step": 9
+    },
+    {
+      "epoch": 0.0028910089621277824,
+      "grad_norm": 2.308546357873222,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 1.7858,
+      "step": 10
+    },
+    {
+      "epoch": 0.003180109858340561,
+      "grad_norm": 1.9146380317808562,
+      "learning_rate": 6.285714285714286e-06,
+      "loss": 1.8005,
+      "step": 11
+    },
+    {
+      "epoch": 0.003469210754553339,
+      "grad_norm": 1.7733597852169312,
+      "learning_rate": 6.857142857142858e-06,
+      "loss": 1.7313,
+      "step": 12
+    },
+    {
+      "epoch": 0.0037583116507661175,
+      "grad_norm": 1.9163487932648622,
+      "learning_rate": 7.428571428571429e-06,
+      "loss": 1.8676,
+      "step": 13
+    },
+    {
+      "epoch": 0.004047412546978896,
+      "grad_norm": 1.81837480862002,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.7696,
+      "step": 14
+    },
+    {
+      "epoch": 0.004336513443191674,
+      "grad_norm": 1.7872590059217233,
+      "learning_rate": 8.571428571428571e-06,
+      "loss": 1.804,
+      "step": 15
+    },
+    {
+      "epoch": 0.004625614339404453,
+      "grad_norm": 1.875901273963478,
+      "learning_rate": 9.142857142857144e-06,
+      "loss": 1.7379,
+      "step": 16
+    },
+    {
+      "epoch": 0.004914715235617231,
+      "grad_norm": 1.6902352819283006,
+      "learning_rate": 9.714285714285715e-06,
+      "loss": 1.7908,
+      "step": 17
+    },
+    {
+      "epoch": 0.005203816131830009,
+      "grad_norm": 1.673145216450599,
+      "learning_rate": 1.0285714285714285e-05,
+      "loss": 1.8525,
+      "step": 18
+    },
+    {
+      "epoch": 0.005492917028042787,
+      "grad_norm": 1.6917810585273594,
+      "learning_rate": 1.0857142857142858e-05,
+      "loss": 1.7779,
+      "step": 19
+    },
+    {
+      "epoch": 0.005782017924255565,
+      "grad_norm": 1.5858495861042294,
+      "learning_rate": 1.1428571428571429e-05,
+      "loss": 1.6791,
+      "step": 20
+    },
+    {
+      "epoch": 0.006071118820468344,
+      "grad_norm": 2.004987370634698,
+      "learning_rate": 1.2e-05,
+      "loss": 1.8019,
+      "step": 21
+    },
+    {
+      "epoch": 0.006360219716681122,
+      "grad_norm": 1.5348568469917172,
+      "learning_rate": 1.2571428571428572e-05,
+      "loss": 1.8565,
+      "step": 22
+    },
+    {
+      "epoch": 0.0066493206128939,
+      "grad_norm": 1.8857981149629766,
+      "learning_rate": 1.3142857142857145e-05,
+      "loss": 1.7292,
+      "step": 23
+    },
+    {
+      "epoch": 0.006938421509106678,
+      "grad_norm": 1.4949439821061543,
+      "learning_rate": 1.3714285714285716e-05,
+      "loss": 1.7357,
+      "step": 24
+    },
+    {
+      "epoch": 0.007227522405319456,
+      "grad_norm": 1.693165357859836,
+      "learning_rate": 1.4285714285714287e-05,
+      "loss": 1.8103,
+      "step": 25
+    },
+    {
+      "epoch": 0.007516623301532235,
+      "grad_norm": 1.6138034582127923,
+      "learning_rate": 1.4857142857142858e-05,
+      "loss": 1.7634,
+      "step": 26
+    },
+    {
+      "epoch": 0.007805724197745013,
+      "grad_norm": 1.6357966310241434,
+      "learning_rate": 1.542857142857143e-05,
+      "loss": 1.7699,
+      "step": 27
+    },
+    {
+      "epoch": 0.008094825093957791,
+      "grad_norm": 1.5038647382371504,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 1.7824,
+      "step": 28
+    },
+    {
+      "epoch": 0.00838392599017057,
+      "grad_norm": 1.4835637524801124,
+      "learning_rate": 1.6571428571428574e-05,
+      "loss": 1.685,
+      "step": 29
+    },
+    {
+      "epoch": 0.008673026886383347,
+      "grad_norm": 1.5071883679399785,
+      "learning_rate": 1.7142857142857142e-05,
+      "loss": 1.8,
+      "step": 30
+    },
+    {
+      "epoch": 0.008962127782596125,
+      "grad_norm": 1.7942262887809508,
+      "learning_rate": 1.7714285714285717e-05,
+      "loss": 1.8129,
+      "step": 31
+    },
+    {
+      "epoch": 0.009251228678808905,
+      "grad_norm": 1.6582976562001455,
+      "learning_rate": 1.8285714285714288e-05,
+      "loss": 1.8308,
+      "step": 32
+    },
+    {
+      "epoch": 0.009540329575021683,
+      "grad_norm": 1.5337861742601142,
+      "learning_rate": 1.885714285714286e-05,
+      "loss": 1.7456,
+      "step": 33
+    },
+    {
+      "epoch": 0.009829430471234461,
+      "grad_norm": 1.8002402808530826,
+      "learning_rate": 1.942857142857143e-05,
+      "loss": 1.9268,
+      "step": 34
+    },
+    {
+      "epoch": 0.01011853136744724,
+      "grad_norm": 1.7514076269246104,
+      "learning_rate": 2e-05,
+      "loss": 1.652,
+      "step": 35
+    },
+    {
+      "epoch": 0.010407632263660017,
+      "grad_norm": 1.8491103433261384,
+      "learning_rate": 1.9999995790774362e-05,
+      "loss": 1.8142,
+      "step": 36
+    },
+    {
+      "epoch": 0.010696733159872795,
+      "grad_norm": 1.8365212131962536,
+      "learning_rate": 1.9999983163100983e-05,
+      "loss": 1.7107,
+      "step": 37
+    },
+    {
+      "epoch": 0.010985834056085574,
+      "grad_norm": 1.5873275727358707,
+      "learning_rate": 1.9999962116990498e-05,
+      "loss": 1.6696,
+      "step": 38
+    },
+    {
+      "epoch": 0.011274934952298352,
+      "grad_norm": 1.7256221512542549,
+      "learning_rate": 1.999993265246062e-05,
+      "loss": 1.6901,
+      "step": 39
+    },
+    {
+      "epoch": 0.01156403584851113,
+      "grad_norm": 1.8002146086893285,
+      "learning_rate": 1.9999894769536163e-05,
+      "loss": 1.8886,
+      "step": 40
+    },
+    {
+      "epoch": 0.01185313674472391,
+      "grad_norm": 1.8367868066313766,
+      "learning_rate": 1.999984846824901e-05,
+      "loss": 1.7876,
+      "step": 41
+    },
+    {
+      "epoch": 0.012142237640936688,
+      "grad_norm": 1.5841989203084583,
+      "learning_rate": 1.999979374863814e-05,
+      "loss": 1.9172,
+      "step": 42
+    },
+    {
+      "epoch": 0.012431338537149466,
+      "grad_norm": 1.5522062286750515,
+      "learning_rate": 1.9999730610749623e-05,
+      "loss": 1.8207,
+      "step": 43
+    },
+    {
+      "epoch": 0.012720439433362244,
+      "grad_norm": 1.6373562793041219,
+      "learning_rate": 1.9999659054636612e-05,
+      "loss": 1.6358,
+      "step": 44
+    },
+    {
+      "epoch": 0.013009540329575022,
+      "grad_norm": 1.4771807978853968,
+      "learning_rate": 1.9999579080359337e-05,
+      "loss": 1.6684,
+      "step": 45
+    },
+    {
+      "epoch": 0.0132986412257878,
+      "grad_norm": 1.6255295191073982,
+      "learning_rate": 1.9999490687985134e-05,
+      "loss": 1.7113,
+      "step": 46
+    },
+    {
+      "epoch": 0.013587742122000578,
+      "grad_norm": 1.4924171597100626,
+      "learning_rate": 1.999939387758841e-05,
+      "loss": 1.7272,
+      "step": 47
+    },
+    {
+      "epoch": 0.013876843018213356,
+      "grad_norm": 1.536593102227256,
+      "learning_rate": 1.9999288649250667e-05,
+      "loss": 1.7532,
+      "step": 48
+    },
+    {
+      "epoch": 0.014165943914426134,
+      "grad_norm": 1.510648288143349,
+      "learning_rate": 1.9999175003060487e-05,
+      "loss": 1.8358,
+      "step": 49
+    },
+    {
+      "epoch": 0.014455044810638912,
+      "grad_norm": 1.4755302580340564,
+      "learning_rate": 1.9999052939113552e-05,
+      "loss": 1.8305,
+      "step": 50
+    },
+    {
+      "epoch": 0.014744145706851692,
+      "grad_norm": 1.659035076331237,
+      "learning_rate": 1.9998922457512608e-05,
+      "loss": 1.7916,
+      "step": 51
+    },
+    {
+      "epoch": 0.01503324660306447,
+      "grad_norm": 1.4194234922554345,
+      "learning_rate": 1.9998783558367506e-05,
+      "loss": 1.6693,
+      "step": 52
+    },
+    {
+      "epoch": 0.015322347499277248,
+      "grad_norm": 1.5709438854610118,
+      "learning_rate": 1.9998636241795184e-05,
+      "loss": 1.7873,
+      "step": 53
+    },
+    {
+      "epoch": 0.015611448395490026,
+      "grad_norm": 1.520579238616479,
+      "learning_rate": 1.999848050791965e-05,
+      "loss": 1.6949,
+      "step": 54
+    },
+    {
+      "epoch": 0.015900549291702804,
+      "grad_norm": 1.6871179175362505,
+      "learning_rate": 1.9998316356872015e-05,
+      "loss": 1.7697,
+      "step": 55
+    },
+    {
+      "epoch": 0.016189650187915582,
+      "grad_norm": 1.3949302668328034,
+      "learning_rate": 1.9998143788790462e-05,
+      "loss": 1.6176,
+      "step": 56
+    },
+    {
+      "epoch": 0.01647875108412836,
+      "grad_norm": 1.76393196989828,
+      "learning_rate": 1.9997962803820274e-05,
+      "loss": 1.7379,
+      "step": 57
+    },
+    {
+      "epoch": 0.01676785198034114,
+      "grad_norm": 2.3952825919806724,
+      "learning_rate": 1.9997773402113804e-05,
+      "loss": 1.8324,
+      "step": 58
+    },
+    {
+      "epoch": 0.017056952876553916,
+      "grad_norm": 1.5638827679672842,
+      "learning_rate": 1.9997575583830508e-05,
+      "loss": 1.783,
+      "step": 59
+    },
+    {
+      "epoch": 0.017346053772766695,
+      "grad_norm": 1.5321008746402438,
+      "learning_rate": 1.999736934913691e-05,
+      "loss": 1.7136,
+      "step": 60
+    },
+    {
+      "epoch": 0.017635154668979473,
+      "grad_norm": 1.46754278560031,
+      "learning_rate": 1.9997154698206636e-05,
+      "loss": 1.7567,
+      "step": 61
+    },
+    {
+      "epoch": 0.01792425556519225,
+      "grad_norm": 1.5842416386679679,
+      "learning_rate": 1.999693163122038e-05,
+      "loss": 1.757,
+      "step": 62
+    },
+    {
+      "epoch": 0.01821335646140503,
+      "grad_norm": 1.5267999996929678,
+      "learning_rate": 1.9996700148365936e-05,
+      "loss": 1.768,
+      "step": 63
+    },
+    {
+      "epoch": 0.01850245735761781,
+      "grad_norm": 1.6371904649667028,
+      "learning_rate": 1.9996460249838176e-05,
+      "loss": 1.8574,
+      "step": 64
+    },
+    {
+      "epoch": 0.01879155825383059,
+      "grad_norm": 1.4572549097707117,
+      "learning_rate": 1.9996211935839053e-05,
+      "loss": 1.8134,
+      "step": 65
+    },
+    {
+      "epoch": 0.019080659150043366,
+      "grad_norm": 1.6026953718941006,
+      "learning_rate": 1.9995955206577617e-05,
+      "loss": 1.7718,
+      "step": 66
+    },
+    {
+      "epoch": 0.019369760046256145,
+      "grad_norm": 1.4731386293380266,
+      "learning_rate": 1.9995690062269985e-05,
+      "loss": 1.7598,
+      "step": 67
+    },
+    {
+      "epoch": 0.019658860942468923,
+      "grad_norm": 1.4280673736098164,
+      "learning_rate": 1.9995416503139372e-05,
+      "loss": 1.7469,
+      "step": 68
+    },
+    {
+      "epoch": 0.0199479618386817,
+      "grad_norm": 1.4976095361519832,
+      "learning_rate": 1.9995134529416077e-05,
+      "loss": 1.7258,
+      "step": 69
+    },
+    {
+      "epoch": 0.02023706273489448,
+      "grad_norm": 1.3734046074406898,
+      "learning_rate": 1.9994844141337467e-05,
+      "loss": 1.6921,
+      "step": 70
+    },
+    {
+      "epoch": 0.020526163631107257,
+      "grad_norm": 1.5285691045400083,
+      "learning_rate": 1.9994545339148017e-05,
+      "loss": 1.755,
+      "step": 71
+    },
+    {
+      "epoch": 0.020815264527320035,
+      "grad_norm": 1.7237565941986015,
+      "learning_rate": 1.9994238123099264e-05,
+      "loss": 1.7254,
+      "step": 72
+    },
+    {
+      "epoch": 0.021104365423532813,
+      "grad_norm": 1.4511698572347005,
+      "learning_rate": 1.9993922493449833e-05,
+      "loss": 1.7339,
+      "step": 73
+    },
+    {
+      "epoch": 0.02139346631974559,
+      "grad_norm": 1.3034115810336904,
+      "learning_rate": 1.9993598450465446e-05,
+      "loss": 1.8015,
+      "step": 74
+    },
+    {
+      "epoch": 0.02168256721595837,
+      "grad_norm": 1.3220719840165054,
+      "learning_rate": 1.9993265994418887e-05,
+      "loss": 1.7794,
+      "step": 75
+    },
+    {
+      "epoch": 0.021971668112171147,
+      "grad_norm": 1.3514726546077938,
+      "learning_rate": 1.999292512559004e-05,
+      "loss": 1.781,
+      "step": 76
+    },
+    {
+      "epoch": 0.022260769008383925,
+      "grad_norm": 1.4854967526025784,
+      "learning_rate": 1.9992575844265857e-05,
+      "loss": 1.8188,
+      "step": 77
+    },
+    {
+      "epoch": 0.022549869904596703,
+      "grad_norm": 1.42527411671424,
+      "learning_rate": 1.9992218150740386e-05,
+      "loss": 1.7858,
+      "step": 78
+    },
+    {
+      "epoch": 0.02283897080080948,
+      "grad_norm": 1.3690019880914763,
+      "learning_rate": 1.9991852045314744e-05,
+      "loss": 1.69,
+      "step": 79
+    },
+    {
+      "epoch": 0.02312807169702226,
+      "grad_norm": 1.4796987548843166,
+      "learning_rate": 1.9991477528297136e-05,
+      "loss": 1.7501,
+      "step": 80
+    },
+    {
+      "epoch": 0.023417172593235037,
+      "grad_norm": 1.4237249543510602,
+      "learning_rate": 1.999109460000285e-05,
+      "loss": 1.7634,
+      "step": 81
+    },
+    {
+      "epoch": 0.02370627348944782,
+      "grad_norm": 1.3087042024971207,
+      "learning_rate": 1.9990703260754248e-05,
+      "loss": 1.8106,
+      "step": 82
+    },
+    {
+      "epoch": 0.023995374385660597,
+      "grad_norm": 1.4158584649264259,
+      "learning_rate": 1.999030351088078e-05,
+      "loss": 1.7503,
+      "step": 83
+    },
+    {
+      "epoch": 0.024284475281873375,
+      "grad_norm": 1.4266437173978106,
+      "learning_rate": 1.9989895350718972e-05,
+      "loss": 1.6435,
+      "step": 84
+    },
+    {
+      "epoch": 0.024573576178086153,
+      "grad_norm": 1.519614177980851,
+      "learning_rate": 1.9989478780612434e-05,
+      "loss": 1.7155,
+      "step": 85
+    },
+    {
+      "epoch": 0.02486267707429893,
+      "grad_norm": 1.4829974498829237,
+      "learning_rate": 1.9989053800911854e-05,
+      "loss": 1.7406,
+      "step": 86
+    },
+    {
+      "epoch": 0.02515177797051171,
+      "grad_norm": 1.5340468112410737,
+      "learning_rate": 1.9988620411974994e-05,
+      "loss": 1.6487,
+      "step": 87
+    },
+    {
+      "epoch": 0.025440878866724487,
+      "grad_norm": 1.3928928308705317,
+      "learning_rate": 1.9988178614166706e-05,
+      "loss": 1.7535,
+      "step": 88
+    },
+    {
+      "epoch": 0.025729979762937265,
+      "grad_norm": 1.4648291742355268,
+      "learning_rate": 1.9987728407858912e-05,
+      "loss": 1.6977,
+      "step": 89
+    },
+    {
+      "epoch": 0.026019080659150044,
+      "grad_norm": 1.5330205874026324,
+      "learning_rate": 1.9987269793430618e-05,
+      "loss": 1.8134,
+      "step": 90
+    },
+    {
+      "epoch": 0.02630818155536282,
+      "grad_norm": 1.4322119486638492,
+      "learning_rate": 1.9986802771267902e-05,
+      "loss": 1.7493,
+      "step": 91
+    },
+    {
+      "epoch": 0.0265972824515756,
+      "grad_norm": 1.3741952459535092,
+      "learning_rate": 1.9986327341763933e-05,
+      "loss": 1.7498,
+      "step": 92
+    },
+    {
+      "epoch": 0.026886383347788378,
+      "grad_norm": 1.4439173869710846,
+      "learning_rate": 1.998584350531894e-05,
+      "loss": 1.8364,
+      "step": 93
+    },
+    {
+      "epoch": 0.027175484244001156,
+      "grad_norm": 1.5238329839004898,
+      "learning_rate": 1.9985351262340242e-05,
+      "loss": 1.7025,
+      "step": 94
+    },
+    {
+      "epoch": 0.027464585140213934,
+      "grad_norm": 1.4211403871339836,
+      "learning_rate": 1.998485061324223e-05,
+      "loss": 1.8061,
+      "step": 95
+    },
+    {
+      "epoch": 0.027753686036426712,
+      "grad_norm": 1.4891220326396055,
+      "learning_rate": 1.998434155844638e-05,
+      "loss": 1.7756,
+      "step": 96
+    },
+    {
+      "epoch": 0.02804278693263949,
+      "grad_norm": 1.4672105325874756,
+      "learning_rate": 1.9983824098381226e-05,
+      "loss": 1.7642,
+      "step": 97
+    },
+    {
+      "epoch": 0.028331887828852268,
+      "grad_norm": 1.4102584299249543,
+      "learning_rate": 1.9983298233482396e-05,
+      "loss": 1.7302,
+      "step": 98
+    },
+    {
+      "epoch": 0.028620988725065046,
+      "grad_norm": 1.4570496244711266,
+      "learning_rate": 1.9982763964192586e-05,
+      "loss": 1.7299,
+      "step": 99
+    },
+    {
+      "epoch": 0.028910089621277824,
+      "grad_norm": 1.4536709099623053,
+      "learning_rate": 1.9982221290961568e-05,
+      "loss": 1.7647,
+      "step": 100
+    },
+    {
+      "epoch": 0.029199190517490606,
+      "grad_norm": 1.427202594686143,
+      "learning_rate": 1.9981670214246186e-05,
+      "loss": 1.7361,
+      "step": 101
+    },
+    {
+      "epoch": 0.029488291413703384,
+      "grad_norm": 1.4166737515686485,
+      "learning_rate": 1.9981110734510368e-05,
+      "loss": 1.7107,
+      "step": 102
+    },
+    {
+      "epoch": 0.029777392309916162,
+      "grad_norm": 1.4291514974937183,
+      "learning_rate": 1.99805428522251e-05,
+      "loss": 1.7162,
+      "step": 103
+    },
+    {
+      "epoch": 0.03006649320612894,
+      "grad_norm": 1.3691457808724619,
+      "learning_rate": 1.997996656786846e-05,
+      "loss": 1.8823,
+      "step": 104
+    },
+    {
+      "epoch": 0.030355594102341718,
+      "grad_norm": 1.38193263647762,
+      "learning_rate": 1.9979381881925587e-05,
+      "loss": 1.6658,
+      "step": 105
+    },
+    {
+      "epoch": 0.030644694998554496,
+      "grad_norm": 1.4331634763444776,
+      "learning_rate": 1.997878879488869e-05,
+      "loss": 1.7695,
+      "step": 106
+    },
+    {
+      "epoch": 0.030933795894767274,
+      "grad_norm": 1.4389434938173666,
+      "learning_rate": 1.9978187307257064e-05,
+      "loss": 1.7356,
+      "step": 107
+    },
+    {
+      "epoch": 0.031222896790980052,
+      "grad_norm": 1.4057807430118254,
+      "learning_rate": 1.9977577419537066e-05,
+      "loss": 1.7191,
+      "step": 108
+    },
+    {
+      "epoch": 0.03151199768719283,
+      "grad_norm": 1.4476225746298625,
+      "learning_rate": 1.9976959132242128e-05,
+      "loss": 1.8028,
+      "step": 109
+    },
+    {
+      "epoch": 0.03180109858340561,
+      "grad_norm": 1.3595624610249801,
+      "learning_rate": 1.997633244589275e-05,
+      "loss": 1.708,
+      "step": 110
+    },
+    {
+      "epoch": 0.032090199479618386,
+      "grad_norm": 1.5622440441344578,
+      "learning_rate": 1.997569736101651e-05,
+      "loss": 1.7138,
+      "step": 111
+    },
+    {
+      "epoch": 0.032379300375831165,
+      "grad_norm": 1.4451837327532462,
+      "learning_rate": 1.9975053878148045e-05,
+      "loss": 1.745,
+      "step": 112
+    },
+    {
+      "epoch": 0.03266840127204394,
+      "grad_norm": 1.3461945134165731,
+      "learning_rate": 1.9974401997829063e-05,
+      "loss": 1.792,
+      "step": 113
+    },
+    {
+      "epoch": 0.03295750216825672,
+      "grad_norm": 1.4916956994354813,
+      "learning_rate": 1.997374172060836e-05,
+      "loss": 1.7312,
+      "step": 114
+    },
+    {
+      "epoch": 0.0332466030644695,
+      "grad_norm": 1.3473635350533286,
+      "learning_rate": 1.997307304704178e-05,
+      "loss": 1.7257,
+      "step": 115
+    },
+    {
+      "epoch": 0.03353570396068228,
+      "grad_norm": 1.5308835567804706,
+      "learning_rate": 1.9972395977692243e-05,
+      "loss": 1.8833,
+      "step": 116
+    },
+    {
+      "epoch": 0.033824804856895055,
+      "grad_norm": 1.4933799858376915,
+      "learning_rate": 1.9971710513129735e-05,
+      "loss": 1.8031,
+      "step": 117
+    },
+    {
+      "epoch": 0.03411390575310783,
+      "grad_norm": 1.4666676384345598,
+      "learning_rate": 1.9971016653931315e-05,
+      "loss": 1.699,
+      "step": 118
+    },
+    {
+      "epoch": 0.03440300664932061,
+      "grad_norm": 1.376396676563015,
+      "learning_rate": 1.9970314400681098e-05,
+      "loss": 1.7561,
+      "step": 119
+    },
+    {
+      "epoch": 0.03469210754553339,
+      "grad_norm": 1.4309738039308064,
+      "learning_rate": 1.9969603753970285e-05,
+      "loss": 1.7812,
+      "step": 120
+    },
+    {
+      "epoch": 0.03498120844174617,
+      "grad_norm": 1.4395837496367765,
+      "learning_rate": 1.9968884714397116e-05,
+      "loss": 1.6863,
+      "step": 121
+    },
+    {
+      "epoch": 0.035270309337958945,
+      "grad_norm": 1.3440476522906626,
+      "learning_rate": 1.9968157282566917e-05,
+      "loss": 1.7026,
+      "step": 122
+    },
+    {
+      "epoch": 0.03555941023417172,
+      "grad_norm": 1.3253598540978417,
+      "learning_rate": 1.9967421459092076e-05,
+      "loss": 1.6773,
+      "step": 123
+    },
+    {
+      "epoch": 0.0358485111303845,
+      "grad_norm": 1.4654991309929606,
+      "learning_rate": 1.996667724459204e-05,
+      "loss": 1.6995,
+      "step": 124
+    },
+    {
+      "epoch": 0.03613761202659728,
+      "grad_norm": 1.3695366677193352,
+      "learning_rate": 1.9965924639693323e-05,
+      "loss": 1.8028,
+      "step": 125
+    },
+    {
+      "epoch": 0.03642671292281006,
+      "grad_norm": 1.5996594030035125,
+      "learning_rate": 1.9965163645029496e-05,
+      "loss": 1.6509,
+      "step": 126
+    },
+    {
+      "epoch": 0.036715813819022836,
+      "grad_norm": 1.3638136094831714,
+      "learning_rate": 1.9964394261241205e-05,
+      "loss": 1.7974,
+      "step": 127
+    },
+    {
+      "epoch": 0.03700491471523562,
+      "grad_norm": 1.5950089807963124,
+      "learning_rate": 1.9963616488976156e-05,
+      "loss": 1.7817,
+      "step": 128
+    },
+    {
+      "epoch": 0.0372940156114484,
+      "grad_norm": 1.3874690940551964,
+      "learning_rate": 1.9962830328889104e-05,
+      "loss": 1.7727,
+      "step": 129
+    },
+    {
+      "epoch": 0.03758311650766118,
+      "grad_norm": 1.428332455733304,
+      "learning_rate": 1.9962035781641877e-05,
+      "loss": 1.8162,
+      "step": 130
+    },
+    {
+      "epoch": 0.037872217403873955,
+      "grad_norm": 1.413377966510162,
+      "learning_rate": 1.996123284790336e-05,
+      "loss": 1.6724,
+      "step": 131
+    },
+    {
+      "epoch": 0.03816131830008673,
+      "grad_norm": 1.4114781128468488,
+      "learning_rate": 1.9960421528349503e-05,
+      "loss": 1.6688,
+      "step": 132
+    },
+    {
+      "epoch": 0.03845041919629951,
+      "grad_norm": 1.4667937866255885,
+      "learning_rate": 1.9959601823663305e-05,
+      "loss": 1.7404,
+      "step": 133
+    },
+    {
+      "epoch": 0.03873952009251229,
+      "grad_norm": 1.3864312896377629,
+      "learning_rate": 1.9958773734534834e-05,
+      "loss": 1.6797,
+      "step": 134
+    },
+    {
+      "epoch": 0.03902862098872507,
+      "grad_norm": 1.5863275696219379,
+      "learning_rate": 1.9957937261661215e-05,
+      "loss": 1.7093,
+      "step": 135
+    },
+    {
+      "epoch": 0.039317721884937845,
+      "grad_norm": 1.3943132521368478,
+      "learning_rate": 1.9957092405746628e-05,
+      "loss": 1.7466,
+      "step": 136
+    },
+    {
+      "epoch": 0.03960682278115062,
+      "grad_norm": 1.4418545458171168,
+      "learning_rate": 1.9956239167502304e-05,
+      "loss": 1.8363,
+      "step": 137
+    },
+    {
+      "epoch": 0.0398959236773634,
+      "grad_norm": 1.6597022379228474,
+      "learning_rate": 1.9955377547646546e-05,
+      "loss": 1.7726,
+      "step": 138
+    },
+    {
+      "epoch": 0.04018502457357618,
+      "grad_norm": 1.374590779273713,
+      "learning_rate": 1.9954507546904697e-05,
+      "loss": 1.7559,
+      "step": 139
+    },
+    {
+      "epoch": 0.04047412546978896,
+      "grad_norm": 1.2866499565497842,
+      "learning_rate": 1.995362916600917e-05,
+      "loss": 1.7478,
+      "step": 140
+    },
+    {
+      "epoch": 0.040763226366001735,
+      "grad_norm": 1.3549533524362687,
+      "learning_rate": 1.9952742405699425e-05,
+      "loss": 1.8389,
+      "step": 141
+    },
+    {
+      "epoch": 0.041052327262214514,
+      "grad_norm": 1.3512745611701633,
+      "learning_rate": 1.995184726672197e-05,
+      "loss": 1.7975,
+      "step": 142
+    },
+    {
+      "epoch": 0.04134142815842729,
+      "grad_norm": 1.3460191452855619,
+      "learning_rate": 1.995094374983038e-05,
+      "loss": 1.7245,
+      "step": 143
+    },
+    {
+      "epoch": 0.04163052905464007,
+      "grad_norm": 1.3305167255129993,
+      "learning_rate": 1.9950031855785276e-05,
+      "loss": 1.6294,
+      "step": 144
+    },
+    {
+      "epoch": 0.04191962995085285,
+      "grad_norm": 1.34284871707125,
+      "learning_rate": 1.9949111585354328e-05,
+      "loss": 1.7124,
+      "step": 145
+    },
+    {
+      "epoch": 0.042208730847065626,
+      "grad_norm": 1.3772925926259127,
+      "learning_rate": 1.9948182939312258e-05,
+      "loss": 1.7095,
+      "step": 146
+    },
+    {
+      "epoch": 0.042497831743278404,
+      "grad_norm": 1.384298315882825,
+      "learning_rate": 1.994724591844085e-05,
+      "loss": 1.8031,
+      "step": 147
+    },
+    {
+      "epoch": 0.04278693263949118,
+      "grad_norm": 1.3740208297445895,
+      "learning_rate": 1.994630052352893e-05,
+      "loss": 1.7814,
+      "step": 148
+    },
+    {
+      "epoch": 0.04307603353570396,
+      "grad_norm": 1.3187990568089156,
+      "learning_rate": 1.9945346755372367e-05,
+      "loss": 1.7486,
+      "step": 149
+    },
+    {
+      "epoch": 0.04336513443191674,
+      "grad_norm": 1.3347832866557865,
+      "learning_rate": 1.9944384614774095e-05,
+      "loss": 1.7225,
+      "step": 150
+    },
+    {
+      "epoch": 0.043654235328129516,
+      "grad_norm": 1.3850913268385951,
+      "learning_rate": 1.9943414102544083e-05,
+      "loss": 1.7608,
+      "step": 151
+    },
+    {
+      "epoch": 0.043943336224342294,
+      "grad_norm": 1.3180540802523992,
+      "learning_rate": 1.994243521949935e-05,
+      "loss": 1.7534,
+      "step": 152
+    },
+    {
+      "epoch": 0.04423243712055507,
+      "grad_norm": 1.3989923351107776,
+      "learning_rate": 1.9941447966463966e-05,
+      "loss": 1.7734,
+      "step": 153
+    },
+    {
+      "epoch": 0.04452153801676785,
+      "grad_norm": 1.2709853505702844,
+      "learning_rate": 1.9940452344269045e-05,
+      "loss": 1.7463,
+      "step": 154
+    },
+    {
+      "epoch": 0.04481063891298063,
+      "grad_norm": 1.3117490865184025,
+      "learning_rate": 1.9939448353752745e-05,
+      "loss": 1.7484,
+      "step": 155
+    },
+    {
+      "epoch": 0.045099739809193407,
+      "grad_norm": 1.3068520843456393,
+      "learning_rate": 1.9938435995760275e-05,
+      "loss": 1.7793,
+      "step": 156
+    },
+    {
+      "epoch": 0.045388840705406185,
+      "grad_norm": 1.2903723753260385,
+      "learning_rate": 1.993741527114388e-05,
+      "loss": 1.7654,
+      "step": 157
+    },
+    {
+      "epoch": 0.04567794160161896,
+      "grad_norm": 1.3281351334689,
+      "learning_rate": 1.993638618076285e-05,
+      "loss": 1.7574,
+      "step": 158
+    },
+    {
+      "epoch": 0.04596704249783174,
+      "grad_norm": 1.331538167128145,
+      "learning_rate": 1.9935348725483526e-05,
+      "loss": 1.7431,
+      "step": 159
+    },
+    {
+      "epoch": 0.04625614339404452,
+      "grad_norm": 1.4516409761685403,
+      "learning_rate": 1.9934302906179285e-05,
+      "loss": 1.7334,
+      "step": 160
+    },
+    {
+      "epoch": 0.0465452442902573,
+      "grad_norm": 1.2977322772074695,
+      "learning_rate": 1.9933248723730536e-05,
+      "loss": 1.6306,
+      "step": 161
+    },
+    {
+      "epoch": 0.046834345186470075,
+      "grad_norm": 1.357062910781076,
+      "learning_rate": 1.9932186179024744e-05,
+      "loss": 1.668,
+      "step": 162
+    },
+    {
+      "epoch": 0.04712344608268285,
+      "grad_norm": 1.3803460903193827,
+      "learning_rate": 1.9931115272956405e-05,
+      "loss": 1.7275,
+      "step": 163
+    },
+    {
+      "epoch": 0.04741254697889564,
+      "grad_norm": 1.315843862445011,
+      "learning_rate": 1.9930036006427058e-05,
+      "loss": 1.6444,
+      "step": 164
+    },
+    {
+      "epoch": 0.047701647875108416,
+      "grad_norm": 1.4617283691426304,
+      "learning_rate": 1.9928948380345274e-05,
+      "loss": 1.8109,
+      "step": 165
+    },
+    {
+      "epoch": 0.047990748771321194,
+      "grad_norm": 1.38383885394047,
+      "learning_rate": 1.9927852395626672e-05,
+      "loss": 1.7686,
+      "step": 166
+    },
+    {
+      "epoch": 0.04827984966753397,
+      "grad_norm": 1.458724964408903,
+      "learning_rate": 1.9926748053193898e-05,
+      "loss": 1.6962,
+      "step": 167
+    },
+    {
+      "epoch": 0.04856895056374675,
+      "grad_norm": 1.3411049270177544,
+      "learning_rate": 1.9925635353976634e-05,
+      "loss": 1.6548,
+      "step": 168
+    },
+    {
+      "epoch": 0.04885805145995953,
+      "grad_norm": 1.4316515117975288,
+      "learning_rate": 1.992451429891161e-05,
+      "loss": 1.7439,
+      "step": 169
+    },
+    {
+      "epoch": 0.049147152356172306,
+      "grad_norm": 1.3310049050553332,
+      "learning_rate": 1.9923384888942568e-05,
+      "loss": 1.7042,
+      "step": 170
+    },
+    {
+      "epoch": 0.049436253252385085,
+      "grad_norm": 1.411118381246596,
+      "learning_rate": 1.9922247125020307e-05,
+      "loss": 1.7172,
+      "step": 171
+    },
+    {
+      "epoch": 0.04972535414859786,
+      "grad_norm": 1.2882521449870807,
+      "learning_rate": 1.992110100810264e-05,
+      "loss": 1.7679,
+      "step": 172
+    },
+    {
+      "epoch": 0.05001445504481064,
+      "grad_norm": 1.3412263735063734,
+      "learning_rate": 1.991994653915443e-05,
+      "loss": 1.6375,
+      "step": 173
+    },
+    {
+      "epoch": 0.05030355594102342,
+      "grad_norm": 1.2875141274832036,
+      "learning_rate": 1.991878371914755e-05,
+      "loss": 1.6816,
+      "step": 174
+    },
+    {
+      "epoch": 0.0505926568372362,
+      "grad_norm": 1.2707786543811448,
+      "learning_rate": 1.991761254906092e-05,
+      "loss": 1.7382,
+      "step": 175
+    },
+    {
+      "epoch": 0.050881757733448975,
+      "grad_norm": 1.2815480979153002,
+      "learning_rate": 1.9916433029880485e-05,
+      "loss": 1.7519,
+      "step": 176
+    },
+    {
+      "epoch": 0.05117085862966175,
+      "grad_norm": 1.358987922639111,
+      "learning_rate": 1.991524516259921e-05,
+      "loss": 1.7895,
+      "step": 177
+    },
+    {
+      "epoch": 0.05145995952587453,
+      "grad_norm": 1.2688791830287245,
+      "learning_rate": 1.9914048948217105e-05,
+      "loss": 1.776,
+      "step": 178
+    },
+    {
+      "epoch": 0.05174906042208731,
+      "grad_norm": 1.2813810920416182,
+      "learning_rate": 1.9912844387741194e-05,
+      "loss": 1.7085,
+      "step": 179
+    },
+    {
+      "epoch": 0.05203816131830009,
+      "grad_norm": 1.2916088597948596,
+      "learning_rate": 1.991163148218553e-05,
+      "loss": 1.7124,
+      "step": 180
+    },
+    {
+      "epoch": 0.052327262214512865,
+      "grad_norm": 1.310527847083729,
+      "learning_rate": 1.9910410232571187e-05,
+      "loss": 1.6333,
+      "step": 181
+    },
+    {
+      "epoch": 0.05261636311072564,
+      "grad_norm": 1.3292526744086122,
+      "learning_rate": 1.9909180639926274e-05,
+      "loss": 1.663,
+      "step": 182
+    },
+    {
+      "epoch": 0.05290546400693842,
+      "grad_norm": 1.4107465041661023,
+      "learning_rate": 1.990794270528592e-05,
+      "loss": 1.6939,
+      "step": 183
+    },
+    {
+      "epoch": 0.0531945649031512,
+      "grad_norm": 1.5697948856294517,
+      "learning_rate": 1.9906696429692265e-05,
+      "loss": 1.79,
+      "step": 184
+    },
+    {
+      "epoch": 0.05348366579936398,
+      "grad_norm": 1.359687672105227,
+      "learning_rate": 1.9905441814194482e-05,
+      "loss": 1.6024,
+      "step": 185
+    },
+    {
+      "epoch": 0.053772766695576756,
+      "grad_norm": 1.435220148856358,
+      "learning_rate": 1.990417885984877e-05,
+      "loss": 1.6539,
+      "step": 186
+    },
+    {
+      "epoch": 0.054061867591789534,
+      "grad_norm": 1.4702625299572514,
+      "learning_rate": 1.990290756771834e-05,
+      "loss": 1.7306,
+      "step": 187
+    },
+    {
+      "epoch": 0.05435096848800231,
+      "grad_norm": 1.5247866204720824,
+      "learning_rate": 1.9901627938873416e-05,
+      "loss": 1.7609,
+      "step": 188
+    },
+    {
+      "epoch": 0.05464006938421509,
+      "grad_norm": 1.3031717794857007,
+      "learning_rate": 1.9900339974391252e-05,
+      "loss": 1.7389,
+      "step": 189
+    },
+    {
+      "epoch": 0.05492917028042787,
+      "grad_norm": 1.3511696476383066,
+      "learning_rate": 1.9899043675356114e-05,
+      "loss": 1.7297,
+      "step": 190
+    },
+    {
+      "epoch": 0.055218271176640646,
+      "grad_norm": 1.4509301331653846,
+      "learning_rate": 1.9897739042859285e-05,
+      "loss": 1.737,
+      "step": 191
+    },
+    {
+      "epoch": 0.055507372072853424,
+      "grad_norm": 1.3380070550795329,
+      "learning_rate": 1.9896426077999062e-05,
+      "loss": 1.6662,
+      "step": 192
+    },
+    {
+      "epoch": 0.0557964729690662,
+      "grad_norm": 1.3684181423939343,
+      "learning_rate": 1.989510478188076e-05,
+      "loss": 1.7297,
+      "step": 193
+    },
+    {
+      "epoch": 0.05608557386527898,
+      "grad_norm": 1.4948293977771248,
+      "learning_rate": 1.9893775155616704e-05,
+      "loss": 1.7733,
+      "step": 194
+    },
+    {
+      "epoch": 0.05637467476149176,
+      "grad_norm": 6.056711246922273,
+      "learning_rate": 1.989243720032624e-05,
+      "loss": 1.8127,
+      "step": 195
+    },
+    {
+      "epoch": 0.056663775657704536,
+      "grad_norm": 1.4031162820625223,
+      "learning_rate": 1.989109091713571e-05,
+      "loss": 1.7476,
+      "step": 196
+    },
+    {
+      "epoch": 0.056952876553917314,
+      "grad_norm": 1.388743000455743,
+      "learning_rate": 1.988973630717848e-05,
+      "loss": 1.7256,
+      "step": 197
+    },
+    {
+      "epoch": 0.05724197745013009,
+      "grad_norm": 1.3228310217145096,
+      "learning_rate": 1.9888373371594923e-05,
+      "loss": 1.7153,
+      "step": 198
+    },
+    {
+      "epoch": 0.05753107834634287,
+      "grad_norm": 1.349390586328493,
+      "learning_rate": 1.9887002111532413e-05,
+      "loss": 1.7913,
+      "step": 199
+    },
+    {
+      "epoch": 0.05782017924255565,
+      "grad_norm": 1.3218835310967434,
+      "learning_rate": 1.9885622528145346e-05,
+      "loss": 1.893,
+      "step": 200
+    },
+    {
+      "epoch": 0.058109280138768434,
+      "grad_norm": 1.3066393320266039,
+      "learning_rate": 1.9884234622595117e-05,
+      "loss": 1.7445,
+      "step": 201
+    },
+    {
+      "epoch": 0.05839838103498121,
+      "grad_norm": 1.3431846081510803,
+      "learning_rate": 1.988283839605013e-05,
+      "loss": 1.764,
+      "step": 202
+    },
+    {
+      "epoch": 0.05868748193119399,
+      "grad_norm": 1.2716836951252142,
+      "learning_rate": 1.988143384968578e-05,
+      "loss": 1.689,
+      "step": 203
+    },
+    {
+      "epoch": 0.05897658282740677,
+      "grad_norm": 1.3463201818447268,
+      "learning_rate": 1.9880020984684486e-05,
+      "loss": 1.6615,
+      "step": 204
+    },
+    {
+      "epoch": 0.059265683723619546,
+      "grad_norm": 1.3067391536442128,
+      "learning_rate": 1.9878599802235662e-05,
+      "loss": 1.683,
+      "step": 205
+    },
+    {
+      "epoch": 0.059554784619832324,
+      "grad_norm": 1.407390031033922,
+      "learning_rate": 1.987717030353572e-05,
+      "loss": 1.7617,
+      "step": 206
+    },
+    {
+      "epoch": 0.0598438855160451,
+      "grad_norm": 1.2785798130045682,
+      "learning_rate": 1.9875732489788082e-05,
+      "loss": 1.6828,
+      "step": 207
+    },
+    {
+      "epoch": 0.06013298641225788,
+      "grad_norm": 1.581333677513783,
+      "learning_rate": 1.9874286362203162e-05,
+      "loss": 1.776,
+      "step": 208
+    },
+    {
+      "epoch": 0.06042208730847066,
+      "grad_norm": 1.3170624490173282,
+      "learning_rate": 1.987283192199837e-05,
+      "loss": 1.7548,
+      "step": 209
+    },
+    {
+      "epoch": 0.060711188204683436,
+      "grad_norm": 1.6773819921976063,
+      "learning_rate": 1.9871369170398126e-05,
+      "loss": 1.7935,
+      "step": 210
+    },
+    {
+      "epoch": 0.061000289100896214,
+      "grad_norm": 1.4063651245701678,
+      "learning_rate": 1.9869898108633834e-05,
+      "loss": 1.7806,
+      "step": 211
+    },
+    {
+      "epoch": 0.06128938999710899,
+      "grad_norm": 1.3241937165367996,
+      "learning_rate": 1.986841873794391e-05,
+      "loss": 1.7262,
+      "step": 212
+    },
+    {
+      "epoch": 0.06157849089332177,
+      "grad_norm": 1.49389208627189,
+      "learning_rate": 1.9866931059573748e-05,
+      "loss": 1.7103,
+      "step": 213
+    },
+    {
+      "epoch": 0.06186759178953455,
+      "grad_norm": 1.219442276366396,
+      "learning_rate": 1.9865435074775744e-05,
+      "loss": 1.6194,
+      "step": 214
+    },
+    {
+      "epoch": 0.062156692685747326,
+      "grad_norm": 1.3453821292169061,
+      "learning_rate": 1.9863930784809284e-05,
+      "loss": 1.8086,
+      "step": 215
+    },
+    {
+      "epoch": 0.062445793581960105,
+      "grad_norm": 1.2961743317843437,
+      "learning_rate": 1.986241819094075e-05,
+      "loss": 1.6352,
+      "step": 216
+    },
+    {
+      "epoch": 0.06273489447817288,
+      "grad_norm": 1.3729904036841547,
+      "learning_rate": 1.986089729444351e-05,
+      "loss": 1.7041,
+      "step": 217
+    },
+    {
+      "epoch": 0.06302399537438566,
+      "grad_norm": 1.3443809779979903,
+      "learning_rate": 1.9859368096597925e-05,
+      "loss": 1.7332,
+      "step": 218
+    },
+    {
+      "epoch": 0.06331309627059845,
+      "grad_norm": 1.3367070791840931,
+      "learning_rate": 1.9857830598691337e-05,
+      "loss": 1.639,
+      "step": 219
+    },
+    {
+      "epoch": 0.06360219716681122,
+      "grad_norm": 1.3799415387024292,
+      "learning_rate": 1.9856284802018087e-05,
+      "loss": 1.7675,
+      "step": 220
+    },
+    {
+      "epoch": 0.063891298063024,
+      "grad_norm": 1.3464389753341524,
+      "learning_rate": 1.9854730707879502e-05,
+      "loss": 1.6702,
+      "step": 221
+    },
+    {
+      "epoch": 0.06418039895923677,
+      "grad_norm": 1.3867842033565003,
+      "learning_rate": 1.9853168317583874e-05,
+      "loss": 1.7216,
+      "step": 222
+    },
+    {
+      "epoch": 0.06446949985544956,
+      "grad_norm": 1.2892516322419894,
+      "learning_rate": 1.9851597632446508e-05,
+      "loss": 1.6755,
+      "step": 223
+    },
+    {
+      "epoch": 0.06475860075166233,
+      "grad_norm": 1.4307425363074875,
+      "learning_rate": 1.9850018653789665e-05,
+      "loss": 1.7618,
+      "step": 224
+    },
+    {
+      "epoch": 0.06504770164787511,
+      "grad_norm": 1.288672937098248,
+      "learning_rate": 1.984843138294261e-05,
+      "loss": 1.756,
+      "step": 225
+    },
+    {
+      "epoch": 0.06533680254408789,
+      "grad_norm": 1.3540206108809039,
+      "learning_rate": 1.9846835821241573e-05,
+      "loss": 1.7373,
+      "step": 226
+    },
+    {
+      "epoch": 0.06562590344030067,
+      "grad_norm": 1.3108759943656667,
+      "learning_rate": 1.9845231970029774e-05,
+      "loss": 1.6429,
+      "step": 227
+    },
+    {
+      "epoch": 0.06591500433651344,
+      "grad_norm": 1.3549691917787245,
+      "learning_rate": 1.9843619830657404e-05,
+      "loss": 1.6692,
+      "step": 228
+    },
+    {
+      "epoch": 0.06620410523272623,
+      "grad_norm": 1.350566839669386,
+      "learning_rate": 1.984199940448164e-05,
+      "loss": 1.6499,
+      "step": 229
+    },
+    {
+      "epoch": 0.066493206128939,
+      "grad_norm": 1.3992430914185707,
+      "learning_rate": 1.9840370692866624e-05,
+      "loss": 1.7279,
+      "step": 230
+    },
+    {
+      "epoch": 0.06678230702515178,
+      "grad_norm": 1.336273448437935,
+      "learning_rate": 1.9838733697183482e-05,
+      "loss": 1.7378,
+      "step": 231
+    },
+    {
+      "epoch": 0.06707140792136455,
+      "grad_norm": 1.3714334108906678,
+      "learning_rate": 1.9837088418810312e-05,
+      "loss": 1.6845,
+      "step": 232
+    },
+    {
+      "epoch": 0.06736050881757734,
+      "grad_norm": 1.331192814515942,
+      "learning_rate": 1.983543485913218e-05,
+      "loss": 1.7605,
+      "step": 233
+    },
+    {
+      "epoch": 0.06764960971379011,
+      "grad_norm": 1.3590047127242209,
+      "learning_rate": 1.983377301954113e-05,
+      "loss": 1.7027,
+      "step": 234
+    },
+    {
+      "epoch": 0.0679387106100029,
+      "grad_norm": 1.2628103031697302,
+      "learning_rate": 1.983210290143617e-05,
+      "loss": 1.6682,
+      "step": 235
+    },
+    {
+      "epoch": 0.06822781150621567,
+      "grad_norm": 1.31345700201241,
+      "learning_rate": 1.9830424506223286e-05,
+      "loss": 1.7787,
+      "step": 236
+    },
+    {
+      "epoch": 0.06851691240242845,
+      "grad_norm": 1.367333629142093,
+      "learning_rate": 1.9828737835315422e-05,
+      "loss": 1.7182,
+      "step": 237
+    },
+    {
+      "epoch": 0.06880601329864122,
+      "grad_norm": 1.3814327986016506,
+      "learning_rate": 1.9827042890132498e-05,
+      "loss": 1.8058,
+      "step": 238
+    },
+    {
+      "epoch": 0.06909511419485401,
+      "grad_norm": 1.2616372891530419,
+      "learning_rate": 1.982533967210139e-05,
+      "loss": 1.7131,
+      "step": 239
+    },
+    {
+      "epoch": 0.06938421509106678,
+      "grad_norm": 1.3695087479555836,
+      "learning_rate": 1.982362818265595e-05,
+      "loss": 1.7163,
+      "step": 240
+    },
+    {
+      "epoch": 0.06967331598727956,
+      "grad_norm": 1.3557868876434347,
+      "learning_rate": 1.9821908423236983e-05,
+      "loss": 1.7421,
+      "step": 241
+    },
+    {
+      "epoch": 0.06996241688349233,
+      "grad_norm": 1.392442344450402,
+      "learning_rate": 1.982018039529226e-05,
+      "loss": 1.747,
+      "step": 242
+    },
+    {
+      "epoch": 0.07025151777970512,
+      "grad_norm": 1.4471159007355705,
+      "learning_rate": 1.9818444100276517e-05,
+      "loss": 1.6452,
+      "step": 243
+    },
+    {
+      "epoch": 0.07054061867591789,
+      "grad_norm": 1.323270395385573,
+      "learning_rate": 1.981669953965144e-05,
+      "loss": 1.795,
+      "step": 244
+    },
+    {
+      "epoch": 0.07082971957213068,
+      "grad_norm": 1.32959668728154,
+      "learning_rate": 1.9814946714885682e-05,
+      "loss": 1.8027,
+      "step": 245
+    },
+    {
+      "epoch": 0.07111882046834345,
+      "grad_norm": 1.3027587139671195,
+      "learning_rate": 1.981318562745485e-05,
+      "loss": 1.7326,
+      "step": 246
+    },
+    {
+      "epoch": 0.07140792136455623,
+      "grad_norm": 1.328453385919723,
+      "learning_rate": 1.9811416278841507e-05,
+      "loss": 1.6839,
+      "step": 247
+    },
+    {
+      "epoch": 0.071697022260769,
+      "grad_norm": 1.2271866822510065,
+      "learning_rate": 1.980963867053517e-05,
+      "loss": 1.6021,
+      "step": 248
+    },
+    {
+      "epoch": 0.07198612315698179,
+      "grad_norm": 1.2908284006992323,
+      "learning_rate": 1.9807852804032306e-05,
+      "loss": 1.6339,
+      "step": 249
+    },
+    {
+      "epoch": 0.07227522405319456,
+      "grad_norm": 1.373522694157661,
+      "learning_rate": 1.9806058680836343e-05,
+      "loss": 1.6902,
+      "step": 250
+    },
+    {
+      "epoch": 0.07256432494940734,
+      "grad_norm": 1.2634235256940958,
+      "learning_rate": 1.9804256302457653e-05,
+      "loss": 1.6517,
+      "step": 251
+    },
+    {
+      "epoch": 0.07285342584562012,
+      "grad_norm": 1.3278784508700443,
+      "learning_rate": 1.9802445670413562e-05,
+      "loss": 1.7972,
+      "step": 252
+    },
+    {
+      "epoch": 0.0731425267418329,
+      "grad_norm": 1.3552386657179452,
+      "learning_rate": 1.980062678622834e-05,
+      "loss": 1.726,
+      "step": 253
+    },
+    {
+      "epoch": 0.07343162763804567,
+      "grad_norm": 1.3742079147545718,
+      "learning_rate": 1.9798799651433204e-05,
+      "loss": 1.7317,
+      "step": 254
+    },
+    {
+      "epoch": 0.07372072853425846,
+      "grad_norm": 1.4074238143735376,
+      "learning_rate": 1.9796964267566322e-05,
+      "loss": 1.7275,
+      "step": 255
+    },
+    {
+      "epoch": 0.07400982943047124,
+      "grad_norm": 1.286876584367334,
+      "learning_rate": 1.9795120636172804e-05,
+      "loss": 1.6061,
+      "step": 256
+    },
+    {
+      "epoch": 0.07429893032668401,
+      "grad_norm": 1.3521349118231711,
+      "learning_rate": 1.9793268758804693e-05,
+      "loss": 1.7237,
+      "step": 257
+    },
+    {
+      "epoch": 0.0745880312228968,
+      "grad_norm": 1.2816583167182087,
+      "learning_rate": 1.979140863702099e-05,
+      "loss": 1.6783,
+      "step": 258
+    },
+    {
+      "epoch": 0.07487713211910957,
+      "grad_norm": 1.3634926357253074,
+      "learning_rate": 1.978954027238763e-05,
+      "loss": 1.6784,
+      "step": 259
+    },
+    {
+      "epoch": 0.07516623301532235,
+      "grad_norm": 1.4038229521295733,
+      "learning_rate": 1.9787663666477486e-05,
+      "loss": 1.6535,
+      "step": 260
+    },
+    {
+      "epoch": 0.07545533391153512,
+      "grad_norm": 1.2842218606303095,
+      "learning_rate": 1.9785778820870366e-05,
+      "loss": 1.7058,
+      "step": 261
+    },
+    {
+      "epoch": 0.07574443480774791,
+      "grad_norm": 1.347020075001843,
+      "learning_rate": 1.9783885737153022e-05,
+      "loss": 1.7832,
+      "step": 262
+    },
+    {
+      "epoch": 0.07603353570396068,
+      "grad_norm": 1.26053283819741,
+      "learning_rate": 1.9781984416919138e-05,
+      "loss": 1.644,
+      "step": 263
+    },
+    {
+      "epoch": 0.07632263660017347,
+      "grad_norm": 1.2495419382455348,
+      "learning_rate": 1.9780074861769328e-05,
+      "loss": 1.6807,
+      "step": 264
+    },
+    {
+      "epoch": 0.07661173749638624,
+      "grad_norm": 1.7139411191235945,
+      "learning_rate": 1.977815707331114e-05,
+      "loss": 1.7222,
+      "step": 265
+    },
+    {
+      "epoch": 0.07690083839259902,
+      "grad_norm": 1.4547863480275938,
+      "learning_rate": 1.9776231053159063e-05,
+      "loss": 1.7774,
+      "step": 266
+    },
+    {
+      "epoch": 0.07718993928881179,
+      "grad_norm": 1.3437347227693044,
+      "learning_rate": 1.97742968029345e-05,
+      "loss": 1.7208,
+      "step": 267
+    },
+    {
+      "epoch": 0.07747904018502458,
+      "grad_norm": 1.354739938343438,
+      "learning_rate": 1.9772354324265796e-05,
+      "loss": 1.6918,
+      "step": 268
+    },
+    {
+      "epoch": 0.07776814108123735,
+      "grad_norm": 1.2544728177655773,
+      "learning_rate": 1.977040361878821e-05,
+      "loss": 1.6351,
+      "step": 269
+    },
+    {
+      "epoch": 0.07805724197745013,
+      "grad_norm": 1.3417761983959458,
+      "learning_rate": 1.9768444688143938e-05,
+      "loss": 1.6564,
+      "step": 270
+    },
+    {
+      "epoch": 0.0783463428736629,
+      "grad_norm": 1.3422116229415715,
+      "learning_rate": 1.9766477533982094e-05,
+      "loss": 1.8033,
+      "step": 271
+    },
+    {
+      "epoch": 0.07863544376987569,
+      "grad_norm": 1.2706457122008528,
+      "learning_rate": 1.976450215795872e-05,
+      "loss": 1.6505,
+      "step": 272
+    },
+    {
+      "epoch": 0.07892454466608846,
+      "grad_norm": 1.267352186352676,
+      "learning_rate": 1.9762518561736782e-05,
+      "loss": 1.6591,
+      "step": 273
+    },
+    {
+      "epoch": 0.07921364556230125,
+      "grad_norm": 1.4466694988557338,
+      "learning_rate": 1.976052674698615e-05,
+      "loss": 1.7952,
+      "step": 274
+    },
+    {
+      "epoch": 0.07950274645851402,
+      "grad_norm": 1.321832538613218,
+      "learning_rate": 1.975852671538363e-05,
+      "loss": 1.7061,
+      "step": 275
+    },
+    {
+      "epoch": 0.0797918473547268,
+      "grad_norm": 1.2829050605849048,
+      "learning_rate": 1.9756518468612934e-05,
+      "loss": 1.6605,
+      "step": 276
+    },
+    {
+      "epoch": 0.08008094825093957,
+      "grad_norm": 1.2987764926867447,
+      "learning_rate": 1.9754502008364702e-05,
+      "loss": 1.7047,
+      "step": 277
+    },
+    {
+      "epoch": 0.08037004914715236,
+      "grad_norm": 1.2552473311885344,
+      "learning_rate": 1.9752477336336473e-05,
+      "loss": 1.7031,
+      "step": 278
+    },
+    {
+      "epoch": 0.08065915004336513,
+      "grad_norm": 1.1985068633897615,
+      "learning_rate": 1.9750444454232714e-05,
+      "loss": 1.7382,
+      "step": 279
+    },
+    {
+      "epoch": 0.08094825093957791,
+      "grad_norm": 1.314553485624065,
+      "learning_rate": 1.9748403363764795e-05,
+      "loss": 1.6491,
+      "step": 280
+    },
+    {
+      "epoch": 0.08123735183579069,
+      "grad_norm": 1.294998222736541,
+      "learning_rate": 1.9746354066650995e-05,
+      "loss": 1.7044,
+      "step": 281
+    },
+    {
+      "epoch": 0.08152645273200347,
+      "grad_norm": 1.3584436916798586,
+      "learning_rate": 1.974429656461651e-05,
+      "loss": 1.6672,
+      "step": 282
+    },
+    {
+      "epoch": 0.08181555362821624,
+      "grad_norm": 1.3102650044903212,
+      "learning_rate": 1.9742230859393432e-05,
+      "loss": 1.6954,
+      "step": 283
+    },
+    {
+      "epoch": 0.08210465452442903,
+      "grad_norm": 2.693221641429579,
+      "learning_rate": 1.974015695272077e-05,
+      "loss": 1.6296,
+      "step": 284
+    },
+    {
+      "epoch": 0.0823937554206418,
+      "grad_norm": 1.3351424022555927,
+      "learning_rate": 1.9738074846344427e-05,
+      "loss": 1.6594,
+      "step": 285
+    },
+    {
+      "epoch": 0.08268285631685458,
+      "grad_norm": 1.2766910430053877,
+      "learning_rate": 1.973598454201722e-05,
+      "loss": 1.6983,
+      "step": 286
+    },
+    {
+      "epoch": 0.08297195721306735,
+      "grad_norm": 1.4668406259400848,
+      "learning_rate": 1.973388604149886e-05,
+      "loss": 1.6698,
+      "step": 287
+    },
+    {
+      "epoch": 0.08326105810928014,
+      "grad_norm": 1.3104546778761357,
+      "learning_rate": 1.9731779346555957e-05,
+      "loss": 1.6723,
+      "step": 288
+    },
+    {
+      "epoch": 0.08355015900549291,
+      "grad_norm": 1.531005942756331,
+      "learning_rate": 1.9729664458962024e-05,
+      "loss": 1.7863,
+      "step": 289
+    },
+    {
+      "epoch": 0.0838392599017057,
+      "grad_norm": 1.3653225817314174,
+      "learning_rate": 1.9727541380497465e-05,
+      "loss": 1.7107,
+      "step": 290
+    },
+    {
+      "epoch": 0.08412836079791847,
+      "grad_norm": 1.3892591267410797,
+      "learning_rate": 1.972541011294959e-05,
+      "loss": 1.6403,
+      "step": 291
+    },
+    {
+      "epoch": 0.08441746169413125,
+      "grad_norm": 1.4118648874854935,
+      "learning_rate": 1.972327065811259e-05,
+      "loss": 1.6871,
+      "step": 292
+    },
+    {
+      "epoch": 0.08470656259034404,
+      "grad_norm": 1.4198103761024161,
+      "learning_rate": 1.972112301778756e-05,
+      "loss": 1.7095,
+      "step": 293
+    },
+    {
+      "epoch": 0.08499566348655681,
+      "grad_norm": 1.3896634687986935,
+      "learning_rate": 1.9718967193782475e-05,
+      "loss": 1.6347,
+      "step": 294
+    },
+    {
+      "epoch": 0.08528476438276959,
+      "grad_norm": 1.3431516951928908,
+      "learning_rate": 1.971680318791221e-05,
+      "loss": 1.7874,
+      "step": 295
+    },
+    {
+      "epoch": 0.08557386527898236,
+      "grad_norm": 1.3768126854675709,
+      "learning_rate": 1.9714631001998517e-05,
+      "loss": 1.6698,
+      "step": 296
+    },
+    {
+      "epoch": 0.08586296617519515,
+      "grad_norm": 1.452732246792645,
+      "learning_rate": 1.9712450637870048e-05,
+      "loss": 1.7896,
+      "step": 297
+    },
+    {
+      "epoch": 0.08615206707140792,
+      "grad_norm": 1.3268182615670008,
+      "learning_rate": 1.9710262097362323e-05,
+      "loss": 1.7162,
+      "step": 298
+    },
+    {
+      "epoch": 0.0864411679676207,
+      "grad_norm": 1.3451587005617647,
+      "learning_rate": 1.9708065382317763e-05,
+      "loss": 1.6132,
+      "step": 299
+    },
+    {
+      "epoch": 0.08673026886383348,
+      "grad_norm": 1.3567905514483718,
+      "learning_rate": 1.9705860494585653e-05,
+      "loss": 1.7181,
+      "step": 300
+    },
+    {
+      "epoch": 0.08701936976004626,
+      "grad_norm": 1.4178400258624095,
+      "learning_rate": 1.9703647436022174e-05,
+      "loss": 1.7197,
+      "step": 301
+    },
+    {
+      "epoch": 0.08730847065625903,
+      "grad_norm": 1.3176335169067535,
+      "learning_rate": 1.970142620849038e-05,
+      "loss": 1.6433,
+      "step": 302
+    },
+    {
+      "epoch": 0.08759757155247182,
+      "grad_norm": 1.3848318848405352,
+      "learning_rate": 1.9699196813860188e-05,
+      "loss": 1.7975,
+      "step": 303
+    },
+    {
+      "epoch": 0.08788667244868459,
+      "grad_norm": 1.4792413426843254,
+      "learning_rate": 1.9696959254008416e-05,
+      "loss": 1.7221,
+      "step": 304
+    },
+    {
+      "epoch": 0.08817577334489737,
+      "grad_norm": 1.3289471852066919,
+      "learning_rate": 1.9694713530818738e-05,
+      "loss": 1.7631,
+      "step": 305
+    },
+    {
+      "epoch": 0.08846487424111014,
+      "grad_norm": 1.4455763431259683,
+      "learning_rate": 1.9692459646181706e-05,
+      "loss": 1.758,
+      "step": 306
+    },
+    {
+      "epoch": 0.08875397513732293,
+      "grad_norm": 1.4156278432006244,
+      "learning_rate": 1.969019760199474e-05,
+      "loss": 1.7106,
+      "step": 307
+    },
+    {
+      "epoch": 0.0890430760335357,
+      "grad_norm": 1.335012985289393,
+      "learning_rate": 1.9687927400162134e-05,
+      "loss": 1.6303,
+      "step": 308
+    },
+    {
+      "epoch": 0.08933217692974849,
+      "grad_norm": 1.3515063736435737,
+      "learning_rate": 1.9685649042595045e-05,
+      "loss": 1.6907,
+      "step": 309
+    },
+    {
+      "epoch": 0.08962127782596126,
+      "grad_norm": 1.2906659764446389,
+      "learning_rate": 1.96833625312115e-05,
+      "loss": 1.7601,
+      "step": 310
+    },
+    {
+      "epoch": 0.08991037872217404,
+      "grad_norm": 1.4128720604920837,
+      "learning_rate": 1.968106786793638e-05,
+      "loss": 1.685,
+      "step": 311
+    },
+    {
+      "epoch": 0.09019947961838681,
+      "grad_norm": 1.3417671224975973,
+      "learning_rate": 1.967876505470144e-05,
+      "loss": 1.8098,
+      "step": 312
+    },
+    {
+      "epoch": 0.0904885805145996,
+      "grad_norm": 1.2524980838386586,
+      "learning_rate": 1.9676454093445293e-05,
+      "loss": 1.7347,
+      "step": 313
+    },
+    {
+      "epoch": 0.09077768141081237,
+      "grad_norm": 1.2903112596909276,
+      "learning_rate": 1.9674134986113407e-05,
+      "loss": 1.7183,
+      "step": 314
+    },
+    {
+      "epoch": 0.09106678230702515,
+      "grad_norm": 1.315346772447713,
+      "learning_rate": 1.9671807734658122e-05,
+      "loss": 1.666,
+      "step": 315
+    },
+    {
+      "epoch": 0.09135588320323793,
+      "grad_norm": 1.2815003343295048,
+      "learning_rate": 1.966947234103861e-05,
+      "loss": 1.6706,
+      "step": 316
+    },
+    {
+      "epoch": 0.09164498409945071,
+      "grad_norm": 1.3032057492518452,
+      "learning_rate": 1.9667128807220918e-05,
+      "loss": 1.6852,
+      "step": 317
+    },
+    {
+      "epoch": 0.09193408499566348,
+      "grad_norm": 1.399838923906023,
+      "learning_rate": 1.9664777135177936e-05,
+      "loss": 1.6786,
+      "step": 318
+    },
+    {
+      "epoch": 0.09222318589187627,
+      "grad_norm": 1.3183521495611081,
+      "learning_rate": 1.9662417326889408e-05,
+      "loss": 1.7656,
+      "step": 319
+    },
+    {
+      "epoch": 0.09251228678808904,
+      "grad_norm": 1.2919964291273878,
+      "learning_rate": 1.966004938434193e-05,
+      "loss": 1.7994,
+      "step": 320
+    },
+    {
+      "epoch": 0.09280138768430182,
+      "grad_norm": 1.331071824643648,
+      "learning_rate": 1.965767330952894e-05,
+      "loss": 1.7315,
+      "step": 321
+    },
+    {
+      "epoch": 0.0930904885805146,
+      "grad_norm": 1.3127780153887916,
+      "learning_rate": 1.9655289104450724e-05,
+      "loss": 1.58,
+      "step": 322
+    },
+    {
+      "epoch": 0.09337958947672738,
+      "grad_norm": 1.427592788785929,
+      "learning_rate": 1.9652896771114416e-05,
+      "loss": 1.6926,
+      "step": 323
+    },
+    {
+      "epoch": 0.09366869037294015,
+      "grad_norm": 1.3536283403103335,
+      "learning_rate": 1.9650496311533985e-05,
+      "loss": 1.6208,
+      "step": 324
+    },
+    {
+      "epoch": 0.09395779126915293,
+      "grad_norm": 1.3049249531971474,
+      "learning_rate": 1.9648087727730254e-05,
+      "loss": 1.7007,
+      "step": 325
+    },
+    {
+      "epoch": 0.0942468921653657,
+      "grad_norm": 8.074839924042823,
+      "learning_rate": 1.9645671021730876e-05,
+      "loss": 1.7401,
+      "step": 326
+    },
+    {
+      "epoch": 0.09453599306157849,
+      "grad_norm": 1.689958269781092,
+      "learning_rate": 1.9643246195570337e-05,
+      "loss": 1.7612,
+      "step": 327
+    },
+    {
+      "epoch": 0.09482509395779128,
+      "grad_norm": 1.3759192019092044,
+      "learning_rate": 1.9640813251289968e-05,
+      "loss": 1.6256,
+      "step": 328
+    },
+    {
+      "epoch": 0.09511419485400405,
+      "grad_norm": 1.2852662388113483,
+      "learning_rate": 1.9638372190937938e-05,
+      "loss": 1.5961,
+      "step": 329
+    },
+    {
+      "epoch": 0.09540329575021683,
+      "grad_norm": 1.3397296142416237,
+      "learning_rate": 1.963592301656923e-05,
+      "loss": 1.7228,
+      "step": 330
+    },
+    {
+      "epoch": 0.0956923966464296,
+      "grad_norm": 1.3050235554989664,
+      "learning_rate": 1.963346573024568e-05,
+      "loss": 1.631,
+      "step": 331
+    },
+    {
+      "epoch": 0.09598149754264239,
+      "grad_norm": 1.2687361679426568,
+      "learning_rate": 1.9631000334035936e-05,
+      "loss": 1.727,
+      "step": 332
+    },
+    {
+      "epoch": 0.09627059843885516,
+      "grad_norm": 1.4434478962876411,
+      "learning_rate": 1.962852683001548e-05,
+      "loss": 1.6118,
+      "step": 333
+    },
+    {
+      "epoch": 0.09655969933506794,
+      "grad_norm": 1.3835831592110455,
+      "learning_rate": 1.9626045220266626e-05,
+      "loss": 1.7184,
+      "step": 334
+    },
+    {
+      "epoch": 0.09684880023128072,
+      "grad_norm": 1.285345165652217,
+      "learning_rate": 1.96235555068785e-05,
+      "loss": 1.7396,
+      "step": 335
+    },
+    {
+      "epoch": 0.0971379011274935,
+      "grad_norm": 1.2741876410509978,
+      "learning_rate": 1.9621057691947052e-05,
+      "loss": 1.6827,
+      "step": 336
+    },
+    {
+      "epoch": 0.09742700202370627,
+      "grad_norm": 1.3414110861817643,
+      "learning_rate": 1.961855177757506e-05,
+      "loss": 1.7649,
+      "step": 337
+    },
+    {
+      "epoch": 0.09771610291991906,
+      "grad_norm": 1.3155148016487777,
+      "learning_rate": 1.9616037765872115e-05,
+      "loss": 1.6339,
+      "step": 338
+    },
+    {
+      "epoch": 0.09800520381613183,
+      "grad_norm": 1.3422815166211728,
+      "learning_rate": 1.9613515658954627e-05,
+      "loss": 1.6248,
+      "step": 339
+    },
+    {
+      "epoch": 0.09829430471234461,
+      "grad_norm": 1.2928851291446535,
+      "learning_rate": 1.9610985458945813e-05,
+      "loss": 1.6883,
+      "step": 340
+    },
+    {
+      "epoch": 0.09858340560855738,
+      "grad_norm": 1.310011271967797,
+      "learning_rate": 1.9608447167975716e-05,
+      "loss": 1.5808,
+      "step": 341
+    },
+    {
+      "epoch": 0.09887250650477017,
+      "grad_norm": 1.3065951865270455,
+      "learning_rate": 1.960590078818118e-05,
+      "loss": 1.6558,
+      "step": 342
+    },
+    {
+      "epoch": 0.09916160740098294,
+      "grad_norm": 1.3349828850162693,
+      "learning_rate": 1.960334632170587e-05,
+      "loss": 1.7289,
+      "step": 343
+    },
+    {
+      "epoch": 0.09945070829719573,
+      "grad_norm": 1.2943452424584643,
+      "learning_rate": 1.960078377070024e-05,
+      "loss": 1.613,
+      "step": 344
+    },
+    {
+      "epoch": 0.0997398091934085,
+      "grad_norm": 1.4861917307646695,
+      "learning_rate": 1.9598213137321568e-05,
+      "loss": 1.7105,
+      "step": 345
+    },
+    {
+      "epoch": 0.10002891008962128,
+      "grad_norm": 1.3722730686933504,
+      "learning_rate": 1.959563442373393e-05,
+      "loss": 1.7553,
+      "step": 346
+    },
+    {
+      "epoch": 0.10031801098583405,
+      "grad_norm": 1.3453966326217088,
+      "learning_rate": 1.95930476321082e-05,
+      "loss": 1.737,
+      "step": 347
+    },
+    {
+      "epoch": 0.10060711188204684,
+      "grad_norm": 1.3523784458893544,
+      "learning_rate": 1.9590452764622056e-05,
+      "loss": 1.6319,
+      "step": 348
+    },
+    {
+      "epoch": 0.10089621277825961,
+      "grad_norm": 1.3502487223737636,
+      "learning_rate": 1.9587849823459976e-05,
+      "loss": 1.6941,
+      "step": 349
+    },
+    {
+      "epoch": 0.1011853136744724,
+      "grad_norm": 1.398518128752155,
+      "learning_rate": 1.9585238810813233e-05,
+      "loss": 1.6475,
+      "step": 350
+    },
+    {
+      "epoch": 0.10147441457068516,
+      "grad_norm": 1.200679760577495,
+      "learning_rate": 1.9582619728879898e-05,
+      "loss": 1.7337,
+      "step": 351
+    },
+    {
+      "epoch": 0.10176351546689795,
+      "grad_norm": 1.3585311474803605,
+      "learning_rate": 1.9579992579864827e-05,
+      "loss": 1.7192,
+      "step": 352
+    },
+    {
+      "epoch": 0.10205261636311072,
+      "grad_norm": 1.4501829885257667,
+      "learning_rate": 1.9577357365979673e-05,
+      "loss": 1.7227,
+      "step": 353
+    },
+    {
+      "epoch": 0.1023417172593235,
+      "grad_norm": 1.2783896023412555,
+      "learning_rate": 1.9574714089442884e-05,
+      "loss": 1.7873,
+      "step": 354
+    },
+    {
+      "epoch": 0.10263081815553628,
+      "grad_norm": 1.2457865420249044,
+      "learning_rate": 1.9572062752479684e-05,
+      "loss": 1.6078,
+      "step": 355
+    },
+    {
+      "epoch": 0.10291991905174906,
+      "grad_norm": 1.298867064640292,
+      "learning_rate": 1.956940335732209e-05,
+      "loss": 1.7645,
+      "step": 356
+    },
+    {
+      "epoch": 0.10320901994796183,
+      "grad_norm": 1.2599727613590195,
+      "learning_rate": 1.95667359062089e-05,
+      "loss": 1.6698,
+      "step": 357
+    },
+    {
+      "epoch": 0.10349812084417462,
+      "grad_norm": 1.2508214829214046,
+      "learning_rate": 1.9564060401385697e-05,
+      "loss": 1.6279,
+      "step": 358
+    },
+    {
+      "epoch": 0.10378722174038739,
+      "grad_norm": 1.1832707796953907,
+      "learning_rate": 1.956137684510484e-05,
+      "loss": 1.5799,
+      "step": 359
+    },
+    {
+      "epoch": 0.10407632263660017,
+      "grad_norm": 1.3001383458399352,
+      "learning_rate": 1.9558685239625467e-05,
+      "loss": 1.6116,
+      "step": 360
+    },
+    {
+      "epoch": 0.10436542353281295,
+      "grad_norm": 1.2428470322542244,
+      "learning_rate": 1.9555985587213495e-05,
+      "loss": 1.6716,
+      "step": 361
+    },
+    {
+      "epoch": 0.10465452442902573,
+      "grad_norm": 1.3093773655524557,
+      "learning_rate": 1.955327789014161e-05,
+      "loss": 1.6915,
+      "step": 362
+    },
+    {
+      "epoch": 0.1049436253252385,
+      "grad_norm": 1.383021031949293,
+      "learning_rate": 1.9550562150689277e-05,
+      "loss": 1.7317,
+      "step": 363
+    },
+    {
+      "epoch": 0.10523272622145129,
+      "grad_norm": 1.329573981449402,
+      "learning_rate": 1.9547838371142726e-05,
+      "loss": 1.5913,
+      "step": 364
+    },
+    {
+      "epoch": 0.10552182711766407,
+      "grad_norm": 1.4166951368370655,
+      "learning_rate": 1.9545106553794962e-05,
+      "loss": 1.6004,
+      "step": 365
+    },
+    {
+      "epoch": 0.10581092801387684,
+      "grad_norm": 1.3377620749480776,
+      "learning_rate": 1.9542366700945746e-05,
+      "loss": 1.7907,
+      "step": 366
+    },
+    {
+      "epoch": 0.10610002891008963,
+      "grad_norm": 1.3388694333443414,
+      "learning_rate": 1.953961881490161e-05,
+      "loss": 1.6979,
+      "step": 367
+    },
+    {
+      "epoch": 0.1063891298063024,
+      "grad_norm": 1.3581005844418126,
+      "learning_rate": 1.9536862897975852e-05,
+      "loss": 1.6602,
+      "step": 368
+    },
+    {
+      "epoch": 0.10667823070251518,
+      "grad_norm": 1.3242684005167475,
+      "learning_rate": 1.9534098952488527e-05,
+      "loss": 1.7131,
+      "step": 369
+    },
+    {
+      "epoch": 0.10696733159872795,
+      "grad_norm": 1.2765833725967193,
+      "learning_rate": 1.9531326980766444e-05,
+      "loss": 1.7917,
+      "step": 370
+    },
+    {
+      "epoch": 0.10725643249494074,
+      "grad_norm": 1.2957317184999853,
+      "learning_rate": 1.952854698514318e-05,
+      "loss": 1.6569,
+      "step": 371
+    },
+    {
+      "epoch": 0.10754553339115351,
+      "grad_norm": 1.3050908294406351,
+      "learning_rate": 1.952575896795906e-05,
+      "loss": 1.5815,
+      "step": 372
+    },
+    {
+      "epoch": 0.1078346342873663,
+      "grad_norm": 1.4007267378768093,
+      "learning_rate": 1.952296293156116e-05,
+      "loss": 1.6437,
+      "step": 373
+    },
+    {
+      "epoch": 0.10812373518357907,
+      "grad_norm": 1.288112071905893,
+      "learning_rate": 1.952015887830331e-05,
+      "loss": 1.7148,
+      "step": 374
+    },
+    {
+      "epoch": 0.10841283607979185,
+      "grad_norm": 1.4132340648150645,
+      "learning_rate": 1.9517346810546094e-05,
+      "loss": 1.7236,
+      "step": 375
+    },
+    {
+      "epoch": 0.10870193697600462,
+      "grad_norm": 1.3520649123502502,
+      "learning_rate": 1.951452673065683e-05,
+      "loss": 1.7027,
+      "step": 376
+    },
+    {
+      "epoch": 0.10899103787221741,
+      "grad_norm": 1.344289195938231,
+      "learning_rate": 1.951169864100959e-05,
+      "loss": 1.6753,
+      "step": 377
+    },
+    {
+      "epoch": 0.10928013876843018,
+      "grad_norm": 1.2727655897790644,
+      "learning_rate": 1.950886254398519e-05,
+      "loss": 1.7595,
+      "step": 378
+    },
+    {
+      "epoch": 0.10956923966464296,
+      "grad_norm": 1.3488163579726409,
+      "learning_rate": 1.9506018441971186e-05,
+      "loss": 1.6687,
+      "step": 379
+    },
+    {
+      "epoch": 0.10985834056085574,
+      "grad_norm": 1.3474523993403258,
+      "learning_rate": 1.9503166337361863e-05,
+      "loss": 1.6676,
+      "step": 380
+    },
+    {
+      "epoch": 0.11014744145706852,
+      "grad_norm": 1.2758765009101956,
+      "learning_rate": 1.9500306232558263e-05,
+      "loss": 1.6571,
+      "step": 381
+    },
+    {
+      "epoch": 0.11043654235328129,
+      "grad_norm": 1.2196542542467945,
+      "learning_rate": 1.9497438129968144e-05,
+      "loss": 1.6551,
+      "step": 382
+    },
+    {
+      "epoch": 0.11072564324949408,
+      "grad_norm": 1.301338171513902,
+      "learning_rate": 1.9494562032006e-05,
+      "loss": 1.6548,
+      "step": 383
+    },
+    {
+      "epoch": 0.11101474414570685,
+      "grad_norm": 1.3203283874481997,
+      "learning_rate": 1.9491677941093075e-05,
+      "loss": 1.7338,
+      "step": 384
+    },
+    {
+      "epoch": 0.11130384504191963,
+      "grad_norm": 1.3024684784620577,
+      "learning_rate": 1.9488785859657314e-05,
+      "loss": 1.5528,
+      "step": 385
+    },
+    {
+      "epoch": 0.1115929459381324,
+      "grad_norm": 1.2625026147962017,
+      "learning_rate": 1.9485885790133402e-05,
+      "loss": 1.6967,
+      "step": 386
+    },
+    {
+      "epoch": 0.11188204683434519,
+      "grad_norm": 1.49068356270609,
+      "learning_rate": 1.9482977734962753e-05,
+      "loss": 1.7096,
+      "step": 387
+    },
+    {
+      "epoch": 0.11217114773055796,
+      "grad_norm": 1.3130185709588313,
+      "learning_rate": 1.9480061696593502e-05,
+      "loss": 1.7174,
+      "step": 388
+    },
+    {
+      "epoch": 0.11246024862677075,
+      "grad_norm": 1.2742915647508892,
+      "learning_rate": 1.9477137677480493e-05,
+      "loss": 1.6696,
+      "step": 389
+    },
+    {
+      "epoch": 0.11274934952298352,
+      "grad_norm": 1.3627324489936457,
+      "learning_rate": 1.9474205680085302e-05,
+      "loss": 1.6542,
+      "step": 390
+    },
+    {
+      "epoch": 0.1130384504191963,
+      "grad_norm": 1.3437349437643114,
+      "learning_rate": 1.947126570687622e-05,
+      "loss": 1.6747,
+      "step": 391
+    },
+    {
+      "epoch": 0.11332755131540907,
+      "grad_norm": 1.2678997121287945,
+      "learning_rate": 1.9468317760328244e-05,
+      "loss": 1.6918,
+      "step": 392
+    },
+    {
+      "epoch": 0.11361665221162186,
+      "grad_norm": 1.2929245484043537,
+      "learning_rate": 1.9465361842923087e-05,
+      "loss": 1.7217,
+      "step": 393
+    },
+    {
+      "epoch": 0.11390575310783463,
+      "grad_norm": 1.2499932038294317,
+      "learning_rate": 1.9462397957149182e-05,
+      "loss": 1.6128,
+      "step": 394
+    },
+    {
+      "epoch": 0.11419485400404741,
+      "grad_norm": 1.221557019735438,
+      "learning_rate": 1.9459426105501653e-05,
+      "loss": 1.6467,
+      "step": 395
+    },
+    {
+      "epoch": 0.11448395490026018,
+      "grad_norm": 1.2693539163839065,
+      "learning_rate": 1.945644629048234e-05,
+      "loss": 1.7283,
+      "step": 396
+    },
+    {
+      "epoch": 0.11477305579647297,
+      "grad_norm": 1.21274399309353,
+      "learning_rate": 1.9453458514599794e-05,
+      "loss": 1.7008,
+      "step": 397
+    },
+    {
+      "epoch": 0.11506215669268574,
+      "grad_norm": 1.2218147838259905,
+      "learning_rate": 1.945046278036925e-05,
+      "loss": 1.7437,
+      "step": 398
+    },
+    {
+      "epoch": 0.11535125758889853,
+      "grad_norm": 1.2968685896469103,
+      "learning_rate": 1.9447459090312656e-05,
+      "loss": 1.6983,
+      "step": 399
+    },
+    {
+      "epoch": 0.1156403584851113,
+      "grad_norm": 1.2334993392446711,
+      "learning_rate": 1.9444447446958658e-05,
+      "loss": 1.7061,
+      "step": 400
+    },
+    {
+      "epoch": 0.11592945938132408,
+      "grad_norm": 1.2984129611641557,
+      "learning_rate": 1.9441427852842584e-05,
+      "loss": 1.663,
+      "step": 401
+    },
+    {
+      "epoch": 0.11621856027753687,
+      "grad_norm": 1.2475965365995239,
+      "learning_rate": 1.943840031050647e-05,
+      "loss": 1.6427,
+      "step": 402
+    },
+    {
+      "epoch": 0.11650766117374964,
+      "grad_norm": 1.2694332854835886,
+      "learning_rate": 1.943536482249904e-05,
+      "loss": 1.7047,
+      "step": 403
+    },
+    {
+      "epoch": 0.11679676206996242,
+      "grad_norm": 1.2646968431962804,
+      "learning_rate": 1.94323213913757e-05,
+      "loss": 1.6885,
+      "step": 404
+    },
+    {
+      "epoch": 0.1170858629661752,
+      "grad_norm": 1.2293159012021218,
+      "learning_rate": 1.9429270019698553e-05,
+      "loss": 1.757,
+      "step": 405
+    },
+    {
+      "epoch": 0.11737496386238798,
+      "grad_norm": 1.2484085036943657,
+      "learning_rate": 1.9426210710036374e-05,
+      "loss": 1.6741,
+      "step": 406
+    },
+    {
+      "epoch": 0.11766406475860075,
+      "grad_norm": 1.43839701132394,
+      "learning_rate": 1.9423143464964633e-05,
+      "loss": 1.6768,
+      "step": 407
+    },
+    {
+      "epoch": 0.11795316565481354,
+      "grad_norm": 1.292908155403034,
+      "learning_rate": 1.9420068287065476e-05,
+      "loss": 1.6126,
+      "step": 408
+    },
+    {
+      "epoch": 0.1182422665510263,
+      "grad_norm": 1.370089960450461,
+      "learning_rate": 1.9416985178927724e-05,
+      "loss": 1.731,
+      "step": 409
+    },
+    {
+      "epoch": 0.11853136744723909,
+      "grad_norm": 1.3262176752311388,
+      "learning_rate": 1.941389414314688e-05,
+      "loss": 1.6579,
+      "step": 410
+    },
+    {
+      "epoch": 0.11882046834345186,
+      "grad_norm": 1.339305591546624,
+      "learning_rate": 1.9410795182325113e-05,
+      "loss": 1.7495,
+      "step": 411
+    },
+    {
+      "epoch": 0.11910956923966465,
+      "grad_norm": 1.341821261680916,
+      "learning_rate": 1.940768829907127e-05,
+      "loss": 1.7005,
+      "step": 412
+    },
+    {
+      "epoch": 0.11939867013587742,
+      "grad_norm": 1.276085151807024,
+      "learning_rate": 1.9404573496000867e-05,
+      "loss": 1.6043,
+      "step": 413
+    },
+    {
+      "epoch": 0.1196877710320902,
+      "grad_norm": 1.302570146214278,
+      "learning_rate": 1.940145077573608e-05,
+      "loss": 1.6722,
+      "step": 414
+    },
+    {
+      "epoch": 0.11997687192830297,
+      "grad_norm": 1.2582811301580585,
+      "learning_rate": 1.9398320140905765e-05,
+      "loss": 1.6383,
+      "step": 415
+    },
+    {
+      "epoch": 0.12026597282451576,
+      "grad_norm": 1.258733310784144,
+      "learning_rate": 1.9395181594145428e-05,
+      "loss": 1.5806,
+      "step": 416
+    },
+    {
+      "epoch": 0.12055507372072853,
+      "grad_norm": 1.2988890119500425,
+      "learning_rate": 1.9392035138097235e-05,
+      "loss": 1.7048,
+      "step": 417
+    },
+    {
+      "epoch": 0.12084417461694132,
+      "grad_norm": 1.29122939937491,
+      "learning_rate": 1.938888077541002e-05,
+      "loss": 1.5844,
+      "step": 418
+    },
+    {
+      "epoch": 0.12113327551315409,
+      "grad_norm": 1.2756402426612268,
+      "learning_rate": 1.9385718508739263e-05,
+      "loss": 1.6935,
+      "step": 419
+    },
+    {
+      "epoch": 0.12142237640936687,
+      "grad_norm": 1.3542547094512076,
+      "learning_rate": 1.9382548340747105e-05,
+      "loss": 1.6781,
+      "step": 420
+    },
+    {
+      "epoch": 0.12171147730557964,
+      "grad_norm": 1.3120014515081435,
+      "learning_rate": 1.937937027410234e-05,
+      "loss": 1.6601,
+      "step": 421
+    },
+    {
+      "epoch": 0.12200057820179243,
+      "grad_norm": 1.2955651895370563,
+      "learning_rate": 1.9376184311480404e-05,
+      "loss": 1.59,
+      "step": 422
+    },
+    {
+      "epoch": 0.1222896790980052,
+      "grad_norm": 1.3192882851086605,
+      "learning_rate": 1.9372990455563384e-05,
+      "loss": 1.6291,
+      "step": 423
+    },
+    {
+      "epoch": 0.12257877999421798,
+      "grad_norm": 1.4080370752052034,
+      "learning_rate": 1.9369788709040014e-05,
+      "loss": 1.5843,
+      "step": 424
+    },
+    {
+      "epoch": 0.12286788089043076,
+      "grad_norm": 1.309890861391924,
+      "learning_rate": 1.936657907460567e-05,
+      "loss": 1.7576,
+      "step": 425
+    },
+    {
+      "epoch": 0.12315698178664354,
+      "grad_norm": 1.3296661949737203,
+      "learning_rate": 1.936336155496236e-05,
+      "loss": 1.6412,
+      "step": 426
+    },
+    {
+      "epoch": 0.12344608268285631,
+      "grad_norm": 1.2137542022109604,
+      "learning_rate": 1.9360136152818746e-05,
+      "loss": 1.6677,
+      "step": 427
+    },
+    {
+      "epoch": 0.1237351835790691,
+      "grad_norm": 1.304478996513997,
+      "learning_rate": 1.9356902870890114e-05,
+      "loss": 1.6157,
+      "step": 428
+    },
+    {
+      "epoch": 0.12402428447528187,
+      "grad_norm": 1.2362598408529617,
+      "learning_rate": 1.9353661711898382e-05,
+      "loss": 1.6493,
+      "step": 429
+    },
+    {
+      "epoch": 0.12431338537149465,
+      "grad_norm": 1.3868990501411094,
+      "learning_rate": 1.9350412678572114e-05,
+      "loss": 1.7838,
+      "step": 430
+    },
+    {
+      "epoch": 0.12460248626770742,
+      "grad_norm": 1.3811144656508765,
+      "learning_rate": 1.934715577364648e-05,
+      "loss": 1.6164,
+      "step": 431
+    },
+    {
+      "epoch": 0.12489158716392021,
+      "grad_norm": 1.1886802269263785,
+      "learning_rate": 1.93438909998633e-05,
+      "loss": 1.6917,
+      "step": 432
+    },
+    {
+      "epoch": 0.125180688060133,
+      "grad_norm": 1.373861158324828,
+      "learning_rate": 1.9340618359971003e-05,
+      "loss": 1.7273,
+      "step": 433
+    },
+    {
+      "epoch": 0.12546978895634575,
+      "grad_norm": 1.360887411910304,
+      "learning_rate": 1.9337337856724647e-05,
+      "loss": 1.6129,
+      "step": 434
+    },
+    {
+      "epoch": 0.12575888985255854,
+      "grad_norm": 1.2310796622388105,
+      "learning_rate": 1.9334049492885906e-05,
+      "loss": 1.686,
+      "step": 435
+    },
+    {
+      "epoch": 0.12604799074877132,
+      "grad_norm": 1.2150119003782842,
+      "learning_rate": 1.9330753271223073e-05,
+      "loss": 1.7097,
+      "step": 436
+    },
+    {
+      "epoch": 0.1263370916449841,
+      "grad_norm": 1.3137735986745411,
+      "learning_rate": 1.932744919451106e-05,
+      "loss": 1.7312,
+      "step": 437
+    },
+    {
+      "epoch": 0.1266261925411969,
+      "grad_norm": 1.19581010090534,
+      "learning_rate": 1.9324137265531383e-05,
+      "loss": 1.7073,
+      "step": 438
+    },
+    {
+      "epoch": 0.12691529343740965,
+      "grad_norm": 1.1864958252257658,
+      "learning_rate": 1.9320817487072174e-05,
+      "loss": 1.6634,
+      "step": 439
+    },
+    {
+      "epoch": 0.12720439433362243,
+      "grad_norm": 1.2519130079679632,
+      "learning_rate": 1.9317489861928173e-05,
+      "loss": 1.6686,
+      "step": 440
+    },
+    {
+      "epoch": 0.12749349522983522,
+      "grad_norm": 1.2098081719210536,
