diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100755--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,7042 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.4342162396873643,
+  "eval_steps": 500,
+  "global_step": 1000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0004342162396873643,
+      "grad_norm": 0.9606487154960632,
+      "learning_rate": 5e-05,
+      "loss": 9.188,
+      "step": 1
+    },
+    {
+      "epoch": 0.0008684324793747286,
+      "grad_norm": 0.9193044304847717,
+      "learning_rate": 0.0001,
+      "loss": 9.1897,
+      "step": 2
+    },
+    {
+      "epoch": 0.0013026487190620929,
+      "grad_norm": 0.9406704306602478,
+      "learning_rate": 0.00015,
+      "loss": 9.0726,
+      "step": 3
+    },
+    {
+      "epoch": 0.0017368649587494573,
+      "grad_norm": 0.9401040077209473,
+      "learning_rate": 0.0002,
+      "loss": 8.7129,
+      "step": 4
+    },
+    {
+      "epoch": 0.0021710811984368217,
+      "grad_norm": 0.8919561505317688,
+      "learning_rate": 0.00025,
+      "loss": 8.1364,
+      "step": 5
+    },
+    {
+      "epoch": 0.0026052974381241857,
+      "grad_norm": 0.8310884237289429,
+      "learning_rate": 0.0003,
+      "loss": 7.4088,
+      "step": 6
+    },
+    {
+      "epoch": 0.00303951367781155,
+      "grad_norm": 0.748382031917572,
+      "learning_rate": 0.00035,
+      "loss": 6.6918,
+      "step": 7
+    },
+    {
+      "epoch": 0.0034737299174989146,
+      "grad_norm": 0.6353039145469666,
+      "learning_rate": 0.0004,
+      "loss": 5.8753,
+      "step": 8
+    },
+    {
+      "epoch": 0.0039079461571862786,
+      "grad_norm": 0.5517627000808716,
+      "learning_rate": 0.00045000000000000004,
+      "loss": 5.4347,
+      "step": 9
+    },
+    {
+      "epoch": 0.004342162396873643,
+      "grad_norm": 0.5389806628227234,
+      "learning_rate": 0.0005,
+      "loss": 5.0181,
+      "step": 10
+    },
+    {
+      "epoch": 0.004776378636561007,
+      "grad_norm": 0.6018754243850708,
+      "learning_rate": 0.00055,
+      "loss": 4.9487,
+      "step": 11
+    },
+    {
+      "epoch": 0.005210594876248371,
+      "grad_norm": 0.6734316349029541,
+      "learning_rate": 0.0006,
+      "loss": 4.9233,
+      "step": 12
+    },
+    {
+      "epoch": 0.005644811115935736,
+      "grad_norm": 0.6842392086982727,
+      "learning_rate": 0.0006500000000000001,
+      "loss": 4.8078,
+      "step": 13
+    },
+    {
+      "epoch": 0.0060790273556231,
+      "grad_norm": 0.6519020199775696,
+      "learning_rate": 0.0007,
+      "loss": 4.6065,
+      "step": 14
+    },
+    {
+      "epoch": 0.006513243595310464,
+      "grad_norm": 0.5788787603378296,
+      "learning_rate": 0.00075,
+      "loss": 4.3967,
+      "step": 15
+    },
+    {
+      "epoch": 0.006947459834997829,
+      "grad_norm": 0.5370550751686096,
+      "learning_rate": 0.0008,
+      "loss": 4.1971,
+      "step": 16
+    },
+    {
+      "epoch": 0.007381676074685193,
+      "grad_norm": 0.4862042963504791,
+      "learning_rate": 0.00085,
+      "loss": 3.8583,
+      "step": 17
+    },
+    {
+      "epoch": 0.007815892314372557,
+      "grad_norm": 0.4636250436306,
+      "learning_rate": 0.0009000000000000001,
+      "loss": 3.6372,
+      "step": 18
+    },
+    {
+      "epoch": 0.008250108554059922,
+      "grad_norm": 0.4613461494445801,
+      "learning_rate": 0.00095,
+      "loss": 3.5642,
+      "step": 19
+    },
+    {
+      "epoch": 0.008684324793747287,
+      "grad_norm": 0.4593052566051483,
+      "learning_rate": 0.001,
+      "loss": 3.4463,
+      "step": 20
+    },
+    {
+      "epoch": 0.00911854103343465,
+      "grad_norm": 0.4367975890636444,
+      "learning_rate": 0.0009999974308631955,
+      "loss": 3.2621,
+      "step": 21
+    },
+    {
+      "epoch": 0.009552757273122015,
+      "grad_norm": 0.41558611392974854,
+      "learning_rate": 0.000999989723479183,
+      "loss": 3.0598,
+      "step": 22
+    },
+    {
+      "epoch": 0.00998697351280938,
+      "grad_norm": 0.3967452049255371,
+      "learning_rate": 0.0009999768779271685,
+      "loss": 2.936,
+      "step": 23
+    },
+    {
+      "epoch": 0.010421189752496743,
+      "grad_norm": 0.3664601147174835,
+      "learning_rate": 0.0009999588943391596,
+      "loss": 2.6913,
+      "step": 24
+    },
+    {
+      "epoch": 0.010855405992184108,
+      "grad_norm": 0.359183669090271,
+      "learning_rate": 0.0009999357728999656,
+      "loss": 2.5639,
+      "step": 25
+    },
+    {
+      "epoch": 0.011289622231871473,
+      "grad_norm": 0.3342216908931732,
+      "learning_rate": 0.000999907513847195,
+      "loss": 2.401,
+      "step": 26
+    },
+    {
+      "epoch": 0.011723838471558836,
+      "grad_norm": 0.36653760075569153,
+      "learning_rate": 0.0009998741174712534,
+      "loss": 2.6216,
+      "step": 27
+    },
+    {
+      "epoch": 0.0121580547112462,
+      "grad_norm": 0.30254194140434265,
+      "learning_rate": 0.00099983558411534,
+      "loss": 2.2086,
+      "step": 28
+    },
+    {
+      "epoch": 0.012592270950933565,
+      "grad_norm": 0.2917495369911194,
+      "learning_rate": 0.0009997919141754449,
+      "loss": 2.1274,
+      "step": 29
+    },
+    {
+      "epoch": 0.013026487190620929,
+      "grad_norm": 0.2690434455871582,
+      "learning_rate": 0.000999743108100344,
+      "loss": 2.0581,
+      "step": 30
+    },
+    {
+      "epoch": 0.013460703430308293,
+      "grad_norm": 0.26215028762817383,
+      "learning_rate": 0.0009996891663915954,
+      "loss": 1.9764,
+      "step": 31
+    },
+    {
+      "epoch": 0.013894919669995658,
+      "grad_norm": 0.2453673779964447,
+      "learning_rate": 0.0009996300896035338,
+      "loss": 1.9219,
+      "step": 32
+    },
+    {
+      "epoch": 0.014329135909683021,
+      "grad_norm": 0.24287743866443634,
+      "learning_rate": 0.0009995658783432644,
+      "loss": 1.8433,
+      "step": 33
+    },
+    {
+      "epoch": 0.014763352149370386,
+      "grad_norm": 0.2258252203464508,
+      "learning_rate": 0.0009994965332706573,
+      "loss": 1.7392,
+      "step": 34
+    },
+    {
+      "epoch": 0.015197568389057751,
+      "grad_norm": 0.21900461614131927,
+      "learning_rate": 0.0009994220550983404,
+      "loss": 1.7185,
+      "step": 35
+    },
+    {
+      "epoch": 0.015631784628745114,
+      "grad_norm": 0.21046680212020874,
+      "learning_rate": 0.0009993424445916923,
+      "loss": 1.6541,
+      "step": 36
+    },
+    {
+      "epoch": 0.01606600086843248,
+      "grad_norm": 0.19736959040164948,
+      "learning_rate": 0.0009992577025688338,
+      "loss": 1.6321,
+      "step": 37
+    },
+    {
+      "epoch": 0.016500217108119844,
+      "grad_norm": 0.19068720936775208,
+      "learning_rate": 0.0009991678299006205,
+      "loss": 1.6236,
+      "step": 38
+    },
+    {
+      "epoch": 0.01693443334780721,
+      "grad_norm": 0.18818727135658264,
+      "learning_rate": 0.000999072827510633,
+      "loss": 1.5791,
+      "step": 39
+    },
+    {
+      "epoch": 0.017368649587494574,
+      "grad_norm": 0.17485196888446808,
+      "learning_rate": 0.0009989726963751681,
+      "loss": 1.6086,
+      "step": 40
+    },
+    {
+      "epoch": 0.017802865827181935,
+      "grad_norm": 0.1762186735868454,
+      "learning_rate": 0.000998867437523228,
+      "loss": 1.5328,
+      "step": 41
+    },
+    {
+      "epoch": 0.0182370820668693,
+      "grad_norm": 0.19658854603767395,
+      "learning_rate": 0.0009987570520365104,
+      "loss": 1.4491,
+      "step": 42
+    },
+    {
+      "epoch": 0.018671298306556665,
+      "grad_norm": 0.1639322191476822,
+      "learning_rate": 0.0009986415410493965,
+      "loss": 1.4447,
+      "step": 43
+    },
+    {
+      "epoch": 0.01910551454624403,
+      "grad_norm": 0.1552889496088028,
+      "learning_rate": 0.000998520905748941,
+      "loss": 1.4345,
+      "step": 44
+    },
+    {
+      "epoch": 0.019539730785931395,
+      "grad_norm": 0.14960208535194397,
+      "learning_rate": 0.0009983951473748578,
+      "loss": 1.364,
+      "step": 45
+    },
+    {
+      "epoch": 0.01997394702561876,
+      "grad_norm": 0.14552171528339386,
+      "learning_rate": 0.0009982642672195091,
+      "loss": 1.3422,
+      "step": 46
+    },
+    {
+      "epoch": 0.02040816326530612,
+      "grad_norm": 0.14638382196426392,
+      "learning_rate": 0.0009981282666278908,
+      "loss": 1.325,
+      "step": 47
+    },
+    {
+      "epoch": 0.020842379504993486,
+      "grad_norm": 0.1372004598379135,
+      "learning_rate": 0.0009979871469976197,
+      "loss": 1.2685,
+      "step": 48
+    },
+    {
+      "epoch": 0.02127659574468085,
+      "grad_norm": 0.12978093326091766,
+      "learning_rate": 0.0009978409097789176,
+      "loss": 1.3268,
+      "step": 49
+    },
+    {
+      "epoch": 0.021710811984368215,
+      "grad_norm": 0.13076020777225494,
+      "learning_rate": 0.0009976895564745992,
+      "loss": 1.2211,
+      "step": 50
+    },
+    {
+      "epoch": 0.02214502822405558,
+      "grad_norm": 0.1264386773109436,
+      "learning_rate": 0.000997533088640053,
+      "loss": 1.2597,
+      "step": 51
+    },
+    {
+      "epoch": 0.022579244463742945,
+      "grad_norm": 0.12510421872138977,
+      "learning_rate": 0.0009973715078832287,
+      "loss": 1.178,
+      "step": 52
+    },
+    {
+      "epoch": 0.02301346070343031,
+      "grad_norm": 0.13376790285110474,
+      "learning_rate": 0.0009972048158646183,
+      "loss": 1.1614,
+      "step": 53
+    },
+    {
+      "epoch": 0.02344767694311767,
+      "grad_norm": 0.11965133249759674,
+      "learning_rate": 0.00099703301429724,
+      "loss": 1.1549,
+      "step": 54
+    },
+    {
+      "epoch": 0.023881893182805036,
+      "grad_norm": 0.10872318595647812,
+      "learning_rate": 0.0009968561049466214,
+      "loss": 1.143,
+      "step": 55
+    },
+    {
+      "epoch": 0.0243161094224924,
+      "grad_norm": 0.12483104318380356,
+      "learning_rate": 0.000996674089630779,
+      "loss": 1.2124,
+      "step": 56
+    },
+    {
+      "epoch": 0.024750325662179766,
+      "grad_norm": 0.10931334644556046,
+      "learning_rate": 0.0009964869702202021,
+      "loss": 1.1176,
+      "step": 57
+    },
+    {
+      "epoch": 0.02518454190186713,
+      "grad_norm": 0.12575986981391907,
+      "learning_rate": 0.0009962947486378325,
+      "loss": 1.0858,
+      "step": 58
+    },
+    {
+      "epoch": 0.025618758141554496,
+      "grad_norm": 0.11589810997247696,
+      "learning_rate": 0.0009960974268590438,
+      "loss": 1.0867,
+      "step": 59
+    },
+    {
+      "epoch": 0.026052974381241857,
+      "grad_norm": 0.10477310419082642,
+      "learning_rate": 0.000995895006911623,
+      "loss": 1.1162,
+      "step": 60
+    },
+    {
+      "epoch": 0.026487190620929222,
+      "grad_norm": 0.09671472758054733,
+      "learning_rate": 0.0009956874908757481,
+      "loss": 1.0662,
+      "step": 61
+    },
+    {
+      "epoch": 0.026921406860616587,
+      "grad_norm": 0.10130201280117035,
+      "learning_rate": 0.0009954748808839674,
+      "loss": 1.0826,
+      "step": 62
+    },
+    {
+      "epoch": 0.02735562310030395,
+      "grad_norm": 0.10421687364578247,
+      "learning_rate": 0.0009952571791211776,
+      "loss": 1.0302,
+      "step": 63
+    },
+    {
+      "epoch": 0.027789839339991317,
+      "grad_norm": 0.09363257884979248,
+      "learning_rate": 0.0009950343878246009,
+      "loss": 1.0769,
+      "step": 64
+    },
+    {
+      "epoch": 0.02822405557967868,
+      "grad_norm": 0.10000266879796982,
+      "learning_rate": 0.000994806509283763,
+      "loss": 1.0048,
+      "step": 65
+    },
+    {
+      "epoch": 0.028658271819366043,
+      "grad_norm": 0.091400146484375,
+      "learning_rate": 0.0009945735458404682,
+      "loss": 0.9903,
+      "step": 66
+    },
+    {
+      "epoch": 0.029092488059053408,
+      "grad_norm": 0.09233607351779938,
+      "learning_rate": 0.0009943354998887763,
+      "loss": 1.0563,
+      "step": 67
+    },
+    {
+      "epoch": 0.029526704298740773,
+      "grad_norm": 0.0991148054599762,
+      "learning_rate": 0.0009940923738749779,
+      "loss": 1.0117,
+      "step": 68
+    },
+    {
+      "epoch": 0.029960920538428137,
+      "grad_norm": 0.08926694095134735,
+      "learning_rate": 0.0009938441702975688,
+      "loss": 0.974,
+      "step": 69
+    },
+    {
+      "epoch": 0.030395136778115502,
+      "grad_norm": 0.08934075385332108,
+      "learning_rate": 0.0009935908917072251,
+      "loss": 1.0283,
+      "step": 70
+    },
+    {
+      "epoch": 0.030829353017802867,
+      "grad_norm": 0.11343932151794434,
+      "learning_rate": 0.000993332540706776,
+      "loss": 0.998,
+      "step": 71
+    },
+    {
+      "epoch": 0.03126356925749023,
+      "grad_norm": 0.11343089491128922,
+      "learning_rate": 0.0009930691199511775,
+      "loss": 0.993,
+      "step": 72
+    },
+    {
+      "epoch": 0.0316977854971776,
+      "grad_norm": 0.09531175345182419,
+      "learning_rate": 0.000992800632147486,
+      "loss": 1.0179,
+      "step": 73
+    },
+    {
+      "epoch": 0.03213200173686496,
+      "grad_norm": 0.13536381721496582,
+      "learning_rate": 0.0009925270800548284,
+      "loss": 0.9539,
+      "step": 74
+    },
+    {
+      "epoch": 0.03256621797655232,
+      "grad_norm": 0.09377475082874298,
+      "learning_rate": 0.0009922484664843763,
+      "loss": 0.9811,
+      "step": 75
+    },
+    {
+      "epoch": 0.03300043421623969,
+      "grad_norm": 0.09711369127035141,
+      "learning_rate": 0.0009919647942993148,
+      "loss": 0.9223,
+      "step": 76
+    },
+    {
+      "epoch": 0.03343465045592705,
+      "grad_norm": 0.11353003978729248,
+      "learning_rate": 0.0009916760664148148,
+      "loss": 0.9736,
+      "step": 77
+    },
+    {
+      "epoch": 0.03386886669561442,
+      "grad_norm": 0.09144185483455658,
+      "learning_rate": 0.0009913822857980019,
+      "loss": 0.9576,
+      "step": 78
+    },
+    {
+      "epoch": 0.03430308293530178,
+      "grad_norm": 0.08733920753002167,
+      "learning_rate": 0.0009910834554679266,
+      "loss": 0.9016,
+      "step": 79
+    },
+    {
+      "epoch": 0.03473729917498915,
+      "grad_norm": 0.08889693766832352,
+      "learning_rate": 0.0009907795784955327,
+      "loss": 0.8981,
+      "step": 80
+    },
+    {
+      "epoch": 0.03517151541467651,
+      "grad_norm": 0.081461600959301,
+      "learning_rate": 0.0009904706580036264,
+      "loss": 0.9246,
+      "step": 81
+    },
+    {
+      "epoch": 0.03560573165436387,
+      "grad_norm": 0.08032266050577164,
+      "learning_rate": 0.0009901566971668436,
+      "loss": 0.901,
+      "step": 82
+    },
+    {
+      "epoch": 0.03603994789405124,
+      "grad_norm": 0.08888774365186691,
+      "learning_rate": 0.0009898376992116178,
+      "loss": 0.916,
+      "step": 83
+    },
+    {
+      "epoch": 0.0364741641337386,
+      "grad_norm": 0.08942156285047531,
+      "learning_rate": 0.0009895136674161465,
+      "loss": 0.9122,
+      "step": 84
+    },
+    {
+      "epoch": 0.03690838037342597,
+      "grad_norm": 0.08162346482276917,
+      "learning_rate": 0.0009891846051103576,
+      "loss": 0.8893,
+      "step": 85
+    },
+    {
+      "epoch": 0.03734259661311333,
+      "grad_norm": 0.08491238951683044,
+      "learning_rate": 0.0009888505156758758,
+      "loss": 0.8572,
+      "step": 86
+    },
+    {
+      "epoch": 0.0377768128528007,
+      "grad_norm": 0.07837095111608505,
+      "learning_rate": 0.0009885114025459864,
+      "loss": 0.851,
+      "step": 87
+    },
+    {
+      "epoch": 0.03821102909248806,
+      "grad_norm": 0.08016610145568848,
+      "learning_rate": 0.0009881672692056021,
+      "loss": 0.8985,
+      "step": 88
+    },
+    {
+      "epoch": 0.03864524533217542,
+      "grad_norm": 0.07894952595233917,
+      "learning_rate": 0.000987818119191225,
+      "loss": 0.8719,
+      "step": 89
+    },
+    {
+      "epoch": 0.03907946157186279,
+      "grad_norm": 0.07604355365037918,
+      "learning_rate": 0.0009874639560909118,
+      "loss": 0.883,
+      "step": 90
+    },
+    {
+      "epoch": 0.03951367781155015,
+      "grad_norm": 0.07936527580022812,
+      "learning_rate": 0.0009871047835442364,
+      "loss": 0.8385,
+      "step": 91
+    },
+    {
+      "epoch": 0.03994789405123752,
+      "grad_norm": 0.08259500563144684,
+      "learning_rate": 0.0009867406052422524,
+      "loss": 0.8862,
+      "step": 92
+    },
+    {
+      "epoch": 0.04038211029092488,
+      "grad_norm": 0.09410865604877472,
+      "learning_rate": 0.0009863714249274552,
+      "loss": 0.8654,
+      "step": 93
+    },
+    {
+      "epoch": 0.04081632653061224,
+      "grad_norm": 0.08195285499095917,
+      "learning_rate": 0.000985997246393744,
+      "loss": 0.8616,
+      "step": 94
+    },
+    {
+      "epoch": 0.04125054277029961,
+      "grad_norm": 0.07515832781791687,
+      "learning_rate": 0.000985618073486382,
+      "loss": 0.858,
+      "step": 95
+    },
+    {
+      "epoch": 0.04168475900998697,
+      "grad_norm": 0.07317376881837845,
+      "learning_rate": 0.0009852339101019574,
+      "loss": 0.8699,
+      "step": 96
+    },
+    {
+      "epoch": 0.04211897524967434,
+      "grad_norm": 0.088593028485775,
+      "learning_rate": 0.0009848447601883434,
+      "loss": 0.8326,
+      "step": 97
+    },
+    {
+      "epoch": 0.0425531914893617,
+      "grad_norm": 0.07599225640296936,
+      "learning_rate": 0.0009844506277446577,
+      "loss": 0.7984,
+      "step": 98
+    },
+    {
+      "epoch": 0.04298740772904907,
+      "grad_norm": 0.08955889940261841,
+      "learning_rate": 0.0009840515168212207,
+      "loss": 0.8489,
+      "step": 99
+    },
+    {
+      "epoch": 0.04342162396873643,
+      "grad_norm": 0.07896170765161514,
+      "learning_rate": 0.0009836474315195148,
+      "loss": 0.8133,
+      "step": 100
+    },
+    {
+      "epoch": 0.04385584020842379,
+      "grad_norm": 0.08088196814060211,
+      "learning_rate": 0.0009832383759921415,
+      "loss": 0.8356,
+      "step": 101
+    },
+    {
+      "epoch": 0.04429005644811116,
+      "grad_norm": 0.08699057996273041,
+      "learning_rate": 0.0009828243544427796,
+      "loss": 0.8226,
+      "step": 102
+    },
+    {
+      "epoch": 0.04472427268779852,
+      "grad_norm": 0.08965849131345749,
+      "learning_rate": 0.0009824053711261404,
+      "loss": 0.8137,
+      "step": 103
+    },
+    {
+      "epoch": 0.04515848892748589,
+      "grad_norm": 0.09515663236379623,
+      "learning_rate": 0.0009819814303479266,
+      "loss": 0.7937,
+      "step": 104
+    },
+    {
+      "epoch": 0.04559270516717325,
+      "grad_norm": 0.06896204501390457,
+      "learning_rate": 0.0009815525364647853,
+      "loss": 0.8056,
+      "step": 105
+    },
+    {
+      "epoch": 0.04602692140686062,
+      "grad_norm": 0.08081001043319702,
+      "learning_rate": 0.0009811186938842645,
+      "loss": 0.8465,
+      "step": 106
+    },
+    {
+      "epoch": 0.04646113764654798,
+      "grad_norm": 0.07806787639856339,
+      "learning_rate": 0.000980679907064768,
+      "loss": 0.7897,
+      "step": 107
+    },
+    {
+      "epoch": 0.04689535388623534,
+      "grad_norm": 0.07646839320659637,
+      "learning_rate": 0.0009802361805155097,
+      "loss": 0.7929,
+      "step": 108
+    },
+    {
+      "epoch": 0.04732957012592271,
+      "grad_norm": 0.08139129728078842,
+      "learning_rate": 0.000979787518796466,
+      "loss": 0.8097,
+      "step": 109
+    },
+    {
+      "epoch": 0.04776378636561007,
+      "grad_norm": 0.07595842331647873,
+      "learning_rate": 0.0009793339265183304,
+      "loss": 0.8028,
+      "step": 110
+    },
+    {
+      "epoch": 0.04819800260529744,
+      "grad_norm": 0.07882706820964813,
+      "learning_rate": 0.0009788754083424652,
+      "loss": 0.7825,
+      "step": 111
+    },
+    {
+      "epoch": 0.0486322188449848,
+      "grad_norm": 0.09278842061758041,
+      "learning_rate": 0.0009784119689808543,
+      "loss": 0.7897,
+      "step": 112
+    },
+    {
+      "epoch": 0.049066435084672164,
+      "grad_norm": 0.07860872894525528,
+      "learning_rate": 0.0009779436131960543,
+      "loss": 0.8193,
+      "step": 113
+    },
+    {
+      "epoch": 0.04950065132435953,
+      "grad_norm": 0.09767181426286697,
+      "learning_rate": 0.0009774703458011453,
+      "loss": 0.7653,
+      "step": 114
+    },
+    {
+      "epoch": 0.04993486756404689,
+      "grad_norm": 0.09374384582042694,
+      "learning_rate": 0.000976992171659682,
+      "loss": 0.7541,
+      "step": 115
+    },
+    {
+      "epoch": 0.05036908380373426,
+      "grad_norm": 0.07632128894329071,
+      "learning_rate": 0.0009765090956856436,
+      "loss": 0.7518,
+      "step": 116
+    },
+    {
+      "epoch": 0.05080330004342162,
+      "grad_norm": 0.07544126361608505,
+      "learning_rate": 0.0009760211228433832,
+      "loss": 0.8217,
+      "step": 117
+    },
+    {
+      "epoch": 0.05123751628310899,
+      "grad_norm": 0.07379148155450821,
+      "learning_rate": 0.0009755282581475768,
+      "loss": 0.7902,
+      "step": 118
+    },
+    {
+      "epoch": 0.05167173252279635,
+      "grad_norm": 0.07656335830688477,
+      "learning_rate": 0.0009750305066631716,
+      "loss": 0.7357,
+      "step": 119
+    },
+    {
+      "epoch": 0.052105948762483714,
+      "grad_norm": 0.07699162513017654,
+      "learning_rate": 0.0009745278735053344,
+      "loss": 0.7329,
+      "step": 120
+    },
+    {
+      "epoch": 0.05254016500217108,
+      "grad_norm": 0.07921908050775528,
+      "learning_rate": 0.0009740203638393983,
+      "loss": 0.7368,
+      "step": 121
+    },
+    {
+      "epoch": 0.052974381241858444,
+      "grad_norm": 0.06974276900291443,
+      "learning_rate": 0.0009735079828808107,
+      "loss": 0.7593,
+      "step": 122
+    },
+    {
+      "epoch": 0.05340859748154581,
+      "grad_norm": 0.06961283832788467,
+      "learning_rate": 0.0009729907358950785,
+      "loss": 0.7899,
+      "step": 123
+    },
+    {
+      "epoch": 0.053842813721233174,
+      "grad_norm": 0.07279238104820251,
+      "learning_rate": 0.0009724686281977146,
+      "loss": 0.7455,
+      "step": 124
+    },
+    {
+      "epoch": 0.054277029960920535,
+      "grad_norm": 0.07039442658424377,
+      "learning_rate": 0.0009719416651541838,
+      "loss": 0.7755,
+      "step": 125
+    },
+    {
+      "epoch": 0.0547112462006079,
+      "grad_norm": 0.07618307322263718,
+      "learning_rate": 0.0009714098521798464,
