diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,15302 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9997706947947719,
+  "eval_steps": 500,
+  "global_step": 2180,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00045861041045631735,
+      "grad_norm": 195.7306528530031,
+      "learning_rate": 1.5151515151515153e-05,
+      "loss": 10.2128,
+      "step": 1
+    },
+    {
+      "epoch": 0.0009172208209126347,
+      "grad_norm": 195.42822509560057,
+      "learning_rate": 3.0303030303030306e-05,
+      "loss": 10.3065,
+      "step": 2
+    },
+    {
+      "epoch": 0.001375831231368952,
+      "grad_norm": 164.03152872775192,
+      "learning_rate": 4.545454545454546e-05,
+      "loss": 8.9029,
+      "step": 3
+    },
+    {
+      "epoch": 0.0018344416418252694,
+      "grad_norm": 40.7870411864302,
+      "learning_rate": 6.060606060606061e-05,
+      "loss": 7.177,
+      "step": 4
+    },
+    {
+      "epoch": 0.002293052052281587,
+      "grad_norm": 29.96532614146475,
+      "learning_rate": 7.575757575757576e-05,
+      "loss": 6.2976,
+      "step": 5
+    },
+    {
+      "epoch": 0.002751662462737904,
+      "grad_norm": 19.036432224643686,
+      "learning_rate": 9.090909090909092e-05,
+      "loss": 5.6876,
+      "step": 6
+    },
+    {
+      "epoch": 0.0032102728731942215,
+      "grad_norm": 14.278206789517602,
+      "learning_rate": 0.00010606060606060606,
+      "loss": 5.4017,
+      "step": 7
+    },
+    {
+      "epoch": 0.0036688832836505388,
+      "grad_norm": 10.290741842439477,
+      "learning_rate": 0.00012121212121212122,
+      "loss": 4.9368,
+      "step": 8
+    },
+    {
+      "epoch": 0.0041274936941068565,
+      "grad_norm": 6.7147929347502755,
+      "learning_rate": 0.00013636363636363637,
+      "loss": 4.7543,
+      "step": 9
+    },
+    {
+      "epoch": 0.004586104104563174,
+      "grad_norm": 4.788094217031448,
+      "learning_rate": 0.00015151515151515152,
+      "loss": 4.6118,
+      "step": 10
+    },
+    {
+      "epoch": 0.005044714515019491,
+      "grad_norm": 3.9194363851502847,
+      "learning_rate": 0.00016666666666666666,
+      "loss": 4.4951,
+      "step": 11
+    },
+    {
+      "epoch": 0.005503324925475808,
+      "grad_norm": 3.3605651367715152,
+      "learning_rate": 0.00018181818181818183,
+      "loss": 4.599,
+      "step": 12
+    },
+    {
+      "epoch": 0.005961935335932126,
+      "grad_norm": 2.9057341602199998,
+      "learning_rate": 0.00019696969696969695,
+      "loss": 4.4886,
+      "step": 13
+    },
+    {
+      "epoch": 0.006420545746388443,
+      "grad_norm": 2.5986245666625383,
+      "learning_rate": 0.00021212121212121213,
+      "loss": 4.4607,
+      "step": 14
+    },
+    {
+      "epoch": 0.00687915615684476,
+      "grad_norm": 2.2942472113475016,
+      "learning_rate": 0.00022727272727272727,
+      "loss": 4.3536,
+      "step": 15
+    },
+    {
+      "epoch": 0.0073377665673010775,
+      "grad_norm": 2.1093062367534055,
+      "learning_rate": 0.00024242424242424245,
+      "loss": 4.3355,
+      "step": 16
+    },
+    {
+      "epoch": 0.007796376977757395,
+      "grad_norm": 1.7159708128180031,
+      "learning_rate": 0.00025757575757575756,
+      "loss": 4.2415,
+      "step": 17
+    },
+    {
+      "epoch": 0.008254987388213713,
+      "grad_norm": 1.5690114724269786,
+      "learning_rate": 0.00027272727272727274,
+      "loss": 4.2156,
+      "step": 18
+    },
+    {
+      "epoch": 0.00871359779867003,
+      "grad_norm": 1.377296624938421,
+      "learning_rate": 0.0002878787878787879,
+      "loss": 4.2565,
+      "step": 19
+    },
+    {
+      "epoch": 0.009172208209126348,
+      "grad_norm": 1.3641090813747416,
+      "learning_rate": 0.00030303030303030303,
+      "loss": 4.3025,
+      "step": 20
+    },
+    {
+      "epoch": 0.009630818619582664,
+      "grad_norm": 1.18865082366552,
+      "learning_rate": 0.0003181818181818182,
+      "loss": 4.3244,
+      "step": 21
+    },
+    {
+      "epoch": 0.010089429030038982,
+      "grad_norm": 0.9843888916868967,
+      "learning_rate": 0.0003333333333333333,
+      "loss": 4.1548,
+      "step": 22
+    },
+    {
+      "epoch": 0.010548039440495299,
+      "grad_norm": 1.0020844873205947,
+      "learning_rate": 0.0003484848484848485,
+      "loss": 4.127,
+      "step": 23
+    },
+    {
+      "epoch": 0.011006649850951617,
+      "grad_norm": 0.9536258445781124,
+      "learning_rate": 0.00036363636363636367,
+      "loss": 4.2077,
+      "step": 24
+    },
+    {
+      "epoch": 0.011465260261407933,
+      "grad_norm": 0.9476192396409799,
+      "learning_rate": 0.0003787878787878788,
+      "loss": 4.0926,
+      "step": 25
+    },
+    {
+      "epoch": 0.011923870671864251,
+      "grad_norm": 0.9262272918182922,
+      "learning_rate": 0.0003939393939393939,
+      "loss": 4.0886,
+      "step": 26
+    },
+    {
+      "epoch": 0.01238248108232057,
+      "grad_norm": 0.8038592549251841,
+      "learning_rate": 0.00040909090909090913,
+      "loss": 4.1129,
+      "step": 27
+    },
+    {
+      "epoch": 0.012841091492776886,
+      "grad_norm": 0.7829591161389831,
+      "learning_rate": 0.00042424242424242425,
+      "loss": 4.0739,
+      "step": 28
+    },
+    {
+      "epoch": 0.013299701903233204,
+      "grad_norm": 0.7278601839410145,
+      "learning_rate": 0.0004393939393939394,
+      "loss": 4.1649,
+      "step": 29
+    },
+    {
+      "epoch": 0.01375831231368952,
+      "grad_norm": 0.6309340497449822,
+      "learning_rate": 0.00045454545454545455,
+      "loss": 4.0786,
+      "step": 30
+    },
+    {
+      "epoch": 0.014216922724145839,
+      "grad_norm": 0.6431178848106036,
+      "learning_rate": 0.0004696969696969697,
+      "loss": 4.0686,
+      "step": 31
+    },
+    {
+      "epoch": 0.014675533134602155,
+      "grad_norm": 0.5878956382416325,
+      "learning_rate": 0.0004848484848484849,
+      "loss": 4.0855,
+      "step": 32
+    },
+    {
+      "epoch": 0.015134143545058473,
+      "grad_norm": 0.5870153002786562,
+      "learning_rate": 0.0005,
+      "loss": 4.0078,
+      "step": 33
+    },
+    {
+      "epoch": 0.01559275395551479,
+      "grad_norm": 0.5243591289062026,
+      "learning_rate": 0.0005151515151515151,
+      "loss": 4.0662,
+      "step": 34
+    },
+    {
+      "epoch": 0.016051364365971106,
+      "grad_norm": 0.5812197046460649,
+      "learning_rate": 0.0005303030303030302,
+      "loss": 3.9968,
+      "step": 35
+    },
+    {
+      "epoch": 0.016509974776427426,
+      "grad_norm": 0.4306689579344445,
+      "learning_rate": 0.0005454545454545455,
+      "loss": 3.8704,
+      "step": 36
+    },
+    {
+      "epoch": 0.016968585186883742,
+      "grad_norm": 0.4864064124218161,
+      "learning_rate": 0.0005606060606060606,
+      "loss": 3.9451,
+      "step": 37
+    },
+    {
+      "epoch": 0.01742719559734006,
+      "grad_norm": 0.4595907167276233,
+      "learning_rate": 0.0005757575757575758,
+      "loss": 3.9478,
+      "step": 38
+    },
+    {
+      "epoch": 0.017885806007796375,
+      "grad_norm": 0.44004785457330753,
+      "learning_rate": 0.0005909090909090909,
+      "loss": 3.9712,
+      "step": 39
+    },
+    {
+      "epoch": 0.018344416418252695,
+      "grad_norm": 0.538963314442116,
+      "learning_rate": 0.0006060606060606061,
+      "loss": 3.9339,
+      "step": 40
+    },
+    {
+      "epoch": 0.01880302682870901,
+      "grad_norm": 0.40760408459180053,
+      "learning_rate": 0.0006212121212121212,
+      "loss": 3.8532,
+      "step": 41
+    },
+    {
+      "epoch": 0.019261637239165328,
+      "grad_norm": 0.40600159911693534,
+      "learning_rate": 0.0006363636363636364,
+      "loss": 3.8657,
+      "step": 42
+    },
+    {
+      "epoch": 0.019720247649621648,
+      "grad_norm": 0.43937439209913076,
+      "learning_rate": 0.0006515151515151515,
+      "loss": 3.9358,
+      "step": 43
+    },
+    {
+      "epoch": 0.020178858060077964,
+      "grad_norm": 0.4446810019476093,
+      "learning_rate": 0.0006666666666666666,
+      "loss": 3.8323,
+      "step": 44
+    },
+    {
+      "epoch": 0.02063746847053428,
+      "grad_norm": 0.3845101551911155,
+      "learning_rate": 0.0006818181818181818,
+      "loss": 3.7975,
+      "step": 45
+    },
+    {
+      "epoch": 0.021096078880990597,
+      "grad_norm": 0.36561272314478066,
+      "learning_rate": 0.000696969696969697,
+      "loss": 3.8813,
+      "step": 46
+    },
+    {
+      "epoch": 0.021554689291446917,
+      "grad_norm": 0.36932864211098926,
+      "learning_rate": 0.0007121212121212122,
+      "loss": 3.8072,
+      "step": 47
+    },
+    {
+      "epoch": 0.022013299701903234,
+      "grad_norm": 0.41126935338118276,
+      "learning_rate": 0.0007272727272727273,
+      "loss": 3.8202,
+      "step": 48
+    },
+    {
+      "epoch": 0.02247191011235955,
+      "grad_norm": 0.405657698503483,
+      "learning_rate": 0.0007424242424242425,
+      "loss": 3.917,
+      "step": 49
+    },
+    {
+      "epoch": 0.022930520522815866,
+      "grad_norm": 0.32763977552796913,
+      "learning_rate": 0.0007575757575757576,
+      "loss": 3.8485,
+      "step": 50
+    },
+    {
+      "epoch": 0.023389130933272186,
+      "grad_norm": 0.3525238077511433,
+      "learning_rate": 0.0007727272727272727,
+      "loss": 3.8197,
+      "step": 51
+    },
+    {
+      "epoch": 0.023847741343728503,
+      "grad_norm": 0.35958381264468203,
+      "learning_rate": 0.0007878787878787878,
+      "loss": 3.9413,
+      "step": 52
+    },
+    {
+      "epoch": 0.02430635175418482,
+      "grad_norm": 0.3790913735234885,
+      "learning_rate": 0.000803030303030303,
+      "loss": 3.9376,
+      "step": 53
+    },
+    {
+      "epoch": 0.02476496216464114,
+      "grad_norm": 0.3249322252574824,
+      "learning_rate": 0.0008181818181818183,
+      "loss": 3.8398,
+      "step": 54
+    },
+    {
+      "epoch": 0.025223572575097455,
+      "grad_norm": 0.34859361522332905,
+      "learning_rate": 0.0008333333333333334,
+      "loss": 3.819,
+      "step": 55
+    },
+    {
+      "epoch": 0.025682182985553772,
+      "grad_norm": 0.36869718589466227,
+      "learning_rate": 0.0008484848484848485,
+      "loss": 3.9164,
+      "step": 56
+    },
+    {
+      "epoch": 0.02614079339601009,
+      "grad_norm": 0.34619945636161886,
+      "learning_rate": 0.0008636363636363636,
+      "loss": 3.8633,
+      "step": 57
+    },
+    {
+      "epoch": 0.026599403806466408,
+      "grad_norm": 0.3567302213564318,
+      "learning_rate": 0.0008787878787878789,
+      "loss": 3.9811,
+      "step": 58
+    },
+    {
+      "epoch": 0.027058014216922725,
+      "grad_norm": 0.43280280348456207,
+      "learning_rate": 0.000893939393939394,
+      "loss": 3.7865,
+      "step": 59
+    },
+    {
+      "epoch": 0.02751662462737904,
+      "grad_norm": 0.3696545113895953,
+      "learning_rate": 0.0009090909090909091,
+      "loss": 3.7954,
+      "step": 60
+    },
+    {
+      "epoch": 0.027975235037835357,
+      "grad_norm": 0.3844371705132838,
+      "learning_rate": 0.0009242424242424242,
+      "loss": 3.7121,
+      "step": 61
+    },
+    {
+      "epoch": 0.028433845448291677,
+      "grad_norm": 0.4880042446407562,
+      "learning_rate": 0.0009393939393939394,
+      "loss": 3.8272,
+      "step": 62
+    },
+    {
+      "epoch": 0.028892455858747994,
+      "grad_norm": 0.5209436474998808,
+      "learning_rate": 0.0009545454545454546,
+      "loss": 3.7804,
+      "step": 63
+    },
+    {
+      "epoch": 0.02935106626920431,
+      "grad_norm": 0.438792302391278,
+      "learning_rate": 0.0009696969696969698,
+      "loss": 3.9328,
+      "step": 64
+    },
+    {
+      "epoch": 0.029809676679660627,
+      "grad_norm": 0.507261288107282,
+      "learning_rate": 0.000984848484848485,
+      "loss": 3.7183,
+      "step": 65
+    },
+    {
+      "epoch": 0.030268287090116947,
+      "grad_norm": 0.4210162006376918,
+      "learning_rate": 0.001,
+      "loss": 3.8489,
+      "step": 66
+    },
+    {
+      "epoch": 0.030726897500573263,
+      "grad_norm": 0.44672966620701343,
+      "learning_rate": 0.0009999994478847943,
+      "loss": 3.8555,
+      "step": 67
+    },
+    {
+      "epoch": 0.03118550791102958,
+      "grad_norm": 0.5216322592982295,
+      "learning_rate": 0.0009999977915403962,
+      "loss": 3.7083,
+      "step": 68
+    },
+    {
+      "epoch": 0.0316441183214859,
+      "grad_norm": 0.43522530953859156,
+      "learning_rate": 0.0009999950309704639,
+      "loss": 3.7592,
+      "step": 69
+    },
+    {
+      "epoch": 0.03210272873194221,
+      "grad_norm": 0.46573074517293367,
+      "learning_rate": 0.000999991166181094,
+      "loss": 3.7153,
+      "step": 70
+    },
+    {
+      "epoch": 0.03256133914239853,
+      "grad_norm": 0.4276068142230606,
+      "learning_rate": 0.0009999861971808216,
+      "loss": 3.7825,
+      "step": 71
+    },
+    {
+      "epoch": 0.03301994955285485,
+      "grad_norm": 0.48314356174537304,
+      "learning_rate": 0.0009999801239806208,
+      "loss": 3.7798,
+      "step": 72
+    },
+    {
+      "epoch": 0.033478559963311165,
+      "grad_norm": 0.4438822925002156,
+      "learning_rate": 0.0009999729465939035,
+      "loss": 3.6618,
+      "step": 73
+    },
+    {
+      "epoch": 0.033937170373767485,
+      "grad_norm": 0.5015209861730612,
+      "learning_rate": 0.0009999646650365212,
+      "loss": 3.7399,
+      "step": 74
+    },
+    {
+      "epoch": 0.034395780784223805,
+      "grad_norm": 0.6193602476195937,
+      "learning_rate": 0.0009999552793267634,
+      "loss": 3.7672,
+      "step": 75
+    },
+    {
+      "epoch": 0.03485439119468012,
+      "grad_norm": 0.5202766763705031,
+      "learning_rate": 0.0009999447894853577,
+      "loss": 3.704,
+      "step": 76
+    },
+    {
+      "epoch": 0.03531300160513644,
+      "grad_norm": 0.48923311570783196,
+      "learning_rate": 0.0009999331955354708,
+      "loss": 3.719,
+      "step": 77
+    },
+    {
+      "epoch": 0.03577161201559275,
+      "grad_norm": 0.6205951733984051,
+      "learning_rate": 0.0009999204975027073,
+      "loss": 3.7102,
+      "step": 78
+    },
+    {
+      "epoch": 0.03623022242604907,
+      "grad_norm": 0.6874013726169927,
+      "learning_rate": 0.0009999066954151103,
+      "loss": 3.6796,
+      "step": 79
+    },
+    {
+      "epoch": 0.03668883283650539,
+      "grad_norm": 0.6848357429038243,
+      "learning_rate": 0.0009998917893031614,
+      "loss": 3.7495,
+      "step": 80
+    },
+    {
+      "epoch": 0.0371474432469617,
+      "grad_norm": 0.5154913139479887,
+      "learning_rate": 0.0009998757791997801,
+      "loss": 3.6142,
+      "step": 81
+    },
+    {
+      "epoch": 0.03760605365741802,
+      "grad_norm": 0.6795270300098945,
+      "learning_rate": 0.0009998586651403238,
+      "loss": 3.6673,
+      "step": 82
+    },
+    {
+      "epoch": 0.03806466406787434,
+      "grad_norm": 0.6729504623810377,
+      "learning_rate": 0.0009998404471625885,
+      "loss": 3.6788,
+      "step": 83
+    },
+    {
+      "epoch": 0.038523274478330656,
+      "grad_norm": 0.8261583053997574,
+      "learning_rate": 0.0009998211253068078,
+      "loss": 3.6216,
+      "step": 84
+    },
+    {
+      "epoch": 0.038981884888786976,
+      "grad_norm": 0.9213890649494518,
+      "learning_rate": 0.0009998006996156535,
+      "loss": 3.6963,
+      "step": 85
+    },
+    {
+      "epoch": 0.039440495299243296,
+      "grad_norm": 0.7512210251875971,
+      "learning_rate": 0.0009997791701342347,
+      "loss": 3.6952,
+      "step": 86
+    },
+    {
+      "epoch": 0.03989910570969961,
+      "grad_norm": 0.8364673330758103,
+      "learning_rate": 0.0009997565369100983,
+      "loss": 3.542,
+      "step": 87
+    },
+    {
+      "epoch": 0.04035771612015593,
+      "grad_norm": 0.8556338723386335,
+      "learning_rate": 0.0009997327999932291,
+      "loss": 3.6662,
+      "step": 88
+    },
+    {
+      "epoch": 0.04081632653061224,
+      "grad_norm": 0.7914712770535725,
+      "learning_rate": 0.000999707959436049,
+      "loss": 3.6274,
+      "step": 89
+    },
+    {
+      "epoch": 0.04127493694106856,
+      "grad_norm": 0.910354064713232,
+      "learning_rate": 0.0009996820152934176,
+      "loss": 3.6781,
+      "step": 90
+    },
+    {
+      "epoch": 0.04173354735152488,
+      "grad_norm": 0.8179807761292784,
+      "learning_rate": 0.000999654967622631,
+      "loss": 3.5154,
+      "step": 91
+    },
+    {
+      "epoch": 0.042192157761981194,
+      "grad_norm": 0.747588752021303,
+      "learning_rate": 0.0009996268164834238,
+      "loss": 3.5504,
+      "step": 92
+    },
+    {
+      "epoch": 0.042650768172437514,
+      "grad_norm": 0.7768238197341649,
+      "learning_rate": 0.000999597561937966,
+      "loss": 3.5485,
+      "step": 93
+    },
+    {
+      "epoch": 0.043109378582893834,
+      "grad_norm": 0.7509124378566197,
+      "learning_rate": 0.0009995672040508656,
+      "loss": 3.5134,
+      "step": 94
+    },
+    {
+      "epoch": 0.04356798899335015,
+      "grad_norm": 0.8001360275146927,
+      "learning_rate": 0.0009995357428891662,
+      "loss": 3.5353,
+      "step": 95
+    },
+    {
+      "epoch": 0.04402659940380647,
+      "grad_norm": 0.7054048109583136,
+      "learning_rate": 0.0009995031785223491,
+      "loss": 3.5951,
+      "step": 96
+    },
+    {
+      "epoch": 0.04448520981426279,
+      "grad_norm": 0.7730032433987475,
+      "learning_rate": 0.000999469511022331,
+      "loss": 3.6421,
+      "step": 97
+    },
+    {
+      "epoch": 0.0449438202247191,
+      "grad_norm": 0.6785779691106867,
+      "learning_rate": 0.0009994347404634657,
+      "loss": 3.5513,
+      "step": 98
+    },
+    {
+      "epoch": 0.04540243063517542,
+      "grad_norm": 0.8229994146687332,
+      "learning_rate": 0.0009993988669225423,
+      "loss": 3.5472,
+      "step": 99
+    },
+    {
+      "epoch": 0.04586104104563173,
+      "grad_norm": 0.8647325544539955,
+      "learning_rate": 0.000999361890478786,
+      "loss": 3.5449,
+      "step": 100
+    },
+    {
+      "epoch": 0.04631965145608805,
+      "grad_norm": 0.8609391914038597,
+      "learning_rate": 0.0009993238112138583,
+      "loss": 3.5448,
+      "step": 101
+    },
+    {
+      "epoch": 0.04677826186654437,
+      "grad_norm": 0.8559576425289054,
+      "learning_rate": 0.0009992846292118554,
+      "loss": 3.4045,
+      "step": 102
+    },
+    {
+      "epoch": 0.047236872277000685,
+      "grad_norm": 0.6396722803922795,
+      "learning_rate": 0.000999244344559309,
+      "loss": 3.4061,
+      "step": 103
+    },
+    {
+      "epoch": 0.047695482687457005,
+      "grad_norm": 0.8751986621136766,
+      "learning_rate": 0.0009992029573451869,
+      "loss": 3.4685,
+      "step": 104
+    },
+    {
+      "epoch": 0.048154093097913325,
+      "grad_norm": 0.7223591949294903,
+      "learning_rate": 0.0009991604676608905,
+      "loss": 3.4407,
+      "step": 105
+    },
+    {
+      "epoch": 0.04861270350836964,
+      "grad_norm": 0.7984606325752777,
+      "learning_rate": 0.0009991168756002568,
+      "loss": 3.4995,
+      "step": 106
+    },
+    {
+      "epoch": 0.04907131391882596,
+      "grad_norm": 0.8618507131733196,
+      "learning_rate": 0.0009990721812595574,
+      "loss": 3.3931,
+      "step": 107
+    },
+    {
+      "epoch": 0.04952992432928228,
+      "grad_norm": 0.8295802023006299,
+      "learning_rate": 0.0009990263847374976,
+      "loss": 3.3956,
+      "step": 108
+    },
+    {
+      "epoch": 0.04998853473973859,
+      "grad_norm": 0.8379119391032637,
+      "learning_rate": 0.0009989794861352173,
+      "loss": 3.4612,
+      "step": 109
+    },
+    {
+      "epoch": 0.05044714515019491,
+      "grad_norm": 0.8335807911997553,
+      "learning_rate": 0.0009989314855562905,
+      "loss": 3.4853,
+      "step": 110
+    },
+    {
+      "epoch": 0.050905755560651224,
+      "grad_norm": 1.042131302814139,
+      "learning_rate": 0.0009988823831067245,
+      "loss": 3.4479,
+      "step": 111
+    },
+    {
+      "epoch": 0.051364365971107544,
+      "grad_norm": 0.7940637198810115,
+      "learning_rate": 0.0009988321788949597,
+      "loss": 3.4211,
+      "step": 112
+    },
+    {
+      "epoch": 0.051822976381563864,
+      "grad_norm": 0.7570526291406791,
+      "learning_rate": 0.0009987808730318709,
+      "loss": 3.3237,
+      "step": 113
+    },
+    {
+      "epoch": 0.05228158679202018,
+      "grad_norm": 0.903200409702691,
+      "learning_rate": 0.0009987284656307644,
+      "loss": 3.3718,
+      "step": 114
+    },
+    {
+      "epoch": 0.052740197202476496,
+      "grad_norm": 0.8239978960596677,
+      "learning_rate": 0.0009986749568073802,
+      "loss": 3.4082,
+      "step": 115
+    },
+    {
+      "epoch": 0.053198807612932816,
+      "grad_norm": 1.044973517402376,
+      "learning_rate": 0.0009986203466798905,
+      "loss": 3.4365,
+      "step": 116
+    },
+    {
+      "epoch": 0.05365741802338913,
+      "grad_norm": 0.9121674956539658,
+      "learning_rate": 0.0009985646353688996,
+      "loss": 3.4619,
+      "step": 117
+    },
+    {
+      "epoch": 0.05411602843384545,
+      "grad_norm": 1.08644063709046,
+      "learning_rate": 0.0009985078229974437,
+      "loss": 3.3893,
+      "step": 118
+    },
+    {
+      "epoch": 0.05457463884430177,
+      "grad_norm": 0.7851370332625733,
+      "learning_rate": 0.0009984499096909905,
+      "loss": 3.3502,
+      "step": 119
+    },
+    {
+      "epoch": 0.05503324925475808,
+      "grad_norm": 1.0276709319613944,
+      "learning_rate": 0.0009983908955774397,
+      "loss": 3.3936,
+      "step": 120
+    },
+    {
+      "epoch": 0.0554918596652144,
+      "grad_norm": 0.9896975904360017,
+      "learning_rate": 0.0009983307807871211,
+      "loss": 3.3281,
+      "step": 121
+    },
+    {
+      "epoch": 0.055950470075670715,
+      "grad_norm": 0.8683501465787161,
+      "learning_rate": 0.0009982695654527965,
+      "loss": 3.3332,
+      "step": 122
+    },
+    {
+      "epoch": 0.056409080486127035,
+      "grad_norm": 0.9502874644236643,
+      "learning_rate": 0.0009982072497096571,
+      "loss": 3.3216,
+      "step": 123
+    },
+    {
+      "epoch": 0.056867690896583355,
+      "grad_norm": 0.8937728936066178,
+      "learning_rate": 0.000998143833695325,
+      "loss": 3.2702,
+      "step": 124
+    },
+    {
+      "epoch": 0.05732630130703967,
+      "grad_norm": 1.1375435412510808,
+      "learning_rate": 0.0009980793175498517,
+      "loss": 3.3025,
+      "step": 125
+    },
+    {
+      "epoch": 0.05778491171749599,
+      "grad_norm": 0.8532550753740189,
+      "learning_rate": 0.000998013701415719,
+      "loss": 3.323,
+      "step": 126
+    },
+    {
+      "epoch": 0.05824352212795231,
+      "grad_norm": 0.8596091265969976,
+      "learning_rate": 0.0009979469854378372,
+      "loss": 3.2354,
+      "step": 127
+    },
+    {
+      "epoch": 0.05870213253840862,
+      "grad_norm": 0.7538842057742248,
+      "learning_rate": 0.000997879169763546,
+      "loss": 3.2269,
+      "step": 128
+    },
+    {
+      "epoch": 0.05916074294886494,
+      "grad_norm": 0.8709831615465082,
+      "learning_rate": 0.000997810254542614,
+      "loss": 3.2939,
+      "step": 129
+    },
+    {
+      "epoch": 0.05961935335932125,
+      "grad_norm": 0.7651338064470595,
+      "learning_rate": 0.0009977402399272374,
+      "loss": 3.2218,
+      "step": 130
+    },
+    {
+      "epoch": 0.06007796376977757,
+      "grad_norm": 0.8976472406608559,
+      "learning_rate": 0.0009976691260720407,
+      "loss": 3.1508,
+      "step": 131
+    },
+    {
+      "epoch": 0.06053657418023389,
+      "grad_norm": 1.015154999780063,
+      "learning_rate": 0.0009975969131340763,
+      "loss": 3.2461,
+      "step": 132
+    },
+    {
+      "epoch": 0.060995184590690206,
+      "grad_norm": 0.9294234356631054,
+      "learning_rate": 0.0009975236012728236,
+      "loss": 3.266,
+      "step": 133
+    },
+    {
+      "epoch": 0.061453795001146526,
+      "grad_norm": 1.005484878512385,
+      "learning_rate": 0.0009974491906501886,
+      "loss": 3.2975,
+      "step": 134
+    },
+    {
+      "epoch": 0.061912405411602846,
+      "grad_norm": 0.8787247299794221,
+      "learning_rate": 0.0009973736814305049,
+      "loss": 3.2487,
+      "step": 135
+    },
+    {
+      "epoch": 0.06237101582205916,
+      "grad_norm": 0.961415926605254,
+      "learning_rate": 0.0009972970737805312,
+      "loss": 3.2191,
+      "step": 136
+    },
+    {
+      "epoch": 0.06282962623251548,
+      "grad_norm": 0.9914501025703937,
+      "learning_rate": 0.0009972193678694525,
+      "loss": 3.2227,
+      "step": 137
+    },
+    {
+      "epoch": 0.0632882366429718,
+      "grad_norm": 0.8744026906320149,
+      "learning_rate": 0.0009971405638688794,
+      "loss": 3.2318,
+      "step": 138
+    },
+    {
+      "epoch": 0.06374684705342812,
+      "grad_norm": 0.9201167384546156,
+      "learning_rate": 0.0009970606619528475,
+      "loss": 3.174,
+      "step": 139
+    },
+    {
+      "epoch": 0.06420545746388442,
+      "grad_norm": 0.8111366934577829,
+      "learning_rate": 0.000996979662297817,
+      "loss": 3.2289,
+      "step": 140
+    },
+    {
+      "epoch": 0.06466406787434074,
+      "grad_norm": 1.0010612214882024,
+      "learning_rate": 0.0009968975650826721,
+      "loss": 3.1533,
+      "step": 141
+    },
+    {
+      "epoch": 0.06512267828479706,
+      "grad_norm": 0.9371437553688488,
+      "learning_rate": 0.000996814370488722,
+      "loss": 3.2421,
+      "step": 142
+    },
+    {
+      "epoch": 0.06558128869525338,
+      "grad_norm": 0.8886574715846152,
+      "learning_rate": 0.000996730078699698,
+      "loss": 3.1456,
+      "step": 143
+    },
+    {
+      "epoch": 0.0660398991057097,
+      "grad_norm": 0.7443130771778591,
+      "learning_rate": 0.0009966446899017558,
+      "loss": 3.1076,
+      "step": 144
+    },
+    {
+      "epoch": 0.06649850951616602,
+      "grad_norm": 0.9377023804575672,
+      "learning_rate": 0.0009965582042834728,
+      "loss": 3.2297,
+      "step": 145
+    },
+    {
+      "epoch": 0.06695711992662233,
+      "grad_norm": 0.8538108860455718,
+      "learning_rate": 0.0009964706220358492,
+      "loss": 3.2199,
+      "step": 146
+    },
+    {
+      "epoch": 0.06741573033707865,
+      "grad_norm": 0.8821992299741697,
+      "learning_rate": 0.000996381943352307,
+      "loss": 3.1634,
+      "step": 147
+    },
+    {
+      "epoch": 0.06787434074753497,
+      "grad_norm": 0.9016445264293866,
+      "learning_rate": 0.0009962921684286896,
+      "loss": 3.2017,
+      "step": 148
+    },
+    {
+      "epoch": 0.06833295115799129,
+      "grad_norm": 0.824529186857188,
+      "learning_rate": 0.0009962012974632614,
+      "loss": 3.2461,
+      "step": 149
+    },
+    {
+      "epoch": 0.06879156156844761,
+      "grad_norm": 0.8786966166897391,
+      "learning_rate": 0.0009961093306567075,
+      "loss": 3.1845,
+      "step": 150
+    },
+    {
+      "epoch": 0.06925017197890392,
+      "grad_norm": 0.9481258640646799,
+      "learning_rate": 0.0009960162682121328,
+      "loss": 3.1916,
+      "step": 151
+    },
+    {
+      "epoch": 0.06970878238936024,
+      "grad_norm": 0.7861163426200737,
+      "learning_rate": 0.0009959221103350623,
+      "loss": 3.1733,
+      "step": 152
+    },
+    {
+      "epoch": 0.07016739279981656,
+      "grad_norm": 0.9598587388049913,
+      "learning_rate": 0.0009958268572334394,
+      "loss": 3.0197,
+      "step": 153
+    },
+    {
+      "epoch": 0.07062600321027288,
+      "grad_norm": 0.881366983397273,
+      "learning_rate": 0.0009957305091176274,
+      "loss": 3.1268,
+      "step": 154
+    },
+    {
+      "epoch": 0.0710846136207292,
+      "grad_norm": 0.9130797709164997,
+      "learning_rate": 0.0009956330662004075,
+      "loss": 3.1469,
+      "step": 155
+    },
+    {
+      "epoch": 0.0715432240311855,
+      "grad_norm": 0.8198655118008726,
+      "learning_rate": 0.0009955345286969779,
+      "loss": 3.1788,
+      "step": 156
+    },
+    {
+      "epoch": 0.07200183444164182,
+      "grad_norm": 0.800865689174721,
+      "learning_rate": 0.0009954348968249551,
+      "loss": 3.1142,
+      "step": 157
+    },
+    {
+      "epoch": 0.07246044485209814,
+      "grad_norm": 1.0058872341579066,
+      "learning_rate": 0.0009953341708043724,
+      "loss": 3.1627,
+      "step": 158
+    },
+    {
+      "epoch": 0.07291905526255446,
+      "grad_norm": 0.8395523848761809,
+      "learning_rate": 0.0009952323508576793,
+      "loss": 3.0997,
+      "step": 159
+    },
+    {
+      "epoch": 0.07337766567301078,
+      "grad_norm": 0.9437603199647973,
+      "learning_rate": 0.0009951294372097406,
+      "loss": 3.0811,
+      "step": 160
+    },
+    {
+      "epoch": 0.0738362760834671,
+      "grad_norm": 0.8017421592853877,
+      "learning_rate": 0.0009950254300878378,
+      "loss": 3.1327,
+      "step": 161
+    },
+    {
+      "epoch": 0.0742948864939234,
+      "grad_norm": 0.8470966873391701,
+      "learning_rate": 0.000994920329721666,
+      "loss": 3.0924,
+      "step": 162
+    },
+    {
+      "epoch": 0.07475349690437973,
+      "grad_norm": 0.9417577485420556,
+      "learning_rate": 0.0009948141363433356,
+      "loss": 3.1463,
+      "step": 163
+    },
+    {
+      "epoch": 0.07521210731483605,
+      "grad_norm": 0.9023072308530313,
+      "learning_rate": 0.00099470685018737,
+      "loss": 3.0523,
+      "step": 164
+    },
+    {
+      "epoch": 0.07567071772529237,
+      "grad_norm": 0.8585706373892799,
+      "learning_rate": 0.0009945984714907073,
+      "loss": 3.0607,
+      "step": 165
+    },
+    {
+      "epoch": 0.07612932813574869,
+      "grad_norm": 0.9328103717036563,
+      "learning_rate": 0.000994489000492697,
+      "loss": 3.137,
+      "step": 166
+    },
+    {
+      "epoch": 0.07658793854620499,
+      "grad_norm": 0.8833104373308935,
+      "learning_rate": 0.0009943784374351016,
+      "loss": 3.1749,
+      "step": 167
+    },
+    {
+      "epoch": 0.07704654895666131,
+      "grad_norm": 0.9096183627515776,
+      "learning_rate": 0.0009942667825620951,
+      "loss": 3.0278,
+      "step": 168
+    },
+    {
+      "epoch": 0.07750515936711763,
+      "grad_norm": 0.9583443936327846,
+      "learning_rate": 0.0009941540361202634,
+      "loss": 2.9746,
+      "step": 169
+    },
+    {
+      "epoch": 0.07796376977757395,
+      "grad_norm": 0.8158985260075705,
+      "learning_rate": 0.0009940401983586022,
+      "loss": 3.1277,
+      "step": 170
+    },
+    {
+      "epoch": 0.07842238018803027,
+      "grad_norm": 0.9631764479659928,
+      "learning_rate": 0.000993925269528518,
+      "loss": 3.1284,
+      "step": 171
+    },
+    {
+      "epoch": 0.07888099059848659,
+      "grad_norm": 0.7753667644394078,
+      "learning_rate": 0.0009938092498838265,
+      "loss": 3.1493,
+      "step": 172
+    },
+    {
+      "epoch": 0.0793396010089429,
+      "grad_norm": 0.7998704549372179,
+      "learning_rate": 0.0009936921396807524,
+      "loss": 3.1571,
+      "step": 173
+    },
+    {
+      "epoch": 0.07979821141939922,
+      "grad_norm": 1.006261339874799,
+      "learning_rate": 0.0009935739391779292,
+      "loss": 2.9955,
+      "step": 174
+    },
+    {
+      "epoch": 0.08025682182985554,
+      "grad_norm": 0.8347915481775977,
+      "learning_rate": 0.000993454648636398,
+      "loss": 3.0476,
+      "step": 175
+    },
+    {
+      "epoch": 0.08071543224031186,
+      "grad_norm": 0.7958428745423937,
+      "learning_rate": 0.0009933342683196074,
+      "loss": 3.0147,
+      "step": 176
+    },
+    {
+      "epoch": 0.08117404265076818,
+      "grad_norm": 0.9491652957301248,
+      "learning_rate": 0.0009932127984934125,
+      "loss": 3.0338,
+      "step": 177
+    },
+    {
+      "epoch": 0.08163265306122448,
+      "grad_norm": 0.7965688353957587,
+      "learning_rate": 0.0009930902394260745,
+      "loss": 3.1245,
+      "step": 178
+    },
+    {
+      "epoch": 0.0820912634716808,
+      "grad_norm": 0.9575816603440428,
+      "learning_rate": 0.0009929665913882607,
+      "loss": 2.9929,
+      "step": 179
+    },
+    {
+      "epoch": 0.08254987388213712,
+      "grad_norm": 0.887103629092979,
+      "learning_rate": 0.0009928418546530425,
+      "loss": 2.9941,
+      "step": 180
+    },
+    {
+      "epoch": 0.08300848429259344,
+      "grad_norm": 0.9088047373987485,
+      "learning_rate": 0.0009927160294958964,
+      "loss": 3.1235,
+      "step": 181
+    },
+    {
+      "epoch": 0.08346709470304976,
+      "grad_norm": 0.8124166567189071,
+      "learning_rate": 0.000992589116194702,
+      "loss": 3.0846,
+      "step": 182
+    },
+    {
+      "epoch": 0.08392570511350608,
+      "grad_norm": 0.9119910218205586,
+      "learning_rate": 0.000992461115029743,
+      "loss": 2.9981,
+      "step": 183
+    },
+    {
+      "epoch": 0.08438431552396239,
+      "grad_norm": 0.9444523303085732,
+      "learning_rate": 0.000992332026283704,
+      "loss": 3.0558,
+      "step": 184
+    },
+    {
+      "epoch": 0.08484292593441871,
+      "grad_norm": 0.728763073086222,
+      "learning_rate": 0.0009922018502416736,
+      "loss": 2.9748,
+      "step": 185
+    },
+    {
+      "epoch": 0.08530153634487503,
+      "grad_norm": 0.7922567881975136,
+      "learning_rate": 0.0009920705871911395,
+      "loss": 2.9752,
+      "step": 186
+    },
+    {
+      "epoch": 0.08576014675533135,
+      "grad_norm": 0.9364167120060036,
+      "learning_rate": 0.0009919382374219915,
+      "loss": 2.9992,
+      "step": 187
+    },
+    {
+      "epoch": 0.08621875716578767,
+      "grad_norm": 0.8538079381136541,
+      "learning_rate": 0.0009918048012265187,
+      "loss": 2.9758,
+      "step": 188
+    },
+    {
+      "epoch": 0.08667736757624397,
+      "grad_norm": 0.713199109823494,
+      "learning_rate": 0.0009916702788994097,
+      "loss": 3.0535,
+      "step": 189
+    },
+    {
+      "epoch": 0.0871359779867003,
+      "grad_norm": 0.7317450714125606,
+      "learning_rate": 0.0009915346707377519,
+      "loss": 2.9983,
+      "step": 190
+    },
+    {
+      "epoch": 0.08759458839715661,
+      "grad_norm": 0.9219605549932931,
+      "learning_rate": 0.0009913979770410305,
+      "loss": 3.0516,
+      "step": 191
+    },
+    {
+      "epoch": 0.08805319880761293,
+      "grad_norm": 0.8795784822266292,
+      "learning_rate": 0.0009912601981111285,
+      "loss": 3.1185,
+      "step": 192
+    },
+    {
+      "epoch": 0.08851180921806925,
+      "grad_norm": 0.8626015874362707,
+      "learning_rate": 0.0009911213342523248,
+      "loss": 3.0876,
+      "step": 193
+    },
+    {
+      "epoch": 0.08897041962852557,
+      "grad_norm": 0.9575400555748717,
+      "learning_rate": 0.000990981385771295,
+      "loss": 3.0082,
+      "step": 194
+    },
+    {
+      "epoch": 0.08942903003898188,
+      "grad_norm": 0.8240019997000216,
+      "learning_rate": 0.00099084035297711,
+      "loss": 3.0254,
+      "step": 195
+    },
+    {
+      "epoch": 0.0898876404494382,
+      "grad_norm": 0.8528801449707022,
+      "learning_rate": 0.000990698236181235,
+      "loss": 2.9983,
+      "step": 196
+    },
+    {
+      "epoch": 0.09034625085989452,
+      "grad_norm": 0.8860278518306403,
+      "learning_rate": 0.0009905550356975293,
+      "loss": 2.9959,
+      "step": 197
+    },
+    {
+      "epoch": 0.09080486127035084,
+      "grad_norm": 1.0292926464828644,
+      "learning_rate": 0.0009904107518422457,
+      "loss": 2.9596,
+      "step": 198
+    },
+    {
+      "epoch": 0.09126347168080716,
+      "grad_norm": 0.8645288788556811,
+      "learning_rate": 0.0009902653849340295,
+      "loss": 3.002,
+      "step": 199
+    },
+    {
+      "epoch": 0.09172208209126347,
+      "grad_norm": 0.8182290444539763,
+      "learning_rate": 0.0009901189352939177,
+      "loss": 3.0617,
+      "step": 200
+    },
+    {
+      "epoch": 0.09218069250171979,
+      "grad_norm": 0.8455205460380645,
+      "learning_rate": 0.0009899714032453387,
+      "loss": 3.0378,
+      "step": 201
+    },
+    {
+      "epoch": 0.0926393029121761,
+      "grad_norm": 0.9473655429231509,
+      "learning_rate": 0.000989822789114111,
+      "loss": 2.9497,
+      "step": 202
+    },
+    {
+      "epoch": 0.09309791332263243,
+      "grad_norm": 0.8544989556657595,
+      "learning_rate": 0.0009896730932284434,
+      "loss": 2.9259,
+      "step": 203
+    },
+    {
+      "epoch": 0.09355652373308875,
+      "grad_norm": 0.88958571909881,
+      "learning_rate": 0.0009895223159189332,
+      "loss": 2.9463,
+      "step": 204
+    },
+    {
+      "epoch": 0.09401513414354506,
+      "grad_norm": 0.8144804563977788,
+      "learning_rate": 0.0009893704575185663,
+      "loss": 2.9266,
+      "step": 205
+    },
+    {
+      "epoch": 0.09447374455400137,
+      "grad_norm": 0.916941065253182,
+      "learning_rate": 0.000989217518362716,
+      "loss": 3.0001,
+      "step": 206
+    },
+    {
+      "epoch": 0.09493235496445769,
+      "grad_norm": 1.0134701058582654,
+      "learning_rate": 0.0009890634987891425,
+      "loss": 2.9935,
+      "step": 207
+    },
+    {
+      "epoch": 0.09539096537491401,
+      "grad_norm": 0.8241100936694118,
+      "learning_rate": 0.0009889083991379917,
+      "loss": 3.0307,
+      "step": 208
+    },
+    {
+      "epoch": 0.09584957578537033,
+      "grad_norm": 0.976634413217131,
+      "learning_rate": 0.0009887522197517954,
+      "loss": 3.0526,
+      "step": 209
+    },
+    {
+      "epoch": 0.09630818619582665,
+      "grad_norm": 0.9812771496836543,
+      "learning_rate": 0.0009885949609754693,
+      "loss": 3.0789,
+      "step": 210
+    },
+    {
+      "epoch": 0.09676679660628296,
+      "grad_norm": 1.0099796585633138,
+      "learning_rate": 0.000988436623156314,
+      "loss": 2.9977,
+      "step": 211
+    },
+    {
+      "epoch": 0.09722540701673928,
+      "grad_norm": 0.9927433669515897,
+      "learning_rate": 0.0009882772066440114,
+      "loss": 2.838,
+      "step": 212
+    },
+    {
+      "epoch": 0.0976840174271956,
+      "grad_norm": 0.901887545095234,
+      "learning_rate": 0.0009881167117906276,
+      "loss": 2.8237,
+      "step": 213
+    },
+    {
+      "epoch": 0.09814262783765192,
+      "grad_norm": 0.9011815103364363,
+      "learning_rate": 0.0009879551389506084,
+      "loss": 2.9528,
+      "step": 214
+    },
+    {
+      "epoch": 0.09860123824810824,
+      "grad_norm": 0.9016188929604817,
+      "learning_rate": 0.0009877924884807814,
+      "loss": 3.0426,
+      "step": 215
+    },
+    {
+      "epoch": 0.09905984865856456,
+      "grad_norm": 0.975364104333327,
+      "learning_rate": 0.000987628760740354,
+      "loss": 2.961,
+      "step": 216
+    },
+    {
+      "epoch": 0.09951845906902086,
+      "grad_norm": 0.820196197328104,
+      "learning_rate": 0.0009874639560909118,
+      "loss": 2.8953,
+      "step": 217
+    },
+    {
+      "epoch": 0.09997706947947718,
+      "grad_norm": 0.9082645432344721,
+      "learning_rate": 0.0009872980748964202,
+      "loss": 2.8911,
+      "step": 218
+    },
+    {
+      "epoch": 0.1004356798899335,
+      "grad_norm": 0.9463439478977312,
+      "learning_rate": 0.000987131117523221,
+      "loss": 2.9133,
+      "step": 219
+    },
+    {
+      "epoch": 0.10089429030038982,
+      "grad_norm": 0.9813915601227976,
+      "learning_rate": 0.000986963084340033,
+      "loss": 2.9495,
+      "step": 220
+    },
+    {
+      "epoch": 0.10135290071084614,
+      "grad_norm": 0.88241840169542,
+      "learning_rate": 0.0009867939757179508,
+      "loss": 2.8842,
+      "step": 221
+    },
+    {
+      "epoch": 0.10181151112130245,
+      "grad_norm": 1.0912626026903895,
+      "learning_rate": 0.0009866237920304443,
+      "loss": 2.9591,
+      "step": 222
+    },
+    {
+      "epoch": 0.10227012153175877,
+      "grad_norm": 0.8514562190379549,
+      "learning_rate": 0.0009864525336533577,
+      "loss": 2.8649,
+      "step": 223
+    },
+    {
+      "epoch": 0.10272873194221509,
+      "grad_norm": 0.9120720209168053,
+      "learning_rate": 0.000986280200964908,
+      "loss": 2.9031,
+      "step": 224
+    },
+    {
+      "epoch": 0.10318734235267141,
+      "grad_norm": 0.992480985651104,
+      "learning_rate": 0.0009861067943456856,
+      "loss": 2.9498,
+      "step": 225
+    },
+    {
+      "epoch": 0.10364595276312773,
+      "grad_norm": 0.8471976858335671,
+      "learning_rate": 0.000985932314178652,
+      "loss": 2.8531,
+      "step": 226
+    },
+    {
+      "epoch": 0.10410456317358405,
+      "grad_norm": 1.0055750766838616,
+      "learning_rate": 0.00098575676084914,
+      "loss": 2.8819,
+      "step": 227
+    },
+    {
+      "epoch": 0.10456317358404035,
+      "grad_norm": 0.9362516980722705,
+      "learning_rate": 0.0009855801347448518,
+      "loss": 2.9964,
+      "step": 228
+    },
+    {
+      "epoch": 0.10502178399449667,
+      "grad_norm": 0.866072836069588,
+      "learning_rate": 0.0009854024362558596,
+      "loss": 2.9201,
+      "step": 229
+    },
+    {
+      "epoch": 0.10548039440495299,
+      "grad_norm": 0.9369376626276946,
+      "learning_rate": 0.0009852236657746035,
+      "loss": 2.9205,
+      "step": 230
+    },
+    {
+      "epoch": 0.10593900481540931,
+      "grad_norm": 0.9251219173327379,
+      "learning_rate": 0.0009850438236958911,
+      "loss": 2.8202,
+      "step": 231
+    },
+    {
+      "epoch": 0.10639761522586563,
+      "grad_norm": 0.8903181153572154,
+      "learning_rate": 0.0009848629104168966,
+      "loss": 2.9534,
+      "step": 232
+    },
+    {
+      "epoch": 0.10685622563632194,
+      "grad_norm": 0.8888700691932708,
+      "learning_rate": 0.00098468092633716,
+      "loss": 2.9553,
+      "step": 233
+    },
+    {
+      "epoch": 0.10731483604677826,
+      "grad_norm": 0.8281665612477659,
+      "learning_rate": 0.0009844978718585855,
+      "loss": 2.8825,
+      "step": 234
+    },
+    {
+      "epoch": 0.10777344645723458,
+      "grad_norm": 0.9539897386094597,
+      "learning_rate": 0.0009843137473854423,
+      "loss": 2.8972,
+      "step": 235
+    },
+    {
+      "epoch": 0.1082320568676909,
+      "grad_norm": 0.8495369154077475,
+      "learning_rate": 0.000984128553324362,
+      "loss": 2.8901,
+      "step": 236
+    },
+    {
+      "epoch": 0.10869066727814722,
+      "grad_norm": 0.9202058608389792,
+      "learning_rate": 0.0009839422900843383,
+      "loss": 2.8844,
+      "step": 237
+    },
+    {
+      "epoch": 0.10914927768860354,
+      "grad_norm": 0.9516379548358328,
+      "learning_rate": 0.0009837549580767261,
+      "loss": 2.9709,
+      "step": 238
+    },
+    {
+      "epoch": 0.10960788809905984,
+      "grad_norm": 0.9326670474809671,
+      "learning_rate": 0.0009835665577152411,
+      "loss": 2.8348,
+      "step": 239
+    },
+    {
+      "epoch": 0.11006649850951616,
+      "grad_norm": 0.8321122034758862,
+      "learning_rate": 0.000983377089415958,
+      "loss": 2.9147,
+      "step": 240
+    },
+    {
+      "epoch": 0.11052510891997248,
+      "grad_norm": 0.8865877955226755,
+      "learning_rate": 0.0009831865535973102,
+      "loss": 2.9041,
+      "step": 241
+    },
+    {
+      "epoch": 0.1109837193304288,
+      "grad_norm": 0.8206693589748508,
+      "learning_rate": 0.0009829949506800885,
+      "loss": 2.9101,
+      "step": 242
+    },
+    {
+      "epoch": 0.11144232974088512,
+      "grad_norm": 0.9168410365478116,
+      "learning_rate": 0.0009828022810874405,
+      "loss": 2.8723,
+      "step": 243
+    },
+    {
+      "epoch": 0.11190094015134143,
+      "grad_norm": 0.8779624233319264,
+      "learning_rate": 0.0009826085452448693,
+      "loss": 2.844,
+      "step": 244
+    },
+    {
+      "epoch": 0.11235955056179775,
+      "grad_norm": 0.8419361533441373,
+      "learning_rate": 0.000982413743580233,
+      "loss": 2.8848,
+      "step": 245
+    },
+    {
+      "epoch": 0.11281816097225407,
+      "grad_norm": 0.8903898967465228,
+      "learning_rate": 0.0009822178765237436,
+      "loss": 2.8132,
+      "step": 246
+    },
+    {
+      "epoch": 0.11327677138271039,
+      "grad_norm": 0.8071069688103292,
+      "learning_rate": 0.0009820209445079654,
+      "loss": 2.8834,
+      "step": 247
+    },
+    {
+      "epoch": 0.11373538179316671,
+      "grad_norm": 0.8754403579036916,
+      "learning_rate": 0.0009818229479678158,
+      "loss": 2.7789,
+      "step": 248
+    },
+    {
+      "epoch": 0.11419399220362302,
+      "grad_norm": 0.8603402616625198,
+      "learning_rate": 0.0009816238873405615,
+      "loss": 2.8032,
+      "step": 249
+    },
+    {
+      "epoch": 0.11465260261407934,
+      "grad_norm": 0.8946981098413249,
+      "learning_rate": 0.0009814237630658207,
+      "loss": 2.92,
+      "step": 250
+    },
+    {
+      "epoch": 0.11511121302453566,
+      "grad_norm": 0.9379557120054217,
+      "learning_rate": 0.00098122257558556,
+      "loss": 2.9475,
+      "step": 251
+    },
+    {
+      "epoch": 0.11556982343499198,
+      "grad_norm": 0.9592090514845009,
+      "learning_rate": 0.0009810203253440937,
+      "loss": 2.9349,
+      "step": 252
+    },
+    {
+      "epoch": 0.1160284338454483,
+      "grad_norm": 0.8861294039525018,
+      "learning_rate": 0.0009808170127880837,
+      "loss": 2.8837,
+      "step": 253
+    },
+    {
+      "epoch": 0.11648704425590461,
+      "grad_norm": 0.8729401143534162,
+      "learning_rate": 0.000980612638366538,
+      "loss": 2.9238,
+      "step": 254
+    },
+    {
+      "epoch": 0.11694565466636092,
+      "grad_norm": 0.9199327854010128,
+      "learning_rate": 0.0009804072025308096,
+      "loss": 2.872,
+      "step": 255
+    },
+    {
+      "epoch": 0.11740426507681724,
+      "grad_norm": 0.8791602430305343,
+      "learning_rate": 0.000980200705734595,
+      "loss": 2.873,
+      "step": 256
+    },
+    {
+      "epoch": 0.11786287548727356,
+      "grad_norm": 0.9105347602037588,
+      "learning_rate": 0.0009799931484339344,
+      "loss": 2.7775,
+      "step": 257
+    },
+    {
+      "epoch": 0.11832148589772988,
+      "grad_norm": 0.9273118784344772,
+      "learning_rate": 0.0009797845310872103,
+      "loss": 2.8781,
+      "step": 258
+    },
+    {
+      "epoch": 0.1187800963081862,
+      "grad_norm": 0.9774650132202869,
+      "learning_rate": 0.0009795748541551457,
+      "loss": 2.7575,
+      "step": 259
+    },
+    {
+      "epoch": 0.1192387067186425,
+      "grad_norm": 0.879223835641391,
+      "learning_rate": 0.000979364118100804,
+      "loss": 2.9073,
+      "step": 260
+    },
+    {
+      "epoch": 0.11969731712909883,
+      "grad_norm": 0.8608877680326807,
+      "learning_rate": 0.0009791523233895875,
+      "loss": 2.8769,
+      "step": 261
+    },
+    {
+      "epoch": 0.12015592753955515,
+      "grad_norm": 0.9745283519582287,
+      "learning_rate": 0.0009789394704892364,
+      "loss": 2.8459,
+      "step": 262
+    },
+    {
+      "epoch": 0.12061453795001147,
+      "grad_norm": 0.9012333451811644,
+      "learning_rate": 0.0009787255598698282,
+      "loss": 2.8207,
+      "step": 263
+    },
+    {
+      "epoch": 0.12107314836046779,
+      "grad_norm": 0.9578509788395994,
+      "learning_rate": 0.0009785105920037758,
+      "loss": 2.946,
+      "step": 264
+    },
+    {
+      "epoch": 0.1215317587709241,
+      "grad_norm": 0.9387557385619509,
+      "learning_rate": 0.0009782945673658275,
+      "loss": 2.8424,
+      "step": 265
+    },
+    {
+      "epoch": 0.12199036918138041,
+      "grad_norm": 1.0004924607806316,
+      "learning_rate": 0.0009780774864330654,
+      "loss": 2.8264,
+      "step": 266
+    },
+    {
+      "epoch": 0.12244897959183673,
+      "grad_norm": 0.8149232199521285,
+      "learning_rate": 0.000977859349684904,
+      "loss": 2.8256,
+      "step": 267
+    },
+    {
+      "epoch": 0.12290759000229305,
+      "grad_norm": 0.9014059407287098,
+      "learning_rate": 0.00097764015760309,
+      "loss": 2.7884,
+      "step": 268
+    },
+    {
+      "epoch": 0.12336620041274937,
+      "grad_norm": 1.022452231370555,
+      "learning_rate": 0.0009774199106717004,
+      "loss": 2.8262,
+      "step": 269
+    },
+    {
+      "epoch": 0.12382481082320569,
+      "grad_norm": 0.8810383586172705,
+      "learning_rate": 0.0009771986093771417,
+      "loss": 2.7035,
+      "step": 270
+    },
+    {
+      "epoch": 0.124283421233662,
+      "grad_norm": 1.119624968853302,
+      "learning_rate": 0.0009769762542081496,
+      "loss": 2.7681,
+      "step": 271
+    },
+    {
+      "epoch": 0.12474203164411832,
+      "grad_norm": 0.927092786530175,
+      "learning_rate": 0.000976752845655786,
+      "loss": 2.7902,
+      "step": 272
+    },
+    {
+      "epoch": 0.12520064205457465,