+      "learning_rate": 1.9314154392900728e-05,
+      "loss": 1.697,
+      "step": 441
+    },
+    {
+      "epoch": 0.127782596126048,
+      "grad_norm": 1.297112805486722,
+      "learning_rate": 1.9310811082797784e-05,
+      "loss": 1.5979,
+      "step": 442
+    },
+    {
+      "epoch": 0.12807169702226076,
+      "grad_norm": 1.3006902970607843,
+      "learning_rate": 1.930745993443389e-05,
+      "loss": 1.5987,
+      "step": 443
+    },
+    {
+      "epoch": 0.12836079791847355,
+      "grad_norm": 1.3888684660127295,
+      "learning_rate": 1.9304100950630197e-05,
+      "loss": 1.6047,
+      "step": 444
+    },
+    {
+      "epoch": 0.12864989881468633,
+      "grad_norm": 1.3325945840074427,
+      "learning_rate": 1.9300734134214443e-05,
+      "loss": 1.6813,
+      "step": 445
+    },
+    {
+      "epoch": 0.12893899971089912,
+      "grad_norm": 1.3053169239838074,
+      "learning_rate": 1.9297359488020974e-05,
+      "loss": 1.8416,
+      "step": 446
+    },
+    {
+      "epoch": 0.12922810060711187,
+      "grad_norm": 1.1934876374596244,
+      "learning_rate": 1.9293977014890716e-05,
+      "loss": 1.7005,
+      "step": 447
+    },
+    {
+      "epoch": 0.12951720150332466,
+      "grad_norm": 1.2640391776060138,
+      "learning_rate": 1.929058671767119e-05,
+      "loss": 1.6578,
+      "step": 448
+    },
+    {
+      "epoch": 0.12980630239953744,
+      "grad_norm": 1.2646711203792806,
+      "learning_rate": 1.9287188599216492e-05,
+      "loss": 1.6699,
+      "step": 449
+    },
+    {
+      "epoch": 0.13009540329575023,
+      "grad_norm": 1.2320011875752572,
+      "learning_rate": 1.928378266238732e-05,
+      "loss": 1.6525,
+      "step": 450
+    },
+    {
+      "epoch": 0.13038450419196299,
+      "grad_norm": 1.34573408732214,
+      "learning_rate": 1.9280368910050943e-05,
+      "loss": 1.7015,
+      "step": 451
+    },
+    {
+      "epoch": 0.13067360508817577,
+      "grad_norm": 1.2353671099343633,
+      "learning_rate": 1.9276947345081213e-05,
+      "loss": 1.6397,
+      "step": 452
+    },
+    {
+      "epoch": 0.13096270598438856,
+      "grad_norm": 1.2197831661100311,
+      "learning_rate": 1.9273517970358557e-05,
+      "loss": 1.6836,
+      "step": 453
+    },
+    {
+      "epoch": 0.13125180688060134,
+      "grad_norm": 1.328367741401711,
+      "learning_rate": 1.9270080788769978e-05,
+      "loss": 1.7178,
+      "step": 454
+    },
+    {
+      "epoch": 0.1315409077768141,
+      "grad_norm": 1.2668567068996344,
+      "learning_rate": 1.9266635803209047e-05,
+      "loss": 1.7255,
+      "step": 455
+    },
+    {
+      "epoch": 0.13183000867302688,
+      "grad_norm": 1.2322035718739746,
+      "learning_rate": 1.9263183016575912e-05,
+      "loss": 1.612,
+      "step": 456
+    },
+    {
+      "epoch": 0.13211910956923967,
+      "grad_norm": 1.244606536596135,
+      "learning_rate": 1.9259722431777286e-05,
+      "loss": 1.7994,
+      "step": 457
+    },
+    {
+      "epoch": 0.13240821046545245,
+      "grad_norm": 1.2384126698619529,
+      "learning_rate": 1.9256254051726443e-05,
+      "loss": 1.6401,
+      "step": 458
+    },
+    {
+      "epoch": 0.1326973113616652,
+      "grad_norm": 1.2780953635907069,
+      "learning_rate": 1.925277787934322e-05,
+      "loss": 1.5826,
+      "step": 459
+    },
+    {
+      "epoch": 0.132986412257878,
+      "grad_norm": 1.2798396535904,
+      "learning_rate": 1.9249293917554017e-05,
+      "loss": 1.674,
+      "step": 460
+    },
+    {
+      "epoch": 0.13327551315409078,
+      "grad_norm": 1.2799203698146353,
+      "learning_rate": 1.924580216929179e-05,
+      "loss": 1.6331,
+      "step": 461
+    },
+    {
+      "epoch": 0.13356461405030357,
+      "grad_norm": 1.2111633406900368,
+      "learning_rate": 1.9242302637496057e-05,
+      "loss": 1.6682,
+      "step": 462
+    },
+    {
+      "epoch": 0.13385371494651632,
+      "grad_norm": 1.2112859487688281,
+      "learning_rate": 1.9238795325112867e-05,
+      "loss": 1.6484,
+      "step": 463
+    },
+    {
+      "epoch": 0.1341428158427291,
+      "grad_norm": 1.288135140665184,
+      "learning_rate": 1.9235280235094852e-05,
+      "loss": 1.626,
+      "step": 464
+    },
+    {
+      "epoch": 0.1344319167389419,
+      "grad_norm": 1.308513915152716,
+      "learning_rate": 1.923175737040116e-05,
+      "loss": 1.6597,
+      "step": 465
+    },
+    {
+      "epoch": 0.13472101763515468,
+      "grad_norm": 1.3262569041157928,
+      "learning_rate": 1.9228226733997503e-05,
+      "loss": 1.7642,
+      "step": 466
+    },
+    {
+      "epoch": 0.13501011853136743,
+      "grad_norm": 1.3424290380477188,
+      "learning_rate": 1.9224688328856127e-05,
+      "loss": 1.7465,
+      "step": 467
+    },
+    {
+      "epoch": 0.13529921942758022,
+      "grad_norm": 1.2651279914803906,
+      "learning_rate": 1.9221142157955825e-05,
+      "loss": 1.6126,
+      "step": 468
+    },
+    {
+      "epoch": 0.135588320323793,
+      "grad_norm": 1.3558655043001309,
+      "learning_rate": 1.921758822428192e-05,
+      "loss": 1.7556,
+      "step": 469
+    },
+    {
+      "epoch": 0.1358774212200058,
+      "grad_norm": 1.402001055732072,
+      "learning_rate": 1.921402653082628e-05,
+      "loss": 1.7992,
+      "step": 470
+    },
+    {
+      "epoch": 0.13616652211621855,
+      "grad_norm": 1.3182689558852307,
+      "learning_rate": 1.9210457080587285e-05,
+      "loss": 1.6262,
+      "step": 471
+    },
+    {
+      "epoch": 0.13645562301243133,
+      "grad_norm": 1.2723760551262082,
+      "learning_rate": 1.9206879876569877e-05,
+      "loss": 1.7064,
+      "step": 472
+    },
+    {
+      "epoch": 0.13674472390864412,
+      "grad_norm": 1.2954400935028445,
+      "learning_rate": 1.92032949217855e-05,
+      "loss": 1.6374,
+      "step": 473
+    },
+    {
+      "epoch": 0.1370338248048569,
+      "grad_norm": 1.2402948267042866,
+      "learning_rate": 1.9199702219252128e-05,
+      "loss": 1.5951,
+      "step": 474
+    },
+    {
+      "epoch": 0.1373229257010697,
+      "grad_norm": 1.2739630063945762,
+      "learning_rate": 1.919610177199426e-05,
+      "loss": 1.7167,
+      "step": 475
+    },
+    {
+      "epoch": 0.13761202659728244,
+      "grad_norm": 1.4113213475938031,
+      "learning_rate": 1.9192493583042922e-05,
+      "loss": 1.5751,
+      "step": 476
+    },
+    {
+      "epoch": 0.13790112749349523,
+      "grad_norm": 1.2872821986905296,
+      "learning_rate": 1.9188877655435645e-05,
+      "loss": 1.65,
+      "step": 477
+    },
+    {
+      "epoch": 0.13819022838970801,
+      "grad_norm": 1.2480655610937714,
+      "learning_rate": 1.918525399221648e-05,
+      "loss": 1.6882,
+      "step": 478
+    },
+    {
+      "epoch": 0.1384793292859208,
+      "grad_norm": 1.2681694886992483,
+      "learning_rate": 1.9181622596435993e-05,
+      "loss": 1.7528,
+      "step": 479
+    },
+    {
+      "epoch": 0.13876843018213356,
+      "grad_norm": 1.226404564427387,
+      "learning_rate": 1.917798347115125e-05,
+      "loss": 1.5918,
+      "step": 480
+    },
+    {
+      "epoch": 0.13905753107834634,
+      "grad_norm": 1.2687390154564733,
+      "learning_rate": 1.9174336619425842e-05,
+      "loss": 1.7533,
+      "step": 481
+    },
+    {
+      "epoch": 0.13934663197455913,
+      "grad_norm": 1.2254053559809934,
+      "learning_rate": 1.9170682044329845e-05,
+      "loss": 1.5968,
+      "step": 482
+    },
+    {
+      "epoch": 0.1396357328707719,
+      "grad_norm": 1.2732869875336201,
+      "learning_rate": 1.9167019748939847e-05,
+      "loss": 1.6944,
+      "step": 483
+    },
+    {
+      "epoch": 0.13992483376698467,
+      "grad_norm": 1.2541243190490785,
+      "learning_rate": 1.9163349736338935e-05,
+      "loss": 1.6237,
+      "step": 484
+    },
+    {
+      "epoch": 0.14021393466319745,
+      "grad_norm": 1.3291823171118335,
+      "learning_rate": 1.915967200961669e-05,
+      "loss": 1.5639,
+      "step": 485
+    },
+    {
+      "epoch": 0.14050303555941024,
+      "grad_norm": 1.2528167731427482,
+      "learning_rate": 1.9155986571869184e-05,
+      "loss": 1.7458,
+      "step": 486
+    },
+    {
+      "epoch": 0.14079213645562302,
+      "grad_norm": 1.2107095174484985,
+      "learning_rate": 1.9152293426198997e-05,
+      "loss": 1.6838,
+      "step": 487
+    },
+    {
+      "epoch": 0.14108123735183578,
+      "grad_norm": 1.2070210238330805,
+      "learning_rate": 1.9148592575715175e-05,
+      "loss": 1.6957,
+      "step": 488
+    },
+    {
+      "epoch": 0.14137033824804857,
+      "grad_norm": 1.3501466427051918,
+      "learning_rate": 1.9144884023533262e-05,
+      "loss": 1.6848,
+      "step": 489
+    },
+    {
+      "epoch": 0.14165943914426135,
+      "grad_norm": 1.289837942023481,
+      "learning_rate": 1.914116777277529e-05,
+      "loss": 1.6398,
+      "step": 490
+    },
+    {
+      "epoch": 0.14194854004047414,
+      "grad_norm": 1.2072886210991958,
+      "learning_rate": 1.9137443826569758e-05,
+      "loss": 1.6803,
+      "step": 491
+    },
+    {
+      "epoch": 0.1422376409366869,
+      "grad_norm": 1.2261256594899907,
+      "learning_rate": 1.9133712188051664e-05,
+      "loss": 1.8143,
+      "step": 492
+    },
+    {
+      "epoch": 0.14252674183289968,
+      "grad_norm": 1.191555412915695,
+      "learning_rate": 1.912997286036246e-05,
+      "loss": 1.7273,
+      "step": 493
+    },
+    {
+      "epoch": 0.14281584272911246,
+      "grad_norm": 1.2668769555610104,
+      "learning_rate": 1.9126225846650083e-05,
+      "loss": 1.6729,
+      "step": 494
+    },
+    {
+      "epoch": 0.14310494362532525,
+      "grad_norm": 1.2310680992910144,
+      "learning_rate": 1.912247115006894e-05,
+      "loss": 1.5509,
+      "step": 495
+    },
+    {
+      "epoch": 0.143394044521538,
+      "grad_norm": 1.2540456207457256,
+      "learning_rate": 1.91187087737799e-05,
+      "loss": 1.6512,
+      "step": 496
+    },
+    {
+      "epoch": 0.1436831454177508,
+      "grad_norm": 1.2983726556931725,
+      "learning_rate": 1.9114938720950307e-05,
+      "loss": 1.6434,
+      "step": 497
+    },
+    {
+      "epoch": 0.14397224631396358,
+      "grad_norm": 1.2424942795398506,
+      "learning_rate": 1.9111160994753957e-05,
+      "loss": 1.712,
+      "step": 498
+    },
+    {
+      "epoch": 0.14426134721017636,
+      "grad_norm": 1.2549250776078968,
+      "learning_rate": 1.9107375598371112e-05,
+      "loss": 1.5818,
+      "step": 499
+    },
+    {
+      "epoch": 0.14455044810638912,
+      "grad_norm": 1.171699859687026,
+      "learning_rate": 1.910358253498849e-05,
+      "loss": 1.6075,
+      "step": 500
+    },
+    {
+      "epoch": 0.1448395490026019,
+      "grad_norm": 1.2660287905231888,
+      "learning_rate": 1.9099781807799263e-05,
+      "loss": 1.6403,
+      "step": 501
+    },
+    {
+      "epoch": 0.1451286498988147,
+      "grad_norm": 1.2138215200444276,
+      "learning_rate": 1.9095973420003056e-05,
+      "loss": 1.6214,
+      "step": 502
+    },
+    {
+      "epoch": 0.14541775079502747,
+      "grad_norm": 1.2379532082411215,
+      "learning_rate": 1.909215737480594e-05,
+      "loss": 1.5915,
+      "step": 503
+    },
+    {
+      "epoch": 0.14570685169124023,
+      "grad_norm": 1.2691642883216934,
+      "learning_rate": 1.9088333675420432e-05,
+      "loss": 1.6913,
+      "step": 504
+    },
+    {
+      "epoch": 0.14599595258745302,
+      "grad_norm": 1.3365845490222084,
+      "learning_rate": 1.9084502325065494e-05,
+      "loss": 1.7174,
+      "step": 505
+    },
+    {
+      "epoch": 0.1462850534836658,
+      "grad_norm": 1.2933988368966916,
+      "learning_rate": 1.9080663326966536e-05,
+      "loss": 1.6947,
+      "step": 506
+    },
+    {
+      "epoch": 0.14657415437987859,
+      "grad_norm": 1.1997391314057113,
+      "learning_rate": 1.9076816684355398e-05,
+      "loss": 1.6454,
+      "step": 507
+    },
+    {
+      "epoch": 0.14686325527609134,
+      "grad_norm": 1.2253205782269874,
+      "learning_rate": 1.9072962400470348e-05,
+      "loss": 1.6475,
+      "step": 508
+    },
+    {
+      "epoch": 0.14715235617230413,
+      "grad_norm": 1.149822466017807,
+      "learning_rate": 1.9069100478556112e-05,
+      "loss": 1.5374,
+      "step": 509
+    },
+    {
+      "epoch": 0.1474414570685169,
+      "grad_norm": 1.2247546090753016,
+      "learning_rate": 1.906523092186382e-05,
+      "loss": 1.7436,
+      "step": 510
+    },
+    {
+      "epoch": 0.1477305579647297,
+      "grad_norm": 1.1999484360154251,
+      "learning_rate": 1.9061353733651036e-05,
+      "loss": 1.6747,
+      "step": 511
+    },
+    {
+      "epoch": 0.14801965886094248,
+      "grad_norm": 1.1947780812062576,
+      "learning_rate": 1.905746891718176e-05,
+      "loss": 1.59,
+      "step": 512
+    },
+    {
+      "epoch": 0.14830875975715524,
+      "grad_norm": 1.2483481364125735,
+      "learning_rate": 1.9053576475726406e-05,
+      "loss": 1.6343,
+      "step": 513
+    },
+    {
+      "epoch": 0.14859786065336802,
+      "grad_norm": 1.2303596001336339,
+      "learning_rate": 1.9049676412561803e-05,
+      "loss": 1.701,
+      "step": 514
+    },
+    {
+      "epoch": 0.1488869615495808,
+      "grad_norm": 1.2116845736311206,
+      "learning_rate": 1.9045768730971198e-05,
+      "loss": 1.7507,
+      "step": 515
+    },
+    {
+      "epoch": 0.1491760624457936,
+      "grad_norm": 1.299089146099033,
+      "learning_rate": 1.904185343424426e-05,
+      "loss": 1.7314,
+      "step": 516
+    },
+    {
+      "epoch": 0.14946516334200635,
+      "grad_norm": 1.2754215966625346,
+      "learning_rate": 1.9037930525677058e-05,
+      "loss": 1.6955,
+      "step": 517
+    },
+    {
+      "epoch": 0.14975426423821914,
+      "grad_norm": 1.2864098329743727,
+      "learning_rate": 1.9034000008572073e-05,
+      "loss": 1.6926,
+      "step": 518
+    },
+    {
+      "epoch": 0.15004336513443192,
+      "grad_norm": 1.2920675731077655,
+      "learning_rate": 1.9030061886238195e-05,
+      "loss": 1.6528,
+      "step": 519
+    },
+    {
+      "epoch": 0.1503324660306447,
+      "grad_norm": 1.2265090425552296,
+      "learning_rate": 1.9026116161990712e-05,
+      "loss": 1.647,
+      "step": 520
+    },
+    {
+      "epoch": 0.15062156692685746,
+      "grad_norm": 1.305532335335852,
+      "learning_rate": 1.902216283915131e-05,
+      "loss": 1.6085,
+      "step": 521
+    },
+    {
+      "epoch": 0.15091066782307025,
+      "grad_norm": 1.2996647218299158,
+      "learning_rate": 1.901820192104808e-05,
+      "loss": 1.6858,
+      "step": 522
+    },
+    {
+      "epoch": 0.15119976871928303,
+      "grad_norm": 1.2728823004232528,
+      "learning_rate": 1.9014233411015496e-05,
+      "loss": 1.6152,
+      "step": 523
+    },
+    {
+      "epoch": 0.15148886961549582,
+      "grad_norm": 1.237846119035289,
+      "learning_rate": 1.901025731239443e-05,
+      "loss": 1.7852,
+      "step": 524
+    },
+    {
+      "epoch": 0.15177797051170858,
+      "grad_norm": 1.263777772165666,
+      "learning_rate": 1.9006273628532146e-05,
+      "loss": 1.6714,
+      "step": 525
+    },
+    {
+      "epoch": 0.15206707140792136,
+      "grad_norm": 1.1966029829151998,
+      "learning_rate": 1.9002282362782284e-05,
+      "loss": 1.5772,
+      "step": 526
+    },
+    {
+      "epoch": 0.15235617230413415,
+      "grad_norm": 1.2573815442286682,
+      "learning_rate": 1.8998283518504874e-05,
+      "loss": 1.6586,
+      "step": 527
+    },
+    {
+      "epoch": 0.15264527320034693,
+      "grad_norm": 1.3315550643097303,
+      "learning_rate": 1.899427709906632e-05,
+      "loss": 1.5776,
+      "step": 528
+    },
+    {
+      "epoch": 0.1529343740965597,
+      "grad_norm": 1.27455167369078,
+      "learning_rate": 1.899026310783941e-05,
+      "loss": 1.6383,
+      "step": 529
+    },
+    {
+      "epoch": 0.15322347499277247,
+      "grad_norm": 1.2177716057481967,
+      "learning_rate": 1.8986241548203305e-05,
+      "loss": 1.7259,
+      "step": 530
+    },
+    {
+      "epoch": 0.15351257588898526,
+      "grad_norm": 1.4172388350158833,
+      "learning_rate": 1.898221242354353e-05,
+      "loss": 1.6084,
+      "step": 531
+    },
+    {
+      "epoch": 0.15380167678519804,
+      "grad_norm": 1.2977973632114062,
+      "learning_rate": 1.8978175737251984e-05,
+      "loss": 1.7148,
+      "step": 532
+    },
+    {
+      "epoch": 0.1540907776814108,
+      "grad_norm": 1.3288320048402222,
+      "learning_rate": 1.897413149272694e-05,
+      "loss": 1.7092,
+      "step": 533
+    },
+    {
+      "epoch": 0.15437987857762359,
+      "grad_norm": 1.362022094342137,
+      "learning_rate": 1.8970079693373014e-05,
+      "loss": 1.6727,
+      "step": 534
+    },
+    {
+      "epoch": 0.15466897947383637,
+      "grad_norm": 1.2868416836773502,
+      "learning_rate": 1.8966020342601204e-05,
+      "loss": 1.6544,
+      "step": 535
+    },
+    {
+      "epoch": 0.15495808037004916,
+      "grad_norm": 1.2255959793529978,
+      "learning_rate": 1.8961953443828853e-05,
+      "loss": 1.595,
+      "step": 536
+    },
+    {
+      "epoch": 0.1552471812662619,
+      "grad_norm": 1.2265187293683357,
+      "learning_rate": 1.8957879000479647e-05,
+      "loss": 1.6067,
+      "step": 537
+    },
+    {
+      "epoch": 0.1555362821624747,
+      "grad_norm": 1.198940426728034,
+      "learning_rate": 1.8953797015983654e-05,
+      "loss": 1.6146,
+      "step": 538
+    },
+    {
+      "epoch": 0.15582538305868748,
+      "grad_norm": 1.190556858810557,
+      "learning_rate": 1.8949707493777262e-05,
+      "loss": 1.699,
+      "step": 539
+    },
+    {
+      "epoch": 0.15611448395490027,
+      "grad_norm": 1.1336765732790504,
+      "learning_rate": 1.894561043730322e-05,
+      "loss": 1.5966,
+      "step": 540
+    },
+    {
+      "epoch": 0.15640358485111303,
+      "grad_norm": 1.1667137603744793,
+      "learning_rate": 1.8941505850010612e-05,
+      "loss": 1.6644,
+      "step": 541
+    },
+    {
+      "epoch": 0.1566926857473258,
+      "grad_norm": 1.3524401243741684,
+      "learning_rate": 1.8937393735354865e-05,
+      "loss": 1.6905,
+      "step": 542
+    },
+    {
+      "epoch": 0.1569817866435386,
+      "grad_norm": 1.2665686572396084,
+      "learning_rate": 1.8933274096797745e-05,
+      "loss": 1.5214,
+      "step": 543
+    },
+    {
+      "epoch": 0.15727088753975138,
+      "grad_norm": 1.2333567103961114,
+      "learning_rate": 1.892914693780735e-05,
+      "loss": 1.6044,
+      "step": 544
+    },
+    {
+      "epoch": 0.15755998843596414,
+      "grad_norm": 1.2700672132910338,
+      "learning_rate": 1.8925012261858103e-05,
+      "loss": 1.6295,
+      "step": 545
+    },
+    {
+      "epoch": 0.15784908933217692,
+      "grad_norm": 1.3269513720335189,
+      "learning_rate": 1.892087007243077e-05,
+      "loss": 1.5688,
+      "step": 546
+    },
+    {
+      "epoch": 0.1581381902283897,
+      "grad_norm": 1.444704489940057,
+      "learning_rate": 1.8916720373012425e-05,
+      "loss": 1.6048,
+      "step": 547
+    },
+    {
+      "epoch": 0.1584272911246025,
+      "grad_norm": 1.2362771720222352,
+      "learning_rate": 1.8912563167096478e-05,
+      "loss": 1.6548,
+      "step": 548
+    },
+    {
+      "epoch": 0.15871639202081528,
+      "grad_norm": 1.3841466139534755,
+      "learning_rate": 1.8908398458182647e-05,
+      "loss": 1.5571,
+      "step": 549
+    },
+    {
+      "epoch": 0.15900549291702804,
+      "grad_norm": 1.3571467399932051,
+      "learning_rate": 1.8904226249776977e-05,
+      "loss": 1.741,
+      "step": 550
+    },
+    {
+      "epoch": 0.15929459381324082,
+      "grad_norm": 1.3013296148852647,
+      "learning_rate": 1.8900046545391817e-05,
+      "loss": 1.6925,
+      "step": 551
+    },
+    {
+      "epoch": 0.1595836947094536,
+      "grad_norm": 1.4245052366468887,
+      "learning_rate": 1.8895859348545832e-05,
+      "loss": 1.6373,
+      "step": 552
+    },
+    {
+      "epoch": 0.1598727956056664,
+      "grad_norm": 1.4862334656509957,
+      "learning_rate": 1.8891664662763998e-05,
+      "loss": 1.7684,
+      "step": 553
+    },
+    {
+      "epoch": 0.16016189650187915,
+      "grad_norm": 1.4591040164361495,
+      "learning_rate": 1.8887462491577585e-05,
+      "loss": 1.6048,
+      "step": 554
+    },
+    {
+      "epoch": 0.16045099739809193,
+      "grad_norm": 1.3905790801269649,
+      "learning_rate": 1.8883252838524167e-05,
+      "loss": 1.7057,
+      "step": 555
+    },
+    {
+      "epoch": 0.16074009829430472,
+      "grad_norm": 1.296590703552253,
+      "learning_rate": 1.8879035707147634e-05,
+      "loss": 1.5708,
+      "step": 556
+    },
+    {
+      "epoch": 0.1610291991905175,
+      "grad_norm": 1.4224940976099907,
+      "learning_rate": 1.887481110099814e-05,
+      "loss": 1.6613,
+      "step": 557
+    },
+    {
+      "epoch": 0.16131830008673026,
+      "grad_norm": 1.209416335942298,
+      "learning_rate": 1.8870579023632163e-05,
+      "loss": 1.6734,
+      "step": 558
+    },
+    {
+      "epoch": 0.16160740098294304,
+      "grad_norm": 1.3856482467534985,
+      "learning_rate": 1.886633947861245e-05,
+      "loss": 1.6075,
+      "step": 559
+    },
+    {
+      "epoch": 0.16189650187915583,
+      "grad_norm": 1.298197860303536,
+      "learning_rate": 1.8862092469508043e-05,
+      "loss": 1.7309,
+      "step": 560
+    },
+    {
+      "epoch": 0.16218560277536861,
+      "grad_norm": 1.2884008702670706,
+      "learning_rate": 1.8857837999894265e-05,
+      "loss": 1.7367,
+      "step": 561
+    },
+    {
+      "epoch": 0.16247470367158137,
+      "grad_norm": 1.3295149830237742,
+      "learning_rate": 1.8853576073352724e-05,
+      "loss": 1.6373,
+      "step": 562
+    },
+    {
+      "epoch": 0.16276380456779416,
+      "grad_norm": 1.244943878603924,
+      "learning_rate": 1.8849306693471297e-05,
+      "loss": 1.7586,
+      "step": 563
+    },
+    {
+      "epoch": 0.16305290546400694,
+      "grad_norm": 1.4571917293027963,
+      "learning_rate": 1.884502986384414e-05,
+      "loss": 1.6308,
+      "step": 564
+    },
+    {
+      "epoch": 0.16334200636021973,
+      "grad_norm": 1.2273508652400458,
+      "learning_rate": 1.884074558807169e-05,
+      "loss": 1.7033,
+      "step": 565
+    },
+    {
+      "epoch": 0.16363110725643248,
+      "grad_norm": 1.3596367181968156,
+      "learning_rate": 1.8836453869760633e-05,
+      "loss": 1.7688,
+      "step": 566
+    },
+    {
+      "epoch": 0.16392020815264527,
+      "grad_norm": 1.1782222979069559,
+      "learning_rate": 1.883215471252394e-05,
+      "loss": 1.6864,
+      "step": 567
+    },
+    {
+      "epoch": 0.16420930904885805,
+      "grad_norm": 1.1876724035966852,
+      "learning_rate": 1.882784811998083e-05,
+      "loss": 1.651,
+      "step": 568
+    },
+    {
+      "epoch": 0.16449840994507084,
+      "grad_norm": 1.2807588292559342,
+      "learning_rate": 1.8823534095756787e-05,
+      "loss": 1.7127,
+      "step": 569
+    },
+    {
+      "epoch": 0.1647875108412836,
+      "grad_norm": 1.1959776867333827,
+      "learning_rate": 1.881921264348355e-05,
+      "loss": 1.6117,
+      "step": 570
+    },
+    {
+      "epoch": 0.16507661173749638,
+      "grad_norm": 1.1977710735486093,
+      "learning_rate": 1.881488376679912e-05,
+      "loss": 1.6682,
+      "step": 571
+    },
+    {
+      "epoch": 0.16536571263370917,
+      "grad_norm": 1.2133712471676765,
+      "learning_rate": 1.8810547469347737e-05,
+      "loss": 1.701,
+      "step": 572
+    },
+    {
+      "epoch": 0.16565481352992195,
+      "grad_norm": 1.1615194296116027,
+      "learning_rate": 1.880620375477989e-05,
+      "loss": 1.6957,
+      "step": 573
+    },
+    {
+      "epoch": 0.1659439144261347,
+      "grad_norm": 1.2434508263741582,
+      "learning_rate": 1.880185262675231e-05,
+      "loss": 1.6334,
+      "step": 574
+    },
+    {
+      "epoch": 0.1662330153223475,
+      "grad_norm": 1.1996625528866651,
+      "learning_rate": 1.8797494088927984e-05,
+      "loss": 1.7306,
+      "step": 575
+    },
+    {
+      "epoch": 0.16652211621856028,
+      "grad_norm": 1.25369862209721,
+      "learning_rate": 1.8793128144976114e-05,
+      "loss": 1.6611,
+      "step": 576
+    },
+    {
+      "epoch": 0.16681121711477306,
+      "grad_norm": 1.2478485036785194,
+      "learning_rate": 1.878875479857216e-05,
+      "loss": 1.6418,
+      "step": 577
+    },
+    {
+      "epoch": 0.16710031801098582,
+      "grad_norm": 1.3329254058696818,
+      "learning_rate": 1.8784374053397792e-05,
+      "loss": 1.6929,
+      "step": 578
+    },
+    {
+      "epoch": 0.1673894189071986,
+      "grad_norm": 1.3690716864028842,
+      "learning_rate": 1.8779985913140927e-05,
+      "loss": 1.6979,
+      "step": 579
+    },
+    {
+      "epoch": 0.1676785198034114,
+      "grad_norm": 1.2637256897353144,
+      "learning_rate": 1.877559038149569e-05,
+      "loss": 1.6229,
+      "step": 580
+    },
+    {
+      "epoch": 0.16796762069962418,
+      "grad_norm": 1.3196301095363738,
+      "learning_rate": 1.8771187462162452e-05,
+      "loss": 1.6806,
+      "step": 581
+    },
+    {
+      "epoch": 0.16825672159583693,
+      "grad_norm": 1.2751675322567242,
+      "learning_rate": 1.8766777158847777e-05,
+      "loss": 1.7331,
+      "step": 582
+    },
+    {
+      "epoch": 0.16854582249204972,
+      "grad_norm": 1.2217555466284777,
+      "learning_rate": 1.8762359475264464e-05,
+      "loss": 1.6439,
+      "step": 583
+    },
+    {
+      "epoch": 0.1688349233882625,
+      "grad_norm": 1.2138119129265534,
+      "learning_rate": 1.8757934415131513e-05,
+      "loss": 1.6229,
+      "step": 584
+    },
+    {
+      "epoch": 0.1691240242844753,
+      "grad_norm": 1.2481504467799762,
+      "learning_rate": 1.8753501982174142e-05,
+      "loss": 1.6527,
+      "step": 585
+    },
+    {
+      "epoch": 0.16941312518068807,
+      "grad_norm": 1.1551738364013855,
+      "learning_rate": 1.8749062180123778e-05,
+      "loss": 1.6202,
+      "step": 586
+    },
+    {
+      "epoch": 0.16970222607690083,
+      "grad_norm": 1.221158314927715,
+      "learning_rate": 1.8744615012718038e-05,
+      "loss": 1.674,
+      "step": 587
+    },
+    {
+      "epoch": 0.16999132697311362,
+      "grad_norm": 8.33538007172163,
+      "learning_rate": 1.8740160483700754e-05,
+      "loss": 1.7745,
+      "step": 588
+    },
+    {
+      "epoch": 0.1702804278693264,
+      "grad_norm": 1.2594866391024326,
+      "learning_rate": 1.8735698596821948e-05,
+      "loss": 1.7114,
+      "step": 589
+    },
+    {
+      "epoch": 0.17056952876553919,
+      "grad_norm": 1.1762832459510764,
+      "learning_rate": 1.8731229355837834e-05,
+      "loss": 1.6011,
+      "step": 590
+    },
+    {
+      "epoch": 0.17085862966175194,
+      "grad_norm": 1.2410876448625578,
+      "learning_rate": 1.8726752764510827e-05,
+      "loss": 1.5289,
+      "step": 591
+    },
+    {
+      "epoch": 0.17114773055796473,
+      "grad_norm": 1.2509478918932246,
+      "learning_rate": 1.8722268826609524e-05,
+      "loss": 1.6921,
+      "step": 592
+    },
+    {
+      "epoch": 0.1714368314541775,
+      "grad_norm": 1.2876270619453627,
+      "learning_rate": 1.87177775459087e-05,
+      "loss": 1.6632,
+      "step": 593
+    },
+    {
+      "epoch": 0.1717259323503903,
+      "grad_norm": 1.236444222374534,
+      "learning_rate": 1.8713278926189324e-05,
+      "loss": 1.628,
+      "step": 594
+    },
+    {
+      "epoch": 0.17201503324660306,
+      "grad_norm": 1.3325085992804107,
+      "learning_rate": 1.8708772971238528e-05,
+      "loss": 1.7139,
+      "step": 595
+    },
+    {
+      "epoch": 0.17230413414281584,
+      "grad_norm": 1.3054803605234377,
+      "learning_rate": 1.8704259684849636e-05,
+      "loss": 1.6715,
+      "step": 596
+    },
+    {
+      "epoch": 0.17259323503902863,
+      "grad_norm": 1.2114622828234745,
+      "learning_rate": 1.869973907082214e-05,
+      "loss": 1.6289,
+      "step": 597
+    },
+    {
+      "epoch": 0.1728823359352414,
+      "grad_norm": 1.3255943035899682,
+      "learning_rate": 1.8695211132961686e-05,
+      "loss": 1.5875,
+      "step": 598
+    },
+    {
+      "epoch": 0.17317143683145417,
+      "grad_norm": 1.2436093130787584,
+      "learning_rate": 1.86906758750801e-05,
+      "loss": 1.6891,
+      "step": 599
+    },
+    {
+      "epoch": 0.17346053772766695,
+      "grad_norm": 1.1789555197737465,
+      "learning_rate": 1.8686133300995374e-05,
+      "loss": 1.7097,
+      "step": 600
+    },
+    {
+      "epoch": 0.17374963862387974,
+      "grad_norm": 1.3205013238390153,
+      "learning_rate": 1.8681583414531644e-05,
+      "loss": 1.6392,
+      "step": 601
+    },
+    {
+      "epoch": 0.17403873952009252,
+      "grad_norm": 1.1648007546672043,
+      "learning_rate": 1.8677026219519212e-05,
+      "loss": 1.675,
+      "step": 602
+    },
+    {
+      "epoch": 0.17432784041630528,
+      "grad_norm": 1.2633263020507286,
+      "learning_rate": 1.8672461719794532e-05,
+      "loss": 1.5957,
+      "step": 603
+    },
+    {
+      "epoch": 0.17461694131251806,
+      "grad_norm": 1.2570430438589553,
+      "learning_rate": 1.8667889919200203e-05,
+      "loss": 1.7026,
+      "step": 604
+    },
+    {
+      "epoch": 0.17490604220873085,
+      "grad_norm": 1.203370562367749,
+      "learning_rate": 1.8663310821584976e-05,
+      "loss": 1.7054,
+      "step": 605
+    },
+    {
+      "epoch": 0.17519514310494363,
+      "grad_norm": 1.2387868298918792,
+      "learning_rate": 1.865872443080374e-05,
+      "loss": 1.7107,
+      "step": 606
+    },
+    {
+      "epoch": 0.1754842440011564,
+      "grad_norm": 1.197249289822212,
+      "learning_rate": 1.8654130750717526e-05,
+      "loss": 1.5961,
+      "step": 607
+    },
+    {
+      "epoch": 0.17577334489736918,
+      "grad_norm": 1.1922137775006185,
+      "learning_rate": 1.8649529785193502e-05,
+      "loss": 1.6349,
+      "step": 608
+    },
+    {
+      "epoch": 0.17606244579358196,
+      "grad_norm": 1.238013689830228,
+      "learning_rate": 1.8644921538104967e-05,
+      "loss": 1.6487,
+      "step": 609
+    },
+    {
+      "epoch": 0.17635154668979475,
+      "grad_norm": 1.198711220073165,
+      "learning_rate": 1.864030601333135e-05,
+      "loss": 1.6152,
+      "step": 610
+    },
+    {
+      "epoch": 0.1766406475860075,
+      "grad_norm": 1.1608845614417607,
+      "learning_rate": 1.8635683214758213e-05,
+      "loss": 1.6114,
+      "step": 611
+    },
+    {
+      "epoch": 0.1769297484822203,
+      "grad_norm": 1.223527743047107,
+      "learning_rate": 1.8631053146277233e-05,
+      "loss": 1.5324,
+      "step": 612
+    },
+    {
+      "epoch": 0.17721884937843307,
+      "grad_norm": 1.2244849270886753,
+      "learning_rate": 1.8626415811786215e-05,
+      "loss": 1.7019,
+      "step": 613
+    },
+    {
+      "epoch": 0.17750795027464586,
+      "grad_norm": 1.2422110644907864,
+      "learning_rate": 1.8621771215189065e-05,
+      "loss": 1.6713,
+      "step": 614
+    },
+    {
+      "epoch": 0.17779705117085862,
+      "grad_norm": 1.210470089961754,
+      "learning_rate": 1.861711936039583e-05,
+      "loss": 1.6569,
+      "step": 615
+    },
+    {
+      "epoch": 0.1780861520670714,
+      "grad_norm": 1.207776683125208,
+      "learning_rate": 1.8612460251322637e-05,
+      "loss": 1.5186,
+      "step": 616
+    },
+    {
+      "epoch": 0.1783752529632842,
+      "grad_norm": 1.2462777938743828,
+      "learning_rate": 1.8607793891891744e-05,
+      "loss": 1.7027,
+      "step": 617
+    },
+    {
+      "epoch": 0.17866435385949697,
+      "grad_norm": 1.2187274782133524,
+      "learning_rate": 1.8603120286031496e-05,
+      "loss": 1.7697,
+      "step": 618
+    },
+    {
+      "epoch": 0.17895345475570976,
+      "grad_norm": 1.2542678538432044,
+      "learning_rate": 1.859843943767635e-05,
+      "loss": 1.7017,
+      "step": 619
+    },
+    {
+      "epoch": 0.1792425556519225,
+      "grad_norm": 1.1778318765224127,
+      "learning_rate": 1.8593751350766854e-05,
+      "loss": 1.6475,
+      "step": 620
+    },
+    {
+      "epoch": 0.1795316565481353,
+      "grad_norm": 1.2711298413207304,
+      "learning_rate": 1.858905602924965e-05,
+      "loss": 1.6394,
+      "step": 621
+    },
+    {
+      "epoch": 0.17982075744434808,
+      "grad_norm": 1.232207257276827,
+      "learning_rate": 1.8584353477077476e-05,
+      "loss": 1.6074,
+      "step": 622
+    },
+    {
+      "epoch": 0.18010985834056087,
+      "grad_norm": 1.2540297946919436,
+      "learning_rate": 1.857964369820915e-05,
+      "loss": 1.6683,
+      "step": 623
+    },
+    {
+      "epoch": 0.18039895923677363,
+      "grad_norm": 1.2340677553615835,
+      "learning_rate": 1.857492669660957e-05,
+      "loss": 1.6716,
+      "step": 624
+    },
+    {
+      "epoch": 0.1806880601329864,
+      "grad_norm": 1.2065806455145902,
+      "learning_rate": 1.857020247624973e-05,
+      "loss": 1.5365,
+      "step": 625
+    },
+    {
+      "epoch": 0.1809771610291992,
+      "grad_norm": 1.2263400735274184,
+      "learning_rate": 1.856547104110669e-05,
+      "loss": 1.5683,
+      "step": 626
+    },
+    {
+      "epoch": 0.18126626192541198,
+      "grad_norm": 1.347386958745164,
+      "learning_rate": 1.8560732395163585e-05,
+      "loss": 1.7436,
+      "step": 627
+    },
+    {
+      "epoch": 0.18155536282162474,
+      "grad_norm": 1.3051300929582412,
+      "learning_rate": 1.8555986542409615e-05,
+      "loss": 1.7476,
+      "step": 628
+    },
+    {
+      "epoch": 0.18184446371783752,
+      "grad_norm": 1.3479071566939564,
+      "learning_rate": 1.855123348684006e-05,
+      "loss": 1.6623,
+      "step": 629
+    },
+    {
+      "epoch": 0.1821335646140503,
+      "grad_norm": 1.2568804583760171,
+      "learning_rate": 1.8546473232456255e-05,
+      "loss": 1.6444,
+      "step": 630
+    },
+    {
+      "epoch": 0.1824226655102631,
+      "grad_norm": 1.2426342867282067,
+      "learning_rate": 1.85417057832656e-05,
+      "loss": 1.5964,
+      "step": 631
+    },
+    {
+      "epoch": 0.18271176640647585,
+      "grad_norm": 1.2638177775731838,
+      "learning_rate": 1.853693114328154e-05,
+      "loss": 1.5322,
+      "step": 632
+    },
+    {
+      "epoch": 0.18300086730268864,
+      "grad_norm": 1.164069058129512,
+      "learning_rate": 1.8532149316523587e-05,
+      "loss": 1.5849,
+      "step": 633
+    },
+    {
+      "epoch": 0.18328996819890142,
+      "grad_norm": 1.297248418393498,
+      "learning_rate": 1.8527360307017306e-05,
+      "loss": 1.7388,
+      "step": 634
+    },
+    {
+      "epoch": 0.1835790690951142,
+      "grad_norm": 1.1341179063142641,
+      "learning_rate": 1.852256411879429e-05,
+      "loss": 1.7354,
+      "step": 635
+    },
+    {
+      "epoch": 0.18386816999132696,
+      "grad_norm": 1.215237098350405,
+      "learning_rate": 1.8517760755892197e-05,
+      "loss": 1.6552,
+      "step": 636
+    },
+    {
+      "epoch": 0.18415727088753975,
+      "grad_norm": 1.1982757446916235,
+      "learning_rate": 1.8512950222354704e-05,
+      "loss": 1.6313,
+      "step": 637
+    },
+    {
+      "epoch": 0.18444637178375253,
+      "grad_norm": 1.2063061953012788,
+      "learning_rate": 1.850813252223154e-05,
+      "loss": 1.6108,
+      "step": 638
+    },
+    {
+      "epoch": 0.18473547267996532,
+      "grad_norm": 1.2120232512101812,
+      "learning_rate": 1.8503307659578466e-05,
+      "loss": 1.6357,
+      "step": 639
+    },
+    {
+      "epoch": 0.18502457357617808,
+      "grad_norm": 1.2479587047666476,
+      "learning_rate": 1.8498475638457265e-05,
+      "loss": 1.6444,
+      "step": 640
+    },
+    {
+      "epoch": 0.18531367447239086,
+      "grad_norm": 1.2758952934748575,
+      "learning_rate": 1.849363646293575e-05,
+      "loss": 1.632,
+      "step": 641
+    },
+    {
+      "epoch": 0.18560277536860365,
+      "grad_norm": 1.1490779391882489,
+      "learning_rate": 1.8488790137087764e-05,
+      "loss": 1.6316,
+      "step": 642
+    },
+    {
+      "epoch": 0.18589187626481643,
+      "grad_norm": 1.195110289142274,
+      "learning_rate": 1.8483936664993152e-05,
+      "loss": 1.6018,
+      "step": 643
+    },
+    {
+      "epoch": 0.1861809771610292,
+      "grad_norm": 1.283174200034784,
+      "learning_rate": 1.847907605073779e-05,
+      "loss": 1.6488,
+      "step": 644
+    },
+    {
+      "epoch": 0.18647007805724197,
+      "grad_norm": 1.2662030686633299,
+      "learning_rate": 1.847420829841357e-05,
+      "loss": 1.6538,
+      "step": 645
+    },
+    {
+      "epoch": 0.18675917895345476,
+      "grad_norm": 1.2237194592018654,
+      "learning_rate": 1.8469333412118375e-05,
+      "loss": 1.6171,
+      "step": 646
+    },
+    {
+      "epoch": 0.18704827984966754,
+      "grad_norm": 1.155988769874379,
+      "learning_rate": 1.846445139595611e-05,
+      "loss": 1.5969,
+      "step": 647
+    },
+    {
+      "epoch": 0.1873373807458803,
+      "grad_norm": 1.2416911437703613,
+      "learning_rate": 1.845956225403667e-05,
+      "loss": 1.6744,
+      "step": 648
+    },
+    {
+      "epoch": 0.18762648164209308,
+      "grad_norm": 1.1875358817766466,
+      "learning_rate": 1.8454665990475967e-05,
+      "loss": 1.5515,
+      "step": 649
+    },
+    {
+      "epoch": 0.18791558253830587,
+      "grad_norm": 1.1818241708803934,
+      "learning_rate": 1.844976260939589e-05,
+      "loss": 1.5487,
+      "step": 650
+    },
+    {
+      "epoch": 0.18820468343451865,
+      "grad_norm": 1.1840647734438647,
+      "learning_rate": 1.8444852114924325e-05,
+      "loss": 1.6148,
+      "step": 651
+    },
+    {
+      "epoch": 0.1884937843307314,
+      "grad_norm": 1.2706827502060694,
+      "learning_rate": 1.8439934511195147e-05,
+      "loss": 1.6112,
+      "step": 652
+    },
+    {
+      "epoch": 0.1887828852269442,
+      "grad_norm": 1.220131649175792,
+      "learning_rate": 1.843500980234822e-05,
+      "loss": 1.6449,
+      "step": 653
+    },
+    {
+      "epoch": 0.18907198612315698,
+      "grad_norm": 1.103922527658955,
+      "learning_rate": 1.8430077992529388e-05,
+      "loss": 1.6444,
+      "step": 654
+    },
+    {
+      "epoch": 0.18936108701936977,
+      "grad_norm": 1.1899809619524633,
+      "learning_rate": 1.8425139085890466e-05,
+      "loss": 1.5967,
+      "step": 655
+    },
+    {
+      "epoch": 0.18965018791558255,
+      "grad_norm": 1.2244806367070153,
+      "learning_rate": 1.8420193086589252e-05,
+      "loss": 1.6504,
+      "step": 656
+    },
+    {
+      "epoch": 0.1899392888117953,
+      "grad_norm": 1.1912067775204556,
+      "learning_rate": 1.841523999878951e-05,
+      "loss": 1.718,
+      "step": 657
+    },
+    {
+      "epoch": 0.1902283897080081,
+      "grad_norm": 1.2117478146838403,
+      "learning_rate": 1.8410279826660976e-05,
+      "loss": 1.5804,
+      "step": 658
+    },
+    {
+      "epoch": 0.19051749060422088,
+      "grad_norm": 1.1341524351911334,
+      "learning_rate": 1.840531257437934e-05,
+      "loss": 1.6287,
+      "step": 659
+    },
+    {
+      "epoch": 0.19080659150043366,
+      "grad_norm": 1.1914356537636899,
+      "learning_rate": 1.8400338246126265e-05,
+      "loss": 1.6563,
+      "step": 660
+    },
+    {
+      "epoch": 0.19109569239664642,
+      "grad_norm": 1.4105624776944174,
+      "learning_rate": 1.839535684608936e-05,
+      "loss": 1.7139,
+      "step": 661
+    },
+    {
+      "epoch": 0.1913847932928592,
+      "grad_norm": 1.1645723203046823,
+      "learning_rate": 1.8390368378462197e-05,
+      "loss": 1.7166,
+      "step": 662
+    },
+    {
+      "epoch": 0.191673894189072,
+      "grad_norm": 1.2936969076152187,
+      "learning_rate": 1.8385372847444296e-05,
+      "loss": 1.6482,
+      "step": 663
+    },
+    {
+      "epoch": 0.19196299508528478,
+      "grad_norm": 1.1569033962294648,
+      "learning_rate": 1.838037025724111e-05,
+      "loss": 1.6843,
+      "step": 664
+    },
+    {
+      "epoch": 0.19225209598149753,
+      "grad_norm": 1.1405851915704683,
+      "learning_rate": 1.837536061206405e-05,
+      "loss": 1.6033,
+      "step": 665
+    },
+    {
+      "epoch": 0.19254119687771032,
+      "grad_norm": 1.2542382381366322,
+      "learning_rate": 1.8370343916130467e-05,
+      "loss": 1.6706,
+      "step": 666
+    },
+    {
+      "epoch": 0.1928302977739231,
+      "grad_norm": 1.2159546224332614,
+      "learning_rate": 1.8365320173663638e-05,
+      "loss": 1.6918,
+      "step": 667
+    },
+    {
+      "epoch": 0.1931193986701359,
+      "grad_norm": 1.2532817994557603,
+      "learning_rate": 1.8360289388892773e-05,
+      "loss": 1.6683,
+      "step": 668
+    },
+    {
+      "epoch": 0.19340849956634865,
+      "grad_norm": 1.2395321203333016,
+      "learning_rate": 1.8355251566053016e-05,
+      "loss": 1.629,
+      "step": 669
+    },
+    {
+      "epoch": 0.19369760046256143,
+      "grad_norm": 1.1689572668191017,
+      "learning_rate": 1.8350206709385437e-05,
+      "loss": 1.6462,
+      "step": 670
+    },
+    {
+      "epoch": 0.19398670135877422,
+      "grad_norm": 1.1981131507977363,
+      "learning_rate": 1.8345154823137015e-05,
+      "loss": 1.5874,
+      "step": 671
+    },
+    {
+      "epoch": 0.194275802254987,
+      "grad_norm": 1.1854854067275211,
+      "learning_rate": 1.834009591156067e-05,
+      "loss": 1.5968,
+      "step": 672
+    },
+    {
+      "epoch": 0.19456490315119976,
+      "grad_norm": 1.1769715429764516,
+      "learning_rate": 1.8335029978915207e-05,
+      "loss": 1.5856,
+      "step": 673
+    },
+    {
+      "epoch": 0.19485400404741254,
+      "grad_norm": 1.1759193873525724,
+      "learning_rate": 1.8329957029465367e-05,
+      "loss": 1.676,
+      "step": 674
+    },
+    {
+      "epoch": 0.19514310494362533,
+      "grad_norm": 1.2034916407402159,
+      "learning_rate": 1.8324877067481782e-05,
+      "loss": 1.6385,
+      "step": 675
+    },
+    {
+      "epoch": 0.1954322058398381,
+      "grad_norm": 1.1498736100928149,
+      "learning_rate": 1.8319790097240998e-05,
+      "loss": 1.5622,
+      "step": 676
+    },
+    {
+      "epoch": 0.19572130673605087,
+      "grad_norm": 1.1833854630281373,
+      "learning_rate": 1.8314696123025456e-05,
+      "loss": 1.5906,
+      "step": 677
+    },
+    {
+      "epoch": 0.19601040763226366,
+      "grad_norm": 1.3064848599083938,
+      "learning_rate": 1.8309595149123486e-05,
+      "loss": 1.5784,
+      "step": 678
+    },
+    {
+      "epoch": 0.19629950852847644,
+      "grad_norm": 1.207598354497605,
+      "learning_rate": 1.8304487179829325e-05,
+      "loss": 1.6724,
+      "step": 679
+    },
+    {
+      "epoch": 0.19658860942468923,
+      "grad_norm": 1.3657596862796557,
+      "learning_rate": 1.829937221944309e-05,
+      "loss": 1.6808,
+      "step": 680
+    },
+    {
+      "epoch": 0.19687771032090198,
+      "grad_norm": 1.2300109610049414,
+      "learning_rate": 1.8294250272270787e-05,
+      "loss": 1.701,
+      "step": 681
+    },
+    {
+      "epoch": 0.19716681121711477,
+      "grad_norm": 1.1932414325288254,
+      "learning_rate": 1.82891213426243e-05,
+      "loss": 1.6844,
+      "step": 682
+    },
+    {
+      "epoch": 0.19745591211332755,
+      "grad_norm": 1.2040899317538816,
+      "learning_rate": 1.8283985434821394e-05,
+      "loss": 1.6674,
+      "step": 683
+    },
+    {
+      "epoch": 0.19774501300954034,
+      "grad_norm": 1.20051829187703,
+      "learning_rate": 1.827884255318571e-05,
+      "loss": 1.6988,
+      "step": 684
+    },
+    {
+      "epoch": 0.1980341139057531,
+      "grad_norm": 1.1808862977245562,
+      "learning_rate": 1.8273692702046754e-05,
+      "loss": 1.5823,
+      "step": 685
+    },
+    {
+      "epoch": 0.19832321480196588,
+      "grad_norm": 1.226726509340714,
+      "learning_rate": 1.8268535885739905e-05,
+      "loss": 1.6371,
+      "step": 686
+    },
+    {
+      "epoch": 0.19861231569817867,
+      "grad_norm": 1.254783885869565,
+      "learning_rate": 1.8263372108606404e-05,
+      "loss": 1.5849,
+      "step": 687
+    },
+    {
+      "epoch": 0.19890141659439145,
+      "grad_norm": 1.2210257028134839,
+      "learning_rate": 1.8258201374993355e-05,
+      "loss": 1.7316,
+      "step": 688
+    },
+    {
+      "epoch": 0.1991905174906042,
+      "grad_norm": 1.27099587011664,
+      "learning_rate": 1.8253023689253707e-05,
+      "loss": 1.592,
+      "step": 689
+    },
+    {
+      "epoch": 0.199479618386817,
+      "grad_norm": 1.366689582202721,
+      "learning_rate": 1.8247839055746276e-05,
+      "loss": 1.5789,
+      "step": 690
+    },
+    {
+      "epoch": 0.19976871928302978,
+      "grad_norm": 1.1755442721898772,
+      "learning_rate": 1.8242647478835717e-05,
+      "loss": 1.6069,
+      "step": 691
+    },
+    {
+      "epoch": 0.20005782017924256,
+      "grad_norm": 1.2600721970141062,
+      "learning_rate": 1.8237448962892537e-05,
+      "loss": 1.7384,
+      "step": 692
+    },
+    {
+      "epoch": 0.20034692107545535,
+      "grad_norm": 1.22621954725729,
+      "learning_rate": 1.823224351229308e-05,
+      "loss": 1.6226,
+      "step": 693
+    },
+    {
+      "epoch": 0.2006360219716681,
+      "grad_norm": 1.2182097639154552,
+      "learning_rate": 1.822703113141953e-05,
+      "loss": 1.613,
+      "step": 694
+    },
+    {
+      "epoch": 0.2009251228678809,
+      "grad_norm": 1.2785953541313604,
+      "learning_rate": 1.82218118246599e-05,
+      "loss": 1.7014,
+      "step": 695
+    },
+    {
+      "epoch": 0.20121422376409367,
+      "grad_norm": 1.2338324878833378,
+      "learning_rate": 1.821658559640804e-05,
+      "loss": 1.6298,
+      "step": 696
+    },
+    {
+      "epoch": 0.20150332466030646,
+      "grad_norm": 1.1768453406356614,
+      "learning_rate": 1.821135245106363e-05,
+      "loss": 1.5619,
+      "step": 697
+    },
+    {
+      "epoch": 0.20179242555651922,
+      "grad_norm": 1.1771764186726315,
+      "learning_rate": 1.820611239303216e-05,
+      "loss": 1.68,
+      "step": 698
+    },
+    {
+      "epoch": 0.202081526452732,
+      "grad_norm": 1.1751259370580003,
+      "learning_rate": 1.8200865426724955e-05,
+      "loss": 1.6353,
+      "step": 699
+    },
+    {
+      "epoch": 0.2023706273489448,
+      "grad_norm": 1.234717870608629,
+      "learning_rate": 1.819561155655914e-05,
+      "loss": 1.6222,
+      "step": 700
+    },
+    {
+      "epoch": 0.20265972824515757,
+      "grad_norm": 1.3036974701284907,
+      "learning_rate": 1.819035078695767e-05,
+      "loss": 1.6272,
+      "step": 701
+    },
+    {
+      "epoch": 0.20294882914137033,
+      "grad_norm": 1.201996899703672,
+      "learning_rate": 1.8185083122349288e-05,
+      "loss": 1.6977,
+      "step": 702
+    },
+    {
+      "epoch": 0.20323793003758311,
+      "grad_norm": 1.2513029094302637,
+      "learning_rate": 1.8179808567168556e-05,
+      "loss": 1.7145,
+      "step": 703
+    },
+    {
+      "epoch": 0.2035270309337959,
+      "grad_norm": 1.2829160939047408,
+      "learning_rate": 1.8174527125855836e-05,
+      "loss": 1.7475,
+      "step": 704
+    },
+    {
+      "epoch": 0.20381613183000868,
+      "grad_norm": 1.2634189347114115,
+      "learning_rate": 1.8169238802857275e-05,
+      "loss": 1.7567,
+      "step": 705
+    },
+    {
+      "epoch": 0.20410523272622144,
+      "grad_norm": 1.1500402066275863,
+      "learning_rate": 1.8163943602624834e-05,
+      "loss": 1.568,
+      "step": 706
+    },
+    {
+      "epoch": 0.20439433362243423,
+      "grad_norm": 1.2208940822026542,
+      "learning_rate": 1.815864152961624e-05,
+      "loss": 1.6101,
+      "step": 707
+    },
+    {
+      "epoch": 0.204683434518647,
+      "grad_norm": 1.2920340236775505,
+      "learning_rate": 1.8153332588295023e-05,
+      "loss": 1.6039,
+      "step": 708
+    },
+    {
+      "epoch": 0.2049725354148598,
+      "grad_norm": 1.3082533794058193,
+      "learning_rate": 1.814801678313049e-05,
+      "loss": 1.66,
+      "step": 709
+    },
+    {
+      "epoch": 0.20526163631107255,
+      "grad_norm": 1.2031595547172316,
+      "learning_rate": 1.8142694118597727e-05,
+      "loss": 1.6835,
+      "step": 710
+    },
+    {
+      "epoch": 0.20555073720728534,
+      "grad_norm": 1.1799641826600815,
+      "learning_rate": 1.8137364599177587e-05,
+      "loss": 1.6254,
+      "step": 711
+    },
+    {
+      "epoch": 0.20583983810349812,
+      "grad_norm": 1.208426243281953,
+      "learning_rate": 1.8132028229356703e-05,
+      "loss": 1.6405,
+      "step": 712
+    },
+    {
+      "epoch": 0.2061289389997109,
+      "grad_norm": 1.1403284646294907,
+      "learning_rate": 1.8126685013627475e-05,
+      "loss": 1.7065,
+      "step": 713
+    },
+    {
+      "epoch": 0.20641803989592367,
+      "grad_norm": 1.2249537048855292,
+      "learning_rate": 1.8121334956488058e-05,
+      "loss": 1.6447,
+      "step": 714
+    },
+    {
+      "epoch": 0.20670714079213645,
+      "grad_norm": 1.189614154915905,
+      "learning_rate": 1.8115978062442375e-05,
+      "loss": 1.5539,
+      "step": 715
+    },
+    {
+      "epoch": 0.20699624168834924,
+      "grad_norm": 1.1643490997811285,
+      "learning_rate": 1.81106143360001e-05,
+      "loss": 1.6068,
+      "step": 716
+    },
+    {
+      "epoch": 0.20728534258456202,
+      "grad_norm": 1.1646757540028276,
+      "learning_rate": 1.8105243781676663e-05,
+      "loss": 1.5891,
+      "step": 717
+    },
+    {
+      "epoch": 0.20757444348077478,
+      "grad_norm": 1.2441086225997955,
+      "learning_rate": 1.809986640399323e-05,
+      "loss": 1.529,
+      "step": 718
+    },
+    {
+      "epoch": 0.20786354437698756,
+      "grad_norm": 1.3159032556741779,
+      "learning_rate": 1.809448220747673e-05,
+      "loss": 1.5943,
+      "step": 719
+    },
+    {
+      "epoch": 0.20815264527320035,
+      "grad_norm": 1.1977613211999474,
+      "learning_rate": 1.8089091196659822e-05,
+      "loss": 1.6186,
+      "step": 720
+    },
+    {
+      "epoch": 0.20844174616941313,
+      "grad_norm": 1.1611286191966523,
+      "learning_rate": 1.8083693376080896e-05,
+      "loss": 1.609,
+      "step": 721
+    },
+    {
+      "epoch": 0.2087308470656259,
+      "grad_norm": 1.2479677009723977,
+      "learning_rate": 1.807828875028408e-05,
+      "loss": 1.6632,
+      "step": 722
+    },
+    {
+      "epoch": 0.20901994796183868,
+      "grad_norm": 1.2198060446129206,
+      "learning_rate": 1.8072877323819247e-05,
+      "loss": 1.6938,
+      "step": 723
+    },
+    {
+      "epoch": 0.20930904885805146,
+      "grad_norm": 1.1959812561872514,
+      "learning_rate": 1.806745910124196e-05,
+      "loss": 1.6715,
+      "step": 724
+    },
+    {
+      "epoch": 0.20959814975426425,
+      "grad_norm": 1.1644575837628937,
+      "learning_rate": 1.8062034087113538e-05,
+      "loss": 1.654,
+      "step": 725
+    },
+    {
+      "epoch": 0.209887250650477,
+      "grad_norm": 1.1952595776641282,
+      "learning_rate": 1.8056602286000995e-05,
+      "loss": 1.7361,
+      "step": 726
+    },
+    {
+      "epoch": 0.2101763515466898,
+      "grad_norm": 1.1187248069836293,
+      "learning_rate": 1.8051163702477067e-05,
+      "loss": 1.4882,
+      "step": 727
+    },
+    {
+      "epoch": 0.21046545244290257,
+      "grad_norm": 1.23860503080743,
+      "learning_rate": 1.80457183411202e-05,
+      "loss": 1.6494,
+      "step": 728
+    },
+    {
+      "epoch": 0.21075455333911536,
+      "grad_norm": 1.1198629149288368,
+      "learning_rate": 1.8040266206514548e-05,
+      "loss": 1.6625,
+      "step": 729
+    },
+    {
+      "epoch": 0.21104365423532814,
+      "grad_norm": 1.1312334064032668,
+      "learning_rate": 1.8034807303249956e-05,
+      "loss": 1.644,
+      "step": 730
+    },
+    {
+      "epoch": 0.2113327551315409,
+      "grad_norm": 1.178023621775733,
+      "learning_rate": 1.8029341635921985e-05,
+      "loss": 1.6237,
+      "step": 731
+    },
+    {
+      "epoch": 0.21162185602775369,
+      "grad_norm": 1.1831551646009177,
+      "learning_rate": 1.8023869209131876e-05,
+      "loss": 1.684,
+      "step": 732
+    },
+    {
+      "epoch": 0.21191095692396647,
+      "grad_norm": 1.1269798393522688,
+      "learning_rate": 1.8018390027486564e-05,
+      "loss": 1.6614,
+      "step": 733
+    },
+    {
+      "epoch": 0.21220005782017926,
+      "grad_norm": 1.163617004294811,
+      "learning_rate": 1.8012904095598672e-05,
+      "loss": 1.5863,
+      "step": 734
+    },
+    {
+      "epoch": 0.212489158716392,
+      "grad_norm": 1.1666239101175062,
+      "learning_rate": 1.8007411418086504e-05,
+      "loss": 1.6286,
+      "step": 735
+    },
+    {
+      "epoch": 0.2127782596126048,
+      "grad_norm": 1.1783363837507064,
+      "learning_rate": 1.8001911999574048e-05,
+      "loss": 1.6024,
+      "step": 736
+    },
+    {
+      "epoch": 0.21306736050881758,
+      "grad_norm": 1.1324429879608466,
+      "learning_rate": 1.7996405844690955e-05,
+      "loss": 1.6086,
+      "step": 737
+    },
+    {
+      "epoch": 0.21335646140503037,
+      "grad_norm": 1.1857513831899675,
+      "learning_rate": 1.7990892958072562e-05,
+      "loss": 1.6637,
+      "step": 738
+    },
+    {
+      "epoch": 0.21364556230124312,
+      "grad_norm": 1.20546977750714,
+      "learning_rate": 1.798537334435986e-05,
+      "loss": 1.651,
+      "step": 739
+    },
+    {
+      "epoch": 0.2139346631974559,
+      "grad_norm": 1.214143713272185,
+      "learning_rate": 1.7979847008199515e-05,
+      "loss": 1.5663,
+      "step": 740
+    },
+    {
+      "epoch": 0.2142237640936687,
+      "grad_norm": 1.2482881773540793,
+      "learning_rate": 1.7974313954243846e-05,
+      "loss": 1.6634,
+      "step": 741
+    },
+    {
+      "epoch": 0.21451286498988148,
+      "grad_norm": 1.3788124938610393,
+      "learning_rate": 1.796877418715082e-05,
+      "loss": 1.7858,
+      "step": 742
+    },
+    {
+      "epoch": 0.21480196588609424,
+      "grad_norm": 1.210426850026346,
+      "learning_rate": 1.796322771158407e-05,
+      "loss": 1.5901,
+      "step": 743
+    },
+    {
+      "epoch": 0.21509106678230702,
+      "grad_norm": 1.1818842755255796,
+      "learning_rate": 1.7957674532212862e-05,
+      "loss": 1.5516,
+      "step": 744
+    },
+    {
+      "epoch": 0.2153801676785198,
+      "grad_norm": 1.2331450769496057,
+      "learning_rate": 1.7952114653712123e-05,
+      "loss": 1.6224,
+      "step": 745
+    },
+    {
+      "epoch": 0.2156692685747326,
+      "grad_norm": 1.1415088042700048,
+      "learning_rate": 1.7946548080762402e-05,
+      "loss": 1.7262,
+      "step": 746
+    },
+    {
+      "epoch": 0.21595836947094535,
+      "grad_norm": 1.0990551729161935,
+      "learning_rate": 1.7940974818049898e-05,
+      "loss": 1.674,
+      "step": 747
+    },
+    {
+      "epoch": 0.21624747036715813,
+      "grad_norm": 1.217412085009134,
+      "learning_rate": 1.7935394870266425e-05,
+      "loss": 1.6138,
+      "step": 748
+    },
+    {
+      "epoch": 0.21653657126337092,
+      "grad_norm": 1.167638566385817,
+      "learning_rate": 1.7929808242109444e-05,
+      "loss": 1.6414,
+      "step": 749
+    },
+    {
+      "epoch": 0.2168256721595837,
+      "grad_norm": 1.2003093360263695,
+      "learning_rate": 1.7924214938282026e-05,
+      "loss": 1.6289,
+      "step": 750
+    },
+    {
+      "epoch": 0.21711477305579646,
+      "grad_norm": 1.2685172755141476,
+      "learning_rate": 1.7918614963492872e-05,
+      "loss": 1.7019,
+      "step": 751
+    },
+    {
+      "epoch": 0.21740387395200925,
+      "grad_norm": 1.173005539771057,
+      "learning_rate": 1.7913008322456283e-05,
+      "loss": 1.6812,
+      "step": 752
+    },
+    {
+      "epoch": 0.21769297484822203,
+      "grad_norm": 1.2159228008333007,
+      "learning_rate": 1.7907395019892196e-05,
+      "loss": 1.6077,
+      "step": 753
+    },
+    {
+      "epoch": 0.21798207574443482,
+      "grad_norm": 1.1742389210088622,
+      "learning_rate": 1.7901775060526132e-05,
+      "loss": 1.622,
+      "step": 754
+    },
+    {
+      "epoch": 0.21827117664064757,
+      "grad_norm": 1.2128889995971577,
+      "learning_rate": 1.789614844908923e-05,
+      "loss": 1.6743,
+      "step": 755
+    },
+    {
+      "epoch": 0.21856027753686036,
+      "grad_norm": 1.1667752847351143,
+      "learning_rate": 1.7890515190318222e-05,
+      "loss": 1.5754,
+      "step": 756
+    },
+    {
+      "epoch": 0.21884937843307314,
+      "grad_norm": 1.1295223300664317,
+      "learning_rate": 1.7884875288955448e-05,
+      "loss": 1.6738,
+      "step": 757
+    },
+    {
+      "epoch": 0.21913847932928593,
+      "grad_norm": 1.261068879100486,
+      "learning_rate": 1.7879228749748824e-05,
+      "loss": 1.6835,
+      "step": 758
+    },
+    {
+      "epoch": 0.2194275802254987,
+      "grad_norm": 1.1728226481913466,
+      "learning_rate": 1.7873575577451863e-05,
+      "loss": 1.6891,
+      "step": 759
+    },
+    {
+      "epoch": 0.21971668112171147,
+      "grad_norm": 1.1787350920697617,
+      "learning_rate": 1.7867915776823663e-05,
+      "loss": 1.6257,
+      "step": 760
+    },
+    {
+      "epoch": 0.22000578201792426,
+      "grad_norm": 1.1620635102842405,
+      "learning_rate": 1.7862249352628897e-05,
+      "loss": 1.6228,
+      "step": 761
+    },
+    {
+      "epoch": 0.22029488291413704,
+      "grad_norm": 1.174838883761586,
+      "learning_rate": 1.785657630963782e-05,
+      "loss": 1.6795,
+      "step": 762
+    },
+    {
+      "epoch": 0.2205839838103498,
+      "grad_norm": 1.2540807508404137,
+      "learning_rate": 1.785089665262625e-05,
+      "loss": 1.6921,
+      "step": 763
+    },
+    {
+      "epoch": 0.22087308470656258,
+      "grad_norm": 1.1328480065063387,
+      "learning_rate": 1.784521038637558e-05,
+      "loss": 1.6261,
+      "step": 764
+    },
+    {
+      "epoch": 0.22116218560277537,
+      "grad_norm": 1.1913711901780004,
+      "learning_rate": 1.783951751567277e-05,
+      "loss": 1.7268,
+      "step": 765
+    },
+    {
+      "epoch": 0.22145128649898815,
+      "grad_norm": 1.1168162011581586,
+      "learning_rate": 1.7833818045310335e-05,
+      "loss": 1.6462,
+      "step": 766
+    },
+    {
+      "epoch": 0.22174038739520094,
+      "grad_norm": 1.1597448679252735,
+      "learning_rate": 1.782811198008634e-05,
+      "loss": 1.6369,
+      "step": 767
+    },
+    {
+      "epoch": 0.2220294882914137,
+      "grad_norm": 1.1983400772535562,
+      "learning_rate": 1.782239932480441e-05,
+      "loss": 1.6555,
+      "step": 768
+    },
+    {
+      "epoch": 0.22231858918762648,
+      "grad_norm": 1.1771360817965915,
+      "learning_rate": 1.7816680084273724e-05,
+      "loss": 1.6722,
+      "step": 769
+    },
+    {
+      "epoch": 0.22260769008383927,
+      "grad_norm": 1.202362947433072,
+      "learning_rate": 1.7810954263308987e-05,
+      "loss": 1.6489,
+      "step": 770
+    },
+    {
+      "epoch": 0.22289679098005205,
+      "grad_norm": 1.1514930888062975,
+      "learning_rate": 1.780522186673046e-05,
+      "loss": 1.5065,
+      "step": 771
+    },
+    {
+      "epoch": 0.2231858918762648,
+      "grad_norm": 1.2543833283472365,
+      "learning_rate": 1.779948289936393e-05,
+      "loss": 1.6361,
+      "step": 772
+    },
+    {
+      "epoch": 0.2234749927724776,
+      "grad_norm": 1.1103075126565838,
+      "learning_rate": 1.779373736604072e-05,
+      "loss": 1.5877,
+      "step": 773
+    },
+    {
+      "epoch": 0.22376409366869038,
+      "grad_norm": 1.2145730232558314,
+      "learning_rate": 1.7787985271597678e-05,
+      "loss": 1.586,
+      "step": 774
+    },
+    {
+      "epoch": 0.22405319456490316,
+      "grad_norm": 1.2583750794569701,
+      "learning_rate": 1.7782226620877178e-05,
+      "loss": 1.6928,
+      "step": 775
+    },
+    {
+      "epoch": 0.22434229546111592,
+      "grad_norm": 1.192600045488377,
+      "learning_rate": 1.777646141872711e-05,
+      "loss": 1.6711,
+      "step": 776
+    },
+    {
+      "epoch": 0.2246313963573287,
+      "grad_norm": 1.2325135805846623,
+      "learning_rate": 1.7770689670000882e-05,
+      "loss": 1.6051,
+      "step": 777
+    },
+    {
+      "epoch": 0.2249204972535415,
+      "grad_norm": 1.2445267162996139,
+      "learning_rate": 1.7764911379557415e-05,
+      "loss": 1.6704,
+      "step": 778
+    },
+    {
+      "epoch": 0.22520959814975428,
+      "grad_norm": 1.164473435396523,
+      "learning_rate": 1.7759126552261133e-05,
+      "loss": 1.6788,
+      "step": 779
+    },
+    {
+      "epoch": 0.22549869904596703,
+      "grad_norm": 1.3267375597230986,
+      "learning_rate": 1.7753335192981963e-05,
+      "loss": 1.6578,
+      "step": 780
+    },
+    {
+      "epoch": 0.22578779994217982,
+      "grad_norm": 1.138551289696566,
+      "learning_rate": 1.7747537306595336e-05,
+      "loss": 1.6774,
+      "step": 781
+    },
+    {
+      "epoch": 0.2260769008383926,
+      "grad_norm": 1.2577334291491808,
+      "learning_rate": 1.774173289798217e-05,
+      "loss": 1.6842,
+      "step": 782
+    },
+    {
+      "epoch": 0.2263660017346054,
+      "grad_norm": 1.1496680909323838,
+      "learning_rate": 1.7735921972028883e-05,
+      "loss": 1.628,
+      "step": 783
+    },
+    {
+      "epoch": 0.22665510263081814,
+      "grad_norm": 1.179850018613425,
+      "learning_rate": 1.773010453362737e-05,
+      "loss": 1.6444,
+      "step": 784
+    },
+    {
+      "epoch": 0.22694420352703093,
+      "grad_norm": 1.2654130215050725,
+      "learning_rate": 1.7724280587675017e-05,
+      "loss": 1.4592,
+      "step": 785
+    },
+    {
+      "epoch": 0.22723330442324371,
+      "grad_norm": 1.1907665269492804,
+      "learning_rate": 1.7718450139074685e-05,
+      "loss": 1.58,
+      "step": 786
+    },
+    {
+      "epoch": 0.2275224053194565,
+      "grad_norm": 1.217021995188848,
+      "learning_rate": 1.7712613192734703e-05,
+      "loss": 1.575,
+      "step": 787
+    },
+    {
+      "epoch": 0.22781150621566926,
+      "grad_norm": 1.1808590593353434,
+      "learning_rate": 1.7706769753568882e-05,
+      "loss": 1.6172,
+      "step": 788
+    },
+    {
+      "epoch": 0.22810060711188204,
+      "grad_norm": 1.196849426701046,
+      "learning_rate": 1.770091982649649e-05,
+      "loss": 1.6436,
+      "step": 789
+    },
+    {
+      "epoch": 0.22838970800809483,
+      "grad_norm": 1.1412097565678294,
+      "learning_rate": 1.7695063416442263e-05,
+      "loss": 1.6423,
+      "step": 790
+    },
+    {
+      "epoch": 0.2286788089043076,
+      "grad_norm": 1.1352763403847086,
+      "learning_rate": 1.7689200528336384e-05,
+      "loss": 1.6991,
+      "step": 791
+    },
+    {
+      "epoch": 0.22896790980052037,
+      "grad_norm": 1.1832029724179747,
+      "learning_rate": 1.7683331167114503e-05,
+      "loss": 1.6004,
+      "step": 792
+    },
+    {
+      "epoch": 0.22925701069673315,
+      "grad_norm": 1.190404880041848,
+      "learning_rate": 1.767745533771771e-05,
+      "loss": 1.6232,
+      "step": 793
+    },
+    {
+      "epoch": 0.22954611159294594,
+      "grad_norm": 1.1425994526274166,
+      "learning_rate": 1.7671573045092546e-05,
+      "loss": 1.6601,
+      "step": 794
+    },
+    {
+      "epoch": 0.22983521248915872,
+      "grad_norm": 4.306902746171457,
+      "learning_rate": 1.766568429419099e-05,
+      "loss": 1.6119,
+      "step": 795
+    },
+    {
+      "epoch": 0.23012431338537148,
+      "grad_norm": 1.2116557080614034,
+      "learning_rate": 1.7659789089970457e-05,
+      "loss": 1.553,
+      "step": 796
+    },
+    {
+      "epoch": 0.23041341428158427,
+      "grad_norm": 1.133998906444825,
+      "learning_rate": 1.7653887437393792e-05,
+      "loss": 1.5345,
+      "step": 797
+    },
+    {
+      "epoch": 0.23070251517779705,
+      "grad_norm": 1.2512101232473924,
+      "learning_rate": 1.764797934142928e-05,
+      "loss": 1.6369,
+      "step": 798
+    },
+    {
+      "epoch": 0.23099161607400984,
+      "grad_norm": 1.1088774895792184,
+      "learning_rate": 1.7642064807050617e-05,
+      "loss": 1.6432,
+      "step": 799
+    },
+    {
+      "epoch": 0.2312807169702226,
+      "grad_norm": 1.2101553368194318,
+      "learning_rate": 1.7636143839236934e-05,
+      "loss": 1.6479,
+      "step": 800
+    },
+    {
+      "epoch": 0.23156981786643538,
+      "grad_norm": 1.191660567807356,
+      "learning_rate": 1.763021644297276e-05,
+      "loss": 1.5965,
+      "step": 801
+    },
+    {
+      "epoch": 0.23185891876264816,
+      "grad_norm": 1.143724831739925,
+      "learning_rate": 1.7624282623248043e-05,
+      "loss": 1.6417,
+      "step": 802
+    },
+    {
+      "epoch": 0.23214801965886095,
+      "grad_norm": 1.143027015736506,
+      "learning_rate": 1.7618342385058147e-05,
+      "loss": 1.6015,
+      "step": 803
+    },
+    {
+      "epoch": 0.23243712055507373,
+      "grad_norm": 1.1346799180515472,
+      "learning_rate": 1.761239573340383e-05,
+      "loss": 1.6608,
+      "step": 804
+    },
+    {
+      "epoch": 0.2327262214512865,
+      "grad_norm": 1.1117892300100156,
+      "learning_rate": 1.760644267329125e-05,
+      "loss": 1.6395,
+      "step": 805
+    },
+    {
+      "epoch": 0.23301532234749928,
+      "grad_norm": 1.180786293910612,
+      "learning_rate": 1.7600483209731963e-05,
+      "loss": 1.7766,
+      "step": 806
+    },
+    {
+      "epoch": 0.23330442324371206,
+      "grad_norm": 1.1174984255388902,
+      "learning_rate": 1.7594517347742918e-05,
+      "loss": 1.5313,
+      "step": 807
+    },
+    {
+      "epoch": 0.23359352413992485,
+      "grad_norm": 1.17896186676479,
+      "learning_rate": 1.7588545092346438e-05,
+      "loss": 1.7089,
+      "step": 808
+    },
+    {
+      "epoch": 0.2338826250361376,
+      "grad_norm": 1.149670644485711,
+      "learning_rate": 1.7582566448570245e-05,
+      "loss": 1.6145,
+      "step": 809
+    },
+    {
+      "epoch": 0.2341717259323504,
+      "grad_norm": 1.2037877986724042,
+      "learning_rate": 1.7576581421447428e-05,
+      "loss": 1.663,
+      "step": 810
+    },
+    {
+      "epoch": 0.23446082682856317,
+      "grad_norm": 1.1979774494369526,
+      "learning_rate": 1.7570590016016455e-05,
+      "loss": 1.6597,
+      "step": 811
+    },
+    {
+      "epoch": 0.23474992772477596,
+      "grad_norm": 1.198242854855127,
+      "learning_rate": 1.7564592237321162e-05,
+      "loss": 1.6047,
+      "step": 812
+    },
+    {
+      "epoch": 0.23503902862098872,
+      "grad_norm": 1.2006909754615687,
+      "learning_rate": 1.7558588090410747e-05,
+      "loss": 1.7242,
+      "step": 813
+    },
+    {
+      "epoch": 0.2353281295172015,
+      "grad_norm": 1.1396286336459736,
+      "learning_rate": 1.7552577580339768e-05,
+      "loss": 1.5758,
+      "step": 814
+    },
+    {
+      "epoch": 0.23561723041341429,
+      "grad_norm": 1.1676518798162923,
+      "learning_rate": 1.7546560712168154e-05,
+      "loss": 1.6103,
+      "step": 815
+    },
+    {
+      "epoch": 0.23590633130962707,
+      "grad_norm": 1.1417358311927566,
+      "learning_rate": 1.7540537490961168e-05,
+      "loss": 1.6361,
+      "step": 816
+    },
+    {
+      "epoch": 0.23619543220583983,
+      "grad_norm": 1.1842710928884372,
+      "learning_rate": 1.753450792178943e-05,
+      "loss": 1.6522,
+      "step": 817
+    },
+    {
+      "epoch": 0.2364845331020526,
+      "grad_norm": 1.2334219525738506,
+      "learning_rate": 1.752847200972891e-05,
+      "loss": 1.6686,
+      "step": 818
+    },
+    {
+      "epoch": 0.2367736339982654,
+      "grad_norm": 1.1595381193664076,
+      "learning_rate": 1.7522429759860903e-05,
+      "loss": 1.6584,
+      "step": 819
+    },
+    {
+      "epoch": 0.23706273489447818,
+      "grad_norm": 1.1998920618458444,
+      "learning_rate": 1.751638117727205e-05,
+      "loss": 1.5457,
+      "step": 820
+    },
+    {
+      "epoch": 0.23735183579069094,
+      "grad_norm": 1.2098085468599757,
+      "learning_rate": 1.7510326267054325e-05,
+      "loss": 1.611,
+      "step": 821
+    },
+    {
+      "epoch": 0.23764093668690373,
+      "grad_norm": 1.145177051100393,
+      "learning_rate": 1.7504265034305023e-05,
+      "loss": 1.5325,
+      "step": 822
+    },
+    {
+      "epoch": 0.2379300375831165,
+      "grad_norm": 1.1017627071573495,
+      "learning_rate": 1.7498197484126756e-05,
+      "loss": 1.5977,
+      "step": 823
+    },
+    {
+      "epoch": 0.2382191384793293,
+      "grad_norm": 1.2445774010012118,
+      "learning_rate": 1.749212362162747e-05,
+      "loss": 1.7187,
+      "step": 824
+    },
+    {
+      "epoch": 0.23850823937554205,
+      "grad_norm": 1.2507033146298203,
+      "learning_rate": 1.7486043451920415e-05,
+      "loss": 1.6508,
+      "step": 825
+    },
+    {
+      "epoch": 0.23879734027175484,
+      "grad_norm": 1.1841673566917597,
+      "learning_rate": 1.747995698012415e-05,
+      "loss": 1.6084,
+      "step": 826
+    },
+    {
+      "epoch": 0.23908644116796762,
+      "grad_norm": 1.1396731986156217,
+      "learning_rate": 1.7473864211362544e-05,
+      "loss": 1.6588,
+      "step": 827
+    },
+    {
+      "epoch": 0.2393755420641804,
+      "grad_norm": 1.204843676859692,
+      "learning_rate": 1.7467765150764763e-05,
+      "loss": 1.6455,
+      "step": 828
+    },
+    {
+      "epoch": 0.23966464296039316,
+      "grad_norm": 1.2736891694731587,
+      "learning_rate": 1.7461659803465273e-05,
+      "loss": 1.638,
+      "step": 829
+    },
+    {
+      "epoch": 0.23995374385660595,
+      "grad_norm": 1.2082297185144963,
+      "learning_rate": 1.7455548174603828e-05,
+      "loss": 1.6357,
+      "step": 830
+    },
+    {
+      "epoch": 0.24024284475281873,
+      "grad_norm": 1.1833363832715826,
+      "learning_rate": 1.7449430269325474e-05,
+      "loss": 1.6351,
+      "step": 831
+    },
+    {
+      "epoch": 0.24053194564903152,
+      "grad_norm": 1.1586148600596027,
+      "learning_rate": 1.7443306092780543e-05,
+      "loss": 1.6456,
+      "step": 832
+    },
+    {
+      "epoch": 0.24082104654524428,
+      "grad_norm": 1.1151118614534092,
+      "learning_rate": 1.743717565012464e-05,
+      "loss": 1.6143,
+      "step": 833
+    },
+    {
+      "epoch": 0.24111014744145706,
+      "grad_norm": 1.1764063110634089,
+      "learning_rate": 1.7431038946518648e-05,
+      "loss": 1.6603,
+      "step": 834
+    },
+    {
+      "epoch": 0.24139924833766985,
+      "grad_norm": 1.1417362079110664,
+      "learning_rate": 1.7424895987128723e-05,
+      "loss": 1.6167,
+      "step": 835
+    },
+    {
+      "epoch": 0.24168834923388263,
+      "grad_norm": 1.1257893632889384,
+      "learning_rate": 1.7418746777126284e-05,
+      "loss": 1.62,
+      "step": 836
+    },
+    {
+      "epoch": 0.2419774501300954,
+      "grad_norm": 1.3179549442602967,
+      "learning_rate": 1.741259132168801e-05,
+      "loss": 1.5686,
+      "step": 837
+    },
+    {
+      "epoch": 0.24226655102630817,
+      "grad_norm": 1.1244904370304307,
+      "learning_rate": 1.7406429625995853e-05,
+      "loss": 1.6366,
+      "step": 838
+    },
+    {
+      "epoch": 0.24255565192252096,
+      "grad_norm": 1.2065775280910245,
+      "learning_rate": 1.7400261695236996e-05,
+      "loss": 1.7549,
+      "step": 839
+    },
+    {
+      "epoch": 0.24284475281873374,
+      "grad_norm": 1.1396498478056365,
+      "learning_rate": 1.739408753460388e-05,
+      "loss": 1.552,
+      "step": 840
+    },
+    {
+      "epoch": 0.24313385371494653,
+      "grad_norm": 1.1240397902793318,
+      "learning_rate": 1.73879071492942e-05,
+      "loss": 1.6687,
+      "step": 841
+    },
+    {
+      "epoch": 0.2434229546111593,
+      "grad_norm": 1.179163925851606,
+      "learning_rate": 1.738172054451088e-05,
+      "loss": 1.6504,
+      "step": 842
+    },
+    {
+      "epoch": 0.24371205550737207,
+      "grad_norm": 1.1030610833520664,
+      "learning_rate": 1.737552772546208e-05,
+      "loss": 1.5912,
+      "step": 843
+    },
+    {
+      "epoch": 0.24400115640358486,
+      "grad_norm": 1.1838602443986301,
+      "learning_rate": 1.73693286973612e-05,
+      "loss": 1.5732,
+      "step": 844
+    },
+    {
+      "epoch": 0.24429025729979764,
+      "grad_norm": 1.143992740058496,
+      "learning_rate": 1.7363123465426853e-05,
+      "loss": 1.6071,
+      "step": 845
+    },
+    {
+      "epoch": 0.2445793581960104,
+      "grad_norm": 1.174590060660084,
+      "learning_rate": 1.7356912034882893e-05,
+      "loss": 1.5759,
+      "step": 846
+    },
+    {
+      "epoch": 0.24486845909222318,
+      "grad_norm": 1.1329692315930469,
+      "learning_rate": 1.735069441095838e-05,
+      "loss": 1.6088,
+      "step": 847
+    },
+    {
+      "epoch": 0.24515755998843597,
+      "grad_norm": 1.1350450062350554,
+      "learning_rate": 1.7344470598887585e-05,
+      "loss": 1.7202,
+      "step": 848
+    },
+    {
+      "epoch": 0.24544666088464875,
+      "grad_norm": 1.181601786302941,
+      "learning_rate": 1.733824060391e-05,
+      "loss": 1.7271,
+      "step": 849
+    },
+    {
+      "epoch": 0.2457357617808615,
+      "grad_norm": 1.1583792292333024,
+      "learning_rate": 1.7332004431270312e-05,
+      "loss": 1.5625,
+      "step": 850
+    },
+    {
+      "epoch": 0.2460248626770743,
+      "grad_norm": 1.1568878076104787,
+      "learning_rate": 1.7325762086218415e-05,
+      "loss": 1.6503,
+      "step": 851
+    },
+    {
+      "epoch": 0.24631396357328708,
+      "grad_norm": 1.2339640973176655,
+      "learning_rate": 1.7319513574009398e-05,
+      "loss": 1.7399,
+      "step": 852
+    },
+    {
+      "epoch": 0.24660306446949987,
+      "grad_norm": 1.2009335734006834,
+      "learning_rate": 1.731325889990354e-05,
+      "loss": 1.5677,
+      "step": 853
+    },
+    {
+      "epoch": 0.24689216536571262,
+      "grad_norm": 1.1688999648294327,
+      "learning_rate": 1.73069980691663e-05,
+      "loss": 1.6136,
+      "step": 854
+    },
+    {
+      "epoch": 0.2471812662619254,
+      "grad_norm": 1.1635714362599143,
+      "learning_rate": 1.730073108706834e-05,
+      "loss": 1.6131,
+      "step": 855
+    },
+    {
+      "epoch": 0.2474703671581382,
+      "grad_norm": 1.2123420765200588,
+      "learning_rate": 1.7294457958885482e-05,
+      "loss": 1.7155,
+      "step": 856
+    },
+    {
+      "epoch": 0.24775946805435098,
+      "grad_norm": 1.1434228753959954,
+      "learning_rate": 1.7288178689898725e-05,
+      "loss": 1.7325,
+      "step": 857
+    },
+    {
+      "epoch": 0.24804856895056374,
+      "grad_norm": 1.2246804177847264,
+      "learning_rate": 1.728189328539425e-05,
+      "loss": 1.6505,
+      "step": 858
+    },
+    {
+      "epoch": 0.24833766984677652,
+      "grad_norm": 1.1585117368059001,
+      "learning_rate": 1.7275601750663387e-05,
+      "loss": 1.5769,
+      "step": 859
+    },
+    {
+      "epoch": 0.2486267707429893,
+      "grad_norm": 1.3882387244074696,
+      "learning_rate": 1.7269304091002633e-05,
+      "loss": 1.6452,
+      "step": 860
+    },
+    {
+      "epoch": 0.2489158716392021,
+      "grad_norm": 1.160203748747366,
+      "learning_rate": 1.726300031171365e-05,
+      "loss": 1.5732,
+      "step": 861
+    },
+    {
+      "epoch": 0.24920497253541485,
+      "grad_norm": 1.157523734644445,
+      "learning_rate": 1.7256690418103238e-05,
+      "loss": 1.6401,
+      "step": 862
+    },
+    {
+      "epoch": 0.24949407343162763,
+      "grad_norm": 1.1890152342635811,
+      "learning_rate": 1.725037441548335e-05,
+      "loss": 1.6668,
+      "step": 863
+    },
+    {
+      "epoch": 0.24978317432784042,
+      "grad_norm": 1.2257536935692366,
+      "learning_rate": 1.7244052309171087e-05,
+      "loss": 1.6516,
+      "step": 864
+    },
+    {
+      "epoch": 0.2500722752240532,
+      "grad_norm": 1.2023469701707792,
+      "learning_rate": 1.7237724104488675e-05,
+      "loss": 1.5962,
+      "step": 865
+    },
+    {
+      "epoch": 0.250361376120266,
+      "grad_norm": 1.2280068128337949,
+      "learning_rate": 1.723138980676349e-05,
+      "loss": 1.6962,
+      "step": 866
+    },
+    {
+      "epoch": 0.25065047701647875,
+      "grad_norm": 1.217449764349956,
+      "learning_rate": 1.7225049421328024e-05,
+      "loss": 1.6377,
+      "step": 867
+    },
+    {
+      "epoch": 0.2509395779126915,
+      "grad_norm": 1.2841045771510446,
+      "learning_rate": 1.7218702953519904e-05,
+      "loss": 1.6582,
+      "step": 868
+    },
+    {
+      "epoch": 0.2512286788089043,
+      "grad_norm": 1.2601843986629024,
+      "learning_rate": 1.7212350408681872e-05,
+      "loss": 1.6102,
+      "step": 869
+    },
+    {
+      "epoch": 0.2515177797051171,
+      "grad_norm": 1.145255827066236,
+      "learning_rate": 1.7205991792161783e-05,
+      "loss": 1.6581,
+      "step": 870
+    },
+    {
+      "epoch": 0.2518068806013299,
+      "grad_norm": 1.2362667478605363,
+      "learning_rate": 1.7199627109312615e-05,
+      "loss": 1.6043,
+      "step": 871
+    },
+    {
+      "epoch": 0.25209598149754264,
+      "grad_norm": 1.2401716127785924,
+      "learning_rate": 1.7193256365492436e-05,
+      "loss": 1.632,
+      "step": 872
+    },
+    {
+      "epoch": 0.2523850823937554,