+      "loss": 0.7631,
+      "step": 126
+    },
+    {
+      "epoch": 0.055145462440295265,
+      "grad_norm": 0.06525314599275589,
+      "learning_rate": 0.0009708731947399038,
+      "loss": 0.7605,
+      "step": 127
+    },
+    {
+      "epoch": 0.05557967867998263,
+      "grad_norm": 0.07441145926713943,
+      "learning_rate": 0.0009703316983493413,
+      "loss": 0.7402,
+      "step": 128
+    },
+    {
+      "epoch": 0.056013894919669995,
+      "grad_norm": 0.06484754383563995,
+      "learning_rate": 0.0009697853685728721,
+      "loss": 0.7238,
+      "step": 129
+    },
+    {
+      "epoch": 0.05644811115935736,
+      "grad_norm": 0.07414862513542175,
+      "learning_rate": 0.0009692342110248802,
+      "loss": 0.7633,
+      "step": 130
+    },
+    {
+      "epoch": 0.056882327399044724,
+      "grad_norm": 0.07470294088125229,
+      "learning_rate": 0.0009686782313693621,
+      "loss": 0.7467,
+      "step": 131
+    },
+    {
+      "epoch": 0.057316543638732086,
+      "grad_norm": 0.07042668759822845,
+      "learning_rate": 0.0009681174353198686,
+      "loss": 0.7309,
+      "step": 132
+    },
+    {
+      "epoch": 0.057750759878419454,
+      "grad_norm": 0.07646633684635162,
+      "learning_rate": 0.0009675518286394472,
+      "loss": 0.722,
+      "step": 133
+    },
+    {
+      "epoch": 0.058184976118106815,
+      "grad_norm": 0.07386059314012527,
+      "learning_rate": 0.0009669814171405816,
+      "loss": 0.7028,
+      "step": 134
+    },
+    {
+      "epoch": 0.058619192357794184,
+      "grad_norm": 0.07046350091695786,
+      "learning_rate": 0.0009664062066851325,
+      "loss": 0.711,
+      "step": 135
+    },
+    {
+      "epoch": 0.059053408597481545,
+      "grad_norm": 0.07790421694517136,
+      "learning_rate": 0.000965826203184277,
+      "loss": 0.7183,
+      "step": 136
+    },
+    {
+      "epoch": 0.05948762483716891,
+      "grad_norm": 0.08006292581558228,
+      "learning_rate": 0.000965241412598449,
+      "loss": 0.7534,
+      "step": 137
+    },
+    {
+      "epoch": 0.059921841076856275,
+      "grad_norm": 0.06924112141132355,
+      "learning_rate": 0.0009646518409372759,
+      "loss": 0.7401,
+      "step": 138
+    },
+    {
+      "epoch": 0.060356057316543636,
+      "grad_norm": 0.07095640152692795,
+      "learning_rate": 0.0009640574942595195,
+      "loss": 0.7056,
+      "step": 139
+    },
+    {
+      "epoch": 0.060790273556231005,
+      "grad_norm": 0.0727936401963234,
+      "learning_rate": 0.0009634583786730109,
+      "loss": 0.6811,
+      "step": 140
+    },
+    {
+      "epoch": 0.061224489795918366,
+      "grad_norm": 0.0759989321231842,
+      "learning_rate": 0.0009628545003345899,
+      "loss": 0.719,
+      "step": 141
+    },
+    {
+      "epoch": 0.061658706035605734,
+      "grad_norm": 0.07433141767978668,
+      "learning_rate": 0.0009622458654500408,
+      "loss": 0.724,
+      "step": 142
+    },
+    {
+      "epoch": 0.062092922275293096,
+      "grad_norm": 0.06923756748437881,
+      "learning_rate": 0.0009616324802740286,
+      "loss": 0.7562,
+      "step": 143
+    },
+    {
+      "epoch": 0.06252713851498046,
+      "grad_norm": 0.10080388188362122,
+      "learning_rate": 0.0009610143511100353,
+      "loss": 0.7116,
+      "step": 144
+    },
+    {
+      "epoch": 0.06296135475466783,
+      "grad_norm": 0.07774697989225388,
+      "learning_rate": 0.0009603914843102941,
+      "loss": 0.7277,
+      "step": 145
+    },
+    {
+      "epoch": 0.0633955709943552,
+      "grad_norm": 0.07297523319721222,
+      "learning_rate": 0.0009597638862757254,
+      "loss": 0.6817,
+      "step": 146
+    },
+    {
+      "epoch": 0.06382978723404255,
+      "grad_norm": 0.07876230776309967,
+      "learning_rate": 0.0009591315634558697,
+      "loss": 0.7015,
+      "step": 147
+    },
+    {
+      "epoch": 0.06426400347372992,
+      "grad_norm": 0.07196466624736786,
+      "learning_rate": 0.0009584945223488226,
+      "loss": 0.6995,
+      "step": 148
+    },
+    {
+      "epoch": 0.06469821971341728,
+      "grad_norm": 0.07307913899421692,
+      "learning_rate": 0.0009578527695011669,
+      "loss": 0.6931,
+      "step": 149
+    },
+    {
+      "epoch": 0.06513243595310464,
+      "grad_norm": 0.0770203247666359,
+      "learning_rate": 0.0009572063115079062,
+      "loss": 0.7135,
+      "step": 150
+    },
+    {
+      "epoch": 0.06556665219279201,
+      "grad_norm": 0.06902311742305756,
+      "learning_rate": 0.0009565551550123967,
+      "loss": 0.6906,
+      "step": 151
+    },
+    {
+      "epoch": 0.06600086843247938,
+      "grad_norm": 0.10059111565351486,
+      "learning_rate": 0.0009558993067062785,
+      "loss": 0.7237,
+      "step": 152
+    },
+    {
+      "epoch": 0.06643508467216674,
+      "grad_norm": 0.07180771976709366,
+      "learning_rate": 0.000955238773329408,
+      "loss": 0.6929,
+      "step": 153
+    },
+    {
+      "epoch": 0.0668693009118541,
+      "grad_norm": 0.07852017134428024,
+      "learning_rate": 0.0009545735616697875,
+      "loss": 0.6531,
+      "step": 154
+    },
+    {
+      "epoch": 0.06730351715154147,
+      "grad_norm": 0.07011571526527405,
+      "learning_rate": 0.0009539036785634961,
+      "loss": 0.6949,
+      "step": 155
+    },
+    {
+      "epoch": 0.06773773339122884,
+      "grad_norm": 0.07384242117404938,
+      "learning_rate": 0.000953229130894619,
+      "loss": 0.688,
+      "step": 156
+    },
+    {
+      "epoch": 0.06817194963091619,
+      "grad_norm": 0.07481442391872406,
+      "learning_rate": 0.0009525499255951775,
+      "loss": 0.6969,
+      "step": 157
+    },
+    {
+      "epoch": 0.06860616587060356,
+      "grad_norm": 0.06441578269004822,
+      "learning_rate": 0.0009518660696450568,
+      "loss": 0.6478,
+      "step": 158
+    },
+    {
+      "epoch": 0.06904038211029093,
+      "grad_norm": 0.06720245629549026,
+      "learning_rate": 0.0009511775700719346,
+      "loss": 0.677,
+      "step": 159
+    },
+    {
+      "epoch": 0.0694745983499783,
+      "grad_norm": 0.08833543956279755,
+      "learning_rate": 0.0009504844339512095,
+      "loss": 0.6912,
+      "step": 160
+    },
+    {
+      "epoch": 0.06990881458966565,
+      "grad_norm": 0.08247049152851105,
+      "learning_rate": 0.0009497866684059277,
+      "loss": 0.6833,
+      "step": 161
+    },
+    {
+      "epoch": 0.07034303082935302,
+      "grad_norm": 0.06711483001708984,
+      "learning_rate": 0.0009490842806067095,
+      "loss": 0.6695,
+      "step": 162
+    },
+    {
+      "epoch": 0.07077724706904039,
+      "grad_norm": 0.06512755900621414,
+      "learning_rate": 0.0009483772777716766,
+      "loss": 0.695,
+      "step": 163
+    },
+    {
+      "epoch": 0.07121146330872774,
+      "grad_norm": 0.06651882827281952,
+      "learning_rate": 0.0009476656671663766,
+      "loss": 0.6608,
+      "step": 164
+    },
+    {
+      "epoch": 0.07164567954841511,
+      "grad_norm": 0.06416794657707214,
+      "learning_rate": 0.0009469494561037098,
+      "loss": 0.6918,
+      "step": 165
+    },
+    {
+      "epoch": 0.07207989578810248,
+      "grad_norm": 0.06808705627918243,
+      "learning_rate": 0.000946228651943853,
+      "loss": 0.6959,
+      "step": 166
+    },
+    {
+      "epoch": 0.07251411202778985,
+      "grad_norm": 0.077076256275177,
+      "learning_rate": 0.0009455032620941839,
+      "loss": 0.6737,
+      "step": 167
+    },
+    {
+      "epoch": 0.0729483282674772,
+      "grad_norm": 0.09289199858903885,
+      "learning_rate": 0.000944773294009206,
+      "loss": 0.651,
+      "step": 168
+    },
+    {
+      "epoch": 0.07338254450716457,
+      "grad_norm": 0.07365357875823975,
+      "learning_rate": 0.0009440387551904703,
+      "loss": 0.6778,
+      "step": 169
+    },
+    {
+      "epoch": 0.07381676074685194,
+      "grad_norm": 0.0635438859462738,
+      "learning_rate": 0.0009432996531865001,
+      "loss": 0.7081,
+      "step": 170
+    },
+    {
+      "epoch": 0.07425097698653929,
+      "grad_norm": 0.07002374529838562,
+      "learning_rate": 0.0009425559955927117,
+      "loss": 0.6505,
+      "step": 171
+    },
+    {
+      "epoch": 0.07468519322622666,
+      "grad_norm": 0.08253403007984161,
+      "learning_rate": 0.0009418077900513376,
+      "loss": 0.6752,
+      "step": 172
+    },
+    {
+      "epoch": 0.07511940946591403,
+      "grad_norm": 0.06753169745206833,
+      "learning_rate": 0.0009410550442513475,
+      "loss": 0.6695,
+      "step": 173
+    },
+    {
+      "epoch": 0.0755536257056014,
+      "grad_norm": 0.06907963007688522,
+      "learning_rate": 0.000940297765928369,
+      "loss": 0.6441,
+      "step": 174
+    },
+    {
+      "epoch": 0.07598784194528875,
+      "grad_norm": 0.0881882980465889,
+      "learning_rate": 0.0009395359628646086,
+      "loss": 0.6456,
+      "step": 175
+    },
+    {
+      "epoch": 0.07642205818497612,
+      "grad_norm": 0.06529023498296738,
+      "learning_rate": 0.0009387696428887715,
+      "loss": 0.6689,
+      "step": 176
+    },
+    {
+      "epoch": 0.07685627442466349,
+      "grad_norm": 0.06664346903562546,
+      "learning_rate": 0.0009379988138759809,
+      "loss": 0.6485,
+      "step": 177
+    },
+    {
+      "epoch": 0.07729049066435084,
+      "grad_norm": 0.06951016932725906,
+      "learning_rate": 0.0009372234837476979,
+      "loss": 0.68,
+      "step": 178
+    },
+    {
+      "epoch": 0.07772470690403821,
+      "grad_norm": 0.060999054461717606,
+      "learning_rate": 0.0009364436604716389,
+      "loss": 0.6706,
+      "step": 179
+    },
+    {
+      "epoch": 0.07815892314372558,
+      "grad_norm": 0.07197262346744537,
+      "learning_rate": 0.0009356593520616947,
+      "loss": 0.6277,
+      "step": 180
+    },
+    {
+      "epoch": 0.07859313938341293,
+      "grad_norm": 0.06676187366247177,
+      "learning_rate": 0.0009348705665778478,
+      "loss": 0.6492,
+      "step": 181
+    },
+    {
+      "epoch": 0.0790273556231003,
+      "grad_norm": 0.07378092408180237,
+      "learning_rate": 0.0009340773121260893,
+      "loss": 0.6375,
+      "step": 182
+    },
+    {
+      "epoch": 0.07946157186278767,
+      "grad_norm": 0.06367610394954681,
+      "learning_rate": 0.000933279596858336,
+      "loss": 0.6221,
+      "step": 183
+    },
+    {
+      "epoch": 0.07989578810247504,
+      "grad_norm": 0.0669325664639473,
+      "learning_rate": 0.0009324774289723468,
+      "loss": 0.6294,
+      "step": 184
+    },
+    {
+      "epoch": 0.08033000434216239,
+      "grad_norm": 0.06983687728643417,
+      "learning_rate": 0.0009316708167116377,
+      "loss": 0.6487,
+      "step": 185
+    },
+    {
+      "epoch": 0.08076422058184976,
+      "grad_norm": 0.06569714099168777,
+      "learning_rate": 0.0009308597683653976,
+      "loss": 0.651,
+      "step": 186
+    },
+    {
+      "epoch": 0.08119843682153713,
+      "grad_norm": 0.06825820356607437,
+      "learning_rate": 0.0009300442922684032,
+      "loss": 0.6193,
+      "step": 187
+    },
+    {
+      "epoch": 0.08163265306122448,
+      "grad_norm": 0.07318625599145889,
+      "learning_rate": 0.000929224396800933,
+      "loss": 0.7024,
+      "step": 188
+    },
+    {
+      "epoch": 0.08206686930091185,
+      "grad_norm": 0.1047971323132515,
+      "learning_rate": 0.0009284000903886818,
+      "loss": 0.6637,
+      "step": 189
+    },
+    {
+      "epoch": 0.08250108554059922,
+      "grad_norm": 0.06583527475595474,
+      "learning_rate": 0.0009275713815026732,
+      "loss": 0.6419,
+      "step": 190
+    },
+    {
+      "epoch": 0.08293530178028659,
+      "grad_norm": 0.09143144637346268,
+      "learning_rate": 0.000926738278659173,
+      "loss": 0.6295,
+      "step": 191
+    },
+    {
+      "epoch": 0.08336951801997394,
+      "grad_norm": 0.06756502389907837,
+      "learning_rate": 0.0009259007904196022,
+      "loss": 0.6718,
+      "step": 192
+    },
+    {
+      "epoch": 0.08380373425966131,
+      "grad_norm": 0.06964406371116638,
+      "learning_rate": 0.000925058925390448,
+      "loss": 0.6542,
+      "step": 193
+    },
+    {
+      "epoch": 0.08423795049934868,
+      "grad_norm": 0.07476377487182617,
+      "learning_rate": 0.0009242126922231762,
+      "loss": 0.6266,
+      "step": 194
+    },
+    {
+      "epoch": 0.08467216673903603,
+      "grad_norm": 0.07210344076156616,
+      "learning_rate": 0.0009233620996141421,
+      "loss": 0.6721,
+      "step": 195
+    },
+    {
+      "epoch": 0.0851063829787234,
+      "grad_norm": 0.07828934490680695,
+      "learning_rate": 0.0009225071563045006,
+      "loss": 0.6223,
+      "step": 196
+    },
+    {
+      "epoch": 0.08554059921841077,
+      "grad_norm": 0.07536309212446213,
+      "learning_rate": 0.000921647871080117,
+      "loss": 0.6563,
+      "step": 197
+    },
+    {
+      "epoch": 0.08597481545809814,
+      "grad_norm": 0.07678169757127762,
+      "learning_rate": 0.0009207842527714766,
+      "loss": 0.6646,
+      "step": 198
+    },
+    {
+      "epoch": 0.0864090316977855,
+      "grad_norm": 0.07808982580900192,
+      "learning_rate": 0.0009199163102535937,
+      "loss": 0.6497,
+      "step": 199
+    },
+    {
+      "epoch": 0.08684324793747286,
+      "grad_norm": 0.07775306701660156,
+      "learning_rate": 0.0009190440524459203,
+      "loss": 0.6083,
+      "step": 200
+    },
+    {
+      "epoch": 0.08727746417716023,
+      "grad_norm": 0.06569099426269531,
+      "learning_rate": 0.0009181674883122553,
+      "loss": 0.6424,
+      "step": 201
+    },
+    {
+      "epoch": 0.08771168041684758,
+      "grad_norm": 0.08122856914997101,
+      "learning_rate": 0.0009172866268606513,
+      "loss": 0.6054,
+      "step": 202
+    },
+    {
+      "epoch": 0.08814589665653495,
+      "grad_norm": 0.08045452833175659,
+      "learning_rate": 0.0009164014771433226,
+      "loss": 0.6372,
+      "step": 203
+    },
+    {
+      "epoch": 0.08858011289622232,
+      "grad_norm": 0.0744997188448906,
+      "learning_rate": 0.000915512048256552,
+      "loss": 0.6341,
+      "step": 204
+    },
+    {
+      "epoch": 0.08901432913590969,
+      "grad_norm": 0.07247951626777649,
+      "learning_rate": 0.0009146183493405975,
+      "loss": 0.5922,
+      "step": 205
+    },
+    {
+      "epoch": 0.08944854537559704,
+      "grad_norm": 0.08652166277170181,
+      "learning_rate": 0.0009137203895795982,
+      "loss": 0.6111,
+      "step": 206
+    },
+    {
+      "epoch": 0.08988276161528441,
+      "grad_norm": 0.06636743992567062,
+      "learning_rate": 0.0009128181782014801,
+      "loss": 0.6287,
+      "step": 207
+    },
+    {
+      "epoch": 0.09031697785497178,
+      "grad_norm": 0.08643815666437149,
+      "learning_rate": 0.0009119117244778608,
+      "loss": 0.6091,
+      "step": 208
+    },
+    {
+      "epoch": 0.09075119409465914,
+      "grad_norm": 0.08439099788665771,
+      "learning_rate": 0.0009110010377239551,
+      "loss": 0.6735,
+      "step": 209
+    },
+    {
+      "epoch": 0.0911854103343465,
+      "grad_norm": 0.0863455981016159,
+      "learning_rate": 0.0009100861272984779,
+      "loss": 0.6049,
+      "step": 210
+    },
+    {
+      "epoch": 0.09161962657403387,
+      "grad_norm": 0.06976283341646194,
+      "learning_rate": 0.0009091670026035499,
+      "loss": 0.5934,
+      "step": 211
+    },
+    {
+      "epoch": 0.09205384281372124,
+      "grad_norm": 0.10158892720937729,
+      "learning_rate": 0.0009082436730845993,
+      "loss": 0.6314,
+      "step": 212
+    },
+    {
+      "epoch": 0.0924880590534086,
+      "grad_norm": 0.06954232603311539,
+      "learning_rate": 0.0009073161482302654,
+      "loss": 0.6461,
+      "step": 213
+    },
+    {
+      "epoch": 0.09292227529309596,
+      "grad_norm": 0.06838048249483109,
+      "learning_rate": 0.0009063844375723014,
+      "loss": 0.5993,
+      "step": 214
+    },
+    {
+      "epoch": 0.09335649153278333,
+      "grad_norm": 0.06874913722276688,
+      "learning_rate": 0.0009054485506854755,
+      "loss": 0.6035,
+      "step": 215
+    },
+    {
+      "epoch": 0.09379070777247069,
+      "grad_norm": 0.07478207349777222,
+      "learning_rate": 0.0009045084971874737,
+      "loss": 0.6123,
+      "step": 216
+    },
+    {
+      "epoch": 0.09422492401215805,
+      "grad_norm": 0.06408224254846573,
+      "learning_rate": 0.0009035642867388002,
+      "loss": 0.6242,
+      "step": 217
+    },
+    {
+      "epoch": 0.09465914025184542,
+      "grad_norm": 0.07433444261550903,
+      "learning_rate": 0.0009026159290426781,
+      "loss": 0.6055,
+      "step": 218
+    },
+    {
+      "epoch": 0.09509335649153278,
+      "grad_norm": 0.06954223662614822,
+      "learning_rate": 0.0009016634338449503,
+      "loss": 0.6257,
+      "step": 219
+    },
+    {
+      "epoch": 0.09552757273122015,
+      "grad_norm": 0.06509328633546829,
+      "learning_rate": 0.0009007068109339783,
+      "loss": 0.5749,
+      "step": 220
+    },
+    {
+      "epoch": 0.09596178897090751,
+      "grad_norm": 0.08184398710727692,
+      "learning_rate": 0.000899746070140543,
+      "loss": 0.6166,
+      "step": 221
+    },
+    {
+      "epoch": 0.09639600521059488,
+      "grad_norm": 0.06200498715043068,
+      "learning_rate": 0.0008987812213377423,
+      "loss": 0.615,
+      "step": 222
+    },
+    {
+      "epoch": 0.09683022145028224,
+      "grad_norm": 0.06602746993303299,
+      "learning_rate": 0.0008978122744408905,
+      "loss": 0.6087,
+      "step": 223
+    },
+    {
+      "epoch": 0.0972644376899696,
+      "grad_norm": 0.0869906097650528,
+      "learning_rate": 0.0008968392394074163,
+      "loss": 0.5725,
+      "step": 224
+    },
+    {
+      "epoch": 0.09769865392965697,
+      "grad_norm": 0.064040407538414,
+      "learning_rate": 0.0008958621262367599,
+      "loss": 0.6116,
+      "step": 225
+    },
+    {
+      "epoch": 0.09813287016934433,
+      "grad_norm": 0.0657719075679779,
+      "learning_rate": 0.0008948809449702712,
+      "loss": 0.6219,
+      "step": 226
+    },
+    {
+      "epoch": 0.0985670864090317,
+      "grad_norm": 0.07608803361654282,
+      "learning_rate": 0.0008938957056911057,
+      "loss": 0.6018,
+      "step": 227
+    },
+    {
+      "epoch": 0.09900130264871906,
+      "grad_norm": 0.07249824702739716,
+      "learning_rate": 0.0008929064185241212,
+      "loss": 0.6027,
+      "step": 228
+    },
+    {
+      "epoch": 0.09943551888840643,
+      "grad_norm": 0.06748249381780624,
+      "learning_rate": 0.0008919130936357742,
+      "loss": 0.6032,
+      "step": 229
+    },
+    {
+      "epoch": 0.09986973512809379,
+      "grad_norm": 0.08020893484354019,
+      "learning_rate": 0.000890915741234015,
+      "loss": 0.607,
+      "step": 230
+    },
+    {
+      "epoch": 0.10030395136778116,
+      "grad_norm": 0.0669545903801918,
+      "learning_rate": 0.0008899143715681822,
+      "loss": 0.5974,
+      "step": 231
+    },
+    {
+      "epoch": 0.10073816760746852,
+      "grad_norm": 0.06186607852578163,
+      "learning_rate": 0.0008889089949288987,
+      "loss": 0.6186,
+      "step": 232
+    },
+    {
+      "epoch": 0.10117238384715588,
+      "grad_norm": 0.06386630982160568,
+      "learning_rate": 0.0008878996216479651,
+      "loss": 0.5995,
+      "step": 233
+    },
+    {
+      "epoch": 0.10160660008684325,
+      "grad_norm": 0.06557858735322952,
+      "learning_rate": 0.0008868862620982534,
+      "loss": 0.6066,
+      "step": 234
+    },
+    {
+      "epoch": 0.10204081632653061,
+      "grad_norm": 0.06579785794019699,
+      "learning_rate": 0.0008858689266936008,
+      "loss": 0.6144,
+      "step": 235
+    },
+    {
+      "epoch": 0.10247503256621798,
+      "grad_norm": 0.06238356605172157,
+      "learning_rate": 0.0008848476258887031,
+      "loss": 0.5899,
+      "step": 236
+    },
+    {
+      "epoch": 0.10290924880590534,
+      "grad_norm": 0.061628226190805435,
+      "learning_rate": 0.0008838223701790055,
+      "loss": 0.6092,
+      "step": 237
+    },
+    {
+      "epoch": 0.1033434650455927,
+      "grad_norm": 0.06545189768075943,
+      "learning_rate": 0.0008827931701005973,
+      "loss": 0.6014,
+      "step": 238
+    },
+    {
+      "epoch": 0.10377768128528007,
+      "grad_norm": 0.056564707309007645,
+      "learning_rate": 0.0008817600362301017,
+      "loss": 0.5959,
+      "step": 239
+    },
+    {
+      "epoch": 0.10421189752496743,
+      "grad_norm": 0.0715203583240509,
+      "learning_rate": 0.0008807229791845672,
+      "loss": 0.5942,
+      "step": 240
+    },
+    {
+      "epoch": 0.1046461137646548,
+      "grad_norm": 0.06504496932029724,
+      "learning_rate": 0.00087968200962136,
+      "loss": 0.5652,
+      "step": 241
+    },
+    {
+      "epoch": 0.10508033000434217,
+      "grad_norm": 0.06740335375070572,
+      "learning_rate": 0.0008786371382380527,
+      "loss": 0.6159,
+      "step": 242
+    },
+    {
+      "epoch": 0.10551454624402953,
+      "grad_norm": 0.06163683906197548,
+      "learning_rate": 0.0008775883757723155,
+      "loss": 0.5913,
+      "step": 243
+    },
+    {
+      "epoch": 0.10594876248371689,
+      "grad_norm": 0.07384572178125381,
+      "learning_rate": 0.0008765357330018055,
+      "loss": 0.5829,
+      "step": 244
+    },
+    {
+      "epoch": 0.10638297872340426,
+      "grad_norm": 0.06352236866950989,
+      "learning_rate": 0.0008754792207440556,
+      "loss": 0.6069,
+      "step": 245
+    },
+    {
+      "epoch": 0.10681719496309162,
+      "grad_norm": 0.0605890229344368,
+      "learning_rate": 0.0008744188498563641,
+      "loss": 0.5685,
+      "step": 246
+    },
+    {
+      "epoch": 0.10725141120277898,
+      "grad_norm": 0.06994660943746567,
+      "learning_rate": 0.0008733546312356824,
+      "loss": 0.581,
+      "step": 247
+    },
+    {
+      "epoch": 0.10768562744246635,
+      "grad_norm": 0.08128967881202698,
+      "learning_rate": 0.0008722865758185036,
+      "loss": 0.5754,
+      "step": 248
+    },
+    {
+      "epoch": 0.10811984368215372,
+      "grad_norm": 0.06497329473495483,
+      "learning_rate": 0.0008712146945807493,
+      "loss": 0.598,
+      "step": 249
+    },
+    {
+      "epoch": 0.10855405992184107,
+      "grad_norm": 0.06091492623090744,
+      "learning_rate": 0.0008701389985376577,
+      "loss": 0.5807,
+      "step": 250
+    },
+    {
+      "epoch": 0.10898827616152844,
+      "grad_norm": 0.06555697321891785,
+      "learning_rate": 0.0008690594987436704,
+      "loss": 0.5768,
+      "step": 251
+    },
+    {
+      "epoch": 0.1094224924012158,