+      "grad_norm": 0.8887260865546257,
+      "learning_rate": 0.0009765283842134411,
+      "loss": 2.7496,
+      "step": 273
+    },
+    {
+      "epoch": 0.12565925246503096,
+      "grad_norm": 1.0419618122062857,
+      "learning_rate": 0.0009763028703768282,
+      "loss": 2.8347,
+      "step": 274
+    },
+    {
+      "epoch": 0.12611786287548726,
+      "grad_norm": 1.0613287396340563,
+      "learning_rate": 0.0009760763046439862,
+      "loss": 2.8571,
+      "step": 275
+    },
+    {
+      "epoch": 0.1265764732859436,
+      "grad_norm": 0.8823701970104146,
+      "learning_rate": 0.0009758486875152766,
+      "loss": 2.8418,
+      "step": 276
+    },
+    {
+      "epoch": 0.1270350836963999,
+      "grad_norm": 0.8987113163825041,
+      "learning_rate": 0.0009756200194933829,
+      "loss": 2.8373,
+      "step": 277
+    },
+    {
+      "epoch": 0.12749369410685624,
+      "grad_norm": 0.9582824410611981,
+      "learning_rate": 0.0009753903010833094,
+      "loss": 2.8097,
+      "step": 278
+    },
+    {
+      "epoch": 0.12795230451731254,
+      "grad_norm": 0.8360030434606588,
+      "learning_rate": 0.0009751595327923803,
+      "loss": 2.8008,
+      "step": 279
+    },
+    {
+      "epoch": 0.12841091492776885,
+      "grad_norm": 0.9792042150961556,
+      "learning_rate": 0.0009749277151302382,
+      "loss": 2.87,
+      "step": 280
+    },
+    {
+      "epoch": 0.12886952533822518,
+      "grad_norm": 0.8991667589553546,
+      "learning_rate": 0.0009746948486088435,
+      "loss": 2.8684,
+      "step": 281
+    },
+    {
+      "epoch": 0.1293281357486815,
+      "grad_norm": 0.8505008457998828,
+      "learning_rate": 0.0009744609337424727,
+      "loss": 2.8596,
+      "step": 282
+    },
+    {
+      "epoch": 0.12978674615913782,
+      "grad_norm": 0.9230130525797575,
+      "learning_rate": 0.0009742259710477177,
+      "loss": 2.7664,
+      "step": 283
+    },
+    {
+      "epoch": 0.13024535656959413,
+      "grad_norm": 0.9714365127723572,
+      "learning_rate": 0.0009739899610434841,
+      "loss": 2.8842,
+      "step": 284
+    },
+    {
+      "epoch": 0.13070396698005043,
+      "grad_norm": 1.0223551500077062,
+      "learning_rate": 0.0009737529042509913,
+      "loss": 2.7561,
+      "step": 285
+    },
+    {
+      "epoch": 0.13116257739050677,
+      "grad_norm": 1.1845472171702751,
+      "learning_rate": 0.0009735148011937693,
+      "loss": 2.7587,
+      "step": 286
+    },
+    {
+      "epoch": 0.13162118780096307,
+      "grad_norm": 0.8896347080104572,
+      "learning_rate": 0.00097327565239766,
+      "loss": 2.8298,
+      "step": 287
+    },
+    {
+      "epoch": 0.1320797982114194,
+      "grad_norm": 1.0068054710296361,
+      "learning_rate": 0.0009730354583908136,
+      "loss": 2.8278,
+      "step": 288
+    },
+    {
+      "epoch": 0.13253840862187571,
+      "grad_norm": 0.9911361747179588,
+      "learning_rate": 0.0009727942197036895,
+      "loss": 2.8456,
+      "step": 289
+    },
+    {
+      "epoch": 0.13299701903233205,
+      "grad_norm": 0.896243627335785,
+      "learning_rate": 0.0009725519368690539,
+      "loss": 2.7968,
+      "step": 290
+    },
+    {
+      "epoch": 0.13345562944278835,
+      "grad_norm": 0.9135570268692881,
+      "learning_rate": 0.0009723086104219787,
+      "loss": 2.7377,
+      "step": 291
+    },
+    {
+      "epoch": 0.13391423985324466,
+      "grad_norm": 1.1318735146476573,
+      "learning_rate": 0.0009720642408998409,
+      "loss": 2.7841,
+      "step": 292
+    },
+    {
+      "epoch": 0.134372850263701,
+      "grad_norm": 0.8877410716227283,
+      "learning_rate": 0.0009718188288423211,
+      "loss": 2.7121,
+      "step": 293
+    },
+    {
+      "epoch": 0.1348314606741573,
+      "grad_norm": 1.0759377324362764,
+      "learning_rate": 0.0009715723747914022,
+      "loss": 2.7603,
+      "step": 294
+    },
+    {
+      "epoch": 0.13529007108461363,
+      "grad_norm": 0.8688271128861191,
+      "learning_rate": 0.0009713248792913685,
+      "loss": 2.8254,
+      "step": 295
+    },
+    {
+      "epoch": 0.13574868149506994,
+      "grad_norm": 0.8940556088168119,
+      "learning_rate": 0.0009710763428888037,
+      "loss": 2.7674,
+      "step": 296
+    },
+    {
+      "epoch": 0.13620729190552625,
+      "grad_norm": 1.0066427733276648,
+      "learning_rate": 0.0009708267661325909,
+      "loss": 2.8475,
+      "step": 297
+    },
+    {
+      "epoch": 0.13666590231598258,
+      "grad_norm": 0.9676872077508448,
+      "learning_rate": 0.0009705761495739107,
+      "loss": 2.8205,
+      "step": 298
+    },
+    {
+      "epoch": 0.13712451272643889,
+      "grad_norm": 1.145680643785185,
+      "learning_rate": 0.0009703244937662399,
+      "loss": 2.7345,
+      "step": 299
+    },
+    {
+      "epoch": 0.13758312313689522,
+      "grad_norm": 0.9349242036000033,
+      "learning_rate": 0.0009700717992653505,
+      "loss": 2.7972,
+      "step": 300
+    },
+    {
+      "epoch": 0.13804173354735153,
+      "grad_norm": 0.9155949604814292,
+      "learning_rate": 0.0009698180666293083,
+      "loss": 2.8448,
+      "step": 301
+    },
+    {
+      "epoch": 0.13850034395780783,
+      "grad_norm": 0.9753885277067499,
+      "learning_rate": 0.000969563296418472,
+      "loss": 2.7864,
+      "step": 302
+    },
+    {
+      "epoch": 0.13895895436826416,
+      "grad_norm": 0.864333505097425,
+      "learning_rate": 0.0009693074891954914,
+      "loss": 2.7682,
+      "step": 303
+    },
+    {
+      "epoch": 0.13941756477872047,
+      "grad_norm": 0.8193926311356593,
+      "learning_rate": 0.0009690506455253072,
+      "loss": 2.7678,
+      "step": 304
+    },
+    {
+      "epoch": 0.1398761751891768,
+      "grad_norm": 0.9680034555845736,
+      "learning_rate": 0.0009687927659751481,
+      "loss": 2.743,
+      "step": 305
+    },
+    {
+      "epoch": 0.1403347855996331,
+      "grad_norm": 0.9166262398103566,
+      "learning_rate": 0.0009685338511145312,
+      "loss": 2.746,
+      "step": 306
+    },
+    {
+      "epoch": 0.14079339601008942,
+      "grad_norm": 0.9586246296076695,
+      "learning_rate": 0.0009682739015152598,
+      "loss": 2.7296,
+      "step": 307
+    },
+    {
+      "epoch": 0.14125200642054575,
+      "grad_norm": 0.9165868560300923,
+      "learning_rate": 0.0009680129177514226,
+      "loss": 2.708,
+      "step": 308
+    },
+    {
+      "epoch": 0.14171061683100206,
+      "grad_norm": 1.016069799232525,
+      "learning_rate": 0.0009677509003993915,
+      "loss": 2.7818,
+      "step": 309
+    },
+    {
+      "epoch": 0.1421692272414584,
+      "grad_norm": 0.935984647179847,
+      "learning_rate": 0.0009674878500378221,
+      "loss": 2.7656,
+      "step": 310
+    },
+    {
+      "epoch": 0.1426278376519147,
+      "grad_norm": 0.8913564099504865,
+      "learning_rate": 0.0009672237672476505,
+      "loss": 2.7426,
+      "step": 311
+    },
+    {
+      "epoch": 0.143086448062371,
+      "grad_norm": 0.9574485676342409,
+      "learning_rate": 0.0009669586526120935,
+      "loss": 2.7188,
+      "step": 312
+    },
+    {
+      "epoch": 0.14354505847282734,
+      "grad_norm": 0.8608092883724667,
+      "learning_rate": 0.0009666925067166459,
+      "loss": 2.7561,
+      "step": 313
+    },
+    {
+      "epoch": 0.14400366888328364,
+      "grad_norm": 0.9503437039816225,
+      "learning_rate": 0.000966425330149081,
+      "loss": 2.8121,
+      "step": 314
+    },
+    {
+      "epoch": 0.14446227929373998,
+      "grad_norm": 0.9495459666694431,
+      "learning_rate": 0.0009661571234994475,
+      "loss": 2.6957,
+      "step": 315
+    },
+    {
+      "epoch": 0.14492088970419628,
+      "grad_norm": 0.8352006755585702,
+      "learning_rate": 0.0009658878873600691,
+      "loss": 2.735,
+      "step": 316
+    },
+    {
+      "epoch": 0.14537950011465262,
+      "grad_norm": 0.9893966004467123,
+      "learning_rate": 0.0009656176223255438,
+      "loss": 2.8221,
+      "step": 317
+    },
+    {
+      "epoch": 0.14583811052510892,
+      "grad_norm": 1.0320206758133763,
+      "learning_rate": 0.000965346328992741,
+      "loss": 2.8208,
+      "step": 318
+    },
+    {
+      "epoch": 0.14629672093556523,
+      "grad_norm": 0.9506338611132933,
+      "learning_rate": 0.0009650740079608014,
+      "loss": 2.7822,
+      "step": 319
+    },
+    {
+      "epoch": 0.14675533134602156,
+      "grad_norm": 0.905130157170755,
+      "learning_rate": 0.0009648006598311353,
+      "loss": 2.6783,
+      "step": 320
+    },
+    {
+      "epoch": 0.14721394175647787,
+      "grad_norm": 0.8849970626577102,
+      "learning_rate": 0.0009645262852074214,
+      "loss": 2.7218,
+      "step": 321
+    },
+    {
+      "epoch": 0.1476725521669342,
+      "grad_norm": 0.8786943488514374,
+      "learning_rate": 0.0009642508846956053,
+      "loss": 2.7062,
+      "step": 322
+    },
+    {
+      "epoch": 0.1481311625773905,
+      "grad_norm": 0.9039598113505151,
+      "learning_rate": 0.0009639744589038983,
+      "loss": 2.7215,
+      "step": 323
+    },
+    {
+      "epoch": 0.1485897729878468,
+      "grad_norm": 1.0494562340898312,
+      "learning_rate": 0.0009636970084427759,
+      "loss": 2.8121,
+      "step": 324
+    },
+    {
+      "epoch": 0.14904838339830315,
+      "grad_norm": 0.9992586874548868,
+      "learning_rate": 0.0009634185339249766,
+      "loss": 2.7763,
+      "step": 325
+    },
+    {
+      "epoch": 0.14950699380875945,
+      "grad_norm": 1.005209409069846,
+      "learning_rate": 0.0009631390359655003,
+      "loss": 2.7582,
+      "step": 326
+    },
+    {
+      "epoch": 0.1499656042192158,
+      "grad_norm": 1.0046232025906898,
+      "learning_rate": 0.0009628585151816074,
+      "loss": 2.639,
+      "step": 327
+    },
+    {
+      "epoch": 0.1504242146296721,
+      "grad_norm": 0.9868200628009041,
+      "learning_rate": 0.0009625769721928172,
+      "loss": 2.6996,
+      "step": 328
+    },
+    {
+      "epoch": 0.1508828250401284,
+      "grad_norm": 0.9466405488928296,
+      "learning_rate": 0.0009622944076209061,
+      "loss": 2.6824,
+      "step": 329
+    },
+    {
+      "epoch": 0.15134143545058473,
+      "grad_norm": 0.9826336622835588,
+      "learning_rate": 0.0009620108220899071,
+      "loss": 2.7472,
+      "step": 330
+    },
+    {
+      "epoch": 0.15180004586104104,
+      "grad_norm": 0.9295384451968925,
+      "learning_rate": 0.0009617262162261075,
+      "loss": 2.6887,
+      "step": 331
+    },
+    {
+      "epoch": 0.15225865627149737,
+      "grad_norm": 0.9196204398209462,
+      "learning_rate": 0.0009614405906580486,
+      "loss": 2.7353,
+      "step": 332
+    },
+    {
+      "epoch": 0.15271726668195368,
+      "grad_norm": 0.9072573768401977,
+      "learning_rate": 0.000961153946016523,
+      "loss": 2.7481,
+      "step": 333
+    },
+    {
+      "epoch": 0.15317587709240998,
+      "grad_norm": 0.8840830372280872,
+      "learning_rate": 0.000960866282934574,
+      "loss": 2.6525,
+      "step": 334
+    },
+    {
+      "epoch": 0.15363448750286632,
+      "grad_norm": 0.8595137412677537,
+      "learning_rate": 0.0009605776020474945,
+      "loss": 2.7109,
+      "step": 335
+    },
+    {
+      "epoch": 0.15409309791332262,
+      "grad_norm": 0.8671778940676147,
+      "learning_rate": 0.0009602879039928249,
+      "loss": 2.7313,
+      "step": 336
+    },
+    {
+      "epoch": 0.15455170832377896,
+      "grad_norm": 0.9179400982976182,
+      "learning_rate": 0.0009599971894103521,
+      "loss": 2.7185,
+      "step": 337
+    },
+    {
+      "epoch": 0.15501031873423526,
+      "grad_norm": 0.8831876266928437,
+      "learning_rate": 0.0009597054589421077,
+      "loss": 2.7579,
+      "step": 338
+    },
+    {
+      "epoch": 0.1554689291446916,
+      "grad_norm": 0.855083766448017,
+      "learning_rate": 0.0009594127132323669,
+      "loss": 2.6832,
+      "step": 339
+    },
+    {
+      "epoch": 0.1559275395551479,
+      "grad_norm": 0.9324522258566154,
+      "learning_rate": 0.0009591189529276474,
+      "loss": 2.763,
+      "step": 340
+    },
+    {
+      "epoch": 0.1563861499656042,
+      "grad_norm": 0.9668108773843086,
+      "learning_rate": 0.0009588241786767072,
+      "loss": 2.7104,
+      "step": 341
+    },
+    {
+      "epoch": 0.15684476037606054,
+      "grad_norm": 0.9009714400831985,
+      "learning_rate": 0.0009585283911305436,
+      "loss": 2.5645,
+      "step": 342
+    },
+    {
+      "epoch": 0.15730337078651685,
+      "grad_norm": 0.9952687390482042,
+      "learning_rate": 0.000958231590942392,
+      "loss": 2.7386,
+      "step": 343
+    },
+    {
+      "epoch": 0.15776198119697318,
+      "grad_norm": 0.9362018189769761,
+      "learning_rate": 0.0009579337787677238,
+      "loss": 2.7508,
+      "step": 344
+    },
+    {
+      "epoch": 0.1582205916074295,
+      "grad_norm": 1.028102690518439,
+      "learning_rate": 0.0009576349552642456,
+      "loss": 2.7739,
+      "step": 345
+    },
+    {
+      "epoch": 0.1586792020178858,
+      "grad_norm": 0.9354939642089415,
+      "learning_rate": 0.0009573351210918975,
+      "loss": 2.605,
+      "step": 346
+    },
+    {
+      "epoch": 0.15913781242834213,
+      "grad_norm": 0.9647521906590177,
+      "learning_rate": 0.0009570342769128514,
+      "loss": 2.6969,
+      "step": 347
+    },
+    {
+      "epoch": 0.15959642283879844,
+      "grad_norm": 0.9725294778594562,
+      "learning_rate": 0.0009567324233915099,
+      "loss": 2.6469,
+      "step": 348
+    },
+    {
+      "epoch": 0.16005503324925477,
+      "grad_norm": 0.958846616464893,
+      "learning_rate": 0.0009564295611945047,
+      "loss": 2.7808,
+      "step": 349
+    },
+    {
+      "epoch": 0.16051364365971107,
+      "grad_norm": 0.9231613992305402,
+      "learning_rate": 0.000956125690990695,
+      "loss": 2.6393,
+      "step": 350
+    },
+    {
+      "epoch": 0.16097225407016738,
+      "grad_norm": 0.8664922741518483,
+      "learning_rate": 0.0009558208134511665,
+      "loss": 2.7795,
+      "step": 351
+    },
+    {
+      "epoch": 0.16143086448062371,
+      "grad_norm": 0.9901569124425272,
+      "learning_rate": 0.0009555149292492289,
+      "loss": 2.6415,
+      "step": 352
+    },
+    {
+      "epoch": 0.16188947489108002,
+      "grad_norm": 0.9105996096527891,
+      "learning_rate": 0.0009552080390604159,
+      "loss": 2.6484,
+      "step": 353
+    },
+    {
+      "epoch": 0.16234808530153635,
+      "grad_norm": 0.9204018330640393,
+      "learning_rate": 0.0009549001435624823,
+      "loss": 2.658,
+      "step": 354
+    },
+    {
+      "epoch": 0.16280669571199266,
+      "grad_norm": 0.9484425693897214,
+      "learning_rate": 0.0009545912434354029,
+      "loss": 2.6723,
+      "step": 355
+    },
+    {
+      "epoch": 0.16326530612244897,
+      "grad_norm": 0.9476573231028919,
+      "learning_rate": 0.0009542813393613721,
+      "loss": 2.7197,
+      "step": 356
+    },
+    {
+      "epoch": 0.1637239165329053,
+      "grad_norm": 0.90457086127769,
+      "learning_rate": 0.0009539704320248006,
+      "loss": 2.6823,
+      "step": 357
+    },
+    {
+      "epoch": 0.1641825269433616,
+      "grad_norm": 1.0577227695740639,
+      "learning_rate": 0.0009536585221123151,
+      "loss": 2.6008,
+      "step": 358
+    },
+    {
+      "epoch": 0.16464113735381794,
+      "grad_norm": 0.9368807042654839,
+      "learning_rate": 0.0009533456103127565,
+      "loss": 2.8108,
+      "step": 359
+    },
+    {
+      "epoch": 0.16509974776427425,
+      "grad_norm": 0.8996542150235647,
+      "learning_rate": 0.000953031697317178,
+      "loss": 2.6392,
+      "step": 360
+    },
+    {
+      "epoch": 0.16555835817473058,
+      "grad_norm": 0.9088718526536299,
+      "learning_rate": 0.0009527167838188445,
+      "loss": 2.707,
+      "step": 361
+    },
+    {
+      "epoch": 0.16601696858518689,
+      "grad_norm": 0.814696013551401,
+      "learning_rate": 0.0009524008705132299,
+      "loss": 2.6876,
+      "step": 362
+    },
+    {
+      "epoch": 0.1664755789956432,
+      "grad_norm": 0.8910430611417197,
+      "learning_rate": 0.0009520839580980166,
+      "loss": 2.6405,
+      "step": 363
+    },
+    {
+      "epoch": 0.16693418940609953,
+      "grad_norm": 0.950966727653377,
+      "learning_rate": 0.0009517660472730929,
+      "loss": 2.5719,
+      "step": 364
+    },
+    {
+      "epoch": 0.16739279981655583,
+      "grad_norm": 1.2628708292039117,
+      "learning_rate": 0.0009514471387405526,
+      "loss": 2.7181,
+      "step": 365
+    },
+    {
+      "epoch": 0.16785141022701217,
+      "grad_norm": 0.8373275145081241,
+      "learning_rate": 0.0009511272332046926,
+      "loss": 2.6296,
+      "step": 366
+    },
+    {
+      "epoch": 0.16831002063746847,
+      "grad_norm": 0.9230497019799034,
+      "learning_rate": 0.0009508063313720119,
+      "loss": 2.7153,
+      "step": 367
+    },
+    {
+      "epoch": 0.16876863104792478,
+      "grad_norm": 0.9485399573493368,
+      "learning_rate": 0.0009504844339512095,
+      "loss": 2.6743,
+      "step": 368
+    },
+    {
+      "epoch": 0.1692272414583811,
+      "grad_norm": 0.9167780928380227,
+      "learning_rate": 0.0009501615416531835,
+      "loss": 2.7691,
+      "step": 369
+    },
+    {
+      "epoch": 0.16968585186883742,
+      "grad_norm": 0.8968006517074619,
+      "learning_rate": 0.0009498376551910285,
+      "loss": 2.719,
+      "step": 370
+    },
+    {
+      "epoch": 0.17014446227929375,
+      "grad_norm": 0.9279872136597834,
+      "learning_rate": 0.0009495127752800352,
+      "loss": 2.7149,
+      "step": 371
+    },
+    {
+      "epoch": 0.17060307268975006,
+      "grad_norm": 0.9760139120143462,
+      "learning_rate": 0.0009491869026376882,
+      "loss": 2.6917,
+      "step": 372
+    },
+    {
+      "epoch": 0.17106168310020636,
+      "grad_norm": 0.9216482887767333,
+      "learning_rate": 0.0009488600379836648,
+      "loss": 2.6384,
+      "step": 373
+    },
+    {
+      "epoch": 0.1715202935106627,
+      "grad_norm": 0.8428830894105103,
+      "learning_rate": 0.0009485321820398321,
+      "loss": 2.5639,
+      "step": 374
+    },
+    {
+      "epoch": 0.171978903921119,
+      "grad_norm": 0.9873482189193917,
+      "learning_rate": 0.0009482033355302475,
+      "loss": 2.6981,
+      "step": 375
+    },
+    {
+      "epoch": 0.17243751433157534,
+      "grad_norm": 0.8378301858639329,
+      "learning_rate": 0.0009478734991811556,
+      "loss": 2.6321,
+      "step": 376
+    },
+    {
+      "epoch": 0.17289612474203164,
+      "grad_norm": 0.9247499259837642,
+      "learning_rate": 0.0009475426737209871,
+      "loss": 2.7437,
+      "step": 377
+    },
+    {
+      "epoch": 0.17335473515248795,
+      "grad_norm": 0.9017509329255388,
+      "learning_rate": 0.000947210859880357,
+      "loss": 2.7036,
+      "step": 378
+    },
+    {
+      "epoch": 0.17381334556294428,
+      "grad_norm": 0.9964385422312817,
+      "learning_rate": 0.0009468780583920631,
+      "loss": 2.5867,
+      "step": 379
+    },
+    {
+      "epoch": 0.1742719559734006,
+      "grad_norm": 0.9446766886932328,
+      "learning_rate": 0.0009465442699910846,
+      "loss": 2.5892,
+      "step": 380
+    },
+    {
+      "epoch": 0.17473056638385692,
+      "grad_norm": 0.9523413341091993,
+      "learning_rate": 0.0009462094954145801,
+      "loss": 2.783,
+      "step": 381
+    },
+    {
+      "epoch": 0.17518917679431323,
+      "grad_norm": 1.0747365307843044,
+      "learning_rate": 0.0009458737354018859,
+      "loss": 2.6096,
+      "step": 382
+    },
+    {
+      "epoch": 0.17564778720476956,
+      "grad_norm": 0.9775163633169596,
+      "learning_rate": 0.000945536990694515,
+      "loss": 2.6588,
+      "step": 383
+    },
+    {
+      "epoch": 0.17610639761522587,
+      "grad_norm": 0.908462959948242,
+      "learning_rate": 0.0009451992620361551,
+      "loss": 2.7395,
+      "step": 384
+    },
+    {
+      "epoch": 0.17656500802568217,
+      "grad_norm": 0.9279381991146389,
+      "learning_rate": 0.0009448605501726664,
+      "loss": 2.6218,
+      "step": 385
+    },
+    {
+      "epoch": 0.1770236184361385,
+      "grad_norm": 0.8148534095083383,
+      "learning_rate": 0.000944520855852081,
+      "loss": 2.7152,
+      "step": 386
+    },
+    {
+      "epoch": 0.17748222884659481,
+      "grad_norm": 0.9269241485465336,
+      "learning_rate": 0.0009441801798246002,
+      "loss": 2.6044,
+      "step": 387
+    },
+    {
+      "epoch": 0.17794083925705115,
+      "grad_norm": 0.9708355396798388,
+      "learning_rate": 0.0009438385228425939,
+      "loss": 2.6986,
+      "step": 388
+    },
+    {
+      "epoch": 0.17839944966750745,
+      "grad_norm": 0.9275670737385878,
+      "learning_rate": 0.0009434958856605982,
+      "loss": 2.7432,
+      "step": 389
+    },
+    {
+      "epoch": 0.17885806007796376,
+      "grad_norm": 0.8814011472032192,
+      "learning_rate": 0.0009431522690353137,
+      "loss": 2.7177,
+      "step": 390
+    },
+    {
+      "epoch": 0.1793166704884201,
+      "grad_norm": 0.8953983369057182,
+      "learning_rate": 0.0009428076737256044,
+      "loss": 2.5708,
+      "step": 391
+    },
+    {
+      "epoch": 0.1797752808988764,
+      "grad_norm": 0.9121797657203198,
+      "learning_rate": 0.0009424621004924954,
+      "loss": 2.6452,
+      "step": 392
+    },
+    {
+      "epoch": 0.18023389130933273,
+      "grad_norm": 0.9044112442212356,
+      "learning_rate": 0.0009421155500991719,
+      "loss": 2.6103,
+      "step": 393
+    },
+    {
+      "epoch": 0.18069250171978904,
+      "grad_norm": 0.8695820757180386,
+      "learning_rate": 0.0009417680233109767,
+      "loss": 2.7817,
+      "step": 394
+    },
+    {
+      "epoch": 0.18115111213024535,
+      "grad_norm": 0.8943740611160251,
+      "learning_rate": 0.000941419520895409,
+      "loss": 2.5907,
+      "step": 395
+    },
+    {
+      "epoch": 0.18160972254070168,
+      "grad_norm": 0.8627807765287066,
+      "learning_rate": 0.0009410700436221229,
+      "loss": 2.5868,
+      "step": 396
+    },
+    {
+      "epoch": 0.18206833295115799,
+      "grad_norm": 0.9917538369643955,
+      "learning_rate": 0.0009407195922629252,
+      "loss": 2.5677,
+      "step": 397
+    },
+    {
+      "epoch": 0.18252694336161432,
+      "grad_norm": 0.8941757703404072,
+      "learning_rate": 0.000940368167591774,
+      "loss": 2.5141,
+      "step": 398
+    },
+    {
+      "epoch": 0.18298555377207062,
+      "grad_norm": 0.9647030005588895,
+      "learning_rate": 0.0009400157703847769,
+      "loss": 2.6025,
+      "step": 399
+    },
+    {
+      "epoch": 0.18344416418252693,
+      "grad_norm": 0.9009611902741215,
+      "learning_rate": 0.0009396624014201895,
+      "loss": 2.6184,
+      "step": 400
+    },
+    {
+      "epoch": 0.18390277459298326,
+      "grad_norm": 0.9123580615165425,
+      "learning_rate": 0.000939308061478413,
+      "loss": 2.5803,
+      "step": 401
+    },
+    {
+      "epoch": 0.18436138500343957,
+      "grad_norm": 0.9879064841202294,
+      "learning_rate": 0.0009389527513419935,
+      "loss": 2.6634,
+      "step": 402
+    },
+    {
+      "epoch": 0.1848199954138959,
+      "grad_norm": 0.8244695239934074,
+      "learning_rate": 0.0009385964717956195,
+      "loss": 2.578,
+      "step": 403
+    },
+    {
+      "epoch": 0.1852786058243522,
+      "grad_norm": 0.8516866270589686,
+      "learning_rate": 0.0009382392236261201,
+      "loss": 2.5757,
+      "step": 404
+    },
+    {
+      "epoch": 0.18573721623480852,
+      "grad_norm": 0.8584678113500517,
+      "learning_rate": 0.0009378810076224644,
+      "loss": 2.6184,
+      "step": 405
+    },
+    {
+      "epoch": 0.18619582664526485,
+      "grad_norm": 0.8788706609024027,
+      "learning_rate": 0.0009375218245757582,
+      "loss": 2.6502,
+      "step": 406
+    },
+    {
+      "epoch": 0.18665443705572116,
+      "grad_norm": 0.8923618163257608,
+      "learning_rate": 0.0009371616752792432,
+      "loss": 2.6388,
+      "step": 407
+    },
+    {
+      "epoch": 0.1871130474661775,
+      "grad_norm": 0.8609612022206576,
+      "learning_rate": 0.0009368005605282949,
+      "loss": 2.6003,
+      "step": 408
+    },
+    {
+      "epoch": 0.1875716578766338,
+      "grad_norm": 0.8925720566852662,
+      "learning_rate": 0.0009364384811204212,
+      "loss": 2.6326,
+      "step": 409
+    },
+    {
+      "epoch": 0.18803026828709013,
+      "grad_norm": 0.9052682416575433,
+      "learning_rate": 0.00093607543785526,
+      "loss": 2.5956,
+      "step": 410
+    },
+    {
+      "epoch": 0.18848887869754644,
+      "grad_norm": 0.8824968175462476,
+      "learning_rate": 0.0009357114315345787,
+      "loss": 2.6454,
+      "step": 411
+    },
+    {
+      "epoch": 0.18894748910800274,
+      "grad_norm": 1.005261068671487,
+      "learning_rate": 0.0009353464629622705,
+      "loss": 2.6014,
+      "step": 412
+    },
+    {
+      "epoch": 0.18940609951845908,
+      "grad_norm": 0.9372507689475151,
+      "learning_rate": 0.0009349805329443544,
+      "loss": 2.5925,
+      "step": 413
+    },
+    {
+      "epoch": 0.18986470992891538,
+      "grad_norm": 0.8678943137657749,
+      "learning_rate": 0.0009346136422889724,
+      "loss": 2.5199,
+      "step": 414
+    },
+    {
+      "epoch": 0.19032332033937172,
+      "grad_norm": 0.959845056984446,
+      "learning_rate": 0.0009342457918063882,
+      "loss": 2.5783,
+      "step": 415
+    },
+    {
+      "epoch": 0.19078193074982802,
+      "grad_norm": 0.8897800079031656,
+      "learning_rate": 0.0009338769823089853,
+      "loss": 2.6373,
+      "step": 416
+    },
+    {
+      "epoch": 0.19124054116028433,
+      "grad_norm": 0.8421652442466547,
+      "learning_rate": 0.0009335072146112648,
+      "loss": 2.63,
+      "step": 417
+    },
+    {
+      "epoch": 0.19169915157074066,
+      "grad_norm": 0.9439774938643902,
+      "learning_rate": 0.0009331364895298444,
+      "loss": 2.7128,
+      "step": 418
+    },
+    {
+      "epoch": 0.19215776198119697,
+      "grad_norm": 0.9022526838926115,
+      "learning_rate": 0.0009327648078834559,
+      "loss": 2.4872,
+      "step": 419
+    },
+    {
+      "epoch": 0.1926163723916533,
+      "grad_norm": 1.0027609629409304,
+      "learning_rate": 0.0009323921704929434,
+      "loss": 2.6944,
+      "step": 420
+    },
+    {
+      "epoch": 0.1930749828021096,
+      "grad_norm": 1.022782292493877,
+      "learning_rate": 0.0009320185781812623,
+      "loss": 2.6569,
+      "step": 421
+    },
+    {
+      "epoch": 0.1935335932125659,
+      "grad_norm": 0.8958502453924275,
+      "learning_rate": 0.0009316440317734762,
+      "loss": 2.5428,
+      "step": 422
+    },
+    {
+      "epoch": 0.19399220362302225,
+      "grad_norm": 0.9539752622389781,
+      "learning_rate": 0.0009312685320967565,
+      "loss": 2.5901,
+      "step": 423
+    },
+    {
+      "epoch": 0.19445081403347855,
+      "grad_norm": 0.9366119413365693,
+      "learning_rate": 0.0009308920799803793,
+      "loss": 2.6433,
+      "step": 424
+    },
+    {
+      "epoch": 0.1949094244439349,
+      "grad_norm": 0.8753654071430201,
+      "learning_rate": 0.0009305146762557246,
+      "loss": 2.5398,
+      "step": 425
+    },
+    {
+      "epoch": 0.1953680348543912,
+      "grad_norm": 0.8647382865496251,
+      "learning_rate": 0.0009301363217562736,
+      "loss": 2.6064,
+      "step": 426
+    },
+    {
+      "epoch": 0.1958266452648475,
+      "grad_norm": 0.8663153182969454,
+      "learning_rate": 0.0009297570173176074,
+      "loss": 2.5916,
+      "step": 427
+    },
+    {
+      "epoch": 0.19628525567530383,
+      "grad_norm": 0.9087240913312226,
+      "learning_rate": 0.000929376763777405,
+      "loss": 2.6538,
+      "step": 428
+    },
+    {
+      "epoch": 0.19674386608576014,
+      "grad_norm": 0.8569107206802382,
+      "learning_rate": 0.0009289955619754413,
+      "loss": 2.5513,
+      "step": 429
+    },
+    {
+      "epoch": 0.19720247649621647,
+      "grad_norm": 0.8960060118105513,
+      "learning_rate": 0.0009286134127535859,
+      "loss": 2.6581,
+      "step": 430
+    },
+    {
+      "epoch": 0.19766108690667278,
+      "grad_norm": 0.851847453573282,
+      "learning_rate": 0.0009282303169558,
+      "loss": 2.5427,
+      "step": 431
+    },
+    {
+      "epoch": 0.1981196973171291,
+      "grad_norm": 0.8777908858857298,
+      "learning_rate": 0.0009278462754281359,
+      "loss": 2.5536,
+      "step": 432
+    },
+    {
+      "epoch": 0.19857830772758542,
+      "grad_norm": 0.9263915457670844,
+      "learning_rate": 0.0009274612890187342,
+      "loss": 2.5917,
+      "step": 433
+    },
+    {
+      "epoch": 0.19903691813804172,
+      "grad_norm": 0.916192986962901,
+      "learning_rate": 0.0009270753585778222,
+      "loss": 2.6448,
+      "step": 434
+    },
+    {
+      "epoch": 0.19949552854849806,
+      "grad_norm": 0.973838250428051,
+      "learning_rate": 0.0009266884849577124,
+      "loss": 2.4447,
+      "step": 435
+    },
+    {
+      "epoch": 0.19995413895895436,
+      "grad_norm": 0.8934514375610153,
+      "learning_rate": 0.0009263006690127998,
+      "loss": 2.534,
+      "step": 436
+    },
+    {
+      "epoch": 0.2004127493694107,
+      "grad_norm": 0.8487907914982208,
+      "learning_rate": 0.0009259119115995609,
+      "loss": 2.5696,
+      "step": 437
+    },
+    {
+      "epoch": 0.200871359779867,
+      "grad_norm": 0.8672389286582282,
+      "learning_rate": 0.0009255222135765511,
+      "loss": 2.5578,
+      "step": 438
+    },
+    {
+      "epoch": 0.2013299701903233,
+      "grad_norm": 0.957532811322003,
+      "learning_rate": 0.0009251315758044032,
+      "loss": 2.6767,
+      "step": 439
+    },
+    {
+      "epoch": 0.20178858060077964,
+      "grad_norm": 0.8842501244800263,
+      "learning_rate": 0.0009247399991458255,
+      "loss": 2.6044,
+      "step": 440
+    },
+    {
+      "epoch": 0.20224719101123595,
+      "grad_norm": 0.8929468461573081,
+      "learning_rate": 0.0009243474844655996,
+      "loss": 2.6143,
+      "step": 441
+    },
+    {
+      "epoch": 0.20270580142169228,
+      "grad_norm": 0.9715202036523445,
+      "learning_rate": 0.0009239540326305791,
+      "loss": 2.5341,
+      "step": 442
+    },
+    {
+      "epoch": 0.2031644118321486,
+      "grad_norm": 0.8656441364958021,
+      "learning_rate": 0.0009235596445096864,
+      "loss": 2.658,
+      "step": 443
+    },
+    {
+      "epoch": 0.2036230222426049,
+      "grad_norm": 0.8372345074641093,
+      "learning_rate": 0.0009231643209739127,
+      "loss": 2.6386,
+      "step": 444
+    },
+    {
+      "epoch": 0.20408163265306123,
+      "grad_norm": 0.8919079761382981,
+      "learning_rate": 0.0009227680628963145,
+      "loss": 2.5839,
+      "step": 445
+    },
+    {
+      "epoch": 0.20454024306351754,
+      "grad_norm": 0.8373007883419241,
+      "learning_rate": 0.000922370871152012,
+      "loss": 2.6355,
+      "step": 446
+    },
+    {
+      "epoch": 0.20499885347397387,
+      "grad_norm": 0.8541504777493839,
+      "learning_rate": 0.0009219727466181877,
+      "loss": 2.6011,
+      "step": 447
+    },
+    {
+      "epoch": 0.20545746388443017,
+      "grad_norm": 0.8769480731267422,
+      "learning_rate": 0.0009215736901740841,
+      "loss": 2.5995,
+      "step": 448
+    },
+    {
+      "epoch": 0.20591607429488648,
+      "grad_norm": 0.9146920352272172,
+      "learning_rate": 0.0009211737027010016,
+      "loss": 2.6979,
+      "step": 449
+    },
+    {
+      "epoch": 0.20637468470534281,
+      "grad_norm": 0.9392592672240379,
+      "learning_rate": 0.0009207727850822971,
+      "loss": 2.5632,
+      "step": 450
+    },
+    {
+      "epoch": 0.20683329511579912,
+      "grad_norm": 0.8104274547241686,
+      "learning_rate": 0.0009203709382033814,
+      "loss": 2.6275,
+      "step": 451
+    },
+    {
+      "epoch": 0.20729190552625545,
+      "grad_norm": 0.8677941805324159,
+      "learning_rate": 0.0009199681629517173,
+      "loss": 2.5284,
+      "step": 452
+    },
+    {
+      "epoch": 0.20775051593671176,
+      "grad_norm": 0.9782676900195534,
+      "learning_rate": 0.0009195644602168184,
+      "loss": 2.5944,
+      "step": 453
+    },
+    {
+      "epoch": 0.2082091263471681,
+      "grad_norm": 0.9270939866989588,
+      "learning_rate": 0.0009191598308902464,
+      "loss": 2.4919,
+      "step": 454
+    },
+    {
+      "epoch": 0.2086677367576244,
+      "grad_norm": 1.01982665098646,
+      "learning_rate": 0.0009187542758656091,
+      "loss": 2.57,
+      "step": 455
+    },
+    {
+      "epoch": 0.2091263471680807,
+      "grad_norm": 0.9368728909562595,
+      "learning_rate": 0.0009183477960385591,
+      "loss": 2.4898,
+      "step": 456
+    },
+    {
+      "epoch": 0.20958495757853704,
+      "grad_norm": 0.8803488674494734,
+      "learning_rate": 0.0009179403923067912,
+      "loss": 2.631,
+      "step": 457
+    },
+    {
+      "epoch": 0.21004356798899335,
+      "grad_norm": 0.9589618910389339,
+      "learning_rate": 0.0009175320655700406,
+      "loss": 2.5002,
+      "step": 458
+    },
+    {
+      "epoch": 0.21050217839944968,
+      "grad_norm": 0.920343393432829,
+      "learning_rate": 0.0009171228167300805,
+      "loss": 2.5202,
+      "step": 459
+    },
+    {
+      "epoch": 0.21096078880990599,
+      "grad_norm": 0.8738524172733814,
+      "learning_rate": 0.0009167126466907215,
+      "loss": 2.5868,
+      "step": 460
+    },
+    {
+      "epoch": 0.2114193992203623,
+      "grad_norm": 0.9123938823724274,
+      "learning_rate": 0.0009163015563578074,
+      "loss": 2.5776,
+      "step": 461
+    },
+    {
+      "epoch": 0.21187800963081863,
+      "grad_norm": 0.8626820745860205,
+      "learning_rate": 0.0009158895466392158,
+      "loss": 2.7009,
+      "step": 462
+    },
+    {
+      "epoch": 0.21233662004127493,
+      "grad_norm": 0.9262635737198887,
+      "learning_rate": 0.0009154766184448535,
+      "loss": 2.5134,
+      "step": 463
+    },
+    {
+      "epoch": 0.21279523045173127,
+      "grad_norm": 1.0450195812731176,
+      "learning_rate": 0.0009150627726866568,
+      "loss": 2.4693,
+      "step": 464
+    },
+    {
+      "epoch": 0.21325384086218757,
+      "grad_norm": 1.0115200488177754,
+      "learning_rate": 0.000914648010278587,
+      "loss": 2.4724,
+      "step": 465
+    },
+    {
+      "epoch": 0.21371245127264388,
+      "grad_norm": 1.0198412546984914,
+      "learning_rate": 0.0009142323321366315,
+      "loss": 2.655,
+      "step": 466
+    },
+    {
+      "epoch": 0.2141710616831002,
+      "grad_norm": 0.9783962067076114,
+      "learning_rate": 0.0009138157391787986,
+      "loss": 2.6702,
+      "step": 467
+    },
+    {
+      "epoch": 0.21462967209355652,
+      "grad_norm": 0.882681016373271,
+      "learning_rate": 0.0009133982323251177,
+      "loss": 2.5443,
+      "step": 468
+    },
+    {
+      "epoch": 0.21508828250401285,
+      "grad_norm": 0.942893512639318,
+      "learning_rate": 0.0009129798124976365,
+      "loss": 2.6357,
+      "step": 469
+    },
+    {
+      "epoch": 0.21554689291446916,
+      "grad_norm": 0.8252427070948415,
+      "learning_rate": 0.0009125604806204187,
+      "loss": 2.5717,
+      "step": 470
+    },
+    {
+      "epoch": 0.21600550332492546,
+      "grad_norm": 0.8920570130902375,
+      "learning_rate": 0.0009121402376195421,
+      "loss": 2.4422,
+      "step": 471
+    },
+    {
+      "epoch": 0.2164641137353818,
+      "grad_norm": 0.9110451608675965,
+      "learning_rate": 0.0009117190844230972,
+      "loss": 2.6003,
+      "step": 472
+    },
+    {
+      "epoch": 0.2169227241458381,
+      "grad_norm": 0.898226875742386,
+      "learning_rate": 0.0009112970219611841,
+      "loss": 2.4852,
+      "step": 473
+    },
+    {
+      "epoch": 0.21738133455629444,
+      "grad_norm": 0.8493207038950716,
+      "learning_rate": 0.0009108740511659115,
+      "loss": 2.6116,
+      "step": 474
+    },
+    {
+      "epoch": 0.21783994496675074,
+      "grad_norm": 0.8654432135875829,
+      "learning_rate": 0.0009104501729713935,
+      "loss": 2.551,
+      "step": 475
+    },
+    {
+      "epoch": 0.21829855537720708,
+      "grad_norm": 0.8523043847385097,
+      "learning_rate": 0.0009100253883137488,
+      "loss": 2.5069,
+      "step": 476
+    },
+    {
+      "epoch": 0.21875716578766338,
+      "grad_norm": 0.8807346071993731,
+      "learning_rate": 0.0009095996981310974,
+      "loss": 2.5247,
+      "step": 477
+    },
+    {
+      "epoch": 0.2192157761981197,
+      "grad_norm": 0.8535314362472305,
+      "learning_rate": 0.0009091731033635596,
+      "loss": 2.5142,
+      "step": 478
+    },
+    {
+      "epoch": 0.21967438660857602,
+      "grad_norm": 0.9308444465756007,
+      "learning_rate": 0.0009087456049532529,
+      "loss": 2.5575,
+      "step": 479
+    },
+    {
+      "epoch": 0.22013299701903233,
+      "grad_norm": 0.9302282114909308,
+      "learning_rate": 0.0009083172038442914,
+      "loss": 2.5577,
+      "step": 480
+    },
+    {
+      "epoch": 0.22059160742948866,
+      "grad_norm": 0.8454881102862536,
+      "learning_rate": 0.0009078879009827817,
+      "loss": 2.4668,
+      "step": 481
+    },
+    {
+      "epoch": 0.22105021783994497,
+      "grad_norm": 0.9629803077692896,
+      "learning_rate": 0.0009074576973168223,
+      "loss": 2.5709,
+      "step": 482
+    },
+    {
+      "epoch": 0.22150882825040127,
+      "grad_norm": 0.8728284707643172,
+      "learning_rate": 0.0009070265937965015,
+      "loss": 2.545,
+      "step": 483
+    },
+    {
+      "epoch": 0.2219674386608576,
+      "grad_norm": 0.9735232828342999,
+      "learning_rate": 0.0009065945913738942,
+      "loss": 2.6356,
+      "step": 484
+    },
+    {
+      "epoch": 0.2224260490713139,
+      "grad_norm": 0.9010786029668294,
+      "learning_rate": 0.0009061616910030609,
+      "loss": 2.5578,
+      "step": 485
+    },
+    {
+      "epoch": 0.22288465948177025,
+      "grad_norm": 0.9868292004323551,
+      "learning_rate": 0.0009057278936400453,
+      "loss": 2.5281,
+      "step": 486
+    },
+    {
+      "epoch": 0.22334326989222655,
+      "grad_norm": 0.9017639968951041,
+      "learning_rate": 0.0009052932002428715,
+      "loss": 2.5371,
+      "step": 487
+    },
+    {
+      "epoch": 0.22380188030268286,
+      "grad_norm": 0.8352890186075146,
+      "learning_rate": 0.0009048576117715435,
+      "loss": 2.5075,
+      "step": 488
+    },
+    {
+      "epoch": 0.2242604907131392,
+      "grad_norm": 0.9071619604420831,
+      "learning_rate": 0.0009044211291880407,
+      "loss": 2.4913,
+      "step": 489
+    },
+    {
+      "epoch": 0.2247191011235955,
+      "grad_norm": 0.9515268709611919,
+      "learning_rate": 0.000903983753456318,
+      "loss": 2.4501,
+      "step": 490
+    },
+    {
+      "epoch": 0.22517771153405183,
+      "grad_norm": 0.9270549472505569,
+      "learning_rate": 0.0009035454855423026,
+      "loss": 2.5299,
+      "step": 491
+    },
+    {
+      "epoch": 0.22563632194450814,
+      "grad_norm": 0.8735797425848812,
+      "learning_rate": 0.0009031063264138922,
+      "loss": 2.4819,
+      "step": 492
+    },
+    {
+      "epoch": 0.22609493235496445,
+      "grad_norm": 0.9257576096346717,
+      "learning_rate": 0.0009026662770409522,
+      "loss": 2.4974,
+      "step": 493
+    },
+    {
+      "epoch": 0.22655354276542078,
+      "grad_norm": 0.9518973204307429,
+      "learning_rate": 0.0009022253383953147,
+      "loss": 2.5344,
+      "step": 494
+    },
+    {
+      "epoch": 0.22701215317587708,
+      "grad_norm": 1.0197897307424462,
+      "learning_rate": 0.0009017835114507753,
+      "loss": 2.534,
+      "step": 495
+    },
+    {
+      "epoch": 0.22747076358633342,
+      "grad_norm": 0.8506151664723305,
+      "learning_rate": 0.0009013407971830914,
+      "loss": 2.5606,
+      "step": 496
+    },
+    {
+      "epoch": 0.22792937399678972,
+      "grad_norm": 0.9133050475276138,
+      "learning_rate": 0.0009008971965699801,
+      "loss": 2.6326,
+      "step": 497
+    },
+    {
+      "epoch": 0.22838798440724603,
+      "grad_norm": 0.8886623069947645,
+      "learning_rate": 0.0009004527105911163,
+      "loss": 2.6235,
+      "step": 498
+    },
+    {
+      "epoch": 0.22884659481770236,
+      "grad_norm": 0.9588193806729778,
+      "learning_rate": 0.0009000073402281295,
+      "loss": 2.5656,
+      "step": 499
+    },
+    {
+      "epoch": 0.22930520522815867,
+      "grad_norm": 0.8555508832989588,
+      "learning_rate": 0.0008995610864646028,
+      "loss": 2.4688,
+      "step": 500
+    },
+    {
+      "epoch": 0.229763815638615,
+      "grad_norm": 0.9374531184087398,
+      "learning_rate": 0.0008991139502860703,
+      "loss": 2.4749,
+      "step": 501
+    },
+    {
+      "epoch": 0.2302224260490713,
+      "grad_norm": 0.9137085147414329,
+      "learning_rate": 0.0008986659326800146,
+      "loss": 2.4868,
+      "step": 502
+    },
+    {
+      "epoch": 0.23068103645952764,
+      "grad_norm": 0.9515463021213008,
+      "learning_rate": 0.0008982170346358651,
+      "loss": 2.497,
+      "step": 503
+    },
+    {
+      "epoch": 0.23113964686998395,
+      "grad_norm": 0.838627153460125,
+      "learning_rate": 0.0008977672571449956,
+      "loss": 2.4872,
+      "step": 504
+    },
+    {
+      "epoch": 0.23159825728044026,
+      "grad_norm": 0.9149033287127447,
+      "learning_rate": 0.0008973166012007217,
+      "loss": 2.6279,
+      "step": 505
+    },
+    {
+      "epoch": 0.2320568676908966,
+      "grad_norm": 0.8585549854028185,
+      "learning_rate": 0.0008968650677982998,
+      "loss": 2.4223,
+      "step": 506
+    },
+    {
+      "epoch": 0.2325154781013529,
+      "grad_norm": 0.8907279345409636,
+      "learning_rate": 0.0008964126579349236,
+      "loss": 2.4684,
+      "step": 507
+    },
+    {
+      "epoch": 0.23297408851180923,
+      "grad_norm": 0.8292335410405693,
+      "learning_rate": 0.0008959593726097226,
+      "loss": 2.4929,
+      "step": 508
+    },
+    {
+      "epoch": 0.23343269892226554,
+      "grad_norm": 0.8925431342682446,
+      "learning_rate": 0.0008955052128237596,
+      "loss": 2.4667,
+      "step": 509
+    },
+    {
+      "epoch": 0.23389130933272184,
+      "grad_norm": 0.8968056772280529,
+      "learning_rate": 0.0008950501795800288,
+      "loss": 2.5965,
+      "step": 510
+    },
+    {
+      "epoch": 0.23434991974317818,
+      "grad_norm": 0.833787516325598,
+      "learning_rate": 0.0008945942738834532,
+      "loss": 2.4791,
+      "step": 511
+    },
+    {
+      "epoch": 0.23480853015363448,
+      "grad_norm": 0.8325577308680987,
+      "learning_rate": 0.0008941374967408826,
+      "loss": 2.5115,
+      "step": 512
+    },
+    {
+      "epoch": 0.23526714056409082,
+      "grad_norm": 0.9076490294037137,
+      "learning_rate": 0.0008936798491610916,
+      "loss": 2.5111,
+      "step": 513
+    },
+    {
+      "epoch": 0.23572575097454712,
+      "grad_norm": 0.8729797787266882,
+      "learning_rate": 0.0008932213321547768,
+      "loss": 2.5529,
+      "step": 514
+    },
+    {
+      "epoch": 0.23618436138500343,
+      "grad_norm": 0.9480905974873385,
+      "learning_rate": 0.0008927619467345554,
+      "loss": 2.4835,
+      "step": 515
+    },
+    {
+      "epoch": 0.23664297179545976,
+      "grad_norm": 0.9952809944885523,
+      "learning_rate": 0.0008923016939149615,
+      "loss": 2.5389,
+      "step": 516
+    },
+    {
+      "epoch": 0.23710158220591607,
+      "grad_norm": 0.8343881220839964,
+      "learning_rate": 0.0008918405747124458,
+      "loss": 2.392,
+      "step": 517
+    },
+    {
+      "epoch": 0.2375601926163724,
+      "grad_norm": 0.8347440130034386,
+      "learning_rate": 0.0008913785901453721,
+      "loss": 2.5125,
+      "step": 518
+    },
+    {
+      "epoch": 0.2380188030268287,
+      "grad_norm": 0.8487726040922298,
+      "learning_rate": 0.000890915741234015,
+      "loss": 2.5193,
+      "step": 519
+    },
+    {
+      "epoch": 0.238477413437285,
+      "grad_norm": 0.9238437282427097,
+      "learning_rate": 0.0008904520290005582,
+      "loss": 2.5528,
+      "step": 520
+    },
+    {
+      "epoch": 0.23893602384774135,
+      "grad_norm": 0.8508797585570004,
+      "learning_rate": 0.000889987454469092,
+      "loss": 2.5651,
+      "step": 521
+    },
+    {
+      "epoch": 0.23939463425819765,
+      "grad_norm": 0.8860570019941831,
+      "learning_rate": 0.0008895220186656111,
+      "loss": 2.4221,
+      "step": 522
+    },
+    {
+      "epoch": 0.239853244668654,
+      "grad_norm": 0.8875302395301381,
+      "learning_rate": 0.0008890557226180122,
+      "loss": 2.4744,
+      "step": 523
+    },
+    {
+      "epoch": 0.2403118550791103,
+      "grad_norm": 0.9197784852630804,
+      "learning_rate": 0.0008885885673560921,
+      "loss": 2.4423,
+      "step": 524
+    },
+    {
+      "epoch": 0.24077046548956663,
+      "grad_norm": 0.9767154812793389,
+      "learning_rate": 0.0008881205539115444,
+      "loss": 2.5751,
+      "step": 525
+    },
+    {
+      "epoch": 0.24122907590002293,
+      "grad_norm": 0.8745996082545586,
+      "learning_rate": 0.0008876516833179589,
+      "loss": 2.4858,
+      "step": 526
+    },
+    {
+      "epoch": 0.24168768631047924,
+      "grad_norm": 0.8913014922217518,
+      "learning_rate": 0.0008871819566108177,
+      "loss": 2.4083,
+      "step": 527
+    },
+    {
+      "epoch": 0.24214629672093557,
+      "grad_norm": 0.9040564248655867,
+      "learning_rate": 0.000886711374827494,
+      "loss": 2.4386,
+      "step": 528
+    },
+    {
+      "epoch": 0.24260490713139188,
+      "grad_norm": 0.8867157073680314,
+      "learning_rate": 0.0008862399390072491,
+      "loss": 2.5129,
+      "step": 529
+    },
+    {
+      "epoch": 0.2430635175418482,
+      "grad_norm": 0.9240996823785252,
+      "learning_rate": 0.0008857676501912305,
+      "loss": 2.4719,
+      "step": 530
+    },
+    {
+      "epoch": 0.24352212795230452,
+      "grad_norm": 0.8622236556332544,
+      "learning_rate": 0.0008852945094224697,
+      "loss": 2.5124,
+      "step": 531
+    },
+    {
+      "epoch": 0.24398073836276082,
+      "grad_norm": 0.8763055359086055,
+      "learning_rate": 0.0008848205177458795,
+      "loss": 2.5607,
+      "step": 532
+    },
+    {
+      "epoch": 0.24443934877321716,
+      "grad_norm": 0.8283967664692398,
+      "learning_rate": 0.0008843456762082518,
+      "loss": 2.4217,
+      "step": 533
+    },
+    {
+      "epoch": 0.24489795918367346,
+      "grad_norm": 0.8930720310010587,
+      "learning_rate": 0.0008838699858582557,
+      "loss": 2.5243,
+      "step": 534
+    },
+    {
+      "epoch": 0.2453565695941298,
+      "grad_norm": 0.8939547154368259,
+      "learning_rate": 0.0008833934477464347,
+      "loss": 2.5977,
+      "step": 535
+    },
+    {
+      "epoch": 0.2458151800045861,
+      "grad_norm": 0.8609522805526152,
+      "learning_rate": 0.0008829160629252045,
+      "loss": 2.4801,
+      "step": 536
+    },
+    {
+      "epoch": 0.2462737904150424,
+      "grad_norm": 0.9094283476979644,
+      "learning_rate": 0.0008824378324488509,
+      "loss": 2.5803,
+      "step": 537
+    },
+    {
+      "epoch": 0.24673240082549874,
+      "grad_norm": 0.8743690270675394,
+      "learning_rate": 0.0008819587573735268,
+      "loss": 2.5557,
+      "step": 538
+    },
+    {
+      "epoch": 0.24719101123595505,
+      "grad_norm": 0.8324836222254035,
+      "learning_rate": 0.0008814788387572513,
+      "loss": 2.4112,
+      "step": 539
+    },
+    {
+      "epoch": 0.24764962164641138,
+      "grad_norm": 0.8714459941857308,
+      "learning_rate": 0.0008809980776599053,
+      "loss": 2.5018,
+      "step": 540
+    },
+    {
+      "epoch": 0.2481082320568677,
+      "grad_norm": 0.9750223978703612,
+      "learning_rate": 0.0008805164751432312,
+      "loss": 2.3782,
+      "step": 541
+    },
+    {
+      "epoch": 0.248566842467324,
+      "grad_norm": 0.8427658690229318,
+      "learning_rate": 0.0008800340322708292,
+      "loss": 2.5722,
+      "step": 542
+    },
+    {
+      "epoch": 0.24902545287778033,
+      "grad_norm": 0.8245358459180856,
+      "learning_rate": 0.0008795507501081555,
+      "loss": 2.4209,
+      "step": 543
+    },
+    {
+      "epoch": 0.24948406328823663,
+      "grad_norm": 0.8826645718560236,
+      "learning_rate": 0.0008790666297225196,
+      "loss": 2.4762,
+      "step": 544
+    },
+    {
+      "epoch": 0.24994267369869297,
+      "grad_norm": 0.9290548670127707,
+      "learning_rate": 0.0008785816721830829,
+      "loss": 2.534,
+      "step": 545
+    },
+    {
+      "epoch": 0.2504012841091493,
+      "grad_norm": 0.8940951136760907,
+      "learning_rate": 0.0008780958785608546,