+      "grad_norm": 1.119783771494612,
+      "learning_rate": 1.7186879566064433e-05,
+      "loss": 1.5714,
+      "step": 873
+    },
+    {
+      "epoch": 0.2526741832899682,
+      "grad_norm": 1.103085959008315,
+      "learning_rate": 1.718049671639688e-05,
+      "loss": 1.6384,
+      "step": 874
+    },
+    {
+      "epoch": 0.25296328418618097,
+      "grad_norm": 1.133467746444829,
+      "learning_rate": 1.717410782186315e-05,
+      "loss": 1.6366,
+      "step": 875
+    },
+    {
+      "epoch": 0.2532523850823938,
+      "grad_norm": 1.175009710061539,
+      "learning_rate": 1.7167712887841705e-05,
+      "loss": 1.6428,
+      "step": 876
+    },
+    {
+      "epoch": 0.25354148597860654,
+      "grad_norm": 1.1563638245548544,
+      "learning_rate": 1.716131191971608e-05,
+      "loss": 1.5624,
+      "step": 877
+    },
+    {
+      "epoch": 0.2538305868748193,
+      "grad_norm": 1.116419831991319,
+      "learning_rate": 1.715490492287491e-05,
+      "loss": 1.5829,
+      "step": 878
+    },
+    {
+      "epoch": 0.2541196877710321,
+      "grad_norm": 1.1370821345076967,
+      "learning_rate": 1.7148491902711884e-05,
+      "loss": 1.6631,
+      "step": 879
+    },
+    {
+      "epoch": 0.25440878866724487,
+      "grad_norm": 1.1611842225988047,
+      "learning_rate": 1.7142072864625777e-05,
+      "loss": 1.5913,
+      "step": 880
+    },
+    {
+      "epoch": 0.2546978895634576,
+      "grad_norm": 1.1019030805813086,
+      "learning_rate": 1.7135647814020425e-05,
+      "loss": 1.6758,
+      "step": 881
+    },
+    {
+      "epoch": 0.25498699045967044,
+      "grad_norm": 1.04613261420445,
+      "learning_rate": 1.7129216756304723e-05,
+      "loss": 1.5948,
+      "step": 882
+    },
+    {
+      "epoch": 0.2552760913558832,
+      "grad_norm": 1.1936376971157356,
+      "learning_rate": 1.7122779696892628e-05,
+      "loss": 1.6552,
+      "step": 883
+    },
+    {
+      "epoch": 0.255565192252096,
+      "grad_norm": 1.1593207052719288,
+      "learning_rate": 1.7116336641203147e-05,
+      "loss": 1.6959,
+      "step": 884
+    },
+    {
+      "epoch": 0.25585429314830876,
+      "grad_norm": 1.1061563744857061,
+      "learning_rate": 1.7109887594660336e-05,
+      "loss": 1.6163,
+      "step": 885
+    },
+    {
+      "epoch": 0.2561433940445215,
+      "grad_norm": 1.134065834530409,
+      "learning_rate": 1.710343256269329e-05,
+      "loss": 1.6726,
+      "step": 886
+    },
+    {
+      "epoch": 0.25643249494073433,
+      "grad_norm": 1.1359039495462206,
+      "learning_rate": 1.7096971550736153e-05,
+      "loss": 1.5318,
+      "step": 887
+    },
+    {
+      "epoch": 0.2567215958369471,
+      "grad_norm": 1.3387153646158823,
+      "learning_rate": 1.7090504564228087e-05,
+      "loss": 1.6655,
+      "step": 888
+    },
+    {
+      "epoch": 0.25701069673315985,
+      "grad_norm": 1.093747361454369,
+      "learning_rate": 1.7084031608613297e-05,
+      "loss": 1.7196,
+      "step": 889
+    },
+    {
+      "epoch": 0.25729979762937266,
+      "grad_norm": 1.114071184877325,
+      "learning_rate": 1.7077552689341012e-05,
+      "loss": 1.5488,
+      "step": 890
+    },
+    {
+      "epoch": 0.2575888985255854,
+      "grad_norm": 1.1616291244384729,
+      "learning_rate": 1.7071067811865477e-05,
+      "loss": 1.5431,
+      "step": 891
+    },
+    {
+      "epoch": 0.25787799942179823,
+      "grad_norm": 1.1354100602957258,
+      "learning_rate": 1.7064576981645954e-05,
+      "loss": 1.6055,
+      "step": 892
+    },
+    {
+      "epoch": 0.258167100318011,
+      "grad_norm": 1.1522440785073058,
+      "learning_rate": 1.7058080204146718e-05,
+      "loss": 1.629,
+      "step": 893
+    },
+    {
+      "epoch": 0.25845620121422375,
+      "grad_norm": 1.1026615179112005,
+      "learning_rate": 1.7051577484837043e-05,
+      "loss": 1.5596,
+      "step": 894
+    },
+    {
+      "epoch": 0.25874530211043656,
+      "grad_norm": 1.1687730289765201,
+      "learning_rate": 1.7045068829191218e-05,
+      "loss": 1.6721,
+      "step": 895
+    },
+    {
+      "epoch": 0.2590344030066493,
+      "grad_norm": 1.1074104029137306,
+      "learning_rate": 1.7038554242688526e-05,
+      "loss": 1.6789,
+      "step": 896
+    },
+    {
+      "epoch": 0.2593235039028621,
+      "grad_norm": 1.1026942328183036,
+      "learning_rate": 1.7032033730813236e-05,
+      "loss": 1.558,
+      "step": 897
+    },
+    {
+      "epoch": 0.2596126047990749,
+      "grad_norm": 1.1646081156981725,
+      "learning_rate": 1.7025507299054605e-05,
+      "loss": 1.601,
+      "step": 898
+    },
+    {
+      "epoch": 0.25990170569528764,
+      "grad_norm": 1.1271463774298178,
+      "learning_rate": 1.7018974952906885e-05,
+      "loss": 1.6538,
+      "step": 899
+    },
+    {
+      "epoch": 0.26019080659150046,
+      "grad_norm": 1.1091390731718525,
+      "learning_rate": 1.7012436697869295e-05,
+      "loss": 1.5819,
+      "step": 900
+    },
+    {
+      "epoch": 0.2604799074877132,
+      "grad_norm": 1.173447822417614,
+      "learning_rate": 1.7005892539446035e-05,
+      "loss": 1.6451,
+      "step": 901
+    },
+    {
+      "epoch": 0.26076900838392597,
+      "grad_norm": 1.1350959293058749,
+      "learning_rate": 1.699934248314627e-05,
+      "loss": 1.6962,
+      "step": 902
+    },
+    {
+      "epoch": 0.2610581092801388,
+      "grad_norm": 1.1162197648231467,
+      "learning_rate": 1.699278653448414e-05,
+      "loss": 1.6747,
+      "step": 903
+    },
+    {
+      "epoch": 0.26134721017635154,
+      "grad_norm": 1.2139920201386476,
+      "learning_rate": 1.698622469897873e-05,
+      "loss": 1.7454,
+      "step": 904
+    },
+    {
+      "epoch": 0.2616363110725643,
+      "grad_norm": 1.0734616221718876,
+      "learning_rate": 1.6979656982154095e-05,
+      "loss": 1.6061,
+      "step": 905
+    },
+    {
+      "epoch": 0.2619254119687771,
+      "grad_norm": 1.1328541465259365,
+      "learning_rate": 1.6973083389539233e-05,
+      "loss": 1.585,
+      "step": 906
+    },
+    {
+      "epoch": 0.26221451286498987,
+      "grad_norm": 1.1568303460362723,
+      "learning_rate": 1.6966503926668094e-05,
+      "loss": 1.6825,
+      "step": 907
+    },
+    {
+      "epoch": 0.2625036137612027,
+      "grad_norm": 1.196990001143863,
+      "learning_rate": 1.6959918599079562e-05,
+      "loss": 1.6538,
+      "step": 908
+    },
+    {
+      "epoch": 0.26279271465741544,
+      "grad_norm": 1.1342440267026168,
+      "learning_rate": 1.6953327412317468e-05,
+      "loss": 1.6075,
+      "step": 909
+    },
+    {
+      "epoch": 0.2630818155536282,
+      "grad_norm": 1.1596789456175167,
+      "learning_rate": 1.6946730371930565e-05,
+      "loss": 1.6929,
+      "step": 910
+    },
+    {
+      "epoch": 0.263370916449841,
+      "grad_norm": 1.1373065436703054,
+      "learning_rate": 1.6940127483472543e-05,
+      "loss": 1.6481,
+      "step": 911
+    },
+    {
+      "epoch": 0.26366001734605377,
+      "grad_norm": 1.1610606493879128,
+      "learning_rate": 1.6933518752502013e-05,
+      "loss": 1.7379,
+      "step": 912
+    },
+    {
+      "epoch": 0.2639491182422666,
+      "grad_norm": 1.125687311211295,
+      "learning_rate": 1.69269041845825e-05,
+      "loss": 1.5849,
+      "step": 913
+    },
+    {
+      "epoch": 0.26423821913847934,
+      "grad_norm": 1.1489558599701768,
+      "learning_rate": 1.692028378528244e-05,
+      "loss": 1.7226,
+      "step": 914
+    },
+    {
+      "epoch": 0.2645273200346921,
+      "grad_norm": 1.188601400208746,
+      "learning_rate": 1.6913657560175197e-05,
+      "loss": 1.6326,
+      "step": 915
+    },
+    {
+      "epoch": 0.2648164209309049,
+      "grad_norm": 1.113219452373414,
+      "learning_rate": 1.6907025514839013e-05,
+      "loss": 1.6478,
+      "step": 916
+    },
+    {
+      "epoch": 0.26510552182711766,
+      "grad_norm": 1.1593190572859267,
+      "learning_rate": 1.6900387654857055e-05,
+      "loss": 1.6093,
+      "step": 917
+    },
+    {
+      "epoch": 0.2653946227233304,
+      "grad_norm": 1.2134229994183705,
+      "learning_rate": 1.6893743985817366e-05,
+      "loss": 1.6903,
+      "step": 918
+    },
+    {
+      "epoch": 0.26568372361954323,
+      "grad_norm": 1.1993364856075077,
+      "learning_rate": 1.688709451331289e-05,
+      "loss": 1.6092,
+      "step": 919
+    },
+    {
+      "epoch": 0.265972824515756,
+      "grad_norm": 1.0919639751411407,
+      "learning_rate": 1.688043924294145e-05,
+      "loss": 1.6078,
+      "step": 920
+    },
+    {
+      "epoch": 0.2662619254119688,
+      "grad_norm": 1.2013320201783475,
+      "learning_rate": 1.687377818030575e-05,
+      "loss": 1.6899,
+      "step": 921
+    },
+    {
+      "epoch": 0.26655102630818156,
+      "grad_norm": 1.1806015487593893,
+      "learning_rate": 1.6867111331013384e-05,
+      "loss": 1.6197,
+      "step": 922
+    },
+    {
+      "epoch": 0.2668401272043943,
+      "grad_norm": 1.1020975960157413,
+      "learning_rate": 1.6860438700676797e-05,
+      "loss": 1.5893,
+      "step": 923
+    },
+    {
+      "epoch": 0.26712922810060713,
+      "grad_norm": 1.094951519307646,
+      "learning_rate": 1.685376029491331e-05,
+      "loss": 1.5254,
+      "step": 924
+    },
+    {
+      "epoch": 0.2674183289968199,
+      "grad_norm": 1.1163275611243435,
+      "learning_rate": 1.684707611934511e-05,
+      "loss": 1.5477,
+      "step": 925
+    },
+    {
+      "epoch": 0.26770742989303264,
+      "grad_norm": 1.2307806882838923,
+      "learning_rate": 1.6840386179599242e-05,
+      "loss": 1.6667,
+      "step": 926
+    },
+    {
+      "epoch": 0.26799653078924546,
+      "grad_norm": 1.171373289157806,
+      "learning_rate": 1.683369048130759e-05,
+      "loss": 1.6172,
+      "step": 927
+    },
+    {
+      "epoch": 0.2682856316854582,
+      "grad_norm": 1.1886579501994778,
+      "learning_rate": 1.6826989030106902e-05,
+      "loss": 1.6133,
+      "step": 928
+    },
+    {
+      "epoch": 0.268574732581671,
+      "grad_norm": 1.1934353770717314,
+      "learning_rate": 1.6820281831638757e-05,
+      "loss": 1.559,
+      "step": 929
+    },
+    {
+      "epoch": 0.2688638334778838,
+      "grad_norm": 1.1614491161854386,
+      "learning_rate": 1.6813568891549584e-05,
+      "loss": 1.5769,
+      "step": 930
+    },
+    {
+      "epoch": 0.26915293437409654,
+      "grad_norm": 1.1397915171658068,
+      "learning_rate": 1.680685021549063e-05,
+      "loss": 1.5537,
+      "step": 931
+    },
+    {
+      "epoch": 0.26944203527030935,
+      "grad_norm": 1.149992178338205,
+      "learning_rate": 1.6800125809117984e-05,
+      "loss": 1.6231,
+      "step": 932
+    },
+    {
+      "epoch": 0.2697311361665221,
+      "grad_norm": 1.1401360403507486,
+      "learning_rate": 1.6793395678092555e-05,
+      "loss": 1.6182,
+      "step": 933
+    },
+    {
+      "epoch": 0.27002023706273487,
+      "grad_norm": 1.1209347016808902,
+      "learning_rate": 1.6786659828080078e-05,
+      "loss": 1.6145,
+      "step": 934
+    },
+    {
+      "epoch": 0.2703093379589477,
+      "grad_norm": 1.159654576616978,
+      "learning_rate": 1.6779918264751082e-05,
+      "loss": 1.707,
+      "step": 935
+    },
+    {
+      "epoch": 0.27059843885516044,
+      "grad_norm": 1.1184081607191674,
+      "learning_rate": 1.6773170993780926e-05,
+      "loss": 1.6816,
+      "step": 936
+    },
+    {
+      "epoch": 0.27088753975137325,
+      "grad_norm": 1.0985004602757438,
+      "learning_rate": 1.676641802084977e-05,
+      "loss": 1.6155,
+      "step": 937
+    },
+    {
+      "epoch": 0.271176640647586,
+      "grad_norm": 1.1086581131090065,
+      "learning_rate": 1.6759659351642563e-05,
+      "loss": 1.5973,
+      "step": 938
+    },
+    {
+      "epoch": 0.27146574154379877,
+      "grad_norm": 1.1262795371219783,
+      "learning_rate": 1.6752894991849066e-05,
+      "loss": 1.6819,
+      "step": 939
+    },
+    {
+      "epoch": 0.2717548424400116,
+      "grad_norm": 1.1042181463660192,
+      "learning_rate": 1.674612494716382e-05,
+      "loss": 1.6242,
+      "step": 940
+    },
+    {
+      "epoch": 0.27204394333622434,
+      "grad_norm": 1.1225263911796381,
+      "learning_rate": 1.6739349223286153e-05,
+      "loss": 1.688,
+      "step": 941
+    },
+    {
+      "epoch": 0.2723330442324371,
+      "grad_norm": 1.0507156428201587,
+      "learning_rate": 1.673256782592018e-05,
+      "loss": 1.6293,
+      "step": 942
+    },
+    {
+      "epoch": 0.2726221451286499,
+      "grad_norm": 1.1687937216638025,
+      "learning_rate": 1.6725780760774775e-05,
+      "loss": 1.5753,
+      "step": 943
+    },
+    {
+      "epoch": 0.27291124602486266,
+      "grad_norm": 1.0675977484748183,
+      "learning_rate": 1.671898803356361e-05,
+      "loss": 1.5865,
+      "step": 944
+    },
+    {
+      "epoch": 0.2732003469210755,
+      "grad_norm": 1.1223391674880248,
+      "learning_rate": 1.67121896500051e-05,
+      "loss": 1.6529,
+      "step": 945
+    },
+    {
+      "epoch": 0.27348944781728823,
+      "grad_norm": 1.1326572315570196,
+      "learning_rate": 1.6705385615822436e-05,
+      "loss": 1.6463,
+      "step": 946
+    },
+    {
+      "epoch": 0.273778548713501,
+      "grad_norm": 1.2086326669717427,
+      "learning_rate": 1.669857593674356e-05,
+      "loss": 1.6377,
+      "step": 947
+    },
+    {
+      "epoch": 0.2740676496097138,
+      "grad_norm": 1.193825641421145,
+      "learning_rate": 1.6691760618501163e-05,
+      "loss": 1.5547,
+      "step": 948
+    },
+    {
+      "epoch": 0.27435675050592656,
+      "grad_norm": 1.2866576631032545,
+      "learning_rate": 1.6684939666832695e-05,
+      "loss": 1.6659,
+      "step": 949
+    },
+    {
+      "epoch": 0.2746458514021394,
+      "grad_norm": 1.2166777753299653,
+      "learning_rate": 1.6678113087480334e-05,
+      "loss": 1.5869,
+      "step": 950
+    },
+    {
+      "epoch": 0.27493495229835213,
+      "grad_norm": 1.192007342337803,
+      "learning_rate": 1.6671280886191005e-05,
+      "loss": 1.6139,
+      "step": 951
+    },
+    {
+      "epoch": 0.2752240531945649,
+      "grad_norm": 1.1110607953589973,
+      "learning_rate": 1.6664443068716365e-05,
+      "loss": 1.5512,
+      "step": 952
+    },
+    {
+      "epoch": 0.2755131540907777,
+      "grad_norm": 1.1519768318705965,
+      "learning_rate": 1.6657599640812793e-05,
+      "loss": 1.5454,
+      "step": 953
+    },
+    {
+      "epoch": 0.27580225498699046,
+      "grad_norm": 1.1479317865206893,
+      "learning_rate": 1.6650750608241403e-05,
+      "loss": 1.6038,
+      "step": 954
+    },
+    {
+      "epoch": 0.2760913558832032,
+      "grad_norm": 1.0950679969123849,
+      "learning_rate": 1.664389597676801e-05,
+      "loss": 1.5197,
+      "step": 955
+    },
+    {
+      "epoch": 0.27638045677941603,
+      "grad_norm": 1.1421306294859326,
+      "learning_rate": 1.6637035752163166e-05,
+      "loss": 1.7071,
+      "step": 956
+    },
+    {
+      "epoch": 0.2766695576756288,
+      "grad_norm": 1.0604403244112353,
+      "learning_rate": 1.66301699402021e-05,
+      "loss": 1.6066,
+      "step": 957
+    },
+    {
+      "epoch": 0.2769586585718416,
+      "grad_norm": 1.1297323126616883,
+      "learning_rate": 1.6623298546664773e-05,
+      "loss": 1.7718,
+      "step": 958
+    },
+    {
+      "epoch": 0.27724775946805436,
+      "grad_norm": 1.0910744186175767,
+      "learning_rate": 1.6616421577335834e-05,
+      "loss": 1.6962,
+      "step": 959
+    },
+    {
+      "epoch": 0.2775368603642671,
+      "grad_norm": 1.12747611212375,
+      "learning_rate": 1.6609539038004624e-05,
+      "loss": 1.7292,
+      "step": 960
+    },
+    {
+      "epoch": 0.2778259612604799,
+      "grad_norm": 1.093202597412715,
+      "learning_rate": 1.6602650934465178e-05,
+      "loss": 1.6183,
+      "step": 961
+    },
+    {
+      "epoch": 0.2781150621566927,
+      "grad_norm": 1.072162280018483,
+      "learning_rate": 1.659575727251621e-05,
+      "loss": 1.5427,
+      "step": 962
+    },
+    {
+      "epoch": 0.27840416305290544,
+      "grad_norm": 1.1781301893929088,
+      "learning_rate": 1.658885805796111e-05,
+      "loss": 1.6114,
+      "step": 963
+    },
+    {
+      "epoch": 0.27869326394911825,
+      "grad_norm": 1.153766159493101,
+      "learning_rate": 1.6581953296607963e-05,
+      "loss": 1.6674,
+      "step": 964
+    },
+    {
+      "epoch": 0.278982364845331,
+      "grad_norm": 1.1586767056723473,
+      "learning_rate": 1.657504299426949e-05,
+      "loss": 1.5837,
+      "step": 965
+    },
+    {
+      "epoch": 0.2792714657415438,
+      "grad_norm": 1.0755813818211648,
+      "learning_rate": 1.656812715676311e-05,
+      "loss": 1.6153,
+      "step": 966
+    },
+    {
+      "epoch": 0.2795605666377566,
+      "grad_norm": 1.2638124252742915,
+      "learning_rate": 1.6561205789910885e-05,
+      "loss": 1.5646,
+      "step": 967
+    },
+    {
+      "epoch": 0.27984966753396934,
+      "grad_norm": 1.0904862968899787,
+      "learning_rate": 1.655427889953953e-05,
+      "loss": 1.6286,
+      "step": 968
+    },
+    {
+      "epoch": 0.28013876843018215,
+      "grad_norm": 1.1573153681833863,
+      "learning_rate": 1.6547346491480414e-05,
+      "loss": 1.6379,
+      "step": 969
+    },
+    {
+      "epoch": 0.2804278693263949,
+      "grad_norm": 1.1816273912861235,
+      "learning_rate": 1.654040857156955e-05,
+      "loss": 1.5763,
+      "step": 970
+    },
+    {
+      "epoch": 0.28071697022260766,
+      "grad_norm": 1.1643158633524044,
+      "learning_rate": 1.6533465145647598e-05,
+      "loss": 1.6681,
+      "step": 971
+    },
+    {
+      "epoch": 0.2810060711188205,
+      "grad_norm": 1.156818680500087,
+      "learning_rate": 1.6526516219559843e-05,
+      "loss": 1.5255,
+      "step": 972
+    },
+    {
+      "epoch": 0.28129517201503323,
+      "grad_norm": 1.1426079298705811,
+      "learning_rate": 1.6519561799156202e-05,
+      "loss": 1.5607,
+      "step": 973
+    },
+    {
+      "epoch": 0.28158427291124605,
+      "grad_norm": 1.0995905315062742,
+      "learning_rate": 1.6512601890291226e-05,
+      "loss": 1.5456,
+      "step": 974
+    },
+    {
+      "epoch": 0.2818733738074588,
+      "grad_norm": 1.1899775942969868,
+      "learning_rate": 1.6505636498824074e-05,
+      "loss": 1.6684,
+      "step": 975
+    },
+    {
+      "epoch": 0.28216247470367156,
+      "grad_norm": 1.155528745251187,
+      "learning_rate": 1.6498665630618532e-05,
+      "loss": 1.5868,
+      "step": 976
+    },
+    {
+      "epoch": 0.2824515755998844,
+      "grad_norm": 1.1411174440976906,
+      "learning_rate": 1.649168929154299e-05,
+      "loss": 1.4712,
+      "step": 977
+    },
+    {
+      "epoch": 0.28274067649609713,
+      "grad_norm": 1.0860228511069425,
+      "learning_rate": 1.6484707487470447e-05,
+      "loss": 1.662,
+      "step": 978
+    },
+    {
+      "epoch": 0.2830297773923099,
+      "grad_norm": 1.1466851119233112,
+      "learning_rate": 1.6477720224278493e-05,
+      "loss": 1.6142,
+      "step": 979
+    },
+    {
+      "epoch": 0.2833188782885227,
+      "grad_norm": 1.3078811221895994,
+      "learning_rate": 1.6470727507849332e-05,
+      "loss": 1.6702,
+      "step": 980
+    },
+    {
+      "epoch": 0.28360797918473546,
+      "grad_norm": 1.1996795619215384,
+      "learning_rate": 1.6463729344069736e-05,
+      "loss": 1.6115,
+      "step": 981
+    },
+    {
+      "epoch": 0.28389708008094827,
+      "grad_norm": 1.2042463666525114,
+      "learning_rate": 1.645672573883109e-05,
+      "loss": 1.6333,
+      "step": 982
+    },
+    {
+      "epoch": 0.28418618097716103,
+      "grad_norm": 1.1063042405661339,
+      "learning_rate": 1.6449716698029335e-05,
+      "loss": 1.6195,
+      "step": 983
+    },
+    {
+      "epoch": 0.2844752818733738,
+      "grad_norm": 1.1072288580101002,
+      "learning_rate": 1.6442702227564996e-05,
+      "loss": 1.6116,
+      "step": 984
+    },
+    {
+      "epoch": 0.2847643827695866,
+      "grad_norm": 1.2243676300508977,
+      "learning_rate": 1.643568233334318e-05,
+      "loss": 1.5821,
+      "step": 985
+    },
+    {
+      "epoch": 0.28505348366579936,
+      "grad_norm": 1.1664544235823486,
+      "learning_rate": 1.642865702127354e-05,
+      "loss": 1.671,
+      "step": 986
+    },
+    {
+      "epoch": 0.28534258456201217,
+      "grad_norm": 1.1167634055353126,
+      "learning_rate": 1.6421626297270316e-05,
+      "loss": 1.6361,
+      "step": 987
+    },
+    {
+      "epoch": 0.2856316854582249,
+      "grad_norm": 1.1612972833032917,
+      "learning_rate": 1.6414590167252273e-05,
+      "loss": 1.5859,
+      "step": 988
+    },
+    {
+      "epoch": 0.2859207863544377,
+      "grad_norm": 1.213672115046863,
+      "learning_rate": 1.6407548637142746e-05,
+      "loss": 1.5757,
+      "step": 989
+    },
+    {
+      "epoch": 0.2862098872506505,
+      "grad_norm": 1.1300532059127562,
+      "learning_rate": 1.640050171286962e-05,
+      "loss": 1.6489,
+      "step": 990
+    },
+    {
+      "epoch": 0.28649898814686325,
+      "grad_norm": 1.1475160145568735,
+      "learning_rate": 1.6393449400365307e-05,
+      "loss": 1.5792,
+      "step": 991
+    },
+    {
+      "epoch": 0.286788089043076,
+      "grad_norm": 1.171976265527746,
+      "learning_rate": 1.6386391705566768e-05,
+      "loss": 1.5027,
+      "step": 992
+    },
+    {
+      "epoch": 0.2870771899392888,
+      "grad_norm": 1.1698090854087668,
+      "learning_rate": 1.6379328634415484e-05,
+      "loss": 1.681,
+      "step": 993
+    },
+    {
+      "epoch": 0.2873662908355016,
+      "grad_norm": 1.1317960093108423,
+      "learning_rate": 1.6372260192857465e-05,
+      "loss": 1.603,
+      "step": 994
+    },
+    {
+      "epoch": 0.2876553917317144,
+      "grad_norm": 1.179188272729989,
+      "learning_rate": 1.636518638684325e-05,
+      "loss": 1.6851,
+      "step": 995
+    },
+    {
+      "epoch": 0.28794449262792715,
+      "grad_norm": 1.1079735999761406,
+      "learning_rate": 1.6358107222327884e-05,
+      "loss": 1.6323,
+      "step": 996
+    },
+    {
+      "epoch": 0.2882335935241399,
+      "grad_norm": 1.1148610947340667,
+      "learning_rate": 1.635102270527093e-05,
+      "loss": 1.5511,
+      "step": 997
+    },
+    {
+      "epoch": 0.2885226944203527,
+      "grad_norm": 1.184333908207028,
+      "learning_rate": 1.6343932841636455e-05,
+      "loss": 1.6631,
+      "step": 998
+    },
+    {
+      "epoch": 0.2888117953165655,
+      "grad_norm": 1.1154370991272007,
+      "learning_rate": 1.6336837637393025e-05,
+      "loss": 1.6557,
+      "step": 999
+    },
+    {
+      "epoch": 0.28910089621277824,
+      "grad_norm": 1.1055845000132498,
+      "learning_rate": 1.63297370985137e-05,
+      "loss": 1.623,
+      "step": 1000
+    },
+    {
+      "epoch": 0.28938999710899105,
+      "grad_norm": 1.1478774668713727,
+      "learning_rate": 1.6322631230976035e-05,
+      "loss": 1.6303,
+      "step": 1001
+    },
+    {
+      "epoch": 0.2896790980052038,
+      "grad_norm": 1.1260832092736042,
+      "learning_rate": 1.6315520040762073e-05,
+      "loss": 1.6627,
+      "step": 1002
+    },
+    {
+      "epoch": 0.2899681989014166,
+      "grad_norm": 1.1328217900445974,
+      "learning_rate": 1.6308403533858334e-05,
+      "loss": 1.6185,
+      "step": 1003
+    },
+    {
+      "epoch": 0.2902572997976294,
+      "grad_norm": 1.1217019928925396,
+      "learning_rate": 1.6301281716255814e-05,
+      "loss": 1.5887,
+      "step": 1004
+    },
+    {
+      "epoch": 0.29054640069384213,
+      "grad_norm": 1.0662817504674387,
+      "learning_rate": 1.629415459394998e-05,
+      "loss": 1.6385,
+      "step": 1005
+    },
+    {
+      "epoch": 0.29083550159005495,
+      "grad_norm": 1.0503730461501608,
+      "learning_rate": 1.6287022172940762e-05,
+      "loss": 1.6164,
+      "step": 1006
+    },
+    {
+      "epoch": 0.2911246024862677,
+      "grad_norm": 1.0733421544977206,
+      "learning_rate": 1.627988445923256e-05,
+      "loss": 1.6689,
+      "step": 1007
+    },
+    {
+      "epoch": 0.29141370338248046,
+      "grad_norm": 1.0667066239811298,
+      "learning_rate": 1.6272741458834224e-05,
+      "loss": 1.6697,
+      "step": 1008
+    },
+    {
+      "epoch": 0.2917028042786933,
+      "grad_norm": 1.087739247251025,
+      "learning_rate": 1.6265593177759046e-05,
+      "loss": 1.6863,
+      "step": 1009
+    },
+    {
+      "epoch": 0.29199190517490603,
+      "grad_norm": 1.0713166761734145,
+      "learning_rate": 1.625843962202478e-05,
+      "loss": 1.5898,
+      "step": 1010
+    },
+    {
+      "epoch": 0.29228100607111884,
+      "grad_norm": 1.1902495194222464,
+      "learning_rate": 1.6251280797653606e-05,
+      "loss": 1.6618,
+      "step": 1011
+    },
+    {
+      "epoch": 0.2925701069673316,
+      "grad_norm": 1.125478902548264,
+      "learning_rate": 1.624411671067215e-05,
+      "loss": 1.5907,
+      "step": 1012
+    },
+    {
+      "epoch": 0.29285920786354436,
+      "grad_norm": 1.1078237145953016,
+      "learning_rate": 1.623694736711146e-05,
+      "loss": 1.7451,
+      "step": 1013
+    },
+    {
+      "epoch": 0.29314830875975717,
+      "grad_norm": 1.1051877093548326,
+      "learning_rate": 1.6229772773007016e-05,
+      "loss": 1.6274,
+      "step": 1014
+    },
+    {
+      "epoch": 0.2934374096559699,
+      "grad_norm": 1.121714398039267,
+      "learning_rate": 1.6222592934398715e-05,
+      "loss": 1.4989,
+      "step": 1015
+    },
+    {
+      "epoch": 0.2937265105521827,
+      "grad_norm": 1.0925330870668526,
+      "learning_rate": 1.6215407857330862e-05,
+      "loss": 1.5753,
+      "step": 1016
+    },
+    {
+      "epoch": 0.2940156114483955,
+      "grad_norm": 1.086137363905831,
+      "learning_rate": 1.620821754785219e-05,
+      "loss": 1.5898,
+      "step": 1017
+    },
+    {
+      "epoch": 0.29430471234460825,
+      "grad_norm": 1.1237702975359138,
+      "learning_rate": 1.6201022012015823e-05,
+      "loss": 1.7232,
+      "step": 1018
+    },
+    {
+      "epoch": 0.29459381324082107,
+      "grad_norm": 1.1254688780372208,
+      "learning_rate": 1.619382125587928e-05,
+      "loss": 1.6816,
+      "step": 1019
+    },
+    {
+      "epoch": 0.2948829141370338,
+      "grad_norm": 1.123440253692758,
+      "learning_rate": 1.6186615285504496e-05,
+      "loss": 1.6185,
+      "step": 1020
+    },
+    {
+      "epoch": 0.2951720150332466,
+      "grad_norm": 1.0947286679961508,
+      "learning_rate": 1.617940410695777e-05,
+      "loss": 1.6673,
+      "step": 1021
+    },
+    {
+      "epoch": 0.2954611159294594,
+      "grad_norm": 1.1147257497620007,
+      "learning_rate": 1.6172187726309802e-05,
+      "loss": 1.5691,
+      "step": 1022
+    },
+    {
+      "epoch": 0.29575021682567215,
+      "grad_norm": 1.2128326700223733,
+      "learning_rate": 1.6164966149635664e-05,
+      "loss": 1.5773,
+      "step": 1023
+    },
+    {
+      "epoch": 0.29603931772188496,
+      "grad_norm": 1.1251444723700124,
+      "learning_rate": 1.615773938301481e-05,
+      "loss": 1.6696,
+      "step": 1024
+    },
+    {
+      "epoch": 0.2963284186180977,
+      "grad_norm": 1.1144709330002471,
+      "learning_rate": 1.615050743253106e-05,
+      "loss": 1.5903,
+      "step": 1025
+    },
+    {
+      "epoch": 0.2966175195143105,
+      "grad_norm": 1.114572930209085,
+      "learning_rate": 1.6143270304272588e-05,
+      "loss": 1.5259,
+      "step": 1026
+    },
+    {
+      "epoch": 0.2969066204105233,
+      "grad_norm": 1.1041968816612577,
+      "learning_rate": 1.613602800433194e-05,
+      "loss": 1.594,
+      "step": 1027
+    },
+    {
+      "epoch": 0.29719572130673605,
+      "grad_norm": 1.1309087615970796,
+      "learning_rate": 1.6128780538806012e-05,
+      "loss": 1.6098,
+      "step": 1028
+    },
+    {
+      "epoch": 0.2974848222029488,
+      "grad_norm": 1.0655925882162895,
+      "learning_rate": 1.6121527913796045e-05,
+      "loss": 1.5776,
+      "step": 1029
+    },
+    {
+      "epoch": 0.2977739230991616,
+      "grad_norm": 1.153738047328653,
+      "learning_rate": 1.611427013540763e-05,
+      "loss": 1.6305,
+      "step": 1030
+    },
+    {
+      "epoch": 0.2980630239953744,
+      "grad_norm": 1.1277739002652585,
+      "learning_rate": 1.6107007209750687e-05,
+      "loss": 1.5917,
+      "step": 1031
+    },
+    {
+      "epoch": 0.2983521248915872,
+      "grad_norm": 1.1292106867822904,
+      "learning_rate": 1.6099739142939477e-05,
+      "loss": 1.644,
+      "step": 1032
+    },
+    {
+      "epoch": 0.29864122578779995,
+      "grad_norm": 1.1236415518463128,
+      "learning_rate": 1.6092465941092585e-05,
+      "loss": 1.6763,
+      "step": 1033
+    },
+    {
+      "epoch": 0.2989303266840127,
+      "grad_norm": 1.119641797142331,
+      "learning_rate": 1.6085187610332925e-05,
+      "loss": 1.6655,
+      "step": 1034
+    },
+    {
+      "epoch": 0.2992194275802255,
+      "grad_norm": 1.1300850404882916,
+      "learning_rate": 1.607790415678772e-05,
+      "loss": 1.6187,
+      "step": 1035
+    },
+    {
+      "epoch": 0.2995085284764383,
+      "grad_norm": 1.1047280483132156,
+      "learning_rate": 1.6070615586588516e-05,
+      "loss": 1.6469,
+      "step": 1036
+    },
+    {
+      "epoch": 0.29979762937265103,
+      "grad_norm": 1.152691255287852,
+      "learning_rate": 1.6063321905871152e-05,
+      "loss": 1.665,
+      "step": 1037
+    },
+    {
+      "epoch": 0.30008673026886384,
+      "grad_norm": 1.169423494531767,
+      "learning_rate": 1.605602312077578e-05,
+      "loss": 1.6282,
+      "step": 1038
+    },
+    {
+      "epoch": 0.3003758311650766,
+      "grad_norm": 1.2677425058482785,
+      "learning_rate": 1.6048719237446852e-05,
+      "loss": 1.6314,
+      "step": 1039
+    },
+    {
+      "epoch": 0.3006649320612894,
+      "grad_norm": 1.1668503548868798,
+      "learning_rate": 1.60414102620331e-05,
+      "loss": 1.6291,
+      "step": 1040
+    },
+    {
+      "epoch": 0.30095403295750217,
+      "grad_norm": 1.1367727503019929,
+      "learning_rate": 1.6034096200687552e-05,
+      "loss": 1.4852,
+      "step": 1041
+    },
+    {
+      "epoch": 0.30124313385371493,
+      "grad_norm": 1.1323984808855299,
+      "learning_rate": 1.6026777059567515e-05,
+      "loss": 1.5893,
+      "step": 1042
+    },
+    {
+      "epoch": 0.30153223474992774,
+      "grad_norm": 1.1609097650981914,
+      "learning_rate": 1.601945284483457e-05,
+      "loss": 1.5886,
+      "step": 1043
+    },
+    {
+      "epoch": 0.3018213356461405,
+      "grad_norm": 1.1415713259757136,
+      "learning_rate": 1.601212356265458e-05,
+      "loss": 1.7002,
+      "step": 1044
+    },
+    {
+      "epoch": 0.30211043654235326,
+      "grad_norm": 1.227231895248573,
+      "learning_rate": 1.6004789219197656e-05,
+      "loss": 1.5416,
+      "step": 1045
+    },
+    {
+      "epoch": 0.30239953743856607,
+      "grad_norm": 1.163043113422494,
+      "learning_rate": 1.599744982063818e-05,
+      "loss": 1.6469,
+      "step": 1046
+    },
+    {
+      "epoch": 0.3026886383347788,
+      "grad_norm": 1.1126723523837783,
+      "learning_rate": 1.5990105373154793e-05,
+      "loss": 1.6411,
+      "step": 1047
+    },
+    {
+      "epoch": 0.30297773923099164,
+      "grad_norm": 1.122459840411915,
+      "learning_rate": 1.5982755882930382e-05,
+      "loss": 1.6519,
+      "step": 1048
+    },
+    {
+      "epoch": 0.3032668401272044,
+      "grad_norm": 1.5840082766344463,
+      "learning_rate": 1.5975401356152074e-05,
+      "loss": 1.6135,
+      "step": 1049
+    },
+    {
+      "epoch": 0.30355594102341715,
+      "grad_norm": 1.1225058692910097,
+      "learning_rate": 1.596804179901125e-05,
+      "loss": 1.5674,
+      "step": 1050
+    },
+    {
+      "epoch": 0.30384504191962997,
+      "grad_norm": 1.1620815072786224,
+      "learning_rate": 1.5960677217703512e-05,
+      "loss": 1.5796,
+      "step": 1051
+    },
+    {
+      "epoch": 0.3041341428158427,
+      "grad_norm": 1.1236988795621041,
+      "learning_rate": 1.59533076184287e-05,
+      "loss": 1.6219,
+      "step": 1052
+    },
+    {
+      "epoch": 0.3044232437120555,
+      "grad_norm": 1.1118641361809816,
+      "learning_rate": 1.594593300739087e-05,
+      "loss": 1.6241,
+      "step": 1053
+    },
+    {
+      "epoch": 0.3047123446082683,
+      "grad_norm": 1.1947464614628307,
+      "learning_rate": 1.5938553390798306e-05,
+      "loss": 1.5793,
+      "step": 1054
+    },
+    {
+      "epoch": 0.30500144550448105,
+      "grad_norm": 1.0683464597665702,
+      "learning_rate": 1.59311687748635e-05,
+      "loss": 1.59,
+      "step": 1055
+    },
+    {
+      "epoch": 0.30529054640069386,
+      "grad_norm": 1.148680753466922,
+      "learning_rate": 1.5923779165803163e-05,
+      "loss": 1.5368,
+      "step": 1056
+    },
+    {
+      "epoch": 0.3055796472969066,
+      "grad_norm": 1.0975947872426266,
+      "learning_rate": 1.591638456983819e-05,
+      "loss": 1.5634,
+      "step": 1057
+    },
+    {
+      "epoch": 0.3058687481931194,
+      "grad_norm": 1.0758466082542466,
+      "learning_rate": 1.5908984993193696e-05,
+      "loss": 1.5855,
+      "step": 1058
+    },
+    {
+      "epoch": 0.3061578490893322,
+      "grad_norm": 1.140426199545744,
+      "learning_rate": 1.590158044209897e-05,
+      "loss": 1.6559,
+      "step": 1059
+    },
+    {
+      "epoch": 0.30644694998554495,
+      "grad_norm": 1.1060228312495028,
+      "learning_rate": 1.58941709227875e-05,
+      "loss": 1.6378,
+      "step": 1060
+    },
+    {
+      "epoch": 0.30673605088175776,
+      "grad_norm": 1.113526257443253,
+      "learning_rate": 1.588675644149695e-05,
+      "loss": 1.6179,
+      "step": 1061
+    },
+    {
+      "epoch": 0.3070251517779705,
+      "grad_norm": 1.2173138936643453,
+      "learning_rate": 1.587933700446918e-05,
+      "loss": 1.6806,
+      "step": 1062
+    },
+    {
+      "epoch": 0.3073142526741833,
+      "grad_norm": 1.1400857891987628,
+      "learning_rate": 1.587191261795019e-05,
+      "loss": 1.6742,
+      "step": 1063
+    },
+    {
+      "epoch": 0.3076033535703961,
+      "grad_norm": 1.1305997594773096,
+      "learning_rate": 1.5864483288190174e-05,
+      "loss": 1.5272,
+      "step": 1064
+    },
+    {
+      "epoch": 0.30789245446660884,
+      "grad_norm": 1.228007225532647,
+      "learning_rate": 1.5857049021443467e-05,
+      "loss": 1.7073,
+      "step": 1065
+    },
+    {
+      "epoch": 0.3081815553628216,
+      "grad_norm": 1.162916715072549,
+      "learning_rate": 1.584960982396858e-05,
+      "loss": 1.633,
+      "step": 1066
+    },
+    {
+      "epoch": 0.3084706562590344,
+      "grad_norm": 1.1597744213298864,
+      "learning_rate": 1.5842165702028157e-05,
+      "loss": 1.4521,
+      "step": 1067
+    },
+    {
+      "epoch": 0.30875975715524717,
+      "grad_norm": 1.091472479025558,
+      "learning_rate": 1.5834716661889004e-05,
+      "loss": 1.5902,
+      "step": 1068
+    },
+    {
+      "epoch": 0.30904885805146,
+      "grad_norm": 1.1086742754929058,
+      "learning_rate": 1.5827262709822056e-05,
+      "loss": 1.5939,
+      "step": 1069
+    },
+    {
+      "epoch": 0.30933795894767274,
+      "grad_norm": 1.1593052089546423,
+      "learning_rate": 1.581980385210238e-05,
+      "loss": 1.5925,
+      "step": 1070
+    },
+    {
+      "epoch": 0.3096270598438855,
+      "grad_norm": 1.1650634297424989,
+      "learning_rate": 1.5812340095009192e-05,
+      "loss": 1.6855,
+      "step": 1071
+    },
+    {
+      "epoch": 0.3099161607400983,
+      "grad_norm": 1.099899286022252,
+      "learning_rate": 1.580487144482581e-05,
+      "loss": 1.5522,
+      "step": 1072
+    },
+    {
+      "epoch": 0.31020526163631107,
+      "grad_norm": 1.1052999796757328,
+      "learning_rate": 1.5797397907839678e-05,
+      "loss": 1.5938,
+      "step": 1073
+    },
+    {
+      "epoch": 0.3104943625325238,
+      "grad_norm": 1.1060803844176064,
+      "learning_rate": 1.5789919490342367e-05,
+      "loss": 1.5835,
+      "step": 1074
+    },
+    {
+      "epoch": 0.31078346342873664,
+      "grad_norm": 1.1187707670977645,
+      "learning_rate": 1.578243619862954e-05,
+      "loss": 1.5327,
+      "step": 1075
+    },
+    {
+      "epoch": 0.3110725643249494,
+      "grad_norm": 1.1521574761443172,
+      "learning_rate": 1.5774948039000966e-05,
+      "loss": 1.5998,
+      "step": 1076
+    },
+    {
+      "epoch": 0.3113616652211622,
+      "grad_norm": 1.109080218315599,
+      "learning_rate": 1.5767455017760524e-05,
+      "loss": 1.6479,
+      "step": 1077
+    },
+    {
+      "epoch": 0.31165076611737497,
+      "grad_norm": 1.1370919220061753,
+      "learning_rate": 1.5759957141216175e-05,
+      "loss": 1.5246,
+      "step": 1078
+    },
+    {
+      "epoch": 0.3119398670135877,
+      "grad_norm": 1.113530924054974,
+      "learning_rate": 1.5752454415679968e-05,
+      "loss": 1.6073,
+      "step": 1079
+    },
+    {
+      "epoch": 0.31222896790980054,
+      "grad_norm": 1.0973243348785886,
+      "learning_rate": 1.5744946847468036e-05,
+      "loss": 1.694,
+      "step": 1080
+    },
+    {
+      "epoch": 0.3125180688060133,
+      "grad_norm": 1.1647355097669179,
+      "learning_rate": 1.5737434442900585e-05,
+      "loss": 1.6135,
+      "step": 1081
+    },
+    {
+      "epoch": 0.31280716970222605,
+      "grad_norm": 1.1314628251550698,
+      "learning_rate": 1.5729917208301905e-05,
+      "loss": 1.7268,
+      "step": 1082
+    },
+    {
+      "epoch": 0.31309627059843886,
+      "grad_norm": 1.1557900599799165,
+      "learning_rate": 1.5722395150000336e-05,
+      "loss": 1.6416,
+      "step": 1083
+    },
+    {
+      "epoch": 0.3133853714946516,
+      "grad_norm": 1.0866193980050511,
+      "learning_rate": 1.5714868274328293e-05,
+      "loss": 1.5172,
+      "step": 1084
+    },
+    {
+      "epoch": 0.31367447239086443,
+      "grad_norm": 1.158830944861789,
+      "learning_rate": 1.5707336587622232e-05,
+      "loss": 1.6342,
+      "step": 1085
+    },
+    {
+      "epoch": 0.3139635732870772,
+      "grad_norm": 1.1695982419960425,
+      "learning_rate": 1.569980009622267e-05,
+      "loss": 1.6118,
+      "step": 1086
+    },
+    {
+      "epoch": 0.31425267418328995,
+      "grad_norm": 1.095428236784029,
+      "learning_rate": 1.5692258806474168e-05,
+      "loss": 1.5944,
+      "step": 1087
+    },
+    {
+      "epoch": 0.31454177507950276,
+      "grad_norm": 1.1019964357061058,
+      "learning_rate": 1.568471272472532e-05,
+      "loss": 1.573,
+      "step": 1088
+    },
+    {
+      "epoch": 0.3148308759757155,
+      "grad_norm": 1.1256111784675789,
+      "learning_rate": 1.567716185732876e-05,
+      "loss": 1.604,
+      "step": 1089
+    },
+    {
+      "epoch": 0.3151199768719283,
+      "grad_norm": 1.1148637100070335,
+      "learning_rate": 1.566960621064115e-05,
+      "loss": 1.617,
+      "step": 1090
+    },
+    {
+      "epoch": 0.3154090777681411,
+      "grad_norm": 1.0804672597638105,
+      "learning_rate": 1.566204579102317e-05,
+      "loss": 1.5047,
+      "step": 1091
+    },
+    {
+      "epoch": 0.31569817866435385,
+      "grad_norm": 1.1419136802627774,
+      "learning_rate": 1.5654480604839532e-05,
+      "loss": 1.5378,
+      "step": 1092
+    },
+    {
+      "epoch": 0.31598727956056666,
+      "grad_norm": 1.1498293898506018,
+      "learning_rate": 1.564691065845894e-05,
+      "loss": 1.5626,
+      "step": 1093
+    },
+    {
+      "epoch": 0.3162763804567794,
+      "grad_norm": 1.1143430311083644,
+      "learning_rate": 1.5639335958254123e-05,
+      "loss": 1.595,
+      "step": 1094
+    },
+    {
+      "epoch": 0.3165654813529922,
+      "grad_norm": 1.0931970046054698,
+      "learning_rate": 1.5631756510601802e-05,
+      "loss": 1.4775,
+      "step": 1095
+    },
+    {
+      "epoch": 0.316854582249205,
+      "grad_norm": 1.105514673067743,
+      "learning_rate": 1.5624172321882703e-05,
+      "loss": 1.6032,
+      "step": 1096
+    },
+    {
+      "epoch": 0.31714368314541774,
+      "grad_norm": 1.0931739627772648,
+      "learning_rate": 1.5616583398481533e-05,
+      "loss": 1.6174,
+      "step": 1097
+    },
+    {
+      "epoch": 0.31743278404163056,
+      "grad_norm": 1.1211144837250133,
+      "learning_rate": 1.560898974678699e-05,
+      "loss": 1.7499,
+      "step": 1098
+    },
+    {
+      "epoch": 0.3177218849378433,
+      "grad_norm": 1.1209699013931562,
+      "learning_rate": 1.5601391373191757e-05,
+      "loss": 1.4995,
+      "step": 1099
+    },
+    {
+      "epoch": 0.31801098583405607,
+      "grad_norm": 1.0922080473768103,
+      "learning_rate": 1.559378828409249e-05,
+      "loss": 1.5329,
+      "step": 1100
+    },
+    {
+      "epoch": 0.3183000867302689,
+      "grad_norm": 1.1365989497096085,
+      "learning_rate": 1.5586180485889803e-05,
+      "loss": 1.6751,
+      "step": 1101
+    },
+    {
+      "epoch": 0.31858918762648164,
+      "grad_norm": 1.1489329053278985,
+      "learning_rate": 1.5578567984988292e-05,
+      "loss": 1.6953,
+      "step": 1102
+    },
+    {
+      "epoch": 0.3188782885226944,
+      "grad_norm": 1.1566994177367973,
+      "learning_rate": 1.5570950787796498e-05,
+      "loss": 1.6297,
+      "step": 1103
+    },
+    {
+      "epoch": 0.3191673894189072,
+      "grad_norm": 1.1580906580853507,
+      "learning_rate": 1.5563328900726925e-05,
+      "loss": 1.5271,
+      "step": 1104
+    },
+    {
+      "epoch": 0.31945649031511997,
+      "grad_norm": 1.0995229482590019,
+      "learning_rate": 1.5555702330196024e-05,
+      "loss": 1.6119,
+      "step": 1105
+    },
+    {
+      "epoch": 0.3197455912113328,
+      "grad_norm": 1.0799883018230365,
+      "learning_rate": 1.554807108262418e-05,
+      "loss": 1.6782,
+      "step": 1106
+    },
+    {
+      "epoch": 0.32003469210754554,
+      "grad_norm": 1.1073029882869814,
+      "learning_rate": 1.5540435164435726e-05,
+      "loss": 1.5397,
+      "step": 1107
+    },
+    {
+      "epoch": 0.3203237930037583,
+      "grad_norm": 1.1329749767157702,
+      "learning_rate": 1.5532794582058922e-05,
+      "loss": 1.7132,
+      "step": 1108
+    },
+    {
+      "epoch": 0.3206128938999711,
+      "grad_norm": 1.1204291824946706,
+      "learning_rate": 1.5525149341925953e-05,
+      "loss": 1.7194,
+      "step": 1109
+    },
+    {
+      "epoch": 0.32090199479618386,
+      "grad_norm": 1.1098026934922063,
+      "learning_rate": 1.5517499450472927e-05,
+      "loss": 1.5843,
+      "step": 1110
+    },
+    {
+      "epoch": 0.3211910956923966,
+      "grad_norm": 1.1014699968821742,
+      "learning_rate": 1.550984491413987e-05,
+      "loss": 1.6016,
+      "step": 1111
+    },
+    {
+      "epoch": 0.32148019658860943,
+      "grad_norm": 1.130886787299436,
+      "learning_rate": 1.5502185739370717e-05,
+      "loss": 1.5793,
+      "step": 1112
+    },
+    {
+      "epoch": 0.3217692974848222,
+      "grad_norm": 1.1290955593656395,
+      "learning_rate": 1.5494521932613307e-05,
+      "loss": 1.6501,
+      "step": 1113
+    },
+    {
+      "epoch": 0.322058398381035,
+      "grad_norm": 1.0883652426866905,
+      "learning_rate": 1.548685350031937e-05,
+      "loss": 1.5592,
+      "step": 1114
+    },
+    {
+      "epoch": 0.32234749927724776,
+      "grad_norm": 1.1464074008441614,
+      "learning_rate": 1.547918044894455e-05,
+      "loss": 1.6295,
+      "step": 1115
+    },
+    {
+      "epoch": 0.3226366001734605,
+      "grad_norm": 1.0848408826768916,
+      "learning_rate": 1.5471502784948364e-05,
+      "loss": 1.6202,
+      "step": 1116
+    },
+    {
+      "epoch": 0.32292570106967333,
+      "grad_norm": 1.150507963992238,
+      "learning_rate": 1.5463820514794215e-05,
+      "loss": 1.5854,
+      "step": 1117
+    },
+    {
+      "epoch": 0.3232148019658861,
+      "grad_norm": 1.0976974430882895,
+      "learning_rate": 1.545613364494938e-05,
+      "loss": 1.5758,
+      "step": 1118
+    },
+    {
+      "epoch": 0.32350390286209885,
+      "grad_norm": 1.1069482166625944,
+      "learning_rate": 1.5448442181885026e-05,
+      "loss": 1.6208,
+      "step": 1119
+    },
+    {
+      "epoch": 0.32379300375831166,
+      "grad_norm": 1.2077191144956074,
+      "learning_rate": 1.5440746132076155e-05,
+      "loss": 1.5925,
+      "step": 1120
+    },
+    {
+      "epoch": 0.3240821046545244,
+      "grad_norm": 1.1541833705064752,
+      "learning_rate": 1.5433045502001668e-05,
+      "loss": 1.6509,
+      "step": 1121
+    },
+    {
+      "epoch": 0.32437120555073723,
+      "grad_norm": 1.1100819739423928,
+      "learning_rate": 1.542534029814429e-05,
+      "loss": 1.5918,
+      "step": 1122
+    },
+    {
+      "epoch": 0.32466030644695,
+      "grad_norm": 1.1765209537252967,
+      "learning_rate": 1.5417630526990613e-05,
+      "loss": 1.6111,
+      "step": 1123
+    },
+    {
+      "epoch": 0.32494940734316274,
+      "grad_norm": 1.092648262023873,
+      "learning_rate": 1.540991619503107e-05,
+      "loss": 1.5288,
+      "step": 1124
+    },
+    {
+      "epoch": 0.32523850823937556,
+      "grad_norm": 1.1276180188457035,
+      "learning_rate": 1.540219730875994e-05,
+      "loss": 1.5382,
+      "step": 1125
+    },
+    {
+      "epoch": 0.3255276091355883,
+      "grad_norm": 1.571590596776389,
+      "learning_rate": 1.5394473874675325e-05,
+      "loss": 1.6576,
+      "step": 1126
+    },
+    {
+      "epoch": 0.32581671003180107,
+      "grad_norm": 1.118162272885637,
+      "learning_rate": 1.5386745899279155e-05,
+      "loss": 1.5767,
+      "step": 1127
+    },
+    {
+      "epoch": 0.3261058109280139,
+      "grad_norm": 1.0863276376658753,
+      "learning_rate": 1.5379013389077193e-05,
+      "loss": 1.5751,
+      "step": 1128
+    },
+    {
+      "epoch": 0.32639491182422664,
+      "grad_norm": 1.1100894550632194,
+      "learning_rate": 1.537127635057902e-05,
+      "loss": 1.4374,
+      "step": 1129
+    },
+    {
+      "epoch": 0.32668401272043945,
+      "grad_norm": 1.086813502970389,
+      "learning_rate": 1.5363534790298012e-05,
+      "loss": 1.6202,
+      "step": 1130
+    },
+    {
+      "epoch": 0.3269731136166522,
+      "grad_norm": 1.1464824890024201,
+      "learning_rate": 1.5355788714751378e-05,
+      "loss": 1.5852,
+      "step": 1131
+    },
+    {
+      "epoch": 0.32726221451286497,
+      "grad_norm": 1.1665253096605663,
+      "learning_rate": 1.5348038130460106e-05,
+      "loss": 1.6235,
+      "step": 1132
+    },
+    {
+      "epoch": 0.3275513154090778,
+      "grad_norm": 1.137364189884065,
+      "learning_rate": 1.534028304394899e-05,
+      "loss": 1.5424,
+      "step": 1133
+    },
+    {
+      "epoch": 0.32784041630529054,
+      "grad_norm": 1.1333000047594932,
+      "learning_rate": 1.5332523461746607e-05,
+      "loss": 1.5362,
+      "step": 1134
+    },
+    {
+      "epoch": 0.32812951720150335,
+      "grad_norm": 1.189175584600463,
+      "learning_rate": 1.5324759390385326e-05,
+      "loss": 1.6452,
+      "step": 1135
+    },
+    {
+      "epoch": 0.3284186180977161,
+      "grad_norm": 1.8321030345914788,
+      "learning_rate": 1.53169908364013e-05,
+      "loss": 1.7045,
+      "step": 1136
+    },
+    {
+      "epoch": 0.32870771899392887,
+      "grad_norm": 1.1179090119797042,
+      "learning_rate": 1.530921780633444e-05,
+      "loss": 1.5984,
+      "step": 1137
+    },
+    {
+      "epoch": 0.3289968198901417,
+      "grad_norm": 1.1414953551831817,
+      "learning_rate": 1.5301440306728433e-05,
+      "loss": 1.6185,
+      "step": 1138
+    },
+    {
+      "epoch": 0.32928592078635444,
+      "grad_norm": 1.036221101321721,
+      "learning_rate": 1.5293658344130734e-05,
+      "loss": 1.5744,
+      "step": 1139
+    },
+    {
+      "epoch": 0.3295750216825672,
+      "grad_norm": 1.5416356009448273,
+      "learning_rate": 1.5285871925092545e-05,
+      "loss": 1.6635,
+      "step": 1140
+    },
+    {
+      "epoch": 0.32986412257878,
+      "grad_norm": 1.2162733684436655,
+      "learning_rate": 1.527808105616883e-05,
+      "loss": 1.6407,
+      "step": 1141
+    },
+    {
+      "epoch": 0.33015322347499276,
+      "grad_norm": 1.1033030184596553,
+      "learning_rate": 1.5270285743918294e-05,
+      "loss": 1.6466,
+      "step": 1142
+    },
+    {
+      "epoch": 0.3304423243712056,
+      "grad_norm": 1.109124489223676,
+      "learning_rate": 1.526248599490338e-05,
+      "loss": 1.7746,
+      "step": 1143
+    },
+    {
+      "epoch": 0.33073142526741833,
+      "grad_norm": 1.092288021297295,
+      "learning_rate": 1.5254681815690271e-05,
+      "loss": 1.526,
+      "step": 1144
+    },
+    {
+      "epoch": 0.3310205261636311,
+      "grad_norm": 1.1575352020094662,
+      "learning_rate": 1.5246873212848874e-05,
+      "loss": 1.5878,
+      "step": 1145
+    },
+    {
+      "epoch": 0.3313096270598439,
+      "grad_norm": 1.1405141791110793,
+      "learning_rate": 1.5239060192952821e-05,
+      "loss": 1.5774,
+      "step": 1146
+    },
+    {
+      "epoch": 0.33159872795605666,
+      "grad_norm": 1.122962325771197,
+      "learning_rate": 1.5231242762579474e-05,
+      "loss": 1.5913,
+      "step": 1147
+    },
+    {
+      "epoch": 0.3318878288522694,
+      "grad_norm": 1.2373901697470753,
+      "learning_rate": 1.5223420928309889e-05,
+      "loss": 1.6305,
+      "step": 1148
+    },
+    {
+      "epoch": 0.33217692974848223,
+      "grad_norm": 1.091007710090704,
+      "learning_rate": 1.5215594696728849e-05,
+      "loss": 1.6023,
+      "step": 1149
+    },
+    {
+      "epoch": 0.332466030644695,
+      "grad_norm": 1.1020358370583065,
+      "learning_rate": 1.5207764074424818e-05,
+      "loss": 1.5833,
+      "step": 1150
+    },
+    {
+      "epoch": 0.3327551315409078,
+      "grad_norm": 1.1069948831247647,
+      "learning_rate": 1.5199929067989975e-05,
+      "loss": 1.6207,
+      "step": 1151
+    },
+    {
+      "epoch": 0.33304423243712056,
+      "grad_norm": 1.15867755196439,
+      "learning_rate": 1.5192089684020179e-05,
+      "loss": 1.6841,
+      "step": 1152
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 1.06051478046549,
+      "learning_rate": 1.518424592911498e-05,
+      "loss": 1.6747,
+      "step": 1153
+    },
+    {
+      "epoch": 0.3336224342295461,
+      "grad_norm": 1.0941722623539405,
+      "learning_rate": 1.5176397809877599e-05,
+      "loss": 1.5649,
+      "step": 1154
+    },
+    {
+      "epoch": 0.3339115351257589,
+      "grad_norm": 1.065308694755837,
+      "learning_rate": 1.5168545332914942e-05,
+      "loss": 1.5981,
+      "step": 1155
+    },
+    {
+      "epoch": 0.33420063602197164,
+      "grad_norm": 1.103949052624009,
+      "learning_rate": 1.5160688504837577e-05,
+      "loss": 1.6367,
+      "step": 1156
+    },
+    {
+      "epoch": 0.33448973691818445,
+      "grad_norm": 1.108624129747703,
+      "learning_rate": 1.5152827332259738e-05,
+      "loss": 1.5322,
+      "step": 1157
+    },
+    {
+      "epoch": 0.3347788378143972,
+      "grad_norm": 1.171699585963367,
+      "learning_rate": 1.5144961821799314e-05,
+      "loss": 1.6257,
+      "step": 1158
+    },
+    {
+      "epoch": 0.33506793871061,
+      "grad_norm": 1.1189026609089276,
+      "learning_rate": 1.5137091980077842e-05,
+      "loss": 1.6154,
+      "step": 1159
+    },
+    {
+      "epoch": 0.3353570396068228,
+      "grad_norm": 1.0695939123636227,
+      "learning_rate": 1.5129217813720516e-05,
+      "loss": 1.6156,
+      "step": 1160
+    },
+    {
+      "epoch": 0.33564614050303554,
+      "grad_norm": 1.1139701552322105,
+      "learning_rate": 1.5121339329356163e-05,
+      "loss": 1.6639,
+      "step": 1161
+    },
+    {
+      "epoch": 0.33593524139924835,
+      "grad_norm": 1.0899522793859586,
+      "learning_rate": 1.5113456533617243e-05,
+      "loss": 1.5977,
+      "step": 1162
+    },
+    {
+      "epoch": 0.3362243422954611,
+      "grad_norm": 1.1231652645265944,
+      "learning_rate": 1.5105569433139859e-05,
+      "loss": 1.5887,
+      "step": 1163
+    },
+    {
+      "epoch": 0.33651344319167387,
+      "grad_norm": 1.1311200256557958,
+      "learning_rate": 1.5097678034563716e-05,
+      "loss": 1.6756,
+      "step": 1164
+    },
+    {
+      "epoch": 0.3368025440878867,
+      "grad_norm": 1.0808514264048,
+      "learning_rate": 1.5089782344532155e-05,
+      "loss": 1.649,
+      "step": 1165
+    },
+    {
+      "epoch": 0.33709164498409944,
+      "grad_norm": 1.1031304339670052,
+      "learning_rate": 1.5081882369692129e-05,
+      "loss": 1.6388,
+      "step": 1166
+    },
+    {
+      "epoch": 0.33738074588031225,
+      "grad_norm": 1.0793600842313416,
+      "learning_rate": 1.507397811669418e-05,
+      "loss": 1.5254,
+      "step": 1167
+    },
+    {
+      "epoch": 0.337669846776525,
+      "grad_norm": 1.2081295628157755,
+      "learning_rate": 1.5066069592192483e-05,
+      "loss": 1.6703,
+      "step": 1168
+    },
+    {
+      "epoch": 0.33795894767273776,
+      "grad_norm": 1.0883959306735564,
+      "learning_rate": 1.5058156802844774e-05,
+      "loss": 1.579,
+      "step": 1169
+    },
+    {
+      "epoch": 0.3382480485689506,
+      "grad_norm": 1.1328433977876502,
+      "learning_rate": 1.5050239755312405e-05,
+      "loss": 1.5702,
+      "step": 1170
+    },
+    {
+      "epoch": 0.33853714946516333,
+      "grad_norm": 1.1355213706257294,
+      "learning_rate": 1.5042318456260305e-05,
+      "loss": 1.6607,
+      "step": 1171
+    },
+    {
+      "epoch": 0.33882625036137615,
+      "grad_norm": 1.1058142406765221,
+      "learning_rate": 1.5034392912356973e-05,
+      "loss": 1.6631,
+      "step": 1172
+    },
+    {
+      "epoch": 0.3391153512575889,
+      "grad_norm": 1.063531469536154,
+      "learning_rate": 1.50264631302745e-05,
+      "loss": 1.6077,
+      "step": 1173
+    },
+    {
+      "epoch": 0.33940445215380166,
+      "grad_norm": 1.127116435978979,
+      "learning_rate": 1.5018529116688526e-05,
+      "loss": 1.585,
+      "step": 1174
+    },
+    {
+      "epoch": 0.3396935530500145,
+      "grad_norm": 1.0953876410839178,
+      "learning_rate": 1.5010590878278266e-05,
+      "loss": 1.5929,
+      "step": 1175
+    },
+    {
+      "epoch": 0.33998265394622723,
+      "grad_norm": 1.0550371976819637,
+      "learning_rate": 1.5002648421726487e-05,
+      "loss": 1.5637,
+      "step": 1176
+    },
+    {
+      "epoch": 0.34027175484244,
+      "grad_norm": 1.0884184120480185,
+      "learning_rate": 1.4994701753719505e-05,
+      "loss": 1.5982,
+      "step": 1177
+    },
+    {
+      "epoch": 0.3405608557386528,
+      "grad_norm": 1.114897410077523,
+      "learning_rate": 1.4986750880947184e-05,
+      "loss": 1.6062,
+      "step": 1178
+    },
+    {
+      "epoch": 0.34084995663486556,
+      "grad_norm": 1.066891980414234,
+      "learning_rate": 1.497879581010293e-05,
+      "loss": 1.606,
+      "step": 1179
+    },
+    {
+      "epoch": 0.34113905753107837,
+      "grad_norm": 1.0880910736641207,
+      "learning_rate": 1.4970836547883681e-05,
+      "loss": 1.5708,
+      "step": 1180
+    },
+    {
+      "epoch": 0.34142815842729113,
+      "grad_norm": 1.1167707203657147,
+      "learning_rate": 1.49628731009899e-05,
+      "loss": 1.5817,
+      "step": 1181
+    },
+    {
+      "epoch": 0.3417172593235039,
+      "grad_norm": 1.1339896953914148,
+      "learning_rate": 1.4954905476125578e-05,
+      "loss": 1.6754,
+      "step": 1182
+    },
+    {
+      "epoch": 0.3420063602197167,
+      "grad_norm": 1.1010220576391134,
+      "learning_rate": 1.494693367999822e-05,
+      "loss": 1.5575,
+      "step": 1183
+    },
+    {
+      "epoch": 0.34229546111592946,
+      "grad_norm": 1.2739889022885835,
+      "learning_rate": 1.4938957719318845e-05,
+      "loss": 1.5497,
+      "step": 1184
+    },
+    {
+      "epoch": 0.3425845620121422,
+      "grad_norm": 1.148232650591867,
+      "learning_rate": 1.4930977600801977e-05,
+      "loss": 1.5948,
+      "step": 1185
+    },
+    {
+      "epoch": 0.342873662908355,
+      "grad_norm": 1.1837885057183477,
+      "learning_rate": 1.4922993331165639e-05,
+      "loss": 1.6017,
+      "step": 1186
+    },
+    {
+      "epoch": 0.3431627638045678,
+      "grad_norm": 1.1050078395263963,
+      "learning_rate": 1.4915004917131345e-05,
+      "loss": 1.6225,
+      "step": 1187
+    },
+    {
+      "epoch": 0.3434518647007806,
+      "grad_norm": 1.0565673964453821,
+      "learning_rate": 1.4907012365424112e-05,
+      "loss": 1.6723,
+      "step": 1188
+    },
+    {
+      "epoch": 0.34374096559699335,
+      "grad_norm": 1.0426217055278246,
+      "learning_rate": 1.4899015682772422e-05,
+      "loss": 1.5567,
+      "step": 1189
+    },
+    {
+      "epoch": 0.3440300664932061,
+      "grad_norm": 1.0750133699973452,
+      "learning_rate": 1.489101487590825e-05,
+      "loss": 1.5984,
+      "step": 1190
+    },
+    {
+      "epoch": 0.3443191673894189,
+      "grad_norm": 1.0444023205829724,
+      "learning_rate": 1.488300995156703e-05,
+      "loss": 1.524,
+      "step": 1191
+    },
+    {
+      "epoch": 0.3446082682856317,
+      "grad_norm": 1.1113297257235197,
+      "learning_rate": 1.4875000916487674e-05,
+      "loss": 1.6477,
+      "step": 1192
+    },
+    {
+      "epoch": 0.34489736918184444,
+      "grad_norm": 1.0700189539478588,
+      "learning_rate": 1.4866987777412548e-05,
+      "loss": 1.6747,
+      "step": 1193
+    },
+    {
+      "epoch": 0.34518647007805725,
+      "grad_norm": 1.0791046704423155,
+      "learning_rate": 1.485897054108747e-05,
+      "loss": 1.5364,
+      "step": 1194
+    },
+    {
+      "epoch": 0.34547557097427,
+      "grad_norm": 1.1152714216930488,
+      "learning_rate": 1.4850949214261717e-05,
+      "loss": 1.6436,
+      "step": 1195
+    },
+    {
+      "epoch": 0.3457646718704828,
+      "grad_norm": 1.1163524885739,
+      "learning_rate": 1.4842923803687999e-05,
+      "loss": 1.6719,
+      "step": 1196
+    },
+    {
+      "epoch": 0.3460537727666956,
+      "grad_norm": 1.1470433328057497,
+      "learning_rate": 1.483489431612247e-05,
+      "loss": 1.6618,
+      "step": 1197
+    },
+    {
+      "epoch": 0.34634287366290833,
+      "grad_norm": 1.1152658216672549,
+      "learning_rate": 1.482686075832472e-05,
+      "loss": 1.6082,
+      "step": 1198
+    },
+    {
+      "epoch": 0.34663197455912115,
+      "grad_norm": 1.087512082279524,
+      "learning_rate": 1.4818823137057751e-05,
+      "loss": 1.6302,
+      "step": 1199
+    },
+    {
+      "epoch": 0.3469210754553339,
+      "grad_norm": 1.1269711018498454,
+      "learning_rate": 1.4810781459088009e-05,
+      "loss": 1.6265,
+      "step": 1200
+    },
+    {
+      "epoch": 0.34721017635154666,
+      "grad_norm": 1.1738148200035237,
+      "learning_rate": 1.4802735731185327e-05,
+      "loss": 1.7348,
+      "step": 1201
+    },
+    {
+      "epoch": 0.3474992772477595,
+      "grad_norm": 1.0479144362222594,
+      "learning_rate": 1.4794685960122967e-05,
+      "loss": 1.5731,
+      "step": 1202
+    },
+    {
+      "epoch": 0.34778837814397223,
+      "grad_norm": 1.1183507786337057,
+      "learning_rate": 1.4786632152677596e-05,
+      "loss": 1.6627,
+      "step": 1203
+    },
+    {
+      "epoch": 0.34807747904018504,
+      "grad_norm": 1.0560834768468748,
+      "learning_rate": 1.4778574315629262e-05,
+      "loss": 1.5118,
+      "step": 1204
+    },
+    {
+      "epoch": 0.3483665799363978,
+      "grad_norm": 1.1504897519151327,
+      "learning_rate": 1.4770512455761425e-05,
+      "loss": 1.6384,
+      "step": 1205
+    },
+    {
+      "epoch": 0.34865568083261056,
+      "grad_norm": 1.128971905615821,
+      "learning_rate": 1.476244657986092e-05,
+      "loss": 1.6184,
+      "step": 1206
+    },
+    {
+      "epoch": 0.34894478172882337,
+      "grad_norm": 1.1323411355011963,
+      "learning_rate": 1.4754376694717958e-05,
+      "loss": 1.742,
+      "step": 1207
+    },
+    {
+      "epoch": 0.34923388262503613,
+      "grad_norm": 1.0865319922492744,
+      "learning_rate": 1.4746302807126142e-05,
+      "loss": 1.5589,
+      "step": 1208
+    },
+    {
+      "epoch": 0.34952298352124894,
+      "grad_norm": 1.1013771826316385,
+      "learning_rate": 1.473822492388243e-05,
+      "loss": 1.5835,
+      "step": 1209
+    },
+    {
+      "epoch": 0.3498120844174617,
+      "grad_norm": 1.133500512314171,
+      "learning_rate": 1.4730143051787151e-05,
+      "loss": 1.5905,
+      "step": 1210
+    },
+    {
+      "epoch": 0.35010118531367446,
+      "grad_norm": 1.0512972757957857,
+      "learning_rate": 1.4722057197643986e-05,
+      "loss": 1.5644,
+      "step": 1211
+    },
+    {
+      "epoch": 0.35039028620988727,
+      "grad_norm": 1.0982556984479055,
+      "learning_rate": 1.4713967368259981e-05,
+      "loss": 1.617,
+      "step": 1212
+    },
+    {
+      "epoch": 0.3506793871061,
+      "grad_norm": 1.1056551102940748,
+      "learning_rate": 1.4705873570445506e-05,
+      "loss": 1.586,
+      "step": 1213
+    },
+    {
+      "epoch": 0.3509684880023128,
+      "grad_norm": 1.1194856012781043,
+      "learning_rate": 1.4697775811014294e-05,
+      "loss": 1.6349,
+      "step": 1214
+    },
+    {
+      "epoch": 0.3512575888985256,
+      "grad_norm": 1.1635585093324126,
+      "learning_rate": 1.4689674096783404e-05,
+      "loss": 1.7592,
+      "step": 1215
+    },
+    {
+      "epoch": 0.35154668979473835,
+      "grad_norm": 1.0761660420536625,
+      "learning_rate": 1.4681568434573221e-05,
+      "loss": 1.6587,
+      "step": 1216
+    },
+    {
+      "epoch": 0.35183579069095117,
+      "grad_norm": 1.0862955186287286,
+      "learning_rate": 1.4673458831207463e-05,
+      "loss": 1.6006,
+      "step": 1217
+    },
+    {
+      "epoch": 0.3521248915871639,
+      "grad_norm": 1.0784753202499149,
+      "learning_rate": 1.4665345293513154e-05,
+      "loss": 1.5591,
+      "step": 1218
+    },
+    {
+      "epoch": 0.3524139924833767,
+      "grad_norm": 1.1153278904671513,
+      "learning_rate": 1.4657227828320637e-05,
+      "loss": 1.5676,
+      "step": 1219
+    },
+    {
+      "epoch": 0.3527030933795895,
+      "grad_norm": 1.0768215016106648,
+      "learning_rate": 1.4649106442463565e-05,
+      "loss": 1.5335,
+      "step": 1220
+    },
+    {
+      "epoch": 0.35299219427580225,
+      "grad_norm": 1.1314763789684028,
+      "learning_rate": 1.4640981142778883e-05,
+      "loss": 1.6171,
+      "step": 1221
+    },
+    {
+      "epoch": 0.353281295172015,
+      "grad_norm": 1.0990493963931165,
+      "learning_rate": 1.4632851936106838e-05,
+      "loss": 1.6048,
+      "step": 1222
+    },
+    {
+      "epoch": 0.3535703960682278,
+      "grad_norm": 1.085058079934556,
+      "learning_rate": 1.462471882929096e-05,
+      "loss": 1.6296,
+      "step": 1223
+    },
+    {
+      "epoch": 0.3538594969644406,
+      "grad_norm": 1.132646612791557,
+      "learning_rate": 1.4616581829178065e-05,
+      "loss": 1.6962,
+      "step": 1224
+    },
+    {
+      "epoch": 0.3541485978606534,
+      "grad_norm": 1.08766264140374,
+      "learning_rate": 1.4608440942618252e-05,
+      "loss": 1.5844,
+      "step": 1225
+    },
+    {
+      "epoch": 0.35443769875686615,
+      "grad_norm": 1.0655552716344114,
+      "learning_rate": 1.4600296176464878e-05,
+      "loss": 1.6268,
+      "step": 1226
+    },
+    {
+      "epoch": 0.3547267996530789,
+      "grad_norm": 1.0715746687267362,
+      "learning_rate": 1.4592147537574585e-05,
+      "loss": 1.7251,
+      "step": 1227
+    },
+    {
+      "epoch": 0.3550159005492917,
+      "grad_norm": 1.153862024978196,
+      "learning_rate": 1.4583995032807259e-05,
+      "loss": 1.5546,
+      "step": 1228
+    },
+    {
+      "epoch": 0.3553050014455045,
+      "grad_norm": 1.0940461925978813,
+      "learning_rate": 1.4575838669026045e-05,
+      "loss": 1.5783,
+      "step": 1229
+    },
+    {
+      "epoch": 0.35559410234171723,
+      "grad_norm": 1.088784612051587,
+      "learning_rate": 1.4567678453097345e-05,
+      "loss": 1.6016,
+      "step": 1230
+    },
+    {
+      "epoch": 0.35588320323793005,
+      "grad_norm": 1.1233954959369794,
+      "learning_rate": 1.4559514391890788e-05,
+      "loss": 1.642,
+      "step": 1231
+    },
+    {
+      "epoch": 0.3561723041341428,
+      "grad_norm": 1.0795613752451976,
+      "learning_rate": 1.4551346492279257e-05,
+      "loss": 1.5359,
+      "step": 1232
+    },
+    {
+      "epoch": 0.3564614050303556,
+      "grad_norm": 1.1164982431881545,
+      "learning_rate": 1.4543174761138852e-05,
+      "loss": 1.537,
+      "step": 1233
+    },
+    {
+      "epoch": 0.3567505059265684,
+      "grad_norm": 1.0439396898471973,
+      "learning_rate": 1.4534999205348909e-05,
+      "loss": 1.5957,
+      "step": 1234
+    },
+    {
+      "epoch": 0.35703960682278113,
+      "grad_norm": 1.094478459597531,
+      "learning_rate": 1.4526819831791983e-05,
+      "loss": 1.5261,
+      "step": 1235
+    },
+    {
+      "epoch": 0.35732870771899394,
+      "grad_norm": 1.0523765417826885,
+      "learning_rate": 1.451863664735383e-05,
+      "loss": 1.7109,
+      "step": 1236
+    },
+    {
+      "epoch": 0.3576178086152067,
+      "grad_norm": 1.0697972681513481,
+      "learning_rate": 1.4510449658923431e-05,
+      "loss": 1.5079,
+      "step": 1237
+    },
+    {
+      "epoch": 0.3579069095114195,
+      "grad_norm": 1.0776837202479659,
+      "learning_rate": 1.4502258873392963e-05,
+      "loss": 1.5373,
+      "step": 1238
+    },
+    {
+      "epoch": 0.35819601040763227,
+      "grad_norm": 1.0948923626052913,
+      "learning_rate": 1.4494064297657795e-05,
+      "loss": 1.5953,
+      "step": 1239
+    },
+    {
+      "epoch": 0.358485111303845,
+      "grad_norm": 1.0851502014434444,
+      "learning_rate": 1.4485865938616495e-05,
+      "loss": 1.6227,
+      "step": 1240
+    },
+    {
+      "epoch": 0.35877421220005784,
+      "grad_norm": 1.1036122017363585,
+      "learning_rate": 1.4477663803170807e-05,
+      "loss": 1.6181,
+      "step": 1241
+    },
+    {
+      "epoch": 0.3590633130962706,
+      "grad_norm": 1.073741412650114,
+      "learning_rate": 1.446945789822566e-05,
+      "loss": 1.6449,
+      "step": 1242
+    },
+    {
+      "epoch": 0.35935241399248335,
+      "grad_norm": 1.1071745305977874,
+      "learning_rate": 1.446124823068916e-05,
+      "loss": 1.6455,
+      "step": 1243
+    },
+    {
+      "epoch": 0.35964151488869617,
+      "grad_norm": 1.1469504663679086,
+      "learning_rate": 1.445303480747257e-05,
+      "loss": 1.5989,
+      "step": 1244
+    },
+    {
+      "epoch": 0.3599306157849089,
+      "grad_norm": 1.1129037111462516,
+      "learning_rate": 1.4444817635490322e-05,
+      "loss": 1.6613,
+      "step": 1245
+    },
+    {
+      "epoch": 0.36021971668112174,
+      "grad_norm": 1.0968054047932645,
+      "learning_rate": 1.4436596721660002e-05,
+      "loss": 1.601,
+      "step": 1246
+    },
+    {
+      "epoch": 0.3605088175773345,
+      "grad_norm": 1.0927563331941559,
+      "learning_rate": 1.4428372072902344e-05,
+      "loss": 1.7164,
+      "step": 1247
+    },
+    {
+      "epoch": 0.36079791847354725,
+      "grad_norm": 1.1272073503734452,
+      "learning_rate": 1.4420143696141233e-05,
+      "loss": 1.5526,
+      "step": 1248
+    },
+    {
+      "epoch": 0.36108701936976006,
+      "grad_norm": 1.0615706091543697,
+      "learning_rate": 1.4411911598303688e-05,
+      "loss": 1.5344,
+      "step": 1249
+    },
+    {
+      "epoch": 0.3613761202659728,
+      "grad_norm": 1.0434496363748602,
+      "learning_rate": 1.4403675786319856e-05,
+      "loss": 1.5325,
+      "step": 1250
+    },
+    {
+      "epoch": 0.3616652211621856,
+      "grad_norm": 1.1158624940443362,
+      "learning_rate": 1.4395436267123017e-05,
+      "loss": 1.6085,
+      "step": 1251
+    },
+    {
+      "epoch": 0.3619543220583984,
+      "grad_norm": 1.1545246739231232,
+      "learning_rate": 1.4387193047649572e-05,
+      "loss": 1.6314,
+      "step": 1252
+    },
+    {
+      "epoch": 0.36224342295461115,
+      "grad_norm": 1.1500739858412437,
+      "learning_rate": 1.437894613483903e-05,
+      "loss": 1.5861,
+      "step": 1253
+    },
+    {
+      "epoch": 0.36253252385082396,
+      "grad_norm": 1.021310863594055,
+      "learning_rate": 1.4370695535634024e-05,
+      "loss": 1.509,
+      "step": 1254
+    },
+    {
+      "epoch": 0.3628216247470367,
+      "grad_norm": 1.086222036655496,
+      "learning_rate": 1.4362441256980271e-05,
+      "loss": 1.6284,
+      "step": 1255
+    },
+    {
+      "epoch": 0.3631107256432495,
+      "grad_norm": 1.120109032505505,
+      "learning_rate": 1.4354183305826601e-05,
+      "loss": 1.5861,
+      "step": 1256
+    },
+    {
+      "epoch": 0.3633998265394623,
+      "grad_norm": 1.0968623723093054,
+      "learning_rate": 1.4345921689124928e-05,
+      "loss": 1.6583,
+      "step": 1257
+    },
+    {
+      "epoch": 0.36368892743567505,
+      "grad_norm": 1.0757768358923314,
+      "learning_rate": 1.4337656413830252e-05,
+      "loss": 1.6732,
+      "step": 1258
+    },
+    {
+      "epoch": 0.3639780283318878,
+      "grad_norm": 1.1072185273450617,
+      "learning_rate": 1.4329387486900659e-05,
+      "loss": 1.5613,
+      "step": 1259
+    },
+    {
+      "epoch": 0.3642671292281006,
+      "grad_norm": 1.1152570699608948,
+      "learning_rate": 1.4321114915297298e-05,
+      "loss": 1.7207,
+      "step": 1260
+    },
+    {
+      "epoch": 0.3645562301243134,
+      "grad_norm": 1.1181703343533083,
+      "learning_rate": 1.4312838705984403e-05,
+      "loss": 1.5774,
+      "step": 1261
+    },
+    {
+      "epoch": 0.3648453310205262,
+      "grad_norm": 1.1234819294312621,
+      "learning_rate": 1.4304558865929251e-05,
+      "loss": 1.5807,
+      "step": 1262
+    },
+    {
+      "epoch": 0.36513443191673894,
+      "grad_norm": 1.1065221175707145,
+      "learning_rate": 1.429627540210219e-05,
+      "loss": 1.5762,
+      "step": 1263
+    },
+    {
+      "epoch": 0.3654235328129517,
+      "grad_norm": 1.061140154200443,
+      "learning_rate": 1.4287988321476615e-05,
+      "loss": 1.6927,
+      "step": 1264
+    },
+    {
+      "epoch": 0.3657126337091645,
+      "grad_norm": 1.0999054498119034,
+      "learning_rate": 1.4279697631028961e-05,
+      "loss": 1.5921,
+      "step": 1265
+    },
+    {
+      "epoch": 0.36600173460537727,
+      "grad_norm": 1.1096047939134852,
+      "learning_rate": 1.4271403337738706e-05,
+      "loss": 1.5883,
+      "step": 1266
+    },
+    {
+      "epoch": 0.36629083550159003,
+      "grad_norm": 1.117311000405351,
+      "learning_rate": 1.4263105448588362e-05,
+      "loss": 1.4585,
+      "step": 1267
+    },
+    {
+      "epoch": 0.36657993639780284,
+      "grad_norm": 1.1164044045072312,
+      "learning_rate": 1.4254803970563464e-05,
+      "loss": 1.6306,
+      "step": 1268
+    },
+    {
+      "epoch": 0.3668690372940156,
+      "grad_norm": 1.0628514308268988,
+      "learning_rate": 1.4246498910652573e-05,
+      "loss": 1.5824,
+      "step": 1269
+    },
+    {
+      "epoch": 0.3671581381902284,
+      "grad_norm": 1.181064137470451,
+      "learning_rate": 1.4238190275847262e-05,
+      "loss": 1.6634,
+      "step": 1270
+    },
+    {
+      "epoch": 0.36744723908644117,
+      "grad_norm": 1.0887767912574196,
+      "learning_rate": 1.4229878073142114e-05,
+      "loss": 1.5741,
+      "step": 1271
+    },
+    {
+      "epoch": 0.3677363399826539,
+      "grad_norm": 1.140612642718485,
+      "learning_rate": 1.4221562309534722e-05,
+      "loss": 1.6199,
+      "step": 1272
+    },
+    {
+      "epoch": 0.36802544087886674,
+      "grad_norm": 1.1783854536403464,
+      "learning_rate": 1.4213242992025662e-05,
+      "loss": 1.6202,
+      "step": 1273
+    },
+    {
+      "epoch": 0.3683145417750795,
+      "grad_norm": 1.2403828595576905,
+      "learning_rate": 1.4204920127618517e-05,
+      "loss": 1.652,
+      "step": 1274
+    },
+    {
+      "epoch": 0.3686036426712923,
+      "grad_norm": 1.14066458994363,
+      "learning_rate": 1.419659372331985e-05,
+      "loss": 1.7794,
+      "step": 1275
+    },
+    {
+      "epoch": 0.36889274356750507,
+      "grad_norm": 1.093343920812435,
+      "learning_rate": 1.41882637861392e-05,
+      "loss": 1.5567,
+      "step": 1276
+    },
+    {
+      "epoch": 0.3691818444637178,
+      "grad_norm": 1.0808707505493302,
+      "learning_rate": 1.4179930323089088e-05,
+      "loss": 1.5926,
+      "step": 1277
+    },
+    {
+      "epoch": 0.36947094535993064,
+      "grad_norm": 1.0509905823110333,
+      "learning_rate": 1.4171593341184997e-05,
+      "loss": 1.5042,
+      "step": 1278
+    },
+    {
+      "epoch": 0.3697600462561434,
+      "grad_norm": 1.1944099137159172,
+      "learning_rate": 1.4163252847445375e-05,
+      "loss": 1.6389,
+      "step": 1279
+    },
+    {
+      "epoch": 0.37004914715235615,
+      "grad_norm": 1.0550018481082393,
+      "learning_rate": 1.4154908848891626e-05,
+      "loss": 1.6389,
+      "step": 1280
+    },
+    {
+      "epoch": 0.37033824804856896,
+      "grad_norm": 1.144736846632199,
+      "learning_rate": 1.4146561352548109e-05,
+      "loss": 1.5658,
+      "step": 1281
+    },
+    {
+      "epoch": 0.3706273489447817,
+      "grad_norm": 1.1625366943570545,
+      "learning_rate": 1.4138210365442113e-05,
+      "loss": 1.6057,
+      "step": 1282
+    },
+    {
+      "epoch": 0.37091644984099453,
+      "grad_norm": 1.1240282635958534,
+      "learning_rate": 1.4129855894603885e-05,
+      "loss": 1.6278,
+      "step": 1283
+    },
+    {
+      "epoch": 0.3712055507372073,
+      "grad_norm": 1.120499316106146,
+      "learning_rate": 1.4121497947066593e-05,
+      "loss": 1.5968,
+      "step": 1284
+    },
+    {
+      "epoch": 0.37149465163342005,
+      "grad_norm": 1.1248421514320672,
+      "learning_rate": 1.4113136529866331e-05,
+      "loss": 1.6972,
+      "step": 1285
+    },
+    {
+      "epoch": 0.37178375252963286,
+      "grad_norm": 1.1229415098527773,
+      "learning_rate": 1.4104771650042122e-05,
+      "loss": 1.5947,
+      "step": 1286
+    },
+    {
+      "epoch": 0.3720728534258456,
+      "grad_norm": 1.0792013166290457,
+      "learning_rate": 1.4096403314635897e-05,
+      "loss": 1.5431,
+      "step": 1287
+    },
+    {
+      "epoch": 0.3723619543220584,
+      "grad_norm": 1.1584266878751586,
+      "learning_rate": 1.4088031530692497e-05,
+      "loss": 1.579,
+      "step": 1288
+    },
+    {
+      "epoch": 0.3726510552182712,
+      "grad_norm": 1.1181226193149765,
+      "learning_rate": 1.4079656305259673e-05,
+      "loss": 1.5989,
+      "step": 1289
+    },
+    {
+      "epoch": 0.37294015611448394,
+      "grad_norm": 1.139645588978396,
+      "learning_rate": 1.407127764538806e-05,
+      "loss": 1.5868,
+      "step": 1290
+    },
+    {
+      "epoch": 0.37322925701069676,
+      "grad_norm": 1.126265249898816,
+      "learning_rate": 1.40628955581312e-05,
+      "loss": 1.6933,
+      "step": 1291
+    },
+    {
+      "epoch": 0.3735183579069095,
+      "grad_norm": 1.0295263942219146,
+      "learning_rate": 1.4054510050545505e-05,
+      "loss": 1.587,
+      "step": 1292
+    },
+    {
+      "epoch": 0.3738074588031223,
+      "grad_norm": 1.1795604406769806,
+      "learning_rate": 1.4046121129690278e-05,
+      "loss": 1.6097,
+      "step": 1293
+    },
+    {
+      "epoch": 0.3740965596993351,
+      "grad_norm": 1.1403420688505401,
+      "learning_rate": 1.4037728802627693e-05,
+      "loss": 1.6267,
+      "step": 1294
+    },
+    {
+      "epoch": 0.37438566059554784,
+      "grad_norm": 1.0993080419061012,
+      "learning_rate": 1.4029333076422782e-05,
+      "loss": 1.671,
+      "step": 1295
+    },
+    {
+      "epoch": 0.3746747614917606,
+      "grad_norm": 1.1780757010441563,
+      "learning_rate": 1.4020933958143455e-05,
+      "loss": 1.7137,
+      "step": 1296
+    },
+    {
+      "epoch": 0.3749638623879734,
+      "grad_norm": 1.1243870924066819,
+      "learning_rate": 1.4012531454860464e-05,
+      "loss": 1.7021,
+      "step": 1297
+    },
+    {
+      "epoch": 0.37525296328418617,
+      "grad_norm": 1.0999261176344182,
+      "learning_rate": 1.4004125573647414e-05,
+      "loss": 1.5531,
+      "step": 1298
+    },
+    {
+      "epoch": 0.375542064180399,
+      "grad_norm": 1.1292673013365175,
+      "learning_rate": 1.399571632158076e-05,
+      "loss": 1.5297,
+      "step": 1299
+    },
+    {
+      "epoch": 0.37583116507661174,
+      "grad_norm": 1.0839707027340657,
+      "learning_rate": 1.3987303705739787e-05,
+      "loss": 1.6223,
+      "step": 1300
+    },
+    {
+      "epoch": 0.3761202659728245,
+      "grad_norm": 1.1689455340590824,
+      "learning_rate": 1.3978887733206614e-05,
+      "loss": 1.602,
+      "step": 1301
+    },
+    {
+      "epoch": 0.3764093668690373,
+      "grad_norm": 1.037131778360143,
+      "learning_rate": 1.3970468411066187e-05,
+      "loss": 1.5195,
+      "step": 1302
+    },
+    {
+      "epoch": 0.37669846776525007,
+      "grad_norm": 1.099003065082928,
+      "learning_rate": 1.3962045746406273e-05,
+      "loss": 1.5807,
+      "step": 1303
