diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8780 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9998000399920016,
+  "eval_steps": 500,
+  "global_step": 1250,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0007998400319936012,
+      "grad_norm": 4.033497738180381,
+      "learning_rate": 0.0,
+      "loss": -0.0349,
+      "step": 1
+    },
+    {
+      "epoch": 0.0015996800639872025,
+      "grad_norm": 3.51489589336218,
+      "learning_rate": 3.8110282485354675e-07,
+      "loss": -0.0542,
+      "step": 2
+    },
+    {
+      "epoch": 0.0023995200959808036,
+      "grad_norm": 3.185205419075282,
+      "learning_rate": 6.040336863117743e-07,
+      "loss": 0.0277,
+      "step": 3
+    },
+    {
+      "epoch": 0.003199360127974405,
+      "grad_norm": 2.6795442503962392,
+      "learning_rate": 7.622056497070935e-07,
+      "loss": 0.0346,
+      "step": 4
+    },
+    {
+      "epoch": 0.003999200159968006,
+      "grad_norm": 4.306548503833335,
+      "learning_rate": 8.84893356068388e-07,
+      "loss": -0.0123,
+      "step": 5
+    },
+    {
+      "epoch": 0.004799040191961607,
+      "grad_norm": 3.232420399829724,
+      "learning_rate": 9.85136511165321e-07,
+      "loss": -0.0277,
+      "step": 6
+    },
+    {
+      "epoch": 0.005598880223955209,
+      "grad_norm": 4.556647322380419,
+      "learning_rate": 1.0698908911626617e-06,
+      "loss": -0.0824,
+      "step": 7
+    },
+    {
+      "epoch": 0.00639872025594881,
+      "grad_norm": 4.75835426891596,
+      "learning_rate": 1.1433084745606403e-06,
+      "loss": -0.0487,
+      "step": 8
+    },
+    {
+      "epoch": 0.007198560287942412,
+      "grad_norm": 3.671579551985109,
+      "learning_rate": 1.2080673726235485e-06,
+      "loss": -0.101,
+      "step": 9
+    },
+    {
+      "epoch": 0.007998400319936013,
+      "grad_norm": 5.478829821928343,
+      "learning_rate": 1.2659961809219347e-06,
+      "loss": 0.0194,
+      "step": 10
+    },
+    {
+      "epoch": 0.008798240351929614,
+      "grad_norm": 4.697076805586441,
+      "learning_rate": 1.318399162250352e-06,
+      "loss": -0.0713,
+      "step": 11
+    },
+    {
+      "epoch": 0.009598080383923215,
+      "grad_norm": 4.668341585855885,
+      "learning_rate": 1.366239336018868e-06,
+      "loss": -0.0365,
+      "step": 12
+    },
+    {
+      "epoch": 0.010397920415916816,
+      "grad_norm": 4.0841275175924725,
+      "learning_rate": 1.4102480297838326e-06,
+      "loss": -0.0814,
+      "step": 13
+    },
+    {
+      "epoch": 0.011197760447910418,
+      "grad_norm": 3.5676724545793244,
+      "learning_rate": 1.4509937160162082e-06,
+      "loss": -0.0815,
+      "step": 14
+    },
+    {
+      "epoch": 0.01199760047990402,
+      "grad_norm": 3.467762437060136,
+      "learning_rate": 1.4889270423801623e-06,
+      "loss": -0.0654,
+      "step": 15
+    },
+    {
+      "epoch": 0.01279744051189762,
+      "grad_norm": 2.8225474286212457,
+      "learning_rate": 1.524411299414187e-06,
+      "loss": -0.062,
+      "step": 16
+    },
+    {
+      "epoch": 0.013597280543891222,
+      "grad_norm": 5.079798214426631,
+      "learning_rate": 1.5577436352844088e-06,
+      "loss": -0.1394,
+      "step": 17
+    },
+    {
+      "epoch": 0.014397120575884824,
+      "grad_norm": 4.339673002027499,
+      "learning_rate": 1.5891701974770953e-06,
+      "loss": -0.042,
+      "step": 18
+    },
+    {
+      "epoch": 0.015196960607878424,
+      "grad_norm": 4.245593535465187,
+      "learning_rate": 1.6188971751464532e-06,
+      "loss": 0.0082,
+      "step": 19
+    },
+    {
+      "epoch": 0.015996800639872025,
+      "grad_norm": 2.975134158916166,
+      "learning_rate": 1.6470990057754815e-06,
+      "loss": -0.0294,
+      "step": 20
+    },
+    {
+      "epoch": 0.016796640671865627,
+      "grad_norm": 3.331212004539707,
+      "learning_rate": 1.673924577474436e-06,
+      "loss": -0.0544,
+      "step": 21
+    },
+    {
+      "epoch": 0.01759648070385923,
+      "grad_norm": 4.545480625771349,
+      "learning_rate": 1.6995019871038986e-06,
+      "loss": -0.0731,
+      "step": 22
+    },
+    {
+      "epoch": 0.01839632073585283,
+      "grad_norm": 3.7461986348064595,
+      "learning_rate": 1.7239422398533632e-06,
+      "loss": -0.1278,
+      "step": 23
+    },
+    {
+      "epoch": 0.01919616076784643,
+      "grad_norm": 2.8335266795291543,
+      "learning_rate": 1.7473421608724147e-06,
+      "loss": -0.083,
+      "step": 24
+    },
+    {
+      "epoch": 0.01999600079984003,
+      "grad_norm": 3.68991966387542,
+      "learning_rate": 1.769786712136776e-06,
+      "loss": -0.0695,
+      "step": 25
+    },
+    {
+      "epoch": 0.020795840831833633,
+      "grad_norm": 4.133751803923456,
+      "learning_rate": 1.7913508546373795e-06,
+      "loss": -0.0445,
+      "step": 26
+    },
+    {
+      "epoch": 0.021595680863827234,
+      "grad_norm": 5.000412923928387,
+      "learning_rate": 1.812101058935323e-06,
+      "loss": 0.1318,
+      "step": 27
+    },
+    {
+      "epoch": 0.022395520895820836,
+      "grad_norm": 3.5992183048553557,
+      "learning_rate": 1.832096540869755e-06,
+      "loss": -0.1249,
+      "step": 28
+    },
+    {
+      "epoch": 0.023195360927814438,
+      "grad_norm": 3.427019545745099,
+      "learning_rate": 1.8513902803279621e-06,
+      "loss": -0.1743,
+      "step": 29
+    },
+    {
+      "epoch": 0.02399520095980804,
+      "grad_norm": 4.725372315931298,
+      "learning_rate": 1.8700298672337092e-06,
+      "loss": 0.0091,
+      "step": 30
+    },
+    {
+      "epoch": 0.024795040991801638,
+      "grad_norm": 5.254587258373332,
+      "learning_rate": 1.888058208767457e-06,
+      "loss": -0.0653,
+      "step": 31
+    },
+    {
+      "epoch": 0.02559488102379524,
+      "grad_norm": 8.849971672055872,
+      "learning_rate": 1.905514124267734e-06,
+      "loss": -0.0049,
+      "step": 32
+    },
+    {
+      "epoch": 0.026394721055788842,
+      "grad_norm": 3.4635929572339874,
+      "learning_rate": 1.922432848562126e-06,
+      "loss": -0.089,
+      "step": 33
+    },
+    {
+      "epoch": 0.027194561087782444,
+      "grad_norm": 5.9169701870000715,
+      "learning_rate": 1.9388464601379558e-06,
+      "loss": -0.1119,
+      "step": 34
+    },
+    {
+      "epoch": 0.027994401119776045,
+      "grad_norm": 5.188979707658736,
+      "learning_rate": 1.9547842472310495e-06,
+      "loss": -0.1121,
+      "step": 35
+    },
+    {
+      "epoch": 0.028794241151769647,
+      "grad_norm": 4.896152090964042,
+      "learning_rate": 1.970273022330642e-06,
+      "loss": -0.0525,
+      "step": 36
+    },
+    {
+      "epoch": 0.02959408118376325,
+      "grad_norm": 5.994962166737604,
+      "learning_rate": 1.9853373935840096e-06,
+      "loss": -0.1089,
+      "step": 37
+    },
+    {
+      "epoch": 0.030393921215756847,
+      "grad_norm": 5.576677749054259,
+      "learning_rate": 2e-06,
+      "loss": 0.0304,
+      "step": 38
+    },
+    {
+      "epoch": 0.03119376124775045,
+      "grad_norm": 4.313822225686198,
+      "learning_rate": 2e-06,
+      "loss": -0.1328,
+      "step": 39
+    },
+    {
+      "epoch": 0.03199360127974405,
+      "grad_norm": 3.81381102986674,
+      "learning_rate": 1.998349834983498e-06,
+      "loss": -0.0792,
+      "step": 40
+    },
+    {
+      "epoch": 0.03279344131173765,
+      "grad_norm": 4.839236808694631,
+      "learning_rate": 1.996699669966997e-06,
+      "loss": -0.056,
+      "step": 41
+    },
+    {
+      "epoch": 0.033593281343731254,
+      "grad_norm": 3.77305458921913,
+      "learning_rate": 1.995049504950495e-06,
+      "loss": -0.0339,
+      "step": 42
+    },
+    {
+      "epoch": 0.03439312137572485,
+      "grad_norm": 5.867440087459917,
+      "learning_rate": 1.9933993399339932e-06,
+      "loss": -0.0176,
+      "step": 43
+    },
+    {
+      "epoch": 0.03519296140771846,
+      "grad_norm": 6.4685114572374465,
+      "learning_rate": 1.991749174917492e-06,
+      "loss": 0.0402,
+      "step": 44
+    },
+    {
+      "epoch": 0.035992801439712056,
+      "grad_norm": 11.48483869174446,
+      "learning_rate": 1.99009900990099e-06,
+      "loss": -0.0702,
+      "step": 45
+    },
+    {
+      "epoch": 0.03679264147170566,
+      "grad_norm": 4.2414264593452735,
+      "learning_rate": 1.9884488448844884e-06,
+      "loss": -0.1241,
+      "step": 46
+    },
+    {
+      "epoch": 0.03759248150369926,
+      "grad_norm": 4.949735715342123,
+      "learning_rate": 1.9867986798679866e-06,
+      "loss": -0.0996,
+      "step": 47
+    },
+    {
+      "epoch": 0.03839232153569286,
+      "grad_norm": 5.06186964090094,
+      "learning_rate": 1.9851485148514852e-06,
+      "loss": -0.1133,
+      "step": 48
+    },
+    {
+      "epoch": 0.039192161567686463,
+      "grad_norm": 3.9056723205659183,
+      "learning_rate": 1.9834983498349835e-06,
+      "loss": -0.1631,
+      "step": 49
+    },
+    {
+      "epoch": 0.03999200159968006,
+      "grad_norm": 5.51223763254555,
+      "learning_rate": 1.9818481848184817e-06,
+      "loss": 0.0191,
+      "step": 50
+    },
+    {
+      "epoch": 0.04079184163167367,
+      "grad_norm": 3.7107878002289,
+      "learning_rate": 1.98019801980198e-06,
+      "loss": -0.0271,
+      "step": 51
+    },
+    {
+      "epoch": 0.041591681663667265,
+      "grad_norm": 5.618046340756691,
+      "learning_rate": 1.9785478547854786e-06,
+      "loss": -0.0227,
+      "step": 52
+    },
+    {
+      "epoch": 0.04239152169566087,
+      "grad_norm": 3.7602961019841468,
+      "learning_rate": 1.976897689768977e-06,
+      "loss": -0.126,
+      "step": 53
+    },
+    {
+      "epoch": 0.04319136172765447,
+      "grad_norm": 4.322826902384424,
+      "learning_rate": 1.975247524752475e-06,
+      "loss": -0.1196,
+      "step": 54
+    },
+    {
+      "epoch": 0.04399120175964807,
+      "grad_norm": 3.6276654934086565,
+      "learning_rate": 1.9735973597359733e-06,
+      "loss": -0.1098,
+      "step": 55
+    },
+    {
+      "epoch": 0.04479104179164167,
+      "grad_norm": 3.729759012982189,
+      "learning_rate": 1.971947194719472e-06,
+      "loss": -0.2248,
+      "step": 56
+    },
+    {
+      "epoch": 0.04559088182363527,
+      "grad_norm": 5.552584092947439,
+      "learning_rate": 1.97029702970297e-06,
+      "loss": -0.0821,
+      "step": 57
+    },
+    {
+      "epoch": 0.046390721855628876,
+      "grad_norm": 6.016002296406734,
+      "learning_rate": 1.9686468646864684e-06,
+      "loss": -0.1302,
+      "step": 58
+    },
+    {
+      "epoch": 0.047190561887622474,
+      "grad_norm": 6.7453871612622995,
+      "learning_rate": 1.966996699669967e-06,
+      "loss": -0.0652,
+      "step": 59
+    },
+    {
+      "epoch": 0.04799040191961608,
+      "grad_norm": 4.874246979289447,
+      "learning_rate": 1.9653465346534653e-06,
+      "loss": -0.0409,
+      "step": 60
+    },
+    {
+      "epoch": 0.04879024195160968,
+      "grad_norm": 3.894046979082966,
+      "learning_rate": 1.9636963696369635e-06,
+      "loss": 0.021,
+      "step": 61
+    },
+    {
+      "epoch": 0.049590081983603276,
+      "grad_norm": 3.829546481539617,
+      "learning_rate": 1.962046204620462e-06,
+      "loss": -0.246,
+      "step": 62
+    },
+    {
+      "epoch": 0.05038992201559688,
+      "grad_norm": 5.021080021581999,
+      "learning_rate": 1.9603960396039604e-06,
+      "loss": -0.0029,
+      "step": 63
+    },
+    {
+      "epoch": 0.05118976204759048,
+      "grad_norm": 4.084832649304883,
+      "learning_rate": 1.9587458745874586e-06,
+      "loss": -0.1911,
+      "step": 64
+    },
+    {
+      "epoch": 0.051989602079584085,
+      "grad_norm": 3.4567077830219595,
+      "learning_rate": 1.9570957095709572e-06,
+      "loss": 0.0388,
+      "step": 65
+    },
+    {
+      "epoch": 0.052789442111577684,
+      "grad_norm": 4.523151395245523,
+      "learning_rate": 1.9554455445544555e-06,
+      "loss": -0.0422,
+      "step": 66
+    },
+    {
+      "epoch": 0.05358928214357129,
+      "grad_norm": 4.574942149645985,
+      "learning_rate": 1.9537953795379537e-06,
+      "loss": 0.0052,
+      "step": 67
+    },
+    {
+      "epoch": 0.05438912217556489,
+      "grad_norm": 5.884212332415378,
+      "learning_rate": 1.952145214521452e-06,
+      "loss": -0.0622,
+      "step": 68
+    },
+    {
+      "epoch": 0.055188962207558485,
+      "grad_norm": 3.172106804857128,
+      "learning_rate": 1.95049504950495e-06,
+      "loss": -0.0936,
+      "step": 69
+    },
+    {
+      "epoch": 0.05598880223955209,
+      "grad_norm": 4.882587885458746,
+      "learning_rate": 1.948844884488449e-06,
+      "loss": -0.1034,
+      "step": 70
+    },
+    {
+      "epoch": 0.05678864227154569,
+      "grad_norm": 3.290096020906111,
+      "learning_rate": 1.947194719471947e-06,
+      "loss": -0.126,
+      "step": 71
+    },
+    {
+      "epoch": 0.057588482303539294,
+      "grad_norm": 4.175847937084437,
+      "learning_rate": 1.9455445544554453e-06,
+      "loss": -0.0295,
+      "step": 72
+    },
+    {
+      "epoch": 0.05838832233553289,
+      "grad_norm": 4.774862772782205,
+      "learning_rate": 1.943894389438944e-06,
+      "loss": -0.0737,
+      "step": 73
+    },
+    {
+      "epoch": 0.0591881623675265,
+      "grad_norm": 4.866413673374395,
+      "learning_rate": 1.942244224422442e-06,
+      "loss": 0.0154,
+      "step": 74
+    },
+    {
+      "epoch": 0.059988002399520096,
+      "grad_norm": 3.244110640100742,
+      "learning_rate": 1.9405940594059404e-06,
+      "loss": -0.0876,
+      "step": 75
+    },
+    {
+      "epoch": 0.060787842431513694,
+      "grad_norm": 4.94642971249312,
+      "learning_rate": 1.938943894389439e-06,
+      "loss": -0.0634,
+      "step": 76
+    },
+    {
+      "epoch": 0.0615876824635073,
+      "grad_norm": 3.1477348357592705,
+      "learning_rate": 1.9372937293729373e-06,
+      "loss": -0.0934,
+      "step": 77
+    },
+    {
+      "epoch": 0.0623875224955009,
+      "grad_norm": 4.180278871715678,
+      "learning_rate": 1.9356435643564355e-06,
+      "loss": -0.0158,
+      "step": 78
+    },
+    {
+      "epoch": 0.0631873625274945,
+      "grad_norm": 4.73751736841566,
+      "learning_rate": 1.933993399339934e-06,
+      "loss": 0.0419,
+      "step": 79
+    },
+    {
+      "epoch": 0.0639872025594881,
+      "grad_norm": 3.7318625198178577,
+      "learning_rate": 1.9323432343234324e-06,
+      "loss": -0.0981,
+      "step": 80
+    },
+    {
+      "epoch": 0.0647870425914817,
+      "grad_norm": 3.5344903982736016,
+      "learning_rate": 1.9306930693069306e-06,
+      "loss": 0.028,
+      "step": 81
+    },
+    {
+      "epoch": 0.0655868826234753,
+      "grad_norm": 3.6655427915390653,
+      "learning_rate": 1.9290429042904292e-06,
+      "loss": -0.0343,
+      "step": 82
+    },
+    {
+      "epoch": 0.06638672265546891,
+      "grad_norm": 3.8402537750787817,
+      "learning_rate": 1.9273927392739275e-06,
+      "loss": -0.0479,
+      "step": 83
+    },
+    {
+      "epoch": 0.06718656268746251,
+      "grad_norm": 6.194776167870759,
+      "learning_rate": 1.9257425742574257e-06,
+      "loss": 0.0046,
+      "step": 84
+    },
+    {
+      "epoch": 0.06798640271945611,
+      "grad_norm": 5.366879383554931,
+      "learning_rate": 1.924092409240924e-06,
+      "loss": -0.1113,
+      "step": 85
+    },
+    {
+      "epoch": 0.0687862427514497,
+      "grad_norm": 3.9997732575047547,
+      "learning_rate": 1.922442244224422e-06,
+      "loss": 0.0446,
+      "step": 86
+    },
+    {
+      "epoch": 0.06958608278344332,
+      "grad_norm": 9.73377697425672,
+      "learning_rate": 1.920792079207921e-06,
+      "loss": -0.0569,
+      "step": 87
+    },
+    {
+      "epoch": 0.07038592281543692,
+      "grad_norm": 5.0689420802437875,
+      "learning_rate": 1.919141914191419e-06,
+      "loss": -0.0352,
+      "step": 88
+    },
+    {
+      "epoch": 0.07118576284743051,
+      "grad_norm": 8.98640262446026,
+      "learning_rate": 1.9174917491749173e-06,
+      "loss": 0.041,
+      "step": 89
+    },
+    {
+      "epoch": 0.07198560287942411,
+      "grad_norm": 5.63457538673664,
+      "learning_rate": 1.9158415841584155e-06,
+      "loss": -0.0193,
+      "step": 90
+    },
+    {
+      "epoch": 0.07278544291141771,
+      "grad_norm": 4.290130537843607,
+      "learning_rate": 1.914191419141914e-06,
+      "loss": -0.0742,
+      "step": 91
+    },
+    {
+      "epoch": 0.07358528294341132,
+      "grad_norm": 4.0945792486692465,
+      "learning_rate": 1.9125412541254124e-06,
+      "loss": 0.029,
+      "step": 92
+    },
+    {
+      "epoch": 0.07438512297540492,
+      "grad_norm": 4.96670528541929,
+      "learning_rate": 1.9108910891089106e-06,
+      "loss": -0.1134,
+      "step": 93
+    },
+    {
+      "epoch": 0.07518496300739852,
+      "grad_norm": 5.027466862141088,
+      "learning_rate": 1.9092409240924093e-06,
+      "loss": -0.0508,
+      "step": 94
+    },
+    {
+      "epoch": 0.07598480303939212,
+      "grad_norm": 9.573722686429775,
+      "learning_rate": 1.9075907590759075e-06,
+      "loss": -0.1592,
+      "step": 95
+    },
+    {
+      "epoch": 0.07678464307138572,
+      "grad_norm": 5.764961349212166,
+      "learning_rate": 1.9059405940594057e-06,
+      "loss": -0.0122,
+      "step": 96
+    },
+    {
+      "epoch": 0.07758448310337933,
+      "grad_norm": 3.3502664260820247,
+      "learning_rate": 1.9042904290429044e-06,
+      "loss": -0.2179,
+      "step": 97
+    },
+    {
+      "epoch": 0.07838432313537293,
+      "grad_norm": 3.967491851586746,
+      "learning_rate": 1.9026402640264026e-06,
+      "loss": -0.0234,
+      "step": 98
+    },
+    {
+      "epoch": 0.07918416316736653,
+      "grad_norm": 4.552540817957586,
+      "learning_rate": 1.9009900990099008e-06,
+      "loss": -0.1591,
+      "step": 99
+    },
+    {
+      "epoch": 0.07998400319936012,
+      "grad_norm": 10.823710544953496,
+      "learning_rate": 1.8993399339933993e-06,
+      "loss": 0.0374,
+      "step": 100
+    },
+    {
+      "epoch": 0.08078384323135374,
+      "grad_norm": 4.806468007236691,
+      "learning_rate": 1.8976897689768975e-06,
+      "loss": -0.0565,
+      "step": 101
+    },
+    {
+      "epoch": 0.08158368326334733,
+      "grad_norm": 3.634545747480329,
+      "learning_rate": 1.896039603960396e-06,
+      "loss": -0.0762,
+      "step": 102
+    },
+    {
+      "epoch": 0.08238352329534093,
+      "grad_norm": 3.2837047295849597,
+      "learning_rate": 1.8943894389438944e-06,
+      "loss": -0.0491,
+      "step": 103
+    },
+    {
+      "epoch": 0.08318336332733453,
+      "grad_norm": 5.176653817957751,
+      "learning_rate": 1.8927392739273926e-06,
+      "loss": -0.1934,
+      "step": 104
+    },
+    {
+      "epoch": 0.08398320335932813,
+      "grad_norm": 6.107024303996945,
+      "learning_rate": 1.8910891089108908e-06,
+      "loss": -0.1129,
+      "step": 105
+    },
+    {
+      "epoch": 0.08478304339132174,
+      "grad_norm": 4.489176343037952,
+      "learning_rate": 1.8894389438943895e-06,
+      "loss": 0.0883,
+      "step": 106
+    },
+    {
+      "epoch": 0.08558288342331534,
+      "grad_norm": 4.318302618280909,
+      "learning_rate": 1.8877887788778877e-06,
+      "loss": -0.1609,
+      "step": 107
+    },
+    {
+      "epoch": 0.08638272345530894,
+      "grad_norm": 4.634209202008312,
+      "learning_rate": 1.886138613861386e-06,
+      "loss": -0.0457,
+      "step": 108
+    },
+    {
+      "epoch": 0.08718256348730254,
+      "grad_norm": 3.630881832190838,
+      "learning_rate": 1.8844884488448844e-06,
+      "loss": -0.1382,
+      "step": 109
+    },
+    {
+      "epoch": 0.08798240351929613,
+      "grad_norm": 3.886065281514502,
+      "learning_rate": 1.8828382838283828e-06,
+      "loss": -0.0535,
+      "step": 110
+    },
+    {
+      "epoch": 0.08878224355128975,
+      "grad_norm": 3.647392695144741,
+      "learning_rate": 1.881188118811881e-06,
+      "loss": -0.0809,
+      "step": 111
+    },
+    {
+      "epoch": 0.08958208358328335,
+      "grad_norm": 3.9753438884802463,
+      "learning_rate": 1.8795379537953795e-06,
+      "loss": -0.0791,
+      "step": 112
+    },
+    {
+      "epoch": 0.09038192361527694,
+      "grad_norm": 4.473252382488765,
+      "learning_rate": 1.8778877887788777e-06,
+      "loss": -0.0723,
+      "step": 113
+    },
+    {
+      "epoch": 0.09118176364727054,
+      "grad_norm": 4.928253206993449,
+      "learning_rate": 1.876237623762376e-06,
+      "loss": 0.0125,
+      "step": 114
+    },
+    {
+      "epoch": 0.09198160367926415,
+      "grad_norm": 4.557945800338486,
+      "learning_rate": 1.8745874587458746e-06,
+      "loss": -0.0889,
+      "step": 115
+    },
+    {
+      "epoch": 0.09278144371125775,
+      "grad_norm": 5.830924417742841,
+      "learning_rate": 1.8729372937293728e-06,
+      "loss": 0.0504,
+      "step": 116
+    },
+    {
+      "epoch": 0.09358128374325135,
+      "grad_norm": 6.4722171650631655,
+      "learning_rate": 1.8712871287128713e-06,
+      "loss": -0.008,
+      "step": 117
+    },
+    {
+      "epoch": 0.09438112377524495,
+      "grad_norm": 3.1676413558574428,
+      "learning_rate": 1.8696369636963695e-06,
+      "loss": -0.0483,
+      "step": 118
+    },
+    {
+      "epoch": 0.09518096380723855,
+      "grad_norm": 5.1310816710504845,
+      "learning_rate": 1.867986798679868e-06,
+      "loss": -0.0513,
+      "step": 119
+    },
+    {
+      "epoch": 0.09598080383923216,
+      "grad_norm": 3.620114816562482,
+      "learning_rate": 1.8663366336633664e-06,
+      "loss": -0.1453,
+      "step": 120
+    },
+    {
+      "epoch": 0.09678064387122576,
+      "grad_norm": 6.0676794834569865,
+      "learning_rate": 1.8646864686468646e-06,
+      "loss": -0.0194,
+      "step": 121
+    },
+    {
+      "epoch": 0.09758048390321936,
+      "grad_norm": 6.414733331241253,
+      "learning_rate": 1.8630363036303628e-06,
+      "loss": -0.0488,
+      "step": 122
+    },
+    {
+      "epoch": 0.09838032393521295,
+      "grad_norm": 4.6846628376767905,
+      "learning_rate": 1.8613861386138615e-06,
+      "loss": -0.0195,
+      "step": 123
+    },
+    {
+      "epoch": 0.09918016396720655,
+      "grad_norm": 3.235246476419315,
+      "learning_rate": 1.8597359735973597e-06,
+      "loss": -0.0942,
+      "step": 124
+    },
+    {
+      "epoch": 0.09998000399920016,
+      "grad_norm": 5.3470459527801495,
+      "learning_rate": 1.858085808580858e-06,
+      "loss": 0.0276,
+      "step": 125
+    },
+    {
+      "epoch": 0.10077984403119376,
+      "grad_norm": 3.9287996597379995,
+      "learning_rate": 1.8564356435643564e-06,
+      "loss": 0.0306,
+      "step": 126
+    },
+    {
+      "epoch": 0.10157968406318736,
+      "grad_norm": 4.995425229535215,
+      "learning_rate": 1.8547854785478546e-06,
+      "loss": 0.0087,
+      "step": 127
+    },
+    {
+      "epoch": 0.10237952409518096,
+      "grad_norm": 4.573732944820577,
+      "learning_rate": 1.853135313531353e-06,
+      "loss": -0.1424,
+      "step": 128
+    },
+    {
+      "epoch": 0.10317936412717456,
+      "grad_norm": 4.55020470630308,
+      "learning_rate": 1.8514851485148515e-06,
+      "loss": -0.0559,
+      "step": 129
+    },
+    {
+      "epoch": 0.10397920415916817,
+      "grad_norm": 4.41698840906731,
+      "learning_rate": 1.8498349834983497e-06,
+      "loss": -0.0188,
+      "step": 130
+    },
+    {
+      "epoch": 0.10477904419116177,
+      "grad_norm": 5.223224115420677,
+      "learning_rate": 1.848184818481848e-06,
+      "loss": -0.1098,
+      "step": 131
+    },
+    {
+      "epoch": 0.10557888422315537,
+      "grad_norm": 3.8011698979898005,
+      "learning_rate": 1.8465346534653466e-06,
+      "loss": -0.0328,
+      "step": 132
+    },
+    {
+      "epoch": 0.10637872425514897,
+      "grad_norm": 3.1746565925932835,
+      "learning_rate": 1.8448844884488448e-06,
+      "loss": -0.0525,
+      "step": 133
+    },
+    {
+      "epoch": 0.10717856428714258,
+      "grad_norm": 3.9995360105342903,
+      "learning_rate": 1.843234323432343e-06,
+      "loss": -0.0665,
+      "step": 134
+    },
+    {
+      "epoch": 0.10797840431913618,
+      "grad_norm": 4.722040708955319,
+      "learning_rate": 1.8415841584158415e-06,
+      "loss": -0.1398,
+      "step": 135
+    },
+    {
+      "epoch": 0.10877824435112977,
+      "grad_norm": 4.002530013096379,
+      "learning_rate": 1.83993399339934e-06,
+      "loss": -0.0609,
+      "step": 136
+    },
+    {
+      "epoch": 0.10957808438312337,
+      "grad_norm": 4.23131853552439,
+      "learning_rate": 1.8382838283828382e-06,
+      "loss": -0.0109,
+      "step": 137
+    },
+    {
+      "epoch": 0.11037792441511697,
+      "grad_norm": 5.324154803758963,
+      "learning_rate": 1.8366336633663366e-06,
+      "loss": -0.0251,
+      "step": 138
+    },
+    {
+      "epoch": 0.11117776444711058,
+      "grad_norm": 4.297402311394241,
+      "learning_rate": 1.8349834983498348e-06,
+      "loss": -0.0589,
+      "step": 139
+    },
+    {
+      "epoch": 0.11197760447910418,
+      "grad_norm": 4.454960816079389,
+      "learning_rate": 1.833333333333333e-06,
+      "loss": -0.1049,
+      "step": 140
+    },
+    {
+      "epoch": 0.11277744451109778,
+      "grad_norm": 4.709849875744532,
+      "learning_rate": 1.8316831683168317e-06,
+      "loss": -0.0279,
+      "step": 141
+    },
+    {
+      "epoch": 0.11357728454309138,
+      "grad_norm": 3.9184959414442724,
+      "learning_rate": 1.83003300330033e-06,
+      "loss": -0.1333,
+      "step": 142
+    },
+    {
+      "epoch": 0.11437712457508498,
+      "grad_norm": 4.362164005140024,
+      "learning_rate": 1.8283828382838282e-06,
+      "loss": -0.0821,
+      "step": 143
+    },
+    {
+      "epoch": 0.11517696460707859,
+      "grad_norm": 3.814336740776002,
+      "learning_rate": 1.8267326732673266e-06,
+      "loss": -0.0764,
+      "step": 144
+    },
+    {
+      "epoch": 0.11597680463907219,
+      "grad_norm": 4.1087265373281925,
+      "learning_rate": 1.825082508250825e-06,
+      "loss": 0.0595,
+      "step": 145
+    },
+    {
+      "epoch": 0.11677664467106579,
+      "grad_norm": 5.05463448309474,
+      "learning_rate": 1.8234323432343233e-06,
+      "loss": -0.0749,
+      "step": 146
+    },
+    {
+      "epoch": 0.11757648470305938,
+      "grad_norm": 7.009438010420224,
+      "learning_rate": 1.8217821782178217e-06,
+      "loss": -0.1623,
+      "step": 147
+    },
+    {
+      "epoch": 0.118376324735053,
+      "grad_norm": 5.86862518535322,
+      "learning_rate": 1.82013201320132e-06,
+      "loss": -0.1914,
+      "step": 148
+    },
+    {
+      "epoch": 0.1191761647670466,
+      "grad_norm": 8.568812361586986,
+      "learning_rate": 1.8184818481848184e-06,
+      "loss": -0.0496,
+      "step": 149
+    },
+    {
+      "epoch": 0.11997600479904019,
+      "grad_norm": 9.02774053582229,
+      "learning_rate": 1.8168316831683168e-06,
+      "loss": -0.022,
+      "step": 150
+    },
+    {
+      "epoch": 0.12077584483103379,
+      "grad_norm": 5.51491933312306,
+      "learning_rate": 1.815181518151815e-06,
+      "loss": -0.1322,
+      "step": 151
+    },
+    {
+      "epoch": 0.12157568486302739,
+      "grad_norm": 5.304215018308479,
+      "learning_rate": 1.8135313531353133e-06,
+      "loss": -0.0676,
+      "step": 152
+    },
+    {
+      "epoch": 0.122375524895021,
+      "grad_norm": 3.9922542678415565,
+      "learning_rate": 1.811881188118812e-06,
+      "loss": 0.0184,
+      "step": 153
+    },
+    {
+      "epoch": 0.1231753649270146,
+      "grad_norm": 4.724197779779715,
+      "learning_rate": 1.8102310231023102e-06,
+      "loss": -0.1204,
+      "step": 154
+    },
+    {
+      "epoch": 0.1239752049590082,
+      "grad_norm": 5.76455405608935,
+      "learning_rate": 1.8085808580858084e-06,
+      "loss": -0.1421,
+      "step": 155
+    },
+    {
+      "epoch": 0.1247750449910018,
+      "grad_norm": 10.161753692062435,