+      "grad_norm": 0.07119280844926834,
+      "learning_rate": 0.0008679762062923176,
+      "loss": 0.6088,
+      "step": 252
+    },
+    {
+      "epoch": 0.10985670864090318,
+      "grad_norm": 0.06837710738182068,
+      "learning_rate": 0.0008668891323161053,
+      "loss": 0.5905,
+      "step": 253
+    },
+    {
+      "epoch": 0.11029092488059053,
+      "grad_norm": 0.067451111972332,
+      "learning_rate": 0.0008657982879864007,
+      "loss": 0.5974,
+      "step": 254
+    },
+    {
+      "epoch": 0.1107251411202779,
+      "grad_norm": 0.06649015098810196,
+      "learning_rate": 0.0008647036845133172,
+      "loss": 0.5684,
+      "step": 255
+    },
+    {
+      "epoch": 0.11115935735996527,
+      "grad_norm": 0.06311678886413574,
+      "learning_rate": 0.0008636053331455986,
+      "loss": 0.58,
+      "step": 256
+    },
+    {
+      "epoch": 0.11159357359965262,
+      "grad_norm": 0.06087026745080948,
+      "learning_rate": 0.0008625032451705052,
+      "loss": 0.5735,
+      "step": 257
+    },
+    {
+      "epoch": 0.11202778983933999,
+      "grad_norm": 0.061150964349508286,
+      "learning_rate": 0.0008613974319136957,
+      "loss": 0.6044,
+      "step": 258
+    },
+    {
+      "epoch": 0.11246200607902736,
+      "grad_norm": 0.06023947522044182,
+      "learning_rate": 0.0008602879047391126,
+      "loss": 0.551,
+      "step": 259
+    },
+    {
+      "epoch": 0.11289622231871473,
+      "grad_norm": 0.06919502466917038,
+      "learning_rate": 0.0008591746750488638,
+      "loss": 0.5744,
+      "step": 260
+    },
+    {
+      "epoch": 0.11333043855840208,
+      "grad_norm": 0.06585050374269485,
+      "learning_rate": 0.0008580577542831072,
+      "loss": 0.6053,
+      "step": 261
+    },
+    {
+      "epoch": 0.11376465479808945,
+      "grad_norm": 0.06967171281576157,
+      "learning_rate": 0.0008569371539199316,
+      "loss": 0.5896,
+      "step": 262
+    },
+    {
+      "epoch": 0.11419887103777682,
+      "grad_norm": 0.062438324093818665,
+      "learning_rate": 0.0008558128854752396,
+      "loss": 0.585,
+      "step": 263
+    },
+    {
+      "epoch": 0.11463308727746417,
+      "grad_norm": 0.06600722670555115,
+      "learning_rate": 0.0008546849605026289,
+      "loss": 0.5809,
+      "step": 264
+    },
+    {
+      "epoch": 0.11506730351715154,
+      "grad_norm": 0.06594278663396835,
+      "learning_rate": 0.0008535533905932737,
+      "loss": 0.5854,
+      "step": 265
+    },
+    {
+      "epoch": 0.11550151975683891,
+      "grad_norm": 0.06402723491191864,
+      "learning_rate": 0.0008524181873758059,
+      "loss": 0.5612,
+      "step": 266
+    },
+    {
+      "epoch": 0.11593573599652628,
+      "grad_norm": 0.06593722105026245,
+      "learning_rate": 0.0008512793625161946,
+      "loss": 0.5825,
+      "step": 267
+    },
+    {
+      "epoch": 0.11636995223621363,
+      "grad_norm": 0.057960718870162964,
+      "learning_rate": 0.0008501369277176275,
+      "loss": 0.5972,
+      "step": 268
+    },
+    {
+      "epoch": 0.116804168475901,
+      "grad_norm": 0.06126386299729347,
+      "learning_rate": 0.0008489908947203897,
+      "loss": 0.5675,
+      "step": 269
+    },
+    {
+      "epoch": 0.11723838471558837,
+      "grad_norm": 0.06619458645582199,
+      "learning_rate": 0.0008478412753017432,
+      "loss": 0.5724,
+      "step": 270
+    },
+    {
+      "epoch": 0.11767260095527572,
+      "grad_norm": 0.05950764939188957,
+      "learning_rate": 0.0008466880812758065,
+      "loss": 0.5732,
+      "step": 271
+    },
+    {
+      "epoch": 0.11810681719496309,
+      "grad_norm": 0.0596248060464859,
+      "learning_rate": 0.0008455313244934324,
+      "loss": 0.5844,
+      "step": 272
+    },
+    {
+      "epoch": 0.11854103343465046,
+      "grad_norm": 0.06557915359735489,
+      "learning_rate": 0.0008443710168420865,
+      "loss": 0.5706,
+      "step": 273
+    },
+    {
+      "epoch": 0.11897524967433783,
+      "grad_norm": 0.06171106547117233,
+      "learning_rate": 0.0008432071702457253,
+      "loss": 0.561,
+      "step": 274
+    },
+    {
+      "epoch": 0.11940946591402518,
+      "grad_norm": 0.06009051948785782,
+      "learning_rate": 0.0008420397966646731,
+      "loss": 0.5764,
+      "step": 275
+    },
+    {
+      "epoch": 0.11984368215371255,
+      "grad_norm": 0.06843849271535873,
+      "learning_rate": 0.0008408689080954998,
+      "loss": 0.5833,
+      "step": 276
+    },
+    {
+      "epoch": 0.12027789839339992,
+      "grad_norm": 0.06061152368783951,
+      "learning_rate": 0.0008396945165708972,
+      "loss": 0.5962,
+      "step": 277
+    },
+    {
+      "epoch": 0.12071211463308727,
+      "grad_norm": 0.05997680127620697,
+      "learning_rate": 0.0008385166341595548,
+      "loss": 0.5738,
+      "step": 278
+    },
+    {
+      "epoch": 0.12114633087277464,
+      "grad_norm": 0.06045431271195412,
+      "learning_rate": 0.0008373352729660373,
+      "loss": 0.5738,
+      "step": 279
+    },
+    {
+      "epoch": 0.12158054711246201,
+      "grad_norm": 0.06007954850792885,
+      "learning_rate": 0.0008361504451306584,
+      "loss": 0.5753,
+      "step": 280
+    },
+    {
+      "epoch": 0.12201476335214936,
+      "grad_norm": 0.05806737020611763,
+      "learning_rate": 0.0008349621628293577,
+      "loss": 0.5703,
+      "step": 281
+    },
+    {
+      "epoch": 0.12244897959183673,
+      "grad_norm": 0.06177588179707527,
+      "learning_rate": 0.0008337704382735741,
+      "loss": 0.5533,
+      "step": 282
+    },
+    {
+      "epoch": 0.1228831958315241,
+      "grad_norm": 0.06487903743982315,
+      "learning_rate": 0.0008325752837101213,
+      "loss": 0.5655,
+      "step": 283
+    },
+    {
+      "epoch": 0.12331741207121147,
+      "grad_norm": 0.06531866639852524,
+      "learning_rate": 0.0008313767114210615,
+      "loss": 0.5694,
+      "step": 284
+    },
+    {
+      "epoch": 0.12375162831089882,
+      "grad_norm": 0.08110279589891434,
+      "learning_rate": 0.0008301747337235797,
+      "loss": 0.5517,
+      "step": 285
+    },
+    {
+      "epoch": 0.12418584455058619,
+      "grad_norm": 0.06367824971675873,
+      "learning_rate": 0.0008289693629698564,
+      "loss": 0.5474,
+      "step": 286
+    },
+    {
+      "epoch": 0.12462006079027356,
+      "grad_norm": 0.07917732000350952,
+      "learning_rate": 0.0008277606115469409,
+      "loss": 0.5685,
+      "step": 287
+    },
+    {
+      "epoch": 0.12505427702996091,
+      "grad_norm": 0.06585974991321564,
+      "learning_rate": 0.0008265484918766243,
+      "loss": 0.5547,
+      "step": 288
+    },
+    {
+      "epoch": 0.12548849326964828,
+      "grad_norm": 0.0627543181180954,
+      "learning_rate": 0.0008253330164153117,
+      "loss": 0.5795,
+      "step": 289
+    },
+    {
+      "epoch": 0.12592270950933565,
+      "grad_norm": 0.06639768183231354,
+      "learning_rate": 0.0008241141976538943,
+      "loss": 0.5669,
+      "step": 290
+    },
+    {
+      "epoch": 0.12635692574902302,
+      "grad_norm": 0.06217719987034798,
+      "learning_rate": 0.0008228920481176202,
+      "loss": 0.5552,
+      "step": 291
+    },
+    {
+      "epoch": 0.1267911419887104,
+      "grad_norm": 0.06806106120347977,
+      "learning_rate": 0.0008216665803659671,
+      "loss": 0.5553,
+      "step": 292
+    },
+    {
+      "epoch": 0.12722535822839776,
+      "grad_norm": 0.058773159980773926,
+      "learning_rate": 0.000820437806992512,
+      "loss": 0.5678,
+      "step": 293
+    },
+    {
+      "epoch": 0.1276595744680851,
+      "grad_norm": 0.06487017124891281,
+      "learning_rate": 0.0008192057406248028,
+      "loss": 0.5493,
+      "step": 294
+    },
+    {
+      "epoch": 0.12809379070777246,
+      "grad_norm": 0.05514706298708916,
+      "learning_rate": 0.0008179703939242276,
+      "loss": 0.5834,
+      "step": 295
+    },
+    {
+      "epoch": 0.12852800694745983,
+      "grad_norm": 0.05893951281905174,
+      "learning_rate": 0.0008167317795858851,
+      "loss": 0.5668,
+      "step": 296
+    },
+    {
+      "epoch": 0.1289622231871472,
+      "grad_norm": 0.06080016866326332,
+      "learning_rate": 0.0008154899103384537,
+      "loss": 0.6112,
+      "step": 297
+    },
+    {
+      "epoch": 0.12939643942683457,
+      "grad_norm": 0.06676195561885834,
+      "learning_rate": 0.0008142447989440618,
+      "loss": 0.5897,
+      "step": 298
+    },
+    {
+      "epoch": 0.12983065566652194,
+      "grad_norm": 0.055440496653318405,
+      "learning_rate": 0.0008129964581981553,
+      "loss": 0.5199,
+      "step": 299
+    },
+    {
+      "epoch": 0.13026487190620928,
+      "grad_norm": 0.0642077848315239,
+      "learning_rate": 0.0008117449009293668,
+      "loss": 0.5688,
+      "step": 300
+    },
+    {
+      "epoch": 0.13069908814589665,
+      "grad_norm": 0.056085068732500076,
+      "learning_rate": 0.0008104901399993836,
+      "loss": 0.5626,
+      "step": 301
+    },
+    {
+      "epoch": 0.13113330438558402,
+      "grad_norm": 0.07240674644708633,
+      "learning_rate": 0.0008092321883028157,
+      "loss": 0.5705,
+      "step": 302
+    },
+    {
+      "epoch": 0.13156752062527138,
+      "grad_norm": 0.06402257084846497,
+      "learning_rate": 0.0008079710587670633,
+      "loss": 0.5608,
+      "step": 303
+    },
+    {
+      "epoch": 0.13200173686495875,
+      "grad_norm": 0.05889822542667389,
+      "learning_rate": 0.0008067067643521834,
+      "loss": 0.5552,
+      "step": 304
+    },
+    {
+      "epoch": 0.13243595310464612,
+      "grad_norm": 0.07787332683801651,
+      "learning_rate": 0.0008054393180507572,
+      "loss": 0.5526,
+      "step": 305
+    },
+    {
+      "epoch": 0.1328701693443335,
+      "grad_norm": 0.06866924464702606,
+      "learning_rate": 0.0008041687328877566,
+      "loss": 0.5726,
+      "step": 306
+    },
+    {
+      "epoch": 0.13330438558402083,
+      "grad_norm": 0.06494660675525665,
+      "learning_rate": 0.00080289502192041,
+      "loss": 0.5605,
+      "step": 307
+    },
+    {
+      "epoch": 0.1337386018237082,
+      "grad_norm": 0.06333691626787186,
+      "learning_rate": 0.0008016181982380681,
+      "loss": 0.5668,
+      "step": 308
+    },
+    {
+      "epoch": 0.13417281806339557,
+      "grad_norm": 0.05925397574901581,
+      "learning_rate": 0.0008003382749620702,
+      "loss": 0.539,
+      "step": 309
+    },
+    {
+      "epoch": 0.13460703430308293,
+      "grad_norm": 0.06211424991488457,
+      "learning_rate": 0.000799055265245608,
+      "loss": 0.5581,
+      "step": 310
+    },
+    {
+      "epoch": 0.1350412505427703,
+      "grad_norm": 0.06610213220119476,
+      "learning_rate": 0.0007977691822735914,
+      "loss": 0.5726,
+      "step": 311
+    },
+    {
+      "epoch": 0.13547546678245767,
+      "grad_norm": 0.06418436765670776,
+      "learning_rate": 0.0007964800392625129,
+      "loss": 0.5616,
+      "step": 312
+    },
+    {
+      "epoch": 0.13590968302214504,
+      "grad_norm": 0.05993996933102608,
+      "learning_rate": 0.0007951878494603115,
+      "loss": 0.5392,
+      "step": 313
+    },
+    {
+      "epoch": 0.13634389926183238,
+      "grad_norm": 0.05776641145348549,
+      "learning_rate": 0.0007938926261462366,
+      "loss": 0.5402,
+      "step": 314
+    },
+    {
+      "epoch": 0.13677811550151975,
+      "grad_norm": 0.06915484368801117,
+      "learning_rate": 0.0007925943826307118,
+      "loss": 0.5336,
+      "step": 315
+    },
+    {
+      "epoch": 0.13721233174120712,
+      "grad_norm": 0.06394239515066147,
+      "learning_rate": 0.0007912931322551981,
+      "loss": 0.5647,
+      "step": 316
+    },
+    {
+      "epoch": 0.13764654798089448,
+      "grad_norm": 0.05997829511761665,
+      "learning_rate": 0.000789988888392056,
+      "loss": 0.5588,
+      "step": 317
+    },
+    {
+      "epoch": 0.13808076422058185,
+      "grad_norm": 0.07337518781423569,
+      "learning_rate": 0.0007886816644444098,
+      "loss": 0.5825,
+      "step": 318
+    },
+    {
+      "epoch": 0.13851498046026922,
+      "grad_norm": 0.07288537919521332,
+      "learning_rate": 0.0007873714738460075,
+      "loss": 0.5615,
+      "step": 319
+    },
+    {
+      "epoch": 0.1389491966999566,
+      "grad_norm": 0.06601582467556,
+      "learning_rate": 0.0007860583300610849,
+      "loss": 0.5545,
+      "step": 320
+    },
+    {
+      "epoch": 0.13938341293964393,
+      "grad_norm": 0.06303299218416214,
+      "learning_rate": 0.000784742246584226,
+      "loss": 0.5579,
+      "step": 321
+    },
+    {
+      "epoch": 0.1398176291793313,
+      "grad_norm": 0.06565733999013901,
+      "learning_rate": 0.000783423236940225,
+      "loss": 0.5728,
+      "step": 322
+    },
+    {
+      "epoch": 0.14025184541901867,
+      "grad_norm": 0.06590873748064041,
+      "learning_rate": 0.0007821013146839467,
+      "loss": 0.5552,
+      "step": 323
+    },
+    {
+      "epoch": 0.14068606165870604,
+      "grad_norm": 0.06782017648220062,
+      "learning_rate": 0.0007807764934001874,
+      "loss": 0.5299,
+      "step": 324
+    },
+    {
+      "epoch": 0.1411202778983934,
+      "grad_norm": 0.06369118392467499,
+      "learning_rate": 0.0007794487867035358,
+      "loss": 0.524,
+      "step": 325
+    },
+    {
+      "epoch": 0.14155449413808077,
+      "grad_norm": 0.06337650865316391,
+      "learning_rate": 0.0007781182082382324,
+      "loss": 0.5802,
+      "step": 326
+    },
+    {
+      "epoch": 0.14198871037776814,
+      "grad_norm": 0.07011637836694717,
+      "learning_rate": 0.0007767847716780296,
+      "loss": 0.5569,
+      "step": 327
+    },
+    {
+      "epoch": 0.14242292661745548,
+      "grad_norm": 0.06176460534334183,
+      "learning_rate": 0.0007754484907260512,
+      "loss": 0.5266,
+      "step": 328
+    },
+    {
+      "epoch": 0.14285714285714285,
+      "grad_norm": 0.06650613993406296,
+      "learning_rate": 0.0007741093791146516,
+      "loss": 0.5628,
+      "step": 329
+    },
+    {
+      "epoch": 0.14329135909683022,
+      "grad_norm": 0.07864295691251755,
+      "learning_rate": 0.0007727674506052743,
+      "loss": 0.5683,
+      "step": 330
+    },
+    {
+      "epoch": 0.14372557533651759,
+      "grad_norm": 0.06307731568813324,
+      "learning_rate": 0.0007714227189883112,
+      "loss": 0.5705,
+      "step": 331
+    },
+    {
+      "epoch": 0.14415979157620495,
+      "grad_norm": 0.07688986510038376,
+      "learning_rate": 0.0007700751980829601,
+      "loss": 0.5714,
+      "step": 332
+    },
+    {
+      "epoch": 0.14459400781589232,
+      "grad_norm": 0.06042364984750748,
+      "learning_rate": 0.0007687249017370832,
+      "loss": 0.5629,
+      "step": 333
+    },
+    {
+      "epoch": 0.1450282240555797,
+      "grad_norm": 0.06250355392694473,
+      "learning_rate": 0.0007673718438270648,
+      "loss": 0.5318,
+      "step": 334
+    },
+    {
+      "epoch": 0.14546244029526703,
+      "grad_norm": 0.06661161035299301,
+      "learning_rate": 0.0007660160382576683,
+      "loss": 0.5261,
+      "step": 335
+    },
+    {
+      "epoch": 0.1458966565349544,
+      "grad_norm": 0.07277819514274597,
+      "learning_rate": 0.0007646574989618937,
+      "loss": 0.5427,
+      "step": 336
+    },
+    {
+      "epoch": 0.14633087277464177,
+      "grad_norm": 0.05938050523400307,
+      "learning_rate": 0.0007632962399008341,
+      "loss": 0.5469,
+      "step": 337
+    },
+    {
+      "epoch": 0.14676508901432914,
+      "grad_norm": 0.0626835823059082,
+      "learning_rate": 0.0007619322750635327,
+      "loss": 0.5321,
+      "step": 338
+    },
+    {
+      "epoch": 0.1471993052540165,
+      "grad_norm": 0.06853507459163666,
+      "learning_rate": 0.0007605656184668384,
+      "loss": 0.5319,
+      "step": 339
+    },
+    {
+      "epoch": 0.14763352149370387,
+      "grad_norm": 0.06294318288564682,
+      "learning_rate": 0.0007591962841552626,
+      "loss": 0.5735,
+      "step": 340
+    },
+    {
+      "epoch": 0.14806773773339124,
+      "grad_norm": 0.06769460439682007,
+      "learning_rate": 0.0007578242862008336,
+      "loss": 0.5734,
+      "step": 341
+    },
+    {
+      "epoch": 0.14850195397307858,
+      "grad_norm": 0.06945870816707611,
+      "learning_rate": 0.0007564496387029531,
+      "loss": 0.5486,
+      "step": 342
+    },
+    {
+      "epoch": 0.14893617021276595,
+      "grad_norm": 0.055839743465185165,
+      "learning_rate": 0.0007550723557882513,
+      "loss": 0.5646,
+      "step": 343
+    },
+    {
+      "epoch": 0.14937038645245332,
+      "grad_norm": 0.05766444280743599,
+      "learning_rate": 0.0007536924516104411,
+      "loss": 0.5671,
+      "step": 344
+    },
+    {
+      "epoch": 0.1498046026921407,
+      "grad_norm": 0.05658617615699768,
+      "learning_rate": 0.000752309940350173,
+      "loss": 0.5573,
+      "step": 345
+    },
+    {
+      "epoch": 0.15023881893182806,
+      "grad_norm": 0.06456328183412552,
+      "learning_rate": 0.0007509248362148888,
+      "loss": 0.5413,
+      "step": 346
+    },
+    {
+      "epoch": 0.15067303517151542,
+      "grad_norm": 0.05971672758460045,
+      "learning_rate": 0.000749537153438677,
+      "loss": 0.528,
+      "step": 347
+    },
+    {
+      "epoch": 0.1511072514112028,
+      "grad_norm": 0.059805817902088165,
+      "learning_rate": 0.0007481469062821251,
+      "loss": 0.5517,
+      "step": 348
+    },
+    {
+      "epoch": 0.15154146765089013,
+      "grad_norm": 0.05796041712164879,
+      "learning_rate": 0.0007467541090321735,
+      "loss": 0.5651,
+      "step": 349
+    },
+    {
+      "epoch": 0.1519756838905775,
+      "grad_norm": 0.0807691365480423,
+      "learning_rate": 0.000745358776001969,
+      "loss": 0.5632,
+      "step": 350
+    },
+    {
+      "epoch": 0.15240990013026487,
+      "grad_norm": 0.056193944066762924,
+      "learning_rate": 0.0007439609215307173,
+      "loss": 0.5544,
+      "step": 351
+    },
+    {
+      "epoch": 0.15284411636995224,
+      "grad_norm": 0.06092338636517525,
+      "learning_rate": 0.000742560559983536,
+      "loss": 0.5378,
+      "step": 352
+    },
+    {
+      "epoch": 0.1532783326096396,
+      "grad_norm": 0.06231565400958061,
+      "learning_rate": 0.0007411577057513066,
+      "loss": 0.5584,
+      "step": 353
+    },
+    {
+      "epoch": 0.15371254884932697,
+      "grad_norm": 0.06663519144058228,
+      "learning_rate": 0.000739752373250527,
+      "loss": 0.5246,
+      "step": 354
+    },
+    {
+      "epoch": 0.15414676508901434,
+      "grad_norm": 0.060423221439123154,
+      "learning_rate": 0.0007383445769231627,
+      "loss": 0.5203,
+      "step": 355
+    },
+    {
+      "epoch": 0.15458098132870168,
+      "grad_norm": 0.06870024651288986,
+      "learning_rate": 0.0007369343312364993,
+      "loss": 0.52,
+      "step": 356
+    },
+    {
+      "epoch": 0.15501519756838905,
+      "grad_norm": 0.06956614553928375,
+      "learning_rate": 0.0007355216506829932,
+      "loss": 0.5816,
+      "step": 357
+    },
+    {
+      "epoch": 0.15544941380807642,
+      "grad_norm": 0.060565389692783356,
+      "learning_rate": 0.0007341065497801229,
+      "loss": 0.5508,
+      "step": 358
+    },
+    {
+      "epoch": 0.1558836300477638,
+      "grad_norm": 0.06745211035013199,
+      "learning_rate": 0.0007326890430702396,
+      "loss": 0.5589,
+      "step": 359
+    },
+    {
+      "epoch": 0.15631784628745116,
+      "grad_norm": 0.06009297072887421,
+      "learning_rate": 0.0007312691451204177,
+      "loss": 0.5259,
+      "step": 360
+    },
+    {
+      "epoch": 0.15675206252713852,
+      "grad_norm": 0.05682894587516785,
+      "learning_rate": 0.0007298468705223059,
+      "loss": 0.5689,
+      "step": 361
+    },
+    {
+      "epoch": 0.15718627876682587,
+      "grad_norm": 0.06062963977456093,
+      "learning_rate": 0.0007284222338919758,
+      "loss": 0.5339,
+      "step": 362
+    },
+    {
+      "epoch": 0.15762049500651323,
+      "grad_norm": 0.058002885431051254,
+      "learning_rate": 0.0007269952498697733,
+      "loss": 0.5714,
+      "step": 363
+    },
+    {
+      "epoch": 0.1580547112462006,
+      "grad_norm": 0.056898582726716995,
+      "learning_rate": 0.0007255659331201672,
+      "loss": 0.5811,
+      "step": 364
+    },
+    {
+      "epoch": 0.15848892748588797,
+      "grad_norm": 0.06246396526694298,
+      "learning_rate": 0.0007241342983315984,
+      "loss": 0.5554,
+      "step": 365
+    },
+    {
+      "epoch": 0.15892314372557534,
+      "grad_norm": 0.05486922711133957,
+      "learning_rate": 0.0007227003602163296,
+      "loss": 0.5567,
+      "step": 366
+    },
+    {
+      "epoch": 0.1593573599652627,
+      "grad_norm": 0.057243555784225464,
+      "learning_rate": 0.0007212641335102932,
+      "loss": 0.5193,
+      "step": 367
+    },
+    {
+      "epoch": 0.15979157620495008,
+      "grad_norm": 0.05640234053134918,
+      "learning_rate": 0.0007198256329729411,
+      "loss": 0.5098,
+      "step": 368
+    },
+    {
+      "epoch": 0.16022579244463742,
+      "grad_norm": 0.05819845199584961,
+      "learning_rate": 0.0007183848733870917,
+      "loss": 0.5366,
+      "step": 369
+    },
+    {
+      "epoch": 0.16066000868432478,
+      "grad_norm": 0.05703158304095268,
+      "learning_rate": 0.0007169418695587791,
+      "loss": 0.5374,
+      "step": 370
+    },
+    {
+      "epoch": 0.16109422492401215,
+      "grad_norm": 0.05363311246037483,
+      "learning_rate": 0.0007154966363171003,
+      "loss": 0.555,
+      "step": 371
+    },
+    {
+      "epoch": 0.16152844116369952,
+      "grad_norm": 0.060146309435367584,
+      "learning_rate": 0.0007140491885140629,
+      "loss": 0.5395,
+      "step": 372
+    },
+    {
+      "epoch": 0.1619626574033869,
+      "grad_norm": 0.05488353967666626,
+      "learning_rate": 0.0007125995410244324,
+      "loss": 0.5117,
+      "step": 373
+    },
+    {
+      "epoch": 0.16239687364307426,
+      "grad_norm": 0.05473377928137779,
+      "learning_rate": 0.00071114770874558,
+      "loss": 0.5463,
+      "step": 374
+    },
+    {
+      "epoch": 0.16283108988276163,
+      "grad_norm": 0.05373527109622955,
+      "learning_rate": 0.0007096937065973285,
+      "loss": 0.522,
+      "step": 375
+    },
+    {
+      "epoch": 0.16326530612244897,
+      "grad_norm": 0.054411571472883224,
+      "learning_rate": 0.0007082375495217995,
+      "loss": 0.519,
+      "step": 376
+    },
+    {
+      "epoch": 0.16369952236213633,
+      "grad_norm": 0.059973325580358505,