+    },
+    {
+      "epoch": 0.3769875686614628,
+      "grad_norm": 1.1281144891295398,
+      "learning_rate": 1.395361974631745e-05,
+      "loss": 1.6533,
+      "step": 1304
+    },
+    {
+      "epoch": 0.37727666955767564,
+      "grad_norm": 1.1355144650861142,
+      "learning_rate": 1.3945190417893106e-05,
+      "loss": 1.5925,
+      "step": 1305
+    },
+    {
+      "epoch": 0.3775657704538884,
+      "grad_norm": 1.095859210884497,
+      "learning_rate": 1.3936757768229428e-05,
+      "loss": 1.6053,
+      "step": 1306
+    },
+    {
+      "epoch": 0.3778548713501012,
+      "grad_norm": 1.1282184354527713,
+      "learning_rate": 1.39283218044254e-05,
+      "loss": 1.6052,
+      "step": 1307
+    },
+    {
+      "epoch": 0.37814397224631396,
+      "grad_norm": 1.0831491499806725,
+      "learning_rate": 1.3919882533582804e-05,
+      "loss": 1.5087,
+      "step": 1308
+    },
+    {
+      "epoch": 0.3784330731425267,
+      "grad_norm": 1.116233937615999,
+      "learning_rate": 1.3911439962806192e-05,
+      "loss": 1.5589,
+      "step": 1309
+    },
+    {
+      "epoch": 0.37872217403873953,
+      "grad_norm": 1.0633282902202397,
+      "learning_rate": 1.3902994099202904e-05,
+      "loss": 1.6164,
+      "step": 1310
+    },
+    {
+      "epoch": 0.3790112749349523,
+      "grad_norm": 1.0952692825870283,
+      "learning_rate": 1.3894544949883046e-05,
+      "loss": 1.7074,
+      "step": 1311
+    },
+    {
+      "epoch": 0.3793003758311651,
+      "grad_norm": 1.0824385592753576,
+      "learning_rate": 1.3886092521959497e-05,
+      "loss": 1.5499,
+      "step": 1312
+    },
+    {
+      "epoch": 0.37958947672737786,
+      "grad_norm": 1.0480969133383606,
+      "learning_rate": 1.3877636822547894e-05,
+      "loss": 1.6255,
+      "step": 1313
+    },
+    {
+      "epoch": 0.3798785776235906,
+      "grad_norm": 1.1345684718964264,
+      "learning_rate": 1.386917785876662e-05,
+      "loss": 1.6102,
+      "step": 1314
+    },
+    {
+      "epoch": 0.38016767851980343,
+      "grad_norm": 1.1080824270324598,
+      "learning_rate": 1.3860715637736817e-05,
+      "loss": 1.5263,
+      "step": 1315
+    },
+    {
+      "epoch": 0.3804567794160162,
+      "grad_norm": 1.059203635427344,
+      "learning_rate": 1.3852250166582364e-05,
+      "loss": 1.6089,
+      "step": 1316
+    },
+    {
+      "epoch": 0.38074588031222895,
+      "grad_norm": 1.087510027682475,
+      "learning_rate": 1.3843781452429875e-05,
+      "loss": 1.541,
+      "step": 1317
+    },
+    {
+      "epoch": 0.38103498120844176,
+      "grad_norm": 1.086247443147945,
+      "learning_rate": 1.3835309502408696e-05,
+      "loss": 1.632,
+      "step": 1318
+    },
+    {
+      "epoch": 0.3813240821046545,
+      "grad_norm": 1.1441906470265728,
+      "learning_rate": 1.3826834323650899e-05,
+      "loss": 1.6891,
+      "step": 1319
+    },
+    {
+      "epoch": 0.38161318300086733,
+      "grad_norm": 1.1166198316687277,
+      "learning_rate": 1.381835592329127e-05,
+      "loss": 1.5583,
+      "step": 1320
+    },
+    {
+      "epoch": 0.3819022838970801,
+      "grad_norm": 1.0406271039215775,
+      "learning_rate": 1.380987430846731e-05,
+      "loss": 1.5878,
+      "step": 1321
+    },
+    {
+      "epoch": 0.38219138479329284,
+      "grad_norm": 1.1782259448538297,
+      "learning_rate": 1.3801389486319223e-05,
+      "loss": 1.5041,
+      "step": 1322
+    },
+    {
+      "epoch": 0.38248048568950566,
+      "grad_norm": 1.119306255364395,
+      "learning_rate": 1.3792901463989918e-05,
+      "loss": 1.668,
+      "step": 1323
+    },
+    {
+      "epoch": 0.3827695865857184,
+      "grad_norm": 1.0695978855808004,
+      "learning_rate": 1.3784410248624996e-05,
+      "loss": 1.6487,
+      "step": 1324
+    },
+    {
+      "epoch": 0.38305868748193117,
+      "grad_norm": 1.0616454533037485,
+      "learning_rate": 1.377591584737274e-05,
+      "loss": 1.5401,
+      "step": 1325
+    },
+    {
+      "epoch": 0.383347788378144,
+      "grad_norm": 1.0774865869206818,
+      "learning_rate": 1.3767418267384128e-05,
+      "loss": 1.6066,
+      "step": 1326
+    },
+    {
+      "epoch": 0.38363688927435674,
+      "grad_norm": 1.060694994399504,
+      "learning_rate": 1.3758917515812798e-05,
+      "loss": 1.6073,
+      "step": 1327
+    },
+    {
+      "epoch": 0.38392599017056955,
+      "grad_norm": 1.0903641939801798,
+      "learning_rate": 1.3750413599815074e-05,
+      "loss": 1.6114,
+      "step": 1328
+    },
+    {
+      "epoch": 0.3842150910667823,
+      "grad_norm": 1.114947234018338,
+      "learning_rate": 1.3741906526549928e-05,
+      "loss": 1.6328,
+      "step": 1329
+    },
+    {
+      "epoch": 0.38450419196299507,
+      "grad_norm": 1.1067463745265371,
+      "learning_rate": 1.3733396303179005e-05,
+      "loss": 1.6358,
+      "step": 1330
+    },
+    {
+      "epoch": 0.3847932928592079,
+      "grad_norm": 1.1577228624063416,
+      "learning_rate": 1.3724882936866596e-05,
+      "loss": 1.5758,
+      "step": 1331
+    },
+    {
+      "epoch": 0.38508239375542064,
+      "grad_norm": 1.1558637576676867,
+      "learning_rate": 1.3716366434779628e-05,
+      "loss": 1.572,
+      "step": 1332
+    },
+    {
+      "epoch": 0.3853714946516334,
+      "grad_norm": 1.1143169382091267,
+      "learning_rate": 1.3707846804087686e-05,
+      "loss": 1.6113,
+      "step": 1333
+    },
+    {
+      "epoch": 0.3856605955478462,
+      "grad_norm": 1.130544176598161,
+      "learning_rate": 1.3699324051962974e-05,
+      "loss": 1.5409,
+      "step": 1334
+    },
+    {
+      "epoch": 0.38594969644405897,
+      "grad_norm": 1.0366446457013492,
+      "learning_rate": 1.3690798185580337e-05,
+      "loss": 1.5622,
+      "step": 1335
+    },
+    {
+      "epoch": 0.3862387973402718,
+      "grad_norm": 1.1288006310703098,
+      "learning_rate": 1.3682269212117223e-05,
+      "loss": 1.6841,
+      "step": 1336
+    },
+    {
+      "epoch": 0.38652789823648454,
+      "grad_norm": 1.1063560145710902,
+      "learning_rate": 1.3673737138753713e-05,
+      "loss": 1.6939,
+      "step": 1337
+    },
+    {
+      "epoch": 0.3868169991326973,
+      "grad_norm": 1.1459823915842076,
+      "learning_rate": 1.3665201972672492e-05,
+      "loss": 1.6378,
+      "step": 1338
+    },
+    {
+      "epoch": 0.3871061000289101,
+      "grad_norm": 1.1039738007856277,
+      "learning_rate": 1.3656663721058844e-05,
+      "loss": 1.5349,
+      "step": 1339
+    },
+    {
+      "epoch": 0.38739520092512286,
+      "grad_norm": 1.1237038119054343,
+      "learning_rate": 1.3648122391100663e-05,
+      "loss": 1.5913,
+      "step": 1340
+    },
+    {
+      "epoch": 0.3876843018213356,
+      "grad_norm": 1.0705587484764698,
+      "learning_rate": 1.3639577989988417e-05,
+      "loss": 1.5015,
+      "step": 1341
+    },
+    {
+      "epoch": 0.38797340271754843,
+      "grad_norm": 1.1425898222093245,
+      "learning_rate": 1.3631030524915171e-05,
+      "loss": 1.638,
+      "step": 1342
+    },
+    {
+      "epoch": 0.3882625036137612,
+      "grad_norm": 1.138358366597531,
+      "learning_rate": 1.362248000307657e-05,
+      "loss": 1.5921,
+      "step": 1343
+    },
+    {
+      "epoch": 0.388551604509974,
+      "grad_norm": 1.0937645241618539,
+      "learning_rate": 1.3613926431670824e-05,
+      "loss": 1.479,
+      "step": 1344
+    },
+    {
+      "epoch": 0.38884070540618676,
+      "grad_norm": 1.113613288350207,
+      "learning_rate": 1.360536981789872e-05,
+      "loss": 1.618,
+      "step": 1345
+    },
+    {
+      "epoch": 0.3891298063023995,
+      "grad_norm": 1.1367651938150571,
+      "learning_rate": 1.3596810168963599e-05,
+      "loss": 1.5512,
+      "step": 1346
+    },
+    {
+      "epoch": 0.38941890719861233,
+      "grad_norm": 1.079517622534508,
+      "learning_rate": 1.358824749207136e-05,
+      "loss": 1.5395,
+      "step": 1347
+    },
+    {
+      "epoch": 0.3897080080948251,
+      "grad_norm": 1.143875281810773,
+      "learning_rate": 1.3579681794430454e-05,
+      "loss": 1.6185,
+      "step": 1348
+    },
+    {
+      "epoch": 0.3899971089910379,
+      "grad_norm": 1.1732842322479438,
+      "learning_rate": 1.3571113083251865e-05,
+      "loss": 1.6675,
+      "step": 1349
+    },
+    {
+      "epoch": 0.39028620988725066,
+      "grad_norm": 1.1246289227681265,
+      "learning_rate": 1.356254136574913e-05,
+      "loss": 1.6558,
+      "step": 1350
+    },
+    {
+      "epoch": 0.3905753107834634,
+      "grad_norm": 1.165298085541357,
+      "learning_rate": 1.3553966649138297e-05,
+      "loss": 1.5468,
+      "step": 1351
+    },
+    {
+      "epoch": 0.3908644116796762,
+      "grad_norm": 1.1344138977255307,
+      "learning_rate": 1.3545388940637958e-05,
+      "loss": 1.5932,
+      "step": 1352
+    },
+    {
+      "epoch": 0.391153512575889,
+      "grad_norm": 1.0690031880578144,
+      "learning_rate": 1.3536808247469213e-05,
+      "loss": 1.5697,
+      "step": 1353
+    },
+    {
+      "epoch": 0.39144261347210174,
+      "grad_norm": 1.0975749907120596,
+      "learning_rate": 1.352822457685567e-05,
+      "loss": 1.6245,
+      "step": 1354
+    },
+    {
+      "epoch": 0.39173171436831455,
+      "grad_norm": 1.0927297214562863,
+      "learning_rate": 1.3519637936023462e-05,
+      "loss": 1.6026,
+      "step": 1355
+    },
+    {
+      "epoch": 0.3920208152645273,
+      "grad_norm": 1.0537160535977657,
+      "learning_rate": 1.3511048332201204e-05,
+      "loss": 1.4856,
+      "step": 1356
+    },
+    {
+      "epoch": 0.3923099161607401,
+      "grad_norm": 1.0920768107193122,
+      "learning_rate": 1.3502455772620012e-05,
+      "loss": 1.4882,
+      "step": 1357
+    },
+    {
+      "epoch": 0.3925990170569529,
+      "grad_norm": 1.0757741953402,
+      "learning_rate": 1.3493860264513492e-05,
+      "loss": 1.6938,
+      "step": 1358
+    },
+    {
+      "epoch": 0.39288811795316564,
+      "grad_norm": 1.1184693762965427,
+      "learning_rate": 1.348526181511773e-05,
+      "loss": 1.6371,
+      "step": 1359
+    },
+    {
+      "epoch": 0.39317721884937845,
+      "grad_norm": 1.0887972408619184,
+      "learning_rate": 1.3476660431671288e-05,
+      "loss": 1.5472,
+      "step": 1360
+    },
+    {
+      "epoch": 0.3934663197455912,
+      "grad_norm": 1.1259522274979648,
+      "learning_rate": 1.34680561214152e-05,
+      "loss": 1.5498,
+      "step": 1361
+    },
+    {
+      "epoch": 0.39375542064180397,
+      "grad_norm": 1.1133474449428702,
+      "learning_rate": 1.3459448891592962e-05,
+      "loss": 1.5723,
+      "step": 1362
+    },
+    {
+      "epoch": 0.3940445215380168,
+      "grad_norm": 1.1499692550396208,
+      "learning_rate": 1.3450838749450529e-05,
+      "loss": 1.484,
+      "step": 1363
+    },
+    {
+      "epoch": 0.39433362243422954,
+      "grad_norm": 1.0715825700296768,
+      "learning_rate": 1.3442225702236307e-05,
+      "loss": 1.5612,
+      "step": 1364
+    },
+    {
+      "epoch": 0.39462272333044235,
+      "grad_norm": 1.077029585331996,
+      "learning_rate": 1.3433609757201147e-05,
+      "loss": 1.5889,
+      "step": 1365
+    },
+    {
+      "epoch": 0.3949118242266551,
+      "grad_norm": 1.1501089531692315,
+      "learning_rate": 1.342499092159834e-05,
+      "loss": 1.4946,
+      "step": 1366
+    },
+    {
+      "epoch": 0.39520092512286786,
+      "grad_norm": 1.096740007082763,
+      "learning_rate": 1.3416369202683615e-05,
+      "loss": 1.5497,
+      "step": 1367
+    },
+    {
+      "epoch": 0.3954900260190807,
+      "grad_norm": 1.0987454926481957,
+      "learning_rate": 1.3407744607715119e-05,
+      "loss": 1.5969,
+      "step": 1368
+    },
+    {
+      "epoch": 0.39577912691529343,
+      "grad_norm": 1.060961357189651,
+      "learning_rate": 1.3399117143953426e-05,
+      "loss": 1.5441,
+      "step": 1369
+    },
+    {
+      "epoch": 0.3960682278115062,
+      "grad_norm": 1.1032994325908025,
+      "learning_rate": 1.3390486818661526e-05,
+      "loss": 1.6306,
+      "step": 1370
+    },
+    {
+      "epoch": 0.396357328707719,
+      "grad_norm": 1.1047680776182227,
+      "learning_rate": 1.3381853639104817e-05,
+      "loss": 1.5242,
+      "step": 1371
+    },
+    {
+      "epoch": 0.39664642960393176,
+      "grad_norm": 1.1022023946061226,
+      "learning_rate": 1.33732176125511e-05,
+      "loss": 1.6319,
+      "step": 1372
+    },
+    {
+      "epoch": 0.3969355305001446,
+      "grad_norm": 1.0789182866424254,
+      "learning_rate": 1.3364578746270564e-05,
+      "loss": 1.6542,
+      "step": 1373
+    },
+    {
+      "epoch": 0.39722463139635733,
+      "grad_norm": 1.13836495763387,
+      "learning_rate": 1.3355937047535806e-05,
+      "loss": 1.5158,
+      "step": 1374
+    },
+    {
+      "epoch": 0.3975137322925701,
+      "grad_norm": 1.1151693144482542,
+      "learning_rate": 1.3347292523621795e-05,
+      "loss": 1.6513,
+      "step": 1375
+    },
+    {
+      "epoch": 0.3978028331887829,
+      "grad_norm": 1.0532473348160287,
+      "learning_rate": 1.3338645181805876e-05,
+      "loss": 1.5503,
+      "step": 1376
+    },
+    {
+      "epoch": 0.39809193408499566,
+      "grad_norm": 1.119386196223983,
+      "learning_rate": 1.3329995029367782e-05,
+      "loss": 1.5879,
+      "step": 1377
+    },
+    {
+      "epoch": 0.3983810349812084,
+      "grad_norm": 1.0973825255595782,
+      "learning_rate": 1.3321342073589595e-05,
+      "loss": 1.5115,
+      "step": 1378
+    },
+    {
+      "epoch": 0.39867013587742123,
+      "grad_norm": 1.0693127903553905,
+      "learning_rate": 1.331268632175576e-05,
+      "loss": 1.6588,
+      "step": 1379
+    },
+    {
+      "epoch": 0.398959236773634,
+      "grad_norm": 1.0823638073171304,
+      "learning_rate": 1.330402778115309e-05,
+      "loss": 1.589,
+      "step": 1380
+    },
+    {
+      "epoch": 0.3992483376698468,
+      "grad_norm": 1.0805775803416322,
+      "learning_rate": 1.3295366459070724e-05,
+      "loss": 1.5052,
+      "step": 1381
+    },
+    {
+      "epoch": 0.39953743856605956,
+      "grad_norm": 1.0443203230920821,
+      "learning_rate": 1.328670236280016e-05,
+      "loss": 1.4888,
+      "step": 1382
+    },
+    {
+      "epoch": 0.3998265394622723,
+      "grad_norm": 1.1511681114820926,
+      "learning_rate": 1.3278035499635226e-05,
+      "loss": 1.6123,
+      "step": 1383
+    },
+    {
+      "epoch": 0.4001156403584851,
+      "grad_norm": 1.0863214241630808,
+      "learning_rate": 1.3269365876872078e-05,
+      "loss": 1.5493,
+      "step": 1384
+    },
+    {
+      "epoch": 0.4004047412546979,
+      "grad_norm": 1.0838040505052533,
+      "learning_rate": 1.3260693501809193e-05,
+      "loss": 1.6153,
+      "step": 1385
+    },
+    {
+      "epoch": 0.4006938421509107,
+      "grad_norm": 1.1500580767514397,
+      "learning_rate": 1.3252018381747367e-05,
+      "loss": 1.5897,
+      "step": 1386
+    },
+    {
+      "epoch": 0.40098294304712345,
+      "grad_norm": 1.0845655171544064,
+      "learning_rate": 1.3243340523989714e-05,
+      "loss": 1.5783,
+      "step": 1387
+    },
+    {
+      "epoch": 0.4012720439433362,
+      "grad_norm": 1.1970928864429935,
+      "learning_rate": 1.3234659935841641e-05,
+      "loss": 1.6129,
+      "step": 1388
+    },
+    {
+      "epoch": 0.401561144839549,
+      "grad_norm": 1.101894564995156,
+      "learning_rate": 1.3225976624610856e-05,
+      "loss": 1.5264,
+      "step": 1389
+    },
+    {
+      "epoch": 0.4018502457357618,
+      "grad_norm": 1.1047369726724903,
+      "learning_rate": 1.3217290597607373e-05,
+      "loss": 1.6217,
+      "step": 1390
+    },
+    {
+      "epoch": 0.40213934663197454,
+      "grad_norm": 1.1482189531968385,
+      "learning_rate": 1.3208601862143468e-05,
+      "loss": 1.6808,
+      "step": 1391
+    },
+    {
+      "epoch": 0.40242844752818735,
+      "grad_norm": 1.1210248814188173,
+      "learning_rate": 1.3199910425533721e-05,
+      "loss": 1.5766,
+      "step": 1392
+    },
+    {
+      "epoch": 0.4027175484244001,
+      "grad_norm": 1.1685854469228043,
+      "learning_rate": 1.3191216295094971e-05,
+      "loss": 1.4749,
+      "step": 1393
+    },
+    {
+      "epoch": 0.4030066493206129,
+      "grad_norm": 1.0797246071856748,
+      "learning_rate": 1.318251947814633e-05,
+      "loss": 1.6377,
+      "step": 1394
+    },
+    {
+      "epoch": 0.4032957502168257,
+      "grad_norm": 1.0635773691483372,
+      "learning_rate": 1.3173819982009173e-05,
+      "loss": 1.4949,
+      "step": 1395
+    },
+    {
+      "epoch": 0.40358485111303843,
+      "grad_norm": 1.115897033244827,
+      "learning_rate": 1.3165117814007124e-05,
+      "loss": 1.6583,
+      "step": 1396
+    },
+    {
+      "epoch": 0.40387395200925125,
+      "grad_norm": 1.0989308444609356,
+      "learning_rate": 1.3156412981466064e-05,
+      "loss": 1.649,
+      "step": 1397
+    },
+    {
+      "epoch": 0.404163052905464,
+      "grad_norm": 1.0797697429688853,
+      "learning_rate": 1.3147705491714113e-05,
+      "loss": 1.5827,
+      "step": 1398
+    },
+    {
+      "epoch": 0.40445215380167676,
+      "grad_norm": 1.0834261852077045,
+      "learning_rate": 1.3138995352081634e-05,
+      "loss": 1.5503,
+      "step": 1399
+    },
+    {
+      "epoch": 0.4047412546978896,
+      "grad_norm": 1.0787814057307044,
+      "learning_rate": 1.3130282569901205e-05,
+      "loss": 1.5232,
+      "step": 1400
+    },
+    {
+      "epoch": 0.40503035559410233,
+      "grad_norm": 1.1461026515480623,
+      "learning_rate": 1.3121567152507646e-05,
+      "loss": 1.5824,
+      "step": 1401
+    },
+    {
+      "epoch": 0.40531945649031514,
+      "grad_norm": 1.0691604966712713,
+      "learning_rate": 1.3112849107237988e-05,
+      "loss": 1.6224,
+      "step": 1402
+    },
+    {
+      "epoch": 0.4056085573865279,
+      "grad_norm": 1.0987697223521924,
+      "learning_rate": 1.3104128441431473e-05,
+      "loss": 1.5653,
+      "step": 1403
+    },
+    {
+      "epoch": 0.40589765828274066,
+      "grad_norm": 1.1957843405058255,
+      "learning_rate": 1.3095405162429556e-05,
+      "loss": 1.6489,
+      "step": 1404
+    },
+    {
+      "epoch": 0.40618675917895347,
+      "grad_norm": 1.1012887872938155,
+      "learning_rate": 1.308667927757588e-05,
+      "loss": 1.4554,
+      "step": 1405
+    },
+    {
+      "epoch": 0.40647586007516623,
+      "grad_norm": 1.2503495892515055,
+      "learning_rate": 1.3077950794216295e-05,
+      "loss": 1.623,
+      "step": 1406
+    },
+    {
+      "epoch": 0.406764960971379,
+      "grad_norm": 1.8219906920012625,
+      "learning_rate": 1.3069219719698829e-05,
+      "loss": 1.6721,
+      "step": 1407
+    },
+    {
+      "epoch": 0.4070540618675918,
+      "grad_norm": 1.160323746535951,
+      "learning_rate": 1.306048606137369e-05,
+      "loss": 1.6361,
+      "step": 1408
+    },
+    {
+      "epoch": 0.40734316276380456,
+      "grad_norm": 1.1049438869652184,
+      "learning_rate": 1.3051749826593276e-05,
+      "loss": 1.6377,
+      "step": 1409
+    },
+    {
+      "epoch": 0.40763226366001737,
+      "grad_norm": 1.129823576698373,
+      "learning_rate": 1.3043011022712135e-05,
+      "loss": 1.5416,
+      "step": 1410
+    },
+    {
+      "epoch": 0.4079213645562301,
+      "grad_norm": 1.1208828506448014,
+      "learning_rate": 1.3034269657086993e-05,
+      "loss": 1.6168,
+      "step": 1411
+    },
+    {
+      "epoch": 0.4082104654524429,
+      "grad_norm": 1.0884563842602648,
+      "learning_rate": 1.3025525737076721e-05,
+      "loss": 1.6246,
+      "step": 1412
+    },
+    {
+      "epoch": 0.4084995663486557,
+      "grad_norm": 1.165052279127997,
+      "learning_rate": 1.3016779270042345e-05,
+      "loss": 1.5437,
+      "step": 1413
+    },
+    {
+      "epoch": 0.40878866724486845,
+      "grad_norm": 1.1628829699708714,
+      "learning_rate": 1.3008030263347043e-05,
+      "loss": 1.5522,
+      "step": 1414
+    },
+    {
+      "epoch": 0.4090777681410812,
+      "grad_norm": 1.1305635940899177,
+      "learning_rate": 1.2999278724356114e-05,
+      "loss": 1.5452,
+      "step": 1415
+    },
+    {
+      "epoch": 0.409366869037294,
+      "grad_norm": 1.1711599188563901,
+      "learning_rate": 1.2990524660437004e-05,
+      "loss": 1.6864,
+      "step": 1416
+    },
+    {
+      "epoch": 0.4096559699335068,
+      "grad_norm": 1.1580090440810888,
+      "learning_rate": 1.2981768078959282e-05,
+      "loss": 1.6319,
+      "step": 1417
+    },
+    {
+      "epoch": 0.4099450708297196,
+      "grad_norm": 1.0758152877920675,
+      "learning_rate": 1.2973008987294623e-05,
+      "loss": 1.6597,
+      "step": 1418
+    },
+    {
+      "epoch": 0.41023417172593235,
+      "grad_norm": 1.0542033973311213,
+      "learning_rate": 1.296424739281684e-05,
+      "loss": 1.5925,
+      "step": 1419
+    },
+    {
+      "epoch": 0.4105232726221451,
+      "grad_norm": 1.1134676224663176,
+      "learning_rate": 1.2955483302901826e-05,
+      "loss": 1.535,
+      "step": 1420
+    },
+    {
+      "epoch": 0.4108123735183579,
+      "grad_norm": 1.1138074722685556,
+      "learning_rate": 1.2946716724927591e-05,
+      "loss": 1.6683,
+      "step": 1421
+    },
+    {
+      "epoch": 0.4111014744145707,
+      "grad_norm": 1.1006812046260228,
+      "learning_rate": 1.293794766627424e-05,
+      "loss": 1.5808,
+      "step": 1422
+    },
+    {
+      "epoch": 0.4113905753107835,
+      "grad_norm": 1.0970946299864843,
+      "learning_rate": 1.292917613432396e-05,
+      "loss": 1.6019,
+      "step": 1423
+    },
+    {
+      "epoch": 0.41167967620699625,
+      "grad_norm": 1.056758252104355,
+      "learning_rate": 1.2920402136461019e-05,
+      "loss": 1.6119,
+      "step": 1424
+    },
+    {
+      "epoch": 0.411968777103209,
+      "grad_norm": 1.0707418940490092,
+      "learning_rate": 1.291162568007177e-05,
+      "loss": 1.6301,
+      "step": 1425
+    },
+    {
+      "epoch": 0.4122578779994218,
+      "grad_norm": 1.0603435006625261,
+      "learning_rate": 1.2902846772544625e-05,
+      "loss": 1.6555,
+      "step": 1426
+    },
+    {
+      "epoch": 0.4125469788956346,
+      "grad_norm": 1.1069343943255046,
+      "learning_rate": 1.289406542127007e-05,
+      "loss": 1.5739,
+      "step": 1427
+    },
+    {
+      "epoch": 0.41283607979184733,
+      "grad_norm": 1.163321808960078,
+      "learning_rate": 1.2885281633640636e-05,
+      "loss": 1.6046,
+      "step": 1428
+    },
+    {
+      "epoch": 0.41312518068806015,
+      "grad_norm": 1.1250069740600672,
+      "learning_rate": 1.2876495417050916e-05,
+      "loss": 1.6445,
+      "step": 1429
+    },
+    {
+      "epoch": 0.4134142815842729,
+      "grad_norm": 1.1031184767643845,
+      "learning_rate": 1.2867706778897544e-05,
+      "loss": 1.5502,
+      "step": 1430
+    },
+    {
+      "epoch": 0.4137033824804857,
+      "grad_norm": 1.0593709318010174,
+      "learning_rate": 1.2858915726579193e-05,
+      "loss": 1.6201,
+      "step": 1431
+    },
+    {
+      "epoch": 0.4139924833766985,
+      "grad_norm": 1.117978458594976,
+      "learning_rate": 1.2850122267496565e-05,
+      "loss": 1.5369,
+      "step": 1432
+    },
+    {
+      "epoch": 0.41428158427291123,
+      "grad_norm": 1.0813007664614265,
+      "learning_rate": 1.284132640905239e-05,
+      "loss": 1.5524,
+      "step": 1433
+    },
+    {
+      "epoch": 0.41457068516912404,
+      "grad_norm": 1.086633499163814,
+      "learning_rate": 1.2832528158651422e-05,
+      "loss": 1.6368,
+      "step": 1434
+    },
+    {
+      "epoch": 0.4148597860653368,
+      "grad_norm": 1.1150674277569264,
+      "learning_rate": 1.282372752370042e-05,
+      "loss": 1.6081,
+      "step": 1435
+    },
+    {
+      "epoch": 0.41514888696154956,
+      "grad_norm": 1.0932357843791627,
+      "learning_rate": 1.2814924511608165e-05,
+      "loss": 1.5927,
+      "step": 1436
+    },
+    {
+      "epoch": 0.41543798785776237,
+      "grad_norm": 1.1543836930682887,
+      "learning_rate": 1.280611912978542e-05,
+      "loss": 1.6581,
+      "step": 1437
+    },
+    {
+      "epoch": 0.4157270887539751,
+      "grad_norm": 1.0505935915592688,
+      "learning_rate": 1.2797311385644955e-05,
+      "loss": 1.5999,
+      "step": 1438
+    },
+    {
+      "epoch": 0.41601618965018794,
+      "grad_norm": 1.1050482535412185,
+      "learning_rate": 1.2788501286601533e-05,
+      "loss": 1.6392,
+      "step": 1439
+    },
+    {
+      "epoch": 0.4163052905464007,
+      "grad_norm": 1.1042656400088005,
+      "learning_rate": 1.2779688840071884e-05,
+      "loss": 1.6577,
+      "step": 1440
+    },
+    {
+      "epoch": 0.41659439144261345,
+      "grad_norm": 1.1720930062334785,
+      "learning_rate": 1.2770874053474729e-05,
+      "loss": 1.5184,
+      "step": 1441
+    },
+    {
+      "epoch": 0.41688349233882627,
+      "grad_norm": 1.0715728062394607,
+      "learning_rate": 1.2762056934230751e-05,
+      "loss": 1.6084,
+      "step": 1442
+    },
+    {
+      "epoch": 0.417172593235039,
+      "grad_norm": 1.1096413637769356,
+      "learning_rate": 1.27532374897626e-05,
+      "loss": 1.5706,
+      "step": 1443
+    },
+    {
+      "epoch": 0.4174616941312518,
+      "grad_norm": 1.1034545699709775,
+      "learning_rate": 1.2744415727494883e-05,
+      "loss": 1.5417,
+      "step": 1444
+    },
+    {
+      "epoch": 0.4177507950274646,
+      "grad_norm": 1.117597930092238,
+      "learning_rate": 1.273559165485415e-05,
+      "loss": 1.6484,
+      "step": 1445
+    },
+    {
+      "epoch": 0.41803989592367735,
+      "grad_norm": 1.052774365499576,
+      "learning_rate": 1.2726765279268918e-05,
+      "loss": 1.4995,
+      "step": 1446
+    },
+    {
+      "epoch": 0.41832899681989016,
+      "grad_norm": 1.0616106111534722,
+      "learning_rate": 1.2717936608169616e-05,
+      "loss": 1.5902,
+      "step": 1447
+    },
+    {
+      "epoch": 0.4186180977161029,
+      "grad_norm": 1.15892114485715,
+      "learning_rate": 1.2709105648988621e-05,
+      "loss": 1.6165,
+      "step": 1448
+    },
+    {
+      "epoch": 0.4189071986123157,
+      "grad_norm": 1.073981070440038,
+      "learning_rate": 1.2700272409160239e-05,
+      "loss": 1.6253,
+      "step": 1449
+    },
+    {
+      "epoch": 0.4191962995085285,
+      "grad_norm": 1.0288115268099387,
+      "learning_rate": 1.2691436896120679e-05,
+      "loss": 1.5277,
+      "step": 1450
+    },
+    {
+      "epoch": 0.41948540040474125,
+      "grad_norm": 1.0775586542877098,
+      "learning_rate": 1.2682599117308083e-05,
+      "loss": 1.5199,
+      "step": 1451
+    },
+    {
+      "epoch": 0.419774501300954,
+      "grad_norm": 1.1234512281897384,
+      "learning_rate": 1.2673759080162488e-05,
+      "loss": 1.6388,
+      "step": 1452
+    },
+    {
+      "epoch": 0.4200636021971668,
+      "grad_norm": 1.0771902747466777,
+      "learning_rate": 1.2664916792125839e-05,
+      "loss": 1.7336,
+      "step": 1453
+    },
+    {
+      "epoch": 0.4203527030933796,
+      "grad_norm": 1.0730469726282257,
+      "learning_rate": 1.2656072260641971e-05,
+      "loss": 1.571,
+      "step": 1454
+    },
+    {
+      "epoch": 0.4206418039895924,
+      "grad_norm": 1.1069931610853274,
+      "learning_rate": 1.2647225493156612e-05,
+      "loss": 1.6053,
+      "step": 1455
+    },
+    {
+      "epoch": 0.42093090488580515,
+      "grad_norm": 1.119206942375949,
+      "learning_rate": 1.2638376497117365e-05,
+      "loss": 1.6341,
+      "step": 1456
+    },
+    {
+      "epoch": 0.4212200057820179,
+      "grad_norm": 1.0564767917365292,
+      "learning_rate": 1.262952527997372e-05,
+      "loss": 1.5695,
+      "step": 1457
+    },
+    {
+      "epoch": 0.4215091066782307,
+      "grad_norm": 1.2051190858764724,
+      "learning_rate": 1.2620671849177032e-05,
+      "loss": 1.6273,
+      "step": 1458
+    },
+    {
+      "epoch": 0.4217982075744435,
+      "grad_norm": 1.0807107576270414,
+      "learning_rate": 1.261181621218051e-05,
+      "loss": 1.5298,
+      "step": 1459
+    },
+    {
+      "epoch": 0.4220873084706563,
+      "grad_norm": 1.0384330371676334,
+      "learning_rate": 1.2602958376439237e-05,
+      "loss": 1.5346,
+      "step": 1460
+    },
+    {
+      "epoch": 0.42237640936686904,
+      "grad_norm": 1.1116407967162223,
+      "learning_rate": 1.2594098349410135e-05,
+      "loss": 1.6088,
+      "step": 1461
+    },
+    {
+      "epoch": 0.4226655102630818,
+      "grad_norm": 1.0704948684121718,
+      "learning_rate": 1.2585236138551977e-05,
+      "loss": 1.6557,
+      "step": 1462
+    },
+    {
+      "epoch": 0.4229546111592946,
+      "grad_norm": 1.0419470708792398,
+      "learning_rate": 1.2576371751325371e-05,
+      "loss": 1.5803,
+      "step": 1463
+    },
+    {
+      "epoch": 0.42324371205550737,
+      "grad_norm": 1.1619481908104672,
+      "learning_rate": 1.2567505195192756e-05,
+      "loss": 1.561,
+      "step": 1464
+    },
+    {
+      "epoch": 0.42353281295172013,
+      "grad_norm": 1.0516826278608193,
+      "learning_rate": 1.2558636477618401e-05,
+      "loss": 1.4679,
+      "step": 1465
+    },
+    {
+      "epoch": 0.42382191384793294,
+      "grad_norm": 1.0656875867110747,
+      "learning_rate": 1.2549765606068397e-05,
+      "loss": 1.6001,
+      "step": 1466
+    },
+    {
+      "epoch": 0.4241110147441457,
+      "grad_norm": 1.047384742603988,
+      "learning_rate": 1.2540892588010633e-05,
+      "loss": 1.5896,
+      "step": 1467
+    },
+    {
+      "epoch": 0.4244001156403585,
+      "grad_norm": 1.02455606013114,
+      "learning_rate": 1.253201743091483e-05,
+      "loss": 1.5398,
+      "step": 1468
+    },
+    {
+      "epoch": 0.42468921653657127,
+      "grad_norm": 1.0484608793173646,
+      "learning_rate": 1.2523140142252482e-05,
+      "loss": 1.576,
+      "step": 1469
+    },
+    {
+      "epoch": 0.424978317432784,
+      "grad_norm": 1.044013000417675,
+      "learning_rate": 1.2514260729496899e-05,
+      "loss": 1.5462,
+      "step": 1470
+    },
+    {
+      "epoch": 0.42526741832899684,
+      "grad_norm": 1.1104946112450396,
+      "learning_rate": 1.2505379200123174e-05,
+      "loss": 1.5837,
+      "step": 1471
+    },
+    {
+      "epoch": 0.4255565192252096,
+      "grad_norm": 1.0483681936376463,
+      "learning_rate": 1.2496495561608168e-05,
+      "loss": 1.6044,
+      "step": 1472
+    },
+    {
+      "epoch": 0.42584562012142235,
+      "grad_norm": 1.0922289548826742,
+      "learning_rate": 1.2487609821430545e-05,
+      "loss": 1.6107,
+      "step": 1473
+    },
+    {
+      "epoch": 0.42613472101763517,
+      "grad_norm": 1.1185333876459898,
+      "learning_rate": 1.2478721987070708e-05,
+      "loss": 1.5502,
+      "step": 1474
+    },
+    {
+      "epoch": 0.4264238219138479,
+      "grad_norm": 1.0615836651946364,
+      "learning_rate": 1.2469832066010843e-05,
+      "loss": 1.6939,
+      "step": 1475
+    },
+    {
+      "epoch": 0.42671292281006074,
+      "grad_norm": 1.0514960710260826,
+      "learning_rate": 1.2460940065734887e-05,
+      "loss": 1.6892,
+      "step": 1476
+    },
+    {
+      "epoch": 0.4270020237062735,
+      "grad_norm": 1.0403687405148012,
+      "learning_rate": 1.2452045993728524e-05,
+      "loss": 1.4727,
+      "step": 1477
+    },
+    {
+      "epoch": 0.42729112460248625,
+      "grad_norm": 1.0708368297716742,
+      "learning_rate": 1.2443149857479188e-05,
+      "loss": 1.5973,
+      "step": 1478
+    },
+    {
+      "epoch": 0.42758022549869906,
+      "grad_norm": 1.0470771269092312,
+      "learning_rate": 1.2434251664476049e-05,
+      "loss": 1.5966,
+      "step": 1479
+    },
+    {
+      "epoch": 0.4278693263949118,
+      "grad_norm": 1.0901772826316531,
+      "learning_rate": 1.2425351422210004e-05,
+      "loss": 1.6223,
+      "step": 1480
+    },
+    {
+      "epoch": 0.4281584272911246,
+      "grad_norm": 1.1708156048786722,
+      "learning_rate": 1.241644913817368e-05,
+      "loss": 1.6334,
+      "step": 1481
+    },
+    {
+      "epoch": 0.4284475281873374,
+      "grad_norm": 1.077107996211494,
+      "learning_rate": 1.2407544819861423e-05,
+      "loss": 1.5245,
+      "step": 1482
+    },
+    {
+      "epoch": 0.42873662908355015,
+      "grad_norm": 1.0861799485708443,
+      "learning_rate": 1.239863847476929e-05,
+      "loss": 1.6777,
+      "step": 1483
+    },
+    {
+      "epoch": 0.42902572997976296,
+      "grad_norm": 1.0772294705942125,
+      "learning_rate": 1.2389730110395037e-05,
+      "loss": 1.5956,
+      "step": 1484
+    },
+    {
+      "epoch": 0.4293148308759757,
+      "grad_norm": 1.0741128234042523,
+      "learning_rate": 1.2380819734238139e-05,
+      "loss": 1.7122,
+      "step": 1485
+    },
+    {
+      "epoch": 0.4296039317721885,
+      "grad_norm": 1.090387270682844,
+      "learning_rate": 1.2371907353799746e-05,
+      "loss": 1.5043,
+      "step": 1486
+    },
+    {
+      "epoch": 0.4298930326684013,
+      "grad_norm": 1.0854173206181026,
+      "learning_rate": 1.23629929765827e-05,
+      "loss": 1.5586,
+      "step": 1487
+    },
+    {
+      "epoch": 0.43018213356461404,
+      "grad_norm": 1.1418669006309328,
+      "learning_rate": 1.235407661009153e-05,
+      "loss": 1.5505,
+      "step": 1488
+    },
+    {
+      "epoch": 0.4304712344608268,
+      "grad_norm": 1.102426065770944,
+      "learning_rate": 1.2345158261832434e-05,
+      "loss": 1.6935,
+      "step": 1489
+    },
+    {
+      "epoch": 0.4307603353570396,
+      "grad_norm": 1.0652935971586843,
+      "learning_rate": 1.2336237939313286e-05,
+      "loss": 1.6352,
+      "step": 1490
+    },
+    {
+      "epoch": 0.43104943625325237,
+      "grad_norm": 1.0261532618112403,
+      "learning_rate": 1.2327315650043605e-05,
+      "loss": 1.6559,
+      "step": 1491
+    },
+    {
+      "epoch": 0.4313385371494652,
+      "grad_norm": 1.0947736897259985,
+      "learning_rate": 1.2318391401534582e-05,
+      "loss": 1.6539,
+      "step": 1492
+    },
+    {
+      "epoch": 0.43162763804567794,
+      "grad_norm": 1.0629425110661714,
+      "learning_rate": 1.2309465201299056e-05,
+      "loss": 1.5578,
+      "step": 1493
+    },
+    {
+      "epoch": 0.4319167389418907,
+      "grad_norm": 1.0415136895327566,
+      "learning_rate": 1.2300537056851498e-05,
+      "loss": 1.5458,
+      "step": 1494
+    },
+    {
+      "epoch": 0.4322058398381035,
+      "grad_norm": 1.0466462987324217,
+      "learning_rate": 1.229160697570803e-05,
+      "loss": 1.5221,
+      "step": 1495
+    },
+    {
+      "epoch": 0.43249494073431627,
+      "grad_norm": 1.0602361418832027,
+      "learning_rate": 1.2282674965386393e-05,
+      "loss": 1.6181,
+      "step": 1496
+    },
+    {
+      "epoch": 0.4327840416305291,
+      "grad_norm": 1.0744486154949497,
+      "learning_rate": 1.2273741033405957e-05,
+      "loss": 1.5255,
+      "step": 1497
+    },
+    {
+      "epoch": 0.43307314252674184,
+      "grad_norm": 1.0630727798369484,
+      "learning_rate": 1.2264805187287711e-05,
+      "loss": 1.6328,
+      "step": 1498
+    },
+    {
+      "epoch": 0.4333622434229546,
+      "grad_norm": 1.0430349528228378,
+      "learning_rate": 1.2255867434554247e-05,
+      "loss": 1.581,
+      "step": 1499
+    },
+    {
+      "epoch": 0.4336513443191674,
+      "grad_norm": 1.053209608808372,
+      "learning_rate": 1.2246927782729779e-05,
+      "loss": 1.5736,
+      "step": 1500
+    },
+    {
+      "epoch": 0.43394044521538017,
+      "grad_norm": 1.0519674787190658,
+      "learning_rate": 1.2237986239340103e-05,
+      "loss": 1.4759,
+      "step": 1501
+    },
+    {
+      "epoch": 0.4342295461115929,
+      "grad_norm": 1.0850086283201288,
+      "learning_rate": 1.2229042811912612e-05,
+      "loss": 1.5456,
+      "step": 1502
+    },
+    {
+      "epoch": 0.43451864700780574,
+      "grad_norm": 1.0443660257904965,
+      "learning_rate": 1.2220097507976292e-05,
+      "loss": 1.6928,
+      "step": 1503
+    },
+    {
+      "epoch": 0.4348077479040185,
+      "grad_norm": 1.0854953744136557,
+      "learning_rate": 1.2211150335061696e-05,
+      "loss": 1.6119,
+      "step": 1504
+    },
+    {
+      "epoch": 0.4350968488002313,
+      "grad_norm": 1.0530061842078442,
+      "learning_rate": 1.2202201300700971e-05,
+      "loss": 1.6598,
+      "step": 1505
+    },
+    {
+      "epoch": 0.43538594969644406,
+      "grad_norm": 1.0692975442013706,
+      "learning_rate": 1.2193250412427803e-05,
+      "loss": 1.656,
+      "step": 1506
+    },
+    {
+      "epoch": 0.4356750505926568,
+      "grad_norm": 1.068620318121473,
+      "learning_rate": 1.2184297677777463e-05,
+      "loss": 1.5694,
+      "step": 1507
+    },
+    {
+      "epoch": 0.43596415148886963,
+      "grad_norm": 1.141527968294171,
+      "learning_rate": 1.2175343104286767e-05,
+      "loss": 1.5666,
+      "step": 1508
+    },
+    {
+      "epoch": 0.4362532523850824,
+      "grad_norm": 1.0351517063749145,
+      "learning_rate": 1.2166386699494073e-05,
+      "loss": 1.6804,
+      "step": 1509
+    },
+    {
+      "epoch": 0.43654235328129515,
+      "grad_norm": 1.0346004411813414,
+      "learning_rate": 1.2157428470939293e-05,
+      "loss": 1.4791,
+      "step": 1510
+    },
+    {
+      "epoch": 0.43683145417750796,
+      "grad_norm": 1.050743573711537,
+      "learning_rate": 1.2148468426163865e-05,
+      "loss": 1.6977,
+      "step": 1511
+    },
+    {
+      "epoch": 0.4371205550737207,
+      "grad_norm": 1.0962825504730682,
+      "learning_rate": 1.2139506572710757e-05,
+      "loss": 1.6005,
+      "step": 1512
+    },
+    {
+      "epoch": 0.43740965596993353,
+      "grad_norm": 1.0623153529126912,
+      "learning_rate": 1.2130542918124468e-05,
+      "loss": 1.6745,
+      "step": 1513
+    },
+    {
+      "epoch": 0.4376987568661463,
+      "grad_norm": 1.1166027202760946,
+      "learning_rate": 1.2121577469951002e-05,
+      "loss": 1.6193,
+      "step": 1514
+    },
+    {
+      "epoch": 0.43798785776235905,
+      "grad_norm": 1.0646497920960403,
+      "learning_rate": 1.2112610235737879e-05,
+      "loss": 1.5542,
+      "step": 1515
+    },
+    {
+      "epoch": 0.43827695865857186,
+      "grad_norm": 1.0690756772994965,
+      "learning_rate": 1.210364122303412e-05,
+      "loss": 1.5353,
+      "step": 1516
+    },
+    {
+      "epoch": 0.4385660595547846,
+      "grad_norm": 1.0384977840300798,
+      "learning_rate": 1.2094670439390246e-05,
+      "loss": 1.5782,
+      "step": 1517
+    },
+    {
+      "epoch": 0.4388551604509974,
+      "grad_norm": 1.0870315782690028,
+      "learning_rate": 1.208569789235827e-05,
+      "loss": 1.5273,
+      "step": 1518
+    },
+    {
+      "epoch": 0.4391442613472102,
+      "grad_norm": 1.04371396962298,
+      "learning_rate": 1.207672358949168e-05,
+      "loss": 1.6347,
+      "step": 1519
+    },
+    {
+      "epoch": 0.43943336224342294,
+      "grad_norm": 1.1521176377861757,
+      "learning_rate": 1.2067747538345457e-05,
+      "loss": 1.6074,
+      "step": 1520
+    },
+    {
+      "epoch": 0.43972246313963576,
+      "grad_norm": 1.1220399345751644,
+      "learning_rate": 1.2058769746476042e-05,
+      "loss": 1.5763,
+      "step": 1521
+    },
+    {
+      "epoch": 0.4400115640358485,
+      "grad_norm": 1.1221946688637574,
+      "learning_rate": 1.2049790221441348e-05,
+      "loss": 1.5683,
+      "step": 1522
+    },
+    {
+      "epoch": 0.44030066493206127,
+      "grad_norm": 1.1015672560776733,
+      "learning_rate": 1.2040808970800741e-05,
+      "loss": 1.711,
+      "step": 1523
+    },
+    {
+      "epoch": 0.4405897658282741,
+      "grad_norm": 1.0707174556673726,
+      "learning_rate": 1.2031826002115046e-05,
+      "loss": 1.5592,
+      "step": 1524
+    },
+    {
+      "epoch": 0.44087886672448684,
+      "grad_norm": 1.0387254594897035,
+      "learning_rate": 1.2022841322946527e-05,
+      "loss": 1.6025,
+      "step": 1525
+    },
+    {
+      "epoch": 0.4411679676206996,
+      "grad_norm": 1.0127412572061887,
+      "learning_rate": 1.2013854940858896e-05,
+      "loss": 1.5284,
+      "step": 1526
+    },
+    {
+      "epoch": 0.4414570685169124,
+      "grad_norm": 1.0833551950366158,
+      "learning_rate": 1.2004866863417297e-05,
+      "loss": 1.5966,
+      "step": 1527
+    },
+    {
+      "epoch": 0.44174616941312517,
+      "grad_norm": 1.0753089166033192,
+      "learning_rate": 1.1995877098188293e-05,
+      "loss": 1.5485,
+      "step": 1528
+    },
+    {
+      "epoch": 0.442035270309338,
+      "grad_norm": 1.0611449887541602,
+      "learning_rate": 1.1986885652739879e-05,
+      "loss": 1.6244,
+      "step": 1529
+    },
+    {
+      "epoch": 0.44232437120555074,
+      "grad_norm": 1.0850293339458401,
+      "learning_rate": 1.197789253464146e-05,
+      "loss": 1.5783,
+      "step": 1530
+    },
+    {
+      "epoch": 0.4426134721017635,
+      "grad_norm": 1.6872765547722357,
+      "learning_rate": 1.1968897751463841e-05,
+      "loss": 1.574,
+      "step": 1531
+    },
+    {
+      "epoch": 0.4429025729979763,
+      "grad_norm": 1.0616045199694535,
+      "learning_rate": 1.195990131077925e-05,
+      "loss": 1.6466,
+      "step": 1532
+    },
+    {
+      "epoch": 0.44319167389418906,
+      "grad_norm": 1.065948095505724,
+      "learning_rate": 1.1950903220161286e-05,
+      "loss": 1.5698,
+      "step": 1533
+    },
+    {
+      "epoch": 0.4434807747904019,
+      "grad_norm": 1.09263405795792,
+      "learning_rate": 1.1941903487184947e-05,
+      "loss": 1.5391,
+      "step": 1534
+    },
+    {
+      "epoch": 0.44376987568661463,
+      "grad_norm": 1.1624092720055494,
+      "learning_rate": 1.1932902119426624e-05,
+      "loss": 1.685,
+      "step": 1535
+    },
+    {
+      "epoch": 0.4440589765828274,
+      "grad_norm": 1.1226602102372294,
+      "learning_rate": 1.1923899124464061e-05,
+      "loss": 1.6046,
+      "step": 1536
+    },
+    {
+      "epoch": 0.4443480774790402,
+      "grad_norm": 1.1136669895243547,
+      "learning_rate": 1.19148945098764e-05,
+      "loss": 1.5789,
+      "step": 1537
+    },
+    {
+      "epoch": 0.44463717837525296,
+      "grad_norm": 1.0447646723045432,
+      "learning_rate": 1.1905888283244122e-05,
+      "loss": 1.6321,
+      "step": 1538
+    },
+    {
+      "epoch": 0.4449262792714657,
+      "grad_norm": 1.1015513334515599,
+      "learning_rate": 1.1896880452149077e-05,
+      "loss": 1.6385,
+      "step": 1539
+    },
+    {
+      "epoch": 0.44521538016767853,
+      "grad_norm": 1.1064446715247294,
+      "learning_rate": 1.1887871024174469e-05,
+      "loss": 1.6045,
+      "step": 1540
+    },
+    {
+      "epoch": 0.4455044810638913,
+      "grad_norm": 1.0640093495277747,
+      "learning_rate": 1.1878860006904833e-05,
+      "loss": 1.6169,
+      "step": 1541
+    },
+    {
+      "epoch": 0.4457935819601041,
+      "grad_norm": 1.0822730820316895,
+      "learning_rate": 1.1869847407926056e-05,
+      "loss": 1.5586,
+      "step": 1542
+    },
+    {
+      "epoch": 0.44608268285631686,
+      "grad_norm": 1.1814795030964806,
+      "learning_rate": 1.1860833234825346e-05,
+      "loss": 1.5969,
+      "step": 1543
+    },
+    {
+      "epoch": 0.4463717837525296,
+      "grad_norm": 1.08674581015306,
+      "learning_rate": 1.1851817495191245e-05,
+      "loss": 1.5683,
+      "step": 1544
+    },
+    {
+      "epoch": 0.44666088464874243,
+      "grad_norm": 1.0464321955579985,
+      "learning_rate": 1.1842800196613609e-05,
+      "loss": 1.5782,
+      "step": 1545
+    },
+    {
+      "epoch": 0.4469499855449552,
+      "grad_norm": 1.111594268877052,
+      "learning_rate": 1.1833781346683602e-05,
+      "loss": 1.5542,
+      "step": 1546
+    },
+    {
+      "epoch": 0.44723908644116794,
+      "grad_norm": 1.082640700956203,
+      "learning_rate": 1.18247609529937e-05,
+      "loss": 1.5979,
+      "step": 1547
+    },
+    {
+      "epoch": 0.44752818733738076,
+      "grad_norm": 1.0888423501633429,
+      "learning_rate": 1.1815739023137685e-05,
+      "loss": 1.7122,
+      "step": 1548
+    },
+    {
+      "epoch": 0.4478172882335935,
+      "grad_norm": 1.0678109212755142,
+      "learning_rate": 1.1806715564710621e-05,
+      "loss": 1.5567,
+      "step": 1549
+    },
+    {
+      "epoch": 0.4481063891298063,
+      "grad_norm": 1.0663194066367874,
+      "learning_rate": 1.1797690585308857e-05,
+      "loss": 1.516,
+      "step": 1550
+    },
+    {
+      "epoch": 0.4483954900260191,
+      "grad_norm": 1.02306443546824,
+      "learning_rate": 1.1788664092530031e-05,
+      "loss": 1.564,
+      "step": 1551
+    },
+    {
+      "epoch": 0.44868459092223184,
+      "grad_norm": 1.056552748135682,
+      "learning_rate": 1.1779636093973052e-05,
+      "loss": 1.4619,
+      "step": 1552
+    },
+    {
+      "epoch": 0.44897369181844465,
+      "grad_norm": 1.0234373195230095,
+      "learning_rate": 1.1770606597238102e-05,
+      "loss": 1.6078,
+      "step": 1553
+    },
+    {
+      "epoch": 0.4492627927146574,
+      "grad_norm": 1.149560071892715,
+      "learning_rate": 1.1761575609926612e-05,
+      "loss": 1.485,
+      "step": 1554
+    },
+    {
+      "epoch": 0.44955189361087017,
+      "grad_norm": 1.1069601273348477,
+      "learning_rate": 1.1752543139641276e-05,
+      "loss": 1.5623,
+      "step": 1555
+    },
+    {
+      "epoch": 0.449840994507083,
+      "grad_norm": 1.0178266134407419,
+      "learning_rate": 1.1743509193986039e-05,
+      "loss": 1.646,
+      "step": 1556
+    },
+    {
+      "epoch": 0.45013009540329574,
+      "grad_norm": 1.0446257604264455,
+      "learning_rate": 1.173447378056608e-05,
+      "loss": 1.5142,
+      "step": 1557
+    },
+    {
+      "epoch": 0.45041919629950855,
+      "grad_norm": 1.059494719950429,
+      "learning_rate": 1.1725436906987814e-05,
+      "loss": 1.5714,
+      "step": 1558
+    },
+    {
+      "epoch": 0.4507082971957213,
+      "grad_norm": 1.091998768943903,
+      "learning_rate": 1.1716398580858903e-05,
+      "loss": 1.5684,
+      "step": 1559
+    },
+    {
+      "epoch": 0.45099739809193407,
+      "grad_norm": 1.016683438935399,
+      "learning_rate": 1.1707358809788203e-05,
+      "loss": 1.5855,
+      "step": 1560
+    },
+    {
+      "epoch": 0.4512864989881469,
+      "grad_norm": 1.050061191728689,
+      "learning_rate": 1.1698317601385808e-05,
+      "loss": 1.5567,
+      "step": 1561
+    },
+    {
+      "epoch": 0.45157559988435964,
+      "grad_norm": 1.0389514452949107,
+      "learning_rate": 1.1689274963263016e-05,
+      "loss": 1.7099,
+      "step": 1562
+    },
+    {
+      "epoch": 0.4518647007805724,
+      "grad_norm": 1.7849120259460782,
+      "learning_rate": 1.168023090303232e-05,
+      "loss": 1.4716,
+      "step": 1563
+    },
+    {
+      "epoch": 0.4521538016767852,
+      "grad_norm": 1.0651537073162627,
+      "learning_rate": 1.1671185428307433e-05,
+      "loss": 1.5449,
+      "step": 1564
+    },
+    {
+      "epoch": 0.45244290257299796,
+      "grad_norm": 1.0221742073168003,
+      "learning_rate": 1.166213854670323e-05,
+      "loss": 1.5334,
+      "step": 1565
+    },
+    {
+      "epoch": 0.4527320034692108,
+      "grad_norm": 1.0748914241769922,
+      "learning_rate": 1.165309026583579e-05,
+      "loss": 1.5314,
+      "step": 1566
+    },
+    {
+      "epoch": 0.45302110436542353,
+      "grad_norm": 1.466573724860024,
+      "learning_rate": 1.1644040593322365e-05,
+      "loss": 1.5139,
+      "step": 1567
+    },
+    {
+      "epoch": 0.4533102052616363,
+      "grad_norm": 1.0931702692217278,
+      "learning_rate": 1.1634989536781372e-05,
+      "loss": 1.5833,
+      "step": 1568
+    },
+    {
+      "epoch": 0.4535993061578491,
+      "grad_norm": 1.0411354682449292,
+      "learning_rate": 1.1625937103832407e-05,
+      "loss": 1.6399,
+      "step": 1569
+    },
+    {
+      "epoch": 0.45388840705406186,
+      "grad_norm": 1.0597888121112922,
+      "learning_rate": 1.1616883302096209e-05,
+      "loss": 1.5814,
+      "step": 1570
+    },
+    {
+      "epoch": 0.4541775079502747,
+      "grad_norm": 1.0740350008228454,
+      "learning_rate": 1.1607828139194683e-05,
+      "loss": 1.5253,
+      "step": 1571
+    },
+    {
+      "epoch": 0.45446660884648743,
+      "grad_norm": 1.0409462958383806,
+      "learning_rate": 1.1598771622750873e-05,
+      "loss": 1.5347,
+      "step": 1572
+    },
+    {
+      "epoch": 0.4547557097427002,
+      "grad_norm": 1.1372194468475538,
+      "learning_rate": 1.1589713760388958e-05,
+      "loss": 1.6414,
+      "step": 1573
+    },
+    {
+      "epoch": 0.455044810638913,
+      "grad_norm": 1.1187189015918522,
+      "learning_rate": 1.1580654559734259e-05,
+      "loss": 1.5682,
+      "step": 1574
+    },
+    {
+      "epoch": 0.45533391153512576,
+      "grad_norm": 1.7328517965634693,
+      "learning_rate": 1.157159402841322e-05,
+      "loss": 1.5551,
+      "step": 1575
+    },
+    {
+      "epoch": 0.4556230124313385,
+      "grad_norm": 1.068455330968356,
+      "learning_rate": 1.1562532174053404e-05,
+      "loss": 1.5969,
+      "step": 1576
+    },
+    {
+      "epoch": 0.4559121133275513,
+      "grad_norm": 1.0707118667885611,
+      "learning_rate": 1.1553469004283494e-05,
+      "loss": 1.5749,
+      "step": 1577
+    },
+    {
+      "epoch": 0.4562012142237641,
+      "grad_norm": 1.0411663720096676,
+      "learning_rate": 1.1544404526733266e-05,
+      "loss": 1.5755,
+      "step": 1578
+    },
+    {
+      "epoch": 0.4564903151199769,
+      "grad_norm": 1.0328702611371707,
+      "learning_rate": 1.1535338749033614e-05,
+      "loss": 1.524,
+      "step": 1579
+    },
+    {
+      "epoch": 0.45677941601618965,
+      "grad_norm": 1.0257612618740342,
+      "learning_rate": 1.1526271678816514e-05,
+      "loss": 1.6144,
+      "step": 1580
+    },
+    {
+      "epoch": 0.4570685169124024,
+      "grad_norm": 1.0956312975841414,
+      "learning_rate": 1.151720332371504e-05,
+      "loss": 1.6069,
+      "step": 1581
+    },
+    {
+      "epoch": 0.4573576178086152,
+      "grad_norm": 1.3713429707098452,
+      "learning_rate": 1.150813369136334e-05,
+      "loss": 1.5297,
+      "step": 1582
+    },
+    {
+      "epoch": 0.457646718704828,
+      "grad_norm": 1.091470175784792,
+      "learning_rate": 1.1499062789396635e-05,
+      "loss": 1.571,
+      "step": 1583
+    },
+    {
+      "epoch": 0.45793581960104074,
+      "grad_norm": 1.1108442385928508,
+      "learning_rate": 1.1489990625451227e-05,
+      "loss": 1.6471,
+      "step": 1584
+    },
+    {
+      "epoch": 0.45822492049725355,
+      "grad_norm": 1.1023999808783733,
+      "learning_rate": 1.1480917207164467e-05,
+      "loss": 1.5595,
+      "step": 1585
+    },
+    {
+      "epoch": 0.4585140213934663,
+      "grad_norm": 1.1260762972365308,
+      "learning_rate": 1.1471842542174775e-05,
+      "loss": 1.5006,
+      "step": 1586
+    },
+    {
+      "epoch": 0.4588031222896791,
+      "grad_norm": 1.1053624122096908,
+      "learning_rate": 1.146276663812161e-05,
+      "loss": 1.6,
+      "step": 1587
+    },
+    {
+      "epoch": 0.4590922231858919,
+      "grad_norm": 1.104581041077756,
+      "learning_rate": 1.1453689502645473e-05,
+      "loss": 1.5319,
+      "step": 1588
+    },
+    {
+      "epoch": 0.45938132408210464,
+      "grad_norm": 1.0676860284145446,
+      "learning_rate": 1.1444611143387916e-05,
+      "loss": 1.5868,
+      "step": 1589
+    },
+    {
+      "epoch": 0.45967042497831745,
+      "grad_norm": 1.0959161374504272,
+      "learning_rate": 1.1435531567991498e-05,
+      "loss": 1.5595,
+      "step": 1590
+    },
+    {
+      "epoch": 0.4599595258745302,
+      "grad_norm": 1.1214747192669348,
+      "learning_rate": 1.1426450784099829e-05,
+      "loss": 1.5731,
+      "step": 1591
+    },
+    {
+      "epoch": 0.46024862677074296,
+      "grad_norm": 1.113641845373453,
+      "learning_rate": 1.1417368799357516e-05,
+      "loss": 1.5992,
+      "step": 1592
+    },
+    {
+      "epoch": 0.4605377276669558,
+      "grad_norm": 1.1149988699433708,
+      "learning_rate": 1.1408285621410185e-05,
+      "loss": 1.6042,
+      "step": 1593
+    },
+    {
+      "epoch": 0.46082682856316853,
+      "grad_norm": 1.0287401302335453,
+      "learning_rate": 1.1399201257904465e-05,
+      "loss": 1.5863,
+      "step": 1594
+    },
+    {
+      "epoch": 0.46111592945938135,
+      "grad_norm": 1.0633324078858208,
+      "learning_rate": 1.1390115716487978e-05,
+      "loss": 1.5317,
+      "step": 1595
+    },
+    {
+      "epoch": 0.4614050303555941,
+      "grad_norm": 1.068392380788327,
+      "learning_rate": 1.1381029004809352e-05,
+      "loss": 1.5892,
+      "step": 1596
+    },
+    {
+      "epoch": 0.46169413125180686,
+      "grad_norm": 1.0318195365557383,
+      "learning_rate": 1.1371941130518185e-05,
+      "loss": 1.5024,
+      "step": 1597
+    },
+    {
+      "epoch": 0.4619832321480197,
+      "grad_norm": 1.0103161704296233,
+      "learning_rate": 1.1362852101265064e-05,
+      "loss": 1.6208,
+      "step": 1598
+    },
+    {
+      "epoch": 0.46227233304423243,
+      "grad_norm": 1.0086510586157338,
+      "learning_rate": 1.1353761924701538e-05,
+      "loss": 1.5261,
+      "step": 1599
+    },
+    {
+      "epoch": 0.4625614339404452,
+      "grad_norm": 1.0455937364849,
+      "learning_rate": 1.134467060848013e-05,
+      "loss": 1.6005,
+      "step": 1600
+    },
+    {
+      "epoch": 0.462850534836658,
+      "grad_norm": 0.9613626742669176,
+      "learning_rate": 1.1335578160254326e-05,
+      "loss": 1.4253,
+      "step": 1601
+    },
+    {
+      "epoch": 0.46313963573287076,
+      "grad_norm": 1.071290759393101,
+      "learning_rate": 1.132648458767855e-05,
+      "loss": 1.6138,
+      "step": 1602
+    },
+    {
+      "epoch": 0.46342873662908357,
+      "grad_norm": 1.051140600560518,
+      "learning_rate": 1.1317389898408188e-05,
+      "loss": 1.5134,
+      "step": 1603
+    },
+    {
+      "epoch": 0.46371783752529633,
+      "grad_norm": 1.0821930565958098,
+      "learning_rate": 1.1308294100099562e-05,
+      "loss": 1.5665,
+      "step": 1604
+    },
+    {
+      "epoch": 0.4640069384215091,
+      "grad_norm": 1.1003437952657174,
+      "learning_rate": 1.1299197200409917e-05,
+      "loss": 1.5263,
+      "step": 1605
+    },
+    {
+      "epoch": 0.4642960393177219,
+      "grad_norm": 1.0391985647731157,
+      "learning_rate": 1.1290099206997438e-05,
+      "loss": 1.5042,
+      "step": 1606
+    },
+    {
+      "epoch": 0.46458514021393466,
+      "grad_norm": 1.0363219491434108,
+      "learning_rate": 1.128100012752123e-05,
+      "loss": 1.5536,
+      "step": 1607
+    },
+    {
+      "epoch": 0.46487424111014747,
+      "grad_norm": 1.0932615951967841,
+      "learning_rate": 1.1271899969641302e-05,
+      "loss": 1.6483,
+      "step": 1608
+    },
+    {
+      "epoch": 0.4651633420063602,
+      "grad_norm": 1.099797469753584,
+      "learning_rate": 1.1262798741018584e-05,
+      "loss": 1.491,
+      "step": 1609
+    },
+    {
+      "epoch": 0.465452442902573,
+      "grad_norm": 1.0510689443346386,
+      "learning_rate": 1.1253696449314898e-05,
+      "loss": 1.7005,
+      "step": 1610
+    },
+    {
+      "epoch": 0.4657415437987858,
+      "grad_norm": 1.0234538344317012,
+      "learning_rate": 1.1244593102192961e-05,
+      "loss": 1.5703,
+      "step": 1611
+    },
+    {
+      "epoch": 0.46603064469499855,
+      "grad_norm": 1.0790829714844545,
+      "learning_rate": 1.1235488707316387e-05,
+      "loss": 1.6022,
+      "step": 1612
+    },
+    {
+      "epoch": 0.4663197455912113,
+      "grad_norm": 1.0451514730579405,
+      "learning_rate": 1.1226383272349666e-05,
+      "loss": 1.5938,
+      "step": 1613
+    },
+    {
+      "epoch": 0.4666088464874241,
+      "grad_norm": 1.0669930998029822,
+      "learning_rate": 1.1217276804958157e-05,
+      "loss": 1.6017,
+      "step": 1614
+    },
+    {
+      "epoch": 0.4668979473836369,
+      "grad_norm": 1.0196829402343588,
+      "learning_rate": 1.1208169312808104e-05,
+      "loss": 1.5358,
+      "step": 1615
+    },
+    {
+      "epoch": 0.4671870482798497,
+      "grad_norm": 1.066151252104385,
+      "learning_rate": 1.11990608035666e-05,
+      "loss": 1.6618,
+      "step": 1616
+    },
+    {
+      "epoch": 0.46747614917606245,
+      "grad_norm": 1.0334156912896786,
+      "learning_rate": 1.1189951284901599e-05,
+      "loss": 1.6237,
+      "step": 1617
+    },
+    {
+      "epoch": 0.4677652500722752,
+      "grad_norm": 1.0590452187367156,
+      "learning_rate": 1.1180840764481908e-05,
+      "loss": 1.6095,
+      "step": 1618
+    },
+    {
+      "epoch": 0.468054350968488,
+      "grad_norm": 1.0847143400803,
+      "learning_rate": 1.1171729249977171e-05,
+      "loss": 1.5517,
+      "step": 1619
+    },
+    {
+      "epoch": 0.4683434518647008,
+      "grad_norm": 1.134538192785943,
+      "learning_rate": 1.1162616749057874e-05,
+      "loss": 1.6134,
+      "step": 1620
+    },
+    {
+      "epoch": 0.46863255276091353,
+      "grad_norm": 1.0823848589233827,
+      "learning_rate": 1.1153503269395331e-05,
+      "loss": 1.6387,
+      "step": 1621
+    },
+    {
+      "epoch": 0.46892165365712635,
+      "grad_norm": 1.0163144272064613,
+      "learning_rate": 1.1144388818661677e-05,
+      "loss": 1.5782,
+      "step": 1622
+    },
+    {
+      "epoch": 0.4692107545533391,
+      "grad_norm": 1.0776942753371328,
+      "learning_rate": 1.1135273404529875e-05,
+      "loss": 1.5359,
+      "step": 1623
+    },
+    {
+      "epoch": 0.4694998554495519,
+      "grad_norm": 1.003802100143087,
+      "learning_rate": 1.1126157034673688e-05,
+      "loss": 1.5768,
+      "step": 1624
+    },
+    {
+      "epoch": 0.4697889563457647,
+      "grad_norm": 1.0873382162677923,
+      "learning_rate": 1.1117039716767688e-05,
+      "loss": 1.533,
+      "step": 1625
+    },
+    {
+      "epoch": 0.47007805724197743,
+      "grad_norm": 1.0531425928521536,
+      "learning_rate": 1.1107921458487245e-05,
+      "loss": 1.6297,
+      "step": 1626
+    },
+    {
+      "epoch": 0.47036715813819024,
+      "grad_norm": 1.0262632213830907,
+      "learning_rate": 1.1098802267508517e-05,
+      "loss": 1.5157,
+      "step": 1627
+    },
+    {
+      "epoch": 0.470656259034403,
+      "grad_norm": 1.2067725310773727,
+      "learning_rate": 1.1089682151508459e-05,
+      "loss": 1.6202,
+      "step": 1628
+    },
+    {
+      "epoch": 0.47094535993061576,
+      "grad_norm": 1.0753317731977976,
+      "learning_rate": 1.1080561118164786e-05,
+      "loss": 1.5209,
+      "step": 1629
+    },
+    {
+      "epoch": 0.47123446082682857,
+      "grad_norm": 1.0580197927206985,
+      "learning_rate": 1.1071439175156e-05,
+      "loss": 1.6924,
+      "step": 1630
+    },
+    {
+      "epoch": 0.47152356172304133,
+      "grad_norm": 1.0725638768761314,
+      "learning_rate": 1.1062316330161367e-05,
+      "loss": 1.5522,
+      "step": 1631
+    },
+    {
+      "epoch": 0.47181266261925414,
+      "grad_norm": 1.083581600630059,
+      "learning_rate": 1.1053192590860901e-05,
+      "loss": 1.5417,
+      "step": 1632
+    },
+    {
+      "epoch": 0.4721017635154669,
+      "grad_norm": 1.0278129394551416,
+      "learning_rate": 1.104406796493539e-05,
+      "loss": 1.5825,
+      "step": 1633
+    },
+    {
+      "epoch": 0.47239086441167966,
+      "grad_norm": 1.1023462927154302,
+      "learning_rate": 1.1034942460066347e-05,
+      "loss": 1.6218,
+      "step": 1634
+    },
+    {
+      "epoch": 0.47267996530789247,
+      "grad_norm": 1.0682546633316112,
+      "learning_rate": 1.1025816083936036e-05,
+      "loss": 1.6491,
+      "step": 1635
+    },
+    {
+      "epoch": 0.4729690662041052,
+      "grad_norm": 1.1175374736941341,
+      "learning_rate": 1.1016688844227455e-05,
+      "loss": 1.8207,
+      "step": 1636
+    },
+    {
+      "epoch": 0.473258167100318,
+      "grad_norm": 1.0827764333069483,
+      "learning_rate": 1.1007560748624319e-05,
+      "loss": 1.5792,
+      "step": 1637
+    },
+    {
+      "epoch": 0.4735472679965308,
+      "grad_norm": 1.0382523916497624,
+      "learning_rate": 1.0998431804811078e-05,
+      "loss": 1.6044,
+      "step": 1638
+    },
+    {
+      "epoch": 0.47383636889274355,
+      "grad_norm": 1.1216922278728259,
+      "learning_rate": 1.0989302020472885e-05,
+      "loss": 1.7668,
+      "step": 1639
+    },
+    {
+      "epoch": 0.47412546978895637,
+      "grad_norm": 1.0371382338349069,
+      "learning_rate": 1.098017140329561e-05,
+      "loss": 1.5395,
+      "step": 1640
+    },
+    {
+      "epoch": 0.4744145706851691,
+      "grad_norm": 1.0830426854918693,
+      "learning_rate": 1.097103996096581e-05,
+      "loss": 1.6564,
+      "step": 1641
+    },
+    {
+      "epoch": 0.4747036715813819,
+      "grad_norm": 1.0990995956529412,
+      "learning_rate": 1.0961907701170752e-05,
+      "loss": 1.5566,
+      "step": 1642
+    },
+    {
+      "epoch": 0.4749927724775947,
+      "grad_norm": 1.031100248282213,
+      "learning_rate": 1.0952774631598377e-05,
+      "loss": 1.5668,
+      "step": 1643
+    },
+    {
+      "epoch": 0.47528187337380745,
+      "grad_norm": 1.0435477534831432,
+      "learning_rate": 1.0943640759937324e-05,
+      "loss": 1.5172,
+      "step": 1644
+    },
+    {
+      "epoch": 0.47557097427002026,
+      "grad_norm": 1.091509007714218,
+      "learning_rate": 1.0934506093876897e-05,
+      "loss": 1.5794,
+      "step": 1645
+    },
+    {
+      "epoch": 0.475860075166233,
+      "grad_norm": 1.0743563095457511,
+      "learning_rate": 1.0925370641107064e-05,
+      "loss": 1.5385,
+      "step": 1646
+    },
+    {
+      "epoch": 0.4761491760624458,
+      "grad_norm": 1.0608942539179838,
+      "learning_rate": 1.0916234409318467e-05,
+      "loss": 1.5736,
+      "step": 1647
+    },
+    {
+      "epoch": 0.4764382769586586,
+      "grad_norm": 1.0681582090957384,
+      "learning_rate": 1.0907097406202398e-05,
+      "loss": 1.6182,
+      "step": 1648
+    },
+    {
+      "epoch": 0.47672737785487135,
+      "grad_norm": 1.0213182678474237,
+      "learning_rate": 1.0897959639450792e-05,
+      "loss": 1.5024,
+      "step": 1649
+    },
+    {
+      "epoch": 0.4770164787510841,
+      "grad_norm": 1.0647485643920613,
+      "learning_rate": 1.0888821116756243e-05,
+      "loss": 1.5716,
+      "step": 1650
+    },
+    {
+      "epoch": 0.4773055796472969,
+      "grad_norm": 1.0849697889257992,
+      "learning_rate": 1.0879681845811964e-05,
+      "loss": 1.5062,
+      "step": 1651
+    },
+    {
+      "epoch": 0.4775946805435097,
+      "grad_norm": 1.0913180326235399,
+      "learning_rate": 1.087054183431181e-05,
+      "loss": 1.5116,
+      "step": 1652
+    },
+    {
+      "epoch": 0.4778837814397225,
+      "grad_norm": 1.039888326533737,
+      "learning_rate": 1.0861401089950254e-05,
+      "loss": 1.4923,
+      "step": 1653
+    },
+    {
+      "epoch": 0.47817288233593525,
+      "grad_norm": 1.0342399032651195,
+      "learning_rate": 1.0852259620422385e-05,
+      "loss": 1.5947,
+      "step": 1654
+    },
+    {
+      "epoch": 0.478461983232148,
+      "grad_norm": 1.1748596521831836,
+      "learning_rate": 1.084311743342391e-05,
+      "loss": 1.6407,
+      "step": 1655
+    },
+    {
+      "epoch": 0.4787510841283608,
+      "grad_norm": 1.087895550052869,
+      "learning_rate": 1.0833974536651131e-05,
+      "loss": 1.6236,
+      "step": 1656
+    },
+    {
+      "epoch": 0.4790401850245736,
+      "grad_norm": 1.1343355813196676,
+      "learning_rate": 1.0824830937800948e-05,
+      "loss": 1.5471,
+      "step": 1657
+    },
+    {
+      "epoch": 0.47932928592078633,
+      "grad_norm": 1.0843029348020614,
+      "learning_rate": 1.0815686644570862e-05,
+      "loss": 1.6355,
+      "step": 1658
+    },
+    {
+      "epoch": 0.47961838681699914,
+      "grad_norm": 1.0697500928143135,
+      "learning_rate": 1.0806541664658945e-05,
+      "loss": 1.6009,
+      "step": 1659
+    },
+    {
+      "epoch": 0.4799074877132119,
+      "grad_norm": 1.0719292107565337,
+      "learning_rate": 1.0797396005763861e-05,
+      "loss": 1.6623,
+      "step": 1660
+    },
+    {
+      "epoch": 0.4801965886094247,
+      "grad_norm": 1.0532747692042155,
+      "learning_rate": 1.0788249675584832e-05,
+      "loss": 1.6036,
+      "step": 1661
+    },
+    {
+      "epoch": 0.48048568950563747,
+      "grad_norm": 1.0424651572707906,
+      "learning_rate": 1.0779102681821653e-05,
+      "loss": 1.4984,
+      "step": 1662
+    },
+    {
+      "epoch": 0.4807747904018502,
+      "grad_norm": 1.0886550395567078,
+      "learning_rate": 1.0769955032174679e-05,
+      "loss": 1.6285,
+      "step": 1663
+    },
+    {
+      "epoch": 0.48106389129806304,
+      "grad_norm": 1.0267085704151517,
+      "learning_rate": 1.076080673434481e-05,
+      "loss": 1.5405,
+      "step": 1664
+    },
+    {
+      "epoch": 0.4813529921942758,
+      "grad_norm": 1.0559905124719542,
+      "learning_rate": 1.0751657796033502e-05,
+      "loss": 1.573,
+      "step": 1665
+    },
+    {
+      "epoch": 0.48164209309048855,
+      "grad_norm": 1.131594500921746,
+      "learning_rate": 1.0742508224942736e-05,
+      "loss": 1.5752,
+      "step": 1666
+    },
+    {
+      "epoch": 0.48193119398670137,
+      "grad_norm": 1.150084532194081,
+      "learning_rate": 1.073335802877504e-05,
+      "loss": 1.4852,
+      "step": 1667
+    },
+    {
+      "epoch": 0.4822202948829141,
+      "grad_norm": 1.1194896204289335,
+      "learning_rate": 1.0724207215233463e-05,
+      "loss": 1.6601,
+      "step": 1668
+    },
+    {
+      "epoch": 0.48250939577912694,
+      "grad_norm": 1.1290501541078468,
+      "learning_rate": 1.0715055792021566e-05,
+      "loss": 1.5427,
+      "step": 1669
+    },
+    {
+      "epoch": 0.4827984966753397,
+      "grad_norm": 1.104539874976553,
+      "learning_rate": 1.0705903766843435e-05,
+      "loss": 1.6067,
+      "step": 1670
+    },
+    {
+      "epoch": 0.48308759757155245,
+      "grad_norm": 1.1073235244566444,
+      "learning_rate": 1.0696751147403658e-05,
+      "loss": 1.5524,
+      "step": 1671
+    },
+    {
+      "epoch": 0.48337669846776526,
+      "grad_norm": 1.074472848144284,
+      "learning_rate": 1.0687597941407325e-05,
+      "loss": 1.6004,
+      "step": 1672
+    },
+    {
+      "epoch": 0.483665799363978,
+      "grad_norm": 1.071105271509608,
+      "learning_rate": 1.0678444156560011e-05,
+      "loss": 1.6048,
+      "step": 1673
+    },
+    {
+      "epoch": 0.4839549002601908,
+      "grad_norm": 1.0641380352478016,
+      "learning_rate": 1.0669289800567787e-05,
+      "loss": 1.5769,
+      "step": 1674
+    },
+    {
+      "epoch": 0.4842440011564036,
+      "grad_norm": 1.060800400572137,
+      "learning_rate": 1.066013488113721e-05,
+      "loss": 1.5111,
+      "step": 1675
+    },
+    {
+      "epoch": 0.48453310205261635,
+      "grad_norm": 1.0474552536762647,
+      "learning_rate": 1.0650979405975297e-05,
+      "loss": 1.5754,
+      "step": 1676
+    },
+    {
+      "epoch": 0.48482220294882916,
+      "grad_norm": 1.007603839207087,
+      "learning_rate": 1.0641823382789544e-05,
+      "loss": 1.4866,
+      "step": 1677
+    },
+    {
+      "epoch": 0.4851113038450419,
+      "grad_norm": 1.2159391237328345,
+      "learning_rate": 1.06326668192879e-05,
+      "loss": 1.6593,
+      "step": 1678
+    },
+    {
+      "epoch": 0.4854004047412547,
+      "grad_norm": 1.0903022925597354,
+      "learning_rate": 1.062350972317878e-05,
+      "loss": 1.6951,
+      "step": 1679
+    },
+    {
+      "epoch": 0.4856895056374675,
+      "grad_norm": 1.0985715524256023,
+      "learning_rate": 1.0614352102171037e-05,
+      "loss": 1.5879,
+      "step": 1680
+    },
+    {
+      "epoch": 0.48597860653368025,
+      "grad_norm": 1.0323836708693017,
+      "learning_rate": 1.0605193963973966e-05,
+      "loss": 1.5354,
+      "step": 1681
+    },
+    {
+      "epoch": 0.48626770742989306,
+      "grad_norm": 1.0543021504639354,
+      "learning_rate": 1.0596035316297309e-05,
+      "loss": 1.6213,
+      "step": 1682
+    },
+    {
+      "epoch": 0.4865568083261058,
+      "grad_norm": 1.0339397180093894,
+      "learning_rate": 1.0586876166851223e-05,
+      "loss": 1.5267,
+      "step": 1683
+    },
+    {
+      "epoch": 0.4868459092223186,
+      "grad_norm": 1.0406151485219466,
+      "learning_rate": 1.0577716523346295e-05,
+      "loss": 1.6136,
+      "step": 1684
+    },
+    {
+      "epoch": 0.4871350101185314,
+      "grad_norm": 0.9951731314723519,
+      "learning_rate": 1.0568556393493527e-05,
+      "loss": 1.5507,
+      "step": 1685
+    },
+    {
+      "epoch": 0.48742411101474414,
+      "grad_norm": 1.0701913301244408,
+      "learning_rate": 1.0559395785004327e-05,
+      "loss": 1.6703,
+      "step": 1686
+    },
+    {
+      "epoch": 0.4877132119109569,
+      "grad_norm": 1.0625542392304808,
+      "learning_rate": 1.0550234705590514e-05,
+      "loss": 1.5405,
+      "step": 1687
+    },
+    {
+      "epoch": 0.4880023128071697,
+      "grad_norm": 1.0211406487410304,
+      "learning_rate": 1.0541073162964293e-05,
+      "loss": 1.5652,
+      "step": 1688
+    },
+    {
+      "epoch": 0.48829141370338247,
+      "grad_norm": 1.0745108247557758,
+      "learning_rate": 1.0531911164838266e-05,
+      "loss": 1.6121,
+      "step": 1689
+    },
+    {
+      "epoch": 0.4885805145995953,
+      "grad_norm": 1.0807661898879897,
+      "learning_rate": 1.0522748718925417e-05,
+      "loss": 1.5362,
+      "step": 1690
+    },
+    {
+      "epoch": 0.48886961549580804,
+      "grad_norm": 1.064890656037219,
+      "learning_rate": 1.0513585832939104e-05,
+      "loss": 1.5601,
+      "step": 1691
+    },
+    {
+      "epoch": 0.4891587163920208,
+      "grad_norm": 1.0625374581866778,
+      "learning_rate": 1.0504422514593062e-05,
+      "loss": 1.5926,
+      "step": 1692
+    },
+    {
+      "epoch": 0.4894478172882336,
+      "grad_norm": 1.0724240407997907,
+      "learning_rate": 1.0495258771601381e-05,
+      "loss": 1.6645,
+      "step": 1693
+    },
+    {
+      "epoch": 0.48973691818444637,
+      "grad_norm": 1.0525580007409545,
+      "learning_rate": 1.0486094611678519e-05,
+      "loss": 1.5948,
+      "step": 1694
+    },
+    {
+      "epoch": 0.4900260190806591,
+      "grad_norm": 1.0441702294744506,
+      "learning_rate": 1.0476930042539275e-05,
+      "loss": 1.5587,
+      "step": 1695
+    },
+    {
+      "epoch": 0.49031511997687194,
+      "grad_norm": 1.029299960987906,
+      "learning_rate": 1.0467765071898799e-05,
+      "loss": 1.5573,
+      "step": 1696
+    },
+    {
+      "epoch": 0.4906042208730847,
+      "grad_norm": 1.0709375351662775,
+      "learning_rate": 1.0458599707472574e-05,
+      "loss": 1.5501,
+      "step": 1697
+    },
+    {
+      "epoch": 0.4908933217692975,
+      "grad_norm": 1.101995131098102,
+      "learning_rate": 1.044943395697642e-05,
+      "loss": 1.4952,
+      "step": 1698
+    },
+    {
+      "epoch": 0.49118242266551027,
+      "grad_norm": 1.1811156883022564,
+      "learning_rate": 1.0440267828126478e-05,
+      "loss": 1.568,
+      "step": 1699
+    },
+    {
+      "epoch": 0.491471523561723,
+      "grad_norm": 1.0335261079152926,
+      "learning_rate": 1.0431101328639216e-05,
+      "loss": 1.5828,
+      "step": 1700
+    },
+    {
+      "epoch": 0.49176062445793584,
+      "grad_norm": 1.0638854019675978,
+      "learning_rate": 1.0421934466231394e-05,
+      "loss": 1.5773,
+      "step": 1701
+    },
+    {
+      "epoch": 0.4920497253541486,
+      "grad_norm": 1.115464345515217,
+      "learning_rate": 1.0412767248620099e-05,
+      "loss": 1.6548,
+      "step": 1702
+    },
+    {
+      "epoch": 0.49233882625036135,
+      "grad_norm": 1.0626812242641446,
+      "learning_rate": 1.0403599683522705e-05,
+      "loss": 1.591,
+      "step": 1703
+    },
+    {
+      "epoch": 0.49262792714657416,
+      "grad_norm": 1.0671233733865448,
+      "learning_rate": 1.0394431778656888e-05,
+      "loss": 1.5684,
+      "step": 1704
+    },
+    {
+      "epoch": 0.4929170280427869,
+      "grad_norm": 1.1021113399785827,
+      "learning_rate": 1.0385263541740598e-05,
+      "loss": 1.5779,
+      "step": 1705
+    },
+    {
+      "epoch": 0.49320612893899973,
+      "grad_norm": 1.043379736588795,
+      "learning_rate": 1.037609498049207e-05,
+      "loss": 1.5176,
+      "step": 1706
+    },
+    {
+      "epoch": 0.4934952298352125,
+      "grad_norm": 1.0365564313352789,
+      "learning_rate": 1.0366926102629817e-05,
+      "loss": 1.4853,
+      "step": 1707
+    },
+    {
+      "epoch": 0.49378433073142525,
+      "grad_norm": 1.0858259034748727,
+      "learning_rate": 1.0357756915872611e-05,
+      "loss": 1.6476,
+      "step": 1708
+    },
+    {
+      "epoch": 0.49407343162763806,
+      "grad_norm": 1.1233478878548293,
+      "learning_rate": 1.0348587427939492e-05,
+      "loss": 1.5527,
+      "step": 1709
+    },
+    {
+      "epoch": 0.4943625325238508,
+      "grad_norm": 1.0779713429219018,
+      "learning_rate": 1.0339417646549742e-05,
+      "loss": 1.5684,
+      "step": 1710
+    },
+    {
+      "epoch": 0.4946516334200636,
+      "grad_norm": 1.0761334573518275,
+      "learning_rate": 1.0330247579422898e-05,
+      "loss": 1.6561,
+      "step": 1711
+    },
+    {
+      "epoch": 0.4949407343162764,
+      "grad_norm": 1.1089608578875743,
+      "learning_rate": 1.0321077234278743e-05,
+      "loss": 1.646,
+      "step": 1712
+    },
+    {
+      "epoch": 0.49522983521248914,
+      "grad_norm": 1.0560166622435565,
+      "learning_rate": 1.0311906618837277e-05,
+      "loss": 1.5802,
+      "step": 1713
+    },
+    {
+      "epoch": 0.49551893610870196,
+      "grad_norm": 1.0529065288869814,
+      "learning_rate": 1.0302735740818747e-05,
+      "loss": 1.5121,
+      "step": 1714
+    },
+    {
+      "epoch": 0.4958080370049147,
+      "grad_norm": 1.0481800085113295,
+      "learning_rate": 1.0293564607943609e-05,
+      "loss": 1.6393,
+      "step": 1715
+    },
+    {
+      "epoch": 0.49609713790112747,
+      "grad_norm": 1.0835132231615257,
+      "learning_rate": 1.0284393227932534e-05,
+      "loss": 1.6045,
+      "step": 1716
+    },
+    {
+      "epoch": 0.4963862387973403,
+      "grad_norm": 1.0747400304049686,
+      "learning_rate": 1.027522160850641e-05,
+      "loss": 1.5536,
+      "step": 1717
+    },
+    {
+      "epoch": 0.49667533969355304,
+      "grad_norm": 1.1192474440904863,
+      "learning_rate": 1.026604975738631e-05,
+      "loss": 1.5686,
+      "step": 1718
+    },
+    {
+      "epoch": 0.49696444058976585,
+      "grad_norm": 1.094863591839747,
+      "learning_rate": 1.025687768229352e-05,
+      "loss": 1.643,
+      "step": 1719
+    },
+    {
+      "epoch": 0.4972535414859786,
+      "grad_norm": 1.0439631447432807,
+      "learning_rate": 1.0247705390949506e-05,
+      "loss": 1.5885,
+      "step": 1720
+    },
+    {
+      "epoch": 0.49754264238219137,
+      "grad_norm": 1.029775520918905,
+      "learning_rate": 1.0238532891075912e-05,
+      "loss": 1.5019,
+      "step": 1721
+    },
+    {
+      "epoch": 0.4978317432784042,
+      "grad_norm": 1.0638879666958196,
+      "learning_rate": 1.0229360190394571e-05,
+      "loss": 1.5869,
+      "step": 1722
+    },
+    {
+      "epoch": 0.49812084417461694,
+      "grad_norm": 1.067588321237544,
+      "learning_rate": 1.0220187296627464e-05,
+      "loss": 1.608,
+      "step": 1723
+    },
+    {
+      "epoch": 0.4984099450708297,
+      "grad_norm": 1.033910733516868,
+      "learning_rate": 1.0211014217496762e-05,
+      "loss": 1.6418,
+      "step": 1724
+    },
+    {
+      "epoch": 0.4986990459670425,
+      "grad_norm": 1.051519685707975,
+      "learning_rate": 1.0201840960724764e-05,
+      "loss": 1.5798,
+      "step": 1725
+    },
+    {
+      "epoch": 0.49898814686325527,
+      "grad_norm": 1.0708890626344802,
+      "learning_rate": 1.0192667534033937e-05,
+      "loss": 1.6132,
+      "step": 1726
+    },
+    {
+      "epoch": 0.4992772477594681,
+      "grad_norm": 1.0576483170475506,
+      "learning_rate": 1.0183493945146887e-05,
+      "loss": 1.5514,
+      "step": 1727
+    },
+    {
+      "epoch": 0.49956634865568084,
+      "grad_norm": 1.064150983535629,
+      "learning_rate": 1.0174320201786352e-05,
+      "loss": 1.5687,
+      "step": 1728
+    },
+    {
+      "epoch": 0.4998554495518936,
+      "grad_norm": 1.0973736743884872,
+      "learning_rate": 1.0165146311675205e-05,
+      "loss": 1.6397,
+      "step": 1729
+    },
+    {
+      "epoch": 0.5001445504481064,
+      "grad_norm": 1.0596298933286115,
+      "learning_rate": 1.015597228253644e-05,
+      "loss": 1.5329,
+      "step": 1730
+    },
+    {
+      "epoch": 0.5004336513443192,
+      "grad_norm": 1.0758177562594944,
+      "learning_rate": 1.0146798122093167e-05,
+      "loss": 1.6121,
+      "step": 1731
+    },
+    {
+      "epoch": 0.500722752240532,
+      "grad_norm": 1.0911173133762655,
+      "learning_rate": 1.0137623838068613e-05,
+      "loss": 1.554,
+      "step": 1732
+    },
+    {
+      "epoch": 0.5010118531367447,
+      "grad_norm": 1.1389678927745486,
+      "learning_rate": 1.0128449438186098e-05,