+      "loss": 2.6153,
+      "step": 546
+    },
+    {
+      "epoch": 0.2508598945196056,
+      "grad_norm": 0.9609493160030156,
+      "learning_rate": 0.0008776092499286912,
+      "loss": 2.4522,
+      "step": 547
+    },
+    {
+      "epoch": 0.2513185049300619,
+      "grad_norm": 0.8826395949163448,
+      "learning_rate": 0.0008771217873612929,
+      "loss": 2.4383,
+      "step": 548
+    },
+    {
+      "epoch": 0.2517771153405182,
+      "grad_norm": 0.9476779897389547,
+      "learning_rate": 0.0008766334919352017,
+      "loss": 2.5836,
+      "step": 549
+    },
+    {
+      "epoch": 0.2522357257509745,
+      "grad_norm": 0.8840552621650305,
+      "learning_rate": 0.0008761443647287987,
+      "loss": 2.4735,
+      "step": 550
+    },
+    {
+      "epoch": 0.2526943361614309,
+      "grad_norm": 0.9087923984504278,
+      "learning_rate": 0.0008756544068223026,
+      "loss": 2.5395,
+      "step": 551
+    },
+    {
+      "epoch": 0.2531529465718872,
+      "grad_norm": 0.8512949023421487,
+      "learning_rate": 0.0008751636192977659,
+      "loss": 2.5258,
+      "step": 552
+    },
+    {
+      "epoch": 0.2536115569823435,
+      "grad_norm": 0.873922737410195,
+      "learning_rate": 0.0008746720032390737,
+      "loss": 2.5752,
+      "step": 553
+    },
+    {
+      "epoch": 0.2540701673927998,
+      "grad_norm": 0.8439708346284819,
+      "learning_rate": 0.0008741795597319408,
+      "loss": 2.5326,
+      "step": 554
+    },
+    {
+      "epoch": 0.2545287778032561,
+      "grad_norm": 0.9096947407596574,
+      "learning_rate": 0.0008736862898639095,
+      "loss": 2.5143,
+      "step": 555
+    },
+    {
+      "epoch": 0.2549873882137125,
+      "grad_norm": 0.8696114052331981,
+      "learning_rate": 0.0008731921947243468,
+      "loss": 2.5423,
+      "step": 556
+    },
+    {
+      "epoch": 0.2554459986241688,
+      "grad_norm": 0.8939893570127709,
+      "learning_rate": 0.0008726972754044427,
+      "loss": 2.4814,
+      "step": 557
+    },
+    {
+      "epoch": 0.2559046090346251,
+      "grad_norm": 0.8279169675770515,
+      "learning_rate": 0.0008722015329972069,
+      "loss": 2.5652,
+      "step": 558
+    },
+    {
+      "epoch": 0.2563632194450814,
+      "grad_norm": 0.8728283739992327,
+      "learning_rate": 0.0008717049685974672,
+      "loss": 2.5315,
+      "step": 559
+    },
+    {
+      "epoch": 0.2568218298555377,
+      "grad_norm": 0.8598309199559669,
+      "learning_rate": 0.0008712075833018665,
+      "loss": 2.4893,
+      "step": 560
+    },
+    {
+      "epoch": 0.25728044026599406,
+      "grad_norm": 0.9097058699877446,
+      "learning_rate": 0.0008707093782088608,
+      "loss": 2.4396,
+      "step": 561
+    },
+    {
+      "epoch": 0.25773905067645037,
+      "grad_norm": 0.8651350278293674,
+      "learning_rate": 0.0008702103544187167,
+      "loss": 2.4288,
+      "step": 562
+    },
+    {
+      "epoch": 0.25819766108690667,
+      "grad_norm": 0.9270676412069445,
+      "learning_rate": 0.0008697105130335085,
+      "loss": 2.5412,
+      "step": 563
+    },
+    {
+      "epoch": 0.258656271497363,
+      "grad_norm": 0.8693701174885781,
+      "learning_rate": 0.0008692098551571164,
+      "loss": 2.5141,
+      "step": 564
+    },
+    {
+      "epoch": 0.2591148819078193,
+      "grad_norm": 0.8886117221836541,
+      "learning_rate": 0.0008687083818952235,
+      "loss": 2.5363,
+      "step": 565
+    },
+    {
+      "epoch": 0.25957349231827564,
+      "grad_norm": 0.8977534573782201,
+      "learning_rate": 0.0008682060943553143,
+      "loss": 2.5197,
+      "step": 566
+    },
+    {
+      "epoch": 0.26003210272873195,
+      "grad_norm": 0.9696031543350739,
+      "learning_rate": 0.0008677029936466707,
+      "loss": 2.4961,
+      "step": 567
+    },
+    {
+      "epoch": 0.26049071313918826,
+      "grad_norm": 0.8580052301604953,
+      "learning_rate": 0.0008671990808803711,
+      "loss": 2.5983,
+      "step": 568
+    },
+    {
+      "epoch": 0.26094932354964456,
+      "grad_norm": 0.9642304115753816,
+      "learning_rate": 0.0008666943571692871,
+      "loss": 2.5043,
+      "step": 569
+    },
+    {
+      "epoch": 0.26140793396010087,
+      "grad_norm": 0.8747450470954847,
+      "learning_rate": 0.0008661888236280813,
+      "loss": 2.5385,
+      "step": 570
+    },
+    {
+      "epoch": 0.26186654437055723,
+      "grad_norm": 0.8244723935309609,
+      "learning_rate": 0.0008656824813732045,
+      "loss": 2.3611,
+      "step": 571
+    },
+    {
+      "epoch": 0.26232515478101354,
+      "grad_norm": 0.8164473391675382,
+      "learning_rate": 0.000865175331522894,
+      "loss": 2.4564,
+      "step": 572
+    },
+    {
+      "epoch": 0.26278376519146984,
+      "grad_norm": 0.8344591361399222,
+      "learning_rate": 0.0008646673751971703,
+      "loss": 2.4057,
+      "step": 573
+    },
+    {
+      "epoch": 0.26324237560192615,
+      "grad_norm": 0.8595253149497238,
+      "learning_rate": 0.000864158613517835,
+      "loss": 2.4082,
+      "step": 574
+    },
+    {
+      "epoch": 0.26370098601238245,
+      "grad_norm": 0.8505227711339117,
+      "learning_rate": 0.0008636490476084681,
+      "loss": 2.4913,
+      "step": 575
+    },
+    {
+      "epoch": 0.2641595964228388,
+      "grad_norm": 0.8490168814697231,
+      "learning_rate": 0.0008631386785944264,
+      "loss": 2.5311,
+      "step": 576
+    },
+    {
+      "epoch": 0.2646182068332951,
+      "grad_norm": 0.922737766211867,
+      "learning_rate": 0.0008626275076028397,
+      "loss": 2.5494,
+      "step": 577
+    },
+    {
+      "epoch": 0.26507681724375143,
+      "grad_norm": 0.8560880402577419,
+      "learning_rate": 0.0008621155357626091,
+      "loss": 2.5175,
+      "step": 578
+    },
+    {
+      "epoch": 0.26553542765420773,
+      "grad_norm": 0.8771132504722752,
+      "learning_rate": 0.0008616027642044042,
+      "loss": 2.4681,
+      "step": 579
+    },
+    {
+      "epoch": 0.2659940380646641,
+      "grad_norm": 0.8205514694225141,
+      "learning_rate": 0.000861089194060661,
+      "loss": 2.5444,
+      "step": 580
+    },
+    {
+      "epoch": 0.2664526484751204,
+      "grad_norm": 0.8208307313461489,
+      "learning_rate": 0.000860574826465579,
+      "loss": 2.4165,
+      "step": 581
+    },
+    {
+      "epoch": 0.2669112588855767,
+      "grad_norm": 0.8570858012574288,
+      "learning_rate": 0.0008600596625551191,
+      "loss": 2.5486,
+      "step": 582
+    },
+    {
+      "epoch": 0.267369869296033,
+      "grad_norm": 0.8640845137816535,
+      "learning_rate": 0.0008595437034670006,
+      "loss": 2.4808,
+      "step": 583
+    },
+    {
+      "epoch": 0.2678284797064893,
+      "grad_norm": 0.857588085416462,
+      "learning_rate": 0.0008590269503406985,
+      "loss": 2.4602,
+      "step": 584
+    },
+    {
+      "epoch": 0.2682870901169457,
+      "grad_norm": 0.7952818036654671,
+      "learning_rate": 0.0008585094043174423,
+      "loss": 2.4825,
+      "step": 585
+    },
+    {
+      "epoch": 0.268745700527402,
+      "grad_norm": 0.8555030560718313,
+      "learning_rate": 0.0008579910665402118,
+      "loss": 2.5366,
+      "step": 586
+    },
+    {
+      "epoch": 0.2692043109378583,
+      "grad_norm": 0.7927482122793815,
+      "learning_rate": 0.000857471938153736,
+      "loss": 2.5852,
+      "step": 587
+    },
+    {
+      "epoch": 0.2696629213483146,
+      "grad_norm": 0.8642606065053011,
+      "learning_rate": 0.0008569520203044892,
+      "loss": 2.4842,
+      "step": 588
+    },
+    {
+      "epoch": 0.2701215317587709,
+      "grad_norm": 0.876485856841674,
+      "learning_rate": 0.0008564313141406901,
+      "loss": 2.4341,
+      "step": 589
+    },
+    {
+      "epoch": 0.27058014216922727,
+      "grad_norm": 0.8133630859099057,
+      "learning_rate": 0.0008559098208122973,
+      "loss": 2.4486,
+      "step": 590
+    },
+    {
+      "epoch": 0.2710387525796836,
+      "grad_norm": 0.8507858736590277,
+      "learning_rate": 0.0008553875414710089,
+      "loss": 2.5178,
+      "step": 591
+    },
+    {
+      "epoch": 0.2714973629901399,
+      "grad_norm": 0.8594197776694759,
+      "learning_rate": 0.0008548644772702579,
+      "loss": 2.427,
+      "step": 592
+    },
+    {
+      "epoch": 0.2719559734005962,
+      "grad_norm": 0.8160885386295565,
+      "learning_rate": 0.0008543406293652116,
+      "loss": 2.5016,
+      "step": 593
+    },
+    {
+      "epoch": 0.2724145838110525,
+      "grad_norm": 0.8711321396905496,
+      "learning_rate": 0.0008538159989127671,
+      "loss": 2.4613,
+      "step": 594
+    },
+    {
+      "epoch": 0.27287319422150885,
+      "grad_norm": 0.8213920061333921,
+      "learning_rate": 0.0008532905870715505,
+      "loss": 2.4819,
+      "step": 595
+    },
+    {
+      "epoch": 0.27333180463196516,
+      "grad_norm": 0.8170070214467201,
+      "learning_rate": 0.0008527643950019131,
+      "loss": 2.5199,
+      "step": 596
+    },
+    {
+      "epoch": 0.27379041504242146,
+      "grad_norm": 0.7971696358981301,
+      "learning_rate": 0.0008522374238659296,
+      "loss": 2.4444,
+      "step": 597
+    },
+    {
+      "epoch": 0.27424902545287777,
+      "grad_norm": 0.8452188950655856,
+      "learning_rate": 0.0008517096748273951,
+      "loss": 2.4059,
+      "step": 598
+    },
+    {
+      "epoch": 0.2747076358633341,
+      "grad_norm": 0.8475077233971497,
+      "learning_rate": 0.0008511811490518227,
+      "loss": 2.4942,
+      "step": 599
+    },
+    {
+      "epoch": 0.27516624627379044,
+      "grad_norm": 0.9301390662857157,
+      "learning_rate": 0.0008506518477064405,
+      "loss": 2.5738,
+      "step": 600
+    },
+    {
+      "epoch": 0.27562485668424674,
+      "grad_norm": 0.946815898747415,
+      "learning_rate": 0.0008501217719601903,
+      "loss": 2.4087,
+      "step": 601
+    },
+    {
+      "epoch": 0.27608346709470305,
+      "grad_norm": 0.9323512311012846,
+      "learning_rate": 0.0008495909229837233,
+      "loss": 2.5485,
+      "step": 602
+    },
+    {
+      "epoch": 0.27654207750515936,
+      "grad_norm": 0.9466645770965414,
+      "learning_rate": 0.000849059301949399,
+      "loss": 2.4637,
+      "step": 603
+    },
+    {
+      "epoch": 0.27700068791561566,
+      "grad_norm": 0.8746210350895612,
+      "learning_rate": 0.0008485269100312812,
+      "loss": 2.4407,
+      "step": 604
+    },
+    {
+      "epoch": 0.277459298326072,
+      "grad_norm": 0.8240855378609937,
+      "learning_rate": 0.0008479937484051368,
+      "loss": 2.5037,
+      "step": 605
+    },
+    {
+      "epoch": 0.27791790873652833,
+      "grad_norm": 0.8473704285098753,
+      "learning_rate": 0.0008474598182484323,
+      "loss": 2.563,
+      "step": 606
+    },
+    {
+      "epoch": 0.27837651914698464,
+      "grad_norm": 0.9017138599733469,
+      "learning_rate": 0.0008469251207403317,
+      "loss": 2.4611,
+      "step": 607
+    },
+    {
+      "epoch": 0.27883512955744094,
+      "grad_norm": 0.8462842048712057,
+      "learning_rate": 0.0008463896570616934,
+      "loss": 2.4549,
+      "step": 608
+    },
+    {
+      "epoch": 0.27929373996789725,
+      "grad_norm": 0.799612209383647,
+      "learning_rate": 0.0008458534283950678,
+      "loss": 2.4726,
+      "step": 609
+    },
+    {
+      "epoch": 0.2797523503783536,
+      "grad_norm": 0.8929586493791423,
+      "learning_rate": 0.0008453164359246952,
+      "loss": 2.4583,
+      "step": 610
+    },
+    {
+      "epoch": 0.2802109607888099,
+      "grad_norm": 0.8367234745257335,
+      "learning_rate": 0.0008447786808365022,
+      "loss": 2.4371,
+      "step": 611
+    },
+    {
+      "epoch": 0.2806695711992662,
+      "grad_norm": 0.8487260363488516,
+      "learning_rate": 0.0008442401643181,
+      "loss": 2.5943,
+      "step": 612
+    },
+    {
+      "epoch": 0.2811281816097225,
+      "grad_norm": 0.8230776104381492,
+      "learning_rate": 0.0008437008875587811,
+      "loss": 2.4547,
+      "step": 613
+    },
+    {
+      "epoch": 0.28158679202017883,
+      "grad_norm": 0.9026997698478793,
+      "learning_rate": 0.0008431608517495171,
+      "loss": 2.4306,
+      "step": 614
+    },
+    {
+      "epoch": 0.2820454024306352,
+      "grad_norm": 0.8480884227904713,
+      "learning_rate": 0.0008426200580829561,
+      "loss": 2.3973,
+      "step": 615
+    },
+    {
+      "epoch": 0.2825040128410915,
+      "grad_norm": 0.805841110998099,
+      "learning_rate": 0.0008420785077534195,
+      "loss": 2.4531,
+      "step": 616
+    },
+    {
+      "epoch": 0.2829626232515478,
+      "grad_norm": 0.9035588123595888,
+      "learning_rate": 0.0008415362019569001,
+      "loss": 2.5385,
+      "step": 617
+    },
+    {
+      "epoch": 0.2834212336620041,
+      "grad_norm": 0.8619394243948544,
+      "learning_rate": 0.0008409931418910591,
+      "loss": 2.4444,
+      "step": 618
+    },
+    {
+      "epoch": 0.2838798440724604,
+      "grad_norm": 0.9835818493103157,
+      "learning_rate": 0.0008404493287552232,
+      "loss": 2.5446,
+      "step": 619
+    },
+    {
+      "epoch": 0.2843384544829168,
+      "grad_norm": 0.8344430717675383,
+      "learning_rate": 0.0008399047637503825,
+      "loss": 2.4666,
+      "step": 620
+    },
+    {
+      "epoch": 0.2847970648933731,
+      "grad_norm": 0.8311665662924184,
+      "learning_rate": 0.0008393594480791875,
+      "loss": 2.4406,
+      "step": 621
+    },
+    {
+      "epoch": 0.2852556753038294,
+      "grad_norm": 0.840625602424926,
+      "learning_rate": 0.0008388133829459463,
+      "loss": 2.5047,
+      "step": 622
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 0.8313131589536529,
+      "learning_rate": 0.0008382665695566227,
+      "loss": 2.5429,
+      "step": 623
+    },
+    {
+      "epoch": 0.286172896124742,
+      "grad_norm": 0.8106725390670106,
+      "learning_rate": 0.0008377190091188324,
+      "loss": 2.4661,
+      "step": 624
+    },
+    {
+      "epoch": 0.28663150653519837,
+      "grad_norm": 0.8387854448369019,
+      "learning_rate": 0.0008371707028418413,
+      "loss": 2.4696,
+      "step": 625
+    },
+    {
+      "epoch": 0.28709011694565467,
+      "grad_norm": 0.7964579796936718,
+      "learning_rate": 0.0008366216519365621,
+      "loss": 2.5121,
+      "step": 626
+    },
+    {
+      "epoch": 0.287548727356111,
+      "grad_norm": 0.8363698349395409,
+      "learning_rate": 0.0008360718576155525,
+      "loss": 2.4493,
+      "step": 627
+    },
+    {
+      "epoch": 0.2880073377665673,
+      "grad_norm": 0.8550141532208351,
+      "learning_rate": 0.0008355213210930118,
+      "loss": 2.4646,
+      "step": 628
+    },
+    {
+      "epoch": 0.28846594817702365,
+      "grad_norm": 0.7988318152330779,
+      "learning_rate": 0.0008349700435847778,
+      "loss": 2.5372,
+      "step": 629
+    },
+    {
+      "epoch": 0.28892455858747995,
+      "grad_norm": 0.8522235820152463,
+      "learning_rate": 0.0008344180263083256,
+      "loss": 2.4847,
+      "step": 630
+    },
+    {
+      "epoch": 0.28938316899793626,
+      "grad_norm": 0.8204037772709851,
+      "learning_rate": 0.000833865270482764,
+      "loss": 2.3885,
+      "step": 631
+    },
+    {
+      "epoch": 0.28984177940839256,
+      "grad_norm": 0.7672530250869493,
+      "learning_rate": 0.0008333117773288324,
+      "loss": 2.4418,
+      "step": 632
+    },
+    {
+      "epoch": 0.29030038981884887,
+      "grad_norm": 0.8093987721824453,
+      "learning_rate": 0.0008327575480688985,
+      "loss": 2.3824,
+      "step": 633
+    },
+    {
+      "epoch": 0.29075900022930523,
+      "grad_norm": 0.8046100165245534,
+      "learning_rate": 0.000832202583926956,
+      "loss": 2.4683,
+      "step": 634
+    },
+    {
+      "epoch": 0.29121761063976154,
+      "grad_norm": 0.8410078145211307,
+      "learning_rate": 0.0008316468861286217,
+      "loss": 2.4881,
+      "step": 635
+    },
+    {
+      "epoch": 0.29167622105021784,
+      "grad_norm": 0.7978392020278359,
+      "learning_rate": 0.0008310904559011323,
+      "loss": 2.4417,
+      "step": 636
+    },
+    {
+      "epoch": 0.29213483146067415,
+      "grad_norm": 0.8684283382689134,
+      "learning_rate": 0.0008305332944733419,
+      "loss": 2.4854,
+      "step": 637
+    },
+    {
+      "epoch": 0.29259344187113046,
+      "grad_norm": 0.8624887410747771,
+      "learning_rate": 0.0008299754030757202,
+      "loss": 2.4893,
+      "step": 638
+    },
+    {
+      "epoch": 0.2930520522815868,
+      "grad_norm": 0.8014988885158734,
+      "learning_rate": 0.0008294167829403481,
+      "loss": 2.5018,
+      "step": 639
+    },
+    {
+      "epoch": 0.2935106626920431,
+      "grad_norm": 0.8531174046505593,
+      "learning_rate": 0.0008288574353009164,
+      "loss": 2.4167,
+      "step": 640
+    },
+    {
+      "epoch": 0.29396927310249943,
+      "grad_norm": 0.847360442782349,
+      "learning_rate": 0.0008282973613927225,
+      "loss": 2.4467,
+      "step": 641
+    },
+    {
+      "epoch": 0.29442788351295573,
+      "grad_norm": 0.8058998118451676,
+      "learning_rate": 0.0008277365624526675,
+      "loss": 2.4849,
+      "step": 642
+    },
+    {
+      "epoch": 0.29488649392341204,
+      "grad_norm": 0.828520558209685,
+      "learning_rate": 0.0008271750397192541,
+      "loss": 2.3976,
+      "step": 643
+    },
+    {
+      "epoch": 0.2953451043338684,
+      "grad_norm": 0.8321205778760585,
+      "learning_rate": 0.0008266127944325832,
+      "loss": 2.4097,
+      "step": 644
+    },
+    {
+      "epoch": 0.2958037147443247,
+      "grad_norm": 0.8687765477001388,
+      "learning_rate": 0.0008260498278343513,
+      "loss": 2.4822,
+      "step": 645
+    },
+    {
+      "epoch": 0.296262325154781,
+      "grad_norm": 0.7990839937225285,
+      "learning_rate": 0.0008254861411678485,
+      "loss": 2.4461,
+      "step": 646
+    },
+    {
+      "epoch": 0.2967209355652373,
+      "grad_norm": 0.9073383003410214,
+      "learning_rate": 0.0008249217356779544,
+      "loss": 2.39,
+      "step": 647
+    },
+    {
+      "epoch": 0.2971795459756936,
+      "grad_norm": 0.7735072179234885,
+      "learning_rate": 0.0008243566126111363,
+      "loss": 2.4752,
+      "step": 648
+    },
+    {
+      "epoch": 0.29763815638615,
+      "grad_norm": 0.7883105378226373,
+      "learning_rate": 0.0008237907732154466,
+      "loss": 2.5279,
+      "step": 649
+    },
+    {
+      "epoch": 0.2980967667966063,
+      "grad_norm": 0.8560873965714537,
+      "learning_rate": 0.0008232242187405194,
+      "loss": 2.4485,
+      "step": 650
+    },
+    {
+      "epoch": 0.2985553772070626,
+      "grad_norm": 0.7846483333916547,
+      "learning_rate": 0.000822656950437568,
+      "loss": 2.4952,
+      "step": 651
+    },
+    {
+      "epoch": 0.2990139876175189,
+      "grad_norm": 0.8244448548776563,
+      "learning_rate": 0.0008220889695593823,
+      "loss": 2.4587,
+      "step": 652
+    },
+    {
+      "epoch": 0.2994725980279752,
+      "grad_norm": 0.8189008765634253,
+      "learning_rate": 0.0008215202773603259,
+      "loss": 2.3748,
+      "step": 653
+    },
+    {
+      "epoch": 0.2999312084384316,
+      "grad_norm": 0.8055293999101186,
+      "learning_rate": 0.0008209508750963328,
+      "loss": 2.3947,
+      "step": 654
+    },
+    {
+      "epoch": 0.3003898188488879,
+      "grad_norm": 0.8539470813329245,
+      "learning_rate": 0.0008203807640249062,
+      "loss": 2.4891,
+      "step": 655
+    },
+    {
+      "epoch": 0.3008484292593442,
+      "grad_norm": 0.853904671769252,
+      "learning_rate": 0.0008198099454051136,
+      "loss": 2.5613,
+      "step": 656
+    },
+    {
+      "epoch": 0.3013070396698005,
+      "grad_norm": 0.8340749334434291,
+      "learning_rate": 0.0008192384204975857,
+      "loss": 2.4341,
+      "step": 657
+    },
+    {
+      "epoch": 0.3017656500802568,
+      "grad_norm": 0.8018352362842517,
+      "learning_rate": 0.000818666190564513,
+      "loss": 2.4166,
+      "step": 658
+    },
+    {
+      "epoch": 0.30222426049071316,
+      "grad_norm": 0.8370439813261786,
+      "learning_rate": 0.0008180932568696426,
+      "loss": 2.3779,
+      "step": 659
+    },
+    {
+      "epoch": 0.30268287090116947,
+      "grad_norm": 0.957828307687405,
+      "learning_rate": 0.0008175196206782764,
+      "loss": 2.4322,
+      "step": 660
+    },
+    {
+      "epoch": 0.30314148131162577,
+      "grad_norm": 0.8897895040327779,
+      "learning_rate": 0.0008169452832572675,
+      "loss": 2.3969,
+      "step": 661
+    },
+    {
+      "epoch": 0.3036000917220821,
+      "grad_norm": 0.7900141164203233,
+      "learning_rate": 0.0008163702458750173,
+      "loss": 2.5595,
+      "step": 662
+    },
+    {
+      "epoch": 0.3040587021325384,
+      "grad_norm": 0.8451218838343026,
+      "learning_rate": 0.0008157945098014734,
+      "loss": 2.4475,
+      "step": 663
+    },
+    {
+      "epoch": 0.30451731254299474,
+      "grad_norm": 0.9410537267068779,
+      "learning_rate": 0.0008152180763081267,
+      "loss": 2.5274,
+      "step": 664
+    },
+    {
+      "epoch": 0.30497592295345105,
+      "grad_norm": 0.7671316741432139,
+      "learning_rate": 0.0008146409466680076,
+      "loss": 2.4575,
+      "step": 665
+    },
+    {
+      "epoch": 0.30543453336390736,
+      "grad_norm": 0.806604361325125,
+      "learning_rate": 0.0008140631221556845,
+      "loss": 2.4902,
+      "step": 666
+    },
+    {
+      "epoch": 0.30589314377436366,
+      "grad_norm": 0.8499387729076465,
+      "learning_rate": 0.0008134846040472599,
+      "loss": 2.5017,
+      "step": 667
+    },
+    {
+      "epoch": 0.30635175418481997,
+      "grad_norm": 0.8409280800333327,
+      "learning_rate": 0.0008129053936203688,
+      "loss": 2.4226,
+      "step": 668
+    },
+    {
+      "epoch": 0.30681036459527633,
+      "grad_norm": 0.818334120662905,
+      "learning_rate": 0.0008123254921541745,
+      "loss": 2.4821,
+      "step": 669
+    },
+    {
+      "epoch": 0.30726897500573264,
+      "grad_norm": 0.8144653206336405,
+      "learning_rate": 0.0008117449009293668,
+      "loss": 2.3762,
+      "step": 670
+    },
+    {
+      "epoch": 0.30772758541618894,
+      "grad_norm": 0.8578195972580046,
+      "learning_rate": 0.0008111636212281586,
+      "loss": 2.4903,
+      "step": 671
+    },
+    {
+      "epoch": 0.30818619582664525,
+      "grad_norm": 0.8775369654686542,
+      "learning_rate": 0.0008105816543342833,
+      "loss": 2.4357,
+      "step": 672
+    },
+    {
+      "epoch": 0.3086448062371016,
+      "grad_norm": 0.8049543446305762,
+      "learning_rate": 0.0008099990015329919,
+      "loss": 2.478,
+      "step": 673
+    },
+    {
+      "epoch": 0.3091034166475579,
+      "grad_norm": 0.8596401314960987,
+      "learning_rate": 0.0008094156641110504,
+      "loss": 2.4505,
+      "step": 674
+    },
+    {
+      "epoch": 0.3095620270580142,
+      "grad_norm": 0.7729068899342302,
+      "learning_rate": 0.0008088316433567369,
+      "loss": 2.5424,
+      "step": 675
+    },
+    {
+      "epoch": 0.31002063746847053,
+      "grad_norm": 0.75552238774419,
+      "learning_rate": 0.0008082469405598378,
+      "loss": 2.5031,
+      "step": 676
+    },
+    {
+      "epoch": 0.31047924787892683,
+      "grad_norm": 0.8143819497632271,
+      "learning_rate": 0.0008076615570116468,
+      "loss": 2.4654,
+      "step": 677
+    },
+    {
+      "epoch": 0.3109378582893832,
+      "grad_norm": 0.7848389699373199,
+      "learning_rate": 0.0008070754940049603,
+      "loss": 2.5207,
+      "step": 678
+    },
+    {
+      "epoch": 0.3113964686998395,
+      "grad_norm": 0.7511489370994087,
+      "learning_rate": 0.0008064887528340756,
+      "loss": 2.4232,
+      "step": 679
+    },
+    {
+      "epoch": 0.3118550791102958,
+      "grad_norm": 0.787306680543128,
+      "learning_rate": 0.0008059013347947874,
+      "loss": 2.5014,
+      "step": 680
+    },
+    {
+      "epoch": 0.3123136895207521,
+      "grad_norm": 0.806496741744841,
+      "learning_rate": 0.0008053132411843857,
+      "loss": 2.4247,
+      "step": 681
+    },
+    {
+      "epoch": 0.3127722999312084,
+      "grad_norm": 0.827807423265918,
+      "learning_rate": 0.0008047244733016521,
+      "loss": 2.5261,
+      "step": 682
+    },
+    {
+      "epoch": 0.3132309103416648,
+      "grad_norm": 0.874616002489615,
+      "learning_rate": 0.0008041350324468573,
+      "loss": 2.4242,
+      "step": 683
+    },
+    {
+      "epoch": 0.3136895207521211,
+      "grad_norm": 0.7968130586152199,
+      "learning_rate": 0.0008035449199217583,
+      "loss": 2.4482,
+      "step": 684
+    },
+    {
+      "epoch": 0.3141481311625774,
+      "grad_norm": 0.8241994261100684,
+      "learning_rate": 0.0008029541370295957,
+      "loss": 2.4767,
+      "step": 685
+    },
+    {
+      "epoch": 0.3146067415730337,
+      "grad_norm": 0.8793152906352687,
+      "learning_rate": 0.0008023626850750903,
+      "loss": 2.5035,
+      "step": 686
+    },
+    {
+      "epoch": 0.31506535198349,
+      "grad_norm": 0.818147518146519,
+      "learning_rate": 0.0008017705653644406,
+      "loss": 2.4824,
+      "step": 687
+    },
+    {
+      "epoch": 0.31552396239394637,
+      "grad_norm": 0.8830867497349973,
+      "learning_rate": 0.0008011777792053195,
+      "loss": 2.3969,
+      "step": 688
+    },
+    {
+      "epoch": 0.3159825728044027,
+      "grad_norm": 0.8159319240006138,
+      "learning_rate": 0.0008005843279068725,
+      "loss": 2.5264,
+      "step": 689
+    },
+    {
+      "epoch": 0.316441183214859,
+      "grad_norm": 0.7999829916949298,
+      "learning_rate": 0.000799990212779713,
+      "loss": 2.5443,
+      "step": 690
+    },
+    {
+      "epoch": 0.3168997936253153,
+      "grad_norm": 0.7941603584893945,
+      "learning_rate": 0.0007993954351359214,
+      "loss": 2.4572,
+      "step": 691
+    },
+    {
+      "epoch": 0.3173584040357716,
+      "grad_norm": 0.8534593238622884,
+      "learning_rate": 0.0007987999962890406,
+      "loss": 2.6325,
+      "step": 692
+    },
+    {
+      "epoch": 0.31781701444622795,
+      "grad_norm": 0.7938476703852975,
+      "learning_rate": 0.0007982038975540742,
+      "loss": 2.416,
+      "step": 693
+    },
+    {
+      "epoch": 0.31827562485668426,
+      "grad_norm": 0.8092588421129193,
+      "learning_rate": 0.0007976071402474826,
+      "loss": 2.41,
+      "step": 694
+    },
+    {
+      "epoch": 0.31873423526714056,
+      "grad_norm": 0.8511129994592941,
+      "learning_rate": 0.0007970097256871811,
+      "loss": 2.5191,
+      "step": 695
+    },
+    {
+      "epoch": 0.31919284567759687,
+      "grad_norm": 0.8330961443808304,
+      "learning_rate": 0.0007964116551925364,
+      "loss": 2.4616,
+      "step": 696
+    },
+    {
+      "epoch": 0.3196514560880532,
+      "grad_norm": 0.7776436474331077,
+      "learning_rate": 0.0007958129300843637,
+      "loss": 2.4298,
+      "step": 697
+    },
+    {
+      "epoch": 0.32011006649850954,
+      "grad_norm": 0.8155619916900446,
+      "learning_rate": 0.0007952135516849239,
+      "loss": 2.3461,
+      "step": 698
+    },
+    {
+      "epoch": 0.32056867690896584,
+      "grad_norm": 0.8373982929529357,
+      "learning_rate": 0.0007946135213179207,
+      "loss": 2.5259,
+      "step": 699
+    },
+    {
+      "epoch": 0.32102728731942215,
+      "grad_norm": 0.8155064877160451,
+      "learning_rate": 0.0007940128403084977,
+      "loss": 2.5914,
+      "step": 700
+    },
+    {
+      "epoch": 0.32148589772987846,
+      "grad_norm": 0.7992039906481014,
+      "learning_rate": 0.0007934115099832355,
+      "loss": 2.458,
+      "step": 701
+    },
+    {
+      "epoch": 0.32194450814033476,
+      "grad_norm": 0.8328916531162547,
+      "learning_rate": 0.0007928095316701483,
+      "loss": 2.4307,
+      "step": 702
+    },
+    {
+      "epoch": 0.3224031185507911,
+      "grad_norm": 0.7983251773000322,
+      "learning_rate": 0.0007922069066986819,
+      "loss": 2.5192,
+      "step": 703
+    },
+    {
+      "epoch": 0.32286172896124743,
+      "grad_norm": 0.827246755077729,
+      "learning_rate": 0.0007916036363997097,
+      "loss": 2.4048,
+      "step": 704
+    },
+    {
+      "epoch": 0.32332033937170374,
+      "grad_norm": 0.8636047606139854,
+      "learning_rate": 0.0007909997221055308,
+      "loss": 2.5106,
+      "step": 705
+    },
+    {
+      "epoch": 0.32377894978216004,
+      "grad_norm": 0.7924011183984959,
+      "learning_rate": 0.0007903951651498658,
+      "loss": 2.481,
+      "step": 706
+    },
+    {
+      "epoch": 0.32423756019261635,
+      "grad_norm": 0.8214533366000466,
+      "learning_rate": 0.0007897899668678557,
+      "loss": 2.3596,
+      "step": 707
+    },
+    {
+      "epoch": 0.3246961706030727,
+      "grad_norm": 0.7748391153650843,
+      "learning_rate": 0.0007891841285960566,
+      "loss": 2.486,
+      "step": 708
+    },
+    {
+      "epoch": 0.325154781013529,
+      "grad_norm": 0.8064957452091662,
+      "learning_rate": 0.0007885776516724388,
+      "loss": 2.3667,
+      "step": 709
+    },
+    {
+      "epoch": 0.3256133914239853,
+      "grad_norm": 0.8330627688249628,
+      "learning_rate": 0.0007879705374363831,
+      "loss": 2.3352,
+      "step": 710
+    },
+    {
+      "epoch": 0.3260720018344416,
+      "grad_norm": 0.8266832678479776,
+      "learning_rate": 0.000787362787228677,
+      "loss": 2.4832,
+      "step": 711
+    },
+    {
+      "epoch": 0.32653061224489793,
+      "grad_norm": 0.8393430355936471,
+      "learning_rate": 0.0007867544023915134,
+      "loss": 2.5113,
+      "step": 712
+    },
+    {
+      "epoch": 0.3269892226553543,
+      "grad_norm": 0.8032773951456234,
+      "learning_rate": 0.0007861453842684861,
+      "loss": 2.3433,
+      "step": 713
+    },
+    {
+      "epoch": 0.3274478330658106,
+      "grad_norm": 0.776471083364939,
+      "learning_rate": 0.0007855357342045882,
+      "loss": 2.4196,
+      "step": 714
+    },
+    {
+      "epoch": 0.3279064434762669,
+      "grad_norm": 0.8224345594894464,
+      "learning_rate": 0.0007849254535462074,
+      "loss": 2.3506,
+      "step": 715
+    },
+    {
+      "epoch": 0.3283650538867232,
+      "grad_norm": 0.723465328338354,
+      "learning_rate": 0.0007843145436411252,
+      "loss": 2.4282,
+      "step": 716
+    },
+    {
+      "epoch": 0.3288236642971795,
+      "grad_norm": 0.8887360627250959,
+      "learning_rate": 0.0007837030058385117,
+      "loss": 2.4248,
+      "step": 717
+    },
+    {
+      "epoch": 0.3292822747076359,
+      "grad_norm": 0.9379174003027203,
+      "learning_rate": 0.0007830908414889246,
+      "loss": 2.6025,
+      "step": 718
+    },
+    {
+      "epoch": 0.3297408851180922,
+      "grad_norm": 0.8356389329724314,
+      "learning_rate": 0.0007824780519443046,
+      "loss": 2.4715,
+      "step": 719
+    },
+    {
+      "epoch": 0.3301994955285485,
+      "grad_norm": 0.8396477294827605,
+      "learning_rate": 0.0007818646385579735,
+      "loss": 2.3986,
+      "step": 720
+    },
+    {
+      "epoch": 0.3306581059390048,
+      "grad_norm": 0.778543644078544,
+      "learning_rate": 0.0007812506026846307,
+      "loss": 2.4466,
+      "step": 721
+    },
+    {
+      "epoch": 0.33111671634946116,
+      "grad_norm": 0.8365353368070481,
+      "learning_rate": 0.0007806359456803504,
+      "loss": 2.499,
+      "step": 722
+    },
+    {
+      "epoch": 0.33157532675991747,
+      "grad_norm": 0.8485848463128066,
+      "learning_rate": 0.0007800206689025785,
+      "loss": 2.3581,
+      "step": 723
+    },
+    {
+      "epoch": 0.33203393717037377,
+      "grad_norm": 0.7701918826224182,
+      "learning_rate": 0.0007794047737101297,
+      "loss": 2.5851,
+      "step": 724
+    },
+    {
+      "epoch": 0.3324925475808301,
+      "grad_norm": 0.7941094153536816,
+      "learning_rate": 0.0007787882614631843,
+      "loss": 2.4217,
+      "step": 725
+    },
+    {
+      "epoch": 0.3329511579912864,
+      "grad_norm": 0.7907747932631611,
+      "learning_rate": 0.0007781711335232856,
+      "loss": 2.4164,
+      "step": 726
+    },
+    {
+      "epoch": 0.33340976840174275,
+      "grad_norm": 0.7952947981149343,
+      "learning_rate": 0.0007775533912533363,
+      "loss": 2.4374,
+      "step": 727
+    },
+    {
+      "epoch": 0.33386837881219905,
+      "grad_norm": 0.8163937215208793,
+      "learning_rate": 0.0007769350360175962,
+      "loss": 2.522,
+      "step": 728
+    },
+    {
+      "epoch": 0.33432698922265536,
+      "grad_norm": 0.8394675486509481,
+      "learning_rate": 0.0007763160691816784,
+      "loss": 2.4929,
+      "step": 729
+    },
+    {
+      "epoch": 0.33478559963311166,
+      "grad_norm": 0.8741238444477554,
+      "learning_rate": 0.000775696492112547,
+      "loss": 2.4073,
+      "step": 730
+    },
+    {
+      "epoch": 0.33524421004356797,
+      "grad_norm": 0.8832613203474191,
+      "learning_rate": 0.0007750763061785137,
+      "loss": 2.4354,
+      "step": 731
+    },
+    {
+      "epoch": 0.33570282045402433,
+      "grad_norm": 0.7608336446412473,
+      "learning_rate": 0.000774455512749235,
+      "loss": 2.4743,
+      "step": 732
+    },
+    {
+      "epoch": 0.33616143086448064,
+      "grad_norm": 0.8457159973716534,
+      "learning_rate": 0.0007738341131957085,
+      "loss": 2.424,
+      "step": 733
+    },
+    {
+      "epoch": 0.33662004127493694,
+      "grad_norm": 0.7938811605009136,
+      "learning_rate": 0.000773212108890271,
+      "loss": 2.4065,
+      "step": 734
+    },
+    {
+      "epoch": 0.33707865168539325,
+      "grad_norm": 0.7589173441720612,
+      "learning_rate": 0.0007725895012065947,
+      "loss": 2.2978,
+      "step": 735
+    },
+    {
+      "epoch": 0.33753726209584956,
+      "grad_norm": 0.7580585768424598,
+      "learning_rate": 0.0007719662915196844,
+      "loss": 2.4511,
+      "step": 736
+    },
+    {
+      "epoch": 0.3379958725063059,
+      "grad_norm": 0.7893474892692919,
+      "learning_rate": 0.0007713424812058736,
+      "loss": 2.4212,
+      "step": 737
+    },
+    {
+      "epoch": 0.3384544829167622,
+      "grad_norm": 0.8355137988097591,
+      "learning_rate": 0.0007707180716428237,
+      "loss": 2.5188,
+      "step": 738
+    },
+    {
+      "epoch": 0.33891309332721853,
+      "grad_norm": 0.8344552460015516,
+      "learning_rate": 0.0007700930642095184,
+      "loss": 2.4838,
+      "step": 739
+    },
+    {
+      "epoch": 0.33937170373767483,
+      "grad_norm": 0.7644293216211167,
+      "learning_rate": 0.0007694674602862621,
+      "loss": 2.3461,
+      "step": 740
+    },
+    {
+      "epoch": 0.33983031414813114,
+      "grad_norm": 0.7986420019134315,
+      "learning_rate": 0.0007688412612546769,
+      "loss": 2.4142,
+      "step": 741
+    },
+    {
+      "epoch": 0.3402889245585875,
+      "grad_norm": 0.7870569795161684,
+      "learning_rate": 0.0007682144684976983,
+      "loss": 2.3185,
+      "step": 742
+    },
+    {
+      "epoch": 0.3407475349690438,
+      "grad_norm": 0.8265280835291952,
+      "learning_rate": 0.0007675870833995739,
+      "loss": 2.3982,
+      "step": 743
+    },
+    {
+      "epoch": 0.3412061453795001,
+      "grad_norm": 0.7671471647724579,
+      "learning_rate": 0.0007669591073458592,
+      "loss": 2.5267,
+      "step": 744
+    },
+    {
+      "epoch": 0.3416647557899564,
+      "grad_norm": 0.8308224314622327,
+      "learning_rate": 0.0007663305417234146,
+      "loss": 2.3643,
+      "step": 745
+    },
+    {
+      "epoch": 0.3421233662004127,
+      "grad_norm": 0.8175682224826435,
+      "learning_rate": 0.0007657013879204022,
+      "loss": 2.4071,
+      "step": 746
+    },
+    {
+      "epoch": 0.3425819766108691,
+      "grad_norm": 0.8124435874392487,
+      "learning_rate": 0.0007650716473262842,
+      "loss": 2.3652,
+      "step": 747
+    },
+    {
+      "epoch": 0.3430405870213254,
+      "grad_norm": 0.8711540683467301,
+      "learning_rate": 0.0007644413213318177,
+      "loss": 2.4466,
+      "step": 748
+    },
+    {
+      "epoch": 0.3434991974317817,
+      "grad_norm": 0.8678857039427678,
+      "learning_rate": 0.0007638104113290531,
+      "loss": 2.4681,
+      "step": 749
+    },
+    {
+      "epoch": 0.343957807842238,
+      "grad_norm": 0.8068598758563714,
+      "learning_rate": 0.0007631789187113303,
+      "loss": 2.4092,
+      "step": 750
+    },
+    {
+      "epoch": 0.3444164182526943,
+      "grad_norm": 0.807353955751985,
+      "learning_rate": 0.000762546844873276,
+      "loss": 2.4176,
+      "step": 751
+    },
+    {
+      "epoch": 0.3448750286631507,
+      "grad_norm": 0.7970375903622561,
+      "learning_rate": 0.0007619141912108007,
+      "loss": 2.5419,
+      "step": 752
+    },
+    {
+      "epoch": 0.345333639073607,
+      "grad_norm": 0.7715925236692133,
+      "learning_rate": 0.000761280959121095,
+      "loss": 2.418,
+      "step": 753
+    },
+    {
+      "epoch": 0.3457922494840633,
+      "grad_norm": 0.8316117250976489,
+      "learning_rate": 0.0007606471500026273,
+      "loss": 2.4646,
+      "step": 754
+    },
+    {
+      "epoch": 0.3462508598945196,
+      "grad_norm": 0.7590714446474525,
+      "learning_rate": 0.0007600127652551401,
+      "loss": 2.4039,
+      "step": 755
+    },
+    {
+      "epoch": 0.3467094703049759,
+      "grad_norm": 0.7980038521325648,
+      "learning_rate": 0.0007593778062796472,
+      "loss": 2.4704,
+      "step": 756
+    },
+    {
+      "epoch": 0.34716808071543226,
+      "grad_norm": 0.8293049479873018,
+      "learning_rate": 0.000758742274478431,
+      "loss": 2.5761,
+      "step": 757
+    },
+    {
+      "epoch": 0.34762669112588857,
+      "grad_norm": 0.8983571494106052,
+      "learning_rate": 0.0007581061712550381,
+      "loss": 2.4401,
+      "step": 758
+    },
+    {
+      "epoch": 0.34808530153634487,
+      "grad_norm": 0.8060773118736914,
+      "learning_rate": 0.0007574694980142779,
+      "loss": 2.4448,
+      "step": 759
+    },
+    {
+      "epoch": 0.3485439119468012,
+      "grad_norm": 0.8011695995930871,
+      "learning_rate": 0.0007568322561622183,
+      "loss": 2.441,
+      "step": 760
+    },
+    {
+      "epoch": 0.3490025223572575,
+      "grad_norm": 0.8372353157812975,
+      "learning_rate": 0.0007561944471061826,
+      "loss": 2.4548,
+      "step": 761
+    },
+    {
+      "epoch": 0.34946113276771384,
+      "grad_norm": 0.7964893067560599,
+      "learning_rate": 0.0007555560722547475,
+      "loss": 2.2992,
+      "step": 762
+    },
+    {
+      "epoch": 0.34991974317817015,
+      "grad_norm": 0.890357447075298,
+      "learning_rate": 0.0007549171330177387,
+      "loss": 2.5008,
+      "step": 763
+    },
+    {
+      "epoch": 0.35037835358862646,
+      "grad_norm": 0.7741352571728309,
+      "learning_rate": 0.0007542776308062285,
+      "loss": 2.3898,
+      "step": 764
+    },
+    {
+      "epoch": 0.35083696399908276,
+      "grad_norm": 0.8129884103396889,
+      "learning_rate": 0.0007536375670325325,
+      "loss": 2.3477,
+      "step": 765
+    },
+    {
+      "epoch": 0.3512955744095391,
+      "grad_norm": 0.7784466092068377,
+      "learning_rate": 0.0007529969431102063,
+      "loss": 2.4366,
+      "step": 766
+    },
+    {
+      "epoch": 0.35175418481999543,
+      "grad_norm": 0.7946468274014368,
+      "learning_rate": 0.000752355760454043,
+      "loss": 2.4698,
+      "step": 767
+    },
+    {
+      "epoch": 0.35221279523045174,
+      "grad_norm": 0.8659633217172661,
+      "learning_rate": 0.0007517140204800693,
+      "loss": 2.4709,
+      "step": 768
+    },
+    {
+      "epoch": 0.35267140564090804,
+      "grad_norm": 0.765743312844718,
+      "learning_rate": 0.0007510717246055425,
+      "loss": 2.4213,
+      "step": 769
+    },
+    {
+      "epoch": 0.35313001605136435,
+      "grad_norm": 0.8274509911475575,
+      "learning_rate": 0.0007504288742489482,
+      "loss": 2.4666,
+      "step": 770
+    },
+    {
+      "epoch": 0.3535886264618207,
+      "grad_norm": 0.7796303232439672,
+      "learning_rate": 0.0007497854708299963,
+      "loss": 2.3797,
+      "step": 771
+    },
+    {
+      "epoch": 0.354047236872277,
+      "grad_norm": 0.7618400267422832,
+      "learning_rate": 0.0007491415157696178,
+      "loss": 2.4301,
+      "step": 772
+    },
+    {
+      "epoch": 0.3545058472827333,
+      "grad_norm": 0.8172391348397618,
+      "learning_rate": 0.0007484970104899623,
+      "loss": 2.4403,
+      "step": 773
+    },
+    {
+      "epoch": 0.35496445769318963,
+      "grad_norm": 0.7976624104913311,
+      "learning_rate": 0.0007478519564143945,
+      "loss": 2.4695,
+      "step": 774
+    },
+    {
+      "epoch": 0.35542306810364593,
+      "grad_norm": 0.8047010538663365,
+      "learning_rate": 0.000747206354967491,
+      "loss": 2.4179,
+      "step": 775
+    },
+    {
+      "epoch": 0.3558816785141023,
+      "grad_norm": 0.8377958929283431,
+      "learning_rate": 0.0007465602075750373,
+      "loss": 2.3499,
+      "step": 776
+    },
+    {
+      "epoch": 0.3563402889245586,
+      "grad_norm": 0.7769265833894681,
+      "learning_rate": 0.0007459135156640247,
+      "loss": 2.472,
+      "step": 777
+    },
+    {
+      "epoch": 0.3567988993350149,
+      "grad_norm": 0.8011268966624387,
+      "learning_rate": 0.0007452662806626468,
+      "loss": 2.4245,
+      "step": 778
+    },
+    {
+      "epoch": 0.3572575097454712,
+      "grad_norm": 0.7816974364444979,
+      "learning_rate": 0.0007446185040002967,
+      "loss": 2.2992,
+      "step": 779
+    },
+    {
+      "epoch": 0.3577161201559275,
+      "grad_norm": 0.8019846011451859,
+      "learning_rate": 0.0007439701871075642,
+      "loss": 2.5029,
+      "step": 780
+    },
+    {
+      "epoch": 0.3581747305663839,
+      "grad_norm": 0.7724822627453368,
+      "learning_rate": 0.0007433213314162313,
+      "loss": 2.432,
+      "step": 781
+    },
+    {
+      "epoch": 0.3586333409768402,
+      "grad_norm": 0.8148067645914069,
+      "learning_rate": 0.0007426719383592705,
+      "loss": 2.4055,
+      "step": 782
+    },
+    {
+      "epoch": 0.3590919513872965,
+      "grad_norm": 0.7877172712500303,
+      "learning_rate": 0.000742022009370841,
+      "loss": 2.3428,
+      "step": 783
+    },
+    {
+      "epoch": 0.3595505617977528,
+      "grad_norm": 0.7623833427037912,
+      "learning_rate": 0.0007413715458862855,
+      "loss": 2.415,
+      "step": 784
+    },
+    {
+      "epoch": 0.3600091722082091,
+      "grad_norm": 0.8013113974612299,
+      "learning_rate": 0.0007407205493421272,
+      "loss": 2.4331,
+      "step": 785
+    },
+    {
+      "epoch": 0.36046778261866547,
+      "grad_norm": 0.7832647275079162,
+      "learning_rate": 0.0007400690211760661,
+      "loss": 2.3312,
+      "step": 786
+    },
+    {
+      "epoch": 0.3609263930291218,
+      "grad_norm": 0.8262354215347212,
+      "learning_rate": 0.0007394169628269771,
+      "loss": 2.3232,
+      "step": 787
+    },
+    {
+      "epoch": 0.3613850034395781,
+      "grad_norm": 0.8156517867039222,
+      "learning_rate": 0.0007387643757349051,
+      "loss": 2.3945,
+      "step": 788
+    },
+    {
+      "epoch": 0.3618436138500344,
+      "grad_norm": 0.7999554278514246,
+      "learning_rate": 0.0007381112613410635,
+      "loss": 2.4227,
+      "step": 789
+    },
+    {
+      "epoch": 0.3623022242604907,
+      "grad_norm": 0.8461764900303819,
+      "learning_rate": 0.0007374576210878298,
+      "loss": 2.5017,
+      "step": 790
+    },
+    {
+      "epoch": 0.36276083467094705,
+      "grad_norm": 0.7831985372869158,
+      "learning_rate": 0.0007368034564187425,
+      "loss": 2.3855,
+      "step": 791
+    },
+    {
+      "epoch": 0.36321944508140336,
+      "grad_norm": 0.834761694474432,
+      "learning_rate": 0.0007361487687784989,
+      "loss": 2.4808,
+      "step": 792
+    },
+    {
+      "epoch": 0.36367805549185966,
+      "grad_norm": 0.8426140905869519,
+      "learning_rate": 0.0007354935596129513,
+      "loss": 2.5269,
+      "step": 793
+    },
+    {
+      "epoch": 0.36413666590231597,
+      "grad_norm": 0.8456489038554248,
+      "learning_rate": 0.000734837830369103,
+      "loss": 2.3973,
+      "step": 794
+    },
+    {
+      "epoch": 0.3645952763127723,
+      "grad_norm": 0.799831069980138,
+      "learning_rate": 0.0007341815824951066,
+      "loss": 2.324,
+      "step": 795
+    },
+    {
+      "epoch": 0.36505388672322864,
+      "grad_norm": 0.7220687176169198,
+      "learning_rate": 0.0007335248174402597,
+      "loss": 2.4254,
+      "step": 796
+    },
+    {
+      "epoch": 0.36551249713368494,
+      "grad_norm": 0.7479853010234035,
+      "learning_rate": 0.0007328675366550023,
+      "loss": 2.3067,
+      "step": 797
+    },
+    {
+      "epoch": 0.36597110754414125,
+      "grad_norm": 0.8604188821961822,
+      "learning_rate": 0.0007322097415909134,
+      "loss": 2.4422,
+      "step": 798
+    },
+    {
+      "epoch": 0.36642971795459756,
+      "grad_norm": 0.8218122289741904,
+      "learning_rate": 0.0007315514337007071,
+      "loss": 2.3655,
+      "step": 799
+    },
+    {
+      "epoch": 0.36688832836505386,
+      "grad_norm": 0.8043149785983855,
+      "learning_rate": 0.0007308926144382312,
+      "loss": 2.3562,
+      "step": 800
+    },
+    {
+      "epoch": 0.3673469387755102,
+      "grad_norm": 0.8524014718605185,
+      "learning_rate": 0.0007302332852584619,
+      "loss": 2.4424,
+      "step": 801
+    },
+    {
+      "epoch": 0.36780554918596653,
+      "grad_norm": 0.827935290424514,
+      "learning_rate": 0.0007295734476175018,
+      "loss": 2.4352,
+      "step": 802
+    },
+    {
+      "epoch": 0.36826415959642284,
+      "grad_norm": 0.798869420333097,
+      "learning_rate": 0.0007289131029725768,
+      "loss": 2.3931,
+      "step": 803
+    },
+    {
+      "epoch": 0.36872277000687914,
+      "grad_norm": 0.7669871025542838,
+      "learning_rate": 0.0007282522527820319,
+      "loss": 2.426,
+      "step": 804
+    },
+    {
+      "epoch": 0.36918138041733545,
+      "grad_norm": 0.7830605293105295,
+      "learning_rate": 0.000727590898505329,
+      "loss": 2.3833,
+      "step": 805
+    },
+    {
+      "epoch": 0.3696399908277918,
+      "grad_norm": 0.7492852094337122,
+      "learning_rate": 0.0007269290416030429,
+      "loss": 2.4092,
+      "step": 806
+    },
+    {
+      "epoch": 0.3700986012382481,
+      "grad_norm": 0.726770121341186,
+      "learning_rate": 0.000726266683536859,
+      "loss": 2.3645,
+      "step": 807
+    },
+    {
+      "epoch": 0.3705572116487044,
+      "grad_norm": 0.776874146658042,
+      "learning_rate": 0.0007256038257695687,
+      "loss": 2.3911,
+      "step": 808
+    },
+    {
+      "epoch": 0.3710158220591607,
+      "grad_norm": 0.7406030762226419,
+      "learning_rate": 0.0007249404697650678,
+      "loss": 2.4094,
+      "step": 809
+    },
+    {
+      "epoch": 0.37147443246961703,
+      "grad_norm": 0.7942326944671352,
+      "learning_rate": 0.0007242766169883518,
+      "loss": 2.397,
+      "step": 810
+    },
+    {
+      "epoch": 0.3719330428800734,
+      "grad_norm": 0.7923080253926893,
+      "learning_rate": 0.0007236122689055138,
+      "loss": 2.4094,
+      "step": 811
+    },
+    {
+      "epoch": 0.3723916532905297,
+      "grad_norm": 0.7662258723047741,
+      "learning_rate": 0.0007229474269837401,
+      "loss": 2.4553,
+      "step": 812
+    },
+    {
+      "epoch": 0.372850263700986,
+      "grad_norm": 0.7736457000514669,
+      "learning_rate": 0.0007222820926913085,
+      "loss": 2.3136,
+      "step": 813
+    },
+    {
+      "epoch": 0.3733088741114423,
+      "grad_norm": 0.7868478090239894,
+      "learning_rate": 0.0007216162674975833,
+      "loss": 2.4446,
+      "step": 814
+    },
+    {
+      "epoch": 0.3737674845218987,
+      "grad_norm": 0.7561957486972944,
+      "learning_rate": 0.0007209499528730138,
+      "loss": 2.4321,
+      "step": 815
+    },
+    {
+      "epoch": 0.374226094932355,
+      "grad_norm": 0.7774792422380222,
+      "learning_rate": 0.0007202831502891294,
+      "loss": 2.3756,
+      "step": 816
+    },
+    {
+      "epoch": 0.3746847053428113,
+      "grad_norm": 0.7675065631789207,
+      "learning_rate": 0.0007196158612185375,
+      "loss": 2.4854,
+      "step": 817
+    },
+    {
+      "epoch": 0.3751433157532676,
+      "grad_norm": 0.7237217960396513,
+      "learning_rate": 0.0007189480871349201,
+      "loss": 2.3313,
+      "step": 818
+    },
+    {
+      "epoch": 0.3756019261637239,
+      "grad_norm": 0.7379563682095128,
+      "learning_rate": 0.0007182798295130299,
+      "loss": 2.3642,
+      "step": 819
+    },
+    {
+      "epoch": 0.37606053657418026,
+      "grad_norm": 0.7321795057378356,
+      "learning_rate": 0.0007176110898286878,