+      "learning_rate": 0.0007067792524832604,
+      "loss": 0.5546,
+      "step": 377
+    },
+    {
+      "epoch": 0.1641337386018237,
+      "grad_norm": 0.07386507093906403,
+      "learning_rate": 0.000705318830467969,
+      "loss": 0.548,
+      "step": 378
+    },
+    {
+      "epoch": 0.16456795484151107,
+      "grad_norm": 0.05318816378712654,
+      "learning_rate": 0.0007038562984840216,
+      "loss": 0.5325,
+      "step": 379
+    },
+    {
+      "epoch": 0.16500217108119844,
+      "grad_norm": 0.06362178921699524,
+      "learning_rate": 0.0007023916715611969,
+      "loss": 0.5537,
+      "step": 380
+    },
+    {
+      "epoch": 0.1654363873208858,
+      "grad_norm": 0.05186508968472481,
+      "learning_rate": 0.0007009249647508028,
+      "loss": 0.5534,
+      "step": 381
+    },
+    {
+      "epoch": 0.16587060356057318,
+      "grad_norm": 0.05456273630261421,
+      "learning_rate": 0.0006994561931255209,
+      "loss": 0.533,
+      "step": 382
+    },
+    {
+      "epoch": 0.16630481980026052,
+      "grad_norm": 0.07069144397974014,
+      "learning_rate": 0.0006979853717792523,
+      "loss": 0.5591,
+      "step": 383
+    },
+    {
+      "epoch": 0.16673903603994789,
+      "grad_norm": 0.13737140595912933,
+      "learning_rate": 0.0006965125158269618,
+      "loss": 0.8425,
+      "step": 384
+    },
+    {
+      "epoch": 0.16717325227963525,
+      "grad_norm": 0.08646462857723236,
+      "learning_rate": 0.0006950376404045235,
+      "loss": 0.5356,
+      "step": 385
+    },
+    {
+      "epoch": 0.16760746851932262,
+      "grad_norm": 0.059485238045454025,
+      "learning_rate": 0.0006935607606685641,
+      "loss": 0.571,
+      "step": 386
+    },
+    {
+      "epoch": 0.16804168475901,
+      "grad_norm": 0.07368376851081848,
+      "learning_rate": 0.000692081891796308,
+      "loss": 0.5346,
+      "step": 387
+    },
+    {
+      "epoch": 0.16847590099869736,
+      "grad_norm": 0.0615500770509243,
+      "learning_rate": 0.0006906010489854209,
+      "loss": 0.5257,
+      "step": 388
+    },
+    {
+      "epoch": 0.16891011723838473,
+      "grad_norm": 0.064671091735363,
+      "learning_rate": 0.0006891182474538539,
+      "loss": 0.5492,
+      "step": 389
+    },
+    {
+      "epoch": 0.16934433347807207,
+      "grad_norm": 0.06681560724973679,
+      "learning_rate": 0.0006876335024396872,
+      "loss": 0.5595,
+      "step": 390
+    },
+    {
+      "epoch": 0.16977854971775944,
+      "grad_norm": 0.0575130432844162,
+      "learning_rate": 0.0006861468292009726,
+      "loss": 0.5895,
+      "step": 391
+    },
+    {
+      "epoch": 0.1702127659574468,
+      "grad_norm": 0.05460560694336891,
+      "learning_rate": 0.0006846582430155782,
+      "loss": 0.5374,
+      "step": 392
+    },
+    {
+      "epoch": 0.17064698219713417,
+      "grad_norm": 0.06030447408556938,
+      "learning_rate": 0.0006831677591810301,
+      "loss": 0.6073,
+      "step": 393
+    },
+    {
+      "epoch": 0.17108119843682154,
+      "grad_norm": 0.0582847036421299,
+      "learning_rate": 0.0006816753930143557,
+      "loss": 0.5347,
+      "step": 394
+    },
+    {
+      "epoch": 0.1715154146765089,
+      "grad_norm": 0.05572697892785072,
+      "learning_rate": 0.0006801811598519267,
+      "loss": 0.5042,
+      "step": 395
+    },
+    {
+      "epoch": 0.17194963091619628,
+      "grad_norm": 0.05724466219544411,
+      "learning_rate": 0.0006786850750493006,
+      "loss": 0.5613,
+      "step": 396
+    },
+    {
+      "epoch": 0.17238384715588362,
+      "grad_norm": 0.06665828824043274,
+      "learning_rate": 0.0006771871539810632,
+      "loss": 0.5841,
+      "step": 397
+    },
+    {
+      "epoch": 0.172818063395571,
+      "grad_norm": 0.05465397238731384,
+      "learning_rate": 0.0006756874120406714,
+      "loss": 0.5636,
+      "step": 398
+    },
+    {
+      "epoch": 0.17325227963525835,
+      "grad_norm": 0.07157056033611298,
+      "learning_rate": 0.000674185864640294,
+      "loss": 0.561,
+      "step": 399
+    },
+    {
+      "epoch": 0.17368649587494572,
+      "grad_norm": 0.05561946704983711,
+      "learning_rate": 0.0006726825272106538,
+      "loss": 0.5351,
+      "step": 400
+    },
+    {
+      "epoch": 0.1741207121146331,
+      "grad_norm": 0.05475137010216713,
+      "learning_rate": 0.0006711774152008689,
+      "loss": 0.5445,
+      "step": 401
+    },
+    {
+      "epoch": 0.17455492835432046,
+      "grad_norm": 0.07090818136930466,
+      "learning_rate": 0.0006696705440782938,
+      "loss": 0.5279,
+      "step": 402
+    },
+    {
+      "epoch": 0.17498914459400783,
+      "grad_norm": 0.051751721650362015,
+      "learning_rate": 0.0006681619293283609,
+      "loss": 0.4934,
+      "step": 403
+    },
+    {
+      "epoch": 0.17542336083369517,
+      "grad_norm": 0.06089828908443451,
+      "learning_rate": 0.0006666515864544209,
+      "loss": 0.5413,
+      "step": 404
+    },
+    {
+      "epoch": 0.17585757707338254,
+      "grad_norm": 0.052899301052093506,
+      "learning_rate": 0.0006651395309775837,
+      "loss": 0.5187,
+      "step": 405
+    },
+    {
+      "epoch": 0.1762917933130699,
+      "grad_norm": 0.05137433856725693,
+      "learning_rate": 0.0006636257784365584,
+      "loss": 0.5636,
+      "step": 406
+    },
+    {
+      "epoch": 0.17672600955275727,
+      "grad_norm": 0.060585830360651016,
+      "learning_rate": 0.0006621103443874949,
+      "loss": 0.5311,
+      "step": 407
+    },
+    {
+      "epoch": 0.17716022579244464,
+      "grad_norm": 0.052542008459568024,
+      "learning_rate": 0.0006605932444038228,
+      "loss": 0.5344,
+      "step": 408
+    },
+    {
+      "epoch": 0.177594442032132,
+      "grad_norm": 0.052282754331827164,
+      "learning_rate": 0.0006590744940760914,
+      "loss": 0.5616,
+      "step": 409
+    },
+    {
+      "epoch": 0.17802865827181938,
+      "grad_norm": 0.055943962186574936,
+      "learning_rate": 0.0006575541090118104,
+      "loss": 0.5657,
+      "step": 410
+    },
+    {
+      "epoch": 0.17846287451150672,
+      "grad_norm": 0.05422157794237137,
+      "learning_rate": 0.0006560321048352886,
+      "loss": 0.5482,
+      "step": 411
+    },
+    {
+      "epoch": 0.1788970907511941,
+      "grad_norm": 0.053238000720739365,
+      "learning_rate": 0.0006545084971874737,
+      "loss": 0.5477,
+      "step": 412
+    },
+    {
+      "epoch": 0.17933130699088146,
+      "grad_norm": 0.05666307359933853,
+      "learning_rate": 0.0006529833017257919,
+      "loss": 0.5593,
+      "step": 413
+    },
+    {
+      "epoch": 0.17976552323056882,
+      "grad_norm": 0.05702383071184158,
+      "learning_rate": 0.000651456534123986,
+      "loss": 0.5229,
+      "step": 414
+    },
+    {
+      "epoch": 0.1801997394702562,
+      "grad_norm": 0.05041234940290451,
+      "learning_rate": 0.0006499282100719558,
+      "loss": 0.5197,
+      "step": 415
+    },
+    {
+      "epoch": 0.18063395570994356,
+      "grad_norm": 0.05691606178879738,
+      "learning_rate": 0.0006483983452755952,
+      "loss": 0.5282,
+      "step": 416
+    },
+    {
+      "epoch": 0.18106817194963093,
+      "grad_norm": 0.05236973240971565,
+      "learning_rate": 0.0006468669554566324,
+      "loss": 0.574,
+      "step": 417
+    },
+    {
+      "epoch": 0.18150238818931827,
+      "grad_norm": 0.05194111168384552,
+      "learning_rate": 0.0006453340563524669,
+      "loss": 0.521,
+      "step": 418
+    },
+    {
+      "epoch": 0.18193660442900564,
+      "grad_norm": 0.05307863652706146,
+      "learning_rate": 0.0006437996637160086,
+      "loss": 0.5474,
+      "step": 419
+    },
+    {
+      "epoch": 0.182370820668693,
+      "grad_norm": 0.05186399817466736,
+      "learning_rate": 0.0006422637933155162,
+      "loss": 0.5584,
+      "step": 420
+    },
+    {
+      "epoch": 0.18280503690838038,
+      "grad_norm": 0.053380340337753296,
+      "learning_rate": 0.0006407264609344343,
+      "loss": 0.5309,
+      "step": 421
+    },
+    {
+      "epoch": 0.18323925314806774,
+      "grad_norm": 0.05926523730158806,
+      "learning_rate": 0.0006391876823712317,
+      "loss": 0.5362,
+      "step": 422
+    },
+    {
+      "epoch": 0.1836734693877551,
+      "grad_norm": 0.053086958825588226,
+      "learning_rate": 0.0006376474734392387,
+      "loss": 0.5345,
+      "step": 423
+    },
+    {
+      "epoch": 0.18410768562744248,
+      "grad_norm": 0.05506280064582825,
+      "learning_rate": 0.0006361058499664855,
+      "loss": 0.5301,
+      "step": 424
+    },
+    {
+      "epoch": 0.18454190186712982,
+      "grad_norm": 0.0567692294716835,
+      "learning_rate": 0.0006345628277955385,
+      "loss": 0.5336,
+      "step": 425
+    },
+    {
+      "epoch": 0.1849761181068172,
+      "grad_norm": 0.04989070072770119,
+      "learning_rate": 0.0006330184227833376,
+      "loss": 0.53,
+      "step": 426
+    },
+    {
+      "epoch": 0.18541033434650456,
+      "grad_norm": 0.05129670351743698,
+      "learning_rate": 0.000631472650801034,
+      "loss": 0.523,
+      "step": 427
+    },
+    {
+      "epoch": 0.18584455058619193,
+      "grad_norm": 0.052287109196186066,
+      "learning_rate": 0.0006299255277338265,
+      "loss": 0.5309,
+      "step": 428
+    },
+    {
+      "epoch": 0.1862787668258793,
+      "grad_norm": 0.051617227494716644,
+      "learning_rate": 0.0006283770694807982,
+      "loss": 0.5406,
+      "step": 429
+    },
+    {
+      "epoch": 0.18671298306556666,
+      "grad_norm": 0.06112902611494064,
+      "learning_rate": 0.0006268272919547536,
+      "loss": 0.5148,
+      "step": 430
+    },
+    {
+      "epoch": 0.187147199305254,
+      "grad_norm": 0.04904035106301308,
+      "learning_rate": 0.0006252762110820547,
+      "loss": 0.5277,
+      "step": 431
+    },
+    {
+      "epoch": 0.18758141554494137,
+      "grad_norm": 0.05905110388994217,
+      "learning_rate": 0.0006237238428024572,
+      "loss": 0.5148,
+      "step": 432
+    },
+    {
+      "epoch": 0.18801563178462874,
+      "grad_norm": 0.05047300457954407,
+      "learning_rate": 0.000622170203068947,
+      "loss": 0.5119,
+      "step": 433
+    },
+    {
+      "epoch": 0.1884498480243161,
+      "grad_norm": 0.0518287792801857,
+      "learning_rate": 0.0006206153078475762,
+      "loss": 0.5276,
+      "step": 434
+    },
+    {
+      "epoch": 0.18888406426400348,
+      "grad_norm": 0.04822736606001854,
+      "learning_rate": 0.0006190591731172991,
+      "loss": 0.5347,
+      "step": 435
+    },
+    {
+      "epoch": 0.18931828050369084,
+      "grad_norm": 0.052807264029979706,
+      "learning_rate": 0.0006175018148698076,
+      "loss": 0.5343,
+      "step": 436
+    },
+    {
+      "epoch": 0.1897524967433782,
+      "grad_norm": 0.054650746285915375,
+      "learning_rate": 0.0006159432491093672,
+      "loss": 0.5079,
+      "step": 437
+    },
+    {
+      "epoch": 0.19018671298306555,
+      "grad_norm": 0.047450825572013855,
+      "learning_rate": 0.0006143834918526527,
+      "loss": 0.5532,
+      "step": 438
+    },
+    {
+      "epoch": 0.19062092922275292,
+      "grad_norm": 0.047222550958395004,
+      "learning_rate": 0.0006128225591285831,
+      "loss": 0.5376,
+      "step": 439
+    },
+    {
+      "epoch": 0.1910551454624403,
+      "grad_norm": 0.05076510086655617,
+      "learning_rate": 0.0006112604669781572,
+      "loss": 0.5152,
+      "step": 440
+    },
+    {
+      "epoch": 0.19148936170212766,
+      "grad_norm": 0.06690964847803116,
+      "learning_rate": 0.0006096972314542889,
+      "loss": 0.5395,
+      "step": 441
+    },
+    {
+      "epoch": 0.19192357794181503,
+      "grad_norm": 0.049380868673324585,
+      "learning_rate": 0.0006081328686216418,
+      "loss": 0.4996,
+      "step": 442
+    },
+    {
+      "epoch": 0.1923577941815024,
+      "grad_norm": 0.06767404824495316,
+      "learning_rate": 0.0006065673945564642,
+      "loss": 0.5217,
+      "step": 443
+    },
+    {
+      "epoch": 0.19279201042118976,
+      "grad_norm": 0.054420776665210724,
+      "learning_rate": 0.0006050008253464246,
+      "loss": 0.5587,
+      "step": 444
+    },
+    {
+      "epoch": 0.1932262266608771,
+      "grad_norm": 0.0557849258184433,
+      "learning_rate": 0.0006034331770904454,
+      "loss": 0.5441,
+      "step": 445
+    },
+    {
+      "epoch": 0.19366044290056447,
+      "grad_norm": 0.06514862924814224,
+      "learning_rate": 0.0006018644658985379,
+      "loss": 0.5666,
+      "step": 446
+    },
+    {
+      "epoch": 0.19409465914025184,
+      "grad_norm": 0.061305928975343704,
+      "learning_rate": 0.0006002947078916364,
+      "loss": 0.5576,
+      "step": 447
+    },
+    {
+      "epoch": 0.1945288753799392,
+      "grad_norm": 0.05214373767375946,
+      "learning_rate": 0.0005987239192014335,
+      "loss": 0.5424,
+      "step": 448
+    },
+    {
+      "epoch": 0.19496309161962658,
+      "grad_norm": 0.05650056526064873,
+      "learning_rate": 0.0005971521159702136,
+      "loss": 0.5361,
+      "step": 449
+    },
+    {
+      "epoch": 0.19539730785931395,
+      "grad_norm": 0.05745421350002289,
+      "learning_rate": 0.0005955793143506863,
+      "loss": 0.5273,
+      "step": 450
+    },
+    {
+      "epoch": 0.19583152409900131,
+      "grad_norm": 0.05364866554737091,
+      "learning_rate": 0.0005940055305058219,
+      "loss": 0.5026,
+      "step": 451
+    },
+    {
+      "epoch": 0.19626574033868865,
+      "grad_norm": 0.047814685851335526,
+      "learning_rate": 0.0005924307806086844,
+      "loss": 0.5439,
+      "step": 452
+    },
+    {
+      "epoch": 0.19669995657837602,
+      "grad_norm": 0.051681190729141235,
+      "learning_rate": 0.0005908550808422655,
+      "loss": 0.5271,
+      "step": 453
+    },
+    {
+      "epoch": 0.1971341728180634,
+      "grad_norm": 0.05405402183532715,
+      "learning_rate": 0.0005892784473993184,
+      "loss": 0.4996,
+      "step": 454
+    },
+    {
+      "epoch": 0.19756838905775076,
+      "grad_norm": 0.04617345333099365,
+      "learning_rate": 0.0005877008964821908,
+      "loss": 0.5409,
+      "step": 455
+    },
+    {
+      "epoch": 0.19800260529743813,
+      "grad_norm": 0.05039280652999878,
+      "learning_rate": 0.0005861224443026595,
+      "loss": 0.5433,
+      "step": 456
+    },
+    {
+      "epoch": 0.1984368215371255,
+      "grad_norm": 0.05184314772486687,
+      "learning_rate": 0.0005845431070817626,
+      "loss": 0.5422,
+      "step": 457
+    },
+    {
+      "epoch": 0.19887103777681286,
+      "grad_norm": 0.056363269686698914,
+      "learning_rate": 0.000582962901049634,
+      "loss": 0.5202,
+      "step": 458
+    },
+    {
+      "epoch": 0.1993052540165002,
+      "grad_norm": 0.04747706279158592,
+      "learning_rate": 0.0005813818424453351,
+      "loss": 0.5287,
+      "step": 459
+    },
+    {
+      "epoch": 0.19973947025618757,
+      "grad_norm": 0.05112398415803909,
+      "learning_rate": 0.0005797999475166897,
+      "loss": 0.563,
+      "step": 460
+    },
+    {
+      "epoch": 0.20017368649587494,
+      "grad_norm": 0.050983451306819916,
+      "learning_rate": 0.0005782172325201155,
+      "loss": 0.5309,
+      "step": 461
+    },
+    {
+      "epoch": 0.2006079027355623,
+      "grad_norm": 0.05590641498565674,
+      "learning_rate": 0.000576633713720458,
+      "loss": 0.54,
+      "step": 462
+    },
+    {
+      "epoch": 0.20104211897524968,
+      "grad_norm": 0.04705791547894478,
+      "learning_rate": 0.000575049407390823,
+      "loss": 0.5291,
+      "step": 463
+    },
+    {
+      "epoch": 0.20147633521493705,
+      "grad_norm": 0.05234816297888756,
+      "learning_rate": 0.000573464329812409,
+      "loss": 0.5331,
+      "step": 464
+    },
+    {
+      "epoch": 0.20191055145462442,
+      "grad_norm": 0.04806197062134743,
+      "learning_rate": 0.0005718784972743409,
+      "loss": 0.5141,
+      "step": 465
+    },
+    {
+      "epoch": 0.20234476769431176,
+      "grad_norm": 0.0483107753098011,
+      "learning_rate": 0.0005702919260735014,
+      "loss": 0.5066,
+      "step": 466
+    },
+    {
+      "epoch": 0.20277898393399912,
+      "grad_norm": 0.05739190801978111,
+      "learning_rate": 0.0005687046325143647,
+      "loss": 0.5282,
+      "step": 467
+    },
+    {
+      "epoch": 0.2032132001736865,
+      "grad_norm": 0.04906076937913895,
+      "learning_rate": 0.0005671166329088278,
+      "loss": 0.5257,
+      "step": 468
+    },
+    {
+      "epoch": 0.20364741641337386,
+      "grad_norm": 0.04784254729747772,
+      "learning_rate": 0.0005655279435760435,
+      "loss": 0.5193,
+      "step": 469
+    },
+    {
+      "epoch": 0.20408163265306123,
+      "grad_norm": 0.06473550200462341,
+      "learning_rate": 0.000563938580842253,
+      "loss": 0.5257,
+      "step": 470
+    },
+    {
+      "epoch": 0.2045158488927486,
+      "grad_norm": 0.047775350511074066,
+      "learning_rate": 0.0005623485610406173,
+      "loss": 0.5008,
+      "step": 471
+    },
+    {
+      "epoch": 0.20495006513243597,
+      "grad_norm": 0.05449296161532402,
+      "learning_rate": 0.0005607579005110502,
+      "loss": 0.5316,
+      "step": 472
+    },
+    {
+      "epoch": 0.2053842813721233,
+      "grad_norm": 0.05296491086483002,
+      "learning_rate": 0.0005591666156000494,
+      "loss": 0.5493,
+      "step": 473
+    },
+    {
+      "epoch": 0.20581849761181067,
+      "grad_norm": 0.045328423380851746,
+      "learning_rate": 0.0005575747226605297,
+      "loss": 0.508,
+      "step": 474
+    },
+    {
+      "epoch": 0.20625271385149804,
+      "grad_norm": 0.05045296251773834,
+      "learning_rate": 0.0005559822380516539,
+      "loss": 0.5118,
+      "step": 475
+    },
+    {
+      "epoch": 0.2066869300911854,
+      "grad_norm": 0.04571797326207161,
+      "learning_rate": 0.0005543891781386656,
+      "loss": 0.5168,
+      "step": 476
+    },
+    {
+      "epoch": 0.20712114633087278,
+      "grad_norm": 0.05250009894371033,
+      "learning_rate": 0.0005527955592927197,
+      "loss": 0.5175,
+      "step": 477
+    },
+    {
+      "epoch": 0.20755536257056015,
+      "grad_norm": 0.048049163073301315,
+      "learning_rate": 0.0005512013978907157,
+      "loss": 0.5279,
+      "step": 478
+    },
+    {
+      "epoch": 0.20798957881024752,
+      "grad_norm": 0.04621399939060211,
+      "learning_rate": 0.0005496067103151288,
+      "loss": 0.5304,
+      "step": 479
+    },
+    {
+      "epoch": 0.20842379504993486,
+      "grad_norm": 0.04643552750349045,
+      "learning_rate": 0.0005480115129538409,
+      "loss": 0.5137,
+      "step": 480
+    },
+    {
+      "epoch": 0.20885801128962223,
+      "grad_norm": 0.05422172695398331,
+      "learning_rate": 0.0005464158221999731,
+      "loss": 0.525,
+      "step": 481
+    },
+    {
+      "epoch": 0.2092922275293096,
+      "grad_norm": 0.04593772441148758,
+      "learning_rate": 0.0005448196544517168,
+      "loss": 0.5229,
+      "step": 482
+    },
+    {
+      "epoch": 0.20972644376899696,
+      "grad_norm": 0.05278801918029785,
+      "learning_rate": 0.0005432230261121651,
+      "loss": 0.5256,
+      "step": 483
+    },
+    {
+      "epoch": 0.21016066000868433,
+      "grad_norm": 0.04598787799477577,
+      "learning_rate": 0.0005416259535891447,
+      "loss": 0.5275,
+      "step": 484
+    },
+    {
+      "epoch": 0.2105948762483717,
+      "grad_norm": 0.054117828607559204,
+      "learning_rate": 0.0005400284532950467,
+      "loss": 0.5179,
+      "step": 485
+    },
+    {
+      "epoch": 0.21102909248805907,
+      "grad_norm": 0.046420540660619736,
+      "learning_rate": 0.0005384305416466584,
+      "loss": 0.5399,
+      "step": 486
+    },
+    {
+      "epoch": 0.2114633087277464,
+      "grad_norm": 0.048214301466941833,
+      "learning_rate": 0.0005368322350649942,
+      "loss": 0.5202,
+      "step": 487
+    },
+    {
+      "epoch": 0.21189752496743378,
+      "grad_norm": 0.05419298633933067,
+      "learning_rate": 0.0005352335499751269,
+      "loss": 0.4925,
+      "step": 488
+    },
+    {
+      "epoch": 0.21233174120712114,
+      "grad_norm": 0.04759565740823746,
+      "learning_rate": 0.0005336345028060199,
+      "loss": 0.5162,
+      "step": 489
+    },
+    {
+      "epoch": 0.2127659574468085,
+      "grad_norm": 0.05174829810857773,
+      "learning_rate": 0.0005320351099903565,
+      "loss": 0.5043,
+      "step": 490
+    },
+    {
+      "epoch": 0.21320017368649588,
+      "grad_norm": 0.06595347821712494,
+      "learning_rate": 0.0005304353879643726,
+      "loss": 0.5029,
+      "step": 491
+    },
+    {
+      "epoch": 0.21363438992618325,
+      "grad_norm": 0.04908424988389015,
+      "learning_rate": 0.0005288353531676872,
+      "loss": 0.5043,
+      "step": 492
+    },
+    {
+      "epoch": 0.21406860616587062,
+      "grad_norm": 0.05055024474859238,
+      "learning_rate": 0.0005272350220431334,
+      "loss": 0.5192,
+      "step": 493
+    },
+    {
+      "epoch": 0.21450282240555796,
+      "grad_norm": 0.049286942929029465,
+      "learning_rate": 0.0005256344110365896,
+      "loss": 0.5041,
+      "step": 494
+    },
+    {
+      "epoch": 0.21493703864524533,
+      "grad_norm": 0.04700294882059097,
+      "learning_rate": 0.0005240335365968104,
+      "loss": 0.5258,
+      "step": 495
+    },
+    {
+      "epoch": 0.2153712548849327,
+      "grad_norm": 0.0560920424759388,
+      "learning_rate": 0.0005224324151752575,
+      "loss": 0.5388,
+      "step": 496
+    },
+    {
+      "epoch": 0.21580547112462006,
+      "grad_norm": 0.05053974315524101,
+      "learning_rate": 0.0005208310632259308,
+      "loss": 0.5045,
+      "step": 497
+    },
+    {
+      "epoch": 0.21623968736430743,
+      "grad_norm": 0.06237909570336342,
+      "learning_rate": 0.0005192294972051992,
+      "loss": 0.463,
+      "step": 498
+    },
+    {
+      "epoch": 0.2166739036039948,
+      "grad_norm": 0.05128193646669388,
+      "learning_rate": 0.0005176277335716317,
+      "loss": 0.5209,
+      "step": 499
+    },
+    {
+      "epoch": 0.21710811984368214,
+      "grad_norm": 0.04433543235063553,
+      "learning_rate": 0.0005160257887858277,
+      "loss": 0.5046,
+      "step": 500
+    },
+    {
+      "epoch": 0.2175423360833695,
+      "grad_norm": 0.05125072970986366,
+      "learning_rate": 0.0005144236793102484,
+      "loss": 0.4894,
+      "step": 501
+    },
+    {
+      "epoch": 0.21797655232305688,
+      "grad_norm": 0.04465992748737335,
+      "learning_rate": 0.0005128214216090478,
+      "loss": 0.4987,