+      "loss": 1.6776,
+      "step": 1733
+    },
+    {
+      "epoch": 0.5013009540329575,
+      "grad_norm": 1.1608598708897635,
+      "learning_rate": 1.0119274930169052e-05,
+      "loss": 1.5955,
+      "step": 1734
+    },
+    {
+      "epoch": 0.5015900549291703,
+      "grad_norm": 1.0512239995362633,
+      "learning_rate": 1.0110100321740986e-05,
+      "loss": 1.53,
+      "step": 1735
+    },
+    {
+      "epoch": 0.501879155825383,
+      "grad_norm": 1.0040385125504452,
+      "learning_rate": 1.0100925620625503e-05,
+      "loss": 1.5813,
+      "step": 1736
+    },
+    {
+      "epoch": 0.5021682567215958,
+      "grad_norm": 1.077983886477995,
+      "learning_rate": 1.0091750834546274e-05,
+      "loss": 1.6012,
+      "step": 1737
+    },
+    {
+      "epoch": 0.5024573576178086,
+      "grad_norm": 1.0488937847075557,
+      "learning_rate": 1.0082575971227055e-05,
+      "loss": 1.5804,
+      "step": 1738
+    },
+    {
+      "epoch": 0.5027464585140214,
+      "grad_norm": 1.0309491977598126,
+      "learning_rate": 1.0073401038391655e-05,
+      "loss": 1.5519,
+      "step": 1739
+    },
+    {
+      "epoch": 0.5030355594102341,
+      "grad_norm": 1.0259986169098292,
+      "learning_rate": 1.006422604376395e-05,
+      "loss": 1.5864,
+      "step": 1740
+    },
+    {
+      "epoch": 0.503324660306447,
+      "grad_norm": 1.1212168811852075,
+      "learning_rate": 1.0055050995067863e-05,
+      "loss": 1.5996,
+      "step": 1741
+    },
+    {
+      "epoch": 0.5036137612026598,
+      "grad_norm": 1.0764774481934771,
+      "learning_rate": 1.0045875900027362e-05,
+      "loss": 1.594,
+      "step": 1742
+    },
+    {
+      "epoch": 0.5039028620988725,
+      "grad_norm": 1.0673768912814041,
+      "learning_rate": 1.0036700766366459e-05,
+      "loss": 1.5524,
+      "step": 1743
+    },
+    {
+      "epoch": 0.5041919629950853,
+      "grad_norm": 1.0326097508296919,
+      "learning_rate": 1.0027525601809196e-05,
+      "loss": 1.5714,
+      "step": 1744
+    },
+    {
+      "epoch": 0.5044810638912981,
+      "grad_norm": 2.2681512679091975,
+      "learning_rate": 1.0018350414079636e-05,
+      "loss": 1.6576,
+      "step": 1745
+    },
+    {
+      "epoch": 0.5047701647875108,
+      "grad_norm": 1.0772275670890892,
+      "learning_rate": 1.0009175210901874e-05,
+      "loss": 1.5079,
+      "step": 1746
+    },
+    {
+      "epoch": 0.5050592656837236,
+      "grad_norm": 1.0207418254960374,
+      "learning_rate": 1e-05,
+      "loss": 1.4982,
+      "step": 1747
+    },
+    {
+      "epoch": 0.5053483665799364,
+      "grad_norm": 1.116188922272078,
+      "learning_rate": 9.99082478909813e-06,
+      "loss": 1.6675,
+      "step": 1748
+    },
+    {
+      "epoch": 0.5056374674761491,
+      "grad_norm": 1.0372714347178416,
+      "learning_rate": 9.981649585920367e-06,
+      "loss": 1.6048,
+      "step": 1749
+    },
+    {
+      "epoch": 0.5059265683723619,
+      "grad_norm": 1.0514621069166015,
+      "learning_rate": 9.972474398190809e-06,
+      "loss": 1.59,
+      "step": 1750
+    },
+    {
+      "epoch": 0.5062156692685748,
+      "grad_norm": 1.0822214504080134,
+      "learning_rate": 9.963299233633544e-06,
+      "loss": 1.5685,
+      "step": 1751
+    },
+    {
+      "epoch": 0.5065047701647876,
+      "grad_norm": 1.0494123652176268,
+      "learning_rate": 9.954124099972643e-06,
+      "loss": 1.6974,
+      "step": 1752
+    },
+    {
+      "epoch": 0.5067938710610003,
+      "grad_norm": 1.0451781287433801,
+      "learning_rate": 9.94494900493214e-06,
+      "loss": 1.5801,
+      "step": 1753
+    },
+    {
+      "epoch": 0.5070829719572131,
+      "grad_norm": 1.1206042060450756,
+      "learning_rate": 9.935773956236052e-06,
+      "loss": 1.5897,
+      "step": 1754
+    },
+    {
+      "epoch": 0.5073720728534259,
+      "grad_norm": 1.0557251169995299,
+      "learning_rate": 9.926598961608347e-06,
+      "loss": 1.5872,
+      "step": 1755
+    },
+    {
+      "epoch": 0.5076611737496386,
+      "grad_norm": 1.0356049682750146,
+      "learning_rate": 9.917424028772949e-06,
+      "loss": 1.6173,
+      "step": 1756
+    },
+    {
+      "epoch": 0.5079502746458514,
+      "grad_norm": 1.0375178090579145,
+      "learning_rate": 9.908249165453727e-06,
+      "loss": 1.5486,
+      "step": 1757
+    },
+    {
+      "epoch": 0.5082393755420642,
+      "grad_norm": 1.0599583237889871,
+      "learning_rate": 9.8990743793745e-06,
+      "loss": 1.6356,
+      "step": 1758
+    },
+    {
+      "epoch": 0.5085284764382769,
+      "grad_norm": 1.0844697954651266,
+      "learning_rate": 9.889899678259016e-06,
+      "loss": 1.6366,
+      "step": 1759
+    },
+    {
+      "epoch": 0.5088175773344897,
+      "grad_norm": 1.078668064066197,
+      "learning_rate": 9.880725069830951e-06,
+      "loss": 1.5257,
+      "step": 1760
+    },
+    {
+      "epoch": 0.5091066782307025,
+      "grad_norm": 1.0884035286214844,
+      "learning_rate": 9.871550561813905e-06,
+      "loss": 1.6437,
+      "step": 1761
+    },
+    {
+      "epoch": 0.5093957791269152,
+      "grad_norm": 1.0672691706862036,
+      "learning_rate": 9.862376161931392e-06,
+      "loss": 1.6464,
+      "step": 1762
+    },
+    {
+      "epoch": 0.5096848800231281,
+      "grad_norm": 1.0748288001430242,
+      "learning_rate": 9.853201877906836e-06,
+      "loss": 1.5667,
+      "step": 1763
+    },
+    {
+      "epoch": 0.5099739809193409,
+      "grad_norm": 1.0901544889035424,
+      "learning_rate": 9.844027717463562e-06,
+      "loss": 1.6553,
+      "step": 1764
+    },
+    {
+      "epoch": 0.5102630818155536,
+      "grad_norm": 1.063095348861724,
+      "learning_rate": 9.834853688324795e-06,
+      "loss": 1.6022,
+      "step": 1765
+    },
+    {
+      "epoch": 0.5105521827117664,
+      "grad_norm": 1.0791104600024521,
+      "learning_rate": 9.82567979821365e-06,
+      "loss": 1.6797,
+      "step": 1766
+    },
+    {
+      "epoch": 0.5108412836079792,
+      "grad_norm": 1.015019294504856,
+      "learning_rate": 9.816506054853115e-06,
+      "loss": 1.5354,
+      "step": 1767
+    },
+    {
+      "epoch": 0.511130384504192,
+      "grad_norm": 1.0391023612280914,
+      "learning_rate": 9.807332465966066e-06,
+      "loss": 1.6162,
+      "step": 1768
+    },
+    {
+      "epoch": 0.5114194854004047,
+      "grad_norm": 1.0538986164846234,
+      "learning_rate": 9.798159039275239e-06,
+      "loss": 1.6539,
+      "step": 1769
+    },
+    {
+      "epoch": 0.5117085862966175,
+      "grad_norm": 1.093595104973077,
+      "learning_rate": 9.788985782503243e-06,
+      "loss": 1.6175,
+      "step": 1770
+    },
+    {
+      "epoch": 0.5119976871928303,
+      "grad_norm": 1.1275132158110948,
+      "learning_rate": 9.779812703372538e-06,
+      "loss": 1.625,
+      "step": 1771
+    },
+    {
+      "epoch": 0.512286788089043,
+      "grad_norm": 1.043366569269239,
+      "learning_rate": 9.770639809605434e-06,
+      "loss": 1.5885,
+      "step": 1772
+    },
+    {
+      "epoch": 0.5125758889852559,
+      "grad_norm": 1.0472391225316844,
+      "learning_rate": 9.76146710892409e-06,
+      "loss": 1.4726,
+      "step": 1773
+    },
+    {
+      "epoch": 0.5128649898814687,
+      "grad_norm": 1.0626332832610101,
+      "learning_rate": 9.752294609050496e-06,
+      "loss": 1.4796,
+      "step": 1774
+    },
+    {
+      "epoch": 0.5131540907776814,
+      "grad_norm": 1.0641004990587684,
+      "learning_rate": 9.743122317706483e-06,
+      "loss": 1.6595,
+      "step": 1775
+    },
+    {
+      "epoch": 0.5134431916738942,
+      "grad_norm": 1.0704734853586566,
+      "learning_rate": 9.733950242613692e-06,
+      "loss": 1.5468,
+      "step": 1776
+    },
+    {
+      "epoch": 0.513732292570107,
+      "grad_norm": 1.022707148566278,
+      "learning_rate": 9.724778391493594e-06,
+      "loss": 1.5835,
+      "step": 1777
+    },
+    {
+      "epoch": 0.5140213934663197,
+      "grad_norm": 1.0572401294799512,
+      "learning_rate": 9.715606772067469e-06,
+      "loss": 1.6003,
+      "step": 1778
+    },
+    {
+      "epoch": 0.5143104943625325,
+      "grad_norm": 1.0630295086299928,
+      "learning_rate": 9.706435392056394e-06,
+      "loss": 1.5064,
+      "step": 1779
+    },
+    {
+      "epoch": 0.5145995952587453,
+      "grad_norm": 1.0448017550933613,
+      "learning_rate": 9.697264259181254e-06,
+      "loss": 1.5824,
+      "step": 1780
+    },
+    {
+      "epoch": 0.514888696154958,
+      "grad_norm": 1.0384395784785245,
+      "learning_rate": 9.688093381162726e-06,
+      "loss": 1.5373,
+      "step": 1781
+    },
+    {
+      "epoch": 0.5151777970511708,
+      "grad_norm": 1.060673008956403,
+      "learning_rate": 9.678922765721262e-06,
+      "loss": 1.5941,
+      "step": 1782
+    },
+    {
+      "epoch": 0.5154668979473837,
+      "grad_norm": 1.0583093942489816,
+      "learning_rate": 9.669752420577104e-06,
+      "loss": 1.6185,
+      "step": 1783
+    },
+    {
+      "epoch": 0.5157559988435965,
+      "grad_norm": 1.0610011200901694,
+      "learning_rate": 9.660582353450263e-06,
+      "loss": 1.5685,
+      "step": 1784
+    },
+    {
+      "epoch": 0.5160450997398092,
+      "grad_norm": 1.0588149003284895,
+      "learning_rate": 9.651412572060511e-06,
+      "loss": 1.5742,
+      "step": 1785
+    },
+    {
+      "epoch": 0.516334200636022,
+      "grad_norm": 1.0416721198608683,
+      "learning_rate": 9.642243084127388e-06,
+      "loss": 1.616,
+      "step": 1786
+    },
+    {
+      "epoch": 0.5166233015322348,
+      "grad_norm": 1.0143802566853162,
+      "learning_rate": 9.633073897370183e-06,
+      "loss": 1.5624,
+      "step": 1787
+    },
+    {
+      "epoch": 0.5169124024284475,
+      "grad_norm": 1.0598542456097253,
+      "learning_rate": 9.623905019507933e-06,
+      "loss": 1.5889,
+      "step": 1788
+    },
+    {
+      "epoch": 0.5172015033246603,
+      "grad_norm": 1.0846874746777628,
+      "learning_rate": 9.614736458259405e-06,
+      "loss": 1.4884,
+      "step": 1789
+    },
+    {
+      "epoch": 0.5174906042208731,
+      "grad_norm": 1.094555318865235,
+      "learning_rate": 9.605568221343115e-06,
+      "loss": 1.5693,
+      "step": 1790
+    },
+    {
+      "epoch": 0.5177797051170858,
+      "grad_norm": 1.035439783806554,
+      "learning_rate": 9.596400316477296e-06,
+      "loss": 1.5783,
+      "step": 1791
+    },
+    {
+      "epoch": 0.5180688060132986,
+      "grad_norm": 1.0795071236926523,
+      "learning_rate": 9.587232751379905e-06,
+      "loss": 1.5815,
+      "step": 1792
+    },
+    {
+      "epoch": 0.5183579069095114,
+      "grad_norm": 1.0263573356842672,
+      "learning_rate": 9.578065533768611e-06,
+      "loss": 1.585,
+      "step": 1793
+    },
+    {
+      "epoch": 0.5186470078057241,
+      "grad_norm": 1.0780919729406528,
+      "learning_rate": 9.568898671360791e-06,
+      "loss": 1.5207,
+      "step": 1794
+    },
+    {
+      "epoch": 0.518936108701937,
+      "grad_norm": 1.0405802902124344,
+      "learning_rate": 9.559732171873524e-06,
+      "loss": 1.6851,
+      "step": 1795
+    },
+    {
+      "epoch": 0.5192252095981498,
+      "grad_norm": 1.0546047645488434,
+      "learning_rate": 9.550566043023582e-06,
+      "loss": 1.5562,
+      "step": 1796
+    },
+    {
+      "epoch": 0.5195143104943626,
+      "grad_norm": 1.062453052212825,
+      "learning_rate": 9.541400292527428e-06,
+      "loss": 1.5297,
+      "step": 1797
+    },
+    {
+      "epoch": 0.5198034113905753,
+      "grad_norm": 1.1179476725169573,
+      "learning_rate": 9.532234928101205e-06,
+      "loss": 1.6953,
+      "step": 1798
+    },
+    {
+      "epoch": 0.5200925122867881,
+      "grad_norm": 1.086649952070679,
+      "learning_rate": 9.523069957460728e-06,
+      "loss": 1.561,
+      "step": 1799
+    },
+    {
+      "epoch": 0.5203816131830009,
+      "grad_norm": 1.0636689733975708,
+      "learning_rate": 9.513905388321485e-06,
+      "loss": 1.5703,
+      "step": 1800
+    },
+    {
+      "epoch": 0.5206707140792136,
+      "grad_norm": 1.102799053024275,
+      "learning_rate": 9.50474122839862e-06,
+      "loss": 1.5309,
+      "step": 1801
+    },
+    {
+      "epoch": 0.5209598149754264,
+      "grad_norm": 1.0040492423131084,
+      "learning_rate": 9.49557748540694e-06,
+      "loss": 1.5523,
+      "step": 1802
+    },
+    {
+      "epoch": 0.5212489158716392,
+      "grad_norm": 1.1178296865511104,
+      "learning_rate": 9.486414167060899e-06,
+      "loss": 1.5297,
+      "step": 1803
+    },
+    {
+      "epoch": 0.5215380167678519,
+      "grad_norm": 1.07004298175949,
+      "learning_rate": 9.477251281074586e-06,
+      "loss": 1.6225,
+      "step": 1804
+    },
+    {
+      "epoch": 0.5218271176640648,
+      "grad_norm": 1.042451604935498,
+      "learning_rate": 9.468088835161737e-06,
+      "loss": 1.6386,
+      "step": 1805
+    },
+    {
+      "epoch": 0.5221162185602776,
+      "grad_norm": 1.0511803993934579,
+      "learning_rate": 9.458926837035707e-06,
+      "loss": 1.5489,
+      "step": 1806
+    },
+    {
+      "epoch": 0.5224053194564903,
+      "grad_norm": 1.0297924134961818,
+      "learning_rate": 9.449765294409489e-06,
+      "loss": 1.648,
+      "step": 1807
+    },
+    {
+      "epoch": 0.5226944203527031,
+      "grad_norm": 1.0387231795659488,
+      "learning_rate": 9.440604214995673e-06,
+      "loss": 1.6004,
+      "step": 1808
+    },
+    {
+      "epoch": 0.5229835212489159,
+      "grad_norm": 1.0711941630280286,
+      "learning_rate": 9.431443606506475e-06,
+      "loss": 1.5854,
+      "step": 1809
+    },
+    {
+      "epoch": 0.5232726221451286,
+      "grad_norm": 1.028541126192078,
+      "learning_rate": 9.422283476653708e-06,
+      "loss": 1.5798,
+      "step": 1810
+    },
+    {
+      "epoch": 0.5235617230413414,
+      "grad_norm": 1.038279049187084,
+      "learning_rate": 9.41312383314878e-06,
+      "loss": 1.6557,
+      "step": 1811
+    },
+    {
+      "epoch": 0.5238508239375542,
+      "grad_norm": 1.028337631141712,
+      "learning_rate": 9.403964683702694e-06,
+      "loss": 1.5164,
+      "step": 1812
+    },
+    {
+      "epoch": 0.524139924833767,
+      "grad_norm": 1.0368782522066862,
+      "learning_rate": 9.394806036026037e-06,
+      "loss": 1.5377,
+      "step": 1813
+    },
+    {
+      "epoch": 0.5244290257299797,
+      "grad_norm": 1.0053184662150028,
+      "learning_rate": 9.385647897828968e-06,
+      "loss": 1.5433,
+      "step": 1814
+    },
+    {
+      "epoch": 0.5247181266261925,
+      "grad_norm": 1.0733960342131295,
+      "learning_rate": 9.376490276821225e-06,
+      "loss": 1.5997,
+      "step": 1815
+    },
+    {
+      "epoch": 0.5250072275224054,
+      "grad_norm": 1.0357357408238628,
+      "learning_rate": 9.367333180712104e-06,
+      "loss": 1.5989,
+      "step": 1816
+    },
+    {
+      "epoch": 0.5252963284186181,
+      "grad_norm": 1.0576891997008908,
+      "learning_rate": 9.35817661721046e-06,
+      "loss": 1.6155,
+      "step": 1817
+    },
+    {
+      "epoch": 0.5255854293148309,
+      "grad_norm": 1.0924226886771382,
+      "learning_rate": 9.349020594024705e-06,
+      "loss": 1.63,
+      "step": 1818
+    },
+    {
+      "epoch": 0.5258745302110437,
+      "grad_norm": 1.020216998725467,
+      "learning_rate": 9.339865118862791e-06,
+      "loss": 1.6428,
+      "step": 1819
+    },
+    {
+      "epoch": 0.5261636311072564,
+      "grad_norm": 1.0058140627007404,
+      "learning_rate": 9.330710199432214e-06,
+      "loss": 1.5178,
+      "step": 1820
+    },
+    {
+      "epoch": 0.5264527320034692,
+      "grad_norm": 1.0870776866818141,
+      "learning_rate": 9.321555843439994e-06,
+      "loss": 1.5567,
+      "step": 1821
+    },
+    {
+      "epoch": 0.526741832899682,
+      "grad_norm": 1.1051914783136976,
+      "learning_rate": 9.312402058592679e-06,
+      "loss": 1.6565,
+      "step": 1822
+    },
+    {
+      "epoch": 0.5270309337958947,
+      "grad_norm": 1.013803770580375,
+      "learning_rate": 9.303248852596345e-06,
+      "loss": 1.5426,
+      "step": 1823
+    },
+    {
+      "epoch": 0.5273200346921075,
+      "grad_norm": 1.0500781703061208,
+      "learning_rate": 9.294096233156566e-06,
+      "loss": 1.5784,
+      "step": 1824
+    },
+    {
+      "epoch": 0.5276091355883203,
+      "grad_norm": 1.0582023895860582,
+      "learning_rate": 9.284944207978439e-06,
+      "loss": 1.6611,
+      "step": 1825
+    },
+    {
+      "epoch": 0.5278982364845332,
+      "grad_norm": 1.075298277804485,
+      "learning_rate": 9.275792784766542e-06,
+      "loss": 1.5343,
+      "step": 1826
+    },
+    {
+      "epoch": 0.5281873373807459,
+      "grad_norm": 1.0384130773603029,
+      "learning_rate": 9.266641971224963e-06,
+      "loss": 1.5374,
+      "step": 1827
+    },
+    {
+      "epoch": 0.5284764382769587,
+      "grad_norm": 1.044902120232085,
+      "learning_rate": 9.257491775057264e-06,
+      "loss": 1.6471,
+      "step": 1828
+    },
+    {
+      "epoch": 0.5287655391731715,
+      "grad_norm": 1.031603118321479,
+      "learning_rate": 9.2483422039665e-06,
+      "loss": 1.5717,
+      "step": 1829
+    },
+    {
+      "epoch": 0.5290546400693842,
+      "grad_norm": 1.015466069348422,
+      "learning_rate": 9.239193265655191e-06,
+      "loss": 1.5384,
+      "step": 1830
+    },
+    {
+      "epoch": 0.529343740965597,
+      "grad_norm": 1.0468149788864793,
+      "learning_rate": 9.230044967825324e-06,
+      "loss": 1.6353,
+      "step": 1831
+    },
+    {
+      "epoch": 0.5296328418618098,
+      "grad_norm": 1.0350378737902792,
+      "learning_rate": 9.22089731817835e-06,
+      "loss": 1.6765,
+      "step": 1832
+    },
+    {
+      "epoch": 0.5299219427580225,
+      "grad_norm": 1.032304212684832,
+      "learning_rate": 9.211750324415173e-06,
+      "loss": 1.5547,
+      "step": 1833
+    },
+    {
+      "epoch": 0.5302110436542353,
+      "grad_norm": 1.062810271339594,
+      "learning_rate": 9.202603994236142e-06,
+      "loss": 1.5704,
+      "step": 1834
+    },
+    {
+      "epoch": 0.5305001445504481,
+      "grad_norm": 1.048501201195202,
+      "learning_rate": 9.193458335341058e-06,
+      "loss": 1.5655,
+      "step": 1835
+    },
+    {
+      "epoch": 0.5307892454466608,
+      "grad_norm": 1.0161011827887958,
+      "learning_rate": 9.184313355429141e-06,
+      "loss": 1.5682,
+      "step": 1836
+    },
+    {
+      "epoch": 0.5310783463428737,
+      "grad_norm": 1.0536878115204331,
+      "learning_rate": 9.175169062199054e-06,
+      "loss": 1.5526,
+      "step": 1837
+    },
+    {
+      "epoch": 0.5313674472390865,
+      "grad_norm": 1.0222917475435525,
+      "learning_rate": 9.16602546334887e-06,
+      "loss": 1.5799,
+      "step": 1838
+    },
+    {
+      "epoch": 0.5316565481352992,
+      "grad_norm": 1.0694403910171348,
+      "learning_rate": 9.156882566576093e-06,
+      "loss": 1.637,
+      "step": 1839
+    },
+    {
+      "epoch": 0.531945649031512,
+      "grad_norm": 1.012723962660485,
+      "learning_rate": 9.147740379577615e-06,
+      "loss": 1.4992,
+      "step": 1840
+    },
+    {
+      "epoch": 0.5322347499277248,
+      "grad_norm": 1.0948448873885346,
+      "learning_rate": 9.138598910049748e-06,
+      "loss": 1.6737,
+      "step": 1841
+    },
+    {
+      "epoch": 0.5325238508239376,
+      "grad_norm": 1.0411552195081186,
+      "learning_rate": 9.129458165688192e-06,
+      "loss": 1.5131,
+      "step": 1842
+    },
+    {
+      "epoch": 0.5328129517201503,
+      "grad_norm": 1.0354782390870787,
+      "learning_rate": 9.120318154188039e-06,
+      "loss": 1.5528,
+      "step": 1843
+    },
+    {
+      "epoch": 0.5331020526163631,
+      "grad_norm": 1.0312926804037754,
+      "learning_rate": 9.111178883243762e-06,
+      "loss": 1.6279,
+      "step": 1844
+    },
+    {
+      "epoch": 0.5333911535125759,
+      "grad_norm": 1.063477800410858,
+      "learning_rate": 9.102040360549211e-06,
+      "loss": 1.5809,
+      "step": 1845
+    },
+    {
+      "epoch": 0.5336802544087886,
+      "grad_norm": 1.045053643373545,
+      "learning_rate": 9.092902593797607e-06,
+      "loss": 1.4449,
+      "step": 1846
+    },
+    {
+      "epoch": 0.5339693553050014,
+      "grad_norm": 1.0434408444594696,
+      "learning_rate": 9.083765590681538e-06,
+      "loss": 1.5653,
+      "step": 1847
+    },
+    {
+      "epoch": 0.5342584562012143,
+      "grad_norm": 1.0887490627748158,
+      "learning_rate": 9.07462935889294e-06,
+      "loss": 1.659,
+      "step": 1848
+    },
+    {
+      "epoch": 0.534547557097427,
+      "grad_norm": 1.0643171354261491,
+      "learning_rate": 9.065493906123105e-06,
+      "loss": 1.5933,
+      "step": 1849
+    },
+    {
+      "epoch": 0.5348366579936398,
+      "grad_norm": 1.0162513597173648,
+      "learning_rate": 9.056359240062676e-06,
+      "loss": 1.5428,
+      "step": 1850
+    },
+    {
+      "epoch": 0.5351257588898526,
+      "grad_norm": 1.0694375460230756,
+      "learning_rate": 9.047225368401622e-06,
+      "loss": 1.5763,
+      "step": 1851
+    },
+    {
+      "epoch": 0.5354148597860653,
+      "grad_norm": 1.008672036890399,
+      "learning_rate": 9.038092298829253e-06,
+      "loss": 1.54,
+      "step": 1852
+    },
+    {
+      "epoch": 0.5357039606822781,
+      "grad_norm": 1.0389865129553266,
+      "learning_rate": 9.028960039034194e-06,
+      "loss": 1.5908,
+      "step": 1853
+    },
+    {
+      "epoch": 0.5359930615784909,
+      "grad_norm": 1.1190225223061014,
+      "learning_rate": 9.019828596704394e-06,
+      "loss": 1.6748,
+      "step": 1854
+    },
+    {
+      "epoch": 0.5362821624747036,
+      "grad_norm": 1.071648331417543,
+      "learning_rate": 9.010697979527118e-06,
+      "loss": 1.4834,
+      "step": 1855
+    },
+    {
+      "epoch": 0.5365712633709164,
+      "grad_norm": 1.0484164927771769,
+      "learning_rate": 9.001568195188925e-06,
+      "loss": 1.6085,
+      "step": 1856
+    },
+    {
+      "epoch": 0.5368603642671292,
+      "grad_norm": 1.0810152414924428,
+      "learning_rate": 8.992439251375686e-06,
+      "loss": 1.5757,
+      "step": 1857
+    },
+    {
+      "epoch": 0.537149465163342,
+      "grad_norm": 1.0711571574852117,
+      "learning_rate": 8.983311155772552e-06,
+      "loss": 1.5078,
+      "step": 1858
+    },
+    {
+      "epoch": 0.5374385660595548,
+      "grad_norm": 1.058169838113003,
+      "learning_rate": 8.974183916063967e-06,
+      "loss": 1.5907,
+      "step": 1859
+    },
+    {
+      "epoch": 0.5377276669557676,
+      "grad_norm": 1.0045985632957348,
+      "learning_rate": 8.965057539933653e-06,
+      "loss": 1.5722,
+      "step": 1860
+    },
+    {
+      "epoch": 0.5380167678519804,
+      "grad_norm": 1.05027426073653,
+      "learning_rate": 8.95593203506461e-06,
+      "loss": 1.5971,
+      "step": 1861
+    },
+    {
+      "epoch": 0.5383058687481931,
+      "grad_norm": 1.0400951671530732,
+      "learning_rate": 8.946807409139099e-06,
+      "loss": 1.4927,
+      "step": 1862
+    },
+    {
+      "epoch": 0.5385949696444059,
+      "grad_norm": 1.0845982085338564,
+      "learning_rate": 8.937683669838636e-06,
+      "loss": 1.5772,
+      "step": 1863
+    },
+    {
+      "epoch": 0.5388840705406187,
+      "grad_norm": 1.092832554768011,
+      "learning_rate": 8.928560824844003e-06,
+      "loss": 1.6774,
+      "step": 1864
+    },
+    {
+      "epoch": 0.5391731714368314,
+      "grad_norm": 1.097409465405629,
+      "learning_rate": 8.919438881835217e-06,
+      "loss": 1.5756,
+      "step": 1865
+    },
+    {
+      "epoch": 0.5394622723330442,
+      "grad_norm": 1.1135616686297214,
+      "learning_rate": 8.910317848491545e-06,
+      "loss": 1.6092,
+      "step": 1866
+    },
+    {
+      "epoch": 0.539751373229257,
+      "grad_norm": 1.0629539882867265,
+      "learning_rate": 8.901197732491486e-06,
+      "loss": 1.5739,
+      "step": 1867
+    },
+    {
+      "epoch": 0.5400404741254697,
+      "grad_norm": 1.1011417173783966,
+      "learning_rate": 8.892078541512758e-06,
+      "loss": 1.5833,
+      "step": 1868
+    },
+    {
+      "epoch": 0.5403295750216826,
+      "grad_norm": 1.043383500998138,
+      "learning_rate": 8.882960283232315e-06,
+      "loss": 1.5533,
+      "step": 1869
+    },
+    {
+      "epoch": 0.5406186759178954,
+      "grad_norm": 1.1582137604085936,
+      "learning_rate": 8.873842965326316e-06,
+      "loss": 1.5789,
+      "step": 1870
+    },
+    {
+      "epoch": 0.5409077768141082,
+      "grad_norm": 1.0706372308004994,
+      "learning_rate": 8.864726595470126e-06,
+      "loss": 1.6032,
+      "step": 1871
+    },
+    {
+      "epoch": 0.5411968777103209,
+      "grad_norm": 1.0929319074400397,
+      "learning_rate": 8.855611181338323e-06,
+      "loss": 1.4867,
+      "step": 1872
+    },
+    {
+      "epoch": 0.5414859786065337,
+      "grad_norm": 1.052649765474697,
+      "learning_rate": 8.846496730604672e-06,
+      "loss": 1.5104,
+      "step": 1873
+    },
+    {
+      "epoch": 0.5417750795027465,
+      "grad_norm": 1.1069093642908276,
+      "learning_rate": 8.83738325094213e-06,
+      "loss": 1.6257,
+      "step": 1874
+    },
+    {
+      "epoch": 0.5420641803989592,
+      "grad_norm": 1.0697610216966835,
+      "learning_rate": 8.828270750022832e-06,
+      "loss": 1.5467,
+      "step": 1875
+    },
+    {
+      "epoch": 0.542353281295172,
+      "grad_norm": 1.0437650036942958,
+      "learning_rate": 8.819159235518095e-06,
+      "loss": 1.5564,
+      "step": 1876
+    },
+    {
+      "epoch": 0.5426423821913848,
+      "grad_norm": 1.031078898238742,
+      "learning_rate": 8.810048715098404e-06,
+      "loss": 1.6038,
+      "step": 1877
+    },
+    {
+      "epoch": 0.5429314830875975,
+      "grad_norm": 1.051957143482446,
+      "learning_rate": 8.800939196433404e-06,
+      "loss": 1.5693,
+      "step": 1878
+    },
+    {
+      "epoch": 0.5432205839838103,
+      "grad_norm": 1.0512174929867104,
+      "learning_rate": 8.791830687191902e-06,
+      "loss": 1.6365,
+      "step": 1879
+    },
+    {
+      "epoch": 0.5435096848800232,
+      "grad_norm": 1.0572864729439442,
+      "learning_rate": 8.782723195041846e-06,
+      "loss": 1.5679,
+      "step": 1880
+    },
+    {
+      "epoch": 0.5437987857762359,
+      "grad_norm": 1.0520213325941057,
+      "learning_rate": 8.773616727650339e-06,
+      "loss": 1.5948,
+      "step": 1881
+    },
+    {
+      "epoch": 0.5440878866724487,
+      "grad_norm": 1.0566252890590055,
+      "learning_rate": 8.764511292683614e-06,
+      "loss": 1.581,
+      "step": 1882
+    },
+    {
+      "epoch": 0.5443769875686615,
+      "grad_norm": 1.0230286252812237,
+      "learning_rate": 8.755406897807039e-06,
+      "loss": 1.5128,
+      "step": 1883
+    },
+    {
+      "epoch": 0.5446660884648742,
+      "grad_norm": 1.1146581133875821,
+      "learning_rate": 8.746303550685106e-06,
+      "loss": 1.6534,
+      "step": 1884
+    },
+    {
+      "epoch": 0.544955189361087,
+      "grad_norm": 1.0496978249756113,
+      "learning_rate": 8.737201258981417e-06,
+      "loss": 1.5549,
+      "step": 1885
+    },
+    {
+      "epoch": 0.5452442902572998,
+      "grad_norm": 1.0254947555463505,
+      "learning_rate": 8.728100030358701e-06,
+      "loss": 1.7059,
+      "step": 1886
+    },
+    {
+      "epoch": 0.5455333911535126,
+      "grad_norm": 1.0310833440172291,
+      "learning_rate": 8.718999872478775e-06,
+      "loss": 1.561,
+      "step": 1887
+    },
+    {
+      "epoch": 0.5458224920497253,
+      "grad_norm": 1.0722377867940174,
+      "learning_rate": 8.709900793002563e-06,
+      "loss": 1.5517,
+      "step": 1888
+    },
+    {
+      "epoch": 0.5461115929459381,
+      "grad_norm": 1.0220824461113878,
+      "learning_rate": 8.700802799590088e-06,
+      "loss": 1.4631,
+      "step": 1889
+    },
+    {
+      "epoch": 0.546400693842151,
+      "grad_norm": 1.0505873661536551,
+      "learning_rate": 8.691705899900445e-06,
+      "loss": 1.6354,
+      "step": 1890
+    },
+    {
+      "epoch": 0.5466897947383637,
+      "grad_norm": 1.1359980557606923,
+      "learning_rate": 8.682610101591813e-06,
+      "loss": 1.6107,
+      "step": 1891
+    },
+    {
+      "epoch": 0.5469788956345765,
+      "grad_norm": 1.1026471656094516,
+      "learning_rate": 8.67351541232145e-06,
+      "loss": 1.5594,
+      "step": 1892
+    },
+    {
+      "epoch": 0.5472679965307893,
+      "grad_norm": 1.0449410278292073,
+      "learning_rate": 8.664421839745678e-06,
+      "loss": 1.5034,
+      "step": 1893
+    },
+    {
+      "epoch": 0.547557097427002,
+      "grad_norm": 1.030330449635608,
+      "learning_rate": 8.655329391519871e-06,
+      "loss": 1.5489,
+      "step": 1894
+    },
+    {
+      "epoch": 0.5478461983232148,
+      "grad_norm": 1.0215095749557113,
+      "learning_rate": 8.646238075298464e-06,
+      "loss": 1.6313,
+      "step": 1895
+    },
+    {
+      "epoch": 0.5481352992194276,
+      "grad_norm": 1.0795910665812256,
+      "learning_rate": 8.637147898734941e-06,
+      "loss": 1.5269,
+      "step": 1896
+    },
+    {
+      "epoch": 0.5484244001156403,
+      "grad_norm": 1.0367138800877507,
+      "learning_rate": 8.628058869481817e-06,
+      "loss": 1.5723,
+      "step": 1897
+    },
+    {
+      "epoch": 0.5487135010118531,
+      "grad_norm": 1.071020440304501,
+      "learning_rate": 8.61897099519065e-06,
+      "loss": 1.4967,
+      "step": 1898
+    },
+    {
+      "epoch": 0.5490026019080659,
+      "grad_norm": 1.0954202953254553,
+      "learning_rate": 8.609884283512024e-06,
+      "loss": 1.5833,
+      "step": 1899
+    },
+    {
+      "epoch": 0.5492917028042787,
+      "grad_norm": 1.0715762816413257,
+      "learning_rate": 8.60079874209554e-06,
+      "loss": 1.6525,
+      "step": 1900
+    },
+    {
+      "epoch": 0.5495808037004914,
+      "grad_norm": 1.037142198908277,
+      "learning_rate": 8.59171437858982e-06,
+      "loss": 1.5432,
+      "step": 1901
+    },
+    {
+      "epoch": 0.5498699045967043,
+      "grad_norm": 1.029454308876277,
+      "learning_rate": 8.58263120064249e-06,
+      "loss": 1.5822,
+      "step": 1902
+    },
+    {
+      "epoch": 0.5501590054929171,
+      "grad_norm": 1.0452906564592603,
+      "learning_rate": 8.573549215900173e-06,
+      "loss": 1.5034,
+      "step": 1903
+    },
+    {
+      "epoch": 0.5504481063891298,
+      "grad_norm": 1.0539409616389446,
+      "learning_rate": 8.564468432008502e-06,
+      "loss": 1.6563,
+      "step": 1904
+    },
+    {
+      "epoch": 0.5507372072853426,
+      "grad_norm": 1.062727425669123,
+      "learning_rate": 8.555388856612089e-06,
+      "loss": 1.5339,
+      "step": 1905
+    },
+    {
+      "epoch": 0.5510263081815554,
+      "grad_norm": 1.030684199512718,
+      "learning_rate": 8.54631049735453e-06,
+      "loss": 1.5321,
+      "step": 1906
+    },
+    {
+      "epoch": 0.5513154090777681,
+      "grad_norm": 1.0479535508067723,
+      "learning_rate": 8.537233361878395e-06,
+      "loss": 1.4625,
+      "step": 1907
+    },
+    {
+      "epoch": 0.5516045099739809,
+      "grad_norm": 1.0519306558195942,
+      "learning_rate": 8.528157457825226e-06,
+      "loss": 1.5735,
+      "step": 1908
+    },
+    {
+      "epoch": 0.5518936108701937,
+      "grad_norm": 1.057684473855353,
+      "learning_rate": 8.519082792835534e-06,
+      "loss": 1.5486,
+      "step": 1909
+    },
+    {
+      "epoch": 0.5521827117664064,
+      "grad_norm": 1.1392942098438261,
+      "learning_rate": 8.510009374548777e-06,
+      "loss": 1.5855,
+      "step": 1910
+    },
+    {
+      "epoch": 0.5524718126626192,
+      "grad_norm": 1.177323036330576,
+      "learning_rate": 8.50093721060337e-06,
+      "loss": 1.5244,
+      "step": 1911
+    },
+    {
+      "epoch": 0.5527609135588321,
+      "grad_norm": 1.0354093609075838,
+      "learning_rate": 8.491866308636667e-06,
+      "loss": 1.463,
+      "step": 1912
+    },
+    {
+      "epoch": 0.5530500144550448,
+      "grad_norm": 1.0692782433141481,
+      "learning_rate": 8.482796676284963e-06,
+      "loss": 1.5936,
+      "step": 1913
+    },
+    {
+      "epoch": 0.5533391153512576,
+      "grad_norm": 1.0435163197392727,
+      "learning_rate": 8.473728321183486e-06,
+      "loss": 1.6304,
+      "step": 1914
+    },
+    {
+      "epoch": 0.5536282162474704,
+      "grad_norm": 1.0337426046327096,
+      "learning_rate": 8.464661250966388e-06,
+      "loss": 1.5662,
+      "step": 1915
+    },
+    {
+      "epoch": 0.5539173171436832,
+      "grad_norm": 1.0703771270160636,
+      "learning_rate": 8.455595473266736e-06,
+      "loss": 1.57,
+      "step": 1916
+    },
+    {
+      "epoch": 0.5542064180398959,
+      "grad_norm": 1.0300120361216238,
+      "learning_rate": 8.44653099571651e-06,
+      "loss": 1.631,
+      "step": 1917
+    },
+    {
+      "epoch": 0.5544955189361087,
+      "grad_norm": 1.0652782937103356,
+      "learning_rate": 8.437467825946599e-06,
+      "loss": 1.7476,
+      "step": 1918
+    },
+    {
+      "epoch": 0.5547846198323215,
+      "grad_norm": 1.031332243048047,
+      "learning_rate": 8.428405971586783e-06,
+      "loss": 1.5216,
+      "step": 1919
+    },
+    {
+      "epoch": 0.5550737207285342,
+      "grad_norm": 1.1872978937847811,
+      "learning_rate": 8.419345440265744e-06,
+      "loss": 1.5472,
+      "step": 1920
+    },
+    {
+      "epoch": 0.555362821624747,
+      "grad_norm": 1.0860697300690225,
+      "learning_rate": 8.410286239611048e-06,
+      "loss": 1.4993,
+      "step": 1921
+    },
+    {
+      "epoch": 0.5556519225209599,
+      "grad_norm": 1.1079631833389385,
+      "learning_rate": 8.401228377249134e-06,
+      "loss": 1.5462,
+      "step": 1922
+    },
+    {
+      "epoch": 0.5559410234171726,
+      "grad_norm": 1.043717478340039,
+      "learning_rate": 8.39217186080532e-06,
+      "loss": 1.605,
+      "step": 1923
+    },
+    {
+      "epoch": 0.5562301243133854,
+      "grad_norm": 1.0835118920387927,
+      "learning_rate": 8.383116697903791e-06,
+      "loss": 1.6688,
+      "step": 1924
+    },
+    {
+      "epoch": 0.5565192252095982,
+      "grad_norm": 1.0705831254099958,
+      "learning_rate": 8.374062896167597e-06,
+      "loss": 1.636,
+      "step": 1925
+    },
+    {
+      "epoch": 0.5568083261058109,
+      "grad_norm": 1.0919009362091248,
+      "learning_rate": 8.365010463218628e-06,
+      "loss": 1.4397,
+      "step": 1926
+    },
+    {
+      "epoch": 0.5570974270020237,
+      "grad_norm": 1.0410701910777804,
+      "learning_rate": 8.355959406677639e-06,
+      "loss": 1.5403,
+      "step": 1927
+    },
+    {
+      "epoch": 0.5573865278982365,
+      "grad_norm": 1.043559502606014,
+      "learning_rate": 8.346909734164214e-06,
+      "loss": 1.6419,
+      "step": 1928
+    },
+    {
+      "epoch": 0.5576756287944492,
+      "grad_norm": 1.088305429593252,
+      "learning_rate": 8.337861453296773e-06,
+      "loss": 1.5822,
+      "step": 1929
+    },
+    {
+      "epoch": 0.557964729690662,
+      "grad_norm": 1.0829470219652828,
+      "learning_rate": 8.32881457169257e-06,
+      "loss": 1.5701,
+      "step": 1930
+    },
+    {
+      "epoch": 0.5582538305868748,
+      "grad_norm": 1.0213951976162692,
+      "learning_rate": 8.319769096967681e-06,
+      "loss": 1.5171,
+      "step": 1931
+    },
+    {
+      "epoch": 0.5585429314830876,
+      "grad_norm": 1.014367499801957,
+      "learning_rate": 8.310725036736988e-06,
+      "loss": 1.5532,
+      "step": 1932
+    },
+    {
+      "epoch": 0.5588320323793003,
+      "grad_norm": 1.0606529257465926,
+      "learning_rate": 8.301682398614194e-06,
+      "loss": 1.5224,
+      "step": 1933
+    },
+    {
+      "epoch": 0.5591211332755132,
+      "grad_norm": 1.0173171716241347,
+      "learning_rate": 8.292641190211802e-06,
+      "loss": 1.536,
+      "step": 1934
+    },
+    {
+      "epoch": 0.559410234171726,
+      "grad_norm": 1.0016158736003675,
+      "learning_rate": 8.2836014191411e-06,
+      "loss": 1.4436,
+      "step": 1935
+    },
+    {
+      "epoch": 0.5596993350679387,
+      "grad_norm": 1.082732124789314,
+      "learning_rate": 8.274563093012184e-06,
+      "loss": 1.5456,
+      "step": 1936
+    },
+    {
+      "epoch": 0.5599884359641515,
+      "grad_norm": 1.0167290635529262,
+      "learning_rate": 8.265526219433923e-06,
+      "loss": 1.6223,
+      "step": 1937
+    },
+    {
+      "epoch": 0.5602775368603643,
+      "grad_norm": 1.0733194052819095,
+      "learning_rate": 8.256490806013965e-06,
+      "loss": 1.576,
+      "step": 1938
+    },
+    {
+      "epoch": 0.560566637756577,
+      "grad_norm": 1.0205253943615036,
+      "learning_rate": 8.247456860358725e-06,
+      "loss": 1.595,
+      "step": 1939
+    },
+    {
+      "epoch": 0.5608557386527898,
+      "grad_norm": 1.0419411503759481,
+      "learning_rate": 8.23842439007339e-06,
+      "loss": 1.5736,
+      "step": 1940
+    },
+    {
+      "epoch": 0.5611448395490026,
+      "grad_norm": 1.068964807431337,
+      "learning_rate": 8.229393402761902e-06,
+      "loss": 1.5516,
+      "step": 1941
+    },
+    {
+      "epoch": 0.5614339404452153,
+      "grad_norm": 1.403758242188754,
+      "learning_rate": 8.22036390602695e-06,
+      "loss": 1.5815,
+      "step": 1942
+    },
+    {
+      "epoch": 0.5617230413414281,
+      "grad_norm": 1.0201136890006297,
+      "learning_rate": 8.211335907469976e-06,
+      "loss": 1.5186,
+      "step": 1943
+    },
+    {
+      "epoch": 0.562012142237641,
+      "grad_norm": 1.0311704857810622,
+      "learning_rate": 8.20230941469115e-06,
+      "loss": 1.5096,
+      "step": 1944
+    },
+    {
+      "epoch": 0.5623012431338538,
+      "grad_norm": 1.0409221945444354,
+      "learning_rate": 8.193284435289384e-06,
+      "loss": 1.5683,
+      "step": 1945
+    },
+    {
+      "epoch": 0.5625903440300665,
+      "grad_norm": 1.0516798428015424,
+      "learning_rate": 8.184260976862315e-06,
+      "loss": 1.6381,
+      "step": 1946
+    },
+    {
+      "epoch": 0.5628794449262793,
+      "grad_norm": 1.0334051732711151,
+      "learning_rate": 8.175239047006298e-06,
+      "loss": 1.5846,
+      "step": 1947
+    },
+    {
+      "epoch": 0.5631685458224921,
+      "grad_norm": 1.0432625400705864,
+      "learning_rate": 8.166218653316401e-06,
+      "loss": 1.5435,
+      "step": 1948
+    },
+    {
+      "epoch": 0.5634576467187048,
+      "grad_norm": 1.0069078845713315,
+      "learning_rate": 8.157199803386396e-06,
+      "loss": 1.5136,
+      "step": 1949
+    },
+    {
+      "epoch": 0.5637467476149176,
+      "grad_norm": 1.0227073691734319,
+      "learning_rate": 8.14818250480876e-06,
+      "loss": 1.5295,
+      "step": 1950
+    },
+    {
+      "epoch": 0.5640358485111304,
+      "grad_norm": 1.007144621633634,
+      "learning_rate": 8.139166765174657e-06,
+      "loss": 1.5809,
+      "step": 1951
+    },
+    {
+      "epoch": 0.5643249494073431,
+      "grad_norm": 1.0245837603008965,
+      "learning_rate": 8.130152592073947e-06,
+      "loss": 1.5368,
+      "step": 1952
+    },
+    {
+      "epoch": 0.5646140503035559,
+      "grad_norm": 1.053806464399692,
+      "learning_rate": 8.12113999309517e-06,
+      "loss": 1.6457,
+      "step": 1953
+    },
+    {
+      "epoch": 0.5649031511997687,
+      "grad_norm": 1.012076951001509,
+      "learning_rate": 8.112128975825535e-06,
+      "loss": 1.5668,
+      "step": 1954
+    },
+    {
+      "epoch": 0.5651922520959815,
+      "grad_norm": 1.088739641040976,
+      "learning_rate": 8.103119547850924e-06,
+      "loss": 1.5508,
+      "step": 1955
+    },
+    {
+      "epoch": 0.5654813529921943,
+      "grad_norm": 1.065035850650176,
+      "learning_rate": 8.094111716755878e-06,
+      "loss": 1.5083,
+      "step": 1956
+    },
+    {
+      "epoch": 0.5657704538884071,
+      "grad_norm": 1.0267896811252029,
+      "learning_rate": 8.085105490123603e-06,
+      "loss": 1.6599,
+      "step": 1957
+    },
+    {
+      "epoch": 0.5660595547846198,
+      "grad_norm": 1.0449827874853526,
+      "learning_rate": 8.07610087553594e-06,
+      "loss": 1.5941,
+      "step": 1958
+    },
+    {
+      "epoch": 0.5663486556808326,
+      "grad_norm": 1.0163938307676907,
+      "learning_rate": 8.06709788057338e-06,
+      "loss": 1.6274,
+      "step": 1959
+    },
+    {
+      "epoch": 0.5666377565770454,
+      "grad_norm": 1.0319501641130382,
+      "learning_rate": 8.058096512815055e-06,
+      "loss": 1.5715,
+      "step": 1960
+    },
+    {
+      "epoch": 0.5669268574732582,
+      "grad_norm": 1.0759765391966305,
+      "learning_rate": 8.04909677983872e-06,
+      "loss": 1.5852,
+      "step": 1961
+    },
+    {
+      "epoch": 0.5672159583694709,
+      "grad_norm": 1.006556253184911,
+      "learning_rate": 8.040098689220754e-06,
+      "loss": 1.5102,
+      "step": 1962
+    },
+    {
+      "epoch": 0.5675050592656837,
+      "grad_norm": 1.0718842695744195,
+      "learning_rate": 8.03110224853616e-06,
+      "loss": 1.5632,
+      "step": 1963
+    },
+    {
+      "epoch": 0.5677941601618965,
+      "grad_norm": 1.0244702702295507,
+      "learning_rate": 8.022107465358546e-06,
+      "loss": 1.6309,
+      "step": 1964
+    },
+    {
+      "epoch": 0.5680832610581092,
+      "grad_norm": 1.1016438396635402,
+      "learning_rate": 8.013114347260123e-06,
+      "loss": 1.5104,
+      "step": 1965
+    },
+    {
+      "epoch": 0.5683723619543221,
+      "grad_norm": 1.1105155688219368,
+      "learning_rate": 8.004122901811712e-06,
+      "loss": 1.5597,
+      "step": 1966
+    },
+    {
+      "epoch": 0.5686614628505349,
+      "grad_norm": 1.0277073589635268,
+      "learning_rate": 7.995133136582707e-06,
+      "loss": 1.6471,
+      "step": 1967
+    },
+    {
+      "epoch": 0.5689505637467476,
+      "grad_norm": 1.0501616924995187,
+      "learning_rate": 7.986145059141105e-06,
+      "loss": 1.5752,
+      "step": 1968
+    },
+    {
+      "epoch": 0.5692396646429604,
+      "grad_norm": 1.0178557692313934,
+      "learning_rate": 7.977158677053474e-06,
+      "loss": 1.6291,
+      "step": 1969
+    },
+    {
+      "epoch": 0.5695287655391732,
+      "grad_norm": 19.769701162628678,
+      "learning_rate": 7.968173997884959e-06,
+      "loss": 1.5988,
+      "step": 1970
+    },
+    {
+      "epoch": 0.5698178664353859,
+      "grad_norm": 1.118496109621781,
+      "learning_rate": 7.95919102919926e-06,
+      "loss": 1.5804,
+      "step": 1971
+    },
+    {
+      "epoch": 0.5701069673315987,
+      "grad_norm": 1.0267514682577998,
+      "learning_rate": 7.950209778558653e-06,
+      "loss": 1.5376,
+      "step": 1972
+    },
+    {
+      "epoch": 0.5703960682278115,
+      "grad_norm": 1.080345159282818,
+      "learning_rate": 7.94123025352396e-06,
+      "loss": 1.5562,
+      "step": 1973
+    },
+    {
+      "epoch": 0.5706851691240243,
+      "grad_norm": 1.014121815420302,
+      "learning_rate": 7.932252461654544e-06,
+      "loss": 1.5893,
+      "step": 1974
+    },
+    {
+      "epoch": 0.570974270020237,
+      "grad_norm": 1.0408402100878356,
+      "learning_rate": 7.923276410508323e-06,
+      "loss": 1.5496,
+      "step": 1975
+    },
+    {
+      "epoch": 0.5712633709164499,
+      "grad_norm": 1.0614729668663685,
+      "learning_rate": 7.914302107641737e-06,
+      "loss": 1.5891,
+      "step": 1976
+    },
+    {
+      "epoch": 0.5715524718126627,
+      "grad_norm": 1.0507221024649747,
+      "learning_rate": 7.905329560609759e-06,
+      "loss": 1.5388,
+      "step": 1977
+    },
+    {
+      "epoch": 0.5718415727088754,
+      "grad_norm": 1.069136192799548,
+      "learning_rate": 7.896358776965882e-06,
+      "loss": 1.5993,
+      "step": 1978
+    },
+    {
+      "epoch": 0.5721306736050882,
+      "grad_norm": 1.031503377526069,
+      "learning_rate": 7.887389764262123e-06,
+      "loss": 1.6032,
+      "step": 1979
+    },
+    {
+      "epoch": 0.572419774501301,
+      "grad_norm": 1.0155418737802535,
+      "learning_rate": 7.878422530049001e-06,
+      "loss": 1.4378,
+      "step": 1980
+    },
+    {
+      "epoch": 0.5727088753975137,
+      "grad_norm": 1.003844854616889,
+      "learning_rate": 7.869457081875533e-06,
+      "loss": 1.5057,
+      "step": 1981
+    },
+    {
+      "epoch": 0.5729979762937265,
+      "grad_norm": 1.032500247027364,
+      "learning_rate": 7.860493427289246e-06,
+      "loss": 1.6058,
+      "step": 1982
+    },
+    {
+      "epoch": 0.5732870771899393,
+      "grad_norm": 1.0579281948970534,
+      "learning_rate": 7.851531573836139e-06,
+      "loss": 1.55,
+      "step": 1983
+    },
+    {
+      "epoch": 0.573576178086152,
+      "grad_norm": 1.0336865358378893,
+      "learning_rate": 7.84257152906071e-06,
+      "loss": 1.5947,
+      "step": 1984
+    },
+    {
+      "epoch": 0.5738652789823648,
+      "grad_norm": 1.022691929430556,
+      "learning_rate": 7.83361330050593e-06,
+      "loss": 1.5701,
+      "step": 1985
+    },
+    {
+      "epoch": 0.5741543798785776,
+      "grad_norm": 1.050577644782535,
+      "learning_rate": 7.824656895713238e-06,
+      "loss": 1.576,
+      "step": 1986
+    },
+    {
+      "epoch": 0.5744434807747903,
+      "grad_norm": 1.0353817137840815,
+      "learning_rate": 7.815702322222539e-06,
+      "loss": 1.5722,
+      "step": 1987
+    },
+    {
+      "epoch": 0.5747325816710032,
+      "grad_norm": 1.0431597870619023,
+      "learning_rate": 7.806749587572195e-06,
+      "loss": 1.631,
+      "step": 1988
+    },
+    {
+      "epoch": 0.575021682567216,
+      "grad_norm": 1.090291540562683,
+      "learning_rate": 7.797798699299034e-06,
+      "loss": 1.5737,
+      "step": 1989
+    },
+    {
+      "epoch": 0.5753107834634288,
+      "grad_norm": 1.381999333014581,
+      "learning_rate": 7.788849664938302e-06,
+      "loss": 1.5452,
+      "step": 1990
+    },
+    {
+      "epoch": 0.5755998843596415,
+      "grad_norm": 1.082687889971779,
+      "learning_rate": 7.77990249202371e-06,
+      "loss": 1.5955,
+      "step": 1991
+    },
+    {
+      "epoch": 0.5758889852558543,
+      "grad_norm": 1.0576076927971625,
+      "learning_rate": 7.770957188087391e-06,
+      "loss": 1.5846,
+      "step": 1992
+    },
+    {
+      "epoch": 0.5761780861520671,
+      "grad_norm": 1.0339741136387337,
+      "learning_rate": 7.762013760659902e-06,
+      "loss": 1.5303,
+      "step": 1993
+    },
+    {
+      "epoch": 0.5764671870482798,
+      "grad_norm": 1.020525142563875,
+      "learning_rate": 7.753072217270223e-06,
+      "loss": 1.5276,
+      "step": 1994
+    },
+    {
+      "epoch": 0.5767562879444926,
+      "grad_norm": 1.1067702665348997,
+      "learning_rate": 7.744132565445755e-06,
+      "loss": 1.5618,
+      "step": 1995
+    },
+    {
+      "epoch": 0.5770453888407054,
+      "grad_norm": 1.0741249996088822,
+      "learning_rate": 7.735194812712294e-06,
+      "loss": 1.5309,
+      "step": 1996
+    },
+    {
+      "epoch": 0.5773344897369181,
+      "grad_norm": 1.0689237612057823,
+      "learning_rate": 7.726258966594046e-06,
+      "loss": 1.5289,
+      "step": 1997
+    },
+    {
+      "epoch": 0.577623590633131,
+      "grad_norm": 1.0518310185029454,
+      "learning_rate": 7.717325034613612e-06,
+      "loss": 1.5855,
+      "step": 1998
+    },
+    {
+      "epoch": 0.5779126915293438,
+      "grad_norm": 1.0414586145745626,
+      "learning_rate": 7.708393024291971e-06,
+      "loss": 1.4916,
+      "step": 1999
+    },
+    {
+      "epoch": 0.5782017924255565,
+      "grad_norm": 1.0586468524067412,
+      "learning_rate": 7.699462943148502e-06,
+      "loss": 1.6108,
+      "step": 2000
+    },
+    {
+      "epoch": 0.5784908933217693,
+      "grad_norm": 1.0671608576954175,
+      "learning_rate": 7.690534798700946e-06,
+      "loss": 1.582,
+      "step": 2001
+    },
+    {
+      "epoch": 0.5787799942179821,
+      "grad_norm": 1.007831397654129,
+      "learning_rate": 7.68160859846542e-06,
+      "loss": 1.5368,
+      "step": 2002
+    },
+    {
+      "epoch": 0.5790690951141949,
+      "grad_norm": 1.027058366441966,
+      "learning_rate": 7.672684349956399e-06,
+      "loss": 1.5801,
+      "step": 2003
+    },
+    {
+      "epoch": 0.5793581960104076,
+      "grad_norm": 1.0489543178700635,
+      "learning_rate": 7.663762060686718e-06,
+      "loss": 1.5527,
+      "step": 2004
+    },
+    {
+      "epoch": 0.5796472969066204,
+      "grad_norm": 1.0691464938854618,
+      "learning_rate": 7.654841738167567e-06,
+      "loss": 1.5547,
+      "step": 2005
+    },
+    {
+      "epoch": 0.5799363978028332,
+      "grad_norm": 1.0802509712208035,
+      "learning_rate": 7.645923389908473e-06,
+      "loss": 1.6196,
+      "step": 2006
+    },
+    {
+      "epoch": 0.5802254986990459,
+      "grad_norm": 1.029539836246475,
+      "learning_rate": 7.637007023417304e-06,
+      "loss": 1.6457,
+      "step": 2007
+    },
+    {
+      "epoch": 0.5805145995952588,
+      "grad_norm": 1.0343058194415873,
+      "learning_rate": 7.62809264620026e-06,
+      "loss": 1.5931,
+      "step": 2008
+    },
+    {
+      "epoch": 0.5808037004914716,
+      "grad_norm": 1.0683690321009465,
+      "learning_rate": 7.619180265761865e-06,
+      "loss": 1.537,
+      "step": 2009
+    },
+    {
+      "epoch": 0.5810928013876843,
+      "grad_norm": 1.0607848244141498,
+      "learning_rate": 7.610269889604962e-06,
+      "loss": 1.6254,
+      "step": 2010
+    },
+    {
+      "epoch": 0.5813819022838971,
+      "grad_norm": 1.0504440465452836,
+      "learning_rate": 7.601361525230713e-06,
+      "loss": 1.6922,
+      "step": 2011
+    },
+    {
+      "epoch": 0.5816710031801099,
+      "grad_norm": 1.0161034794373591,
+      "learning_rate": 7.592455180138579e-06,
+      "loss": 1.4799,
+      "step": 2012
+    },
+    {
+      "epoch": 0.5819601040763226,
+      "grad_norm": 1.0373576500989061,
+      "learning_rate": 7.583550861826321e-06,
+      "loss": 1.5758,
+      "step": 2013
+    },
+    {
+      "epoch": 0.5822492049725354,
+      "grad_norm": 1.0817550047407012,
+      "learning_rate": 7.57464857779e-06,
+      "loss": 1.5452,
+      "step": 2014
+    },
+    {
+      "epoch": 0.5825383058687482,
+      "grad_norm": 1.083128827732501,
+      "learning_rate": 7.565748335523955e-06,
+      "loss": 1.5828,
+      "step": 2015
+    },
+    {
+      "epoch": 0.5828274067649609,
+      "grad_norm": 1.0843870504382112,
+      "learning_rate": 7.556850142520814e-06,
+      "loss": 1.5899,
+      "step": 2016
+    },
+    {
+      "epoch": 0.5831165076611737,
+      "grad_norm": 1.1059854296355742,
+      "learning_rate": 7.547954006271481e-06,
+      "loss": 1.602,
+      "step": 2017
+    },
+    {
+      "epoch": 0.5834056085573865,
+      "grad_norm": 1.0687711938668183,
+      "learning_rate": 7.539059934265118e-06,
+      "loss": 1.504,
+      "step": 2018
+    },
+    {
+      "epoch": 0.5836947094535994,
+      "grad_norm": 1.0848853251402641,
+      "learning_rate": 7.530167933989161e-06,
+      "loss": 1.5815,
+      "step": 2019
+    },
+    {
+      "epoch": 0.5839838103498121,
+      "grad_norm": 1.0682406432894893,
+      "learning_rate": 7.521278012929293e-06,
+      "loss": 1.5258,
+      "step": 2020
+    },
+    {
+      "epoch": 0.5842729112460249,
+      "grad_norm": 1.0086721072982225,
+      "learning_rate": 7.5123901785694586e-06,
+      "loss": 1.5107,
+      "step": 2021
+    },
+    {
+      "epoch": 0.5845620121422377,
+      "grad_norm": 1.033149721911076,
+      "learning_rate": 7.503504438391831e-06,
+      "loss": 1.5532,
+      "step": 2022
+    },
+    {
+      "epoch": 0.5848511130384504,
+      "grad_norm": 1.0586778156563794,
+      "learning_rate": 7.49462079987683e-06,
+      "loss": 1.5555,
+      "step": 2023
+    },
+    {
+      "epoch": 0.5851402139346632,
+      "grad_norm": 1.0620534880626484,
+      "learning_rate": 7.4857392705031025e-06,
+      "loss": 1.4613,
+      "step": 2024
+    },
+    {
+      "epoch": 0.585429314830876,
+      "grad_norm": 1.0592696568931845,
+      "learning_rate": 7.47685985774752e-06,
+      "loss": 1.5497,
+      "step": 2025
+    },
+    {
+      "epoch": 0.5857184157270887,
+      "grad_norm": 1.1081129124508828,
+      "learning_rate": 7.467982569085175e-06,
+      "loss": 1.6219,
+      "step": 2026
+    },
+    {
+      "epoch": 0.5860075166233015,
+      "grad_norm": 1.0513460586577323,
+      "learning_rate": 7.45910741198937e-06,
+      "loss": 1.6091,
+      "step": 2027
+    },
+    {
+      "epoch": 0.5862966175195143,
+      "grad_norm": 1.029602506895878,
+      "learning_rate": 7.450234393931609e-06,
+      "loss": 1.6163,
+      "step": 2028
+    },
+    {
+      "epoch": 0.586585718415727,
+      "grad_norm": 1.0483199075459788,
+      "learning_rate": 7.441363522381602e-06,
+      "loss": 1.4142,
+      "step": 2029
+    },
+    {
+      "epoch": 0.5868748193119399,
+      "grad_norm": 1.0408307350087287,
+      "learning_rate": 7.4324948048072484e-06,
+      "loss": 1.5666,
+      "step": 2030
+    },
+    {
+      "epoch": 0.5871639202081527,
+      "grad_norm": 1.041657058430387,
+      "learning_rate": 7.4236282486746325e-06,
+      "loss": 1.5757,
+      "step": 2031
+    },
+    {
+      "epoch": 0.5874530211043654,
+      "grad_norm": 1.0503560558611809,
+      "learning_rate": 7.414763861448024e-06,
+      "loss": 1.6007,
+      "step": 2032
+    },
+    {
+      "epoch": 0.5877421220005782,
+      "grad_norm": 1.0183071825405716,
+      "learning_rate": 7.4059016505898665e-06,
+      "loss": 1.5635,
+      "step": 2033
+    },
+    {
+      "epoch": 0.588031222896791,
+      "grad_norm": 1.0192827549972139,
+      "learning_rate": 7.397041623560766e-06,
+      "loss": 1.5609,
+      "step": 2034
+    },
+    {
+      "epoch": 0.5883203237930038,
+      "grad_norm": 1.034906387221898,
+      "learning_rate": 7.388183787819492e-06,
+      "loss": 1.698,
+      "step": 2035
+    },
+    {
+      "epoch": 0.5886094246892165,
+      "grad_norm": 1.0185385273747416,
+      "learning_rate": 7.379328150822971e-06,
+      "loss": 1.5656,
+      "step": 2036
+    },
+    {
+      "epoch": 0.5888985255854293,
+      "grad_norm": 1.0554495191296656,
+      "learning_rate": 7.3704747200262815e-06,
+      "loss": 1.6439,
+      "step": 2037
+    },
+    {
+      "epoch": 0.5891876264816421,
+      "grad_norm": 1.0671015919575622,
+      "learning_rate": 7.361623502882637e-06,
+      "loss": 1.6786,
+      "step": 2038
+    },
+    {
+      "epoch": 0.5894767273778548,
+      "grad_norm": 1.0305222951460096,
+      "learning_rate": 7.352774506843394e-06,
+      "loss": 1.5179,
+      "step": 2039
+    },
+    {
+      "epoch": 0.5897658282740676,
+      "grad_norm": 1.046052574199978,
+      "learning_rate": 7.343927739358032e-06,
+      "loss": 1.5263,
+      "step": 2040
+    },
+    {
+      "epoch": 0.5900549291702805,
+      "grad_norm": 1.0395664020829851,
+      "learning_rate": 7.335083207874165e-06,
+      "loss": 1.523,
+      "step": 2041
+    },
+    {
+      "epoch": 0.5903440300664932,
+      "grad_norm": 0.9990519884747506,
+      "learning_rate": 7.3262409198375124e-06,
+      "loss": 1.486,
+      "step": 2042
+    },
+    {
+      "epoch": 0.590633130962706,
+      "grad_norm": 1.003600653676946,
+      "learning_rate": 7.317400882691918e-06,
+      "loss": 1.5043,
+      "step": 2043
+    },
+    {
+      "epoch": 0.5909222318589188,
+      "grad_norm": 1.0272308031802884,
+      "learning_rate": 7.308563103879322e-06,
+      "loss": 1.5785,
+      "step": 2044
+    },
+    {
+      "epoch": 0.5912113327551315,
+      "grad_norm": 1.020667241456248,
+      "learning_rate": 7.299727590839765e-06,
+      "loss": 1.5475,
+      "step": 2045
+    },
+    {
+      "epoch": 0.5915004336513443,
+      "grad_norm": 1.0067772541329505,
+      "learning_rate": 7.290894351011382e-06,
+      "loss": 1.5978,
+      "step": 2046
+    },
+    {
+      "epoch": 0.5917895345475571,
+      "grad_norm": 1.0158497229440513,
+      "learning_rate": 7.282063391830387e-06,
+      "loss": 1.4825,
+      "step": 2047
+    },
+    {
+      "epoch": 0.5920786354437699,
+      "grad_norm": 1.0460776993727472,
+      "learning_rate": 7.273234720731085e-06,
+      "loss": 1.6044,
+      "step": 2048
+    },
+    {
+      "epoch": 0.5923677363399826,
+      "grad_norm": 1.0599530946966202,
+      "learning_rate": 7.264408345145851e-06,
+      "loss": 1.5229,
+      "step": 2049
+    },
+    {
+      "epoch": 0.5926568372361954,
+      "grad_norm": 1.0870814845010357,
+      "learning_rate": 7.255584272505123e-06,
+      "loss": 1.6537,
+      "step": 2050
+    },
+    {
+      "epoch": 0.5929459381324083,
+      "grad_norm": 1.034391690332542,
+      "learning_rate": 7.246762510237404e-06,
+      "loss": 1.6302,
+      "step": 2051
+    },
+    {
+      "epoch": 0.593235039028621,
+      "grad_norm": 6.544227220737374,
+      "learning_rate": 7.2379430657692495e-06,
+      "loss": 1.6717,
+      "step": 2052
+    },
+    {
+      "epoch": 0.5935241399248338,
+      "grad_norm": 1.0416198636995884,
+      "learning_rate": 7.229125946525273e-06,
+      "loss": 1.5583,
+      "step": 2053
+    },
+    {
+      "epoch": 0.5938132408210466,
+      "grad_norm": 1.0436452531269607,
+      "learning_rate": 7.220311159928117e-06,
+      "loss": 1.5085,
+      "step": 2054
+    },
+    {
+      "epoch": 0.5941023417172593,
+      "grad_norm": 1.0114598418528282,
+      "learning_rate": 7.211498713398471e-06,
+      "loss": 1.4908,
+      "step": 2055
+    },
+    {
+      "epoch": 0.5943914426134721,
+      "grad_norm": 1.0751530887949694,
+      "learning_rate": 7.202688614355047e-06,
+      "loss": 1.5879,
+      "step": 2056
+    },
+    {
+      "epoch": 0.5946805435096849,
+      "grad_norm": 1.0273276992594014,
+      "learning_rate": 7.193880870214583e-06,
+      "loss": 1.5579,
+      "step": 2057
+    },
+    {
+      "epoch": 0.5949696444058976,
+      "grad_norm": 1.0871466248828028,
+      "learning_rate": 7.185075488391838e-06,
+      "loss": 1.6878,
+      "step": 2058
+    },
+    {
+      "epoch": 0.5952587453021104,
+      "grad_norm": 1.0020537603273587,
+      "learning_rate": 7.176272476299582e-06,
+      "loss": 1.6083,
+      "step": 2059
+    },
+    {
+      "epoch": 0.5955478461983232,
+      "grad_norm": 1.0214279397341675,
+      "learning_rate": 7.167471841348582e-06,
+      "loss": 1.554,
+      "step": 2060
+    },
+    {
+      "epoch": 0.5958369470945359,
+      "grad_norm": 1.067675983759507,
+      "learning_rate": 7.158673590947615e-06,
+      "loss": 1.5558,
+      "step": 2061
+    },
+    {
+      "epoch": 0.5961260479907488,
+      "grad_norm": 1.0232682050016655,
+      "learning_rate": 7.149877732503441e-06,
+      "loss": 1.5884,
+      "step": 2062
+    },
+    {
+      "epoch": 0.5964151488869616,
+      "grad_norm": 1.1004682407987547,
+      "learning_rate": 7.141084273420809e-06,
+      "loss": 1.6706,
+      "step": 2063
+    },
+    {
+      "epoch": 0.5967042497831744,
+      "grad_norm": 1.046034346667384,
+      "learning_rate": 7.1322932211024564e-06,
+      "loss": 1.4823,
+      "step": 2064
+    },
+    {
+      "epoch": 0.5969933506793871,
+      "grad_norm": 1.0405086920381532,
+      "learning_rate": 7.123504582949083e-06,
+      "loss": 1.4963,
+      "step": 2065
+    },
+    {
+      "epoch": 0.5972824515755999,
+      "grad_norm": 1.0343438942672538,
+      "learning_rate": 7.114718366359367e-06,
+      "loss": 1.5459,
+      "step": 2066
+    },
+    {
+      "epoch": 0.5975715524718127,
+      "grad_norm": 1.0355991392953394,
+      "learning_rate": 7.105934578729934e-06,
+      "loss": 1.5478,
+      "step": 2067
+    },
+    {
+      "epoch": 0.5978606533680254,
+      "grad_norm": 1.0530221181790307,
+      "learning_rate": 7.097153227455379e-06,
+      "loss": 1.6123,
+      "step": 2068
+    },
+    {
+      "epoch": 0.5981497542642382,
+      "grad_norm": 1.0416676262772353,
+      "learning_rate": 7.088374319928235e-06,
+      "loss": 1.5972,
+      "step": 2069
+    },
+    {
+      "epoch": 0.598438855160451,
+      "grad_norm": 1.0517615111288614,
+      "learning_rate": 7.079597863538983e-06,
+      "loss": 1.5659,
+      "step": 2070
+    },
+    {
+      "epoch": 0.5987279560566637,
+      "grad_norm": 1.029849984832984,
+      "learning_rate": 7.070823865676046e-06,
+      "loss": 1.557,
+      "step": 2071
+    },
+    {
+      "epoch": 0.5990170569528765,
+      "grad_norm": 1.0454670203799596,
+      "learning_rate": 7.062052333725764e-06,
+      "loss": 1.5664,
+      "step": 2072
+    },
+    {
+      "epoch": 0.5993061578490894,
+      "grad_norm": 1.1010786994595503,
+      "learning_rate": 7.0532832750724115e-06,
+      "loss": 1.6529,
+      "step": 2073
+    },
+    {
+      "epoch": 0.5995952587453021,
+      "grad_norm": 1.0520669960752511,
+      "learning_rate": 7.044516697098176e-06,
+      "loss": 1.4954,
+      "step": 2074
+    },
+    {
+      "epoch": 0.5998843596415149,
+      "grad_norm": 1.1108434302528232,
+      "learning_rate": 7.035752607183162e-06,
+      "loss": 1.6228,
+      "step": 2075
+    },
+    {
+      "epoch": 0.6001734605377277,
+      "grad_norm": 1.0727153788294717,
+      "learning_rate": 7.026991012705377e-06,
+      "loss": 1.6826,
+      "step": 2076
+    },
+    {
+      "epoch": 0.6004625614339405,
+      "grad_norm": 1.0422121104611137,
+      "learning_rate": 7.018231921040721e-06,
+      "loss": 1.5317,
+      "step": 2077
+    },
+    {
+      "epoch": 0.6007516623301532,
+      "grad_norm": 1.1332711748925963,
+      "learning_rate": 7.0094753395629965e-06,
+      "loss": 1.5219,
+      "step": 2078
+    },
+    {
+      "epoch": 0.601040763226366,
+      "grad_norm": 0.9993819863744031,
+      "learning_rate": 7.000721275643887e-06,
+      "loss": 1.4795,
+      "step": 2079
+    },
+    {
+      "epoch": 0.6013298641225788,
+      "grad_norm": 1.0682971031807795,
+      "learning_rate": 6.9919697366529595e-06,
+      "loss": 1.5385,
+      "step": 2080
+    },
+    {
+      "epoch": 0.6016189650187915,
+      "grad_norm": 1.043339833550436,
+      "learning_rate": 6.983220729957657e-06,
+      "loss": 1.5054,
+      "step": 2081
+    },
+    {
+      "epoch": 0.6019080659150043,
+      "grad_norm": 1.0352639592585875,
+      "learning_rate": 6.974474262923283e-06,
+      "loss": 1.6307,
+      "step": 2082
+    },
+    {
+      "epoch": 0.6021971668112172,
+      "grad_norm": 1.0240397117319717,
+      "learning_rate": 6.965730342913011e-06,
+      "loss": 1.5804,
+      "step": 2083
+    },
+    {
+      "epoch": 0.6024862677074299,
+      "grad_norm": 1.0342351404644785,
+      "learning_rate": 6.956988977287868e-06,
+      "loss": 1.5606,
+      "step": 2084
+    },
+    {
+      "epoch": 0.6027753686036427,
+      "grad_norm": 1.0649993792818435,
+      "learning_rate": 6.948250173406726e-06,
+      "loss": 1.5965,
+      "step": 2085
+    },
+    {
+      "epoch": 0.6030644694998555,
+      "grad_norm": 1.0616975611219477,
+      "learning_rate": 6.93951393862631e-06,
+      "loss": 1.5625,
+      "step": 2086
+    },
+    {
+      "epoch": 0.6033535703960682,
+      "grad_norm": 1.1371568565307015,
+      "learning_rate": 6.930780280301175e-06,
+      "loss": 1.7321,
+      "step": 2087
+    },
+    {
+      "epoch": 0.603642671292281,
+      "grad_norm": 1.0395995909926687,
+      "learning_rate": 6.922049205783709e-06,
+      "loss": 1.5379,
+      "step": 2088
+    },
+    {
+      "epoch": 0.6039317721884938,
+      "grad_norm": 1.0343222060687443,
+      "learning_rate": 6.913320722424122e-06,
+      "loss": 1.4898,
+      "step": 2089
+    },
+    {
+      "epoch": 0.6042208730847065,
+      "grad_norm": 1.0184578303890766,
+      "learning_rate": 6.904594837570448e-06,
+      "loss": 1.5799,
+      "step": 2090
+    },
+    {
+      "epoch": 0.6045099739809193,
+      "grad_norm": 1.0239502665631421,
+      "learning_rate": 6.89587155856853e-06,
+      "loss": 1.5177,
+      "step": 2091
+    },
+    {
+      "epoch": 0.6047990748771321,
+      "grad_norm": 1.1211126262100752,
+      "learning_rate": 6.8871508927620155e-06,
+      "loss": 1.6699,
+      "step": 2092
+    },
+    {
+      "epoch": 0.605088175773345,
+      "grad_norm": 1.0345579261622053,
+      "learning_rate": 6.87843284749236e-06,
+      "loss": 1.5575,
+      "step": 2093
+    },
+    {
+      "epoch": 0.6053772766695577,
+      "grad_norm": 1.038826348616125,
+      "learning_rate": 6.869717430098801e-06,
+      "loss": 1.5907,
+      "step": 2094
+    },
+    {
+      "epoch": 0.6056663775657705,
+      "grad_norm": 1.0032888557348199,
+      "learning_rate": 6.86100464791837e-06,
+      "loss": 1.5448,
+      "step": 2095
+    },
+    {
+      "epoch": 0.6059554784619833,
+      "grad_norm": 1.0945122516770431,
+      "learning_rate": 6.852294508285886e-06,
+      "loss": 1.5855,
+      "step": 2096
+    },
+    {
+      "epoch": 0.606244579358196,
+      "grad_norm": 1.0275634929745217,
+      "learning_rate": 6.843587018533936e-06,
+      "loss": 1.5363,
+      "step": 2097
+    },
+    {
+      "epoch": 0.6065336802544088,
+      "grad_norm": 1.0294290705012428,
+      "learning_rate": 6.834882185992878e-06,
+      "loss": 1.5912,
+      "step": 2098
+    },
+    {
+      "epoch": 0.6068227811506216,
+      "grad_norm": 1.0584674675673924,
+      "learning_rate": 6.8261800179908284e-06,
+      "loss": 1.5429,
+      "step": 2099
+    },
+    {
+      "epoch": 0.6071118820468343,
+      "grad_norm": 1.0925055442187697,
+      "learning_rate": 6.817480521853672e-06,
+      "loss": 1.6922,
+      "step": 2100
+    },
+    {
+      "epoch": 0.6074009829430471,
+      "grad_norm": 1.0116809307214536,
+      "learning_rate": 6.808783704905031e-06,
+      "loss": 1.4906,
+      "step": 2101
+    },
+    {
+      "epoch": 0.6076900838392599,
+      "grad_norm": 1.028585779235767,
+      "learning_rate": 6.800089574466281e-06,
+      "loss": 1.5869,
+      "step": 2102
+    },
+    {
+      "epoch": 0.6079791847354726,
+      "grad_norm": 1.025077051602754,
+      "learning_rate": 6.791398137856535e-06,
+      "loss": 1.5686,
+      "step": 2103
+    },
+    {
+      "epoch": 0.6082682856316854,
+      "grad_norm": 1.049820969279426,
+      "learning_rate": 6.782709402392632e-06,
+      "loss": 1.4974,
+      "step": 2104
+    },
+    {
+      "epoch": 0.6085573865278983,
+      "grad_norm": 1.021054660445814,
+      "learning_rate": 6.774023375389145e-06,
+      "loss": 1.5471,
+      "step": 2105
+    },
+    {
+      "epoch": 0.608846487424111,
+      "grad_norm": 1.0416371973276695,
+      "learning_rate": 6.765340064158361e-06,
+      "loss": 1.535,
+      "step": 2106
+    },
+    {
+      "epoch": 0.6091355883203238,
+      "grad_norm": 0.9960649466877588,
+      "learning_rate": 6.756659476010288e-06,
+      "loss": 1.5594,
+      "step": 2107
+    },
+    {
+      "epoch": 0.6094246892165366,
+      "grad_norm": 1.0529741218044932,
+      "learning_rate": 6.747981618252632e-06,
+      "loss": 1.577,
+      "step": 2108
+    },
+    {
+      "epoch": 0.6097137901127494,
+      "grad_norm": 1.0122060944007447,
+      "learning_rate": 6.73930649819081e-06,
+      "loss": 1.6607,
+      "step": 2109
+    },
+    {
+      "epoch": 0.6100028910089621,
+      "grad_norm": 1.0235671226448881,
+      "learning_rate": 6.730634123127926e-06,
+      "loss": 1.6113,
+      "step": 2110
+    },
+    {
+      "epoch": 0.6102919919051749,
+      "grad_norm": 1.0375275780491622,
+      "learning_rate": 6.7219645003647745e-06,
+      "loss": 1.5755,
+      "step": 2111
+    },
+    {
+      "epoch": 0.6105810928013877,
+      "grad_norm": 1.039784965236839,
+      "learning_rate": 6.71329763719984e-06,
+      "loss": 1.4914,
+      "step": 2112
+    },
+    {
+      "epoch": 0.6108701936976004,
+      "grad_norm": 1.0617046264649195,
+      "learning_rate": 6.704633540929279e-06,
+      "loss": 1.6213,
+      "step": 2113
+    },
+    {
+      "epoch": 0.6111592945938132,
+      "grad_norm": 1.0710943179900063,
+      "learning_rate": 6.695972218846916e-06,
+      "loss": 1.6019,
+      "step": 2114
+    },
+    {
+      "epoch": 0.611448395490026,
+      "grad_norm": 1.0175849039381482,
+      "learning_rate": 6.687313678244243e-06,
+      "loss": 1.5539,
+      "step": 2115
+    },
+    {
+      "epoch": 0.6117374963862388,
+      "grad_norm": 1.077129210840068,
+      "learning_rate": 6.678657926410412e-06,
+      "loss": 1.6442,
+      "step": 2116
+    },
+    {
+      "epoch": 0.6120265972824516,
+      "grad_norm": 1.0447968418724285,
+      "learning_rate": 6.67000497063222e-06,
+      "loss": 1.5636,
+      "step": 2117
+    },
+    {
+      "epoch": 0.6123156981786644,
+      "grad_norm": 1.0092988805933885,
+      "learning_rate": 6.6613548181941235e-06,
+      "loss": 1.4989,
+      "step": 2118
+    },
+    {
+      "epoch": 0.6126047990748771,
+      "grad_norm": 1.0291683612465103,
+      "learning_rate": 6.652707476378207e-06,
+      "loss": 1.5319,
+      "step": 2119
+    },
+    {
+      "epoch": 0.6128938999710899,
+      "grad_norm": 1.052008946164791,
+      "learning_rate": 6.644062952464196e-06,
+      "loss": 1.4764,
+      "step": 2120
+    },
+    {
+      "epoch": 0.6131830008673027,
+      "grad_norm": 1.0717729972358117,
+      "learning_rate": 6.635421253729437e-06,
+      "loss": 1.535,
+      "step": 2121
+    },
+    {
+      "epoch": 0.6134721017635155,
+      "grad_norm": 1.0469518433590574,
+      "learning_rate": 6.626782387448904e-06,
+      "loss": 1.5004,
+      "step": 2122
+    },
+    {
+      "epoch": 0.6137612026597282,
+      "grad_norm": 1.0506052056692963,
+      "learning_rate": 6.618146360895186e-06,
+      "loss": 1.5185,
+      "step": 2123
+    },
+    {
+      "epoch": 0.614050303555941,
+      "grad_norm": 1.0864360309540992,
+      "learning_rate": 6.609513181338476e-06,
+      "loss": 1.5926,
+      "step": 2124
+    },
+    {
+      "epoch": 0.6143394044521538,
+      "grad_norm": 1.3449471172392797,
+      "learning_rate": 6.600882856046579e-06,
+      "loss": 1.5991,
+      "step": 2125
+    },
+    {
+      "epoch": 0.6146285053483665,
+      "grad_norm": 1.045840879588071,
+      "learning_rate": 6.592255392284886e-06,
+      "loss": 1.6447,
+      "step": 2126
+    },
+    {
+      "epoch": 0.6149176062445794,
+      "grad_norm": 1.0435706292954865,
+      "learning_rate": 6.583630797316387e-06,
+      "loss": 1.5452,
+      "step": 2127
+    },
+    {
+      "epoch": 0.6152067071407922,
+      "grad_norm": 1.0556700544481026,
+      "learning_rate": 6.5750090784016595e-06,
+      "loss": 1.4714,
+      "step": 2128
+    },
+    {
+      "epoch": 0.6154958080370049,
+      "grad_norm": 1.0212397946806842,
+      "learning_rate": 6.566390242798855e-06,
+      "loss": 1.4563,
+      "step": 2129
+    },
+    {
+      "epoch": 0.6157849089332177,
+      "grad_norm": 1.0448817927594793,
+      "learning_rate": 6.557774297763696e-06,
+      "loss": 1.6282,
+      "step": 2130
+    },
+    {
+      "epoch": 0.6160740098294305,
+      "grad_norm": 1.086620466396412,
+      "learning_rate": 6.549161250549474e-06,
+      "loss": 1.5813,
+      "step": 2131
+    },
+    {
+      "epoch": 0.6163631107256432,
+      "grad_norm": 1.0672257846815265,
+      "learning_rate": 6.5405511084070415e-06,
+      "loss": 1.6048,
+      "step": 2132
+    },
+    {
+      "epoch": 0.616652211621856,
+      "grad_norm": 1.0746898138104755,
+      "learning_rate": 6.531943878584804e-06,
+      "loss": 1.493,
+      "step": 2133
+    },
+    {
+      "epoch": 0.6169413125180688,
+      "grad_norm": 1.108725873065191,
+      "learning_rate": 6.523339568328715e-06,
+      "loss": 1.5696,
+      "step": 2134
+    },
+    {
+      "epoch": 0.6172304134142815,
+      "grad_norm": 1.0096209339776385,
+      "learning_rate": 6.514738184882275e-06,
+      "loss": 1.4642,
+      "step": 2135
+    },
+    {
+      "epoch": 0.6175195143104943,
+      "grad_norm": 1.0293654749090704,
+      "learning_rate": 6.506139735486512e-06,
+      "loss": 1.6007,
+      "step": 2136
+    },
+    {
+      "epoch": 0.6178086152067072,
+      "grad_norm": 1.0501515375376458,
+      "learning_rate": 6.4975442273799905e-06,
+      "loss": 1.4841,
+      "step": 2137
+    },
+    {
+      "epoch": 0.61809771610292,
+      "grad_norm": 1.0278409768897196,
+      "learning_rate": 6.488951667798796e-06,
+      "loss": 1.4278,
+      "step": 2138
+    },
+    {
+      "epoch": 0.6183868169991327,
+      "grad_norm": 1.044468873452407,
+      "learning_rate": 6.48036206397654e-06,
+      "loss": 1.5552,
+      "step": 2139
+    },
+    {
+      "epoch": 0.6186759178953455,
+      "grad_norm": 0.9961967291159364,
+      "learning_rate": 6.471775423144328e-06,
+      "loss": 1.4214,
+      "step": 2140
+    },
+    {
+      "epoch": 0.6189650187915583,
+      "grad_norm": 0.9953322258186914,
+      "learning_rate": 6.463191752530791e-06,
+      "loss": 1.5986,
+      "step": 2141
+    },
+    {
+      "epoch": 0.619254119687771,
+      "grad_norm": 1.0534611770014588,
+      "learning_rate": 6.454611059362046e-06,
+      "loss": 1.5033,
+      "step": 2142
+    },
+    {
+      "epoch": 0.6195432205839838,
+      "grad_norm": 1.0355826989579244,
+      "learning_rate": 6.446033350861706e-06,
+      "loss": 1.485,
+      "step": 2143
+    },
+    {
+      "epoch": 0.6198323214801966,
+      "grad_norm": 1.0235085183628148,
+      "learning_rate": 6.4374586342508745e-06,
+      "loss": 1.3951,
+      "step": 2144
+    },
+    {
+      "epoch": 0.6201214223764093,
+      "grad_norm": 1.0365612544529943,
+      "learning_rate": 6.428886916748138e-06,
+      "loss": 1.538,
+      "step": 2145
+    },
+    {
+      "epoch": 0.6204105232726221,
+      "grad_norm": 1.0409195912532607,
+      "learning_rate": 6.420318205569551e-06,
+      "loss": 1.5605,
+      "step": 2146
+    },
+    {
+      "epoch": 0.620699624168835,
+      "grad_norm": 1.060660067125822,
+      "learning_rate": 6.411752507928643e-06,
+      "loss": 1.6794,
+      "step": 2147
+    },
+    {
+      "epoch": 0.6209887250650477,
+      "grad_norm": 1.1032226210731721,
+      "learning_rate": 6.403189831036405e-06,
+      "loss": 1.5769,
+      "step": 2148
+    },
+    {
+      "epoch": 0.6212778259612605,
+      "grad_norm": 1.069433192636476,
+      "learning_rate": 6.3946301821012824e-06,
+      "loss": 1.5545,
+      "step": 2149
+    },
+    {
+      "epoch": 0.6215669268574733,
+      "grad_norm": 1.0789399321327287,
+      "learning_rate": 6.386073568329176e-06,
+      "loss": 1.5061,
+      "step": 2150
+    },
+    {
+      "epoch": 0.6218560277536861,
+      "grad_norm": 1.0936939699597317,
+      "learning_rate": 6.3775199969234315e-06,
+      "loss": 1.5997,
+      "step": 2151
+    },
+    {
+      "epoch": 0.6221451286498988,
+      "grad_norm": 1.0294263609540737,
+      "learning_rate": 6.368969475084831e-06,
+      "loss": 1.5805,
+      "step": 2152
+    },
+    {
+      "epoch": 0.6224342295461116,
+      "grad_norm": 1.0784697001369723,
+      "learning_rate": 6.360422010011586e-06,
+      "loss": 1.5364,
+      "step": 2153
+    },
+    {
+      "epoch": 0.6227233304423244,
+      "grad_norm": 1.0187372868017754,
+      "learning_rate": 6.351877608899339e-06,
+      "loss": 1.5126,
+      "step": 2154
+    },
+    {
+      "epoch": 0.6230124313385371,
+      "grad_norm": 1.0309643932537929,
+      "learning_rate": 6.3433362789411565e-06,
+      "loss": 1.5597,
+      "step": 2155
+    },
+    {
+      "epoch": 0.6233015322347499,
+      "grad_norm": 1.0521482397276203,
+      "learning_rate": 6.334798027327511e-06,
+      "loss": 1.6182,
+      "step": 2156
+    },
+    {
+      "epoch": 0.6235906331309627,
+      "grad_norm": 1.0550535077539223,
+      "learning_rate": 6.326262861246291e-06,
+      "loss": 1.5906,
+      "step": 2157
+    },
+    {
+      "epoch": 0.6238797340271754,
+      "grad_norm": 1.054034379896794,
+      "learning_rate": 6.3177307878827825e-06,
+      "loss": 1.4888,
+      "step": 2158
+    },
+    {
+      "epoch": 0.6241688349233883,
+      "grad_norm": 1.0794202005586107,
+      "learning_rate": 6.309201814419667e-06,
+      "loss": 1.5323,
+      "step": 2159
+    },
+    {
+      "epoch": 0.6244579358196011,
+      "grad_norm": 1.040727345447366,
+      "learning_rate": 6.3006759480370264e-06,
+      "loss": 1.4654,
+      "step": 2160
+    },
+    {
+      "epoch": 0.6247470367158138,
+      "grad_norm": 1.0447751591620802,
+      "learning_rate": 6.292153195912315e-06,
+      "loss": 1.6222,
+      "step": 2161
+    },
+    {
+      "epoch": 0.6250361376120266,
+      "grad_norm": 1.0196645212223057,
+      "learning_rate": 6.283633565220374e-06,
+      "loss": 1.5337,
+      "step": 2162
+    },
+    {
+      "epoch": 0.6253252385082394,
+      "grad_norm": 1.080196724463442,
+      "learning_rate": 6.275117063133409e-06,
+      "loss": 1.5483,
+      "step": 2163
+    },
+    {
+      "epoch": 0.6256143394044521,
+      "grad_norm": 1.0181279359329012,
+      "learning_rate": 6.266603696820999e-06,
+      "loss": 1.6225,
+      "step": 2164
+    },
+    {
+      "epoch": 0.6259034403006649,
+      "grad_norm": 1.0622962778185558,
+      "learning_rate": 6.258093473450075e-06,
+      "loss": 1.5548,
+      "step": 2165
+    },
+    {
+      "epoch": 0.6261925411968777,
+      "grad_norm": 1.0749118976975562,
+      "learning_rate": 6.24958640018493e-06,