+      "loss": 2.4145,
+      "step": 820
+    },
+    {
+      "epoch": 0.37651914698463657,
+      "grad_norm": 0.7556860428389097,
+      "learning_rate": 0.0007169418695587791,
+      "loss": 2.3841,
+      "step": 821
+    },
+    {
+      "epoch": 0.37697775739509287,
+      "grad_norm": 0.7808894678168462,
+      "learning_rate": 0.0007162721701812506,
+      "loss": 2.3176,
+      "step": 822
+    },
+    {
+      "epoch": 0.3774363678055492,
+      "grad_norm": 0.7703574182876359,
+      "learning_rate": 0.0007156019931751072,
+      "loss": 2.293,
+      "step": 823
+    },
+    {
+      "epoch": 0.3778949782160055,
+      "grad_norm": 0.8576840707049628,
+      "learning_rate": 0.0007149313400204082,
+      "loss": 2.3896,
+      "step": 824
+    },
+    {
+      "epoch": 0.37835358862646185,
+      "grad_norm": 0.7935463520539894,
+      "learning_rate": 0.0007142602121982653,
+      "loss": 2.4776,
+      "step": 825
+    },
+    {
+      "epoch": 0.37881219903691815,
+      "grad_norm": 0.8152052995328517,
+      "learning_rate": 0.0007135886111908379,
+      "loss": 2.4877,
+      "step": 826
+    },
+    {
+      "epoch": 0.37927080944737446,
+      "grad_norm": 0.8506101953548971,
+      "learning_rate": 0.0007129165384813303,
+      "loss": 2.4163,
+      "step": 827
+    },
+    {
+      "epoch": 0.37972941985783076,
+      "grad_norm": 0.7751659485483297,
+      "learning_rate": 0.0007122439955539888,
+      "loss": 2.4336,
+      "step": 828
+    },
+    {
+      "epoch": 0.38018803026828707,
+      "grad_norm": 0.7480228377179761,
+      "learning_rate": 0.0007115709838940983,
+      "loss": 2.4933,
+      "step": 829
+    },
+    {
+      "epoch": 0.38064664067874343,
+      "grad_norm": 0.8183791589152181,
+      "learning_rate": 0.0007108975049879785,
+      "loss": 2.3453,
+      "step": 830
+    },
+    {
+      "epoch": 0.38110525108919974,
+      "grad_norm": 0.7278953561486289,
+      "learning_rate": 0.0007102235603229814,
+      "loss": 2.4554,
+      "step": 831
+    },
+    {
+      "epoch": 0.38156386149965604,
+      "grad_norm": 0.8101029425593707,
+      "learning_rate": 0.000709549151387487,
+      "loss": 2.4454,
+      "step": 832
+    },
+    {
+      "epoch": 0.38202247191011235,
+      "grad_norm": 0.8319722631316053,
+      "learning_rate": 0.0007088742796709013,
+      "loss": 2.4679,
+      "step": 833
+    },
+    {
+      "epoch": 0.38248108232056865,
+      "grad_norm": 0.7547124073365196,
+      "learning_rate": 0.000708198946663652,
+      "loss": 2.4109,
+      "step": 834
+    },
+    {
+      "epoch": 0.382939692731025,
+      "grad_norm": 0.8335398846244974,
+      "learning_rate": 0.0007075231538571856,
+      "loss": 2.4706,
+      "step": 835
+    },
+    {
+      "epoch": 0.3833983031414813,
+      "grad_norm": 0.7722622172936391,
+      "learning_rate": 0.0007068469027439641,
+      "loss": 2.4518,
+      "step": 836
+    },
+    {
+      "epoch": 0.38385691355193763,
+      "grad_norm": 0.7663551156142071,
+      "learning_rate": 0.0007061701948174613,
+      "loss": 2.402,
+      "step": 837
+    },
+    {
+      "epoch": 0.38431552396239393,
+      "grad_norm": 0.7494683337003081,
+      "learning_rate": 0.0007054930315721606,
+      "loss": 2.3193,
+      "step": 838
+    },
+    {
+      "epoch": 0.38477413437285024,
+      "grad_norm": 0.7598321734677372,
+      "learning_rate": 0.0007048154145035501,
+      "loss": 2.3335,
+      "step": 839
+    },
+    {
+      "epoch": 0.3852327447833066,
+      "grad_norm": 0.7551825686662414,
+      "learning_rate": 0.0007041373451081207,
+      "loss": 2.3732,
+      "step": 840
+    },
+    {
+      "epoch": 0.3856913551937629,
+      "grad_norm": 0.7847442226042368,
+      "learning_rate": 0.0007034588248833621,
+      "loss": 2.3558,
+      "step": 841
+    },
+    {
+      "epoch": 0.3861499656042192,
+      "grad_norm": 0.8107591785097529,
+      "learning_rate": 0.0007027798553277595,
+      "loss": 2.513,
+      "step": 842
+    },
+    {
+      "epoch": 0.3866085760146755,
+      "grad_norm": 0.796783695683336,
+      "learning_rate": 0.0007021004379407909,
+      "loss": 2.4566,
+      "step": 843
+    },
+    {
+      "epoch": 0.3870671864251318,
+      "grad_norm": 0.738852054885363,
+      "learning_rate": 0.0007014205742229227,
+      "loss": 2.3493,
+      "step": 844
+    },
+    {
+      "epoch": 0.3875257968355882,
+      "grad_norm": 0.7775826469705701,
+      "learning_rate": 0.0007007402656756072,
+      "loss": 2.3992,
+      "step": 845
+    },
+    {
+      "epoch": 0.3879844072460445,
+      "grad_norm": 0.7832252532833714,
+      "learning_rate": 0.0007000595138012797,
+      "loss": 2.5319,
+      "step": 846
+    },
+    {
+      "epoch": 0.3884430176565008,
+      "grad_norm": 0.7834331993597842,
+      "learning_rate": 0.0006993783201033535,
+      "loss": 2.4395,
+      "step": 847
+    },
+    {
+      "epoch": 0.3889016280669571,
+      "grad_norm": 0.8057964714657259,
+      "learning_rate": 0.0006986966860862182,
+      "loss": 2.446,
+      "step": 848
+    },
+    {
+      "epoch": 0.3893602384774134,
+      "grad_norm": 0.8036787294884629,
+      "learning_rate": 0.000698014613255236,
+      "loss": 2.3749,
+      "step": 849
+    },
+    {
+      "epoch": 0.3898188488878698,
+      "grad_norm": 0.7761529298449101,
+      "learning_rate": 0.0006973321031167382,
+      "loss": 2.382,
+      "step": 850
+    },
+    {
+      "epoch": 0.3902774592983261,
+      "grad_norm": 0.7597772717579167,
+      "learning_rate": 0.0006966491571780216,
+      "loss": 2.4692,
+      "step": 851
+    },
+    {
+      "epoch": 0.3907360697087824,
+      "grad_norm": 0.7965279420471847,
+      "learning_rate": 0.0006959657769473453,
+      "loss": 2.3512,
+      "step": 852
+    },
+    {
+      "epoch": 0.3911946801192387,
+      "grad_norm": 0.7407779021554329,
+      "learning_rate": 0.000695281963933928,
+      "loss": 2.4833,
+      "step": 853
+    },
+    {
+      "epoch": 0.391653290529695,
+      "grad_norm": 0.7918835778913079,
+      "learning_rate": 0.0006945977196479438,
+      "loss": 2.4186,
+      "step": 854
+    },
+    {
+      "epoch": 0.39211190094015136,
+      "grad_norm": 0.7773278620548337,
+      "learning_rate": 0.0006939130456005196,
+      "loss": 2.4678,
+      "step": 855
+    },
+    {
+      "epoch": 0.39257051135060766,
+      "grad_norm": 0.8259130798537223,
+      "learning_rate": 0.0006932279433037311,
+      "loss": 2.4351,
+      "step": 856
+    },
+    {
+      "epoch": 0.39302912176106397,
+      "grad_norm": 0.7809018284975712,
+      "learning_rate": 0.0006925424142705997,
+      "loss": 2.4022,
+      "step": 857
+    },
+    {
+      "epoch": 0.3934877321715203,
+      "grad_norm": 0.78379912765028,
+      "learning_rate": 0.0006918564600150896,
+      "loss": 2.4085,
+      "step": 858
+    },
+    {
+      "epoch": 0.39394634258197664,
+      "grad_norm": 0.7996718427860734,
+      "learning_rate": 0.0006911700820521042,
+      "loss": 2.3263,
+      "step": 859
+    },
+    {
+      "epoch": 0.39440495299243294,
+      "grad_norm": 0.8327977402914121,
+      "learning_rate": 0.0006904832818974818,
+      "loss": 2.4348,
+      "step": 860
+    },
+    {
+      "epoch": 0.39486356340288925,
+      "grad_norm": 0.8274268519325751,
+      "learning_rate": 0.0006897960610679939,
+      "loss": 2.5125,
+      "step": 861
+    },
+    {
+      "epoch": 0.39532217381334556,
+      "grad_norm": 0.864125715717045,
+      "learning_rate": 0.0006891084210813407,
+      "loss": 2.462,
+      "step": 862
+    },
+    {
+      "epoch": 0.39578078422380186,
+      "grad_norm": 0.7768837066022344,
+      "learning_rate": 0.0006884203634561483,
+      "loss": 2.3931,
+      "step": 863
+    },
+    {
+      "epoch": 0.3962393946342582,
+      "grad_norm": 0.7559525681247987,
+      "learning_rate": 0.0006877318897119651,
+      "loss": 2.3934,
+      "step": 864
+    },
+    {
+      "epoch": 0.39669800504471453,
+      "grad_norm": 0.7958448120639392,
+      "learning_rate": 0.0006870430013692579,
+      "loss": 2.3468,
+      "step": 865
+    },
+    {
+      "epoch": 0.39715661545517084,
+      "grad_norm": 0.7751102182960715,
+      "learning_rate": 0.0006863536999494101,
+      "loss": 2.4015,
+      "step": 866
+    },
+    {
+      "epoch": 0.39761522586562714,
+      "grad_norm": 0.7490951930817289,
+      "learning_rate": 0.0006856639869747167,
+      "loss": 2.4678,
+      "step": 867
+    },
+    {
+      "epoch": 0.39807383627608345,
+      "grad_norm": 0.7986202386284662,
+      "learning_rate": 0.0006849738639683818,
+      "loss": 2.3005,
+      "step": 868
+    },
+    {
+      "epoch": 0.3985324466865398,
+      "grad_norm": 0.849517349008425,
+      "learning_rate": 0.000684283332454515,
+      "loss": 2.4516,
+      "step": 869
+    },
+    {
+      "epoch": 0.3989910570969961,
+      "grad_norm": 0.7805948008941463,
+      "learning_rate": 0.0006835923939581281,
+      "loss": 2.3695,
+      "step": 870
+    },
+    {
+      "epoch": 0.3994496675074524,
+      "grad_norm": 0.790258638227324,
+      "learning_rate": 0.0006829010500051318,
+      "loss": 2.5036,
+      "step": 871
+    },
+    {
+      "epoch": 0.3999082779179087,
+      "grad_norm": 0.8533833766669168,
+      "learning_rate": 0.0006822093021223321,
+      "loss": 2.3874,
+      "step": 872
+    },
+    {
+      "epoch": 0.40036688832836503,
+      "grad_norm": 0.8082743808523055,
+      "learning_rate": 0.0006815171518374268,
+      "loss": 2.4305,
+      "step": 873
+    },
+    {
+      "epoch": 0.4008254987388214,
+      "grad_norm": 0.7868674388745728,
+      "learning_rate": 0.0006808246006790031,
+      "loss": 2.3833,
+      "step": 874
+    },
+    {
+      "epoch": 0.4012841091492777,
+      "grad_norm": 0.7329903888435764,
+      "learning_rate": 0.0006801316501765329,
+      "loss": 2.4034,
+      "step": 875
+    },
+    {
+      "epoch": 0.401742719559734,
+      "grad_norm": 0.7524056976563231,
+      "learning_rate": 0.0006794383018603704,
+      "loss": 2.4428,
+      "step": 876
+    },
+    {
+      "epoch": 0.4022013299701903,
+      "grad_norm": 0.8043246997906253,
+      "learning_rate": 0.0006787445572617481,
+      "loss": 2.4995,
+      "step": 877
+    },
+    {
+      "epoch": 0.4026599403806466,
+      "grad_norm": 0.7433248024073857,
+      "learning_rate": 0.0006780504179127734,
+      "loss": 2.3932,
+      "step": 878
+    },
+    {
+      "epoch": 0.403118550791103,
+      "grad_norm": 0.769432189365245,
+      "learning_rate": 0.0006773558853464265,
+      "loss": 2.3795,
+      "step": 879
+    },
+    {
+      "epoch": 0.4035771612015593,
+      "grad_norm": 0.747286419761717,
+      "learning_rate": 0.000676660961096555,
+      "loss": 2.341,
+      "step": 880
+    },
+    {
+      "epoch": 0.4040357716120156,
+      "grad_norm": 0.8312577446827969,
+      "learning_rate": 0.000675965646697872,
+      "loss": 2.4319,
+      "step": 881
+    },
+    {
+      "epoch": 0.4044943820224719,
+      "grad_norm": 0.7994930176937499,
+      "learning_rate": 0.0006752699436859519,
+      "loss": 2.3939,
+      "step": 882
+    },
+    {
+      "epoch": 0.4049529924329282,
+      "grad_norm": 0.8131036993670067,
+      "learning_rate": 0.0006745738535972279,
+      "loss": 2.3493,
+      "step": 883
+    },
+    {
+      "epoch": 0.40541160284338457,
+      "grad_norm": 0.7924015985153806,
+      "learning_rate": 0.0006738773779689874,
+      "loss": 2.3823,
+      "step": 884
+    },
+    {
+      "epoch": 0.4058702132538409,
+      "grad_norm": 0.7503038737776756,
+      "learning_rate": 0.0006731805183393696,
+      "loss": 2.3827,
+      "step": 885
+    },
+    {
+      "epoch": 0.4063288236642972,
+      "grad_norm": 0.7403684925269056,
+      "learning_rate": 0.0006724832762473618,
+      "loss": 2.3716,
+      "step": 886
+    },
+    {
+      "epoch": 0.4067874340747535,
+      "grad_norm": 0.7781047740230017,
+      "learning_rate": 0.0006717856532327956,
+      "loss": 2.3294,
+      "step": 887
+    },
+    {
+      "epoch": 0.4072460444852098,
+      "grad_norm": 0.7873778617037155,
+      "learning_rate": 0.0006710876508363444,
+      "loss": 2.3498,
+      "step": 888
+    },
+    {
+      "epoch": 0.40770465489566615,
+      "grad_norm": 0.8040778090018359,
+      "learning_rate": 0.0006703892705995189,
+      "loss": 2.4305,
+      "step": 889
+    },
+    {
+      "epoch": 0.40816326530612246,
+      "grad_norm": 0.7925881048947717,
+      "learning_rate": 0.0006696905140646647,
+      "loss": 2.4848,
+      "step": 890
+    },
+    {
+      "epoch": 0.40862187571657876,
+      "grad_norm": 0.7711531709053264,
+      "learning_rate": 0.0006689913827749581,
+      "loss": 2.3793,
+      "step": 891
+    },
+    {
+      "epoch": 0.40908048612703507,
+      "grad_norm": 0.7921806440783266,
+      "learning_rate": 0.0006682918782744032,
+      "loss": 2.3915,
+      "step": 892
+    },
+    {
+      "epoch": 0.4095390965374914,
+      "grad_norm": 0.8403717143997614,
+      "learning_rate": 0.0006675920021078282,
+      "loss": 2.217,
+      "step": 893
+    },
+    {
+      "epoch": 0.40999770694794774,
+      "grad_norm": 0.7874228027818527,
+      "learning_rate": 0.0006668917558208823,
+      "loss": 2.5134,
+      "step": 894
+    },
+    {
+      "epoch": 0.41045631735840404,
+      "grad_norm": 0.7860740359765793,
+      "learning_rate": 0.0006661911409600321,
+      "loss": 2.4097,
+      "step": 895
+    },
+    {
+      "epoch": 0.41091492776886035,
+      "grad_norm": 0.7425557381393602,
+      "learning_rate": 0.0006654901590725577,
+      "loss": 2.3866,
+      "step": 896
+    },
+    {
+      "epoch": 0.41137353817931666,
+      "grad_norm": 0.774723229269881,
+      "learning_rate": 0.0006647888117065507,
+      "loss": 2.4338,
+      "step": 897
+    },
+    {
+      "epoch": 0.41183214858977296,
+      "grad_norm": 0.7718129758949019,
+      "learning_rate": 0.0006640871004109086,
+      "loss": 2.3324,
+      "step": 898
+    },
+    {
+      "epoch": 0.4122907590002293,
+      "grad_norm": 0.7816137221161352,
+      "learning_rate": 0.000663385026735334,
+      "loss": 2.3608,
+      "step": 899
+    },
+    {
+      "epoch": 0.41274936941068563,
+      "grad_norm": 0.7538965830992294,
+      "learning_rate": 0.0006626825922303285,
+      "loss": 2.4752,
+      "step": 900
+    },
+    {
+      "epoch": 0.41320797982114194,
+      "grad_norm": 0.769590181193084,
+      "learning_rate": 0.0006619797984471915,
+      "loss": 2.3813,
+      "step": 901
+    },
+    {
+      "epoch": 0.41366659023159824,
+      "grad_norm": 0.7997473551287376,
+      "learning_rate": 0.0006612766469380158,
+      "loss": 2.44,
+      "step": 902
+    },
+    {
+      "epoch": 0.41412520064205455,
+      "grad_norm": 0.7723832808763097,
+      "learning_rate": 0.0006605731392556833,
+      "loss": 2.5689,
+      "step": 903
+    },
+    {
+      "epoch": 0.4145838110525109,
+      "grad_norm": 0.7639340612371525,
+      "learning_rate": 0.0006598692769538637,
+      "loss": 2.4265,
+      "step": 904
+    },
+    {
+      "epoch": 0.4150424214629672,
+      "grad_norm": 0.7500109154091948,
+      "learning_rate": 0.0006591650615870091,
+      "loss": 2.4336,
+      "step": 905
+    },
+    {
+      "epoch": 0.4155010318734235,
+      "grad_norm": 0.7557021978819943,
+      "learning_rate": 0.0006584604947103514,
+      "loss": 2.3986,
+      "step": 906
+    },
+    {
+      "epoch": 0.4159596422838798,
+      "grad_norm": 0.8087448799382427,
+      "learning_rate": 0.0006577555778798993,
+      "loss": 2.3277,
+      "step": 907
+    },
+    {
+      "epoch": 0.4164182526943362,
+      "grad_norm": 0.7855967878865526,
+      "learning_rate": 0.0006570503126524336,
+      "loss": 2.3279,
+      "step": 908
+    },
+    {
+      "epoch": 0.4168768631047925,
+      "grad_norm": 0.7442604500215989,
+      "learning_rate": 0.0006563447005855054,
+      "loss": 2.3756,
+      "step": 909
+    },
+    {
+      "epoch": 0.4173354735152488,
+      "grad_norm": 0.8064028112487832,
+      "learning_rate": 0.000655638743237431,
+      "loss": 2.429,
+      "step": 910
+    },
+    {
+      "epoch": 0.4177940839257051,
+      "grad_norm": 0.771210176595237,
+      "learning_rate": 0.0006549324421672894,
+      "loss": 2.4373,
+      "step": 911
+    },
+    {
+      "epoch": 0.4182526943361614,
+      "grad_norm": 0.777522302051713,
+      "learning_rate": 0.0006542257989349194,
+      "loss": 2.4226,
+      "step": 912
+    },
+    {
+      "epoch": 0.4187113047466178,
+      "grad_norm": 0.7934910473868008,
+      "learning_rate": 0.0006535188151009142,
+      "loss": 2.3967,
+      "step": 913
+    },
+    {
+      "epoch": 0.4191699151570741,
+      "grad_norm": 0.8075775462310806,
+      "learning_rate": 0.0006528114922266204,
+      "loss": 2.4262,
+      "step": 914
+    },
+    {
+      "epoch": 0.4196285255675304,
+      "grad_norm": 0.7578538093094995,
+      "learning_rate": 0.0006521038318741327,
+      "loss": 2.4361,
+      "step": 915
+    },
+    {
+      "epoch": 0.4200871359779867,
+      "grad_norm": 0.7548132363696239,
+      "learning_rate": 0.0006513958356062912,
+      "loss": 2.3673,
+      "step": 916
+    },
+    {
+      "epoch": 0.420545746388443,
+      "grad_norm": 0.7374564867195003,
+      "learning_rate": 0.0006506875049866781,
+      "loss": 2.3494,
+      "step": 917
+    },
+    {
+      "epoch": 0.42100435679889936,
+      "grad_norm": 0.8156954993794754,
+      "learning_rate": 0.0006499788415796137,
+      "loss": 2.3877,
+      "step": 918
+    },
+    {
+      "epoch": 0.42146296720935567,
+      "grad_norm": 0.7447189099996389,
+      "learning_rate": 0.0006492698469501532,
+      "loss": 2.325,
+      "step": 919
+    },
+    {
+      "epoch": 0.42192157761981197,
+      "grad_norm": 0.7915514022992308,
+      "learning_rate": 0.0006485605226640837,
+      "loss": 2.4039,
+      "step": 920
+    },
+    {
+      "epoch": 0.4223801880302683,
+      "grad_norm": 0.7781321200534322,
+      "learning_rate": 0.00064785087028792,
+      "loss": 2.4362,
+      "step": 921
+    },
+    {
+      "epoch": 0.4228387984407246,
+      "grad_norm": 0.760931392803422,
+      "learning_rate": 0.0006471408913889019,
+      "loss": 2.3275,
+      "step": 922
+    },
+    {
+      "epoch": 0.42329740885118095,
+      "grad_norm": 0.733367894801162,
+      "learning_rate": 0.0006464305875349892,
+      "loss": 2.3551,
+      "step": 923
+    },
+    {
+      "epoch": 0.42375601926163725,
+      "grad_norm": 0.7314057871862766,
+      "learning_rate": 0.000645719960294861,
+      "loss": 2.3514,
+      "step": 924
+    },
+    {
+      "epoch": 0.42421462967209356,
+      "grad_norm": 0.700985252552568,
+      "learning_rate": 0.0006450090112379092,
+      "loss": 2.329,
+      "step": 925
+    },
+    {
+      "epoch": 0.42467324008254986,
+      "grad_norm": 0.7662114391011058,
+      "learning_rate": 0.0006442977419342371,
+      "loss": 2.3395,
+      "step": 926
+    },
+    {
+      "epoch": 0.42513185049300617,
+      "grad_norm": 0.7842510256334313,
+      "learning_rate": 0.000643586153954655,
+      "loss": 2.4326,
+      "step": 927
+    },
+    {
+      "epoch": 0.42559046090346253,
+      "grad_norm": 0.7549133649681778,
+      "learning_rate": 0.0006428742488706772,
+      "loss": 2.3894,
+      "step": 928
+    },
+    {
+      "epoch": 0.42604907131391884,
+      "grad_norm": 0.771110832269478,
+      "learning_rate": 0.0006421620282545182,
+      "loss": 2.2771,
+      "step": 929
+    },
+    {
+      "epoch": 0.42650768172437514,
+      "grad_norm": 0.750916858789119,
+      "learning_rate": 0.0006414494936790892,
+      "loss": 2.4898,
+      "step": 930
+    },
+    {
+      "epoch": 0.42696629213483145,
+      "grad_norm": 0.745243746394856,
+      "learning_rate": 0.0006407366467179951,
+      "loss": 2.3798,
+      "step": 931
+    },
+    {
+      "epoch": 0.42742490254528775,
+      "grad_norm": 0.7988635290101133,
+      "learning_rate": 0.0006400234889455301,
+      "loss": 2.4764,
+      "step": 932
+    },
+    {
+      "epoch": 0.4278835129557441,
+      "grad_norm": 0.7409468740443218,
+      "learning_rate": 0.0006393100219366755,
+      "loss": 2.4188,
+      "step": 933
+    },
+    {
+      "epoch": 0.4283421233662004,
+      "grad_norm": 0.7509593950809499,
+      "learning_rate": 0.0006385962472670953,
+      "loss": 2.3054,
+      "step": 934
+    },
+    {
+      "epoch": 0.42880073377665673,
+      "grad_norm": 0.7089000581078662,
+      "learning_rate": 0.0006378821665131328,
+      "loss": 2.3657,
+      "step": 935
+    },
+    {
+      "epoch": 0.42925934418711303,
+      "grad_norm": 0.7114712829298058,
+      "learning_rate": 0.0006371677812518072,
+      "loss": 2.4513,
+      "step": 936
+    },
+    {
+      "epoch": 0.42971795459756934,
+      "grad_norm": 0.8331786269296507,
+      "learning_rate": 0.0006364530930608107,
+      "loss": 2.3813,
+      "step": 937
+    },
+    {
+      "epoch": 0.4301765650080257,
+      "grad_norm": 0.7472149678892605,
+      "learning_rate": 0.0006357381035185038,
+      "loss": 2.4121,
+      "step": 938
+    },
+    {
+      "epoch": 0.430635175418482,
+      "grad_norm": 0.8037719970611584,
+      "learning_rate": 0.0006350228142039131,
+      "loss": 2.3987,
+      "step": 939
+    },
+    {
+      "epoch": 0.4310937858289383,
+      "grad_norm": 0.7315251338044012,
+      "learning_rate": 0.000634307226696727,
+      "loss": 2.3964,
+      "step": 940
+    },
+    {
+      "epoch": 0.4315523962393946,
+      "grad_norm": 0.7413070652736419,
+      "learning_rate": 0.0006335913425772926,
+      "loss": 2.4907,
+      "step": 941
+    },
+    {
+      "epoch": 0.4320110066498509,
+      "grad_norm": 0.7377562765042559,
+      "learning_rate": 0.0006328751634266117,
+      "loss": 2.2817,
+      "step": 942
+    },
+    {
+      "epoch": 0.4324696170603073,
+      "grad_norm": 0.7690327754934616,
+      "learning_rate": 0.0006321586908263382,
+      "loss": 2.3668,
+      "step": 943
+    },
+    {
+      "epoch": 0.4329282274707636,
+      "grad_norm": 0.8064782120063628,
+      "learning_rate": 0.0006314419263587732,
+      "loss": 2.4562,
+      "step": 944
+    },
+    {
+      "epoch": 0.4333868378812199,
+      "grad_norm": 0.7687691904738112,
+      "learning_rate": 0.0006307248716068637,
+      "loss": 2.3628,
+      "step": 945
+    },
+    {
+      "epoch": 0.4338454482916762,
+      "grad_norm": 0.8036878484799449,
+      "learning_rate": 0.0006300075281541964,
+      "loss": 2.284,
+      "step": 946
+    },
+    {
+      "epoch": 0.4343040587021325,
+      "grad_norm": 0.8647337521685259,
+      "learning_rate": 0.0006292898975849966,
+      "loss": 2.5158,
+      "step": 947
+    },
+    {
+      "epoch": 0.4347626691125889,
+      "grad_norm": 0.8307657477759944,
+      "learning_rate": 0.000628571981484123,
+      "loss": 2.3808,
+      "step": 948
+    },
+    {
+      "epoch": 0.4352212795230452,
+      "grad_norm": 0.7685168635595864,
+      "learning_rate": 0.0006278537814370654,
+      "loss": 2.4794,
+      "step": 949
+    },
+    {
+      "epoch": 0.4356798899335015,
+      "grad_norm": 0.7845795111387591,
+      "learning_rate": 0.0006271352990299406,
+      "loss": 2.4077,
+      "step": 950
+    },
+    {
+      "epoch": 0.4361385003439578,
+      "grad_norm": 0.7202325573052176,
+      "learning_rate": 0.0006264165358494885,
+      "loss": 2.3299,
+      "step": 951
+    },
+    {
+      "epoch": 0.43659711075441415,
+      "grad_norm": 0.8492844821931691,
+      "learning_rate": 0.0006256974934830694,
+      "loss": 2.4327,
+      "step": 952
+    },
+    {
+      "epoch": 0.43705572116487046,
+      "grad_norm": 0.7575066814046433,
+      "learning_rate": 0.0006249781735186606,
+      "loss": 2.3504,
+      "step": 953
+    },
+    {
+      "epoch": 0.43751433157532676,
+      "grad_norm": 0.7926516183556723,
+      "learning_rate": 0.0006242585775448518,
+      "loss": 2.4122,
+      "step": 954
+    },
+    {
+      "epoch": 0.43797294198578307,
+      "grad_norm": 0.7883554044682614,
+      "learning_rate": 0.0006235387071508427,
+      "loss": 2.32,
+      "step": 955
+    },
+    {
+      "epoch": 0.4384315523962394,
+      "grad_norm": 0.7698677128914114,
+      "learning_rate": 0.0006228185639264384,
+      "loss": 2.4119,
+      "step": 956
+    },
+    {
+      "epoch": 0.43889016280669574,
+      "grad_norm": 0.8003805450376746,
+      "learning_rate": 0.0006220981494620475,
+      "loss": 2.3819,
+      "step": 957
+    },
+    {
+      "epoch": 0.43934877321715204,
+      "grad_norm": 0.8083809448610174,
+      "learning_rate": 0.000621377465348677,
+      "loss": 2.4269,
+      "step": 958
+    },
+    {
+      "epoch": 0.43980738362760835,
+      "grad_norm": 0.7323701208823488,
+      "learning_rate": 0.0006206565131779293,
+      "loss": 2.3778,
+      "step": 959
+    },
+    {
+      "epoch": 0.44026599403806466,
+      "grad_norm": 0.7619756410996217,
+      "learning_rate": 0.0006199352945419994,
+      "loss": 2.4079,
+      "step": 960
+    },
+    {
+      "epoch": 0.44072460444852096,
+      "grad_norm": 0.787601822245211,
+      "learning_rate": 0.00061921381103367,
+      "loss": 2.4063,
+      "step": 961
+    },
+    {
+      "epoch": 0.4411832148589773,
+      "grad_norm": 0.746259891385525,
+      "learning_rate": 0.0006184920642463094,
+      "loss": 2.4325,
+      "step": 962
+    },
+    {
+      "epoch": 0.44164182526943363,
+      "grad_norm": 0.7565463349177101,
+      "learning_rate": 0.0006177700557738672,
+      "loss": 2.4344,
+      "step": 963
+    },
+    {
+      "epoch": 0.44210043567988994,
+      "grad_norm": 0.7673619320534943,
+      "learning_rate": 0.0006170477872108706,
+      "loss": 2.3154,
+      "step": 964
+    },
+    {
+      "epoch": 0.44255904609034624,
+      "grad_norm": 0.7743678143035738,
+      "learning_rate": 0.0006163252601524216,
+      "loss": 2.4014,
+      "step": 965
+    },
+    {
+      "epoch": 0.44301765650080255,
+      "grad_norm": 0.7748604308988115,
+      "learning_rate": 0.0006156024761941925,
+      "loss": 2.3992,
+      "step": 966
+    },
+    {
+      "epoch": 0.4434762669112589,
+      "grad_norm": 0.7502538721984661,
+      "learning_rate": 0.000614879436932424,
+      "loss": 2.4334,
+      "step": 967
+    },
+    {
+      "epoch": 0.4439348773217152,
+      "grad_norm": 0.7694471325456645,
+      "learning_rate": 0.0006141561439639196,
+      "loss": 2.3762,
+      "step": 968
+    },
+    {
+      "epoch": 0.4443934877321715,
+      "grad_norm": 0.7384485008078905,
+      "learning_rate": 0.0006134325988860433,
+      "loss": 2.313,
+      "step": 969
+    },
+    {
+      "epoch": 0.4448520981426278,
+      "grad_norm": 0.7560429662750795,
+      "learning_rate": 0.0006127088032967165,
+      "loss": 2.3451,
+      "step": 970
+    },
+    {
+      "epoch": 0.44531070855308413,
+      "grad_norm": 0.7155836599769017,
+      "learning_rate": 0.0006119847587944131,
+      "loss": 2.3804,
+      "step": 971
+    },
+    {
+      "epoch": 0.4457693189635405,
+      "grad_norm": 0.7484792388225292,
+      "learning_rate": 0.0006112604669781572,
+      "loss": 2.3545,
+      "step": 972
+    },
+    {
+      "epoch": 0.4462279293739968,
+      "grad_norm": 0.7792283693715785,
+      "learning_rate": 0.0006105359294475188,
+      "loss": 2.5049,
+      "step": 973
+    },
+    {
+      "epoch": 0.4466865397844531,
+      "grad_norm": 0.7127458506625264,
+      "learning_rate": 0.0006098111478026107,
+      "loss": 2.4063,
+      "step": 974
+    },
+    {
+      "epoch": 0.4471451501949094,
+      "grad_norm": 0.7033469353202575,
+      "learning_rate": 0.0006090861236440848,
+      "loss": 2.4946,
+      "step": 975
+    },
+    {
+      "epoch": 0.4476037606053657,
+      "grad_norm": 0.7822961773669381,
+      "learning_rate": 0.0006083608585731282,
+      "loss": 2.3834,
+      "step": 976
+    },
+    {
+      "epoch": 0.4480623710158221,
+      "grad_norm": 0.7436797721823993,
+      "learning_rate": 0.0006076353541914609,
+      "loss": 2.3639,
+      "step": 977
+    },
+    {
+      "epoch": 0.4485209814262784,
+      "grad_norm": 0.7108335801397705,
+      "learning_rate": 0.0006069096121013307,
+      "loss": 2.3834,
+      "step": 978
+    },
+    {
+      "epoch": 0.4489795918367347,
+      "grad_norm": 0.777138276191316,
+      "learning_rate": 0.0006061836339055105,
+      "loss": 2.3632,
+      "step": 979
+    },
+    {
+      "epoch": 0.449438202247191,
+      "grad_norm": 0.7586062022211917,
+      "learning_rate": 0.0006054574212072948,
+      "loss": 2.4064,
+      "step": 980
+    },
+    {
+      "epoch": 0.4498968126576473,
+      "grad_norm": 0.7270150396811638,
+      "learning_rate": 0.0006047309756104958,
+      "loss": 2.353,
+      "step": 981
+    },
+    {
+      "epoch": 0.45035542306810367,
+      "grad_norm": 0.8102033897725549,
+      "learning_rate": 0.00060400429871944,
+      "loss": 2.3694,
+      "step": 982
+    },
+    {
+      "epoch": 0.45081403347856,
+      "grad_norm": 0.7436265510070513,
+      "learning_rate": 0.0006032773921389654,
+      "loss": 2.4047,
+      "step": 983
+    },
+    {
+      "epoch": 0.4512726438890163,
+      "grad_norm": 0.7620450465674881,
+      "learning_rate": 0.0006025502574744162,
+      "loss": 2.4062,
+      "step": 984
+    },
+    {
+      "epoch": 0.4517312542994726,
+      "grad_norm": 0.7448678769013799,
+      "learning_rate": 0.000601822896331641,
+      "loss": 2.3726,
+      "step": 985
+    },
+    {
+      "epoch": 0.4521898647099289,
+      "grad_norm": 0.7031494578025107,
+      "learning_rate": 0.0006010953103169883,
+      "loss": 2.3444,
+      "step": 986
+    },
+    {
+      "epoch": 0.45264847512038525,
+      "grad_norm": 0.781354424866334,
+      "learning_rate": 0.0006003675010373034,
+      "loss": 2.4233,
+      "step": 987
+    },
+    {
+      "epoch": 0.45310708553084156,
+      "grad_norm": 0.7545460934530757,
+      "learning_rate": 0.0005996394700999246,
+      "loss": 2.4209,
+      "step": 988
+    },
+    {
+      "epoch": 0.45356569594129786,
+      "grad_norm": 0.6581466772904306,
+      "learning_rate": 0.0005989112191126794,
+      "loss": 2.3166,
+      "step": 989
+    },
+    {
+      "epoch": 0.45402430635175417,
+      "grad_norm": 0.7268461665395977,
+      "learning_rate": 0.0005981827496838822,
+      "loss": 2.2821,
+      "step": 990
+    },
+    {
+      "epoch": 0.4544829167622105,
+      "grad_norm": 0.7888906668442421,
+      "learning_rate": 0.0005974540634223286,
+      "loss": 2.4272,
+      "step": 991
+    },
+    {
+      "epoch": 0.45494152717266684,
+      "grad_norm": 0.7566514246737435,
+      "learning_rate": 0.0005967251619372939,
+      "loss": 2.456,
+      "step": 992
+    },
+    {
+      "epoch": 0.45540013758312314,
+      "grad_norm": 0.7994753504273014,
+      "learning_rate": 0.0005959960468385284,
+      "loss": 2.4161,
+      "step": 993
+    },
+    {
+      "epoch": 0.45585874799357945,
+      "grad_norm": 0.8022555872299741,
+      "learning_rate": 0.0005952667197362542,
+      "loss": 2.3799,
+      "step": 994
+    },
+    {
+      "epoch": 0.45631735840403576,
+      "grad_norm": 0.8047052623136152,
+      "learning_rate": 0.0005945371822411621,
+      "loss": 2.3891,
+      "step": 995
+    },
+    {
+      "epoch": 0.45677596881449206,
+      "grad_norm": 0.7752346815410472,
+      "learning_rate": 0.0005938074359644063,
+      "loss": 2.2675,
+      "step": 996
+    },
+    {
+      "epoch": 0.4572345792249484,
+      "grad_norm": 0.7845243265086606,
+      "learning_rate": 0.0005930774825176034,
+      "loss": 2.4108,
+      "step": 997
+    },
+    {
+      "epoch": 0.45769318963540473,
+      "grad_norm": 0.7565721038133862,
+      "learning_rate": 0.0005923473235128268,
+      "loss": 2.3996,
+      "step": 998
+    },
+    {
+      "epoch": 0.45815180004586104,
+      "grad_norm": 0.7665827030546911,
+      "learning_rate": 0.0005916169605626042,
+      "loss": 2.3092,
+      "step": 999
+    },
+    {
+      "epoch": 0.45861041045631734,
+      "grad_norm": 0.7367256104579301,
+      "learning_rate": 0.0005908863952799134,
+      "loss": 2.3886,
+      "step": 1000
+    },
+    {
+      "epoch": 0.4590690208667737,
+      "grad_norm": 0.7304382378407054,
+      "learning_rate": 0.0005901556292781793,
+      "loss": 2.374,
+      "step": 1001
+    },
+    {
+      "epoch": 0.45952763127723,
+      "grad_norm": 0.7458931105222593,
+      "learning_rate": 0.0005894246641712698,
+      "loss": 2.3557,
+      "step": 1002
+    },
+    {
+      "epoch": 0.4599862416876863,
+      "grad_norm": 0.7246199869991191,
+      "learning_rate": 0.0005886935015734931,
+      "loss": 2.3578,
+      "step": 1003
+    },
+    {
+      "epoch": 0.4604448520981426,
+      "grad_norm": 0.7159188317509113,
+      "learning_rate": 0.0005879621430995928,
+      "loss": 2.3088,
+      "step": 1004
+    },
+    {
+      "epoch": 0.4609034625085989,
+      "grad_norm": 0.7596300655854938,
+      "learning_rate": 0.0005872305903647455,
+      "loss": 2.3943,
+      "step": 1005
+    },
+    {
+      "epoch": 0.4613620729190553,
+      "grad_norm": 0.7267759079645525,
+      "learning_rate": 0.0005864988449845569,
+      "loss": 2.3774,
+      "step": 1006
+    },
+    {
+      "epoch": 0.4618206833295116,
+      "grad_norm": 0.7681573434207174,
+      "learning_rate": 0.0005857669085750578,
+      "loss": 2.428,
+      "step": 1007
+    },
+    {
+      "epoch": 0.4622792937399679,
+      "grad_norm": 0.790179047888998,
+      "learning_rate": 0.0005850347827527013,
+      "loss": 2.4097,
+      "step": 1008
+    },
+    {
+      "epoch": 0.4627379041504242,
+      "grad_norm": 0.7127586113506821,
+      "learning_rate": 0.0005843024691343584,
+      "loss": 2.3417,
+      "step": 1009
+    },
+    {
+      "epoch": 0.4631965145608805,
+      "grad_norm": 0.7093880849762619,
+      "learning_rate": 0.000583569969337315,
+      "loss": 2.368,
+      "step": 1010
+    },
+    {
+      "epoch": 0.4636551249713369,
+      "grad_norm": 0.7960512972160929,
+      "learning_rate": 0.0005828372849792686,
+      "loss": 2.3746,
+      "step": 1011
+    },
+    {
+      "epoch": 0.4641137353817932,
+      "grad_norm": 0.7291705700907861,
+      "learning_rate": 0.0005821044176783234,
+      "loss": 2.3863,
+      "step": 1012
+    },
+    {
+      "epoch": 0.4645723457922495,
+      "grad_norm": 0.7303265929548203,
+      "learning_rate": 0.0005813713690529886,
+      "loss": 2.3577,
+      "step": 1013
+    },
+    {
+      "epoch": 0.4650309562027058,
+      "grad_norm": 0.7239149405304421,
+      "learning_rate": 0.0005806381407221729,
+      "loss": 2.4175,
+      "step": 1014
+    },
+    {
+      "epoch": 0.4654895666131621,
+      "grad_norm": 0.7310582048248462,
+      "learning_rate": 0.0005799047343051826,
+      "loss": 2.3685,
+      "step": 1015
+    },
+    {
+      "epoch": 0.46594817702361846,
+      "grad_norm": 0.7694189134814546,
+      "learning_rate": 0.0005791711514217171,
+      "loss": 2.4301,
+      "step": 1016
+    },
+    {
+      "epoch": 0.46640678743407477,
+      "grad_norm": 0.8038386834580727,
+      "learning_rate": 0.0005784373936918654,
+      "loss": 2.3953,
+      "step": 1017
+    },
+    {
+      "epoch": 0.46686539784453107,
+      "grad_norm": 0.7509886131990852,
+      "learning_rate": 0.0005777034627361025,
+      "loss": 2.4599,
+      "step": 1018
+    },
+    {
+      "epoch": 0.4673240082549874,
+      "grad_norm": 0.7619528346955009,
+      "learning_rate": 0.0005769693601752864,
+      "loss": 2.2942,
+      "step": 1019
+    },
+    {
+      "epoch": 0.4677826186654437,
+      "grad_norm": 0.7556070958994932,
+      "learning_rate": 0.0005762350876306537,
+      "loss": 2.4123,
+      "step": 1020
+    },
+    {
+      "epoch": 0.46824122907590005,
+      "grad_norm": 0.7577303396069439,
+      "learning_rate": 0.0005755006467238168,
+      "loss": 2.3704,
+      "step": 1021
+    },
+    {
+      "epoch": 0.46869983948635635,
+      "grad_norm": 0.7274023868924113,
+      "learning_rate": 0.0005747660390767593,
+      "loss": 2.3742,
+      "step": 1022
+    },
+    {
+      "epoch": 0.46915844989681266,
+      "grad_norm": 0.7772198441206507,
+      "learning_rate": 0.0005740312663118338,
+      "loss": 2.3023,
+      "step": 1023
+    },
+    {
+      "epoch": 0.46961706030726896,
+      "grad_norm": 0.7743233597767718,
+      "learning_rate": 0.0005732963300517568,
+      "loss": 2.351,
+      "step": 1024
+    },
+    {
+      "epoch": 0.47007567071772527,
+      "grad_norm": 0.7484793581962509,
+      "learning_rate": 0.0005725612319196064,
+      "loss": 2.501,
+      "step": 1025
+    },
+    {
+      "epoch": 0.47053428112818163,
+      "grad_norm": 0.7091653980633634,
+      "learning_rate": 0.000571825973538818,
+      "loss": 2.375,
+      "step": 1026
+    },
+    {
+      "epoch": 0.47099289153863794,
+      "grad_norm": 0.719409041761202,
+      "learning_rate": 0.0005710905565331811,
+      "loss": 2.3568,
+      "step": 1027
+    },
+    {
+      "epoch": 0.47145150194909424,
+      "grad_norm": 0.725957380557862,
+      "learning_rate": 0.0005703549825268353,
+      "loss": 2.4561,
+      "step": 1028
+    },
+    {
+      "epoch": 0.47191011235955055,
+      "grad_norm": 0.7675433182832567,
+      "learning_rate": 0.0005696192531442667,
+      "loss": 2.3591,
+      "step": 1029
+    },
+    {
+      "epoch": 0.47236872277000685,
+      "grad_norm": 0.7411109328499789,
+      "learning_rate": 0.000568883370010305,
+      "loss": 2.4507,
+      "step": 1030
+    },
+    {
+      "epoch": 0.4728273331804632,
+      "grad_norm": 0.7113044602760203,
+      "learning_rate": 0.0005681473347501192,
+      "loss": 2.6098,
+      "step": 1031
+    },
+    {
+      "epoch": 0.4732859435909195,
+      "grad_norm": 0.7712264529560997,
+      "learning_rate": 0.0005674111489892144,
+      "loss": 2.3352,
+      "step": 1032
+    },
+    {
+      "epoch": 0.47374455400137583,
+      "grad_norm": 0.7723316557737319,
+      "learning_rate": 0.0005666748143534282,
+      "loss": 2.4255,
+      "step": 1033
+    },
+    {
+      "epoch": 0.47420316441183213,
+      "grad_norm": 0.7672311856285584,
+      "learning_rate": 0.0005659383324689266,
+      "loss": 2.3998,
+      "step": 1034
+    },
+    {
+      "epoch": 0.47466177482228844,
+      "grad_norm": 0.8132915767340191,
+      "learning_rate": 0.0005652017049622007,
+      "loss": 2.3452,
+      "step": 1035
+    },
+    {
+      "epoch": 0.4751203852327448,
+      "grad_norm": 0.8037279285550349,
+      "learning_rate": 0.0005644649334600641,
+      "loss": 2.4182,
+      "step": 1036
+    },
+    {
+      "epoch": 0.4755789956432011,
+      "grad_norm": 0.7439348797690696,
+      "learning_rate": 0.0005637280195896474,
+      "loss": 2.4656,
+      "step": 1037
+    },
+    {
+      "epoch": 0.4760376060536574,
+      "grad_norm": 0.7329438329344977,
+      "learning_rate": 0.0005629909649783961,
+      "loss": 2.4258,
+      "step": 1038
+    },
+    {
+      "epoch": 0.4764962164641137,
+      "grad_norm": 0.7284931509457525,
+      "learning_rate": 0.0005622537712540664,
+      "loss": 2.535,
+      "step": 1039
+    },
+    {
+      "epoch": 0.47695482687457,
+      "grad_norm": 0.740829634225936,
+      "learning_rate": 0.0005615164400447218,
+      "loss": 2.4169,
+      "step": 1040
+    },
+    {
+      "epoch": 0.4774134372850264,
+      "grad_norm": 0.7162768136930373,
+      "learning_rate": 0.0005607789729787294,
+      "loss": 2.4225,
+      "step": 1041
+    },
+    {
+      "epoch": 0.4778720476954827,
+      "grad_norm": 0.7150892102693966,
+      "learning_rate": 0.0005600413716847564,
+      "loss": 2.3362,
+      "step": 1042
+    },
+    {
+      "epoch": 0.478330658105939,
+      "grad_norm": 0.739895685181595,
+      "learning_rate": 0.000559303637791766,
+      "loss": 2.354,
+      "step": 1043
+    },
+    {
+      "epoch": 0.4787892685163953,
+      "grad_norm": 0.7390293606670287,
+      "learning_rate": 0.0005585657729290151,
+      "loss": 2.3683,
+      "step": 1044
+    },
+    {
+      "epoch": 0.47924787892685167,
+      "grad_norm": 0.7494603215528397,
+      "learning_rate": 0.000557827778726049,
+      "loss": 2.3606,
+      "step": 1045
+    },
+    {
+      "epoch": 0.479706489337308,
+      "grad_norm": 0.8064461304300613,
+      "learning_rate": 0.0005570896568126993,
+      "loss": 2.3055,
+      "step": 1046
+    },
+    {
+      "epoch": 0.4801650997477643,
+      "grad_norm": 0.773004342232383,
+      "learning_rate": 0.0005563514088190788,
+      "loss": 2.3632,
+      "step": 1047
+    },
+    {
+      "epoch": 0.4806237101582206,
+      "grad_norm": 0.7717629556685554,
+      "learning_rate": 0.0005556130363755798,
+      "loss": 2.3049,
+      "step": 1048
+    },
+    {
+      "epoch": 0.4810823205686769,
+      "grad_norm": 0.8144850613501192,
+      "learning_rate": 0.0005548745411128688,
+      "loss": 2.2248,
+      "step": 1049
+    },
+    {
+      "epoch": 0.48154093097913325,
+      "grad_norm": 0.7462790645080902,
+      "learning_rate": 0.0005541359246618835,
+      "loss": 2.3196,
+      "step": 1050
+    },
+    {
+      "epoch": 0.48199954138958956,
+      "grad_norm": 0.8171885040057193,
+      "learning_rate": 0.0005533971886538293,
+      "loss": 2.3953,
+      "step": 1051
+    },
+    {
+      "epoch": 0.48245815180004586,
+      "grad_norm": 0.7770329169490751,
+      "learning_rate": 0.000552658334720176,
+      "loss": 2.3825,
+      "step": 1052
+    },
+    {
+      "epoch": 0.48291676221050217,
+      "grad_norm": 0.7660951999733349,
+      "learning_rate": 0.0005519193644926535,
+      "loss": 2.3081,
+      "step": 1053
+    },
+    {
+      "epoch": 0.4833753726209585,
+      "grad_norm": 0.7932087594564895,
+      "learning_rate": 0.0005511802796032485,
+      "loss": 2.4008,
+      "step": 1054
+    },
+    {
+      "epoch": 0.48383398303141484,
+      "grad_norm": 0.774994524643612,
+      "learning_rate": 0.0005504410816842009,
+      "loss": 2.339,
+      "step": 1055
+    },
+    {
+      "epoch": 0.48429259344187114,
+      "grad_norm": 0.7556797507310413,
+      "learning_rate": 0.0005497017723680009,
+      "loss": 2.2903,
+      "step": 1056
+    },
+    {
+      "epoch": 0.48475120385232745,
+      "grad_norm": 0.7388899714283187,
+      "learning_rate": 0.0005489623532873836,
+      "loss": 2.3501,
+      "step": 1057
+    },
+    {
+      "epoch": 0.48520981426278376,
+      "grad_norm": 0.7534517991310166,
+      "learning_rate": 0.0005482228260753273,
+      "loss": 2.407,
+      "step": 1058
+    },
+    {
+      "epoch": 0.48566842467324006,
+      "grad_norm": 0.7409077834914009,
+      "learning_rate": 0.0005474831923650488,
+      "loss": 2.4169,
+      "step": 1059
+    },
+    {
+      "epoch": 0.4861270350836964,
+      "grad_norm": 0.7159827992916848,
+      "learning_rate": 0.00054674345379,
+      "loss": 2.3387,
+      "step": 1060
+    },
+    {
+      "epoch": 0.48658564549415273,
+      "grad_norm": 0.7028165324403611,
+      "learning_rate": 0.000546003611983865,
+      "loss": 2.3742,
+      "step": 1061
+    },
+    {
+      "epoch": 0.48704425590460904,
+      "grad_norm": 0.7208409837791206,
+      "learning_rate": 0.0005452636685805552,
+      "loss": 2.3147,
+      "step": 1062
+    },
+    {
+      "epoch": 0.48750286631506534,
+      "grad_norm": 0.7278277112176054,
+      "learning_rate": 0.0005445236252142066,
+      "loss": 2.5194,
+      "step": 1063
+    },
+    {
+      "epoch": 0.48796147672552165,
+      "grad_norm": 0.7445740673321137,
+      "learning_rate": 0.000543783483519176,
+      "loss": 2.3814,
+      "step": 1064
+    },
+    {
+      "epoch": 0.488420087135978,
+      "grad_norm": 0.7848762561491107,
+      "learning_rate": 0.0005430432451300374,
+      "loss": 2.4033,
+      "step": 1065
+    },
+    {
+      "epoch": 0.4888786975464343,
+      "grad_norm": 0.693792151598291,
+      "learning_rate": 0.0005423029116815781,
+      "loss": 2.3222,
+      "step": 1066
+    },
+    {
+      "epoch": 0.4893373079568906,
+      "grad_norm": 0.7854851811683814,
+      "learning_rate": 0.0005415624848087959,
+      "loss": 2.3916,
+      "step": 1067
+    },
+    {
+      "epoch": 0.4897959183673469,
+      "grad_norm": 0.7419161600943543,
+      "learning_rate": 0.000540821966146894,
+      "loss": 2.4411,
+      "step": 1068
+    },
+    {
+      "epoch": 0.49025452877780323,
+      "grad_norm": 0.7556481334198433,
+      "learning_rate": 0.0005400813573312793,
+      "loss": 2.3466,
+      "step": 1069
+    },
+    {
+      "epoch": 0.4907131391882596,
+      "grad_norm": 0.7500830326289408,
+      "learning_rate": 0.0005393406599975572,
+      "loss": 2.4095,
+      "step": 1070
+    },
+    {
+      "epoch": 0.4911717495987159,
+      "grad_norm": 0.7352433628482031,
+      "learning_rate": 0.0005385998757815287,
+      "loss": 2.365,
+      "step": 1071
+    },
+    {
+      "epoch": 0.4916303600091722,
+      "grad_norm": 0.7260167522270558,
+      "learning_rate": 0.0005378590063191867,
+      "loss": 2.3065,
+      "step": 1072
+    },
+    {
+      "epoch": 0.4920889704196285,
+      "grad_norm": 0.7679019901371453,
+      "learning_rate": 0.0005371180532467124,
+      "loss": 2.3514,
+      "step": 1073
+    },
+    {
+      "epoch": 0.4925475808300848,
+      "grad_norm": 0.730759788117719,
+      "learning_rate": 0.000536377018200472,
+      "loss": 2.3579,
+      "step": 1074
+    },
+    {
+      "epoch": 0.4930061912405412,
+      "grad_norm": 0.7914300424264641,
+      "learning_rate": 0.0005356359028170118,
+      "loss": 2.355,
+      "step": 1075
+    },
+    {
+      "epoch": 0.4934648016509975,
+      "grad_norm": 0.7296658015940123,
+      "learning_rate": 0.0005348947087330564,
+      "loss": 2.3822,
+      "step": 1076
+    },
+    {
+      "epoch": 0.4939234120614538,
+      "grad_norm": 0.7852859898445731,
+      "learning_rate": 0.0005341534375855037,
+      "loss": 2.4281,
+      "step": 1077
+    },
+    {
+      "epoch": 0.4943820224719101,
+      "grad_norm": 0.7698807568262943,
+      "learning_rate": 0.0005334120910114222,
+      "loss": 2.3183,
+      "step": 1078
+    },
+    {
+      "epoch": 0.4948406328823664,
+      "grad_norm": 0.7240585504635293,
+      "learning_rate": 0.0005326706706480467,
+      "loss": 2.3475,
+      "step": 1079
+    },
+    {
+      "epoch": 0.49529924329282277,
+      "grad_norm": 0.7961532175104188,
+      "learning_rate": 0.0005319291781327749,
+      "loss": 2.304,
+      "step": 1080
+    },
+    {
+      "epoch": 0.49575785370327907,
+      "grad_norm": 0.7308151610954473,
+      "learning_rate": 0.0005311876151031642,
+      "loss": 2.4083,
+      "step": 1081
+    },
+    {
+      "epoch": 0.4962164641137354,
+      "grad_norm": 0.734861436513632,
+      "learning_rate": 0.0005304459831969274,
+      "loss": 2.3457,
+      "step": 1082
+    },
+    {
+      "epoch": 0.4966750745241917,
+      "grad_norm": 0.7284862352424569,
+      "learning_rate": 0.0005297042840519294,
+      "loss": 2.297,
+      "step": 1083
+    },
+    {
+      "epoch": 0.497133684934648,
+      "grad_norm": 0.7372863313698329,
+      "learning_rate": 0.0005289625193061838,
+      "loss": 2.33,
+      "step": 1084
+    },
+    {
+      "epoch": 0.49759229534510435,
+      "grad_norm": 0.8433038017543675,
+      "learning_rate": 0.0005282206905978489,
+      "loss": 2.3796,
+      "step": 1085
+    },
+    {
+      "epoch": 0.49805090575556066,
+      "grad_norm": 0.753622541201416,
+      "learning_rate": 0.0005274787995652246,
+      "loss": 2.3333,
+      "step": 1086
+    },
+    {
+      "epoch": 0.49850951616601696,
+      "grad_norm": 0.7301525676197143,
+      "learning_rate": 0.000526736847846748,
+      "loss": 2.2781,
+      "step": 1087
+    },
+    {
+      "epoch": 0.49896812657647327,
+      "grad_norm": 0.7501682846911957,
+      "learning_rate": 0.0005259948370809901,
+      "loss": 2.4212,
+      "step": 1088
+    },
+    {
+      "epoch": 0.4994267369869296,
+      "grad_norm": 0.7696862364604142,
+      "learning_rate": 0.0005252527689066533,
+      "loss": 2.37,
+      "step": 1089
+    },
+    {
+      "epoch": 0.49988534739738594,
+      "grad_norm": 0.7430745744055722,
+      "learning_rate": 0.0005245106449625654,
+      "loss": 2.3639,
+      "step": 1090
+    },
+    {
+      "epoch": 0.5003439578078422,
+      "grad_norm": 0.7712527896638628,
+      "learning_rate": 0.0005237684668876785,
+      "loss": 2.3218,
+      "step": 1091
+    },
+    {
+      "epoch": 0.5008025682182986,
+      "grad_norm": 0.8559784890611765,
+      "learning_rate": 0.0005230262363210637,
+      "loss": 2.4653,
+      "step": 1092
+    },
+    {
+      "epoch": 0.5012611786287549,
+      "grad_norm": 0.7971472245610001,
+      "learning_rate": 0.0005222839549019079,
+      "loss": 2.4613,
+      "step": 1093
+    },
+    {