+      "step": 502
+    },
+    {
+      "epoch": 0.21841076856274425,
+      "grad_norm": 0.04684034362435341,
+      "learning_rate": 0.0005112190321479025,
+      "loss": 0.5498,
+      "step": 503
+    },
+    {
+      "epoch": 0.2188449848024316,
+      "grad_norm": 0.04876565933227539,
+      "learning_rate": 0.0005096165273938436,
+      "loss": 0.5071,
+      "step": 504
+    },
+    {
+      "epoch": 0.21927920104211898,
+      "grad_norm": 0.04982059448957443,
+      "learning_rate": 0.0005080139238150869,
+      "loss": 0.5084,
+      "step": 505
+    },
+    {
+      "epoch": 0.21971341728180635,
+      "grad_norm": 0.052202560007572174,
+      "learning_rate": 0.0005064112378808637,
+      "loss": 0.5233,
+      "step": 506
+    },
+    {
+      "epoch": 0.2201476335214937,
+      "grad_norm": 0.04747169837355614,
+      "learning_rate": 0.0005048084860612516,
+      "loss": 0.5263,
+      "step": 507
+    },
+    {
+      "epoch": 0.22058184976118106,
+      "grad_norm": 0.0462409108877182,
+      "learning_rate": 0.0005032056848270056,
+      "loss": 0.5237,
+      "step": 508
+    },
+    {
+      "epoch": 0.22101606600086843,
+      "grad_norm": 0.04612501338124275,
+      "learning_rate": 0.000501602850649388,
+      "loss": 0.5507,
+      "step": 509
+    },
+    {
+      "epoch": 0.2214502822405558,
+      "grad_norm": 0.04581817239522934,
+      "learning_rate": 0.0005,
+      "loss": 0.4986,
+      "step": 510
+    },
+    {
+      "epoch": 0.22188449848024316,
+      "grad_norm": 0.0477612242102623,
+      "learning_rate": 0.0004983971493506121,
+      "loss": 0.51,
+      "step": 511
+    },
+    {
+      "epoch": 0.22231871471993053,
+      "grad_norm": 0.05176355689764023,
+      "learning_rate": 0.0004967943151729944,
+      "loss": 0.5115,
+      "step": 512
+    },
+    {
+      "epoch": 0.2227529309596179,
+      "grad_norm": 0.04731719568371773,
+      "learning_rate": 0.0004951915139387483,
+      "loss": 0.5564,
+      "step": 513
+    },
+    {
+      "epoch": 0.22318714719930524,
+      "grad_norm": 0.04438960924744606,
+      "learning_rate": 0.0004935887621191363,
+      "loss": 0.5502,
+      "step": 514
+    },
+    {
+      "epoch": 0.2236213634389926,
+      "grad_norm": 0.06116556003689766,
+      "learning_rate": 0.0004919860761849132,
+      "loss": 0.5417,
+      "step": 515
+    },
+    {
+      "epoch": 0.22405557967867998,
+      "grad_norm": 0.046353522688150406,
+      "learning_rate": 0.0004903834726061564,
+      "loss": 0.5285,
+      "step": 516
+    },
+    {
+      "epoch": 0.22448979591836735,
+      "grad_norm": 0.04643326997756958,
+      "learning_rate": 0.0004887809678520976,
+      "loss": 0.4774,
+      "step": 517
+    },
+    {
+      "epoch": 0.22492401215805471,
+      "grad_norm": 0.0530591681599617,
+      "learning_rate": 0.0004871785783909523,
+      "loss": 0.5076,
+      "step": 518
+    },
+    {
+      "epoch": 0.22535822839774208,
+      "grad_norm": 0.04789597541093826,
+      "learning_rate": 0.0004855763206897516,
+      "loss": 0.4989,
+      "step": 519
+    },
+    {
+      "epoch": 0.22579244463742945,
+      "grad_norm": 0.047394026070833206,
+      "learning_rate": 0.0004839742112141724,
+      "loss": 0.5069,
+      "step": 520
+    },
+    {
+      "epoch": 0.2262266608771168,
+      "grad_norm": 0.044425006955862045,
+      "learning_rate": 0.0004823722664283684,
+      "loss": 0.5212,
+      "step": 521
+    },
+    {
+      "epoch": 0.22666087711680416,
+      "grad_norm": 0.044534265995025635,
+      "learning_rate": 0.0004807705027948008,
+      "loss": 0.5184,
+      "step": 522
+    },
+    {
+      "epoch": 0.22709509335649153,
+      "grad_norm": 0.04293319582939148,
+      "learning_rate": 0.0004791689367740692,
+      "loss": 0.4807,
+      "step": 523
+    },
+    {
+      "epoch": 0.2275293095961789,
+      "grad_norm": 0.04485393315553665,
+      "learning_rate": 0.0004775675848247427,
+      "loss": 0.4746,
+      "step": 524
+    },
+    {
+      "epoch": 0.22796352583586627,
+      "grad_norm": 0.043829191476106644,
+      "learning_rate": 0.0004759664634031897,
+      "loss": 0.4936,
+      "step": 525
+    },
+    {
+      "epoch": 0.22839774207555363,
+      "grad_norm": 0.05110118165612221,
+      "learning_rate": 0.00047436558896341046,
+      "loss": 0.5849,
+      "step": 526
+    },
+    {
+      "epoch": 0.228831958315241,
+      "grad_norm": 0.04783101752400398,
+      "learning_rate": 0.0004727649779568666,
+      "loss": 0.5034,
+      "step": 527
+    },
+    {
+      "epoch": 0.22926617455492834,
+      "grad_norm": 0.047072965651750565,
+      "learning_rate": 0.00047116464683231285,
+      "loss": 0.5155,
+      "step": 528
+    },
+    {
+      "epoch": 0.2297003907946157,
+      "grad_norm": 0.0427449531853199,
+      "learning_rate": 0.0004695646120356274,
+      "loss": 0.5138,
+      "step": 529
+    },
+    {
+      "epoch": 0.23013460703430308,
+      "grad_norm": 0.04664051905274391,
+      "learning_rate": 0.0004679648900096436,
+      "loss": 0.4727,
+      "step": 530
+    },
+    {
+      "epoch": 0.23056882327399045,
+      "grad_norm": 0.04555191844701767,
+      "learning_rate": 0.0004663654971939802,
+      "loss": 0.5386,
+      "step": 531
+    },
+    {
+      "epoch": 0.23100303951367782,
+      "grad_norm": 0.05113929882645607,
+      "learning_rate": 0.00046476645002487297,
+      "loss": 0.5018,
+      "step": 532
+    },
+    {
+      "epoch": 0.23143725575336518,
+      "grad_norm": 0.04245986044406891,
+      "learning_rate": 0.00046316776493500613,
+      "loss": 0.4949,
+      "step": 533
+    },
+    {
+      "epoch": 0.23187147199305255,
+      "grad_norm": 0.047864172607660294,
+      "learning_rate": 0.0004615694583533418,
+      "loss": 0.5032,
+      "step": 534
+    },
+    {
+      "epoch": 0.2323056882327399,
+      "grad_norm": 0.06898235529661179,
+      "learning_rate": 0.0004599715467049534,
+      "loss": 0.4922,
+      "step": 535
+    },
+    {
+      "epoch": 0.23273990447242726,
+      "grad_norm": 0.04230332002043724,
+      "learning_rate": 0.0004583740464108554,
+      "loss": 0.5164,
+      "step": 536
+    },
+    {
+      "epoch": 0.23317412071211463,
+      "grad_norm": 0.043854959309101105,
+      "learning_rate": 0.00045677697388783495,
+      "loss": 0.4894,
+      "step": 537
+    },
+    {
+      "epoch": 0.233608336951802,
+      "grad_norm": 0.05595362186431885,
+      "learning_rate": 0.0004551803455482833,
+      "loss": 0.5107,
+      "step": 538
+    },
+    {
+      "epoch": 0.23404255319148937,
+      "grad_norm": 0.05336504802107811,
+      "learning_rate": 0.0004535841778000269,
+      "loss": 0.4845,
+      "step": 539
+    },
+    {
+      "epoch": 0.23447676943117673,
+      "grad_norm": 0.052477382123470306,
+      "learning_rate": 0.0004519884870461591,
+      "loss": 0.5215,
+      "step": 540
+    },
+    {
+      "epoch": 0.2349109856708641,
+      "grad_norm": 0.04965275526046753,
+      "learning_rate": 0.00045039328968487125,
+      "loss": 0.499,
+      "step": 541
+    },
+    {
+      "epoch": 0.23534520191055144,
+      "grad_norm": 0.07106837630271912,
+      "learning_rate": 0.00044879860210928434,
+      "loss": 0.5159,
+      "step": 542
+    },
+    {
+      "epoch": 0.2357794181502388,
+      "grad_norm": 0.06256967037916183,
+      "learning_rate": 0.0004472044407072805,
+      "loss": 0.505,
+      "step": 543
+    },
+    {
+      "epoch": 0.23621363438992618,
+      "grad_norm": 0.04907793179154396,
+      "learning_rate": 0.0004456108218613346,
+      "loss": 0.5117,
+      "step": 544
+    },
+    {
+      "epoch": 0.23664785062961355,
+      "grad_norm": 0.06353382021188736,
+      "learning_rate": 0.0004440177619483461,
+      "loss": 0.4941,
+      "step": 545
+    },
+    {
+      "epoch": 0.23708206686930092,
+      "grad_norm": 0.05265835300087929,
+      "learning_rate": 0.0004424252773394704,
+      "loss": 0.5179,
+      "step": 546
+    },
+    {
+      "epoch": 0.23751628310898829,
+      "grad_norm": 0.05525769665837288,
+      "learning_rate": 0.0004408333843999506,
+      "loss": 0.4938,
+      "step": 547
+    },
+    {
+      "epoch": 0.23795049934867565,
+      "grad_norm": 0.06166142225265503,
+      "learning_rate": 0.00043924209948894985,
+      "loss": 0.5287,
+      "step": 548
+    },
+    {
+      "epoch": 0.238384715588363,
+      "grad_norm": 0.04812607169151306,
+      "learning_rate": 0.0004376514389593826,
+      "loss": 0.5401,
+      "step": 549
+    },
+    {
+      "epoch": 0.23881893182805036,
+      "grad_norm": 0.048272911459207535,
+      "learning_rate": 0.00043606141915774693,
+      "loss": 0.5258,
+      "step": 550
+    },
+    {
+      "epoch": 0.23925314806773773,
+      "grad_norm": 0.04806293919682503,
+      "learning_rate": 0.00043447205642395664,
+      "loss": 0.5023,
+      "step": 551
+    },
+    {
+      "epoch": 0.2396873643074251,
+      "grad_norm": 0.04957396164536476,
+      "learning_rate": 0.0004328833670911724,
+      "loss": 0.5015,
+      "step": 552
+    },
+    {
+      "epoch": 0.24012158054711247,
+      "grad_norm": 0.04413224756717682,
+      "learning_rate": 0.00043129536748563545,
+      "loss": 0.5336,
+      "step": 553
+    },
+    {
+      "epoch": 0.24055579678679984,
+      "grad_norm": 0.04619096592068672,
+      "learning_rate": 0.0004297080739264987,
+      "loss": 0.4832,
+      "step": 554
+    },
+    {
+      "epoch": 0.2409900130264872,
+      "grad_norm": 0.05157000198960304,
+      "learning_rate": 0.0004281215027256592,
+      "loss": 0.529,
+      "step": 555
+    },
+    {
+      "epoch": 0.24142422926617454,
+      "grad_norm": 0.04697950556874275,
+      "learning_rate": 0.00042653567018759105,
+      "loss": 0.5034,
+      "step": 556
+    },
+    {
+      "epoch": 0.2418584455058619,
+      "grad_norm": 0.04245249181985855,
+      "learning_rate": 0.0004249505926091771,
+      "loss": 0.4839,
+      "step": 557
+    },
+    {
+      "epoch": 0.24229266174554928,
+      "grad_norm": 0.042621515691280365,
+      "learning_rate": 0.000423366286279542,
+      "loss": 0.5639,
+      "step": 558
+    },
+    {
+      "epoch": 0.24272687798523665,
+      "grad_norm": 0.054264895617961884,
+      "learning_rate": 0.0004217827674798845,
+      "loss": 0.5077,
+      "step": 559
+    },
+    {
+      "epoch": 0.24316109422492402,
+      "grad_norm": 0.04171125963330269,
+      "learning_rate": 0.0004202000524833105,
+      "loss": 0.5001,
+      "step": 560
+    },
+    {
+      "epoch": 0.2435953104646114,
+      "grad_norm": 0.047443173825740814,
+      "learning_rate": 0.00041861815755466506,
+      "loss": 0.5153,
+      "step": 561
+    },
+    {
+      "epoch": 0.24402952670429873,
+      "grad_norm": 0.04359521344304085,
+      "learning_rate": 0.0004170370989503662,
+      "loss": 0.516,
+      "step": 562
+    },
+    {
+      "epoch": 0.2444637429439861,
+      "grad_norm": 0.042067334055900574,
+      "learning_rate": 0.0004154568929182374,
+      "loss": 0.4783,
+      "step": 563
+    },
+    {
+      "epoch": 0.24489795918367346,
+      "grad_norm": 0.042515527456998825,
+      "learning_rate": 0.00041387755569734057,
+      "loss": 0.5112,
+      "step": 564
+    },
+    {
+      "epoch": 0.24533217542336083,
+      "grad_norm": 0.04880267009139061,
+      "learning_rate": 0.00041229910351780926,
+      "loss": 0.5108,
+      "step": 565
+    },
+    {
+      "epoch": 0.2457663916630482,
+      "grad_norm": 0.048264361917972565,
+      "learning_rate": 0.0004107215526006817,
+      "loss": 0.4802,
+      "step": 566
+    },
+    {
+      "epoch": 0.24620060790273557,
+      "grad_norm": 0.0446770153939724,
+      "learning_rate": 0.00040914491915773453,
+      "loss": 0.5461,
+      "step": 567
+    },
+    {
+      "epoch": 0.24663482414242294,
+      "grad_norm": 0.05128175765275955,
+      "learning_rate": 0.00040756921939131565,
+      "loss": 0.5288,
+      "step": 568
+    },
+    {
+      "epoch": 0.24706904038211028,
+      "grad_norm": 0.05804765224456787,
+      "learning_rate": 0.0004059944694941783,
+      "loss": 0.4981,
+      "step": 569
+    },
+    {
+      "epoch": 0.24750325662179765,
+      "grad_norm": 0.04847874119877815,
+      "learning_rate": 0.00040442068564931397,
+      "loss": 0.5185,
+      "step": 570
+    },
+    {
+      "epoch": 0.24793747286148501,
+      "grad_norm": 0.04028409346938133,
+      "learning_rate": 0.0004028478840297866,
+      "loss": 0.493,
+      "step": 571
+    },
+    {
+      "epoch": 0.24837168910117238,
+      "grad_norm": 0.04243363067507744,
+      "learning_rate": 0.00040127608079856646,
+      "loss": 0.5079,
+      "step": 572
+    },
+    {
+      "epoch": 0.24880590534085975,
+      "grad_norm": 0.04515422508120537,
+      "learning_rate": 0.00039970529210836363,
+      "loss": 0.5121,
+      "step": 573
+    },
+    {
+      "epoch": 0.24924012158054712,
+      "grad_norm": 0.05250799283385277,
+      "learning_rate": 0.00039813553410146226,
+      "loss": 0.4975,
+      "step": 574
+    },
+    {
+      "epoch": 0.2496743378202345,
+      "grad_norm": 0.043340008705854416,
+      "learning_rate": 0.00039656682290955457,
+      "loss": 0.5094,
+      "step": 575
+    },
+    {
+      "epoch": 0.25010855405992183,
+      "grad_norm": 0.04954323545098305,
+      "learning_rate": 0.00039499917465357534,
+      "loss": 0.5327,
+      "step": 576
+    },
+    {
+      "epoch": 0.2505427702996092,
+      "grad_norm": 0.04032951593399048,
+      "learning_rate": 0.00039343260544353573,
+      "loss": 0.5035,
+      "step": 577
+    },
+    {
+      "epoch": 0.25097698653929656,
+      "grad_norm": 0.04386662319302559,
+      "learning_rate": 0.0003918671313783583,
+      "loss": 0.5142,
+      "step": 578
+    },
+    {
+      "epoch": 0.25141120277898393,
+      "grad_norm": 0.04662942886352539,
+      "learning_rate": 0.00039030276854571115,
+      "loss": 0.4996,
+      "step": 579
+    },
+    {
+      "epoch": 0.2518454190186713,
+      "grad_norm": 0.046797946095466614,
+      "learning_rate": 0.00038873953302184284,
+      "loss": 0.4728,
+      "step": 580
+    },
+    {
+      "epoch": 0.25227963525835867,
+      "grad_norm": 0.041140951216220856,
+      "learning_rate": 0.000387177440871417,
+      "loss": 0.546,
+      "step": 581
+    },
+    {
+      "epoch": 0.25271385149804604,
+      "grad_norm": 0.04658018797636032,
+      "learning_rate": 0.00038561650814734736,
+      "loss": 0.4804,
+      "step": 582
+    },
+    {
+      "epoch": 0.2531480677377334,
+      "grad_norm": 0.04261363297700882,
+      "learning_rate": 0.0003840567508906328,
+      "loss": 0.4833,
+      "step": 583
+    },
+    {
+      "epoch": 0.2535822839774208,
+      "grad_norm": 0.046267736703157425,
+      "learning_rate": 0.0003824981851301924,
+      "loss": 0.4809,
+      "step": 584
+    },
+    {
+      "epoch": 0.25401650021710814,
+      "grad_norm": 0.04030013084411621,
+      "learning_rate": 0.0003809408268827009,
+      "loss": 0.4878,
+      "step": 585
+    },
+    {
+      "epoch": 0.2544507164567955,
+      "grad_norm": 0.041512105613946915,
+      "learning_rate": 0.0003793846921524237,
+      "loss": 0.5165,
+      "step": 586
+    },
+    {
+      "epoch": 0.2548849326964828,
+      "grad_norm": 0.04772355034947395,
+      "learning_rate": 0.00037782979693105293,
+      "loss": 0.5193,
+      "step": 587
+    },
+    {
+      "epoch": 0.2553191489361702,
+      "grad_norm": 0.040986839681863785,
+      "learning_rate": 0.00037627615719754295,
+      "loss": 0.5028,
+      "step": 588
+    },
+    {
+      "epoch": 0.25575336517585756,
+      "grad_norm": 0.04139196127653122,
+      "learning_rate": 0.00037472378891794533,
+      "loss": 0.4998,
+      "step": 589
+    },
+    {
+      "epoch": 0.25618758141554493,
+      "grad_norm": 0.04017603024840355,
+      "learning_rate": 0.0003731727080452464,
+      "loss": 0.5239,
+      "step": 590
+    },
+    {
+      "epoch": 0.2566217976552323,
+      "grad_norm": 0.04039851576089859,
+      "learning_rate": 0.00037162293051920184,
+      "loss": 0.4739,
+      "step": 591
+    },
+    {
+      "epoch": 0.25705601389491967,
+      "grad_norm": 0.04436818137764931,
+      "learning_rate": 0.0003700744722661736,
+      "loss": 0.5153,
+      "step": 592
+    },
+    {
+      "epoch": 0.25749023013460703,
+      "grad_norm": 0.03977528214454651,
+      "learning_rate": 0.0003685273491989661,
+      "loss": 0.5217,
+      "step": 593
+    },
+    {
+      "epoch": 0.2579244463742944,
+      "grad_norm": 0.038483258336782455,
+      "learning_rate": 0.0003669815772166625,
+      "loss": 0.4904,
+      "step": 594
+    },
+    {
+      "epoch": 0.25835866261398177,
+      "grad_norm": 0.04520029202103615,
+      "learning_rate": 0.0003654371722044616,
+      "loss": 0.5303,
+      "step": 595
+    },
+    {
+      "epoch": 0.25879287885366914,
+      "grad_norm": 0.04642047360539436,
+      "learning_rate": 0.00036389415003351444,
+      "loss": 0.4783,
+      "step": 596
+    },
+    {
+      "epoch": 0.2592270950933565,
+      "grad_norm": 0.037170182913541794,
+      "learning_rate": 0.00036235252656076133,
+      "loss": 0.5254,
+      "step": 597
+    },
+    {
+      "epoch": 0.2596613113330439,
+      "grad_norm": 0.037870265543460846,
+      "learning_rate": 0.00036081231762876846,
+      "loss": 0.4933,
+      "step": 598
+    },
+    {
+      "epoch": 0.26009552757273124,
+      "grad_norm": 0.03957228735089302,
+      "learning_rate": 0.0003592735390655658,
+      "loss": 0.5194,
+      "step": 599
+    },
+    {
+      "epoch": 0.26052974381241856,
+      "grad_norm": 0.04471513256430626,
+      "learning_rate": 0.0003577362066844838,
+      "loss": 0.4737,
+      "step": 600
+    },
+    {
+      "epoch": 0.2609639600521059,
+      "grad_norm": 0.0370667465031147,
+      "learning_rate": 0.0003562003362839914,
+      "loss": 0.5072,
+      "step": 601
+    },
+    {
+      "epoch": 0.2613981762917933,
+      "grad_norm": 0.04277738928794861,
+      "learning_rate": 0.00035466594364753326,
+      "loss": 0.5093,
+      "step": 602
+    },
+    {
+      "epoch": 0.26183239253148066,
+      "grad_norm": 0.04622077941894531,
+      "learning_rate": 0.00035313304454336763,
+      "loss": 0.5008,
+      "step": 603
+    },
+    {
+      "epoch": 0.26226660877116803,
+      "grad_norm": 0.0380844846367836,
+      "learning_rate": 0.0003516016547244047,
+      "loss": 0.4899,
+      "step": 604
+    },
+    {
+      "epoch": 0.2627008250108554,
+      "grad_norm": 0.03964554890990257,
+      "learning_rate": 0.0003500717899280442,
+      "loss": 0.544,
+      "step": 605
+    },
+    {
+      "epoch": 0.26313504125054277,
+      "grad_norm": 0.041507843881845474,
+      "learning_rate": 0.00034854346587601397,
+      "loss": 0.5341,
+      "step": 606
+    },
+    {
+      "epoch": 0.26356925749023014,
+      "grad_norm": 0.04041410982608795,
+      "learning_rate": 0.0003470166982742082,
+      "loss": 0.5325,
+      "step": 607
+    },
+    {
+      "epoch": 0.2640034737299175,
+      "grad_norm": 0.03915121778845787,
+      "learning_rate": 0.00034549150281252633,
+      "loss": 0.5164,
+      "step": 608
+    },
+    {
+      "epoch": 0.26443768996960487,
+      "grad_norm": 0.04141729697585106,
+      "learning_rate": 0.0003439678951647115,
+      "loss": 0.4988,
+      "step": 609
+    },
+    {
+      "epoch": 0.26487190620929224,
+      "grad_norm": 0.043251313269138336,
+      "learning_rate": 0.0003424458909881897,
+      "loss": 0.5319,
+      "step": 610
+    },
+    {
+      "epoch": 0.2653061224489796,
+      "grad_norm": 0.045615166425704956,
+      "learning_rate": 0.00034092550592390857,
+      "loss": 0.4668,
+      "step": 611
+    },
+    {
+      "epoch": 0.265740338688667,
+      "grad_norm": 0.03943486511707306,
+      "learning_rate": 0.00033940675559617726,
+      "loss": 0.4991,
+      "step": 612
+    },
+    {
+      "epoch": 0.26617455492835435,
+      "grad_norm": 0.0382043793797493,
+      "learning_rate": 0.000337889655612505,
+      "loss": 0.5117,
+      "step": 613
+    },
+    {
+      "epoch": 0.26660877116804166,
+      "grad_norm": 0.039549313485622406,
+      "learning_rate": 0.00033637422156344155,
+      "loss": 0.5029,
+      "step": 614
+    },
+    {
+      "epoch": 0.267042987407729,
+      "grad_norm": 0.03972559794783592,
+      "learning_rate": 0.00033486046902241664,
+      "loss": 0.5073,
+      "step": 615
+    },
+    {
+      "epoch": 0.2674772036474164,
+      "grad_norm": 0.03974078595638275,
+      "learning_rate": 0.0003333484135455792,
+      "loss": 0.5083,
+      "step": 616
+    },
+    {
+      "epoch": 0.26791141988710376,
+      "grad_norm": 0.039242375642061234,
+      "learning_rate": 0.00033183807067163916,
+      "loss": 0.4979,
+      "step": 617
+    },
+    {
+      "epoch": 0.26834563612679113,
+      "grad_norm": 0.045767784118652344,
+      "learning_rate": 0.0003303294559217063,
+      "loss": 0.4835,
+      "step": 618
+    },
+    {
+      "epoch": 0.2687798523664785,
+      "grad_norm": 0.03927744925022125,
+      "learning_rate": 0.0003288225847991312,
+      "loss": 0.5103,
+      "step": 619
+    },
+    {
+      "epoch": 0.26921406860616587,
+      "grad_norm": 0.03815273195505142,
+      "learning_rate": 0.0003273174727893463,
+      "loss": 0.5097,
+      "step": 620
+    },
+    {
+      "epoch": 0.26964828484585324,
+      "grad_norm": 0.0421292670071125,
+      "learning_rate": 0.00032581413535970593,
+      "loss": 0.4996,
+      "step": 621
+    },
+    {
+      "epoch": 0.2700825010855406,
+      "grad_norm": 0.0482180081307888,
+      "learning_rate": 0.0003243125879593286,
+      "loss": 0.4785,
+      "step": 622
+    },
+    {
+      "epoch": 0.270516717325228,
+      "grad_norm": 0.03935736045241356,
+      "learning_rate": 0.0003228128460189368,
+      "loss": 0.5063,
+      "step": 623
+    },
+    {
+      "epoch": 0.27095093356491534,
+      "grad_norm": 0.037722837179899216,
+      "learning_rate": 0.0003213149249506997,
+      "loss": 0.4979,
+      "step": 624
+    },
+    {
+      "epoch": 0.2713851498046027,
+      "grad_norm": 0.03764335811138153,
+      "learning_rate": 0.0003198188401480734,
+      "loss": 0.4818,
+      "step": 625
+    },
+    {
+      "epoch": 0.2718193660442901,
+      "grad_norm": 0.04136700555682182,
+      "learning_rate": 0.0003183246069856443,
+      "loss": 0.5144,
+      "step": 626
+    },
+    {
+      "epoch": 0.27225358228397745,
+      "grad_norm": 0.04152214527130127,
+      "learning_rate": 0.00031683224081897,
+      "loss": 0.5217,
+      "step": 627
+    },
+    {