+      "learning_rate": 1.8069306930693068e-06,
+      "loss": 0.0592,
+      "step": 156
+    },
+    {
+      "epoch": 0.1255748850229954,
+      "grad_norm": 3.667923249601308,
+      "learning_rate": 1.805280528052805e-06,
+      "loss": -0.0988,
+      "step": 157
+    },
+    {
+      "epoch": 0.126374725054989,
+      "grad_norm": 4.515737987543515,
+      "learning_rate": 1.8036303630363035e-06,
+      "loss": 0.0522,
+      "step": 158
+    },
+    {
+      "epoch": 0.1271745650869826,
+      "grad_norm": 3.336996513422035,
+      "learning_rate": 1.801980198019802e-06,
+      "loss": -0.09,
+      "step": 159
+    },
+    {
+      "epoch": 0.1279744051189762,
+      "grad_norm": 3.969953099317271,
+      "learning_rate": 1.8003300330033002e-06,
+      "loss": -0.0435,
+      "step": 160
+    },
+    {
+      "epoch": 0.1287742451509698,
+      "grad_norm": 4.549949209747214,
+      "learning_rate": 1.7986798679867984e-06,
+      "loss": -0.0613,
+      "step": 161
+    },
+    {
+      "epoch": 0.1295740851829634,
+      "grad_norm": 3.759639050223288,
+      "learning_rate": 1.797029702970297e-06,
+      "loss": -0.0784,
+      "step": 162
+    },
+    {
+      "epoch": 0.130373925214957,
+      "grad_norm": 4.619365249559499,
+      "learning_rate": 1.7953795379537953e-06,
+      "loss": -0.0111,
+      "step": 163
+    },
+    {
+      "epoch": 0.1311737652469506,
+      "grad_norm": 4.114791027895229,
+      "learning_rate": 1.7937293729372935e-06,
+      "loss": -0.0327,
+      "step": 164
+    },
+    {
+      "epoch": 0.13197360527894422,
+      "grad_norm": 3.8956026767168836,
+      "learning_rate": 1.792079207920792e-06,
+      "loss": -0.1106,
+      "step": 165
+    },
+    {
+      "epoch": 0.13277344531093782,
+      "grad_norm": 4.818435179721396,
+      "learning_rate": 1.7904290429042904e-06,
+      "loss": -0.0034,
+      "step": 166
+    },
+    {
+      "epoch": 0.13357328534293142,
+      "grad_norm": 6.763152130893218,
+      "learning_rate": 1.7887788778877888e-06,
+      "loss": 0.0651,
+      "step": 167
+    },
+    {
+      "epoch": 0.13437312537492502,
+      "grad_norm": 4.097132792098502,
+      "learning_rate": 1.787128712871287e-06,
+      "loss": -0.0269,
+      "step": 168
+    },
+    {
+      "epoch": 0.13517296540691862,
+      "grad_norm": 4.706830462846675,
+      "learning_rate": 1.7854785478547853e-06,
+      "loss": 0.0558,
+      "step": 169
+    },
+    {
+      "epoch": 0.13597280543891221,
+      "grad_norm": 4.254134691338051,
+      "learning_rate": 1.783828382838284e-06,
+      "loss": -0.0046,
+      "step": 170
+    },
+    {
+      "epoch": 0.1367726454709058,
+      "grad_norm": 5.457939580250951,
+      "learning_rate": 1.7821782178217822e-06,
+      "loss": -0.0379,
+      "step": 171
+    },
+    {
+      "epoch": 0.1375724855028994,
+      "grad_norm": 3.2577166280201544,
+      "learning_rate": 1.7805280528052804e-06,
+      "loss": -0.0993,
+      "step": 172
+    },
+    {
+      "epoch": 0.138372325534893,
+      "grad_norm": 5.551040160162887,
+      "learning_rate": 1.7788778877887789e-06,
+      "loss": -0.0543,
+      "step": 173
+    },
+    {
+      "epoch": 0.13917216556688664,
+      "grad_norm": 3.69149537962834,
+      "learning_rate": 1.777227722772277e-06,
+      "loss": -0.0443,
+      "step": 174
+    },
+    {
+      "epoch": 0.13997200559888023,
+      "grad_norm": 4.4643620642536455,
+      "learning_rate": 1.7755775577557755e-06,
+      "loss": -0.0449,
+      "step": 175
+    },
+    {
+      "epoch": 0.14077184563087383,
+      "grad_norm": 3.5240643064279977,
+      "learning_rate": 1.773927392739274e-06,
+      "loss": -0.0928,
+      "step": 176
+    },
+    {
+      "epoch": 0.14157168566286743,
+      "grad_norm": 5.981016645991625,
+      "learning_rate": 1.7722772277227722e-06,
+      "loss": 0.0686,
+      "step": 177
+    },
+    {
+      "epoch": 0.14237152569486103,
+      "grad_norm": 4.336791468199441,
+      "learning_rate": 1.7706270627062704e-06,
+      "loss": -0.0617,
+      "step": 178
+    },
+    {
+      "epoch": 0.14317136572685463,
+      "grad_norm": 3.678032699373225,
+      "learning_rate": 1.768976897689769e-06,
+      "loss": -0.1058,
+      "step": 179
+    },
+    {
+      "epoch": 0.14397120575884823,
+      "grad_norm": 8.431078918847803,
+      "learning_rate": 1.7673267326732673e-06,
+      "loss": -0.034,
+      "step": 180
+    },
+    {
+      "epoch": 0.14477104579084182,
+      "grad_norm": 4.90238148952107,
+      "learning_rate": 1.7656765676567655e-06,
+      "loss": 0.0263,
+      "step": 181
+    },
+    {
+      "epoch": 0.14557088582283542,
+      "grad_norm": 4.1587161441545115,
+      "learning_rate": 1.764026402640264e-06,
+      "loss": -0.0128,
+      "step": 182
+    },
+    {
+      "epoch": 0.14637072585482905,
+      "grad_norm": 4.255313468888732,
+      "learning_rate": 1.7623762376237624e-06,
+      "loss": -0.0138,
+      "step": 183
+    },
+    {
+      "epoch": 0.14717056588682265,
+      "grad_norm": 6.24454443290786,
+      "learning_rate": 1.7607260726072606e-06,
+      "loss": -0.0941,
+      "step": 184
+    },
+    {
+      "epoch": 0.14797040591881624,
+      "grad_norm": 4.293655354485335,
+      "learning_rate": 1.759075907590759e-06,
+      "loss": -0.0621,
+      "step": 185
+    },
+    {
+      "epoch": 0.14877024595080984,
+      "grad_norm": 4.224321769134034,
+      "learning_rate": 1.7574257425742573e-06,
+      "loss": -0.0214,
+      "step": 186
+    },
+    {
+      "epoch": 0.14957008598280344,
+      "grad_norm": 3.7629471117165827,
+      "learning_rate": 1.7557755775577555e-06,
+      "loss": -0.0735,
+      "step": 187
+    },
+    {
+      "epoch": 0.15036992601479704,
+      "grad_norm": 4.511985288731285,
+      "learning_rate": 1.7541254125412542e-06,
+      "loss": -0.1376,
+      "step": 188
+    },
+    {
+      "epoch": 0.15116976604679064,
+      "grad_norm": 4.701449783409153,
+      "learning_rate": 1.7524752475247524e-06,
+      "loss": -0.1208,
+      "step": 189
+    },
+    {
+      "epoch": 0.15196960607878424,
+      "grad_norm": 7.169693891516351,
+      "learning_rate": 1.7508250825082506e-06,
+      "loss": 0.0103,
+      "step": 190
+    },
+    {
+      "epoch": 0.15276944611077783,
+      "grad_norm": 3.6302391864591126,
+      "learning_rate": 1.749174917491749e-06,
+      "loss": -0.036,
+      "step": 191
+    },
+    {
+      "epoch": 0.15356928614277143,
+      "grad_norm": 8.15707311459662,
+      "learning_rate": 1.7475247524752475e-06,
+      "loss": -0.0226,
+      "step": 192
+    },
+    {
+      "epoch": 0.15436912617476506,
+      "grad_norm": 4.001526302961896,
+      "learning_rate": 1.7458745874587458e-06,
+      "loss": -0.0587,
+      "step": 193
+    },
+    {
+      "epoch": 0.15516896620675866,
+      "grad_norm": 4.468601251007179,
+      "learning_rate": 1.7442244224422442e-06,
+      "loss": -0.1388,
+      "step": 194
+    },
+    {
+      "epoch": 0.15596880623875226,
+      "grad_norm": 4.107118632559092,
+      "learning_rate": 1.7425742574257424e-06,
+      "loss": -0.0961,
+      "step": 195
+    },
+    {
+      "epoch": 0.15676864627074585,
+      "grad_norm": 3.4961373949789665,
+      "learning_rate": 1.7409240924092409e-06,
+      "loss": -0.0567,
+      "step": 196
+    },
+    {
+      "epoch": 0.15756848630273945,
+      "grad_norm": 4.144654814148264,
+      "learning_rate": 1.7392739273927393e-06,
+      "loss": -0.1309,
+      "step": 197
+    },
+    {
+      "epoch": 0.15836832633473305,
+      "grad_norm": 3.6625054473315664,
+      "learning_rate": 1.7376237623762375e-06,
+      "loss": -0.0208,
+      "step": 198
+    },
+    {
+      "epoch": 0.15916816636672665,
+      "grad_norm": 4.664494531197071,
+      "learning_rate": 1.7359735973597358e-06,
+      "loss": 0.0178,
+      "step": 199
+    },
+    {
+      "epoch": 0.15996800639872025,
+      "grad_norm": 6.383022272218445,
+      "learning_rate": 1.7343234323432342e-06,
+      "loss": -0.0616,
+      "step": 200
+    },
+    {
+      "epoch": 0.16076784643071385,
+      "grad_norm": 5.505206158317875,
+      "learning_rate": 1.7326732673267326e-06,
+      "loss": -0.0452,
+      "step": 201
+    },
+    {
+      "epoch": 0.16156768646270747,
+      "grad_norm": 3.5601606217056765,
+      "learning_rate": 1.7310231023102309e-06,
+      "loss": 0.0225,
+      "step": 202
+    },
+    {
+      "epoch": 0.16236752649470107,
+      "grad_norm": 4.408138222273653,
+      "learning_rate": 1.7293729372937293e-06,
+      "loss": -0.1139,
+      "step": 203
+    },
+    {
+      "epoch": 0.16316736652669467,
+      "grad_norm": 3.2562884601218087,
+      "learning_rate": 1.7277227722772275e-06,
+      "loss": 0.0087,
+      "step": 204
+    },
+    {
+      "epoch": 0.16396720655868827,
+      "grad_norm": 4.350781355214131,
+      "learning_rate": 1.726072607260726e-06,
+      "loss": -0.0885,
+      "step": 205
+    },
+    {
+      "epoch": 0.16476704659068186,
+      "grad_norm": 3.3568949216522475,
+      "learning_rate": 1.7244224422442244e-06,
+      "loss": -0.0134,
+      "step": 206
+    },
+    {
+      "epoch": 0.16556688662267546,
+      "grad_norm": 6.798474914966856,
+      "learning_rate": 1.7227722772277227e-06,
+      "loss": -0.0945,
+      "step": 207
+    },
+    {
+      "epoch": 0.16636672665466906,
+      "grad_norm": 4.577665859282248,
+      "learning_rate": 1.7211221122112209e-06,
+      "loss": -0.1607,
+      "step": 208
+    },
+    {
+      "epoch": 0.16716656668666266,
+      "grad_norm": 6.460632243204499,
+      "learning_rate": 1.7194719471947195e-06,
+      "loss": -0.0235,
+      "step": 209
+    },
+    {
+      "epoch": 0.16796640671865626,
+      "grad_norm": 4.306267256349224,
+      "learning_rate": 1.7178217821782178e-06,
+      "loss": -0.0059,
+      "step": 210
+    },
+    {
+      "epoch": 0.16876624675064986,
+      "grad_norm": 3.0483507543879105,
+      "learning_rate": 1.716171617161716e-06,
+      "loss": -0.111,
+      "step": 211
+    },
+    {
+      "epoch": 0.16956608678264348,
+      "grad_norm": 5.737336519193611,
+      "learning_rate": 1.7145214521452144e-06,
+      "loss": 0.018,
+      "step": 212
+    },
+    {
+      "epoch": 0.17036592681463708,
+      "grad_norm": 3.7845990191052734,
+      "learning_rate": 1.7128712871287127e-06,
+      "loss": -0.0926,
+      "step": 213
+    },
+    {
+      "epoch": 0.17116576684663068,
+      "grad_norm": 3.669531800966666,
+      "learning_rate": 1.711221122112211e-06,
+      "loss": -0.0776,
+      "step": 214
+    },
+    {
+      "epoch": 0.17196560687862428,
+      "grad_norm": 4.005323920134325,
+      "learning_rate": 1.7095709570957095e-06,
+      "loss": -0.0399,
+      "step": 215
+    },
+    {
+      "epoch": 0.17276544691061788,
+      "grad_norm": 2.8598435648570186,
+      "learning_rate": 1.7079207920792078e-06,
+      "loss": -0.0548,
+      "step": 216
+    },
+    {
+      "epoch": 0.17356528694261147,
+      "grad_norm": 4.139220262158334,
+      "learning_rate": 1.7062706270627062e-06,
+      "loss": 0.0686,
+      "step": 217
+    },
+    {
+      "epoch": 0.17436512697460507,
+      "grad_norm": 4.988425208682803,
+      "learning_rate": 1.7046204620462046e-06,
+      "loss": -0.0028,
+      "step": 218
+    },
+    {
+      "epoch": 0.17516496700659867,
+      "grad_norm": 3.4806124639328164,
+      "learning_rate": 1.7029702970297029e-06,
+      "loss": -0.0914,
+      "step": 219
+    },
+    {
+      "epoch": 0.17596480703859227,
+      "grad_norm": 6.013581164060899,
+      "learning_rate": 1.7013201320132013e-06,
+      "loss": -0.0207,
+      "step": 220
+    },
+    {
+      "epoch": 0.1767646470705859,
+      "grad_norm": 6.048232130793178,
+      "learning_rate": 1.6996699669966995e-06,
+      "loss": 0.0301,
+      "step": 221
+    },
+    {
+      "epoch": 0.1775644871025795,
+      "grad_norm": 4.206288334982141,
+      "learning_rate": 1.698019801980198e-06,
+      "loss": -0.0503,
+      "step": 222
+    },
+    {
+      "epoch": 0.1783643271345731,
+      "grad_norm": 4.383148234898824,
+      "learning_rate": 1.6963696369636964e-06,
+      "loss": 0.0425,
+      "step": 223
+    },
+    {
+      "epoch": 0.1791641671665667,
+      "grad_norm": 4.013900208301416,
+      "learning_rate": 1.6947194719471947e-06,
+      "loss": -0.0873,
+      "step": 224
+    },
+    {
+      "epoch": 0.1799640071985603,
+      "grad_norm": 3.729807083009099,
+      "learning_rate": 1.6930693069306929e-06,
+      "loss": -0.0124,
+      "step": 225
+    },
+    {
+      "epoch": 0.1807638472305539,
+      "grad_norm": 4.739805223350201,
+      "learning_rate": 1.6914191419141915e-06,
+      "loss": -0.0965,
+      "step": 226
+    },
+    {
+      "epoch": 0.18156368726254749,
+      "grad_norm": 3.684225018193131,
+      "learning_rate": 1.6897689768976898e-06,
+      "loss": -0.0899,
+      "step": 227
+    },
+    {
+      "epoch": 0.18236352729454108,
+      "grad_norm": 4.647773349022286,
+      "learning_rate": 1.688118811881188e-06,
+      "loss": -0.1433,
+      "step": 228
+    },
+    {
+      "epoch": 0.18316336732653468,
+      "grad_norm": 4.314549940205055,
+      "learning_rate": 1.6864686468646864e-06,
+      "loss": -0.0987,
+      "step": 229
+    },
+    {
+      "epoch": 0.1839632073585283,
+      "grad_norm": 6.602144366923463,
+      "learning_rate": 1.6848184818481847e-06,
+      "loss": -0.0855,
+      "step": 230
+    },
+    {
+      "epoch": 0.1847630473905219,
+      "grad_norm": 4.611073533381248,
+      "learning_rate": 1.683168316831683e-06,
+      "loss": -0.1262,
+      "step": 231
+    },
+    {
+      "epoch": 0.1855628874225155,
+      "grad_norm": 4.9020247032635655,
+      "learning_rate": 1.6815181518151815e-06,
+      "loss": -0.1706,
+      "step": 232
+    },
+    {
+      "epoch": 0.1863627274545091,
+      "grad_norm": 4.16092482080365,
+      "learning_rate": 1.6798679867986798e-06,
+      "loss": -0.009,
+      "step": 233
+    },
+    {
+      "epoch": 0.1871625674865027,
+      "grad_norm": 3.5906088992190277,
+      "learning_rate": 1.678217821782178e-06,
+      "loss": 0.0999,
+      "step": 234
+    },
+    {
+      "epoch": 0.1879624075184963,
+      "grad_norm": 4.005270108795308,
+      "learning_rate": 1.6765676567656767e-06,
+      "loss": -0.0993,
+      "step": 235
+    },
+    {
+      "epoch": 0.1887622475504899,
+      "grad_norm": 6.563769408476828,
+      "learning_rate": 1.6749174917491749e-06,
+      "loss": -0.0193,
+      "step": 236
+    },
+    {
+      "epoch": 0.1895620875824835,
+      "grad_norm": 3.380070162840573,
+      "learning_rate": 1.6732673267326731e-06,
+      "loss": -0.0809,
+      "step": 237
+    },
+    {
+      "epoch": 0.1903619276144771,
+      "grad_norm": 4.931354996369631,
+      "learning_rate": 1.6716171617161716e-06,
+      "loss": -0.0658,
+      "step": 238
+    },
+    {
+      "epoch": 0.1911617676464707,
+      "grad_norm": 4.710207450817461,
+      "learning_rate": 1.66996699669967e-06,
+      "loss": -0.1167,
+      "step": 239
+    },
+    {
+      "epoch": 0.19196160767846432,
+      "grad_norm": 3.361685025176525,
+      "learning_rate": 1.6683168316831682e-06,
+      "loss": -0.1245,
+      "step": 240
+    },
+    {
+      "epoch": 0.19276144771045792,
+      "grad_norm": 3.767676589968502,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": -0.1638,
+      "step": 241
+    },
+    {
+      "epoch": 0.19356128774245152,
+      "grad_norm": 3.7460434704410575,
+      "learning_rate": 1.6650165016501649e-06,
+      "loss": -0.1207,
+      "step": 242
+    },
+    {
+      "epoch": 0.1943611277744451,
+      "grad_norm": 3.7655100191535413,
+      "learning_rate": 1.6633663366336631e-06,
+      "loss": 0.0038,
+      "step": 243
+    },
+    {
+      "epoch": 0.1951609678064387,
+      "grad_norm": 4.3387270640143685,
+      "learning_rate": 1.6617161716171618e-06,
+      "loss": -0.0234,
+      "step": 244
+    },
+    {
+      "epoch": 0.1959608078384323,
+      "grad_norm": 4.729420704117281,
+      "learning_rate": 1.66006600660066e-06,
+      "loss": -0.1446,
+      "step": 245
+    },
+    {
+      "epoch": 0.1967606478704259,
+      "grad_norm": 11.46352939122658,
+      "learning_rate": 1.6584158415841582e-06,
+      "loss": -0.0447,
+      "step": 246
+    },
+    {
+      "epoch": 0.1975604879024195,
+      "grad_norm": 4.6392172787916355,
+      "learning_rate": 1.6567656765676567e-06,
+      "loss": -0.1279,
+      "step": 247
+    },
+    {
+      "epoch": 0.1983603279344131,
+      "grad_norm": 7.81945174107532,
+      "learning_rate": 1.6551155115511551e-06,
+      "loss": -0.1788,
+      "step": 248
+    },
+    {
+      "epoch": 0.19916016796640673,
+      "grad_norm": 4.257894476705108,
+      "learning_rate": 1.6534653465346533e-06,
+      "loss": -0.0386,
+      "step": 249
+    },
+    {
+      "epoch": 0.19996000799840033,
+      "grad_norm": 3.9255930993081094,
+      "learning_rate": 1.6518151815181518e-06,
+      "loss": -0.0204,
+      "step": 250
+    },
+    {
+      "epoch": 0.20075984803039393,
+      "grad_norm": 8.61324493331346,
+      "learning_rate": 1.65016501650165e-06,
+      "loss": 0.0872,
+      "step": 251
+    },
+    {
+      "epoch": 0.20155968806238753,
+      "grad_norm": 3.7965562474708525,
+      "learning_rate": 1.6485148514851484e-06,
+      "loss": -0.0834,
+      "step": 252
+    },
+    {
+      "epoch": 0.20235952809438112,
+      "grad_norm": 4.327305685228189,
+      "learning_rate": 1.6468646864686469e-06,
+      "loss": -0.0639,
+      "step": 253
+    },
+    {
+      "epoch": 0.20315936812637472,
+      "grad_norm": 3.461407011747761,
+      "learning_rate": 1.6452145214521451e-06,
+      "loss": -0.1243,
+      "step": 254
+    },
+    {
+      "epoch": 0.20395920815836832,
+      "grad_norm": 5.164636623307167,
+      "learning_rate": 1.6435643564356433e-06,
+      "loss": -0.0318,
+      "step": 255
+    },
+    {
+      "epoch": 0.20475904819036192,
+      "grad_norm": 4.411537190722961,
+      "learning_rate": 1.641914191419142e-06,
+      "loss": -0.1533,
+      "step": 256
+    },
+    {
+      "epoch": 0.20555888822235552,
+      "grad_norm": 4.832045065537041,
+      "learning_rate": 1.6402640264026402e-06,
+      "loss": -0.0931,
+      "step": 257
+    },
+    {
+      "epoch": 0.20635872825434912,
+      "grad_norm": 4.133203614158014,
+      "learning_rate": 1.6386138613861385e-06,
+      "loss": -0.0951,
+      "step": 258
+    },
+    {
+      "epoch": 0.20715856828634274,
+      "grad_norm": 4.649558155027992,
+      "learning_rate": 1.636963696369637e-06,
+      "loss": -0.0459,
+      "step": 259
+    },
+    {
+      "epoch": 0.20795840831833634,
+      "grad_norm": 3.7050574045200126,
+      "learning_rate": 1.6353135313531351e-06,
+      "loss": -0.1324,
+      "step": 260
+    },
+    {
+      "epoch": 0.20875824835032994,
+      "grad_norm": 4.406446520163225,
+      "learning_rate": 1.6336633663366336e-06,
+      "loss": -0.0903,
+      "step": 261
+    },
+    {
+      "epoch": 0.20955808838232354,
+      "grad_norm": 4.150658998676116,
+      "learning_rate": 1.632013201320132e-06,
+      "loss": 0.052,
+      "step": 262
+    },
+    {
+      "epoch": 0.21035792841431714,
+      "grad_norm": 4.637643800546993,
+      "learning_rate": 1.6303630363036302e-06,
+      "loss": -0.1008,
+      "step": 263
+    },
+    {
+      "epoch": 0.21115776844631073,
+      "grad_norm": 4.356392007316505,
+      "learning_rate": 1.6287128712871285e-06,
+      "loss": -0.0666,
+      "step": 264
+    },
+    {
+      "epoch": 0.21195760847830433,
+      "grad_norm": 4.2232050225914675,
+      "learning_rate": 1.6270627062706271e-06,
+      "loss": -0.034,
+      "step": 265
+    },
+    {
+      "epoch": 0.21275744851029793,
+      "grad_norm": 4.621467065766651,
+      "learning_rate": 1.6254125412541253e-06,
+      "loss": -0.1166,
+      "step": 266
+    },
+    {
+      "epoch": 0.21355728854229153,
+      "grad_norm": 3.6996328893459385,
+      "learning_rate": 1.6237623762376238e-06,
+      "loss": -0.0651,
+      "step": 267
+    },
+    {
+      "epoch": 0.21435712857428516,
+      "grad_norm": 6.476232651431598,
+      "learning_rate": 1.622112211221122e-06,
+      "loss": -0.087,
+      "step": 268
+    },
+    {
+      "epoch": 0.21515696860627875,
+      "grad_norm": 3.3774511125642115,
+      "learning_rate": 1.6204620462046205e-06,
+      "loss": -0.0541,
+      "step": 269
+    },
+    {
+      "epoch": 0.21595680863827235,
+      "grad_norm": 8.039893341875281,
+      "learning_rate": 1.6188118811881189e-06,
+      "loss": 0.0426,
+      "step": 270
+    },
+    {
+      "epoch": 0.21675664867026595,
+      "grad_norm": 3.45288250792369,
+      "learning_rate": 1.6171617161716171e-06,
+      "loss": -0.071,
+      "step": 271
+    },
+    {
+      "epoch": 0.21755648870225955,
+      "grad_norm": 5.813145099240533,
+      "learning_rate": 1.6155115511551154e-06,
+      "loss": 0.0774,
+      "step": 272
+    },
+    {
+      "epoch": 0.21835632873425315,
+      "grad_norm": 3.4988010260216202,
+      "learning_rate": 1.6138613861386138e-06,
+      "loss": -0.0378,
+      "step": 273
+    },
+    {
+      "epoch": 0.21915616876624675,
+      "grad_norm": 4.136529473287242,
+      "learning_rate": 1.6122112211221122e-06,
+      "loss": 0.0215,
+      "step": 274
+    },
+    {
+      "epoch": 0.21995600879824034,
+      "grad_norm": 3.9538185204867884,
+      "learning_rate": 1.6105610561056105e-06,
+      "loss": -0.1014,
+      "step": 275
+    },
+    {
+      "epoch": 0.22075584883023394,
+      "grad_norm": 4.987429074808495,
+      "learning_rate": 1.608910891089109e-06,
+      "loss": -0.0326,
+      "step": 276
+    },
+    {
+      "epoch": 0.22155568886222757,
+      "grad_norm": 3.1011116987800595,
+      "learning_rate": 1.6072607260726071e-06,
+      "loss": -0.0742,
+      "step": 277
+    },
+    {
+      "epoch": 0.22235552889422117,
+      "grad_norm": 4.87646247250274,
+      "learning_rate": 1.6056105610561056e-06,
+      "loss": -0.0488,
+      "step": 278
+    },
+    {
+      "epoch": 0.22315536892621476,
+      "grad_norm": 5.406703510997709,
+      "learning_rate": 1.603960396039604e-06,
+      "loss": -0.0998,
+      "step": 279
+    },
+    {
+      "epoch": 0.22395520895820836,
+      "grad_norm": 4.199602090060885,
+      "learning_rate": 1.6023102310231022e-06,
+      "loss": 0.0787,
+      "step": 280
+    },
+    {
+      "epoch": 0.22475504899020196,
+      "grad_norm": 6.262166502823287,
+      "learning_rate": 1.6006600660066005e-06,
+      "loss": 0.0123,
+      "step": 281
+    },
+    {
+      "epoch": 0.22555488902219556,
+      "grad_norm": 5.4085402736640225,
+      "learning_rate": 1.5990099009900991e-06,
+      "loss": -0.1219,
+      "step": 282
+    },
+    {
+      "epoch": 0.22635472905418916,
+      "grad_norm": 14.35407252989058,
+      "learning_rate": 1.5973597359735973e-06,
+      "loss": -0.0427,
+      "step": 283
+    },
+    {
+      "epoch": 0.22715456908618276,
+      "grad_norm": 5.560237467243524,
+      "learning_rate": 1.5957095709570956e-06,
+      "loss": -0.0363,
+      "step": 284
+    },
+    {
+      "epoch": 0.22795440911817635,
+      "grad_norm": 5.376214533362693,
+      "learning_rate": 1.594059405940594e-06,
+      "loss": -0.1198,
+      "step": 285
+    },
+    {
+      "epoch": 0.22875424915016995,
+      "grad_norm": 7.872347430401011,
+      "learning_rate": 1.5924092409240922e-06,
+      "loss": 0.02,
+      "step": 286
+    },
+    {
+      "epoch": 0.22955408918216358,
+      "grad_norm": 4.079731942515135,
+      "learning_rate": 1.5907590759075907e-06,
+      "loss": -0.1465,
+      "step": 287
+    },
+    {
+      "epoch": 0.23035392921415718,
+      "grad_norm": 4.054081807256331,
+      "learning_rate": 1.5891089108910891e-06,
+      "loss": -0.097,
+      "step": 288
+    },
+    {
+      "epoch": 0.23115376924615078,
+      "grad_norm": 5.668828140611865,
+      "learning_rate": 1.5874587458745874e-06,
+      "loss": -0.0113,
+      "step": 289
+    },
+    {
+      "epoch": 0.23195360927814437,
+      "grad_norm": 4.222209049226612,
+      "learning_rate": 1.5858085808580856e-06,
+      "loss": -0.0565,
+      "step": 290
+    },
+    {
+      "epoch": 0.23275344931013797,
+      "grad_norm": 3.7308714963795735,
+      "learning_rate": 1.5841584158415842e-06,
+      "loss": -0.0311,
+      "step": 291
+    },
+    {
+      "epoch": 0.23355328934213157,
+      "grad_norm": 3.9384379405107914,
+      "learning_rate": 1.5825082508250825e-06,
+      "loss": -0.0921,
+      "step": 292
+    },
+    {
+      "epoch": 0.23435312937412517,
+      "grad_norm": 4.208635426370359,
+      "learning_rate": 1.5808580858085807e-06,
+      "loss": -0.1037,
+      "step": 293
+    },
+    {
+      "epoch": 0.23515296940611877,
+      "grad_norm": 4.471661666164002,
+      "learning_rate": 1.5792079207920791e-06,
+      "loss": -0.048,
+      "step": 294
+    },
+    {
+      "epoch": 0.23595280943811237,
+      "grad_norm": 5.922123322526879,
+      "learning_rate": 1.5775577557755776e-06,
+      "loss": -0.0822,
+      "step": 295
+    },
+    {
+      "epoch": 0.236752649470106,
+      "grad_norm": 3.9336004171911596,
+      "learning_rate": 1.5759075907590758e-06,
+      "loss": -0.0751,
+      "step": 296
+    },
+    {
+      "epoch": 0.2375524895020996,
+      "grad_norm": 2.9881202405051086,
+      "learning_rate": 1.5742574257425742e-06,
+      "loss": -0.0694,
+      "step": 297
+    },
+    {
+      "epoch": 0.2383523295340932,
+      "grad_norm": 7.04293625200489,
+      "learning_rate": 1.5726072607260725e-06,
+      "loss": -0.1209,
+      "step": 298
+    },
+    {
+      "epoch": 0.2391521695660868,
+      "grad_norm": 4.3791375350104165,
+      "learning_rate": 1.5709570957095707e-06,
+      "loss": -0.0704,
+      "step": 299
+    },
+    {
+      "epoch": 0.23995200959808038,
+      "grad_norm": 4.4299869604327835,
+      "learning_rate": 1.5693069306930694e-06,
+      "loss": 0.0578,
+      "step": 300
+    },
+    {
+      "epoch": 0.24075184963007398,
+      "grad_norm": 5.721847612449816,
+      "learning_rate": 1.5676567656765676e-06,
+      "loss": 0.0124,
+      "step": 301
+    },
+    {
+      "epoch": 0.24155168966206758,
+      "grad_norm": 4.346519849517093,
+      "learning_rate": 1.5660066006600658e-06,
+      "loss": -0.0676,
+      "step": 302
+    },
+    {
+      "epoch": 0.24235152969406118,
+      "grad_norm": 4.166900068739509,
+      "learning_rate": 1.5643564356435643e-06,
+      "loss": -0.0379,
+      "step": 303
+    },
+    {
+      "epoch": 0.24315136972605478,
+      "grad_norm": 4.171740126126224,
+      "learning_rate": 1.5627062706270627e-06,
+      "loss": -0.0369,
+      "step": 304
+    },
+    {
+      "epoch": 0.2439512097580484,
+      "grad_norm": 4.571373866809776,
+      "learning_rate": 1.561056105610561e-06,
+      "loss": -0.0423,
+      "step": 305
+    },
+    {
+      "epoch": 0.244751049790042,
+      "grad_norm": 4.687528076087793,
+      "learning_rate": 1.5594059405940594e-06,
+      "loss": -0.0427,
+      "step": 306
+    },
+    {
+      "epoch": 0.2455508898220356,
+      "grad_norm": 4.099266935733802,
+      "learning_rate": 1.5577557755775576e-06,
+      "loss": -0.1168,
+      "step": 307
+    },
+    {
+      "epoch": 0.2463507298540292,
+      "grad_norm": 4.76705500925925,
+      "learning_rate": 1.556105610561056e-06,
+      "loss": -0.0726,
+      "step": 308
+    },
+    {
+      "epoch": 0.2471505698860228,
+      "grad_norm": 7.726027050692815,
+      "learning_rate": 1.5544554455445545e-06,
+      "loss": -0.0858,
+      "step": 309
+    },
+    {
+      "epoch": 0.2479504099180164,
+      "grad_norm": 4.588817621333979,
+      "learning_rate": 1.5528052805280527e-06,
+      "loss": 0.0889,
+      "step": 310
+    },
+    {
+      "epoch": 0.24875024995001,
+      "grad_norm": 5.351566242300243,
+      "learning_rate": 1.551155115511551e-06,
+      "loss": -0.143,
+      "step": 311
+    },
+    {
+      "epoch": 0.2495500899820036,
+      "grad_norm": 6.279051438632601,
+      "learning_rate": 1.5495049504950496e-06,