+      "epoch": 0.5017197890392112,
+      "grad_norm": 0.7647126444217153,
+      "learning_rate": 0.0005215416242695108,
+      "loss": 2.3231,
+      "step": 1094
+    },
+    {
+      "epoch": 0.5021783994496675,
+      "grad_norm": 0.695561572541439,
+      "learning_rate": 0.0005207992460632804,
+      "loss": 2.381,
+      "step": 1095
+    },
+    {
+      "epoch": 0.5026370098601238,
+      "grad_norm": 0.7310941601939802,
+      "learning_rate": 0.0005200568219227299,
+      "loss": 2.2842,
+      "step": 1096
+    },
+    {
+      "epoch": 0.5030956202705802,
+      "grad_norm": 0.7573287826227789,
+      "learning_rate": 0.000519314353487474,
+      "loss": 2.3786,
+      "step": 1097
+    },
+    {
+      "epoch": 0.5035542306810364,
+      "grad_norm": 0.7444881148595078,
+      "learning_rate": 0.0005185718423972251,
+      "loss": 2.3801,
+      "step": 1098
+    },
+    {
+      "epoch": 0.5040128410914928,
+      "grad_norm": 0.769672165964386,
+      "learning_rate": 0.0005178292902917898,
+      "loss": 2.3969,
+      "step": 1099
+    },
+    {
+      "epoch": 0.504471451501949,
+      "grad_norm": 0.7162093230754122,
+      "learning_rate": 0.0005170866988110656,
+      "loss": 2.3817,
+      "step": 1100
+    },
+    {
+      "epoch": 0.5049300619124054,
+      "grad_norm": 0.7415141937236666,
+      "learning_rate": 0.0005163440695950362,
+      "loss": 2.2881,
+      "step": 1101
+    },
+    {
+      "epoch": 0.5053886723228618,
+      "grad_norm": 0.7325529357156322,
+      "learning_rate": 0.0005156014042837695,
+      "loss": 2.318,
+      "step": 1102
+    },
+    {
+      "epoch": 0.505847282733318,
+      "grad_norm": 0.7636805875563523,
+      "learning_rate": 0.0005148587045174128,
+      "loss": 2.434,
+      "step": 1103
+    },
+    {
+      "epoch": 0.5063058931437744,
+      "grad_norm": 0.7550819372686122,
+      "learning_rate": 0.0005141159719361891,
+      "loss": 2.344,
+      "step": 1104
+    },
+    {
+      "epoch": 0.5067645035542306,
+      "grad_norm": 0.7397852753501325,
+      "learning_rate": 0.0005133732081803945,
+      "loss": 2.367,
+      "step": 1105
+    },
+    {
+      "epoch": 0.507223113964687,
+      "grad_norm": 0.7768327374687205,
+      "learning_rate": 0.0005126304148903936,
+      "loss": 2.3173,
+      "step": 1106
+    },
+    {
+      "epoch": 0.5076817243751434,
+      "grad_norm": 0.7457860899316522,
+      "learning_rate": 0.0005118875937066161,
+      "loss": 2.3878,
+      "step": 1107
+    },
+    {
+      "epoch": 0.5081403347855996,
+      "grad_norm": 0.8089874803125691,
+      "learning_rate": 0.0005111447462695537,
+      "loss": 2.4024,
+      "step": 1108
+    },
+    {
+      "epoch": 0.508598945196056,
+      "grad_norm": 0.7380355451013945,
+      "learning_rate": 0.0005104018742197557,
+      "loss": 2.3884,
+      "step": 1109
+    },
+    {
+      "epoch": 0.5090575556065122,
+      "grad_norm": 0.7665265016193711,
+      "learning_rate": 0.0005096589791978261,
+      "loss": 2.3048,
+      "step": 1110
+    },
+    {
+      "epoch": 0.5095161660169686,
+      "grad_norm": 0.795672089330085,
+      "learning_rate": 0.0005089160628444192,
+      "loss": 2.4433,
+      "step": 1111
+    },
+    {
+      "epoch": 0.509974776427425,
+      "grad_norm": 0.7370844848488245,
+      "learning_rate": 0.0005081731268002371,
+      "loss": 2.3604,
+      "step": 1112
+    },
+    {
+      "epoch": 0.5104333868378812,
+      "grad_norm": 0.9069133945346299,
+      "learning_rate": 0.0005074301727060243,
+      "loss": 2.3008,
+      "step": 1113
+    },
+    {
+      "epoch": 0.5108919972483376,
+      "grad_norm": 0.7406802746027433,
+      "learning_rate": 0.0005066872022025663,
+      "loss": 2.3554,
+      "step": 1114
+    },
+    {
+      "epoch": 0.5113506076587938,
+      "grad_norm": 0.7377045114955911,
+      "learning_rate": 0.0005059442169306844,
+      "loss": 2.3962,
+      "step": 1115
+    },
+    {
+      "epoch": 0.5118092180692502,
+      "grad_norm": 0.7630141900650005,
+      "learning_rate": 0.0005052012185312321,
+      "loss": 2.3359,
+      "step": 1116
+    },
+    {
+      "epoch": 0.5122678284797065,
+      "grad_norm": 0.7301294784697385,
+      "learning_rate": 0.0005044582086450925,
+      "loss": 2.4325,
+      "step": 1117
+    },
+    {
+      "epoch": 0.5127264388901628,
+      "grad_norm": 0.7276988370880062,
+      "learning_rate": 0.0005037151889131737,
+      "loss": 2.3961,
+      "step": 1118
+    },
+    {
+      "epoch": 0.5131850493006191,
+      "grad_norm": 0.7384128217185172,
+      "learning_rate": 0.0005029721609764059,
+      "loss": 2.3499,
+      "step": 1119
+    },
+    {
+      "epoch": 0.5136436597110754,
+      "grad_norm": 0.7522188974471854,
+      "learning_rate": 0.000502229126475737,
+      "loss": 2.4035,
+      "step": 1120
+    },
+    {
+      "epoch": 0.5141022701215318,
+      "grad_norm": 0.7695932620875284,
+      "learning_rate": 0.0005014860870521293,
+      "loss": 2.295,
+      "step": 1121
+    },
+    {
+      "epoch": 0.5145608805319881,
+      "grad_norm": 0.7948174856946855,
+      "learning_rate": 0.0005007430443465569,
+      "loss": 2.4264,
+      "step": 1122
+    },
+    {
+      "epoch": 0.5150194909424444,
+      "grad_norm": 0.7356673610548223,
+      "learning_rate": 0.0005,
+      "loss": 2.3504,
+      "step": 1123
+    },
+    {
+      "epoch": 0.5154781013529007,
+      "grad_norm": 0.709715655733214,
+      "learning_rate": 0.0004992569556534432,
+      "loss": 2.4041,
+      "step": 1124
+    },
+    {
+      "epoch": 0.515936711763357,
+      "grad_norm": 0.7876346644328811,
+      "learning_rate": 0.0004985139129478707,
+      "loss": 2.364,
+      "step": 1125
+    },
+    {
+      "epoch": 0.5163953221738133,
+      "grad_norm": 0.803866581298956,
+      "learning_rate": 0.0004977708735242633,
+      "loss": 2.3145,
+      "step": 1126
+    },
+    {
+      "epoch": 0.5168539325842697,
+      "grad_norm": 0.7483032660002727,
+      "learning_rate": 0.0004970278390235942,
+      "loss": 2.4083,
+      "step": 1127
+    },
+    {
+      "epoch": 0.517312542994726,
+      "grad_norm": 0.7676123368486729,
+      "learning_rate": 0.0004962848110868262,
+      "loss": 2.3563,
+      "step": 1128
+    },
+    {
+      "epoch": 0.5177711534051823,
+      "grad_norm": 0.740856785675299,
+      "learning_rate": 0.0004955417913549074,
+      "loss": 2.3295,
+      "step": 1129
+    },
+    {
+      "epoch": 0.5182297638156386,
+      "grad_norm": 0.7688894143148632,
+      "learning_rate": 0.0004947987814687679,
+      "loss": 2.4612,
+      "step": 1130
+    },
+    {
+      "epoch": 0.5186883742260949,
+      "grad_norm": 0.7570272397332097,
+      "learning_rate": 0.0004940557830693157,
+      "loss": 2.2857,
+      "step": 1131
+    },
+    {
+      "epoch": 0.5191469846365513,
+      "grad_norm": 0.7942109788501072,
+      "learning_rate": 0.0004933127977974338,
+      "loss": 2.3452,
+      "step": 1132
+    },
+    {
+      "epoch": 0.5196055950470075,
+      "grad_norm": 0.7487341333437423,
+      "learning_rate": 0.0004925698272939757,
+      "loss": 2.4219,
+      "step": 1133
+    },
+    {
+      "epoch": 0.5200642054574639,
+      "grad_norm": 0.773095591676767,
+      "learning_rate": 0.0004918268731997632,
+      "loss": 2.3593,
+      "step": 1134
+    },
+    {
+      "epoch": 0.5205228158679202,
+      "grad_norm": 0.7325258556704362,
+      "learning_rate": 0.0004910839371555809,
+      "loss": 2.3629,
+      "step": 1135
+    },
+    {
+      "epoch": 0.5209814262783765,
+      "grad_norm": 0.7280924997845466,
+      "learning_rate": 0.0004903410208021739,
+      "loss": 2.3371,
+      "step": 1136
+    },
+    {
+      "epoch": 0.5214400366888329,
+      "grad_norm": 0.7377009517437975,
+      "learning_rate": 0.0004895981257802443,
+      "loss": 2.3573,
+      "step": 1137
+    },
+    {
+      "epoch": 0.5218986470992891,
+      "grad_norm": 0.7771162532928951,
+      "learning_rate": 0.0004888552537304463,
+      "loss": 2.3343,
+      "step": 1138
+    },
+    {
+      "epoch": 0.5223572575097455,
+      "grad_norm": 0.7426234769398069,
+      "learning_rate": 0.00048811240629338394,
+      "loss": 2.2758,
+      "step": 1139
+    },
+    {
+      "epoch": 0.5228158679202017,
+      "grad_norm": 0.7431087183489627,
+      "learning_rate": 0.00048736958510960663,
+      "loss": 2.3716,
+      "step": 1140
+    },
+    {
+      "epoch": 0.5232744783306581,
+      "grad_norm": 0.7448874407008623,
+      "learning_rate": 0.00048662679181960564,
+      "loss": 2.4287,
+      "step": 1141
+    },
+    {
+      "epoch": 0.5237330887411145,
+      "grad_norm": 0.7446964090268252,
+      "learning_rate": 0.00048588402806381094,
+      "loss": 2.3874,
+      "step": 1142
+    },
+    {
+      "epoch": 0.5241916991515707,
+      "grad_norm": 0.7206286005205238,
+      "learning_rate": 0.0004851412954825874,
+      "loss": 2.3147,
+      "step": 1143
+    },
+    {
+      "epoch": 0.5246503095620271,
+      "grad_norm": 0.718418904566847,
+      "learning_rate": 0.00048439859571623034,
+      "loss": 2.3429,
+      "step": 1144
+    },
+    {
+      "epoch": 0.5251089199724833,
+      "grad_norm": 0.7407666933828703,
+      "learning_rate": 0.00048365593040496373,
+      "loss": 2.4245,
+      "step": 1145
+    },
+    {
+      "epoch": 0.5255675303829397,
+      "grad_norm": 0.7270566623512651,
+      "learning_rate": 0.00048291330118893443,
+      "loss": 2.3728,
+      "step": 1146
+    },
+    {
+      "epoch": 0.526026140793396,
+      "grad_norm": 0.7385987548013462,
+      "learning_rate": 0.0004821707097082102,
+      "loss": 2.369,
+      "step": 1147
+    },
+    {
+      "epoch": 0.5264847512038523,
+      "grad_norm": 0.742495498424516,
+      "learning_rate": 0.0004814281576027749,
+      "loss": 2.4496,
+      "step": 1148
+    },
+    {
+      "epoch": 0.5269433616143087,
+      "grad_norm": 0.7267018918386104,
+      "learning_rate": 0.000480685646512526,
+      "loss": 2.3353,
+      "step": 1149
+    },
+    {
+      "epoch": 0.5274019720247649,
+      "grad_norm": 0.7304883960403039,
+      "learning_rate": 0.00047994317807727025,
+      "loss": 2.448,
+      "step": 1150
+    },
+    {
+      "epoch": 0.5278605824352213,
+      "grad_norm": 0.7619507544802935,
+      "learning_rate": 0.00047920075393671974,
+      "loss": 2.2606,
+      "step": 1151
+    },
+    {
+      "epoch": 0.5283191928456776,
+      "grad_norm": 0.7260587716030704,
+      "learning_rate": 0.0004784583757304893,
+      "loss": 2.3955,
+      "step": 1152
+    },
+    {
+      "epoch": 0.5287778032561339,
+      "grad_norm": 0.7229764181589751,
+      "learning_rate": 0.00047771604509809214,
+      "loss": 2.3856,
+      "step": 1153
+    },
+    {
+      "epoch": 0.5292364136665902,
+      "grad_norm": 0.7806976624138776,
+      "learning_rate": 0.0004769737636789364,
+      "loss": 2.3842,
+      "step": 1154
+    },
+    {
+      "epoch": 0.5296950240770465,
+      "grad_norm": 0.7375687197872717,
+      "learning_rate": 0.00047623153311232157,
+      "loss": 2.2224,
+      "step": 1155
+    },
+    {
+      "epoch": 0.5301536344875029,
+      "grad_norm": 0.7208499019984996,
+      "learning_rate": 0.0004754893550374346,
+      "loss": 2.2736,
+      "step": 1156
+    },
+    {
+      "epoch": 0.5306122448979592,
+      "grad_norm": 0.7662572880747237,
+      "learning_rate": 0.00047474723109334685,
+      "loss": 2.3215,
+      "step": 1157
+    },
+    {
+      "epoch": 0.5310708553084155,
+      "grad_norm": 0.7127434493692193,
+      "learning_rate": 0.00047400516291900993,
+      "loss": 2.3542,
+      "step": 1158
+    },
+    {
+      "epoch": 0.5315294657188718,
+      "grad_norm": 0.7674347778002436,
+      "learning_rate": 0.0004732631521532522,
+      "loss": 2.372,
+      "step": 1159
+    },
+    {
+      "epoch": 0.5319880761293282,
+      "grad_norm": 0.745467331147685,
+      "learning_rate": 0.0004725212004347755,
+      "loss": 2.3403,
+      "step": 1160
+    },
+    {
+      "epoch": 0.5324466865397844,
+      "grad_norm": 0.7089544854730525,
+      "learning_rate": 0.00047177930940215095,
+      "loss": 2.3235,
+      "step": 1161
+    },
+    {
+      "epoch": 0.5329052969502408,
+      "grad_norm": 0.6998827494671123,
+      "learning_rate": 0.00047103748069381624,
+      "loss": 2.297,
+      "step": 1162
+    },
+    {
+      "epoch": 0.533363907360697,
+      "grad_norm": 0.6970260330192072,
+      "learning_rate": 0.0004702957159480707,
+      "loss": 2.4538,
+      "step": 1163
+    },
+    {
+      "epoch": 0.5338225177711534,
+      "grad_norm": 0.7348196905853637,
+      "learning_rate": 0.00046955401680307267,
+      "loss": 2.5275,
+      "step": 1164
+    },
+    {
+      "epoch": 0.5342811281816098,
+      "grad_norm": 0.6924372043517905,
+      "learning_rate": 0.0004688123848968359,
+      "loss": 2.3174,
+      "step": 1165
+    },
+    {
+      "epoch": 0.534739738592066,
+      "grad_norm": 0.7247426794510489,
+      "learning_rate": 0.00046807082186722516,
+      "loss": 2.3516,
+      "step": 1166
+    },
+    {
+      "epoch": 0.5351983490025224,
+      "grad_norm": 0.6910984321324076,
+      "learning_rate": 0.0004673293293519535,
+      "loss": 2.3545,
+      "step": 1167
+    },
+    {
+      "epoch": 0.5356569594129786,
+      "grad_norm": 0.7901924335353723,
+      "learning_rate": 0.00046658790898857806,
+      "loss": 2.4228,
+      "step": 1168
+    },
+    {
+      "epoch": 0.536115569823435,
+      "grad_norm": 0.7241900803683647,
+      "learning_rate": 0.0004658465624144963,
+      "loss": 2.3165,
+      "step": 1169
+    },
+    {
+      "epoch": 0.5365741802338914,
+      "grad_norm": 0.74237133254411,
+      "learning_rate": 0.0004651052912669438,
+      "loss": 2.4545,
+      "step": 1170
+    },
+    {
+      "epoch": 0.5370327906443476,
+      "grad_norm": 0.6883046283425929,
+      "learning_rate": 0.0004643640971829883,
+      "loss": 2.3191,
+      "step": 1171
+    },
+    {
+      "epoch": 0.537491401054804,
+      "grad_norm": 0.7406236153916099,
+      "learning_rate": 0.0004636229817995281,
+      "loss": 2.3272,
+      "step": 1172
+    },
+    {
+      "epoch": 0.5379500114652602,
+      "grad_norm": 0.727967801391972,
+      "learning_rate": 0.0004628819467532876,
+      "loss": 2.269,
+      "step": 1173
+    },
+    {
+      "epoch": 0.5384086218757166,
+      "grad_norm": 0.7353314837324119,
+      "learning_rate": 0.00046214099368081335,
+      "loss": 2.3988,
+      "step": 1174
+    },
+    {
+      "epoch": 0.538867232286173,
+      "grad_norm": 0.7640778414964691,
+      "learning_rate": 0.0004614001242184714,
+      "loss": 2.2856,
+      "step": 1175
+    },
+    {
+      "epoch": 0.5393258426966292,
+      "grad_norm": 0.7253295418478604,
+      "learning_rate": 0.000460659340002443,
+      "loss": 2.354,
+      "step": 1176
+    },
+    {
+      "epoch": 0.5397844531070856,
+      "grad_norm": 0.774006891136923,
+      "learning_rate": 0.00045991864266872073,
+      "loss": 2.3574,
+      "step": 1177
+    },
+    {
+      "epoch": 0.5402430635175418,
+      "grad_norm": 0.7421656922615532,
+      "learning_rate": 0.00045917803385310595,
+      "loss": 2.3343,
+      "step": 1178
+    },
+    {
+      "epoch": 0.5407016739279982,
+      "grad_norm": 0.7047310083525883,
+      "learning_rate": 0.00045843751519120417,
+      "loss": 2.3963,
+      "step": 1179
+    },
+    {
+      "epoch": 0.5411602843384545,
+      "grad_norm": 0.7433918202604156,
+      "learning_rate": 0.00045769708831842193,
+      "loss": 2.452,
+      "step": 1180
+    },
+    {
+      "epoch": 0.5416188947489108,
+      "grad_norm": 0.8014054908769067,
+      "learning_rate": 0.00045695675486996266,
+      "loss": 2.3225,
+      "step": 1181
+    },
+    {
+      "epoch": 0.5420775051593671,
+      "grad_norm": 0.7133761537829727,
+      "learning_rate": 0.00045621651648082405,
+      "loss": 2.2485,
+      "step": 1182
+    },
+    {
+      "epoch": 0.5425361155698234,
+      "grad_norm": 0.6903666998001154,
+      "learning_rate": 0.00045547637478579356,
+      "loss": 2.2707,
+      "step": 1183
+    },
+    {
+      "epoch": 0.5429947259802798,
+      "grad_norm": 0.7340416985738734,
+      "learning_rate": 0.0004547363314194449,
+      "loss": 2.3989,
+      "step": 1184
+    },
+    {
+      "epoch": 0.5434533363907361,
+      "grad_norm": 0.69324158248045,
+      "learning_rate": 0.000453996388016135,
+      "loss": 2.4459,
+      "step": 1185
+    },
+    {
+      "epoch": 0.5439119468011924,
+      "grad_norm": 0.7419933999742144,
+      "learning_rate": 0.0004532565462099999,
+      "loss": 2.303,
+      "step": 1186
+    },
+    {
+      "epoch": 0.5443705572116487,
+      "grad_norm": 0.7256938623344287,
+      "learning_rate": 0.0004525168076349513,
+      "loss": 2.4408,
+      "step": 1187
+    },
+    {
+      "epoch": 0.544829167622105,
+      "grad_norm": 0.7109581723996822,
+      "learning_rate": 0.0004517771739246729,
+      "loss": 2.4189,
+      "step": 1188
+    },
+    {
+      "epoch": 0.5452877780325613,
+      "grad_norm": 0.7275448928109757,
+      "learning_rate": 0.0004510376467126165,
+      "loss": 2.3767,
+      "step": 1189
+    },
+    {
+      "epoch": 0.5457463884430177,
+      "grad_norm": 0.7130556354056917,
+      "learning_rate": 0.0004502982276319992,
+      "loss": 2.3438,
+      "step": 1190
+    },
+    {
+      "epoch": 0.546204998853474,
+      "grad_norm": 0.7315049686665561,
+      "learning_rate": 0.0004495589183157991,
+      "loss": 2.3165,
+      "step": 1191
+    },
+    {
+      "epoch": 0.5466636092639303,
+      "grad_norm": 0.7170790771627839,
+      "learning_rate": 0.0004488197203967517,
+      "loss": 2.3389,
+      "step": 1192
+    },
+    {
+      "epoch": 0.5471222196743866,
+      "grad_norm": 0.7022216368512322,
+      "learning_rate": 0.0004480806355073467,
+      "loss": 2.3354,
+      "step": 1193
+    },
+    {
+      "epoch": 0.5475808300848429,
+      "grad_norm": 0.7576003391755007,
+      "learning_rate": 0.000447341665279824,
+      "loss": 2.3298,
+      "step": 1194
+    },
+    {
+      "epoch": 0.5480394404952993,
+      "grad_norm": 0.7874769830373499,
+      "learning_rate": 0.0004466028113461708,
+      "loss": 2.3574,
+      "step": 1195
+    },
+    {
+      "epoch": 0.5484980509057555,
+      "grad_norm": 0.7198140103030686,
+      "learning_rate": 0.0004458640753381167,
+      "loss": 2.3381,
+      "step": 1196
+    },
+    {
+      "epoch": 0.5489566613162119,
+      "grad_norm": 0.6960069374180375,
+      "learning_rate": 0.0004451254588871313,
+      "loss": 2.3613,
+      "step": 1197
+    },
+    {
+      "epoch": 0.5494152717266682,
+      "grad_norm": 0.7089537609911952,
+      "learning_rate": 0.0004443869636244203,
+      "loss": 2.3969,
+      "step": 1198
+    },
+    {
+      "epoch": 0.5498738821371245,
+      "grad_norm": 0.7003311499475385,
+      "learning_rate": 0.0004436485911809212,
+      "loss": 2.3006,
+      "step": 1199
+    },
+    {
+      "epoch": 0.5503324925475809,
+      "grad_norm": 0.7498982109051444,
+      "learning_rate": 0.00044291034318730087,
+      "loss": 2.3536,
+      "step": 1200
+    },
+    {
+      "epoch": 0.5507911029580371,
+      "grad_norm": 0.7275870185366435,
+      "learning_rate": 0.0004421722212739511,
+      "loss": 2.3134,
+      "step": 1201
+    },
+    {
+      "epoch": 0.5512497133684935,
+      "grad_norm": 0.7735153990467714,
+      "learning_rate": 0.0004414342270709848,
+      "loss": 2.3725,
+      "step": 1202
+    },
+    {
+      "epoch": 0.5517083237789497,
+      "grad_norm": 0.7245285071874594,
+      "learning_rate": 0.00044069636220823397,
+      "loss": 2.3292,
+      "step": 1203
+    },
+    {
+      "epoch": 0.5521669341894061,
+      "grad_norm": 0.7419624809811807,
+      "learning_rate": 0.0004399586283152437,
+      "loss": 2.4194,
+      "step": 1204
+    },
+    {
+      "epoch": 0.5526255445998625,
+      "grad_norm": 0.7334450926826922,
+      "learning_rate": 0.0004392210270212706,
+      "loss": 2.3552,
+      "step": 1205
+    },
+    {
+      "epoch": 0.5530841550103187,
+      "grad_norm": 0.7317436478693055,
+      "learning_rate": 0.00043848355995527825,
+      "loss": 2.4112,
+      "step": 1206
+    },
+    {
+      "epoch": 0.5535427654207751,
+      "grad_norm": 0.7763719542075121,
+      "learning_rate": 0.00043774622874593374,
+      "loss": 2.2915,
+      "step": 1207
+    },
+    {
+      "epoch": 0.5540013758312313,
+      "grad_norm": 0.7586352594879103,
+      "learning_rate": 0.000437009035021604,
+      "loss": 2.2993,
+      "step": 1208
+    },
+    {
+      "epoch": 0.5544599862416877,
+      "grad_norm": 0.7354357206695545,
+      "learning_rate": 0.00043627198041035274,
+      "loss": 2.3939,
+      "step": 1209
+    },
+    {
+      "epoch": 0.554918596652144,
+      "grad_norm": 0.7220071472239609,
+      "learning_rate": 0.00043553506653993597,
+      "loss": 2.4038,
+      "step": 1210
+    },
+    {
+      "epoch": 0.5553772070626003,
+      "grad_norm": 0.7788719942182383,
+      "learning_rate": 0.0004347982950377992,
+      "loss": 2.2641,
+      "step": 1211
+    },
+    {
+      "epoch": 0.5558358174730567,
+      "grad_norm": 0.7451537393276483,
+      "learning_rate": 0.0004340616675310735,
+      "loss": 2.3036,
+      "step": 1212
+    },
+    {
+      "epoch": 0.5562944278835129,
+      "grad_norm": 0.7737467324034026,
+      "learning_rate": 0.00043332518564657193,
+      "loss": 2.2949,
+      "step": 1213
+    },
+    {
+      "epoch": 0.5567530382939693,
+      "grad_norm": 0.7952065513135844,
+      "learning_rate": 0.0004325888510107856,
+      "loss": 2.4984,
+      "step": 1214
+    },
+    {
+      "epoch": 0.5572116487044256,
+      "grad_norm": 0.7436181406989724,
+      "learning_rate": 0.0004318526652498809,
+      "loss": 2.322,
+      "step": 1215
+    },
+    {
+      "epoch": 0.5576702591148819,
+      "grad_norm": 0.7759878733313123,
+      "learning_rate": 0.00043111662998969523,
+      "loss": 2.409,
+      "step": 1216
+    },
+    {
+      "epoch": 0.5581288695253382,
+      "grad_norm": 0.7725066315860007,
+      "learning_rate": 0.0004303807468557335,
+      "loss": 2.4267,
+      "step": 1217
+    },
+    {
+      "epoch": 0.5585874799357945,
+      "grad_norm": 0.7293598141999955,
+      "learning_rate": 0.0004296450174731648,
+      "loss": 2.3573,
+      "step": 1218
+    },
+    {
+      "epoch": 0.5590460903462509,
+      "grad_norm": 0.7312348687984134,
+      "learning_rate": 0.0004289094434668188,
+      "loss": 2.3164,
+      "step": 1219
+    },
+    {
+      "epoch": 0.5595047007567072,
+      "grad_norm": 0.7260128867480268,
+      "learning_rate": 0.00042817402646118185,
+      "loss": 2.4281,
+      "step": 1220
+    },
+    {
+      "epoch": 0.5599633111671635,
+      "grad_norm": 0.7509038956128925,
+      "learning_rate": 0.0004274387680803936,
+      "loss": 2.3865,
+      "step": 1221
+    },
+    {
+      "epoch": 0.5604219215776198,
+      "grad_norm": 0.7398411878518288,
+      "learning_rate": 0.00042670366994824327,
+      "loss": 2.3427,
+      "step": 1222
+    },
+    {
+      "epoch": 0.5608805319880761,
+      "grad_norm": 0.7321439337663757,
+      "learning_rate": 0.0004259687336881663,
+      "loss": 2.358,
+      "step": 1223
+    },
+    {
+      "epoch": 0.5613391423985324,
+      "grad_norm": 0.7068646591430642,
+      "learning_rate": 0.0004252339609232408,
+      "loss": 2.3718,
+      "step": 1224
+    },
+    {
+      "epoch": 0.5617977528089888,
+      "grad_norm": 0.7680993219729434,
+      "learning_rate": 0.0004244993532761834,
+      "loss": 2.3385,
+      "step": 1225
+    },
+    {
+      "epoch": 0.562256363219445,
+      "grad_norm": 0.7585426793788155,
+      "learning_rate": 0.00042376491236934634,
+      "loss": 2.3996,
+      "step": 1226
+    },
+    {
+      "epoch": 0.5627149736299014,
+      "grad_norm": 0.7208830861869326,
+      "learning_rate": 0.0004230306398247136,
+      "loss": 2.2506,
+      "step": 1227
+    },
+    {
+      "epoch": 0.5631735840403577,
+      "grad_norm": 0.7354093046666302,
+      "learning_rate": 0.0004222965372638976,
+      "loss": 2.3879,
+      "step": 1228
+    },
+    {
+      "epoch": 0.563632194450814,
+      "grad_norm": 0.7196834801136469,
+      "learning_rate": 0.0004215626063081348,
+      "loss": 2.437,
+      "step": 1229
+    },
+    {
+      "epoch": 0.5640908048612704,
+      "grad_norm": 0.7791778326311881,
+      "learning_rate": 0.000420828848578283,
+      "loss": 2.3515,
+      "step": 1230
+    },
+    {
+      "epoch": 0.5645494152717266,
+      "grad_norm": 0.7332810320720887,
+      "learning_rate": 0.0004200952656948175,
+      "loss": 2.4722,
+      "step": 1231
+    },
+    {
+      "epoch": 0.565008025682183,
+      "grad_norm": 0.7316440058866585,
+      "learning_rate": 0.0004193618592778272,
+      "loss": 2.3921,
+      "step": 1232
+    },
+    {
+      "epoch": 0.5654666360926393,
+      "grad_norm": 0.6880544811851171,
+      "learning_rate": 0.0004186286309470116,
+      "loss": 2.3263,
+      "step": 1233
+    },
+    {
+      "epoch": 0.5659252465030956,
+      "grad_norm": 0.6923080450288638,
+      "learning_rate": 0.0004178955823216767,
+      "loss": 2.4077,
+      "step": 1234
+    },
+    {
+      "epoch": 0.566383856913552,
+      "grad_norm": 0.7905353078697231,
+      "learning_rate": 0.00041716271502073137,
+      "loss": 2.4194,
+      "step": 1235
+    },
+    {
+      "epoch": 0.5668424673240082,
+      "grad_norm": 0.7266601504008786,
+      "learning_rate": 0.000416430030662685,
+      "loss": 2.2604,
+      "step": 1236
+    },
+    {
+      "epoch": 0.5673010777344646,
+      "grad_norm": 0.7275474704666222,
+      "learning_rate": 0.00041569753086564173,
+      "loss": 2.2999,
+      "step": 1237
+    },
+    {
+      "epoch": 0.5677596881449208,
+      "grad_norm": 0.7717321130897403,
+      "learning_rate": 0.0004149652172472988,
+      "loss": 2.4104,
+      "step": 1238
+    },
+    {
+      "epoch": 0.5682182985553772,
+      "grad_norm": 0.786835542971422,
+      "learning_rate": 0.00041423309142494234,
+      "loss": 2.3661,
+      "step": 1239
+    },
+    {
+      "epoch": 0.5686769089658336,
+      "grad_norm": 0.7469045056506918,
+      "learning_rate": 0.0004135011550154433,
+      "loss": 2.3541,
+      "step": 1240
+    },
+    {
+      "epoch": 0.5691355193762898,
+      "grad_norm": 0.7151403242660005,
+      "learning_rate": 0.0004127694096352546,
+      "loss": 2.3862,
+      "step": 1241
+    },
+    {
+      "epoch": 0.5695941297867462,
+      "grad_norm": 0.7106755622822535,
+      "learning_rate": 0.00041203785690040743,
+      "loss": 2.3899,
+      "step": 1242
+    },
+    {
+      "epoch": 0.5700527401972024,
+      "grad_norm": 0.8072931608752169,
+      "learning_rate": 0.00041130649842650694,
+      "loss": 2.3678,
+      "step": 1243
+    },
+    {
+      "epoch": 0.5705113506076588,
+      "grad_norm": 0.7864701409552527,
+      "learning_rate": 0.00041057533582873016,
+      "loss": 2.4124,
+      "step": 1244
+    },
+    {
+      "epoch": 0.5709699610181151,
+      "grad_norm": 0.7573576251676304,
+      "learning_rate": 0.0004098443707218208,
+      "loss": 2.4128,
+      "step": 1245
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 0.7744839070495754,
+      "learning_rate": 0.00040911360472008673,
+      "loss": 2.4178,
+      "step": 1246
+    },
+    {
+      "epoch": 0.5718871818390278,
+      "grad_norm": 0.7382538051777024,
+      "learning_rate": 0.0004083830394373959,
+      "loss": 2.3202,
+      "step": 1247
+    },
+    {
+      "epoch": 0.572345792249484,
+      "grad_norm": 0.8298298839721426,
+      "learning_rate": 0.00040765267648717324,
+      "loss": 2.3954,
+      "step": 1248
+    },
+    {
+      "epoch": 0.5728044026599404,
+      "grad_norm": 0.7893966119993506,
+      "learning_rate": 0.00040692251748239677,
+      "loss": 2.3467,
+      "step": 1249
+    },
+    {
+      "epoch": 0.5732630130703967,
+      "grad_norm": 0.7913639548569216,
+      "learning_rate": 0.00040619256403559383,
+      "loss": 2.2962,
+      "step": 1250
+    },
+    {
+      "epoch": 0.573721623480853,
+      "grad_norm": 0.739298408762412,
+      "learning_rate": 0.000405462817758838,
+      "loss": 2.3328,
+      "step": 1251
+    },
+    {
+      "epoch": 0.5741802338913093,
+      "grad_norm": 0.7289057292721081,
+      "learning_rate": 0.0004047332802637457,
+      "loss": 2.2883,
+      "step": 1252
+    },
+    {
+      "epoch": 0.5746388443017657,
+      "grad_norm": 0.7120685330498755,
+      "learning_rate": 0.00040400395316147157,
+      "loss": 2.3433,
+      "step": 1253
+    },
+    {
+      "epoch": 0.575097454712222,
+      "grad_norm": 0.8119706495018589,
+      "learning_rate": 0.00040327483806270627,
+      "loss": 2.418,
+      "step": 1254
+    },
+    {
+      "epoch": 0.5755560651226783,
+      "grad_norm": 0.7264885397297092,
+      "learning_rate": 0.0004025459365776715,
+      "loss": 2.2944,
+      "step": 1255
+    },
+    {
+      "epoch": 0.5760146755331346,
+      "grad_norm": 0.6937239494351415,
+      "learning_rate": 0.00040181725031611794,
+      "loss": 2.3908,
+      "step": 1256
+    },
+    {
+      "epoch": 0.5764732859435909,
+      "grad_norm": 0.744158770364705,
+      "learning_rate": 0.0004010887808873206,
+      "loss": 2.2943,
+      "step": 1257
+    },
+    {
+      "epoch": 0.5769318963540473,
+      "grad_norm": 0.7295812743276878,
+      "learning_rate": 0.00040036052990007553,
+      "loss": 2.3846,
+      "step": 1258
+    },
+    {
+      "epoch": 0.5773905067645035,
+      "grad_norm": 0.7560319710771977,
+      "learning_rate": 0.0003996324989626967,
+      "loss": 2.3467,
+      "step": 1259
+    },
+    {
+      "epoch": 0.5778491171749599,
+      "grad_norm": 0.7324209379349118,
+      "learning_rate": 0.00039890468968301166,
+      "loss": 2.3689,
+      "step": 1260
+    },
+    {
+      "epoch": 0.5783077275854162,
+      "grad_norm": 0.7257715777281594,
+      "learning_rate": 0.0003981771036683591,
+      "loss": 2.3965,
+      "step": 1261
+    },
+    {
+      "epoch": 0.5787663379958725,
+      "grad_norm": 0.6902321727887025,
+      "learning_rate": 0.00039744974252558385,
+      "loss": 2.2582,
+      "step": 1262
+    },
+    {
+      "epoch": 0.5792249484063289,
+      "grad_norm": 0.7292296352527049,
+      "learning_rate": 0.00039672260786103463,
+      "loss": 2.3571,
+      "step": 1263
+    },
+    {
+      "epoch": 0.5796835588167851,
+      "grad_norm": 0.7384316305669932,
+      "learning_rate": 0.00039599570128055994,
+      "loss": 2.4385,
+      "step": 1264
+    },
+    {
+      "epoch": 0.5801421692272415,
+      "grad_norm": 0.7549724192952563,
+      "learning_rate": 0.0003952690243895044,
+      "loss": 2.3027,
+      "step": 1265
+    },
+    {
+      "epoch": 0.5806007796376977,
+      "grad_norm": 0.7570111704820458,
+      "learning_rate": 0.0003945425787927054,
+      "loss": 2.4868,
+      "step": 1266
+    },
+    {
+      "epoch": 0.5810593900481541,
+      "grad_norm": 0.736091880178726,
+      "learning_rate": 0.00039381636609448975,
+      "loss": 2.2765,
+      "step": 1267
+    },
+    {
+      "epoch": 0.5815180004586105,
+      "grad_norm": 0.7092514442892016,
+      "learning_rate": 0.0003930903878986693,
+      "loss": 2.3808,
+      "step": 1268
+    },
+    {
+      "epoch": 0.5819766108690667,
+      "grad_norm": 0.7073897947409455,
+      "learning_rate": 0.00039236464580853916,
+      "loss": 2.4262,
+      "step": 1269
+    },
+    {
+      "epoch": 0.5824352212795231,
+      "grad_norm": 0.7097499752802617,
+      "learning_rate": 0.0003916391414268718,
+      "loss": 2.3145,
+      "step": 1270
+    },
+    {
+      "epoch": 0.5828938316899793,
+      "grad_norm": 0.7549808550028023,
+      "learning_rate": 0.00039091387635591536,
+      "loss": 2.4351,
+      "step": 1271
+    },
+    {
+      "epoch": 0.5833524421004357,
+      "grad_norm": 0.7463697763035759,
+      "learning_rate": 0.0003901888521973894,
+      "loss": 2.2752,
+      "step": 1272
+    },
+    {
+      "epoch": 0.583811052510892,
+      "grad_norm": 0.7437507151207823,
+      "learning_rate": 0.0003894640705524813,
+      "loss": 2.4005,
+      "step": 1273
+    },
+    {
+      "epoch": 0.5842696629213483,
+      "grad_norm": 0.7250662676041209,
+      "learning_rate": 0.00038873953302184284,
+      "loss": 2.2589,
+      "step": 1274
+    },
+    {
+      "epoch": 0.5847282733318047,
+      "grad_norm": 0.7195841665851344,
+      "learning_rate": 0.000388015241205587,
+      "loss": 2.2574,
+      "step": 1275
+    },
+    {
+      "epoch": 0.5851868837422609,
+      "grad_norm": 0.7621223972724288,
+      "learning_rate": 0.00038729119670328355,
+      "loss": 2.4038,
+      "step": 1276
+    },
+    {
+      "epoch": 0.5856454941527173,
+      "grad_norm": 0.7357402433432864,
+      "learning_rate": 0.00038656740111395665,
+      "loss": 2.3557,
+      "step": 1277
+    },
+    {
+      "epoch": 0.5861041045631736,
+      "grad_norm": 0.8051298292397154,
+      "learning_rate": 0.00038584385603608053,
+      "loss": 2.4091,
+      "step": 1278
+    },
+    {
+      "epoch": 0.5865627149736299,
+      "grad_norm": 0.7628003485705646,
+      "learning_rate": 0.00038512056306757615,
+      "loss": 2.4235,
+      "step": 1279
+    },
+    {
+      "epoch": 0.5870213253840862,
+      "grad_norm": 0.7203706496024823,
+      "learning_rate": 0.0003843975238058075,
+      "loss": 2.3358,
+      "step": 1280
+    },
+    {
+      "epoch": 0.5874799357945425,
+      "grad_norm": 0.6848178224259719,
+      "learning_rate": 0.00038367473984757863,
+      "loss": 2.3439,
+      "step": 1281
+    },
+    {
+      "epoch": 0.5879385462049989,
+      "grad_norm": 0.747536174086349,
+      "learning_rate": 0.0003829522127891296,
+      "loss": 2.4331,
+      "step": 1282
+    },
+    {
+      "epoch": 0.5883971566154552,
+      "grad_norm": 0.7295279385192088,
+      "learning_rate": 0.0003822299442261329,
+      "loss": 2.2501,
+      "step": 1283
+    },
+    {
+      "epoch": 0.5888557670259115,
+      "grad_norm": 0.7330717067822509,
+      "learning_rate": 0.00038150793575369063,
+      "loss": 2.3615,
+      "step": 1284
+    },
+    {
+      "epoch": 0.5893143774363678,
+      "grad_norm": 0.7547994877789342,
+      "learning_rate": 0.0003807861889663299,
+      "loss": 2.2962,
+      "step": 1285
+    },
+    {
+      "epoch": 0.5897729878468241,
+      "grad_norm": 0.766029346892157,
+      "learning_rate": 0.0003800647054580006,
+      "loss": 2.3157,
+      "step": 1286
+    },
+    {
+      "epoch": 0.5902315982572804,
+      "grad_norm": 0.7309019468778013,
+      "learning_rate": 0.00037934348682207064,
+      "loss": 2.3157,
+      "step": 1287
+    },
+    {
+      "epoch": 0.5906902086677368,
+      "grad_norm": 0.7127771856817978,
+      "learning_rate": 0.00037862253465132306,
+      "loss": 2.2819,
+      "step": 1288
+    },
+    {
+      "epoch": 0.591148819078193,
+      "grad_norm": 0.7419629567075842,
+      "learning_rate": 0.00037790185053795245,
+      "loss": 2.4238,
+      "step": 1289
+    },
+    {
+      "epoch": 0.5916074294886494,
+      "grad_norm": 0.824698113972859,
+      "learning_rate": 0.0003771814360735616,
+      "loss": 2.3734,
+      "step": 1290
+    },
+    {
+      "epoch": 0.5920660398991057,
+      "grad_norm": 0.769201230650809,
+      "learning_rate": 0.00037646129284915755,
+      "loss": 2.3634,
+      "step": 1291
+    },
+    {
+      "epoch": 0.592524650309562,
+      "grad_norm": 0.7963060295978823,
+      "learning_rate": 0.00037574142245514825,
+      "loss": 2.3223,
+      "step": 1292
+    },
+    {
+      "epoch": 0.5929832607200184,
+      "grad_norm": 0.7214457819385294,
+      "learning_rate": 0.0003750218264813393,
+      "loss": 2.3961,
+      "step": 1293
+    },
+    {
+      "epoch": 0.5934418711304746,
+      "grad_norm": 0.7447714852409056,
+      "learning_rate": 0.0003743025065169305,
+      "loss": 2.3864,
+      "step": 1294
+    },
+    {
+      "epoch": 0.593900481540931,
+      "grad_norm": 0.7173991841398818,
+      "learning_rate": 0.0003735834641505116,
+      "loss": 2.4617,
+      "step": 1295
+    },
+    {
+      "epoch": 0.5943590919513873,
+      "grad_norm": 0.6971216237157616,
+      "learning_rate": 0.00037286470097005954,
+      "loss": 2.2722,
+      "step": 1296
+    },
+    {
+      "epoch": 0.5948177023618436,
+      "grad_norm": 0.7046043990704357,
+      "learning_rate": 0.0003721462185629347,
+      "loss": 2.3183,
+      "step": 1297
+    },
+    {
+      "epoch": 0.5952763127723,
+      "grad_norm": 0.7477714594771898,
+      "learning_rate": 0.00037142801851587707,
+      "loss": 2.3973,
+      "step": 1298
+    },
+    {
+      "epoch": 0.5957349231827562,
+      "grad_norm": 0.6911238230153542,
+      "learning_rate": 0.00037071010241500357,
+      "loss": 2.3788,
+      "step": 1299
+    },
+    {
+      "epoch": 0.5961935335932126,
+      "grad_norm": 0.7376427082455745,
+      "learning_rate": 0.00036999247184580383,
+      "loss": 2.2972,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5966521440036688,
+      "grad_norm": 0.6716378153375236,
+      "learning_rate": 0.00036927512839313636,
+      "loss": 2.3445,
+      "step": 1301
+    },
+    {
+      "epoch": 0.5971107544141252,
+      "grad_norm": 0.6715921005259583,
+      "learning_rate": 0.0003685580736412268,
+      "loss": 2.299,
+      "step": 1302
+    },
+    {
+      "epoch": 0.5975693648245816,
+      "grad_norm": 0.7463760818163089,
+      "learning_rate": 0.000367841309173662,
+      "loss": 2.3119,
+      "step": 1303
+    },
+    {
+      "epoch": 0.5980279752350378,
+      "grad_norm": 0.7175379520499743,
+      "learning_rate": 0.0003671248365733883,
+      "loss": 2.2713,
+      "step": 1304
+    },
+    {
+      "epoch": 0.5984865856454942,
+      "grad_norm": 0.7188034576689705,
+      "learning_rate": 0.0003664086574227075,
+      "loss": 2.2417,
+      "step": 1305
+    },
+    {
+      "epoch": 0.5989451960559504,
+      "grad_norm": 0.7186972561740652,
+      "learning_rate": 0.000365692773303273,
+      "loss": 2.3415,
+      "step": 1306
+    },
+    {
+      "epoch": 0.5994038064664068,
+      "grad_norm": 0.7206876959948625,
+      "learning_rate": 0.00036497718579608696,
+      "loss": 2.2928,
+      "step": 1307
+    },
+    {
+      "epoch": 0.5998624168768631,
+      "grad_norm": 0.7249008181665882,
+      "learning_rate": 0.0003642618964814964,
+      "loss": 2.3331,
+      "step": 1308
+    },
+    {
+      "epoch": 0.6003210272873194,
+      "grad_norm": 0.7983039757125537,
+      "learning_rate": 0.00036354690693918946,
+      "loss": 2.424,
+      "step": 1309
+    },
+    {
+      "epoch": 0.6007796376977758,
+      "grad_norm": 0.7235192684474422,
+      "learning_rate": 0.00036283221874819284,
+      "loss": 2.3609,
+      "step": 1310
+    },
+    {
+      "epoch": 0.601238248108232,
+      "grad_norm": 0.7890243748268781,
+      "learning_rate": 0.0003621178334868672,
+      "loss": 2.363,
+      "step": 1311
+    },
+    {
+      "epoch": 0.6016968585186884,
+      "grad_norm": 0.7265571686093398,
+      "learning_rate": 0.00036140375273290476,
+      "loss": 2.3468,
+      "step": 1312
+    },
+    {
+      "epoch": 0.6021554689291447,
+      "grad_norm": 0.7248139387066259,
+      "learning_rate": 0.0003606899780633245,
+      "loss": 2.2684,
+      "step": 1313
+    },
+    {
+      "epoch": 0.602614079339601,
+      "grad_norm": 0.7565347064732125,
+      "learning_rate": 0.0003599765110544699,
+      "loss": 2.2574,
+      "step": 1314
+    },
+    {
+      "epoch": 0.6030726897500573,
+      "grad_norm": 0.71916520811086,
+      "learning_rate": 0.0003592633532820052,
+      "loss": 2.2643,
+      "step": 1315
+    },
+    {
+      "epoch": 0.6035313001605136,
+      "grad_norm": 0.72935238341795,
+      "learning_rate": 0.0003585505063209109,
+      "loss": 2.3468,
+      "step": 1316
+    },
+    {
+      "epoch": 0.60398991057097,
+      "grad_norm": 0.7343058777504048,
+      "learning_rate": 0.00035783797174548194,
+      "loss": 2.4433,
+      "step": 1317
+    },
+    {
+      "epoch": 0.6044485209814263,
+      "grad_norm": 0.6854776043652366,
+      "learning_rate": 0.00035712575112932277,
+      "loss": 2.2341,
+      "step": 1318
+    },
+    {
+      "epoch": 0.6049071313918826,
+      "grad_norm": 0.7388526234885471,
+      "learning_rate": 0.000356413846045345,
+      "loss": 2.4864,
+      "step": 1319
+    },
+    {
+      "epoch": 0.6053657418023389,
+      "grad_norm": 0.7080175250719067,
+      "learning_rate": 0.000355702258065763,
+      "loss": 2.2998,
+      "step": 1320
+    },
+    {
+      "epoch": 0.6058243522127952,
+      "grad_norm": 0.7626004008126402,
+      "learning_rate": 0.0003549909887620909,
+      "loss": 2.2976,
+      "step": 1321
+    },
+    {
+      "epoch": 0.6062829626232515,
+      "grad_norm": 0.7636938064504112,
+      "learning_rate": 0.00035428003970513914,
+      "loss": 2.378,
+      "step": 1322
+    },
+    {
+      "epoch": 0.6067415730337079,
+      "grad_norm": 0.7639860219490429,
+      "learning_rate": 0.00035356941246501085,
+      "loss": 2.3883,
+      "step": 1323
+    },
+    {
+      "epoch": 0.6072001834441642,
+      "grad_norm": 0.6790859988820608,
+      "learning_rate": 0.0003528591086110984,
+      "loss": 2.3567,
+      "step": 1324
+    },
+    {
+      "epoch": 0.6076587938546205,
+      "grad_norm": 0.7778513860928605,
+      "learning_rate": 0.00035214912971208,
+      "loss": 2.275,
+      "step": 1325
+    },
+    {
+      "epoch": 0.6081174042650768,
+      "grad_norm": 0.6958326800881296,
+      "learning_rate": 0.0003514394773359163,
+      "loss": 2.2065,
+      "step": 1326
+    },
+    {
+      "epoch": 0.6085760146755331,
+      "grad_norm": 0.7662650397788965,
+      "learning_rate": 0.0003507301530498469,
+      "loss": 2.3681,
+      "step": 1327
+    },
+    {
+      "epoch": 0.6090346250859895,
+      "grad_norm": 0.7574953670417077,
+      "learning_rate": 0.00035002115842038646,
+      "loss": 2.2995,
+      "step": 1328
+    },
+    {
+      "epoch": 0.6094932354964457,
+      "grad_norm": 0.7649105727325206,
+      "learning_rate": 0.00034931249501332195,
+      "loss": 2.3456,
+      "step": 1329
+    },
+    {
+      "epoch": 0.6099518459069021,
+      "grad_norm": 0.7964732893709734,
+      "learning_rate": 0.00034860416439370885,
+      "loss": 2.3837,
+      "step": 1330
+    },
+    {
+      "epoch": 0.6104104563173584,
+      "grad_norm": 0.7403403441588019,
+      "learning_rate": 0.0003478961681258674,
+      "loss": 2.3857,
+      "step": 1331
+    },
+    {
+      "epoch": 0.6108690667278147,
+      "grad_norm": 0.6975670677441976,
+      "learning_rate": 0.0003471885077733796,
+      "loss": 2.3094,
+      "step": 1332
+    },
+    {
+      "epoch": 0.6113276771382711,
+      "grad_norm": 0.7289956160471281,
+      "learning_rate": 0.0003464811848990859,
+      "loss": 2.3692,
+      "step": 1333
+    },
+    {
+      "epoch": 0.6117862875487273,
+      "grad_norm": 0.7191573812196363,
+      "learning_rate": 0.00034577420106508063,
+      "loss": 2.4324,
+      "step": 1334
+    },
+    {
+      "epoch": 0.6122448979591837,
+      "grad_norm": 0.6967590249593395,
+      "learning_rate": 0.0003450675578327105,
+      "loss": 2.2871,
+      "step": 1335
+    },
+    {
+      "epoch": 0.6127035083696399,
+      "grad_norm": 0.7159501621265978,
+      "learning_rate": 0.000344361256762569,
+      "loss": 2.3187,
+      "step": 1336
+    },
+    {
+      "epoch": 0.6131621187800963,
+      "grad_norm": 0.7086282245569108,
+      "learning_rate": 0.00034365529941449456,
+      "loss": 2.2103,
+      "step": 1337
+    },
+    {
+      "epoch": 0.6136207291905527,
+      "grad_norm": 0.756610442609131,
+      "learning_rate": 0.0003429496873475664,
+      "loss": 2.3008,
+      "step": 1338
+    },
+    {
+      "epoch": 0.6140793396010089,
+      "grad_norm": 0.7133854872088641,
+      "learning_rate": 0.0003422444221201009,
+      "loss": 2.2656,
+      "step": 1339
+    },
+    {
+      "epoch": 0.6145379500114653,
+      "grad_norm": 0.7120255431966003,
+      "learning_rate": 0.0003415395052896487,
+      "loss": 2.369,
+      "step": 1340
+    },
+    {
+      "epoch": 0.6149965604219215,
+      "grad_norm": 0.7249331217387454,
+      "learning_rate": 0.0003408349384129912,
+      "loss": 2.3891,
+      "step": 1341
+    },
+    {
+      "epoch": 0.6154551708323779,
+      "grad_norm": 0.7206253752696861,
+      "learning_rate": 0.00034013072304613643,
+      "loss": 2.2972,
+      "step": 1342
+    },
+    {
+      "epoch": 0.6159137812428342,
+      "grad_norm": 0.7231818433182353,
+      "learning_rate": 0.00033942686074431674,
+      "loss": 2.3614,
+      "step": 1343
+    },
+    {
+      "epoch": 0.6163723916532905,
+      "grad_norm": 0.6848043503500569,
+      "learning_rate": 0.0003387233530619843,
+      "loss": 2.3212,
+      "step": 1344
+    },
+    {
+      "epoch": 0.6168310020637469,
+      "grad_norm": 0.7129702747688489,
+      "learning_rate": 0.0003380202015528084,
+      "loss": 2.46,
+      "step": 1345
+    },
+    {
+      "epoch": 0.6172896124742032,
+      "grad_norm": 0.6809186285365622,
+      "learning_rate": 0.0003373174077696715,
+      "loss": 2.3533,
+      "step": 1346
+    },
+    {
+      "epoch": 0.6177482228846595,
+      "grad_norm": 0.7329989994309805,
+      "learning_rate": 0.0003366149732646661,
+      "loss": 2.301,
+      "step": 1347
+    },
+    {
+      "epoch": 0.6182068332951158,
+      "grad_norm": 0.7325333277224166,
+      "learning_rate": 0.00033591289958909143,
+      "loss": 2.2888,
+      "step": 1348
+    },
+    {
+      "epoch": 0.6186654437055721,
+      "grad_norm": 0.7210347387432782,
+      "learning_rate": 0.00033521118829344954,
+      "loss": 2.3034,
+      "step": 1349
+    },
+    {
+      "epoch": 0.6191240541160284,
+      "grad_norm": 0.7264862346583063,
+      "learning_rate": 0.0003345098409274423,
+      "loss": 2.4944,
+      "step": 1350
+    },
+    {
+      "epoch": 0.6195826645264848,
+      "grad_norm": 0.7234593797851477,
+      "learning_rate": 0.00033380885903996796,
+      "loss": 2.2759,
+      "step": 1351
+    },
+    {
+      "epoch": 0.6200412749369411,
+      "grad_norm": 0.7071129192886292,
+      "learning_rate": 0.00033310824417911766,
+      "loss": 2.3938,
+      "step": 1352
+    },
+    {
+      "epoch": 0.6204998853473974,
+      "grad_norm": 0.7319716528511366,
+      "learning_rate": 0.00033240799789217184,
+      "loss": 2.4167,
+      "step": 1353
+    },
+    {
+      "epoch": 0.6209584957578537,
+      "grad_norm": 0.7177646354213443,
+      "learning_rate": 0.00033170812172559694,
+      "loss": 2.2761,
+      "step": 1354
+    },
+    {
+      "epoch": 0.62141710616831,
+      "grad_norm": 0.7306680240966001,
+      "learning_rate": 0.000331008617225042,
+      "loss": 2.3067,
+      "step": 1355
+    },
+    {
+      "epoch": 0.6218757165787664,
+      "grad_norm": 0.6876538762999067,
+      "learning_rate": 0.0003303094859353355,
+      "loss": 2.3259,
+      "step": 1356
+    },
+    {
+      "epoch": 0.6223343269892226,
+      "grad_norm": 0.6983937354118633,
+      "learning_rate": 0.0003296107294004812,
+      "loss": 2.2446,
+      "step": 1357
+    },
+    {
+      "epoch": 0.622792937399679,
+      "grad_norm": 0.7097167362246571,
+      "learning_rate": 0.0003289123491636559,
+      "loss": 2.3555,
+      "step": 1358
+    },
+    {
+      "epoch": 0.6232515478101353,
+      "grad_norm": 0.6765003000183394,
+      "learning_rate": 0.00032821434676720443,
+      "loss": 2.419,
+      "step": 1359
+    },
+    {
+      "epoch": 0.6237101582205916,
+      "grad_norm": 0.7009804519222963,
+      "learning_rate": 0.00032751672375263836,
+      "loss": 2.3241,
+      "step": 1360
+    },
+    {
+      "epoch": 0.624168768631048,
+      "grad_norm": 0.7051169051755481,
+      "learning_rate": 0.0003268194816606305,
+      "loss": 2.3501,
+      "step": 1361
+    },
+    {
+      "epoch": 0.6246273790415042,
+      "grad_norm": 0.7559206100158674,
+      "learning_rate": 0.00032612262203101267,
+      "loss": 2.3309,
+      "step": 1362
+    },
+    {
+      "epoch": 0.6250859894519606,
+      "grad_norm": 0.6976323689686499,
+      "learning_rate": 0.00032542614640277225,
+      "loss": 2.2673,
+      "step": 1363
+    },
+    {
+      "epoch": 0.6255445998624168,
+      "grad_norm": 0.7250439542092385,
+      "learning_rate": 0.0003247300563140481,
+      "loss": 2.291,
+      "step": 1364
+    },
+    {
+      "epoch": 0.6260032102728732,
+      "grad_norm": 0.7386373690762259,
+      "learning_rate": 0.00032403435330212807,
+      "loss": 2.3663,
+      "step": 1365
+    },
+    {
+      "epoch": 0.6264618206833296,
+      "grad_norm": 0.7393725957569647,
+      "learning_rate": 0.00032333903890344515,
+      "loss": 2.2557,