+      "loss": 1.493,
+      "step": 2166
+    },
+    {
+      "epoch": 0.6264816420930905,
+      "grad_norm": 1.027656025937481,
+      "learning_rate": 6.241082484187207e-06,
+      "loss": 1.4837,
+      "step": 2167
+    },
+    {
+      "epoch": 0.6267707429893032,
+      "grad_norm": 1.0382679535531123,
+      "learning_rate": 6.2325817326158764e-06,
+      "loss": 1.5858,
+      "step": 2168
+    },
+    {
+      "epoch": 0.627059843885516,
+      "grad_norm": 1.0589472743959976,
+      "learning_rate": 6.224084152627262e-06,
+      "loss": 1.5104,
+      "step": 2169
+    },
+    {
+      "epoch": 0.6273489447817289,
+      "grad_norm": 1.1113654314267245,
+      "learning_rate": 6.215589751375005e-06,
+      "loss": 1.5716,
+      "step": 2170
+    },
+    {
+      "epoch": 0.6276380456779416,
+      "grad_norm": 1.0079031973881047,
+      "learning_rate": 6.207098536010083e-06,
+      "loss": 1.489,
+      "step": 2171
+    },
+    {
+      "epoch": 0.6279271465741544,
+      "grad_norm": 1.0787949826145822,
+      "learning_rate": 6.198610513680779e-06,
+      "loss": 1.5873,
+      "step": 2172
+    },
+    {
+      "epoch": 0.6282162474703672,
+      "grad_norm": 1.0593205391551248,
+      "learning_rate": 6.190125691532691e-06,
+      "loss": 1.5697,
+      "step": 2173
+    },
+    {
+      "epoch": 0.6285053483665799,
+      "grad_norm": 1.020759638844432,
+      "learning_rate": 6.181644076708732e-06,
+      "loss": 1.4957,
+      "step": 2174
+    },
+    {
+      "epoch": 0.6287944492627927,
+      "grad_norm": 0.9960794050747592,
+      "learning_rate": 6.173165676349103e-06,
+      "loss": 1.5647,
+      "step": 2175
+    },
+    {
+      "epoch": 0.6290835501590055,
+      "grad_norm": 1.0532455266610188,
+      "learning_rate": 6.164690497591305e-06,
+      "loss": 1.5383,
+      "step": 2176
+    },
+    {
+      "epoch": 0.6293726510552182,
+      "grad_norm": 1.0370109281714925,
+      "learning_rate": 6.1562185475701295e-06,
+      "loss": 1.628,
+      "step": 2177
+    },
+    {
+      "epoch": 0.629661751951431,
+      "grad_norm": 1.0475104494837768,
+      "learning_rate": 6.14774983341764e-06,
+      "loss": 1.5314,
+      "step": 2178
+    },
+    {
+      "epoch": 0.6299508528476438,
+      "grad_norm": 1.0166223741437819,
+      "learning_rate": 6.139284362263185e-06,
+      "loss": 1.5838,
+      "step": 2179
+    },
+    {
+      "epoch": 0.6302399537438566,
+      "grad_norm": 1.0396123019346366,
+      "learning_rate": 6.130822141233383e-06,
+      "loss": 1.5689,
+      "step": 2180
+    },
+    {
+      "epoch": 0.6305290546400694,
+      "grad_norm": 1.00382088086992,
+      "learning_rate": 6.122363177452109e-06,
+      "loss": 1.5648,
+      "step": 2181
+    },
+    {
+      "epoch": 0.6308181555362822,
+      "grad_norm": 1.0245625105388183,
+      "learning_rate": 6.113907478040503e-06,
+      "loss": 1.5885,
+      "step": 2182
+    },
+    {
+      "epoch": 0.631107256432495,
+      "grad_norm": 1.0191349868091641,
+      "learning_rate": 6.105455050116954e-06,
+      "loss": 1.6424,
+      "step": 2183
+    },
+    {
+      "epoch": 0.6313963573287077,
+      "grad_norm": 1.0420545221608395,
+      "learning_rate": 6.0970059007970996e-06,
+      "loss": 1.55,
+      "step": 2184
+    },
+    {
+      "epoch": 0.6316854582249205,
+      "grad_norm": 1.0420364452023971,
+      "learning_rate": 6.0885600371938115e-06,
+      "loss": 1.525,
+      "step": 2185
+    },
+    {
+      "epoch": 0.6319745591211333,
+      "grad_norm": 1.0808021330873194,
+      "learning_rate": 6.080117466417198e-06,
+      "loss": 1.5821,
+      "step": 2186
+    },
+    {
+      "epoch": 0.632263660017346,
+      "grad_norm": 1.0433369061895272,
+      "learning_rate": 6.071678195574602e-06,
+      "loss": 1.635,
+      "step": 2187
+    },
+    {
+      "epoch": 0.6325527609135588,
+      "grad_norm": 1.0107387332886495,
+      "learning_rate": 6.063242231770577e-06,
+      "loss": 1.5684,
+      "step": 2188
+    },
+    {
+      "epoch": 0.6328418618097716,
+      "grad_norm": 1.0396142531260353,
+      "learning_rate": 6.0548095821069e-06,
+      "loss": 1.5664,
+      "step": 2189
+    },
+    {
+      "epoch": 0.6331309627059843,
+      "grad_norm": 1.0253936532239378,
+      "learning_rate": 6.046380253682553e-06,
+      "loss": 1.5491,
+      "step": 2190
+    },
+    {
+      "epoch": 0.6334200636021972,
+      "grad_norm": 1.019756114675134,
+      "learning_rate": 6.03795425359373e-06,
+      "loss": 1.5733,
+      "step": 2191
+    },
+    {
+      "epoch": 0.63370916449841,
+      "grad_norm": 0.9962552981657559,
+      "learning_rate": 6.0295315889338134e-06,
+      "loss": 1.6141,
+      "step": 2192
+    },
+    {
+      "epoch": 0.6339982653946227,
+      "grad_norm": 1.0683305987903917,
+      "learning_rate": 6.021112266793387e-06,
+      "loss": 1.5367,
+      "step": 2193
+    },
+    {
+      "epoch": 0.6342873662908355,
+      "grad_norm": 1.0509682698783864,
+      "learning_rate": 6.0126962942602165e-06,
+      "loss": 1.625,
+      "step": 2194
+    },
+    {
+      "epoch": 0.6345764671870483,
+      "grad_norm": 1.0485155494152818,
+      "learning_rate": 6.004283678419243e-06,
+      "loss": 1.6497,
+      "step": 2195
+    },
+    {
+      "epoch": 0.6348655680832611,
+      "grad_norm": 1.0371089571108287,
+      "learning_rate": 5.995874426352589e-06,
+      "loss": 1.5283,
+      "step": 2196
+    },
+    {
+      "epoch": 0.6351546689794738,
+      "grad_norm": 1.005436280223932,
+      "learning_rate": 5.987468545139541e-06,
+      "loss": 1.5026,
+      "step": 2197
+    },
+    {
+      "epoch": 0.6354437698756866,
+      "grad_norm": 1.028579296169917,
+      "learning_rate": 5.9790660418565485e-06,
+      "loss": 1.5449,
+      "step": 2198
+    },
+    {
+      "epoch": 0.6357328707718994,
+      "grad_norm": 1.022039282722136,
+      "learning_rate": 5.970666923577222e-06,
+      "loss": 1.558,
+      "step": 2199
+    },
+    {
+      "epoch": 0.6360219716681121,
+      "grad_norm": 1.0499266746837728,
+      "learning_rate": 5.962271197372313e-06,
+      "loss": 1.5776,
+      "step": 2200
+    },
+    {
+      "epoch": 0.636311072564325,
+      "grad_norm": 1.1051549910969307,
+      "learning_rate": 5.953878870309724e-06,
+      "loss": 1.543,
+      "step": 2201
+    },
+    {
+      "epoch": 0.6366001734605378,
+      "grad_norm": 1.014205877490355,
+      "learning_rate": 5.945489949454495e-06,
+      "loss": 1.5534,
+      "step": 2202
+    },
+    {
+      "epoch": 0.6368892743567505,
+      "grad_norm": 1.0331962149575853,
+      "learning_rate": 5.937104441868804e-06,
+      "loss": 1.5442,
+      "step": 2203
+    },
+    {
+      "epoch": 0.6371783752529633,
+      "grad_norm": 1.030305816463569,
+      "learning_rate": 5.928722354611941e-06,
+      "loss": 1.5284,
+      "step": 2204
+    },
+    {
+      "epoch": 0.6374674761491761,
+      "grad_norm": 1.0264819550750173,
+      "learning_rate": 5.920343694740329e-06,
+      "loss": 1.5905,
+      "step": 2205
+    },
+    {
+      "epoch": 0.6377565770453888,
+      "grad_norm": 1.0166911741428681,
+      "learning_rate": 5.911968469307505e-06,
+      "loss": 1.4995,
+      "step": 2206
+    },
+    {
+      "epoch": 0.6380456779416016,
+      "grad_norm": 1.0982495480065115,
+      "learning_rate": 5.903596685364106e-06,
+      "loss": 1.5096,
+      "step": 2207
+    },
+    {
+      "epoch": 0.6383347788378144,
+      "grad_norm": 1.0114601344735037,
+      "learning_rate": 5.89522834995788e-06,
+      "loss": 1.5241,
+      "step": 2208
+    },
+    {
+      "epoch": 0.6386238797340271,
+      "grad_norm": 1.0510019026867954,
+      "learning_rate": 5.886863470133672e-06,
+      "loss": 1.5554,
+      "step": 2209
+    },
+    {
+      "epoch": 0.6389129806302399,
+      "grad_norm": 1.0466425092619458,
+      "learning_rate": 5.878502052933412e-06,
+      "loss": 1.5094,
+      "step": 2210
+    },
+    {
+      "epoch": 0.6392020815264527,
+      "grad_norm": 0.9980482801105709,
+      "learning_rate": 5.8701441053961185e-06,
+      "loss": 1.5735,
+      "step": 2211
+    },
+    {
+      "epoch": 0.6394911824226656,
+      "grad_norm": 1.0022814754424962,
+      "learning_rate": 5.8617896345578904e-06,
+      "loss": 1.611,
+      "step": 2212
+    },
+    {
+      "epoch": 0.6397802833188783,
+      "grad_norm": 1.060301824097519,
+      "learning_rate": 5.8534386474518945e-06,
+      "loss": 1.656,
+      "step": 2213
+    },
+    {
+      "epoch": 0.6400693842150911,
+      "grad_norm": 1.0596587088620473,
+      "learning_rate": 5.845091151108373e-06,
+      "loss": 1.6156,
+      "step": 2214
+    },
+    {
+      "epoch": 0.6403584851113039,
+      "grad_norm": 1.0293992884981398,
+      "learning_rate": 5.836747152554625e-06,
+      "loss": 1.5242,
+      "step": 2215
+    },
+    {
+      "epoch": 0.6406475860075166,
+      "grad_norm": 1.0143102098808645,
+      "learning_rate": 5.828406658815004e-06,
+      "loss": 1.6187,
+      "step": 2216
+    },
+    {
+      "epoch": 0.6409366869037294,
+      "grad_norm": 1.0814172498625096,
+      "learning_rate": 5.820069676910915e-06,
+      "loss": 1.6579,
+      "step": 2217
+    },
+    {
+      "epoch": 0.6412257877999422,
+      "grad_norm": 1.0263493825302894,
+      "learning_rate": 5.811736213860801e-06,
+      "loss": 1.5604,
+      "step": 2218
+    },
+    {
+      "epoch": 0.6415148886961549,
+      "grad_norm": 1.2763706327996165,
+      "learning_rate": 5.8034062766801544e-06,
+      "loss": 1.5271,
+      "step": 2219
+    },
+    {
+      "epoch": 0.6418039895923677,
+      "grad_norm": 1.0306070061974366,
+      "learning_rate": 5.795079872381488e-06,
+      "loss": 1.5561,
+      "step": 2220
+    },
+    {
+      "epoch": 0.6420930904885805,
+      "grad_norm": 1.0154882488346175,
+      "learning_rate": 5.786757007974341e-06,
+      "loss": 1.5975,
+      "step": 2221
+    },
+    {
+      "epoch": 0.6423821913847932,
+      "grad_norm": 1.019756711690662,
+      "learning_rate": 5.778437690465284e-06,
+      "loss": 1.4048,
+      "step": 2222
+    },
+    {
+      "epoch": 0.6426712922810061,
+      "grad_norm": 1.0275267551765046,
+      "learning_rate": 5.77012192685789e-06,
+      "loss": 1.4779,
+      "step": 2223
+    },
+    {
+      "epoch": 0.6429603931772189,
+      "grad_norm": 1.0205756543110946,
+      "learning_rate": 5.761809724152742e-06,
+      "loss": 1.5125,
+      "step": 2224
+    },
+    {
+      "epoch": 0.6432494940734317,
+      "grad_norm": 1.052441520796048,
+      "learning_rate": 5.7535010893474284e-06,
+      "loss": 1.4749,
+      "step": 2225
+    },
+    {
+      "epoch": 0.6435385949696444,
+      "grad_norm": 1.0248149607415977,
+      "learning_rate": 5.745196029436539e-06,
+      "loss": 1.4762,
+      "step": 2226
+    },
+    {
+      "epoch": 0.6438276958658572,
+      "grad_norm": 1.0711692926454333,
+      "learning_rate": 5.736894551411642e-06,
+      "loss": 1.5347,
+      "step": 2227
+    },
+    {
+      "epoch": 0.64411679676207,
+      "grad_norm": 1.011095118859186,
+      "learning_rate": 5.728596662261296e-06,
+      "loss": 1.5371,
+      "step": 2228
+    },
+    {
+      "epoch": 0.6444058976582827,
+      "grad_norm": 1.0508552730811824,
+      "learning_rate": 5.720302368971042e-06,
+      "loss": 1.5704,
+      "step": 2229
+    },
+    {
+      "epoch": 0.6446949985544955,
+      "grad_norm": 1.0572676578512272,
+      "learning_rate": 5.712011678523389e-06,
+      "loss": 1.4886,
+      "step": 2230
+    },
+    {
+      "epoch": 0.6449840994507083,
+      "grad_norm": 1.0077951405710113,
+      "learning_rate": 5.70372459789781e-06,
+      "loss": 1.5751,
+      "step": 2231
+    },
+    {
+      "epoch": 0.645273200346921,
+      "grad_norm": 0.9990816976270711,
+      "learning_rate": 5.695441134070752e-06,
+      "loss": 1.4933,
+      "step": 2232
+    },
+    {
+      "epoch": 0.6455623012431339,
+      "grad_norm": 1.1135987810384929,
+      "learning_rate": 5.687161294015603e-06,
+      "loss": 1.5239,
+      "step": 2233
+    },
+    {
+      "epoch": 0.6458514021393467,
+      "grad_norm": 1.0451462949364874,
+      "learning_rate": 5.678885084702703e-06,
+      "loss": 1.5518,
+      "step": 2234
+    },
+    {
+      "epoch": 0.6461405030355594,
+      "grad_norm": 1.0743438889754726,
+      "learning_rate": 5.670612513099343e-06,
+      "loss": 1.5345,
+      "step": 2235
+    },
+    {
+      "epoch": 0.6464296039317722,
+      "grad_norm": 1.0176277914271357,
+      "learning_rate": 5.66234358616975e-06,
+      "loss": 1.5341,
+      "step": 2236
+    },
+    {
+      "epoch": 0.646718704827985,
+      "grad_norm": 1.0505974969198069,
+      "learning_rate": 5.6540783108750775e-06,
+      "loss": 1.5187,
+      "step": 2237
+    },
+    {
+      "epoch": 0.6470078057241977,
+      "grad_norm": 1.0744831698740367,
+      "learning_rate": 5.6458166941734e-06,
+      "loss": 1.6101,
+      "step": 2238
+    },
+    {
+      "epoch": 0.6472969066204105,
+      "grad_norm": 1.0529559391043364,
+      "learning_rate": 5.637558743019731e-06,
+      "loss": 1.6057,
+      "step": 2239
+    },
+    {
+      "epoch": 0.6475860075166233,
+      "grad_norm": 1.0659425372971514,
+      "learning_rate": 5.6293044643659815e-06,
+      "loss": 1.5856,
+      "step": 2240
+    },
+    {
+      "epoch": 0.6478751084128361,
+      "grad_norm": 1.02713030529779,
+      "learning_rate": 5.62105386516097e-06,
+      "loss": 1.5281,
+      "step": 2241
+    },
+    {
+      "epoch": 0.6481642093090488,
+      "grad_norm": 1.1162961469298274,
+      "learning_rate": 5.612806952350432e-06,
+      "loss": 1.5655,
+      "step": 2242
+    },
+    {
+      "epoch": 0.6484533102052616,
+      "grad_norm": 1.053344215427422,
+      "learning_rate": 5.604563732876989e-06,
+      "loss": 1.4743,
+      "step": 2243
+    },
+    {
+      "epoch": 0.6487424111014745,
+      "grad_norm": 1.0103964589220165,
+      "learning_rate": 5.596324213680147e-06,
+      "loss": 1.4839,
+      "step": 2244
+    },
+    {
+      "epoch": 0.6490315119976872,
+      "grad_norm": 1.0330538313365736,
+      "learning_rate": 5.588088401696313e-06,
+      "loss": 1.5762,
+      "step": 2245
+    },
+    {
+      "epoch": 0.6493206128939,
+      "grad_norm": 1.0359109012589718,
+      "learning_rate": 5.579856303858769e-06,
+      "loss": 1.5126,
+      "step": 2246
+    },
+    {
+      "epoch": 0.6496097137901128,
+      "grad_norm": 1.0459290799665912,
+      "learning_rate": 5.571627927097658e-06,
+      "loss": 1.5695,
+      "step": 2247
+    },
+    {
+      "epoch": 0.6498988146863255,
+      "grad_norm": 1.1229243040710744,
+      "learning_rate": 5.56340327834e-06,
+      "loss": 1.625,
+      "step": 2248
+    },
+    {
+      "epoch": 0.6501879155825383,
+      "grad_norm": 1.0493741529046865,
+      "learning_rate": 5.5551823645096815e-06,
+      "loss": 1.6494,
+      "step": 2249
+    },
+    {
+      "epoch": 0.6504770164787511,
+      "grad_norm": 1.0241854062871758,
+      "learning_rate": 5.546965192527433e-06,
+      "loss": 1.5266,
+      "step": 2250
+    },
+    {
+      "epoch": 0.6507661173749638,
+      "grad_norm": 1.0392713563824427,
+      "learning_rate": 5.538751769310842e-06,
+      "loss": 1.5222,
+      "step": 2251
+    },
+    {
+      "epoch": 0.6510552182711766,
+      "grad_norm": 0.9726384476275192,
+      "learning_rate": 5.530542101774341e-06,
+      "loss": 1.5151,
+      "step": 2252
+    },
+    {
+      "epoch": 0.6513443191673894,
+      "grad_norm": 0.9840961105459071,
+      "learning_rate": 5.522336196829197e-06,
+      "loss": 1.5393,
+      "step": 2253
+    },
+    {
+      "epoch": 0.6516334200636021,
+      "grad_norm": 1.045692507375498,
+      "learning_rate": 5.514134061383507e-06,
+      "loss": 1.5749,
+      "step": 2254
+    },
+    {
+      "epoch": 0.651922520959815,
+      "grad_norm": 1.0446231572810856,
+      "learning_rate": 5.5059357023422075e-06,
+      "loss": 1.5864,
+      "step": 2255
+    },
+    {
+      "epoch": 0.6522116218560278,
+      "grad_norm": 1.073258389730826,
+      "learning_rate": 5.497741126607038e-06,
+      "loss": 1.5247,
+      "step": 2256
+    },
+    {
+      "epoch": 0.6525007227522406,
+      "grad_norm": 1.0802104082606006,
+      "learning_rate": 5.48955034107657e-06,
+      "loss": 1.5361,
+      "step": 2257
+    },
+    {
+      "epoch": 0.6527898236484533,
+      "grad_norm": 1.0329904583145144,
+      "learning_rate": 5.48136335264617e-06,
+      "loss": 1.601,
+      "step": 2258
+    },
+    {
+      "epoch": 0.6530789245446661,
+      "grad_norm": 1.0337099079890086,
+      "learning_rate": 5.473180168208021e-06,
+      "loss": 1.5326,
+      "step": 2259
+    },
+    {
+      "epoch": 0.6533680254408789,
+      "grad_norm": 0.9929243648424316,
+      "learning_rate": 5.465000794651093e-06,
+      "loss": 1.4794,
+      "step": 2260
+    },
+    {
+      "epoch": 0.6536571263370916,
+      "grad_norm": 1.0358348027055488,
+      "learning_rate": 5.456825238861147e-06,
+      "loss": 1.7051,
+      "step": 2261
+    },
+    {
+      "epoch": 0.6539462272333044,
+      "grad_norm": 1.0129596353460375,
+      "learning_rate": 5.448653507720747e-06,
+      "loss": 1.5184,
+      "step": 2262
+    },
+    {
+      "epoch": 0.6542353281295172,
+      "grad_norm": 1.0218966948409314,
+      "learning_rate": 5.440485608109215e-06,
+      "loss": 1.4542,
+      "step": 2263
+    },
+    {
+      "epoch": 0.6545244290257299,
+      "grad_norm": 1.0464331743959716,
+      "learning_rate": 5.4323215469026615e-06,
+      "loss": 1.513,
+      "step": 2264
+    },
+    {
+      "epoch": 0.6548135299219427,
+      "grad_norm": 1.0178698188075115,
+      "learning_rate": 5.424161330973956e-06,
+      "loss": 1.4295,
+      "step": 2265
+    },
+    {
+      "epoch": 0.6551026308181556,
+      "grad_norm": 1.036811268056617,
+      "learning_rate": 5.416004967192742e-06,
+      "loss": 1.5682,
+      "step": 2266
+    },
+    {
+      "epoch": 0.6553917317143683,
+      "grad_norm": 1.0465495450404148,
+      "learning_rate": 5.407852462425416e-06,
+      "loss": 1.6019,
+      "step": 2267
+    },
+    {
+      "epoch": 0.6556808326105811,
+      "grad_norm": 1.0566604853117803,
+      "learning_rate": 5.399703823535119e-06,
+      "loss": 1.558,
+      "step": 2268
+    },
+    {
+      "epoch": 0.6559699335067939,
+      "grad_norm": 0.9819346657725794,
+      "learning_rate": 5.39155905738175e-06,
+      "loss": 1.5153,
+      "step": 2269
+    },
+    {
+      "epoch": 0.6562590344030067,
+      "grad_norm": 1.0307554535129084,
+      "learning_rate": 5.383418170821937e-06,
+      "loss": 1.5941,
+      "step": 2270
+    },
+    {
+      "epoch": 0.6565481352992194,
+      "grad_norm": 1.0342359311455165,
+      "learning_rate": 5.3752811707090455e-06,
+      "loss": 1.5115,
+      "step": 2271
+    },
+    {
+      "epoch": 0.6568372361954322,
+      "grad_norm": 1.0510658034778178,
+      "learning_rate": 5.367148063893166e-06,
+      "loss": 1.5484,
+      "step": 2272
+    },
+    {
+      "epoch": 0.657126337091645,
+      "grad_norm": 1.0416620525018032,
+      "learning_rate": 5.359018857221121e-06,
+      "loss": 1.6382,
+      "step": 2273
+    },
+    {
+      "epoch": 0.6574154379878577,
+      "grad_norm": 1.0337411373541736,
+      "learning_rate": 5.3508935575364405e-06,
+      "loss": 1.5191,
+      "step": 2274
+    },
+    {
+      "epoch": 0.6577045388840705,
+      "grad_norm": 1.038231107638723,
+      "learning_rate": 5.342772171679364e-06,
+      "loss": 1.6216,
+      "step": 2275
+    },
+    {
+      "epoch": 0.6579936397802834,
+      "grad_norm": 1.0045538537278222,
+      "learning_rate": 5.334654706486852e-06,
+      "loss": 1.5586,
+      "step": 2276
+    },
+    {
+      "epoch": 0.6582827406764961,
+      "grad_norm": 1.03407803641689,
+      "learning_rate": 5.326541168792541e-06,
+      "loss": 1.5025,
+      "step": 2277
+    },
+    {
+      "epoch": 0.6585718415727089,
+      "grad_norm": 1.0374305648905928,
+      "learning_rate": 5.3184315654267824e-06,
+      "loss": 1.5467,
+      "step": 2278
+    },
+    {
+      "epoch": 0.6588609424689217,
+      "grad_norm": 1.0306231531253396,
+      "learning_rate": 5.310325903216598e-06,
+      "loss": 1.5317,
+      "step": 2279
+    },
+    {
+      "epoch": 0.6591500433651344,
+      "grad_norm": 1.0551572797555222,
+      "learning_rate": 5.302224188985708e-06,
+      "loss": 1.5211,
+      "step": 2280
+    },
+    {
+      "epoch": 0.6594391442613472,
+      "grad_norm": 2.828616797296591,
+      "learning_rate": 5.294126429554498e-06,
+      "loss": 1.4956,
+      "step": 2281
+    },
+    {
+      "epoch": 0.65972824515756,
+      "grad_norm": 1.024793297488049,
+      "learning_rate": 5.286032631740023e-06,
+      "loss": 1.6235,
+      "step": 2282
+    },
+    {
+      "epoch": 0.6600173460537727,
+      "grad_norm": 1.0392615786594692,
+      "learning_rate": 5.2779428023560155e-06,
+      "loss": 1.7145,
+      "step": 2283
+    },
+    {
+      "epoch": 0.6603064469499855,
+      "grad_norm": 1.0305689227356953,
+      "learning_rate": 5.2698569482128545e-06,
+      "loss": 1.5058,
+      "step": 2284
+    },
+    {
+      "epoch": 0.6605955478461983,
+      "grad_norm": 1.0666739856448908,
+      "learning_rate": 5.261775076117572e-06,
+      "loss": 1.5765,
+      "step": 2285
+    },
+    {
+      "epoch": 0.6608846487424112,
+      "grad_norm": 1.1093333890523078,
+      "learning_rate": 5.253697192873862e-06,
+      "loss": 1.5596,
+      "step": 2286
+    },
+    {
+      "epoch": 0.6611737496386239,
+      "grad_norm": 1.08041140619824,
+      "learning_rate": 5.245623305282047e-06,
+      "loss": 1.5187,
+      "step": 2287
+    },
+    {
+      "epoch": 0.6614628505348367,
+      "grad_norm": 1.040175779794725,
+      "learning_rate": 5.237553420139085e-06,
+      "loss": 1.4988,
+      "step": 2288
+    },
+    {
+      "epoch": 0.6617519514310495,
+      "grad_norm": 1.0401372057234441,
+      "learning_rate": 5.229487544238575e-06,
+      "loss": 1.5651,
+      "step": 2289
+    },
+    {
+      "epoch": 0.6620410523272622,
+      "grad_norm": 1.0350179165707336,
+      "learning_rate": 5.221425684370738e-06,
+      "loss": 1.5661,
+      "step": 2290
+    },
+    {
+      "epoch": 0.662330153223475,
+      "grad_norm": 1.0355019022237686,
+      "learning_rate": 5.213367847322408e-06,
+      "loss": 1.5656,
+      "step": 2291
+    },
+    {
+      "epoch": 0.6626192541196878,
+      "grad_norm": 1.0160043691774807,
+      "learning_rate": 5.205314039877032e-06,
+      "loss": 1.4874,
+      "step": 2292
+    },
+    {
+      "epoch": 0.6629083550159005,
+      "grad_norm": 1.0451799240433048,
+      "learning_rate": 5.197264268814677e-06,
+      "loss": 1.5759,
+      "step": 2293
+    },
+    {
+      "epoch": 0.6631974559121133,
+      "grad_norm": 1.0462695205922974,
+      "learning_rate": 5.189218540911998e-06,
+      "loss": 1.6112,
+      "step": 2294
+    },
+    {
+      "epoch": 0.6634865568083261,
+      "grad_norm": 1.0391344011346086,
+      "learning_rate": 5.181176862942248e-06,
+      "loss": 1.5796,
+      "step": 2295
+    },
+    {
+      "epoch": 0.6637756577045388,
+      "grad_norm": 1.0178243998506182,
+      "learning_rate": 5.173139241675284e-06,
+      "loss": 1.6569,
+      "step": 2296
+    },
+    {
+      "epoch": 0.6640647586007516,
+      "grad_norm": 1.0572676943518013,
+      "learning_rate": 5.165105683877534e-06,
+      "loss": 1.5945,
+      "step": 2297
+    },
+    {
+      "epoch": 0.6643538594969645,
+      "grad_norm": 1.0367646477827481,
+      "learning_rate": 5.157076196312004e-06,
+      "loss": 1.6672,
+      "step": 2298
+    },
+    {
+      "epoch": 0.6646429603931773,
+      "grad_norm": 1.0295421455972824,
+      "learning_rate": 5.149050785738286e-06,
+      "loss": 1.5596,
+      "step": 2299
+    },
+    {
+      "epoch": 0.66493206128939,
+      "grad_norm": 1.022416135838012,
+      "learning_rate": 5.141029458912532e-06,
+      "loss": 1.5207,
+      "step": 2300
+    },
+    {
+      "epoch": 0.6652211621856028,
+      "grad_norm": 0.9957736037252973,
+      "learning_rate": 5.133012222587457e-06,
+      "loss": 1.4816,
+      "step": 2301
+    },
+    {
+      "epoch": 0.6655102630818156,
+      "grad_norm": 1.0450746953349128,
+      "learning_rate": 5.124999083512327e-06,
+      "loss": 1.5953,
+      "step": 2302
+    },
+    {
+      "epoch": 0.6657993639780283,
+      "grad_norm": 0.9932883921083772,
+      "learning_rate": 5.116990048432972e-06,
+      "loss": 1.4171,
+      "step": 2303
+    },
+    {
+      "epoch": 0.6660884648742411,
+      "grad_norm": 1.1721358139530453,
+      "learning_rate": 5.108985124091754e-06,
+      "loss": 1.5822,
+      "step": 2304
+    },
+    {
+      "epoch": 0.6663775657704539,
+      "grad_norm": 1.041027464141664,
+      "learning_rate": 5.10098431722758e-06,
+      "loss": 1.4945,
+      "step": 2305
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 1.0108807575296979,
+      "learning_rate": 5.092987634575892e-06,
+      "loss": 1.5891,
+      "step": 2306
+    },
+    {
+      "epoch": 0.6669557675628794,
+      "grad_norm": 1.0474468855646906,
+      "learning_rate": 5.084995082868658e-06,
+      "loss": 1.5036,
+      "step": 2307
+    },
+    {
+      "epoch": 0.6672448684590923,
+      "grad_norm": 1.0360251112993981,
+      "learning_rate": 5.0770066688343655e-06,
+      "loss": 1.5251,
+      "step": 2308
+    },
+    {
+      "epoch": 0.667533969355305,
+      "grad_norm": 1.0346909583162316,
+      "learning_rate": 5.069022399198023e-06,
+      "loss": 1.5584,
+      "step": 2309
+    },
+    {
+      "epoch": 0.6678230702515178,
+      "grad_norm": 1.0445088955190482,
+      "learning_rate": 5.061042280681156e-06,
+      "loss": 1.5882,
+      "step": 2310
+    },
+    {
+      "epoch": 0.6681121711477306,
+      "grad_norm": 1.0419868379942245,
+      "learning_rate": 5.053066320001783e-06,
+      "loss": 1.5797,
+      "step": 2311
+    },
+    {
+      "epoch": 0.6684012720439433,
+      "grad_norm": 1.0689415965408784,
+      "learning_rate": 5.045094523874423e-06,
+      "loss": 1.5901,
+      "step": 2312
+    },
+    {
+      "epoch": 0.6686903729401561,
+      "grad_norm": 1.054062687432979,
+      "learning_rate": 5.037126899010102e-06,
+      "loss": 1.6123,
+      "step": 2313
+    },
+    {
+      "epoch": 0.6689794738363689,
+      "grad_norm": 1.010200152456534,
+      "learning_rate": 5.029163452116322e-06,
+      "loss": 1.6666,
+      "step": 2314
+    },
+    {
+      "epoch": 0.6692685747325817,
+      "grad_norm": 1.0562511421382814,
+      "learning_rate": 5.02120418989707e-06,
+      "loss": 1.5017,
+      "step": 2315
+    },
+    {
+      "epoch": 0.6695576756287944,
+      "grad_norm": 1.0242903368576202,
+      "learning_rate": 5.013249119052817e-06,
+      "loss": 1.4597,
+      "step": 2316
+    },
+    {
+      "epoch": 0.6698467765250072,
+      "grad_norm": 1.0476073821384742,
+      "learning_rate": 5.0052982462805e-06,
+      "loss": 1.5468,
+      "step": 2317
+    },
+    {
+      "epoch": 0.67013587742122,
+      "grad_norm": 1.0232225615743717,
+      "learning_rate": 4.997351578273516e-06,
+      "loss": 1.5295,
+      "step": 2318
+    },
+    {
+      "epoch": 0.6704249783174328,
+      "grad_norm": 1.0306163669472685,
+      "learning_rate": 4.989409121721738e-06,
+      "loss": 1.5855,
+      "step": 2319
+    },
+    {
+      "epoch": 0.6707140792136456,
+      "grad_norm": 1.0768066289400315,
+      "learning_rate": 4.981470883311474e-06,
+      "loss": 1.567,
+      "step": 2320
+    },
+    {
+      "epoch": 0.6710031801098584,
+      "grad_norm": 1.0435000495999243,
+      "learning_rate": 4.973536869725503e-06,
+      "loss": 1.5563,
+      "step": 2321
+    },
+    {
+      "epoch": 0.6712922810060711,
+      "grad_norm": 0.9893580173479523,
+      "learning_rate": 4.965607087643025e-06,
+      "loss": 1.4767,
+      "step": 2322
+    },
+    {
+      "epoch": 0.6715813819022839,
+      "grad_norm": 1.0096952606451866,
+      "learning_rate": 4.957681543739697e-06,
+      "loss": 1.6457,
+      "step": 2323
+    },
+    {
+      "epoch": 0.6718704827984967,
+      "grad_norm": 1.0251399506142898,
+      "learning_rate": 4.949760244687597e-06,
+      "loss": 1.6713,
+      "step": 2324
+    },
+    {
+      "epoch": 0.6721595836947094,
+      "grad_norm": 1.052617768972847,
+      "learning_rate": 4.941843197155226e-06,
+      "loss": 1.6225,
+      "step": 2325
+    },
+    {
+      "epoch": 0.6724486845909222,
+      "grad_norm": 1.0165209467121914,
+      "learning_rate": 4.933930407807522e-06,
+      "loss": 1.4723,
+      "step": 2326
+    },
+    {
+      "epoch": 0.672737785487135,
+      "grad_norm": 1.019574386937902,
+      "learning_rate": 4.926021883305822e-06,
+      "loss": 1.5238,
+      "step": 2327
+    },
+    {
+      "epoch": 0.6730268863833477,
+      "grad_norm": 1.036542043454047,
+      "learning_rate": 4.918117630307878e-06,
+      "loss": 1.5995,
+      "step": 2328
+    },
+    {
+      "epoch": 0.6733159872795605,
+      "grad_norm": 1.0183836724208408,
+      "learning_rate": 4.910217655467847e-06,
+      "loss": 1.5024,
+      "step": 2329
+    },
+    {
+      "epoch": 0.6736050881757734,
+      "grad_norm": 1.003401896724112,
+      "learning_rate": 4.9023219654362885e-06,
+      "loss": 1.4684,
+      "step": 2330
+    },
+    {
+      "epoch": 0.6738941890719862,
+      "grad_norm": 1.0320062806366481,
+      "learning_rate": 4.894430566860144e-06,
+      "loss": 1.5274,
+      "step": 2331
+    },
+    {
+      "epoch": 0.6741832899681989,
+      "grad_norm": 1.042311002731924,
+      "learning_rate": 4.886543466382755e-06,
+      "loss": 1.6325,
+      "step": 2332
+    },
+    {
+      "epoch": 0.6744723908644117,
+      "grad_norm": 1.2642107854161773,
+      "learning_rate": 4.8786606706438385e-06,
+      "loss": 1.5938,
+      "step": 2333
+    },
+    {
+      "epoch": 0.6747614917606245,
+      "grad_norm": 1.03231809082626,
+      "learning_rate": 4.8707821862794866e-06,
+      "loss": 1.5958,
+      "step": 2334
+    },
+    {
+      "epoch": 0.6750505926568372,
+      "grad_norm": 0.9912462119062858,
+      "learning_rate": 4.862908019922162e-06,
+      "loss": 1.5314,
+      "step": 2335
+    },
+    {
+      "epoch": 0.67533969355305,
+      "grad_norm": 1.0380736449714523,
+      "learning_rate": 4.8550381782006895e-06,
+      "loss": 1.5151,
+      "step": 2336
+    },
+    {
+      "epoch": 0.6756287944492628,
+      "grad_norm": 0.9875258341933627,
+      "learning_rate": 4.847172667740265e-06,
+      "loss": 1.5915,
+      "step": 2337
+    },
+    {
+      "epoch": 0.6759178953454755,
+      "grad_norm": 1.0105730190287259,
+      "learning_rate": 4.8393114951624265e-06,
+      "loss": 1.5753,
+      "step": 2338
+    },
+    {
+      "epoch": 0.6762069962416883,
+      "grad_norm": 1.0170294961097692,
+      "learning_rate": 4.831454667085059e-06,
+      "loss": 1.6208,
+      "step": 2339
+    },
+    {
+      "epoch": 0.6764960971379012,
+      "grad_norm": 1.0193982194578417,
+      "learning_rate": 4.823602190122405e-06,
+      "loss": 1.5311,
+      "step": 2340
+    },
+    {
+      "epoch": 0.6767851980341139,
+      "grad_norm": 1.0242416204521136,
+      "learning_rate": 4.8157540708850225e-06,
+      "loss": 1.4932,
+      "step": 2341
+    },
+    {
+      "epoch": 0.6770742989303267,
+      "grad_norm": 1.002395706889193,
+      "learning_rate": 4.8079103159798236e-06,
+      "loss": 1.4969,
+      "step": 2342
+    },
+    {
+      "epoch": 0.6773633998265395,
+      "grad_norm": 1.0101478170816376,
+      "learning_rate": 4.8000709320100255e-06,
+      "loss": 1.4534,
+      "step": 2343
+    },
+    {
+      "epoch": 0.6776525007227523,
+      "grad_norm": 1.0290520037774804,
+      "learning_rate": 4.792235925575184e-06,
+      "loss": 1.5145,
+      "step": 2344
+    },
+    {
+      "epoch": 0.677941601618965,
+      "grad_norm": 1.0876931851183416,
+      "learning_rate": 4.784405303271157e-06,
+      "loss": 1.5995,
+      "step": 2345
+    },
+    {
+      "epoch": 0.6782307025151778,
+      "grad_norm": 1.005962861032748,
+      "learning_rate": 4.776579071690111e-06,
+      "loss": 1.5286,
+      "step": 2346
+    },
+    {
+      "epoch": 0.6785198034113906,
+      "grad_norm": 1.018535408137834,
+      "learning_rate": 4.76875723742053e-06,
+      "loss": 1.5271,
+      "step": 2347
+    },
+    {
+      "epoch": 0.6788089043076033,
+      "grad_norm": 1.0258674495175568,
+      "learning_rate": 4.760939807047183e-06,
+      "loss": 1.5129,
+      "step": 2348
+    },
+    {
+      "epoch": 0.6790980052038161,
+      "grad_norm": 1.0437520512227207,
+      "learning_rate": 4.75312678715113e-06,
+      "loss": 1.5793,
+      "step": 2349
+    },
+    {
+      "epoch": 0.679387106100029,
+      "grad_norm": 1.0244665741629266,
+      "learning_rate": 4.745318184309734e-06,
+      "loss": 1.6247,
+      "step": 2350
+    },
+    {
+      "epoch": 0.6796762069962416,
+      "grad_norm": 0.9921278985377041,
+      "learning_rate": 4.737514005096624e-06,
+      "loss": 1.4425,
+      "step": 2351
+    },
+    {
+      "epoch": 0.6799653078924545,
+      "grad_norm": 1.0225104346123235,
+      "learning_rate": 4.729714256081708e-06,
+      "loss": 1.4479,
+      "step": 2352
+    },
+    {
+      "epoch": 0.6802544087886673,
+      "grad_norm": 1.0084711222827276,
+      "learning_rate": 4.7219189438311674e-06,
+      "loss": 1.5562,
+      "step": 2353
+    },
+    {
+      "epoch": 0.68054350968488,
+      "grad_norm": 1.0742092881303418,
+      "learning_rate": 4.714128074907455e-06,
+      "loss": 1.4834,
+      "step": 2354
+    },
+    {
+      "epoch": 0.6808326105810928,
+      "grad_norm": 1.0503287838138804,
+      "learning_rate": 4.7063416558692694e-06,
+      "loss": 1.484,
+      "step": 2355
+    },
+    {
+      "epoch": 0.6811217114773056,
+      "grad_norm": 1.0193473690969417,
+      "learning_rate": 4.6985596932715685e-06,
+      "loss": 1.5557,
+      "step": 2356
+    },
+    {
+      "epoch": 0.6814108123735183,
+      "grad_norm": 1.0125686030172614,
+      "learning_rate": 4.690782193665564e-06,
+      "loss": 1.5907,
+      "step": 2357
+    },
+    {
+      "epoch": 0.6816999132697311,
+      "grad_norm": 1.0318830598950361,
+      "learning_rate": 4.683009163598705e-06,
+      "loss": 1.5467,
+      "step": 2358
+    },
+    {
+      "epoch": 0.6819890141659439,
+      "grad_norm": 1.021276150831182,
+      "learning_rate": 4.675240609614673e-06,
+      "loss": 1.5457,
+      "step": 2359
+    },
+    {
+      "epoch": 0.6822781150621567,
+      "grad_norm": 1.0060655627936956,
+      "learning_rate": 4.667476538253397e-06,
+      "loss": 1.5569,
+      "step": 2360
+    },
+    {
+      "epoch": 0.6825672159583694,
+      "grad_norm": 1.0561498386360748,
+      "learning_rate": 4.6597169560510165e-06,
+      "loss": 1.5813,
+      "step": 2361
+    },
+    {
+      "epoch": 0.6828563168545823,
+      "grad_norm": 1.0603230402667239,
+      "learning_rate": 4.651961869539897e-06,
+      "loss": 1.5291,
+      "step": 2362
+    },
+    {
+      "epoch": 0.6831454177507951,
+      "grad_norm": 1.0385844283945216,
+      "learning_rate": 4.6442112852486225e-06,
+      "loss": 1.4986,
+      "step": 2363
+    },
+    {
+      "epoch": 0.6834345186470078,
+      "grad_norm": 1.0417106932902274,
+      "learning_rate": 4.636465209701987e-06,
+      "loss": 1.6389,
+      "step": 2364
+    },
+    {
+      "epoch": 0.6837236195432206,
+      "grad_norm": 1.016124378153035,
+      "learning_rate": 4.6287236494209855e-06,
+      "loss": 1.595,
+      "step": 2365
+    },
+    {
+      "epoch": 0.6840127204394334,
+      "grad_norm": 1.0643419311389555,
+      "learning_rate": 4.6209866109228075e-06,
+      "loss": 1.572,
+      "step": 2366
+    },
+    {
+      "epoch": 0.6843018213356461,
+      "grad_norm": 1.10830958567818,
+      "learning_rate": 4.613254100720849e-06,
+      "loss": 1.6202,
+      "step": 2367
+    },
+    {
+      "epoch": 0.6845909222318589,
+      "grad_norm": 0.9997640054722354,
+      "learning_rate": 4.605526125324682e-06,
+      "loss": 1.5104,
+      "step": 2368
+    },
+    {
+      "epoch": 0.6848800231280717,
+      "grad_norm": 1.02237845941241,
+      "learning_rate": 4.597802691240061e-06,
+      "loss": 1.5857,
+      "step": 2369
+    },
+    {
+      "epoch": 0.6851691240242844,
+      "grad_norm": 1.0668584175008946,
+      "learning_rate": 4.59008380496893e-06,
+      "loss": 1.742,
+      "step": 2370
+    },
+    {
+      "epoch": 0.6854582249204972,
+      "grad_norm": 1.0391168802614548,
+      "learning_rate": 4.58236947300939e-06,
+      "loss": 1.6097,
+      "step": 2371
+    },
+    {
+      "epoch": 0.68574732581671,
+      "grad_norm": 1.0528072948194604,
+      "learning_rate": 4.574659701855712e-06,
+      "loss": 1.6592,
+      "step": 2372
+    },
+    {
+      "epoch": 0.6860364267129229,
+      "grad_norm": 1.0445602174201212,
+      "learning_rate": 4.566954497998333e-06,
+      "loss": 1.5357,
+      "step": 2373
+    },
+    {
+      "epoch": 0.6863255276091356,
+      "grad_norm": 1.039940126769689,
+      "learning_rate": 4.559253867923844e-06,
+      "loss": 1.5413,
+      "step": 2374
+    },
+    {
+      "epoch": 0.6866146285053484,
+      "grad_norm": 1.0300329456252075,
+      "learning_rate": 4.55155781811498e-06,
+      "loss": 1.5712,
+      "step": 2375
+    },
+    {
+      "epoch": 0.6869037294015612,
+      "grad_norm": 1.1126942824755646,
+      "learning_rate": 4.543866355050618e-06,
+      "loss": 1.5542,
+      "step": 2376
+    },
+    {
+      "epoch": 0.6871928302977739,
+      "grad_norm": 1.1170430159096856,
+      "learning_rate": 4.536179485205788e-06,
+      "loss": 1.5524,
+      "step": 2377
+    },
+    {
+      "epoch": 0.6874819311939867,
+      "grad_norm": 1.0385144697871138,
+      "learning_rate": 4.528497215051638e-06,
+      "loss": 1.4586,
+      "step": 2378
+    },
+    {
+      "epoch": 0.6877710320901995,
+      "grad_norm": 1.0449199693612592,
+      "learning_rate": 4.520819551055449e-06,
+      "loss": 1.6578,
+      "step": 2379
+    },
+    {
+      "epoch": 0.6880601329864122,
+      "grad_norm": 1.085297632870415,
+      "learning_rate": 4.513146499680629e-06,
+      "loss": 1.6205,
+      "step": 2380
+    },
+    {
+      "epoch": 0.688349233882625,
+      "grad_norm": 1.0670124175732796,
+      "learning_rate": 4.505478067386698e-06,
+      "loss": 1.5948,
+      "step": 2381
+    },
+    {
+      "epoch": 0.6886383347788378,
+      "grad_norm": 0.9951490533296982,
+      "learning_rate": 4.497814260629287e-06,
+      "loss": 1.5572,
+      "step": 2382
+    },
+    {
+      "epoch": 0.6889274356750505,
+      "grad_norm": 1.0502242826848907,
+      "learning_rate": 4.490155085860131e-06,
+      "loss": 1.5831,
+      "step": 2383
+    },
+    {
+      "epoch": 0.6892165365712634,
+      "grad_norm": 1.0646448629034688,
+      "learning_rate": 4.482500549527072e-06,
+      "loss": 1.4825,
+      "step": 2384
+    },
+    {
+      "epoch": 0.6895056374674762,
+      "grad_norm": 1.0237924438956325,
+      "learning_rate": 4.474850658074049e-06,
+      "loss": 1.5236,
+      "step": 2385
+    },
+    {
+      "epoch": 0.6897947383636889,
+      "grad_norm": 1.04519663463218,
+      "learning_rate": 4.467205417941079e-06,
+      "loss": 1.5349,
+      "step": 2386
+    },
+    {
+      "epoch": 0.6900838392599017,
+      "grad_norm": 1.0560012002364279,
+      "learning_rate": 4.459564835564275e-06,
+      "loss": 1.5522,
+      "step": 2387
+    },
+    {
+      "epoch": 0.6903729401561145,
+      "grad_norm": 1.0549679390875804,
+      "learning_rate": 4.451928917375822e-06,
+      "loss": 1.6093,
+      "step": 2388
+    },
+    {
+      "epoch": 0.6906620410523273,
+      "grad_norm": 1.047470187669572,
+      "learning_rate": 4.444297669803981e-06,
+      "loss": 1.5184,
+      "step": 2389
+    },
+    {
+      "epoch": 0.69095114194854,
+      "grad_norm": 1.0073355487874749,
+      "learning_rate": 4.4366710992730755e-06,
+      "loss": 1.5699,
+      "step": 2390
+    },
+    {
+      "epoch": 0.6912402428447528,
+      "grad_norm": 1.0586454089185726,
+      "learning_rate": 4.429049212203505e-06,
+      "loss": 1.5931,
+      "step": 2391
+    },
+    {
+      "epoch": 0.6915293437409656,
+      "grad_norm": 1.0131150636575983,
+      "learning_rate": 4.421432015011714e-06,
+      "loss": 1.5454,
+      "step": 2392
+    },
+    {
+      "epoch": 0.6918184446371783,
+      "grad_norm": 1.0899466632803332,
+      "learning_rate": 4.4138195141102e-06,
+      "loss": 1.6228,
+      "step": 2393
+    },
+    {
+      "epoch": 0.6921075455333912,
+      "grad_norm": 0.9945022757384525,
+      "learning_rate": 4.406211715907516e-06,
+      "loss": 1.5439,
+      "step": 2394
+    },
+    {
+      "epoch": 0.692396646429604,
+      "grad_norm": 1.0200980586300865,
+      "learning_rate": 4.3986086268082425e-06,
+      "loss": 1.5775,
+      "step": 2395
+    },
+    {
+      "epoch": 0.6926857473258167,
+      "grad_norm": 1.0338287794153778,
+      "learning_rate": 4.391010253213008e-06,
+      "loss": 1.5849,
+      "step": 2396
+    },
+    {
+      "epoch": 0.6929748482220295,
+      "grad_norm": 1.0469232828651016,
+      "learning_rate": 4.3834166015184695e-06,
+      "loss": 1.6229,
+      "step": 2397
+    },
+    {
+      "epoch": 0.6932639491182423,
+      "grad_norm": 1.0227040850405649,
+      "learning_rate": 4.375827678117302e-06,
+      "loss": 1.4579,
+      "step": 2398
+    },
+    {
+      "epoch": 0.693553050014455,
+      "grad_norm": 1.0548718404349842,
+      "learning_rate": 4.368243489398202e-06,
+      "loss": 1.5667,
+      "step": 2399
+    },
+    {
+      "epoch": 0.6938421509106678,
+      "grad_norm": 1.0707551462593596,
+      "learning_rate": 4.360664041745879e-06,
+      "loss": 1.5685,
+      "step": 2400
+    },
+    {
+      "epoch": 0.6941312518068806,
+      "grad_norm": 1.0649182303277598,
+      "learning_rate": 4.3530893415410635e-06,
+      "loss": 1.5323,
+      "step": 2401
+    },
+    {
+      "epoch": 0.6944203527030933,
+      "grad_norm": 1.0929937639269471,
+      "learning_rate": 4.345519395160473e-06,
+      "loss": 1.6142,
+      "step": 2402
+    },
+    {
+      "epoch": 0.6947094535993061,
+      "grad_norm": 4.354929899798579,
+      "learning_rate": 4.33795420897683e-06,
+      "loss": 1.6584,
+      "step": 2403
+    },
+    {
+      "epoch": 0.694998554495519,
+      "grad_norm": 1.0658327982988558,
+      "learning_rate": 4.330393789358854e-06,
+      "loss": 1.5474,
+      "step": 2404
+    },
+    {
+      "epoch": 0.6952876553917318,
+      "grad_norm": 1.028922864063671,
+      "learning_rate": 4.322838142671245e-06,
+      "loss": 1.4661,
+      "step": 2405
+    },
+    {
+      "epoch": 0.6955767562879445,
+      "grad_norm": 1.0404592981349243,
+      "learning_rate": 4.315287275274683e-06,
+      "loss": 1.5933,
+      "step": 2406
+    },
+    {
+      "epoch": 0.6958658571841573,
+      "grad_norm": 1.0417977770127251,
+      "learning_rate": 4.307741193525834e-06,
+      "loss": 1.5458,
+      "step": 2407
+    },
+    {
+      "epoch": 0.6961549580803701,
+      "grad_norm": 1.0365204058301623,
+      "learning_rate": 4.300199903777332e-06,
+      "loss": 1.5067,
+      "step": 2408
+    },
+    {
+      "epoch": 0.6964440589765828,
+      "grad_norm": 1.5908003186470059,
+      "learning_rate": 4.292663412377772e-06,
+      "loss": 1.6107,
+      "step": 2409
+    },
+    {
+      "epoch": 0.6967331598727956,
+      "grad_norm": 1.0512627606443825,
+      "learning_rate": 4.2851317256717095e-06,
+      "loss": 1.653,
+      "step": 2410
+    },
+    {
+      "epoch": 0.6970222607690084,
+      "grad_norm": 1.0185904812927407,
+      "learning_rate": 4.277604849999666e-06,
+      "loss": 1.5899,
+      "step": 2411
+    },
+    {
+      "epoch": 0.6973113616652211,
+      "grad_norm": 1.0133930937344175,
+      "learning_rate": 4.2700827916981e-06,
+      "loss": 1.5608,
+      "step": 2412
+    },
+    {
+      "epoch": 0.6976004625614339,
+      "grad_norm": 0.9934284745282551,
+      "learning_rate": 4.262565557099416e-06,
+      "loss": 1.516,
+      "step": 2413
+    },
+    {
+      "epoch": 0.6978895634576467,
+      "grad_norm": 1.035388323831418,
+      "learning_rate": 4.25505315253197e-06,
+      "loss": 1.5493,
+      "step": 2414
+    },
+    {
+      "epoch": 0.6981786643538594,
+      "grad_norm": 1.0308956524031305,
+      "learning_rate": 4.2475455843200384e-06,
+      "loss": 1.5737,
+      "step": 2415
+    },
+    {
+      "epoch": 0.6984677652500723,
+      "grad_norm": 1.0255838677426103,
+      "learning_rate": 4.2400428587838285e-06,
+      "loss": 1.567,
+      "step": 2416
+    },
+    {
+      "epoch": 0.6987568661462851,
+      "grad_norm": 1.0341432040273755,
+      "learning_rate": 4.232544982239476e-06,
+      "loss": 1.537,
+      "step": 2417
+    },
+    {
+      "epoch": 0.6990459670424979,
+      "grad_norm": 1.0420460422111224,
+      "learning_rate": 4.2250519609990345e-06,
+      "loss": 1.6168,
+      "step": 2418
+    },
+    {
+      "epoch": 0.6993350679387106,
+      "grad_norm": 0.9993193265147904,
+      "learning_rate": 4.2175638013704655e-06,
+      "loss": 1.5017,
+      "step": 2419
+    },
+    {
+      "epoch": 0.6996241688349234,
+      "grad_norm": 0.9972628886488659,
+      "learning_rate": 4.210080509657635e-06,
+      "loss": 1.6023,
+      "step": 2420
+    },
+    {
+      "epoch": 0.6999132697311362,
+      "grad_norm": 1.03747460891003,
+      "learning_rate": 4.202602092160324e-06,
+      "loss": 1.581,
+      "step": 2421
+    },
+    {
+      "epoch": 0.7002023706273489,
+      "grad_norm": 1.0055218420280065,
+      "learning_rate": 4.195128555174196e-06,
+      "loss": 1.4819,
+      "step": 2422
+    },
+    {
+      "epoch": 0.7004914715235617,
+      "grad_norm": 1.0607000100302022,
+      "learning_rate": 4.187659904990811e-06,
+      "loss": 1.5735,
+      "step": 2423
+    },
+    {
+      "epoch": 0.7007805724197745,
+      "grad_norm": 1.0720978373634167,
+      "learning_rate": 4.18019614789762e-06,
+      "loss": 1.491,
+      "step": 2424
+    },
+    {
+      "epoch": 0.7010696733159872,
+      "grad_norm": 1.0398723457906462,
+      "learning_rate": 4.172737290177949e-06,
+      "loss": 1.5094,
+      "step": 2425
+    },
+    {
+      "epoch": 0.7013587742122,
+      "grad_norm": 1.0554189593131018,
+      "learning_rate": 4.165283338110998e-06,
+      "loss": 1.5263,
+      "step": 2426
+    },
+    {
+      "epoch": 0.7016478751084129,
+      "grad_norm": 0.9907071661041562,
+      "learning_rate": 4.157834297971842e-06,
+      "loss": 1.4625,
+      "step": 2427
+    },
+    {
+      "epoch": 0.7019369760046256,
+      "grad_norm": 1.012480334544803,
+      "learning_rate": 4.150390176031423e-06,
+      "loss": 1.4523,
+      "step": 2428
+    },
+    {
+      "epoch": 0.7022260769008384,
+      "grad_norm": 1.0082727758787573,
+      "learning_rate": 4.142950978556537e-06,
+      "loss": 1.4929,
+      "step": 2429
+    },
+    {
+      "epoch": 0.7025151777970512,
+      "grad_norm": 1.0093576002362032,
+      "learning_rate": 4.135516711809829e-06,
+      "loss": 1.3949,
+      "step": 2430
+    },
+    {
+      "epoch": 0.7028042786932639,
+      "grad_norm": 1.024550178487323,
+      "learning_rate": 4.128087382049813e-06,
+      "loss": 1.5691,
+      "step": 2431
+    },
+    {
+      "epoch": 0.7030933795894767,
+      "grad_norm": 1.034850391716068,
+      "learning_rate": 4.1206629955308244e-06,
+      "loss": 1.466,
+      "step": 2432
+    },
+    {
+      "epoch": 0.7033824804856895,
+      "grad_norm": 14.4072134447479,
+      "learning_rate": 4.113243558503046e-06,
+      "loss": 1.4912,
+      "step": 2433
+    },
+    {
+      "epoch": 0.7036715813819023,
+      "grad_norm": 1.0114171822065283,
+      "learning_rate": 4.105829077212503e-06,
+      "loss": 1.5595,
+      "step": 2434
+    },
+    {
+      "epoch": 0.703960682278115,
+      "grad_norm": 1.051652587230225,
+      "learning_rate": 4.098419557901036e-06,
+      "loss": 1.5612,
+      "step": 2435
+    },
+    {
+      "epoch": 0.7042497831743278,
+      "grad_norm": 1.0525582170224568,
+      "learning_rate": 4.091015006806307e-06,
+      "loss": 1.5588,
+      "step": 2436
+    },
+    {
+      "epoch": 0.7045388840705407,
+      "grad_norm": 1.0088652557518278,
+      "learning_rate": 4.083615430161812e-06,
+      "loss": 1.5189,
+      "step": 2437
+    },
+    {
+      "epoch": 0.7048279849667534,
+      "grad_norm": 1.0234156287984635,
+      "learning_rate": 4.076220834196838e-06,
+      "loss": 1.4895,
+      "step": 2438
+    },
+    {
+      "epoch": 0.7051170858629662,
+      "grad_norm": 1.0601428555500987,
+      "learning_rate": 4.068831225136499e-06,
+      "loss": 1.5216,
+      "step": 2439
+    },
+    {
+      "epoch": 0.705406186759179,
+      "grad_norm": 1.018235988570744,
+      "learning_rate": 4.0614466092016935e-06,
+      "loss": 1.5853,
+      "step": 2440
+    },
+    {
+      "epoch": 0.7056952876553917,
+      "grad_norm": 1.01900144875426,
+      "learning_rate": 4.054066992609132e-06,
+      "loss": 1.5535,
+      "step": 2441
+    },
+    {
+      "epoch": 0.7059843885516045,
+      "grad_norm": 1.0534554446252693,
+      "learning_rate": 4.046692381571304e-06,
+      "loss": 1.5674,
+      "step": 2442
+    },
+    {
+      "epoch": 0.7062734894478173,
+      "grad_norm": 1.0458125786452797,
+      "learning_rate": 4.039322782296489e-06,
+      "loss": 1.5137,
+      "step": 2443
+    },
+    {
+      "epoch": 0.70656259034403,
+      "grad_norm": 1.0337416398933583,
+      "learning_rate": 4.031958200988752e-06,
+      "loss": 1.5917,
+      "step": 2444
+    },
+    {
+      "epoch": 0.7068516912402428,
+      "grad_norm": 1.021652215495229,
+      "learning_rate": 4.024598643847928e-06,
+      "loss": 1.5819,
+      "step": 2445
+    },
+    {
+      "epoch": 0.7071407921364556,
+      "grad_norm": 1.008965303229469,
+      "learning_rate": 4.017244117069625e-06,
+      "loss": 1.6137,
+      "step": 2446
+    },
+    {
+      "epoch": 0.7074298930326685,
+      "grad_norm": 1.0341798587977602,
+      "learning_rate": 4.009894626845211e-06,
+      "loss": 1.643,
+      "step": 2447
+    },
+    {
+      "epoch": 0.7077189939288812,
+      "grad_norm": 1.0236808116840943,
+      "learning_rate": 4.002550179361821e-06,
+      "loss": 1.5508,
+      "step": 2448
+    },
+    {
+      "epoch": 0.708008094825094,
+      "grad_norm": 1.0177730066925021,
+      "learning_rate": 3.995210780802349e-06,
+      "loss": 1.5428,
+      "step": 2449
+    },
+    {
+      "epoch": 0.7082971957213068,
+      "grad_norm": 1.0162363315570373,
+      "learning_rate": 3.987876437345421e-06,
+      "loss": 1.5158,
+      "step": 2450
+    },
+    {
+      "epoch": 0.7085862966175195,
+      "grad_norm": 1.0701796601116178,
+      "learning_rate": 3.980547155165429e-06,
+      "loss": 1.6247,
+      "step": 2451
+    },
+    {
+      "epoch": 0.7088753975137323,
+      "grad_norm": 1.0491617189094606,
+      "learning_rate": 3.973222940432488e-06,
+      "loss": 1.4697,
+      "step": 2452
+    },
+    {
+      "epoch": 0.7091644984099451,
+      "grad_norm": 1.057283541789337,
+      "learning_rate": 3.965903799312453e-06,
+      "loss": 1.6398,
+      "step": 2453
+    },
+    {
+      "epoch": 0.7094535993061578,
+      "grad_norm": 1.0014478307417924,
+      "learning_rate": 3.9585897379669025e-06,
+      "loss": 1.517,
+      "step": 2454
+    },
+    {
+      "epoch": 0.7097427002023706,
+      "grad_norm": 1.0447885323285497,
+      "learning_rate": 3.951280762553152e-06,
+      "loss": 1.4844,
+      "step": 2455
+    },
+    {
+      "epoch": 0.7100318010985834,
+      "grad_norm": 1.0728991878954843,
+      "learning_rate": 3.9439768792242235e-06,
+      "loss": 1.6121,
+      "step": 2456
+    },
+    {
+      "epoch": 0.7103209019947961,
+      "grad_norm": 1.0611080554814958,
+      "learning_rate": 3.936678094128852e-06,
+      "loss": 1.5063,
+      "step": 2457
+    },
+    {
+      "epoch": 0.710610002891009,
+      "grad_norm": 1.0040281932516926,
+      "learning_rate": 3.929384413411489e-06,
+      "loss": 1.4476,
+      "step": 2458
+    },
+    {
+      "epoch": 0.7108991037872218,
+      "grad_norm": 1.0072014227405155,
+      "learning_rate": 3.922095843212279e-06,
+      "loss": 1.5214,
+      "step": 2459
+    },
+    {
+      "epoch": 0.7111882046834345,
+      "grad_norm": 1.013766761773577,
+      "learning_rate": 3.9148123896670775e-06,
+      "loss": 1.5536,
+      "step": 2460
+    },
+    {
+      "epoch": 0.7114773055796473,
+      "grad_norm": 1.0376900792620245,
+      "learning_rate": 3.907534058907414e-06,
+      "loss": 1.5628,
+      "step": 2461
+    },
+    {
+      "epoch": 0.7117664064758601,
+      "grad_norm": 0.9980947843802235,
+      "learning_rate": 3.900260857060526e-06,
+      "loss": 1.4988,
+      "step": 2462
+    },
+    {
+      "epoch": 0.7120555073720729,
+      "grad_norm": 1.0008480966360283,
+      "learning_rate": 3.892992790249318e-06,
+      "loss": 1.571,
+      "step": 2463
+    },
+    {
+      "epoch": 0.7123446082682856,
+      "grad_norm": 1.02879987669725,
+      "learning_rate": 3.885729864592373e-06,
+      "loss": 1.5297,
+      "step": 2464
+    },
+    {
+      "epoch": 0.7126337091644984,
+      "grad_norm": 0.9826481627071425,
+      "learning_rate": 3.878472086203957e-06,
+      "loss": 1.4891,
+      "step": 2465
+    },
+    {
+      "epoch": 0.7129228100607112,
+      "grad_norm": 1.004707560778291,
+      "learning_rate": 3.871219461193992e-06,
+      "loss": 1.5048,
+      "step": 2466
+    },
+    {
+      "epoch": 0.7132119109569239,
+      "grad_norm": 1.8389650068118726,
+      "learning_rate": 3.8639719956680624e-06,
+      "loss": 1.6513,
+      "step": 2467
+    },
+    {
+      "epoch": 0.7135010118531367,
+      "grad_norm": 1.0315093554955308,
+      "learning_rate": 3.856729695727417e-06,
+      "loss": 1.5104,
+      "step": 2468
+    },
+    {
+      "epoch": 0.7137901127493496,
+      "grad_norm": 1.0371727685227374,
+      "learning_rate": 3.849492567468948e-06,
+      "loss": 1.6025,
+      "step": 2469
+    },
+    {
+      "epoch": 0.7140792136455623,
+      "grad_norm": 1.0713861606670556,
+      "learning_rate": 3.8422606169851915e-06,
+      "loss": 1.5189,
+      "step": 2470
+    },
+    {
+      "epoch": 0.7143683145417751,
+      "grad_norm": 1.0344594832831375,
+      "learning_rate": 3.835033850364337e-06,
+      "loss": 1.4872,
+      "step": 2471
+    },
+    {
+      "epoch": 0.7146574154379879,
+      "grad_norm": 0.996876994334722,
+      "learning_rate": 3.827812273690202e-06,
+      "loss": 1.5337,
+      "step": 2472
+    },
+    {
+      "epoch": 0.7149465163342006,
+      "grad_norm": 1.03307381738908,
+      "learning_rate": 3.820595893042236e-06,
+      "loss": 1.5189,
+      "step": 2473
+    },
+    {
+      "epoch": 0.7152356172304134,
+      "grad_norm": 0.9900381568891868,
+      "learning_rate": 3.813384714495507e-06,
+      "loss": 1.5414,
+      "step": 2474
+    },
+    {
+      "epoch": 0.7155247181266262,
+      "grad_norm": 1.004458136818192,
+      "learning_rate": 3.806178744120721e-06,
+      "loss": 1.5031,
+      "step": 2475
+    },
+    {
+      "epoch": 0.715813819022839,
+      "grad_norm": 1.0262988664310408,
+      "learning_rate": 3.7989779879841827e-06,
+      "loss": 1.5316,
+      "step": 2476
+    },
+    {
+      "epoch": 0.7161029199190517,
+      "grad_norm": 0.9973121434776985,
+      "learning_rate": 3.7917824521478108e-06,
+      "loss": 1.5452,
+      "step": 2477
+    },
+    {
+      "epoch": 0.7163920208152645,
+      "grad_norm": 1.0011205707294142,
+      "learning_rate": 3.784592142669139e-06,
+      "loss": 1.5803,
+      "step": 2478
+    },
+    {
+      "epoch": 0.7166811217114774,
+      "grad_norm": 1.0528592538677195,
+      "learning_rate": 3.777407065601292e-06,
+      "loss": 1.5338,
+      "step": 2479
+    },
+    {
+      "epoch": 0.71697022260769,
+      "grad_norm": 0.9878688023075959,
+      "learning_rate": 3.7702272269929875e-06,
+      "loss": 1.5155,
+      "step": 2480
+    },
+    {
+      "epoch": 0.7172593235039029,
+      "grad_norm": 1.0199998419122094,
+      "learning_rate": 3.7630526328885407e-06,
+      "loss": 1.5846,
+      "step": 2481
+    },
+    {
+      "epoch": 0.7175484244001157,
+      "grad_norm": 1.0663555861780545,
+      "learning_rate": 3.755883289327853e-06,
+      "loss": 1.6109,
+      "step": 2482
+    },
+    {
+      "epoch": 0.7178375252963284,
+      "grad_norm": 1.0051325776958036,
+      "learning_rate": 3.748719202346397e-06,
+      "loss": 1.5402,
+      "step": 2483
+    },
+    {
+      "epoch": 0.7181266261925412,
+      "grad_norm": 1.0276296931046631,
+      "learning_rate": 3.7415603779752217e-06,
+      "loss": 1.5976,
+      "step": 2484
+    },
+    {
+      "epoch": 0.718415727088754,
+      "grad_norm": 1.0068347986433557,
+      "learning_rate": 3.7344068222409556e-06,
+      "loss": 1.556,
+      "step": 2485
+    },
+    {
+      "epoch": 0.7187048279849667,
+      "grad_norm": 0.9972497886907018,
+      "learning_rate": 3.7272585411657814e-06,
+      "loss": 1.4689,
+      "step": 2486
+    },
+    {
+      "epoch": 0.7189939288811795,
+      "grad_norm": 1.009067797605547,
+      "learning_rate": 3.7201155407674404e-06,
+      "loss": 1.5533,
+      "step": 2487
+    },
+    {
+      "epoch": 0.7192830297773923,
+      "grad_norm": 1.0215002102381199,
+      "learning_rate": 3.7129778270592398e-06,
+      "loss": 1.4602,
+      "step": 2488
+    },
+    {
+      "epoch": 0.719572130673605,
+      "grad_norm": 1.0151445321031933,
+      "learning_rate": 3.705845406050026e-06,
+      "loss": 1.5534,
+      "step": 2489
+    },
+    {
+      "epoch": 0.7198612315698178,
+      "grad_norm": 1.0518088740769975,
+      "learning_rate": 3.6987182837441894e-06,
+      "loss": 1.4838,
+      "step": 2490
+    },
+    {
+      "epoch": 0.7201503324660307,
+      "grad_norm": 1.0444571177644306,
+      "learning_rate": 3.691596466141666e-06,
+      "loss": 1.6451,
+      "step": 2491
+    },
+    {
+      "epoch": 0.7204394333622435,
+      "grad_norm": 1.0053936343615846,
+      "learning_rate": 3.6844799592379276e-06,
+      "loss": 1.58,
+      "step": 2492
+    },
+    {
+      "epoch": 0.7207285342584562,
+      "grad_norm": 1.0304316527500628,
+      "learning_rate": 3.6773687690239666e-06,
+      "loss": 1.5751,
+      "step": 2493
+    },
+    {
+      "epoch": 0.721017635154669,
+      "grad_norm": 1.0392223422551006,
+      "learning_rate": 3.6702629014863013e-06,
+      "loss": 1.5415,
+      "step": 2494
+    },
+    {
+      "epoch": 0.7213067360508818,
+      "grad_norm": 1.000721731114243,
+      "learning_rate": 3.663162362606978e-06,
+      "loss": 1.6011,
+      "step": 2495
+    },
+    {
+      "epoch": 0.7215958369470945,
+      "grad_norm": 1.0068472839407272,
+      "learning_rate": 3.6560671583635467e-06,
+      "loss": 1.4912,
+      "step": 2496
+    },
+    {
+      "epoch": 0.7218849378433073,
+      "grad_norm": 0.9686648514825775,
+      "learning_rate": 3.648977294729068e-06,
+      "loss": 1.4818,
+      "step": 2497
+    },
+    {
+      "epoch": 0.7221740387395201,
+      "grad_norm": 1.024537399663901,
+      "learning_rate": 3.641892777672117e-06,
+      "loss": 1.5836,
+      "step": 2498
+    },
+    {
+      "epoch": 0.7224631396357328,
+      "grad_norm": 1.0352735039139256,
+      "learning_rate": 3.6348136131567537e-06,
+      "loss": 1.5478,
+      "step": 2499
+    },
+    {
+      "epoch": 0.7227522405319456,
+      "grad_norm": 1.0169598645200595,
+      "learning_rate": 3.6277398071425363e-06,
+      "loss": 1.5094,
+      "step": 2500
+    },
+    {
+      "epoch": 0.7230413414281585,
+      "grad_norm": 1.0351459689844578,
+      "learning_rate": 3.6206713655845215e-06,
+      "loss": 1.5376,
+      "step": 2501
+    },
+    {
+      "epoch": 0.7233304423243712,
+      "grad_norm": 1.0494640005289906,
+      "learning_rate": 3.613608294433235e-06,
+      "loss": 1.5344,
+      "step": 2502
+    },
+    {
+      "epoch": 0.723619543220584,
+      "grad_norm": 1.033938530556738,
+      "learning_rate": 3.606550599634695e-06,
+      "loss": 1.5527,
+      "step": 2503
+    },
+    {
+      "epoch": 0.7239086441167968,
+      "grad_norm": 1.0200415095207411,
+      "learning_rate": 3.5994982871303806e-06,
+      "loss": 1.5289,
+      "step": 2504
+    },
+    {
+      "epoch": 0.7241977450130095,
+      "grad_norm": 1.0730185405143184,
+      "learning_rate": 3.5924513628572544e-06,
+      "loss": 1.4905,
+      "step": 2505
+    },
+    {
+      "epoch": 0.7244868459092223,
+      "grad_norm": 1.0026625331352448,
+      "learning_rate": 3.5854098327477316e-06,
+      "loss": 1.4439,
+      "step": 2506
+    },
+    {
+      "epoch": 0.7247759468054351,
+      "grad_norm": 1.0155627481742289,
+      "learning_rate": 3.578373702729687e-06,
+      "loss": 1.5129,
+      "step": 2507
+    },
+    {
+      "epoch": 0.7250650477016479,
+      "grad_norm": 1.0226774488932953,
+      "learning_rate": 3.5713429787264585e-06,
+      "loss": 1.5387,
+      "step": 2508
+    },
+    {
+      "epoch": 0.7253541485978606,
+      "grad_norm": 1.0747846637918208,
+      "learning_rate": 3.5643176666568235e-06,
+      "loss": 1.6028,
+      "step": 2509
+    },
+    {
+      "epoch": 0.7256432494940734,
+      "grad_norm": 1.0424257328593651,
+      "learning_rate": 3.5572977724350067e-06,
+      "loss": 1.5514,
+      "step": 2510
+    },
+    {
+      "epoch": 0.7259323503902863,
+      "grad_norm": 1.0208982899831545,
+      "learning_rate": 3.5502833019706697e-06,
+      "loss": 1.476,
+      "step": 2511
+    },
+    {
+      "epoch": 0.726221451286499,
+      "grad_norm": 1.0279802964437839,
+      "learning_rate": 3.5432742611689155e-06,
+      "loss": 1.4997,
+      "step": 2512
+    },
+    {
+      "epoch": 0.7265105521827118,
+      "grad_norm": 1.026736267243685,
+      "learning_rate": 3.536270655930264e-06,
+      "loss": 1.4832,
+      "step": 2513
+    },
+    {
+      "epoch": 0.7267996530789246,
+      "grad_norm": 1.0565889677519076,
+      "learning_rate": 3.5292724921506706e-06,
+      "loss": 1.5368,
+      "step": 2514
+    },
+    {
+      "epoch": 0.7270887539751373,
+      "grad_norm": 1.020449696835888,
+      "learning_rate": 3.5222797757215087e-06,
+      "loss": 1.4796,
+      "step": 2515
+    },
+    {
+      "epoch": 0.7273778548713501,
+      "grad_norm": 1.0457699627697372,
+      "learning_rate": 3.515292512529559e-06,
+      "loss": 1.485,
+      "step": 2516
+    },
+    {
+      "epoch": 0.7276669557675629,
+      "grad_norm": 1.0424976185761812,
+      "learning_rate": 3.5083107084570143e-06,
+      "loss": 1.5801,
+      "step": 2517
+    },
+    {
+      "epoch": 0.7279560566637756,
+      "grad_norm": 1.0001102136598656,
+      "learning_rate": 3.5013343693814692e-06,
+      "loss": 1.4164,
+      "step": 2518
+    },
+    {
+      "epoch": 0.7282451575599884,
+      "grad_norm": 1.0779555885639696,
+      "learning_rate": 3.494363501175929e-06,
+      "loss": 1.5407,
+      "step": 2519
+    },
+    {
+      "epoch": 0.7285342584562012,
+      "grad_norm": 1.0432982151049046,
+      "learning_rate": 3.4873981097087807e-06,
+      "loss": 1.495,
+      "step": 2520
+    },
+    {
+      "epoch": 0.728823359352414,
+      "grad_norm": 1.0366651780237193,
+      "learning_rate": 3.480438200843801e-06,
+      "loss": 1.6006,
+      "step": 2521
+    },
+    {
+      "epoch": 0.7291124602486267,
+      "grad_norm": 0.9876779827204685,
+      "learning_rate": 3.4734837804401623e-06,
+      "loss": 1.5698,
+      "step": 2522
+    },
+    {
+      "epoch": 0.7294015611448396,
+      "grad_norm": 1.0169861079651417,
+      "learning_rate": 3.4665348543524036e-06,
+      "loss": 1.5253,
+      "step": 2523
+    },
+    {
+      "epoch": 0.7296906620410524,
+      "grad_norm": 1.0613969962732868,
+      "learning_rate": 3.459591428430451e-06,
+      "loss": 1.5139,
+      "step": 2524
+    },
+    {
+      "epoch": 0.7299797629372651,
+      "grad_norm": 1.0201132003229008,
+      "learning_rate": 3.4526535085195877e-06,
+      "loss": 1.4839,
+      "step": 2525
+    },
+    {
+      "epoch": 0.7302688638334779,
+      "grad_norm": 1.0247716907925635,
+      "learning_rate": 3.4457211004604727e-06,
+      "loss": 1.5447,
+      "step": 2526
+    },
+    {
+      "epoch": 0.7305579647296907,
+      "grad_norm": 1.014969957880706,
+      "learning_rate": 3.438794210089118e-06,
+      "loss": 1.6541,
+      "step": 2527
+    },
+    {
+      "epoch": 0.7308470656259034,
+      "grad_norm": 1.04474095025419,
+      "learning_rate": 3.431872843236889e-06,
+      "loss": 1.5539,
+      "step": 2528
+    },
+    {
+      "epoch": 0.7311361665221162,
+      "grad_norm": 1.0233172962530088,
+      "learning_rate": 3.4249570057305105e-06,
+      "loss": 1.5655,
+      "step": 2529
+    },
+    {
+      "epoch": 0.731425267418329,
+      "grad_norm": 1.0271311312695852,
+      "learning_rate": 3.418046703392044e-06,
+      "loss": 1.5598,
+      "step": 2530
+    },
+    {
+      "epoch": 0.7317143683145417,
+      "grad_norm": 1.0767414513993074,
+      "learning_rate": 3.4111419420388904e-06,
+      "loss": 1.5715,
+      "step": 2531
+    },
+    {
+      "epoch": 0.7320034692107545,
+      "grad_norm": 1.0033214013418974,
+      "learning_rate": 3.404242727483795e-06,
+      "loss": 1.591,
+      "step": 2532
+    },
+    {
+      "epoch": 0.7322925701069674,
+      "grad_norm": 1.0367521779247184,
+      "learning_rate": 3.3973490655348273e-06,
+      "loss": 1.5909,
+      "step": 2533
+    },
+    {
+      "epoch": 0.7325816710031801,
+      "grad_norm": 1.0288704463224445,
+      "learning_rate": 3.3904609619953778e-06,
+      "loss": 1.5208,
+      "step": 2534
+    },
+    {
+      "epoch": 0.7328707718993929,
+      "grad_norm": 1.0037114329662404,
+      "learning_rate": 3.3835784226641653e-06,
+      "loss": 1.5759,
+      "step": 2535
+    },
+    {
+      "epoch": 0.7331598727956057,
+      "grad_norm": 0.9991016286225819,
+      "learning_rate": 3.3767014533352284e-06,
+      "loss": 1.5605,
+      "step": 2536
+    },
+    {
+      "epoch": 0.7334489736918185,
+      "grad_norm": 1.0314286623091882,
+      "learning_rate": 3.369830059797904e-06,
+      "loss": 1.5052,
+      "step": 2537
+    },
+    {
+      "epoch": 0.7337380745880312,
+      "grad_norm": 1.0247283980720958,
+      "learning_rate": 3.3629642478368383e-06,
+      "loss": 1.6186,
+      "step": 2538
+    },
+    {
+      "epoch": 0.734027175484244,
+      "grad_norm": 1.0108751281270694,
+      "learning_rate": 3.3561040232319907e-06,
+      "loss": 1.4812,
+      "step": 2539
+    },
+    {
+      "epoch": 0.7343162763804568,
+      "grad_norm": 0.9982437433471867,
+      "learning_rate": 3.349249391758602e-06,
+      "loss": 1.5332,
+      "step": 2540
+    },
+    {
+      "epoch": 0.7346053772766695,
+      "grad_norm": 1.006223116842084,
+      "learning_rate": 3.342400359187208e-06,
+      "loss": 1.5673,
+      "step": 2541
+    },
+    {
+      "epoch": 0.7348944781728823,
+      "grad_norm": 1.0582532667470306,
+      "learning_rate": 3.3355569312836388e-06,
+      "loss": 1.5841,
+      "step": 2542
+    },
+    {
+      "epoch": 0.7351835790690951,
+      "grad_norm": 1.0293862397160103,
+      "learning_rate": 3.328719113809e-06,
+      "loss": 1.6917,
+      "step": 2543
+    },
+    {
+      "epoch": 0.7354726799653079,
+      "grad_norm": 1.0388579357994043,
+      "learning_rate": 3.3218869125196694e-06,
+      "loss": 1.4731,
+      "step": 2544
+    },
+    {
+      "epoch": 0.7357617808615207,
+      "grad_norm": 1.00608344525251,
+      "learning_rate": 3.315060333167307e-06,
+      "loss": 1.5497,
+      "step": 2545
+    },
+    {
+      "epoch": 0.7360508817577335,
+      "grad_norm": 1.021619930745868,
+      "learning_rate": 3.308239381498838e-06,
+      "loss": 1.5367,
+      "step": 2546
+    },
+    {
+      "epoch": 0.7363399826539462,
+      "grad_norm": 1.0207376727961825,
+      "learning_rate": 3.3014240632564443e-06,
+      "loss": 1.5074,
+      "step": 2547
+    },
+    {
+      "epoch": 0.736629083550159,
+      "grad_norm": 1.0168306398579021,
+      "learning_rate": 3.2946143841775646e-06,
+      "loss": 1.541,
+      "step": 2548
+    },
+    {
+      "epoch": 0.7369181844463718,
+      "grad_norm": 1.0336734374018308,
+      "learning_rate": 3.287810349994901e-06,
+      "loss": 1.5916,
+      "step": 2549
+    },
+    {
+      "epoch": 0.7372072853425846,
+      "grad_norm": 1.077764485344871,
+      "learning_rate": 3.281011966436394e-06,
+      "loss": 1.5263,
+      "step": 2550
+    },
+    {
+      "epoch": 0.7374963862387973,
+      "grad_norm": 1.0080424040437506,
+      "learning_rate": 3.2742192392252247e-06,
+      "loss": 1.6345,
+      "step": 2551
+    },
+    {
+      "epoch": 0.7377854871350101,
+      "grad_norm": 0.9953584626415105,
+      "learning_rate": 3.2674321740798255e-06,
+      "loss": 1.5432,
+      "step": 2552
+    },
+    {
+      "epoch": 0.738074588031223,
+      "grad_norm": 1.095697853670159,
+      "learning_rate": 3.2606507767138494e-06,
+      "loss": 1.6241,
+      "step": 2553
+    },
+    {
+      "epoch": 0.7383636889274356,
+      "grad_norm": 1.0308303820036064,
+      "learning_rate": 3.2538750528361816e-06,
+      "loss": 1.5393,
+      "step": 2554
+    },
+    {
+      "epoch": 0.7386527898236485,
+      "grad_norm": 0.9848164761009687,
+      "learning_rate": 3.247105008150934e-06,
+      "loss": 1.4844,
+      "step": 2555
+    },
+    {
+      "epoch": 0.7389418907198613,
+      "grad_norm": 1.0679668005057916,
+      "learning_rate": 3.2403406483574384e-06,
+      "loss": 1.5563,
+      "step": 2556
+    },
+    {
+      "epoch": 0.739230991616074,
+      "grad_norm": 1.04767019364261,
+      "learning_rate": 3.2335819791502355e-06,
+      "loss": 1.5754,
+      "step": 2557
+    },
+    {
+      "epoch": 0.7395200925122868,
+      "grad_norm": 1.0321798795478945,
+      "learning_rate": 3.2268290062190755e-06,
+      "loss": 1.5799,
+      "step": 2558
+    },
+    {
+      "epoch": 0.7398091934084996,
+      "grad_norm": 1.0211663941763789,
+      "learning_rate": 3.220081735248922e-06,
+      "loss": 1.5741,
+      "step": 2559
+    },
+    {
+      "epoch": 0.7400982943047123,
+      "grad_norm": 1.0191122170864415,
+      "learning_rate": 3.2133401719199285e-06,
+      "loss": 1.5203,
+      "step": 2560
+    },
+    {
+      "epoch": 0.7403873952009251,
+      "grad_norm": 1.0545046651852399,
+      "learning_rate": 3.206604321907444e-06,
+      "loss": 1.545,
+      "step": 2561
+    },
+    {
+      "epoch": 0.7406764960971379,
+      "grad_norm": 1.0124757830593905,
+      "learning_rate": 3.1998741908820173e-06,
+      "loss": 1.5693,
+      "step": 2562
+    },
+    {
+      "epoch": 0.7409655969933506,
+      "grad_norm": 1.0638281745437896,
+      "learning_rate": 3.1931497845093753e-06,
+      "loss": 1.5345,
+      "step": 2563
+    },
+    {
+      "epoch": 0.7412546978895634,
+      "grad_norm": 1.0267298711572816,
+      "learning_rate": 3.1864311084504194e-06,
+      "loss": 1.5283,
+      "step": 2564
+    },
+    {
+      "epoch": 0.7415437987857763,
+      "grad_norm": 1.0176123819474867,
+      "learning_rate": 3.1797181683612444e-06,
+      "loss": 1.6881,
+      "step": 2565
+    },
+    {
+      "epoch": 0.7418328996819891,
+      "grad_norm": 1.0830415082615363,
+      "learning_rate": 3.173010969893099e-06,
+      "loss": 1.4895,
+      "step": 2566
+    },
+    {
+      "epoch": 0.7421220005782018,
+      "grad_norm": 1.007831085722778,
+      "learning_rate": 3.166309518692411e-06,
+      "loss": 1.6018,
+      "step": 2567
+    },
+    {
+      "epoch": 0.7424111014744146,
+      "grad_norm": 1.059967467087403,
+      "learning_rate": 3.1596138204007588e-06,
+      "loss": 1.5672,
+      "step": 2568
+    },
+    {
+      "epoch": 0.7427002023706274,
+      "grad_norm": 1.0468068017941161,
+      "learning_rate": 3.152923880654889e-06,
+      "loss": 1.6229,
+      "step": 2569
+    },
+    {
+      "epoch": 0.7429893032668401,
+      "grad_norm": 1.0391781844425465,
+      "learning_rate": 3.146239705086692e-06,
+      "loss": 1.5437,
+      "step": 2570
+    },
+    {
+      "epoch": 0.7432784041630529,
+      "grad_norm": 0.9994576541571422,
+      "learning_rate": 3.139561299323206e-06,
+      "loss": 1.5123,
+      "step": 2571
+    },
+    {
+      "epoch": 0.7435675050592657,
+      "grad_norm": 1.0730086820884217,
+      "learning_rate": 3.132888668986619e-06,
+      "loss": 1.6265,
+      "step": 2572
+    },
+    {
+      "epoch": 0.7438566059554784,
+      "grad_norm": 1.0927337601062659,
+      "learning_rate": 3.126221819694252e-06,
+      "loss": 1.5959,
+      "step": 2573
+    },
+    {
+      "epoch": 0.7441457068516912,
+      "grad_norm": 1.0221825064039038,
+      "learning_rate": 3.1195607570585562e-06,
+      "loss": 1.5618,
+      "step": 2574
+    },
+    {
+      "epoch": 0.744434807747904,
+      "grad_norm": 1.0460709268343236,
+      "learning_rate": 3.112905486687113e-06,
+      "loss": 1.5593,
+      "step": 2575
+    },
+    {
+      "epoch": 0.7447239086441167,
+      "grad_norm": 1.0149424746986917,
+      "learning_rate": 3.106256014182637e-06,
+      "loss": 1.5754,
+      "step": 2576
+    },
+    {
+      "epoch": 0.7450130095403296,
+      "grad_norm": 1.0097165006814974,
+      "learning_rate": 3.0996123451429463e-06,
+      "loss": 1.569,
+      "step": 2577
+    },
+    {
+      "epoch": 0.7453021104365424,
+      "grad_norm": 1.0287236043208767,
+      "learning_rate": 3.0929744851609853e-06,
+      "loss": 1.6208,
+      "step": 2578
+    },
+    {
+      "epoch": 0.7455912113327551,
+      "grad_norm": 1.0146999241856058,
+      "learning_rate": 3.0863424398248067e-06,
+      "loss": 1.5653,
+      "step": 2579
+    },
+    {
+      "epoch": 0.7458803122289679,
+      "grad_norm": 1.0103848506561606,
+      "learning_rate": 3.079716214717562e-06,
+      "loss": 1.511,
+      "step": 2580
+    },
+    {
+      "epoch": 0.7461694131251807,
+      "grad_norm": 0.994356312610372,
+      "learning_rate": 3.0730958154175072e-06,
+      "loss": 1.4918,
+      "step": 2581
+    },
+    {
+      "epoch": 0.7464585140213935,
+      "grad_norm": 1.0269505420547047,
+      "learning_rate": 3.066481247497991e-06,
+      "loss": 1.541,
+      "step": 2582
+    },
+    {
+      "epoch": 0.7467476149176062,
+      "grad_norm": 1.0429684865684299,
+      "learning_rate": 3.05987251652746e-06,
+      "loss": 1.6264,
+      "step": 2583
+    },
+    {
+      "epoch": 0.747036715813819,
+      "grad_norm": 1.0183976322023274,
+      "learning_rate": 3.0532696280694385e-06,
+      "loss": 1.5561,
+      "step": 2584
+    },
+    {
+      "epoch": 0.7473258167100318,
+      "grad_norm": 0.9888969814555285,
+      "learning_rate": 3.0466725876825342e-06,
+      "loss": 1.6166,
+      "step": 2585
+    },
+    {
+      "epoch": 0.7476149176062445,
+      "grad_norm": 1.0861070351112132,
+      "learning_rate": 3.0400814009204404e-06,
+      "loss": 1.5974,
+      "step": 2586
+    },
+    {
+      "epoch": 0.7479040185024574,
+      "grad_norm": 1.00956236328955,
+      "learning_rate": 3.0334960733319076e-06,
+      "loss": 1.5121,
+      "step": 2587
+    },
+    {
+      "epoch": 0.7481931193986702,
+      "grad_norm": 1.0104832403331987,
+      "learning_rate": 3.0269166104607674e-06,
+      "loss": 1.4693,
+      "step": 2588
+    },
+    {
+      "epoch": 0.7484822202948829,
+      "grad_norm": 1.0587319123776202,
+      "learning_rate": 3.020343017845905e-06,
+      "loss": 1.469,
+      "step": 2589
+    },
+    {
+      "epoch": 0.7487713211910957,
+      "grad_norm": 1.0392006519671793,
+      "learning_rate": 3.013775301021271e-06,
+      "loss": 1.6003,
+      "step": 2590
+    },
+    {
+      "epoch": 0.7490604220873085,
+      "grad_norm": 1.0288868505872355,
+      "learning_rate": 3.007213465515865e-06,
+      "loss": 1.4686,
+      "step": 2591
+    },
+    {
+      "epoch": 0.7493495229835212,
+      "grad_norm": 1.0136337812775904,
+      "learning_rate": 3.00065751685373e-06,
+      "loss": 1.5557,
+      "step": 2592
+    },
+    {
+      "epoch": 0.749638623879734,
+      "grad_norm": 0.9899670706961841,
+      "learning_rate": 2.9941074605539687e-06,
+      "loss": 1.5311,
+      "step": 2593
+    },
+    {
+      "epoch": 0.7499277247759468,
+      "grad_norm": 1.0474648033847238,
+      "learning_rate": 2.9875633021307104e-06,
+      "loss": 1.5114,
+      "step": 2594
+    },
+    {
+      "epoch": 0.7502168256721596,
+      "grad_norm": 1.0103295201616964,
+      "learning_rate": 2.981025047093118e-06,
+      "loss": 1.5653,
+      "step": 2595
+    },
+    {
+      "epoch": 0.7505059265683723,
+      "grad_norm": 1.0100337912783712,
+      "learning_rate": 2.9744927009453983e-06,
+      "loss": 1.6234,
+      "step": 2596
+    },
+    {
+      "epoch": 0.7507950274645852,
+      "grad_norm": 1.0555628085614217,
+      "learning_rate": 2.9679662691867705e-06,
+      "loss": 1.6073,
+      "step": 2597
+    },
+    {
+      "epoch": 0.751084128360798,
+      "grad_norm": 1.005756871384676,
+      "learning_rate": 2.9614457573114765e-06,
+      "loss": 1.4883,
+      "step": 2598
+    },
+    {