+      "loss": 0.0312,
+      "step": 312
+    },
+    {
+      "epoch": 0.2503499300139972,
+      "grad_norm": 4.251123392069971,
+      "learning_rate": 1.5478547854785478e-06,
+      "loss": -0.0477,
+      "step": 313
+    },
+    {
+      "epoch": 0.2511497700459908,
+      "grad_norm": 4.255617580398947,
+      "learning_rate": 1.546204620462046e-06,
+      "loss": -0.1445,
+      "step": 314
+    },
+    {
+      "epoch": 0.2519496100779844,
+      "grad_norm": 3.979778076387235,
+      "learning_rate": 1.5445544554455445e-06,
+      "loss": -0.0937,
+      "step": 315
+    },
+    {
+      "epoch": 0.252749450109978,
+      "grad_norm": 5.547095237980292,
+      "learning_rate": 1.5429042904290427e-06,
+      "loss": -0.0091,
+      "step": 316
+    },
+    {
+      "epoch": 0.2535492901419716,
+      "grad_norm": 5.863554498962612,
+      "learning_rate": 1.5412541254125414e-06,
+      "loss": -0.0883,
+      "step": 317
+    },
+    {
+      "epoch": 0.2543491301739652,
+      "grad_norm": 3.9832799266815533,
+      "learning_rate": 1.5396039603960396e-06,
+      "loss": -0.1173,
+      "step": 318
+    },
+    {
+      "epoch": 0.2551489702059588,
+      "grad_norm": 4.961222194402448,
+      "learning_rate": 1.5379537953795378e-06,
+      "loss": -0.0647,
+      "step": 319
+    },
+    {
+      "epoch": 0.2559488102379524,
+      "grad_norm": 2.7901444246654945,
+      "learning_rate": 1.5363036303630363e-06,
+      "loss": -0.0873,
+      "step": 320
+    },
+    {
+      "epoch": 0.25674865026994603,
+      "grad_norm": 4.6616454131308265,
+      "learning_rate": 1.5346534653465347e-06,
+      "loss": -0.1283,
+      "step": 321
+    },
+    {
+      "epoch": 0.2575484903019396,
+      "grad_norm": 4.32603696177896,
+      "learning_rate": 1.533003300330033e-06,
+      "loss": -0.0748,
+      "step": 322
+    },
+    {
+      "epoch": 0.25834833033393323,
+      "grad_norm": 4.653928241866685,
+      "learning_rate": 1.5313531353135314e-06,
+      "loss": -0.1215,
+      "step": 323
+    },
+    {
+      "epoch": 0.2591481703659268,
+      "grad_norm": 4.476046494175142,
+      "learning_rate": 1.5297029702970296e-06,
+      "loss": -0.0247,
+      "step": 324
+    },
+    {
+      "epoch": 0.2599480103979204,
+      "grad_norm": 5.41605277862076,
+      "learning_rate": 1.528052805280528e-06,
+      "loss": 0.0055,
+      "step": 325
+    },
+    {
+      "epoch": 0.260747850429914,
+      "grad_norm": 7.359939974664472,
+      "learning_rate": 1.5264026402640265e-06,
+      "loss": -0.0994,
+      "step": 326
+    },
+    {
+      "epoch": 0.2615476904619076,
+      "grad_norm": 3.7953460503418794,
+      "learning_rate": 1.5247524752475247e-06,
+      "loss": 0.0777,
+      "step": 327
+    },
+    {
+      "epoch": 0.2623475304939012,
+      "grad_norm": 4.375620334787856,
+      "learning_rate": 1.523102310231023e-06,
+      "loss": -0.1532,
+      "step": 328
+    },
+    {
+      "epoch": 0.2631473705258948,
+      "grad_norm": 5.690054518936246,
+      "learning_rate": 1.5214521452145214e-06,
+      "loss": -0.0744,
+      "step": 329
+    },
+    {
+      "epoch": 0.26394721055788845,
+      "grad_norm": 4.095859129867475,
+      "learning_rate": 1.5198019801980198e-06,
+      "loss": -0.1342,
+      "step": 330
+    },
+    {
+      "epoch": 0.264747050589882,
+      "grad_norm": 5.261928086906211,
+      "learning_rate": 1.518151815181518e-06,
+      "loss": -0.0327,
+      "step": 331
+    },
+    {
+      "epoch": 0.26554689062187564,
+      "grad_norm": 3.723958703243353,
+      "learning_rate": 1.5165016501650165e-06,
+      "loss": -0.0919,
+      "step": 332
+    },
+    {
+      "epoch": 0.2663467306538692,
+      "grad_norm": 7.064342427249925,
+      "learning_rate": 1.5148514851485147e-06,
+      "loss": -0.3055,
+      "step": 333
+    },
+    {
+      "epoch": 0.26714657068586284,
+      "grad_norm": 5.094162064249706,
+      "learning_rate": 1.5132013201320131e-06,
+      "loss": -0.1551,
+      "step": 334
+    },
+    {
+      "epoch": 0.2679464107178564,
+      "grad_norm": 5.182464177568643,
+      "learning_rate": 1.5115511551155116e-06,
+      "loss": -0.1656,
+      "step": 335
+    },
+    {
+      "epoch": 0.26874625074985004,
+      "grad_norm": 4.205631232130195,
+      "learning_rate": 1.5099009900990098e-06,
+      "loss": -0.1339,
+      "step": 336
+    },
+    {
+      "epoch": 0.2695460907818436,
+      "grad_norm": 3.4595951551287234,
+      "learning_rate": 1.508250825082508e-06,
+      "loss": -0.0259,
+      "step": 337
+    },
+    {
+      "epoch": 0.27034593081383723,
+      "grad_norm": 4.040524953973991,
+      "learning_rate": 1.5066006600660067e-06,
+      "loss": 0.0424,
+      "step": 338
+    },
+    {
+      "epoch": 0.27114577084583086,
+      "grad_norm": 3.3792775209230044,
+      "learning_rate": 1.504950495049505e-06,
+      "loss": -0.0252,
+      "step": 339
+    },
+    {
+      "epoch": 0.27194561087782443,
+      "grad_norm": 5.329937489556339,
+      "learning_rate": 1.5033003300330032e-06,
+      "loss": -0.1064,
+      "step": 340
+    },
+    {
+      "epoch": 0.27274545090981805,
+      "grad_norm": 3.8366500907383,
+      "learning_rate": 1.5016501650165016e-06,
+      "loss": -0.1327,
+      "step": 341
+    },
+    {
+      "epoch": 0.2735452909418116,
+      "grad_norm": 3.4211397121327334,
+      "learning_rate": 1.5e-06,
+      "loss": -0.0469,
+      "step": 342
+    },
+    {
+      "epoch": 0.27434513097380525,
+      "grad_norm": 7.190396728605877,
+      "learning_rate": 1.4983498349834983e-06,
+      "loss": -0.0455,
+      "step": 343
+    },
+    {
+      "epoch": 0.2751449710057988,
+      "grad_norm": 5.208941899667468,
+      "learning_rate": 1.4966996699669967e-06,
+      "loss": -0.0118,
+      "step": 344
+    },
+    {
+      "epoch": 0.27594481103779245,
+      "grad_norm": 5.0666099160345635,
+      "learning_rate": 1.495049504950495e-06,
+      "loss": -0.08,
+      "step": 345
+    },
+    {
+      "epoch": 0.276744651069786,
+      "grad_norm": 3.445240945570377,
+      "learning_rate": 1.4933993399339932e-06,
+      "loss": 0.0043,
+      "step": 346
+    },
+    {
+      "epoch": 0.27754449110177964,
+      "grad_norm": 6.719396089938185,
+      "learning_rate": 1.4917491749174918e-06,
+      "loss": -0.0528,
+      "step": 347
+    },
+    {
+      "epoch": 0.27834433113377327,
+      "grad_norm": 4.948551220275233,
+      "learning_rate": 1.49009900990099e-06,
+      "loss": -0.0351,
+      "step": 348
+    },
+    {
+      "epoch": 0.27914417116576684,
+      "grad_norm": 4.198757242081244,
+      "learning_rate": 1.4884488448844883e-06,
+      "loss": -0.1767,
+      "step": 349
+    },
+    {
+      "epoch": 0.27994401119776047,
+      "grad_norm": 4.020517893591624,
+      "learning_rate": 1.4867986798679867e-06,
+      "loss": -0.0777,
+      "step": 350
+    },
+    {
+      "epoch": 0.28074385122975404,
+      "grad_norm": 7.665385125345826,
+      "learning_rate": 1.4851485148514852e-06,
+      "loss": -0.0675,
+      "step": 351
+    },
+    {
+      "epoch": 0.28154369126174766,
+      "grad_norm": 4.359035902610134,
+      "learning_rate": 1.4834983498349834e-06,
+      "loss": 0.1022,
+      "step": 352
+    },
+    {
+      "epoch": 0.28234353129374123,
+      "grad_norm": 4.515833866344382,
+      "learning_rate": 1.4818481848184818e-06,
+      "loss": -0.1318,
+      "step": 353
+    },
+    {
+      "epoch": 0.28314337132573486,
+      "grad_norm": 10.48643307447715,
+      "learning_rate": 1.48019801980198e-06,
+      "loss": -0.0624,
+      "step": 354
+    },
+    {
+      "epoch": 0.28394321135772843,
+      "grad_norm": 3.9055137245563167,
+      "learning_rate": 1.4785478547854785e-06,
+      "loss": 0.0131,
+      "step": 355
+    },
+    {
+      "epoch": 0.28474305138972206,
+      "grad_norm": 5.064555563223541,
+      "learning_rate": 1.476897689768977e-06,
+      "loss": -0.0848,
+      "step": 356
+    },
+    {
+      "epoch": 0.2855428914217157,
+      "grad_norm": 6.403904331900866,
+      "learning_rate": 1.4752475247524752e-06,
+      "loss": -0.1231,
+      "step": 357
+    },
+    {
+      "epoch": 0.28634273145370925,
+      "grad_norm": 4.4680198659839405,
+      "learning_rate": 1.4735973597359734e-06,
+      "loss": -0.0751,
+      "step": 358
+    },
+    {
+      "epoch": 0.2871425714857029,
+      "grad_norm": 7.88048544071049,
+      "learning_rate": 1.4719471947194718e-06,
+      "loss": -0.0111,
+      "step": 359
+    },
+    {
+      "epoch": 0.28794241151769645,
+      "grad_norm": 4.041245481168213,
+      "learning_rate": 1.4702970297029703e-06,
+      "loss": -0.0219,
+      "step": 360
+    },
+    {
+      "epoch": 0.2887422515496901,
+      "grad_norm": 3.2378522821181748,
+      "learning_rate": 1.4686468646864685e-06,
+      "loss": 0.0154,
+      "step": 361
+    },
+    {
+      "epoch": 0.28954209158168365,
+      "grad_norm": 5.187324980575399,
+      "learning_rate": 1.466996699669967e-06,
+      "loss": 0.0768,
+      "step": 362
+    },
+    {
+      "epoch": 0.2903419316136773,
+      "grad_norm": 3.892629574264858,
+      "learning_rate": 1.4653465346534652e-06,
+      "loss": -0.1757,
+      "step": 363
+    },
+    {
+      "epoch": 0.29114177164567084,
+      "grad_norm": 4.66291997005039,
+      "learning_rate": 1.4636963696369636e-06,
+      "loss": -0.0897,
+      "step": 364
+    },
+    {
+      "epoch": 0.29194161167766447,
+      "grad_norm": 4.490266870328807,
+      "learning_rate": 1.462046204620462e-06,
+      "loss": -0.1638,
+      "step": 365
+    },
+    {
+      "epoch": 0.2927414517096581,
+      "grad_norm": 7.248644471878,
+      "learning_rate": 1.4603960396039603e-06,
+      "loss": -0.1413,
+      "step": 366
+    },
+    {
+      "epoch": 0.29354129174165167,
+      "grad_norm": 13.65496051906939,
+      "learning_rate": 1.4587458745874585e-06,
+      "loss": -0.0144,
+      "step": 367
+    },
+    {
+      "epoch": 0.2943411317736453,
+      "grad_norm": 3.0336626027850593,
+      "learning_rate": 1.4570957095709572e-06,
+      "loss": -0.008,
+      "step": 368
+    },
+    {
+      "epoch": 0.29514097180563886,
+      "grad_norm": 7.873854054225054,
+      "learning_rate": 1.4554455445544554e-06,
+      "loss": 0.1034,
+      "step": 369
+    },
+    {
+      "epoch": 0.2959408118376325,
+      "grad_norm": 3.727594324731175,
+      "learning_rate": 1.4537953795379538e-06,
+      "loss": -0.1401,
+      "step": 370
+    },
+    {
+      "epoch": 0.29674065186962606,
+      "grad_norm": 5.229701446706082,
+      "learning_rate": 1.452145214521452e-06,
+      "loss": -0.1203,
+      "step": 371
+    },
+    {
+      "epoch": 0.2975404919016197,
+      "grad_norm": 4.492128268970922,
+      "learning_rate": 1.4504950495049503e-06,
+      "loss": 0.0004,
+      "step": 372
+    },
+    {
+      "epoch": 0.29834033193361326,
+      "grad_norm": 5.077090301738471,
+      "learning_rate": 1.448844884488449e-06,
+      "loss": 0.0219,
+      "step": 373
+    },
+    {
+      "epoch": 0.2991401719656069,
+      "grad_norm": 4.796744776644939,
+      "learning_rate": 1.4471947194719472e-06,
+      "loss": -0.0064,
+      "step": 374
+    },
+    {
+      "epoch": 0.29994001199760045,
+      "grad_norm": 7.42447528462134,
+      "learning_rate": 1.4455445544554454e-06,
+      "loss": 0.07,
+      "step": 375
+    },
+    {
+      "epoch": 0.3007398520295941,
+      "grad_norm": 3.848638759590051,
+      "learning_rate": 1.4438943894389438e-06,
+      "loss": -0.0777,
+      "step": 376
+    },
+    {
+      "epoch": 0.3015396920615877,
+      "grad_norm": 4.256980996790008,
+      "learning_rate": 1.4422442244224423e-06,
+      "loss": -0.1766,
+      "step": 377
+    },
+    {
+      "epoch": 0.3023395320935813,
+      "grad_norm": 3.961327287203466,
+      "learning_rate": 1.4405940594059405e-06,
+      "loss": -0.0571,
+      "step": 378
+    },
+    {
+      "epoch": 0.3031393721255749,
+      "grad_norm": 5.478690567895318,
+      "learning_rate": 1.438943894389439e-06,
+      "loss": 0.0013,
+      "step": 379
+    },
+    {
+      "epoch": 0.3039392121575685,
+      "grad_norm": 3.8685538296119106,
+      "learning_rate": 1.4372937293729372e-06,
+      "loss": -0.0135,
+      "step": 380
+    },
+    {
+      "epoch": 0.3047390521895621,
+      "grad_norm": 3.712350805091167,
+      "learning_rate": 1.4356435643564356e-06,
+      "loss": -0.0965,
+      "step": 381
+    },
+    {
+      "epoch": 0.30553889222155567,
+      "grad_norm": 4.12545866294737,
+      "learning_rate": 1.433993399339934e-06,
+      "loss": 0.0192,
+      "step": 382
+    },
+    {
+      "epoch": 0.3063387322535493,
+      "grad_norm": 3.9826126090375085,
+      "learning_rate": 1.4323432343234323e-06,
+      "loss": 0.0096,
+      "step": 383
+    },
+    {
+      "epoch": 0.30713857228554287,
+      "grad_norm": 5.253969236088526,
+      "learning_rate": 1.4306930693069305e-06,
+      "loss": 0.0596,
+      "step": 384
+    },
+    {
+      "epoch": 0.3079384123175365,
+      "grad_norm": 4.369221167744991,
+      "learning_rate": 1.4290429042904292e-06,
+      "loss": -0.0586,
+      "step": 385
+    },
+    {
+      "epoch": 0.3087382523495301,
+      "grad_norm": 3.386456014084215,
+      "learning_rate": 1.4273927392739274e-06,
+      "loss": -0.1952,
+      "step": 386
+    },
+    {
+      "epoch": 0.3095380923815237,
+      "grad_norm": 4.175162288229841,
+      "learning_rate": 1.4257425742574256e-06,
+      "loss": -0.1559,
+      "step": 387
+    },
+    {
+      "epoch": 0.3103379324135173,
+      "grad_norm": 4.07269720996871,
+      "learning_rate": 1.424092409240924e-06,
+      "loss": -0.0591,
+      "step": 388
+    },
+    {
+      "epoch": 0.3111377724455109,
+      "grad_norm": 3.873233515579836,
+      "learning_rate": 1.4224422442244223e-06,
+      "loss": -0.0649,
+      "step": 389
+    },
+    {
+      "epoch": 0.3119376124775045,
+      "grad_norm": 5.33165026969968,
+      "learning_rate": 1.4207920792079207e-06,
+      "loss": -0.0568,
+      "step": 390
+    },
+    {
+      "epoch": 0.3127374525094981,
+      "grad_norm": 5.644618937197355,
+      "learning_rate": 1.4191419141914192e-06,
+      "loss": -0.0425,
+      "step": 391
+    },
+    {
+      "epoch": 0.3135372925414917,
+      "grad_norm": 4.609038777130941,
+      "learning_rate": 1.4174917491749174e-06,
+      "loss": -0.0991,
+      "step": 392
+    },
+    {
+      "epoch": 0.3143371325734853,
+      "grad_norm": 5.362814464107483,
+      "learning_rate": 1.4158415841584156e-06,
+      "loss": -0.0377,
+      "step": 393
+    },
+    {
+      "epoch": 0.3151369726054789,
+      "grad_norm": 4.1100020129716315,
+      "learning_rate": 1.4141914191419143e-06,
+      "loss": -0.0176,
+      "step": 394
+    },
+    {
+      "epoch": 0.31593681263747253,
+      "grad_norm": 3.6462471572713198,
+      "learning_rate": 1.4125412541254125e-06,
+      "loss": 0.0183,
+      "step": 395
+    },
+    {
+      "epoch": 0.3167366526694661,
+      "grad_norm": 3.425535847868438,
+      "learning_rate": 1.4108910891089107e-06,
+      "loss": -0.1166,
+      "step": 396
+    },
+    {
+      "epoch": 0.3175364927014597,
+      "grad_norm": 4.023065583159361,
+      "learning_rate": 1.4092409240924092e-06,
+      "loss": -0.069,
+      "step": 397
+    },
+    {
+      "epoch": 0.3183363327334533,
+      "grad_norm": 4.435192529053884,
+      "learning_rate": 1.4075907590759076e-06,
+      "loss": -0.1024,
+      "step": 398
+    },
+    {
+      "epoch": 0.3191361727654469,
+      "grad_norm": 4.351874787170239,
+      "learning_rate": 1.4059405940594058e-06,
+      "loss": -0.1381,
+      "step": 399
+    },
+    {
+      "epoch": 0.3199360127974405,
+      "grad_norm": 5.114118048590294,
+      "learning_rate": 1.4042904290429043e-06,
+      "loss": -0.027,
+      "step": 400
+    },
+    {
+      "epoch": 0.3207358528294341,
+      "grad_norm": 6.62264310550409,
+      "learning_rate": 1.4026402640264025e-06,
+      "loss": -0.1372,
+      "step": 401
+    },
+    {
+      "epoch": 0.3215356928614277,
+      "grad_norm": 5.515472496453124,
+      "learning_rate": 1.4009900990099007e-06,
+      "loss": 0.0028,
+      "step": 402
+    },
+    {
+      "epoch": 0.3223355328934213,
+      "grad_norm": 5.43524070368167,
+      "learning_rate": 1.3993399339933994e-06,
+      "loss": -0.1192,
+      "step": 403
+    },
+    {
+      "epoch": 0.32313537292541494,
+      "grad_norm": 4.309916510249054,
+      "learning_rate": 1.3976897689768976e-06,
+      "loss": -0.049,
+      "step": 404
+    },
+    {
+      "epoch": 0.3239352129574085,
+      "grad_norm": 4.392826058059571,
+      "learning_rate": 1.3960396039603959e-06,
+      "loss": -0.1248,
+      "step": 405
+    },
+    {
+      "epoch": 0.32473505298940214,
+      "grad_norm": 5.384606404349416,
+      "learning_rate": 1.3943894389438943e-06,
+      "loss": -0.0248,
+      "step": 406
+    },
+    {
+      "epoch": 0.3255348930213957,
+      "grad_norm": 5.369884451931867,
+      "learning_rate": 1.3927392739273927e-06,
+      "loss": 0.0453,
+      "step": 407
+    },
+    {
+      "epoch": 0.32633473305338934,
+      "grad_norm": 3.799887635426884,
+      "learning_rate": 1.391089108910891e-06,
+      "loss": 0.0924,
+      "step": 408
+    },
+    {
+      "epoch": 0.3271345730853829,
+      "grad_norm": 5.151153079821819,
+      "learning_rate": 1.3894389438943894e-06,
+      "loss": -0.1524,
+      "step": 409
+    },
+    {
+      "epoch": 0.32793441311737653,
+      "grad_norm": 4.9429474730234935,
+      "learning_rate": 1.3877887788778876e-06,
+      "loss": -0.0066,
+      "step": 410
+    },
+    {
+      "epoch": 0.3287342531493701,
+      "grad_norm": 3.8669767688401637,
+      "learning_rate": 1.386138613861386e-06,
+      "loss": -0.0998,
+      "step": 411
+    },
+    {
+      "epoch": 0.32953409318136373,
+      "grad_norm": 4.1249285605053165,
+      "learning_rate": 1.3844884488448845e-06,
+      "loss": -0.1198,
+      "step": 412
+    },
+    {
+      "epoch": 0.33033393321335736,
+      "grad_norm": 4.264021911092433,
+      "learning_rate": 1.3828382838283827e-06,
+      "loss": -0.027,
+      "step": 413
+    },
+    {
+      "epoch": 0.3311337732453509,
+      "grad_norm": 8.192155984704781,
+      "learning_rate": 1.381188118811881e-06,
+      "loss": 0.0037,
+      "step": 414
+    },
+    {
+      "epoch": 0.33193361327734455,
+      "grad_norm": 4.842071045333458,
+      "learning_rate": 1.3795379537953794e-06,
+      "loss": -0.0183,
+      "step": 415
+    },
+    {
+      "epoch": 0.3327334533093381,
+      "grad_norm": 5.69008602834876,
+      "learning_rate": 1.3778877887788779e-06,
+      "loss": -0.037,
+      "step": 416
+    },
+    {
+      "epoch": 0.33353329334133175,
+      "grad_norm": 3.4506588237689044,
+      "learning_rate": 1.376237623762376e-06,
+      "loss": -0.1827,
+      "step": 417
+    },
+    {
+      "epoch": 0.3343331333733253,
+      "grad_norm": 4.276859677588479,
+      "learning_rate": 1.3745874587458745e-06,
+      "loss": -0.1397,
+      "step": 418
+    },
+    {
+      "epoch": 0.33513297340531895,
+      "grad_norm": 5.137955642134524,
+      "learning_rate": 1.3729372937293728e-06,
+      "loss": 0.0313,
+      "step": 419
+    },
+    {
+      "epoch": 0.3359328134373125,
+      "grad_norm": 5.626427193889533,
+      "learning_rate": 1.3712871287128714e-06,
+      "loss": 0.0585,
+      "step": 420
+    },
+    {
+      "epoch": 0.33673265346930614,
+      "grad_norm": 4.259015114708382,
+      "learning_rate": 1.3696369636963696e-06,
+      "loss": -0.1352,
+      "step": 421
+    },
+    {
+      "epoch": 0.3375324935012997,
+      "grad_norm": 3.900501996524311,
+      "learning_rate": 1.3679867986798679e-06,
+      "loss": -0.1541,
+      "step": 422
+    },
+    {
+      "epoch": 0.33833233353329334,
+      "grad_norm": 21.342155698599925,
+      "learning_rate": 1.3663366336633663e-06,
+      "loss": 0.0163,
+      "step": 423
+    },
+    {
+      "epoch": 0.33913217356528697,
+      "grad_norm": 9.376314198251674,
+      "learning_rate": 1.3646864686468647e-06,
+      "loss": -0.1147,
+      "step": 424
+    },
+    {
+      "epoch": 0.33993201359728054,
+      "grad_norm": 3.9556694436435773,
+      "learning_rate": 1.363036303630363e-06,
+      "loss": 0.0607,
+      "step": 425
+    },
+    {
+      "epoch": 0.34073185362927416,
+      "grad_norm": 4.413407376716041,
+      "learning_rate": 1.3613861386138614e-06,
+      "loss": 0.0269,
+      "step": 426
+    },
+    {
+      "epoch": 0.34153169366126773,
+      "grad_norm": 4.5745629523971285,
+      "learning_rate": 1.3597359735973596e-06,
+      "loss": -0.1232,
+      "step": 427
+    },
+    {
+      "epoch": 0.34233153369326136,
+      "grad_norm": 6.482169711595175,
+      "learning_rate": 1.3580858085808579e-06,
+      "loss": -0.0933,
+      "step": 428
+    },
+    {
+      "epoch": 0.34313137372525493,
+      "grad_norm": 4.614948794989073,
+      "learning_rate": 1.3564356435643565e-06,
+      "loss": -0.0881,
+      "step": 429
+    },
+    {
+      "epoch": 0.34393121375724856,
+      "grad_norm": 4.902443350581836,
+      "learning_rate": 1.3547854785478547e-06,
+      "loss": -0.0492,
+      "step": 430
+    },
+    {
+      "epoch": 0.3447310537892421,
+      "grad_norm": 4.293832374460016,
+      "learning_rate": 1.353135313531353e-06,
+      "loss": -0.0988,
+      "step": 431
+    },
+    {
+      "epoch": 0.34553089382123575,
+      "grad_norm": 4.239300667652253,
+      "learning_rate": 1.3514851485148514e-06,
+      "loss": -0.0395,
+      "step": 432
+    },
+    {
+      "epoch": 0.3463307338532294,
+      "grad_norm": 6.000658634911202,
+      "learning_rate": 1.3498349834983499e-06,
+      "loss": -0.0687,
+      "step": 433
+    },
+    {
+      "epoch": 0.34713057388522295,
+      "grad_norm": 4.533327665512432,
+      "learning_rate": 1.348184818481848e-06,
+      "loss": -0.0457,
+      "step": 434
+    },
+    {
+      "epoch": 0.3479304139172166,
+      "grad_norm": 4.966203144811649,
+      "learning_rate": 1.3465346534653465e-06,
+      "loss": -0.0891,
+      "step": 435
+    },
+    {
+      "epoch": 0.34873025394921014,
+      "grad_norm": 3.160979702375991,
+      "learning_rate": 1.3448844884488448e-06,
+      "loss": 0.0509,
+      "step": 436
+    },
+    {
+      "epoch": 0.34953009398120377,
+      "grad_norm": 7.163626654459487,
+      "learning_rate": 1.3432343234323432e-06,
+      "loss": -0.1026,
+      "step": 437
+    },
+    {
+      "epoch": 0.35032993401319734,
+      "grad_norm": 4.10929586240042,
+      "learning_rate": 1.3415841584158416e-06,
+      "loss": -0.0346,
+      "step": 438
+    },
+    {
+      "epoch": 0.35112977404519097,
+      "grad_norm": 7.444864169509166,
+      "learning_rate": 1.3399339933993399e-06,
+      "loss": -0.084,
+      "step": 439
+    },
+    {
+      "epoch": 0.35192961407718454,
+      "grad_norm": 4.279436158804133,
+      "learning_rate": 1.338283828382838e-06,
+      "loss": -0.0753,
+      "step": 440
+    },
+    {
+      "epoch": 0.35272945410917816,
+      "grad_norm": 7.0310221317242965,
+      "learning_rate": 1.3366336633663367e-06,
+      "loss": 0.0822,
+      "step": 441
+    },
+    {
+      "epoch": 0.3535292941411718,
+      "grad_norm": 3.546380500099962,
+      "learning_rate": 1.334983498349835e-06,
+      "loss": -0.0826,
+      "step": 442
+    },
+    {
+      "epoch": 0.35432913417316536,
+      "grad_norm": 3.978575910618056,
+      "learning_rate": 1.3333333333333332e-06,
+      "loss": -0.0183,
+      "step": 443
+    },
+    {
+      "epoch": 0.355128974205159,
+      "grad_norm": 4.893702894351932,
+      "learning_rate": 1.3316831683168316e-06,
+      "loss": 0.0513,
+      "step": 444
+    },
+    {
+      "epoch": 0.35592881423715256,
+      "grad_norm": 4.712476792012751,
+      "learning_rate": 1.3300330033003299e-06,
+      "loss": 0.0161,
+      "step": 445
+    },
+    {
+      "epoch": 0.3567286542691462,
+      "grad_norm": 4.363095681693482,
+      "learning_rate": 1.3283828382838283e-06,
+      "loss": -0.0878,
+      "step": 446
+    },
+    {
+      "epoch": 0.35752849430113975,
+      "grad_norm": 3.6779713769559206,
+      "learning_rate": 1.3267326732673268e-06,
+      "loss": -0.0884,
+      "step": 447
+    },
+    {
+      "epoch": 0.3583283343331334,
+      "grad_norm": 4.691244638726057,
+      "learning_rate": 1.325082508250825e-06,
+      "loss": 0.0164,
+      "step": 448
+    },
+    {
+      "epoch": 0.35912817436512695,
+      "grad_norm": 3.9918624835208574,
+      "learning_rate": 1.3234323432343232e-06,
+      "loss": -0.0623,
+      "step": 449
+    },
+    {
+      "epoch": 0.3599280143971206,
+      "grad_norm": 4.3423857158760475,
+      "learning_rate": 1.3217821782178219e-06,
+      "loss": -0.0166,
+      "step": 450
+    },
+    {
+      "epoch": 0.3607278544291142,
+      "grad_norm": 3.3557272335230266,
+      "learning_rate": 1.32013201320132e-06,
+      "loss": -0.1243,
+      "step": 451
+    },
+    {
+      "epoch": 0.3615276944611078,
+      "grad_norm": 4.121010209091045,
+      "learning_rate": 1.3184818481848183e-06,
+      "loss": 0.0157,
+      "step": 452
+    },
+    {
+      "epoch": 0.3623275344931014,
+      "grad_norm": 5.61777014754645,
+      "learning_rate": 1.3168316831683168e-06,
+      "loss": 0.0009,
+      "step": 453
+    },
+    {
+      "epoch": 0.36312737452509497,
+      "grad_norm": 3.9497241442966673,
+      "learning_rate": 1.3151815181518152e-06,
+      "loss": -0.0748,
+      "step": 454
+    },
+    {
+      "epoch": 0.3639272145570886,
+      "grad_norm": 3.78165099484685,
+      "learning_rate": 1.3135313531353134e-06,
+      "loss": 0.0137,
+      "step": 455
+    },
+    {
+      "epoch": 0.36472705458908217,
+      "grad_norm": 3.8365555088656573,
+      "learning_rate": 1.3118811881188119e-06,
+      "loss": 0.0,
+      "step": 456
+    },
+    {
+      "epoch": 0.3655268946210758,
+      "grad_norm": 3.9613296946642933,
+      "learning_rate": 1.31023102310231e-06,
+      "loss": -0.068,
+      "step": 457
+    },
+    {
+      "epoch": 0.36632673465306936,
+      "grad_norm": 3.558717962079936,
+      "learning_rate": 1.3085808580858083e-06,
+      "loss": -0.1112,
+      "step": 458
+    },
+    {
+      "epoch": 0.367126574685063,
+      "grad_norm": 4.93902023669042,
+      "learning_rate": 1.306930693069307e-06,
+      "loss": -0.0433,
+      "step": 459
+    },
+    {
+      "epoch": 0.3679264147170566,
+      "grad_norm": 4.69421251966819,
+      "learning_rate": 1.3052805280528052e-06,
+      "loss": 0.0994,
+      "step": 460
+    },
+    {
+      "epoch": 0.3687262547490502,
+      "grad_norm": 6.3721851791610336,
+      "learning_rate": 1.3036303630363034e-06,
+      "loss": -0.1078,
+      "step": 461
+    },
+    {
+      "epoch": 0.3695260947810438,
+      "grad_norm": 3.389060929800596,
+      "learning_rate": 1.3019801980198019e-06,
+      "loss": -0.088,
+      "step": 462
+    },
+    {
+      "epoch": 0.3703259348130374,
+      "grad_norm": 5.748513070947605,
+      "learning_rate": 1.3003300330033003e-06,
+      "loss": -0.029,
+      "step": 463
+    },
+    {
+      "epoch": 0.371125774845031,
+      "grad_norm": 4.690511727792042,
+      "learning_rate": 1.2986798679867985e-06,
+      "loss": -0.0756,
+      "step": 464
+    },
+    {
+      "epoch": 0.3719256148770246,
+      "grad_norm": 5.217441052047622,
+      "learning_rate": 1.297029702970297e-06,
+      "loss": -0.0748,
+      "step": 465
+    },
+    {
+      "epoch": 0.3727254549090182,
+      "grad_norm": 4.240980113487688,
+      "learning_rate": 1.2953795379537952e-06,
+      "loss": -0.0008,
+      "step": 466
+    },
+    {
+      "epoch": 0.3735252949410118,
+      "grad_norm": 4.743889341456478,
+      "learning_rate": 1.2937293729372937e-06,
+      "loss": -0.0671,
+      "step": 467
+    },
+    {
+      "epoch": 0.3743251349730054,
+      "grad_norm": 4.473362389672442,
+      "learning_rate": 1.292079207920792e-06,
+      "loss": 0.0101,
+      "step": 468
+    },
+    {
+      "epoch": 0.375124975004999,