+      "epoch": 0.27268779852366476,
+      "grad_norm": 0.036686960607767105,
+      "learning_rate": 0.00031534175698442194,
+      "loss": 0.4938,
+      "step": 628
+    },
+    {
+      "epoch": 0.27312201476335213,
+      "grad_norm": 0.0376245342195034,
+      "learning_rate": 0.00031385317079902743,
+      "loss": 0.5255,
+      "step": 629
+    },
+    {
+      "epoch": 0.2735562310030395,
+      "grad_norm": 0.04143936559557915,
+      "learning_rate": 0.000312366497560313,
+      "loss": 0.5005,
+      "step": 630
+    },
+    {
+      "epoch": 0.27399044724272686,
+      "grad_norm": 0.036747369915246964,
+      "learning_rate": 0.00031088175254614616,
+      "loss": 0.4621,
+      "step": 631
+    },
+    {
+      "epoch": 0.27442466348241423,
+      "grad_norm": 0.039392732083797455,
+      "learning_rate": 0.00030939895101457916,
+      "loss": 0.5333,
+      "step": 632
+    },
+    {
+      "epoch": 0.2748588797221016,
+      "grad_norm": 0.03783806785941124,
+      "learning_rate": 0.0003079181082036922,
+      "loss": 0.5075,
+      "step": 633
+    },
+    {
+      "epoch": 0.27529309596178897,
+      "grad_norm": 0.03799246624112129,
+      "learning_rate": 0.000306439239331436,
+      "loss": 0.5086,
+      "step": 634
+    },
+    {
+      "epoch": 0.27572731220147634,
+      "grad_norm": 0.040941022336483,
+      "learning_rate": 0.0003049623595954766,
+      "loss": 0.5125,
+      "step": 635
+    },
+    {
+      "epoch": 0.2761615284411637,
+      "grad_norm": 0.045769453048706055,
+      "learning_rate": 0.0003034874841730382,
+      "loss": 0.4913,
+      "step": 636
+    },
+    {
+      "epoch": 0.2765957446808511,
+      "grad_norm": 0.044418562203645706,
+      "learning_rate": 0.00030201462822074786,
+      "loss": 0.5225,
+      "step": 637
+    },
+    {
+      "epoch": 0.27702996092053844,
+      "grad_norm": 0.036289479583501816,
+      "learning_rate": 0.0003005438068744792,
+      "loss": 0.5016,
+      "step": 638
+    },
+    {
+      "epoch": 0.2774641771602258,
+      "grad_norm": 0.03927507996559143,
+      "learning_rate": 0.0002990750352491973,
+      "loss": 0.497,
+      "step": 639
+    },
+    {
+      "epoch": 0.2778983933999132,
+      "grad_norm": 0.052069611847400665,
+      "learning_rate": 0.0002976083284388031,
+      "loss": 0.5026,
+      "step": 640
+    },
+    {
+      "epoch": 0.27833260963960055,
+      "grad_norm": 0.0408078134059906,
+      "learning_rate": 0.00029614370151597835,
+      "loss": 0.5069,
+      "step": 641
+    },
+    {
+      "epoch": 0.27876682587928786,
+      "grad_norm": 0.040615539997816086,
+      "learning_rate": 0.0002946811695320311,
+      "loss": 0.4878,
+      "step": 642
+    },
+    {
+      "epoch": 0.27920104211897523,
+      "grad_norm": 0.04145561158657074,
+      "learning_rate": 0.00029322074751673977,
+      "loss": 0.4547,
+      "step": 643
+    },
+    {
+      "epoch": 0.2796352583586626,
+      "grad_norm": 0.044274091720581055,
+      "learning_rate": 0.00029176245047820063,
+      "loss": 0.5051,
+      "step": 644
+    },
+    {
+      "epoch": 0.28006947459834997,
+      "grad_norm": 0.055839963257312775,
+      "learning_rate": 0.0002903062934026716,
+      "loss": 0.495,
+      "step": 645
+    },
+    {
+      "epoch": 0.28050369083803733,
+      "grad_norm": 0.038132019340991974,
+      "learning_rate": 0.0002888522912544202,
+      "loss": 0.4746,
+      "step": 646
+    },
+    {
+      "epoch": 0.2809379070777247,
+      "grad_norm": 0.039124827831983566,
+      "learning_rate": 0.00028740045897556767,
+      "loss": 0.4978,
+      "step": 647
+    },
+    {
+      "epoch": 0.28137212331741207,
+      "grad_norm": 0.03960909694433212,
+      "learning_rate": 0.0002859508114859374,
+      "loss": 0.5043,
+      "step": 648
+    },
+    {
+      "epoch": 0.28180633955709944,
+      "grad_norm": 0.03705562651157379,
+      "learning_rate": 0.00028450336368289974,
+      "loss": 0.5041,
+      "step": 649
+    },
+    {
+      "epoch": 0.2822405557967868,
+      "grad_norm": 0.03920963406562805,
+      "learning_rate": 0.00028305813044122096,
+      "loss": 0.5055,
+      "step": 650
+    },
+    {
+      "epoch": 0.2826747720364742,
+      "grad_norm": 0.04196755215525627,
+      "learning_rate": 0.00028161512661290845,
+      "loss": 0.4962,
+      "step": 651
+    },
+    {
+      "epoch": 0.28310898827616154,
+      "grad_norm": 0.03515279293060303,
+      "learning_rate": 0.00028017436702705903,
+      "loss": 0.4708,
+      "step": 652
+    },
+    {
+      "epoch": 0.2835432045158489,
+      "grad_norm": 0.042391590774059296,
+      "learning_rate": 0.0002787358664897068,
+      "loss": 0.5268,
+      "step": 653
+    },
+    {
+      "epoch": 0.2839774207555363,
+      "grad_norm": 0.03707970678806305,
+      "learning_rate": 0.0002772996397836704,
+      "loss": 0.5252,
+      "step": 654
+    },
+    {
+      "epoch": 0.28441163699522365,
+      "grad_norm": 0.039679594337940216,
+      "learning_rate": 0.0002758657016684015,
+      "loss": 0.4954,
+      "step": 655
+    },
+    {
+      "epoch": 0.28484585323491096,
+      "grad_norm": 0.03491951525211334,
+      "learning_rate": 0.0002744340668798326,
+      "loss": 0.4859,
+      "step": 656
+    },
+    {
+      "epoch": 0.28528006947459833,
+      "grad_norm": 0.033995699137449265,
+      "learning_rate": 0.00027300475013022663,
+      "loss": 0.4917,
+      "step": 657
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 0.0408964566886425,
+      "learning_rate": 0.00027157776610802414,
+      "loss": 0.5176,
+      "step": 658
+    },
+    {
+      "epoch": 0.28614850195397307,
+      "grad_norm": 0.035840053111314774,
+      "learning_rate": 0.0002701531294776943,
+      "loss": 0.4757,
+      "step": 659
+    },
+    {
+      "epoch": 0.28658271819366044,
+      "grad_norm": 0.03554424270987511,
+      "learning_rate": 0.0002687308548795825,
+      "loss": 0.5043,
+      "step": 660
+    },
+    {
+      "epoch": 0.2870169344333478,
+      "grad_norm": 0.03549895063042641,
+      "learning_rate": 0.00026731095692976073,
+      "loss": 0.4924,
+      "step": 661
+    },
+    {
+      "epoch": 0.28745115067303517,
+      "grad_norm": 0.0438251756131649,
+      "learning_rate": 0.00026589345021987723,
+      "loss": 0.4986,
+      "step": 662
+    },
+    {
+      "epoch": 0.28788536691272254,
+      "grad_norm": 0.037839341908693314,
+      "learning_rate": 0.00026447834931700686,
+      "loss": 0.4716,
+      "step": 663
+    },
+    {
+      "epoch": 0.2883195831524099,
+      "grad_norm": 0.03682737424969673,
+      "learning_rate": 0.0002630656687635007,
+      "loss": 0.4991,
+      "step": 664
+    },
+    {
+      "epoch": 0.2887537993920973,
+      "grad_norm": 0.03697109594941139,
+      "learning_rate": 0.0002616554230768374,
+      "loss": 0.5104,
+      "step": 665
+    },
+    {
+      "epoch": 0.28918801563178465,
+      "grad_norm": 0.036725230515003204,
+      "learning_rate": 0.0002602476267494731,
+      "loss": 0.5163,
+      "step": 666
+    },
+    {
+      "epoch": 0.289622231871472,
+      "grad_norm": 0.04177143797278404,
+      "learning_rate": 0.0002588422942486932,
+      "loss": 0.4567,
+      "step": 667
+    },
+    {
+      "epoch": 0.2900564481111594,
+      "grad_norm": 0.04078909382224083,
+      "learning_rate": 0.0002574394400164639,
+      "loss": 0.4946,
+      "step": 668
+    },
+    {
+      "epoch": 0.2904906643508467,
+      "grad_norm": 0.04146347567439079,
+      "learning_rate": 0.0002560390784692828,
+      "loss": 0.4947,
+      "step": 669
+    },
+    {
+      "epoch": 0.29092488059053406,
+      "grad_norm": 0.03871840611100197,
+      "learning_rate": 0.00025464122399803123,
+      "loss": 0.5238,
+      "step": 670
+    },
+    {
+      "epoch": 0.29135909683022143,
+      "grad_norm": 0.04148663207888603,
+      "learning_rate": 0.00025324589096782657,
+      "loss": 0.5246,
+      "step": 671
+    },
+    {
+      "epoch": 0.2917933130699088,
+      "grad_norm": 0.03910430520772934,
+      "learning_rate": 0.0002518530937178751,
+      "loss": 0.4846,
+      "step": 672
+    },
+    {
+      "epoch": 0.29222752930959617,
+      "grad_norm": 0.036824680864810944,
+      "learning_rate": 0.000250462846561323,
+      "loss": 0.4654,
+      "step": 673
+    },
+    {
+      "epoch": 0.29266174554928354,
+      "grad_norm": 0.036250337958335876,
+      "learning_rate": 0.00024907516378511136,
+      "loss": 0.5106,
+      "step": 674
+    },
+    {
+      "epoch": 0.2930959617889709,
+      "grad_norm": 0.03372564911842346,
+      "learning_rate": 0.00024769005964982713,
+      "loss": 0.5222,
+      "step": 675
+    },
+    {
+      "epoch": 0.2935301780286583,
+      "grad_norm": 0.03601228445768356,
+      "learning_rate": 0.000246307548389559,
+      "loss": 0.5099,
+      "step": 676
+    },
+    {
+      "epoch": 0.29396439426834564,
+      "grad_norm": 0.039937473833560944,
+      "learning_rate": 0.0002449276442117486,
+      "loss": 0.5262,
+      "step": 677
+    },
+    {
+      "epoch": 0.294398610508033,
+      "grad_norm": 0.03719143569469452,
+      "learning_rate": 0.000243550361297047,
+      "loss": 0.5089,
+      "step": 678
+    },
+    {
+      "epoch": 0.2948328267477204,
+      "grad_norm": 0.034734416753053665,
+      "learning_rate": 0.00024217571379916668,
+      "loss": 0.4962,
+      "step": 679
+    },
+    {
+      "epoch": 0.29526704298740775,
+      "grad_norm": 0.03447417914867401,
+      "learning_rate": 0.00024080371584473748,
+      "loss": 0.4952,
+      "step": 680
+    },
+    {
+      "epoch": 0.2957012592270951,
+      "grad_norm": 0.03631613776087761,
+      "learning_rate": 0.00023943438153316155,
+      "loss": 0.497,
+      "step": 681
+    },
+    {
+      "epoch": 0.2961354754667825,
+      "grad_norm": 0.03350459039211273,
+      "learning_rate": 0.00023806772493646723,
+      "loss": 0.4775,
+      "step": 682
+    },
+    {
+      "epoch": 0.2965696917064698,
+      "grad_norm": 0.032677456736564636,
+      "learning_rate": 0.00023670376009916595,
+      "loss": 0.5001,
+      "step": 683
+    },
+    {
+      "epoch": 0.29700390794615716,
+      "grad_norm": 0.03544562682509422,
+      "learning_rate": 0.00023534250103810628,
+      "loss": 0.4796,
+      "step": 684
+    },
+    {
+      "epoch": 0.29743812418584453,
+      "grad_norm": 0.03991761803627014,
+      "learning_rate": 0.00023398396174233177,
+      "loss": 0.5307,
+      "step": 685
+    },
+    {
+      "epoch": 0.2978723404255319,
+      "grad_norm": 0.037654612213373184,
+      "learning_rate": 0.00023262815617293515,
+      "loss": 0.4727,
+      "step": 686
+    },
+    {
+      "epoch": 0.29830655666521927,
+      "grad_norm": 0.03447496145963669,
+      "learning_rate": 0.00023127509826291698,
+      "loss": 0.5037,
+      "step": 687
+    },
+    {
+      "epoch": 0.29874077290490664,
+      "grad_norm": 0.03626209869980812,
+      "learning_rate": 0.00022992480191704002,
+      "loss": 0.4758,
+      "step": 688
+    },
+    {
+      "epoch": 0.299174989144594,
+      "grad_norm": 0.03519487380981445,
+      "learning_rate": 0.000228577281011689,
+      "loss": 0.5046,
+      "step": 689
+    },
+    {
+      "epoch": 0.2996092053842814,
+      "grad_norm": 0.036514561623334885,
+      "learning_rate": 0.0002272325493947257,
+      "loss": 0.5014,
+      "step": 690
+    },
+    {
+      "epoch": 0.30004342162396874,
+      "grad_norm": 0.036256395280361176,
+      "learning_rate": 0.00022589062088534834,
+      "loss": 0.5311,
+      "step": 691
+    },
+    {
+      "epoch": 0.3004776378636561,
+      "grad_norm": 0.036770399659872055,
+      "learning_rate": 0.0002245515092739488,
+      "loss": 0.4922,
+      "step": 692
+    },
+    {
+      "epoch": 0.3009118541033435,
+      "grad_norm": 0.037295542657375336,
+      "learning_rate": 0.00022321522832197032,
+      "loss": 0.4817,
+      "step": 693
+    },
+    {
+      "epoch": 0.30134607034303085,
+      "grad_norm": 0.03913348540663719,
+      "learning_rate": 0.00022188179176176764,
+      "loss": 0.4565,
+      "step": 694
+    },
+    {
+      "epoch": 0.3017802865827182,
+      "grad_norm": 0.04223249852657318,
+      "learning_rate": 0.00022055121329646416,
+      "loss": 0.5297,
+      "step": 695
+    },
+    {
+      "epoch": 0.3022145028224056,
+      "grad_norm": 0.03583509102463722,
+      "learning_rate": 0.0002192235065998126,
+      "loss": 0.4808,
+      "step": 696
+    },
+    {
+      "epoch": 0.3026487190620929,
+      "grad_norm": 0.04437342286109924,
+      "learning_rate": 0.0002178986853160535,
+      "loss": 0.4794,
+      "step": 697
+    },
+    {
+      "epoch": 0.30308293530178027,
+      "grad_norm": 0.043945055454969406,
+      "learning_rate": 0.0002165767630597752,
+      "loss": 0.4959,
+      "step": 698
+    },
+    {
+      "epoch": 0.30351715154146763,
+      "grad_norm": 0.04786218702793121,
+      "learning_rate": 0.00021525775341577403,
+      "loss": 0.5047,
+      "step": 699
+    },
+    {
+      "epoch": 0.303951367781155,
+      "grad_norm": 0.041139792650938034,
+      "learning_rate": 0.0002139416699389153,
+      "loss": 0.5134,
+      "step": 700
+    },
+    {
+      "epoch": 0.30438558402084237,
+      "grad_norm": 0.034917134791612625,
+      "learning_rate": 0.00021262852615399258,
+      "loss": 0.4924,
+      "step": 701
+    },
+    {
+      "epoch": 0.30481980026052974,
+      "grad_norm": 0.03301383554935455,
+      "learning_rate": 0.0002113183355555904,
+      "loss": 0.4862,
+      "step": 702
+    },
+    {
+      "epoch": 0.3052540165002171,
+      "grad_norm": 0.0419352687895298,
+      "learning_rate": 0.00021001111160794383,
+      "loss": 0.5147,
+      "step": 703
+    },
+    {
+      "epoch": 0.3056882327399045,
+      "grad_norm": 0.03364208713173866,
+      "learning_rate": 0.00020870686774480197,
+      "loss": 0.5236,
+      "step": 704
+    },
+    {
+      "epoch": 0.30612244897959184,
+      "grad_norm": 0.03912827745079994,
+      "learning_rate": 0.0002074056173692881,
+      "loss": 0.5042,
+      "step": 705
+    },
+    {
+      "epoch": 0.3065566652192792,
+      "grad_norm": 0.03657132387161255,
+      "learning_rate": 0.00020610737385376348,
+      "loss": 0.5011,
+      "step": 706
+    },
+    {
+      "epoch": 0.3069908814589666,
+      "grad_norm": 0.03518640622496605,
+      "learning_rate": 0.00020481215053968872,
+      "loss": 0.4798,
+      "step": 707
+    },
+    {
+      "epoch": 0.30742509769865395,
+      "grad_norm": 0.03507522493600845,
+      "learning_rate": 0.00020351996073748714,
+      "loss": 0.5313,
+      "step": 708
+    },
+    {
+      "epoch": 0.3078593139383413,
+      "grad_norm": 0.03609062731266022,
+      "learning_rate": 0.00020223081772640866,
+      "loss": 0.4785,
+      "step": 709
+    },
+    {
+      "epoch": 0.3082935301780287,
+      "grad_norm": 0.03479000926017761,
+      "learning_rate": 0.000200944734754392,
+      "loss": 0.4835,
+      "step": 710
+    },
+    {
+      "epoch": 0.308727746417716,
+      "grad_norm": 0.044929634779691696,
+      "learning_rate": 0.00019966172503792985,
+      "loss": 0.5122,
+      "step": 711
+    },
+    {
+      "epoch": 0.30916196265740337,
+      "grad_norm": 0.03376319631934166,
+      "learning_rate": 0.00019838180176193177,
+      "loss": 0.4876,
+      "step": 712
+    },
+    {
+      "epoch": 0.30959617889709073,
+      "grad_norm": 0.03246736153960228,
+      "learning_rate": 0.0001971049780795901,
+      "loss": 0.4804,
+      "step": 713
+    },
+    {
+      "epoch": 0.3100303951367781,
+      "grad_norm": 0.031609781086444855,
+      "learning_rate": 0.00019583126711224342,
+      "loss": 0.545,
+      "step": 714
+    },
+    {
+      "epoch": 0.31046461137646547,
+      "grad_norm": 0.034195881336927414,
+      "learning_rate": 0.00019456068194924288,
+      "loss": 0.5096,
+      "step": 715
+    },
+    {
+      "epoch": 0.31089882761615284,
+      "grad_norm": 0.03344978764653206,
+      "learning_rate": 0.0001932932356478168,
+      "loss": 0.5079,
+      "step": 716
+    },
+    {
+      "epoch": 0.3113330438558402,
+      "grad_norm": 0.03219503536820412,
+      "learning_rate": 0.00019202894123293674,
+      "loss": 0.5339,
+      "step": 717
+    },
+    {
+      "epoch": 0.3117672600955276,
+      "grad_norm": 0.03737034648656845,
+      "learning_rate": 0.00019076781169718426,
+      "loss": 0.5131,
+      "step": 718
+    },
+    {
+      "epoch": 0.31220147633521494,
+      "grad_norm": 0.03394628316164017,
+      "learning_rate": 0.00018950986000061638,
+      "loss": 0.4969,
+      "step": 719
+    },
+    {
+      "epoch": 0.3126356925749023,
+      "grad_norm": 0.032426606863737106,
+      "learning_rate": 0.00018825509907063325,
+      "loss": 0.49,
+      "step": 720
+    },
+    {
+      "epoch": 0.3130699088145897,
+      "grad_norm": 0.03836345672607422,
+      "learning_rate": 0.00018700354180184464,
+      "loss": 0.4925,
+      "step": 721
+    },
+    {
+      "epoch": 0.31350412505427705,
+      "grad_norm": 0.035867128521203995,
+      "learning_rate": 0.0001857552010559382,
+      "loss": 0.5466,
+      "step": 722
+    },
+    {
+      "epoch": 0.3139383412939644,
+      "grad_norm": 0.03811186924576759,
+      "learning_rate": 0.0001845100896615462,
+      "loss": 0.4825,
+      "step": 723
+    },
+    {
+      "epoch": 0.31437255753365173,
+      "grad_norm": 0.032081685960292816,
+      "learning_rate": 0.00018326822041411523,
+      "loss": 0.4858,
+      "step": 724
+    },
+    {
+      "epoch": 0.3148067737733391,
+      "grad_norm": 0.03794045001268387,
+      "learning_rate": 0.00018202960607577247,
+      "loss": 0.5023,
+      "step": 725
+    },
+    {
+      "epoch": 0.31524099001302647,
+      "grad_norm": 0.033024583011865616,
+      "learning_rate": 0.00018079425937519728,
+      "loss": 0.4744,
+      "step": 726
+    },
+    {
+      "epoch": 0.31567520625271384,
+      "grad_norm": 0.03800756111741066,
+      "learning_rate": 0.00017956219300748795,
+      "loss": 0.4813,
+      "step": 727
+    },
+    {
+      "epoch": 0.3161094224924012,
+      "grad_norm": 0.03822485730051994,
+      "learning_rate": 0.0001783334196340331,
+      "loss": 0.4984,
+      "step": 728
+    },
+    {
+      "epoch": 0.3165436387320886,
+      "grad_norm": 0.03531699255108833,
+      "learning_rate": 0.00017710795188237988,
+      "loss": 0.5095,
+      "step": 729
+    },
+    {
+      "epoch": 0.31697785497177594,
+      "grad_norm": 0.03231954947113991,
+      "learning_rate": 0.0001758858023461059,
+      "loss": 0.4795,
+      "step": 730
+    },
+    {
+      "epoch": 0.3174120712114633,
+      "grad_norm": 0.04151546210050583,
+      "learning_rate": 0.00017466698358468825,
+      "loss": 0.4852,
+      "step": 731
+    },
+    {
+      "epoch": 0.3178462874511507,
+      "grad_norm": 0.03277713060379028,
+      "learning_rate": 0.00017345150812337563,
+      "loss": 0.4736,
+      "step": 732
+    },
+    {
+      "epoch": 0.31828050369083805,
+      "grad_norm": 0.03749159723520279,
+      "learning_rate": 0.0001722393884530593,
+      "loss": 0.5042,
+      "step": 733
+    },
+    {
+      "epoch": 0.3187147199305254,
+      "grad_norm": 0.029695043340325356,
+      "learning_rate": 0.00017103063703014372,
+      "loss": 0.511,
+      "step": 734
+    },
+    {
+      "epoch": 0.3191489361702128,
+      "grad_norm": 0.033079251646995544,
+      "learning_rate": 0.00016982526627642042,
+      "loss": 0.4871,
+      "step": 735
+    },
+    {
+      "epoch": 0.31958315240990015,
+      "grad_norm": 0.03277754411101341,
+      "learning_rate": 0.00016862328857893854,
+      "loss": 0.4805,
+      "step": 736
+    },
+    {
+      "epoch": 0.3200173686495875,
+      "grad_norm": 0.031276337802410126,
+      "learning_rate": 0.00016742471628987892,
+      "loss": 0.5195,
+      "step": 737
+    },
+    {
+      "epoch": 0.32045158488927483,
+      "grad_norm": 0.031646616756916046,
+      "learning_rate": 0.000166229561726426,
+      "loss": 0.4899,
+      "step": 738
+    },
+    {
+      "epoch": 0.3208858011289622,
+      "grad_norm": 0.03364792838692665,
+      "learning_rate": 0.00016503783717064247,
+      "loss": 0.5011,
+      "step": 739
+    },
+    {
+      "epoch": 0.32132001736864957,
+      "grad_norm": 0.032123491168022156,
+      "learning_rate": 0.00016384955486934156,
+      "loss": 0.5083,
+      "step": 740
+    },
+    {
+      "epoch": 0.32175423360833694,
+      "grad_norm": 0.03388531878590584,
+      "learning_rate": 0.00016266472703396284,
+      "loss": 0.5127,
+      "step": 741
+    },
+    {
+      "epoch": 0.3221884498480243,
+      "grad_norm": 0.03697797656059265,
+      "learning_rate": 0.00016148336584044537,
+      "loss": 0.5167,
+      "step": 742
+    },
+    {
+      "epoch": 0.3226226660877117,
+      "grad_norm": 0.03302115574479103,
+      "learning_rate": 0.000160305483429103,
+      "loss": 0.483,
+      "step": 743
+    },
+    {
+      "epoch": 0.32305688232739904,
+      "grad_norm": 0.03180578723549843,
+      "learning_rate": 0.0001591310919045003,
+      "loss": 0.5144,
+      "step": 744
+    },
+    {
+      "epoch": 0.3234910985670864,
+      "grad_norm": 0.033885449171066284,
+      "learning_rate": 0.00015796020333532697,
+      "loss": 0.484,
+      "step": 745
+    },
+    {
+      "epoch": 0.3239253148067738,
+      "grad_norm": 0.03529027849435806,
+      "learning_rate": 0.0001567928297542749,
+      "loss": 0.4967,
+      "step": 746
+    },
+    {
+      "epoch": 0.32435953104646115,
+      "grad_norm": 0.030900994315743446,
+      "learning_rate": 0.00015562898315791353,
+      "loss": 0.5177,
+      "step": 747
+    },
+    {
+      "epoch": 0.3247937472861485,
+      "grad_norm": 0.03337928652763367,
+      "learning_rate": 0.00015446867550656767,
+      "loss": 0.4902,
+      "step": 748
+    },
+    {
+      "epoch": 0.3252279635258359,
+      "grad_norm": 0.028887873515486717,
+      "learning_rate": 0.00015331191872419348,
+      "loss": 0.4841,
+      "step": 749
+    },
+    {
+      "epoch": 0.32566217976552325,
+      "grad_norm": 0.03361840918660164,
+      "learning_rate": 0.0001521587246982568,
+      "loss": 0.5171,
+      "step": 750
+    },
+    {
+      "epoch": 0.3260963960052106,
+      "grad_norm": 0.030776534229516983,
+      "learning_rate": 0.00015100910527961049,
+      "loss": 0.4592,
+      "step": 751
+    },
+    {
+      "epoch": 0.32653061224489793,
+      "grad_norm": 0.028537072241306305,
+      "learning_rate": 0.00014986307228237266,
+      "loss": 0.4894,
+      "step": 752
+    },
+    {
+      "epoch": 0.3269648284845853,