+      "step": 1366
+    },
+    {
+      "epoch": 0.6269204310937858,
+      "grad_norm": 0.7165746391025397,
+      "learning_rate": 0.00032264411465357333,
+      "loss": 2.4115,
+      "step": 1367
+    },
+    {
+      "epoch": 0.6273790415042422,
+      "grad_norm": 0.7574228992431871,
+      "learning_rate": 0.00032194958208722654,
+      "loss": 2.2772,
+      "step": 1368
+    },
+    {
+      "epoch": 0.6278376519146984,
+      "grad_norm": 0.7255232737867822,
+      "learning_rate": 0.00032125544273825204,
+      "loss": 2.2563,
+      "step": 1369
+    },
+    {
+      "epoch": 0.6282962623251548,
+      "grad_norm": 0.7906138464890484,
+      "learning_rate": 0.0003205616981396297,
+      "loss": 2.4707,
+      "step": 1370
+    },
+    {
+      "epoch": 0.6287548727356111,
+      "grad_norm": 0.7364100490030853,
+      "learning_rate": 0.00031986834982346713,
+      "loss": 2.2816,
+      "step": 1371
+    },
+    {
+      "epoch": 0.6292134831460674,
+      "grad_norm": 0.7495672886436979,
+      "learning_rate": 0.00031917539932099694,
+      "loss": 2.262,
+      "step": 1372
+    },
+    {
+      "epoch": 0.6296720935565238,
+      "grad_norm": 0.6867995693730848,
+      "learning_rate": 0.00031848284816257336,
+      "loss": 2.3458,
+      "step": 1373
+    },
+    {
+      "epoch": 0.63013070396698,
+      "grad_norm": 0.7373570199024532,
+      "learning_rate": 0.0003177906978776682,
+      "loss": 2.3328,
+      "step": 1374
+    },
+    {
+      "epoch": 0.6305893143774364,
+      "grad_norm": 0.7276037063057947,
+      "learning_rate": 0.0003170989499948683,
+      "loss": 2.4325,
+      "step": 1375
+    },
+    {
+      "epoch": 0.6310479247878927,
+      "grad_norm": 0.746053524284194,
+      "learning_rate": 0.0003164076060418719,
+      "loss": 2.3852,
+      "step": 1376
+    },
+    {
+      "epoch": 0.631506535198349,
+      "grad_norm": 0.7365786408248687,
+      "learning_rate": 0.000315716667545485,
+      "loss": 2.2298,
+      "step": 1377
+    },
+    {
+      "epoch": 0.6319651456088053,
+      "grad_norm": 0.6724183330622866,
+      "learning_rate": 0.00031502613603161836,
+      "loss": 2.3789,
+      "step": 1378
+    },
+    {
+      "epoch": 0.6324237560192616,
+      "grad_norm": 0.7105556534483881,
+      "learning_rate": 0.00031433601302528335,
+      "loss": 2.344,
+      "step": 1379
+    },
+    {
+      "epoch": 0.632882366429718,
+      "grad_norm": 0.6924767074925289,
+      "learning_rate": 0.00031364630005058995,
+      "loss": 2.2883,
+      "step": 1380
+    },
+    {
+      "epoch": 0.6333409768401743,
+      "grad_norm": 0.7139888217782515,
+      "learning_rate": 0.0003129569986307422,
+      "loss": 2.323,
+      "step": 1381
+    },
+    {
+      "epoch": 0.6337995872506306,
+      "grad_norm": 0.7293462875495649,
+      "learning_rate": 0.00031226811028803515,
+      "loss": 2.306,
+      "step": 1382
+    },
+    {
+      "epoch": 0.6342581976610869,
+      "grad_norm": 0.7097109405458358,
+      "learning_rate": 0.00031157963654385173,
+      "loss": 2.3998,
+      "step": 1383
+    },
+    {
+      "epoch": 0.6347168080715432,
+      "grad_norm": 0.7287313549216252,
+      "learning_rate": 0.0003108915789186592,
+      "loss": 2.4745,
+      "step": 1384
+    },
+    {
+      "epoch": 0.6351754184819995,
+      "grad_norm": 0.7423883553364887,
+      "learning_rate": 0.00031020393893200604,
+      "loss": 2.3337,
+      "step": 1385
+    },
+    {
+      "epoch": 0.6356340288924559,
+      "grad_norm": 0.7115193443261759,
+      "learning_rate": 0.00030951671810251823,
+      "loss": 2.2915,
+      "step": 1386
+    },
+    {
+      "epoch": 0.6360926393029122,
+      "grad_norm": 0.6989803448464067,
+      "learning_rate": 0.0003088299179478959,
+      "loss": 2.3139,
+      "step": 1387
+    },
+    {
+      "epoch": 0.6365512497133685,
+      "grad_norm": 0.7298398024286988,
+      "learning_rate": 0.0003081435399849104,
+      "loss": 2.4055,
+      "step": 1388
+    },
+    {
+      "epoch": 0.6370098601238248,
+      "grad_norm": 0.7197813593230928,
+      "learning_rate": 0.0003074575857294004,
+      "loss": 2.3315,
+      "step": 1389
+    },
+    {
+      "epoch": 0.6374684705342811,
+      "grad_norm": 0.7384768592636004,
+      "learning_rate": 0.0003067720566962691,
+      "loss": 2.3195,
+      "step": 1390
+    },
+    {
+      "epoch": 0.6379270809447375,
+      "grad_norm": 0.7287865695704606,
+      "learning_rate": 0.0003060869543994806,
+      "loss": 2.3946,
+      "step": 1391
+    },
+    {
+      "epoch": 0.6383856913551937,
+      "grad_norm": 0.7179811219773088,
+      "learning_rate": 0.0003054022803520562,
+      "loss": 2.3122,
+      "step": 1392
+    },
+    {
+      "epoch": 0.6388443017656501,
+      "grad_norm": 0.7518626406727406,
+      "learning_rate": 0.0003047180360660721,
+      "loss": 2.3324,
+      "step": 1393
+    },
+    {
+      "epoch": 0.6393029121761064,
+      "grad_norm": 0.7116298821310875,
+      "learning_rate": 0.00030403422305265475,
+      "loss": 2.296,
+      "step": 1394
+    },
+    {
+      "epoch": 0.6397615225865627,
+      "grad_norm": 0.7477024212610466,
+      "learning_rate": 0.0003033508428219785,
+      "loss": 2.3508,
+      "step": 1395
+    },
+    {
+      "epoch": 0.6402201329970191,
+      "grad_norm": 0.7246054366244431,
+      "learning_rate": 0.00030266789688326184,
+      "loss": 2.4199,
+      "step": 1396
+    },
+    {
+      "epoch": 0.6406787434074753,
+      "grad_norm": 0.731726103992136,
+      "learning_rate": 0.00030198538674476393,
+      "loss": 2.3278,
+      "step": 1397
+    },
+    {
+      "epoch": 0.6411373538179317,
+      "grad_norm": 0.7485300324103513,
+      "learning_rate": 0.00030130331391378185,
+      "loss": 2.2805,
+      "step": 1398
+    },
+    {
+      "epoch": 0.6415959642283879,
+      "grad_norm": 0.7311965818054108,
+      "learning_rate": 0.0003006216798966468,
+      "loss": 2.2703,
+      "step": 1399
+    },
+    {
+      "epoch": 0.6420545746388443,
+      "grad_norm": 0.7236400939630404,
+      "learning_rate": 0.00029994048619872034,
+      "loss": 2.3291,
+      "step": 1400
+    },
+    {
+      "epoch": 0.6425131850493007,
+      "grad_norm": 0.754050912652918,
+      "learning_rate": 0.0002992597343243927,
+      "loss": 2.3219,
+      "step": 1401
+    },
+    {
+      "epoch": 0.6429717954597569,
+      "grad_norm": 0.699101546165523,
+      "learning_rate": 0.0002985794257770773,
+      "loss": 2.4222,
+      "step": 1402
+    },
+    {
+      "epoch": 0.6434304058702133,
+      "grad_norm": 0.7567033736973187,
+      "learning_rate": 0.0002978995620592092,
+      "loss": 2.3944,
+      "step": 1403
+    },
+    {
+      "epoch": 0.6438890162806695,
+      "grad_norm": 0.7611491665785218,
+      "learning_rate": 0.0002972201446722405,
+      "loss": 2.3112,
+      "step": 1404
+    },
+    {
+      "epoch": 0.6443476266911259,
+      "grad_norm": 0.7024613521370716,
+      "learning_rate": 0.00029654117511663803,
+      "loss": 2.3413,
+      "step": 1405
+    },
+    {
+      "epoch": 0.6448062371015822,
+      "grad_norm": 0.7147915482270434,
+      "learning_rate": 0.0002958626548918795,
+      "loss": 2.2903,
+      "step": 1406
+    },
+    {
+      "epoch": 0.6452648475120385,
+      "grad_norm": 0.721909498679481,
+      "learning_rate": 0.00029518458549645014,
+      "loss": 2.4098,
+      "step": 1407
+    },
+    {
+      "epoch": 0.6457234579224949,
+      "grad_norm": 0.7211943990264993,
+      "learning_rate": 0.00029450696842783954,
+      "loss": 2.2812,
+      "step": 1408
+    },
+    {
+      "epoch": 0.6461820683329511,
+      "grad_norm": 0.6981087740156533,
+      "learning_rate": 0.00029382980518253865,
+      "loss": 2.2835,
+      "step": 1409
+    },
+    {
+      "epoch": 0.6466406787434075,
+      "grad_norm": 0.7349020481181259,
+      "learning_rate": 0.00029315309725603595,
+      "loss": 2.3544,
+      "step": 1410
+    },
+    {
+      "epoch": 0.6470992891538638,
+      "grad_norm": 0.7502869726616941,
+      "learning_rate": 0.00029247684614281446,
+      "loss": 2.3498,
+      "step": 1411
+    },
+    {
+      "epoch": 0.6475578995643201,
+      "grad_norm": 0.7414017426299883,
+      "learning_rate": 0.0002918010533363481,
+      "loss": 2.3133,
+      "step": 1412
+    },
+    {
+      "epoch": 0.6480165099747764,
+      "grad_norm": 0.7118582963687152,
+      "learning_rate": 0.0002911257203290987,
+      "loss": 2.3269,
+      "step": 1413
+    },
+    {
+      "epoch": 0.6484751203852327,
+      "grad_norm": 0.7093081968755305,
+      "learning_rate": 0.00029045084861251314,
+      "loss": 2.352,
+      "step": 1414
+    },
+    {
+      "epoch": 0.6489337307956891,
+      "grad_norm": 0.708685641768302,
+      "learning_rate": 0.00028977643967701897,
+      "loss": 2.3698,
+      "step": 1415
+    },
+    {
+      "epoch": 0.6493923412061454,
+      "grad_norm": 0.7352625927131423,
+      "learning_rate": 0.00028910249501202156,
+      "loss": 2.3077,
+      "step": 1416
+    },
+    {
+      "epoch": 0.6498509516166017,
+      "grad_norm": 0.8017424842676608,
+      "learning_rate": 0.00028842901610590165,
+      "loss": 2.3792,
+      "step": 1417
+    },
+    {
+      "epoch": 0.650309562027058,
+      "grad_norm": 0.7330423556889765,
+      "learning_rate": 0.00028775600444601123,
+      "loss": 2.3838,
+      "step": 1418
+    },
+    {
+      "epoch": 0.6507681724375143,
+      "grad_norm": 0.7110192517076982,
+      "learning_rate": 0.00028708346151866973,
+      "loss": 2.2634,
+      "step": 1419
+    },
+    {
+      "epoch": 0.6512267828479706,
+      "grad_norm": 0.7528763246087339,
+      "learning_rate": 0.0002864113888091622,
+      "loss": 2.4198,
+      "step": 1420
+    },
+    {
+      "epoch": 0.651685393258427,
+      "grad_norm": 0.7144881353109057,
+      "learning_rate": 0.0002857397878017348,
+      "loss": 2.3562,
+      "step": 1421
+    },
+    {
+      "epoch": 0.6521440036688833,
+      "grad_norm": 0.7309409813784762,
+      "learning_rate": 0.00028506865997959173,
+      "loss": 2.304,
+      "step": 1422
+    },
+    {
+      "epoch": 0.6526026140793396,
+      "grad_norm": 0.7327862776397656,
+      "learning_rate": 0.000284398006824893,
+      "loss": 2.3991,
+      "step": 1423
+    },
+    {
+      "epoch": 0.6530612244897959,
+      "grad_norm": 0.7524024557145078,
+      "learning_rate": 0.00028372782981874963,
+      "loss": 2.3319,
+      "step": 1424
+    },
+    {
+      "epoch": 0.6535198349002522,
+      "grad_norm": 0.7691091580139142,
+      "learning_rate": 0.00028305813044122096,
+      "loss": 2.2968,
+      "step": 1425
+    },
+    {
+      "epoch": 0.6539784453107086,
+      "grad_norm": 0.7508142496111878,
+      "learning_rate": 0.0002823889101713122,
+      "loss": 2.3185,
+      "step": 1426
+    },
+    {
+      "epoch": 0.6544370557211648,
+      "grad_norm": 0.715616931750071,
+      "learning_rate": 0.0002817201704869701,
+      "loss": 2.3807,
+      "step": 1427
+    },
+    {
+      "epoch": 0.6548956661316212,
+      "grad_norm": 0.754140362313214,
+      "learning_rate": 0.00028105191286508,
+      "loss": 2.4413,
+      "step": 1428
+    },
+    {
+      "epoch": 0.6553542765420775,
+      "grad_norm": 0.769176571268719,
+      "learning_rate": 0.00028038413878146245,
+      "loss": 2.2709,
+      "step": 1429
+    },
+    {
+      "epoch": 0.6558128869525338,
+      "grad_norm": 0.7761959203036439,
+      "learning_rate": 0.00027971684971087073,
+      "loss": 2.2569,
+      "step": 1430
+    },
+    {
+      "epoch": 0.6562714973629902,
+      "grad_norm": 0.7131416760573112,
+      "learning_rate": 0.00027905004712698643,
+      "loss": 2.2995,
+      "step": 1431
+    },
+    {
+      "epoch": 0.6567301077734464,
+      "grad_norm": 0.7265593336966315,
+      "learning_rate": 0.0002783837325024167,
+      "loss": 2.3467,
+      "step": 1432
+    },
+    {
+      "epoch": 0.6571887181839028,
+      "grad_norm": 0.6904184688367537,
+      "learning_rate": 0.00027771790730869153,
+      "loss": 2.383,
+      "step": 1433
+    },
+    {
+      "epoch": 0.657647328594359,
+      "grad_norm": 0.735345555031758,
+      "learning_rate": 0.0002770525730162599,
+      "loss": 2.2945,
+      "step": 1434
+    },
+    {
+      "epoch": 0.6581059390048154,
+      "grad_norm": 0.7637801745625725,
+      "learning_rate": 0.00027638773109448645,
+      "loss": 2.3113,
+      "step": 1435
+    },
+    {
+      "epoch": 0.6585645494152718,
+      "grad_norm": 0.7194127062640134,
+      "learning_rate": 0.00027572338301164824,
+      "loss": 2.4066,
+      "step": 1436
+    },
+    {
+      "epoch": 0.659023159825728,
+      "grad_norm": 0.7593629178200951,
+      "learning_rate": 0.0002750595302349324,
+      "loss": 2.2321,
+      "step": 1437
+    },
+    {
+      "epoch": 0.6594817702361844,
+      "grad_norm": 0.7108596207103615,
+      "learning_rate": 0.00027439617423043145,
+      "loss": 2.2481,
+      "step": 1438
+    },
+    {
+      "epoch": 0.6599403806466407,
+      "grad_norm": 0.7211067852888381,
+      "learning_rate": 0.00027373331646314114,
+      "loss": 2.2338,
+      "step": 1439
+    },
+    {
+      "epoch": 0.660398991057097,
+      "grad_norm": 0.7336647521458878,
+      "learning_rate": 0.0002730709583969572,
+      "loss": 2.2775,
+      "step": 1440
+    },
+    {
+      "epoch": 0.6608576014675533,
+      "grad_norm": 0.705362526020865,
+      "learning_rate": 0.0002724091014946711,
+      "loss": 2.4054,
+      "step": 1441
+    },
+    {
+      "epoch": 0.6613162118780096,
+      "grad_norm": 0.7554900109519131,
+      "learning_rate": 0.00027174774721796824,
+      "loss": 2.3299,
+      "step": 1442
+    },
+    {
+      "epoch": 0.661774822288466,
+      "grad_norm": 0.7187558719648468,
+      "learning_rate": 0.0002710868970274232,
+      "loss": 2.2258,
+      "step": 1443
+    },
+    {
+      "epoch": 0.6622334326989223,
+      "grad_norm": 0.6946064468869274,
+      "learning_rate": 0.0002704265523824982,
+      "loss": 2.2412,
+      "step": 1444
+    },
+    {
+      "epoch": 0.6626920431093786,
+      "grad_norm": 0.7252071465324326,
+      "learning_rate": 0.00026976671474153826,
+      "loss": 2.3626,
+      "step": 1445
+    },
+    {
+      "epoch": 0.6631506535198349,
+      "grad_norm": 0.7290074982893177,
+      "learning_rate": 0.00026910738556176886,
+      "loss": 2.373,
+      "step": 1446
+    },
+    {
+      "epoch": 0.6636092639302912,
+      "grad_norm": 0.7601018459872515,
+      "learning_rate": 0.0002684485662992929,
+      "loss": 2.3697,
+      "step": 1447
+    },
+    {
+      "epoch": 0.6640678743407475,
+      "grad_norm": 0.7344212561465886,
+      "learning_rate": 0.0002677902584090869,
+      "loss": 2.3984,
+      "step": 1448
+    },
+    {
+      "epoch": 0.6645264847512039,
+      "grad_norm": 0.7425781535936111,
+      "learning_rate": 0.00026713246334499774,
+      "loss": 2.2459,
+      "step": 1449
+    },
+    {
+      "epoch": 0.6649850951616602,
+      "grad_norm": 0.7496324828257173,
+      "learning_rate": 0.00026647518255974023,
+      "loss": 2.2967,
+      "step": 1450
+    },
+    {
+      "epoch": 0.6654437055721165,
+      "grad_norm": 0.6815440724710059,
+      "learning_rate": 0.0002658184175048934,
+      "loss": 2.2476,
+      "step": 1451
+    },
+    {
+      "epoch": 0.6659023159825728,
+      "grad_norm": 0.7531888235383093,
+      "learning_rate": 0.00026516216963089694,
+      "loss": 2.3229,
+      "step": 1452
+    },
+    {
+      "epoch": 0.6663609263930291,
+      "grad_norm": 0.7207938542363947,
+      "learning_rate": 0.0002645064403870488,
+      "loss": 2.3378,
+      "step": 1453
+    },
+    {
+      "epoch": 0.6668195368034855,
+      "grad_norm": 0.7588623431023215,
+      "learning_rate": 0.0002638512312215011,
+      "loss": 2.346,
+      "step": 1454
+    },
+    {
+      "epoch": 0.6672781472139417,
+      "grad_norm": 0.7026998158625682,
+      "learning_rate": 0.0002631965435812575,
+      "loss": 2.28,
+      "step": 1455
+    },
+    {
+      "epoch": 0.6677367576243981,
+      "grad_norm": 0.6962252906350248,
+      "learning_rate": 0.00026254237891217046,
+      "loss": 2.2844,
+      "step": 1456
+    },
+    {
+      "epoch": 0.6681953680348544,
+      "grad_norm": 0.7174731097023104,
+      "learning_rate": 0.0002618887386589367,
+      "loss": 2.3307,
+      "step": 1457
+    },
+    {
+      "epoch": 0.6686539784453107,
+      "grad_norm": 0.7429680021697286,
+      "learning_rate": 0.0002612356242650949,
+      "loss": 2.2802,
+      "step": 1458
+    },
+    {
+      "epoch": 0.6691125888557671,
+      "grad_norm": 0.7200230398005955,
+      "learning_rate": 0.0002605830371730229,
+      "loss": 2.2052,
+      "step": 1459
+    },
+    {
+      "epoch": 0.6695711992662233,
+      "grad_norm": 0.7340529993697791,
+      "learning_rate": 0.0002599309788239339,
+      "loss": 2.3302,
+      "step": 1460
+    },
+    {
+      "epoch": 0.6700298096766797,
+      "grad_norm": 0.7245523705144222,
+      "learning_rate": 0.00025927945065787306,
+      "loss": 2.3156,
+      "step": 1461
+    },
+    {
+      "epoch": 0.6704884200871359,
+      "grad_norm": 0.7206607381072233,
+      "learning_rate": 0.0002586284541137145,
+      "loss": 2.2538,
+      "step": 1462
+    },
+    {
+      "epoch": 0.6709470304975923,
+      "grad_norm": 0.6749119579107242,
+      "learning_rate": 0.00025797799062915905,
+      "loss": 2.3594,
+      "step": 1463
+    },
+    {
+      "epoch": 0.6714056409080487,
+      "grad_norm": 0.7056789562518191,
+      "learning_rate": 0.00025732806164072966,
+      "loss": 2.3396,
+      "step": 1464
+    },
+    {
+      "epoch": 0.6718642513185049,
+      "grad_norm": 0.7102357048677442,
+      "learning_rate": 0.00025667866858376874,
+      "loss": 2.3741,
+      "step": 1465
+    },
+    {
+      "epoch": 0.6723228617289613,
+      "grad_norm": 0.6995418919641851,
+      "learning_rate": 0.0002560298128924358,
+      "loss": 2.3893,
+      "step": 1466
+    },
+    {
+      "epoch": 0.6727814721394175,
+      "grad_norm": 0.8040365887838391,
+      "learning_rate": 0.0002553814959997032,
+      "loss": 2.4374,
+      "step": 1467
+    },
+    {
+      "epoch": 0.6732400825498739,
+      "grad_norm": 0.7431651868808552,
+      "learning_rate": 0.00025473371933735334,
+      "loss": 2.2411,
+      "step": 1468
+    },
+    {
+      "epoch": 0.6736986929603302,
+      "grad_norm": 0.7557768173267818,
+      "learning_rate": 0.00025408648433597534,
+      "loss": 2.3652,
+      "step": 1469
+    },
+    {
+      "epoch": 0.6741573033707865,
+      "grad_norm": 0.7547411700164544,
+      "learning_rate": 0.00025343979242496283,
+      "loss": 2.2294,
+      "step": 1470
+    },
+    {
+      "epoch": 0.6746159137812429,
+      "grad_norm": 0.7614800285817364,
+      "learning_rate": 0.00025279364503250925,
+      "loss": 2.4134,
+      "step": 1471
+    },
+    {
+      "epoch": 0.6750745241916991,
+      "grad_norm": 0.7218479032957226,
+      "learning_rate": 0.0002521480435856056,
+      "loss": 2.3238,
+      "step": 1472
+    },
+    {
+      "epoch": 0.6755331346021555,
+      "grad_norm": 0.730716663969642,
+      "learning_rate": 0.0002515029895100378,
+      "loss": 2.2766,
+      "step": 1473
+    },
+    {
+      "epoch": 0.6759917450126118,
+      "grad_norm": 0.7379977115263566,
+      "learning_rate": 0.0002508584842303822,
+      "loss": 2.3956,
+      "step": 1474
+    },
+    {
+      "epoch": 0.6764503554230681,
+      "grad_norm": 0.6846767128281769,
+      "learning_rate": 0.0002502145291700038,
+      "loss": 2.2881,
+      "step": 1475
+    },
+    {
+      "epoch": 0.6769089658335244,
+      "grad_norm": 0.7336349821332186,
+      "learning_rate": 0.0002495711257510517,
+      "loss": 2.3287,
+      "step": 1476
+    },
+    {
+      "epoch": 0.6773675762439807,
+      "grad_norm": 0.7232053890158836,
+      "learning_rate": 0.0002489282753944575,
+      "loss": 2.2885,
+      "step": 1477
+    },
+    {
+      "epoch": 0.6778261866544371,
+      "grad_norm": 0.7444495787199806,
+      "learning_rate": 0.00024828597951993093,
+      "loss": 2.2934,
+      "step": 1478
+    },
+    {
+      "epoch": 0.6782847970648934,
+      "grad_norm": 0.6889070675506592,
+      "learning_rate": 0.00024764423954595706,
+      "loss": 2.3256,
+      "step": 1479
+    },
+    {
+      "epoch": 0.6787434074753497,
+      "grad_norm": 0.7039224153721773,
+      "learning_rate": 0.0002470030568897938,
+      "loss": 2.2675,
+      "step": 1480
+    },
+    {
+      "epoch": 0.679202017885806,
+      "grad_norm": 0.7080436630873566,
+      "learning_rate": 0.00024636243296746773,
+      "loss": 2.3652,
+      "step": 1481
+    },
+    {
+      "epoch": 0.6796606282962623,
+      "grad_norm": 0.7238665806363301,
+      "learning_rate": 0.0002457223691937716,
+      "loss": 2.2766,
+      "step": 1482
+    },
+    {
+      "epoch": 0.6801192387067186,
+      "grad_norm": 0.7009075987606646,
+      "learning_rate": 0.0002450828669822613,
+      "loss": 2.3046,
+      "step": 1483
+    },
+    {
+      "epoch": 0.680577849117175,
+      "grad_norm": 0.7574255145848914,
+      "learning_rate": 0.00024444392774525253,
+      "loss": 2.3295,
+      "step": 1484
+    },
+    {
+      "epoch": 0.6810364595276313,
+      "grad_norm": 0.7097739347819487,
+      "learning_rate": 0.00024380555289381733,
+      "loss": 2.3471,
+      "step": 1485
+    },
+    {
+      "epoch": 0.6814950699380876,
+      "grad_norm": 0.7036362316007437,
+      "learning_rate": 0.00024316774383778184,
+      "loss": 2.2637,
+      "step": 1486
+    },
+    {
+      "epoch": 0.6819536803485439,
+      "grad_norm": 0.6960685493740015,
+      "learning_rate": 0.0002425305019857222,
+      "loss": 2.3077,
+      "step": 1487
+    },
+    {
+      "epoch": 0.6824122907590002,
+      "grad_norm": 0.7356626683582227,
+      "learning_rate": 0.00024189382874496184,
+      "loss": 2.3377,
+      "step": 1488
+    },
+    {
+      "epoch": 0.6828709011694566,
+      "grad_norm": 0.721096067988944,
+      "learning_rate": 0.00024125772552156916,
+      "loss": 2.3366,
+      "step": 1489
+    },
+    {
+      "epoch": 0.6833295115799128,
+      "grad_norm": 0.736601236231682,
+      "learning_rate": 0.00024062219372035292,
+      "loss": 2.3939,
+      "step": 1490
+    },
+    {
+      "epoch": 0.6837881219903692,
+      "grad_norm": 0.7059843536075242,
+      "learning_rate": 0.00023998723474486007,
+      "loss": 2.3375,
+      "step": 1491
+    },
+    {
+      "epoch": 0.6842467324008255,
+      "grad_norm": 0.7372174145823464,
+      "learning_rate": 0.00023935284999737272,
+      "loss": 2.3883,
+      "step": 1492
+    },
+    {
+      "epoch": 0.6847053428112818,
+      "grad_norm": 0.7753585081661647,
+      "learning_rate": 0.00023871904087890505,
+      "loss": 2.3555,
+      "step": 1493
+    },
+    {
+      "epoch": 0.6851639532217382,
+      "grad_norm": 0.722086032903188,
+      "learning_rate": 0.00023808580878919945,
+      "loss": 2.364,
+      "step": 1494
+    },
+    {
+      "epoch": 0.6856225636321944,
+      "grad_norm": 0.727085977441199,
+      "learning_rate": 0.00023745315512672398,
+      "loss": 2.2728,
+      "step": 1495
+    },
+    {
+      "epoch": 0.6860811740426508,
+      "grad_norm": 0.7243495668826077,
+      "learning_rate": 0.0002368210812886698,
+      "loss": 2.306,
+      "step": 1496
+    },
+    {
+      "epoch": 0.686539784453107,
+      "grad_norm": 0.7030806057213503,
+      "learning_rate": 0.0002361895886709471,
+      "loss": 2.3394,
+      "step": 1497
+    },
+    {
+      "epoch": 0.6869983948635634,
+      "grad_norm": 0.7290508900434413,
+      "learning_rate": 0.0002355586786681823,
+      "loss": 2.2922,
+      "step": 1498
+    },
+    {
+      "epoch": 0.6874570052740198,
+      "grad_norm": 0.7088487613399788,
+      "learning_rate": 0.00023492835267371575,
+      "loss": 2.2913,
+      "step": 1499
+    },
+    {
+      "epoch": 0.687915615684476,
+      "grad_norm": 0.6949168943992716,
+      "learning_rate": 0.0002342986120795978,
+      "loss": 2.3601,
+      "step": 1500
+    },
+    {
+      "epoch": 0.6883742260949324,
+      "grad_norm": 0.6940569444115177,
+      "learning_rate": 0.0002336694582765857,
+      "loss": 2.3386,
+      "step": 1501
+    },
+    {
+      "epoch": 0.6888328365053886,
+      "grad_norm": 0.6977390667463202,
+      "learning_rate": 0.00023304089265414085,
+      "loss": 2.3833,
+      "step": 1502
+    },
+    {
+      "epoch": 0.689291446915845,
+      "grad_norm": 0.6930366663812657,
+      "learning_rate": 0.00023241291660042613,
+      "loss": 2.346,
+      "step": 1503
+    },
+    {
+      "epoch": 0.6897500573263013,
+      "grad_norm": 0.7275246224714506,
+      "learning_rate": 0.0002317855315023018,
+      "loss": 2.3843,
+      "step": 1504
+    },
+    {
+      "epoch": 0.6902086677367576,
+      "grad_norm": 0.7224332310669006,
+      "learning_rate": 0.00023115873874532324,
+      "loss": 2.3303,
+      "step": 1505
+    },
+    {
+      "epoch": 0.690667278147214,
+      "grad_norm": 0.722584009063506,
+      "learning_rate": 0.00023053253971373796,
+      "loss": 2.3968,
+      "step": 1506
+    },
+    {
+      "epoch": 0.6911258885576702,
+      "grad_norm": 0.7361094129971699,
+      "learning_rate": 0.00022990693579048166,
+      "loss": 2.2554,
+      "step": 1507
+    },
+    {
+      "epoch": 0.6915844989681266,
+      "grad_norm": 0.7369196791286953,
+      "learning_rate": 0.00022928192835717644,
+      "loss": 2.3502,
+      "step": 1508
+    },
+    {
+      "epoch": 0.6920431093785829,
+      "grad_norm": 0.7550657746978839,
+      "learning_rate": 0.00022865751879412634,
+      "loss": 2.3109,
+      "step": 1509
+    },
+    {
+      "epoch": 0.6925017197890392,
+      "grad_norm": 0.7831952014068041,
+      "learning_rate": 0.00022803370848031585,
+      "loss": 2.4108,
+      "step": 1510
+    },
+    {
+      "epoch": 0.6929603301994955,
+      "grad_norm": 0.7053350505300641,
+      "learning_rate": 0.00022741049879340542,
+      "loss": 2.2994,
+      "step": 1511
+    },
+    {
+      "epoch": 0.6934189406099518,
+      "grad_norm": 0.7072257015259377,
+      "learning_rate": 0.00022678789110972897,
+      "loss": 2.2873,
+      "step": 1512
+    },
+    {
+      "epoch": 0.6938775510204082,
+      "grad_norm": 0.7432478315587612,
+      "learning_rate": 0.00022616588680429155,
+      "loss": 2.328,
+      "step": 1513
+    },
+    {
+      "epoch": 0.6943361614308645,
+      "grad_norm": 0.7539143084354185,
+      "learning_rate": 0.00022554448725076526,
+      "loss": 2.3212,
+      "step": 1514
+    },
+    {
+      "epoch": 0.6947947718413208,
+      "grad_norm": 0.7039696977749935,
+      "learning_rate": 0.0002249236938214863,
+      "loss": 2.2977,
+      "step": 1515
+    },
+    {
+      "epoch": 0.6952533822517771,
+      "grad_norm": 0.7471646518528324,
+      "learning_rate": 0.00022430350788745296,
+      "loss": 2.3416,
+      "step": 1516
+    },
+    {
+      "epoch": 0.6957119926622334,
+      "grad_norm": 0.7437675984998676,
+      "learning_rate": 0.00022368393081832166,
+      "loss": 2.4259,
+      "step": 1517
+    },
+    {
+      "epoch": 0.6961706030726897,
+      "grad_norm": 0.6850545973382509,
+      "learning_rate": 0.00022306496398240383,
+      "loss": 2.3306,
+      "step": 1518
+    },
+    {
+      "epoch": 0.6966292134831461,
+      "grad_norm": 0.7408897392827607,
+      "learning_rate": 0.00022244660874666373,
+      "loss": 2.2942,
+      "step": 1519
+    },
+    {
+      "epoch": 0.6970878238936024,
+      "grad_norm": 0.7144927231869553,
+      "learning_rate": 0.00022182886647671452,
+      "loss": 2.2803,
+      "step": 1520
+    },
+    {
+      "epoch": 0.6975464343040587,
+      "grad_norm": 0.7517655505584642,
+      "learning_rate": 0.0002212117385368157,
+      "loss": 2.2643,
+      "step": 1521
+    },
+    {
+      "epoch": 0.698005044714515,
+      "grad_norm": 0.735473524505164,
+      "learning_rate": 0.00022059522628987038,
+      "loss": 2.2988,
+      "step": 1522
+    },
+    {
+      "epoch": 0.6984636551249713,
+      "grad_norm": 0.7223397226251516,
+      "learning_rate": 0.00021997933109742162,
+      "loss": 2.2976,
+      "step": 1523
+    },
+    {
+      "epoch": 0.6989222655354277,
+      "grad_norm": 0.711044957416005,
+      "learning_rate": 0.00021936405431964969,
+      "loss": 2.2712,
+      "step": 1524
+    },
+    {
+      "epoch": 0.6993808759458839,
+      "grad_norm": 0.7377950999975822,
+      "learning_rate": 0.00021874939731536926,
+      "loss": 2.2043,
+      "step": 1525
+    },
+    {
+      "epoch": 0.6998394863563403,
+      "grad_norm": 0.7303441836005071,
+      "learning_rate": 0.00021813536144202656,
+      "loss": 2.2942,
+      "step": 1526
+    },
+    {
+      "epoch": 0.7002980967667966,
+      "grad_norm": 0.6999406115445806,
+      "learning_rate": 0.00021752194805569553,
+      "loss": 2.2778,
+      "step": 1527
+    },
+    {
+      "epoch": 0.7007567071772529,
+      "grad_norm": 0.7223224654775442,
+      "learning_rate": 0.0002169091585110754,
+      "loss": 2.2629,
+      "step": 1528
+    },
+    {
+      "epoch": 0.7012153175877093,
+      "grad_norm": 0.7481776494266595,
+      "learning_rate": 0.00021629699416148828,
+      "loss": 2.2569,
+      "step": 1529
+    },
+    {
+      "epoch": 0.7016739279981655,
+      "grad_norm": 0.7657013795424286,
+      "learning_rate": 0.000215685456358875,
+      "loss": 2.4281,
+      "step": 1530
+    },
+    {
+      "epoch": 0.7021325384086219,
+      "grad_norm": 0.7299421482170702,
+      "learning_rate": 0.00021507454645379258,
+      "loss": 2.3961,
+      "step": 1531
+    },
+    {
+      "epoch": 0.7025911488190782,
+      "grad_norm": 0.7327411377778809,
+      "learning_rate": 0.00021446426579541184,
+      "loss": 2.3662,
+      "step": 1532
+    },
+    {
+      "epoch": 0.7030497592295345,
+      "grad_norm": 0.7496493752481614,
+      "learning_rate": 0.00021385461573151387,
+      "loss": 2.296,
+      "step": 1533
+    },
+    {
+      "epoch": 0.7035083696399909,
+      "grad_norm": 0.7328852175361853,
+      "learning_rate": 0.00021324559760848677,
+      "loss": 2.3412,
+      "step": 1534
+    },
+    {
+      "epoch": 0.7039669800504471,
+      "grad_norm": 0.7148362726314901,
+      "learning_rate": 0.00021263721277132303,
+      "loss": 2.2446,
+      "step": 1535
+    },
+    {
+      "epoch": 0.7044255904609035,
+      "grad_norm": 0.7079593237353757,
+      "learning_rate": 0.0002120294625636171,
+      "loss": 2.2905,
+      "step": 1536
+    },
+    {
+      "epoch": 0.7048842008713598,
+      "grad_norm": 0.6839707182838743,
+      "learning_rate": 0.0002114223483275613,
+      "loss": 2.3351,
+      "step": 1537
+    },
+    {
+      "epoch": 0.7053428112818161,
+      "grad_norm": 0.7224740686368016,
+      "learning_rate": 0.0002108158714039435,
+      "loss": 2.3567,
+      "step": 1538
+    },
+    {
+      "epoch": 0.7058014216922724,
+      "grad_norm": 0.7278817521430339,
+      "learning_rate": 0.00021021003313214455,
+      "loss": 2.445,
+      "step": 1539
+    },
+    {
+      "epoch": 0.7062600321027287,
+      "grad_norm": 0.6841095062484636,
+      "learning_rate": 0.00020960483485013432,
+      "loss": 2.3309,
+      "step": 1540
+    },
+    {
+      "epoch": 0.7067186425131851,
+      "grad_norm": 0.7002193938171691,
+      "learning_rate": 0.0002090002778944694,
+      "loss": 2.3103,
+      "step": 1541
+    },
+    {
+      "epoch": 0.7071772529236414,
+      "grad_norm": 0.7099540792433043,
+      "learning_rate": 0.00020839636360029025,
+      "loss": 2.3094,
+      "step": 1542
+    },
+    {
+      "epoch": 0.7076358633340977,
+      "grad_norm": 0.7305590781229436,
+      "learning_rate": 0.00020779309330131818,
+      "loss": 2.3191,
+      "step": 1543
+    },
+    {
+      "epoch": 0.708094473744554,
+      "grad_norm": 0.7315867673427574,
+      "learning_rate": 0.00020719046832985184,
+      "loss": 2.2957,
+      "step": 1544
+    },
+    {
+      "epoch": 0.7085530841550103,
+      "grad_norm": 0.6790501883668474,
+      "learning_rate": 0.0002065884900167646,
+      "loss": 2.2913,
+      "step": 1545
+    },
+    {
+      "epoch": 0.7090116945654666,
+      "grad_norm": 0.7368579389705057,
+      "learning_rate": 0.0002059871596915024,
+      "loss": 2.3869,
+      "step": 1546
+    },
+    {
+      "epoch": 0.709470304975923,
+      "grad_norm": 0.7228824775667203,
+      "learning_rate": 0.0002053864786820795,
+      "loss": 2.4105,
+      "step": 1547
+    },
+    {
+      "epoch": 0.7099289153863793,
+      "grad_norm": 0.682497107303838,
+      "learning_rate": 0.00020478644831507627,
+      "loss": 2.19,
+      "step": 1548
+    },
+    {
+      "epoch": 0.7103875257968356,
+      "grad_norm": 0.7077090131481432,
+      "learning_rate": 0.00020418706991563634,
+      "loss": 2.3914,
+      "step": 1549
+    },
+    {
+      "epoch": 0.7108461362072919,
+      "grad_norm": 0.7316795927359594,
+      "learning_rate": 0.00020358834480746363,
+      "loss": 2.3446,
+      "step": 1550
+    },
+    {
+      "epoch": 0.7113047466177482,
+      "grad_norm": 0.7371253180292694,
+      "learning_rate": 0.0002029902743128188,
+      "loss": 2.3372,
+      "step": 1551
+    },
+    {
+      "epoch": 0.7117633570282046,
+      "grad_norm": 0.7425668283900206,
+      "learning_rate": 0.0002023928597525174,
+      "loss": 2.3718,
+      "step": 1552
+    },
+    {
+      "epoch": 0.7122219674386608,
+      "grad_norm": 0.6835821091989103,
+      "learning_rate": 0.00020179610244592595,
+      "loss": 2.3099,
+      "step": 1553
+    },
+    {
+      "epoch": 0.7126805778491172,
+      "grad_norm": 0.703997546489184,
+      "learning_rate": 0.00020120000371095937,
+      "loss": 2.3113,
+      "step": 1554
+    },
+    {
+      "epoch": 0.7131391882595735,
+      "grad_norm": 0.7211605948330991,
+      "learning_rate": 0.0002006045648640787,
+      "loss": 2.3022,
+      "step": 1555
+    },
+    {
+      "epoch": 0.7135977986700298,
+      "grad_norm": 0.7013931068187292,
+      "learning_rate": 0.00020000978722028713,
+      "loss": 2.3175,
+      "step": 1556
+    },
+    {
+      "epoch": 0.7140564090804862,
+      "grad_norm": 0.7245242487430905,
+      "learning_rate": 0.00019941567209312767,
+      "loss": 2.4459,
+      "step": 1557
+    },
+    {
+      "epoch": 0.7145150194909424,
+      "grad_norm": 0.7176493400386776,
+      "learning_rate": 0.00019882222079468036,
+      "loss": 2.3275,
+      "step": 1558
+    },
+    {
+      "epoch": 0.7149736299013988,
+      "grad_norm": 0.699947730639048,
+      "learning_rate": 0.0001982294346355595,
+      "loss": 2.2544,
+      "step": 1559
+    },
+    {
+      "epoch": 0.715432240311855,
+      "grad_norm": 0.7003495925589657,
+      "learning_rate": 0.00019763731492490976,
+      "loss": 2.3234,
+      "step": 1560
+    },
+    {
+      "epoch": 0.7158908507223114,
+      "grad_norm": 0.7114622747202772,
+      "learning_rate": 0.00019704586297040422,
+      "loss": 2.2953,
+      "step": 1561
+    },
+    {
+      "epoch": 0.7163494611327678,
+      "grad_norm": 0.6999787441038189,
+      "learning_rate": 0.0001964550800782417,
+      "loss": 2.3052,
+      "step": 1562
+    },
+    {
+      "epoch": 0.716808071543224,
+      "grad_norm": 0.699150092234492,
+      "learning_rate": 0.00019586496755314288,
+      "loss": 2.2595,
+      "step": 1563
+    },
+    {
+      "epoch": 0.7172666819536804,
+      "grad_norm": 0.7162355719003398,
+      "learning_rate": 0.00019527552669834798,
+      "loss": 2.2339,
+      "step": 1564
+    },
+    {
+      "epoch": 0.7177252923641366,
+      "grad_norm": 0.7260817967976917,
+      "learning_rate": 0.0001946867588156142,
+      "loss": 2.3177,
+      "step": 1565
+    },
+    {
+      "epoch": 0.718183902774593,
+      "grad_norm": 0.7771211972670156,
+      "learning_rate": 0.00019409866520521258,
+      "loss": 2.2731,
+      "step": 1566
+    },
+    {
+      "epoch": 0.7186425131850493,
+      "grad_norm": 0.7592890696754723,
+      "learning_rate": 0.00019351124716592455,
+      "loss": 2.3152,
+      "step": 1567
+    },
+    {
+      "epoch": 0.7191011235955056,
+      "grad_norm": 0.7799968286631157,
+      "learning_rate": 0.0001929245059950397,
+      "loss": 2.204,
+      "step": 1568
+    },
+    {
+      "epoch": 0.719559734005962,
+      "grad_norm": 0.7645278837687474,
+      "learning_rate": 0.0001923384429883533,
+      "loss": 2.4082,
+      "step": 1569
+    },
+    {
+      "epoch": 0.7200183444164182,
+      "grad_norm": 0.7430909082963685,
+      "learning_rate": 0.00019175305944016237,
+      "loss": 2.2449,
+      "step": 1570
+    },
+    {
+      "epoch": 0.7204769548268746,
+      "grad_norm": 0.7204585225246969,
+      "learning_rate": 0.00019116835664326326,
+      "loss": 2.3406,
+      "step": 1571
+    },
+    {
+      "epoch": 0.7209355652373309,
+      "grad_norm": 0.7455366913030028,
+      "learning_rate": 0.0001905843358889497,
+      "loss": 2.3347,
+      "step": 1572
+    },
+    {
+      "epoch": 0.7213941756477872,
+      "grad_norm": 0.6865097125666818,
+      "learning_rate": 0.00019000099846700836,
+      "loss": 2.3213,
+      "step": 1573
+    },
+    {
+      "epoch": 0.7218527860582435,
+      "grad_norm": 0.7213389954125653,
+      "learning_rate": 0.00018941834566571692,
+      "loss": 2.3185,
+      "step": 1574
+    },
+    {
+      "epoch": 0.7223113964686998,
+      "grad_norm": 0.7607994091353033,
+      "learning_rate": 0.00018883637877184145,
+      "loss": 2.2437,
+      "step": 1575
+    },
+    {
+      "epoch": 0.7227700068791562,
+      "grad_norm": 0.6921425072708365,
+      "learning_rate": 0.00018825509907063325,
+      "loss": 2.3418,
+      "step": 1576
+    },
+    {
+      "epoch": 0.7232286172896125,
+      "grad_norm": 0.7352560764670343,
+      "learning_rate": 0.00018767450784582557,
+      "loss": 2.308,
+      "step": 1577
+    },
+    {
+      "epoch": 0.7236872277000688,
+      "grad_norm": 0.7173298755839285,
+      "learning_rate": 0.00018709460637963122,
+      "loss": 2.3106,
+      "step": 1578
+    },
+    {
+      "epoch": 0.7241458381105251,
+      "grad_norm": 0.7745862181456696,
+      "learning_rate": 0.00018651539595274013,
+      "loss": 2.2785,
+      "step": 1579
+    },
+    {
+      "epoch": 0.7246044485209814,
+      "grad_norm": 0.7479908346136227,
+      "learning_rate": 0.00018593687784431578,
+      "loss": 2.2426,
+      "step": 1580
+    },
+    {
+      "epoch": 0.7250630589314377,
+      "grad_norm": 0.7193859082208186,
+      "learning_rate": 0.00018535905333199248,
+      "loss": 2.2442,
+      "step": 1581
+    },
+    {
+      "epoch": 0.7255216693418941,
+      "grad_norm": 0.6965124779494938,
+      "learning_rate": 0.0001847819236918733,
+      "loss": 2.3011,
+      "step": 1582
+    },
+    {
+      "epoch": 0.7259802797523504,
+      "grad_norm": 0.7372843136404025,
+      "learning_rate": 0.00018420549019852655,
+      "loss": 2.3762,
+      "step": 1583
+    },
+    {
+      "epoch": 0.7264388901628067,
+      "grad_norm": 0.7579684014955799,
+      "learning_rate": 0.00018362975412498266,
+      "loss": 2.4084,
+      "step": 1584
+    },
+    {
+      "epoch": 0.726897500573263,
+      "grad_norm": 0.7502101985712183,
+      "learning_rate": 0.00018305471674273261,
+      "loss": 2.2572,
+      "step": 1585
+    },
+    {
+      "epoch": 0.7273561109837193,
+      "grad_norm": 0.7233390452324019,
+      "learning_rate": 0.0001824803793217237,
+      "loss": 2.2504,
+      "step": 1586
+    },
+    {
+      "epoch": 0.7278147213941757,
+      "grad_norm": 0.7181760566212755,
+      "learning_rate": 0.00018190674313035737,
+      "loss": 2.328,
+      "step": 1587
+    },
+    {
+      "epoch": 0.7282733318046319,
+      "grad_norm": 0.7098856333680259,
+      "learning_rate": 0.00018133380943548716,
+      "loss": 2.3413,
+      "step": 1588
+    },
+    {
+      "epoch": 0.7287319422150883,
+      "grad_norm": 0.7246647451978225,
+      "learning_rate": 0.00018076157950241452,
+      "loss": 2.2241,
+      "step": 1589
+    },
+    {
+      "epoch": 0.7291905526255446,
+      "grad_norm": 0.7213975920189527,
+      "learning_rate": 0.00018019005459488652,
+      "loss": 2.3447,
+      "step": 1590
+    },
+    {
+      "epoch": 0.7296491630360009,
+      "grad_norm": 0.7918179451688911,
+      "learning_rate": 0.00017961923597509388,
+      "loss": 2.408,
+      "step": 1591
+    },
+    {
+      "epoch": 0.7301077734464573,
+      "grad_norm": 0.7425832479545157,
+      "learning_rate": 0.00017904912490366722,
+      "loss": 2.2842,
+      "step": 1592
+    },
+    {
+      "epoch": 0.7305663838569135,
+      "grad_norm": 0.708359746651067,
+      "learning_rate": 0.00017847972263967433,
+      "loss": 2.3526,
+      "step": 1593
+    },
+    {
+      "epoch": 0.7310249942673699,
+      "grad_norm": 0.7687611616095341,
+      "learning_rate": 0.0001779110304406177,
+      "loss": 2.2793,
+      "step": 1594
+    },
+    {
+      "epoch": 0.7314836046778261,
+      "grad_norm": 0.7379486662015635,
+      "learning_rate": 0.000177343049562432,
+      "loss": 2.2194,
+      "step": 1595
+    },
+    {
+      "epoch": 0.7319422150882825,
+      "grad_norm": 0.6991004364248443,
+      "learning_rate": 0.0001767757812594807,
+      "loss": 2.3304,
+      "step": 1596
+    },
+    {
+      "epoch": 0.7324008254987389,
+      "grad_norm": 0.7568612684348488,
+      "learning_rate": 0.0001762092267845534,
+      "loss": 2.2738,
+      "step": 1597
+    },
+    {
+      "epoch": 0.7328594359091951,
+      "grad_norm": 0.754672463538874,
+      "learning_rate": 0.00017564338738886365,
+      "loss": 2.2972,
+      "step": 1598
+    },
+    {
+      "epoch": 0.7333180463196515,
+      "grad_norm": 0.7432197885213252,
+      "learning_rate": 0.0001750782643220457,
+      "loss": 2.3685,
+      "step": 1599
+    },
+    {
+      "epoch": 0.7337766567301077,
+      "grad_norm": 0.7735271934303021,
+      "learning_rate": 0.00017451385883215166,
+      "loss": 2.4635,
+      "step": 1600
+    },
+    {
+      "epoch": 0.7342352671405641,
+      "grad_norm": 0.7409854565663724,
+      "learning_rate": 0.00017395017216564863,
+      "loss": 2.2286,
+      "step": 1601
+    },
+    {
+      "epoch": 0.7346938775510204,
+      "grad_norm": 0.721938950291884,
+      "learning_rate": 0.00017338720556741687,
+      "loss": 2.3376,
+      "step": 1602
+    },
+    {
+      "epoch": 0.7351524879614767,
+      "grad_norm": 0.7306793611776132,
+      "learning_rate": 0.00017282496028074606,
+      "loss": 2.3816,
+      "step": 1603
+    },
+    {
+      "epoch": 0.7356110983719331,
+      "grad_norm": 0.7021692573673161,
+      "learning_rate": 0.00017226343754733254,
+      "loss": 2.29,
+      "step": 1604
+    },
+    {
+      "epoch": 0.7360697087823893,
+      "grad_norm": 0.7519745043402944,
+      "learning_rate": 0.00017170263860727769,
+      "loss": 2.2634,
+      "step": 1605
+    },
+    {
+      "epoch": 0.7365283191928457,
+      "grad_norm": 0.7368345368592865,
+      "learning_rate": 0.0001711425646990838,
+      "loss": 2.3137,
+      "step": 1606
+    },
+    {
+      "epoch": 0.736986929603302,
+      "grad_norm": 0.7593107025677773,
+      "learning_rate": 0.00017058321705965202,
+      "loss": 2.3291,
+      "step": 1607
+    },
+    {
+      "epoch": 0.7374455400137583,
+      "grad_norm": 0.7037250751297562,
+      "learning_rate": 0.0001700245969242798,
+      "loss": 2.2933,
+      "step": 1608
+    },
+    {
+      "epoch": 0.7379041504242146,
+      "grad_norm": 0.7279183942577285,
+      "learning_rate": 0.00016946670552665804,
+      "loss": 2.323,
+      "step": 1609
+    },
+    {
+      "epoch": 0.7383627608346709,
+      "grad_norm": 0.6871747393781154,
+      "learning_rate": 0.00016890954409886795,
+      "loss": 2.3542,
+      "step": 1610
+    },
+    {
+      "epoch": 0.7388213712451273,
+      "grad_norm": 0.7211469499843058,
+      "learning_rate": 0.00016835311387137836,
+      "loss": 2.3455,
+      "step": 1611
+    },
+    {
+      "epoch": 0.7392799816555836,
+      "grad_norm": 0.7444803714348671,
+      "learning_rate": 0.0001677974160730441,
+      "loss": 2.2786,
+      "step": 1612
+    },
+    {
+      "epoch": 0.7397385920660399,
+      "grad_norm": 0.7040494209723243,
+      "learning_rate": 0.00016724245193110176,
+      "loss": 2.3354,
+      "step": 1613
+    },
+    {
+      "epoch": 0.7401972024764962,
+      "grad_norm": 0.7253301398301311,
+      "learning_rate": 0.00016668822267116784,
+      "loss": 2.3338,
+      "step": 1614
+    },
+    {
+      "epoch": 0.7406558128869525,
+      "grad_norm": 0.6775759777590966,
+      "learning_rate": 0.00016613472951723597,
+      "loss": 2.2776,
+      "step": 1615
+    },
+    {
+      "epoch": 0.7411144232974088,
+      "grad_norm": 0.739888015083308,
+      "learning_rate": 0.00016558197369167434,
+      "loss": 2.2884,
+      "step": 1616
+    },
+    {
+      "epoch": 0.7415730337078652,
+      "grad_norm": 0.7520757327505853,
+      "learning_rate": 0.00016502995641522216,
+      "loss": 2.3471,
+      "step": 1617
+    },
+    {
+      "epoch": 0.7420316441183215,
+      "grad_norm": 0.7204276516466549,
+      "learning_rate": 0.00016447867890698843,
+      "loss": 2.3526,
+      "step": 1618
+    },
+    {
+      "epoch": 0.7424902545287778,
+      "grad_norm": 0.7316255249971231,
+      "learning_rate": 0.00016392814238444753,
+      "loss": 2.3337,
+      "step": 1619
+    },
+    {
+      "epoch": 0.7429488649392341,
+      "grad_norm": 0.744416569663432,
+      "learning_rate": 0.00016337834806343782,
+      "loss": 2.2912,
+      "step": 1620
+    },
+    {
+      "epoch": 0.7434074753496904,
+      "grad_norm": 0.7385458921313917,
+      "learning_rate": 0.0001628292971581588,
+      "loss": 2.2344,
+      "step": 1621
+    },
+    {
+      "epoch": 0.7438660857601468,
+      "grad_norm": 0.7392695657417512,
+      "learning_rate": 0.00016228099088116772,
+      "loss": 2.3756,
+      "step": 1622
+    },
+    {
+      "epoch": 0.744324696170603,
+      "grad_norm": 0.7423684435267772,
+      "learning_rate": 0.00016173343044337734,
+      "loss": 2.3874,
+      "step": 1623
+    },
+    {
+      "epoch": 0.7447833065810594,
+      "grad_norm": 0.7734669741986436,
+      "learning_rate": 0.00016118661705405356,
+      "loss": 2.193,
+      "step": 1624
+    },
+    {
+      "epoch": 0.7452419169915158,
+      "grad_norm": 0.7189517036990479,
+      "learning_rate": 0.00016064055192081255,
+      "loss": 2.3509,
+      "step": 1625
+    },
+    {
+      "epoch": 0.745700527401972,
+      "grad_norm": 0.6819578960264154,
+      "learning_rate": 0.00016009523624961757,
+      "loss": 2.2632,
+      "step": 1626
+    },
+    {
+      "epoch": 0.7461591378124284,
+      "grad_norm": 0.692689091656934,
+      "learning_rate": 0.00015955067124477678,
+      "loss": 2.1779,
+      "step": 1627
+    },
+    {
+      "epoch": 0.7466177482228846,
+      "grad_norm": 0.6846336917296338,
+      "learning_rate": 0.000159006858108941,
+      "loss": 2.298,
+      "step": 1628
+    },
+    {
+      "epoch": 0.747076358633341,
+      "grad_norm": 0.7589427060313152,
+      "learning_rate": 0.00015846379804310002,
+      "loss": 2.2934,
+      "step": 1629
+    },
+    {
+      "epoch": 0.7475349690437973,
+      "grad_norm": 0.7355400186470221,
+      "learning_rate": 0.00015792149224658054,
+      "loss": 2.3087,
+      "step": 1630
+    },
+    {
+      "epoch": 0.7479935794542536,
+      "grad_norm": 0.7250276918726055,
+      "learning_rate": 0.00015737994191704385,
+      "loss": 2.3717,
+      "step": 1631
+    },
+    {
+      "epoch": 0.74845218986471,
+      "grad_norm": 0.729736531965279,
+      "learning_rate": 0.0001568391482504829,
+      "loss": 2.2539,
+      "step": 1632
+    },
+    {
+      "epoch": 0.7489108002751662,
+      "grad_norm": 1.0046952224750938,
+      "learning_rate": 0.00015629911244121903,
+      "loss": 2.2414,
+      "step": 1633
+    },
+    {
+      "epoch": 0.7493694106856226,
+      "grad_norm": 0.7153080991575871,
+      "learning_rate": 0.0001557598356819,
+      "loss": 2.4193,
+      "step": 1634
+    },
+    {
+      "epoch": 0.7498280210960789,
+      "grad_norm": 0.7227733826029701,
+      "learning_rate": 0.00015522131916349786,
+      "loss": 2.2983,
+      "step": 1635
+    },
+    {
+      "epoch": 0.7502866315065352,
+      "grad_norm": 0.7857899204202784,
+      "learning_rate": 0.00015468356407530493,
+      "loss": 2.3463,
+      "step": 1636
+    },
+    {
+      "epoch": 0.7507452419169915,
+      "grad_norm": 0.7645441269764669,
+      "learning_rate": 0.00015414657160493217,
+      "loss": 2.3718,
+      "step": 1637
+    },
+    {
+      "epoch": 0.7512038523274478,
+      "grad_norm": 0.7212345349136542,
+      "learning_rate": 0.00015361034293830673,
+      "loss": 2.3423,
+      "step": 1638