+      "grad_norm": 4.197750015674087,
+      "learning_rate": 1.2904290429042903e-06,
+      "loss": 0.0175,
+      "step": 469
+    },
+    {
+      "epoch": 0.3759248150369926,
+      "grad_norm": 6.425414954415456,
+      "learning_rate": 1.2887788778877888e-06,
+      "loss": -0.0783,
+      "step": 470
+    },
+    {
+      "epoch": 0.3767246550689862,
+      "grad_norm": 2.9864850798252855,
+      "learning_rate": 1.2871287128712872e-06,
+      "loss": -0.0884,
+      "step": 471
+    },
+    {
+      "epoch": 0.3775244951009798,
+      "grad_norm": 5.261385424958508,
+      "learning_rate": 1.2854785478547854e-06,
+      "loss": -0.09,
+      "step": 472
+    },
+    {
+      "epoch": 0.3783243351329734,
+      "grad_norm": 4.958045993888585,
+      "learning_rate": 1.2838283828382839e-06,
+      "loss": -0.0682,
+      "step": 473
+    },
+    {
+      "epoch": 0.379124175164967,
+      "grad_norm": 4.7190018138263605,
+      "learning_rate": 1.282178217821782e-06,
+      "loss": -0.1395,
+      "step": 474
+    },
+    {
+      "epoch": 0.3799240151969606,
+      "grad_norm": 5.882864051380202,
+      "learning_rate": 1.2805280528052803e-06,
+      "loss": -0.1829,
+      "step": 475
+    },
+    {
+      "epoch": 0.3807238552289542,
+      "grad_norm": 3.7556665205378352,
+      "learning_rate": 1.278877887788779e-06,
+      "loss": -0.021,
+      "step": 476
+    },
+    {
+      "epoch": 0.3815236952609478,
+      "grad_norm": 4.086321431606577,
+      "learning_rate": 1.2772277227722772e-06,
+      "loss": -0.0382,
+      "step": 477
+    },
+    {
+      "epoch": 0.3823235352929414,
+      "grad_norm": 4.616776862820448,
+      "learning_rate": 1.2755775577557754e-06,
+      "loss": -0.1779,
+      "step": 478
+    },
+    {
+      "epoch": 0.383123375324935,
+      "grad_norm": 4.004332580198827,
+      "learning_rate": 1.2739273927392739e-06,
+      "loss": -0.0252,
+      "step": 479
+    },
+    {
+      "epoch": 0.38392321535692864,
+      "grad_norm": 4.624789258949781,
+      "learning_rate": 1.2722772277227723e-06,
+      "loss": -0.0274,
+      "step": 480
+    },
+    {
+      "epoch": 0.3847230553889222,
+      "grad_norm": 4.107644532644881,
+      "learning_rate": 1.2706270627062705e-06,
+      "loss": -0.0706,
+      "step": 481
+    },
+    {
+      "epoch": 0.38552289542091583,
+      "grad_norm": 5.606536912327608,
+      "learning_rate": 1.268976897689769e-06,
+      "loss": -0.1579,
+      "step": 482
+    },
+    {
+      "epoch": 0.3863227354529094,
+      "grad_norm": 3.661768864377637,
+      "learning_rate": 1.2673267326732672e-06,
+      "loss": -0.0483,
+      "step": 483
+    },
+    {
+      "epoch": 0.38712257548490303,
+      "grad_norm": 4.163789722318428,
+      "learning_rate": 1.2656765676567657e-06,
+      "loss": -0.1628,
+      "step": 484
+    },
+    {
+      "epoch": 0.3879224155168966,
+      "grad_norm": 5.862521290689618,
+      "learning_rate": 1.264026402640264e-06,
+      "loss": -0.0378,
+      "step": 485
+    },
+    {
+      "epoch": 0.3887222555488902,
+      "grad_norm": 4.451191371926914,
+      "learning_rate": 1.2623762376237623e-06,
+      "loss": 0.041,
+      "step": 486
+    },
+    {
+      "epoch": 0.3895220955808838,
+      "grad_norm": 6.554041470323983,
+      "learning_rate": 1.2607260726072606e-06,
+      "loss": -0.0089,
+      "step": 487
+    },
+    {
+      "epoch": 0.3903219356128774,
+      "grad_norm": 4.958459911280161,
+      "learning_rate": 1.259075907590759e-06,
+      "loss": -0.0351,
+      "step": 488
+    },
+    {
+      "epoch": 0.39112177564487105,
+      "grad_norm": 5.5754285433841595,
+      "learning_rate": 1.2574257425742574e-06,
+      "loss": -0.0866,
+      "step": 489
+    },
+    {
+      "epoch": 0.3919216156768646,
+      "grad_norm": 4.927561354349523,
+      "learning_rate": 1.2557755775577557e-06,
+      "loss": 0.0114,
+      "step": 490
+    },
+    {
+      "epoch": 0.39272145570885825,
+      "grad_norm": 4.275369657183623,
+      "learning_rate": 1.2541254125412541e-06,
+      "loss": 0.0731,
+      "step": 491
+    },
+    {
+      "epoch": 0.3935212957408518,
+      "grad_norm": 4.553288397020381,
+      "learning_rate": 1.2524752475247523e-06,
+      "loss": -0.0366,
+      "step": 492
+    },
+    {
+      "epoch": 0.39432113577284544,
+      "grad_norm": 4.3640356820358415,
+      "learning_rate": 1.2508250825082508e-06,
+      "loss": -0.026,
+      "step": 493
+    },
+    {
+      "epoch": 0.395120975804839,
+      "grad_norm": 6.781778763227194,
+      "learning_rate": 1.2491749174917492e-06,
+      "loss": 0.0272,
+      "step": 494
+    },
+    {
+      "epoch": 0.39592081583683264,
+      "grad_norm": 4.147600624744722,
+      "learning_rate": 1.2475247524752474e-06,
+      "loss": -0.0533,
+      "step": 495
+    },
+    {
+      "epoch": 0.3967206558688262,
+      "grad_norm": 7.925587764087279,
+      "learning_rate": 1.2458745874587457e-06,
+      "loss": -0.0023,
+      "step": 496
+    },
+    {
+      "epoch": 0.39752049590081984,
+      "grad_norm": 3.9471683782785267,
+      "learning_rate": 1.2442244224422443e-06,
+      "loss": -0.0624,
+      "step": 497
+    },
+    {
+      "epoch": 0.39832033593281346,
+      "grad_norm": 5.046330000323796,
+      "learning_rate": 1.2425742574257426e-06,
+      "loss": -0.1152,
+      "step": 498
+    },
+    {
+      "epoch": 0.39912017596480703,
+      "grad_norm": 3.797212185428219,
+      "learning_rate": 1.2409240924092408e-06,
+      "loss": -0.0566,
+      "step": 499
+    },
+    {
+      "epoch": 0.39992001599680066,
+      "grad_norm": 5.219397955775355,
+      "learning_rate": 1.2392739273927392e-06,
+      "loss": -0.0473,
+      "step": 500
+    },
+    {
+      "epoch": 0.40071985602879423,
+      "grad_norm": 4.888043487068187,
+      "learning_rate": 1.2376237623762375e-06,
+      "loss": -0.0766,
+      "step": 501
+    },
+    {
+      "epoch": 0.40151969606078786,
+      "grad_norm": 4.812490488018197,
+      "learning_rate": 1.2359735973597359e-06,
+      "loss": -0.0063,
+      "step": 502
+    },
+    {
+      "epoch": 0.4023195360927814,
+      "grad_norm": 4.137421603194797,
+      "learning_rate": 1.2343234323432343e-06,
+      "loss": -0.0894,
+      "step": 503
+    },
+    {
+      "epoch": 0.40311937612477505,
+      "grad_norm": 4.446466976487968,
+      "learning_rate": 1.2326732673267326e-06,
+      "loss": -0.1032,
+      "step": 504
+    },
+    {
+      "epoch": 0.4039192161567686,
+      "grad_norm": 4.517828908408806,
+      "learning_rate": 1.2310231023102308e-06,
+      "loss": -0.0366,
+      "step": 505
+    },
+    {
+      "epoch": 0.40471905618876225,
+      "grad_norm": 4.025363379714323,
+      "learning_rate": 1.2293729372937294e-06,
+      "loss": -0.0322,
+      "step": 506
+    },
+    {
+      "epoch": 0.4055188962207559,
+      "grad_norm": 3.0065985273378026,
+      "learning_rate": 1.2277227722772277e-06,
+      "loss": -0.1451,
+      "step": 507
+    },
+    {
+      "epoch": 0.40631873625274945,
+      "grad_norm": 6.102992165053075,
+      "learning_rate": 1.226072607260726e-06,
+      "loss": -0.0947,
+      "step": 508
+    },
+    {
+      "epoch": 0.4071185762847431,
+      "grad_norm": 5.373085923158729,
+      "learning_rate": 1.2244224422442243e-06,
+      "loss": -0.038,
+      "step": 509
+    },
+    {
+      "epoch": 0.40791841631673664,
+      "grad_norm": 6.507842394701745,
+      "learning_rate": 1.2227722772277228e-06,
+      "loss": -0.0416,
+      "step": 510
+    },
+    {
+      "epoch": 0.40871825634873027,
+      "grad_norm": 6.211682775156014,
+      "learning_rate": 1.221122112211221e-06,
+      "loss": -0.0155,
+      "step": 511
+    },
+    {
+      "epoch": 0.40951809638072384,
+      "grad_norm": 3.253438304578216,
+      "learning_rate": 1.2194719471947194e-06,
+      "loss": -0.0971,
+      "step": 512
+    },
+    {
+      "epoch": 0.41031793641271747,
+      "grad_norm": 3.562574820820311,
+      "learning_rate": 1.2178217821782177e-06,
+      "loss": -0.0982,
+      "step": 513
+    },
+    {
+      "epoch": 0.41111777644471104,
+      "grad_norm": 3.5746182911507067,
+      "learning_rate": 1.216171617161716e-06,
+      "loss": -0.023,
+      "step": 514
+    },
+    {
+      "epoch": 0.41191761647670466,
+      "grad_norm": 3.488381447372906,
+      "learning_rate": 1.2145214521452146e-06,
+      "loss": 0.1215,
+      "step": 515
+    },
+    {
+      "epoch": 0.41271745650869823,
+      "grad_norm": 3.1641673883077788,
+      "learning_rate": 1.2128712871287128e-06,
+      "loss": -0.0269,
+      "step": 516
+    },
+    {
+      "epoch": 0.41351729654069186,
+      "grad_norm": 4.13780306256476,
+      "learning_rate": 1.211221122112211e-06,
+      "loss": -0.08,
+      "step": 517
+    },
+    {
+      "epoch": 0.4143171365726855,
+      "grad_norm": 5.23630954806011,
+      "learning_rate": 1.2095709570957095e-06,
+      "loss": 0.0683,
+      "step": 518
+    },
+    {
+      "epoch": 0.41511697660467906,
+      "grad_norm": 5.08041039318514,
+      "learning_rate": 1.207920792079208e-06,
+      "loss": -0.0634,
+      "step": 519
+    },
+    {
+      "epoch": 0.4159168166366727,
+      "grad_norm": 4.328106931793288,
+      "learning_rate": 1.2062706270627063e-06,
+      "loss": -0.1251,
+      "step": 520
+    },
+    {
+      "epoch": 0.41671665666866625,
+      "grad_norm": 3.8685318893058978,
+      "learning_rate": 1.2046204620462046e-06,
+      "loss": -0.0397,
+      "step": 521
+    },
+    {
+      "epoch": 0.4175164967006599,
+      "grad_norm": 4.228873306898751,
+      "learning_rate": 1.2029702970297028e-06,
+      "loss": -0.0097,
+      "step": 522
+    },
+    {
+      "epoch": 0.41831633673265345,
+      "grad_norm": 8.545101114191558,
+      "learning_rate": 1.2013201320132014e-06,
+      "loss": 0.0257,
+      "step": 523
+    },
+    {
+      "epoch": 0.4191161767646471,
+      "grad_norm": 4.940963303084406,
+      "learning_rate": 1.1996699669966997e-06,
+      "loss": -0.0218,
+      "step": 524
+    },
+    {
+      "epoch": 0.41991601679664065,
+      "grad_norm": 4.486186548838174,
+      "learning_rate": 1.198019801980198e-06,
+      "loss": -0.1274,
+      "step": 525
+    },
+    {
+      "epoch": 0.42071585682863427,
+      "grad_norm": 6.117424152809813,
+      "learning_rate": 1.1963696369636963e-06,
+      "loss": -0.0412,
+      "step": 526
+    },
+    {
+      "epoch": 0.4215156968606279,
+      "grad_norm": 6.026088581435606,
+      "learning_rate": 1.1947194719471948e-06,
+      "loss": -0.1461,
+      "step": 527
+    },
+    {
+      "epoch": 0.42231553689262147,
+      "grad_norm": 4.391032244166686,
+      "learning_rate": 1.193069306930693e-06,
+      "loss": 0.0956,
+      "step": 528
+    },
+    {
+      "epoch": 0.4231153769246151,
+      "grad_norm": 5.139889742785653,
+      "learning_rate": 1.1914191419141915e-06,
+      "loss": -0.0998,
+      "step": 529
+    },
+    {
+      "epoch": 0.42391521695660866,
+      "grad_norm": 5.63979191849408,
+      "learning_rate": 1.1897689768976897e-06,
+      "loss": -0.0672,
+      "step": 530
+    },
+    {
+      "epoch": 0.4247150569886023,
+      "grad_norm": 8.323803093358931,
+      "learning_rate": 1.188118811881188e-06,
+      "loss": -0.1988,
+      "step": 531
+    },
+    {
+      "epoch": 0.42551489702059586,
+      "grad_norm": 3.5224991117629263,
+      "learning_rate": 1.1864686468646866e-06,
+      "loss": -0.0976,
+      "step": 532
+    },
+    {
+      "epoch": 0.4263147370525895,
+      "grad_norm": 3.3222865799787407,
+      "learning_rate": 1.1848184818481848e-06,
+      "loss": -0.1625,
+      "step": 533
+    },
+    {
+      "epoch": 0.42711457708458306,
+      "grad_norm": 4.473219337166838,
+      "learning_rate": 1.183168316831683e-06,
+      "loss": -0.102,
+      "step": 534
+    },
+    {
+      "epoch": 0.4279144171165767,
+      "grad_norm": 4.255445918061684,
+      "learning_rate": 1.1815181518151815e-06,
+      "loss": 0.0347,
+      "step": 535
+    },
+    {
+      "epoch": 0.4287142571485703,
+      "grad_norm": 5.273596279438336,
+      "learning_rate": 1.17986798679868e-06,
+      "loss": -0.0233,
+      "step": 536
+    },
+    {
+      "epoch": 0.4295140971805639,
+      "grad_norm": 3.2928329204452167,
+      "learning_rate": 1.1782178217821781e-06,
+      "loss": -0.1795,
+      "step": 537
+    },
+    {
+      "epoch": 0.4303139372125575,
+      "grad_norm": 3.1943187707330676,
+      "learning_rate": 1.1765676567656766e-06,
+      "loss": -0.1193,
+      "step": 538
+    },
+    {
+      "epoch": 0.4311137772445511,
+      "grad_norm": 4.457107636902936,
+      "learning_rate": 1.1749174917491748e-06,
+      "loss": -0.0256,
+      "step": 539
+    },
+    {
+      "epoch": 0.4319136172765447,
+      "grad_norm": 4.508728040150466,
+      "learning_rate": 1.1732673267326732e-06,
+      "loss": -0.1272,
+      "step": 540
+    },
+    {
+      "epoch": 0.4327134573085383,
+      "grad_norm": 4.679176366336832,
+      "learning_rate": 1.1716171617161717e-06,
+      "loss": -0.0107,
+      "step": 541
+    },
+    {
+      "epoch": 0.4335132973405319,
+      "grad_norm": 4.572704243632147,
+      "learning_rate": 1.16996699669967e-06,
+      "loss": -0.0189,
+      "step": 542
+    },
+    {
+      "epoch": 0.43431313737252547,
+      "grad_norm": 3.823996049360206,
+      "learning_rate": 1.1683168316831681e-06,
+      "loss": 0.0071,
+      "step": 543
+    },
+    {
+      "epoch": 0.4351129774045191,
+      "grad_norm": 4.448417665137879,
+      "learning_rate": 1.1666666666666668e-06,
+      "loss": 0.0018,
+      "step": 544
+    },
+    {
+      "epoch": 0.4359128174365127,
+      "grad_norm": 3.7067784825161625,
+      "learning_rate": 1.165016501650165e-06,
+      "loss": -0.0643,
+      "step": 545
+    },
+    {
+      "epoch": 0.4367126574685063,
+      "grad_norm": 4.304960211061566,
+      "learning_rate": 1.1633663366336632e-06,
+      "loss": 0.0434,
+      "step": 546
+    },
+    {
+      "epoch": 0.4375124975004999,
+      "grad_norm": 4.809624741919171,
+      "learning_rate": 1.1617161716171617e-06,
+      "loss": -0.1175,
+      "step": 547
+    },
+    {
+      "epoch": 0.4383123375324935,
+      "grad_norm": 3.572402442577118,
+      "learning_rate": 1.16006600660066e-06,
+      "loss": 0.0641,
+      "step": 548
+    },
+    {
+      "epoch": 0.4391121775644871,
+      "grad_norm": 3.1323439706728173,
+      "learning_rate": 1.1584158415841584e-06,
+      "loss": -0.1315,
+      "step": 549
+    },
+    {
+      "epoch": 0.4399120175964807,
+      "grad_norm": 6.63310206919076,
+      "learning_rate": 1.1567656765676568e-06,
+      "loss": -0.1572,
+      "step": 550
+    },
+    {
+      "epoch": 0.4407118576284743,
+      "grad_norm": 5.7194336862922475,
+      "learning_rate": 1.155115511551155e-06,
+      "loss": -0.0498,
+      "step": 551
+    },
+    {
+      "epoch": 0.4415116976604679,
+      "grad_norm": 5.0355458371512976,
+      "learning_rate": 1.1534653465346533e-06,
+      "loss": -0.0343,
+      "step": 552
+    },
+    {
+      "epoch": 0.4423115376924615,
+      "grad_norm": 4.479813545297925,
+      "learning_rate": 1.151815181518152e-06,
+      "loss": -0.052,
+      "step": 553
+    },
+    {
+      "epoch": 0.44311137772445514,
+      "grad_norm": 5.515043665694904,
+      "learning_rate": 1.1501650165016501e-06,
+      "loss": -0.0962,
+      "step": 554
+    },
+    {
+      "epoch": 0.4439112177564487,
+      "grad_norm": 4.35349503345848,
+      "learning_rate": 1.1485148514851484e-06,
+      "loss": -0.1718,
+      "step": 555
+    },
+    {
+      "epoch": 0.44471105778844233,
+      "grad_norm": 4.324313582265245,
+      "learning_rate": 1.1468646864686468e-06,
+      "loss": -0.1059,
+      "step": 556
+    },
+    {
+      "epoch": 0.4455108978204359,
+      "grad_norm": 6.603357917371442,
+      "learning_rate": 1.1452145214521452e-06,
+      "loss": -0.0179,
+      "step": 557
+    },
+    {
+      "epoch": 0.44631073785242953,
+      "grad_norm": 4.731847308612818,
+      "learning_rate": 1.1435643564356435e-06,
+      "loss": -0.0285,
+      "step": 558
+    },
+    {
+      "epoch": 0.4471105778844231,
+      "grad_norm": 3.2387655192879,
+      "learning_rate": 1.141914191419142e-06,
+      "loss": 0.0157,
+      "step": 559
+    },
+    {
+      "epoch": 0.4479104179164167,
+      "grad_norm": 4.11046977381839,
+      "learning_rate": 1.1402640264026401e-06,
+      "loss": -0.0738,
+      "step": 560
+    },
+    {
+      "epoch": 0.4487102579484103,
+      "grad_norm": 4.7767170879491765,
+      "learning_rate": 1.1386138613861384e-06,
+      "loss": -0.1826,
+      "step": 561
+    },
+    {
+      "epoch": 0.4495100979804039,
+      "grad_norm": 4.712736226037487,
+      "learning_rate": 1.136963696369637e-06,
+      "loss": -0.0453,
+      "step": 562
+    },
+    {
+      "epoch": 0.45030993801239755,
+      "grad_norm": 5.2823586287855795,
+      "learning_rate": 1.1353135313531353e-06,
+      "loss": -0.0112,
+      "step": 563
+    },
+    {
+      "epoch": 0.4511097780443911,
+      "grad_norm": 4.313375854007458,
+      "learning_rate": 1.1336633663366335e-06,
+      "loss": -0.1529,
+      "step": 564
+    },
+    {
+      "epoch": 0.45190961807638474,
+      "grad_norm": 3.661113275988152,
+      "learning_rate": 1.132013201320132e-06,
+      "loss": 0.0182,
+      "step": 565
+    },
+    {
+      "epoch": 0.4527094581083783,
+      "grad_norm": 3.3641831166871015,
+      "learning_rate": 1.1303630363036304e-06,
+      "loss": -0.0967,
+      "step": 566
+    },
+    {
+      "epoch": 0.45350929814037194,
+      "grad_norm": 3.5187461731044634,
+      "learning_rate": 1.1287128712871286e-06,
+      "loss": -0.0115,
+      "step": 567
+    },
+    {
+      "epoch": 0.4543091381723655,
+      "grad_norm": 5.055238201441692,
+      "learning_rate": 1.127062706270627e-06,
+      "loss": -0.0701,
+      "step": 568
+    },
+    {
+      "epoch": 0.45510897820435914,
+      "grad_norm": 5.048998878882335,
+      "learning_rate": 1.1254125412541253e-06,
+      "loss": -0.0982,
+      "step": 569
+    },
+    {
+      "epoch": 0.4559088182363527,
+      "grad_norm": 6.706995417966311,
+      "learning_rate": 1.123762376237624e-06,
+      "loss": -0.0323,
+      "step": 570
+    },
+    {
+      "epoch": 0.45670865826834633,
+      "grad_norm": 5.717361031375047,
+      "learning_rate": 1.1221122112211221e-06,
+      "loss": -0.0251,
+      "step": 571
+    },
+    {
+      "epoch": 0.4575084983003399,
+      "grad_norm": 5.073568794377317,
+      "learning_rate": 1.1204620462046204e-06,
+      "loss": -0.2165,
+      "step": 572
+    },
+    {
+      "epoch": 0.45830833833233353,
+      "grad_norm": 4.340499430104141,
+      "learning_rate": 1.1188118811881188e-06,
+      "loss": -0.1008,
+      "step": 573
+    },
+    {
+      "epoch": 0.45910817836432716,
+      "grad_norm": 11.188522206922801,
+      "learning_rate": 1.117161716171617e-06,
+      "loss": -0.0362,
+      "step": 574
+    },
+    {
+      "epoch": 0.45990801839632073,
+      "grad_norm": 8.96889120914533,
+      "learning_rate": 1.1155115511551155e-06,
+      "loss": 0.0492,
+      "step": 575
+    },
+    {
+      "epoch": 0.46070785842831435,
+      "grad_norm": 4.7623433805729825,
+      "learning_rate": 1.113861386138614e-06,
+      "loss": -0.0462,
+      "step": 576
+    },
+    {
+      "epoch": 0.4615076984603079,
+      "grad_norm": 4.809552169223393,
+      "learning_rate": 1.1122112211221121e-06,
+      "loss": -0.0298,
+      "step": 577
+    },
+    {
+      "epoch": 0.46230753849230155,
+      "grad_norm": 4.199314588295882,
+      "learning_rate": 1.1105610561056104e-06,
+      "loss": -0.1501,
+      "step": 578
+    },
+    {
+      "epoch": 0.4631073785242951,
+      "grad_norm": 3.531450924178731,
+      "learning_rate": 1.108910891089109e-06,
+      "loss": -0.0589,
+      "step": 579
+    },
+    {
+      "epoch": 0.46390721855628875,
+      "grad_norm": 4.247443159328166,
+      "learning_rate": 1.1072607260726073e-06,
+      "loss": -0.0362,
+      "step": 580
+    },
+    {
+      "epoch": 0.4647070585882823,
+      "grad_norm": 3.4845953217374346,
+      "learning_rate": 1.1056105610561055e-06,
+      "loss": -0.0171,
+      "step": 581
+    },
+    {
+      "epoch": 0.46550689862027594,
+      "grad_norm": 6.269891746353069,
+      "learning_rate": 1.103960396039604e-06,
+      "loss": -0.0552,
+      "step": 582
+    },
+    {
+      "epoch": 0.46630673865226957,
+      "grad_norm": 4.110925186237378,
+      "learning_rate": 1.1023102310231024e-06,
+      "loss": -0.1525,
+      "step": 583
+    },
+    {
+      "epoch": 0.46710657868426314,
+      "grad_norm": 4.027706037445169,
+      "learning_rate": 1.1006600660066006e-06,
+      "loss": -0.1283,
+      "step": 584
+    },
+    {
+      "epoch": 0.46790641871625677,
+      "grad_norm": 6.085436348609565,
+      "learning_rate": 1.099009900990099e-06,
+      "loss": 0.0152,
+      "step": 585
+    },
+    {
+      "epoch": 0.46870625874825034,
+      "grad_norm": 4.403688809188704,
+      "learning_rate": 1.0973597359735973e-06,
+      "loss": -0.0462,
+      "step": 586
+    },
+    {
+      "epoch": 0.46950609878024396,
+      "grad_norm": 3.7204758942669924,
+      "learning_rate": 1.0957095709570955e-06,
+      "loss": -0.0472,
+      "step": 587
+    },
+    {
+      "epoch": 0.47030593881223753,
+      "grad_norm": 4.16282077045935,
+      "learning_rate": 1.0940594059405941e-06,
+      "loss": -0.1098,
+      "step": 588
+    },
+    {
+      "epoch": 0.47110577884423116,
+      "grad_norm": 4.490618781800236,
+      "learning_rate": 1.0924092409240924e-06,
+      "loss": -0.0323,
+      "step": 589
+    },
+    {
+      "epoch": 0.47190561887622473,
+      "grad_norm": 3.83450769320751,
+      "learning_rate": 1.0907590759075906e-06,
+      "loss": -0.0482,
+      "step": 590
+    },
+    {
+      "epoch": 0.47270545890821836,
+      "grad_norm": 3.951497100856045,
+      "learning_rate": 1.089108910891089e-06,
+      "loss": -0.1206,
+      "step": 591
+    },
+    {
+      "epoch": 0.473505298940212,
+      "grad_norm": 4.798754196622245,
+      "learning_rate": 1.0874587458745875e-06,
+      "loss": 0.0584,
+      "step": 592
+    },
+    {
+      "epoch": 0.47430513897220555,
+      "grad_norm": 4.437199971976538,
+      "learning_rate": 1.0858085808580857e-06,
+      "loss": -0.0666,
+      "step": 593
+    },
+    {
+      "epoch": 0.4751049790041992,
+      "grad_norm": 5.877945506525689,
+      "learning_rate": 1.0841584158415842e-06,
+      "loss": -0.0795,
+      "step": 594
+    },
+    {
+      "epoch": 0.47590481903619275,
+      "grad_norm": 4.9827179740392165,
+      "learning_rate": 1.0825082508250824e-06,
+      "loss": -0.145,
+      "step": 595
+    },
+    {
+      "epoch": 0.4767046590681864,
+      "grad_norm": 4.579893011650921,
+      "learning_rate": 1.0808580858085808e-06,
+      "loss": 0.0438,
+      "step": 596
+    },
+    {
+      "epoch": 0.47750449910017995,
+      "grad_norm": 4.6741113317873975,
+      "learning_rate": 1.0792079207920793e-06,
+      "loss": -0.0471,
+      "step": 597
+    },
+    {
+      "epoch": 0.4783043391321736,
+      "grad_norm": 4.45102253856279,
+      "learning_rate": 1.0775577557755775e-06,
+      "loss": -0.0705,
+      "step": 598
+    },
+    {
+      "epoch": 0.47910417916416714,
+      "grad_norm": 7.735572429403455,
+      "learning_rate": 1.0759075907590757e-06,
+      "loss": -0.0004,
+      "step": 599
+    },
+    {
+      "epoch": 0.47990401919616077,
+      "grad_norm": 4.3890790831054645,
+      "learning_rate": 1.0742574257425744e-06,
+      "loss": -0.11,
+      "step": 600
+    },
+    {
+      "epoch": 0.4807038592281544,
+      "grad_norm": 3.750715441802859,
+      "learning_rate": 1.0726072607260726e-06,
+      "loss": -0.0943,
+      "step": 601
+    },
+    {
+      "epoch": 0.48150369926014797,
+      "grad_norm": 3.65650652124133,
+      "learning_rate": 1.0709570957095708e-06,
+      "loss": -0.169,
+      "step": 602
+    },
+    {
+      "epoch": 0.4823035392921416,
+      "grad_norm": 4.55201247427091,
+      "learning_rate": 1.0693069306930693e-06,
+      "loss": -0.0753,
+      "step": 603
+    },
+    {
+      "epoch": 0.48310337932413516,
+      "grad_norm": 4.55776208974576,
+      "learning_rate": 1.0676567656765675e-06,
+      "loss": -0.1347,
+      "step": 604
+    },
+    {
+      "epoch": 0.4839032193561288,
+      "grad_norm": 4.48864117831728,
+      "learning_rate": 1.066006600660066e-06,
+      "loss": -0.0974,
+      "step": 605
+    },
+    {
+      "epoch": 0.48470305938812236,
+      "grad_norm": 3.9363005491507668,
+      "learning_rate": 1.0643564356435644e-06,
+      "loss": -0.0079,
+      "step": 606
+    },
+    {
+      "epoch": 0.485502899420116,
+      "grad_norm": 4.252862078789536,
+      "learning_rate": 1.0627062706270626e-06,
+      "loss": -0.0136,
+      "step": 607
+    },
+    {
+      "epoch": 0.48630273945210956,
+      "grad_norm": 4.543019341197776,
+      "learning_rate": 1.0610561056105608e-06,
+      "loss": -0.1351,
+      "step": 608
+    },
+    {
+      "epoch": 0.4871025794841032,
+      "grad_norm": 5.655845163937271,
+      "learning_rate": 1.0594059405940595e-06,
+      "loss": -0.0266,
+      "step": 609
+    },
+    {
+      "epoch": 0.4879024195160968,
+      "grad_norm": 11.681523554473147,
+      "learning_rate": 1.0577557755775577e-06,
+      "loss": -0.1126,
+      "step": 610
+    },
+    {
+      "epoch": 0.4887022595480904,
+      "grad_norm": 4.307554643653013,
+      "learning_rate": 1.056105610561056e-06,
+      "loss": -0.0647,
+      "step": 611
+    },
+    {
+      "epoch": 0.489502099580084,
+      "grad_norm": 5.732899988046993,
+      "learning_rate": 1.0544554455445544e-06,
+      "loss": -0.0934,
+      "step": 612
+    },
+    {
+      "epoch": 0.4903019396120776,
+      "grad_norm": 4.150445541916088,
+      "learning_rate": 1.0528052805280528e-06,
+      "loss": -0.1406,
+      "step": 613
+    },
+    {
+      "epoch": 0.4911017796440712,
+      "grad_norm": 5.1082669144378725,
+      "learning_rate": 1.051155115511551e-06,
+      "loss": 0.0558,
+      "step": 614
+    },
+    {
+      "epoch": 0.49190161967606477,
+      "grad_norm": 7.67678160227598,
+      "learning_rate": 1.0495049504950495e-06,
+      "loss": 0.0088,
+      "step": 615
+    },
+    {
+      "epoch": 0.4927014597080584,
+      "grad_norm": 3.354349182723264,
+      "learning_rate": 1.0478547854785477e-06,
+      "loss": -0.1359,
+      "step": 616
+    },
+    {
+      "epoch": 0.49350129974005197,
+      "grad_norm": 4.725187848913473,
+      "learning_rate": 1.046204620462046e-06,
+      "loss": -0.0762,
+      "step": 617
+    },
+    {
+      "epoch": 0.4943011397720456,
+      "grad_norm": 5.7805399690672825,
+      "learning_rate": 1.0445544554455446e-06,
+      "loss": 0.0578,
+      "step": 618
+    },
+    {
+      "epoch": 0.49510097980403917,
+      "grad_norm": 4.179300735975294,
+      "learning_rate": 1.0429042904290428e-06,
+      "loss": 0.005,
+      "step": 619
+    },
+    {
+      "epoch": 0.4959008198360328,
+      "grad_norm": 5.098901682751034,
+      "learning_rate": 1.0412541254125413e-06,
+      "loss": -0.0156,
+      "step": 620
+    },
+    {
+      "epoch": 0.4967006598680264,
+      "grad_norm": 8.248543813099444,
+      "learning_rate": 1.0396039603960395e-06,
+      "loss": -0.0457,
+      "step": 621
+    },
+    {
+      "epoch": 0.49750049990002,
+      "grad_norm": 4.864094199749138,
+      "learning_rate": 1.037953795379538e-06,
+      "loss": 0.0829,
+      "step": 622
+    },
+    {
+      "epoch": 0.4983003399320136,
+      "grad_norm": 4.692602610129244,
+      "learning_rate": 1.0363036303630364e-06,
+      "loss": -0.0396,
+      "step": 623
+    },
+    {
+      "epoch": 0.4991001799640072,
+      "grad_norm": 3.696934414641692,
+      "learning_rate": 1.0346534653465346e-06,
+      "loss": -0.1939,
+      "step": 624
+    },
+    {
+      "epoch": 0.4999000199960008,
+      "grad_norm": 5.491055932283412,