+      "grad_norm": 0.02929234690964222,
+      "learning_rate": 0.00014872063748380543,
+      "loss": 0.4832,
+      "step": 753
+    },
+    {
+      "epoch": 0.32739904472427267,
+      "grad_norm": 0.02816353552043438,
+      "learning_rate": 0.00014758181262419424,
+      "loss": 0.4788,
+      "step": 754
+    },
+    {
+      "epoch": 0.32783326096396004,
+      "grad_norm": 0.03165418654680252,
+      "learning_rate": 0.00014644660940672628,
+      "loss": 0.5128,
+      "step": 755
+    },
+    {
+      "epoch": 0.3282674772036474,
+      "grad_norm": 0.0279870443046093,
+      "learning_rate": 0.00014531503949737106,
+      "loss": 0.4966,
+      "step": 756
+    },
+    {
+      "epoch": 0.3287016934433348,
+      "grad_norm": 0.029440978541970253,
+      "learning_rate": 0.00014418711452476046,
+      "loss": 0.4785,
+      "step": 757
+    },
+    {
+      "epoch": 0.32913590968302214,
+      "grad_norm": 0.032016150653362274,
+      "learning_rate": 0.00014306284608006837,
+      "loss": 0.4813,
+      "step": 758
+    },
+    {
+      "epoch": 0.3295701259227095,
+      "grad_norm": 0.029074901714920998,
+      "learning_rate": 0.00014194224571689284,
+      "loss": 0.5165,
+      "step": 759
+    },
+    {
+      "epoch": 0.3300043421623969,
+      "grad_norm": 0.030839256942272186,
+      "learning_rate": 0.00014082532495113625,
+      "loss": 0.4615,
+      "step": 760
+    },
+    {
+      "epoch": 0.33043855840208425,
+      "grad_norm": 0.03017420321702957,
+      "learning_rate": 0.00013971209526088762,
+      "loss": 0.473,
+      "step": 761
+    },
+    {
+      "epoch": 0.3308727746417716,
+      "grad_norm": 0.03192548081278801,
+      "learning_rate": 0.00013860256808630427,
+      "loss": 0.5315,
+      "step": 762
+    },
+    {
+      "epoch": 0.331306990881459,
+      "grad_norm": 0.029673421755433083,
+      "learning_rate": 0.00013749675482949486,
+      "loss": 0.4978,
+      "step": 763
+    },
+    {
+      "epoch": 0.33174120712114635,
+      "grad_norm": 0.035233914852142334,
+      "learning_rate": 0.00013639466685440134,
+      "loss": 0.5174,
+      "step": 764
+    },
+    {
+      "epoch": 0.3321754233608337,
+      "grad_norm": 0.03196287900209427,
+      "learning_rate": 0.00013529631548668298,
+      "loss": 0.5121,
+      "step": 765
+    },
+    {
+      "epoch": 0.33260963960052103,
+      "grad_norm": 0.028510358184576035,
+      "learning_rate": 0.0001342017120135993,
+      "loss": 0.4629,
+      "step": 766
+    },
+    {
+      "epoch": 0.3330438558402084,
+      "grad_norm": 0.032123226672410965,
+      "learning_rate": 0.0001331108676838948,
+      "loss": 0.5538,
+      "step": 767
+    },
+    {
+      "epoch": 0.33347807207989577,
+      "grad_norm": 0.0301981084048748,
+      "learning_rate": 0.00013202379370768252,
+      "loss": 0.5278,
+      "step": 768
+    },
+    {
+      "epoch": 0.33391228831958314,
+      "grad_norm": 0.029721124097704887,
+      "learning_rate": 0.00013094050125632973,
+      "loss": 0.5353,
+      "step": 769
+    },
+    {
+      "epoch": 0.3343465045592705,
+      "grad_norm": 0.029476812109351158,
+      "learning_rate": 0.00012986100146234231,
+      "loss": 0.488,
+      "step": 770
+    },
+    {
+      "epoch": 0.3347807207989579,
+      "grad_norm": 0.029208241030573845,
+      "learning_rate": 0.00012878530541925076,
+      "loss": 0.5384,
+      "step": 771
+    },
+    {
+      "epoch": 0.33521493703864524,
+      "grad_norm": 0.029901737347245216,
+      "learning_rate": 0.00012771342418149657,
+      "loss": 0.4675,
+      "step": 772
+    },
+    {
+      "epoch": 0.3356491532783326,
+      "grad_norm": 0.030133001506328583,
+      "learning_rate": 0.00012664536876431755,
+      "loss": 0.5175,
+      "step": 773
+    },
+    {
+      "epoch": 0.33608336951802,
+      "grad_norm": 0.028815865516662598,
+      "learning_rate": 0.0001255811501436359,
+      "loss": 0.5012,
+      "step": 774
+    },
+    {
+      "epoch": 0.33651758575770735,
+      "grad_norm": 0.02868812158703804,
+      "learning_rate": 0.00012452077925594434,
+      "loss": 0.521,
+      "step": 775
+    },
+    {
+      "epoch": 0.3369518019973947,
+      "grad_norm": 0.035325054079294205,
+      "learning_rate": 0.00012346426699819457,
+      "loss": 0.5075,
+      "step": 776
+    },
+    {
+      "epoch": 0.3373860182370821,
+      "grad_norm": 0.03588160499930382,
+      "learning_rate": 0.00012241162422768444,
+      "loss": 0.5091,
+      "step": 777
+    },
+    {
+      "epoch": 0.33782023447676945,
+      "grad_norm": 0.03347684070467949,
+      "learning_rate": 0.00012136286176194744,
+      "loss": 0.5039,
+      "step": 778
+    },
+    {
+      "epoch": 0.3382544507164568,
+      "grad_norm": 0.03358198329806328,
+      "learning_rate": 0.00012031799037864011,
+      "loss": 0.5014,
+      "step": 779
+    },
+    {
+      "epoch": 0.33868866695614414,
+      "grad_norm": 0.02983028069138527,
+      "learning_rate": 0.0001192770208154328,
+      "loss": 0.4823,
+      "step": 780
+    },
+    {
+      "epoch": 0.3391228831958315,
+      "grad_norm": 0.028506051748991013,
+      "learning_rate": 0.00011823996376989849,
+      "loss": 0.5135,
+      "step": 781
+    },
+    {
+      "epoch": 0.33955709943551887,
+      "grad_norm": 0.031358085572719574,
+      "learning_rate": 0.00011720682989940262,
+      "loss": 0.4677,
+      "step": 782
+    },
+    {
+      "epoch": 0.33999131567520624,
+      "grad_norm": 0.031006425619125366,
+      "learning_rate": 0.00011617762982099444,
+      "loss": 0.5188,
+      "step": 783
+    },
+    {
+      "epoch": 0.3404255319148936,
+      "grad_norm": 0.0309920571744442,
+      "learning_rate": 0.00011515237411129698,
+      "loss": 0.5087,
+      "step": 784
+    },
+    {
+      "epoch": 0.340859748154581,
+      "grad_norm": 0.027928480878472328,
+      "learning_rate": 0.0001141310733063991,
+      "loss": 0.5005,
+      "step": 785
+    },
+    {
+      "epoch": 0.34129396439426835,
+      "grad_norm": 0.028440937399864197,
+      "learning_rate": 0.00011311373790174655,
+      "loss": 0.5055,
+      "step": 786
+    },
+    {
+      "epoch": 0.3417281806339557,
+      "grad_norm": 0.03135620057582855,
+      "learning_rate": 0.00011210037835203507,
+      "loss": 0.4947,
+      "step": 787
+    },
+    {
+      "epoch": 0.3421623968736431,
+      "grad_norm": 0.029479345306754112,
+      "learning_rate": 0.00011109100507110132,
+      "loss": 0.5048,
+      "step": 788
+    },
+    {
+      "epoch": 0.34259661311333045,
+      "grad_norm": 0.030894558876752853,
+      "learning_rate": 0.00011008562843181796,
+      "loss": 0.4726,
+      "step": 789
+    },
+    {
+      "epoch": 0.3430308293530178,
+      "grad_norm": 0.03545878455042839,
+      "learning_rate": 0.0001090842587659851,
+      "loss": 0.5165,
+      "step": 790
+    },
+    {
+      "epoch": 0.3434650455927052,
+      "grad_norm": 0.028414282947778702,
+      "learning_rate": 0.00010808690636422585,
+      "loss": 0.5008,
+      "step": 791
+    },
+    {
+      "epoch": 0.34389926183239256,
+      "grad_norm": 0.029046185314655304,
+      "learning_rate": 0.00010709358147587883,
+      "loss": 0.5067,
+      "step": 792
+    },
+    {
+      "epoch": 0.34433347807207987,
+      "grad_norm": 0.029067449271678925,
+      "learning_rate": 0.00010610429430889452,
+      "loss": 0.5039,
+      "step": 793
+    },
+    {
+      "epoch": 0.34476769431176724,
+      "grad_norm": 0.03735330328345299,
+      "learning_rate": 0.00010511905502972886,
+      "loss": 0.4874,
+      "step": 794
+    },
+    {
+      "epoch": 0.3452019105514546,
+      "grad_norm": 0.025629056617617607,
+      "learning_rate": 0.00010413787376324018,
+      "loss": 0.4743,
+      "step": 795
+    },
+    {
+      "epoch": 0.345636126791142,
+      "grad_norm": 0.03742838650941849,
+      "learning_rate": 0.00010316076059258389,
+      "loss": 0.4706,
+      "step": 796
+    },
+    {
+      "epoch": 0.34607034303082934,
+      "grad_norm": 0.028506414964795113,
+      "learning_rate": 0.00010218772555910954,
+      "loss": 0.4677,
+      "step": 797
+    },
+    {
+      "epoch": 0.3465045592705167,
+      "grad_norm": 0.028791414573788643,
+      "learning_rate": 0.00010121877866225782,
+      "loss": 0.4786,
+      "step": 798
+    },
+    {
+      "epoch": 0.3469387755102041,
+      "grad_norm": 0.027565713971853256,
+      "learning_rate": 0.00010025392985945703,
+      "loss": 0.4756,
+      "step": 799
+    },
+    {
+      "epoch": 0.34737299174989145,
+      "grad_norm": 0.037437453866004944,
+      "learning_rate": 9.929318906602175e-05,
+      "loss": 0.4941,
+      "step": 800
+    },
+    {
+      "epoch": 0.3478072079895788,
+      "grad_norm": 0.028768004849553108,
+      "learning_rate": 9.833656615504976e-05,
+      "loss": 0.5263,
+      "step": 801
+    },
+    {
+      "epoch": 0.3482414242292662,
+      "grad_norm": 0.030494874343276024,
+      "learning_rate": 9.738407095732193e-05,
+      "loss": 0.4933,
+      "step": 802
+    },
+    {
+      "epoch": 0.34867564046895355,
+      "grad_norm": 0.02905452810227871,
+      "learning_rate": 9.643571326119983e-05,
+      "loss": 0.4751,
+      "step": 803
+    },
+    {
+      "epoch": 0.3491098567086409,
+      "grad_norm": 0.02821405418217182,
+      "learning_rate": 9.549150281252633e-05,
+      "loss": 0.4874,
+      "step": 804
+    },
+    {
+      "epoch": 0.3495440729483283,
+      "grad_norm": 0.02889527939260006,
+      "learning_rate": 9.455144931452459e-05,
+      "loss": 0.4732,
+      "step": 805
+    },
+    {
+      "epoch": 0.34997828918801566,
+      "grad_norm": 0.02988606132566929,
+      "learning_rate": 9.36155624276987e-05,
+      "loss": 0.5044,
+      "step": 806
+    },
+    {
+      "epoch": 0.35041250542770297,
+      "grad_norm": 0.03306758031249046,
+      "learning_rate": 9.26838517697346e-05,
+      "loss": 0.492,
+      "step": 807
+    },
+    {
+      "epoch": 0.35084672166739034,
+      "grad_norm": 0.028606578707695007,
+      "learning_rate": 9.175632691540064e-05,
+      "loss": 0.4696,
+      "step": 808
+    },
+    {
+      "epoch": 0.3512809379070777,
+      "grad_norm": 0.027804942801594734,
+      "learning_rate": 9.083299739645007e-05,
+      "loss": 0.504,
+      "step": 809
+    },
+    {
+      "epoch": 0.3517151541467651,
+      "grad_norm": 0.02756122685968876,
+      "learning_rate": 8.991387270152201e-05,
+      "loss": 0.5129,
+      "step": 810
+    },
+    {
+      "epoch": 0.35214937038645244,
+      "grad_norm": 0.03051147237420082,
+      "learning_rate": 8.899896227604509e-05,
+      "loss": 0.5054,
+      "step": 811
+    },
+    {
+      "epoch": 0.3525835866261398,
+      "grad_norm": 0.032840099185705185,
+      "learning_rate": 8.808827552213916e-05,
+      "loss": 0.4926,
+      "step": 812
+    },
+    {
+      "epoch": 0.3530178028658272,
+      "grad_norm": 0.026586757972836494,
+      "learning_rate": 8.718182179851997e-05,
+      "loss": 0.4992,
+      "step": 813
+    },
+    {
+      "epoch": 0.35345201910551455,
+      "grad_norm": 0.030486930161714554,
+      "learning_rate": 8.627961042040184e-05,
+      "loss": 0.5221,
+      "step": 814
+    },
+    {
+      "epoch": 0.3538862353452019,
+      "grad_norm": 0.03037872351706028,
+      "learning_rate": 8.538165065940262e-05,
+      "loss": 0.4801,
+      "step": 815
+    },
+    {
+      "epoch": 0.3543204515848893,
+      "grad_norm": 0.027922067791223526,
+      "learning_rate": 8.448795174344803e-05,
+      "loss": 0.5011,
+      "step": 816
+    },
+    {
+      "epoch": 0.35475466782457665,
+      "grad_norm": 0.030396727845072746,
+      "learning_rate": 8.359852285667752e-05,
+      "loss": 0.4937,
+      "step": 817
+    },
+    {
+      "epoch": 0.355188884064264,
+      "grad_norm": 0.028359906747937202,
+      "learning_rate": 8.271337313934868e-05,
+      "loss": 0.4905,
+      "step": 818
+    },
+    {
+      "epoch": 0.3556231003039514,
+      "grad_norm": 0.041622817516326904,
+      "learning_rate": 8.183251168774475e-05,
+      "loss": 0.4829,
+      "step": 819
+    },
+    {
+      "epoch": 0.35605731654363876,
+      "grad_norm": 0.030735066160559654,
+      "learning_rate": 8.09559475540797e-05,
+      "loss": 0.5038,
+      "step": 820
+    },
+    {
+      "epoch": 0.35649153278332607,
+      "grad_norm": 0.030377600342035294,
+      "learning_rate": 8.008368974640634e-05,
+      "loss": 0.523,
+      "step": 821
+    },
+    {
+      "epoch": 0.35692574902301344,
+      "grad_norm": 0.027815300971269608,
+      "learning_rate": 7.921574722852342e-05,
+      "loss": 0.4758,
+      "step": 822
+    },
+    {
+      "epoch": 0.3573599652627008,
+      "grad_norm": 0.026980141177773476,
+      "learning_rate": 7.835212891988292e-05,
+      "loss": 0.5151,
+      "step": 823
+    },
+    {
+      "epoch": 0.3577941815023882,
+      "grad_norm": 0.028305258601903915,
+      "learning_rate": 7.749284369549953e-05,
+      "loss": 0.4968,
+      "step": 824
+    },
+    {
+      "epoch": 0.35822839774207554,
+      "grad_norm": 0.028862129896879196,
+      "learning_rate": 7.663790038585794e-05,
+      "loss": 0.5093,
+      "step": 825
+    },
+    {
+      "epoch": 0.3586626139817629,
+      "grad_norm": 0.031168105080723763,
+      "learning_rate": 7.578730777682385e-05,
+      "loss": 0.507,
+      "step": 826
+    },
+    {
+      "epoch": 0.3590968302214503,
+      "grad_norm": 0.02467159368097782,
+      "learning_rate": 7.494107460955207e-05,
+      "loss": 0.5031,
+      "step": 827
+    },
+    {
+      "epoch": 0.35953104646113765,
+      "grad_norm": 0.028088459745049477,
+      "learning_rate": 7.409920958039795e-05,
+      "loss": 0.5031,
+      "step": 828
+    },
+    {
+      "epoch": 0.359965262700825,
+      "grad_norm": 0.027428364381194115,
+      "learning_rate": 7.326172134082704e-05,
+      "loss": 0.5262,
+      "step": 829
+    },
+    {
+      "epoch": 0.3603994789405124,
+      "grad_norm": 0.033019062131643295,
+      "learning_rate": 7.242861849732695e-05,
+      "loss": 0.4722,
+      "step": 830
+    },
+    {
+      "epoch": 0.36083369518019975,
+      "grad_norm": 0.028655540198087692,
+      "learning_rate": 7.159990961131818e-05,
+      "loss": 0.4899,
+      "step": 831
+    },
+    {
+      "epoch": 0.3612679114198871,
+      "grad_norm": 0.02457267791032791,
+      "learning_rate": 7.077560319906695e-05,
+      "loss": 0.4989,
+      "step": 832
+    },
+    {
+      "epoch": 0.3617021276595745,
+      "grad_norm": 0.027305081486701965,
+      "learning_rate": 6.995570773159692e-05,
+      "loss": 0.4885,
+      "step": 833
+    },
+    {
+      "epoch": 0.36213634389926186,
+      "grad_norm": 0.026189139112830162,
+      "learning_rate": 6.914023163460248e-05,
+      "loss": 0.4727,
+      "step": 834
+    },
+    {
+      "epoch": 0.36257056013894917,
+      "grad_norm": 0.028178861364722252,
+      "learning_rate": 6.832918328836246e-05,
+      "loss": 0.5538,
+      "step": 835
+    },
+    {
+      "epoch": 0.36300477637863654,
+      "grad_norm": 0.02753199078142643,
+      "learning_rate": 6.752257102765324e-05,
+      "loss": 0.4991,
+      "step": 836
+    },
+    {
+      "epoch": 0.3634389926183239,
+      "grad_norm": 0.02777719311416149,
+      "learning_rate": 6.6720403141664e-05,
+      "loss": 0.4607,
+      "step": 837
+    },
+    {
+      "epoch": 0.3638732088580113,
+      "grad_norm": 0.026312079280614853,
+      "learning_rate": 6.592268787391076e-05,
+      "loss": 0.5031,
+      "step": 838
+    },
+    {
+      "epoch": 0.36430742509769865,
+      "grad_norm": 0.028296776115894318,
+      "learning_rate": 6.512943342215233e-05,
+      "loss": 0.4825,
+      "step": 839
+    },
+    {
+      "epoch": 0.364741641337386,
+      "grad_norm": 0.026355035603046417,
+      "learning_rate": 6.43406479383053e-05,
+      "loss": 0.4871,
+      "step": 840
+    },
+    {
+      "epoch": 0.3651758575770734,
+      "grad_norm": 0.025023045018315315,
+      "learning_rate": 6.355633952836115e-05,
+      "loss": 0.4797,
+      "step": 841
+    },
+    {
+      "epoch": 0.36561007381676075,
+      "grad_norm": 0.027957437559962273,
+      "learning_rate": 6.277651625230219e-05,
+      "loss": 0.5078,
+      "step": 842
+    },
+    {
+      "epoch": 0.3660442900564481,
+      "grad_norm": 0.028170911595225334,
+      "learning_rate": 6.200118612401917e-05,
+      "loss": 0.4781,
+      "step": 843
+    },
+    {
+      "epoch": 0.3664785062961355,
+      "grad_norm": 0.032493725419044495,
+      "learning_rate": 6.123035711122859e-05,
+      "loss": 0.5078,
+      "step": 844
+    },
+    {
+      "epoch": 0.36691272253582286,
+      "grad_norm": 0.029985696077346802,
+      "learning_rate": 6.046403713539139e-05,
+      "loss": 0.478,
+      "step": 845
+    },
+    {
+      "epoch": 0.3673469387755102,
+      "grad_norm": 0.03212130442261696,
+      "learning_rate": 5.9702234071631e-05,
+      "loss": 0.5146,
+      "step": 846
+    },
+    {
+      "epoch": 0.3677811550151976,
+      "grad_norm": 0.024975216016173363,
+      "learning_rate": 5.89449557486525e-05,
+      "loss": 0.4564,
+      "step": 847
+    },
+    {
+      "epoch": 0.36821537125488496,
+      "grad_norm": 0.026881275698542595,
+      "learning_rate": 5.8192209948662365e-05,
+      "loss": 0.518,
+      "step": 848
+    },
+    {
+      "epoch": 0.3686495874945723,
+      "grad_norm": 0.04243510961532593,
+      "learning_rate": 5.7444004407288256e-05,
+      "loss": 0.5238,
+      "step": 849
+    },
+    {
+      "epoch": 0.36908380373425964,
+      "grad_norm": 0.02634822018444538,
+      "learning_rate": 5.6700346813499945e-05,
+      "loss": 0.5126,
+      "step": 850
+    },
+    {
+      "epoch": 0.369518019973947,
+      "grad_norm": 0.029794232919812202,
+      "learning_rate": 5.5961244809529746e-05,
+      "loss": 0.5017,
+      "step": 851
+    },
+    {
+      "epoch": 0.3699522362136344,
+      "grad_norm": 0.028596797958016396,
+      "learning_rate": 5.5226705990794155e-05,
+      "loss": 0.508,
+      "step": 852
+    },
+    {
+      "epoch": 0.37038645245332175,
+      "grad_norm": 0.04081004112958908,
+      "learning_rate": 5.449673790581611e-05,
+      "loss": 0.4946,
+      "step": 853
+    },
+    {
+      "epoch": 0.3708206686930091,
+      "grad_norm": 0.02494877576828003,
+      "learning_rate": 5.377134805614714e-05,
+      "loss": 0.5008,
+      "step": 854
+    },
+    {
+      "epoch": 0.3712548849326965,
+      "grad_norm": 0.029261523857712746,
+      "learning_rate": 5.305054389629022e-05,
+      "loss": 0.5155,
+      "step": 855
+    },
+    {
+      "epoch": 0.37168910117238385,
+      "grad_norm": 0.026129230856895447,
+      "learning_rate": 5.2334332833623486e-05,
+      "loss": 0.4979,
+      "step": 856
+    },
+    {
+      "epoch": 0.3721233174120712,
+      "grad_norm": 0.03483575955033302,
+      "learning_rate": 5.162272222832348e-05,
+      "loss": 0.4956,
+      "step": 857
+    },
+    {
+      "epoch": 0.3725575336517586,
+      "grad_norm": 0.028554193675518036,
+      "learning_rate": 5.0915719393290485e-05,
+      "loss": 0.4816,
+      "step": 858
+    },
+    {
+      "epoch": 0.37299174989144596,
+      "grad_norm": 0.028302082791924477,
+      "learning_rate": 5.021333159407232e-05,
+      "loss": 0.4734,
+      "step": 859
+    },
+    {
+      "epoch": 0.3734259661311333,
+      "grad_norm": 0.03633783757686615,
+      "learning_rate": 4.9515566048790485e-05,
+      "loss": 0.5152,
+      "step": 860
+    },
+    {
+      "epoch": 0.3738601823708207,
+      "grad_norm": 0.025067532435059547,
+      "learning_rate": 4.882242992806546e-05,
+      "loss": 0.505,
+      "step": 861
+    },
+    {
+      "epoch": 0.374294398610508,
+      "grad_norm": 0.03426215052604675,
+      "learning_rate": 4.813393035494329e-05,
+      "loss": 0.4811,
+      "step": 862
+    },
+    {
+      "epoch": 0.3747286148501954,
+      "grad_norm": 0.024589484557509422,
+      "learning_rate": 4.745007440482252e-05,
+      "loss": 0.4881,
+      "step": 863
+    },
+    {
+      "epoch": 0.37516283108988274,
+      "grad_norm": 0.032448723912239075,
+      "learning_rate": 4.677086910538092e-05,
+      "loss": 0.5057,
+      "step": 864
+    },
+    {
+      "epoch": 0.3755970473295701,
+      "grad_norm": 0.029285568743944168,
+      "learning_rate": 4.6096321436503994e-05,
+      "loss": 0.5035,
+      "step": 865
+    },
+    {
+      "epoch": 0.3760312635692575,
+      "grad_norm": 0.02594076097011566,
+      "learning_rate": 4.5426438330212535e-05,
+      "loss": 0.4789,
+      "step": 866
+    },
+    {
+      "epoch": 0.37646547980894485,
+      "grad_norm": 0.028019646182656288,
+      "learning_rate": 4.476122667059207e-05,
+      "loss": 0.4841,
+      "step": 867
+    },
+    {
+      "epoch": 0.3768996960486322,
+      "grad_norm": 0.022888783365488052,
+      "learning_rate": 4.410069329372152e-05,
+      "loss": 0.4609,
+      "step": 868
+    },
+    {
+      "epoch": 0.3773339122883196,
+      "grad_norm": 0.025522053241729736,
+      "learning_rate": 4.3444844987603426e-05,
+      "loss": 0.5148,
+      "step": 869
+    },
+    {
+      "epoch": 0.37776812852800695,
+      "grad_norm": 0.027691571041941643,
+      "learning_rate": 4.279368849209381e-05,
+      "loss": 0.4689,
+      "step": 870
+    },
+    {
+      "epoch": 0.3782023447676943,
+      "grad_norm": 0.02531832829117775,
+      "learning_rate": 4.214723049883307e-05,
+      "loss": 0.5024,
+      "step": 871
+    },
+    {
+      "epoch": 0.3786365610073817,
+      "grad_norm": 0.025689370930194855,
+      "learning_rate": 4.150547765117746e-05,
+      "loss": 0.4839,
+      "step": 872
+    },
+    {
+      "epoch": 0.37907077724706906,
+      "grad_norm": 0.031849976629018784,
+      "learning_rate": 4.086843654413031e-05,
+      "loss": 0.4888,
+      "step": 873
+    },
+    {
+      "epoch": 0.3795049934867564,
+      "grad_norm": 0.024920670315623283,
+      "learning_rate": 4.0236113724274713e-05,
+      "loss": 0.5088,
+      "step": 874
+    },
+    {
+      "epoch": 0.3799392097264438,
+      "grad_norm": 0.03232395276427269,
+      "learning_rate": 3.960851568970586e-05,
+      "loss": 0.5039,
+      "step": 875
+    },
+    {
+      "epoch": 0.3803734259661311,
+      "grad_norm": 0.025085026398301125,
+      "learning_rate": 3.898564888996475e-05,
+      "loss": 0.4753,
+      "step": 876
+    },
+    {
+      "epoch": 0.3808076422058185,
+      "grad_norm": 0.03306965157389641,
+      "learning_rate": 3.83675197259713e-05,
+      "loss": 0.493,
+      "step": 877
+    },
+    {