+      "epoch": 0.7513732292570107,
+      "grad_norm": 1.0257195563131012,
+      "learning_rate": 2.95493117080878e-06,
+      "loss": 1.5429,
+      "step": 2599
+    },
+    {
+      "epoch": 0.7516623301532235,
+      "grad_norm": 1.0477649876438562,
+      "learning_rate": 2.9484225151629586e-06,
+      "loss": 1.4643,
+      "step": 2600
+    },
+    {
+      "epoch": 0.7519514310494363,
+      "grad_norm": 1.050027335641844,
+      "learning_rate": 2.9419197958532888e-06,
+      "loss": 1.4616,
+      "step": 2601
+    },
+    {
+      "epoch": 0.752240531945649,
+      "grad_norm": 1.013432505899696,
+      "learning_rate": 2.9354230183540476e-06,
+      "loss": 1.5654,
+      "step": 2602
+    },
+    {
+      "epoch": 0.7525296328418618,
+      "grad_norm": 1.037768968045221,
+      "learning_rate": 2.9289321881345257e-06,
+      "loss": 1.5227,
+      "step": 2603
+    },
+    {
+      "epoch": 0.7528187337380746,
+      "grad_norm": 0.9931225217414482,
+      "learning_rate": 2.9224473106589913e-06,
+      "loss": 1.4861,
+      "step": 2604
+    },
+    {
+      "epoch": 0.7531078346342873,
+      "grad_norm": 1.0022781678924897,
+      "learning_rate": 2.915968391386703e-06,
+      "loss": 1.4512,
+      "step": 2605
+    },
+    {
+      "epoch": 0.7533969355305001,
+      "grad_norm": 1.0740225770351521,
+      "learning_rate": 2.9094954357719163e-06,
+      "loss": 1.5652,
+      "step": 2606
+    },
+    {
+      "epoch": 0.753686036426713,
+      "grad_norm": 0.9938008268972566,
+      "learning_rate": 2.9030284492638526e-06,
+      "loss": 1.5106,
+      "step": 2607
+    },
+    {
+      "epoch": 0.7539751373229256,
+      "grad_norm": 1.0158342107235576,
+      "learning_rate": 2.89656743730671e-06,
+      "loss": 1.5079,
+      "step": 2608
+    },
+    {
+      "epoch": 0.7542642382191385,
+      "grad_norm": 1.0191977984899148,
+      "learning_rate": 2.8901124053396633e-06,
+      "loss": 1.498,
+      "step": 2609
+    },
+    {
+      "epoch": 0.7545533391153513,
+      "grad_norm": 1.0042644824460043,
+      "learning_rate": 2.8836633587968533e-06,
+      "loss": 1.4744,
+      "step": 2610
+    },
+    {
+      "epoch": 0.7548424400115641,
+      "grad_norm": 1.0174514532647274,
+      "learning_rate": 2.877220303107373e-06,
+      "loss": 1.487,
+      "step": 2611
+    },
+    {
+      "epoch": 0.7551315409077768,
+      "grad_norm": 0.9891233042599077,
+      "learning_rate": 2.8707832436952777e-06,
+      "loss": 1.5266,
+      "step": 2612
+    },
+    {
+      "epoch": 0.7554206418039896,
+      "grad_norm": 1.0546334222470721,
+      "learning_rate": 2.8643521859795785e-06,
+      "loss": 1.4392,
+      "step": 2613
+    },
+    {
+      "epoch": 0.7557097427002024,
+      "grad_norm": 0.9993952458177704,
+      "learning_rate": 2.8579271353742277e-06,
+      "loss": 1.4897,
+      "step": 2614
+    },
+    {
+      "epoch": 0.7559988435964151,
+      "grad_norm": 1.0529482735096456,
+      "learning_rate": 2.8515080972881193e-06,
+      "loss": 1.5761,
+      "step": 2615
+    },
+    {
+      "epoch": 0.7562879444926279,
+      "grad_norm": 1.0130904537621124,
+      "learning_rate": 2.8450950771250952e-06,
+      "loss": 1.5123,
+      "step": 2616
+    },
+    {
+      "epoch": 0.7565770453888407,
+      "grad_norm": 1.0449293906491497,
+      "learning_rate": 2.8386880802839233e-06,
+      "loss": 1.5278,
+      "step": 2617
+    },
+    {
+      "epoch": 0.7568661462850534,
+      "grad_norm": 1.040110495955749,
+      "learning_rate": 2.8322871121582984e-06,
+      "loss": 1.5059,
+      "step": 2618
+    },
+    {
+      "epoch": 0.7571552471812663,
+      "grad_norm": 0.99137863722194,
+      "learning_rate": 2.825892178136851e-06,
+      "loss": 1.473,
+      "step": 2619
+    },
+    {
+      "epoch": 0.7574443480774791,
+      "grad_norm": 0.9964543317074376,
+      "learning_rate": 2.819503283603119e-06,
+      "loss": 1.5422,
+      "step": 2620
+    },
+    {
+      "epoch": 0.7577334489736918,
+      "grad_norm": 1.022467340372004,
+      "learning_rate": 2.813120433935568e-06,
+      "loss": 1.4921,
+      "step": 2621
+    },
+    {
+      "epoch": 0.7580225498699046,
+      "grad_norm": 1.0515851844439317,
+      "learning_rate": 2.8067436345075627e-06,
+      "loss": 1.514,
+      "step": 2622
+    },
+    {
+      "epoch": 0.7583116507661174,
+      "grad_norm": 1.025927927648296,
+      "learning_rate": 2.8003728906873884e-06,
+      "loss": 1.5791,
+      "step": 2623
+    },
+    {
+      "epoch": 0.7586007516623302,
+      "grad_norm": 1.0674707556695548,
+      "learning_rate": 2.7940082078382192e-06,
+      "loss": 1.6048,
+      "step": 2624
+    },
+    {
+      "epoch": 0.7588898525585429,
+      "grad_norm": 1.0608994681428927,
+      "learning_rate": 2.78764959131813e-06,
+      "loss": 1.4335,
+      "step": 2625
+    },
+    {
+      "epoch": 0.7591789534547557,
+      "grad_norm": 1.1861688170849973,
+      "learning_rate": 2.7812970464800983e-06,
+      "loss": 1.5875,
+      "step": 2626
+    },
+    {
+      "epoch": 0.7594680543509685,
+      "grad_norm": 1.012194530925656,
+      "learning_rate": 2.77495057867198e-06,
+      "loss": 1.4933,
+      "step": 2627
+    },
+    {
+      "epoch": 0.7597571552471812,
+      "grad_norm": 1.020384410894713,
+      "learning_rate": 2.7686101932365162e-06,
+      "loss": 1.5482,
+      "step": 2628
+    },
+    {
+      "epoch": 0.760046256143394,
+      "grad_norm": 1.0422371086855238,
+      "learning_rate": 2.762275895511328e-06,
+      "loss": 1.5735,
+      "step": 2629
+    },
+    {
+      "epoch": 0.7603353570396069,
+      "grad_norm": 1.0547225067684234,
+      "learning_rate": 2.7559476908289163e-06,
+      "loss": 1.6217,
+      "step": 2630
+    },
+    {
+      "epoch": 0.7606244579358196,
+      "grad_norm": 1.042291959541293,
+      "learning_rate": 2.7496255845166507e-06,
+      "loss": 1.5922,
+      "step": 2631
+    },
+    {
+      "epoch": 0.7609135588320324,
+      "grad_norm": 1.0560415146384452,
+      "learning_rate": 2.7433095818967627e-06,
+      "loss": 1.57,
+      "step": 2632
+    },
+    {
+      "epoch": 0.7612026597282452,
+      "grad_norm": 1.0137326992591507,
+      "learning_rate": 2.7369996882863513e-06,
+      "loss": 1.5151,
+      "step": 2633
+    },
+    {
+      "epoch": 0.7614917606244579,
+      "grad_norm": 1.0382506402844722,
+      "learning_rate": 2.7306959089973673e-06,
+      "loss": 1.6437,
+      "step": 2634
+    },
+    {
+      "epoch": 0.7617808615206707,
+      "grad_norm": 1.0238316650505968,
+      "learning_rate": 2.7243982493366185e-06,
+      "loss": 1.5832,
+      "step": 2635
+    },
+    {
+      "epoch": 0.7620699624168835,
+      "grad_norm": 1.0383328379120784,
+      "learning_rate": 2.7181067146057525e-06,
+      "loss": 1.6272,
+      "step": 2636
+    },
+    {
+      "epoch": 0.7623590633130962,
+      "grad_norm": 1.0460459547748595,
+      "learning_rate": 2.711821310101277e-06,
+      "loss": 1.615,
+      "step": 2637
+    },
+    {
+      "epoch": 0.762648164209309,
+      "grad_norm": 1.04030728747329,
+      "learning_rate": 2.705542041114524e-06,
+      "loss": 1.5009,
+      "step": 2638
+    },
+    {
+      "epoch": 0.7629372651055218,
+      "grad_norm": 1.047474919968278,
+      "learning_rate": 2.699268912931663e-06,
+      "loss": 1.4929,
+      "step": 2639
+    },
+    {
+      "epoch": 0.7632263660017347,
+      "grad_norm": 0.9892678300248734,
+      "learning_rate": 2.693001930833702e-06,
+      "loss": 1.5212,
+      "step": 2640
+    },
+    {
+      "epoch": 0.7635154668979474,
+      "grad_norm": 0.9930374148007767,
+      "learning_rate": 2.686741100096464e-06,
+      "loss": 1.4351,
+      "step": 2641
+    },
+    {
+      "epoch": 0.7638045677941602,
+      "grad_norm": 1.0117831988431913,
+      "learning_rate": 2.680486425990604e-06,
+      "loss": 1.5082,
+      "step": 2642
+    },
+    {
+      "epoch": 0.764093668690373,
+      "grad_norm": 1.056057919079606,
+      "learning_rate": 2.674237913781583e-06,
+      "loss": 1.4432,
+      "step": 2643
+    },
+    {
+      "epoch": 0.7643827695865857,
+      "grad_norm": 1.0390779634042422,
+      "learning_rate": 2.667995568729689e-06,
+      "loss": 1.5184,
+      "step": 2644
+    },
+    {
+      "epoch": 0.7646718704827985,
+      "grad_norm": 1.0307588188445542,
+      "learning_rate": 2.6617593960900034e-06,
+      "loss": 1.4731,
+      "step": 2645
+    },
+    {
+      "epoch": 0.7649609713790113,
+      "grad_norm": 1.036288434905917,
+      "learning_rate": 2.6555294011124157e-06,
+      "loss": 1.5053,
+      "step": 2646
+    },
+    {
+      "epoch": 0.765250072275224,
+      "grad_norm": 1.047622378504518,
+      "learning_rate": 2.649305589041624e-06,
+      "loss": 1.6148,
+      "step": 2647
+    },
+    {
+      "epoch": 0.7655391731714368,
+      "grad_norm": 1.1648622507522859,
+      "learning_rate": 2.6430879651171104e-06,
+      "loss": 1.6241,
+      "step": 2648
+    },
+    {
+      "epoch": 0.7658282740676496,
+      "grad_norm": 1.0207626598152555,
+      "learning_rate": 2.6368765345731486e-06,
+      "loss": 1.41,
+      "step": 2649
+    },
+    {
+      "epoch": 0.7661173749638623,
+      "grad_norm": 1.0421403147335728,
+      "learning_rate": 2.6306713026388064e-06,
+      "loss": 1.439,
+      "step": 2650
+    },
+    {
+      "epoch": 0.7664064758600752,
+      "grad_norm": 1.037974003069992,
+      "learning_rate": 2.624472274537925e-06,
+      "loss": 1.6079,
+      "step": 2651
+    },
+    {
+      "epoch": 0.766695576756288,
+      "grad_norm": 1.054758737513161,
+      "learning_rate": 2.6182794554891244e-06,
+      "loss": 1.5178,
+      "step": 2652
+    },
+    {
+      "epoch": 0.7669846776525007,
+      "grad_norm": 1.0336454424264914,
+      "learning_rate": 2.612092850705801e-06,
+      "loss": 1.5773,
+      "step": 2653
+    },
+    {
+      "epoch": 0.7672737785487135,
+      "grad_norm": 1.0336594335577038,
+      "learning_rate": 2.605912465396121e-06,
+      "loss": 1.5423,
+      "step": 2654
+    },
+    {
+      "epoch": 0.7675628794449263,
+      "grad_norm": 1.1046963081188579,
+      "learning_rate": 2.599738304763009e-06,
+      "loss": 1.4679,
+      "step": 2655
+    },
+    {
+      "epoch": 0.7678519803411391,
+      "grad_norm": 1.052301961579788,
+      "learning_rate": 2.5935703740041495e-06,
+      "loss": 1.6023,
+      "step": 2656
+    },
+    {
+      "epoch": 0.7681410812373518,
+      "grad_norm": 1.0091116362720542,
+      "learning_rate": 2.58740867831199e-06,
+      "loss": 1.5645,
+      "step": 2657
+    },
+    {
+      "epoch": 0.7684301821335646,
+      "grad_norm": 1.034178453678895,
+      "learning_rate": 2.581253222873721e-06,
+      "loss": 1.6006,
+      "step": 2658
+    },
+    {
+      "epoch": 0.7687192830297774,
+      "grad_norm": 1.070596406334517,
+      "learning_rate": 2.57510401287128e-06,
+      "loss": 1.558,
+      "step": 2659
+    },
+    {
+      "epoch": 0.7690083839259901,
+      "grad_norm": 1.0392672383344062,
+      "learning_rate": 2.5689610534813556e-06,
+      "loss": 1.5559,
+      "step": 2660
+    },
+    {
+      "epoch": 0.769297484822203,
+      "grad_norm": 0.993264130785009,
+      "learning_rate": 2.5628243498753657e-06,
+      "loss": 1.5859,
+      "step": 2661
+    },
+    {
+      "epoch": 0.7695865857184158,
+      "grad_norm": 1.0360872808641102,
+      "learning_rate": 2.5566939072194608e-06,
+      "loss": 1.5185,
+      "step": 2662
+    },
+    {
+      "epoch": 0.7698756866146285,
+      "grad_norm": 1.0564605880065785,
+      "learning_rate": 2.5505697306745257e-06,
+      "loss": 1.563,
+      "step": 2663
+    },
+    {
+      "epoch": 0.7701647875108413,
+      "grad_norm": 0.9952794578382141,
+      "learning_rate": 2.5444518253961747e-06,
+      "loss": 1.5473,
+      "step": 2664
+    },
+    {
+      "epoch": 0.7704538884070541,
+      "grad_norm": 1.056744557782212,
+      "learning_rate": 2.538340196534731e-06,
+      "loss": 1.5099,
+      "step": 2665
+    },
+    {
+      "epoch": 0.7707429893032668,
+      "grad_norm": 1.0694078921207326,
+      "learning_rate": 2.5322348492352377e-06,
+      "loss": 1.5975,
+      "step": 2666
+    },
+    {
+      "epoch": 0.7710320901994796,
+      "grad_norm": 1.0271996894307336,
+      "learning_rate": 2.5261357886374584e-06,
+      "loss": 1.4618,
+      "step": 2667
+    },
+    {
+      "epoch": 0.7713211910956924,
+      "grad_norm": 1.020648818515057,
+      "learning_rate": 2.520043019875852e-06,
+      "loss": 1.5489,
+      "step": 2668
+    },
+    {
+      "epoch": 0.7716102919919052,
+      "grad_norm": 1.0579955221977166,
+      "learning_rate": 2.513956548079586e-06,
+      "loss": 1.5034,
+      "step": 2669
+    },
+    {
+      "epoch": 0.7718993928881179,
+      "grad_norm": 1.015659487407231,
+      "learning_rate": 2.5078763783725313e-06,
+      "loss": 1.4978,
+      "step": 2670
+    },
+    {
+      "epoch": 0.7721884937843307,
+      "grad_norm": 1.0235083807280219,
+      "learning_rate": 2.501802515873247e-06,
+      "loss": 1.5562,
+      "step": 2671
+    },
+    {
+      "epoch": 0.7724775946805436,
+      "grad_norm": 0.9970511549829404,
+      "learning_rate": 2.4957349656949814e-06,
+      "loss": 1.5173,
+      "step": 2672
+    },
+    {
+      "epoch": 0.7727666955767563,
+      "grad_norm": 1.0054069943853308,
+      "learning_rate": 2.4896737329456742e-06,
+      "loss": 1.3873,
+      "step": 2673
+    },
+    {
+      "epoch": 0.7730557964729691,
+      "grad_norm": 1.0474315853136484,
+      "learning_rate": 2.48361882272795e-06,
+      "loss": 1.5269,
+      "step": 2674
+    },
+    {
+      "epoch": 0.7733448973691819,
+      "grad_norm": 1.0281510473451847,
+      "learning_rate": 2.4775702401391e-06,
+      "loss": 1.5462,
+      "step": 2675
+    },
+    {
+      "epoch": 0.7736339982653946,
+      "grad_norm": 1.024240901086314,
+      "learning_rate": 2.471527990271092e-06,
+      "loss": 1.5289,
+      "step": 2676
+    },
+    {
+      "epoch": 0.7739230991616074,
+      "grad_norm": 1.0284418307040137,
+      "learning_rate": 2.4654920782105696e-06,
+      "loss": 1.5139,
+      "step": 2677
+    },
+    {
+      "epoch": 0.7742122000578202,
+      "grad_norm": 1.009907393770174,
+      "learning_rate": 2.459462509038836e-06,
+      "loss": 1.5075,
+      "step": 2678
+    },
+    {
+      "epoch": 0.7745013009540329,
+      "grad_norm": 1.067358258451308,
+      "learning_rate": 2.453439287831848e-06,
+      "loss": 1.5945,
+      "step": 2679
+    },
+    {
+      "epoch": 0.7747904018502457,
+      "grad_norm": 1.0122367829821395,
+      "learning_rate": 2.4474224196602327e-06,
+      "loss": 1.5136,
+      "step": 2680
+    },
+    {
+      "epoch": 0.7750795027464585,
+      "grad_norm": 1.0110351702625238,
+      "learning_rate": 2.4414119095892575e-06,
+      "loss": 1.5839,
+      "step": 2681
+    },
+    {
+      "epoch": 0.7753686036426712,
+      "grad_norm": 1.0071850894827417,
+      "learning_rate": 2.4354077626788397e-06,
+      "loss": 1.4615,
+      "step": 2682
+    },
+    {
+      "epoch": 0.775657704538884,
+      "grad_norm": 1.0439388834267012,
+      "learning_rate": 2.429409983983546e-06,
+      "loss": 1.5832,
+      "step": 2683
+    },
+    {
+      "epoch": 0.7759468054350969,
+      "grad_norm": 1.0074235671628706,
+      "learning_rate": 2.423418578552571e-06,
+      "loss": 1.512,
+      "step": 2684
+    },
+    {
+      "epoch": 0.7762359063313097,
+      "grad_norm": 1.0275466730873077,
+      "learning_rate": 2.4174335514297563e-06,
+      "loss": 1.5235,
+      "step": 2685
+    },
+    {
+      "epoch": 0.7765250072275224,
+      "grad_norm": 1.045629835800285,
+      "learning_rate": 2.411454907653562e-06,
+      "loss": 1.5636,
+      "step": 2686
+    },
+    {
+      "epoch": 0.7768141081237352,
+      "grad_norm": 0.9919605335151682,
+      "learning_rate": 2.405482652257086e-06,
+      "loss": 1.5052,
+      "step": 2687
+    },
+    {
+      "epoch": 0.777103209019948,
+      "grad_norm": 1.0202920183765987,
+      "learning_rate": 2.3995167902680404e-06,
+      "loss": 1.6233,
+      "step": 2688
+    },
+    {
+      "epoch": 0.7773923099161607,
+      "grad_norm": 1.0389970416465488,
+      "learning_rate": 2.393557326708752e-06,
+      "loss": 1.5091,
+      "step": 2689
+    },
+    {
+      "epoch": 0.7776814108123735,
+      "grad_norm": 1.2516050460023302,
+      "learning_rate": 2.3876042665961742e-06,
+      "loss": 1.5516,
+      "step": 2690
+    },
+    {
+      "epoch": 0.7779705117085863,
+      "grad_norm": 1.023721769923095,
+      "learning_rate": 2.381657614941858e-06,
+      "loss": 1.5869,
+      "step": 2691
+    },
+    {
+      "epoch": 0.778259612604799,
+      "grad_norm": 1.0287184127874531,
+      "learning_rate": 2.3757173767519627e-06,
+      "loss": 1.6773,
+      "step": 2692
+    },
+    {
+      "epoch": 0.7785487135010118,
+      "grad_norm": 1.0213410243191499,
+      "learning_rate": 2.3697835570272454e-06,
+      "loss": 1.576,
+      "step": 2693
+    },
+    {
+      "epoch": 0.7788378143972247,
+      "grad_norm": 1.0295776090887365,
+      "learning_rate": 2.3638561607630684e-06,
+      "loss": 1.5191,
+      "step": 2694
+    },
+    {
+      "epoch": 0.7791269152934374,
+      "grad_norm": 1.0172548941498183,
+      "learning_rate": 2.357935192949382e-06,
+      "loss": 1.5398,
+      "step": 2695
+    },
+    {
+      "epoch": 0.7794160161896502,
+      "grad_norm": 1.0271233173227783,
+      "learning_rate": 2.3520206585707196e-06,
+      "loss": 1.589,
+      "step": 2696
+    },
+    {
+      "epoch": 0.779705117085863,
+      "grad_norm": 1.0157747885132966,
+      "learning_rate": 2.3461125626062087e-06,
+      "loss": 1.5409,
+      "step": 2697
+    },
+    {
+      "epoch": 0.7799942179820758,
+      "grad_norm": 1.009682984628032,
+      "learning_rate": 2.340210910029548e-06,
+      "loss": 1.5139,
+      "step": 2698
+    },
+    {
+      "epoch": 0.7802833188782885,
+      "grad_norm": 0.9979983786246405,
+      "learning_rate": 2.334315705809015e-06,
+      "loss": 1.5734,
+      "step": 2699
+    },
+    {
+      "epoch": 0.7805724197745013,
+      "grad_norm": 1.0140716952490711,
+      "learning_rate": 2.328426954907456e-06,
+      "loss": 1.5299,
+      "step": 2700
+    },
+    {
+      "epoch": 0.7808615206707141,
+      "grad_norm": 1.0051066994795377,
+      "learning_rate": 2.322544662282292e-06,
+      "loss": 1.3366,
+      "step": 2701
+    },
+    {
+      "epoch": 0.7811506215669268,
+      "grad_norm": 1.0068974511936952,
+      "learning_rate": 2.316668832885501e-06,
+      "loss": 1.4143,
+      "step": 2702
+    },
+    {
+      "epoch": 0.7814397224631396,
+      "grad_norm": 1.002168577332565,
+      "learning_rate": 2.3107994716636184e-06,
+      "loss": 1.5001,
+      "step": 2703
+    },
+    {
+      "epoch": 0.7817288233593525,
+      "grad_norm": 1.0087888240443759,
+      "learning_rate": 2.304936583557742e-06,
+      "loss": 1.5567,
+      "step": 2704
+    },
+    {
+      "epoch": 0.7820179242555652,
+      "grad_norm": 1.0217924315200615,
+      "learning_rate": 2.299080173503511e-06,
+      "loss": 1.5603,
+      "step": 2705
+    },
+    {
+      "epoch": 0.782307025151778,
+      "grad_norm": 1.0261137488808436,
+      "learning_rate": 2.2932302464311196e-06,
+      "loss": 1.4599,
+      "step": 2706
+    },
+    {
+      "epoch": 0.7825961260479908,
+      "grad_norm": 1.0880824620147693,
+      "learning_rate": 2.287386807265297e-06,
+      "loss": 1.4913,
+      "step": 2707
+    },
+    {
+      "epoch": 0.7828852269442035,
+      "grad_norm": 1.0323116968162067,
+      "learning_rate": 2.281549860925317e-06,
+      "loss": 1.5627,
+      "step": 2708
+    },
+    {
+      "epoch": 0.7831743278404163,
+      "grad_norm": 0.9798312083859206,
+      "learning_rate": 2.2757194123249847e-06,
+      "loss": 1.5243,
+      "step": 2709
+    },
+    {
+      "epoch": 0.7834634287366291,
+      "grad_norm": 1.045137968128974,
+      "learning_rate": 2.26989546637263e-06,
+      "loss": 1.5304,
+      "step": 2710
+    },
+    {
+      "epoch": 0.7837525296328418,
+      "grad_norm": 1.0279100329909745,
+      "learning_rate": 2.2640780279711204e-06,
+      "loss": 1.534,
+      "step": 2711
+    },
+    {
+      "epoch": 0.7840416305290546,
+      "grad_norm": 1.0329508615037883,
+      "learning_rate": 2.2582671020178338e-06,
+      "loss": 1.5558,
+      "step": 2712
+    },
+    {
+      "epoch": 0.7843307314252674,
+      "grad_norm": 1.0313171182314225,
+      "learning_rate": 2.2524626934046667e-06,
+      "loss": 1.5393,
+      "step": 2713
+    },
+    {
+      "epoch": 0.7846198323214802,
+      "grad_norm": 1.0459786342678203,
+      "learning_rate": 2.246664807018041e-06,
+      "loss": 1.5502,
+      "step": 2714
+    },
+    {
+      "epoch": 0.784908933217693,
+      "grad_norm": 0.9986793394390265,
+      "learning_rate": 2.2408734477388714e-06,
+      "loss": 1.5721,
+      "step": 2715
+    },
+    {
+      "epoch": 0.7851980341139058,
+      "grad_norm": 1.0175038891382477,
+      "learning_rate": 2.235088620442588e-06,
+      "loss": 1.5822,
+      "step": 2716
+    },
+    {
+      "epoch": 0.7854871350101186,
+      "grad_norm": 1.037856169696711,
+      "learning_rate": 2.2293103299991182e-06,
+      "loss": 1.508,
+      "step": 2717
+    },
+    {
+      "epoch": 0.7857762359063313,
+      "grad_norm": 1.0252571641355601,
+      "learning_rate": 2.2235385812728925e-06,
+      "loss": 1.5276,
+      "step": 2718
+    },
+    {
+      "epoch": 0.7860653368025441,
+      "grad_norm": 0.9927855201591951,
+      "learning_rate": 2.2177733791228264e-06,
+      "loss": 1.4685,
+      "step": 2719
+    },
+    {
+      "epoch": 0.7863544376987569,
+      "grad_norm": 1.0140697724623526,
+      "learning_rate": 2.212014728402323e-06,
+      "loss": 1.5325,
+      "step": 2720
+    },
+    {
+      "epoch": 0.7866435385949696,
+      "grad_norm": 1.03708230200635,
+      "learning_rate": 2.206262633959282e-06,
+      "loss": 1.5278,
+      "step": 2721
+    },
+    {
+      "epoch": 0.7869326394911824,
+      "grad_norm": 0.9841737847014148,
+      "learning_rate": 2.2005171006360728e-06,
+      "loss": 1.5351,
+      "step": 2722
+    },
+    {
+      "epoch": 0.7872217403873952,
+      "grad_norm": 1.0186478608084988,
+      "learning_rate": 2.1947781332695406e-06,
+      "loss": 1.5917,
+      "step": 2723
+    },
+    {
+      "epoch": 0.7875108412836079,
+      "grad_norm": 1.0370585116744402,
+      "learning_rate": 2.1890457366910146e-06,
+      "loss": 1.5397,
+      "step": 2724
+    },
+    {
+      "epoch": 0.7877999421798207,
+      "grad_norm": 1.0945145073941631,
+      "learning_rate": 2.1833199157262798e-06,
+      "loss": 1.523,
+      "step": 2725
+    },
+    {
+      "epoch": 0.7880890430760336,
+      "grad_norm": 1.0962372641409628,
+      "learning_rate": 2.17760067519559e-06,
+      "loss": 1.5633,
+      "step": 2726
+    },
+    {
+      "epoch": 0.7883781439722463,
+      "grad_norm": 1.030202874127414,
+      "learning_rate": 2.1718880199136616e-06,
+      "loss": 1.553,
+      "step": 2727
+    },
+    {
+      "epoch": 0.7886672448684591,
+      "grad_norm": 1.0540126861235477,
+      "learning_rate": 2.166181954689669e-06,
+      "loss": 1.4786,
+      "step": 2728
+    },
+    {
+      "epoch": 0.7889563457646719,
+      "grad_norm": 1.0475977435706332,
+      "learning_rate": 2.160482484327232e-06,
+      "loss": 1.5702,
+      "step": 2729
+    },
+    {
+      "epoch": 0.7892454466608847,
+      "grad_norm": 1.0205569340224987,
+      "learning_rate": 2.1547896136244204e-06,
+      "loss": 1.5767,
+      "step": 2730
+    },
+    {
+      "epoch": 0.7895345475570974,
+      "grad_norm": 1.003309451994158,
+      "learning_rate": 2.149103347373753e-06,
+      "loss": 1.4726,
+      "step": 2731
+    },
+    {
+      "epoch": 0.7898236484533102,
+      "grad_norm": 1.0340675515563924,
+      "learning_rate": 2.143423690362185e-06,
+      "loss": 1.6098,
+      "step": 2732
+    },
+    {
+      "epoch": 0.790112749349523,
+      "grad_norm": 1.023751613786549,
+      "learning_rate": 2.137750647371104e-06,
+      "loss": 1.6044,
+      "step": 2733
+    },
+    {
+      "epoch": 0.7904018502457357,
+      "grad_norm": 1.0389661926802263,
+      "learning_rate": 2.1320842231763395e-06,
+      "loss": 1.5992,
+      "step": 2734
+    },
+    {
+      "epoch": 0.7906909511419485,
+      "grad_norm": 1.035821080983456,
+      "learning_rate": 2.12642442254814e-06,
+      "loss": 1.4436,
+      "step": 2735
+    },
+    {
+      "epoch": 0.7909800520381614,
+      "grad_norm": 1.0177596990047655,
+      "learning_rate": 2.120771250251178e-06,
+      "loss": 1.5733,
+      "step": 2736
+    },
+    {
+      "epoch": 0.791269152934374,
+      "grad_norm": 1.0223613490385248,
+      "learning_rate": 2.115124711044553e-06,
+      "loss": 1.6145,
+      "step": 2737
+    },
+    {
+      "epoch": 0.7915582538305869,
+      "grad_norm": 0.9798560153217744,
+      "learning_rate": 2.109484809681778e-06,
+      "loss": 1.5226,
+      "step": 2738
+    },
+    {
+      "epoch": 0.7918473547267997,
+      "grad_norm": 1.0313791542691273,
+      "learning_rate": 2.1038515509107735e-06,
+      "loss": 1.549,
+      "step": 2739
+    },
+    {
+      "epoch": 0.7921364556230124,
+      "grad_norm": 0.9909766977580767,
+      "learning_rate": 2.098224939473871e-06,
+      "loss": 1.5653,
+      "step": 2740
+    },
+    {
+      "epoch": 0.7924255565192252,
+      "grad_norm": 1.036254236234448,
+      "learning_rate": 2.0926049801078073e-06,
+      "loss": 1.5203,
+      "step": 2741
+    },
+    {
+      "epoch": 0.792714657415438,
+      "grad_norm": 1.0562005096478166,
+      "learning_rate": 2.0869916775437184e-06,
+      "loss": 1.595,
+      "step": 2742
+    },
+    {
+      "epoch": 0.7930037583116508,
+      "grad_norm": 1.047911822577376,
+      "learning_rate": 2.081385036507131e-06,
+      "loss": 1.6768,
+      "step": 2743
+    },
+    {
+      "epoch": 0.7932928592078635,
+      "grad_norm": 1.0179660522581466,
+      "learning_rate": 2.075785061717975e-06,
+      "loss": 1.5552,
+      "step": 2744
+    },
+    {
+      "epoch": 0.7935819601040763,
+      "grad_norm": 1.0349588329717203,
+      "learning_rate": 2.0701917578905596e-06,
+      "loss": 1.5375,
+      "step": 2745
+    },
+    {
+      "epoch": 0.7938710610002891,
+      "grad_norm": 1.034900036094183,
+      "learning_rate": 2.0646051297335767e-06,
+      "loss": 1.5133,
+      "step": 2746
+    },
+    {
+      "epoch": 0.7941601618965018,
+      "grad_norm": 1.0557982889116977,
+      "learning_rate": 2.0590251819501074e-06,
+      "loss": 1.5324,
+      "step": 2747
+    },
+    {
+      "epoch": 0.7944492627927147,
+      "grad_norm": 1.0169676874596782,
+      "learning_rate": 2.053451919237599e-06,
+      "loss": 1.6089,
+      "step": 2748
+    },
+    {
+      "epoch": 0.7947383636889275,
+      "grad_norm": 1.0091371908451983,
+      "learning_rate": 2.04788534628788e-06,
+      "loss": 1.4463,
+      "step": 2749
+    },
+    {
+      "epoch": 0.7950274645851402,
+      "grad_norm": 1.026329672174675,
+      "learning_rate": 2.042325467787137e-06,
+      "loss": 1.561,
+      "step": 2750
+    },
+    {
+      "epoch": 0.795316565481353,
+      "grad_norm": 1.0275505395416442,
+      "learning_rate": 2.0367722884159334e-06,
+      "loss": 1.5193,
+      "step": 2751
+    },
+    {
+      "epoch": 0.7956056663775658,
+      "grad_norm": 13.313648843933533,
+      "learning_rate": 2.0312258128491833e-06,
+      "loss": 1.7386,
+      "step": 2752
+    },
+    {
+      "epoch": 0.7958947672737785,
+      "grad_norm": 1.0582130274906696,
+      "learning_rate": 2.0256860457561565e-06,
+      "loss": 1.4538,
+      "step": 2753
+    },
+    {
+      "epoch": 0.7961838681699913,
+      "grad_norm": 1.043433377192559,
+      "learning_rate": 2.0201529918004847e-06,
+      "loss": 1.4754,
+      "step": 2754
+    },
+    {
+      "epoch": 0.7964729690662041,
+      "grad_norm": 1.0003980112616504,
+      "learning_rate": 2.0146266556401405e-06,
+      "loss": 1.5319,
+      "step": 2755
+    },
+    {
+      "epoch": 0.7967620699624168,
+      "grad_norm": 1.0209309019054675,
+      "learning_rate": 2.0091070419274418e-06,
+      "loss": 1.5078,
+      "step": 2756
+    },
+    {
+      "epoch": 0.7970511708586296,
+      "grad_norm": 0.9971703157805758,
+      "learning_rate": 2.0035941553090466e-06,
+      "loss": 1.5064,
+      "step": 2757
+    },
+    {
+      "epoch": 0.7973402717548425,
+      "grad_norm": 1.0031007867253694,
+      "learning_rate": 1.9980880004259573e-06,
+      "loss": 1.5047,
+      "step": 2758
+    },
+    {
+      "epoch": 0.7976293726510553,
+      "grad_norm": 0.998407863104387,
+      "learning_rate": 1.9925885819134973e-06,
+      "loss": 1.4317,
+      "step": 2759
+    },
+    {
+      "epoch": 0.797918473547268,
+      "grad_norm": 1.0055192748415027,
+      "learning_rate": 1.9870959044013284e-06,
+      "loss": 1.4929,
+      "step": 2760
+    },
+    {
+      "epoch": 0.7982075744434808,
+      "grad_norm": 1.044423063556292,
+      "learning_rate": 1.9816099725134375e-06,
+      "loss": 1.5663,
+      "step": 2761
+    },
+    {
+      "epoch": 0.7984966753396936,
+      "grad_norm": 1.043165472856723,
+      "learning_rate": 1.976130790868126e-06,
+      "loss": 1.4963,
+      "step": 2762
+    },
+    {
+      "epoch": 0.7987857762359063,
+      "grad_norm": 1.056999410192017,
+      "learning_rate": 1.9706583640780174e-06,
+      "loss": 1.5257,
+      "step": 2763
+    },
+    {
+      "epoch": 0.7990748771321191,
+      "grad_norm": 1.0070270410066195,
+      "learning_rate": 1.9651926967500435e-06,
+      "loss": 1.4628,
+      "step": 2764
+    },
+    {
+      "epoch": 0.7993639780283319,
+      "grad_norm": 1.0426642173662626,
+      "learning_rate": 1.9597337934854567e-06,
+      "loss": 1.6614,
+      "step": 2765
+    },
+    {
+      "epoch": 0.7996530789245446,
+      "grad_norm": 1.0243864383426298,
+      "learning_rate": 1.954281658879804e-06,
+      "loss": 1.5584,
+      "step": 2766
+    },
+    {
+      "epoch": 0.7999421798207574,
+      "grad_norm": 1.0038255587894498,
+      "learning_rate": 1.948836297522937e-06,
+      "loss": 1.4904,
+      "step": 2767
+    },
+    {
+      "epoch": 0.8002312807169703,
+      "grad_norm": 1.027827226485763,
+      "learning_rate": 1.94339771399901e-06,
+      "loss": 1.6162,
+      "step": 2768
+    },
+    {
+      "epoch": 0.800520381613183,
+      "grad_norm": 1.0066764680680251,
+      "learning_rate": 1.937965912886465e-06,
+      "loss": 1.5841,
+      "step": 2769
+    },
+    {
+      "epoch": 0.8008094825093958,
+      "grad_norm": 1.0232114716394152,
+      "learning_rate": 1.932540898758042e-06,
+      "loss": 1.4868,
+      "step": 2770
+    },
+    {
+      "epoch": 0.8010985834056086,
+      "grad_norm": 1.0619591014893468,
+      "learning_rate": 1.927122676180756e-06,
+      "loss": 1.4854,
+      "step": 2771
+    },
+    {
+      "epoch": 0.8013876843018214,
+      "grad_norm": 1.0267232228645986,
+      "learning_rate": 1.9217112497159175e-06,
+      "loss": 1.4977,
+      "step": 2772
+    },
+    {
+      "epoch": 0.8016767851980341,
+      "grad_norm": 1.0359171976552342,
+      "learning_rate": 1.9163066239191076e-06,
+      "loss": 1.4673,
+      "step": 2773
+    },
+    {
+      "epoch": 0.8019658860942469,
+      "grad_norm": 1.080370746923145,
+      "learning_rate": 1.9109088033401803e-06,
+      "loss": 1.4257,
+      "step": 2774
+    },
+    {
+      "epoch": 0.8022549869904597,
+      "grad_norm": 1.0117605162135948,
+      "learning_rate": 1.9055177925232704e-06,
+      "loss": 1.4826,
+      "step": 2775
+    },
+    {
+      "epoch": 0.8025440878866724,
+      "grad_norm": 1.0827644426535612,
+      "learning_rate": 1.9001335960067724e-06,
+      "loss": 1.6365,
+      "step": 2776
+    },
+    {
+      "epoch": 0.8028331887828852,
+      "grad_norm": 1.0152873876127775,
+      "learning_rate": 1.8947562183233415e-06,
+      "loss": 1.5482,
+      "step": 2777
+    },
+    {
+      "epoch": 0.803122289679098,
+      "grad_norm": 1.0441615031382376,
+      "learning_rate": 1.8893856639999032e-06,
+      "loss": 1.4753,
+      "step": 2778
+    },
+    {
+      "epoch": 0.8034113905753107,
+      "grad_norm": 1.027042159998036,
+      "learning_rate": 1.8840219375576297e-06,
+      "loss": 1.5594,
+      "step": 2779
+    },
+    {
+      "epoch": 0.8037004914715236,
+      "grad_norm": 1.0035389815048292,
+      "learning_rate": 1.8786650435119446e-06,
+      "loss": 1.4271,
+      "step": 2780
+    },
+    {
+      "epoch": 0.8039895923677364,
+      "grad_norm": 0.9836029464324794,
+      "learning_rate": 1.8733149863725275e-06,
+      "loss": 1.4881,
+      "step": 2781
+    },
+    {
+      "epoch": 0.8042786932639491,
+      "grad_norm": 1.0112823345897357,
+      "learning_rate": 1.8679717706432987e-06,
+      "loss": 1.6043,
+      "step": 2782
+    },
+    {
+      "epoch": 0.8045677941601619,
+      "grad_norm": 1.003811564258474,
+      "learning_rate": 1.8626354008224167e-06,
+      "loss": 1.502,
+      "step": 2783
+    },
+    {
+      "epoch": 0.8048568950563747,
+      "grad_norm": 0.9653043845335706,
+      "learning_rate": 1.8573058814022759e-06,
+      "loss": 1.4656,
+      "step": 2784
+    },
+    {
+      "epoch": 0.8051459959525874,
+      "grad_norm": 1.0316187091079672,
+      "learning_rate": 1.8519832168695118e-06,
+      "loss": 1.6002,
+      "step": 2785
+    },
+    {
+      "epoch": 0.8054350968488002,
+      "grad_norm": 1.0846367745851,
+      "learning_rate": 1.8466674117049788e-06,
+      "loss": 1.6449,
+      "step": 2786
+    },
+    {
+      "epoch": 0.805724197745013,
+      "grad_norm": 1.0087445495914147,
+      "learning_rate": 1.8413584703837618e-06,
+      "loss": 1.5094,
+      "step": 2787
+    },
+    {
+      "epoch": 0.8060132986412258,
+      "grad_norm": 1.0274192745042854,
+      "learning_rate": 1.83605639737517e-06,
+      "loss": 1.6316,
+      "step": 2788
+    },
+    {
+      "epoch": 0.8063023995374385,
+      "grad_norm": 1.030673066626127,
+      "learning_rate": 1.8307611971427275e-06,
+      "loss": 1.4976,
+      "step": 2789
+    },
+    {
+      "epoch": 0.8065915004336514,
+      "grad_norm": 1.0321067570295515,
+      "learning_rate": 1.8254728741441686e-06,
+      "loss": 1.5581,
+      "step": 2790
+    },
+    {
+      "epoch": 0.8068806013298642,
+      "grad_norm": 1.0135259548640718,
+      "learning_rate": 1.8201914328314452e-06,
+      "loss": 1.5194,
+      "step": 2791
+    },
+    {
+      "epoch": 0.8071697022260769,
+      "grad_norm": 1.0582542854965402,
+      "learning_rate": 1.8149168776507154e-06,
+      "loss": 1.5602,
+      "step": 2792
+    },
+    {
+      "epoch": 0.8074588031222897,
+      "grad_norm": 1.0225082082488506,
+      "learning_rate": 1.8096492130423348e-06,
+      "loss": 1.5286,
+      "step": 2793
+    },
+    {
+      "epoch": 0.8077479040185025,
+      "grad_norm": 1.0140222605606406,
+      "learning_rate": 1.8043884434408598e-06,
+      "loss": 1.4777,
+      "step": 2794
+    },
+    {
+      "epoch": 0.8080370049147152,
+      "grad_norm": 1.019590451699603,
+      "learning_rate": 1.7991345732750475e-06,
+      "loss": 1.6325,
+      "step": 2795
+    },
+    {
+      "epoch": 0.808326105810928,
+      "grad_norm": 1.0321746485602898,
+      "learning_rate": 1.7938876069678423e-06,
+      "loss": 1.4944,
+      "step": 2796
+    },
+    {
+      "epoch": 0.8086152067071408,
+      "grad_norm": 1.0251884646740983,
+      "learning_rate": 1.7886475489363731e-06,
+      "loss": 1.5311,
+      "step": 2797
+    },
+    {
+      "epoch": 0.8089043076033535,
+      "grad_norm": 1.0699877049006887,
+      "learning_rate": 1.783414403591962e-06,
+      "loss": 1.5943,
+      "step": 2798
+    },
+    {
+      "epoch": 0.8091934084995663,
+      "grad_norm": 1.0346112763294042,
+      "learning_rate": 1.7781881753401042e-06,
+      "loss": 1.545,
+      "step": 2799
+    },
+    {
+      "epoch": 0.8094825093957791,
+      "grad_norm": 1.013358529487172,
+      "learning_rate": 1.7729688685804746e-06,
+      "loss": 1.595,
+      "step": 2800
+    },
+    {
+      "epoch": 0.8097716102919919,
+      "grad_norm": 1.0483003830437474,
+      "learning_rate": 1.7677564877069208e-06,
+      "loss": 1.608,
+      "step": 2801
+    },
+    {
+      "epoch": 0.8100607111882047,
+      "grad_norm": 1.0265311038920306,
+      "learning_rate": 1.762551037107464e-06,
+      "loss": 1.4789,
+      "step": 2802
+    },
+    {
+      "epoch": 0.8103498120844175,
+      "grad_norm": 1.0282427992280434,
+      "learning_rate": 1.7573525211642839e-06,
+      "loss": 1.6618,
+      "step": 2803
+    },
+    {
+      "epoch": 0.8106389129806303,
+      "grad_norm": 1.0389562177226028,
+      "learning_rate": 1.7521609442537246e-06,
+      "loss": 1.482,
+      "step": 2804
+    },
+    {
+      "epoch": 0.810928013876843,
+      "grad_norm": 1.0188612287256182,
+      "learning_rate": 1.7469763107462955e-06,
+      "loss": 1.4504,
+      "step": 2805
+    },
+    {
+      "epoch": 0.8112171147730558,
+      "grad_norm": 1.028721054156086,
+      "learning_rate": 1.7417986250066498e-06,
+      "loss": 1.5395,
+      "step": 2806
+    },
+    {
+      "epoch": 0.8115062156692686,
+      "grad_norm": 1.0592233112913134,
+      "learning_rate": 1.7366278913935974e-06,
+      "loss": 1.4805,
+      "step": 2807
+    },
+    {
+      "epoch": 0.8117953165654813,
+      "grad_norm": 1.023432708935403,
+      "learning_rate": 1.7314641142600975e-06,
+      "loss": 1.55,
+      "step": 2808
+    },
+    {
+      "epoch": 0.8120844174616941,
+      "grad_norm": 1.006088669649468,
+      "learning_rate": 1.7263072979532513e-06,
+      "loss": 1.4976,
+      "step": 2809
+    },
+    {
+      "epoch": 0.8123735183579069,
+      "grad_norm": 1.0118739789625844,
+      "learning_rate": 1.7211574468142965e-06,
+      "loss": 1.5105,
+      "step": 2810
+    },
+    {
+      "epoch": 0.8126626192541196,
+      "grad_norm": 1.1196909515606084,
+      "learning_rate": 1.7160145651786098e-06,
+      "loss": 1.5756,
+      "step": 2811
+    },
+    {
+      "epoch": 0.8129517201503325,
+      "grad_norm": 1.0315260152179635,
+      "learning_rate": 1.7108786573757019e-06,
+      "loss": 1.5017,
+      "step": 2812
+    },
+    {
+      "epoch": 0.8132408210465453,
+      "grad_norm": 1.0340138088572837,
+      "learning_rate": 1.7057497277292156e-06,
+      "loss": 1.5386,
+      "step": 2813
+    },
+    {
+      "epoch": 0.813529921942758,
+      "grad_norm": 1.0674244395261991,
+      "learning_rate": 1.7006277805569104e-06,
+      "loss": 1.6127,
+      "step": 2814
+    },
+    {
+      "epoch": 0.8138190228389708,
+      "grad_norm": 1.0352685050111419,
+      "learning_rate": 1.6955128201706761e-06,
+      "loss": 1.6168,
+      "step": 2815
+    },
+    {
+      "epoch": 0.8141081237351836,
+      "grad_norm": 1.0274279415529453,
+      "learning_rate": 1.6904048508765169e-06,
+      "loss": 1.5104,
+      "step": 2816
+    },
+    {
+      "epoch": 0.8143972246313964,
+      "grad_norm": 1.0020287599344555,
+      "learning_rate": 1.6853038769745466e-06,
+      "loss": 1.5233,
+      "step": 2817
+    },
+    {
+      "epoch": 0.8146863255276091,
+      "grad_norm": 1.0411144636140115,
+      "learning_rate": 1.6802099027590025e-06,
+      "loss": 1.5059,
+      "step": 2818
+    },
+    {
+      "epoch": 0.8149754264238219,
+      "grad_norm": 1.0345914726341394,
+      "learning_rate": 1.6751229325182194e-06,
+      "loss": 1.5014,
+      "step": 2819
+    },
+    {
+      "epoch": 0.8152645273200347,
+      "grad_norm": 1.0189996962566437,
+      "learning_rate": 1.6700429705346365e-06,
+      "loss": 1.5738,
+      "step": 2820
+    },
+    {
+      "epoch": 0.8155536282162474,
+      "grad_norm": 1.0144054259333344,
+      "learning_rate": 1.664970021084794e-06,
+      "loss": 1.5623,
+      "step": 2821
+    },
+    {
+      "epoch": 0.8158427291124603,
+      "grad_norm": 1.081173342530248,
+      "learning_rate": 1.6599040884393348e-06,
+      "loss": 1.619,
+      "step": 2822
+    },
+    {
+      "epoch": 0.8161318300086731,
+      "grad_norm": 1.008452170224636,
+      "learning_rate": 1.6548451768629847e-06,
+      "loss": 1.5095,
+      "step": 2823
+    },
+    {
+      "epoch": 0.8164209309048858,
+      "grad_norm": 1.0365585151341772,
+      "learning_rate": 1.6497932906145653e-06,
+      "loss": 1.5406,
+      "step": 2824
+    },
+    {
+      "epoch": 0.8167100318010986,
+      "grad_norm": 1.0139363620711301,
+      "learning_rate": 1.6447484339469843e-06,
+      "loss": 1.5082,
+      "step": 2825
+    },
+    {
+      "epoch": 0.8169991326973114,
+      "grad_norm": 1.0560824630714423,
+      "learning_rate": 1.6397106111072304e-06,
+      "loss": 1.5618,
+      "step": 2826
+    },
+    {
+      "epoch": 0.8172882335935241,
+      "grad_norm": 1.0295046071131588,
+      "learning_rate": 1.6346798263363662e-06,
+      "loss": 1.508,
+      "step": 2827
+    },
+    {
+      "epoch": 0.8175773344897369,
+      "grad_norm": 1.0113717230861838,
+      "learning_rate": 1.629656083869534e-06,
+      "loss": 1.5794,
+      "step": 2828
+    },
+    {
+      "epoch": 0.8178664353859497,
+      "grad_norm": 0.9899686145453735,
+      "learning_rate": 1.6246393879359512e-06,
+      "loss": 1.3909,
+      "step": 2829
+    },
+    {
+      "epoch": 0.8181555362821624,
+      "grad_norm": 1.0156876998704936,
+      "learning_rate": 1.6196297427588947e-06,
+      "loss": 1.5197,
+      "step": 2830
+    },
+    {
+      "epoch": 0.8184446371783752,
+      "grad_norm": 1.0386226760505475,
+      "learning_rate": 1.6146271525557089e-06,
+      "loss": 1.5475,
+      "step": 2831
+    },
+    {
+      "epoch": 0.818733738074588,
+      "grad_norm": 0.9887333454264747,
+      "learning_rate": 1.609631621537805e-06,
+      "loss": 1.5934,
+      "step": 2832
+    },
+    {
+      "epoch": 0.8190228389708009,
+      "grad_norm": 1.0159956425775778,
+      "learning_rate": 1.6046431539106432e-06,
+      "loss": 1.5501,
+      "step": 2833
+    },
+    {
+      "epoch": 0.8193119398670136,
+      "grad_norm": 1.0151934739582318,
+      "learning_rate": 1.599661753873739e-06,
+      "loss": 1.5144,
+      "step": 2834
+    },
+    {
+      "epoch": 0.8196010407632264,
+      "grad_norm": 0.9972660246586103,
+      "learning_rate": 1.5946874256206612e-06,
+      "loss": 1.5172,
+      "step": 2835
+    },
+    {
+      "epoch": 0.8198901416594392,
+      "grad_norm": 1.039634402856375,
+      "learning_rate": 1.5897201733390277e-06,
+      "loss": 1.5737,
+      "step": 2836
+    },
+    {
+      "epoch": 0.8201792425556519,
+      "grad_norm": 1.0363509779499416,
+      "learning_rate": 1.5847600012104914e-06,
+      "loss": 1.4994,
+      "step": 2837
+    },
+    {
+      "epoch": 0.8204683434518647,
+      "grad_norm": 1.0251623658124283,
+      "learning_rate": 1.5798069134107486e-06,
+      "loss": 1.5512,
+      "step": 2838
+    },
+    {
+      "epoch": 0.8207574443480775,
+      "grad_norm": 1.0211548732336777,
+      "learning_rate": 1.574860914109535e-06,
+      "loss": 1.5301,
+      "step": 2839
+    },
+    {
+      "epoch": 0.8210465452442902,
+      "grad_norm": 1.0355917724158974,
+      "learning_rate": 1.5699220074706146e-06,
+      "loss": 1.4762,
+      "step": 2840
+    },
+    {
+      "epoch": 0.821335646140503,
+      "grad_norm": 1.017500280316858,
+      "learning_rate": 1.5649901976517802e-06,
+      "loss": 1.4317,
+      "step": 2841
+    },
+    {
+      "epoch": 0.8216247470367158,
+      "grad_norm": 1.0036590886664063,
+      "learning_rate": 1.560065488804855e-06,
+      "loss": 1.5182,
+      "step": 2842
+    },
+    {
+      "epoch": 0.8219138479329285,
+      "grad_norm": 0.9894486689649677,
+      "learning_rate": 1.5551478850756795e-06,
+      "loss": 1.4683,
+      "step": 2843
+    },
+    {
+      "epoch": 0.8222029488291414,
+      "grad_norm": 1.0087180715002666,
+      "learning_rate": 1.5502373906041135e-06,
+      "loss": 1.5754,
+      "step": 2844
+    },
+    {
+      "epoch": 0.8224920497253542,
+      "grad_norm": 1.0187239412167362,
+      "learning_rate": 1.5453340095240333e-06,
+      "loss": 1.5474,
+      "step": 2845
+    },
+    {
+      "epoch": 0.822781150621567,
+      "grad_norm": 1.049889080098156,
+      "learning_rate": 1.540437745963329e-06,
+      "loss": 1.5515,
+      "step": 2846
+    },
+    {
+      "epoch": 0.8230702515177797,
+      "grad_norm": 1.0436940671884496,
+      "learning_rate": 1.535548604043895e-06,
+      "loss": 1.5192,
+      "step": 2847
+    },
+    {
+      "epoch": 0.8233593524139925,
+      "grad_norm": 1.048644944025632,
+      "learning_rate": 1.5306665878816273e-06,
+      "loss": 1.5735,
+      "step": 2848
+    },
+    {
+      "epoch": 0.8236484533102053,
+      "grad_norm": 1.0575875383288051,
+      "learning_rate": 1.5257917015864343e-06,
+      "loss": 1.5497,
+      "step": 2849
+    },
+    {
+      "epoch": 0.823937554206418,
+      "grad_norm": 1.0458675537278164,
+      "learning_rate": 1.5209239492622118e-06,
+      "loss": 1.4702,
+      "step": 2850
+    },
+    {
+      "epoch": 0.8242266551026308,
+      "grad_norm": 1.012883654982487,
+      "learning_rate": 1.516063335006851e-06,
+      "loss": 1.5025,
+      "step": 2851
+    },
+    {
+      "epoch": 0.8245157559988436,
+      "grad_norm": 1.0639291626391147,
+      "learning_rate": 1.5112098629122418e-06,
+      "loss": 1.5346,
+      "step": 2852
+    },
+    {
+      "epoch": 0.8248048568950563,
+      "grad_norm": 0.9997360396410622,
+      "learning_rate": 1.5063635370642515e-06,
+      "loss": 1.519,
+      "step": 2853
+    },
+    {
+      "epoch": 0.8250939577912692,
+      "grad_norm": 1.037208246678323,
+      "learning_rate": 1.5015243615427367e-06,
+      "loss": 1.4712,
+      "step": 2854
+    },
+    {
+      "epoch": 0.825383058687482,
+      "grad_norm": 1.0370712092608567,
+      "learning_rate": 1.496692340421534e-06,
+      "loss": 1.4428,
+      "step": 2855
+    },
+    {
+      "epoch": 0.8256721595836947,
+      "grad_norm": 1.0367520816109999,
+      "learning_rate": 1.4918674777684593e-06,
+      "loss": 1.5668,
+      "step": 2856
+    },
+    {
+      "epoch": 0.8259612604799075,
+      "grad_norm": 1.037898763392011,
+      "learning_rate": 1.4870497776452987e-06,
+      "loss": 1.577,
+      "step": 2857
+    },
+    {
+      "epoch": 0.8262503613761203,
+      "grad_norm": 1.0321330230131815,
+      "learning_rate": 1.482239244107806e-06,
+      "loss": 1.5297,
+      "step": 2858
+    },
+    {
+      "epoch": 0.826539462272333,
+      "grad_norm": 0.9960223711296703,
+      "learning_rate": 1.4774358812057099e-06,
+      "loss": 1.4324,
+      "step": 2859
+    },
+    {
+      "epoch": 0.8268285631685458,
+      "grad_norm": 0.9950074208220431,
+      "learning_rate": 1.472639692982697e-06,
+      "loss": 1.5392,
+      "step": 2860
+    },
+    {
+      "epoch": 0.8271176640647586,
+      "grad_norm": 1.0193842420030303,
+      "learning_rate": 1.4678506834764117e-06,
+      "loss": 1.6206,
+      "step": 2861
+    },
+    {
+      "epoch": 0.8274067649609714,
+      "grad_norm": 1.0274541839793394,
+      "learning_rate": 1.4630688567184626e-06,
+      "loss": 1.6373,
+      "step": 2862
+    },
+    {
+      "epoch": 0.8276958658571841,
+      "grad_norm": 1.0093557416715597,
+      "learning_rate": 1.458294216734406e-06,
+      "loss": 1.4711,
+      "step": 2863
+    },
+    {
+      "epoch": 0.827984966753397,
+      "grad_norm": 0.9847081544781567,
+      "learning_rate": 1.453526767543747e-06,
+      "loss": 1.4931,
+      "step": 2864
+    },
+    {
+      "epoch": 0.8282740676496098,
+      "grad_norm": 0.9992104104592591,
+      "learning_rate": 1.4487665131599427e-06,
+      "loss": 1.4394,
+      "step": 2865
+    },
+    {
+      "epoch": 0.8285631685458225,
+      "grad_norm": 1.0146967628478392,
+      "learning_rate": 1.4440134575903863e-06,
+      "loss": 1.6129,
+      "step": 2866
+    },
+    {
+      "epoch": 0.8288522694420353,
+      "grad_norm": 1.015177086740991,
+      "learning_rate": 1.4392676048364196e-06,
+      "loss": 1.6257,
+      "step": 2867
+    },
+    {
+      "epoch": 0.8291413703382481,
+      "grad_norm": 1.0132087192160308,
+      "learning_rate": 1.4345289588933108e-06,
+      "loss": 1.5016,
+      "step": 2868
+    },
+    {
+      "epoch": 0.8294304712344608,
+      "grad_norm": 0.9818604641929798,
+      "learning_rate": 1.4297975237502703e-06,
+      "loss": 1.4384,
+      "step": 2869
+    },
+    {
+      "epoch": 0.8297195721306736,
+      "grad_norm": 1.046356050195022,
+      "learning_rate": 1.4250733033904307e-06,
+      "loss": 1.5307,
+      "step": 2870
+    },
+    {
+      "epoch": 0.8300086730268864,
+      "grad_norm": 1.0244781276189747,
+      "learning_rate": 1.4203563017908528e-06,
+      "loss": 1.531,
+      "step": 2871
+    },
+    {
+      "epoch": 0.8302977739230991,
+      "grad_norm": 1.0288839813417898,
+      "learning_rate": 1.4156465229225248e-06,
+      "loss": 1.5861,
+      "step": 2872
+    },
+    {
+      "epoch": 0.8305868748193119,
+      "grad_norm": 1.0017444595677945,
+      "learning_rate": 1.4109439707503502e-06,
+      "loss": 1.4851,
+      "step": 2873
+    },
+    {
+      "epoch": 0.8308759757155247,
+      "grad_norm": 1.0149137542784608,
+      "learning_rate": 1.4062486492331483e-06,
+      "loss": 1.6605,
+      "step": 2874
+    },
+    {
+      "epoch": 0.8311650766117374,
+      "grad_norm": 1.0403748224182918,
+      "learning_rate": 1.4015605623236516e-06,
+      "loss": 1.5055,
+      "step": 2875
+    },
+    {
+      "epoch": 0.8314541775079503,
+      "grad_norm": 1.0210038481263386,
+      "learning_rate": 1.3968797139685052e-06,
+      "loss": 1.6049,
+      "step": 2876
+    },
+    {
+      "epoch": 0.8317432784041631,
+      "grad_norm": 1.034845382747551,
+      "learning_rate": 1.3922061081082594e-06,
+      "loss": 1.5339,
+      "step": 2877
+    },
+    {
+      "epoch": 0.8320323793003759,
+      "grad_norm": 1.0024234527705465,
+      "learning_rate": 1.3875397486773634e-06,
+      "loss": 1.5893,
+      "step": 2878
+    },
+    {
+      "epoch": 0.8323214801965886,
+      "grad_norm": 1.014206502260035,
+      "learning_rate": 1.3828806396041739e-06,
+      "loss": 1.5067,
+      "step": 2879
+    },
+    {
+      "epoch": 0.8326105810928014,
+      "grad_norm": 1.0084022998625297,
+      "learning_rate": 1.3782287848109354e-06,
+      "loss": 1.5929,
+      "step": 2880
+    },
+    {
+      "epoch": 0.8328996819890142,
+      "grad_norm": 1.014637050681247,
+      "learning_rate": 1.3735841882137912e-06,
+      "loss": 1.5491,
+      "step": 2881
+    },
+    {
+      "epoch": 0.8331887828852269,
+      "grad_norm": 1.0262885440592302,
+      "learning_rate": 1.3689468537227679e-06,
+      "loss": 1.5471,
+      "step": 2882
+    },
+    {
+      "epoch": 0.8334778837814397,
+      "grad_norm": 0.9731741634018817,
+      "learning_rate": 1.3643167852417894e-06,
+      "loss": 1.5053,
+      "step": 2883
+    },
+    {
+      "epoch": 0.8337669846776525,
+      "grad_norm": 1.0062504115100417,
+      "learning_rate": 1.3596939866686532e-06,
+      "loss": 1.582,
+      "step": 2884
+    },
+    {
+      "epoch": 0.8340560855738652,
+      "grad_norm": 1.0405577614877042,
+      "learning_rate": 1.3550784618950363e-06,
+      "loss": 1.56,
+      "step": 2885
+    },
+    {
+      "epoch": 0.834345186470078,
+      "grad_norm": 1.0315227011882186,
+      "learning_rate": 1.350470214806503e-06,
+      "loss": 1.5449,
+      "step": 2886
+    },
+    {
+      "epoch": 0.8346342873662909,
+      "grad_norm": 0.9956473665010385,
+      "learning_rate": 1.3458692492824764e-06,
+      "loss": 1.4926,
+      "step": 2887
+    },
+    {
+      "epoch": 0.8349233882625036,
+      "grad_norm": 1.029567250929445,
+      "learning_rate": 1.341275569196263e-06,
+      "loss": 1.4903,
+      "step": 2888
+    },
+    {
+      "epoch": 0.8352124891587164,
+      "grad_norm": 1.0244959916815293,
+      "learning_rate": 1.336689178415025e-06,
+      "loss": 1.6724,
+      "step": 2889
+    },
+    {
+      "epoch": 0.8355015900549292,
+      "grad_norm": 1.0497412030945819,
+      "learning_rate": 1.3321100807997978e-06,
+      "loss": 1.6181,
+      "step": 2890
+    },
+    {
+      "epoch": 0.835790690951142,
+      "grad_norm": 1.1047324184605676,
+      "learning_rate": 1.3275382802054704e-06,
+      "loss": 1.4583,
+      "step": 2891
+    },
+    {
+      "epoch": 0.8360797918473547,
+      "grad_norm": 1.0854469030974707,
+      "learning_rate": 1.3229737804807884e-06,
+      "loss": 1.5735,
+      "step": 2892
+    },
+    {
+      "epoch": 0.8363688927435675,
+      "grad_norm": 0.9930359534503734,
+      "learning_rate": 1.3184165854683572e-06,
+      "loss": 1.4468,
+      "step": 2893
+    },
+    {
+      "epoch": 0.8366579936397803,
+      "grad_norm": 1.0098609108288683,
+      "learning_rate": 1.3138666990046289e-06,
+      "loss": 1.4508,
+      "step": 2894
+    },
+    {
+      "epoch": 0.836947094535993,
+      "grad_norm": 1.052751549195703,
+      "learning_rate": 1.3093241249198995e-06,
+      "loss": 1.5559,
+      "step": 2895
+    },
+    {
+      "epoch": 0.8372361954322058,
+      "grad_norm": 0.9921270358924388,
+      "learning_rate": 1.304788867038318e-06,
+      "loss": 1.5005,
+      "step": 2896
+    },
+    {
+      "epoch": 0.8375252963284187,
+      "grad_norm": 1.0207459875082465,
+      "learning_rate": 1.3002609291778656e-06,
+      "loss": 1.5361,
+      "step": 2897
+    },
+    {
+      "epoch": 0.8378143972246314,
+      "grad_norm": 1.0063746581056667,
+      "learning_rate": 1.2957403151503644e-06,
+      "loss": 1.5926,
+      "step": 2898
+    },
+    {
+      "epoch": 0.8381034981208442,
+      "grad_norm": 0.9947361121259657,
+      "learning_rate": 1.2912270287614737e-06,
+      "loss": 1.5027,
+      "step": 2899
+    },
+    {
+      "epoch": 0.838392599017057,
+      "grad_norm": 1.1501066461004779,
+      "learning_rate": 1.2867210738106817e-06,
+      "loss": 1.6237,
+      "step": 2900
+    },
+    {
+      "epoch": 0.8386816999132697,
+      "grad_norm": 1.0092487282840366,
+      "learning_rate": 1.2822224540913041e-06,
+      "loss": 1.5642,
+      "step": 2901
+    },
+    {
+      "epoch": 0.8389708008094825,
+      "grad_norm": 1.0001271494851989,
+      "learning_rate": 1.2777311733904785e-06,
+      "loss": 1.4856,
+      "step": 2902
+    },
+    {
+      "epoch": 0.8392599017056953,
+      "grad_norm": 1.0041383503447296,
+      "learning_rate": 1.2732472354891735e-06,
+      "loss": 1.5536,
+      "step": 2903
+    },
+    {
+      "epoch": 0.839549002601908,
+      "grad_norm": 0.9886387719658303,
+      "learning_rate": 1.2687706441621683e-06,
+      "loss": 1.4771,
+      "step": 2904
+    },
+    {
+      "epoch": 0.8398381034981208,
+      "grad_norm": 0.9949577028979317,
+      "learning_rate": 1.2643014031780555e-06,
+      "loss": 1.4883,
+      "step": 2905
+    },
+    {
+      "epoch": 0.8401272043943336,
+      "grad_norm": 1.0090675484357094,
+      "learning_rate": 1.2598395162992493e-06,
+      "loss": 1.4921,
+      "step": 2906
+    },
+    {
+      "epoch": 0.8404163052905465,
+      "grad_norm": 1.027858870534643,
+      "learning_rate": 1.2553849872819658e-06,
+      "loss": 1.4439,
+      "step": 2907
+    },
+    {
+      "epoch": 0.8407054061867592,
+      "grad_norm": 1.026763547771158,
+      "learning_rate": 1.2509378198762255e-06,
+      "loss": 1.5045,
+      "step": 2908
+    },
+    {
+      "epoch": 0.840994507082972,
+      "grad_norm": 1.0158803680829365,
+      "learning_rate": 1.2464980178258569e-06,
+      "loss": 1.4569,
+      "step": 2909
+    },
+    {
+      "epoch": 0.8412836079791848,
+      "grad_norm": 1.0373169774433542,
+      "learning_rate": 1.242065584868488e-06,
+      "loss": 1.5697,
+      "step": 2910
+    },
+    {
+      "epoch": 0.8415727088753975,
+      "grad_norm": 1.0016791318873441,
+      "learning_rate": 1.2376405247355393e-06,
+      "loss": 1.4602,
+      "step": 2911
+    },
+    {
+      "epoch": 0.8418618097716103,
+      "grad_norm": 0.9799617106571606,
+      "learning_rate": 1.233222841152224e-06,
+      "loss": 1.5081,
+      "step": 2912
+    },
+    {
+      "epoch": 0.8421509106678231,
+      "grad_norm": 1.003411111799833,
+      "learning_rate": 1.2288125378375505e-06,
+      "loss": 1.5255,
+      "step": 2913
+    },
+    {
+      "epoch": 0.8424400115640358,
+      "grad_norm": 0.9898178253370078,
+      "learning_rate": 1.2244096185043097e-06,
+      "loss": 1.539,
+      "step": 2914
+    },
+    {
+      "epoch": 0.8427291124602486,
+      "grad_norm": 1.0436779272515133,
+      "learning_rate": 1.2200140868590759e-06,
+      "loss": 1.5959,
+      "step": 2915
+    },
+    {
+      "epoch": 0.8430182133564614,
+      "grad_norm": 1.0223602748713398,
+      "learning_rate": 1.2156259466022103e-06,
+      "loss": 1.5217,
+      "step": 2916
+    },
+    {
+      "epoch": 0.8433073142526741,
+      "grad_norm": 1.0006949594655865,
+      "learning_rate": 1.2112452014278453e-06,
+      "loss": 1.524,
+      "step": 2917
+    },
+    {
+      "epoch": 0.843596415148887,
+      "grad_norm": 1.0061559229643389,
+      "learning_rate": 1.2068718550238867e-06,
+      "loss": 1.5439,
+      "step": 2918
+    },
+    {
+      "epoch": 0.8438855160450998,
+      "grad_norm": 1.0613494248132136,
+      "learning_rate": 1.2025059110720182e-06,
+      "loss": 1.5761,
+      "step": 2919
+    },
+    {
+      "epoch": 0.8441746169413126,
+      "grad_norm": 1.0231783930308536,
+      "learning_rate": 1.198147373247691e-06,
+      "loss": 1.5197,
+      "step": 2920
+    },
+    {
+      "epoch": 0.8444637178375253,
+      "grad_norm": 0.9938139240665564,
+      "learning_rate": 1.1937962452201156e-06,
+      "loss": 1.503,
+      "step": 2921
+    },
+    {
+      "epoch": 0.8447528187337381,
+      "grad_norm": 1.0098725240423672,
+      "learning_rate": 1.1894525306522653e-06,
+      "loss": 1.4343,
+      "step": 2922
+    },
+    {
+      "epoch": 0.8450419196299509,
+      "grad_norm": 1.0317209285989177,
+      "learning_rate": 1.1851162332008815e-06,
+      "loss": 1.5207,
+      "step": 2923
+    },
+    {
+      "epoch": 0.8453310205261636,
+      "grad_norm": 1.0226960317650402,
+      "learning_rate": 1.1807873565164507e-06,
+      "loss": 1.5511,
+      "step": 2924
+    },
+    {
+      "epoch": 0.8456201214223764,
+      "grad_norm": 1.0031559800609446,
+      "learning_rate": 1.1764659042432158e-06,
+      "loss": 1.6792,
+      "step": 2925
+    },
+    {
+      "epoch": 0.8459092223185892,
+      "grad_norm": 1.0241234239392405,
+      "learning_rate": 1.1721518800191744e-06,
+      "loss": 1.5987,
+      "step": 2926
+    },
+    {
+      "epoch": 0.8461983232148019,
+      "grad_norm": 0.9829164824319113,
+      "learning_rate": 1.1678452874760648e-06,
+      "loss": 1.5474,
+      "step": 2927
+    },
+    {
+      "epoch": 0.8464874241110147,
+      "grad_norm": 1.005267548010226,
+      "learning_rate": 1.1635461302393681e-06,
+      "loss": 1.5114,
+      "step": 2928
+    },
+    {
+      "epoch": 0.8467765250072276,
+      "grad_norm": 1.0160059849648995,
+      "learning_rate": 1.1592544119283133e-06,
+      "loss": 1.5416,
+      "step": 2929
+    },
+    {
+      "epoch": 0.8470656259034403,
+      "grad_norm": 1.0300507226568318,
+      "learning_rate": 1.1549701361558596e-06,
+      "loss": 1.5876,
+      "step": 2930
+    },
+    {
+      "epoch": 0.8473547267996531,
+      "grad_norm": 1.0147766434773908,
+      "learning_rate": 1.1506933065287062e-06,
+      "loss": 1.5744,
+      "step": 2931
+    },
+    {
+      "epoch": 0.8476438276958659,
+      "grad_norm": 1.0217444114749341,
+      "learning_rate": 1.1464239266472777e-06,
+      "loss": 1.4812,
+      "step": 2932
+    },
+    {
+      "epoch": 0.8479329285920786,
+      "grad_norm": 1.0534211807504215,
+      "learning_rate": 1.1421620001057354e-06,
+      "loss": 1.5701,
+      "step": 2933
+    },
+    {
+      "epoch": 0.8482220294882914,
+      "grad_norm": 1.0308488323653489,
+      "learning_rate": 1.1379075304919584e-06,
+      "loss": 1.5535,
+      "step": 2934
+    },
+    {
+      "epoch": 0.8485111303845042,
+      "grad_norm": 1.0228195878875888,
+      "learning_rate": 1.1336605213875507e-06,
+      "loss": 1.5593,
+      "step": 2935
+    },
+    {
+      "epoch": 0.848800231280717,
+      "grad_norm": 0.9924027528897786,
+      "learning_rate": 1.1294209763678387e-06,
+      "loss": 1.5639,
+      "step": 2936
+    },
+    {
+      "epoch": 0.8490893321769297,
+      "grad_norm": 0.9862293519188159,
+      "learning_rate": 1.125188899001861e-06,
+      "loss": 1.4321,
+      "step": 2937
+    },
+    {
+      "epoch": 0.8493784330731425,
+      "grad_norm": 1.082715607666728,
+      "learning_rate": 1.1209642928523722e-06,
+      "loss": 1.5177,
+      "step": 2938
+    },
+    {
+      "epoch": 0.8496675339693553,
+      "grad_norm": 1.0051400663791383,
+      "learning_rate": 1.116747161475833e-06,
+      "loss": 1.4617,
+      "step": 2939
+    },
+    {
+      "epoch": 0.849956634865568,
+      "grad_norm": 1.0193140055623413,
+      "learning_rate": 1.1125375084224198e-06,
+      "loss": 1.4989,
+      "step": 2940
+    },
+    {
+      "epoch": 0.8502457357617809,
+      "grad_norm": 1.0155264670316932,
+      "learning_rate": 1.108335337236005e-06,
+      "loss": 1.5608,
+      "step": 2941
+    },
+    {
+      "epoch": 0.8505348366579937,
+      "grad_norm": 1.0303523303795177,
+      "learning_rate": 1.1041406514541675e-06,
+      "loss": 1.5975,
+      "step": 2942
+    },
+    {
+      "epoch": 0.8508239375542064,
+      "grad_norm": 1.0206445309815315,
+      "learning_rate": 1.0999534546081847e-06,
+      "loss": 1.5811,
+      "step": 2943
+    },
+    {
+      "epoch": 0.8511130384504192,
+      "grad_norm": 1.013055201307031,
+      "learning_rate": 1.095773750223026e-06,
+      "loss": 1.5446,
+      "step": 2944
+    },
+    {
+      "epoch": 0.851402139346632,
+      "grad_norm": 1.011573505087957,
+      "learning_rate": 1.0916015418173553e-06,
+      "loss": 1.4776,
+      "step": 2945
+    },
+    {
+      "epoch": 0.8516912402428447,
+      "grad_norm": 1.0071575798441077,
+      "learning_rate": 1.0874368329035245e-06,
+      "loss": 1.6041,
+      "step": 2946
+    },
+    {
+      "epoch": 0.8519803411390575,
+      "grad_norm": 1.0101287048192047,
+      "learning_rate": 1.0832796269875757e-06,
+      "loss": 1.4334,
+      "step": 2947
+    },
+    {
+      "epoch": 0.8522694420352703,
+      "grad_norm": 1.065304659077651,
+      "learning_rate": 1.0791299275692324e-06,
+      "loss": 1.6138,
+      "step": 2948
+    },
+    {
+      "epoch": 0.852558542931483,
+      "grad_norm": 0.9857957217046613,
+      "learning_rate": 1.074987738141896e-06,
+      "loss": 1.6195,
+      "step": 2949
+    },
+    {
+      "epoch": 0.8528476438276958,
+      "grad_norm": 0.9696234365853619,
+      "learning_rate": 1.0708530621926538e-06,
+      "loss": 1.5222,
+      "step": 2950
+    },
+    {
+      "epoch": 0.8531367447239087,
+      "grad_norm": 1.0019491008972587,
+      "learning_rate": 1.066725903202256e-06,
+      "loss": 1.6274,
+      "step": 2951
+    },
+    {
+      "epoch": 0.8534258456201215,
+      "grad_norm": 1.0447704668105708,
+      "learning_rate": 1.0626062646451374e-06,
+      "loss": 1.5331,
+      "step": 2952
+    },
+    {
+      "epoch": 0.8537149465163342,
+      "grad_norm": 1.0384141651849303,
+      "learning_rate": 1.0584941499893897e-06,
+      "loss": 1.5241,
+      "step": 2953
+    },
+    {
+      "epoch": 0.854004047412547,
+      "grad_norm": 1.038454219850976,
+      "learning_rate": 1.0543895626967827e-06,
+      "loss": 1.5553,
+      "step": 2954
+    },
+    {
+      "epoch": 0.8542931483087598,
+      "grad_norm": 1.0155853284410439,
+      "learning_rate": 1.050292506222741e-06,
+      "loss": 1.5369,
+      "step": 2955
+    },
+    {
+      "epoch": 0.8545822492049725,
+      "grad_norm": 1.0354616967458374,
+      "learning_rate": 1.0462029840163478e-06,
+      "loss": 1.5916,
+      "step": 2956
+    },
+    {
+      "epoch": 0.8548713501011853,
+      "grad_norm": 1.0108477922791064,
+      "learning_rate": 1.0421209995203529e-06,
+      "loss": 1.5188,
+      "step": 2957
+    },
+    {
+      "epoch": 0.8551604509973981,
+      "grad_norm": 1.0567930978985622,
+      "learning_rate": 1.038046556171154e-06,
+      "loss": 1.5632,
+      "step": 2958
+    },
+    {
+      "epoch": 0.8554495518936108,
+      "grad_norm": 1.0007528310348233,
+      "learning_rate": 1.033979657398797e-06,
+      "loss": 1.5908,
+      "step": 2959
+    },
+    {
+      "epoch": 0.8557386527898236,
+      "grad_norm": 1.0080320698182563,
+      "learning_rate": 1.029920306626986e-06,
+      "loss": 1.4954,
+      "step": 2960
+    },
+    {
+      "epoch": 0.8560277536860365,
+      "grad_norm": 1.0177322815656524,
+      "learning_rate": 1.0258685072730635e-06,
+      "loss": 1.5138,
+      "step": 2961
+    },
+    {
+      "epoch": 0.8563168545822492,
+      "grad_norm": 0.9756576818534307,
+      "learning_rate": 1.0218242627480157e-06,
+      "loss": 1.4857,
+      "step": 2962
+    },
+    {
+      "epoch": 0.856605955478462,
+      "grad_norm": 1.0049167194806465,
+      "learning_rate": 1.0177875764564716e-06,
+      "loss": 1.5385,
+      "step": 2963
+    },
+    {
+      "epoch": 0.8568950563746748,
+      "grad_norm": 1.0016428605302516,
+      "learning_rate": 1.0137584517966981e-06,
+      "loss": 1.5842,
+      "step": 2964
+    },
+    {
+      "epoch": 0.8571841572708876,
+      "grad_norm": 1.0114100420633418,
+      "learning_rate": 1.0097368921605922e-06,
+      "loss": 1.5137,
+      "step": 2965
+    },
+    {
+      "epoch": 0.8574732581671003,
+      "grad_norm": 1.009075973345162,
+      "learning_rate": 1.005722900933681e-06,
+      "loss": 1.5199,
+      "step": 2966
+    },
+    {
+      "epoch": 0.8577623590633131,
+      "grad_norm": 1.0431113492001378,
+      "learning_rate": 1.0017164814951298e-06,
+      "loss": 1.5006,
+      "step": 2967
+    },
+    {
+      "epoch": 0.8580514599595259,
+      "grad_norm": 1.0451129291959385,
+      "learning_rate": 9.97717637217719e-07,
+      "loss": 1.5846,
+      "step": 2968
+    },
+    {
+      "epoch": 0.8583405608557386,
+      "grad_norm": 0.9920805724986859,
+      "learning_rate": 9.937263714678559e-07,
+      "loss": 1.4749,
+      "step": 2969
+    },
+    {
+      "epoch": 0.8586296617519514,
+      "grad_norm": 1.057401670995717,
+      "learning_rate": 9.897426876055716e-07,
+      "loss": 1.534,
+      "step": 2970
+    },
+    {
+      "epoch": 0.8589187626481642,
+      "grad_norm": 1.0252581754356553,
+      "learning_rate": 9.85766588984508e-07,
+      "loss": 1.6094,
+      "step": 2971
+    },
+    {
+      "epoch": 0.859207863544377,
+      "grad_norm": 0.9670916146381178,
+      "learning_rate": 9.817980789519232e-07,
+      "loss": 1.4257,
+      "step": 2972
+    },
+    {
+      "epoch": 0.8594969644405898,
+      "grad_norm": 1.0103202999529086,
+      "learning_rate": 9.778371608486913e-07,
+      "loss": 1.4711,
+      "step": 2973
+    },
+    {
+      "epoch": 0.8597860653368026,
+      "grad_norm": 1.0441653887904694,
+      "learning_rate": 9.738838380092908e-07,
+      "loss": 1.6006,
+      "step": 2974
+    },
+    {
+      "epoch": 0.8600751662330153,
+      "grad_norm": 1.001720118210544,
+      "learning_rate": 9.69938113761808e-07,
+      "loss": 1.4338,
+      "step": 2975
+    },
+    {
+      "epoch": 0.8603642671292281,
+      "grad_norm": 1.026387167554004,
+      "learning_rate": 9.659999914279295e-07,
+      "loss": 1.579,
+      "step": 2976
+    },
+    {
+      "epoch": 0.8606533680254409,
+      "grad_norm": 1.0099731871473028,
+      "learning_rate": 9.620694743229464e-07,
+      "loss": 1.5376,
+      "step": 2977
+    },
+    {
+      "epoch": 0.8609424689216536,
+      "grad_norm": 1.0649451735713305,
+      "learning_rate": 9.581465657557442e-07,
+      "loss": 1.6208,
+      "step": 2978
+    },
+    {
+      "epoch": 0.8612315698178664,
+      "grad_norm": 1.0121584813735875,
+      "learning_rate": 9.542312690288035e-07,
+      "loss": 1.5692,
+      "step": 2979
+    },
+    {
+      "epoch": 0.8615206707140792,
+      "grad_norm": 1.048683957469565,
+      "learning_rate": 9.503235874382011e-07,
+      "loss": 1.4944,
+      "step": 2980
+    },
+    {
+      "epoch": 0.861809771610292,
+      "grad_norm": 1.0011313218750106,
+      "learning_rate": 9.464235242735975e-07,
+      "loss": 1.5226,
+      "step": 2981
+    },
+    {
+      "epoch": 0.8620988725065047,
+      "grad_norm": 1.040780472601663,
+      "learning_rate": 9.425310828182399e-07,
+      "loss": 1.5159,
+      "step": 2982
+    },
+    {
+      "epoch": 0.8623879734027176,
+      "grad_norm": 1.0324734761039005,
+      "learning_rate": 9.386462663489637e-07,
+      "loss": 1.4759,
+      "step": 2983
+    },
+    {
+      "epoch": 0.8626770742989304,
+      "grad_norm": 1.0069828571082553,
+      "learning_rate": 9.34769078136184e-07,
+      "loss": 1.548,
+      "step": 2984
+    },
+    {
+      "epoch": 0.8629661751951431,
+      "grad_norm": 0.9791018048109408,
+      "learning_rate": 9.308995214438909e-07,
+      "loss": 1.4899,
+      "step": 2985
+    },
+    {
+      "epoch": 0.8632552760913559,
+      "grad_norm": 0.9933020363123762,
+      "learning_rate": 9.270375995296499e-07,
+      "loss": 1.4768,
+      "step": 2986
+    },
+    {
+      "epoch": 0.8635443769875687,
+      "grad_norm": 0.9919326167985684,
+      "learning_rate": 9.231833156446058e-07,
+      "loss": 1.5336,
+      "step": 2987
+    },
+    {
+      "epoch": 0.8638334778837814,
+      "grad_norm": 1.0256775881006523,
+      "learning_rate": 9.193366730334662e-07,
+      "loss": 1.548,
+      "step": 2988
+    },
+    {
+      "epoch": 0.8641225787799942,
+      "grad_norm": 1.0127103560500617,
+      "learning_rate": 9.154976749345057e-07,
+      "loss": 1.4632,
+      "step": 2989
+    },
+    {
+      "epoch": 0.864411679676207,
+      "grad_norm": 0.9854730936741144,
+      "learning_rate": 9.11666324579572e-07,
+      "loss": 1.5636,
+      "step": 2990
+    },
+    {
+      "epoch": 0.8647007805724197,
+      "grad_norm": 1.0359171062353214,
+      "learning_rate": 9.078426251940642e-07,
+      "loss": 1.497,
+      "step": 2991
+    },
+    {
+      "epoch": 0.8649898814686325,
+      "grad_norm": 1.0102008113494152,
+      "learning_rate": 9.040265799969461e-07,
+      "loss": 1.5441,
+      "step": 2992
+    },
+    {
+      "epoch": 0.8652789823648454,
+      "grad_norm": 1.0231161729266944,
+      "learning_rate": 9.00218192200738e-07,
+      "loss": 1.5781,
+      "step": 2993
+    },
+    {
+      "epoch": 0.8655680832610582,
+      "grad_norm": 0.9934667925003046,
+      "learning_rate": 8.964174650115098e-07,
+      "loss": 1.5193,
+      "step": 2994
+    },
+    {
+      "epoch": 0.8658571841572709,
+      "grad_norm": 1.0134134202669436,
+      "learning_rate": 8.926244016288898e-07,
+      "loss": 1.5442,
+      "step": 2995
+    },
+    {
+      "epoch": 0.8661462850534837,
+      "grad_norm": 1.033324247572779,
+      "learning_rate": 8.888390052460438e-07,
+      "loss": 1.5517,
+      "step": 2996
+    },
+    {
+      "epoch": 0.8664353859496965,
+      "grad_norm": 1.040390607374968,
+      "learning_rate": 8.85061279049696e-07,
+      "loss": 1.5187,
+      "step": 2997
+    },
+    {
+      "epoch": 0.8667244868459092,
+      "grad_norm": 1.0201272252321216,
+      "learning_rate": 8.812912262201023e-07,
+      "loss": 1.4972,
+      "step": 2998
+    },
+    {
+      "epoch": 0.867013587742122,
+      "grad_norm": 1.0135337648014036,
+      "learning_rate": 8.775288499310641e-07,
+      "loss": 1.429,
+      "step": 2999
+    },
+    {
+      "epoch": 0.8673026886383348,
+      "grad_norm": 1.0426692210502864,
+      "learning_rate": 8.737741533499211e-07,
+      "loss": 1.4643,
+      "step": 3000
+    },
+    {
+      "epoch": 0.8675917895345475,
+      "grad_norm": 1.0486677381377651,
+      "learning_rate": 8.700271396375448e-07,
+      "loss": 1.5506,
+      "step": 3001
+    },
+    {
+      "epoch": 0.8678808904307603,
+      "grad_norm": 1.056068244908767,
+      "learning_rate": 8.662878119483409e-07,
+      "loss": 1.5328,
+      "step": 3002
+    },
+    {
+      "epoch": 0.8681699913269731,
+      "grad_norm": 1.0014841023721637,
+      "learning_rate": 8.625561734302423e-07,
+      "loss": 1.5308,
+      "step": 3003
+    },
+    {
+      "epoch": 0.8684590922231858,
+      "grad_norm": 1.0028373873803342,
+      "learning_rate": 8.588322272247141e-07,
+      "loss": 1.631,
+      "step": 3004
+    },
+    {
+      "epoch": 0.8687481931193987,
+      "grad_norm": 1.0770052968945758,
+      "learning_rate": 8.551159764667394e-07,
+      "loss": 1.501,
+      "step": 3005
+    },
+    {
+      "epoch": 0.8690372940156115,
+      "grad_norm": 1.0023234656137985,
+      "learning_rate": 8.51407424284827e-07,
+      "loss": 1.4009,
+      "step": 3006
+    },
+    {
+      "epoch": 0.8693263949118242,
+      "grad_norm": 1.0427208370761107,
+      "learning_rate": 8.47706573801006e-07,
+      "loss": 1.5685,
+      "step": 3007
+    },
+    {
+      "epoch": 0.869615495808037,
+      "grad_norm": 1.0161689983633457,
+      "learning_rate": 8.440134281308154e-07,
+      "loss": 1.4635,
+      "step": 3008
+    },
+    {
+      "epoch": 0.8699045967042498,
+      "grad_norm": 1.0392839979801731,
+      "learning_rate": 8.40327990383315e-07,
+      "loss": 1.4786,
+      "step": 3009
+    },
+    {
+      "epoch": 0.8701936976004626,
+      "grad_norm": 1.0011785129769581,
+      "learning_rate": 8.366502636610685e-07,
+      "loss": 1.54,
+      "step": 3010
+    },
+    {
+      "epoch": 0.8704827984966753,
+      "grad_norm": 0.9986619095878997,
+      "learning_rate": 8.329802510601559e-07,
+      "loss": 1.4873,
+      "step": 3011
+    },
+    {
+      "epoch": 0.8707718993928881,
+      "grad_norm": 1.0254579441354883,
+      "learning_rate": 8.293179556701592e-07,
+      "loss": 1.5356,
+      "step": 3012
+    },
+    {
+      "epoch": 0.8710610002891009,
+      "grad_norm": 0.9740356688368446,
+      "learning_rate": 8.25663380574161e-07,
+      "loss": 1.5339,
+      "step": 3013
+    },
+    {
+      "epoch": 0.8713501011853136,
+      "grad_norm": 0.994362278080706,
+      "learning_rate": 8.220165288487503e-07,
+      "loss": 1.5192,
+      "step": 3014
+    },
+    {
+      "epoch": 0.8716392020815265,
+      "grad_norm": 1.0102193655739509,
+      "learning_rate": 8.183774035640102e-07,
+      "loss": 1.5601,
+      "step": 3015
+    },
+    {
+      "epoch": 0.8719283029777393,
+      "grad_norm": 1.0505846506708176,
+      "learning_rate": 8.147460077835223e-07,
+      "loss": 1.5351,
+      "step": 3016
+    },
+    {
+      "epoch": 0.872217403873952,
+      "grad_norm": 1.0243342424302568,
+      "learning_rate": 8.111223445643568e-07,
+      "loss": 1.5943,
+      "step": 3017
+    },
+    {
+      "epoch": 0.8725065047701648,
+      "grad_norm": 1.0161448514439857,
+      "learning_rate": 8.075064169570801e-07,
+      "loss": 1.5848,
+      "step": 3018
+    },
+    {
+      "epoch": 0.8727956056663776,
+      "grad_norm": 1.0336685111846609,
+      "learning_rate": 8.038982280057417e-07,
+      "loss": 1.4821,
+      "step": 3019
+    },
+    {
+      "epoch": 0.8730847065625903,
+      "grad_norm": 0.9945635850697373,
+      "learning_rate": 8.002977807478751e-07,
+      "loss": 1.6036,
+      "step": 3020
+    },
+    {
+      "epoch": 0.8733738074588031,
+      "grad_norm": 1.0375065309057254,
+      "learning_rate": 7.967050782145036e-07,
+      "loss": 1.5719,
+      "step": 3021
+    },
+    {
+      "epoch": 0.8736629083550159,
+      "grad_norm": 1.1180053860541013,
+      "learning_rate": 7.931201234301256e-07,
+      "loss": 1.5223,
+      "step": 3022
+    },
+    {
+      "epoch": 0.8739520092512287,
+      "grad_norm": 1.0187188804767295,
+      "learning_rate": 7.895429194127146e-07,
+      "loss": 1.4546,
+      "step": 3023
+    },
+    {
+      "epoch": 0.8742411101474414,
+      "grad_norm": 1.0076735681459965,
+      "learning_rate": 7.859734691737264e-07,
+      "loss": 1.6058,
+      "step": 3024
+    },
+    {
+      "epoch": 0.8745302110436542,
+      "grad_norm": 0.9933275272887948,
+      "learning_rate": 7.824117757180838e-07,
+      "loss": 1.5494,
+      "step": 3025
+    },
+    {
+      "epoch": 0.8748193119398671,
+      "grad_norm": 1.0147578838177835,
+      "learning_rate": 7.788578420441783e-07,
+      "loss": 1.5218,
+      "step": 3026
+    },
+    {
+      "epoch": 0.8751084128360798,
+      "grad_norm": 1.0436264710057785,
+      "learning_rate": 7.753116711438746e-07,
+      "loss": 1.6536,
+      "step": 3027
+    },
+    {
+      "epoch": 0.8753975137322926,
+      "grad_norm": 1.0289344817418282,
+      "learning_rate": 7.717732660025001e-07,
+      "loss": 1.48,
+      "step": 3028
+    },
+    {
+      "epoch": 0.8756866146285054,
+      "grad_norm": 1.0291656030279113,
+      "learning_rate": 7.682426295988433e-07,
+      "loss": 1.5087,
+      "step": 3029
+    },
+    {
+      "epoch": 0.8759757155247181,
+      "grad_norm": 1.0170091041927778,
+      "learning_rate": 7.647197649051508e-07,