+      "learning_rate": 1.0330033003300328e-06,
+      "loss": -0.0639,
+      "step": 625
+    },
+    {
+      "epoch": 0.5006998600279944,
+      "grad_norm": 3.92108657024522,
+      "learning_rate": 1.0313531353135315e-06,
+      "loss": -0.0969,
+      "step": 626
+    },
+    {
+      "epoch": 0.501499700059988,
+      "grad_norm": 5.243975134143899,
+      "learning_rate": 1.0297029702970297e-06,
+      "loss": -0.0206,
+      "step": 627
+    },
+    {
+      "epoch": 0.5022995400919816,
+      "grad_norm": 4.328763949057532,
+      "learning_rate": 1.028052805280528e-06,
+      "loss": -0.1081,
+      "step": 628
+    },
+    {
+      "epoch": 0.5030993801239753,
+      "grad_norm": 4.197218364380273,
+      "learning_rate": 1.0264026402640264e-06,
+      "loss": -0.0932,
+      "step": 629
+    },
+    {
+      "epoch": 0.5038992201559688,
+      "grad_norm": 4.300285400855226,
+      "learning_rate": 1.0247524752475248e-06,
+      "loss": -0.1016,
+      "step": 630
+    },
+    {
+      "epoch": 0.5046990601879624,
+      "grad_norm": 4.4977354073528675,
+      "learning_rate": 1.023102310231023e-06,
+      "loss": -0.0846,
+      "step": 631
+    },
+    {
+      "epoch": 0.505498900219956,
+      "grad_norm": 4.303063503427107,
+      "learning_rate": 1.0214521452145215e-06,
+      "loss": -0.0724,
+      "step": 632
+    },
+    {
+      "epoch": 0.5062987402519497,
+      "grad_norm": 6.073276027125735,
+      "learning_rate": 1.0198019801980197e-06,
+      "loss": -0.0326,
+      "step": 633
+    },
+    {
+      "epoch": 0.5070985802839432,
+      "grad_norm": 4.973392453051079,
+      "learning_rate": 1.018151815181518e-06,
+      "loss": 0.0189,
+      "step": 634
+    },
+    {
+      "epoch": 0.5078984203159368,
+      "grad_norm": 4.811288854611089,
+      "learning_rate": 1.0165016501650166e-06,
+      "loss": -0.1415,
+      "step": 635
+    },
+    {
+      "epoch": 0.5086982603479304,
+      "grad_norm": 4.853499316182358,
+      "learning_rate": 1.0148514851485148e-06,
+      "loss": -0.0896,
+      "step": 636
+    },
+    {
+      "epoch": 0.509498100379924,
+      "grad_norm": 3.839359784599905,
+      "learning_rate": 1.013201320132013e-06,
+      "loss": -0.0358,
+      "step": 637
+    },
+    {
+      "epoch": 0.5102979404119176,
+      "grad_norm": 4.783165043297262,
+      "learning_rate": 1.0115511551155115e-06,
+      "loss": -0.0102,
+      "step": 638
+    },
+    {
+      "epoch": 0.5110977804439112,
+      "grad_norm": 5.184943614862812,
+      "learning_rate": 1.00990099009901e-06,
+      "loss": -0.1354,
+      "step": 639
+    },
+    {
+      "epoch": 0.5118976204759048,
+      "grad_norm": 39.49636862757923,
+      "learning_rate": 1.0082508250825082e-06,
+      "loss": 0.0023,
+      "step": 640
+    },
+    {
+      "epoch": 0.5126974605078984,
+      "grad_norm": 4.083973866781674,
+      "learning_rate": 1.0066006600660066e-06,
+      "loss": -0.0357,
+      "step": 641
+    },
+    {
+      "epoch": 0.5134973005398921,
+      "grad_norm": 5.817761080874723,
+      "learning_rate": 1.0049504950495048e-06,
+      "loss": -0.0441,
+      "step": 642
+    },
+    {
+      "epoch": 0.5142971405718856,
+      "grad_norm": 8.811705641420119,
+      "learning_rate": 1.0033003300330033e-06,
+      "loss": 0.1013,
+      "step": 643
+    },
+    {
+      "epoch": 0.5150969806038792,
+      "grad_norm": 4.211618456869653,
+      "learning_rate": 1.0016501650165017e-06,
+      "loss": -0.1781,
+      "step": 644
+    },
+    {
+      "epoch": 0.5158968206358728,
+      "grad_norm": 6.430696021299668,
+      "learning_rate": 1e-06,
+      "loss": -0.0475,
+      "step": 645
+    },
+    {
+      "epoch": 0.5166966606678665,
+      "grad_norm": 4.287574273625528,
+      "learning_rate": 9.983498349834984e-07,
+      "loss": -0.0304,
+      "step": 646
+    },
+    {
+      "epoch": 0.51749650069986,
+      "grad_norm": 2.916483591782696,
+      "learning_rate": 9.966996699669966e-07,
+      "loss": -0.1574,
+      "step": 647
+    },
+    {
+      "epoch": 0.5182963407318536,
+      "grad_norm": 5.071406876730617,
+      "learning_rate": 9.95049504950495e-07,
+      "loss": -0.0227,
+      "step": 648
+    },
+    {
+      "epoch": 0.5190961807638472,
+      "grad_norm": 3.863107212570657,
+      "learning_rate": 9.933993399339933e-07,
+      "loss": -0.0452,
+      "step": 649
+    },
+    {
+      "epoch": 0.5198960207958409,
+      "grad_norm": 4.234640528387632,
+      "learning_rate": 9.917491749174917e-07,
+      "loss": -0.0816,
+      "step": 650
+    },
+    {
+      "epoch": 0.5206958608278345,
+      "grad_norm": 4.255603238118902,
+      "learning_rate": 9.9009900990099e-07,
+      "loss": -0.1311,
+      "step": 651
+    },
+    {
+      "epoch": 0.521495700859828,
+      "grad_norm": 4.723143092604518,
+      "learning_rate": 9.884488448844884e-07,
+      "loss": -0.1076,
+      "step": 652
+    },
+    {
+      "epoch": 0.5222955408918216,
+      "grad_norm": 4.609676855516043,
+      "learning_rate": 9.867986798679866e-07,
+      "loss": -0.0468,
+      "step": 653
+    },
+    {
+      "epoch": 0.5230953809238152,
+      "grad_norm": 4.648497611546731,
+      "learning_rate": 9.85148514851485e-07,
+      "loss": 0.066,
+      "step": 654
+    },
+    {
+      "epoch": 0.5238952209558089,
+      "grad_norm": 4.444890919088204,
+      "learning_rate": 9.834983498349835e-07,
+      "loss": -0.0954,
+      "step": 655
+    },
+    {
+      "epoch": 0.5246950609878024,
+      "grad_norm": 3.789960680030435,
+      "learning_rate": 9.818481848184817e-07,
+      "loss": -0.0846,
+      "step": 656
+    },
+    {
+      "epoch": 0.525494901019796,
+      "grad_norm": 7.20767352956141,
+      "learning_rate": 9.801980198019802e-07,
+      "loss": -0.0689,
+      "step": 657
+    },
+    {
+      "epoch": 0.5262947410517896,
+      "grad_norm": 3.8251645221108883,
+      "learning_rate": 9.785478547854786e-07,
+      "loss": -0.0026,
+      "step": 658
+    },
+    {
+      "epoch": 0.5270945810837833,
+      "grad_norm": 10.44191187281191,
+      "learning_rate": 9.768976897689768e-07,
+      "loss": -0.1835,
+      "step": 659
+    },
+    {
+      "epoch": 0.5278944211157769,
+      "grad_norm": 6.355832474610427,
+      "learning_rate": 9.75247524752475e-07,
+      "loss": -0.0168,
+      "step": 660
+    },
+    {
+      "epoch": 0.5286942611477704,
+      "grad_norm": 3.864986648832606,
+      "learning_rate": 9.735973597359735e-07,
+      "loss": -0.0699,
+      "step": 661
+    },
+    {
+      "epoch": 0.529494101179764,
+      "grad_norm": 9.07974895718424,
+      "learning_rate": 9.71947194719472e-07,
+      "loss": -0.0013,
+      "step": 662
+    },
+    {
+      "epoch": 0.5302939412117577,
+      "grad_norm": 10.325195266318097,
+      "learning_rate": 9.702970297029702e-07,
+      "loss": -0.1641,
+      "step": 663
+    },
+    {
+      "epoch": 0.5310937812437513,
+      "grad_norm": 4.441568363287549,
+      "learning_rate": 9.686468646864686e-07,
+      "loss": 0.0397,
+      "step": 664
+    },
+    {
+      "epoch": 0.5318936212757448,
+      "grad_norm": 5.1229672005813605,
+      "learning_rate": 9.66996699669967e-07,
+      "loss": -0.1918,
+      "step": 665
+    },
+    {
+      "epoch": 0.5326934613077384,
+      "grad_norm": 4.721345907443594,
+      "learning_rate": 9.653465346534653e-07,
+      "loss": -0.0215,
+      "step": 666
+    },
+    {
+      "epoch": 0.533493301339732,
+      "grad_norm": 7.118557074848867,
+      "learning_rate": 9.636963696369637e-07,
+      "loss": -0.0039,
+      "step": 667
+    },
+    {
+      "epoch": 0.5342931413717257,
+      "grad_norm": 4.7832103026691755,
+      "learning_rate": 9.62046204620462e-07,
+      "loss": -0.0221,
+      "step": 668
+    },
+    {
+      "epoch": 0.5350929814037193,
+      "grad_norm": 6.3240275362924505,
+      "learning_rate": 9.603960396039604e-07,
+      "loss": -0.0409,
+      "step": 669
+    },
+    {
+      "epoch": 0.5358928214357128,
+      "grad_norm": 4.027979744126303,
+      "learning_rate": 9.587458745874586e-07,
+      "loss": -0.0503,
+      "step": 670
+    },
+    {
+      "epoch": 0.5366926614677064,
+      "grad_norm": 4.74302864470449,
+      "learning_rate": 9.57095709570957e-07,
+      "loss": -0.0521,
+      "step": 671
+    },
+    {
+      "epoch": 0.5374925014997001,
+      "grad_norm": 4.985380734947703,
+      "learning_rate": 9.554455445544553e-07,
+      "loss": -0.082,
+      "step": 672
+    },
+    {
+      "epoch": 0.5382923415316937,
+      "grad_norm": 4.472849852523853,
+      "learning_rate": 9.537953795379537e-07,
+      "loss": 0.1016,
+      "step": 673
+    },
+    {
+      "epoch": 0.5390921815636872,
+      "grad_norm": 3.308647938974776,
+      "learning_rate": 9.521452145214522e-07,
+      "loss": -0.0457,
+      "step": 674
+    },
+    {
+      "epoch": 0.5398920215956808,
+      "grad_norm": 4.677054809392709,
+      "learning_rate": 9.504950495049504e-07,
+      "loss": -0.1658,
+      "step": 675
+    },
+    {
+      "epoch": 0.5406918616276745,
+      "grad_norm": 4.067875284374342,
+      "learning_rate": 9.488448844884487e-07,
+      "loss": 0.0196,
+      "step": 676
+    },
+    {
+      "epoch": 0.5414917016596681,
+      "grad_norm": 3.730643734647644,
+      "learning_rate": 9.471947194719472e-07,
+      "loss": -0.0894,
+      "step": 677
+    },
+    {
+      "epoch": 0.5422915416916617,
+      "grad_norm": 5.204622803431674,
+      "learning_rate": 9.455445544554454e-07,
+      "loss": -0.0538,
+      "step": 678
+    },
+    {
+      "epoch": 0.5430913817236552,
+      "grad_norm": 5.261361679954622,
+      "learning_rate": 9.438943894389439e-07,
+      "loss": -0.026,
+      "step": 679
+    },
+    {
+      "epoch": 0.5438912217556489,
+      "grad_norm": 4.057248991454938,
+      "learning_rate": 9.422442244224422e-07,
+      "loss": -0.0785,
+      "step": 680
+    },
+    {
+      "epoch": 0.5446910617876425,
+      "grad_norm": 5.580346613410825,
+      "learning_rate": 9.405940594059405e-07,
+      "loss": -0.0335,
+      "step": 681
+    },
+    {
+      "epoch": 0.5454909018196361,
+      "grad_norm": 3.287439521262259,
+      "learning_rate": 9.389438943894389e-07,
+      "loss": -0.014,
+      "step": 682
+    },
+    {
+      "epoch": 0.5462907418516296,
+      "grad_norm": 4.383770462305995,
+      "learning_rate": 9.372937293729373e-07,
+      "loss": -0.1349,
+      "step": 683
+    },
+    {
+      "epoch": 0.5470905818836233,
+      "grad_norm": 3.570830393151609,
+      "learning_rate": 9.356435643564356e-07,
+      "loss": -0.1603,
+      "step": 684
+    },
+    {
+      "epoch": 0.5478904219156169,
+      "grad_norm": 3.9301633755259076,
+      "learning_rate": 9.33993399339934e-07,
+      "loss": -0.0728,
+      "step": 685
+    },
+    {
+      "epoch": 0.5486902619476105,
+      "grad_norm": 4.577722525237458,
+      "learning_rate": 9.323432343234323e-07,
+      "loss": -0.0686,
+      "step": 686
+    },
+    {
+      "epoch": 0.5494901019796041,
+      "grad_norm": 3.4290177162671704,
+      "learning_rate": 9.306930693069307e-07,
+      "loss": -0.0002,
+      "step": 687
+    },
+    {
+      "epoch": 0.5502899420115976,
+      "grad_norm": 3.7664714194362574,
+      "learning_rate": 9.29042904290429e-07,
+      "loss": 0.0181,
+      "step": 688
+    },
+    {
+      "epoch": 0.5510897820435913,
+      "grad_norm": 3.53486758576286,
+      "learning_rate": 9.273927392739273e-07,
+      "loss": 0.0046,
+      "step": 689
+    },
+    {
+      "epoch": 0.5518896220755849,
+      "grad_norm": 5.533353761076084,
+      "learning_rate": 9.257425742574257e-07,
+      "loss": -0.1142,
+      "step": 690
+    },
+    {
+      "epoch": 0.5526894621075785,
+      "grad_norm": 3.618813464410366,
+      "learning_rate": 9.24092409240924e-07,
+      "loss": -0.0533,
+      "step": 691
+    },
+    {
+      "epoch": 0.553489302139572,
+      "grad_norm": 3.2399813948203064,
+      "learning_rate": 9.224422442244224e-07,
+      "loss": -0.2655,
+      "step": 692
+    },
+    {
+      "epoch": 0.5542891421715657,
+      "grad_norm": 4.777938394039966,
+      "learning_rate": 9.207920792079208e-07,
+      "loss": -0.1287,
+      "step": 693
+    },
+    {
+      "epoch": 0.5550889822035593,
+      "grad_norm": 4.753607333838816,
+      "learning_rate": 9.191419141914191e-07,
+      "loss": -0.0197,
+      "step": 694
+    },
+    {
+      "epoch": 0.5558888222355529,
+      "grad_norm": 4.574962995202333,
+      "learning_rate": 9.174917491749174e-07,
+      "loss": -0.0882,
+      "step": 695
+    },
+    {
+      "epoch": 0.5566886622675465,
+      "grad_norm": 5.716789868568477,
+      "learning_rate": 9.158415841584159e-07,
+      "loss": -0.0978,
+      "step": 696
+    },
+    {
+      "epoch": 0.5574885022995401,
+      "grad_norm": 5.1621141783698805,
+      "learning_rate": 9.141914191419141e-07,
+      "loss": -0.0492,
+      "step": 697
+    },
+    {
+      "epoch": 0.5582883423315337,
+      "grad_norm": 5.066721910041668,
+      "learning_rate": 9.125412541254125e-07,
+      "loss": -0.1249,
+      "step": 698
+    },
+    {
+      "epoch": 0.5590881823635273,
+      "grad_norm": 6.6614345364199,
+      "learning_rate": 9.108910891089109e-07,
+      "loss": -0.0784,
+      "step": 699
+    },
+    {
+      "epoch": 0.5598880223955209,
+      "grad_norm": 7.651991800546116,
+      "learning_rate": 9.092409240924092e-07,
+      "loss": 0.0583,
+      "step": 700
+    },
+    {
+      "epoch": 0.5606878624275144,
+      "grad_norm": 5.556872832637128,
+      "learning_rate": 9.075907590759075e-07,
+      "loss": -0.0337,
+      "step": 701
+    },
+    {
+      "epoch": 0.5614877024595081,
+      "grad_norm": 10.302249877529787,
+      "learning_rate": 9.05940594059406e-07,
+      "loss": -0.0318,
+      "step": 702
+    },
+    {
+      "epoch": 0.5622875424915017,
+      "grad_norm": 6.446831319626775,
+      "learning_rate": 9.042904290429042e-07,
+      "loss": -0.032,
+      "step": 703
+    },
+    {
+      "epoch": 0.5630873825234953,
+      "grad_norm": 3.1431425262284156,
+      "learning_rate": 9.026402640264025e-07,
+      "loss": -0.1257,
+      "step": 704
+    },
+    {
+      "epoch": 0.563887222555489,
+      "grad_norm": 6.8613589813033755,
+      "learning_rate": 9.00990099009901e-07,
+      "loss": -0.0899,
+      "step": 705
+    },
+    {
+      "epoch": 0.5646870625874825,
+      "grad_norm": 4.727075571003651,
+      "learning_rate": 8.993399339933992e-07,
+      "loss": -0.0103,
+      "step": 706
+    },
+    {
+      "epoch": 0.5654869026194761,
+      "grad_norm": 3.757033554841126,
+      "learning_rate": 8.976897689768976e-07,
+      "loss": -0.1201,
+      "step": 707
+    },
+    {
+      "epoch": 0.5662867426514697,
+      "grad_norm": 3.6568326567325586,
+      "learning_rate": 8.96039603960396e-07,
+      "loss": -0.0442,
+      "step": 708
+    },
+    {
+      "epoch": 0.5670865826834633,
+      "grad_norm": 4.422118271245446,
+      "learning_rate": 8.943894389438944e-07,
+      "loss": -0.1491,
+      "step": 709
+    },
+    {
+      "epoch": 0.5678864227154569,
+      "grad_norm": 4.090528079313399,
+      "learning_rate": 8.927392739273927e-07,
+      "loss": -0.1213,
+      "step": 710
+    },
+    {
+      "epoch": 0.5686862627474505,
+      "grad_norm": 6.203118235394611,
+      "learning_rate": 8.910891089108911e-07,
+      "loss": -0.0415,
+      "step": 711
+    },
+    {
+      "epoch": 0.5694861027794441,
+      "grad_norm": 4.192997526379617,
+      "learning_rate": 8.894389438943894e-07,
+      "loss": -0.0378,
+      "step": 712
+    },
+    {
+      "epoch": 0.5702859428114377,
+      "grad_norm": 4.532567536428998,
+      "learning_rate": 8.877887788778878e-07,
+      "loss": -0.0275,
+      "step": 713
+    },
+    {
+      "epoch": 0.5710857828434314,
+      "grad_norm": 4.168577112275988,
+      "learning_rate": 8.861386138613861e-07,
+      "loss": -0.0494,
+      "step": 714
+    },
+    {
+      "epoch": 0.5718856228754249,
+      "grad_norm": 7.436541583714728,
+      "learning_rate": 8.844884488448845e-07,
+      "loss": 0.0338,
+      "step": 715
+    },
+    {
+      "epoch": 0.5726854629074185,
+      "grad_norm": 4.3341221213745555,
+      "learning_rate": 8.828382838283828e-07,
+      "loss": -0.1824,
+      "step": 716
+    },
+    {
+      "epoch": 0.5734853029394121,
+      "grad_norm": 5.774496226610818,
+      "learning_rate": 8.811881188118812e-07,
+      "loss": -0.0764,
+      "step": 717
+    },
+    {
+      "epoch": 0.5742851429714058,
+      "grad_norm": 4.53339855951246,
+      "learning_rate": 8.795379537953795e-07,
+      "loss": -0.0494,
+      "step": 718
+    },
+    {
+      "epoch": 0.5750849830033993,
+      "grad_norm": 3.7011611817540118,
+      "learning_rate": 8.778877887788778e-07,
+      "loss": -0.1227,
+      "step": 719
+    },
+    {
+      "epoch": 0.5758848230353929,
+      "grad_norm": 4.117237995569142,
+      "learning_rate": 8.762376237623762e-07,
+      "loss": -0.0782,
+      "step": 720
+    },
+    {
+      "epoch": 0.5766846630673865,
+      "grad_norm": 3.347109381610254,
+      "learning_rate": 8.745874587458745e-07,
+      "loss": -0.1517,
+      "step": 721
+    },
+    {
+      "epoch": 0.5774845030993802,
+      "grad_norm": 3.9587273384893447,
+      "learning_rate": 8.729372937293729e-07,
+      "loss": -0.0647,
+      "step": 722
+    },
+    {
+      "epoch": 0.5782843431313738,
+      "grad_norm": 4.073013317525639,
+      "learning_rate": 8.712871287128712e-07,
+      "loss": -0.0354,
+      "step": 723
+    },
+    {
+      "epoch": 0.5790841831633673,
+      "grad_norm": 4.360786018214523,
+      "learning_rate": 8.696369636963697e-07,
+      "loss": -0.0845,
+      "step": 724
+    },
+    {
+      "epoch": 0.5798840231953609,
+      "grad_norm": 4.911325926898916,
+      "learning_rate": 8.679867986798679e-07,
+      "loss": -0.0581,
+      "step": 725
+    },
+    {
+      "epoch": 0.5806838632273545,
+      "grad_norm": 7.374565499647674,
+      "learning_rate": 8.663366336633663e-07,
+      "loss": 0.0876,
+      "step": 726
+    },
+    {
+      "epoch": 0.5814837032593482,
+      "grad_norm": 5.228899924882716,
+      "learning_rate": 8.646864686468647e-07,
+      "loss": -0.0188,
+      "step": 727
+    },
+    {
+      "epoch": 0.5822835432913417,
+      "grad_norm": 5.076190474367137,
+      "learning_rate": 8.63036303630363e-07,
+      "loss": -0.0482,
+      "step": 728
+    },
+    {
+      "epoch": 0.5830833833233353,
+      "grad_norm": 3.8391392043031067,
+      "learning_rate": 8.613861386138613e-07,
+      "loss": -0.0351,
+      "step": 729
+    },
+    {
+      "epoch": 0.5838832233553289,
+      "grad_norm": 4.720664943150823,
+      "learning_rate": 8.597359735973598e-07,
+      "loss": -0.0919,
+      "step": 730
+    },
+    {
+      "epoch": 0.5846830633873226,
+      "grad_norm": 3.696825036479519,
+      "learning_rate": 8.58085808580858e-07,
+      "loss": -0.0491,
+      "step": 731
+    },
+    {
+      "epoch": 0.5854829034193162,
+      "grad_norm": 4.957633580857494,
+      "learning_rate": 8.564356435643563e-07,
+      "loss": -0.0084,
+      "step": 732
+    },
+    {
+      "epoch": 0.5862827434513097,
+      "grad_norm": 4.145591204807576,
+      "learning_rate": 8.547854785478548e-07,
+      "loss": -0.0777,
+      "step": 733
+    },
+    {
+      "epoch": 0.5870825834833033,
+      "grad_norm": 6.008926427229853,
+      "learning_rate": 8.531353135313531e-07,
+      "loss": -0.0285,
+      "step": 734
+    },
+    {
+      "epoch": 0.587882423515297,
+      "grad_norm": 6.457940104008622,
+      "learning_rate": 8.514851485148514e-07,
+      "loss": -0.1126,
+      "step": 735
+    },
+    {
+      "epoch": 0.5886822635472906,
+      "grad_norm": 4.150286303835989,
+      "learning_rate": 8.498349834983498e-07,
+      "loss": -0.0633,
+      "step": 736
+    },
+    {
+      "epoch": 0.5894821035792841,
+      "grad_norm": 4.765817707105298,
+      "learning_rate": 8.481848184818482e-07,
+      "loss": -0.1274,
+      "step": 737
+    },
+    {
+      "epoch": 0.5902819436112777,
+      "grad_norm": 2.944307272093047,
+      "learning_rate": 8.465346534653464e-07,
+      "loss": -0.1143,
+      "step": 738
+    },
+    {
+      "epoch": 0.5910817836432714,
+      "grad_norm": 4.301068842918969,
+      "learning_rate": 8.448844884488449e-07,
+      "loss": -0.0303,
+      "step": 739
+    },
+    {
+      "epoch": 0.591881623675265,
+      "grad_norm": 3.5345181895694724,
+      "learning_rate": 8.432343234323432e-07,
+      "loss": -0.0369,
+      "step": 740
+    },
+    {
+      "epoch": 0.5926814637072585,
+      "grad_norm": 3.6758939784537477,
+      "learning_rate": 8.415841584158416e-07,
+      "loss": -0.0435,
+      "step": 741
+    },
+    {
+      "epoch": 0.5934813037392521,
+      "grad_norm": 9.280032916175081,
+      "learning_rate": 8.399339933993399e-07,
+      "loss": -0.077,
+      "step": 742
+    },
+    {
+      "epoch": 0.5942811437712457,
+      "grad_norm": 3.691324223005634,
+      "learning_rate": 8.382838283828383e-07,
+      "loss": 0.0384,
+      "step": 743
+    },
+    {
+      "epoch": 0.5950809838032394,
+      "grad_norm": 5.13228312974853,
+      "learning_rate": 8.366336633663366e-07,
+      "loss": -0.0101,
+      "step": 744
+    },
+    {
+      "epoch": 0.595880823835233,
+      "grad_norm": 3.127958499467288,
+      "learning_rate": 8.34983498349835e-07,
+      "loss": 0.0036,
+      "step": 745
+    },
+    {
+      "epoch": 0.5966806638672265,
+      "grad_norm": 4.371581567674568,
+      "learning_rate": 8.333333333333333e-07,
+      "loss": -0.088,
+      "step": 746
+    },
+    {
+      "epoch": 0.5974805038992201,
+      "grad_norm": 3.7498910252313786,
+      "learning_rate": 8.316831683168316e-07,
+      "loss": -0.0249,
+      "step": 747
+    },
+    {
+      "epoch": 0.5982803439312138,
+      "grad_norm": 9.554152491664782,
+      "learning_rate": 8.3003300330033e-07,
+      "loss": -0.0718,
+      "step": 748
+    },
+    {
+      "epoch": 0.5990801839632074,
+      "grad_norm": 4.450073267301403,
+      "learning_rate": 8.283828382838283e-07,
+      "loss": -0.038,
+      "step": 749
+    },
+    {
+      "epoch": 0.5998800239952009,
+      "grad_norm": 4.885796549111672,
+      "learning_rate": 8.267326732673267e-07,
+      "loss": -0.0279,
+      "step": 750
+    },
+    {
+      "epoch": 0.6006798640271945,
+      "grad_norm": 4.729710058959354,
+      "learning_rate": 8.25082508250825e-07,
+      "loss": -0.0415,
+      "step": 751
+    },
+    {
+      "epoch": 0.6014797040591882,
+      "grad_norm": 5.310403387692374,
+      "learning_rate": 8.234323432343234e-07,
+      "loss": -0.1964,
+      "step": 752
+    },
+    {
+      "epoch": 0.6022795440911818,
+      "grad_norm": 4.462129183196535,
+      "learning_rate": 8.217821782178217e-07,
+      "loss": -0.0289,
+      "step": 753
+    },
+    {
+      "epoch": 0.6030793841231754,
+      "grad_norm": 4.343559206058792,
+      "learning_rate": 8.201320132013201e-07,
+      "loss": 0.0155,
+      "step": 754
+    },
+    {
+      "epoch": 0.6038792241551689,
+      "grad_norm": 7.14111860643498,
+      "learning_rate": 8.184818481848184e-07,
+      "loss": 0.076,
+      "step": 755
+    },
+    {
+      "epoch": 0.6046790641871626,
+      "grad_norm": 5.741261351757093,
+      "learning_rate": 8.168316831683168e-07,
+      "loss": -0.0684,
+      "step": 756
+    },
+    {
+      "epoch": 0.6054789042191562,
+      "grad_norm": 4.841100743497433,
+      "learning_rate": 8.151815181518151e-07,
+      "loss": 0.0251,
+      "step": 757
+    },
+    {
+      "epoch": 0.6062787442511498,
+      "grad_norm": 5.776828704222559,
+      "learning_rate": 8.135313531353136e-07,
+      "loss": 0.0922,
+      "step": 758
+    },
+    {
+      "epoch": 0.6070785842831433,
+      "grad_norm": 5.171546395456714,
+      "learning_rate": 8.118811881188119e-07,
+      "loss": 0.0378,
+      "step": 759
+    },
+    {
+      "epoch": 0.607878424315137,
+      "grad_norm": 4.5497593071961475,
+      "learning_rate": 8.102310231023102e-07,
+      "loss": -0.0091,
+      "step": 760
+    },
+    {
+      "epoch": 0.6086782643471306,
+      "grad_norm": 3.4008260116242837,
+      "learning_rate": 8.085808580858086e-07,
+      "loss": -0.0513,
+      "step": 761
+    },
+    {
+      "epoch": 0.6094781043791242,
+      "grad_norm": 5.9528170929025475,
+      "learning_rate": 8.069306930693069e-07,
+      "loss": -0.0846,
+      "step": 762
+    },
+    {
+      "epoch": 0.6102779444111178,
+      "grad_norm": 8.59254741230532,
+      "learning_rate": 8.052805280528052e-07,
+      "loss": 0.0454,
+      "step": 763
+    },
+    {
+      "epoch": 0.6110777844431113,
+      "grad_norm": 5.92388092933109,
+      "learning_rate": 8.036303630363036e-07,
+      "loss": -0.0677,
+      "step": 764
+    },
+    {
+      "epoch": 0.611877624475105,
+      "grad_norm": 4.5071343981279375,
+      "learning_rate": 8.01980198019802e-07,
+      "loss": -0.0598,
+      "step": 765
+    },
+    {
+      "epoch": 0.6126774645070986,
+      "grad_norm": 5.095952967655762,
+      "learning_rate": 8.003300330033002e-07,
+      "loss": -0.1077,
+      "step": 766
+    },
+    {
+      "epoch": 0.6134773045390922,
+      "grad_norm": 3.686767117360266,
+      "learning_rate": 7.986798679867987e-07,
+      "loss": -0.0509,
+      "step": 767
+    },
+    {
+      "epoch": 0.6142771445710857,
+      "grad_norm": 4.709313867244328,
+      "learning_rate": 7.97029702970297e-07,
+      "loss": -0.0384,
+      "step": 768
+    },
+    {
+      "epoch": 0.6150769846030794,
+      "grad_norm": 4.606485786769665,
+      "learning_rate": 7.953795379537953e-07,
+      "loss": -0.0639,
+      "step": 769
+    },
+    {
+      "epoch": 0.615876824635073,
+      "grad_norm": 4.126308811511686,
+      "learning_rate": 7.937293729372937e-07,
+      "loss": -0.1454,
+      "step": 770
+    },
+    {
+      "epoch": 0.6166766646670666,
+      "grad_norm": 5.0891862328923985,
+      "learning_rate": 7.920792079207921e-07,
+      "loss": 0.0552,
+      "step": 771
+    },
+    {
+      "epoch": 0.6174765046990602,
+      "grad_norm": 4.348447825302712,
+      "learning_rate": 7.904290429042903e-07,
+      "loss": -0.0914,
+      "step": 772
+    },
+    {
+      "epoch": 0.6182763447310538,
+      "grad_norm": 4.069062786121328,
+      "learning_rate": 7.887788778877888e-07,
+      "loss": -0.0,
+      "step": 773
+    },
+    {
+      "epoch": 0.6190761847630474,
+      "grad_norm": 4.293180451800697,
+      "learning_rate": 7.871287128712871e-07,
+      "loss": -0.2021,
+      "step": 774
+    },
+    {
+      "epoch": 0.619876024795041,
+      "grad_norm": 6.832958614340714,
+      "learning_rate": 7.854785478547854e-07,
+      "loss": -0.049,
+      "step": 775
+    },
+    {
+      "epoch": 0.6206758648270346,
+      "grad_norm": 4.62295713929861,
+      "learning_rate": 7.838283828382838e-07,
+      "loss": -0.0099,
+      "step": 776
+    },
+    {
+      "epoch": 0.6214757048590281,
+      "grad_norm": 16.021103293232393,
+      "learning_rate": 7.821782178217821e-07,
+      "loss": -0.0229,
+      "step": 777
+    },
+    {
+      "epoch": 0.6222755448910218,
+      "grad_norm": 6.878826117062365,
+      "learning_rate": 7.805280528052805e-07,
+      "loss": 0.0852,
+      "step": 778
+    },
+    {
+      "epoch": 0.6230753849230154,
+      "grad_norm": 3.5925006851057595,
+      "learning_rate": 7.788778877887788e-07,
+      "loss": -0.0838,
+      "step": 779
+    },
+    {
+      "epoch": 0.623875224955009,
+      "grad_norm": 4.312199015410901,
+      "learning_rate": 7.772277227722772e-07,
+      "loss": -0.1938,
+      "step": 780
+    },
+    {
+      "epoch": 0.6246750649870026,
+      "grad_norm": 4.427437777822434,
+      "learning_rate": 7.755775577557755e-07,
+      "loss": -0.1088,
+      "step": 781
+    },
+    {
+      "epoch": 0.6254749050189962,
+      "grad_norm": 3.391720502521415,
+      "learning_rate": 7.739273927392739e-07,
+      "loss": -0.1204,
+      "step": 782
+    },
+    {