+    },
+    {
+      "epoch": 0.7516624627379042,
+      "grad_norm": 0.7198089174160163,
+      "learning_rate": 0.00015307487925966844,
+      "loss": 2.3422,
+      "step": 1639
+    },
+    {
+      "epoch": 0.7521210731483605,
+      "grad_norm": 0.7194839545642331,
+      "learning_rate": 0.00015254018175156776,
+      "loss": 2.3868,
+      "step": 1640
+    },
+    {
+      "epoch": 0.7525796835588168,
+      "grad_norm": 0.7503993586105082,
+      "learning_rate": 0.0001520062515948632,
+      "loss": 2.3727,
+      "step": 1641
+    },
+    {
+      "epoch": 0.7530382939692731,
+      "grad_norm": 0.7198946192835345,
+      "learning_rate": 0.0001514730899687189,
+      "loss": 2.1921,
+      "step": 1642
+    },
+    {
+      "epoch": 0.7534969043797294,
+      "grad_norm": 0.7241987072171926,
+      "learning_rate": 0.00015094069805060122,
+      "loss": 2.3695,
+      "step": 1643
+    },
+    {
+      "epoch": 0.7539555147901857,
+      "grad_norm": 0.6853092511468228,
+      "learning_rate": 0.00015040907701627666,
+      "loss": 2.297,
+      "step": 1644
+    },
+    {
+      "epoch": 0.7544141252006421,
+      "grad_norm": 0.7445492758149922,
+      "learning_rate": 0.00014987822803980976,
+      "loss": 2.2142,
+      "step": 1645
+    },
+    {
+      "epoch": 0.7548727356110984,
+      "grad_norm": 0.6948230104040495,
+      "learning_rate": 0.00014934815229355965,
+      "loss": 2.335,
+      "step": 1646
+    },
+    {
+      "epoch": 0.7553313460215547,
+      "grad_norm": 0.7851287117634383,
+      "learning_rate": 0.00014881885094817748,
+      "loss": 2.3652,
+      "step": 1647
+    },
+    {
+      "epoch": 0.755789956432011,
+      "grad_norm": 0.7406563465919247,
+      "learning_rate": 0.00014829032517260488,
+      "loss": 2.3865,
+      "step": 1648
+    },
+    {
+      "epoch": 0.7562485668424673,
+      "grad_norm": 0.6982021105625565,
+      "learning_rate": 0.0001477625761340704,
+      "loss": 2.3479,
+      "step": 1649
+    },
+    {
+      "epoch": 0.7567071772529237,
+      "grad_norm": 0.6784220239231997,
+      "learning_rate": 0.0001472356049980868,
+      "loss": 2.3553,
+      "step": 1650
+    },
+    {
+      "epoch": 0.7571657876633799,
+      "grad_norm": 0.7525487191005547,
+      "learning_rate": 0.00014670941292844954,
+      "loss": 2.2486,
+      "step": 1651
+    },
+    {
+      "epoch": 0.7576243980738363,
+      "grad_norm": 0.7139290195449878,
+      "learning_rate": 0.00014618400108723295,
+      "loss": 2.2361,
+      "step": 1652
+    },
+    {
+      "epoch": 0.7580830084842926,
+      "grad_norm": 0.7324535880773018,
+      "learning_rate": 0.00014565937063478862,
+      "loss": 2.2486,
+      "step": 1653
+    },
+    {
+      "epoch": 0.7585416188947489,
+      "grad_norm": 0.7373327309915209,
+      "learning_rate": 0.00014513552272974207,
+      "loss": 2.3153,
+      "step": 1654
+    },
+    {
+      "epoch": 0.7590002293052053,
+      "grad_norm": 0.7151814496740836,
+      "learning_rate": 0.0001446124585289913,
+      "loss": 2.2574,
+      "step": 1655
+    },
+    {
+      "epoch": 0.7594588397156615,
+      "grad_norm": 0.7001689627193768,
+      "learning_rate": 0.00014409017918770266,
+      "loss": 2.2796,
+      "step": 1656
+    },
+    {
+      "epoch": 0.7599174501261179,
+      "grad_norm": 0.7092106264963857,
+      "learning_rate": 0.00014356868585930994,
+      "loss": 2.3677,
+      "step": 1657
+    },
+    {
+      "epoch": 0.7603760605365741,
+      "grad_norm": 0.6857961314201686,
+      "learning_rate": 0.00014304797969551077,
+      "loss": 2.2122,
+      "step": 1658
+    },
+    {
+      "epoch": 0.7608346709470305,
+      "grad_norm": 0.6886837034722922,
+      "learning_rate": 0.00014252806184626417,
+      "loss": 2.4916,
+      "step": 1659
+    },
+    {
+      "epoch": 0.7612932813574869,
+      "grad_norm": 0.7640953568552398,
+      "learning_rate": 0.00014200893345978817,
+      "loss": 2.2938,
+      "step": 1660
+    },
+    {
+      "epoch": 0.7617518917679431,
+      "grad_norm": 0.7047996379668184,
+      "learning_rate": 0.00014149059568255778,
+      "loss": 2.312,
+      "step": 1661
+    },
+    {
+      "epoch": 0.7622105021783995,
+      "grad_norm": 0.7330969902177513,
+      "learning_rate": 0.00014097304965930157,
+      "loss": 2.2448,
+      "step": 1662
+    },
+    {
+      "epoch": 0.7626691125888557,
+      "grad_norm": 0.6841063198179199,
+      "learning_rate": 0.00014045629653299953,
+      "loss": 2.3347,
+      "step": 1663
+    },
+    {
+      "epoch": 0.7631277229993121,
+      "grad_norm": 0.705886590182099,
+      "learning_rate": 0.00013994033744488076,
+      "loss": 2.3119,
+      "step": 1664
+    },
+    {
+      "epoch": 0.7635863334097684,
+      "grad_norm": 0.681997777486257,
+      "learning_rate": 0.00013942517353442092,
+      "loss": 2.2548,
+      "step": 1665
+    },
+    {
+      "epoch": 0.7640449438202247,
+      "grad_norm": 0.7158327482139912,
+      "learning_rate": 0.0001389108059393391,
+      "loss": 2.3402,
+      "step": 1666
+    },
+    {
+      "epoch": 0.7645035542306811,
+      "grad_norm": 0.7257340224827228,
+      "learning_rate": 0.00013839723579559581,
+      "loss": 2.3782,
+      "step": 1667
+    },
+    {
+      "epoch": 0.7649621646411373,
+      "grad_norm": 0.71439529116138,
+      "learning_rate": 0.00013788446423739103,
+      "loss": 2.2894,
+      "step": 1668
+    },
+    {
+      "epoch": 0.7654207750515937,
+      "grad_norm": 0.6869862925139286,
+      "learning_rate": 0.00013737249239716042,
+      "loss": 2.3059,
+      "step": 1669
+    },
+    {
+      "epoch": 0.76587938546205,
+      "grad_norm": 0.730371342862147,
+      "learning_rate": 0.00013686132140557355,
+      "loss": 2.2428,
+      "step": 1670
+    },
+    {
+      "epoch": 0.7663379958725063,
+      "grad_norm": 0.7116456411662503,
+      "learning_rate": 0.00013635095239153188,
+      "loss": 2.3229,
+      "step": 1671
+    },
+    {
+      "epoch": 0.7667966062829626,
+      "grad_norm": 0.757872754328761,
+      "learning_rate": 0.00013584138648216527,
+      "loss": 2.258,
+      "step": 1672
+    },
+    {
+      "epoch": 0.7672552166934189,
+      "grad_norm": 0.7133865479651404,
+      "learning_rate": 0.0001353326248028298,
+      "loss": 2.3193,
+      "step": 1673
+    },
+    {
+      "epoch": 0.7677138271038753,
+      "grad_norm": 0.7742187497909928,
+      "learning_rate": 0.00013482466847710594,
+      "loss": 2.3255,
+      "step": 1674
+    },
+    {
+      "epoch": 0.7681724375143316,
+      "grad_norm": 0.7043260380282257,
+      "learning_rate": 0.00013431751862679554,
+      "loss": 2.3144,
+      "step": 1675
+    },
+    {
+      "epoch": 0.7686310479247879,
+      "grad_norm": 0.6975593471663218,
+      "learning_rate": 0.00013381117637191887,
+      "loss": 2.2682,
+      "step": 1676
+    },
+    {
+      "epoch": 0.7690896583352442,
+      "grad_norm": 0.7205302659032837,
+      "learning_rate": 0.00013330564283071293,
+      "loss": 2.238,
+      "step": 1677
+    },
+    {
+      "epoch": 0.7695482687457005,
+      "grad_norm": 0.7527567171014405,
+      "learning_rate": 0.000132800919119629,
+      "loss": 2.2755,
+      "step": 1678
+    },
+    {
+      "epoch": 0.7700068791561568,
+      "grad_norm": 0.6983126679566811,
+      "learning_rate": 0.00013229700635332948,
+      "loss": 2.2407,
+      "step": 1679
+    },
+    {
+      "epoch": 0.7704654895666132,
+      "grad_norm": 0.7143848605686671,
+      "learning_rate": 0.00013179390564468585,
+      "loss": 2.2768,
+      "step": 1680
+    },
+    {
+      "epoch": 0.7709240999770695,
+      "grad_norm": 0.7643863357342355,
+      "learning_rate": 0.00013129161810477641,
+      "loss": 2.267,
+      "step": 1681
+    },
+    {
+      "epoch": 0.7713827103875258,
+      "grad_norm": 0.7262962873695448,
+      "learning_rate": 0.0001307901448428837,
+      "loss": 2.3451,
+      "step": 1682
+    },
+    {
+      "epoch": 0.7718413207979821,
+      "grad_norm": 0.6902179129562658,
+      "learning_rate": 0.0001302894869664916,
+      "loss": 2.2912,
+      "step": 1683
+    },
+    {
+      "epoch": 0.7722999312084384,
+      "grad_norm": 0.731115256334471,
+      "learning_rate": 0.00012978964558128336,
+      "loss": 2.2333,
+      "step": 1684
+    },
+    {
+      "epoch": 0.7727585416188948,
+      "grad_norm": 0.7298964304536596,
+      "learning_rate": 0.00012929062179113925,
+      "loss": 2.2385,
+      "step": 1685
+    },
+    {
+      "epoch": 0.773217152029351,
+      "grad_norm": 0.6968591272994532,
+      "learning_rate": 0.00012879241669813368,
+      "loss": 2.2855,
+      "step": 1686
+    },
+    {
+      "epoch": 0.7736757624398074,
+      "grad_norm": 0.7395579455645211,
+      "learning_rate": 0.00012829503140253295,
+      "loss": 2.3934,
+      "step": 1687
+    },
+    {
+      "epoch": 0.7741343728502637,
+      "grad_norm": 0.7398689834808585,
+      "learning_rate": 0.0001277984670027933,
+      "loss": 2.2962,
+      "step": 1688
+    },
+    {
+      "epoch": 0.77459298326072,
+      "grad_norm": 0.7366343377008299,
+      "learning_rate": 0.00012730272459555737,
+      "loss": 2.3411,
+      "step": 1689
+    },
+    {
+      "epoch": 0.7750515936711764,
+      "grad_norm": 0.7673583297268873,
+      "learning_rate": 0.00012680780527565312,
+      "loss": 2.2652,
+      "step": 1690
+    },
+    {
+      "epoch": 0.7755102040816326,
+      "grad_norm": 0.7287046537854295,
+      "learning_rate": 0.0001263137101360905,
+      "loss": 2.2409,
+      "step": 1691
+    },
+    {
+      "epoch": 0.775968814492089,
+      "grad_norm": 0.7503763487977952,
+      "learning_rate": 0.00012582044026805922,
+      "loss": 2.3032,
+      "step": 1692
+    },
+    {
+      "epoch": 0.7764274249025452,
+      "grad_norm": 0.7445910527035103,
+      "learning_rate": 0.00012532799676092627,
+      "loss": 2.2503,
+      "step": 1693
+    },
+    {
+      "epoch": 0.7768860353130016,
+      "grad_norm": 0.7415129031699724,
+      "learning_rate": 0.00012483638070223414,
+      "loss": 2.3351,
+      "step": 1694
+    },
+    {
+      "epoch": 0.777344645723458,
+      "grad_norm": 0.7821822469027946,
+      "learning_rate": 0.00012434559317769752,
+      "loss": 2.2969,
+      "step": 1695
+    },
+    {
+      "epoch": 0.7778032561339142,
+      "grad_norm": 0.7465473528209117,
+      "learning_rate": 0.0001238556352712012,
+      "loss": 2.2571,
+      "step": 1696
+    },
+    {
+      "epoch": 0.7782618665443706,
+      "grad_norm": 0.7415448912131485,
+      "learning_rate": 0.00012336650806479827,
+      "loss": 2.3761,
+      "step": 1697
+    },
+    {
+      "epoch": 0.7787204769548268,
+      "grad_norm": 0.7115779448807874,
+      "learning_rate": 0.00012287821263870708,
+      "loss": 2.329,
+      "step": 1698
+    },
+    {
+      "epoch": 0.7791790873652832,
+      "grad_norm": 0.7728944359606066,
+      "learning_rate": 0.00012239075007130885,
+      "loss": 2.3786,
+      "step": 1699
+    },
+    {
+      "epoch": 0.7796376977757395,
+      "grad_norm": 0.7259331058918763,
+      "learning_rate": 0.00012190412143914536,
+      "loss": 2.2476,
+      "step": 1700
+    },
+    {
+      "epoch": 0.7800963081861958,
+      "grad_norm": 0.7221705161085847,
+      "learning_rate": 0.0001214183278169172,
+      "loss": 2.1672,
+      "step": 1701
+    },
+    {
+      "epoch": 0.7805549185966522,
+      "grad_norm": 0.6837721720801297,
+      "learning_rate": 0.00012093337027748042,
+      "loss": 2.4087,
+      "step": 1702
+    },
+    {
+      "epoch": 0.7810135290071084,
+      "grad_norm": 0.7521255308292717,
+      "learning_rate": 0.00012044924989184459,
+      "loss": 2.3044,
+      "step": 1703
+    },
+    {
+      "epoch": 0.7814721394175648,
+      "grad_norm": 0.7156657703693917,
+      "learning_rate": 0.0001199659677291709,
+      "loss": 2.2861,
+      "step": 1704
+    },
+    {
+      "epoch": 0.7819307498280211,
+      "grad_norm": 0.6844653192845467,
+      "learning_rate": 0.00011948352485676895,
+      "loss": 2.3707,
+      "step": 1705
+    },
+    {
+      "epoch": 0.7823893602384774,
+      "grad_norm": 0.7195277565849696,
+      "learning_rate": 0.00011900192234009477,
+      "loss": 2.287,
+      "step": 1706
+    },
+    {
+      "epoch": 0.7828479706489337,
+      "grad_norm": 0.6986166858814762,
+      "learning_rate": 0.00011852116124274875,
+      "loss": 2.2782,
+      "step": 1707
+    },
+    {
+      "epoch": 0.78330658105939,
+      "grad_norm": 0.6914350132310244,
+      "learning_rate": 0.00011804124262647314,
+      "loss": 2.3346,
+      "step": 1708
+    },
+    {
+      "epoch": 0.7837651914698464,
+      "grad_norm": 0.7784274375964995,
+      "learning_rate": 0.00011756216755114929,
+      "loss": 2.2632,
+      "step": 1709
+    },
+    {
+      "epoch": 0.7842238018803027,
+      "grad_norm": 0.72171323235965,
+      "learning_rate": 0.00011708393707479548,
+      "loss": 2.4223,
+      "step": 1710
+    },
+    {
+      "epoch": 0.784682412290759,
+      "grad_norm": 0.7441874590491525,
+      "learning_rate": 0.00011660655225356531,
+      "loss": 2.243,
+      "step": 1711
+    },
+    {
+      "epoch": 0.7851410227012153,
+      "grad_norm": 0.6939918477006038,
+      "learning_rate": 0.0001161300141417444,
+      "loss": 2.314,
+      "step": 1712
+    },
+    {
+      "epoch": 0.7855996331116716,
+      "grad_norm": 0.7452294117435677,
+      "learning_rate": 0.00011565432379174823,
+      "loss": 2.2796,
+      "step": 1713
+    },
+    {
+      "epoch": 0.7860582435221279,
+      "grad_norm": 0.7342990384722292,
+      "learning_rate": 0.00011517948225412056,
+      "loss": 2.2859,
+      "step": 1714
+    },
+    {
+      "epoch": 0.7865168539325843,
+      "grad_norm": 0.7472840684700649,
+      "learning_rate": 0.00011470549057753032,
+      "loss": 2.3681,
+      "step": 1715
+    },
+    {
+      "epoch": 0.7869754643430406,
+      "grad_norm": 0.7349304392296393,
+      "learning_rate": 0.00011423234980876957,
+      "loss": 2.3297,
+      "step": 1716
+    },
+    {
+      "epoch": 0.7874340747534969,
+      "grad_norm": 0.7026054365916677,
+      "learning_rate": 0.00011376006099275099,
+      "loss": 2.2785,
+      "step": 1717
+    },
+    {
+      "epoch": 0.7878926851639533,
+      "grad_norm": 0.7396870464027855,
+      "learning_rate": 0.00011328862517250609,
+      "loss": 2.3412,
+      "step": 1718
+    },
+    {
+      "epoch": 0.7883512955744095,
+      "grad_norm": 0.769436565940954,
+      "learning_rate": 0.00011281804338918239,
+      "loss": 2.3459,
+      "step": 1719
+    },
+    {
+      "epoch": 0.7888099059848659,
+      "grad_norm": 0.8443076167032652,
+      "learning_rate": 0.00011234831668204115,
+      "loss": 2.2924,
+      "step": 1720
+    },
+    {
+      "epoch": 0.7892685163953221,
+      "grad_norm": 0.7122823307805229,
+      "learning_rate": 0.00011187944608845569,
+      "loss": 2.2986,
+      "step": 1721
+    },
+    {
+      "epoch": 0.7897271268057785,
+      "grad_norm": 0.7769997330242945,
+      "learning_rate": 0.00011141143264390801,
+      "loss": 2.3424,
+      "step": 1722
+    },
+    {
+      "epoch": 0.7901857372162349,
+      "grad_norm": 0.7249065639457029,
+      "learning_rate": 0.0001109442773819877,
+      "loss": 2.3312,
+      "step": 1723
+    },
+    {
+      "epoch": 0.7906443476266911,
+      "grad_norm": 0.7346310326557223,
+      "learning_rate": 0.0001104779813343889,
+      "loss": 2.2388,
+      "step": 1724
+    },
+    {
+      "epoch": 0.7911029580371475,
+      "grad_norm": 0.6822836273101317,
+      "learning_rate": 0.00011001254553090812,
+      "loss": 2.2801,
+      "step": 1725
+    },
+    {
+      "epoch": 0.7915615684476037,
+      "grad_norm": 0.6858197897418418,
+      "learning_rate": 0.00010954797099944186,
+      "loss": 2.3197,
+      "step": 1726
+    },
+    {
+      "epoch": 0.7920201788580601,
+      "grad_norm": 0.7359876307880447,
+      "learning_rate": 0.0001090842587659851,
+      "loss": 2.2727,
+      "step": 1727
+    },
+    {
+      "epoch": 0.7924787892685164,
+      "grad_norm": 0.7756118148518656,
+      "learning_rate": 0.00010862140985462804,
+      "loss": 2.3529,
+      "step": 1728
+    },
+    {
+      "epoch": 0.7929373996789727,
+      "grad_norm": 0.7224913448020965,
+      "learning_rate": 0.00010815942528755418,
+      "loss": 2.2437,
+      "step": 1729
+    },
+    {
+      "epoch": 0.7933960100894291,
+      "grad_norm": 0.7630786212259069,
+      "learning_rate": 0.00010769830608503844,
+      "loss": 2.3413,
+      "step": 1730
+    },
+    {
+      "epoch": 0.7938546204998853,
+      "grad_norm": 0.7579758609510368,
+      "learning_rate": 0.00010723805326544473,
+      "loss": 2.3801,
+      "step": 1731
+    },
+    {
+      "epoch": 0.7943132309103417,
+      "grad_norm": 0.7297920554027788,
+      "learning_rate": 0.00010677866784522316,
+      "loss": 2.2785,
+      "step": 1732
+    },
+    {
+      "epoch": 0.794771841320798,
+      "grad_norm": 0.7278297518842008,
+      "learning_rate": 0.00010632015083890839,
+      "loss": 2.2258,
+      "step": 1733
+    },
+    {
+      "epoch": 0.7952304517312543,
+      "grad_norm": 0.7197403859136692,
+      "learning_rate": 0.00010586250325911745,
+      "loss": 2.4443,
+      "step": 1734
+    },
+    {
+      "epoch": 0.7956890621417106,
+      "grad_norm": 0.7158546757451448,
+      "learning_rate": 0.00010540572611654697,
+      "loss": 2.3417,
+      "step": 1735
+    },
+    {
+      "epoch": 0.7961476725521669,
+      "grad_norm": 0.7635891699153537,
+      "learning_rate": 0.00010494982041997126,
+      "loss": 2.2849,
+      "step": 1736
+    },
+    {
+      "epoch": 0.7966062829626233,
+      "grad_norm": 0.7591604130867374,
+      "learning_rate": 0.0001044947871762405,
+      "loss": 2.3419,
+      "step": 1737
+    },
+    {
+      "epoch": 0.7970648933730796,
+      "grad_norm": 0.7123345089639284,
+      "learning_rate": 0.00010404062739027753,
+      "loss": 2.3512,
+      "step": 1738
+    },
+    {
+      "epoch": 0.7975235037835359,
+      "grad_norm": 0.7184033737170671,
+      "learning_rate": 0.00010358734206507641,
+      "loss": 2.3993,
+      "step": 1739
+    },
+    {
+      "epoch": 0.7979821141939922,
+      "grad_norm": 0.6921271089398683,
+      "learning_rate": 0.00010313493220170017,
+      "loss": 2.2457,
+      "step": 1740
+    },
+    {
+      "epoch": 0.7984407246044485,
+      "grad_norm": 0.7385848778025248,
+      "learning_rate": 0.00010268339879927836,
+      "loss": 2.2032,
+      "step": 1741
+    },
+    {
+      "epoch": 0.7988993350149048,
+      "grad_norm": 0.6975978551433469,
+      "learning_rate": 0.00010223274285500466,
+      "loss": 2.3278,
+      "step": 1742
+    },
+    {
+      "epoch": 0.7993579454253612,
+      "grad_norm": 0.7434412477537324,
+      "learning_rate": 0.00010178296536413495,
+      "loss": 2.3501,
+      "step": 1743
+    },
+    {
+      "epoch": 0.7998165558358175,
+      "grad_norm": 0.7300347880898207,
+      "learning_rate": 0.00010133406731998546,
+      "loss": 2.2379,
+      "step": 1744
+    },
+    {
+      "epoch": 0.8002751662462738,
+      "grad_norm": 0.7350042850245959,
+      "learning_rate": 0.00010088604971392979,
+      "loss": 2.3122,
+      "step": 1745
+    },
+    {
+      "epoch": 0.8007337766567301,
+      "grad_norm": 0.7389764618477017,
+      "learning_rate": 0.0001004389135353972,
+      "loss": 2.3417,
+      "step": 1746
+    },
+    {
+      "epoch": 0.8011923870671864,
+      "grad_norm": 0.7285825472757897,
+      "learning_rate": 9.999265977187049e-05,
+      "loss": 2.3359,
+      "step": 1747
+    },
+    {
+      "epoch": 0.8016509974776428,
+      "grad_norm": 0.7081580424007522,
+      "learning_rate": 9.95472894088838e-05,
+      "loss": 2.3623,
+      "step": 1748
+    },
+    {
+      "epoch": 0.802109607888099,
+      "grad_norm": 0.7053535724091323,
+      "learning_rate": 9.910280343001993e-05,
+      "loss": 2.2501,
+      "step": 1749
+    },
+    {
+      "epoch": 0.8025682182985554,
+      "grad_norm": 0.6874149404763135,
+      "learning_rate": 9.865920281690866e-05,
+      "loss": 2.3117,
+      "step": 1750
+    },
+    {
+      "epoch": 0.8030268287090117,
+      "grad_norm": 0.7305255596632451,
+      "learning_rate": 9.821648854922482e-05,
+      "loss": 2.3075,
+      "step": 1751
+    },
+    {
+      "epoch": 0.803485439119468,
+      "grad_norm": 0.7322369638118497,
+      "learning_rate": 9.77746616046854e-05,
+      "loss": 2.311,
+      "step": 1752
+    },
+    {
+      "epoch": 0.8039440495299244,
+      "grad_norm": 0.7216315683726877,
+      "learning_rate": 9.733372295904774e-05,
+      "loss": 2.3255,
+      "step": 1753
+    },
+    {
+      "epoch": 0.8044026599403806,
+      "grad_norm": 0.7394276941701603,
+      "learning_rate": 9.68936735861079e-05,
+      "loss": 2.2892,
+      "step": 1754
+    },
+    {
+      "epoch": 0.804861270350837,
+      "grad_norm": 0.7539677185286483,
+      "learning_rate": 9.645451445769737e-05,
+      "loss": 2.4026,
+      "step": 1755
+    },
+    {
+      "epoch": 0.8053198807612932,
+      "grad_norm": 0.7004303450854833,
+      "learning_rate": 9.601624654368196e-05,
+      "loss": 2.2667,
+      "step": 1756
+    },
+    {
+      "epoch": 0.8057784911717496,
+      "grad_norm": 0.7085342737067718,
+      "learning_rate": 9.557887081195938e-05,
+      "loss": 2.2951,
+      "step": 1757
+    },
+    {
+      "epoch": 0.806237101582206,
+      "grad_norm": 0.7477506888159575,
+      "learning_rate": 9.514238822845667e-05,
+      "loss": 2.3701,
+      "step": 1758
+    },
+    {
+      "epoch": 0.8066957119926622,
+      "grad_norm": 0.7351862221183025,
+      "learning_rate": 9.470679975712837e-05,
+      "loss": 2.228,
+      "step": 1759
+    },
+    {
+      "epoch": 0.8071543224031186,
+      "grad_norm": 0.7447260634168392,
+      "learning_rate": 9.427210635995481e-05,
+      "loss": 2.3467,
+      "step": 1760
+    },
+    {
+      "epoch": 0.8076129328135748,
+      "grad_norm": 0.7188216868833555,
+      "learning_rate": 9.383830899693923e-05,
+      "loss": 2.2932,
+      "step": 1761
+    },
+    {
+      "epoch": 0.8080715432240312,
+      "grad_norm": 0.7226890465907573,
+      "learning_rate": 9.340540862610591e-05,
+      "loss": 2.2162,
+      "step": 1762
+    },
+    {
+      "epoch": 0.8085301536344875,
+      "grad_norm": 0.716383935096951,
+      "learning_rate": 9.297340620349854e-05,
+      "loss": 2.3351,
+      "step": 1763
+    },
+    {
+      "epoch": 0.8089887640449438,
+      "grad_norm": 0.7215561307705608,
+      "learning_rate": 9.25423026831777e-05,
+      "loss": 2.2834,
+      "step": 1764
+    },
+    {
+      "epoch": 0.8094473744554002,
+      "grad_norm": 0.7070641455650886,
+      "learning_rate": 9.211209901721846e-05,
+      "loss": 2.2868,
+      "step": 1765
+    },
+    {
+      "epoch": 0.8099059848658564,
+      "grad_norm": 0.7234321784169141,
+      "learning_rate": 9.168279615570863e-05,
+      "loss": 2.2096,
+      "step": 1766
+    },
+    {
+      "epoch": 0.8103645952763128,
+      "grad_norm": 0.7349123828664521,
+      "learning_rate": 9.125439504674699e-05,
+      "loss": 2.2941,
+      "step": 1767
+    },
+    {
+      "epoch": 0.8108232056867691,
+      "grad_norm": 0.6804834139207561,
+      "learning_rate": 9.082689663644057e-05,
+      "loss": 2.2349,
+      "step": 1768
+    },
+    {
+      "epoch": 0.8112818160972254,
+      "grad_norm": 0.6996426134660846,
+      "learning_rate": 9.040030186890264e-05,
+      "loss": 2.2307,
+      "step": 1769
+    },
+    {
+      "epoch": 0.8117404265076817,
+      "grad_norm": 0.6980379252668564,
+      "learning_rate": 8.997461168625138e-05,
+      "loss": 2.3364,
+      "step": 1770
+    },
+    {
+      "epoch": 0.812199036918138,
+      "grad_norm": 0.7171379626317289,
+      "learning_rate": 8.954982702860664e-05,
+      "loss": 2.3236,
+      "step": 1771
+    },
+    {
+      "epoch": 0.8126576473285944,
+      "grad_norm": 0.7258087833766341,
+      "learning_rate": 8.912594883408865e-05,
+      "loss": 2.2348,
+      "step": 1772
+    },
+    {
+      "epoch": 0.8131162577390507,
+      "grad_norm": 0.7758724933413859,
+      "learning_rate": 8.870297803881589e-05,
+      "loss": 2.3312,
+      "step": 1773
+    },
+    {
+      "epoch": 0.813574868149507,
+      "grad_norm": 0.7197382482313714,
+      "learning_rate": 8.828091557690287e-05,
+      "loss": 2.3109,
+      "step": 1774
+    },
+    {
+      "epoch": 0.8140334785599633,
+      "grad_norm": 0.7184076462027718,
+      "learning_rate": 8.785976238045801e-05,
+      "loss": 2.2589,
+      "step": 1775
+    },
+    {
+      "epoch": 0.8144920889704196,
+      "grad_norm": 0.7204952394068871,
+      "learning_rate": 8.743951937958144e-05,
+      "loss": 2.198,
+      "step": 1776
+    },
+    {
+      "epoch": 0.8149506993808759,
+      "grad_norm": 0.7254909953687794,
+      "learning_rate": 8.702018750236357e-05,
+      "loss": 2.2838,
+      "step": 1777
+    },
+    {
+      "epoch": 0.8154093097913323,
+      "grad_norm": 0.7484604560440167,
+      "learning_rate": 8.660176767488237e-05,
+      "loss": 2.3152,
+      "step": 1778
+    },
+    {
+      "epoch": 0.8158679202017886,
+      "grad_norm": 0.730658074918838,
+      "learning_rate": 8.618426082120146e-05,
+      "loss": 2.2695,
+      "step": 1779
+    },
+    {
+      "epoch": 0.8163265306122449,
+      "grad_norm": 0.7593365108797246,
+      "learning_rate": 8.576766786336854e-05,
+      "loss": 2.2512,
+      "step": 1780
+    },
+    {
+      "epoch": 0.8167851410227012,
+      "grad_norm": 0.7702547298039304,
+      "learning_rate": 8.535198972141294e-05,
+      "loss": 2.2519,
+      "step": 1781
+    },
+    {
+      "epoch": 0.8172437514331575,
+      "grad_norm": 0.7572409990224439,
+      "learning_rate": 8.493722731334347e-05,
+      "loss": 2.3581,
+      "step": 1782
+    },
+    {
+      "epoch": 0.8177023618436139,
+      "grad_norm": 0.6951378874896736,
+      "learning_rate": 8.452338155514644e-05,
+      "loss": 2.3824,
+      "step": 1783
+    },
+    {
+      "epoch": 0.8181609722540701,
+      "grad_norm": 0.7025994436922206,
+      "learning_rate": 8.411045336078426e-05,
+      "loss": 2.3499,
+      "step": 1784
+    },
+    {
+      "epoch": 0.8186195826645265,
+      "grad_norm": 0.7181486606680837,
+      "learning_rate": 8.369844364219264e-05,
+      "loss": 2.3254,
+      "step": 1785
+    },
+    {
+      "epoch": 0.8190781930749828,
+      "grad_norm": 0.7236976086156125,
+      "learning_rate": 8.328735330927873e-05,
+      "loss": 2.2463,
+      "step": 1786
+    },
+    {
+      "epoch": 0.8195368034854391,
+      "grad_norm": 0.7213147355922596,
+      "learning_rate": 8.287718326991961e-05,
+      "loss": 2.2692,
+      "step": 1787
+    },
+    {
+      "epoch": 0.8199954138958955,
+      "grad_norm": 0.7536849264290656,
+      "learning_rate": 8.246793442995954e-05,
+      "loss": 2.3901,
+      "step": 1788
+    },
+    {
+      "epoch": 0.8204540243063517,
+      "grad_norm": 0.7059953972267321,
+      "learning_rate": 8.205960769320875e-05,
+      "loss": 2.2981,
+      "step": 1789
+    },
+    {
+      "epoch": 0.8209126347168081,
+      "grad_norm": 0.7051491595912786,
+      "learning_rate": 8.165220396144085e-05,
+      "loss": 2.2223,
+      "step": 1790
+    },
+    {
+      "epoch": 0.8213712451272643,
+      "grad_norm": 0.714963518228223,
+      "learning_rate": 8.12457241343909e-05,
+      "loss": 2.3632,
+      "step": 1791
+    },
+    {
+      "epoch": 0.8218298555377207,
+      "grad_norm": 0.7794668293450567,
+      "learning_rate": 8.084016910975367e-05,
+      "loss": 2.2448,
+      "step": 1792
+    },
+    {
+      "epoch": 0.8222884659481771,
+      "grad_norm": 0.710258730113819,
+      "learning_rate": 8.043553978318169e-05,
+      "loss": 2.2814,
+      "step": 1793
+    },
+    {
+      "epoch": 0.8227470763586333,
+      "grad_norm": 0.7265366583632947,
+      "learning_rate": 8.003183704828281e-05,
+      "loss": 2.2897,
+      "step": 1794
+    },
+    {
+      "epoch": 0.8232056867690897,
+      "grad_norm": 0.7127033223963466,
+      "learning_rate": 7.962906179661872e-05,
+      "loss": 2.3027,
+      "step": 1795
+    },
+    {
+      "epoch": 0.8236642971795459,
+      "grad_norm": 0.7082248851954948,
+      "learning_rate": 7.922721491770296e-05,
+      "loss": 2.3222,
+      "step": 1796
+    },
+    {
+      "epoch": 0.8241229075900023,
+      "grad_norm": 0.713380860267903,
+      "learning_rate": 7.882629729899832e-05,
+      "loss": 2.2903,
+      "step": 1797
+    },
+    {
+      "epoch": 0.8245815180004586,
+      "grad_norm": 0.7330141385679207,
+      "learning_rate": 7.842630982591598e-05,
+      "loss": 2.1641,
+      "step": 1798
+    },
+    {
+      "epoch": 0.8250401284109149,
+      "grad_norm": 0.728985483311404,
+      "learning_rate": 7.802725338181232e-05,
+      "loss": 2.3224,
+      "step": 1799
+    },
+    {
+      "epoch": 0.8254987388213713,
+      "grad_norm": 0.7235507750871983,
+      "learning_rate": 7.762912884798812e-05,
+      "loss": 2.2903,
+      "step": 1800
+    },
+    {
+      "epoch": 0.8259573492318275,
+      "grad_norm": 0.7853064632813476,
+      "learning_rate": 7.723193710368564e-05,
+      "loss": 2.3005,
+      "step": 1801
+    },
+    {
+      "epoch": 0.8264159596422839,
+      "grad_norm": 0.7302679139841116,
+      "learning_rate": 7.683567902608729e-05,
+      "loss": 2.2849,
+      "step": 1802
+    },
+    {
+      "epoch": 0.8268745700527402,
+      "grad_norm": 0.7257757625279996,
+      "learning_rate": 7.644035549031364e-05,
+      "loss": 2.2949,
+      "step": 1803
+    },
+    {
+      "epoch": 0.8273331804631965,
+      "grad_norm": 0.7258109481701551,
+      "learning_rate": 7.604596736942115e-05,
+      "loss": 2.199,
+      "step": 1804
+    },
+    {
+      "epoch": 0.8277917908736528,
+      "grad_norm": 0.7234426745239716,
+      "learning_rate": 7.56525155344004e-05,
+      "loss": 2.2156,
+      "step": 1805
+    },
+    {
+      "epoch": 0.8282504012841091,
+      "grad_norm": 0.7355877179188128,
+      "learning_rate": 7.52600008541745e-05,
+      "loss": 2.2664,
+      "step": 1806
+    },
+    {
+      "epoch": 0.8287090116945655,
+      "grad_norm": 0.7299048442404878,
+      "learning_rate": 7.486842419559681e-05,
+      "loss": 2.3695,
+      "step": 1807
+    },
+    {
+      "epoch": 0.8291676221050218,
+      "grad_norm": 0.7252233278338903,
+      "learning_rate": 7.447778642344898e-05,
+      "loss": 2.2989,
+      "step": 1808
+    },
+    {
+      "epoch": 0.8296262325154781,
+      "grad_norm": 0.7700177195614782,
+      "learning_rate": 7.408808840043912e-05,
+      "loss": 2.3623,
+      "step": 1809
+    },
+    {
+      "epoch": 0.8300848429259344,
+      "grad_norm": 0.7179286800663804,
+      "learning_rate": 7.369933098720021e-05,
+      "loss": 2.301,
+      "step": 1810
+    },
+    {
+      "epoch": 0.8305434533363908,
+      "grad_norm": 0.7639157930171443,
+      "learning_rate": 7.331151504228767e-05,
+      "loss": 2.3297,
+      "step": 1811
+    },
+    {
+      "epoch": 0.831002063746847,
+      "grad_norm": 0.6747538171370899,
+      "learning_rate": 7.292464142217775e-05,
+      "loss": 2.3063,
+      "step": 1812
+    },
+    {
+      "epoch": 0.8314606741573034,
+      "grad_norm": 0.7075973413493241,
+      "learning_rate": 7.25387109812658e-05,
+      "loss": 2.2681,
+      "step": 1813
+    },
+    {
+      "epoch": 0.8319192845677597,
+      "grad_norm": 0.7279113660202674,
+      "learning_rate": 7.215372457186415e-05,
+      "loss": 2.2656,
+      "step": 1814
+    },
+    {
+      "epoch": 0.832377894978216,
+      "grad_norm": 0.6982836322972255,
+      "learning_rate": 7.176968304420007e-05,
+      "loss": 2.2343,
+      "step": 1815
+    },
+    {
+      "epoch": 0.8328365053886724,
+      "grad_norm": 0.7252807323452916,
+      "learning_rate": 7.138658724641417e-05,
+      "loss": 2.2348,
+      "step": 1816
+    },
+    {
+      "epoch": 0.8332951157991286,
+      "grad_norm": 0.7255677898247026,
+      "learning_rate": 7.10044380245587e-05,
+      "loss": 2.3183,
+      "step": 1817
+    },
+    {
+      "epoch": 0.833753726209585,
+      "grad_norm": 0.739136539068194,
+      "learning_rate": 7.062323622259515e-05,
+      "loss": 2.2587,
+      "step": 1818
+    },
+    {
+      "epoch": 0.8342123366200412,
+      "grad_norm": 0.7916801669750044,
+      "learning_rate": 7.024298268239265e-05,
+      "loss": 2.2285,
+      "step": 1819
+    },
+    {
+      "epoch": 0.8346709470304976,
+      "grad_norm": 0.6904195905932508,
+      "learning_rate": 6.986367824372647e-05,
+      "loss": 2.3282,
+      "step": 1820
+    },
+    {
+      "epoch": 0.835129557440954,
+      "grad_norm": 0.6761104322140185,
+      "learning_rate": 6.948532374427541e-05,
+      "loss": 2.2787,
+      "step": 1821
+    },
+    {
+      "epoch": 0.8355881678514102,
+      "grad_norm": 0.7245625641610476,
+      "learning_rate": 6.910792001962063e-05,
+      "loss": 2.3893,
+      "step": 1822
+    },
+    {
+      "epoch": 0.8360467782618666,
+      "grad_norm": 0.7231189631823779,
+      "learning_rate": 6.873146790324358e-05,
+      "loss": 2.306,
+      "step": 1823
+    },
+    {
+      "epoch": 0.8365053886723228,
+      "grad_norm": 0.7440369205930614,
+      "learning_rate": 6.83559682265239e-05,
+      "loss": 2.3042,
+      "step": 1824
+    },
+    {
+      "epoch": 0.8369639990827792,
+      "grad_norm": 0.7429903001682765,
+      "learning_rate": 6.798142181873784e-05,
+      "loss": 2.2718,
+      "step": 1825
+    },
+    {
+      "epoch": 0.8374226094932355,
+      "grad_norm": 0.7168120514893009,
+      "learning_rate": 6.760782950705662e-05,
+      "loss": 2.2786,
+      "step": 1826
+    },
+    {
+      "epoch": 0.8378812199036918,
+      "grad_norm": 0.7294960045810399,
+      "learning_rate": 6.723519211654422e-05,
+      "loss": 2.2996,
+      "step": 1827
+    },
+    {
+      "epoch": 0.8383398303141482,
+      "grad_norm": 0.7306417497253564,
+      "learning_rate": 6.686351047015554e-05,
+      "loss": 2.3133,
+      "step": 1828
+    },
+    {
+      "epoch": 0.8387984407246044,
+      "grad_norm": 0.7254820483521226,
+      "learning_rate": 6.649278538873515e-05,
+      "loss": 2.1985,
+      "step": 1829
+    },
+    {
+      "epoch": 0.8392570511350608,
+      "grad_norm": 0.7249407003933934,
+      "learning_rate": 6.612301769101465e-05,
+      "loss": 2.3171,
+      "step": 1830
+    },
+    {
+      "epoch": 0.8397156615455171,
+      "grad_norm": 0.7468342218295666,
+      "learning_rate": 6.575420819361177e-05,
+      "loss": 2.387,
+      "step": 1831
+    },
+    {
+      "epoch": 0.8401742719559734,
+      "grad_norm": 0.740838479411546,
+      "learning_rate": 6.538635771102757e-05,
+      "loss": 2.242,
+      "step": 1832
+    },
+    {
+      "epoch": 0.8406328823664297,
+      "grad_norm": 0.7401388343839184,
+      "learning_rate": 6.501946705564566e-05,
+      "loss": 2.3902,
+      "step": 1833
+    },
+    {
+      "epoch": 0.841091492776886,
+      "grad_norm": 0.717058907632415,
+      "learning_rate": 6.465353703772959e-05,
+      "loss": 2.2728,
+      "step": 1834
+    },
+    {
+      "epoch": 0.8415501031873424,
+      "grad_norm": 0.6968001501939545,
+      "learning_rate": 6.428856846542136e-05,
+      "loss": 2.283,
+      "step": 1835
+    },
+    {
+      "epoch": 0.8420087135977987,
+      "grad_norm": 0.7220519715357658,
+      "learning_rate": 6.392456214473996e-05,
+      "loss": 2.3374,
+      "step": 1836
+    },
+    {
+      "epoch": 0.842467324008255,
+      "grad_norm": 0.698003524935235,
+      "learning_rate": 6.3561518879579e-05,
+      "loss": 2.2352,
+      "step": 1837
+    },
+    {
+      "epoch": 0.8429259344187113,
+      "grad_norm": 0.7217239112186452,
+      "learning_rate": 6.31994394717052e-05,
+      "loss": 2.1921,
+      "step": 1838
+    },
+    {
+      "epoch": 0.8433845448291676,
+      "grad_norm": 0.7366898542471704,
+      "learning_rate": 6.283832472075685e-05,
+      "loss": 2.3256,
+      "step": 1839
+    },
+    {
+      "epoch": 0.8438431552396239,
+      "grad_norm": 0.76469218944505,
+      "learning_rate": 6.247817542424178e-05,
+      "loss": 2.2994,
+      "step": 1840
+    },
+    {
+      "epoch": 0.8443017656500803,
+      "grad_norm": 0.7428247473948034,
+      "learning_rate": 6.211899237753559e-05,
+      "loss": 2.3201,
+      "step": 1841
+    },
+    {
+      "epoch": 0.8447603760605366,
+      "grad_norm": 0.7771567751114726,
+      "learning_rate": 6.176077637387984e-05,
+      "loss": 2.2923,
+      "step": 1842
+    },
+    {
+      "epoch": 0.8452189864709929,
+      "grad_norm": 0.7365084993291714,
+      "learning_rate": 6.140352820438066e-05,
+      "loss": 2.2541,
+      "step": 1843
+    },
+    {
+      "epoch": 0.8456775968814492,
+      "grad_norm": 0.7482366230582131,
+      "learning_rate": 6.104724865800665e-05,
+      "loss": 2.2874,
+      "step": 1844
+    },
+    {
+      "epoch": 0.8461362072919055,
+      "grad_norm": 0.7057152956891448,
+      "learning_rate": 6.069193852158711e-05,
+      "loss": 2.3007,
+      "step": 1845
+    },
+    {
+      "epoch": 0.8465948177023619,
+      "grad_norm": 0.7365468978701801,
+      "learning_rate": 6.0337598579810584e-05,
+      "loss": 2.2483,
+      "step": 1846
+    },
+    {
+      "epoch": 0.8470534281128181,
+      "grad_norm": 0.7373144455447158,
+      "learning_rate": 5.9984229615223096e-05,
+      "loss": 2.2732,
+      "step": 1847
+    },
+    {
+      "epoch": 0.8475120385232745,
+      "grad_norm": 0.7312924703014861,
+      "learning_rate": 5.963183240822606e-05,
+      "loss": 2.3218,
+      "step": 1848
+    },
+    {
+      "epoch": 0.8479706489337308,
+      "grad_norm": 0.7614697830460637,
+      "learning_rate": 5.9280407737074825e-05,
+      "loss": 2.2402,
+      "step": 1849
+    },
+    {
+      "epoch": 0.8484292593441871,
+      "grad_norm": 0.7091256708896855,
+      "learning_rate": 5.8929956377877125e-05,
+      "loss": 2.2884,
+      "step": 1850
+    },
+    {
+      "epoch": 0.8488878697546435,
+      "grad_norm": 0.7286793115269631,
+      "learning_rate": 5.8580479104591075e-05,
+      "loss": 2.2728,
+      "step": 1851
+    },
+    {
+      "epoch": 0.8493464801650997,
+      "grad_norm": 0.6982000658361596,
+      "learning_rate": 5.823197668902341e-05,
+      "loss": 2.3331,
+      "step": 1852
+    },
+    {
+      "epoch": 0.8498050905755561,
+      "grad_norm": 0.6966513410446892,
+      "learning_rate": 5.78844499008282e-05,
+      "loss": 2.2296,
+      "step": 1853
+    },
+    {
+      "epoch": 0.8502637009860123,
+      "grad_norm": 0.7300249109725828,
+      "learning_rate": 5.753789950750454e-05,
+      "loss": 2.2761,
+      "step": 1854
+    },
+    {
+      "epoch": 0.8507223113964687,
+      "grad_norm": 0.7498997632797101,
+      "learning_rate": 5.719232627439558e-05,
+      "loss": 2.3088,
+      "step": 1855
+    },
+    {
+      "epoch": 0.8511809218069251,
+      "grad_norm": 0.6960200534363328,
+      "learning_rate": 5.6847730964686315e-05,
+      "loss": 2.1799,
+      "step": 1856
+    },
+    {
+      "epoch": 0.8516395322173813,
+      "grad_norm": 0.6761771463043935,
+      "learning_rate": 5.650411433940189e-05,
+      "loss": 2.3581,
+      "step": 1857
+    },
+    {
+      "epoch": 0.8520981426278377,
+      "grad_norm": 0.7191041256729772,
+      "learning_rate": 5.61614771574061e-05,
+      "loss": 2.2464,
+      "step": 1858
+    },
+    {
+      "epoch": 0.8525567530382939,
+      "grad_norm": 0.6892446161789376,
+      "learning_rate": 5.581982017539988e-05,
+      "loss": 2.2158,
+      "step": 1859
+    },
+    {
+      "epoch": 0.8530153634487503,
+      "grad_norm": 0.7386620401823248,
+      "learning_rate": 5.5479144147919216e-05,
+      "loss": 2.2444,
+      "step": 1860
+    },
+    {
+      "epoch": 0.8534739738592066,
+      "grad_norm": 0.72894501140791,
+      "learning_rate": 5.51394498273336e-05,
+      "loss": 2.2491,
+      "step": 1861
+    },
+    {
+      "epoch": 0.8539325842696629,
+      "grad_norm": 0.750738281599297,
+      "learning_rate": 5.480073796384494e-05,
+      "loss": 2.3642,
+      "step": 1862
+    },
+    {
+      "epoch": 0.8543911946801193,
+      "grad_norm": 0.7140305618111831,
+      "learning_rate": 5.446300930548492e-05,
+      "loss": 2.3544,
+      "step": 1863
+    },
+    {
+      "epoch": 0.8548498050905755,
+      "grad_norm": 0.6797981268093745,
+      "learning_rate": 5.412626459811415e-05,
+      "loss": 2.2116,
+      "step": 1864
+    },
+    {
+      "epoch": 0.8553084155010319,
+      "grad_norm": 0.7203577030464694,
+      "learning_rate": 5.3790504585419954e-05,
+      "loss": 2.245,
+      "step": 1865
+    },
+    {
+      "epoch": 0.8557670259114882,
+      "grad_norm": 0.6744176419953988,
+      "learning_rate": 5.345573000891541e-05,
+      "loss": 2.1654,
+      "step": 1866
+    },
+    {
+      "epoch": 0.8562256363219445,
+      "grad_norm": 0.7398105928421597,
+      "learning_rate": 5.312194160793693e-05,
+      "loss": 2.2942,
+      "step": 1867
+    },
+    {
+      "epoch": 0.8566842467324008,
+      "grad_norm": 0.7223713557301462,
+      "learning_rate": 5.278914011964303e-05,
+      "loss": 2.3137,
+      "step": 1868
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 0.7428839672795267,
+      "learning_rate": 5.2457326279013006e-05,
+      "loss": 2.2622,
+      "step": 1869
+    },
+    {
+      "epoch": 0.8576014675533135,
+      "grad_norm": 0.7612774877543238,
+      "learning_rate": 5.2126500818844514e-05,
+      "loss": 2.4027,
+      "step": 1870
+    },
+    {
+      "epoch": 0.8580600779637698,
+      "grad_norm": 0.7273539847785528,
+      "learning_rate": 5.1796664469752566e-05,
+      "loss": 2.3502,
+      "step": 1871
+    },
+    {
+      "epoch": 0.8585186883742261,
+      "grad_norm": 0.7154723287378163,
+      "learning_rate": 5.1467817960167975e-05,
+      "loss": 2.2418,
+      "step": 1872
+    },
+    {
+      "epoch": 0.8589772987846824,
+      "grad_norm": 0.728428559063922,
+      "learning_rate": 5.113996201633536e-05,
+      "loss": 2.316,
+      "step": 1873
+    },
+    {
+      "epoch": 0.8594359091951387,
+      "grad_norm": 0.730587540207258,
+      "learning_rate": 5.0813097362311765e-05,
+      "loss": 2.303,
+      "step": 1874
+    },
+    {
+      "epoch": 0.859894519605595,
+      "grad_norm": 0.685433408380862,
+      "learning_rate": 5.048722471996475e-05,
+      "loss": 2.158,
+      "step": 1875
+    },
+    {
+      "epoch": 0.8603531300160514,
+      "grad_norm": 0.7079953971958118,
+      "learning_rate": 5.016234480897158e-05,
+      "loss": 2.2992,
+      "step": 1876
+    },
+    {
+      "epoch": 0.8608117404265077,
+      "grad_norm": 0.742536711886256,
+      "learning_rate": 4.9838458346816664e-05,
+      "loss": 2.2351,
+      "step": 1877
+    },
+    {
+      "epoch": 0.861270350836964,
+      "grad_norm": 0.7117820799815655,
+      "learning_rate": 4.9515566048790485e-05,
+      "loss": 2.2615,
+      "step": 1878
+    },
+    {
+      "epoch": 0.8617289612474203,
+      "grad_norm": 0.7588457841719125,
+      "learning_rate": 4.9193668627988074e-05,
+      "loss": 2.2439,
+      "step": 1879
+    },
+    {
+      "epoch": 0.8621875716578766,
+      "grad_norm": 0.7161315976880985,
+      "learning_rate": 4.887276679530744e-05,
+      "loss": 2.3039,
+      "step": 1880
+    },
+    {
+      "epoch": 0.862646182068333,
+      "grad_norm": 0.7590503662312474,
+      "learning_rate": 4.855286125944752e-05,
+      "loss": 2.3212,
+      "step": 1881
+    },
+    {
+      "epoch": 0.8631047924787892,
+      "grad_norm": 0.7219140429393168,
+      "learning_rate": 4.8233952726907224e-05,
+      "loss": 2.31,
+      "step": 1882
+    },
+    {
+      "epoch": 0.8635634028892456,
+      "grad_norm": 0.726340818458617,
+      "learning_rate": 4.7916041901983565e-05,
+      "loss": 2.2531,
+      "step": 1883
+    },
+    {
+      "epoch": 0.8640220132997019,
+      "grad_norm": 0.7254154095102977,
+      "learning_rate": 4.7599129486770145e-05,
+      "loss": 2.2548,
+      "step": 1884
+    },
+    {
+      "epoch": 0.8644806237101582,
+      "grad_norm": 0.7155439196869944,
+      "learning_rate": 4.728321618115555e-05,
+      "loss": 2.2714,
+      "step": 1885
+    },
+    {
+      "epoch": 0.8649392341206146,
+      "grad_norm": 0.7233719317278844,
+      "learning_rate": 4.696830268282204e-05,
+      "loss": 2.2219,
+      "step": 1886
+    },
+    {
+      "epoch": 0.8653978445310708,
+      "grad_norm": 0.7273421399061434,
+      "learning_rate": 4.665438968724361e-05,
+      "loss": 2.3329,
+      "step": 1887
+    },
+    {
+      "epoch": 0.8658564549415272,
+      "grad_norm": 0.7200993571390291,
+      "learning_rate": 4.634147788768489e-05,
+      "loss": 2.2686,
+      "step": 1888
+    },
+    {
+      "epoch": 0.8663150653519834,
+      "grad_norm": 0.7334060343940669,
+      "learning_rate": 4.6029567975199414e-05,
+      "loss": 2.3107,
+      "step": 1889
+    },
+    {
+      "epoch": 0.8667736757624398,
+      "grad_norm": 0.7322289592673676,
+      "learning_rate": 4.571866063862795e-05,
+      "loss": 2.3034,
+      "step": 1890
+    },
+    {
+      "epoch": 0.8672322861728962,
+      "grad_norm": 0.7914025355333787,
+      "learning_rate": 4.540875656459703e-05,
+      "loss": 2.4218,
+      "step": 1891
+    },
+    {
+      "epoch": 0.8676908965833524,
+      "grad_norm": 0.6894963980493447,
+      "learning_rate": 4.509985643751785e-05,
+      "loss": 2.2613,
+      "step": 1892
+    },
+    {
+      "epoch": 0.8681495069938088,
+      "grad_norm": 0.7255099850136174,
+      "learning_rate": 4.479196093958421e-05,
+      "loss": 2.2664,
+      "step": 1893
+    },
+    {
+      "epoch": 0.868608117404265,
+      "grad_norm": 0.7115160774974131,
+      "learning_rate": 4.4485070750771187e-05,
+      "loss": 2.3088,
+      "step": 1894
+    },
+    {
+      "epoch": 0.8690667278147214,
+      "grad_norm": 0.783048729961547,
+      "learning_rate": 4.417918654883363e-05,
+      "loss": 2.2641,
+      "step": 1895
+    },
+    {
+      "epoch": 0.8695253382251777,
+      "grad_norm": 0.7337746395398039,
+      "learning_rate": 4.3874309009305e-05,
+      "loss": 2.327,
+      "step": 1896
+    },
+    {
+      "epoch": 0.869983948635634,
+      "grad_norm": 0.7178051341853066,
+      "learning_rate": 4.357043880549538e-05,
+      "loss": 2.3679,
+      "step": 1897
+    },
+    {
+      "epoch": 0.8704425590460904,
+      "grad_norm": 0.8409057461220555,
+      "learning_rate": 4.326757660849012e-05,
+      "loss": 2.3247,
+      "step": 1898
+    },
+    {
+      "epoch": 0.8709011694565466,
+      "grad_norm": 0.7199549091963621,
+      "learning_rate": 4.2965723087148635e-05,
+      "loss": 2.249,
+      "step": 1899
+    },
+    {
+      "epoch": 0.871359779867003,
+      "grad_norm": 0.6984432679939969,
+      "learning_rate": 4.266487890810256e-05,
+      "loss": 2.2427,
+      "step": 1900
+    },
+    {
+      "epoch": 0.8718183902774593,
+      "grad_norm": 0.7129228677300524,
+      "learning_rate": 4.2365044735754365e-05,
+      "loss": 2.2102,
+      "step": 1901
+    },
+    {
+      "epoch": 0.8722770006879156,
+      "grad_norm": 0.7279059553870858,
+      "learning_rate": 4.2066221232276266e-05,
+      "loss": 2.2363,
+      "step": 1902
+    },
+    {
+      "epoch": 0.8727356110983719,
+      "grad_norm": 0.7412289645446063,
+      "learning_rate": 4.176840905760815e-05,
+      "loss": 2.2721,
+      "step": 1903
+    },
+    {
+      "epoch": 0.8731942215088283,
+      "grad_norm": 0.7018166359189186,
+      "learning_rate": 4.1471608869456443e-05,
+      "loss": 2.2317,
+      "step": 1904
+    },
+    {
+      "epoch": 0.8736528319192846,
+      "grad_norm": 0.7350308199568246,
+      "learning_rate": 4.117582132329284e-05,
+      "loss": 2.4067,
+      "step": 1905
+    },
+    {
+      "epoch": 0.8741114423297409,
+      "grad_norm": 0.7046240974134578,
+      "learning_rate": 4.088104707235263e-05,
+      "loss": 2.2998,
+      "step": 1906
+    },
+    {
+      "epoch": 0.8745700527401972,
+      "grad_norm": 0.734503797671813,
+      "learning_rate": 4.058728676763313e-05,
+      "loss": 2.3186,
+      "step": 1907
+    },
+    {
+      "epoch": 0.8750286631506535,
+      "grad_norm": 0.763847848871875,
+      "learning_rate": 4.0294541057892375e-05,
+      "loss": 2.2263,
+      "step": 1908
+    },
+    {
+      "epoch": 0.8754872735611099,
+      "grad_norm": 0.731819858217221,
+      "learning_rate": 4.000281058964794e-05,
+      "loss": 2.2694,
+      "step": 1909
+    },
+    {
+      "epoch": 0.8759458839715661,
+      "grad_norm": 0.7523144507586096,
+      "learning_rate": 3.971209600717507e-05,
+      "loss": 2.2213,
+      "step": 1910
+    },
+    {
+      "epoch": 0.8764044943820225,
+      "grad_norm": 0.7220506406742802,