+      "loss": 1.4962,
+      "step": 3030
+    },
+    {
+      "epoch": 0.8762648164209309,
+      "grad_norm": 1.0416083434093044,
+      "learning_rate": 7.612046748871327e-07,
+      "loss": 1.6043,
+      "step": 3031
+    },
+    {
+      "epoch": 0.8765539173171437,
+      "grad_norm": 1.0174133840166977,
+      "learning_rate": 7.576973625039485e-07,
+      "loss": 1.4911,
+      "step": 3032
+    },
+    {
+      "epoch": 0.8768430182133564,
+      "grad_norm": 1.0158116035424603,
+      "learning_rate": 7.54197830708211e-07,
+      "loss": 1.6487,
+      "step": 3033
+    },
+    {
+      "epoch": 0.8771321191095692,
+      "grad_norm": 1.0233495183990262,
+      "learning_rate": 7.507060824459867e-07,
+      "loss": 1.6041,
+      "step": 3034
+    },
+    {
+      "epoch": 0.877421220005782,
+      "grad_norm": 1.0557269888118457,
+      "learning_rate": 7.472221206567853e-07,
+      "loss": 1.6047,
+      "step": 3035
+    },
+    {
+      "epoch": 0.8777103209019947,
+      "grad_norm": 0.9892449344649002,
+      "learning_rate": 7.437459482735621e-07,
+      "loss": 1.5638,
+      "step": 3036
+    },
+    {
+      "epoch": 0.8779994217982076,
+      "grad_norm": 1.0277514078661083,
+      "learning_rate": 7.402775682227159e-07,
+      "loss": 1.4572,
+      "step": 3037
+    },
+    {
+      "epoch": 0.8782885226944204,
+      "grad_norm": 1.0182905485977924,
+      "learning_rate": 7.36816983424089e-07,
+      "loss": 1.5404,
+      "step": 3038
+    },
+    {
+      "epoch": 0.8785776235906332,
+      "grad_norm": 1.0487186788636744,
+      "learning_rate": 7.333641967909555e-07,
+      "loss": 1.6087,
+      "step": 3039
+    },
+    {
+      "epoch": 0.8788667244868459,
+      "grad_norm": 1.020993983275783,
+      "learning_rate": 7.299192112300246e-07,
+      "loss": 1.5661,
+      "step": 3040
+    },
+    {
+      "epoch": 0.8791558253830587,
+      "grad_norm": 1.0155829604865614,
+      "learning_rate": 7.264820296414444e-07,
+      "loss": 1.5811,
+      "step": 3041
+    },
+    {
+      "epoch": 0.8794449262792715,
+      "grad_norm": 1.0593485320502665,
+      "learning_rate": 7.230526549187888e-07,
+      "loss": 1.5104,
+      "step": 3042
+    },
+    {
+      "epoch": 0.8797340271754842,
+      "grad_norm": 1.0365811015757016,
+      "learning_rate": 7.196310899490577e-07,
+      "loss": 1.5425,
+      "step": 3043
+    },
+    {
+      "epoch": 0.880023128071697,
+      "grad_norm": 0.9870315240811077,
+      "learning_rate": 7.162173376126825e-07,
+      "loss": 1.5179,
+      "step": 3044
+    },
+    {
+      "epoch": 0.8803122289679098,
+      "grad_norm": 1.0221187969869967,
+      "learning_rate": 7.128114007835119e-07,
+      "loss": 1.5674,
+      "step": 3045
+    },
+    {
+      "epoch": 0.8806013298641225,
+      "grad_norm": 1.0185018252578244,
+      "learning_rate": 7.094132823288146e-07,
+      "loss": 1.4914,
+      "step": 3046
+    },
+    {
+      "epoch": 0.8808904307603354,
+      "grad_norm": 1.0119466260063643,
+      "learning_rate": 7.060229851092859e-07,
+      "loss": 1.5393,
+      "step": 3047
+    },
+    {
+      "epoch": 0.8811795316565482,
+      "grad_norm": 1.0401929115351305,
+      "learning_rate": 7.026405119790258e-07,
+      "loss": 1.5434,
+      "step": 3048
+    },
+    {
+      "epoch": 0.8814686325527609,
+      "grad_norm": 1.0225275159784648,
+      "learning_rate": 6.992658657855566e-07,
+      "loss": 1.559,
+      "step": 3049
+    },
+    {
+      "epoch": 0.8817577334489737,
+      "grad_norm": 1.0394559070472247,
+      "learning_rate": 6.958990493698048e-07,
+      "loss": 1.6306,
+      "step": 3050
+    },
+    {
+      "epoch": 0.8820468343451865,
+      "grad_norm": 1.0231021659099169,
+      "learning_rate": 6.925400655661118e-07,
+      "loss": 1.5568,
+      "step": 3051
+    },
+    {
+      "epoch": 0.8823359352413992,
+      "grad_norm": 0.9998189801045234,
+      "learning_rate": 6.891889172022193e-07,
+      "loss": 1.5143,
+      "step": 3052
+    },
+    {
+      "epoch": 0.882625036137612,
+      "grad_norm": 1.019919212007427,
+      "learning_rate": 6.858456070992737e-07,
+      "loss": 1.5019,
+      "step": 3053
+    },
+    {
+      "epoch": 0.8829141370338248,
+      "grad_norm": 1.0092278153316163,
+      "learning_rate": 6.825101380718291e-07,
+      "loss": 1.6136,
+      "step": 3054
+    },
+    {
+      "epoch": 0.8832032379300376,
+      "grad_norm": 1.0234557257644283,
+      "learning_rate": 6.791825129278296e-07,
+      "loss": 1.3689,
+      "step": 3055
+    },
+    {
+      "epoch": 0.8834923388262503,
+      "grad_norm": 1.0076353041740123,
+      "learning_rate": 6.758627344686219e-07,
+      "loss": 1.5324,
+      "step": 3056
+    },
+    {
+      "epoch": 0.8837814397224631,
+      "grad_norm": 1.0583722221175702,
+      "learning_rate": 6.72550805488944e-07,
+      "loss": 1.4989,
+      "step": 3057
+    },
+    {
+      "epoch": 0.884070540618676,
+      "grad_norm": 0.979772673841499,
+      "learning_rate": 6.692467287769277e-07,
+      "loss": 1.5579,
+      "step": 3058
+    },
+    {
+      "epoch": 0.8843596415148887,
+      "grad_norm": 0.9928827625620551,
+      "learning_rate": 6.65950507114096e-07,
+      "loss": 1.4767,
+      "step": 3059
+    },
+    {
+      "epoch": 0.8846487424111015,
+      "grad_norm": 1.0159691995779891,
+      "learning_rate": 6.626621432753543e-07,
+      "loss": 1.4946,
+      "step": 3060
+    },
+    {
+      "epoch": 0.8849378433073143,
+      "grad_norm": 0.9990014486485255,
+      "learning_rate": 6.593816400289987e-07,
+      "loss": 1.5839,
+      "step": 3061
+    },
+    {
+      "epoch": 0.885226944203527,
+      "grad_norm": 1.0034477632551753,
+      "learning_rate": 6.561090001367021e-07,
+      "loss": 1.4548,
+      "step": 3062
+    },
+    {
+      "epoch": 0.8855160450997398,
+      "grad_norm": 0.9963907632538619,
+      "learning_rate": 6.528442263535217e-07,
+      "loss": 1.5365,
+      "step": 3063
+    },
+    {
+      "epoch": 0.8858051459959526,
+      "grad_norm": 1.0520058826432377,
+      "learning_rate": 6.495873214278903e-07,
+      "loss": 1.5109,
+      "step": 3064
+    },
+    {
+      "epoch": 0.8860942468921653,
+      "grad_norm": 1.2791550019167197,
+      "learning_rate": 6.463382881016189e-07,
+      "loss": 1.5708,
+      "step": 3065
+    },
+    {
+      "epoch": 0.8863833477883781,
+      "grad_norm": 0.9814309537327417,
+      "learning_rate": 6.430971291098897e-07,
+      "loss": 1.5255,
+      "step": 3066
+    },
+    {
+      "epoch": 0.8866724486845909,
+      "grad_norm": 1.0427872167986523,
+      "learning_rate": 6.398638471812546e-07,
+      "loss": 1.5057,
+      "step": 3067
+    },
+    {
+      "epoch": 0.8869615495808038,
+      "grad_norm": 1.0233915440576786,
+      "learning_rate": 6.366384450376406e-07,
+      "loss": 1.5838,
+      "step": 3068
+    },
+    {
+      "epoch": 0.8872506504770165,
+      "grad_norm": 1.0303814079769207,
+      "learning_rate": 6.334209253943324e-07,
+      "loss": 1.5357,
+      "step": 3069
+    },
+    {
+      "epoch": 0.8875397513732293,
+      "grad_norm": 1.0293750873094956,
+      "learning_rate": 6.302112909599878e-07,
+      "loss": 1.5364,
+      "step": 3070
+    },
+    {
+      "epoch": 0.8878288522694421,
+      "grad_norm": 1.0178771064929686,
+      "learning_rate": 6.270095444366165e-07,
+      "loss": 1.5995,
+      "step": 3071
+    },
+    {
+      "epoch": 0.8881179531656548,
+      "grad_norm": 81.94836146793712,
+      "learning_rate": 6.238156885195979e-07,
+      "loss": 1.6275,
+      "step": 3072
+    },
+    {
+      "epoch": 0.8884070540618676,
+      "grad_norm": 1.0275244404695443,
+      "learning_rate": 6.206297258976635e-07,
+      "loss": 1.5359,
+      "step": 3073
+    },
+    {
+      "epoch": 0.8886961549580804,
+      "grad_norm": 1.027509403577981,
+      "learning_rate": 6.174516592528957e-07,
+      "loss": 1.4704,
+      "step": 3074
+    },
+    {
+      "epoch": 0.8889852558542931,
+      "grad_norm": 0.9812770223367149,
+      "learning_rate": 6.142814912607409e-07,
+      "loss": 1.4856,
+      "step": 3075
+    },
+    {
+      "epoch": 0.8892743567505059,
+      "grad_norm": 1.0194895670923627,
+      "learning_rate": 6.111192245899855e-07,
+      "loss": 1.6035,
+      "step": 3076
+    },
+    {
+      "epoch": 0.8895634576467187,
+      "grad_norm": 1.0356456287405906,
+      "learning_rate": 6.079648619027678e-07,
+      "loss": 1.4824,
+      "step": 3077
+    },
+    {
+      "epoch": 0.8898525585429314,
+      "grad_norm": 1.0163918527985112,
+      "learning_rate": 6.048184058545769e-07,
+      "loss": 1.5227,
+      "step": 3078
+    },
+    {
+      "epoch": 0.8901416594391443,
+      "grad_norm": 1.043313312687037,
+      "learning_rate": 6.016798590942374e-07,
+      "loss": 1.588,
+      "step": 3079
+    },
+    {
+      "epoch": 0.8904307603353571,
+      "grad_norm": 1.0028608975714564,
+      "learning_rate": 5.985492242639201e-07,
+      "loss": 1.5557,
+      "step": 3080
+    },
+    {
+      "epoch": 0.8907198612315698,
+      "grad_norm": 1.0170111246669946,
+      "learning_rate": 5.954265039991359e-07,
+      "loss": 1.5129,
+      "step": 3081
+    },
+    {
+      "epoch": 0.8910089621277826,
+      "grad_norm": 1.0016841595414188,
+      "learning_rate": 5.923117009287327e-07,
+      "loss": 1.5165,
+      "step": 3082
+    },
+    {
+      "epoch": 0.8912980630239954,
+      "grad_norm": 1.0421592212792306,
+      "learning_rate": 5.892048176748899e-07,
+      "loss": 1.5348,
+      "step": 3083
+    },
+    {
+      "epoch": 0.8915871639202082,
+      "grad_norm": 1.0229924480656314,
+      "learning_rate": 5.861058568531219e-07,
+      "loss": 1.4904,
+      "step": 3084
+    },
+    {
+      "epoch": 0.8918762648164209,
+      "grad_norm": 1.0143908619267754,
+      "learning_rate": 5.830148210722775e-07,
+      "loss": 1.5656,
+      "step": 3085
+    },
+    {
+      "epoch": 0.8921653657126337,
+      "grad_norm": 1.0335352882891098,
+      "learning_rate": 5.799317129345262e-07,
+      "loss": 1.5734,
+      "step": 3086
+    },
+    {
+      "epoch": 0.8924544666088465,
+      "grad_norm": 1.0027356886623935,
+      "learning_rate": 5.768565350353683e-07,
+      "loss": 1.5222,
+      "step": 3087
+    },
+    {
+      "epoch": 0.8927435675050592,
+      "grad_norm": 1.0269341502298412,
+      "learning_rate": 5.737892899636289e-07,
+      "loss": 1.4756,
+      "step": 3088
+    },
+    {
+      "epoch": 0.893032668401272,
+      "grad_norm": 1.0568186708660454,
+      "learning_rate": 5.707299803014522e-07,
+      "loss": 1.5298,
+      "step": 3089
+    },
+    {
+      "epoch": 0.8933217692974849,
+      "grad_norm": 1.0546267248907253,
+      "learning_rate": 5.676786086243024e-07,
+      "loss": 1.6242,
+      "step": 3090
+    },
+    {
+      "epoch": 0.8936108701936976,
+      "grad_norm": 1.0417628404611001,
+      "learning_rate": 5.646351775009617e-07,
+      "loss": 1.5186,
+      "step": 3091
+    },
+    {
+      "epoch": 0.8938999710899104,
+      "grad_norm": 0.9720826711144169,
+      "learning_rate": 5.615996894935305e-07,
+      "loss": 1.5829,
+      "step": 3092
+    },
+    {
+      "epoch": 0.8941890719861232,
+      "grad_norm": 1.0272185962044402,
+      "learning_rate": 5.585721471574191e-07,
+      "loss": 1.5109,
+      "step": 3093
+    },
+    {
+      "epoch": 0.8944781728823359,
+      "grad_norm": 1.018371999796928,
+      "learning_rate": 5.555525530413452e-07,
+      "loss": 1.6655,
+      "step": 3094
+    },
+    {
+      "epoch": 0.8947672737785487,
+      "grad_norm": 1.0515076205758382,
+      "learning_rate": 5.525409096873446e-07,
+      "loss": 1.6168,
+      "step": 3095
+    },
+    {
+      "epoch": 0.8950563746747615,
+      "grad_norm": 1.0079296493622547,
+      "learning_rate": 5.495372196307513e-07,
+      "loss": 1.5634,
+      "step": 3096
+    },
+    {
+      "epoch": 0.8953454755709743,
+      "grad_norm": 1.0191457641326216,
+      "learning_rate": 5.465414854002071e-07,
+      "loss": 1.5148,
+      "step": 3097
+    },
+    {
+      "epoch": 0.895634576467187,
+      "grad_norm": 1.0321740824296401,
+      "learning_rate": 5.435537095176591e-07,
+      "loss": 1.5031,
+      "step": 3098
+    },
+    {
+      "epoch": 0.8959236773633998,
+      "grad_norm": 1.0156913511716648,
+      "learning_rate": 5.405738944983496e-07,
+      "loss": 1.5902,
+      "step": 3099
+    },
+    {
+      "epoch": 0.8962127782596127,
+      "grad_norm": 0.9689649669364573,
+      "learning_rate": 5.37602042850821e-07,
+      "loss": 1.5058,
+      "step": 3100
+    },
+    {
+      "epoch": 0.8965018791558254,
+      "grad_norm": 1.0586419737719637,
+      "learning_rate": 5.346381570769121e-07,
+      "loss": 1.5402,
+      "step": 3101
+    },
+    {
+      "epoch": 0.8967909800520382,
+      "grad_norm": 1.045239073527291,
+      "learning_rate": 5.31682239671758e-07,
+      "loss": 1.5561,
+      "step": 3102
+    },
+    {
+      "epoch": 0.897080080948251,
+      "grad_norm": 1.0339417491457,
+      "learning_rate": 5.287342931237826e-07,
+      "loss": 1.5931,
+      "step": 3103
+    },
+    {
+      "epoch": 0.8973691818444637,
+      "grad_norm": 1.0177229593897175,
+      "learning_rate": 5.257943199146976e-07,
+      "loss": 1.5036,
+      "step": 3104
+    },
+    {
+      "epoch": 0.8976582827406765,
+      "grad_norm": 0.9801097387848228,
+      "learning_rate": 5.228623225195084e-07,
+      "loss": 1.5057,
+      "step": 3105
+    },
+    {
+      "epoch": 0.8979473836368893,
+      "grad_norm": 1.0086094407287949,
+      "learning_rate": 5.199383034065019e-07,
+      "loss": 1.4998,
+      "step": 3106
+    },
+    {
+      "epoch": 0.898236484533102,
+      "grad_norm": 1.0232802242828754,
+      "learning_rate": 5.17022265037247e-07,
+      "loss": 1.6306,
+      "step": 3107
+    },
+    {
+      "epoch": 0.8985255854293148,
+      "grad_norm": 1.0316740801763362,
+      "learning_rate": 5.141142098665997e-07,
+      "loss": 1.4265,
+      "step": 3108
+    },
+    {
+      "epoch": 0.8988146863255276,
+      "grad_norm": 0.9774319703908794,
+      "learning_rate": 5.112141403426907e-07,
+      "loss": 1.5034,
+      "step": 3109
+    },
+    {
+      "epoch": 0.8991037872217403,
+      "grad_norm": 1.0047307415851177,
+      "learning_rate": 5.083220589069282e-07,
+      "loss": 1.5343,
+      "step": 3110
+    },
+    {
+      "epoch": 0.8993928881179531,
+      "grad_norm": 0.9744515561558069,
+      "learning_rate": 5.054379679939991e-07,
+      "loss": 1.4605,
+      "step": 3111
+    },
+    {
+      "epoch": 0.899681989014166,
+      "grad_norm": 1.056136913076778,
+      "learning_rate": 5.025618700318602e-07,
+      "loss": 1.6816,
+      "step": 3112
+    },
+    {
+      "epoch": 0.8999710899103788,
+      "grad_norm": 1.029780120976592,
+      "learning_rate": 4.996937674417402e-07,
+      "loss": 1.4366,
+      "step": 3113
+    },
+    {
+      "epoch": 0.9002601908065915,
+      "grad_norm": 0.9642331448417096,
+      "learning_rate": 4.968336626381376e-07,
+      "loss": 1.4894,
+      "step": 3114
+    },
+    {
+      "epoch": 0.9005492917028043,
+      "grad_norm": 1.0254975400947042,
+      "learning_rate": 4.939815580288176e-07,
+      "loss": 1.6145,
+      "step": 3115
+    },
+    {
+      "epoch": 0.9008383925990171,
+      "grad_norm": 1.0011290504633594,
+      "learning_rate": 4.911374560148119e-07,
+      "loss": 1.4363,
+      "step": 3116
+    },
+    {
+      "epoch": 0.9011274934952298,
+      "grad_norm": 0.9947102394681394,
+      "learning_rate": 4.883013589904118e-07,
+      "loss": 1.5367,
+      "step": 3117
+    },
+    {
+      "epoch": 0.9014165943914426,
+      "grad_norm": 1.0227700871817909,
+      "learning_rate": 4.854732693431729e-07,
+      "loss": 1.4806,
+      "step": 3118
+    },
+    {
+      "epoch": 0.9017056952876554,
+      "grad_norm": 1.0024988957297487,
+      "learning_rate": 4.826531894539088e-07,
+      "loss": 1.5124,
+      "step": 3119
+    },
+    {
+      "epoch": 0.9019947961838681,
+      "grad_norm": 0.9976266039378101,
+      "learning_rate": 4.798411216966902e-07,
+      "loss": 1.549,
+      "step": 3120
+    },
+    {
+      "epoch": 0.9022838970800809,
+      "grad_norm": 1.0426385446063768,
+      "learning_rate": 4.770370684388403e-07,
+      "loss": 1.5672,
+      "step": 3121
+    },
+    {
+      "epoch": 0.9025729979762938,
+      "grad_norm": 0.9952409871492689,
+      "learning_rate": 4.742410320409396e-07,
+      "loss": 1.5454,
+      "step": 3122
+    },
+    {
+      "epoch": 0.9028620988725065,
+      "grad_norm": 0.9983203023415955,
+      "learning_rate": 4.714530148568197e-07,
+      "loss": 1.5098,
+      "step": 3123
+    },
+    {
+      "epoch": 0.9031511997687193,
+      "grad_norm": 1.0099485894297007,
+      "learning_rate": 4.686730192335553e-07,
+      "loss": 1.4783,
+      "step": 3124
+    },
+    {
+      "epoch": 0.9034403006649321,
+      "grad_norm": 1.0377526937232606,
+      "learning_rate": 4.659010475114756e-07,
+      "loss": 1.5532,
+      "step": 3125
+    },
+    {
+      "epoch": 0.9037294015611448,
+      "grad_norm": 1.0227979289108544,
+      "learning_rate": 4.631371020241504e-07,
+      "loss": 1.6264,
+      "step": 3126
+    },
+    {
+      "epoch": 0.9040185024573576,
+      "grad_norm": 1.0325245622645285,
+      "learning_rate": 4.6038118509839305e-07,
+      "loss": 1.5669,
+      "step": 3127
+    },
+    {
+      "epoch": 0.9043076033535704,
+      "grad_norm": 1.0393278864012687,
+      "learning_rate": 4.5763329905425755e-07,
+      "loss": 1.5164,
+      "step": 3128
+    },
+    {
+      "epoch": 0.9045967042497832,
+      "grad_norm": 1.0433452371789977,
+      "learning_rate": 4.548934462050414e-07,
+      "loss": 1.5929,
+      "step": 3129
+    },
+    {
+      "epoch": 0.9048858051459959,
+      "grad_norm": 1.0382522748445147,
+      "learning_rate": 4.52161628857275e-07,
+      "loss": 1.6578,
+      "step": 3130
+    },
+    {
+      "epoch": 0.9051749060422087,
+      "grad_norm": 1.0157948090960203,
+      "learning_rate": 4.494378493107232e-07,
+      "loss": 1.6117,
+      "step": 3131
+    },
+    {
+      "epoch": 0.9054640069384216,
+      "grad_norm": 0.9674398615360374,
+      "learning_rate": 4.467221098583907e-07,
+      "loss": 1.5118,
+      "step": 3132
+    },
+    {
+      "epoch": 0.9057531078346343,
+      "grad_norm": 1.014576300199338,
+      "learning_rate": 4.440144127865065e-07,
+      "loss": 1.4656,
+      "step": 3133
+    },
+    {
+      "epoch": 0.9060422087308471,
+      "grad_norm": 1.03218638859635,
+      "learning_rate": 4.413147603745349e-07,
+      "loss": 1.5533,
+      "step": 3134
+    },
+    {
+      "epoch": 0.9063313096270599,
+      "grad_norm": 0.9898613547768969,
+      "learning_rate": 4.386231548951614e-07,
+      "loss": 1.5152,
+      "step": 3135
+    },
+    {
+      "epoch": 0.9066204105232726,
+      "grad_norm": 1.0615694508569338,
+      "learning_rate": 4.3593959861430445e-07,
+      "loss": 1.463,
+      "step": 3136
+    },
+    {
+      "epoch": 0.9069095114194854,
+      "grad_norm": 0.9747578749602581,
+      "learning_rate": 4.332640937911015e-07,
+      "loss": 1.5475,
+      "step": 3137
+    },
+    {
+      "epoch": 0.9071986123156982,
+      "grad_norm": 0.9984994398332339,
+      "learning_rate": 4.305966426779118e-07,
+      "loss": 1.5346,
+      "step": 3138
+    },
+    {
+      "epoch": 0.9074877132119109,
+      "grad_norm": 1.0331761726255597,
+      "learning_rate": 4.2793724752031807e-07,
+      "loss": 1.4807,
+      "step": 3139
+    },
+    {
+      "epoch": 0.9077768141081237,
+      "grad_norm": 1.0345168077648523,
+      "learning_rate": 4.252859105571194e-07,
+      "loss": 1.524,
+      "step": 3140
+    },
+    {
+      "epoch": 0.9080659150043365,
+      "grad_norm": 1.031267885852081,
+      "learning_rate": 4.226426340203282e-07,
+      "loss": 1.5737,
+      "step": 3141
+    },
+    {
+      "epoch": 0.9083550159005493,
+      "grad_norm": 0.9995804237225647,
+      "learning_rate": 4.200074201351778e-07,
+      "loss": 1.4657,
+      "step": 3142
+    },
+    {
+      "epoch": 0.908644116796762,
+      "grad_norm": 0.999993417482904,
+      "learning_rate": 4.173802711201069e-07,
+      "loss": 1.5305,
+      "step": 3143
+    },
+    {
+      "epoch": 0.9089332176929749,
+      "grad_norm": 1.0647518932296045,
+      "learning_rate": 4.1476118918676866e-07,
+      "loss": 1.5532,
+      "step": 3144
+    },
+    {
+      "epoch": 0.9092223185891877,
+      "grad_norm": 1.0155370872921254,
+      "learning_rate": 4.1215017654002596e-07,
+      "loss": 1.4907,
+      "step": 3145
+    },
+    {
+      "epoch": 0.9095114194854004,
+      "grad_norm": 1.0405090429906123,
+      "learning_rate": 4.0954723537794614e-07,
+      "loss": 1.5287,
+      "step": 3146
+    },
+    {
+      "epoch": 0.9098005203816132,
+      "grad_norm": 0.9829736778456083,
+      "learning_rate": 4.06952367891803e-07,
+      "loss": 1.516,
+      "step": 3147
+    },
+    {
+      "epoch": 0.910089621277826,
+      "grad_norm": 1.014517818242728,
+      "learning_rate": 4.0436557626607254e-07,
+      "loss": 1.6172,
+      "step": 3148
+    },
+    {
+      "epoch": 0.9103787221740387,
+      "grad_norm": 0.9983651245033365,
+      "learning_rate": 4.017868626784327e-07,
+      "loss": 1.7067,
+      "step": 3149
+    },
+    {
+      "epoch": 0.9106678230702515,
+      "grad_norm": 1.0073630531455102,
+      "learning_rate": 3.9921622929976144e-07,
+      "loss": 1.5645,
+      "step": 3150
+    },
+    {
+      "epoch": 0.9109569239664643,
+      "grad_norm": 1.056616852053204,
+      "learning_rate": 3.9665367829413325e-07,
+      "loss": 1.628,
+      "step": 3151
+    },
+    {
+      "epoch": 0.911246024862677,
+      "grad_norm": 1.0650995773132252,
+      "learning_rate": 3.9409921181882025e-07,
+      "loss": 1.5927,
+      "step": 3152
+    },
+    {
+      "epoch": 0.9115351257588898,
+      "grad_norm": 1.0468257561465013,
+      "learning_rate": 3.9155283202428675e-07,
+      "loss": 1.5418,
+      "step": 3153
+    },
+    {
+      "epoch": 0.9118242266551027,
+      "grad_norm": 1.0235763832803293,
+      "learning_rate": 3.89014541054189e-07,
+      "loss": 1.548,
+      "step": 3154
+    },
+    {
+      "epoch": 0.9121133275513154,
+      "grad_norm": 1.0052368455317549,
+      "learning_rate": 3.864843410453767e-07,
+      "loss": 1.5826,
+      "step": 3155
+    },
+    {
+      "epoch": 0.9124024284475282,
+      "grad_norm": 1.0101367954645848,
+      "learning_rate": 3.83962234127887e-07,
+      "loss": 1.4908,
+      "step": 3156
+    },
+    {
+      "epoch": 0.912691529343741,
+      "grad_norm": 1.0380739698026051,
+      "learning_rate": 3.8144822242494163e-07,
+      "loss": 1.5955,
+      "step": 3157
+    },
+    {
+      "epoch": 0.9129806302399538,
+      "grad_norm": 1.0300644014955462,
+      "learning_rate": 3.7894230805294865e-07,
+      "loss": 1.4588,
+      "step": 3158
+    },
+    {
+      "epoch": 0.9132697311361665,
+      "grad_norm": 1.0317748394896384,
+      "learning_rate": 3.764444931215028e-07,
+      "loss": 1.5531,
+      "step": 3159
+    },
+    {
+      "epoch": 0.9135588320323793,
+      "grad_norm": 1.0471297342363408,
+      "learning_rate": 3.739547797333754e-07,
+      "loss": 1.6016,
+      "step": 3160
+    },
+    {
+      "epoch": 0.9138479329285921,
+      "grad_norm": 1.077595141788026,
+      "learning_rate": 3.7147316998451866e-07,
+      "loss": 1.5762,
+      "step": 3161
+    },
+    {
+      "epoch": 0.9141370338248048,
+      "grad_norm": 1.0164944749854854,
+      "learning_rate": 3.689996659640649e-07,
+      "loss": 1.5407,
+      "step": 3162
+    },
+    {
+      "epoch": 0.9144261347210176,
+      "grad_norm": 1.0174794747973057,
+      "learning_rate": 3.665342697543228e-07,
+      "loss": 1.4777,
+      "step": 3163
+    },
+    {
+      "epoch": 0.9147152356172304,
+      "grad_norm": 1.0048638934179828,
+      "learning_rate": 3.6407698343077e-07,
+      "loss": 1.5585,
+      "step": 3164
+    },
+    {
+      "epoch": 0.9150043365134432,
+      "grad_norm": 1.0144756658834138,
+      "learning_rate": 3.6162780906206394e-07,
+      "loss": 1.4362,
+      "step": 3165
+    },
+    {
+      "epoch": 0.915293437409656,
+      "grad_norm": 1.0026593874227705,
+      "learning_rate": 3.591867487100309e-07,
+      "loss": 1.5051,
+      "step": 3166
+    },
+    {
+      "epoch": 0.9155825383058688,
+      "grad_norm": 0.9877821629294984,
+      "learning_rate": 3.5675380442966476e-07,
+      "loss": 1.4665,
+      "step": 3167
+    },
+    {
+      "epoch": 0.9158716392020815,
+      "grad_norm": 1.0434272199417192,
+      "learning_rate": 3.5432897826912616e-07,
+      "loss": 1.542,
+      "step": 3168
+    },
+    {
+      "epoch": 0.9161607400982943,
+      "grad_norm": 1.0035074134086246,
+      "learning_rate": 3.5191227226974655e-07,
+      "loss": 1.4788,
+      "step": 3169
+    },
+    {
+      "epoch": 0.9164498409945071,
+      "grad_norm": 1.0158621160721815,
+      "learning_rate": 3.495036884660152e-07,
+      "loss": 1.5017,
+      "step": 3170
+    },
+    {
+      "epoch": 0.9167389418907199,
+      "grad_norm": 6.106744166150401,
+      "learning_rate": 3.471032288855869e-07,
+      "loss": 1.4726,
+      "step": 3171
+    },
+    {
+      "epoch": 0.9170280427869326,
+      "grad_norm": 1.0475320751286106,
+      "learning_rate": 3.447108955492784e-07,
+      "loss": 1.5852,
+      "step": 3172
+    },
+    {
+      "epoch": 0.9173171436831454,
+      "grad_norm": 1.0364044224368654,
+      "learning_rate": 3.4232669047106317e-07,
+      "loss": 1.5144,
+      "step": 3173
+    },
+    {
+      "epoch": 0.9176062445793582,
+      "grad_norm": 1.0194607746947502,
+      "learning_rate": 3.399506156580712e-07,
+      "loss": 1.5152,
+      "step": 3174
+    },
+    {
+      "epoch": 0.917895345475571,
+      "grad_norm": 1.0049729539873002,
+      "learning_rate": 3.375826731105925e-07,
+      "loss": 1.5241,
+      "step": 3175
+    },
+    {
+      "epoch": 0.9181844463717838,
+      "grad_norm": 1.0203727095399375,
+      "learning_rate": 3.3522286482206455e-07,
+      "loss": 1.6236,
+      "step": 3176
+    },
+    {
+      "epoch": 0.9184735472679966,
+      "grad_norm": 1.0206162170376636,
+      "learning_rate": 3.32871192779084e-07,
+      "loss": 1.5418,
+      "step": 3177
+    },
+    {
+      "epoch": 0.9187626481642093,
+      "grad_norm": 0.9627068073379181,
+      "learning_rate": 3.305276589613904e-07,
+      "loss": 1.5756,
+      "step": 3178
+    },
+    {
+      "epoch": 0.9190517490604221,
+      "grad_norm": 1.029417283532649,
+      "learning_rate": 3.281922653418801e-07,
+      "loss": 1.512,
+      "step": 3179
+    },
+    {
+      "epoch": 0.9193408499566349,
+      "grad_norm": 1.0014141498279212,
+      "learning_rate": 3.258650138865915e-07,
+      "loss": 1.5904,
+      "step": 3180
+    },
+    {
+      "epoch": 0.9196299508528476,
+      "grad_norm": 1.0188974882797537,
+      "learning_rate": 3.235459065547086e-07,
+      "loss": 1.5253,
+      "step": 3181
+    },
+    {
+      "epoch": 0.9199190517490604,
+      "grad_norm": 0.986834168963485,
+      "learning_rate": 3.2123494529856304e-07,
+      "loss": 1.4656,
+      "step": 3182
+    },
+    {
+      "epoch": 0.9202081526452732,
+      "grad_norm": 1.0374677306866014,
+      "learning_rate": 3.1893213206362405e-07,
+      "loss": 1.5336,
+      "step": 3183
+    },
+    {
+      "epoch": 0.9204972535414859,
+      "grad_norm": 1.0263858307455793,
+      "learning_rate": 3.1663746878850544e-07,
+      "loss": 1.569,
+      "step": 3184
+    },
+    {
+      "epoch": 0.9207863544376987,
+      "grad_norm": 1.012502641013397,
+      "learning_rate": 3.1435095740495527e-07,
+      "loss": 1.5038,
+      "step": 3185
+    },
+    {
+      "epoch": 0.9210754553339116,
+      "grad_norm": 0.9804576823402295,
+      "learning_rate": 3.1207259983786706e-07,
+      "loss": 1.5805,
+      "step": 3186
+    },
+    {
+      "epoch": 0.9213645562301244,
+      "grad_norm": 1.0253695348954028,
+      "learning_rate": 3.0980239800526e-07,
+      "loss": 1.572,
+      "step": 3187
+    },
+    {
+      "epoch": 0.9216536571263371,
+      "grad_norm": 1.0350805407673427,
+      "learning_rate": 3.0754035381829415e-07,
+      "loss": 1.4791,
+      "step": 3188
+    },
+    {
+      "epoch": 0.9219427580225499,
+      "grad_norm": 1.0082060993992685,
+      "learning_rate": 3.0528646918126294e-07,
+      "loss": 1.4851,
+      "step": 3189
+    },
+    {
+      "epoch": 0.9222318589187627,
+      "grad_norm": 1.0534255947808198,
+      "learning_rate": 3.0304074599158626e-07,
+      "loss": 1.4853,
+      "step": 3190
+    },
+    {
+      "epoch": 0.9225209598149754,
+      "grad_norm": 1.0380391378093667,
+      "learning_rate": 3.008031861398142e-07,
+      "loss": 1.6083,
+      "step": 3191
+    },
+    {
+      "epoch": 0.9228100607111882,
+      "grad_norm": 0.994413933952922,
+      "learning_rate": 2.985737915096254e-07,
+      "loss": 1.4872,
+      "step": 3192
+    },
+    {
+      "epoch": 0.923099161607401,
+      "grad_norm": 1.0212779048311726,
+      "learning_rate": 2.9635256397782755e-07,
+      "loss": 1.5423,
+      "step": 3193
+    },
+    {
+      "epoch": 0.9233882625036137,
+      "grad_norm": 1.0146747891343744,
+      "learning_rate": 2.941395054143492e-07,
+      "loss": 1.5526,
+      "step": 3194
+    },
+    {
+      "epoch": 0.9236773633998265,
+      "grad_norm": 1.0016189583283832,
+      "learning_rate": 2.9193461768224016e-07,
+      "loss": 1.4722,
+      "step": 3195
+    },
+    {
+      "epoch": 0.9239664642960393,
+      "grad_norm": 1.001052423670681,
+      "learning_rate": 2.8973790263767897e-07,
+      "loss": 1.5394,
+      "step": 3196
+    },
+    {
+      "epoch": 0.924255565192252,
+      "grad_norm": 1.0266914850546907,
+      "learning_rate": 2.875493621299552e-07,
+      "loss": 1.4856,
+      "step": 3197
+    },
+    {
+      "epoch": 0.9245446660884649,
+      "grad_norm": 0.9938666348846271,
+      "learning_rate": 2.85368998001484e-07,
+      "loss": 1.4743,
+      "step": 3198
+    },
+    {
+      "epoch": 0.9248337669846777,
+      "grad_norm": 1.0050534072272532,
+      "learning_rate": 2.831968120877926e-07,
+      "loss": 1.4872,
+      "step": 3199
+    },
+    {
+      "epoch": 0.9251228678808904,
+      "grad_norm": 1.0577319188914276,
+      "learning_rate": 2.8103280621752604e-07,
+      "loss": 1.4582,
+      "step": 3200
+    },
+    {
+      "epoch": 0.9254119687771032,
+      "grad_norm": 0.9970429750752302,
+      "learning_rate": 2.788769822124426e-07,
+      "loss": 1.5551,
+      "step": 3201
+    },
+    {
+      "epoch": 0.925701069673316,
+      "grad_norm": 1.0135509155719788,
+      "learning_rate": 2.7672934188741065e-07,
+      "loss": 1.4848,
+      "step": 3202
+    },
+    {
+      "epoch": 0.9259901705695288,
+      "grad_norm": 1.0333821776249823,
+      "learning_rate": 2.745898870504116e-07,
+      "loss": 1.4389,
+      "step": 3203
+    },
+    {
+      "epoch": 0.9262792714657415,
+      "grad_norm": 0.9881936676805886,
+      "learning_rate": 2.724586195025358e-07,
+      "loss": 1.4996,
+      "step": 3204
+    },
+    {
+      "epoch": 0.9265683723619543,
+      "grad_norm": 1.0134183645535257,
+      "learning_rate": 2.703355410379793e-07,
+      "loss": 1.5977,
+      "step": 3205
+    },
+    {
+      "epoch": 0.9268574732581671,
+      "grad_norm": 1.0117705663453405,
+      "learning_rate": 2.682206534440457e-07,
+      "loss": 1.5159,
+      "step": 3206
+    },
+    {
+      "epoch": 0.9271465741543798,
+      "grad_norm": 1.0293521242057286,
+      "learning_rate": 2.6611395850114297e-07,
+      "loss": 1.5445,
+      "step": 3207
+    },
+    {
+      "epoch": 0.9274356750505927,
+      "grad_norm": 1.0276163364115385,
+      "learning_rate": 2.640154579827814e-07,
+      "loss": 1.552,
+      "step": 3208
+    },
+    {
+      "epoch": 0.9277247759468055,
+      "grad_norm": 1.0469224839215487,
+      "learning_rate": 2.6192515365557334e-07,
+      "loss": 1.5817,
+      "step": 3209
+    },
+    {
+      "epoch": 0.9280138768430182,
+      "grad_norm": 1.0661665811998111,
+      "learning_rate": 2.598430472792335e-07,
+      "loss": 1.4776,
+      "step": 3210
+    },
+    {
+      "epoch": 0.928302977739231,
+      "grad_norm": 1.0143817643107766,
+      "learning_rate": 2.577691406065708e-07,
+      "loss": 1.5386,
+      "step": 3211
+    },
+    {
+      "epoch": 0.9285920786354438,
+      "grad_norm": 1.012406011253722,
+      "learning_rate": 2.5570343538349307e-07,
+      "loss": 1.5859,
+      "step": 3212
+    },
+    {
+      "epoch": 0.9288811795316565,
+      "grad_norm": 0.9858617428045467,
+      "learning_rate": 2.5364593334900713e-07,
+      "loss": 1.5311,
+      "step": 3213
+    },
+    {
+      "epoch": 0.9291702804278693,
+      "grad_norm": 1.0087884469100739,
+      "learning_rate": 2.515966362352085e-07,
+      "loss": 1.4826,
+      "step": 3214
+    },
+    {
+      "epoch": 0.9294593813240821,
+      "grad_norm": 0.9825120615675764,
+      "learning_rate": 2.4955554576728713e-07,
+      "loss": 1.4567,
+      "step": 3215
+    },
+    {
+      "epoch": 0.9297484822202949,
+      "grad_norm": 1.014021558885241,
+      "learning_rate": 2.475226636635275e-07,
+      "loss": 1.4562,
+      "step": 3216
+    },
+    {
+      "epoch": 0.9300375831165076,
+      "grad_norm": 1.0109460186262644,
+      "learning_rate": 2.4549799163530064e-07,
+      "loss": 1.4603,
+      "step": 3217
+    },
+    {
+      "epoch": 0.9303266840127205,
+      "grad_norm": 1.014225625599271,
+      "learning_rate": 2.434815313870664e-07,
+      "loss": 1.6465,
+      "step": 3218
+    },
+    {
+      "epoch": 0.9306157849089333,
+      "grad_norm": 1.082828135603616,
+      "learning_rate": 2.4147328461637143e-07,
+      "loss": 1.4808,
+      "step": 3219
+    },
+    {
+      "epoch": 0.930904885805146,
+      "grad_norm": 1.0365379626646767,
+      "learning_rate": 2.3947325301385106e-07,
+      "loss": 1.5566,
+      "step": 3220
+    },
+    {
+      "epoch": 0.9311939867013588,
+      "grad_norm": 1.0621986761304876,
+      "learning_rate": 2.3748143826321957e-07,
+      "loss": 1.5612,
+      "step": 3221
+    },
+    {
+      "epoch": 0.9314830875975716,
+      "grad_norm": 1.0306543969837023,
+      "learning_rate": 2.3549784204127678e-07,
+      "loss": 1.4962,
+      "step": 3222
+    },
+    {
+      "epoch": 0.9317721884937843,
+      "grad_norm": 0.9879899151176428,
+      "learning_rate": 2.3352246601790584e-07,
+      "loss": 1.561,
+      "step": 3223
+    },
+    {
+      "epoch": 0.9320612893899971,
+      "grad_norm": 1.0452316245996112,
+      "learning_rate": 2.315553118560654e-07,
+      "loss": 1.5566,
+      "step": 3224
+    },
+    {
+      "epoch": 0.9323503902862099,
+      "grad_norm": 1.0031584901010149,
+      "learning_rate": 2.2959638121179296e-07,
+      "loss": 1.5205,
+      "step": 3225
+    },
+    {
+      "epoch": 0.9326394911824226,
+      "grad_norm": 1.0257762099768053,
+      "learning_rate": 2.2764567573420936e-07,
+      "loss": 1.5621,
+      "step": 3226
+    },
+    {
+      "epoch": 0.9329285920786354,
+      "grad_norm": 1.0220449774098996,
+      "learning_rate": 2.2570319706550214e-07,
+      "loss": 1.5706,
+      "step": 3227
+    },
+    {
+      "epoch": 0.9332176929748482,
+      "grad_norm": 0.9977824818631326,
+      "learning_rate": 2.2376894684093874e-07,
+      "loss": 1.4667,
+      "step": 3228
+    },
+    {
+      "epoch": 0.933506793871061,
+      "grad_norm": 0.9979053800931573,
+      "learning_rate": 2.2184292668885998e-07,
+      "loss": 1.5589,
+      "step": 3229
+    },
+    {
+      "epoch": 0.9337958947672738,
+      "grad_norm": 1.0516888873310062,
+      "learning_rate": 2.1992513823067442e-07,
+      "loss": 1.5789,
+      "step": 3230
+    },
+    {
+      "epoch": 0.9340849956634866,
+      "grad_norm": 1.006218442483333,
+      "learning_rate": 2.18015583080865e-07,
+      "loss": 1.4663,
+      "step": 3231
+    },
+    {
+      "epoch": 0.9343740965596994,
+      "grad_norm": 1.00481133419246,
+      "learning_rate": 2.1611426284697924e-07,
+      "loss": 1.4849,
+      "step": 3232
+    },
+    {
+      "epoch": 0.9346631974559121,
+      "grad_norm": 1.0053677167438884,
+      "learning_rate": 2.1422117912963557e-07,
+      "loss": 1.4581,
+      "step": 3233
+    },
+    {
+      "epoch": 0.9349522983521249,
+      "grad_norm": 1.0062136497519727,
+      "learning_rate": 2.1233633352251815e-07,
+      "loss": 1.4862,
+      "step": 3234
+    },
+    {
+      "epoch": 0.9352413992483377,
+      "grad_norm": 1.006997929940114,
+      "learning_rate": 2.104597276123721e-07,
+      "loss": 1.4982,
+      "step": 3235
+    },
+    {
+      "epoch": 0.9355305001445504,
+      "grad_norm": 1.0436371193070748,
+      "learning_rate": 2.085913629790126e-07,
+      "loss": 1.5932,
+      "step": 3236
+    },
+    {
+      "epoch": 0.9358196010407632,
+      "grad_norm": 1.0060281061950849,
+      "learning_rate": 2.067312411953104e-07,
+      "loss": 1.6462,
+      "step": 3237
+    },
+    {
+      "epoch": 0.936108701936976,
+      "grad_norm": 1.0569872526665947,
+      "learning_rate": 2.0487936382720175e-07,
+      "loss": 1.5393,
+      "step": 3238
+    },
+    {
+      "epoch": 0.9363978028331887,
+      "grad_norm": 1.0027944102533757,
+      "learning_rate": 2.0303573243367957e-07,
+      "loss": 1.4506,
+      "step": 3239
+    },
+    {
+      "epoch": 0.9366869037294016,
+      "grad_norm": 1.0091610485091689,
+      "learning_rate": 2.0120034856679572e-07,
+      "loss": 1.5025,
+      "step": 3240
+    },
+    {
+      "epoch": 0.9369760046256144,
+      "grad_norm": 1.0371806788104074,
+      "learning_rate": 1.9937321377166086e-07,
+      "loss": 1.4954,
+      "step": 3241
+    },
+    {
+      "epoch": 0.9372651055218271,
+      "grad_norm": 0.9903007349878045,
+      "learning_rate": 1.9755432958643794e-07,
+      "loss": 1.5573,
+      "step": 3242
+    },
+    {
+      "epoch": 0.9375542064180399,
+      "grad_norm": 1.0014412197521625,
+      "learning_rate": 1.9574369754234656e-07,
+      "loss": 1.5365,
+      "step": 3243
+    },
+    {
+      "epoch": 0.9378433073142527,
+      "grad_norm": 1.0078807596972117,
+      "learning_rate": 1.9394131916365743e-07,
+      "loss": 1.5472,
+      "step": 3244
+    },
+    {
+      "epoch": 0.9381324082104655,
+      "grad_norm": 1.0020389612944207,
+      "learning_rate": 1.921471959676957e-07,
+      "loss": 1.4626,
+      "step": 3245
+    },
+    {
+      "epoch": 0.9384215091066782,
+      "grad_norm": 0.9936340912080557,
+      "learning_rate": 1.903613294648343e-07,
+      "loss": 1.5063,
+      "step": 3246
+    },
+    {
+      "epoch": 0.938710610002891,
+      "grad_norm": 1.0034057663539506,
+      "learning_rate": 1.885837211584962e-07,
+      "loss": 1.4914,
+      "step": 3247
+    },
+    {
+      "epoch": 0.9389997108991038,
+      "grad_norm": 1.0229548285376504,
+      "learning_rate": 1.8681437254515212e-07,
+      "loss": 1.4681,
+      "step": 3248
+    },
+    {
+      "epoch": 0.9392888117953165,
+      "grad_norm": 0.9875602739945066,
+      "learning_rate": 1.8505328511431942e-07,
+      "loss": 1.4686,
+      "step": 3249
+    },
+    {
+      "epoch": 0.9395779126915293,
+      "grad_norm": 1.1053198486015643,
+      "learning_rate": 1.8330046034856108e-07,
+      "loss": 1.5338,
+      "step": 3250
+    },
+    {
+      "epoch": 0.9398670135877422,
+      "grad_norm": 1.0141303696689221,
+      "learning_rate": 1.8155589972348453e-07,
+      "loss": 1.5528,
+      "step": 3251
+    },
+    {
+      "epoch": 0.9401561144839549,
+      "grad_norm": 1.0258259602127857,
+      "learning_rate": 1.798196047077394e-07,
+      "loss": 1.5201,
+      "step": 3252
+    },
+    {
+      "epoch": 0.9404452153801677,
+      "grad_norm": 1.026016703100879,
+      "learning_rate": 1.7809157676301758e-07,
+      "loss": 1.6363,
+      "step": 3253
+    },
+    {
+      "epoch": 0.9407343162763805,
+      "grad_norm": 1.0162339712697583,
+      "learning_rate": 1.7637181734405206e-07,
+      "loss": 1.4253,
+      "step": 3254
+    },
+    {
+      "epoch": 0.9410234171725932,
+      "grad_norm": 1.0008171616014512,
+      "learning_rate": 1.7466032789861143e-07,
+      "loss": 1.3732,
+      "step": 3255
+    },
+    {
+      "epoch": 0.941312518068806,
+      "grad_norm": 1.0314320698319472,
+      "learning_rate": 1.7295710986750425e-07,
+      "loss": 1.4587,
+      "step": 3256
+    },
+    {
+      "epoch": 0.9416016189650188,
+      "grad_norm": 1.0124332489916132,
+      "learning_rate": 1.7126216468458023e-07,
+      "loss": 1.504,
+      "step": 3257
+    },
+    {
+      "epoch": 0.9418907198612315,
+      "grad_norm": 1.0276783812147685,
+      "learning_rate": 1.695754937767169e-07,
+      "loss": 1.5765,
+      "step": 3258
+    },
+    {
+      "epoch": 0.9421798207574443,
+      "grad_norm": 1.0326954140320108,
+      "learning_rate": 1.6789709856383174e-07,
+      "loss": 1.5939,
+      "step": 3259
+    },
+    {
+      "epoch": 0.9424689216536571,
+      "grad_norm": 1.0626201936851463,
+      "learning_rate": 1.6622698045887343e-07,
+      "loss": 1.4838,
+      "step": 3260
+    },
+    {
+      "epoch": 0.94275802254987,
+      "grad_norm": 0.998040950485079,
+      "learning_rate": 1.6456514086782284e-07,
+      "loss": 1.5902,
+      "step": 3261
+    },
+    {
+      "epoch": 0.9430471234460827,
+      "grad_norm": 1.0512121078398298,
+      "learning_rate": 1.6291158118969087e-07,
+      "loss": 1.5589,
+      "step": 3262
+    },
+    {
+      "epoch": 0.9433362243422955,
+      "grad_norm": 1.0132859553736602,
+      "learning_rate": 1.6126630281651845e-07,
+      "loss": 1.5436,
+      "step": 3263
+    },
+    {
+      "epoch": 0.9436253252385083,
+      "grad_norm": 0.9821860739567699,
+      "learning_rate": 1.5962930713337765e-07,
+      "loss": 1.5474,
+      "step": 3264
+    },
+    {
+      "epoch": 0.943914426134721,
+      "grad_norm": 1.0099628768461213,
+      "learning_rate": 1.5800059551836278e-07,
+      "loss": 1.5028,
+      "step": 3265
+    },
+    {
+      "epoch": 0.9442035270309338,
+      "grad_norm": 1.011571076311801,
+      "learning_rate": 1.5638016934259593e-07,
+      "loss": 1.487,
+      "step": 3266
+    },
+    {
+      "epoch": 0.9444926279271466,
+      "grad_norm": 1.012587156035977,
+      "learning_rate": 1.5476802997022812e-07,
+      "loss": 1.5445,
+      "step": 3267
+    },
+    {
+      "epoch": 0.9447817288233593,
+      "grad_norm": 0.9844655673085515,
+      "learning_rate": 1.5316417875842927e-07,
+      "loss": 1.5257,
+      "step": 3268
+    },
+    {
+      "epoch": 0.9450708297195721,
+      "grad_norm": 1.025157890409344,
+      "learning_rate": 1.5156861705739377e-07,
+      "loss": 1.6002,
+      "step": 3269
+    },
+    {
+      "epoch": 0.9453599306157849,
+      "grad_norm": 1.021352354981231,
+      "learning_rate": 1.4998134621033722e-07,
+      "loss": 1.5915,
+      "step": 3270
+    },
+    {
+      "epoch": 0.9456490315119976,
+      "grad_norm": 1.014504601352518,
+      "learning_rate": 1.4840236755349735e-07,
+      "loss": 1.4617,
+      "step": 3271
+    },
+    {
+      "epoch": 0.9459381324082105,
+      "grad_norm": 1.0435403417058664,
+      "learning_rate": 1.4683168241612644e-07,
+      "loss": 1.456,
+      "step": 3272
+    },
+    {
+      "epoch": 0.9462272333044233,
+      "grad_norm": 1.0197424555632464,
+      "learning_rate": 1.4526929212050124e-07,
+      "loss": 1.4182,
+      "step": 3273
+    },
+    {
+      "epoch": 0.946516334200636,
+      "grad_norm": 1.0117732796278047,
+      "learning_rate": 1.4371519798191181e-07,
+      "loss": 1.6158,
+      "step": 3274
+    },
+    {
+      "epoch": 0.9468054350968488,
+      "grad_norm": 1.0032159315000166,
+      "learning_rate": 1.4216940130866387e-07,
+      "loss": 1.5338,
+      "step": 3275
+    },
+    {
+      "epoch": 0.9470945359930616,
+      "grad_norm": 1.0101352744617476,
+      "learning_rate": 1.4063190340207866e-07,
+      "loss": 1.4652,
+      "step": 3276
+    },
+    {
+      "epoch": 0.9473836368892744,
+      "grad_norm": 1.0203216460653535,
+      "learning_rate": 1.391027055564931e-07,
+      "loss": 1.5288,
+      "step": 3277
+    },
+    {
+      "epoch": 0.9476727377854871,
+      "grad_norm": 0.9674860421813775,
+      "learning_rate": 1.3758180905925288e-07,
+      "loss": 1.4711,
+      "step": 3278
+    },
+    {
+      "epoch": 0.9479618386816999,
+      "grad_norm": 1.070001070344823,
+      "learning_rate": 1.3606921519071837e-07,
+      "loss": 1.4916,
+      "step": 3279
+    },
+    {
+      "epoch": 0.9482509395779127,
+      "grad_norm": 1.0061620313579656,
+      "learning_rate": 1.3456492522425868e-07,
+      "loss": 1.5785,
+      "step": 3280
+    },
+    {
+      "epoch": 0.9485400404741254,
+      "grad_norm": 1.0176323455440603,
+      "learning_rate": 1.3306894042625417e-07,
+      "loss": 1.5273,
+      "step": 3281
+    },
+    {
+      "epoch": 0.9488291413703382,
+      "grad_norm": 1.0159146581060772,
+      "learning_rate": 1.3158126205609078e-07,
+      "loss": 1.5557,
+      "step": 3282
+    },
+    {
+      "epoch": 0.9491182422665511,
+      "grad_norm": 1.0127607107980785,
+      "learning_rate": 1.3010189136616447e-07,
+      "loss": 1.4973,
+      "step": 3283
+    },
+    {
+      "epoch": 0.9494073431627638,
+      "grad_norm": 0.9919114917013327,
+      "learning_rate": 1.2863082960187567e-07,
+      "loss": 1.5475,
+      "step": 3284
+    },
+    {
+      "epoch": 0.9496964440589766,
+      "grad_norm": 1.0194658488829158,
+      "learning_rate": 1.2716807800163155e-07,
+      "loss": 1.5419,
+      "step": 3285
+    },
+    {
+      "epoch": 0.9499855449551894,
+      "grad_norm": 0.9911288872635503,
+      "learning_rate": 1.2571363779684044e-07,
+      "loss": 1.4445,
+      "step": 3286
+    },
+    {
+      "epoch": 0.9502746458514021,
+      "grad_norm": 1.0605549733160986,
+      "learning_rate": 1.242675102119184e-07,
+      "loss": 1.5311,
+      "step": 3287
+    },
+    {
+      "epoch": 0.9505637467476149,
+      "grad_norm": 1.0117525940915424,
+      "learning_rate": 1.2282969646427832e-07,
+      "loss": 1.5498,
+      "step": 3288
+    },
+    {
+      "epoch": 0.9508528476438277,
+      "grad_norm": 0.9945210922690929,
+      "learning_rate": 1.2140019776433753e-07,
+      "loss": 1.5773,
+      "step": 3289
+    },
+    {
+      "epoch": 0.9511419485400405,
+      "grad_norm": 0.9928757900621521,
+      "learning_rate": 1.1997901531551447e-07,
+      "loss": 1.4454,
+      "step": 3290
+    },
+    {
+      "epoch": 0.9514310494362532,
+      "grad_norm": 1.0477259096934022,
+      "learning_rate": 1.1856615031422214e-07,
+      "loss": 1.5514,
+      "step": 3291
+    },
+    {
+      "epoch": 0.951720150332466,
+      "grad_norm": 1.0360840626077918,
+      "learning_rate": 1.1716160394987463e-07,
+      "loss": 1.6791,
+      "step": 3292
+    },
+    {
+      "epoch": 0.9520092512286789,
+      "grad_norm": 1.0101276454140806,
+      "learning_rate": 1.157653774048828e-07,
+      "loss": 1.566,
+      "step": 3293
+    },
+    {
+      "epoch": 0.9522983521248916,
+      "grad_norm": 1.0039087342415245,
+      "learning_rate": 1.1437747185465309e-07,
+      "loss": 1.6799,
+      "step": 3294
+    },
+    {
+      "epoch": 0.9525874530211044,
+      "grad_norm": 1.011524565273543,
+      "learning_rate": 1.1299788846758753e-07,
+      "loss": 1.507,
+      "step": 3295
+    },
+    {
+      "epoch": 0.9528765539173172,
+      "grad_norm": 0.9923653770022051,
+      "learning_rate": 1.1162662840508043e-07,
+      "loss": 1.5341,
+      "step": 3296
+    },
+    {
+      "epoch": 0.9531656548135299,
+      "grad_norm": 1.0072229050332544,
+      "learning_rate": 1.1026369282152282e-07,
+      "loss": 1.5225,
+      "step": 3297
+    },
+    {
+      "epoch": 0.9534547557097427,
+      "grad_norm": 1.0177164014707178,
+      "learning_rate": 1.0890908286429247e-07,
+      "loss": 1.4578,
+      "step": 3298
+    },
+    {
+      "epoch": 0.9537438566059555,
+      "grad_norm": 0.9868266730787336,
+      "learning_rate": 1.075627996737627e-07,
+      "loss": 1.4817,
+      "step": 3299
+    },
+    {
+      "epoch": 0.9540329575021682,
+      "grad_norm": 1.033801100640919,
+      "learning_rate": 1.0622484438329472e-07,
+      "loss": 1.3963,
+      "step": 3300
+    },
+    {
+      "epoch": 0.954322058398381,
+      "grad_norm": 1.0079909597786436,
+      "learning_rate": 1.0489521811924086e-07,
+      "loss": 1.4781,
+      "step": 3301
+    },
+    {
+      "epoch": 0.9546111592945938,
+      "grad_norm": 1.0692516544328348,
+      "learning_rate": 1.0357392200094019e-07,
+      "loss": 1.4502,
+      "step": 3302
+    },
+    {
+      "epoch": 0.9549002601908065,
+      "grad_norm": 0.9915491055473519,
+      "learning_rate": 1.0226095714071738e-07,
+      "loss": 1.5173,
+      "step": 3303
+    },
+    {
+      "epoch": 0.9551893610870194,
+      "grad_norm": 1.029981119133024,
+      "learning_rate": 1.0095632464388827e-07,
+      "loss": 1.5582,
+      "step": 3304
+    },
+    {
+      "epoch": 0.9554784619832322,
+      "grad_norm": 1.0062259571029208,
+      "learning_rate": 9.966002560874988e-08,
+      "loss": 1.5162,
+      "step": 3305
+    },
+    {
+      "epoch": 0.955767562879445,
+      "grad_norm": 1.025135218495841,
+      "learning_rate": 9.837206112658592e-08,
+      "loss": 1.5974,
+      "step": 3306
+    },
+    {
+      "epoch": 0.9560566637756577,
+      "grad_norm": 1.005939789694238,
+      "learning_rate": 9.709243228166244e-08,
+      "loss": 1.5303,
+      "step": 3307
+    },
+    {
+      "epoch": 0.9563457646718705,
+      "grad_norm": 1.0421029347289905,
+      "learning_rate": 9.58211401512299e-08,
+      "loss": 1.5335,
+      "step": 3308
+    },
+    {
+      "epoch": 0.9566348655680833,
+      "grad_norm": 0.9891544166411924,
+      "learning_rate": 9.455818580551779e-08,
+      "loss": 1.4834,
+      "step": 3309
+    },
+    {
+      "epoch": 0.956923966464296,
+      "grad_norm": 1.0125879428213824,
+      "learning_rate": 9.330357030773784e-08,
+      "loss": 1.5184,
+      "step": 3310
+    },
+    {
+      "epoch": 0.9572130673605088,
+      "grad_norm": 1.0150626356526298,
+      "learning_rate": 9.20572947140841e-08,
+      "loss": 1.5464,
+      "step": 3311
+    },
+    {
+      "epoch": 0.9575021682567216,
+      "grad_norm": 1.0069840786517743,
+      "learning_rate": 9.081936007372727e-08,
+      "loss": 1.5111,
+      "step": 3312
+    },
+    {
+      "epoch": 0.9577912691529343,
+      "grad_norm": 1.4642838216971452,
+      "learning_rate": 8.958976742881375e-08,
+      "loss": 1.4559,
+      "step": 3313
+    },
+    {
+      "epoch": 0.9580803700491471,
+      "grad_norm": 1.0108339156484936,
+      "learning_rate": 8.836851781447332e-08,
+      "loss": 1.5435,
+      "step": 3314
+    },
+    {
+      "epoch": 0.95836947094536,
+      "grad_norm": 1.053702911203741,
+      "learning_rate": 8.715561225880687e-08,
+      "loss": 1.52,
+      "step": 3315
+    },
+    {
+      "epoch": 0.9586585718415727,
+      "grad_norm": 1.020766039826636,
+      "learning_rate": 8.595105178289543e-08,
+      "loss": 1.564,
+      "step": 3316
+    },
+    {
+      "epoch": 0.9589476727377855,
+      "grad_norm": 0.9907921801105501,
+      "learning_rate": 8.475483740079005e-08,
+      "loss": 1.5343,
+      "step": 3317
+    },
+    {
+      "epoch": 0.9592367736339983,
+      "grad_norm": 1.0017139474397532,
+      "learning_rate": 8.356697011951853e-08,
+      "loss": 1.4641,
+      "step": 3318
+    },
+    {
+      "epoch": 0.9595258745302111,
+      "grad_norm": 1.014063969077958,
+      "learning_rate": 8.238745093908207e-08,
+      "loss": 1.4645,
+      "step": 3319
+    },
+    {
+      "epoch": 0.9598149754264238,
+      "grad_norm": 1.0267612117487268,
+      "learning_rate": 8.121628085245192e-08,
+      "loss": 1.5375,
+      "step": 3320
+    },
+    {
+      "epoch": 0.9601040763226366,
+      "grad_norm": 1.0172232969605,
+      "learning_rate": 8.005346084557386e-08,
+      "loss": 1.5753,
+      "step": 3321
+    },
+    {
+      "epoch": 0.9603931772188494,
+      "grad_norm": 0.9976529360795198,
+      "learning_rate": 7.88989918973604e-08,
+      "loss": 1.4794,
+      "step": 3322
+    },
+    {
+      "epoch": 0.9606822781150621,
+      "grad_norm": 1.0021346951051398,
+      "learning_rate": 7.77528749796952e-08,
+      "loss": 1.6439,
+      "step": 3323
+    },
+    {
+      "epoch": 0.9609713790112749,
+      "grad_norm": 1.0072810639911183,
+      "learning_rate": 7.661511105743313e-08,
+      "loss": 1.5601,
+      "step": 3324
+    },
+    {
+      "epoch": 0.9612604799074878,
+      "grad_norm": 1.0090740056135148,
+      "learning_rate": 7.54857010883947e-08,
+      "loss": 1.5753,
+      "step": 3325
+    },
+    {
+      "epoch": 0.9615495808037005,
+      "grad_norm": 1.0201421923914553,
+      "learning_rate": 7.436464602336602e-08,
+      "loss": 1.617,
+      "step": 3326
+    },
+    {
+      "epoch": 0.9618386816999133,
+      "grad_norm": 1.0226457586468796,
+      "learning_rate": 7.325194680610437e-08,
+      "loss": 1.5816,
+      "step": 3327
+    },
+    {
+      "epoch": 0.9621277825961261,
+      "grad_norm": 1.0143864845201374,
+      "learning_rate": 7.214760437332935e-08,
+      "loss": 1.4547,
+      "step": 3328
+    },
+    {
+      "epoch": 0.9624168834923388,
+      "grad_norm": 1.0173555609276221,
+      "learning_rate": 7.105161965472617e-08,
+      "loss": 1.5577,
+      "step": 3329
+    },
+    {
+      "epoch": 0.9627059843885516,
+      "grad_norm": 1.0195757237728227,
+      "learning_rate": 6.996399357294459e-08,
+      "loss": 1.6033,
+      "step": 3330
+    },
+    {
+      "epoch": 0.9629950852847644,
+      "grad_norm": 1.0458410656089878,
+      "learning_rate": 6.888472704359661e-08,
+      "loss": 1.6399,
+      "step": 3331
+    },
+    {
+      "epoch": 0.9632841861809771,
+      "grad_norm": 0.989335161835818,
+      "learning_rate": 6.781382097525879e-08,
+      "loss": 1.4511,
+      "step": 3332
+    },
+    {
+      "epoch": 0.9635732870771899,
+      "grad_norm": 1.0092505997049857,
+      "learning_rate": 6.675127626946554e-08,
+      "loss": 1.6377,
+      "step": 3333
+    },
+    {
+      "epoch": 0.9638623879734027,
+      "grad_norm": 1.1818165883763694,
+      "learning_rate": 6.569709382071798e-08,
+      "loss": 1.5127,
+      "step": 3334
+    },
+    {
+      "epoch": 0.9641514888696155,
+      "grad_norm": 1.0264360615499033,
+      "learning_rate": 6.46512745164729e-08,
+      "loss": 1.5016,
+      "step": 3335
+    },
+    {
+      "epoch": 0.9644405897658282,
+      "grad_norm": 1.0338827726236335,
+      "learning_rate": 6.361381923714827e-08,
+      "loss": 1.5653,
+      "step": 3336
+    },
+    {
+      "epoch": 0.9647296906620411,
+      "grad_norm": 0.9606621173241008,
+      "learning_rate": 6.2584728856121e-08,
+      "loss": 1.492,
+      "step": 3337
+    },
+    {
+      "epoch": 0.9650187915582539,
+      "grad_norm": 1.0289442128957706,
+      "learning_rate": 6.156400423972697e-08,
+      "loss": 1.4857,
+      "step": 3338
+    },
+    {
+      "epoch": 0.9653078924544666,
+      "grad_norm": 1.052516616677803,
+      "learning_rate": 6.055164624725552e-08,
+      "loss": 1.5634,
+      "step": 3339
+    },
+    {
+      "epoch": 0.9655969933506794,
+      "grad_norm": 1.0111972440267298,
+      "learning_rate": 5.954765573095711e-08,
+      "loss": 1.6212,
+      "step": 3340
+    },
+    {
+      "epoch": 0.9658860942468922,
+      "grad_norm": 1.0231545275520029,
+      "learning_rate": 5.855203353603678e-08,
+      "loss": 1.4007,
+      "step": 3341
+    },
+    {
+      "epoch": 0.9661751951431049,
+      "grad_norm": 1.0349237356762029,
+      "learning_rate": 5.756478050065295e-08,
+      "loss": 1.5794,
+      "step": 3342
+    },
+    {
+      "epoch": 0.9664642960393177,
+      "grad_norm": 1.0464126242294896,
+      "learning_rate": 5.658589745591969e-08,
+      "loss": 1.5732,
+      "step": 3343
+    },
+    {
+      "epoch": 0.9667533969355305,
+      "grad_norm": 0.9947892638787387,
+      "learning_rate": 5.5615385225906704e-08,
+      "loss": 1.5062,
+      "step": 3344
+    },
+    {
+      "epoch": 0.9670424978317432,
+      "grad_norm": 0.9915061592505466,
+      "learning_rate": 5.4653244627632665e-08,
+      "loss": 1.5187,
+      "step": 3345
+    },
+    {
+      "epoch": 0.967331598727956,
+      "grad_norm": 1.033434903691277,
+      "learning_rate": 5.369947647107299e-08,
+      "loss": 1.5937,
+      "step": 3346
+    },
+    {
+      "epoch": 0.9676206996241689,
+      "grad_norm": 1.1783750996898332,
+      "learning_rate": 5.275408155914985e-08,
+      "loss": 1.5262,
+      "step": 3347
+    },
+    {
+      "epoch": 0.9679098005203816,
+      "grad_norm": 1.0062365328888778,
+      "learning_rate": 5.181706068774328e-08,
+      "loss": 1.5282,
+      "step": 3348
+    },
+    {
+      "epoch": 0.9681989014165944,
+      "grad_norm": 1.0232789710903314,
+      "learning_rate": 5.088841464567784e-08,
+      "loss": 1.5214,
+      "step": 3349
+    },
+    {
+      "epoch": 0.9684880023128072,
+      "grad_norm": 1.003554711030437,
+      "learning_rate": 4.996814421472817e-08,
+      "loss": 1.508,
+      "step": 3350
+    },
+    {
+      "epoch": 0.96877710320902,
+      "grad_norm": 1.0134796852238523,
+      "learning_rate": 4.9056250169622344e-08,
+      "loss": 1.5157,
+      "step": 3351
+    },
+    {
+      "epoch": 0.9690662041052327,
+      "grad_norm": 1.032244368486483,
+      "learning_rate": 4.815273327803183e-08,
+      "loss": 1.5703,
+      "step": 3352
+    },
+    {
+      "epoch": 0.9693553050014455,
+      "grad_norm": 0.9862453454366271,
+      "learning_rate": 4.725759430057819e-08,
+      "loss": 1.4643,
+      "step": 3353
+    },
+    {
+      "epoch": 0.9696444058976583,
+      "grad_norm": 1.054886826598327,
+      "learning_rate": 4.6370833990829755e-08,
+      "loss": 1.5997,
+      "step": 3354
+    },
+    {
+      "epoch": 0.969933506793871,
+      "grad_norm": 0.9872443751849136,
+      "learning_rate": 4.5492453095302705e-08,
+      "loss": 1.5232,
+      "step": 3355
+    },
+    {
+      "epoch": 0.9702226076900838,
+      "grad_norm": 0.9659002595534865,
+      "learning_rate": 4.462245235345664e-08,
+      "loss": 1.5167,
+      "step": 3356
+    },
+    {
+      "epoch": 0.9705117085862967,
+      "grad_norm": 1.0066404174755743,
+      "learning_rate": 4.3760832497697915e-08,
+      "loss": 1.5248,
+      "step": 3357
+    },
+    {
+      "epoch": 0.9708008094825094,
+      "grad_norm": 1.0157931291450712,
+      "learning_rate": 4.290759425337521e-08,
+      "loss": 1.5306,
+      "step": 3358
+    },
+    {
+      "epoch": 0.9710899103787222,
+      "grad_norm": 0.9992030995431808,
+      "learning_rate": 4.206273833878616e-08,
+      "loss": 1.4868,
+      "step": 3359
+    },
+    {
+      "epoch": 0.971379011274935,
+      "grad_norm": 1.0486892001220192,
+      "learning_rate": 4.122626546516517e-08,
+      "loss": 1.6459,
+      "step": 3360
+    },
+    {
+      "epoch": 0.9716681121711477,
+      "grad_norm": 1.0054674538534711,
+      "learning_rate": 4.039817633669674e-08,
+      "loss": 1.6193,
+      "step": 3361
+    },
+    {
+      "epoch": 0.9719572130673605,
+      "grad_norm": 0.996361984163474,
+      "learning_rate": 3.957847165049988e-08,
+      "loss": 1.4618,
+      "step": 3362
+    },
+    {
+      "epoch": 0.9722463139635733,
+      "grad_norm": 1.0028545543784153,
+      "learning_rate": 3.8767152096641504e-08,
+      "loss": 1.5294,
+      "step": 3363
+    },
+    {
+      "epoch": 0.9725354148597861,
+      "grad_norm": 1.00024499257164,
+      "learning_rate": 3.796421835812525e-08,
+      "loss": 1.5036,
+      "step": 3364
+    },
+    {
+      "epoch": 0.9728245157559988,
+      "grad_norm": 0.9844024230039832,
+      "learning_rate": 3.716967111089931e-08,
+      "loss": 1.5538,
+      "step": 3365
+    },
+    {
+      "epoch": 0.9731136166522116,
+      "grad_norm": 1.0275669898294235,
+      "learning_rate": 3.638351102384641e-08,
+      "loss": 1.6121,
+      "step": 3366
+    },
+    {
+      "epoch": 0.9734027175484244,
+      "grad_norm": 1.0376454385889713,
+      "learning_rate": 3.5605738758793803e-08,
+      "loss": 1.5645,
+      "step": 3367
+    },
+    {
+      "epoch": 0.9736918184446371,
+      "grad_norm": 1.0067004200981327,
+      "learning_rate": 3.4836354970505524e-08,
+      "loss": 1.5234,
+      "step": 3368
+    },
+    {
+      "epoch": 0.97398091934085,
+      "grad_norm": 1.0208608206289178,
+      "learning_rate": 3.407536030668124e-08,
+      "loss": 1.5686,
+      "step": 3369
+    },
+    {
+      "epoch": 0.9742700202370628,
+      "grad_norm": 0.983336844061033,
+      "learning_rate": 3.3322755407961817e-08,
+      "loss": 1.526,
+      "step": 3370
+    },
+    {
+      "epoch": 0.9745591211332755,
+      "grad_norm": 0.9875406099461627,
+      "learning_rate": 3.2578540907926e-08,
+      "loss": 1.576,
+      "step": 3371
+    },
+    {
+      "epoch": 0.9748482220294883,
+      "grad_norm": 1.0214673748837493,
+      "learning_rate": 3.184271743308376e-08,
+      "loss": 1.5233,
+      "step": 3372
+    },
+    {
+      "epoch": 0.9751373229257011,
+      "grad_norm": 1.0220434422124811,
+      "learning_rate": 3.1115285602887346e-08,
+      "loss": 1.5606,
+      "step": 3373
+    },
+    {
+      "epoch": 0.9754264238219138,
+      "grad_norm": 1.0154997725811554,
+      "learning_rate": 3.039624602971913e-08,
+      "loss": 1.5326,
+      "step": 3374
+    },
+    {
+      "epoch": 0.9757155247181266,
+      "grad_norm": 0.9903388008067455,
+      "learning_rate": 2.968559931890158e-08,
+      "loss": 1.5153,
+      "step": 3375
+    },
+    {
+      "epoch": 0.9760046256143394,
+      "grad_norm": 1.0473180368678654,
+      "learning_rate": 2.8983346068688357e-08,
+      "loss": 1.5103,
+      "step": 3376
+    },
+    {
+      "epoch": 0.9762937265105521,
+      "grad_norm": 1.003427444340405,
+      "learning_rate": 2.8289486870266557e-08,
+      "loss": 1.4164,
+      "step": 3377
+    },
+    {
+      "epoch": 0.9765828274067649,
+      "grad_norm": 1.02031840271782,
+      "learning_rate": 2.7604022307758937e-08,
+      "loss": 1.4828,
+      "step": 3378
+    },
+    {
+      "epoch": 0.9768719283029778,
+      "grad_norm": 1.007161898598709,
+      "learning_rate": 2.6926952958221673e-08,
+      "loss": 1.5335,
+      "step": 3379
+    },
+    {
+      "epoch": 0.9771610291991906,
+      "grad_norm": 1.0292234266831068,
+      "learning_rate": 2.625827939164105e-08,
+      "loss": 1.526,
+      "step": 3380
+    },
+    {
+      "epoch": 0.9774501300954033,
+      "grad_norm": 0.9914736482215937,
+      "learning_rate": 2.5598002170936777e-08,
+      "loss": 1.5889,
+      "step": 3381
+    },
+    {
+      "epoch": 0.9777392309916161,
+      "grad_norm": 1.002773492389994,
+      "learning_rate": 2.494612185195977e-08,
+      "loss": 1.4692,
+      "step": 3382
+    },
+    {
+      "epoch": 0.9780283318878289,
+      "grad_norm": 1.0339958179836686,
+      "learning_rate": 2.4302638983494385e-08,
+      "loss": 1.4461,
+      "step": 3383
+    },
+    {
+      "epoch": 0.9783174327840416,
+      "grad_norm": 1.015412766197476,
+      "learning_rate": 2.3667554107250634e-08,
+      "loss": 1.6069,
+      "step": 3384
+    },
+    {
+      "epoch": 0.9786065336802544,
+      "grad_norm": 1.078696850809058,
+      "learning_rate": 2.3040867757873063e-08,
+      "loss": 1.438,
+      "step": 3385
+    },
+    {
+      "epoch": 0.9788956345764672,
+      "grad_norm": 1.0109883872840963,
+      "learning_rate": 2.2422580462934108e-08,
+      "loss": 1.4309,
+      "step": 3386
+    },
+    {
+      "epoch": 0.9791847354726799,
+      "grad_norm": 0.9876541163371628,
+      "learning_rate": 2.1812692742936293e-08,
+      "loss": 1.5439,
+      "step": 3387
+    },
+    {
+      "epoch": 0.9794738363688927,
+      "grad_norm": 0.999274975041907,
+      "learning_rate": 2.121120511131114e-08,
+      "loss": 1.4618,
+      "step": 3388
+    },
+    {
+      "epoch": 0.9797629372651055,
+      "grad_norm": 1.0065094250128939,
+      "learning_rate": 2.0618118074416937e-08,
+      "loss": 1.4714,
+      "step": 3389
+    },
+    {
+      "epoch": 0.9800520381613183,
+      "grad_norm": 1.0237487423301916,
+      "learning_rate": 2.0033432131542073e-08,
+      "loss": 1.53,
+      "step": 3390
+    },
+    {
+      "epoch": 0.9803411390575311,
+      "grad_norm": 0.984517654331654,
+      "learning_rate": 1.9457147774900598e-08,
+      "loss": 1.3868,
+      "step": 3391
+    },
+    {
+      "epoch": 0.9806302399537439,
+      "grad_norm": 1.0081788742061784,
+      "learning_rate": 1.8889265489635544e-08,
+      "loss": 1.5124,
+      "step": 3392
+    },
+    {
+      "epoch": 0.9809193408499567,
+      "grad_norm": 0.986391237491382,
+      "learning_rate": 1.832978575381561e-08,
+      "loss": 1.5248,
+      "step": 3393
+    },
+    {
+      "epoch": 0.9812084417461694,
+      "grad_norm": 1.015920934266796,
+      "learning_rate": 1.777870903843515e-08,
+      "loss": 1.5534,
+      "step": 3394
+    },
+    {
+      "epoch": 0.9814975426423822,
+      "grad_norm": 1.0204634225617906,
+      "learning_rate": 1.7236035807416397e-08,
+      "loss": 1.6085,
+      "step": 3395
+    },
+    {
+      "epoch": 0.981786643538595,
+      "grad_norm": 1.01958295489975,
+      "learning_rate": 1.6701766517607244e-08,
+      "loss": 1.5547,
+      "step": 3396
+    },
+    {
+      "epoch": 0.9820757444348077,
+      "grad_norm": 1.0553672490182737,
+      "learning_rate": 1.6175901618776802e-08,
+      "loss": 1.4711,
+      "step": 3397
+    },
+    {
+      "epoch": 0.9823648453310205,
+      "grad_norm": 1.0026715307164875,
+      "learning_rate": 1.565844155362539e-08,
+      "loss": 1.5812,
+      "step": 3398
+    },
+    {
+      "epoch": 0.9826539462272333,
+      "grad_norm": 1.0113546412866203,
+      "learning_rate": 1.5149386757771222e-08,
+      "loss": 1.4891,
+      "step": 3399
+    },
+    {
+      "epoch": 0.982943047123446,
+      "grad_norm": 1.0883677114840837,
+      "learning_rate": 1.4648737659760381e-08,
+      "loss": 1.6301,
+      "step": 3400
+    },
+    {
+      "epoch": 0.9832321480196589,
+      "grad_norm": 1.0859509286181708,
+      "learning_rate": 1.4156494681062393e-08,
+      "loss": 1.6455,
+      "step": 3401
+    },
+    {
+      "epoch": 0.9835212489158717,
+      "grad_norm": 1.0238880574304083,
+      "learning_rate": 1.3672658236070224e-08,
+      "loss": 1.5733,
+      "step": 3402
+    },
+    {
+      "epoch": 0.9838103498120844,
+      "grad_norm": 1.004682379101414,
+      "learning_rate": 1.3197228732098055e-08,
+      "loss": 1.5475,
+      "step": 3403
+    },
+    {
+      "epoch": 0.9840994507082972,
+      "grad_norm": 0.9884064630529379,
+      "learning_rate": 1.2730206569384618e-08,
+      "loss": 1.5392,
+      "step": 3404
+    },
+    {
+      "epoch": 0.98438855160451,
+      "grad_norm": 1.0029867798135903,
+      "learning_rate": 1.227159214108986e-08,
+      "loss": 1.4885,
+      "step": 3405
+    },
+    {
+      "epoch": 0.9846776525007227,
+      "grad_norm": 1.0259507364707694,
+      "learning_rate": 1.1821385833296062e-08,
+      "loss": 1.5529,
+      "step": 3406
+    },
+    {
+      "epoch": 0.9849667533969355,
+      "grad_norm": 1.0286869916162014,
+      "learning_rate": 1.1379588025006716e-08,
+      "loss": 1.5776,
+      "step": 3407
+    },
+    {
+      "epoch": 0.9852558542931483,
+      "grad_norm": 1.0163123186599254,
+      "learning_rate": 1.0946199088148757e-08,
+      "loss": 1.4828,
+      "step": 3408
+    },
+    {
+      "epoch": 0.9855449551893611,
+      "grad_norm": 0.9842987463970254,
+      "learning_rate": 1.0521219387567005e-08,
+      "loss": 1.4765,
+      "step": 3409
+    },
+    {
+      "epoch": 0.9858340560855738,
+      "grad_norm": 1.0160236689375794,
+      "learning_rate": 1.0104649281028612e-08,
+      "loss": 1.6089,
+      "step": 3410
+    },
+    {
+      "epoch": 0.9861231569817867,
+      "grad_norm": 1.0015042014949036,
+      "learning_rate": 9.696489119221942e-09,
+      "loss": 1.5566,
+      "step": 3411
+    },
+    {
+      "epoch": 0.9864122578779995,
+      "grad_norm": 1.0140035856553329,
+      "learning_rate": 9.29673924575436e-09,
+      "loss": 1.5773,
+      "step": 3412
+    },
+    {
+      "epoch": 0.9867013587742122,
+      "grad_norm": 0.9790669644843083,
+      "learning_rate": 8.905399997154452e-09,
+      "loss": 1.5154,
+      "step": 3413
+    },
+    {
+      "epoch": 0.986990459670425,
+      "grad_norm": 0.989790421795058,
+      "learning_rate": 8.522471702866464e-09,
+      "loss": 1.4711,
+      "step": 3414
+    },
+    {
+      "epoch": 0.9872795605666378,
+      "grad_norm": 1.0878148172181985,
+      "learning_rate": 8.147954685259196e-09,
+      "loss": 1.5904,
+      "step": 3415
+    },
+    {
+      "epoch": 0.9875686614628505,
+      "grad_norm": 1.0148013204665856,
+      "learning_rate": 7.781849259617113e-09,
+      "loss": 1.5565,
+      "step": 3416
+    },
+    {
+      "epoch": 0.9878577623590633,
+      "grad_norm": 0.9983063127028312,
+      "learning_rate": 7.424155734143679e-09,
+      "loss": 1.4992,
+      "step": 3417
+    },
+    {
+      "epoch": 0.9881468632552761,
+      "grad_norm": 0.9937548775097956,
+      "learning_rate": 7.0748744099624625e-09,
+      "loss": 1.5946,
+      "step": 3418
+    },
+    {
+      "epoch": 0.9884359641514888,
+      "grad_norm": 1.023996069774853,
+      "learning_rate": 6.734005581113812e-09,
+      "loss": 1.4843,
+      "step": 3419
+    },
+    {
+      "epoch": 0.9887250650477016,
+      "grad_norm": 0.9923185965266281,
+      "learning_rate": 6.401549534555962e-09,
+      "loss": 1.538,
+      "step": 3420
+    },
+    {
+      "epoch": 0.9890141659439144,
+      "grad_norm": 0.9925886028598676,
+      "learning_rate": 6.077506550167256e-09,
+      "loss": 1.5145,
+      "step": 3421
+    },
+    {
+      "epoch": 0.9893032668401271,
+      "grad_norm": 1.0067774791104365,
+      "learning_rate": 5.761876900739482e-09,
+      "loss": 1.497,
+      "step": 3422
+    },
+    {
+      "epoch": 0.98959236773634,
+      "grad_norm": 1.0219945361390008,
+      "learning_rate": 5.454660851984539e-09,
+      "loss": 1.5678,
+      "step": 3423
+    },
+    {
+      "epoch": 0.9898814686325528,
+      "grad_norm": 1.001549469328582,
+      "learning_rate": 5.155858662531099e-09,
+      "loss": 1.4998,
+      "step": 3424
+    },
+    {
+      "epoch": 0.9901705695287656,
+      "grad_norm": 0.9973668152632869,
+      "learning_rate": 4.8654705839257246e-09,
+      "loss": 1.4939,
+      "step": 3425
+    },
+    {
+      "epoch": 0.9904596704249783,
+      "grad_norm": 0.9686048595225004,
+      "learning_rate": 4.583496860627312e-09,
+      "loss": 1.505,
+      "step": 3426
+    },
+    {
+      "epoch": 0.9907487713211911,
+      "grad_norm": 1.0024920784278541,
+      "learning_rate": 4.309937730015978e-09,
+      "loss": 1.549,
+      "step": 3427
+    },
+    {
+      "epoch": 0.9910378722174039,
+      "grad_norm": 1.01092625892776,
+      "learning_rate": 4.0447934223852845e-09,
+      "loss": 1.6417,
+      "step": 3428
+    },
+    {
+      "epoch": 0.9913269731136166,
+      "grad_norm": 1.024731527477819,
+      "learning_rate": 3.78806416094668e-09,
+      "loss": 1.5023,
+      "step": 3429
+    },
+    {
+      "epoch": 0.9916160740098294,
+      "grad_norm": 0.9969413180160911,
+      "learning_rate": 3.5397501618261722e-09,
+      "loss": 1.5396,
+      "step": 3430
+    },
+    {
+      "epoch": 0.9919051749060422,
+      "grad_norm": 1.0443636236307239,
+      "learning_rate": 3.2998516340654318e-09,
+      "loss": 1.6323,
+      "step": 3431
+    },
+    {
+      "epoch": 0.9921942758022549,
+      "grad_norm": 1.0382932438086545,
+      "learning_rate": 3.0683687796206895e-09,
+      "loss": 1.5445,
+      "step": 3432
+    },
+    {
+      "epoch": 0.9924833766984678,
+      "grad_norm": 1.0037775420316293,
+      "learning_rate": 2.845301793366062e-09,
+      "loss": 1.5706,
+      "step": 3433
+    },
+    {
+      "epoch": 0.9927724775946806,
+      "grad_norm": 1.0128737541403898,
+      "learning_rate": 2.6306508630902226e-09,
+      "loss": 1.4672,
+      "step": 3434
+    },
+    {
+      "epoch": 0.9930615784908933,
+      "grad_norm": 0.9977769460697357,
+      "learning_rate": 2.4244161694941814e-09,
+      "loss": 1.5046,
+      "step": 3435
+    },
+    {
+      "epoch": 0.9933506793871061,
+      "grad_norm": 0.9975756807682504,
+      "learning_rate": 2.226597886195725e-09,
+      "loss": 1.4956,
+      "step": 3436
+    },
+    {
+      "epoch": 0.9936397802833189,
+      "grad_norm": 1.013831187653767,
+      "learning_rate": 2.037196179728307e-09,
+      "loss": 1.5481,
+      "step": 3437
+    },
+    {
+      "epoch": 0.9939288811795317,
+      "grad_norm": 0.9965931192183253,
+      "learning_rate": 1.8562112095388273e-09,
+      "loss": 1.5102,
+      "step": 3438
+    },
+    {
+      "epoch": 0.9942179820757444,
+      "grad_norm": 0.9635231194983375,
+      "learning_rate": 1.683643127987633e-09,
+      "loss": 1.5787,
+      "step": 3439
+    },
+    {
+      "epoch": 0.9945070829719572,
+      "grad_norm": 1.0108641511249579,
+      "learning_rate": 1.5194920803507375e-09,
+      "loss": 1.5484,
+      "step": 3440
+    },
+    {
+      "epoch": 0.99479618386817,
+      "grad_norm": 1.0551796338579675,
+      "learning_rate": 1.3637582048187104e-09,
+      "loss": 1.5806,
+      "step": 3441
+    },
+    {
+      "epoch": 0.9950852847643827,
+      "grad_norm": 1.0345911204373301,
+      "learning_rate": 1.2164416324933482e-09,
+      "loss": 1.5539,
+      "step": 3442
+    },
+    {
+      "epoch": 0.9953743856605956,
+      "grad_norm": 1.005098988776073,
+      "learning_rate": 1.077542487394334e-09,
+      "loss": 1.598,
+      "step": 3443
+    },
+    {
+      "epoch": 0.9956634865568084,
+      "grad_norm": 1.030673111746361,
+      "learning_rate": 9.470608864525777e-10,
+      "loss": 1.4873,
+      "step": 3444
+    },
+    {
+      "epoch": 0.9959525874530211,
+      "grad_norm": 0.9771102194241225,
+      "learning_rate": 8.249969395135449e-10,
+      "loss": 1.6138,
+      "step": 3445
+    },
+    {
+      "epoch": 0.9962416883492339,
+      "grad_norm": 1.0056666684963416,
+      "learning_rate": 7.113507493361482e-10,
+      "loss": 1.4501,
+      "step": 3446
+    },
+    {
+      "epoch": 0.9965307892454467,
+      "grad_norm": 1.0631977783464415,
+      "learning_rate": 6.061224115927467e-10,
+      "loss": 1.6686,
+      "step": 3447
+    },
+    {
+      "epoch": 0.9968198901416594,
+      "grad_norm": 0.9883535172015726,
+      "learning_rate": 5.093120148680353e-10,
+      "loss": 1.5398,
+      "step": 3448
+    },
+    {
+      "epoch": 0.9971089910378722,
+      "grad_norm": 0.9894895190048201,
+      "learning_rate": 4.209196406645966e-10,
+      "loss": 1.4819,
+      "step": 3449
+    },
+    {
+      "epoch": 0.997398091934085,
+      "grad_norm": 1.0064038764989938,
+      "learning_rate": 3.4094536339179807e-10,
+      "loss": 1.4988,
+      "step": 3450
+    },
+    {
+      "epoch": 0.9976871928302977,
+      "grad_norm": 0.9583584460148167,
+      "learning_rate": 2.6938925037689467e-10,
+      "loss": 1.5026,
+      "step": 3451
+    },
+    {
+      "epoch": 0.9979762937265105,
+      "grad_norm": 1.040258646016716,
+      "learning_rate": 2.0625136185947748e-10,
+      "loss": 1.5319,
+      "step": 3452
+    },
+    {
+      "epoch": 0.9982653946227233,
+      "grad_norm": 1.0293920452207672,
+      "learning_rate": 1.5153175099147376e-10,
+      "loss": 1.5227,
+      "step": 3453
+    },
+    {
+      "epoch": 0.9985544955189362,
+      "grad_norm": 1.0327132025776296,
+      "learning_rate": 1.0523046383825731e-10,
+      "loss": 1.5682,
+      "step": 3454
+    },
+    {
+      "epoch": 0.9988435964151489,
+      "grad_norm": 1.0024850462628732,
+      "learning_rate": 6.734753937864825e-11,
+      "loss": 1.4182,
+      "step": 3455
+    },
+    {
+      "epoch": 0.9991326973113617,
+      "grad_norm": 1.0431751397180047,
+      "learning_rate": 3.788300950380297e-11,
+      "loss": 1.5725,
+      "step": 3456
+    },
+    {
+      "epoch": 0.9994217982075745,
+      "grad_norm": 1.015523866634744,
+      "learning_rate": 1.6836899018324303e-11,
+      "loss": 1.5303,
+      "step": 3457
+    },
+    {
+      "epoch": 0.9997108991037872,
+      "grad_norm": 1.0112011289497136,
+      "learning_rate": 4.209225640261494e-12,
+      "loss": 1.5459,
+      "step": 3458
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.066385150811282,
+      "learning_rate": 0.0,
+      "loss": 1.4635,
+      "step": 3459
+    },
+    {
+      "epoch": 1.0,
+      "step": 3459,
+      "total_flos": 362122114498560.0,
+      "train_loss": 1.5948989816951282,
+      "train_runtime": 22004.1828,
+      "train_samples_per_second": 5.03,
+      "train_steps_per_second": 0.157
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 3459,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 700,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 362122114498560.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}