+      "epoch": 0.6262747450509898,
+      "grad_norm": 3.3125718306904512,
+      "learning_rate": 7.722772277227722e-07,
+      "loss": -0.0828,
+      "step": 783
+    },
+    {
+      "epoch": 0.6270745850829834,
+      "grad_norm": 5.402117257600779,
+      "learning_rate": 7.706270627062707e-07,
+      "loss": -0.054,
+      "step": 784
+    },
+    {
+      "epoch": 0.627874425114977,
+      "grad_norm": 4.489840589382479,
+      "learning_rate": 7.689768976897689e-07,
+      "loss": -0.0368,
+      "step": 785
+    },
+    {
+      "epoch": 0.6286742651469706,
+      "grad_norm": 3.848546702562119,
+      "learning_rate": 7.673267326732673e-07,
+      "loss": -0.0842,
+      "step": 786
+    },
+    {
+      "epoch": 0.6294741051789642,
+      "grad_norm": 4.3003301981629765,
+      "learning_rate": 7.656765676567657e-07,
+      "loss": -0.1575,
+      "step": 787
+    },
+    {
+      "epoch": 0.6302739452109578,
+      "grad_norm": 3.234095759477404,
+      "learning_rate": 7.64026402640264e-07,
+      "loss": -0.1448,
+      "step": 788
+    },
+    {
+      "epoch": 0.6310737852429514,
+      "grad_norm": 4.587363650091271,
+      "learning_rate": 7.623762376237624e-07,
+      "loss": -0.0806,
+      "step": 789
+    },
+    {
+      "epoch": 0.6318736252749451,
+      "grad_norm": 4.91701458923025,
+      "learning_rate": 7.607260726072607e-07,
+      "loss": 0.0289,
+      "step": 790
+    },
+    {
+      "epoch": 0.6326734653069386,
+      "grad_norm": 4.26117605640356,
+      "learning_rate": 7.59075907590759e-07,
+      "loss": -0.0442,
+      "step": 791
+    },
+    {
+      "epoch": 0.6334733053389322,
+      "grad_norm": 4.714328336316601,
+      "learning_rate": 7.574257425742574e-07,
+      "loss": -0.1182,
+      "step": 792
+    },
+    {
+      "epoch": 0.6342731453709258,
+      "grad_norm": 4.8043256421789975,
+      "learning_rate": 7.557755775577558e-07,
+      "loss": -0.0532,
+      "step": 793
+    },
+    {
+      "epoch": 0.6350729854029195,
+      "grad_norm": 4.981158806362152,
+      "learning_rate": 7.54125412541254e-07,
+      "loss": -0.0322,
+      "step": 794
+    },
+    {
+      "epoch": 0.635872825434913,
+      "grad_norm": 6.600526915815604,
+      "learning_rate": 7.524752475247525e-07,
+      "loss": -0.1059,
+      "step": 795
+    },
+    {
+      "epoch": 0.6366726654669066,
+      "grad_norm": 4.989184316503347,
+      "learning_rate": 7.508250825082508e-07,
+      "loss": -0.0888,
+      "step": 796
+    },
+    {
+      "epoch": 0.6374725054989002,
+      "grad_norm": 4.683126825596764,
+      "learning_rate": 7.491749174917491e-07,
+      "loss": -0.0425,
+      "step": 797
+    },
+    {
+      "epoch": 0.6382723455308938,
+      "grad_norm": 2.835338855117646,
+      "learning_rate": 7.475247524752475e-07,
+      "loss": -0.1188,
+      "step": 798
+    },
+    {
+      "epoch": 0.6390721855628875,
+      "grad_norm": 3.334782349034145,
+      "learning_rate": 7.458745874587459e-07,
+      "loss": -0.0085,
+      "step": 799
+    },
+    {
+      "epoch": 0.639872025594881,
+      "grad_norm": 3.8950834045490677,
+      "learning_rate": 7.442244224422441e-07,
+      "loss": -0.0866,
+      "step": 800
+    },
+    {
+      "epoch": 0.6406718656268746,
+      "grad_norm": 3.7954504535302047,
+      "learning_rate": 7.425742574257426e-07,
+      "loss": -0.1718,
+      "step": 801
+    },
+    {
+      "epoch": 0.6414717056588682,
+      "grad_norm": 3.9147882759458907,
+      "learning_rate": 7.409240924092409e-07,
+      "loss": -0.1172,
+      "step": 802
+    },
+    {
+      "epoch": 0.6422715456908619,
+      "grad_norm": 4.14195212922553,
+      "learning_rate": 7.392739273927392e-07,
+      "loss": -0.2055,
+      "step": 803
+    },
+    {
+      "epoch": 0.6430713857228554,
+      "grad_norm": 5.564167929906255,
+      "learning_rate": 7.376237623762376e-07,
+      "loss": -0.0587,
+      "step": 804
+    },
+    {
+      "epoch": 0.643871225754849,
+      "grad_norm": 5.104003509197404,
+      "learning_rate": 7.359735973597359e-07,
+      "loss": -0.1599,
+      "step": 805
+    },
+    {
+      "epoch": 0.6446710657868426,
+      "grad_norm": 3.249110466926901,
+      "learning_rate": 7.343234323432343e-07,
+      "loss": 0.0254,
+      "step": 806
+    },
+    {
+      "epoch": 0.6454709058188363,
+      "grad_norm": 5.330488201062819,
+      "learning_rate": 7.326732673267326e-07,
+      "loss": -0.0932,
+      "step": 807
+    },
+    {
+      "epoch": 0.6462707458508299,
+      "grad_norm": 3.391881050876262,
+      "learning_rate": 7.31023102310231e-07,
+      "loss": -0.0045,
+      "step": 808
+    },
+    {
+      "epoch": 0.6470705858828234,
+      "grad_norm": 4.729176906400958,
+      "learning_rate": 7.293729372937293e-07,
+      "loss": 0.0309,
+      "step": 809
+    },
+    {
+      "epoch": 0.647870425914817,
+      "grad_norm": 4.873305784391995,
+      "learning_rate": 7.277227722772277e-07,
+      "loss": -0.016,
+      "step": 810
+    },
+    {
+      "epoch": 0.6486702659468107,
+      "grad_norm": 4.737325724513948,
+      "learning_rate": 7.26072607260726e-07,
+      "loss": -0.0555,
+      "step": 811
+    },
+    {
+      "epoch": 0.6494701059788043,
+      "grad_norm": 4.700178573137915,
+      "learning_rate": 7.244224422442245e-07,
+      "loss": -0.0054,
+      "step": 812
+    },
+    {
+      "epoch": 0.6502699460107978,
+      "grad_norm": 3.8303301007119375,
+      "learning_rate": 7.227722772277227e-07,
+      "loss": -0.0237,
+      "step": 813
+    },
+    {
+      "epoch": 0.6510697860427914,
+      "grad_norm": 3.9787912774229404,
+      "learning_rate": 7.211221122112211e-07,
+      "loss": -0.0158,
+      "step": 814
+    },
+    {
+      "epoch": 0.651869626074785,
+      "grad_norm": 4.3995808661843805,
+      "learning_rate": 7.194719471947195e-07,
+      "loss": 0.0276,
+      "step": 815
+    },
+    {
+      "epoch": 0.6526694661067787,
+      "grad_norm": 4.052268422009291,
+      "learning_rate": 7.178217821782178e-07,
+      "loss": -0.0178,
+      "step": 816
+    },
+    {
+      "epoch": 0.6534693061387723,
+      "grad_norm": 4.162149014725009,
+      "learning_rate": 7.161716171617161e-07,
+      "loss": -0.0343,
+      "step": 817
+    },
+    {
+      "epoch": 0.6542691461707658,
+      "grad_norm": 4.733808124710197,
+      "learning_rate": 7.145214521452146e-07,
+      "loss": -0.0719,
+      "step": 818
+    },
+    {
+      "epoch": 0.6550689862027594,
+      "grad_norm": 4.367810691084756,
+      "learning_rate": 7.128712871287128e-07,
+      "loss": -0.1031,
+      "step": 819
+    },
+    {
+      "epoch": 0.6558688262347531,
+      "grad_norm": 3.393663864026175,
+      "learning_rate": 7.112211221122111e-07,
+      "loss": -0.0469,
+      "step": 820
+    },
+    {
+      "epoch": 0.6566686662667467,
+      "grad_norm": 6.593078579096907,
+      "learning_rate": 7.095709570957096e-07,
+      "loss": -0.0492,
+      "step": 821
+    },
+    {
+      "epoch": 0.6574685062987402,
+      "grad_norm": 5.90593512770552,
+      "learning_rate": 7.079207920792078e-07,
+      "loss": 0.017,
+      "step": 822
+    },
+    {
+      "epoch": 0.6582683463307338,
+      "grad_norm": 4.280214281867913,
+      "learning_rate": 7.062706270627063e-07,
+      "loss": -0.0432,
+      "step": 823
+    },
+    {
+      "epoch": 0.6590681863627275,
+      "grad_norm": 8.423741404535653,
+      "learning_rate": 7.046204620462046e-07,
+      "loss": -0.0291,
+      "step": 824
+    },
+    {
+      "epoch": 0.6598680263947211,
+      "grad_norm": 5.038317153573228,
+      "learning_rate": 7.029702970297029e-07,
+      "loss": -0.093,
+      "step": 825
+    },
+    {
+      "epoch": 0.6606678664267147,
+      "grad_norm": 6.706344720432834,
+      "learning_rate": 7.013201320132013e-07,
+      "loss": -0.1315,
+      "step": 826
+    },
+    {
+      "epoch": 0.6614677064587082,
+      "grad_norm": 3.716489601764274,
+      "learning_rate": 6.996699669966997e-07,
+      "loss": -0.0825,
+      "step": 827
+    },
+    {
+      "epoch": 0.6622675464907019,
+      "grad_norm": 4.158836764412884,
+      "learning_rate": 6.980198019801979e-07,
+      "loss": -0.0161,
+      "step": 828
+    },
+    {
+      "epoch": 0.6630673865226955,
+      "grad_norm": 3.733283485919958,
+      "learning_rate": 6.963696369636964e-07,
+      "loss": -0.0594,
+      "step": 829
+    },
+    {
+      "epoch": 0.6638672265546891,
+      "grad_norm": 6.968567713247902,
+      "learning_rate": 6.947194719471947e-07,
+      "loss": -0.0441,
+      "step": 830
+    },
+    {
+      "epoch": 0.6646670665866826,
+      "grad_norm": 5.008730323701448,
+      "learning_rate": 6.93069306930693e-07,
+      "loss": -0.1307,
+      "step": 831
+    },
+    {
+      "epoch": 0.6654669066186762,
+      "grad_norm": 4.2175650733942955,
+      "learning_rate": 6.914191419141914e-07,
+      "loss": -0.0771,
+      "step": 832
+    },
+    {
+      "epoch": 0.6662667466506699,
+      "grad_norm": 5.402322742920563,
+      "learning_rate": 6.897689768976897e-07,
+      "loss": -0.0841,
+      "step": 833
+    },
+    {
+      "epoch": 0.6670665866826635,
+      "grad_norm": 5.646942573991696,
+      "learning_rate": 6.88118811881188e-07,
+      "loss": -0.0622,
+      "step": 834
+    },
+    {
+      "epoch": 0.667866426714657,
+      "grad_norm": 13.649360926832344,
+      "learning_rate": 6.864686468646864e-07,
+      "loss": -0.0637,
+      "step": 835
+    },
+    {
+      "epoch": 0.6686662667466506,
+      "grad_norm": 5.461268948386568,
+      "learning_rate": 6.848184818481848e-07,
+      "loss": -0.0241,
+      "step": 836
+    },
+    {
+      "epoch": 0.6694661067786443,
+      "grad_norm": 4.011621520471584,
+      "learning_rate": 6.831683168316831e-07,
+      "loss": -0.0422,
+      "step": 837
+    },
+    {
+      "epoch": 0.6702659468106379,
+      "grad_norm": 20.835014010983784,
+      "learning_rate": 6.815181518151815e-07,
+      "loss": -0.126,
+      "step": 838
+    },
+    {
+      "epoch": 0.6710657868426315,
+      "grad_norm": 5.041368060556288,
+      "learning_rate": 6.798679867986798e-07,
+      "loss": -0.1016,
+      "step": 839
+    },
+    {
+      "epoch": 0.671865626874625,
+      "grad_norm": 5.834292995896152,
+      "learning_rate": 6.782178217821783e-07,
+      "loss": -0.0738,
+      "step": 840
+    },
+    {
+      "epoch": 0.6726654669066187,
+      "grad_norm": 3.5120723151753985,
+      "learning_rate": 6.765676567656765e-07,
+      "loss": -0.0504,
+      "step": 841
+    },
+    {
+      "epoch": 0.6734653069386123,
+      "grad_norm": 2.901517891733533,
+      "learning_rate": 6.749174917491749e-07,
+      "loss": -0.0698,
+      "step": 842
+    },
+    {
+      "epoch": 0.6742651469706059,
+      "grad_norm": 3.465841087435974,
+      "learning_rate": 6.732673267326733e-07,
+      "loss": -0.1227,
+      "step": 843
+    },
+    {
+      "epoch": 0.6750649870025994,
+      "grad_norm": 4.536588693958206,
+      "learning_rate": 6.716171617161716e-07,
+      "loss": -0.0602,
+      "step": 844
+    },
+    {
+      "epoch": 0.675864827034593,
+      "grad_norm": 10.903656834330391,
+      "learning_rate": 6.699669966996699e-07,
+      "loss": -0.1289,
+      "step": 845
+    },
+    {
+      "epoch": 0.6766646670665867,
+      "grad_norm": 7.296365266758308,
+      "learning_rate": 6.683168316831684e-07,
+      "loss": -0.1561,
+      "step": 846
+    },
+    {
+      "epoch": 0.6774645070985803,
+      "grad_norm": 4.412331570876947,
+      "learning_rate": 6.666666666666666e-07,
+      "loss": -0.0958,
+      "step": 847
+    },
+    {
+      "epoch": 0.6782643471305739,
+      "grad_norm": 2.8672230897612345,
+      "learning_rate": 6.650165016501649e-07,
+      "loss": -0.1643,
+      "step": 848
+    },
+    {
+      "epoch": 0.6790641871625674,
+      "grad_norm": 4.5674033793568904,
+      "learning_rate": 6.633663366336634e-07,
+      "loss": -0.1336,
+      "step": 849
+    },
+    {
+      "epoch": 0.6798640271945611,
+      "grad_norm": 4.381540695320094,
+      "learning_rate": 6.617161716171616e-07,
+      "loss": -0.0132,
+      "step": 850
+    },
+    {
+      "epoch": 0.6806638672265547,
+      "grad_norm": 5.664059132542103,
+      "learning_rate": 6.6006600660066e-07,
+      "loss": -0.0285,
+      "step": 851
+    },
+    {
+      "epoch": 0.6814637072585483,
+      "grad_norm": 5.7461780617115,
+      "learning_rate": 6.584158415841584e-07,
+      "loss": 0.0047,
+      "step": 852
+    },
+    {
+      "epoch": 0.6822635472905418,
+      "grad_norm": 4.320720191192789,
+      "learning_rate": 6.567656765676567e-07,
+      "loss": -0.1447,
+      "step": 853
+    },
+    {
+      "epoch": 0.6830633873225355,
+      "grad_norm": 3.108737285911658,
+      "learning_rate": 6.55115511551155e-07,
+      "loss": -0.1611,
+      "step": 854
+    },
+    {
+      "epoch": 0.6838632273545291,
+      "grad_norm": 5.023876212557061,
+      "learning_rate": 6.534653465346535e-07,
+      "loss": -0.1529,
+      "step": 855
+    },
+    {
+      "epoch": 0.6846630673865227,
+      "grad_norm": 4.545192341611211,
+      "learning_rate": 6.518151815181517e-07,
+      "loss": -0.0456,
+      "step": 856
+    },
+    {
+      "epoch": 0.6854629074185163,
+      "grad_norm": 5.267884265007784,
+      "learning_rate": 6.501650165016502e-07,
+      "loss": -0.0288,
+      "step": 857
+    },
+    {
+      "epoch": 0.6862627474505099,
+      "grad_norm": 5.101939165542976,
+      "learning_rate": 6.485148514851485e-07,
+      "loss": -0.0036,
+      "step": 858
+    },
+    {
+      "epoch": 0.6870625874825035,
+      "grad_norm": 3.55089136672625,
+      "learning_rate": 6.468646864686468e-07,
+      "loss": -0.0762,
+      "step": 859
+    },
+    {
+      "epoch": 0.6878624275144971,
+      "grad_norm": 4.121292066096188,
+      "learning_rate": 6.452145214521452e-07,
+      "loss": -0.0858,
+      "step": 860
+    },
+    {
+      "epoch": 0.6886622675464907,
+      "grad_norm": 4.411771304555708,
+      "learning_rate": 6.435643564356436e-07,
+      "loss": -0.0359,
+      "step": 861
+    },
+    {
+      "epoch": 0.6894621075784843,
+      "grad_norm": 4.163024887578695,
+      "learning_rate": 6.419141914191419e-07,
+      "loss": -0.1036,
+      "step": 862
+    },
+    {
+      "epoch": 0.6902619476104779,
+      "grad_norm": 3.9253552561550307,
+      "learning_rate": 6.402640264026402e-07,
+      "loss": -0.092,
+      "step": 863
+    },
+    {
+      "epoch": 0.6910617876424715,
+      "grad_norm": 3.3756432096953906,
+      "learning_rate": 6.386138613861386e-07,
+      "loss": 0.0002,
+      "step": 864
+    },
+    {
+      "epoch": 0.6918616276744651,
+      "grad_norm": 4.4942197763700245,
+      "learning_rate": 6.369636963696369e-07,
+      "loss": 0.0044,
+      "step": 865
+    },
+    {
+      "epoch": 0.6926614677064588,
+      "grad_norm": 4.469003096662444,
+      "learning_rate": 6.353135313531353e-07,
+      "loss": -0.0027,
+      "step": 866
+    },
+    {
+      "epoch": 0.6934613077384523,
+      "grad_norm": 3.212779461175308,
+      "learning_rate": 6.336633663366336e-07,
+      "loss": -0.0255,
+      "step": 867
+    },
+    {
+      "epoch": 0.6942611477704459,
+      "grad_norm": 3.473986685130551,
+      "learning_rate": 6.32013201320132e-07,
+      "loss": -0.0171,
+      "step": 868
+    },
+    {
+      "epoch": 0.6950609878024395,
+      "grad_norm": 4.216176017792983,
+      "learning_rate": 6.303630363036303e-07,
+      "loss": -0.0673,
+      "step": 869
+    },
+    {
+      "epoch": 0.6958608278344331,
+      "grad_norm": 12.098304541476889,
+      "learning_rate": 6.287128712871287e-07,
+      "loss": 0.0742,
+      "step": 870
+    },
+    {
+      "epoch": 0.6966606678664267,
+      "grad_norm": 8.083305542363899,
+      "learning_rate": 6.270627062706271e-07,
+      "loss": 0.0401,
+      "step": 871
+    },
+    {
+      "epoch": 0.6974605078984203,
+      "grad_norm": 4.685734734341067,
+      "learning_rate": 6.254125412541254e-07,
+      "loss": -0.0235,
+      "step": 872
+    },
+    {
+      "epoch": 0.6982603479304139,
+      "grad_norm": 5.85956379904162,
+      "learning_rate": 6.237623762376237e-07,
+      "loss": -0.1556,
+      "step": 873
+    },
+    {
+      "epoch": 0.6990601879624075,
+      "grad_norm": 3.61321328588953,
+      "learning_rate": 6.221122112211222e-07,
+      "loss": -0.093,
+      "step": 874
+    },
+    {
+      "epoch": 0.6998600279944012,
+      "grad_norm": 3.528560458155755,
+      "learning_rate": 6.204620462046204e-07,
+      "loss": -0.0322,
+      "step": 875
+    },
+    {
+      "epoch": 0.7006598680263947,
+      "grad_norm": 6.51982029204985,
+      "learning_rate": 6.188118811881187e-07,
+      "loss": 0.0667,
+      "step": 876
+    },
+    {
+      "epoch": 0.7014597080583883,
+      "grad_norm": 4.530213336089527,
+      "learning_rate": 6.171617161716172e-07,
+      "loss": -0.0434,
+      "step": 877
+    },
+    {
+      "epoch": 0.7022595480903819,
+      "grad_norm": 6.616076811947681,
+      "learning_rate": 6.155115511551154e-07,
+      "loss": 0.0219,
+      "step": 878
+    },
+    {
+      "epoch": 0.7030593881223756,
+      "grad_norm": 4.1562171877376075,
+      "learning_rate": 6.138613861386138e-07,
+      "loss": -0.0648,
+      "step": 879
+    },
+    {
+      "epoch": 0.7038592281543691,
+      "grad_norm": 4.419647626337558,
+      "learning_rate": 6.122112211221122e-07,
+      "loss": -0.1112,
+      "step": 880
+    },
+    {
+      "epoch": 0.7046590681863627,
+      "grad_norm": 4.940388667457944,
+      "learning_rate": 6.105610561056105e-07,
+      "loss": -0.0162,
+      "step": 881
+    },
+    {
+      "epoch": 0.7054589082183563,
+      "grad_norm": 6.023947292183416,
+      "learning_rate": 6.089108910891088e-07,
+      "loss": -0.0435,
+      "step": 882
+    },
+    {
+      "epoch": 0.70625874825035,
+      "grad_norm": 7.1901426018379935,
+      "learning_rate": 6.072607260726073e-07,
+      "loss": -0.001,
+      "step": 883
+    },
+    {
+      "epoch": 0.7070585882823436,
+      "grad_norm": 5.858167518812244,
+      "learning_rate": 6.056105610561055e-07,
+      "loss": -0.0017,
+      "step": 884
+    },
+    {
+      "epoch": 0.7078584283143371,
+      "grad_norm": 4.6643313740940835,
+      "learning_rate": 6.03960396039604e-07,
+      "loss": -0.0808,
+      "step": 885
+    },
+    {
+      "epoch": 0.7086582683463307,
+      "grad_norm": 3.964900755824796,
+      "learning_rate": 6.023102310231023e-07,
+      "loss": -0.1257,
+      "step": 886
+    },
+    {
+      "epoch": 0.7094581083783243,
+      "grad_norm": 3.6312620074127797,
+      "learning_rate": 6.006600660066007e-07,
+      "loss": -0.044,
+      "step": 887
+    },
+    {
+      "epoch": 0.710257948410318,
+      "grad_norm": 4.46804227760141,
+      "learning_rate": 5.99009900990099e-07,
+      "loss": -0.0864,
+      "step": 888
+    },
+    {
+      "epoch": 0.7110577884423115,
+      "grad_norm": 3.9877331513072884,
+      "learning_rate": 5.973597359735974e-07,
+      "loss": -0.0555,
+      "step": 889
+    },
+    {
+      "epoch": 0.7118576284743051,
+      "grad_norm": 5.71452000001456,
+      "learning_rate": 5.957095709570957e-07,
+      "loss": -0.0516,
+      "step": 890
+    },
+    {
+      "epoch": 0.7126574685062987,
+      "grad_norm": 4.729225707280767,
+      "learning_rate": 5.94059405940594e-07,
+      "loss": -0.1252,
+      "step": 891
+    },
+    {
+      "epoch": 0.7134573085382924,
+      "grad_norm": 10.531702048903348,
+      "learning_rate": 5.924092409240924e-07,
+      "loss": -0.0143,
+      "step": 892
+    },
+    {
+      "epoch": 0.714257148570286,
+      "grad_norm": 4.337061940699697,
+      "learning_rate": 5.907590759075907e-07,
+      "loss": -0.0256,
+      "step": 893
+    },
+    {
+      "epoch": 0.7150569886022795,
+      "grad_norm": 4.553139268045056,
+      "learning_rate": 5.891089108910891e-07,
+      "loss": 0.0592,
+      "step": 894
+    },
+    {
+      "epoch": 0.7158568286342731,
+      "grad_norm": 3.65289282929829,
+      "learning_rate": 5.874587458745874e-07,
+      "loss": 0.0011,
+      "step": 895
+    },
+    {
+      "epoch": 0.7166566686662668,
+      "grad_norm": 2.969949901428,
+      "learning_rate": 5.858085808580858e-07,
+      "loss": -0.1288,
+      "step": 896
+    },
+    {
+      "epoch": 0.7174565086982604,
+      "grad_norm": 5.316965178875907,
+      "learning_rate": 5.841584158415841e-07,
+      "loss": -0.0688,
+      "step": 897
+    },
+    {
+      "epoch": 0.7182563487302539,
+      "grad_norm": 6.112915886146603,
+      "learning_rate": 5.825082508250825e-07,
+      "loss": -0.1923,
+      "step": 898
+    },
+    {
+      "epoch": 0.7190561887622475,
+      "grad_norm": 4.072265156624673,
+      "learning_rate": 5.808580858085808e-07,
+      "loss": -0.1749,
+      "step": 899
+    },
+    {
+      "epoch": 0.7198560287942412,
+      "grad_norm": 4.286524287381163,
+      "learning_rate": 5.792079207920792e-07,
+      "loss": -0.0736,
+      "step": 900
+    },
+    {
+      "epoch": 0.7206558688262348,
+      "grad_norm": 6.654813369667659,
+      "learning_rate": 5.775577557755775e-07,
+      "loss": -0.1185,
+      "step": 901
+    },
+    {
+      "epoch": 0.7214557088582284,
+      "grad_norm": 5.075962580453491,
+      "learning_rate": 5.75907590759076e-07,
+      "loss": 0.0889,
+      "step": 902
+    },
+    {
+      "epoch": 0.7222555488902219,
+      "grad_norm": 5.581410015072146,
+      "learning_rate": 5.742574257425742e-07,
+      "loss": -0.0718,
+      "step": 903
+    },
+    {
+      "epoch": 0.7230553889222155,
+      "grad_norm": 7.760040178489886,
+      "learning_rate": 5.726072607260726e-07,
+      "loss": -0.0227,
+      "step": 904
+    },
+    {
+      "epoch": 0.7238552289542092,
+      "grad_norm": 4.491264765964933,
+      "learning_rate": 5.70957095709571e-07,
+      "loss": -0.0838,
+      "step": 905
+    },
+    {
+      "epoch": 0.7246550689862028,
+      "grad_norm": 4.876358038696258,
+      "learning_rate": 5.693069306930692e-07,
+      "loss": -0.172,
+      "step": 906
+    },
+    {
+      "epoch": 0.7254549090181963,
+      "grad_norm": 3.9354286195012422,
+      "learning_rate": 5.676567656765676e-07,
+      "loss": -0.0793,
+      "step": 907
+    },
+    {
+      "epoch": 0.7262547490501899,
+      "grad_norm": 3.51016598192195,
+      "learning_rate": 5.66006600660066e-07,
+      "loss": 0.0052,
+      "step": 908
+    },
+    {
+      "epoch": 0.7270545890821836,
+      "grad_norm": 3.6895775842146166,
+      "learning_rate": 5.643564356435643e-07,
+      "loss": -0.0167,
+      "step": 909
+    },
+    {
+      "epoch": 0.7278544291141772,
+      "grad_norm": 4.773443293841103,
+      "learning_rate": 5.627062706270626e-07,
+      "loss": -0.0359,
+      "step": 910
+    },
+    {
+      "epoch": 0.7286542691461708,
+      "grad_norm": 4.278237278247243,
+      "learning_rate": 5.610561056105611e-07,
+      "loss": -0.054,
+      "step": 911
+    },
+    {
+      "epoch": 0.7294541091781643,
+      "grad_norm": 10.27415077431224,
+      "learning_rate": 5.594059405940594e-07,
+      "loss": -0.0075,
+      "step": 912
+    },
+    {
+      "epoch": 0.730253949210158,
+      "grad_norm": 4.2766288821859755,
+      "learning_rate": 5.577557755775577e-07,
+      "loss": 0.0493,
+      "step": 913
+    },
+    {
+      "epoch": 0.7310537892421516,
+      "grad_norm": 3.5281016400546275,
+      "learning_rate": 5.561056105610561e-07,
+      "loss": -0.0766,
+      "step": 914
+    },
+    {
+      "epoch": 0.7318536292741452,
+      "grad_norm": 3.8674946364382223,
+      "learning_rate": 5.544554455445545e-07,
+      "loss": -0.0442,
+      "step": 915
+    },
+    {
+      "epoch": 0.7326534693061387,
+      "grad_norm": 5.4465871711884395,
+      "learning_rate": 5.528052805280527e-07,
+      "loss": -0.0841,
+      "step": 916
+    },
+    {
+      "epoch": 0.7334533093381324,
+      "grad_norm": 3.530741427097772,
+      "learning_rate": 5.511551155115512e-07,
+      "loss": -0.0629,
+      "step": 917
+    },
+    {
+      "epoch": 0.734253149370126,
+      "grad_norm": 3.965321298788348,
+      "learning_rate": 5.495049504950495e-07,
+      "loss": -0.0573,
+      "step": 918
+    },
+    {
+      "epoch": 0.7350529894021196,
+      "grad_norm": 4.295307109186891,
+      "learning_rate": 5.478547854785477e-07,
+      "loss": -0.1381,
+      "step": 919
+    },
+    {
+      "epoch": 0.7358528294341132,
+      "grad_norm": 3.8500617084264257,
+      "learning_rate": 5.462046204620462e-07,
+      "loss": -0.049,
+      "step": 920
+    },
+    {
+      "epoch": 0.7366526694661067,
+      "grad_norm": 4.8663143152337005,
+      "learning_rate": 5.445544554455445e-07,
+      "loss": -0.0553,
+      "step": 921
+    },
+    {
+      "epoch": 0.7374525094981004,
+      "grad_norm": 3.4389426238655476,
+      "learning_rate": 5.429042904290429e-07,
+      "loss": -0.1896,
+      "step": 922
+    },
+    {
+      "epoch": 0.738252349530094,
+      "grad_norm": 5.019665193069423,
+      "learning_rate": 5.412541254125412e-07,
+      "loss": 0.1007,
+      "step": 923
+    },
+    {
+      "epoch": 0.7390521895620876,
+      "grad_norm": 3.9100586900916126,
+      "learning_rate": 5.396039603960396e-07,
+      "loss": -0.0697,
+      "step": 924
+    },
+    {
+      "epoch": 0.7398520295940811,
+      "grad_norm": 3.919136094051066,
+      "learning_rate": 5.379537953795379e-07,
+      "loss": -0.0907,
+      "step": 925
+    },
+    {
+      "epoch": 0.7406518696260748,
+      "grad_norm": 4.124863593418168,
+      "learning_rate": 5.363036303630363e-07,
+      "loss": -0.0569,
+      "step": 926
+    },
+    {
+      "epoch": 0.7414517096580684,
+      "grad_norm": 3.9668145454046977,
+      "learning_rate": 5.346534653465346e-07,
+      "loss": -0.0045,
+      "step": 927
+    },
+    {
+      "epoch": 0.742251549690062,
+      "grad_norm": 4.794421236275003,
+      "learning_rate": 5.33003300330033e-07,
+      "loss": -0.1284,
+      "step": 928
+    },
+    {
+      "epoch": 0.7430513897220556,
+      "grad_norm": 5.586609443850984,
+      "learning_rate": 5.313531353135313e-07,
+      "loss": -0.0346,
+      "step": 929
+    },
+    {
+      "epoch": 0.7438512297540492,
+      "grad_norm": 5.408030486873039,
+      "learning_rate": 5.297029702970297e-07,
+      "loss": -0.1473,
+      "step": 930
+    },
+    {
+      "epoch": 0.7446510697860428,
+      "grad_norm": 3.506888694865617,
+      "learning_rate": 5.28052805280528e-07,
+      "loss": -0.0018,
+      "step": 931
+    },
+    {
+      "epoch": 0.7454509098180364,
+      "grad_norm": 4.703351899310227,
+      "learning_rate": 5.264026402640264e-07,
+      "loss": 0.0077,
+      "step": 932
+    },
+    {
+      "epoch": 0.74625074985003,
+      "grad_norm": 5.030173808558858,
+      "learning_rate": 5.247524752475247e-07,
+      "loss": -0.1189,
+      "step": 933
+    },
+    {
+      "epoch": 0.7470505898820236,
+      "grad_norm": 6.14459022838033,
+      "learning_rate": 5.23102310231023e-07,
+      "loss": 0.0135,
+      "step": 934
+    },
+    {
+      "epoch": 0.7478504299140172,
+      "grad_norm": 3.162063833925453,
+      "learning_rate": 5.214521452145214e-07,
+      "loss": -0.0648,
+      "step": 935
+    },
+    {
+      "epoch": 0.7486502699460108,
+      "grad_norm": 3.322990934843452,
+      "learning_rate": 5.198019801980198e-07,
+      "loss": -0.0444,
+      "step": 936
+    },
+    {
+      "epoch": 0.7494501099780044,
+      "grad_norm": 7.281580577762579,
+      "learning_rate": 5.181518151815182e-07,
+      "loss": -0.079,
+      "step": 937
+    },
+    {
+      "epoch": 0.750249950009998,
+      "grad_norm": 4.995689441346887,
+      "learning_rate": 5.165016501650164e-07,
+      "loss": -0.0494,
+      "step": 938
+    },
+    {
+      "epoch": 0.7510497900419916,
+      "grad_norm": 3.662981016059356,
+      "learning_rate": 5.148514851485149e-07,
+      "loss": -0.1306,
+      "step": 939
+    },
+    {
+      "epoch": 0.7518496300739852,
+      "grad_norm": 4.564675844346983,
+      "learning_rate": 5.132013201320132e-07,