+      "epoch": 0.38124185844550584,
+      "grad_norm": 0.02512296289205551,
+      "learning_rate": 3.7754134549959293e-05,
+      "loss": 0.5048,
+      "step": 878
+    },
+    {
+      "epoch": 0.3816760746851932,
+      "grad_norm": 0.02608814276754856,
+      "learning_rate": 3.714549966541014e-05,
+      "loss": 0.4893,
+      "step": 879
+    },
+    {
+      "epoch": 0.3821102909248806,
+      "grad_norm": 0.02658063732087612,
+      "learning_rate": 3.654162132698918e-05,
+      "loss": 0.5069,
+      "step": 880
+    },
+    {
+      "epoch": 0.38254450716456795,
+      "grad_norm": 0.027712546288967133,
+      "learning_rate": 3.594250574048058e-05,
+      "loss": 0.4778,
+      "step": 881
+    },
+    {
+      "epoch": 0.3829787234042553,
+      "grad_norm": 0.02457358129322529,
+      "learning_rate": 3.534815906272404e-05,
+      "loss": 0.4946,
+      "step": 882
+    },
+    {
+      "epoch": 0.3834129396439427,
+      "grad_norm": 0.032855454832315445,
+      "learning_rate": 3.475858740155108e-05,
+      "loss": 0.4796,
+      "step": 883
+    },
+    {
+      "epoch": 0.38384715588363005,
+      "grad_norm": 0.02618669532239437,
+      "learning_rate": 3.417379681572297e-05,
+      "loss": 0.4944,
+      "step": 884
+    },
+    {
+      "epoch": 0.3842813721233174,
+      "grad_norm": 0.02916988916695118,
+      "learning_rate": 3.3593793314867616e-05,
+      "loss": 0.4865,
+      "step": 885
+    },
+    {
+      "epoch": 0.3847155883630048,
+      "grad_norm": 0.02979261800646782,
+      "learning_rate": 3.3018582859418446e-05,
+      "loss": 0.5272,
+      "step": 886
+    },
+    {
+      "epoch": 0.38514980460269216,
+      "grad_norm": 0.024905268102884293,
+      "learning_rate": 3.2448171360552835e-05,
+      "loss": 0.4878,
+      "step": 887
+    },
+    {
+      "epoch": 0.3855840208423795,
+      "grad_norm": 0.025121575221419334,
+      "learning_rate": 3.18825646801314e-05,
+      "loss": 0.4887,
+      "step": 888
+    },
+    {
+      "epoch": 0.3860182370820669,
+      "grad_norm": 0.02593301795423031,
+      "learning_rate": 3.132176863063807e-05,
+      "loss": 0.5255,
+      "step": 889
+    },
+    {
+      "epoch": 0.3864524533217542,
+      "grad_norm": 0.02763954922556877,
+      "learning_rate": 3.076578897511978e-05,
+      "loss": 0.4672,
+      "step": 890
+    },
+    {
+      "epoch": 0.3868866695614416,
+      "grad_norm": 0.026684967800974846,
+      "learning_rate": 3.0214631427127882e-05,
+      "loss": 0.5277,
+      "step": 891
+    },
+    {
+      "epoch": 0.38732088580112894,
+      "grad_norm": 0.030234767124056816,
+      "learning_rate": 2.9668301650658758e-05,
+      "loss": 0.502,
+      "step": 892
+    },
+    {
+      "epoch": 0.3877551020408163,
+      "grad_norm": 0.023924171924591064,
+      "learning_rate": 2.9126805260096256e-05,
+      "loss": 0.4813,
+      "step": 893
+    },
+    {
+      "epoch": 0.3881893182805037,
+      "grad_norm": 0.03323635831475258,
+      "learning_rate": 2.859014782015351e-05,
+      "loss": 0.5422,
+      "step": 894
+    },
+    {
+      "epoch": 0.38862353452019105,
+      "grad_norm": 0.02363520860671997,
+      "learning_rate": 2.8058334845816213e-05,
+      "loss": 0.5091,
+      "step": 895
+    },
+    {
+      "epoch": 0.3890577507598784,
+      "grad_norm": 0.024630311876535416,
+      "learning_rate": 2.7531371802285433e-05,
+      "loss": 0.4642,
+      "step": 896
+    },
+    {
+      "epoch": 0.3894919669995658,
+      "grad_norm": 0.025726784020662308,
+      "learning_rate": 2.7009264104921606e-05,
+      "loss": 0.511,
+      "step": 897
+    },
+    {
+      "epoch": 0.38992618323925315,
+      "grad_norm": 0.02972118742763996,
+      "learning_rate": 2.6492017119189414e-05,
+      "loss": 0.4825,
+      "step": 898
+    },
+    {
+      "epoch": 0.3903603994789405,
+      "grad_norm": 0.024715179577469826,
+      "learning_rate": 2.597963616060167e-05,
+      "loss": 0.4879,
+      "step": 899
+    },
+    {
+      "epoch": 0.3907946157186279,
+      "grad_norm": 0.02915225736796856,
+      "learning_rate": 2.547212649466568e-05,
+      "loss": 0.4903,
+      "step": 900
+    },
+    {
+      "epoch": 0.39122883195831526,
+      "grad_norm": 0.02453678846359253,
+      "learning_rate": 2.4969493336828353e-05,
+      "loss": 0.5024,
+      "step": 901
+    },
+    {
+      "epoch": 0.39166304819800263,
+      "grad_norm": 0.025468653067946434,
+      "learning_rate": 2.4471741852423235e-05,
+      "loss": 0.4964,
+      "step": 902
+    },
+    {
+      "epoch": 0.39209726443769,
+      "grad_norm": 0.024076983332633972,
+      "learning_rate": 2.3978877156616784e-05,
+      "loss": 0.4872,
+      "step": 903
+    },
+    {
+      "epoch": 0.3925314806773773,
+      "grad_norm": 0.02444782853126526,
+      "learning_rate": 2.349090431435641e-05,
+      "loss": 0.482,
+      "step": 904
+    },
+    {
+      "epoch": 0.3929656969170647,
+      "grad_norm": 0.025424372404813766,
+      "learning_rate": 2.3007828340318116e-05,
+      "loss": 0.4907,
+      "step": 905
+    },
+    {
+      "epoch": 0.39339991315675205,
+      "grad_norm": 0.023807033896446228,
+      "learning_rate": 2.252965419885483e-05,
+      "loss": 0.5007,
+      "step": 906
+    },
+    {
+      "epoch": 0.3938341293964394,
+      "grad_norm": 0.022182505577802658,
+      "learning_rate": 2.2056386803945726e-05,
+      "loss": 0.4712,
+      "step": 907
+    },
+    {
+      "epoch": 0.3942683456361268,
+      "grad_norm": 0.02266152761876583,
+      "learning_rate": 2.1588031019145637e-05,
+      "loss": 0.479,
+      "step": 908
+    },
+    {
+      "epoch": 0.39470256187581415,
+      "grad_norm": 0.024663135409355164,
+      "learning_rate": 2.1124591657534777e-05,
+      "loss": 0.514,
+      "step": 909
+    },
+    {
+      "epoch": 0.3951367781155015,
+      "grad_norm": 0.023098768666386604,
+      "learning_rate": 2.0666073481669712e-05,
+      "loss": 0.4857,
+      "step": 910
+    },
+    {
+      "epoch": 0.3955709943551889,
+      "grad_norm": 0.025155888870358467,
+      "learning_rate": 2.0212481203534084e-05,
+      "loss": 0.5054,
+      "step": 911
+    },
+    {
+      "epoch": 0.39600521059487626,
+      "grad_norm": 0.030762221664190292,
+      "learning_rate": 1.9763819484490355e-05,
+      "loss": 0.4967,
+      "step": 912
+    },
+    {
+      "epoch": 0.3964394268345636,
+      "grad_norm": 0.029380332678556442,
+      "learning_rate": 1.932009293523196e-05,
+      "loss": 0.5412,
+      "step": 913
+    },
+    {
+      "epoch": 0.396873643074251,
+      "grad_norm": 0.023896733298897743,
+      "learning_rate": 1.888130611573563e-05,
+      "loss": 0.5002,
+      "step": 914
+    },
+    {
+      "epoch": 0.39730785931393836,
+      "grad_norm": 0.024988802149891853,
+      "learning_rate": 1.844746353521487e-05,
+      "loss": 0.4888,
+      "step": 915
+    },
+    {
+      "epoch": 0.39774207555362573,
+      "grad_norm": 0.02422316186130047,
+      "learning_rate": 1.801856965207338e-05,
+      "loss": 0.4818,
+      "step": 916
+    },
+    {
+      "epoch": 0.3981762917933131,
+      "grad_norm": 0.031409528106451035,
+      "learning_rate": 1.7594628873859486e-05,
+      "loss": 0.4923,
+      "step": 917
+    },
+    {
+      "epoch": 0.3986105080330004,
+      "grad_norm": 0.023978758603334427,
+      "learning_rate": 1.7175645557220565e-05,
+      "loss": 0.4798,
+      "step": 918
+    },
+    {
+      "epoch": 0.3990447242726878,
+      "grad_norm": 0.02715397998690605,
+      "learning_rate": 1.6761624007858522e-05,
+      "loss": 0.4474,
+      "step": 919
+    },
+    {
+      "epoch": 0.39947894051237515,
+      "grad_norm": 0.02678792178630829,
+      "learning_rate": 1.6352568480485276e-05,
+      "loss": 0.514,
+      "step": 920
+    },
+    {
+      "epoch": 0.3999131567520625,
+      "grad_norm": 0.02234644442796707,
+      "learning_rate": 1.594848317877934e-05,
+      "loss": 0.4869,
+      "step": 921
+    },
+    {
+      "epoch": 0.4003473729917499,
+      "grad_norm": 0.029807301238179207,
+      "learning_rate": 1.5549372255342365e-05,
+      "loss": 0.5242,
+      "step": 922
+    },
+    {
+      "epoch": 0.40078158923143725,
+      "grad_norm": 0.03037095069885254,
+      "learning_rate": 1.5155239811656562e-05,
+      "loss": 0.5035,
+      "step": 923
+    },
+    {
+      "epoch": 0.4012158054711246,
+      "grad_norm": 0.022767795249819756,
+      "learning_rate": 1.4766089898042678e-05,
+      "loss": 0.4903,
+      "step": 924
+    },
+    {
+      "epoch": 0.401650021710812,
+      "grad_norm": 0.026207391172647476,
+      "learning_rate": 1.4381926513618138e-05,
+      "loss": 0.4676,
+      "step": 925
+    },
+    {
+      "epoch": 0.40208423795049936,
+      "grad_norm": 0.025653038173913956,
+      "learning_rate": 1.400275360625608e-05,
+      "loss": 0.4804,
+      "step": 926
+    },
+    {
+      "epoch": 0.4025184541901867,
+      "grad_norm": 0.024481656029820442,
+      "learning_rate": 1.362857507254478e-05,
+      "loss": 0.4871,
+      "step": 927
+    },
+    {
+      "epoch": 0.4029526704298741,
+      "grad_norm": 0.02475779317319393,
+      "learning_rate": 1.3259394757747678e-05,
+      "loss": 0.5199,
+      "step": 928
+    },
+    {
+      "epoch": 0.40338688666956146,
+      "grad_norm": 0.02371501363813877,
+      "learning_rate": 1.289521645576358e-05,
+      "loss": 0.528,
+      "step": 929
+    },
+    {
+      "epoch": 0.40382110290924883,
+      "grad_norm": 0.023029111325740814,
+      "learning_rate": 1.2536043909088191e-05,
+      "loss": 0.4801,
+      "step": 930
+    },
+    {
+      "epoch": 0.40425531914893614,
+      "grad_norm": 0.025304608047008514,
+      "learning_rate": 1.2181880808775025e-05,
+      "loss": 0.4793,
+      "step": 931
+    },
+    {
+      "epoch": 0.4046895353886235,
+      "grad_norm": 0.02734360471367836,
+      "learning_rate": 1.183273079439795e-05,
+      "loss": 0.5022,
+      "step": 932
+    },
+    {
+      "epoch": 0.4051237516283109,
+      "grad_norm": 0.024530354887247086,
+      "learning_rate": 1.1488597454013538e-05,
+      "loss": 0.5032,
+      "step": 933
+    },
+    {
+      "epoch": 0.40555796786799825,
+      "grad_norm": 0.02369566634297371,
+      "learning_rate": 1.1149484324124327e-05,
+      "loss": 0.4633,
+      "step": 934
+    },
+    {
+      "epoch": 0.4059921841076856,
+      "grad_norm": 0.024855339899659157,
+      "learning_rate": 1.0815394889642338e-05,
+      "loss": 0.5175,
+      "step": 935
+    },
+    {
+      "epoch": 0.406426400347373,
+      "grad_norm": 0.023386195302009583,
+      "learning_rate": 1.0486332583853564e-05,
+      "loss": 0.5014,
+      "step": 936
+    },
+    {
+      "epoch": 0.40686061658706035,
+      "grad_norm": 0.029485292732715607,
+      "learning_rate": 1.0162300788382261e-05,
+      "loss": 0.4942,
+      "step": 937
+    },
+    {
+      "epoch": 0.4072948328267477,
+      "grad_norm": 0.023054329678416252,
+      "learning_rate": 9.843302833156376e-06,
+      "loss": 0.4683,
+      "step": 938
+    },
+    {
+      "epoch": 0.4077290490664351,
+      "grad_norm": 0.031031260266900063,
+      "learning_rate": 9.529341996373674e-06,
+      "loss": 0.5173,
+      "step": 939
+    },
+    {
+      "epoch": 0.40816326530612246,
+      "grad_norm": 0.023358464241027832,
+      "learning_rate": 9.22042150446728e-06,
+      "loss": 0.4722,
+      "step": 940
+    },
+    {
+      "epoch": 0.4085974815458098,
+      "grad_norm": 0.025074800476431847,
+      "learning_rate": 8.916544532073411e-06,
+      "loss": 0.4784,
+      "step": 941
+    },
+    {
+      "epoch": 0.4090316977854972,
+      "grad_norm": 0.02519945055246353,
+      "learning_rate": 8.617714201998083e-06,
+      "loss": 0.5022,
+      "step": 942
+    },
+    {
+      "epoch": 0.40946591402518456,
+      "grad_norm": 0.027328645810484886,
+      "learning_rate": 8.323933585185184e-06,
+      "loss": 0.4896,
+      "step": 943
+    },
+    {
+      "epoch": 0.40990013026487193,
+      "grad_norm": 0.02542622946202755,
+      "learning_rate": 8.035205700685167e-06,
+      "loss": 0.4676,
+      "step": 944
+    },
+    {
+      "epoch": 0.41033434650455924,
+      "grad_norm": 0.027195794507861137,
+      "learning_rate": 7.751533515623799e-06,
+      "loss": 0.4903,
+      "step": 945
+    },
+    {
+      "epoch": 0.4107685627442466,
+      "grad_norm": 0.02717754803597927,
+      "learning_rate": 7.47291994517163e-06,
+      "loss": 0.4812,
+      "step": 946
+    },
+    {
+      "epoch": 0.411202778983934,
+      "grad_norm": 0.02495565079152584,
+      "learning_rate": 7.199367852514238e-06,
+      "loss": 0.5014,
+      "step": 947
+    },
+    {
+      "epoch": 0.41163699522362135,
+      "grad_norm": 0.024513866752386093,
+      "learning_rate": 6.93088004882253e-06,
+      "loss": 0.5071,
+      "step": 948
+    },
+    {
+      "epoch": 0.4120712114633087,
+      "grad_norm": 0.03526155650615692,
+      "learning_rate": 6.667459293224154e-06,
+      "loss": 0.531,
+      "step": 949
+    },
+    {
+      "epoch": 0.4125054277029961,
+      "grad_norm": 0.023175369948148727,
+      "learning_rate": 6.4091082927749125e-06,
+      "loss": 0.4811,
+      "step": 950
+    },
+    {
+      "epoch": 0.41293964394268345,
+      "grad_norm": 0.026199523359537125,
+      "learning_rate": 6.15582970243117e-06,
+      "loss": 0.5087,
+      "step": 951
+    },
+    {
+      "epoch": 0.4133738601823708,
+      "grad_norm": 0.024454379454255104,
+      "learning_rate": 5.907626125022158e-06,
+      "loss": 0.479,
+      "step": 952
+    },
+    {
+      "epoch": 0.4138080764220582,
+      "grad_norm": 0.02492498978972435,
+      "learning_rate": 5.664500111223769e-06,
+      "loss": 0.4977,
+      "step": 953
+    },
+    {
+      "epoch": 0.41424229266174556,
+      "grad_norm": 0.023200612515211105,
+      "learning_rate": 5.426454159531913e-06,
+      "loss": 0.4803,
+      "step": 954
+    },
+    {
+      "epoch": 0.4146765089014329,
+      "grad_norm": 0.024389084428548813,
+      "learning_rate": 5.193490716237037e-06,
+      "loss": 0.4733,
+      "step": 955
+    },
+    {
+      "epoch": 0.4151107251411203,
+      "grad_norm": 0.027189958840608597,
+      "learning_rate": 4.9656121753990924e-06,
+      "loss": 0.4994,
+      "step": 956
+    },
+    {
+      "epoch": 0.41554494138080766,
+      "grad_norm": 0.02239627204835415,
+      "learning_rate": 4.7428208788224955e-06,
+      "loss": 0.4978,
+      "step": 957
+    },
+    {
+      "epoch": 0.41597915762049503,
+      "grad_norm": 0.021820807829499245,
+      "learning_rate": 4.52511911603265e-06,
+      "loss": 0.4871,
+      "step": 958
+    },
+    {
+      "epoch": 0.41641337386018235,
+      "grad_norm": 0.02775476686656475,
+      "learning_rate": 4.312509124251906e-06,
+      "loss": 0.4918,
+      "step": 959
+    },
+    {
+      "epoch": 0.4168475900998697,
+      "grad_norm": 0.023040220141410828,
+      "learning_rate": 4.104993088376974e-06,
+      "loss": 0.4497,
+      "step": 960
+    },
+    {
+      "epoch": 0.4172818063395571,
+      "grad_norm": 0.025357872247695923,
+      "learning_rate": 3.9025731409561e-06,
+      "loss": 0.4945,
+      "step": 961
+    },
+    {
+      "epoch": 0.41771602257924445,
+      "grad_norm": 0.02308405004441738,
+      "learning_rate": 3.7052513621674833e-06,
+      "loss": 0.4945,
+      "step": 962
+    },
+    {
+      "epoch": 0.4181502388189318,
+      "grad_norm": 0.02784230373799801,
+      "learning_rate": 3.513029779797783e-06,
+      "loss": 0.4543,
+      "step": 963
+    },
+    {
+      "epoch": 0.4185844550586192,
+      "grad_norm": 0.026198869571089745,
+      "learning_rate": 3.325910369220975e-06,
+      "loss": 0.4975,
+      "step": 964
+    },
+    {
+      "epoch": 0.41901867129830656,
+      "grad_norm": 0.023575875908136368,
+      "learning_rate": 3.143895053378698e-06,
+      "loss": 0.4987,
+      "step": 965
+    },
+    {
+      "epoch": 0.4194528875379939,
+      "grad_norm": 0.02861775830388069,
+      "learning_rate": 2.966985702759828e-06,
+      "loss": 0.4857,
+      "step": 966
+    },
+    {
+      "epoch": 0.4198871037776813,
+      "grad_norm": 0.033127009868621826,
+      "learning_rate": 2.7951841353817676e-06,
+      "loss": 0.465,
+      "step": 967
+    },
+    {
+      "epoch": 0.42032132001736866,
+      "grad_norm": 0.02451484464108944,
+      "learning_rate": 2.6284921167712973e-06,
+      "loss": 0.4863,
+      "step": 968
+    },
+    {
+      "epoch": 0.42075553625705603,
+      "grad_norm": 0.02445485256612301,
+      "learning_rate": 2.466911359946977e-06,
+      "loss": 0.4896,
+      "step": 969
+    },
+    {
+      "epoch": 0.4211897524967434,
+      "grad_norm": 0.02243492379784584,
+      "learning_rate": 2.3104435254008848e-06,
+      "loss": 0.4839,
+      "step": 970
+    },
+    {
+      "epoch": 0.42162396873643077,
+      "grad_norm": 0.026183119043707848,
+      "learning_rate": 2.1590902210822937e-06,
+      "loss": 0.4594,
+      "step": 971
+    },
+    {
+      "epoch": 0.42205818497611813,
+      "grad_norm": 0.025734947994351387,
+      "learning_rate": 2.012853002380466e-06,
+      "loss": 0.4684,
+      "step": 972
+    },
+    {
+      "epoch": 0.42249240121580545,
+      "grad_norm": 0.027873944491147995,
+      "learning_rate": 1.8717333721091634e-06,
+      "loss": 0.5056,
+      "step": 973
+    },
+    {
+      "epoch": 0.4229266174554928,
+      "grad_norm": 0.02484060265123844,
+      "learning_rate": 1.735732780490884e-06,
+      "loss": 0.5153,
+      "step": 974
+    },
+    {
+      "epoch": 0.4233608336951802,
+      "grad_norm": 0.022599438205361366,
+      "learning_rate": 1.6048526251421502e-06,
+      "loss": 0.4936,
+      "step": 975
+    },
+    {
+      "epoch": 0.42379504993486755,
+      "grad_norm": 0.020465485751628876,
+      "learning_rate": 1.4790942510590765e-06,
+      "loss": 0.4848,
+      "step": 976
+    },
+    {
+      "epoch": 0.4242292661745549,
+      "grad_norm": 0.027462849393486977,
+      "learning_rate": 1.3584589506034362e-06,
+      "loss": 0.4666,
+      "step": 977
+    },
+    {
+      "epoch": 0.4246634824142423,
+      "grad_norm": 0.02553846500813961,
+      "learning_rate": 1.2429479634897267e-06,
+      "loss": 0.4899,
+      "step": 978
+    },
+    {
+      "epoch": 0.42509769865392966,
+      "grad_norm": 0.024199191480875015,
+      "learning_rate": 1.132562476771959e-06,
+      "loss": 0.4792,
+      "step": 979
+    },
+    {
+      "epoch": 0.425531914893617,
+      "grad_norm": 0.023328760638833046,
+      "learning_rate": 1.0273036248318324e-06,
+      "loss": 0.4973,
+      "step": 980
+    },
+    {
+      "epoch": 0.4259661311333044,
+      "grad_norm": 0.02446364425122738,
+      "learning_rate": 9.271724893669121e-07,
+      "loss": 0.4654,
+      "step": 981
+    },
+    {
+      "epoch": 0.42640034737299176,
+      "grad_norm": 0.02676616981625557,
+      "learning_rate": 8.321700993795811e-07,
+      "loss": 0.4881,
+      "step": 982
+    },
+    {
+      "epoch": 0.42683456361267913,
+      "grad_norm": 0.02478119358420372,
+      "learning_rate": 7.422974311662722e-07,
+      "loss": 0.5055,
+      "step": 983
+    },
+    {
+      "epoch": 0.4272687798523665,
+      "grad_norm": 0.03248545154929161,
+      "learning_rate": 6.575554083078084e-07,
+      "loss": 0.4949,
+      "step": 984
+    },
+    {
+      "epoch": 0.42770299609205387,
+      "grad_norm": 0.024504756554961205,
+      "learning_rate": 5.779449016595773e-07,
+      "loss": 0.4986,
+      "step": 985
+    },
+    {
+      "epoch": 0.42813721233174123,
+      "grad_norm": 0.02531745657324791,
+      "learning_rate": 5.034667293427053e-07,
+      "loss": 0.4851,
+      "step": 986
+    },
+    {
+      "epoch": 0.42857142857142855,
+      "grad_norm": 0.024369893595576286,
+      "learning_rate": 4.3412165673556436e-07,
+      "loss": 0.5044,
+      "step": 987
+    },
+    {
+      "epoch": 0.4290056448111159,
+      "grad_norm": 0.02750435471534729,
+      "learning_rate": 3.6991039646616656e-07,
+      "loss": 0.468,
+      "step": 988
+    },
+    {
+      "epoch": 0.4294398610508033,
+      "grad_norm": 0.02262257970869541,
+      "learning_rate": 3.1083360840455934e-07,
+      "loss": 0.498,
+      "step": 989
+    },
+    {
+      "epoch": 0.42987407729049065,
+      "grad_norm": 0.024247439578175545,
+      "learning_rate": 2.568918996560532e-07,
+      "loss": 0.4885,
+      "step": 990
+    },
+    {
+      "epoch": 0.430308293530178,
+      "grad_norm": 0.024190355092287064,
+      "learning_rate": 2.0808582455528192e-07,
+      "loss": 0.541,
+      "step": 991
+    },
+    {
+      "epoch": 0.4307425097698654,
+      "grad_norm": 0.023794230073690414,
+      "learning_rate": 1.6441588466009628e-07,
+      "loss": 0.4616,
+      "step": 992
+    },
+    {
+      "epoch": 0.43117672600955276,
+      "grad_norm": 0.023178860545158386,
+      "learning_rate": 1.2588252874673466e-07,
+      "loss": 0.5077,
+      "step": 993
+    },
+    {
+      "epoch": 0.4316109422492401,
+      "grad_norm": 0.0218358151614666,
+      "learning_rate": 9.248615280499361e-08,
+      "loss": 0.517,
+      "step": 994
+    },
+    {
+      "epoch": 0.4320451584889275,
+      "grad_norm": 0.021379638463258743,
+      "learning_rate": 6.422710003439747e-08,
+      "loss": 0.4589,
+      "step": 995
+    },
+    {
+      "epoch": 0.43247937472861486,
+      "grad_norm": 0.024375783279538155,
+      "learning_rate": 4.1105660840368156e-08,
+      "loss": 0.4963,
+      "step": 996
+    },
+    {
+      "epoch": 0.43291359096830223,
+      "grad_norm": 0.02347472496330738,
+      "learning_rate": 2.3122072831505136e-08,
+      "loss": 0.4855,
+      "step": 997
+    },
+    {
+      "epoch": 0.4333478072079896,
+      "grad_norm": 0.02579214982688427,
+      "learning_rate": 1.0276520816976388e-08,
+      "loss": 0.4786,
+      "step": 998
+    },
+    {
+      "epoch": 0.43378202344767697,
+      "grad_norm": 0.023950695991516113,
+      "learning_rate": 2.5691368046865116e-09,
+      "loss": 0.5068,
+      "step": 999
+    },
+    {
+      "epoch": 0.4342162396873643,
+      "grad_norm": 0.022179216146469116,
+      "learning_rate": 0.0,
+      "loss": 0.4841,
+      "step": 1000
+    },
+    {
+      "epoch": 0.4342162396873643,
+      "step": 1000,
+      "total_flos": 1.0100654119636173e+20,
+      "train_loss": 0.710578792899847,
+      "train_runtime": 20977.6715,
+      "train_samples_per_second": 0.763,
+      "train_steps_per_second": 0.048
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.0100654119636173e+20,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}