+      "learning_rate": 3.9422397952505465e-05,
+      "loss": 2.1755,
+      "step": 1911
+    },
+    {
+      "epoch": 0.8768631047924788,
+      "grad_norm": 0.6803378401519575,
+      "learning_rate": 3.913371706542596e-05,
+      "loss": 2.3044,
+      "step": 1912
+    },
+    {
+      "epoch": 0.8773217152029351,
+      "grad_norm": 0.7225948664089792,
+      "learning_rate": 3.884605398347707e-05,
+      "loss": 2.1722,
+      "step": 1913
+    },
+    {
+      "epoch": 0.8777803256133915,
+      "grad_norm": 0.7127079966132621,
+      "learning_rate": 3.8559409341951456e-05,
+      "loss": 2.3223,
+      "step": 1914
+    },
+    {
+      "epoch": 0.8782389360238477,
+      "grad_norm": 0.9258454962649637,
+      "learning_rate": 3.8273783773892404e-05,
+      "loss": 2.3374,
+      "step": 1915
+    },
+    {
+      "epoch": 0.8786975464343041,
+      "grad_norm": 0.7252260819288117,
+      "learning_rate": 3.798917791009293e-05,
+      "loss": 2.2555,
+      "step": 1916
+    },
+    {
+      "epoch": 0.8791561568447603,
+      "grad_norm": 0.7005809533059819,
+      "learning_rate": 3.770559237909393e-05,
+      "loss": 2.2428,
+      "step": 1917
+    },
+    {
+      "epoch": 0.8796147672552167,
+      "grad_norm": 0.7228488336875963,
+      "learning_rate": 3.742302780718288e-05,
+      "loss": 2.2275,
+      "step": 1918
+    },
+    {
+      "epoch": 0.8800733776656731,
+      "grad_norm": 0.7126735358349339,
+      "learning_rate": 3.7141484818392635e-05,
+      "loss": 2.2117,
+      "step": 1919
+    },
+    {
+      "epoch": 0.8805319880761293,
+      "grad_norm": 0.694108518302772,
+      "learning_rate": 3.686096403449973e-05,
+      "loss": 2.2695,
+      "step": 1920
+    },
+    {
+      "epoch": 0.8809905984865857,
+      "grad_norm": 0.7361283409140226,
+      "learning_rate": 3.658146607502344e-05,
+      "loss": 2.2786,
+      "step": 1921
+    },
+    {
+      "epoch": 0.8814492088970419,
+      "grad_norm": 0.6982635760699415,
+      "learning_rate": 3.630299155722411e-05,
+      "loss": 2.3344,
+      "step": 1922
+    },
+    {
+      "epoch": 0.8819078193074983,
+      "grad_norm": 0.7339759354127735,
+      "learning_rate": 3.6025541096101676e-05,
+      "loss": 2.313,
+      "step": 1923
+    },
+    {
+      "epoch": 0.8823664297179546,
+      "grad_norm": 0.7305717835061438,
+      "learning_rate": 3.574911530439473e-05,
+      "loss": 2.3352,
+      "step": 1924
+    },
+    {
+      "epoch": 0.8828250401284109,
+      "grad_norm": 0.7399534639833846,
+      "learning_rate": 3.5473714792578606e-05,
+      "loss": 2.3142,
+      "step": 1925
+    },
+    {
+      "epoch": 0.8832836505388673,
+      "grad_norm": 0.7556456296674257,
+      "learning_rate": 3.519934016886478e-05,
+      "loss": 2.3792,
+      "step": 1926
+    },
+    {
+      "epoch": 0.8837422609493235,
+      "grad_norm": 0.7132899342622725,
+      "learning_rate": 3.4925992039198776e-05,
+      "loss": 2.3275,
+      "step": 1927
+    },
+    {
+      "epoch": 0.8842008713597799,
+      "grad_norm": 0.7521641563249845,
+      "learning_rate": 3.465367100725908e-05,
+      "loss": 2.2935,
+      "step": 1928
+    },
+    {
+      "epoch": 0.8846594817702362,
+      "grad_norm": 0.7131426364146989,
+      "learning_rate": 3.438237767445618e-05,
+      "loss": 2.239,
+      "step": 1929
+    },
+    {
+      "epoch": 0.8851180921806925,
+      "grad_norm": 0.6845763234447338,
+      "learning_rate": 3.4112112639930804e-05,
+      "loss": 2.3227,
+      "step": 1930
+    },
+    {
+      "epoch": 0.8855767025911488,
+      "grad_norm": 0.7437755923416018,
+      "learning_rate": 3.3842876500552564e-05,
+      "loss": 2.2775,
+      "step": 1931
+    },
+    {
+      "epoch": 0.8860353130016051,
+      "grad_norm": 0.7044107880181344,
+      "learning_rate": 3.357466985091906e-05,
+      "loss": 2.3154,
+      "step": 1932
+    },
+    {
+      "epoch": 0.8864939234120615,
+      "grad_norm": 0.6694017245920526,
+      "learning_rate": 3.330749328335414e-05,
+      "loss": 2.2562,
+      "step": 1933
+    },
+    {
+      "epoch": 0.8869525338225178,
+      "grad_norm": 0.7345227568203369,
+      "learning_rate": 3.304134738790659e-05,
+      "loss": 2.2458,
+      "step": 1934
+    },
+    {
+      "epoch": 0.8874111442329741,
+      "grad_norm": 0.7230924233624242,
+      "learning_rate": 3.277623275234953e-05,
+      "loss": 2.3058,
+      "step": 1935
+    },
+    {
+      "epoch": 0.8878697546434304,
+      "grad_norm": 0.7283444171065788,
+      "learning_rate": 3.2512149962177994e-05,
+      "loss": 2.3592,
+      "step": 1936
+    },
+    {
+      "epoch": 0.8883283650538867,
+      "grad_norm": 0.7089427361973437,
+      "learning_rate": 3.224909960060851e-05,
+      "loss": 2.2366,
+      "step": 1937
+    },
+    {
+      "epoch": 0.888786975464343,
+      "grad_norm": 0.7212290402541809,
+      "learning_rate": 3.198708224857755e-05,
+      "loss": 2.2852,
+      "step": 1938
+    },
+    {
+      "epoch": 0.8892455858747994,
+      "grad_norm": 0.6877176991408143,
+      "learning_rate": 3.172609848474023e-05,
+      "loss": 2.2223,
+      "step": 1939
+    },
+    {
+      "epoch": 0.8897041962852557,
+      "grad_norm": 0.7041919513581185,
+      "learning_rate": 3.1466148885468895e-05,
+      "loss": 2.3635,
+      "step": 1940
+    },
+    {
+      "epoch": 0.890162806695712,
+      "grad_norm": 0.7198247365023949,
+      "learning_rate": 3.120723402485198e-05,
+      "loss": 2.299,
+      "step": 1941
+    },
+    {
+      "epoch": 0.8906214171061683,
+      "grad_norm": 0.7239034924458737,
+      "learning_rate": 3.094935447469294e-05,
+      "loss": 2.1966,
+      "step": 1942
+    },
+    {
+      "epoch": 0.8910800275166246,
+      "grad_norm": 0.7166382189079809,
+      "learning_rate": 3.069251080450863e-05,
+      "loss": 2.3154,
+      "step": 1943
+    },
+    {
+      "epoch": 0.891538637927081,
+      "grad_norm": 0.7237206375103279,
+      "learning_rate": 3.0436703581528113e-05,
+      "loss": 2.2898,
+      "step": 1944
+    },
+    {
+      "epoch": 0.8919972483375372,
+      "grad_norm": 0.7084673178423593,
+      "learning_rate": 3.0181933370691694e-05,
+      "loss": 2.3028,
+      "step": 1945
+    },
+    {
+      "epoch": 0.8924558587479936,
+      "grad_norm": 0.6859938648548278,
+      "learning_rate": 2.9928200734649523e-05,
+      "loss": 2.2208,
+      "step": 1946
+    },
+    {
+      "epoch": 0.8929144691584499,
+      "grad_norm": 0.7001730854754188,
+      "learning_rate": 2.9675506233760142e-05,
+      "loss": 2.2673,
+      "step": 1947
+    },
+    {
+      "epoch": 0.8933730795689062,
+      "grad_norm": 0.7507247917051159,
+      "learning_rate": 2.942385042608925e-05,
+      "loss": 2.3125,
+      "step": 1948
+    },
+    {
+      "epoch": 0.8938316899793626,
+      "grad_norm": 0.7072829726984374,
+      "learning_rate": 2.9173233867409054e-05,
+      "loss": 2.268,
+      "step": 1949
+    },
+    {
+      "epoch": 0.8942903003898188,
+      "grad_norm": 0.7219747411937606,
+      "learning_rate": 2.892365711119638e-05,
+      "loss": 2.2691,
+      "step": 1950
+    },
+    {
+      "epoch": 0.8947489108002752,
+      "grad_norm": 0.7192148303448229,
+      "learning_rate": 2.8675120708631596e-05,
+      "loss": 2.3086,
+      "step": 1951
+    },
+    {
+      "epoch": 0.8952075212107314,
+      "grad_norm": 0.694216122496986,
+      "learning_rate": 2.8427625208597764e-05,
+      "loss": 2.2654,
+      "step": 1952
+    },
+    {
+      "epoch": 0.8956661316211878,
+      "grad_norm": 0.7421163723298023,
+      "learning_rate": 2.8181171157678874e-05,
+      "loss": 2.3128,
+      "step": 1953
+    },
+    {
+      "epoch": 0.8961247420316442,
+      "grad_norm": 0.7271258544479349,
+      "learning_rate": 2.7935759100159053e-05,
+      "loss": 2.3614,
+      "step": 1954
+    },
+    {
+      "epoch": 0.8965833524421004,
+      "grad_norm": 0.7665300824164257,
+      "learning_rate": 2.7691389578021365e-05,
+      "loss": 2.2363,
+      "step": 1955
+    },
+    {
+      "epoch": 0.8970419628525568,
+      "grad_norm": 0.706408385821372,
+      "learning_rate": 2.7448063130946223e-05,
+      "loss": 2.2877,
+      "step": 1956
+    },
+    {
+      "epoch": 0.897500573263013,
+      "grad_norm": 0.7051634765334506,
+      "learning_rate": 2.7205780296310544e-05,
+      "loss": 2.3067,
+      "step": 1957
+    },
+    {
+      "epoch": 0.8979591836734694,
+      "grad_norm": 0.7091112014783713,
+      "learning_rate": 2.6964541609186378e-05,
+      "loss": 2.2569,
+      "step": 1958
+    },
+    {
+      "epoch": 0.8984177940839257,
+      "grad_norm": 0.7329752076831735,
+      "learning_rate": 2.6724347602340104e-05,
+      "loss": 2.2589,
+      "step": 1959
+    },
+    {
+      "epoch": 0.898876404494382,
+      "grad_norm": 0.7113659067033999,
+      "learning_rate": 2.6485198806230682e-05,
+      "loss": 2.2865,
+      "step": 1960
+    },
+    {
+      "epoch": 0.8993350149048384,
+      "grad_norm": 0.7253782670592386,
+      "learning_rate": 2.6247095749008797e-05,
+      "loss": 2.396,
+      "step": 1961
+    },
+    {
+      "epoch": 0.8997936253152946,
+      "grad_norm": 0.7036912218981978,
+      "learning_rate": 2.6010038956515826e-05,
+      "loss": 2.3389,
+      "step": 1962
+    },
+    {
+      "epoch": 0.900252235725751,
+      "grad_norm": 0.7262997530309051,
+      "learning_rate": 2.5774028952282423e-05,
+      "loss": 2.2833,
+      "step": 1963
+    },
+    {
+      "epoch": 0.9007108461362073,
+      "grad_norm": 0.7358238442180726,
+      "learning_rate": 2.5539066257527277e-05,
+      "loss": 2.2691,
+      "step": 1964
+    },
+    {
+      "epoch": 0.9011694565466636,
+      "grad_norm": 0.707408828207914,
+      "learning_rate": 2.530515139115652e-05,
+      "loss": 2.304,
+      "step": 1965
+    },
+    {
+      "epoch": 0.90162806695712,
+      "grad_norm": 0.718650089070979,
+      "learning_rate": 2.5072284869761874e-05,
+      "loss": 2.2616,
+      "step": 1966
+    },
+    {
+      "epoch": 0.9020866773675762,
+      "grad_norm": 0.7055508324775379,
+      "learning_rate": 2.4840467207619786e-05,
+      "loss": 2.2171,
+      "step": 1967
+    },
+    {
+      "epoch": 0.9025452877780326,
+      "grad_norm": 0.7393582477713856,
+      "learning_rate": 2.460969891669068e-05,
+      "loss": 2.3395,
+      "step": 1968
+    },
+    {
+      "epoch": 0.9030038981884889,
+      "grad_norm": 0.693074417113313,
+      "learning_rate": 2.4379980506617272e-05,
+      "loss": 2.2863,
+      "step": 1969
+    },
+    {
+      "epoch": 0.9034625085989452,
+      "grad_norm": 0.7284203043053964,
+      "learning_rate": 2.4151312484723464e-05,
+      "loss": 2.3333,
+      "step": 1970
+    },
+    {
+      "epoch": 0.9039211190094015,
+      "grad_norm": 0.7522291960512043,
+      "learning_rate": 2.3923695356013798e-05,
+      "loss": 2.223,
+      "step": 1971
+    },
+    {
+      "epoch": 0.9043797294198578,
+      "grad_norm": 0.7107496121803782,
+      "learning_rate": 2.3697129623171833e-05,
+      "loss": 2.2484,
+      "step": 1972
+    },
+    {
+      "epoch": 0.9048383398303141,
+      "grad_norm": 0.7336346824605128,
+      "learning_rate": 2.3471615786559042e-05,
+      "loss": 2.3472,
+      "step": 1973
+    },
+    {
+      "epoch": 0.9052969502407705,
+      "grad_norm": 0.7294282666971926,
+      "learning_rate": 2.3247154344213818e-05,
+      "loss": 2.2667,
+      "step": 1974
+    },
+    {
+      "epoch": 0.9057555606512268,
+      "grad_norm": 0.7275537085339051,
+      "learning_rate": 2.3023745791850625e-05,
+      "loss": 2.2424,
+      "step": 1975
+    },
+    {
+      "epoch": 0.9062141710616831,
+      "grad_norm": 0.7191180471346014,
+      "learning_rate": 2.2801390622858354e-05,
+      "loss": 2.3593,
+      "step": 1976
+    },
+    {
+      "epoch": 0.9066727814721394,
+      "grad_norm": 0.7308467522678734,
+      "learning_rate": 2.2580089328299746e-05,
+      "loss": 2.2594,
+      "step": 1977
+    },
+    {
+      "epoch": 0.9071313918825957,
+      "grad_norm": 0.7180580886845207,
+      "learning_rate": 2.235984239690997e-05,
+      "loss": 2.2823,
+      "step": 1978
+    },
+    {
+      "epoch": 0.9075900022930521,
+      "grad_norm": 0.699800271409043,
+      "learning_rate": 2.2140650315095934e-05,
+      "loss": 2.2099,
+      "step": 1979
+    },
+    {
+      "epoch": 0.9080486127035083,
+      "grad_norm": 0.7100141961263668,
+      "learning_rate": 2.192251356693459e-05,
+      "loss": 2.3603,
+      "step": 1980
+    },
+    {
+      "epoch": 0.9085072231139647,
+      "grad_norm": 0.7362508567254464,
+      "learning_rate": 2.170543263417246e-05,
+      "loss": 2.3575,
+      "step": 1981
+    },
+    {
+      "epoch": 0.908965833524421,
+      "grad_norm": 0.7406354639427267,
+      "learning_rate": 2.1489407996224286e-05,
+      "loss": 2.3081,
+      "step": 1982
+    },
+    {
+      "epoch": 0.9094244439348773,
+      "grad_norm": 0.6803030360024486,
+      "learning_rate": 2.127444013017199e-05,
+      "loss": 2.2931,
+      "step": 1983
+    },
+    {
+      "epoch": 0.9098830543453337,
+      "grad_norm": 0.711900488291685,
+      "learning_rate": 2.1060529510763648e-05,
+      "loss": 2.2541,
+      "step": 1984
+    },
+    {
+      "epoch": 0.9103416647557899,
+      "grad_norm": 0.7304920514984573,
+      "learning_rate": 2.084767661041259e-05,
+      "loss": 2.2979,
+      "step": 1985
+    },
+    {
+      "epoch": 0.9108002751662463,
+      "grad_norm": 0.7293969188139222,
+      "learning_rate": 2.063588189919596e-05,
+      "loss": 2.2712,
+      "step": 1986
+    },
+    {
+      "epoch": 0.9112588855767025,
+      "grad_norm": 0.7726288968087385,
+      "learning_rate": 2.0425145844854275e-05,
+      "loss": 2.2766,
+      "step": 1987
+    },
+    {
+      "epoch": 0.9117174959871589,
+      "grad_norm": 0.768601810305481,
+      "learning_rate": 2.0215468912789693e-05,
+      "loss": 2.3033,
+      "step": 1988
+    },
+    {
+      "epoch": 0.9121761063976153,
+      "grad_norm": 0.7635573172942962,
+      "learning_rate": 2.0006851566065575e-05,
+      "loss": 2.3138,
+      "step": 1989
+    },
+    {
+      "epoch": 0.9126347168080715,
+      "grad_norm": 0.7775455325580196,
+      "learning_rate": 1.9799294265405166e-05,
+      "loss": 2.2039,
+      "step": 1990
+    },
+    {
+      "epoch": 0.9130933272185279,
+      "grad_norm": 0.7452547024900242,
+      "learning_rate": 1.9592797469190572e-05,
+      "loss": 2.3201,
+      "step": 1991
+    },
+    {
+      "epoch": 0.9135519376289841,
+      "grad_norm": 0.7168336558065312,
+      "learning_rate": 1.938736163346194e-05,
+      "loss": 2.3229,
+      "step": 1992
+    },
+    {
+      "epoch": 0.9140105480394405,
+      "grad_norm": 0.7350661543088859,
+      "learning_rate": 1.9182987211916246e-05,
+      "loss": 2.237,
+      "step": 1993
+    },
+    {
+      "epoch": 0.9144691584498968,
+      "grad_norm": 0.7555564078663803,
+      "learning_rate": 1.8979674655906332e-05,
+      "loss": 2.3565,
+      "step": 1994
+    },
+    {
+      "epoch": 0.9149277688603531,
+      "grad_norm": 0.7058807701472705,
+      "learning_rate": 1.8777424414440024e-05,
+      "loss": 2.1439,
+      "step": 1995
+    },
+    {
+      "epoch": 0.9153863792708095,
+      "grad_norm": 0.7245271378536041,
+      "learning_rate": 1.8576236934179202e-05,
+      "loss": 2.3117,
+      "step": 1996
+    },
+    {
+      "epoch": 0.9158449896812658,
+      "grad_norm": 0.7259745222752026,
+      "learning_rate": 1.8376112659438393e-05,
+      "loss": 2.3923,
+      "step": 1997
+    },
+    {
+      "epoch": 0.9163036000917221,
+      "grad_norm": 0.7435986267928336,
+      "learning_rate": 1.8177052032184282e-05,
+      "loss": 2.2646,
+      "step": 1998
+    },
+    {
+      "epoch": 0.9167622105021784,
+      "grad_norm": 0.7206769143904086,
+      "learning_rate": 1.7979055492034435e-05,
+      "loss": 2.3805,
+      "step": 1999
+    },
+    {
+      "epoch": 0.9172208209126347,
+      "grad_norm": 0.7518781050192744,
+      "learning_rate": 1.7782123476256407e-05,
+      "loss": 2.2936,
+      "step": 2000
+    },
+    {
+      "epoch": 0.917679431323091,
+      "grad_norm": 0.718142795815218,
+      "learning_rate": 1.7586256419766965e-05,
+      "loss": 2.3748,
+      "step": 2001
+    },
+    {
+      "epoch": 0.9181380417335474,
+      "grad_norm": 0.7332114451213351,
+      "learning_rate": 1.7391454755130766e-05,
+      "loss": 2.2185,
+      "step": 2002
+    },
+    {
+      "epoch": 0.9185966521440037,
+      "grad_norm": 0.7117303957935709,
+      "learning_rate": 1.7197718912559557e-05,
+      "loss": 2.2528,
+      "step": 2003
+    },
+    {
+      "epoch": 0.91905526255446,
+      "grad_norm": 0.7536867066346347,
+      "learning_rate": 1.700504931991148e-05,
+      "loss": 2.3222,
+      "step": 2004
+    },
+    {
+      "epoch": 0.9195138729649163,
+      "grad_norm": 0.7435208421242708,
+      "learning_rate": 1.681344640268978e-05,
+      "loss": 2.2818,
+      "step": 2005
+    },
+    {
+      "epoch": 0.9199724833753726,
+      "grad_norm": 0.7012442400625714,
+      "learning_rate": 1.6622910584041974e-05,
+      "loss": 2.3565,
+      "step": 2006
+    },
+    {
+      "epoch": 0.920431093785829,
+      "grad_norm": 0.7285348447275941,
+      "learning_rate": 1.6433442284758903e-05,
+      "loss": 2.2527,
+      "step": 2007
+    },
+    {
+      "epoch": 0.9208897041962852,
+      "grad_norm": 0.7077741972080454,
+      "learning_rate": 1.624504192327392e-05,
+      "loss": 2.2506,
+      "step": 2008
+    },
+    {
+      "epoch": 0.9213483146067416,
+      "grad_norm": 0.7133672538343535,
+      "learning_rate": 1.6057709915661856e-05,
+      "loss": 2.2634,
+      "step": 2009
+    },
+    {
+      "epoch": 0.9218069250171979,
+      "grad_norm": 0.7256209997870755,
+      "learning_rate": 1.5871446675638057e-05,
+      "loss": 2.3248,
+      "step": 2010
+    },
+    {
+      "epoch": 0.9222655354276542,
+      "grad_norm": 0.715933110566516,
+      "learning_rate": 1.5686252614557638e-05,
+      "loss": 2.2344,
+      "step": 2011
+    },
+    {
+      "epoch": 0.9227241458381106,
+      "grad_norm": 0.7167059244791303,
+      "learning_rate": 1.5502128141414497e-05,
+      "loss": 2.3206,
+      "step": 2012
+    },
+    {
+      "epoch": 0.9231827562485668,
+      "grad_norm": 0.7375497057589727,
+      "learning_rate": 1.5319073662840188e-05,
+      "loss": 2.2481,
+      "step": 2013
+    },
+    {
+      "epoch": 0.9236413666590232,
+      "grad_norm": 0.741896370205967,
+      "learning_rate": 1.5137089583103391e-05,
+      "loss": 2.343,
+      "step": 2014
+    },
+    {
+      "epoch": 0.9240999770694794,
+      "grad_norm": 0.7457354412559366,
+      "learning_rate": 1.4956176304108893e-05,
+      "loss": 2.3927,
+      "step": 2015
+    },
+    {
+      "epoch": 0.9245585874799358,
+      "grad_norm": 0.7611962304068137,
+      "learning_rate": 1.4776334225396481e-05,
+      "loss": 2.3196,
+      "step": 2016
+    },
+    {
+      "epoch": 0.9250171978903922,
+      "grad_norm": 0.7235598989128703,
+      "learning_rate": 1.4597563744140397e-05,
+      "loss": 2.2316,
+      "step": 2017
+    },
+    {
+      "epoch": 0.9254758083008484,
+      "grad_norm": 0.7515712909583367,
+      "learning_rate": 1.4419865255148269e-05,
+      "loss": 2.3661,
+      "step": 2018
+    },
+    {
+      "epoch": 0.9259344187113048,
+      "grad_norm": 0.7111020947703225,
+      "learning_rate": 1.4243239150860122e-05,
+      "loss": 2.1504,
+      "step": 2019
+    },
+    {
+      "epoch": 0.926393029121761,
+      "grad_norm": 0.7138051414237951,
+      "learning_rate": 1.4067685821347932e-05,
+      "loss": 2.1951,
+      "step": 2020
+    },
+    {
+      "epoch": 0.9268516395322174,
+      "grad_norm": 0.7127004427295934,
+      "learning_rate": 1.389320565431429e-05,
+      "loss": 2.3465,
+      "step": 2021
+    },
+    {
+      "epoch": 0.9273102499426737,
+      "grad_norm": 0.7303397368241836,
+      "learning_rate": 1.3719799035091851e-05,
+      "loss": 2.2283,
+      "step": 2022
+    },
+    {
+      "epoch": 0.92776886035313,
+      "grad_norm": 0.6944347151546197,
+      "learning_rate": 1.3547466346642278e-05,
+      "loss": 2.2421,
+      "step": 2023
+    },
+    {
+      "epoch": 0.9282274707635864,
+      "grad_norm": 0.756043957211238,
+      "learning_rate": 1.3376207969555577e-05,
+      "loss": 2.2205,
+      "step": 2024
+    },
+    {
+      "epoch": 0.9286860811740426,
+      "grad_norm": 0.7334818809224479,
+      "learning_rate": 1.32060242820492e-05,
+      "loss": 2.2359,
+      "step": 2025
+    },
+    {
+      "epoch": 0.929144691584499,
+      "grad_norm": 0.6848170142287372,
+      "learning_rate": 1.3036915659967118e-05,
+      "loss": 2.2777,
+      "step": 2026
+    },
+    {
+      "epoch": 0.9296033019949553,
+      "grad_norm": 0.7141240497010237,
+      "learning_rate": 1.2868882476779087e-05,
+      "loss": 2.2585,
+      "step": 2027
+    },
+    {
+      "epoch": 0.9300619124054116,
+      "grad_norm": 0.7401692782825596,
+      "learning_rate": 1.2701925103579815e-05,
+      "loss": 2.2884,
+      "step": 2028
+    },
+    {
+      "epoch": 0.930520522815868,
+      "grad_norm": 0.7292485834163791,
+      "learning_rate": 1.2536043909088191e-05,
+      "loss": 2.3599,
+      "step": 2029
+    },
+    {
+      "epoch": 0.9309791332263242,
+      "grad_norm": 0.734096796205213,
+      "learning_rate": 1.2371239259646228e-05,
+      "loss": 2.362,
+      "step": 2030
+    },
+    {
+      "epoch": 0.9314377436367806,
+      "grad_norm": 0.7242403718713546,
+      "learning_rate": 1.2207511519218672e-05,
+      "loss": 2.2843,
+      "step": 2031
+    },
+    {
+      "epoch": 0.9318963540472369,
+      "grad_norm": 0.711349454164714,
+      "learning_rate": 1.2044861049391676e-05,
+      "loss": 2.2475,
+      "step": 2032
+    },
+    {
+      "epoch": 0.9323549644576932,
+      "grad_norm": 0.7020145615752645,
+      "learning_rate": 1.1883288209372512e-05,
+      "loss": 2.1766,
+      "step": 2033
+    },
+    {
+      "epoch": 0.9328135748681495,
+      "grad_norm": 0.7151050325851365,
+      "learning_rate": 1.1722793355988471e-05,
+      "loss": 2.3382,
+      "step": 2034
+    },
+    {
+      "epoch": 0.9332721852786058,
+      "grad_norm": 0.6782588223242515,
+      "learning_rate": 1.1563376843686135e-05,
+      "loss": 2.3065,
+      "step": 2035
+    },
+    {
+      "epoch": 0.9337307956890621,
+      "grad_norm": 0.7224136460411357,
+      "learning_rate": 1.140503902453055e-05,
+      "loss": 2.234,
+      "step": 2036
+    },
+    {
+      "epoch": 0.9341894060995185,
+      "grad_norm": 0.7123511655434561,
+      "learning_rate": 1.1247780248204665e-05,
+      "loss": 2.2433,
+      "step": 2037
+    },
+    {
+      "epoch": 0.9346480165099748,
+      "grad_norm": 0.7084483122997004,
+      "learning_rate": 1.1091600862008333e-05,
+      "loss": 2.3196,
+      "step": 2038
+    },
+    {
+      "epoch": 0.9351066269204311,
+      "grad_norm": 0.7084535135576943,
+      "learning_rate": 1.0936501210857652e-05,
+      "loss": 2.2407,
+      "step": 2039
+    },
+    {
+      "epoch": 0.9355652373308874,
+      "grad_norm": 0.7232431263071529,
+      "learning_rate": 1.0782481637284013e-05,
+      "loss": 2.3164,
+      "step": 2040
+    },
+    {
+      "epoch": 0.9360238477413437,
+      "grad_norm": 0.7156000745042087,
+      "learning_rate": 1.0629542481433663e-05,
+      "loss": 2.3193,
+      "step": 2041
+    },
+    {
+      "epoch": 0.9364824581518001,
+      "grad_norm": 0.7096861198521895,
+      "learning_rate": 1.0477684081066751e-05,
+      "loss": 2.2926,
+      "step": 2042
+    },
+    {
+      "epoch": 0.9369410685622563,
+      "grad_norm": 0.7390192733120109,
+      "learning_rate": 1.0326906771556566e-05,
+      "loss": 2.3412,
+      "step": 2043
+    },
+    {
+      "epoch": 0.9373996789727127,
+      "grad_norm": 0.6968833903456555,
+      "learning_rate": 1.017721088588891e-05,
+      "loss": 2.277,
+      "step": 2044
+    },
+    {
+      "epoch": 0.937858289383169,
+      "grad_norm": 0.6840450904973493,
+      "learning_rate": 1.0028596754661334e-05,
+      "loss": 2.202,
+      "step": 2045
+    },
+    {
+      "epoch": 0.9383168997936253,
+      "grad_norm": 0.7086437014973221,
+      "learning_rate": 9.881064706082298e-06,
+      "loss": 2.2783,
+      "step": 2046
+    },
+    {
+      "epoch": 0.9387755102040817,
+      "grad_norm": 0.7209254839461962,
+      "learning_rate": 9.734615065970454e-06,
+      "loss": 2.2799,
+      "step": 2047
+    },
+    {
+      "epoch": 0.9392341206145379,
+      "grad_norm": 0.7089041892875391,
+      "learning_rate": 9.58924815775425e-06,
+      "loss": 2.3186,
+      "step": 2048
+    },
+    {
+      "epoch": 0.9396927310249943,
+      "grad_norm": 0.7252508542672064,
+      "learning_rate": 9.444964302470715e-06,
+      "loss": 2.2869,
+      "step": 2049
+    },
+    {
+      "epoch": 0.9401513414354505,
+      "grad_norm": 0.703735167816813,
+      "learning_rate": 9.301763818765018e-06,
+      "loss": 2.2581,
+      "step": 2050
+    },
+    {
+      "epoch": 0.9406099518459069,
+      "grad_norm": 0.717269005522211,
+      "learning_rate": 9.15964702288996e-06,
+      "loss": 2.3389,
+      "step": 2051
+    },
+    {
+      "epoch": 0.9410685622563633,
+      "grad_norm": 0.7210192461554623,
+      "learning_rate": 9.018614228704925e-06,
+      "loss": 2.1746,
+      "step": 2052
+    },
+    {
+      "epoch": 0.9415271726668195,
+      "grad_norm": 0.7150446752394617,
+      "learning_rate": 8.878665747675152e-06,
+      "loss": 2.2318,
+      "step": 2053
+    },
+    {
+      "epoch": 0.9419857830772759,
+      "grad_norm": 0.7515872496374464,
+      "learning_rate": 8.739801888871469e-06,
+      "loss": 2.2613,
+      "step": 2054
+    },
+    {
+      "epoch": 0.9424443934877321,
+      "grad_norm": 0.7066354428266575,
+      "learning_rate": 8.602022958969336e-06,
+      "loss": 2.3132,
+      "step": 2055
+    },
+    {
+      "epoch": 0.9429030038981885,
+      "grad_norm": 0.7114740421116713,
+      "learning_rate": 8.465329262248078e-06,
+      "loss": 2.2666,
+      "step": 2056
+    },
+    {
+      "epoch": 0.9433616143086448,
+      "grad_norm": 0.7500785056469853,
+      "learning_rate": 8.32972110059027e-06,
+      "loss": 2.2597,
+      "step": 2057
+    },
+    {
+      "epoch": 0.9438202247191011,
+      "grad_norm": 0.7116315866519676,
+      "learning_rate": 8.195198773481406e-06,
+      "loss": 2.3083,
+      "step": 2058
+    },
+    {
+      "epoch": 0.9442788351295575,
+      "grad_norm": 0.7117882453546381,
+      "learning_rate": 8.061762578008613e-06,
+      "loss": 2.302,
+      "step": 2059
+    },
+    {
+      "epoch": 0.9447374455400137,
+      "grad_norm": 0.7247617893346484,
+      "learning_rate": 7.929412808860559e-06,
+      "loss": 2.2645,
+      "step": 2060
+    },
+    {
+      "epoch": 0.9451960559504701,
+      "grad_norm": 0.7326889744739357,
+      "learning_rate": 7.79814975832649e-06,
+      "loss": 2.3578,
+      "step": 2061
+    },
+    {
+      "epoch": 0.9456546663609264,
+      "grad_norm": 0.7390848491923252,
+      "learning_rate": 7.667973716295851e-06,
+      "loss": 2.2516,
+      "step": 2062
+    },
+    {
+      "epoch": 0.9461132767713827,
+      "grad_norm": 0.72957153795183,
+      "learning_rate": 7.5388849702571205e-06,
+      "loss": 2.2868,
+      "step": 2063
+    },
+    {
+      "epoch": 0.946571887181839,
+      "grad_norm": 0.7437407845435979,
+      "learning_rate": 7.4108838052979185e-06,
+      "loss": 2.2302,
+      "step": 2064
+    },
+    {
+      "epoch": 0.9470304975922953,
+      "grad_norm": 0.7230687557304496,
+      "learning_rate": 7.283970504103732e-06,
+      "loss": 2.2339,
+      "step": 2065
+    },
+    {
+      "epoch": 0.9474891080027517,
+      "grad_norm": 0.6780648917410009,
+      "learning_rate": 7.1581453469575785e-06,
+      "loss": 2.2103,
+      "step": 2066
+    },
+    {
+      "epoch": 0.947947718413208,
+      "grad_norm": 0.7238111409364941,
+      "learning_rate": 7.033408611739456e-06,
+      "loss": 2.2661,
+      "step": 2067
+    },
+    {
+      "epoch": 0.9484063288236643,
+      "grad_norm": 0.7547794077831144,
+      "learning_rate": 6.909760573925561e-06,
+      "loss": 2.3261,
+      "step": 2068
+    },
+    {
+      "epoch": 0.9488649392341206,
+      "grad_norm": 0.6969730199175775,
+      "learning_rate": 6.787201506587626e-06,
+      "loss": 2.2011,
+      "step": 2069
+    },
+    {
+      "epoch": 0.9493235496445769,
+      "grad_norm": 0.7258396418307412,
+      "learning_rate": 6.66573168039264e-06,
+      "loss": 2.3402,
+      "step": 2070
+    },
+    {
+      "epoch": 0.9497821600550332,
+      "grad_norm": 0.7157839632335483,
+      "learning_rate": 6.545351363601959e-06,
+      "loss": 2.408,
+      "step": 2071
+    },
+    {
+      "epoch": 0.9502407704654896,
+      "grad_norm": 0.725239395647958,
+      "learning_rate": 6.426060822070812e-06,
+      "loss": 2.1961,
+      "step": 2072
+    },
+    {
+      "epoch": 0.9506993808759459,
+      "grad_norm": 0.6838480120614286,
+      "learning_rate": 6.3078603192475716e-06,
+      "loss": 2.364,
+      "step": 2073
+    },
+    {
+      "epoch": 0.9511579912864022,
+      "grad_norm": 0.774086928212201,
+      "learning_rate": 6.1907501161735934e-06,
+      "loss": 2.3159,
+      "step": 2074
+    },
+    {
+      "epoch": 0.9516166016968585,
+      "grad_norm": 0.7283144311825367,
+      "learning_rate": 6.074730471482049e-06,
+      "loss": 2.2302,
+      "step": 2075
+    },
+    {
+      "epoch": 0.9520752121073148,
+      "grad_norm": 0.6988320500897706,
+      "learning_rate": 5.959801641397755e-06,
+      "loss": 2.2185,
+      "step": 2076
+    },
+    {
+      "epoch": 0.9525338225177712,
+      "grad_norm": 0.7223124510685691,
+      "learning_rate": 5.845963879736627e-06,
+      "loss": 2.2592,
+      "step": 2077
+    },
+    {
+      "epoch": 0.9529924329282274,
+      "grad_norm": 0.7564136161296057,
+      "learning_rate": 5.733217437904892e-06,
+      "loss": 2.2552,
+      "step": 2078
+    },
+    {
+      "epoch": 0.9534510433386838,
+      "grad_norm": 0.7075068742200445,
+      "learning_rate": 5.621562564898597e-06,
+      "loss": 2.275,
+      "step": 2079
+    },
+    {
+      "epoch": 0.95390965374914,
+      "grad_norm": 0.7246608829088889,
+      "learning_rate": 5.51099950730316e-06,
+      "loss": 2.2833,
+      "step": 2080
+    },
+    {
+      "epoch": 0.9543682641595964,
+      "grad_norm": 0.6881239351969125,
+      "learning_rate": 5.401528509292763e-06,
+      "loss": 2.3447,
+      "step": 2081
+    },
+    {
+      "epoch": 0.9548268745700528,
+      "grad_norm": 0.7718044270144777,
+      "learning_rate": 5.2931498126298495e-06,
+      "loss": 2.3868,
+      "step": 2082
+    },
+    {
+      "epoch": 0.955285484980509,
+      "grad_norm": 0.7374769976563845,
+      "learning_rate": 5.1858636566645135e-06,
+      "loss": 2.2898,
+      "step": 2083
+    },
+    {
+      "epoch": 0.9557440953909654,
+      "grad_norm": 0.7327134843306785,
+      "learning_rate": 5.0796702783340035e-06,
+      "loss": 2.1123,
+      "step": 2084
+    },
+    {
+      "epoch": 0.9562027058014216,
+      "grad_norm": 0.6987694079084086,
+      "learning_rate": 4.97456991216233e-06,
+      "loss": 2.2325,
+      "step": 2085
+    },
+    {
+      "epoch": 0.956661316211878,
+      "grad_norm": 0.7131744320761021,
+      "learning_rate": 4.870562790259325e-06,
+      "loss": 2.3166,
+      "step": 2086
+    },
+    {
+      "epoch": 0.9571199266223344,
+      "grad_norm": 0.7397688646163888,
+      "learning_rate": 4.7676491423208625e-06,
+      "loss": 2.2245,
+      "step": 2087
+    },
+    {
+      "epoch": 0.9575785370327906,
+      "grad_norm": 0.7278777715767574,
+      "learning_rate": 4.66582919562758e-06,
+      "loss": 2.3322,
+      "step": 2088
+    },
+    {
+      "epoch": 0.958037147443247,
+      "grad_norm": 0.7316936554293261,
+      "learning_rate": 4.5651031750448825e-06,
+      "loss": 2.3781,
+      "step": 2089
+    },
+    {
+      "epoch": 0.9584957578537033,
+      "grad_norm": 0.735328580790398,
+      "learning_rate": 4.465471303022217e-06,
+      "loss": 2.2475,
+      "step": 2090
+    },
+    {
+      "epoch": 0.9589543682641596,
+      "grad_norm": 0.7234355018673867,
+      "learning_rate": 4.366933799592743e-06,
+      "loss": 2.35,
+      "step": 2091
+    },
+    {
+      "epoch": 0.959412978674616,
+      "grad_norm": 0.733396206489197,
+      "learning_rate": 4.269490882372551e-06,
+      "loss": 2.1978,
+      "step": 2092
+    },
+    {
+      "epoch": 0.9598715890850722,
+      "grad_norm": 0.7262883908235934,
+      "learning_rate": 4.1731427665606115e-06,
+      "loss": 2.3252,
+      "step": 2093
+    },
+    {
+      "epoch": 0.9603301994955286,
+      "grad_norm": 0.7148416562616453,
+      "learning_rate": 4.077889664937884e-06,
+      "loss": 2.2511,
+      "step": 2094
+    },
+    {
+      "epoch": 0.9607888099059849,
+      "grad_norm": 0.6926686664642387,
+      "learning_rate": 3.983731787867207e-06,
+      "loss": 2.2111,
+      "step": 2095
+    },
+    {
+      "epoch": 0.9612474203164412,
+      "grad_norm": 0.720412278632776,
+      "learning_rate": 3.890669343292464e-06,
+      "loss": 2.3137,
+      "step": 2096
+    },
+    {
+      "epoch": 0.9617060307268975,
+      "grad_norm": 0.7641870739190394,
+      "learning_rate": 3.7987025367384743e-06,
+      "loss": 2.2936,
+      "step": 2097
+    },
+    {
+      "epoch": 0.9621646411373538,
+      "grad_norm": 0.7287995909826526,
+      "learning_rate": 3.707831571310327e-06,
+      "loss": 2.3103,
+      "step": 2098
+    },
+    {
+      "epoch": 0.9626232515478101,
+      "grad_norm": 0.698199716898602,
+      "learning_rate": 3.6180566476929912e-06,
+      "loss": 2.2431,
+      "step": 2099
+    },
+    {
+      "epoch": 0.9630818619582665,
+      "grad_norm": 0.7266499844575341,
+      "learning_rate": 3.529377964150815e-06,
+      "loss": 2.2513,
+      "step": 2100
+    },
+    {
+      "epoch": 0.9635404723687228,
+      "grad_norm": 0.6937581593471239,
+      "learning_rate": 3.441795716527307e-06,
+      "loss": 2.3229,
+      "step": 2101
+    },
+    {
+      "epoch": 0.9639990827791791,
+      "grad_norm": 0.7336047570211637,
+      "learning_rate": 3.355310098244302e-06,
+      "loss": 2.2766,
+      "step": 2102
+    },
+    {
+      "epoch": 0.9644576931896354,
+      "grad_norm": 0.7306441508693068,
+      "learning_rate": 3.269921300301959e-06,
+      "loss": 2.2149,
+      "step": 2103
+    },
+    {
+      "epoch": 0.9649163036000917,
+      "grad_norm": 0.7578001187621323,
+      "learning_rate": 3.1856295112780988e-06,
+      "loss": 2.2769,
+      "step": 2104
+    },
+    {
+      "epoch": 0.9653749140105481,
+      "grad_norm": 0.756743820144991,
+      "learning_rate": 3.102434917327812e-06,
+      "loss": 2.3244,
+      "step": 2105
+    },
+    {
+      "epoch": 0.9658335244210043,
+      "grad_norm": 0.7318224280533592,
+      "learning_rate": 3.0203377021831292e-06,
+      "loss": 2.3687,
+      "step": 2106
+    },
+    {
+      "epoch": 0.9662921348314607,
+      "grad_norm": 0.7084534934281066,
+      "learning_rate": 2.939338047152573e-06,
+      "loss": 2.1973,
+      "step": 2107
+    },
+    {
+      "epoch": 0.966750745241917,
+      "grad_norm": 0.710516318080791,
+      "learning_rate": 2.8594361311206073e-06,
+      "loss": 2.2536,
+      "step": 2108
+    },
+    {
+      "epoch": 0.9672093556523733,
+      "grad_norm": 0.7027909386811465,
+      "learning_rate": 2.7806321305475225e-06,
+      "loss": 2.2767,
+      "step": 2109
+    },
+    {
+      "epoch": 0.9676679660628297,
+      "grad_norm": 0.6993994237156919,
+      "learning_rate": 2.7029262194688818e-06,
+      "loss": 2.2435,
+      "step": 2110
+    },
+    {
+      "epoch": 0.9681265764732859,
+      "grad_norm": 0.6914561553868835,
+      "learning_rate": 2.626318569495134e-06,
+      "loss": 2.3039,
+      "step": 2111
+    },
+    {
+      "epoch": 0.9685851868837423,
+      "grad_norm": 0.7050431172993622,
+      "learning_rate": 2.550809349811334e-06,
+      "loss": 2.2522,
+      "step": 2112
+    },
+    {
+      "epoch": 0.9690437972941985,
+      "grad_norm": 0.7254853398266254,
+      "learning_rate": 2.476398727176532e-06,
+      "loss": 2.3031,
+      "step": 2113
+    },
+    {
+      "epoch": 0.9695024077046549,
+      "grad_norm": 0.7272142175502959,
+      "learning_rate": 2.4030868659237204e-06,
+      "loss": 2.3363,
+      "step": 2114
+    },
+    {
+      "epoch": 0.9699610181151113,
+      "grad_norm": 0.7189568325330709,
+      "learning_rate": 2.3308739279593317e-06,
+      "loss": 2.2929,
+      "step": 2115
+    },
+    {
+      "epoch": 0.9704196285255675,
+      "grad_norm": 0.6736668026056569,
+      "learning_rate": 2.2597600727626845e-06,
+      "loss": 2.267,
+      "step": 2116
+    },
+    {
+      "epoch": 0.9708782389360239,
+      "grad_norm": 0.7329382137357839,
+      "learning_rate": 2.1897454573860387e-06,
+      "loss": 2.2698,
+      "step": 2117
+    },
+    {
+      "epoch": 0.9713368493464801,
+      "grad_norm": 0.7334419988680935,
+      "learning_rate": 2.1208302364538746e-06,
+      "loss": 2.2064,
+      "step": 2118
+    },
+    {
+      "epoch": 0.9717954597569365,
+      "grad_norm": 0.6948807070804921,
+      "learning_rate": 2.0530145621627804e-06,
+      "loss": 2.2591,
+      "step": 2119
+    },
+    {
+      "epoch": 0.9722540701673928,
+      "grad_norm": 0.7114303311521232,
+      "learning_rate": 1.9862985842810653e-06,
+      "loss": 2.2397,
+      "step": 2120
+    },
+    {
+      "epoch": 0.9727126805778491,
+      "grad_norm": 0.7338462502047637,
+      "learning_rate": 1.920682450148259e-06,
+      "loss": 2.2494,
+      "step": 2121
+    },
+    {
+      "epoch": 0.9731712909883055,
+      "grad_norm": 0.7236603707176735,
+      "learning_rate": 1.856166304675111e-06,
+      "loss": 2.3698,
+      "step": 2122
+    },
+    {
+      "epoch": 0.9736299013987617,
+      "grad_norm": 0.7185411483061852,
+      "learning_rate": 1.792750290342926e-06,
+      "loss": 2.3259,
+      "step": 2123
+    },
+    {
+      "epoch": 0.9740885118092181,
+      "grad_norm": 0.72934146174399,
+      "learning_rate": 1.7304345472035632e-06,
+      "loss": 2.3227,
+      "step": 2124
+    },
+    {
+      "epoch": 0.9745471222196744,
+      "grad_norm": 0.7343610859485526,
+      "learning_rate": 1.6692192128788253e-06,
+      "loss": 2.2653,
+      "step": 2125
+    },
+    {
+      "epoch": 0.9750057326301307,
+      "grad_norm": 0.7723284045466836,
+      "learning_rate": 1.6091044225604035e-06,
+      "loss": 2.3033,
+      "step": 2126
+    },
+    {
+      "epoch": 0.975464343040587,
+      "grad_norm": 0.7089150375942536,
+      "learning_rate": 1.5500903090094888e-06,
+      "loss": 2.2022,
+      "step": 2127
+    },
+    {
+      "epoch": 0.9759229534510433,
+      "grad_norm": 0.7515795690537878,
+      "learning_rate": 1.492177002556383e-06,
+      "loss": 2.2241,
+      "step": 2128
+    },
+    {
+      "epoch": 0.9763815638614997,
+      "grad_norm": 0.7229856027318109,
+      "learning_rate": 1.4353646311004443e-06,
+      "loss": 2.2055,
+      "step": 2129
+    },
+    {
+      "epoch": 0.976840174271956,
+      "grad_norm": 0.7312594919478276,
+      "learning_rate": 1.3796533201094752e-06,
+      "loss": 2.3004,
+      "step": 2130
+    },
+    {
+      "epoch": 0.9772987846824123,
+      "grad_norm": 0.7740351734257102,
+      "learning_rate": 1.3250431926197793e-06,
+      "loss": 2.3377,
+      "step": 2131
+    },
+    {
+      "epoch": 0.9777573950928686,
+      "grad_norm": 0.7173523640129716,
+      "learning_rate": 1.2715343692356607e-06,
+      "loss": 2.3499,
+      "step": 2132
+    },
+    {
+      "epoch": 0.9782160055033249,
+      "grad_norm": 0.7139427043844772,
+      "learning_rate": 1.2191269681292582e-06,
+      "loss": 2.3317,
+      "step": 2133
+    },
+    {
+      "epoch": 0.9786746159137812,
+      "grad_norm": 0.7365581556460079,
+      "learning_rate": 1.1678211050402676e-06,
+      "loss": 2.3367,
+      "step": 2134
+    },
+    {
+      "epoch": 0.9791332263242376,
+      "grad_norm": 0.7703440723610089,
+      "learning_rate": 1.117616893275719e-06,
+      "loss": 2.3765,
+      "step": 2135
+    },
+    {
+      "epoch": 0.9795918367346939,
+      "grad_norm": 0.7200813402078851,
+      "learning_rate": 1.068514443709534e-06,
+      "loss": 2.2858,
+      "step": 2136
+    },
+    {
+      "epoch": 0.9800504471451502,
+      "grad_norm": 0.728338946035438,
+      "learning_rate": 1.0205138647826905e-06,
+      "loss": 2.2864,
+      "step": 2137
+    },
+    {
+      "epoch": 0.9805090575556065,
+      "grad_norm": 0.7407774382402207,
+      "learning_rate": 9.73615262502503e-07,
+      "loss": 2.3831,
+      "step": 2138
+    },
+    {
+      "epoch": 0.9809676679660628,
+      "grad_norm": 0.7457501658873609,
+      "learning_rate": 9.278187404426763e-07,
+      "loss": 2.2605,
+      "step": 2139
+    },
+    {
+      "epoch": 0.9814262783765192,
+      "grad_norm": 0.7262632887262176,
+      "learning_rate": 8.831243997431404e-07,
+      "loss": 2.3687,
+      "step": 2140
+    },
+    {
+      "epoch": 0.9818848887869754,
+      "grad_norm": 0.7069049515430583,
+      "learning_rate": 8.395323391094944e-07,
+      "loss": 2.2603,
+      "step": 2141
+    },
+    {
+      "epoch": 0.9823434991974318,
+      "grad_norm": 0.6921854294771916,
+      "learning_rate": 7.970426548131183e-07,
+      "loss": 2.2338,
+      "step": 2142
+    },
+    {
+      "epoch": 0.982802109607888,
+      "grad_norm": 0.7483821555916882,
+      "learning_rate": 7.556554406908389e-07,
+      "loss": 2.1945,
+      "step": 2143
+    },
+    {
+      "epoch": 0.9832607200183444,
+      "grad_norm": 0.7660526022303513,
+      "learning_rate": 7.153707881446536e-07,
+      "loss": 2.3839,
+      "step": 2144
+    },
+    {
+      "epoch": 0.9837193304288008,
+      "grad_norm": 0.7795192646477059,
+      "learning_rate": 6.761887861417293e-07,
+      "loss": 2.1845,
+      "step": 2145
+    },
+    {
+      "epoch": 0.984177940839257,
+      "grad_norm": 0.7315749204289634,
+      "learning_rate": 6.381095212139032e-07,
+      "loss": 2.2641,
+      "step": 2146
+    },
+    {
+      "epoch": 0.9846365512497134,
+      "grad_norm": 0.7283936753951367,
+      "learning_rate": 6.011330774577384e-07,
+      "loss": 2.3478,
+      "step": 2147
+    },
+    {
+      "epoch": 0.9850951616601696,
+      "grad_norm": 0.7213270612903868,
+      "learning_rate": 5.652595365343016e-07,
+      "loss": 2.2279,
+      "step": 2148
+    },
+    {
+      "epoch": 0.985553772070626,
+      "grad_norm": 0.6999452029859525,
+      "learning_rate": 5.304889776688859e-07,
+      "loss": 2.2892,
+      "step": 2149
+    },
+    {
+      "epoch": 0.9860123824810824,
+      "grad_norm": 0.7314704388878908,
+      "learning_rate": 4.968214776508994e-07,
+      "loss": 2.2881,
+      "step": 2150
+    },
+    {
+      "epoch": 0.9864709928915386,
+      "grad_norm": 0.6745176897092571,
+      "learning_rate": 4.6425711083375454e-07,
+      "loss": 2.2394,
+      "step": 2151
+    },
+    {
+      "epoch": 0.986929603301995,
+      "grad_norm": 0.7117864203427159,
+      "learning_rate": 4.3279594913447906e-07,
+      "loss": 2.2614,
+      "step": 2152
+    },
+    {
+      "epoch": 0.9873882137124512,
+      "grad_norm": 0.7289656247306707,
+      "learning_rate": 4.02438062033883e-07,
+      "loss": 2.2355,
+      "step": 2153
+    },
+    {
+      "epoch": 0.9878468241229076,
+      "grad_norm": 0.7219472171973156,
+      "learning_rate": 3.7318351657616987e-07,
+      "loss": 2.2098,
+      "step": 2154
+    },
+    {
+      "epoch": 0.988305434533364,
+      "grad_norm": 0.7478533906137838,
+      "learning_rate": 3.4503237736882573e-07,
+      "loss": 2.307,
+      "step": 2155
+    },
+    {
+      "epoch": 0.9887640449438202,
+      "grad_norm": 0.749354130567655,
+      "learning_rate": 3.179847065825081e-07,
+      "loss": 2.2691,
+      "step": 2156
+    },
+    {
+      "epoch": 0.9892226553542766,
+      "grad_norm": 0.7067586933033749,
+      "learning_rate": 2.9204056395104594e-07,
+      "loss": 2.2388,
+      "step": 2157
+    },
+    {
+      "epoch": 0.9896812657647328,
+      "grad_norm": 0.7283961099612097,
+      "learning_rate": 2.672000067709956e-07,
+      "loss": 2.2123,
+      "step": 2158
+    },
+    {
+      "epoch": 0.9901398761751892,
+      "grad_norm": 0.7019943842917983,
+      "learning_rate": 2.4346308990175204e-07,
+      "loss": 2.3345,
+      "step": 2159
+    },
+    {
+      "epoch": 0.9905984865856455,
+      "grad_norm": 0.6976289015680712,
+      "learning_rate": 2.208298657653818e-07,
+      "loss": 2.2767,
+      "step": 2160
+    },
+    {
+      "epoch": 0.9910570969961018,
+      "grad_norm": 0.699730945655658,
+      "learning_rate": 1.9930038434645692e-07,
+      "loss": 2.2026,
+      "step": 2161
+    },
+    {
+      "epoch": 0.9915157074065581,
+      "grad_norm": 0.7090674072485939,
+      "learning_rate": 1.7887469319205484e-07,
+      "loss": 2.2904,
+      "step": 2162
+    },
+    {
+      "epoch": 0.9919743178170144,
+      "grad_norm": 0.7022397007874671,
+      "learning_rate": 1.5955283741142523e-07,
+      "loss": 2.375,
+      "step": 2163
+    },
+    {
+      "epoch": 0.9924329282274708,
+      "grad_norm": 0.7309608830456868,
+      "learning_rate": 1.4133485967615655e-07,
+      "loss": 2.2771,
+      "step": 2164
+    },
+    {
+      "epoch": 0.9928915386379271,
+      "grad_norm": 0.750709145309168,
+      "learning_rate": 1.2422080021995407e-07,
+      "loss": 2.376,
+      "step": 2165
+    },
+    {
+      "epoch": 0.9933501490483834,
+      "grad_norm": 0.7293452814893544,
+      "learning_rate": 1.082106968385288e-07,
+      "loss": 2.2163,
+      "step": 2166
+    },
+    {
+      "epoch": 0.9938087594588397,
+      "grad_norm": 0.7087923674474848,
+      "learning_rate": 9.330458488959748e-08,
+      "loss": 2.3373,
+      "step": 2167
+    },
+    {
+      "epoch": 0.994267369869296,
+      "grad_norm": 0.7176256479763542,
+      "learning_rate": 7.950249729271608e-08,
+      "loss": 2.2323,
+      "step": 2168
+    },
+    {
+      "epoch": 0.9947259802797523,
+      "grad_norm": 0.7271164400668896,
+      "learning_rate": 6.680446452922429e-08,
+      "loss": 2.3697,
+      "step": 2169
+    },
+    {
+      "epoch": 0.9951845906902087,
+      "grad_norm": 0.6979688289380547,
+      "learning_rate": 5.521051464230098e-08,
+      "loss": 2.2672,
+      "step": 2170
+    },
+    {
+      "epoch": 0.995643201100665,
+      "grad_norm": 0.7251612492492068,
+      "learning_rate": 4.4720673236631206e-08,
+      "loss": 2.3253,
+      "step": 2171
+    },
+    {
+      "epoch": 0.9961018115111213,
+      "grad_norm": 0.7192635921416226,
+      "learning_rate": 3.53349634786837e-08,
+      "loss": 2.2813,
+      "step": 2172
+    },
+    {
+      "epoch": 0.9965604219215776,
+      "grad_norm": 0.6936977835211854,
+      "learning_rate": 2.7053406096433365e-08,
+      "loss": 2.3367,
+      "step": 2173
+    },
+    {
+      "epoch": 0.9970190323320339,
+      "grad_norm": 0.6817296786411687,
+      "learning_rate": 1.987601937930572e-08,
+      "loss": 2.289,
+      "step": 2174
+    },
+    {
+      "epoch": 0.9974776427424903,
+      "grad_norm": 0.7088704470026743,
+      "learning_rate": 1.3802819178398984e-08,
+      "loss": 2.2937,
+      "step": 2175
+    },
+    {
+      "epoch": 0.9979362531529465,
+      "grad_norm": 0.7393211767263422,
+      "learning_rate": 8.833818906039959e-09,
+      "loss": 2.2399,
+      "step": 2176
+    },
+    {
+      "epoch": 0.9983948635634029,
+      "grad_norm": 0.7195280074870747,
+      "learning_rate": 4.969029536061598e-09,
+      "loss": 2.3103,
+      "step": 2177
+    },
+    {
+      "epoch": 0.9988534739738592,
+      "grad_norm": 0.7145770961403585,
+      "learning_rate": 2.2084596038030037e-09,
+      "loss": 2.2968,
+      "step": 2178
+    },
+    {
+      "epoch": 0.9993120843843155,
+      "grad_norm": 0.7231484201508297,
+      "learning_rate": 5.521152057763601e-10,
+      "loss": 2.2562,
+      "step": 2179
+    },
+    {
+      "epoch": 0.9997706947947719,
+      "grad_norm": 0.7605828271637873,
+      "learning_rate": 0.0,
+      "loss": 2.2285,
+      "step": 2180
+    },
+    {
+      "epoch": 0.9997706947947719,
+      "step": 2180,
+      "total_flos": 6.573991851416617e+17,
+      "train_loss": 2.525755619784014,
+      "train_runtime": 7965.3468,
+      "train_samples_per_second": 70.07,
+      "train_steps_per_second": 0.274
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 2180,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 24000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 6.573991851416617e+17,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}