+      "loss": 0.0849,
+      "step": 940
+    },
+    {
+      "epoch": 0.7526494701059788,
+      "grad_norm": 4.127114880881102,
+      "learning_rate": 5.115511551155115e-07,
+      "loss": -0.1003,
+      "step": 941
+    },
+    {
+      "epoch": 0.7534493101379725,
+      "grad_norm": 7.580284959172436,
+      "learning_rate": 5.099009900990099e-07,
+      "loss": 0.0127,
+      "step": 942
+    },
+    {
+      "epoch": 0.754249150169966,
+      "grad_norm": 3.8477764594959223,
+      "learning_rate": 5.082508250825083e-07,
+      "loss": -0.0247,
+      "step": 943
+    },
+    {
+      "epoch": 0.7550489902019596,
+      "grad_norm": 5.134480523239938,
+      "learning_rate": 5.066006600660065e-07,
+      "loss": -0.064,
+      "step": 944
+    },
+    {
+      "epoch": 0.7558488302339532,
+      "grad_norm": 4.184980403074917,
+      "learning_rate": 5.04950495049505e-07,
+      "loss": -0.0374,
+      "step": 945
+    },
+    {
+      "epoch": 0.7566486702659468,
+      "grad_norm": 4.095109087896494,
+      "learning_rate": 5.033003300330033e-07,
+      "loss": -0.0667,
+      "step": 946
+    },
+    {
+      "epoch": 0.7574485102979404,
+      "grad_norm": 4.493509273952016,
+      "learning_rate": 5.016501650165016e-07,
+      "loss": -0.046,
+      "step": 947
+    },
+    {
+      "epoch": 0.758248350329934,
+      "grad_norm": 4.37283031614021,
+      "learning_rate": 5e-07,
+      "loss": -0.0662,
+      "step": 948
+    },
+    {
+      "epoch": 0.7590481903619276,
+      "grad_norm": 4.509842387488209,
+      "learning_rate": 4.983498349834983e-07,
+      "loss": -0.0951,
+      "step": 949
+    },
+    {
+      "epoch": 0.7598480303939212,
+      "grad_norm": 5.100099502789577,
+      "learning_rate": 4.966996699669966e-07,
+      "loss": -0.1086,
+      "step": 950
+    },
+    {
+      "epoch": 0.7606478704259149,
+      "grad_norm": 4.034320460393353,
+      "learning_rate": 4.95049504950495e-07,
+      "loss": -0.0263,
+      "step": 951
+    },
+    {
+      "epoch": 0.7614477104579084,
+      "grad_norm": 4.593555887493731,
+      "learning_rate": 4.933993399339933e-07,
+      "loss": -0.1636,
+      "step": 952
+    },
+    {
+      "epoch": 0.762247550489902,
+      "grad_norm": 4.526550802808929,
+      "learning_rate": 4.917491749174918e-07,
+      "loss": -0.0163,
+      "step": 953
+    },
+    {
+      "epoch": 0.7630473905218956,
+      "grad_norm": 5.515431862710494,
+      "learning_rate": 4.900990099009901e-07,
+      "loss": 0.0348,
+      "step": 954
+    },
+    {
+      "epoch": 0.7638472305538893,
+      "grad_norm": 4.2009128577609145,
+      "learning_rate": 4.884488448844884e-07,
+      "loss": -0.2542,
+      "step": 955
+    },
+    {
+      "epoch": 0.7646470705858828,
+      "grad_norm": 4.271019064733657,
+      "learning_rate": 4.867986798679868e-07,
+      "loss": -0.1869,
+      "step": 956
+    },
+    {
+      "epoch": 0.7654469106178764,
+      "grad_norm": 2.77377368877347,
+      "learning_rate": 4.851485148514851e-07,
+      "loss": -0.0735,
+      "step": 957
+    },
+    {
+      "epoch": 0.76624675064987,
+      "grad_norm": 6.0608792055595995,
+      "learning_rate": 4.834983498349835e-07,
+      "loss": -0.0099,
+      "step": 958
+    },
+    {
+      "epoch": 0.7670465906818636,
+      "grad_norm": 4.371986441183748,
+      "learning_rate": 4.818481848184819e-07,
+      "loss": -0.0748,
+      "step": 959
+    },
+    {
+      "epoch": 0.7678464307138573,
+      "grad_norm": 3.611071563569357,
+      "learning_rate": 4.801980198019802e-07,
+      "loss": -0.085,
+      "step": 960
+    },
+    {
+      "epoch": 0.7686462707458508,
+      "grad_norm": 5.1052303944455915,
+      "learning_rate": 4.785478547854785e-07,
+      "loss": 0.0156,
+      "step": 961
+    },
+    {
+      "epoch": 0.7694461107778444,
+      "grad_norm": 4.3911870956367505,
+      "learning_rate": 4.768976897689769e-07,
+      "loss": -0.0685,
+      "step": 962
+    },
+    {
+      "epoch": 0.770245950809838,
+      "grad_norm": 4.259527634413816,
+      "learning_rate": 4.752475247524752e-07,
+      "loss": -0.0837,
+      "step": 963
+    },
+    {
+      "epoch": 0.7710457908418317,
+      "grad_norm": 4.970224413650343,
+      "learning_rate": 4.735973597359736e-07,
+      "loss": -0.1512,
+      "step": 964
+    },
+    {
+      "epoch": 0.7718456308738252,
+      "grad_norm": 5.7445596582962395,
+      "learning_rate": 4.7194719471947193e-07,
+      "loss": -0.0713,
+      "step": 965
+    },
+    {
+      "epoch": 0.7726454709058188,
+      "grad_norm": 4.543014932858551,
+      "learning_rate": 4.7029702970297026e-07,
+      "loss": -0.1337,
+      "step": 966
+    },
+    {
+      "epoch": 0.7734453109378124,
+      "grad_norm": 4.451634421927441,
+      "learning_rate": 4.6864686468646865e-07,
+      "loss": -0.1269,
+      "step": 967
+    },
+    {
+      "epoch": 0.7742451509698061,
+      "grad_norm": 5.379063157263456,
+      "learning_rate": 4.66996699669967e-07,
+      "loss": -0.0862,
+      "step": 968
+    },
+    {
+      "epoch": 0.7750449910017997,
+      "grad_norm": 10.70668137767473,
+      "learning_rate": 4.6534653465346537e-07,
+      "loss": 0.0881,
+      "step": 969
+    },
+    {
+      "epoch": 0.7758448310337932,
+      "grad_norm": 3.7167032202960177,
+      "learning_rate": 4.6369636963696365e-07,
+      "loss": -0.017,
+      "step": 970
+    },
+    {
+      "epoch": 0.7766446710657868,
+      "grad_norm": 4.634801839565637,
+      "learning_rate": 4.62046204620462e-07,
+      "loss": -0.1011,
+      "step": 971
+    },
+    {
+      "epoch": 0.7774445110977805,
+      "grad_norm": 3.7911455543889354,
+      "learning_rate": 4.603960396039604e-07,
+      "loss": -0.1149,
+      "step": 972
+    },
+    {
+      "epoch": 0.7782443511297741,
+      "grad_norm": 3.824528464380088,
+      "learning_rate": 4.587458745874587e-07,
+      "loss": 0.0263,
+      "step": 973
+    },
+    {
+      "epoch": 0.7790441911617676,
+      "grad_norm": 3.978602397648478,
+      "learning_rate": 4.5709570957095705e-07,
+      "loss": -0.0273,
+      "step": 974
+    },
+    {
+      "epoch": 0.7798440311937612,
+      "grad_norm": 3.9315702808225206,
+      "learning_rate": 4.5544554455445543e-07,
+      "loss": -0.1133,
+      "step": 975
+    },
+    {
+      "epoch": 0.7806438712257548,
+      "grad_norm": 4.4594711325427845,
+      "learning_rate": 4.5379537953795377e-07,
+      "loss": -0.0891,
+      "step": 976
+    },
+    {
+      "epoch": 0.7814437112577485,
+      "grad_norm": 4.03305817498308,
+      "learning_rate": 4.521452145214521e-07,
+      "loss": -0.1069,
+      "step": 977
+    },
+    {
+      "epoch": 0.7822435512897421,
+      "grad_norm": 4.444308363632601,
+      "learning_rate": 4.504950495049505e-07,
+      "loss": -0.1274,
+      "step": 978
+    },
+    {
+      "epoch": 0.7830433913217356,
+      "grad_norm": 3.8559207670610953,
+      "learning_rate": 4.488448844884488e-07,
+      "loss": -0.1089,
+      "step": 979
+    },
+    {
+      "epoch": 0.7838432313537292,
+      "grad_norm": 6.852327449815295,
+      "learning_rate": 4.471947194719472e-07,
+      "loss": -0.0608,
+      "step": 980
+    },
+    {
+      "epoch": 0.7846430713857229,
+      "grad_norm": 5.172264061437722,
+      "learning_rate": 4.4554455445544555e-07,
+      "loss": -0.0064,
+      "step": 981
+    },
+    {
+      "epoch": 0.7854429114177165,
+      "grad_norm": 4.338134868672705,
+      "learning_rate": 4.438943894389439e-07,
+      "loss": 0.0261,
+      "step": 982
+    },
+    {
+      "epoch": 0.78624275144971,
+      "grad_norm": 4.7120432908120975,
+      "learning_rate": 4.4224422442244227e-07,
+      "loss": -0.006,
+      "step": 983
+    },
+    {
+      "epoch": 0.7870425914817036,
+      "grad_norm": 4.093388665246742,
+      "learning_rate": 4.405940594059406e-07,
+      "loss": -0.1195,
+      "step": 984
+    },
+    {
+      "epoch": 0.7878424315136973,
+      "grad_norm": 3.9779679788074565,
+      "learning_rate": 4.389438943894389e-07,
+      "loss": -0.0318,
+      "step": 985
+    },
+    {
+      "epoch": 0.7886422715456909,
+      "grad_norm": 5.681211915009173,
+      "learning_rate": 4.3729372937293727e-07,
+      "loss": -0.0715,
+      "step": 986
+    },
+    {
+      "epoch": 0.7894421115776845,
+      "grad_norm": 3.6961307708427875,
+      "learning_rate": 4.356435643564356e-07,
+      "loss": -0.1238,
+      "step": 987
+    },
+    {
+      "epoch": 0.790241951609678,
+      "grad_norm": 5.551081220452864,
+      "learning_rate": 4.3399339933993394e-07,
+      "loss": -0.0353,
+      "step": 988
+    },
+    {
+      "epoch": 0.7910417916416717,
+      "grad_norm": 4.1540822277204725,
+      "learning_rate": 4.3234323432343233e-07,
+      "loss": -0.1044,
+      "step": 989
+    },
+    {
+      "epoch": 0.7918416316736653,
+      "grad_norm": 5.690644788629102,
+      "learning_rate": 4.3069306930693066e-07,
+      "loss": -0.0824,
+      "step": 990
+    },
+    {
+      "epoch": 0.7926414717056589,
+      "grad_norm": 5.090727863244342,
+      "learning_rate": 4.29042904290429e-07,
+      "loss": -0.0275,
+      "step": 991
+    },
+    {
+      "epoch": 0.7934413117376524,
+      "grad_norm": 4.9291753971455705,
+      "learning_rate": 4.273927392739274e-07,
+      "loss": -0.0339,
+      "step": 992
+    },
+    {
+      "epoch": 0.794241151769646,
+      "grad_norm": 6.803186332065206,
+      "learning_rate": 4.257425742574257e-07,
+      "loss": -0.0342,
+      "step": 993
+    },
+    {
+      "epoch": 0.7950409918016397,
+      "grad_norm": 5.804775232239933,
+      "learning_rate": 4.240924092409241e-07,
+      "loss": -0.0866,
+      "step": 994
+    },
+    {
+      "epoch": 0.7958408318336333,
+      "grad_norm": 4.750722930505078,
+      "learning_rate": 4.2244224422442244e-07,
+      "loss": -0.1836,
+      "step": 995
+    },
+    {
+      "epoch": 0.7966406718656269,
+      "grad_norm": 6.716244820472627,
+      "learning_rate": 4.207920792079208e-07,
+      "loss": -0.1024,
+      "step": 996
+    },
+    {
+      "epoch": 0.7974405118976204,
+      "grad_norm": 3.722729260908509,
+      "learning_rate": 4.1914191419141916e-07,
+      "loss": -0.0993,
+      "step": 997
+    },
+    {
+      "epoch": 0.7982403519296141,
+      "grad_norm": 3.311724877987371,
+      "learning_rate": 4.174917491749175e-07,
+      "loss": -0.104,
+      "step": 998
+    },
+    {
+      "epoch": 0.7990401919616077,
+      "grad_norm": 5.521959854449801,
+      "learning_rate": 4.158415841584158e-07,
+      "loss": -0.1918,
+      "step": 999
+    },
+    {
+      "epoch": 0.7998400319936013,
+      "grad_norm": 5.022786099690795,
+      "learning_rate": 4.1419141914191417e-07,
+      "loss": -0.044,
+      "step": 1000
+    },
+    {
+      "epoch": 0.8006398720255948,
+      "grad_norm": 3.85176033301189,
+      "learning_rate": 4.125412541254125e-07,
+      "loss": -0.1389,
+      "step": 1001
+    },
+    {
+      "epoch": 0.8014397120575885,
+      "grad_norm": 3.850320054659798,
+      "learning_rate": 4.1089108910891084e-07,
+      "loss": -0.0621,
+      "step": 1002
+    },
+    {
+      "epoch": 0.8022395520895821,
+      "grad_norm": 5.125417738846334,
+      "learning_rate": 4.092409240924092e-07,
+      "loss": 0.024,
+      "step": 1003
+    },
+    {
+      "epoch": 0.8030393921215757,
+      "grad_norm": 5.550789325018317,
+      "learning_rate": 4.0759075907590756e-07,
+      "loss": -0.0596,
+      "step": 1004
+    },
+    {
+      "epoch": 0.8038392321535693,
+      "grad_norm": 2.987341879008025,
+      "learning_rate": 4.0594059405940595e-07,
+      "loss": 0.0174,
+      "step": 1005
+    },
+    {
+      "epoch": 0.8046390721855629,
+      "grad_norm": 3.840920872539408,
+      "learning_rate": 4.042904290429043e-07,
+      "loss": -0.1812,
+      "step": 1006
+    },
+    {
+      "epoch": 0.8054389122175565,
+      "grad_norm": 4.5904346361674495,
+      "learning_rate": 4.026402640264026e-07,
+      "loss": -0.153,
+      "step": 1007
+    },
+    {
+      "epoch": 0.8062387522495501,
+      "grad_norm": 4.716177536008103,
+      "learning_rate": 4.00990099009901e-07,
+      "loss": -0.0347,
+      "step": 1008
+    },
+    {
+      "epoch": 0.8070385922815437,
+      "grad_norm": 3.8819770758540106,
+      "learning_rate": 3.9933993399339934e-07,
+      "loss": -0.0491,
+      "step": 1009
+    },
+    {
+      "epoch": 0.8078384323135372,
+      "grad_norm": 4.337114713855018,
+      "learning_rate": 3.9768976897689767e-07,
+      "loss": -0.076,
+      "step": 1010
+    },
+    {
+      "epoch": 0.8086382723455309,
+      "grad_norm": 5.085958876323165,
+      "learning_rate": 3.9603960396039606e-07,
+      "loss": -0.1292,
+      "step": 1011
+    },
+    {
+      "epoch": 0.8094381123775245,
+      "grad_norm": 4.225428966888881,
+      "learning_rate": 3.943894389438944e-07,
+      "loss": -0.0349,
+      "step": 1012
+    },
+    {
+      "epoch": 0.8102379524095181,
+      "grad_norm": 3.132492848210798,
+      "learning_rate": 3.927392739273927e-07,
+      "loss": 0.0267,
+      "step": 1013
+    },
+    {
+      "epoch": 0.8110377924415118,
+      "grad_norm": 4.728427487496938,
+      "learning_rate": 3.9108910891089106e-07,
+      "loss": 0.0251,
+      "step": 1014
+    },
+    {
+      "epoch": 0.8118376324735053,
+      "grad_norm": 3.231169950869779,
+      "learning_rate": 3.894389438943894e-07,
+      "loss": -0.0048,
+      "step": 1015
+    },
+    {
+      "epoch": 0.8126374725054989,
+      "grad_norm": 5.4315342239443645,
+      "learning_rate": 3.8778877887788773e-07,
+      "loss": 0.0017,
+      "step": 1016
+    },
+    {
+      "epoch": 0.8134373125374925,
+      "grad_norm": 4.1495886173643015,
+      "learning_rate": 3.861386138613861e-07,
+      "loss": -0.0606,
+      "step": 1017
+    },
+    {
+      "epoch": 0.8142371525694861,
+      "grad_norm": 4.571814448385221,
+      "learning_rate": 3.8448844884488445e-07,
+      "loss": 0.0305,
+      "step": 1018
+    },
+    {
+      "epoch": 0.8150369926014797,
+      "grad_norm": 5.376749508040782,
+      "learning_rate": 3.8283828382838284e-07,
+      "loss": -0.0529,
+      "step": 1019
+    },
+    {
+      "epoch": 0.8158368326334733,
+      "grad_norm": 4.234529949886336,
+      "learning_rate": 3.811881188118812e-07,
+      "loss": -0.0884,
+      "step": 1020
+    },
+    {
+      "epoch": 0.8166366726654669,
+      "grad_norm": 5.103075536241735,
+      "learning_rate": 3.795379537953795e-07,
+      "loss": -0.0217,
+      "step": 1021
+    },
+    {
+      "epoch": 0.8174365126974605,
+      "grad_norm": 4.089650205096865,
+      "learning_rate": 3.778877887788779e-07,
+      "loss": -0.1912,
+      "step": 1022
+    },
+    {
+      "epoch": 0.8182363527294542,
+      "grad_norm": 4.315512669100621,
+      "learning_rate": 3.7623762376237623e-07,
+      "loss": -0.1096,
+      "step": 1023
+    },
+    {
+      "epoch": 0.8190361927614477,
+      "grad_norm": 5.730342268116768,
+      "learning_rate": 3.7458745874587457e-07,
+      "loss": -0.0613,
+      "step": 1024
+    },
+    {
+      "epoch": 0.8198360327934413,
+      "grad_norm": 4.482232090208855,
+      "learning_rate": 3.7293729372937295e-07,
+      "loss": -0.0025,
+      "step": 1025
+    },
+    {
+      "epoch": 0.8206358728254349,
+      "grad_norm": 4.242971330310737,
+      "learning_rate": 3.712871287128713e-07,
+      "loss": -0.0786,
+      "step": 1026
+    },
+    {
+      "epoch": 0.8214357128574286,
+      "grad_norm": 4.11563596499524,
+      "learning_rate": 3.696369636963696e-07,
+      "loss": -0.082,
+      "step": 1027
+    },
+    {
+      "epoch": 0.8222355528894221,
+      "grad_norm": 9.332422439085308,
+      "learning_rate": 3.6798679867986796e-07,
+      "loss": 0.0022,
+      "step": 1028
+    },
+    {
+      "epoch": 0.8230353929214157,
+      "grad_norm": 4.067536269426566,
+      "learning_rate": 3.663366336633663e-07,
+      "loss": -0.0361,
+      "step": 1029
+    },
+    {
+      "epoch": 0.8238352329534093,
+      "grad_norm": 6.044500944552922,
+      "learning_rate": 3.6468646864686463e-07,
+      "loss": 0.1496,
+      "step": 1030
+    },
+    {
+      "epoch": 0.824635072985403,
+      "grad_norm": 4.30205099295344,
+      "learning_rate": 3.63036303630363e-07,
+      "loss": -0.045,
+      "step": 1031
+    },
+    {
+      "epoch": 0.8254349130173965,
+      "grad_norm": 4.012282824952302,
+      "learning_rate": 3.6138613861386135e-07,
+      "loss": -0.1718,
+      "step": 1032
+    },
+    {
+      "epoch": 0.8262347530493901,
+      "grad_norm": 4.159359564634324,
+      "learning_rate": 3.5973597359735974e-07,
+      "loss": -0.0454,
+      "step": 1033
+    },
+    {
+      "epoch": 0.8270345930813837,
+      "grad_norm": 3.814387620245282,
+      "learning_rate": 3.5808580858085807e-07,
+      "loss": -0.0466,
+      "step": 1034
+    },
+    {
+      "epoch": 0.8278344331133773,
+      "grad_norm": 6.904216207979962,
+      "learning_rate": 3.564356435643564e-07,
+      "loss": -0.1518,
+      "step": 1035
+    },
+    {
+      "epoch": 0.828634273145371,
+      "grad_norm": 5.250865501632598,
+      "learning_rate": 3.547854785478548e-07,
+      "loss": -0.0118,
+      "step": 1036
+    },
+    {
+      "epoch": 0.8294341131773645,
+      "grad_norm": 5.846194811534598,
+      "learning_rate": 3.5313531353135313e-07,
+      "loss": 0.0351,
+      "step": 1037
+    },
+    {
+      "epoch": 0.8302339532093581,
+      "grad_norm": 4.460630924092106,
+      "learning_rate": 3.5148514851485146e-07,
+      "loss": -0.2443,
+      "step": 1038
+    },
+    {
+      "epoch": 0.8310337932413517,
+      "grad_norm": 4.1447157895119995,
+      "learning_rate": 3.4983498349834985e-07,
+      "loss": -0.0134,
+      "step": 1039
+    },
+    {
+      "epoch": 0.8318336332733454,
+      "grad_norm": 3.5252220895483517,
+      "learning_rate": 3.481848184818482e-07,
+      "loss": -0.0577,
+      "step": 1040
+    },
+    {
+      "epoch": 0.8326334733053389,
+      "grad_norm": 4.361902741479118,
+      "learning_rate": 3.465346534653465e-07,
+      "loss": 0.0104,
+      "step": 1041
+    },
+    {
+      "epoch": 0.8334333133373325,
+      "grad_norm": 4.461915879260683,
+      "learning_rate": 3.4488448844884485e-07,
+      "loss": -0.0156,
+      "step": 1042
+    },
+    {
+      "epoch": 0.8342331533693261,
+      "grad_norm": 4.834838939615413,
+      "learning_rate": 3.432343234323432e-07,
+      "loss": -0.0306,
+      "step": 1043
+    },
+    {
+      "epoch": 0.8350329934013198,
+      "grad_norm": 4.457492333115142,
+      "learning_rate": 3.415841584158416e-07,
+      "loss": -0.0158,
+      "step": 1044
+    },
+    {
+      "epoch": 0.8358328334333134,
+      "grad_norm": 6.418129824325349,
+      "learning_rate": 3.399339933993399e-07,
+      "loss": 0.023,
+      "step": 1045
+    },
+    {
+      "epoch": 0.8366326734653069,
+      "grad_norm": 5.631846859681406,
+      "learning_rate": 3.3828382838283824e-07,
+      "loss": -0.0842,
+      "step": 1046
+    },
+    {
+      "epoch": 0.8374325134973005,
+      "grad_norm": 4.893647743608584,
+      "learning_rate": 3.3663366336633663e-07,
+      "loss": 0.0169,
+      "step": 1047
+    },
+    {
+      "epoch": 0.8382323535292941,
+      "grad_norm": 3.327224537992695,
+      "learning_rate": 3.3498349834983497e-07,
+      "loss": 0.0234,
+      "step": 1048
+    },
+    {
+      "epoch": 0.8390321935612878,
+      "grad_norm": 3.334769765979331,
+      "learning_rate": 3.333333333333333e-07,
+      "loss": -0.0878,
+      "step": 1049
+    },
+    {
+      "epoch": 0.8398320335932813,
+      "grad_norm": 4.036389763362471,
+      "learning_rate": 3.316831683168317e-07,
+      "loss": -0.0402,
+      "step": 1050
+    },
+    {
+      "epoch": 0.8406318736252749,
+      "grad_norm": 3.653874204118681,
+      "learning_rate": 3.3003300330033e-07,
+      "loss": -0.0631,
+      "step": 1051
+    },
+    {
+      "epoch": 0.8414317136572685,
+      "grad_norm": 4.88359310166619,
+      "learning_rate": 3.2838283828382836e-07,
+      "loss": -0.0544,
+      "step": 1052
+    },
+    {
+      "epoch": 0.8422315536892622,
+      "grad_norm": 6.462333703622296,
+      "learning_rate": 3.2673267326732674e-07,
+      "loss": -0.0678,
+      "step": 1053
+    },
+    {
+      "epoch": 0.8430313937212558,
+      "grad_norm": 7.009021395345441,
+      "learning_rate": 3.250825082508251e-07,
+      "loss": 0.0099,
+      "step": 1054
+    },
+    {
+      "epoch": 0.8438312337532493,
+      "grad_norm": 4.2767377032125875,
+      "learning_rate": 3.234323432343234e-07,
+      "loss": 0.0297,
+      "step": 1055
+    },
+    {
+      "epoch": 0.8446310737852429,
+      "grad_norm": 9.79641552202019,
+      "learning_rate": 3.217821782178218e-07,
+      "loss": 0.0968,
+      "step": 1056
+    },
+    {
+      "epoch": 0.8454309138172366,
+      "grad_norm": 4.669605737417231,
+      "learning_rate": 3.201320132013201e-07,
+      "loss": 0.0097,
+      "step": 1057
+    },
+    {
+      "epoch": 0.8462307538492302,
+      "grad_norm": 3.1819061861624807,
+      "learning_rate": 3.1848184818481847e-07,
+      "loss": -0.0776,
+      "step": 1058
+    },
+    {
+      "epoch": 0.8470305938812237,
+      "grad_norm": 4.1774987880629695,
+      "learning_rate": 3.168316831683168e-07,
+      "loss": -0.0806,
+      "step": 1059
+    },
+    {
+      "epoch": 0.8478304339132173,
+      "grad_norm": 4.454569906758588,
+      "learning_rate": 3.1518151815181514e-07,
+      "loss": -0.0526,
+      "step": 1060
+    },
+    {
+      "epoch": 0.848630273945211,
+      "grad_norm": 3.527299815228531,
+      "learning_rate": 3.1353135313531353e-07,
+      "loss": -0.1065,
+      "step": 1061
+    },
+    {
+      "epoch": 0.8494301139772046,
+      "grad_norm": 4.481801002071373,
+      "learning_rate": 3.1188118811881186e-07,
+      "loss": -0.0022,
+      "step": 1062
+    },
+    {
+      "epoch": 0.8502299540091982,
+      "grad_norm": 4.3147168197624755,
+      "learning_rate": 3.102310231023102e-07,
+      "loss": -0.011,
+      "step": 1063
+    },
+    {
+      "epoch": 0.8510297940411917,
+      "grad_norm": 3.812340279657359,
+      "learning_rate": 3.085808580858086e-07,
+      "loss": -0.2093,
+      "step": 1064
+    },
+    {
+      "epoch": 0.8518296340731853,
+      "grad_norm": 5.097295358094463,
+      "learning_rate": 3.069306930693069e-07,
+      "loss": -0.0048,
+      "step": 1065
+    },
+    {
+      "epoch": 0.852629474105179,
+      "grad_norm": 5.088642578790314,
+      "learning_rate": 3.0528052805280525e-07,
+      "loss": -0.1315,
+      "step": 1066
+    },
+    {
+      "epoch": 0.8534293141371726,
+      "grad_norm": 4.052723785754238,
+      "learning_rate": 3.0363036303630364e-07,
+      "loss": -0.1133,
+      "step": 1067
+    },
+    {
+      "epoch": 0.8542291541691661,
+      "grad_norm": 5.193579179546016,
+      "learning_rate": 3.01980198019802e-07,
+      "loss": -0.0787,
+      "step": 1068
+    },
+    {
+      "epoch": 0.8550289942011597,
+      "grad_norm": 3.030054387526671,
+      "learning_rate": 3.0033003300330036e-07,
+      "loss": -0.0677,
+      "step": 1069
+    },
+    {
+      "epoch": 0.8558288342331534,
+      "grad_norm": 9.135373007054904,
+      "learning_rate": 2.986798679867987e-07,
+      "loss": -0.1171,
+      "step": 1070
+    },
+    {
+      "epoch": 0.856628674265147,
+      "grad_norm": 3.3785319743939013,
+      "learning_rate": 2.97029702970297e-07,
+      "loss": -0.0803,
+      "step": 1071
+    },
+    {
+      "epoch": 0.8574285142971406,
+      "grad_norm": 3.9077882713350762,
+      "learning_rate": 2.9537953795379537e-07,
+      "loss": -0.0513,
+      "step": 1072
+    },
+    {
+      "epoch": 0.8582283543291341,
+      "grad_norm": 4.038560493011451,
+      "learning_rate": 2.937293729372937e-07,
+      "loss": -0.0518,
+      "step": 1073
+    },
+    {
+      "epoch": 0.8590281943611278,
+      "grad_norm": 3.800775478942818,
+      "learning_rate": 2.9207920792079203e-07,
+      "loss": -0.1613,
+      "step": 1074
+    },
+    {
+      "epoch": 0.8598280343931214,
+      "grad_norm": 5.948071515082444,
+      "learning_rate": 2.904290429042904e-07,
+      "loss": -0.0834,
+      "step": 1075
+    },
+    {
+      "epoch": 0.860627874425115,
+      "grad_norm": 4.190899280558429,
+      "learning_rate": 2.8877887788778876e-07,
+      "loss": -0.103,
+      "step": 1076
+    },
+    {
+      "epoch": 0.8614277144571085,
+      "grad_norm": 4.910742773988997,
+      "learning_rate": 2.871287128712871e-07,
+      "loss": -0.0388,
+      "step": 1077
+    },
+    {
+      "epoch": 0.8622275544891022,
+      "grad_norm": 7.707730353888358,
+      "learning_rate": 2.854785478547855e-07,
+      "loss": -0.1477,
+      "step": 1078
+    },
+    {
+      "epoch": 0.8630273945210958,
+      "grad_norm": 4.297638426499125,
+      "learning_rate": 2.838283828382838e-07,
+      "loss": 0.0134,
+      "step": 1079
+    },
+    {
+      "epoch": 0.8638272345530894,
+      "grad_norm": 6.145968919540754,
+      "learning_rate": 2.8217821782178215e-07,
+      "loss": 0.0021,
+      "step": 1080
+    },
+    {
+      "epoch": 0.864627074585083,
+      "grad_norm": 3.7854604687393296,
+      "learning_rate": 2.8052805280528054e-07,
+      "loss": -0.1547,
+      "step": 1081
+    },
+    {
+      "epoch": 0.8654269146170765,
+      "grad_norm": 3.991674461171312,
+      "learning_rate": 2.7887788778877887e-07,
+      "loss": -0.0668,
+      "step": 1082
+    },
+    {
+      "epoch": 0.8662267546490702,
+      "grad_norm": 4.795940892228349,
+      "learning_rate": 2.7722772277227726e-07,
+      "loss": 0.0146,
+      "step": 1083
+    },
+    {
+      "epoch": 0.8670265946810638,
+      "grad_norm": 4.944098976058084,
+      "learning_rate": 2.755775577557756e-07,
+      "loss": -0.0523,
+      "step": 1084
+    },
+    {
+      "epoch": 0.8678264347130574,
+      "grad_norm": 5.005504629817167,
+      "learning_rate": 2.7392739273927387e-07,
+      "loss": 0.0065,
+      "step": 1085
+    },
+    {
+      "epoch": 0.8686262747450509,
+      "grad_norm": 4.33635141957305,
+      "learning_rate": 2.7227722772277226e-07,
+      "loss": 0.0051,
+      "step": 1086
+    },
+    {
+      "epoch": 0.8694261147770446,
+      "grad_norm": 6.827221005304679,
+      "learning_rate": 2.706270627062706e-07,
+      "loss": -0.1068,
+      "step": 1087
+    },
+    {
+      "epoch": 0.8702259548090382,
+      "grad_norm": 3.763012118037954,
+      "learning_rate": 2.6897689768976893e-07,
+      "loss": 0.0675,
+      "step": 1088
+    },
+    {
+      "epoch": 0.8710257948410318,
+      "grad_norm": 3.476547412625268,
+      "learning_rate": 2.673267326732673e-07,
+      "loss": -0.1067,
+      "step": 1089
+    },
+    {
+      "epoch": 0.8718256348730254,
+      "grad_norm": 3.622631746348685,
+      "learning_rate": 2.6567656765676565e-07,
+      "loss": -0.0829,
+      "step": 1090
+    },
+    {
+      "epoch": 0.872625474905019,
+      "grad_norm": 4.835180762619133,
+      "learning_rate": 2.64026402640264e-07,
+      "loss": -0.0761,
+      "step": 1091
+    },
+    {
+      "epoch": 0.8734253149370126,
+      "grad_norm": 4.188641976033946,
+      "learning_rate": 2.623762376237624e-07,
+      "loss": -0.114,
+      "step": 1092
+    },
+    {
+      "epoch": 0.8742251549690062,
+      "grad_norm": 4.833712113544916,
+      "learning_rate": 2.607260726072607e-07,
+      "loss": -0.1158,
+      "step": 1093
+    },
+    {
+      "epoch": 0.8750249950009998,
+      "grad_norm": 3.6494839656219935,
+      "learning_rate": 2.590759075907591e-07,
+      "loss": -0.048,
+      "step": 1094
+    },
+    {
+      "epoch": 0.8758248350329934,
+      "grad_norm": 3.9306302162750857,
+      "learning_rate": 2.5742574257425743e-07,
+      "loss": -0.0928,
+      "step": 1095
+    },
+    {
+      "epoch": 0.876624675064987,
+      "grad_norm": 3.7010390446563517,
+      "learning_rate": 2.5577557755775576e-07,
+      "loss": 0.0242,
+      "step": 1096
+    },
+    {
+      "epoch": 0.8774245150969806,
+      "grad_norm": 3.641273539002507,
+      "learning_rate": 2.5412541254125415e-07,
+      "loss": -0.1014,
+      "step": 1097
+    },
+    {
+      "epoch": 0.8782243551289742,
+      "grad_norm": 4.233409363271656,
+      "learning_rate": 2.524752475247525e-07,
+      "loss": -0.0404,
+      "step": 1098
+    },
+    {
+      "epoch": 0.8790241951609679,
+      "grad_norm": 4.188973466495453,
+      "learning_rate": 2.508250825082508e-07,
+      "loss": -0.0684,
+      "step": 1099
+    },
+    {
+      "epoch": 0.8798240351929614,
+      "grad_norm": 5.017584397195866,
+      "learning_rate": 2.4917491749174916e-07,
+      "loss": -0.0368,
+      "step": 1100
+    },
+    {
+      "epoch": 0.880623875224955,
+      "grad_norm": 3.9510700176873566,
+      "learning_rate": 2.475247524752475e-07,
+      "loss": -0.1375,
+      "step": 1101
+    },
+    {
+      "epoch": 0.8814237152569486,
+      "grad_norm": 5.84233851394486,
+      "learning_rate": 2.458745874587459e-07,
+      "loss": -0.1067,
+      "step": 1102
+    },
+    {
+      "epoch": 0.8822235552889423,
+      "grad_norm": 5.406949565806744,
+      "learning_rate": 2.442244224422442e-07,
+      "loss": -0.0962,
+      "step": 1103
+    },
+    {
+      "epoch": 0.8830233953209358,
+      "grad_norm": 5.315262379239265,
+      "learning_rate": 2.4257425742574255e-07,
+      "loss": -0.066,
+      "step": 1104
+    },
+    {
+      "epoch": 0.8838232353529294,
+      "grad_norm": 4.836530658291514,
+      "learning_rate": 2.4092409240924093e-07,
+      "loss": -0.0646,
+      "step": 1105
+    },
+    {
+      "epoch": 0.884623075384923,
+      "grad_norm": 3.293455547222145,
+      "learning_rate": 2.3927392739273927e-07,
+      "loss": -0.1701,
+      "step": 1106
+    },
+    {
+      "epoch": 0.8854229154169166,
+      "grad_norm": 4.709525078481242,
+      "learning_rate": 2.376237623762376e-07,
+      "loss": -0.0504,
+      "step": 1107
+    },
+    {
+      "epoch": 0.8862227554489103,
+      "grad_norm": 4.295657231556702,
+      "learning_rate": 2.3597359735973596e-07,
+      "loss": -0.1419,
+      "step": 1108
+    },
+    {
+      "epoch": 0.8870225954809038,
+      "grad_norm": 3.682029286721376,
+      "learning_rate": 2.3432343234323433e-07,
+      "loss": -0.0927,
+      "step": 1109
+    },
+    {
+      "epoch": 0.8878224355128974,
+      "grad_norm": 7.500929711256007,
+      "learning_rate": 2.3267326732673269e-07,
+      "loss": -0.001,
+      "step": 1110
+    },
+    {
+      "epoch": 0.888622275544891,
+      "grad_norm": 3.370577280876358,
+      "learning_rate": 2.31023102310231e-07,
+      "loss": -0.0219,
+      "step": 1111
+    },
+    {
+      "epoch": 0.8894221155768847,
+      "grad_norm": 4.603247549338215,
+      "learning_rate": 2.2937293729372936e-07,
+      "loss": -0.0407,
+      "step": 1112
+    },
+    {
+      "epoch": 0.8902219556088782,
+      "grad_norm": 3.033292259385364,
+      "learning_rate": 2.2772277227722772e-07,
+      "loss": -0.0612,
+      "step": 1113
+    },
+    {
+      "epoch": 0.8910217956408718,
+      "grad_norm": 5.654397566299044,
+      "learning_rate": 2.2607260726072605e-07,
+      "loss": -0.0081,
+      "step": 1114
+    },
+    {
+      "epoch": 0.8918216356728654,
+      "grad_norm": 4.7463861069291235,
+      "learning_rate": 2.244224422442244e-07,
+      "loss": 0.0324,
+      "step": 1115
+    },
+    {
+      "epoch": 0.8926214757048591,
+      "grad_norm": 3.6795751469461697,
+      "learning_rate": 2.2277227722772277e-07,
+      "loss": -0.082,
+      "step": 1116
+    },
+    {
+      "epoch": 0.8934213157368527,
+      "grad_norm": 6.7304826361036385,
+      "learning_rate": 2.2112211221122113e-07,
+      "loss": -0.1421,
+      "step": 1117
+    },
+    {
+      "epoch": 0.8942211557688462,
+      "grad_norm": 4.6084848352584,
+      "learning_rate": 2.1947194719471944e-07,
+      "loss": -0.0154,
+      "step": 1118
+    },
+    {
+      "epoch": 0.8950209958008398,
+      "grad_norm": 3.3241186250160673,
+      "learning_rate": 2.178217821782178e-07,
+      "loss": -0.0463,
+      "step": 1119
+    },
+    {
+      "epoch": 0.8958208358328335,
+      "grad_norm": 4.285630706698749,
+      "learning_rate": 2.1617161716171616e-07,
+      "loss": -0.0102,
+      "step": 1120
+    },
+    {
+      "epoch": 0.8966206758648271,
+      "grad_norm": 3.7442923024099266,
+      "learning_rate": 2.145214521452145e-07,
+      "loss": -0.0737,
+      "step": 1121
+    },
+    {
+      "epoch": 0.8974205158968206,
+      "grad_norm": 4.067618329578387,
+      "learning_rate": 2.1287128712871286e-07,
+      "loss": -0.0694,
+      "step": 1122
+    },
+    {
+      "epoch": 0.8982203559288142,
+      "grad_norm": 4.012101702069505,
+      "learning_rate": 2.1122112211221122e-07,
+      "loss": 0.013,
+      "step": 1123
+    },
+    {
+      "epoch": 0.8990201959608078,
+      "grad_norm": 3.927004556177739,
+      "learning_rate": 2.0957095709570958e-07,
+      "loss": 0.0032,
+      "step": 1124
+    },
+    {
+      "epoch": 0.8998200359928015,
+      "grad_norm": 4.153485289274271,
+      "learning_rate": 2.079207920792079e-07,
+      "loss": 0.0691,
+      "step": 1125
+    },
+    {
+      "epoch": 0.9006198760247951,
+      "grad_norm": 3.0852405388784936,
+      "learning_rate": 2.0627062706270625e-07,
+      "loss": -0.152,
+      "step": 1126
+    },
+    {
+      "epoch": 0.9014197160567886,
+      "grad_norm": 5.668232907029322,
+      "learning_rate": 2.046204620462046e-07,
+      "loss": -0.0051,
+      "step": 1127
+    },
+    {
+      "epoch": 0.9022195560887822,
+      "grad_norm": 4.016353081034237,
+      "learning_rate": 2.0297029702970297e-07,
+      "loss": 0.004,
+      "step": 1128
+    },
+    {
+      "epoch": 0.9030193961207759,
+      "grad_norm": 5.222331571223233,
+      "learning_rate": 2.013201320132013e-07,
+      "loss": -0.0398,
+      "step": 1129
+    },
+    {
+      "epoch": 0.9038192361527695,
+      "grad_norm": 6.050674073050146,
+      "learning_rate": 1.9966996699669967e-07,
+      "loss": -0.0201,
+      "step": 1130
+    },
+    {
+      "epoch": 0.904619076184763,
+      "grad_norm": 3.6129712440915336,
+      "learning_rate": 1.9801980198019803e-07,
+      "loss": -0.0873,
+      "step": 1131
+    },
+    {
+      "epoch": 0.9054189162167566,
+      "grad_norm": 3.9706646963831527,
+      "learning_rate": 1.9636963696369634e-07,
+      "loss": -0.0425,
+      "step": 1132
+    },
+    {
+      "epoch": 0.9062187562487503,
+      "grad_norm": 3.575800911347383,
+      "learning_rate": 1.947194719471947e-07,
+      "loss": -0.0728,
+      "step": 1133
+    },
+    {
+      "epoch": 0.9070185962807439,
+      "grad_norm": 4.9127144958415165,
+      "learning_rate": 1.9306930693069306e-07,
+      "loss": -0.1456,
+      "step": 1134
+    },
+    {
+      "epoch": 0.9078184363127374,
+      "grad_norm": 4.630716245217967,
+      "learning_rate": 1.9141914191419142e-07,
+      "loss": -0.1385,
+      "step": 1135
+    },
+    {
+      "epoch": 0.908618276344731,
+      "grad_norm": 3.7522928222148413,
+      "learning_rate": 1.8976897689768976e-07,
+      "loss": -0.0473,
+      "step": 1136
+    },
+    {
+      "epoch": 0.9094181163767247,
+      "grad_norm": 3.702942791411621,
+      "learning_rate": 1.8811881188118812e-07,
+      "loss": -0.0115,
+      "step": 1137
+    },
+    {
+      "epoch": 0.9102179564087183,
+      "grad_norm": 4.215778349737591,
+      "learning_rate": 1.8646864686468648e-07,
+      "loss": 0.0369,
+      "step": 1138
+    },
+    {
+      "epoch": 0.9110177964407119,
+      "grad_norm": 4.312725558809124,
+      "learning_rate": 1.848184818481848e-07,
+      "loss": -0.0821,
+      "step": 1139
+    },
+    {
+      "epoch": 0.9118176364727054,
+      "grad_norm": 4.321449833697151,
+      "learning_rate": 1.8316831683168315e-07,
+      "loss": -0.1315,
+      "step": 1140
+    },
+    {
+      "epoch": 0.912617476504699,
+      "grad_norm": 3.2649224360601234,
+      "learning_rate": 1.815181518151815e-07,
+      "loss": -0.1925,
+      "step": 1141
+    },
+    {
+      "epoch": 0.9134173165366927,
+      "grad_norm": 3.760790622253671,
+      "learning_rate": 1.7986798679867987e-07,
+      "loss": -0.0013,
+      "step": 1142
+    },
+    {
+      "epoch": 0.9142171565686863,
+      "grad_norm": 4.157207407424451,
+      "learning_rate": 1.782178217821782e-07,
+      "loss": -0.0581,
+      "step": 1143
+    },
+    {
+      "epoch": 0.9150169966006798,
+      "grad_norm": 5.763265152647982,
+      "learning_rate": 1.7656765676567656e-07,
+      "loss": -0.0789,
+      "step": 1144
+    },
+    {
+      "epoch": 0.9158168366326734,
+      "grad_norm": 5.000391563342132,
+      "learning_rate": 1.7491749174917492e-07,
+      "loss": 0.017,
+      "step": 1145
+    },
+    {
+      "epoch": 0.9166166766646671,
+      "grad_norm": 5.503835207911807,
+      "learning_rate": 1.7326732673267326e-07,
+      "loss": -0.0498,
+      "step": 1146
+    },
+    {
+      "epoch": 0.9174165166966607,
+      "grad_norm": 4.892439922885906,
+      "learning_rate": 1.716171617161716e-07,
+      "loss": -0.0188,
+      "step": 1147
+    },
+    {
+      "epoch": 0.9182163567286543,
+      "grad_norm": 4.1251081972670915,
+      "learning_rate": 1.6996699669966995e-07,
+      "loss": -0.1601,
+      "step": 1148
+    },
+    {
+      "epoch": 0.9190161967606478,
+      "grad_norm": 5.234413078511215,
+      "learning_rate": 1.6831683168316832e-07,
+      "loss": -0.0751,
+      "step": 1149
+    },
+    {
+      "epoch": 0.9198160367926415,
+      "grad_norm": 5.21079968866447,
+      "learning_rate": 1.6666666666666665e-07,
+      "loss": 0.0173,
+      "step": 1150
+    },
+    {
+      "epoch": 0.9206158768246351,
+      "grad_norm": 4.287315805109782,
+      "learning_rate": 1.65016501650165e-07,
+      "loss": -0.0245,
+      "step": 1151
+    },
+    {
+      "epoch": 0.9214157168566287,
+      "grad_norm": 2.89403897319458,
+      "learning_rate": 1.6336633663366337e-07,
+      "loss": -0.1234,
+      "step": 1152
+    },
+    {
+      "epoch": 0.9222155568886222,
+      "grad_norm": 5.421448264794362,
+      "learning_rate": 1.617161716171617e-07,
+      "loss": -0.0738,
+      "step": 1153
+    },
+    {
+      "epoch": 0.9230153969206158,
+      "grad_norm": 5.340841829330172,
+      "learning_rate": 1.6006600660066004e-07,
+      "loss": -0.0978,
+      "step": 1154
+    },
+    {
+      "epoch": 0.9238152369526095,
+      "grad_norm": 5.139659477858416,
+      "learning_rate": 1.584158415841584e-07,
+      "loss": -0.0053,
+      "step": 1155
+    },
+    {
+      "epoch": 0.9246150769846031,
+      "grad_norm": 3.5692046134784676,
+      "learning_rate": 1.5676567656765676e-07,
+      "loss": -0.078,
+      "step": 1156
+    },
+    {
+      "epoch": 0.9254149170165967,
+      "grad_norm": 3.6214949664373624,
+      "learning_rate": 1.551155115511551e-07,
+      "loss": -0.1948,
+      "step": 1157
+    },
+    {
+      "epoch": 0.9262147570485902,
+      "grad_norm": 3.6462100170455516,
+      "learning_rate": 1.5346534653465346e-07,
+      "loss": -0.1466,
+      "step": 1158
+    },
+    {
+      "epoch": 0.9270145970805839,
+      "grad_norm": 4.042490583513813,
+      "learning_rate": 1.5181518151815182e-07,
+      "loss": -0.0494,
+      "step": 1159
+    },
+    {
+      "epoch": 0.9278144371125775,
+      "grad_norm": 3.570518304081052,
+      "learning_rate": 1.5016501650165018e-07,
+      "loss": -0.0203,
+      "step": 1160
+    },
+    {
+      "epoch": 0.9286142771445711,
+      "grad_norm": 4.860609480391736,
+      "learning_rate": 1.485148514851485e-07,
+      "loss": -0.0608,
+      "step": 1161
+    },
+    {
+      "epoch": 0.9294141171765646,
+      "grad_norm": 3.6503782154768336,
+      "learning_rate": 1.4686468646864685e-07,
+      "loss": -0.1209,
+      "step": 1162
+    },
+    {
+      "epoch": 0.9302139572085583,
+      "grad_norm": 5.2679977730736915,
+      "learning_rate": 1.452145214521452e-07,
+      "loss": -0.2231,
+      "step": 1163
+    },
+    {
+      "epoch": 0.9310137972405519,
+      "grad_norm": 4.077205627405786,
+      "learning_rate": 1.4356435643564355e-07,
+      "loss": -0.0918,
+      "step": 1164
+    },
+    {
+      "epoch": 0.9318136372725455,
+      "grad_norm": 4.7895431899614245,
+      "learning_rate": 1.419141914191419e-07,
+      "loss": -0.0527,
+      "step": 1165
+    },
+    {
+      "epoch": 0.9326134773045391,
+      "grad_norm": 5.279018314307402,
+      "learning_rate": 1.4026402640264027e-07,
+      "loss": 0.0147,
+      "step": 1166
+    },
+    {
+      "epoch": 0.9334133173365327,
+      "grad_norm": 4.336258829943017,
+      "learning_rate": 1.3861386138613863e-07,
+      "loss": -0.0938,
+      "step": 1167
+    },
+    {
+      "epoch": 0.9342131573685263,
+      "grad_norm": 4.54870086400182,
+      "learning_rate": 1.3696369636963694e-07,
+      "loss": -0.1337,
+      "step": 1168
+    },
+    {
+      "epoch": 0.9350129974005199,
+      "grad_norm": 4.5996184978678105,
+      "learning_rate": 1.353135313531353e-07,
+      "loss": -0.0564,
+      "step": 1169
+    },
+    {
+      "epoch": 0.9358128374325135,
+      "grad_norm": 3.580175353715861,
+      "learning_rate": 1.3366336633663366e-07,
+      "loss": -0.0917,
+      "step": 1170
+    },
+    {
+      "epoch": 0.936612677464507,
+      "grad_norm": 4.236520821746979,
+      "learning_rate": 1.32013201320132e-07,
+      "loss": -0.0952,
+      "step": 1171
+    },
+    {
+      "epoch": 0.9374125174965007,
+      "grad_norm": 4.45059914145225,
+      "learning_rate": 1.3036303630363035e-07,
+      "loss": -0.0501,
+      "step": 1172
+    },
+    {
+      "epoch": 0.9382123575284943,
+      "grad_norm": 3.895112295625407,
+      "learning_rate": 1.2871287128712872e-07,
+      "loss": -0.0248,
+      "step": 1173
+    },
+    {
+      "epoch": 0.9390121975604879,
+      "grad_norm": 3.446049410323525,
+      "learning_rate": 1.2706270627062708e-07,
+      "loss": -0.0609,
+      "step": 1174
+    },
+    {
+      "epoch": 0.9398120375924816,
+      "grad_norm": 4.440477903266653,
+      "learning_rate": 1.254125412541254e-07,
+      "loss": -0.1183,
+      "step": 1175
+    },
+    {
+      "epoch": 0.9406118776244751,
+      "grad_norm": 8.879875026201496,
+      "learning_rate": 1.2376237623762375e-07,
+      "loss": 0.0741,
+      "step": 1176
+    },
+    {
+      "epoch": 0.9414117176564687,
+      "grad_norm": 3.251289551995566,
+      "learning_rate": 1.221122112211221e-07,
+      "loss": -0.2012,
+      "step": 1177
+    },
+    {
+      "epoch": 0.9422115576884623,
+      "grad_norm": 5.090866489665643,
+      "learning_rate": 1.2046204620462047e-07,
+      "loss": -0.0937,
+      "step": 1178
+    },
+    {
+      "epoch": 0.943011397720456,
+      "grad_norm": 5.634824692335556,
+      "learning_rate": 1.188118811881188e-07,
+      "loss": -0.0674,
+      "step": 1179
+    },
+    {
+      "epoch": 0.9438112377524495,
+      "grad_norm": 7.237946586468722,
+      "learning_rate": 1.1716171617161716e-07,
+      "loss": 0.0063,
+      "step": 1180
+    },
+    {
+      "epoch": 0.9446110777844431,
+      "grad_norm": 5.322093424679851,
+      "learning_rate": 1.155115511551155e-07,
+      "loss": 0.0182,
+      "step": 1181
+    },
+    {
+      "epoch": 0.9454109178164367,
+      "grad_norm": 5.685444219272491,
+      "learning_rate": 1.1386138613861386e-07,
+      "loss": -0.0409,
+      "step": 1182
+    },
+    {
+      "epoch": 0.9462107578484303,
+      "grad_norm": 4.89343356574685,
+      "learning_rate": 1.122112211221122e-07,
+      "loss": -0.1669,
+      "step": 1183
+    },
+    {
+      "epoch": 0.947010597880424,
+      "grad_norm": 2.976462375028197,
+      "learning_rate": 1.1056105610561057e-07,
+      "loss": -0.1014,
+      "step": 1184
+    },
+    {
+      "epoch": 0.9478104379124175,
+      "grad_norm": 5.789816947507454,
+      "learning_rate": 1.089108910891089e-07,
+      "loss": -0.0268,
+      "step": 1185
+    },
+    {
+      "epoch": 0.9486102779444111,
+      "grad_norm": 4.547707602704605,
+      "learning_rate": 1.0726072607260725e-07,
+      "loss": -0.0252,
+      "step": 1186
+    },
+    {
+      "epoch": 0.9494101179764047,
+      "grad_norm": 3.669232395567316,
+      "learning_rate": 1.0561056105610561e-07,
+      "loss": -0.0721,
+      "step": 1187
+    },
+    {
+      "epoch": 0.9502099580083984,
+      "grad_norm": 4.725560206551925,
+      "learning_rate": 1.0396039603960394e-07,
+      "loss": -0.0374,
+      "step": 1188
+    },
+    {
+      "epoch": 0.9510097980403919,
+      "grad_norm": 5.642201380367948,
+      "learning_rate": 1.023102310231023e-07,
+      "loss": -0.0524,
+      "step": 1189
+    },
+    {
+      "epoch": 0.9518096380723855,
+      "grad_norm": 5.1228456388588555,
+      "learning_rate": 1.0066006600660065e-07,
+      "loss": -0.0122,
+      "step": 1190
+    },
+    {
+      "epoch": 0.9526094781043791,
+      "grad_norm": 3.5927188886776995,
+      "learning_rate": 9.900990099009901e-08,
+      "loss": -0.1424,
+      "step": 1191
+    },
+    {
+      "epoch": 0.9534093181363728,
+      "grad_norm": 3.815501173104961,
+      "learning_rate": 9.735973597359735e-08,
+      "loss": -0.1578,
+      "step": 1192
+    },
+    {
+      "epoch": 0.9542091581683664,
+      "grad_norm": 3.474045980664194,
+      "learning_rate": 9.570957095709571e-08,
+      "loss": -0.0597,
+      "step": 1193
+    },
+    {
+      "epoch": 0.9550089982003599,
+      "grad_norm": 4.782969446108987,
+      "learning_rate": 9.405940594059406e-08,
+      "loss": -0.0957,
+      "step": 1194
+    },
+    {
+      "epoch": 0.9558088382323535,
+      "grad_norm": 4.74265001734386,
+      "learning_rate": 9.24092409240924e-08,
+      "loss": -0.0267,
+      "step": 1195
+    },
+    {
+      "epoch": 0.9566086782643471,
+      "grad_norm": 3.661437568587583,
+      "learning_rate": 9.075907590759075e-08,
+      "loss": -0.1227,
+      "step": 1196
+    },
+    {
+      "epoch": 0.9574085182963408,
+      "grad_norm": 4.279576775134146,
+      "learning_rate": 8.91089108910891e-08,
+      "loss": 0.0111,
+      "step": 1197
+    },
+    {
+      "epoch": 0.9582083583283343,
+      "grad_norm": 4.169597103863264,
+      "learning_rate": 8.745874587458746e-08,
+      "loss": 0.0377,
+      "step": 1198
+    },
+    {
+      "epoch": 0.9590081983603279,
+      "grad_norm": 6.207479567569039,
+      "learning_rate": 8.58085808580858e-08,
+      "loss": 0.0014,
+      "step": 1199
+    },
+    {
+      "epoch": 0.9598080383923215,
+      "grad_norm": 3.5423075974898586,
+      "learning_rate": 8.415841584158416e-08,
+      "loss": -0.1664,
+      "step": 1200
+    },
+    {
+      "epoch": 0.9606078784243152,
+      "grad_norm": 3.6165897542858887,
+      "learning_rate": 8.25082508250825e-08,
+      "loss": -0.0884,
+      "step": 1201
+    },
+    {
+      "epoch": 0.9614077184563088,
+      "grad_norm": 4.517734549618362,
+      "learning_rate": 8.085808580858085e-08,
+      "loss": -0.1076,
+      "step": 1202
+    },
+    {
+      "epoch": 0.9622075584883023,
+      "grad_norm": 4.307866136219069,
+      "learning_rate": 7.92079207920792e-08,
+      "loss": -0.115,
+      "step": 1203
+    },
+    {
+      "epoch": 0.9630073985202959,
+      "grad_norm": 4.922270840667124,
+      "learning_rate": 7.755775577557755e-08,
+      "loss": -0.055,
+      "step": 1204
+    },
+    {
+      "epoch": 0.9638072385522896,
+      "grad_norm": 4.385179502669176,
+      "learning_rate": 7.590759075907591e-08,
+      "loss": -0.1254,
+      "step": 1205
+    },
+    {
+      "epoch": 0.9646070785842832,
+      "grad_norm": 3.9178495988004443,
+      "learning_rate": 7.425742574257424e-08,
+      "loss": -0.1016,
+      "step": 1206
+    },
+    {
+      "epoch": 0.9654069186162767,
+      "grad_norm": 3.3585705170911515,
+      "learning_rate": 7.26072607260726e-08,
+      "loss": -0.0209,
+      "step": 1207
+    },
+    {
+      "epoch": 0.9662067586482703,
+      "grad_norm": 3.815494549495067,
+      "learning_rate": 7.095709570957095e-08,
+      "loss": -0.0635,
+      "step": 1208
+    },
+    {
+      "epoch": 0.967006598680264,
+      "grad_norm": 8.403899931437618,
+      "learning_rate": 6.930693069306931e-08,
+      "loss": -0.0465,
+      "step": 1209
+    },
+    {
+      "epoch": 0.9678064387122576,
+      "grad_norm": 4.29568964473992,
+      "learning_rate": 6.765676567656765e-08,
+      "loss": -0.0332,
+      "step": 1210
+    },
+    {
+      "epoch": 0.9686062787442512,
+      "grad_norm": 3.4280415193587235,
+      "learning_rate": 6.6006600660066e-08,
+      "loss": -0.0803,
+      "step": 1211
+    },
+    {
+      "epoch": 0.9694061187762447,
+      "grad_norm": 4.6120423955763625,
+      "learning_rate": 6.435643564356436e-08,
+      "loss": -0.0619,
+      "step": 1212
+    },
+    {
+      "epoch": 0.9702059588082383,
+      "grad_norm": 6.424876752925553,
+      "learning_rate": 6.27062706270627e-08,
+      "loss": -0.1442,
+      "step": 1213
+    },
+    {
+      "epoch": 0.971005798840232,
+      "grad_norm": 5.485217081397391,
+      "learning_rate": 6.105610561056105e-08,
+      "loss": -0.0939,
+      "step": 1214
+    },
+    {
+      "epoch": 0.9718056388722256,
+      "grad_norm": 6.774111317136949,
+      "learning_rate": 5.94059405940594e-08,
+      "loss": -0.0439,
+      "step": 1215
+    },
+    {
+      "epoch": 0.9726054789042191,
+      "grad_norm": 5.096515115630733,
+      "learning_rate": 5.775577557755775e-08,
+      "loss": -0.0734,
+      "step": 1216
+    },
+    {
+      "epoch": 0.9734053189362127,
+      "grad_norm": 4.159248360440637,
+      "learning_rate": 5.61056105610561e-08,
+      "loss": -0.0121,
+      "step": 1217
+    },
+    {
+      "epoch": 0.9742051589682064,
+      "grad_norm": 7.334455086425638,
+      "learning_rate": 5.445544554455445e-08,
+      "loss": -0.1328,
+      "step": 1218
+    },
+    {
+      "epoch": 0.9750049990002,
+      "grad_norm": 5.126425754126674,
+      "learning_rate": 5.2805280528052805e-08,
+      "loss": -0.123,
+      "step": 1219
+    },
+    {
+      "epoch": 0.9758048390321936,
+      "grad_norm": 4.215981033934002,
+      "learning_rate": 5.115511551155115e-08,
+      "loss": -0.0326,
+      "step": 1220
+    },
+    {
+      "epoch": 0.9766046790641871,
+      "grad_norm": 5.727858996419284,
+      "learning_rate": 4.950495049504951e-08,
+      "loss": -0.078,
+      "step": 1221
+    },
+    {
+      "epoch": 0.9774045190961808,
+      "grad_norm": 4.2278775409875475,
+      "learning_rate": 4.7854785478547855e-08,
+      "loss": -0.1444,
+      "step": 1222
+    },
+    {
+      "epoch": 0.9782043591281744,
+      "grad_norm": 5.1475401401668455,
+      "learning_rate": 4.62046204620462e-08,
+      "loss": -0.0758,
+      "step": 1223
+    },
+    {
+      "epoch": 0.979004199160168,
+      "grad_norm": 4.979540737683743,
+      "learning_rate": 4.455445544554455e-08,
+      "loss": -0.0669,
+      "step": 1224
+    },
+    {
+      "epoch": 0.9798040391921615,
+      "grad_norm": 3.4515888370253385,
+      "learning_rate": 4.29042904290429e-08,
+      "loss": -0.0937,
+      "step": 1225
+    },
+    {
+      "epoch": 0.9806038792241552,
+      "grad_norm": 6.288776500193402,
+      "learning_rate": 4.125412541254125e-08,
+      "loss": 0.1102,
+      "step": 1226
+    },
+    {
+      "epoch": 0.9814037192561488,
+      "grad_norm": 5.48415137375722,
+      "learning_rate": 3.96039603960396e-08,
+      "loss": -0.0964,
+      "step": 1227
+    },
+    {
+      "epoch": 0.9822035592881424,
+      "grad_norm": 4.355868334038742,
+      "learning_rate": 3.7953795379537955e-08,
+      "loss": -0.0328,
+      "step": 1228
+    },
+    {
+      "epoch": 0.9830033993201359,
+      "grad_norm": 4.721491339476331,
+      "learning_rate": 3.63036303630363e-08,
+      "loss": -0.0365,
+      "step": 1229
+    },
+    {
+      "epoch": 0.9838032393521295,
+      "grad_norm": 6.339518521675752,
+      "learning_rate": 3.465346534653466e-08,
+      "loss": -0.0573,
+      "step": 1230
+    },
+    {
+      "epoch": 0.9846030793841232,
+      "grad_norm": 4.434922235230731,
+      "learning_rate": 3.3003300330033e-08,
+      "loss": -0.1461,
+      "step": 1231
+    },
+    {
+      "epoch": 0.9854029194161168,
+      "grad_norm": 5.310987908083999,
+      "learning_rate": 3.135313531353135e-08,
+      "loss": -0.0746,
+      "step": 1232
+    },
+    {
+      "epoch": 0.9862027594481104,
+      "grad_norm": 5.686966755780067,
+      "learning_rate": 2.97029702970297e-08,
+      "loss": -0.0133,
+      "step": 1233
+    },
+    {
+      "epoch": 0.9870025994801039,
+      "grad_norm": 4.108463781012627,
+      "learning_rate": 2.805280528052805e-08,
+      "loss": 0.0749,
+      "step": 1234
+    },
+    {
+      "epoch": 0.9878024395120976,
+      "grad_norm": 3.6672659008615764,
+      "learning_rate": 2.6402640264026403e-08,
+      "loss": -0.1047,
+      "step": 1235
+    },
+    {
+      "epoch": 0.9886022795440912,
+      "grad_norm": 3.9834854628962146,
+      "learning_rate": 2.4752475247524754e-08,
+      "loss": -0.0921,
+      "step": 1236
+    },
+    {
+      "epoch": 0.9894021195760848,
+      "grad_norm": 3.5139198067318054,
+      "learning_rate": 2.31023102310231e-08,
+      "loss": -0.1611,
+      "step": 1237
+    },
+    {
+      "epoch": 0.9902019596080783,
+      "grad_norm": 4.1541924223616356,
+      "learning_rate": 2.145214521452145e-08,
+      "loss": -0.0147,
+      "step": 1238
+    },
+    {
+      "epoch": 0.991001799640072,
+      "grad_norm": 6.337275466101498,
+      "learning_rate": 1.98019801980198e-08,
+      "loss": -0.0767,
+      "step": 1239
+    },
+    {
+      "epoch": 0.9918016396720656,
+      "grad_norm": 82.20527671342789,
+      "learning_rate": 1.815181518151815e-08,
+      "loss": -0.0431,
+      "step": 1240
+    },
+    {
+      "epoch": 0.9926014797040592,
+      "grad_norm": 2.6922514851959494,
+      "learning_rate": 1.65016501650165e-08,
+      "loss": -0.0195,
+      "step": 1241
+    },
+    {
+      "epoch": 0.9934013197360528,
+      "grad_norm": 3.6925978256211747,
+      "learning_rate": 1.485148514851485e-08,
+      "loss": -0.1541,
+      "step": 1242
+    },
+    {
+      "epoch": 0.9942011597680463,
+      "grad_norm": 4.79635865852686,
+      "learning_rate": 1.3201320132013201e-08,
+      "loss": -0.1104,
+      "step": 1243
+    },
+    {
+      "epoch": 0.99500099980004,
+      "grad_norm": 4.327272847702339,
+      "learning_rate": 1.155115511551155e-08,
+      "loss": -0.0807,
+      "step": 1244
+    },
+    {
+      "epoch": 0.9958008398320336,
+      "grad_norm": 4.256644720520306,
+      "learning_rate": 9.9009900990099e-09,
+      "loss": -0.0278,
+      "step": 1245
+    },
+    {
+      "epoch": 0.9966006798640272,
+      "grad_norm": 3.586841344680467,
+      "learning_rate": 8.25082508250825e-09,
+      "loss": 0.0033,
+      "step": 1246
+    },
+    {
+      "epoch": 0.9974005198960207,
+      "grad_norm": 4.084312918321821,
+      "learning_rate": 6.600660066006601e-09,
+      "loss": -0.0053,
+      "step": 1247
+    },
+    {
+      "epoch": 0.9982003599280144,
+      "grad_norm": 3.983696015790867,
+      "learning_rate": 4.95049504950495e-09,
+      "loss": -0.0229,
+      "step": 1248
+    },
+    {
+      "epoch": 0.999000199960008,
+      "grad_norm": 4.1770415014644104,
+      "learning_rate": 3.3003300330033003e-09,
+      "loss": -0.1898,
+      "step": 1249
+    },
+    {
+      "epoch": 0.9998000399920016,
+      "grad_norm": 4.645732496504415,
+      "learning_rate": 1.6501650165016502e-09,
+      "loss": 0.0221,
+      "step": 1250
+    },
+    {
+      "epoch": 0.9998000399920016,
+      "step": 1250,
+      "total_flos": 208730583859200.0,
+      "train_loss": -0.05900815903544426,
+      "train_runtime": 14539.7464,
+      "train_samples_per_second": 11.005,
+      "train_steps_per_second": 0.086
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1250,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "total_flos": 208730583859200.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}